Islamic Inscriptions 9781474464482

Citation preview

ISLAMIC INSCRIPTIONS

ISLAMIC INSCRIPTIONS

Sheila S. Blair

Edinburgh University Press

© Sheila S. Blair, 1998 Edinburgh University Press 22 George Square, Edinburgh Transferred to digital print 2008 Designed and typeset in Trump Medieval by Fionna Robson Printed and bound by CPI Antony Rowe, Eastbourne.

A CIP record for this book is available from the British Library ISBN

0 7486 0903 2

The right of Sheila S. Blair to be identified as author of this work has been asserted in accordance with the Copyright, Designs and Patents Act 1988.

Contents List of Illustrations

vii

Acknowledgements

xi

Abbreviations

xii

Note on Transcription

xii

PART I

INTRODUCTION

Chapter 1

PART II

1

Why Read Inscriptions?

MONUMENTAL INSCRIPTIONS

3

19

Chapter 2

The Languages Used in Monumental Inscriptions

21

Chapter 3

Foundation Inscriptions

29

Chapter 4

Other Types of Monumental Inscriptions

43

Chapter 5

Regional Studies

53

Chapter 6

Inscriptions on Various Building Types

68

Chapter 7

Stylistic Development

76

PART III

INSCRIPTIONS ON PORTABLE OBJECTS

95

Chapter 8

Introduction

97

Chapter 9

Metalwares

106

Chapter 10

Woodwork

129

Chapter 11

Ceramics

148

Chapter 12

Textiles

164

Chapter 13

Other Portable Arts

182

Chapter 14

Types of Objects

196

PART IV

READING AND RECORDING INSCRIPTIONS

Chapter 15

Sources, Methods and Conventions

205 207

Bibliography

224

Index

238

— v —

List of Illustrations 1.1 1.2 1.3 1.4 1.5 2.6 2.7 2.8 2.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 4.17 4.18 4.19 5.20 5.21 5.22 5.23 5.24 6.25 6.26

Silk known as the Shroud of St Josse. Two pieces, 52 x 94 cm, 25 x 62 cm. Iran or Central Asia, before 961. Paris, Louvre, 7502. Silver-covered wooden box made for the Umayyad al-Hisham. 27 x 38 x 24 cm. Spain, 976. Treasury Museum, Gerona Cathedral. Slip-covered earthenware bowl. Diameter 39.3 cm. Iran, tenth century. Washington DC, Freer Gallery of Art 57.24. Interior of the Dome of the Rock, Jerusalem, showing the mosaic inscription at the top of the inner face of the octagonal arcade, 692. Inscription around the dome of the mosque in the terminal of the King Khalid International Airport built at Riyadh, Saudi Arabia in 1984. Tomb tower at Lajim in northern Iran, with foundation texts in Pahlavi at the top and in Arabic below, late tenth or early eleventh century. Interior of the tomb at Safid Buland in the Farghana Valley, Uzbekistan, with the first datable inscriptions in Persian, 1055–60. Foundation inscription in Arabic over the portal to the Selimiye mosque at Edirne, giving chronograms with the dates of foundation (976/1568–9) and completion (982/1574–5). Detail of the foundation inscription in Turkish on the fountain near Tophane in Istanbul, early eighteenth century. Detail of the inscription band around the outer face of the arcade in the Dome of the Rock, Jerusalem, with the name of the patron. Copy of van Berchem’s drawing of the patron’s name in the outer face of the arcade in the Dome of the Rock. The area between the dots is al-Maæmun’s restoration. Foundation inscription across the façade of Qalaæun’s tomb complex in Cairo, 682–4/1283–5. Detail of the foundation inscription across the façade of Qalaæun’s tomb complex. Foundation inscription around the portal of the Mosque of Shaykh Lutfallah in Isfahan, dated 1012/1603–4 (after restoration). Detail of the end of the foundation inscription around the portal of the Mosque of Shaykh Lutfallah, showing the date 1012/1603–4 and the signature of the calligrapher ÆAli Riza ÆAbbasi. Umayyad milestone in the name of ÆAbd al-Malik recording the distance of 109 miles to Damascus. Istanbul, Museum of Turkish and Islamic Art 2511. Funerary inscriptions around the interior of the Imamzada Yahya b. Zayd at Sar-i Pul in northern Afghanistan. Commemorative text carved at Persepolis in the name of the Buyid ruler ÆAdud al-Dawla in 344/955–6. Interior of the Mosque of Shaykh Lutfallah showing the mihrab. Wadayas Gate, Rabat, late twelfth century, with an unread inscription around the frame. Foundation text on the Kharput Gate at Diyarbekır dated 297/910. Stone plaque recording the foundation of the Biær al-Watawit in Cairo by the Ikhshidid vizier Ibn al-Furat in 355/966. Flury’s chart of the letter shapes used in the inscription on the mosque at Naæin, mid-tenth century. Beginning of the Koranic band with Chapter 89 (Surat al-Fajr) on the south side of the gateway to the Taj Mahal at Agra, completed 1057/1647. Stucco mihrab added to the Friday Mosque in Isfahan in Safar 710/July 1310. Beginning of the Koranic inscription with Chapter 48:1–6 around the courtyard of the madrasa of Sultan Hasan in Cairo, completed 764/1362.

— vii —

5 6 7 9 10 22 23 26 27 30 30 31 32 33 34 42 46 48 50 55 56 58 63 66 70 71

— ISLAMIC INSCRIPTIONS — 6.27 6.28 6.29 7.30 7.31 7.32 7.33 7.34 7.35 7.36 7.37 7.38 7.39

7.40 7.41 7.42 7.43 7.44 8.45

8.46 8.47 9.48 9.49 9.50

9.51 9.52 9.53

Minaret erected at Dawlatabad, Afghanistan in 502/1108–9. Detail of the tile mosaic inscription with the profession of faith on the north-west minaret erected by al-Nasir Muhammad in his mosque on the citadel in Cairo, 735/1335. Bab al-Nasr in Cairo, begun Muharram 489/April–May 1087 (photo S. R. Peterson). Detail of the mosaic inscription in simple Kufic on the inner face of the octagonal arcade in the Dome of the Rock, Jerusalem. End of the foundation inscription in floriated Kufic on the Aqmar Mosque in Cairo dated 519/1125–6. Kufesque inscription on the façade of the Theotokos church at the monastery of Hosias Loukas in Phocis, Greece, built c. 950. Drawing of the upper band of Kufic with interlaced stems from the minaret erected at Tirmidh in 423/1031–2. Bordered Kufic inscription from the Nizamiyya madrasa at Khargird in north-eastern Iran, constructed 465–70/1072–7. Minaret of MasÆud III at Ghazna, built c. 1100 (photo Catherine B. Asher). Panels in square Kufic on the minaret of MasÆud III at Ghazna. Minaret built at Gar, Iran in 515/1121–2 with a shaft decorated in square Kufic. Interior of the shrine constructed for Pir-i Bakran at Linjan, near Isfahan in the early fourteenth century, showing the north wall with several kinds of square Kufic inscriptions. Interior of the Karatay madrasa at Konya built in 649/1251–2, with sacred names in square Kufic in the Turkish triangles and a Koranic inscription (2:255) in Kufic with interlaced stems around the base of the dome. Marble mosaic panel in square Kufic from the mausoleum of Qalaæun, Cairo. Marble slab with the foundation inscription in naskh for Nur al-Din’s hospital in Damascus, 549/1154–5. Stucco cartouches decorating the Hall of the Ambassadors in the Alhambra palace at Granada, mid-fourteenth-century. Stucco band with Koran 3:185 on the courtyard wall of the tomb complex for Salar and Sanjar in Cairo, begun 703/1303. Underglaze-painted tile lunette from the Uç „erefeli Mosque at Edirne, completed in 851/1447–8. Detail of the inscription around the rim of a lustre plate with the scene of a sleeping groom made by Abu Zayd in Jumada II 607/November 1210. Diameter 35.2 cm. Washington DC, Freer Gallery of Art 41.11. Detail of the underside of the clasp on the Gerona casket with the signature of the artisans Badr and Tarif. Detail of the signature on an inlaid brass ewer in the shape of a bird made by Sulayman in 180/796–7. Ht 38 cm. St Petersburg, Hermitage IR-1567. Inlaid brass basin known as the Baptistère de St Louis made by Muhammad b. al-Zayn. Diameter of rim 50 cm. Paris, Louvre, LP 16. Inlaid steel mirror made by the master Muhammad al-Waziri. Diameter 24 cm. Istanbul, Topkapı Palace Treasury. Bobrinsky bucket made by Muhammad b. ÆAbd al-Wahid and Hajib MasÆud b. Ahmad in Muharram 559/December 1163. Bronze inlaid with copper and silver. Ht 18.5 cm. St Petersburg, Hermitage IR-2268. Inlaid brass pen-box made by Shadhi in 607/1210–11. Length 31.4 cm. Washington DC, Freer Gallery of Art 36.7. Drawing of the inscription around the rim of a bronze ewer made by Ibn Yazid at Basra? in [1]67/783–4. Georgian State Museum, Tbilisi. Blacas ewer made by ShujaÆ b. ManÆa al-Mawsili in Rajab 629/May 1232. Ht 30 cm. London, British Museum 66.12–69.61.

— viii —

72 74 75 77 78 79 81 81 83 84 85 86

87 89 90 91 92 93

99 101 103 107 109

111 115 117 122

— LIST 9.54 9.55

10.56 10.57 10.58

10.59 10.60 10.61 10.62 10.63 11.64

11.65

11.66 11.67 11.68 11.69 12.70 12.71

12.72 12.73 12.74 12.75

12.76

OF

I L L U S T R AT I O N S —

Inlaid brass bowl made in Fars province in the fourteenth century. Diameter 23 cm. Washington DC, Freer Gallery of Art 80.25. Inlaid brass bowl made for the Mamluk sultan Qaæitbay (r. 1468–96). Diameter of rim 32 cm. New York, The Metropolitan Museum of Art, Edward C. Moore Collection, Bequest of Edward C. Moore, 1891 (91.1.565). Detail of the wooden minbar ordered by Nur al-Din Zangi in 564/1168–9 for the Aqsa Mosque in Jerusalem (after van Berchem, MCIA Jerusalem, plate XXIX). Detail with the basmala and Koran 36:1–3 from the wooden frieze around the top of the walls in the tomb built for the Imam al-ShafiÆi in Cairo in 608/1211. Detail of the wooden maqßËra donated to the Great Mosque at Qayrawan in 413–14/1022–3, with the name and genealogy of the Zirid ruler al-MuÆizz b. Badis in Kufic with interlaced stems. Wooden panel from a bier in the Israel Museum, made in Iran in the tenth century, with a Kufic text about the deceased’s testimony. Wooden minbar in the Eski Jami at Bey∞ehir ordered by the E∞refoÌlu amir Sulayman c. 699/1300 (photo Walter Denny). Wooden screen added around the cenotaph in the tomb of Qalaæun in Cairo in 703/1303–4. Wooden revetment in the ÆAttarin Madrasa, Fez, 725/1325–6. Wooden Koran-stand made by Hasan b. Sulayman Isfahani in 761/1360. Ht 130 cm. New York, The Metropolitan Museum of Art, Rogers Rund, 1910 (10.218). Drawing of the inscription on an earthenware bowl painted in blue on an opaque white glaze with blessings to the owner and the signature of the artisan Muhammad al-?. Diameter 23.5 cm. Munich, Staatliches Museum für Völkerkunde no. 28-8-82. Earthenware plate underglaze painted in brown and red on a white slip with the aphorism al-jawd min akhlåq ahl al-janna ‘Generosity is a quality of the people of Paradise’. Eastern Iran, tenth century. Diameter 21 cm. Washington DC, Freer Gallery of Art 65.27. Fragmentary lustre plate signed by Muslim. Egypt, c. 1000. Diameter 38 cm. Athens, Benaki Museum no. 11122. Panel of three lustre tiles decorated with a hanging lamp made by ÆAli b. Muhammad in 709/1309–10. New York, The Metropolitan Museum of Art, Rogers Fund, 1909 (09.87). Fortuny Tablet. Lustre painted terracotta. Ht 108 cm. Madrid, Instituto de Valencia de Don Juan. Underglaze-painted mosque lamp from the Dome of the Rock in Jerusalem made by Musli in 956/1549. Ht 38.5 cm. London, British Museum 87.5-16.1. Fragment of a tiraz made for the ÆAbbasid caliph al-Muktafi at Tinnis in 290/902–3. Linen embroidered with silk. Washington DC, Textile Museum 73.639. Fragment of a tiraz made for the Fatimid caliph al-Zahir (r. 1020–36). Linen tapestry-woven with silk and gold thread. 16.7 x 19.6 cm. The Cleveland Museum of Art, Andrew R. and Martha Holden Jennings Fund, 1982.109. Fragment of a tiraz made under the supervision of the Fatimid vizier al-Afdal c. 500/1106. Linen tapestry-woven with silk. Washington DC, Textile Museum 73.680. Cotton ikat with painted and gilded inscription offering God’s blessing on Muhammad. Yemen, tenth century. Washington, DC, Textile Museum 73.377. Silk cloth inscribed with the name and titles of the Buyid ruler Bahaæ al-Dawla. Width 2.9 m. Mesopotamia or western Iran, c. 1000. Washington DC, Textile Museum 3.116. Front of a lampas-woven silk tunic inscribed with good wishes to the amir Abuæl-Raæis Khalaf b. Mansur al-Tawaghi, the date 203/818–9, and the name of the city, Baghdad. 113.7 x 86.7 cm. Post-1950. The Cleveland Museum of Art, Gift of the Junior Council of the Cleveland Museum of Art, 1982.23. Drawing of the inscription on the ‘Baghdad silk’ made into the shroud of San Pedro de Osma. Spain, late eleventh century. Boston Museum of Fine Arts 33.371.

— ix —

124

126 131 132

134 136 138 140 142 146

151

153 154 157 160 162 167

168 169 170 171

173 174

— ISLAMIC INSCRIPTIONS — 12.77 12.78

13.79 13.80 13.81 13.82 13.83

14.84 14.85 15.86 15.87 15.88 15.89

Silk curtain made for the Nasrids. 4.38 x 2.72 m. Granada, fifteenth century. The Cleveland Museum of Art, Leonard C. Hanna Jr. Bequest, 1982.16. Textile with felines, eagles and pseudo-inscription in Kufic with interlaced stems. Lampas weave, silk and gold thread. 1.70 x 1.09 m. Central Asia, thirteenth century. The Cleveland Museum of Art, Purchase from the J. H. Wade Fund, 1990.2. Blown glass cup with lustre decoration made in Damascus under the supervision of Sunbat. Eighth century. Ht 10.3 cm. Corning Museum of Glass 69.1.1. Glass lamp made for the hospice of Karim al-Din. Egypt, 1310–20. Ht 27.5 cm. Boston Museum of Fine Arts 37.614. Wooden casket with ivory inlay made for the Fatimid caliph al-MuÆizz at al-Mansuriyya c. 970. Length 42 cm. Madrid, National Archaeological Museum 887. Rock-crystal cruet made for the Fatimid caliph al-ÆAziz (r. 977–96). Venice, Treasury of San Marco 80. White jade tankard made for the Timurid prince Ulugh Beg and reinscribed for the Mughal emperors, Jahangir in 1022/1613 and Shah Jahan in 1056/1646. Ht 14.5 cm. Lisbon, Calouse Gulbenkian Foundation 328. Crested grave cover found at Siraf and made for Ibrahim b. ÆAli, who died in 383/993. Length 1.51 m (after Lowick 1985, plate XIII). Detail of a talismanic shirt painted with Koranic verses. Length 122.5 cm. Istanbul, mid-sixteenth century. Istanbul, Topkapı Palace 13/1150. Minaret of Jam, Afghanistan, late twelfth century. Bronze cauldron ordered by Timur on 20 Shawwal 801/25 June 1399 for the shrine of Ahmad Yasavi in Turkestan City. Ht 1.58 m. Lustre tile from a frieze with the date [Shawwa]l 707/March–April 1308. Kashan. 38.1 x 38.1 cm. New York, The Metropolitan Museum of Art, Gift of Emile Rey, 1912 (12.44). Bronze astrolabe made by Bastulus or Nastulus in ÆIraq in 315/927–8. Diameter 17.5 cm. Kuwait, Dar al-Athar al-Islamiyya LNS 36M.

— x —

177

178 184 186 188 191

194 197 202 211 214 218 221

Acknowledgements The research and writing of this book were undertaken in large part thanks to the generous support of the Fondation Max van Berchem in Geneva. The foundation, whose goal is to promote the study of archaeology, history, geography, the history of art, epigraphy, religion, and Islamic and Arabic literature, is named in honour of the Swiss scholar who founded the study of Islamic epigraphy just over a century ago. Van Berchem became the unsurpassed master of the field. He was, in the words of his colleague K. A. C. Creswell, who dedicated the first volume of his massive bibliography on the arts and crafts of Islam to van Berchem, ‘a perfect friend and a perfect scholar’. His material and methods continue to serve as inspiration, and it should come as no surprise that van Berchem’s name occurs more often than any other in this book. Many other friends and scholars have patiently and steadfastly answered my stream of queries and requests for information. In particular I would like to thank Catherine Asher, Carol Bier, Stefano Carboni, Robert and Carole Hillenbrand, Linda Komaroff, Wheeler Thackston, Anne Wardwell, Estelle Whelan and David Whitehouse. Most of all, it is my colleague and husband Jonathan Bloom who kept me from faltering along the long, and sometimes rocky, road.

— xi —

Abbreviations CII = Corpus Inscriptionum Iranicarum EI = Encyclopaedia of Islam EIr = Encyclopaedia Iranica MCIA = Matériaux pour un Corpus Inscriptionum Arabicarum RCEA = Répertoire chronologique d’épigraphie arabe Survey = A. U. Pope and P. Ackerman (eds), A Survey of Persian Art

Volumes in the MCIA Egypte 1 = Max van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, part I: Égypte, fascicules 1–4: Le Caire; Mémoires de la Mission Archéologique française au Caire XIX (Paris, 1894–1903). Egypte 2 = Gaston Wiet, Matériaux pour un Corpus Inscriptionum Arabicarum, part I: Égypte, tome 2: Égypte; Mémoires de l’Institut français d’archéologie orientale du Caire LII (Cairo, 1929–30). Syrie du Nord = Moritz Sobernheim, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Nord, fascicule 1; Mémoires de l’Institut français d’archéologie orientale du Caire XXV (Cairo, 1909). Jerusalem = Max van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Sud, tome 1: Jérusalem, ville; tome 2: Jérusalem, Óaram; tome 3: plates; Mémoires de l’Institut français d’archéologie orientale du Caire XLIII–XLV (Cairo, 1920–2). Alep = Ernst Herzfeld, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Nord; Inscriptions et Monuments d’Alep, tome I (2 vols): text; tome II: plates; Mémoires de l’Institut français d’archéologie orientale du Caire LXXVI–LXXVIII (Cairo, 1954–5). Asie Mineure = Max van Berchem and Halil Edhem, Matériaux pour un Corpus Inscriptionum Arabicarum, part III: Asia Mineure, tome 1: Siwas, Diwrigi; Mémoires de l’Institut français d’archéologie orientale du Caire XXIX (Cairo, 1910–17). Arabie = Hassan Mohammed el-Hawary and Gaston Wiet, Matériaux pour un Corpus Inscriptionum Arabicarum, part IV: Arabie; tome 1: Inscriptions et Monuments de la Mecque; fascicule 1: Haram et KaÆba; ed. and rev. Nikita Elisséeff, Mémoires de l’Institut français d’archéologie orientale du Caire CIX/1 (Cairo, 1985).

Note on Transcription The system of transcription used in this book follows that of the second edition of the Encyclopaedia of Islam, with a few modifications (k, dj, dh, kh and sh are rendered as q, j, dh, kh and sh respectively).

— xii —

Abbreviations CII = Corpus Inscriptionum Iranicarum EI = Encyclopaedia of Islam EIr = Encyclopaedia Iranica MCIA = Matériaux pour un Corpus Inscriptionum Arabicarum RCEA = Répertoire chronologique d’épigraphie arabe Survey = A. U. Pope and P. Ackerman (eds), A Survey of Persian Art

Volumes in the MCIA Egypte 1 = Max van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, part I: Égypte, fascicules 1–4: Le Caire; Mémoires de la Mission Archéologique française au Caire XIX (Paris, 1894–1903). Egypte 2 = Gaston Wiet, Matériaux pour un Corpus Inscriptionum Arabicarum, part I: Égypte, tome 2: Égypte; Mémoires de l’Institut français d’archéologie orientale du Caire LII (Cairo, 1929–30). Syrie du Nord = Moritz Sobernheim, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Nord, fascicule 1; Mémoires de l’Institut français d’archéologie orientale du Caire XXV (Cairo, 1909). Jerusalem = Max van Berchem, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Sud, tome 1: Jérusalem, ville; tome 2: Jérusalem, Óaram; tome 3: plates; Mémoires de l’Institut français d’archéologie orientale du Caire XLIII–XLV (Cairo, 1920–2). Alep = Ernst Herzfeld, Matériaux pour un Corpus Inscriptionum Arabicarum, part II: Syrie du Nord; Inscriptions et Monuments d’Alep, tome I (2 vols): text; tome II: plates; Mémoires de l’Institut français d’archéologie orientale du Caire LXXVI–LXXVIII (Cairo, 1954–5). Asie Mineure = Max van Berchem and Halil Edhem, Matériaux pour un Corpus Inscriptionum Arabicarum, part III: Asia Mineure, tome 1: Siwas, Diwrigi; Mémoires de l’Institut français d’archéologie orientale du Caire XXIX (Cairo, 1910–17). Arabie = Hassan Mohammed el-Hawary and Gaston Wiet, Matériaux pour un Corpus Inscriptionum Arabicarum, part IV: Arabie; tome 1: Inscriptions et Monuments de la Mecque; fascicule 1: Haram et KaÆba; ed. and rev. Nikita Elisséeff, Mémoires de l’Institut français d’archéologie orientale du Caire CIX/1 (Cairo, 1985).

Note on Transcription The system of transcription used in this book follows that of the second edition of the Encyclopaedia of Islam, with a few modifications (k, dj, dh, kh and sh are rendered as q, j, dh, kh and sh respectively).

— xii —

— CHAPTER 1 —

Why Read Inscriptions?

I

nscriptions, like geometric designs and arabesques, are some of the most distinctive and persistent motifs used to decorate works of art and architecture made in the Islamic lands. The use of inscriptions is not unique to Islamic culture. There was a long tradition in the classical world of using inscriptions, particularly to decorate the façades of monumental structures, and modern collections fill several large volumes of the Corpus Inscriptionum Graecarum and the Corpus Inscriptionum Latinarum. The word ‘epigraphy’, the study or science of inscriptions, is derived from the Greek epi graphein (‘to write on’). The Romans recognised the artistic possibilities of monumental inscriptions and developed a system of clear and simple lettering in which the letters and the spaces between them were made to conform to the architectural structure. This innovation had a major impact on Western art, for the Roman system of monumental script is still the basis for modern lettering and book-printing. The Romans used inscriptions to underscore the authority and dignity of their monuments, themselves often vehicles of visual propaganda. A prime example is the Roman triumphal arch, which has a bold band at the top of the arch recording the patron’s name. The arch for the emperor Titus (d. AD 81) in Rome, for example, bears a large four-line band of finely cut letters that were originally gilded. The emperor died during its construction, and the text, which is set over a relief in the vault showing his apotheosis, states `The Roman Senate and People to Deified Titus, Vespasian Augustus, son of Deified Vespasian’.

A tradition of monumental inscriptions accompanying wall reliefs also existed in ancient Iran. The most famous example is the carved cliff at Bisutun (or Behistun) near Kermanshah in south-west Iran. The site lies on the north side of the caravan trail and military route from Babylon and Baghdad over the Zagros Mountains to ancient Ecbatana and medieval Hamadan. Darius I ‘the Great’, Achaemenid king of Persia from 522 to 486 BC, had the cliff decorated with a monumental relief showing his triumph over the usurper Gaumata and the nine rebels. Around the relief is a trilingual inscription in Old Persian, Elamite and Babylonian. The inscription is probably the most important document of the entire Ancient Near East, for it was the key to deciphering cuneiform writing and thus comparable to the Rosetta stone for Egyptology. In the medieval West, images often replaced the writing used in classical times to decorate the façades of major buildings. On Romanesque and Gothic cathedrals, for example, façades are typically decorated with figural reliefs explaining elements of church belief in visual terms. The tympanum over the main (south) portal to the Romanesque church at Moissac, a Cluniac priory built in the early twelfth century on the road to Santiago, for example, shows Christ in Glory, surrounded by seraphim and the symbols of the evangelists and atop the adoring twenty-four elders. Depictions of St Peter and St Paul and the prophets Jeremiah and Isaiah on the door jambs and trumeau (the central post supporting the lintel) support this vision literally (as architectural elements) and figuratively (as representatives of

— 3 —

— ISLAMIC INSCRIPTIONS — the Old and New Testaments). The main (west) façade of the Gothic cathedral at Chartres has a similar scene, with the Old Testament kings and queens of Judea suppporting Christ, the Virgin and the vision of the Second Coming. In the Islamic lands, by contrast, the earlier tradition of monumental writing not only continued but expanded. What sets Islamic epigraphy apart from its precursors is its extent in time, place and media, for inscriptions are used on virtually all kinds of objects created throughout the Islamic lands in all periods. Inscriptions occur on objects of all media and materials, from the humblest, such as oil lamps and other unglazed ceramics, to the finest and most expensive, including rock crystals and jades. Inscriptions were even added in media where the technical limitations of the material make it extremely difficult to incorporate a running text. Take, for example, silk textiles. It is comparatively easy to set up short texts that repeat in mirror reverse, as on a group of compound silk twill bands woven as trimmings for tunics. These tunic bands are generally attributed to eighth- or ninth-century Syria, which had been a centre of silk production since Byzantine times. Many pieces were found in Egyptian graves, especially at Akhmim in Upper Egypt. Some bear Greek or Coptic names, showing that they were made for the Christian market, but others, identical in style, have Arabic inscriptions (for example, Arts of Islam 1976, no. 2). The inscriptions often make no sense. The letters may have been stylised to fit the constraints of weaving, the meaning may have been lost in the repetition of the pattern, or the weavers may not have known how to read Arabic. The demand for inscribed textiles was so great, however, that silk weavers in the Islamic lands soon overcame the confines of the technique, and by the tenth century Persian weavers had figured out how to incorporate long bands of inscriptions on their elaborately patterned silks woven on drawlooms. One of the finest examples to be preserved is the silken shroud of St Josse in the Louvre, (re)used in medieval times to wrap

the bones of St Josse in the abbey of St Josse-surMer, near Caen in northern France (see Figure 1.1). Two fragments are preserved, one large (52 x 94 cm) and one small (24 x 62 cm), and the complete piece can be reconstructed as a square measuring 1.5 metres on a side, with a carpetlike design of borders surrounding a central field. The borders contain a train of two-humped or Bactrian camels, and the field would have had two identical bands of elephants with inscriptions written upside down underneath the elephants’ feet. The text invokes glory and prosperity to the commander, Abu Mansur Bakhtikin (RCEA 1507; Bernus et al., 1971; Arts of Islam 1976, no. 4). Using textual sources, he can be identified as a Turkish commander in the province of Khurasan in north-eastern Iran who was arrested and executed on orders of his Samanid sovereign ÆAbd al-Malik b. Nuh in 349/960–1. The inscription on the silk asks for blessing on a living person and thus provides a terminus ad quem for the piece, which must have been made for Abu Mansur Bakhtikin when he was still alive. Although the only example to survive, this silk must have been one of many identical pieces, for it was extremely time-consuming and expensive to set up the drawloom to weave this complicated design in seven colours, and the cost would have been amortised by weaving the silk squares in multiples. We do not know how silks like the shroud of St Josse were used, but a likely possibility is as saddlecloths. This type of embroidered silk saddlecloth is mentioned by the Persian traveller and IsmaÆili spy Nasir-i Khusraw, who journeyed from his home in north-eastern Iran to the Fatimid court in Cairo in the mid-eleventh century. He describes the saddlecloths worn by the 2,000 horses in the retinue of the Fatimid caliph during festivities for the opening of the canal in Cairo in the spring of 1048. Nasir-i Khusraw specifically comments that the saddlecloths were woven of brocaded silk that was neither cut nor sewn but woven to shape and had inscriptions with the name of the sultan along the borders. It must have been a stunning sight indeed to see the lines of horses, each bedecked with the ruler’s name along its flanks, and the

— 4 —

— WHY READ INSCRIPTIONS? —

1.1 Silk known as the Shroud of St Josse. Two pieces, 52 x 94 cm, 25 x 62 cm. Iran or Central Asia, before 961. Paris, Louvre, 7502.

image obviously struck a chord with the Persian traveller, who might have been familiar with similar saddlecloths at home in north-eastern Iran. If we imagine that the shroud of St Josse and its clones were similarly worn by horses in Abu Mansur Bakhtikin’s retinue, then the saddlecloths would have hung down over the animals’ backs, with the elephants upright to viewers and

the inscriptions legible to the people seated on the horses. The shroud of St Josse is just one example of how artists in medieval Islamic times used inscriptions to decorate objects. Artists did this throughout the Islamic lands, and we can see how common this tradition was by contrasting the shroud to a contemporary object from the

— 5 —

— ISLAMIC INSCRIPTIONS — opposite end of the Islamic lands, a casket made in Islamic Spain (see Figure 1.2). The wooden box, probably designed to hold small and precious materials such as spices or jewels and now preserved as a treasured relic in the Cathedral at Gerona, is covered with hammered plaques of silver decorated with gilt and niello (a black alloy of sulphur, lead and silver). A large inscription, worked in niello that contrasts with the gold ground, encircles the base of the lid. The text first invokes blessings on al-Hakam, the Umayyad caliph of Spain who ruled from AD 961 to 976, and then continues that the box was among the things ordered for Abu Walid Hisham, his son and heirapparent, under the direction of the official Jawdar. Since Hisham was only declared heir-apparent on 5 February 976 and succeeded his father on 1

October of the same year, the inscription allows us to date the box precisely to eight months of the year 976. The inscription further shows that the box was a specific commission, perhaps to commemorate the prince’s elevation to the status of heir-apparent. In addition to the main inscription on the lid of the Gerona casket, another small inscription tucked on the underside of the clasp (see Figure 8.46) gives the signature of the artisans, al-Hakam’s servants Badr and Tarif. The shroud of St Josse, made for a Samanid commander in the mid-tenth century, and the Gerona casket show how precious objects made in medieval Islamic times were inscribed with bands proclaiming the name of the patron or owner. Inscriptions, some with a different message in a different style of script, are also

1.2 Silver-covered wooden box made for the Umayyad al-Hisham. 27 x 38 x 24 cm. Spain, 976. Treasury Museum, Gerona Cathedral.

— 6 —

— WHY READ INSCRIPTIONS? — found on less expensive wares, as can be seen on a deep flaring bowl made in the Samanid domains at the same time as the shroud (see Figure 1.3). Made of buff-coloured earthenware covered with a fine white slip, the bowl is painted in red and dark brown slips and covered with a transparent colourless glaze. In the centre is an abstracted plant motif, but the major decoration is a wide band of elegant angular script encircling the inside walls.

In contrast to the historical texts on the shroud and the casket, the inscription on the bowl contains good wishes to the owner and a proverb. A small decorative motif set at about four o’clock indicates the beginning of the text, which, like all inscriptions written in Arabic script, reads from right to left. The text begins with a phrase offering ‘Blessing to its owner’, a phrase that can be transcribed in Latin characters

1.3 Slip-covered earthenware bowl. Diameter 39.3 cm. Iran, tenth century. Washington DC, Freer Gallery of Art 57.24.

— 7 —

— ISLAMIC INSCRIPTIONS — as baraka li-ßå˙ibihi. A small teardrop motif set at about eight o’clock on the bowl marks the beginning of a proverb, yåqålu qad khå†ara man istaghaniya biråæyah, which translates as `It is said that he who is content with his own opinion runs into danger’. Assuming that the bowl was intended to be held and appreciated with the stem of the plant at the bottom, closest to the viewer, then the most important part of the inscription, the blessing to the owner, is set immediately below the plant motif. To read the following proverb, the viewer has to rotate the bowl completely around in a counter-clockwise direction. The inscription thus invites the holder to handle and turn the bowl. The inscriptions on this bowl and other comparable ceramics are carefully thought out, and the script truly deserves to be called calligraphy, literally `beautiful writing’. Whereas the inscriptions on the shroud and the casket were designed for clarity and immediate comprehension, the inscription on the bowl was designed for aesthetic impact. The dark brown letters are attenuated, and the tall vertical strokes extend high above the bodies of the letters in the lower zone. To fill the empty space, the artist surrounded the letters with a broad band reserved in the white of the slip ground and filled the remaining ground with spots and four-petalled flowers. He encircled the rim with alternating red and black scallops and deliberately skipped a space at the top so that a scallop would not impinge upon the descending tail of the letter nËn in the word man. As with the shroud, we can only speculate about how this inscribed bowl was used. One of the largest and finest ceramics of its type to survive (it measures some 39 cm in diameter), it was probably a serving vessel that was presented to a communal party of diners and filled with food. As the diners reached in and emptied the bowl, the letters would slowly be revealed. Only when they had finished could they pick the bowl up and turn it around to read the text. Even then, the inscription is difficult to read, especially with the stylised letters and lack of dots. Deciphering the inscriptions of these ceramics is something of a speciality today. It is almost as

though the inscriptions on these bowls and plates were composed as puzzles, whose unravelling was designed to entertain the satisfied diners after their meal. The widespread use of inscriptions on these ceramics and other objects from the tenthcentury has wider implications for more general questions about medieval Islamic society. One is the rate of literacy, for the widespread use of inscriptions suggests that many members of the upper classes were literate. The rate of literacy obviously fluctuated, and literacy and verbal skills were prized at some times or in some places more than others. The Samanid period, when written puzzles dominated the decoration of tablewares, for example, was a time of cultural florescence, and the Samanid court was renowned for its patronage of scholarship and learning. Much of the literature composed under Samanid patronage, however, was written in new Persian, and several scholars, beginning with Volov [Golombek] (1966), have addressed the apparent paradox of why ceramics with inscriptions in Arabic were popular at the time that literature in new Persian developed. She suggested that the paradox may be more apparent than real and that the Persian renaissance that took place under the Samanids may have been part of a more general flowering of culture in Iran. Certainly, Samanid administration was modelled on that of the ÆAbbasid caliphs at Baghdad, where earthenware bowls were often painted with Arabic inscriptions written in cobalt blue on the opaque white glaze in the centre of the bowl (see Figure 11.64). Volov also suggested that the Samanid ceramics with Arabic inscriptions might have appealed to a specific segment of the community, and Bulliet (1992) took the argument one step further. Adopting the Annales technique of wedding material culture to other areas of historical enquiry, he connected this type of inscribed ware with the elitist members of the society, descendants of comparatively early converts to Islam who persisted in supporting the legal, theological and spiritual traditions that had been dominant at the time of their ancestors’ conversions. Although somewhat speculative, his analysis is worthy of

— 8 —

— WHY READ INSCRIPTIONS? —

1.4 Interior of the Dome of the Rock, Jerusalem, showing the mosaic inscription at the top of the inner face of the octagonal arcade, 692.

attention, for it shows how inscribed objects can shed light on medieval society. These three examples of inscribed wares – the shroud, the box and the bowl – all date from the tenth century, but inscriptions are found on objects created throughout the history of Islamic culture, from the earliest times to the present. The earliest Islamic building, the Dome of the Rock in Jerusalem, built by the Umayyad caliph

ÆAbd al-Malik in 72/692, already shows a sophisticated use of inscriptions executed in mosaic on the interior (see Figure 1.4). Two long bands of inscriptions, written in gold letters that sparkle against the deep blue ground, encircle the inner and outer faces of the octagonal arcade. Although the bands are similar in technique and style, the texts differ in organisation and content. The band on the inner face has a continuous text

— 9 —

— ISLAMIC INSCRIPTIONS — that begins in the south-west corner and runs counter-clockwise around the open central space. Following the invocation to God, the text describes the omnipotent power of God and extols Muhammad as His Prophet and Jesus as His messenger but not His son. The band on the outer face, by contrast, is divided into six sections by rosettes. The first sections contain Koranic verses and pious phrases of a similar tenor about God’s omnipotence and Muham-

mad’s prophetic mission, but the sixth is a historical text with the name of the patron and the date of construction. Inscriptions continue to play an important role in the decoration of modern buildings in the Islamic lands. A good example is the mosque in the terminal of the King Khalid International Airport built at Riyadh, Saudi Arabia in 1984 (see Figure 1.5). As at the Dome of the Rock, the inscription is written in a large band that encircles

1.5 Inscription around the dome of the mosque in the terminal of the King Khalid International Airport built at Riyadh, Saudi Arabia in 1984.

— 10 —

— WHY READ INSCRIPTIONS? — the base of the dome. This one sits above a band of gold decorated with black geometric patterns and below the light and airy dome constructed of tiers of concrete. The tall white letters written in thuluth script are set against a blue ground. A cartouche at one side encloses the basmala, or invocation to God. The rest of the band is a continuous Koranic text containing the first seven verses of Chapter 57, Surat al-Hadid (‘Iron’). They state that whatever is on the earth or in the heavens declares the glory of God, the almighty who has power over all things. They conclude that whoever spends money on a pious work will be justly rewarded. The Koranic text in the mosque was consciously chosen to fit the site. The opening verses are some of the most lyric evocations in the Koran of God’s majesty and power on heaven and earth. The final verse about pious gifts is undoubtedly an allusion to the patron’s construction of a mosque in the city where modern pilgrims would first set foot in Saudi Arabia en route to Mecca while performing the hajj. Technique and placement reinforce the meaning of the inscription around the base of the dome. The band is set so that it appears to mediate between the earth below, represented by the gold and black band, and the heavens above, represented by the lighted dome. Sometimes the desire to have writing was as important as the execution, and the need to have a specific text outweighed considerations about its style. A good example here is the Massachusetts Avenue entrance to the Islamic Center built in Washington DC in 1368/1949 (Ettinghausen 1974, fig. 8; Thackston 1994: 45). The façade is decorated with a broad band of Koranic text containing part of verse 36 from Chapter 24, Surat al-Nur (‘Light’). The text, which follows the celebrated Light Verse (24:35) in which God is extolled as the light of the heavens and the earth, refers to houses whose construction is permitted by God and in which his name is honoured. Like the verses around the dome of the airport mosque, this verse was specifically and appropriately chosen for its location, the mosque in an Islamic centre. The style of neoKufic, however, is awkward. The tooth of the

letter yåæ in buyËt, the second word, is so tall that it looks like the letter låm, and most passers-by, even those fluent in Arabic, would stumble in reading the text. It is not difficult to understand why writing is such a ubiquitous feature of Islamic art and architecture. There were already strong precedents in the Ancient Near East and the classical world for using written texts with the name of the patron to mark important sites and buildings. More importantly, the word plays a pivotal role in the religion of Islam. The central miracle of the faith is that in the early seventh century God sent down a revelation to the Prophet Muhammad, a revelation that was later written down as the Koran (literally ‘reading’ or ‘recitation’). The first words that God revealed to Muhammad were the opening five verses from Chapter 96, Surat al-ÆAlaq (‘The Clot’): Recite in the name of thy lord who created, Created man from a clot; Recite in the name of thy lord, Who taught by the pen, Taught man what he knew not. These verses are interpreted in various ways, either that writing was relatively new to Arabia or that man is able to learn from writing (that is, from reading books) what he does not otherwise know. Whichever interpretation is correct, these verses underscore the central role of writing (and by extension language) in Islamic culture and art. The central role of writing runs through the Koran. Chapter 68, another early revelation known as Surat al-Qalam (‘The Pen’) or Surat al-Nun (‘The Letter Nun’), opens with the words, ‘Nun. By the pen and what they write …’. Later commentators suggested that the pen was the first thing created by God so that he could write down events to come, and the sentence has inspired poets and mystics throughout the centuries. The Koran puts man’s whole life under the sign of writing. According to another revelation from slightly later, verses 17–18 of Chapter 50, Surat al-Qaf (‘The Letter Qaf’), two noble scribbling angels (kiråm kåtibÈn) sit on man’s shoulders to record his action and thoughts,

— 11 —

— ISLAMIC INSCRIPTIONS — the one on the right noting down his good deeds and the one on the left his evil ones. On Judgment Day, man’s every deed will be totted up in the Book of Reckoning for the final accounting (Chapter 69, Surat al-Haqqa, verses 18–19). Given the importance of writing in the revelation, it is no surprise that writing became such an important feature of Islamic culture. Books and book production became major art forms. Inscriptions became a major motif of decoration. The prolific use of inscriptions in Islamic culture has been noted at least since the eighteenth century by Western scholars, for they recognised the importance of inscriptions in dating objects and works of art. Inscriptions on coins were some of the first to be studied, perhaps because their decoration, unlike that on European coins, was exclusively epigraphic. In the early Middle Ages, Islamic coins had been traded extensively in Scandinavia, northern Germany and Russia, and the large collections in northern Europe stimulated scholarly interest. George Jacob Kehr’s monograph, Monarchiae Asiaticae-Saracenicae Status qualis VIII et IX ... seculo fuit, ex nummis argenteis prisca Arabum scriptura kufica … cusis, et nuper … effossis, illustratus, published in Leipzig in 1724, included correct readings of the Kufic inscriptions on the coins and commentaries on them. It has been called the first scholarly book on Islamic numismatics and on Islamic archeology in its widest sense. From this point on, catalogues of other major European collections of Islamic coins were based on an accurate reading of the inscriptions. Perhaps the most famous example of this group was C. M. Frähn’s classification of the coins in St Petersburg, published in 1821 as Das muhammedanische Münzkabinett des asiatischen Museum der kaiserliche Akademie der Wissenschaften zu St Petersburg. Along with interest in the inscriptions on coins came recording of the inscriptions on buildings. The German traveller and historian Carsten Niebuhr (1733–1815) was one of the first to do so. He joined the scientific expedition sent in 1761 by Frederick V of Denmark to explore Egypt, Arabia and Syria. Niebuhr was the sole

survivor of the rigorous six-year expedition, and the record of his journey became a classic on the geography, people, antiquities and archaeology of the lands he had traversed. The first volume, Beschreibung von Arabien, published in Copenhagen in 1772, was accompanied by abundant illustrations, including drawings of some of the inscriptions he had recorded. Two other volumes, published under the title Reisebeschreibung von Arabien und anderen umliegenden Ländern, appeared in 1774 and 1778, and a fourth was published posthumously in 1837. Niebuhr’s drawings of the inscriptions were some of the first Islamic inscriptions to be published, but they had been collected in a rather piecemeal fashion, and a next step was a more systematic survey of the inscriptions from a particular location. This was one result of Napoleon’s expedition to Egypt carried out between 1798 and 1801. Published in Paris between 1809 and 1828 as the Déscription de l’Égypte, the scientific record of the expedition was one of the greatest achievements of the encyclopaedic tradition of the French Enlightenment. The work comprised nine volumes of text, ten elephant folios of plates, and an atlas. Although somewhat of a hodge-podge, covering everything from Egyptian antiquities to popular music, it included reproductions of inscribed objects and monumental inscriptions, some now destroyed. The plates are still so useful that they were reprinted in 1994, almost two centuries after the original date of publication. They contain, among other things, the Kufic texts engraved on the Nilometer on Rawda Island (plates 744–5) and the long wooden frieze from the Mosque of Ibn Tulun in Cairo (plates 746–8). Scholars next broadened the field of enquiry to inscriptions on other works of Islamic art. The first catalogue of an entire collection of Islamic decorative arts, the two-volume Monumens arabes, persans et turcs, du cabinet de M. le Duc de Blacas et d’autres cabinets, published by the French orientalist J.-T. Reinaud in Paris in 1828, included a significant amount of material on inscriptions. It was the first work on Islamic seals and contains a general introduction to seals

— 12 —

— WHY READ INSCRIPTIONS? — and talismans and their inscriptions (see also Chapter 14). Since the collection also contained one of the most important examples of inscribed Islamic metalware, the Blacas Ewer in the British Museum (see Figure 9.53), the only inlaid brass on which the metalworking centre of Mosul is identified, the catalogue remains a landmark for the study of Islamic epigraphy. Several other scholars of Islamic epigraphy emerged in the early nineteenth century as well. The Viennese historian Joseph von HammerPurgstall (1774–1856) published several inscriptions, including the Kufic inscription on the Hakim Mosque in Cairo (1838), and his lecture on seals was published as a handbook in 1849. The Italian abbot Michelangelo Lanci (1779–1867) wrote a book on Arabic tombstones, published in 1840, and a three-volume work, Trattato delle simboliche rappresentanze arabiche e della varia generazione de’ musulmani caratteri sopra differenti materie operati, published in Paris in 1845–6, that treats various types of inscriptions on objects. The Austrian Joseph von Karabacek (1845–1918) wrote the first study of inscriptions on Islamic textiles, the Arabic inscriptions on the liturgical vestments of the Marienkirche in Danzig (1870). Karabacek recognised the importance of epigraphy in interpreting an object, and as the knighted director of the Hofbibliothek in Vienna and secretary to the Vienna Academy he had a wide impact on other scholars, although most of his own research was devoted to Arabic papyrology. It was only at the turn of the twentieth century, however, that the study of Islamic inscriptions was put on a scientific basis. The real founder of the subject was Karabacek’s Swiss contemporary Max van Berchem (1863–1921). Trained as an orientalist, especially a historian, and imbued with a deep knowledge and love of Islamic art, van Berchem became the unsurpassed master of the field of Islamic epigraphy. Upon returning from his first trip to Egypt, he began to put together the idea for a manual of Arab archaeology and a corpus of Arabic inscriptions. In a series of articles that appeared in the Journal Asiatique beginning in 1891, van Berchem developed his idea of Islamic archaeology as the study of its monuments in

the widest sense, including architecture, the decorative arts, inscriptions, numismatics and seals. By this he meant all the objects and documents, with the exception of manuscripts, that furnish some historical data, either by their forms or by the texts that they present. He laid out a double task for students of Islamic archaeology: to collect the inscriptions of Egypt and Syria so that they can constitute the basis of a corpus of Arabic inscriptions, and to deduce from the study of the monuments a manual of Arab archeology. Taking the first task upon himself, van Berchem persuaded the Académie Française to sponsor the series known as Matériaux pour un Corpus Inscriptionum Arabicarum (typically abbreviated to MCIA or simply CIA), and the first fascicule on the Arabic inscriptions from Egypt appeared in 1894. In the introduction, van Berchem enumerated three main reasons for studying inscriptions: their value for writing, language and history. The first two, their palaeographic and philological value, he judged less important. For van Berchem, the main reason for studying inscriptions lay in their value as historical documents, able to shed light on religious, political, administrative, judicial, military and commercial institutions, customs, ideas, and moral and material civilisation. Van Berchem recognised that while any single inscription could be analysed and could supply specific historical information, a corpus of inscriptions would allow the unusual to be discerned amid the standard. Noting that Arabic inscriptions were most common in a central area – the region bounded by the Mediterranean Sea, the Persian Gulf and the mountains of Persia, Armenia and Cilicia – he limited his Matériaux pour un Corpus Inscriptionum Arabicarum to this central zone, which was subdivided into four areas – Egypt, Syria, Asia Minor and Arabia. Van Berchem himself compiled the volumes on Cairo (1894–1903; later revised and expanded by Wiet, 1929–30) and Jerusalem (1920–2) and collaborated with Halil Edhem on a volume on eastern Anatolia (1910–17). Inscriptions from several cities in northern and southern Syria

— 13 —

— ISLAMIC INSCRIPTIONS — were collected by Sobernheim (1909) and Herzfeld, either in the form of articles (1942) or separate volumes of the corpus (1954–5). A volume on Mecca compiled by el-Hawary and Wiet was published in 1985 under the editorship of Nikita Elisséeff. The weighty tomes of the MCIA went far beyond van Berchem’s initial idea for a corpus listing Arabic inscriptions in chronological order, but forty years after he had proposed it, his original idea was realised with the publication of the Répertoire chronologique d’épigraphie arabe (commonly abbreviated to RCEA). Begun in 1931 under the editorship of Étienne Combe, Jean Sauvaget and Gaston Wiet, this multivolume work aims to amass a corpus of datable historical inscriptions in Arabic. It thus includes inscriptions on portable objects as well as those on buildings. Each volume of the RCEA contains 400 entries, arranged chronologically year by year and then geographically, west to east. The original editors of the RCEA envisaged that sixteen volumes would be sufficient to contain inscriptions up to the year 1250/1834, and the volumes appeared sporadically but steadily over the next three decades. With inscription no. 6400 at the end of volume 16, however, publication had only reached the year 762/1361, about halfway to the cut-off date originally planned. It was clear that the number of recorded inscriptions was far greater than the original editors had imagined and was growing rapidly. At this point, therefore, publication of the RCEA was halted to allow time for review of the project. The first addition, published in 1975, was a geographical index with an alphabetical list of sites mentioned in the first sixteen volumes. The original format was also amended and standardised, although the total of some 400 inscriptions per volume was maintained. To date, two volumes in the new format have appeared, covering inscriptions up to the year 800/1398. The RCEA, the basic reference work for dealing with Islamic epigraphy, is simply a list of inscriptions without commentaries. One of the few

attempts to synthesise this wealth of material is Gaube’s chapter on epigraphy in the Grundriß der Arabischen Philologie (1982). Given the subject of the volume, he limits his survey to inscriptions in Arabic and concentrates on form and vocabulary to the exclusion of style. He discusses seven types of monumental inscriptions as well as inscriptions on eleven different types of portable objects, giving the common Arabic forms used for each type. A table at the end of the essay gives an alphabetical list of Arabic phrases and a numerical list of Koranic citations keyed to the different types of inscriptions. Gaube’s essay is a useful and clear overview of the language used in Arabic inscriptions, but by its reliances on the RCEA it deals mainly with earlier material, and it is unillustrated. Van Berchem had already recognised that inscriptions would vary from region to region and designed his corpus to concentrate on the central Islamic lands. With the increase in knowledge about inscriptions, this regional focus has grown apace, and most works today treat the inscriptions of a particular area. This is clear from the article ‘Kitåbåt’ (‘Inscriptions’) in the second edition of the Encyclopaedia of Islam. The article begins with a general introduction to the subject by the doyenne of epigraphic studies in the mid-twentieth century, Janine Sourdel-Thomine. This is a thoughtful piece about the history of the subject, some of its advantages and disadvantages, and the general state of epigraphic study in the Islamic lands. This general introduction is followed by nine articles by specialists on the different geographical areas: Solange Ory on the Near East, M. Ocaña Jimenez on Muslim Spain, Lucin Golvin on north Africa, A. D. H. Bivar on west Africa, G. S. P. Freeman-Grenville on east Africa, F. Th. Dijkema and A. Alparslan on Turkey, L. Ch. Damais on south-east Asia, A. D. H. Bivar on Iran and Transoxiana, and J. Burton-Page on India. All the articles are useful for their bibliographies and references to individual studies, but the individual articles vary considerably in the quality and quantity of information presented. Some are extremely schematic, others

— 14 —

— WHY READ INSCRIPTIONS? — thoughtful and synthetic. This unevenness results not only from the interests and expertises of the authors, but also from the varying nature of the data and the different questions raised by the material from distinct regions. The increasing trend towards regional studies is also clear from the main bibliographic source on Islamic art in general and on inscriptions in particular, K. A. C. Creswell’s A Bibliography of the Architecture, Arts and Crafts of Islam. The original volume, published under Creswell’s editorship in 1961, included some 500 references (columns 675–714) to articles on monumental inscriptions in the section entitled ‘Calligraphy & Palaeography: Specimens on Stone, &c.’. This arrangement was continued in the first supplement (1973), which included some seventy-five additional entries (columns 215–22) under the same rubric. By the second supplement, which appeared under the editorship of Pearson, Meinecke and Scanlon in 1984, the field had grown so much that the editors changed the heading to calligraphy and epigraphy and subdivided the 700 entries in the section entitled ‘Inscriptions on Stone, etc.’ (columns 325–55) under regional headings. These include not only recent titles but also many that had been omitted from the first two volumes of the bibliography. These works form the standard references for the field of Islamic epigraphy. This book draws upon them, but is somewhat different. It is designed to present the subject to a wide spectrum of people who want to learn about Islamic inscriptions on buildings and objects. It introduces the topic to beginners in the study of Islamic civilisation and history. It shows historians who know texts where inscriptions are recorded and how inscriptions can supplement or even correct the information found in documents. It tells museum curators who possess inscribed objects how to read the inscriptions and what kind of information the inscriptions might provide. It provides archaeologists and researchers with a basic standard on how to record inscriptions. Finally, it includes a lengthy bibliography that provides directions for further reading on many aspects of Islamic epigraphy.

This book is divided into four parts, of which this introductory chapter forms Part I. Parts II and II are interpretive essays covering the two major types of inscriptions found in the Islamic lands: monumental inscriptions and inscriptions on portable objects. The chapters in each part treat the development of content as well as style and include specific examples and illustrations. The techniques for putting inscriptions on buildings and objects and the reasons for doing so differed. Hence, the two parts are organised differently. Part II, on monumental inscriptions, is divided into chapters dealing with different approaches and methodologies used to study the vast corpus of monumental inscriptions, including linguistic, typological, geographical and stylistic considerations. By contrast, Part III, on portable objects, is divided by media (metalwares, woodwork, ceramics, textiles and other arts) and by types of objects (tombstones and other grave markers; arms and armour; and seals, talismans, amulets and other small ornaments). The chapters in both parts introduce the reader to the general literature and the work that has been done. They gives examples of the kinds of information that can be gleaned from the inscriptions. They also point out areas where more work needs to be done (or redone). The reader may be surprised to find that two topics are not included in this handbook of Islamic epigraphy: inscriptions on coins and inscriptions on paper and other written surfaces. This is not to deny the importance of these topics; rather, the reverse. Both topics are so important that they demand separate studies. As the inscriptions on coins and paper offer information about other kinds of Islamic epigraphy, these two subjects are mentioned tangentially in these chapters, and a few general notes about how to use numismatic epigraphy and calligraphy are worthwhile here. Inscriptions are ubiquitous on Islamic coins. Most coins issued in the Islamic lands are exclusively epigraphic. Pictures, images and other symbols are the exception rather than the rule. Since coins are almost always dated, numismatic epigraphy is a useful source for dating the introduction of various titles and scripts, particularly

— 15 —

— ISLAMIC INSCRIPTIONS — on other undated material. The interlaced Kufic script found on slip-painted ceramics associated with the Samanids in tenth-century Central Asia and exemplified by the bowl illustrated in Figure 1.3, for example, can be dated in part based on the introduction of interlaced script on coins. When using numismatic evidence for dating stylistic innovations in epigraphy, it is essential to bear in mind that coins are exceedingly conservative. No-one issues a coin that might be rejected. Hence, the date when any new style or motif appears on a coin must be considered the date of common acceptance, not that of innovation. Another factor to remember in using numismatic evidence for epigraphy is that coins represent the official view. Since they were issued by the state, coins are important evidence in arguments about how bureaucracies purposefully manipulated styles of script. For example, scholars such as Tabbaa (1991, 1994) have speculated about the relationship between the introduction of cursive scripts into epigraphy and attempts to buttress Islamic orthodoxy, but coins are often overlooked as evidence in the argument. Cursive script was used on coins issued in eastern Iran and Central Asia from an early period. A dirham minted at Balkh in 292/904–5, for example, has the name of the Banajirid governor Ahmad b. Muhammad b. Ahmad written in a cursive hand. Similarly, a coin issued at Nishapur in 305/917–18 uses cursive script for the name of the Samanid ruler Nasr b. Ahmad. As elsewhere, cursive hands were probably used for ordinary business correspondence, but the use of cursive on coins to highlight the name of the local ruler shows that cursive was accepted for official purposes in Central Asia at a very early date. Many times there is also a close connection between styles of handwriting and styles of epigraphy. Sometimes the same artists were responsible, and calligraphers designed inscriptions on buildings and other objects. This was probably the case already in early Islamic times, as attested by the inscriptions on the Dome of the Rock and on milestones issued in the name of the Umayyad caliph ÆAbd al-Malik (see Chapter 3). It was certainly the case in later times, for some

inscriptions are signed by a person who calls himself al-kåtib (‘the scribe’). Endowment inscriptions on buildings are usually abridgements of longer endowment deeds that were registered at the judiciary. In some cases, inscriptions on dated monuments can help us date undated manuscripts, and monumental inscriptions are often used to date the large body of fragmentary parchment manuscripts of the Koran made in the early centuries of Islam. As with coins, however, the reader must be careful in using epigraphic evidence to date calligraphy and styles of handwriting. Reasons for writing a document were different from those for erecting an inscription, particularly a monumental one. Documents and inscriptions may have been intended for different audiences. Each had its own tradition, and traditional rules or customs may have affected common practice. The different media also placed different demands on the artist. In general, it is far easier and faster to write a text with a reed pen than to inscribe it, be it on a building or an object. Indeed, the texts that are closest to calligraphy are the ones inscribed on medieval Iranian ceramics, which were painted with a brush. Part IV (Chapter 15) of book is more practical. It begins by showing the reader how to use the basic reference materials. These include the Matériaux pour un Corpus Inscriptionum Arabicarum (MCIA) and the Répertoire chronologique d’épigraphie arabe (RCEA) for historical texts. It also shows how to find, identify and interpret other types of inscriptions, including Koranic texts, hadith and pious phrases, and dates. The second part of Chapter 15 shows how to record inscriptions and what conventions are commonly used. The bibliography does not claim to be comprehensive, for just as there are too many inscriptions to be mentioned individually, there are also too many studies to be cited singly. Rather, the bibliography contains the most important and the most accessible studies. The criteria for inclusion were usefulness and availability. Longer and fuller bibliographies can be found in Creswell’s bibliography on the arts and crafts of

— 16 —

— WHY READ INSCRIPTIONS? — Islam and its various supplements. To facilitate reading, only short references giving the author’s name and appropriate pages, catalogue number or illustration will be included in the text, but the interested reader can find the full citation in the bibliography. Finally, a long and detailed index should help the reader, who is invited to skip about and jump from chapter to chapter. The field of Islamic epigraphy is vast. Thanks to photographs and books, we know far more about it today than any patron, scribe or epigrapher did in earlier times,

when knowledge was probably limited to what a person could see around him. What pertains at any one time or place may not hold at another. Nevertheless, the reader may draw inspiration and method from other contexts. This book is intended to be both an introduction and a provocation: to encourage the reader not only to look at and read inscriptions, but also to analyse and interpret them in order to understand better the rich material and visual world of the Islamic lands over the past 1,400 years.

— 17 —

— CHAPTER 2 —

The Languages Used in Monumental Inscriptions

A

rabic was the standard language for writing monumental inscriptions in the Islamic lands. Arabic had been used for writing monumental texts in Syria several centuries before the coming of Islam. The oldest inscription known in the Arabic language (RCEA 1) is written on a basalt lintel from the site of al-Namara, the fortified post of Safa located some 120 km south-east of Damascus. The five-line funerary text records that the building was the tomb of the Arab prince Imral-Qays, who died in 328 CE; it is written in Arabic words spelled out in Nabatean characters. The oldest examples of Arabic writing date from the fourth century CE. They include a trilingual inscription in Arabic, Greek and Syriac dated 512 CE from Zabad in the Syrian steppe south-east of Aleppo (RCEA 2) and three inscriptions dating between 528 and 568 CE found to the south and south-east of Damascus (RCEA 3–4; Grohmann 1971: 15f.). Arabic continued to be the major language of monumental inscriptions in Islamic times. As the language of the revelation, it acquired special sanctity, and it was always the most common language for religious texts. These include not only citations from the Koran, but also hadith and other pious invocations. Arabic was also the major language for the historical part of the inscription. The first foundation inscription to survive, the record of a dam ordered by the first Umayyad caliph MuÆawiya at Taæif in the Hijaz in the year 58/677–8, comprises six lines of Arabic scratched in the rock wall of the dam (Miles 1948). Foundation inscriptions in Arabic soon assumed a more monumental form. Under the fifth

Umayyad caliph ÆAbd al-Malik (r. 685–705), Arabic replaced Greek and Persian as the standard language of administration and taxation. Arabic inscriptions in the caliph’s name survive on coins, milestones and buildings from the year 72/691–2 onwards, and all use Arabic written in a similar form of the angular script, commonly if misleadingly known as Kufic (Blair 1992b). The letters are well proportioned and well spaced, and attention was paid to making the script both legible and imposing. Both the form of the text and the style of the writing were followed in other inscriptions set up in Syria under the Umayyads, as at Qasr al-BurquÆ in 81/701 (Gaube 1974) and Qasr al-Hayr West in 107/727 (RCEA 27). From early times, Arabic was the main language for inscriptions not only in the heartland of Islam but in other regions as well, particularly for monumental inscriptions prepared under the auspices of the caliphal chancelleries. Relatively few monumental inscriptions survive from the earliest centuries of Islam, and most of them are in greater Syria. Those from other regions have been destroyed, partly because of the less durable materials used, but written reports incorporated by medieval authors suggest that these nowdestroyed inscriptions were similar to those that have survived in Syria. The tenth-century scholar al-Jahshiyari (d. 331/942), for example, records several inscriptions in standard Arabic form on the gates to Acre and Sidon (RCEA 37) and on a treasury in Azarbayjan (RCEA 43). In a few isolated areas in Iran, notably the mountainous region south of the Caspian Sea, older traditions were preserved into Islamic

— 21 —

— ISLAMIC INSCRIPTIONS —

2.6 Tomb tower at Lajim in northern Iran, with foundation texts in Pahlavi at the top and in Arabic below, late tenth or early eleventh century.

— 22 —

— THE LANGUAGES USED

IN

times, and Pahlavi continued to be used there alongside Arabic for monumental inscriptions. This preservation of Pahlavi is not surprising, for many Sasanian traditions remained strong in the area. A few metalwares from the eighth century are inscribed with the name of a local ruler of Mazandaran, and Pahlavi and Arabic were used together on three tomb towers (see Figure 2.6) erected in the region in the tenth and eleventh centuries (Herzfeld 1933; Blair 1992a, nos 31, 32, 79). Generally, the Pahlavi text repeats that in the Arabic, but gives the date in the Yazdigird rather than the hegira calendar (see Chapter 15 for further details on the various calendars used in inscriptions). In medieval times, Persian came to be used for some monumental inscriptions in Iran. Surviving commemorative texts, such as the ones carved

M O N U M E N TA L I N S C R I P T I O N S — by the Buyids at Persepolis, show that in the tenth and eleventh centuries Persian names, dates and other words were introduced slowly. The first Persian inscription to survive on a datable building is at the tomb at Safid Buland in the northern part of the Farghana Valley in Uzbekistan (see Figure 2.7). Recently deciphered by Nastich and Kochnev, the inscriptions include a Persian poem at the top of the zone of transition in mu∂åriÆ akhrab makfËf ma˙dËf metre rhyming in åft that names the deceased, the Qarakhanid Muhammad b. Nasr, and a partly destroyed foundation inscription in Persian at the top of the walls that names the patron, his son MuÆizz al-Dawla ÆAbbas. The inscriptions show that the tomb dates from the years 1055–60, a century earlier than was previously thought (Blair 1992a, no. 47). The Qarakhanids were probably not the

2.7 Interior of the tomb at Safid Buland in the Farghana Valley, Uzbekistan, with the first datable inscriptions in Persian, 1055–60.

— 23 —

— ISLAMIC INSCRIPTIONS — first dynasty to use Persian in monumental inscriptions. Their predecessors in Khurasan and Transoxiana, the Tahirids (821–73), the Samanids (819–1005) and the Saffarids (867–c. 1495), were known for their revival of Persian literature, and their now-lost buildings may well have had Persian inscriptions. Inscriptions in Persian continued to be used on buildings erected by the Qarakhanids and others in the eastern Islamic lands from the late eleventh century onwards. One example is the caravanserai known as Ribat-i Malik in the Zarafshan Valley between Samarqand and Bukhara, which was restored by the Qarakhanid Nasr b. Ibrahim in 471/1078–9. The inscription over the portal seems to be a Persian poem rhyming in åy and describing the construction of the building as a palace worthy of Paradise (Blair 1992a, no. 58). The palace built by the Ghaznavid ruler MasÆud III (r. 1099–1115) at Ghazna had a lengthy Persian poem around the courtyard that extolled the ruler and his palace (Bombaci 1966). The northern tomb at Uzgand is another Qarakhanid building with a foundation inscription in Persian. The cursive inscription on the arch of the portal is a Persian sentence saying that contruction of this palace (dawlatkhåna) began (åghåz karda åmad) on Thursday the 4th of RabiÆ II of the year 547 (9 July 1152), and the text over the door gives the name and titles of the Qarakhanid Husayn b. Hasan b. ÆAli Mujtabi (Yakubovsky 1947). This is also the time when inscriptions in Persian begin to appear on dated metalwares made in the eastern Iranian lands (see Chapter 9). Persian was the literary language of the eastern Islamic lands, and Persian verses were inscribed on both buildings and other objects made in a wide area from the thirteenth century onwards. Persian verses occur on several buildings from Saljuq Anatolia (Meinecke 1976: 81–2). The earliest known was reportedly composed by Sultan Kaykaæus for his tomb in the Shifaiye Madrasa in Sivas shortly before his death in 1220. The Sircali Madrasa in Konya, built in 640/1242–3 under Kaykhusraw’s son Kayqubad, has a Persian verse in praise of the builder

Muhammad al-Tusi. Several verses by Firdawsi decorated the tomb of Sahib Ata in Konya constructed in the late thirteenth century as well as stone buildings erected in Niksar and Tokat in the fourteenth century. Similarly, a lustre frieze with verses from the Shåhnåma was used to decorate the palace erected at the end of the thirteenth century by the Mongol ruler Abaqa at Takht-i Sulayman in north-western Iran (Melikian-Chirvani 1984, 1988, 1991). Despite the common use of Persian verses, either taken from literature or composed for the occasion and eulogising the patron or his works, Arabic continued to be the standard language for foundation texts and signatures. The difference between Persian verses and the Arabic foundation text is clear from the palace built by MasÆud III at Ghazna. In addition to the long Persian poem around the courtyard, the Italian excavators of the site discovered several marble screens with an Arabic inscription saying that it was the work of Muhammad b. Husayn b. Mubarak finished on 1st of Ramadan of the year 505 (3 March 1112). In Ghaznavid times, a stylistic distinction was also made between the different types of text and the two languages used for them, for the lengthy Persian poem is written in floriated Kufic while the brief and straightforward Arabic foundation text is in a readable naskh. Tradition and sanctity are two common explanations for the persistence of Arabic in monumental inscriptions in Iran. Legibility may be another reason for the slow introduction of Persian into monumental epigraphy. Most inscriptions from the early period are written in the angular form of script known as Kufic and have few or no diacritical marks or dots to distinguish letters of the same shape. The root system in Arabic, where most words follow standard forms based on triliteral roots, makes it possible, thought difficult, to read such undotted texts. Persian, by contrast, is an Indo-European language and has no such roots and standard forms to help the reader. Its looser grammatical form, which gives rise to wonderful plays on words in poetry, makes it much more difficult, even impossible, to read undotted Persian texts. Many Persian

— 24 —

— THE LANGUAGES USED

IN

names written in undotted Kufic on early Islamic buildings from Iran, for example, cannot be reconstructed. Persian inscriptions become more common when cursive styles of writing, in which diacritical marks were standard, were introduced. Even after the introduction of Persian into monumental epigraphy, Arabic remained the main language for foundation inscriptions on buildings constructed in Iran. During the Timurid period, foundation inscriptions and pious phrases are in Arabic, while Persian was used for poetry, some composed for the occasion and some from well-known poets. The distinction can be seen on several buildings from Khurasan, such as the Madrasa-yi Do Dar and Masjid-i Shah in Mashhad and the congregational mosque in Nishapur (CII IV). Similarly, the Darb-i Imam, a shrine built in Isfahan during the time of Qara Qoyunlu rule in the mid-fifteenth century, has an Arabic foundation inscription around the portal, but Persian poems on the doorway and interior that were written for the occasion (text in Hunarfar 1977: 341–53; English translation in Golombek and Wilber 1988, no. 170). Both poems contain many commonplace Sufi images and are of rather mundane quality. They refer to the specific parts of the building on which they are inscribed as well as to the historical situation, naming the Qara Qoyunlu ruler Jahanshah and giving the date 857/1453. Even under the Safavids, poetry inscribed in buildings in Iran could still be written in Arabic. The inscriptions on the Masjid-i Shaykh Lutfallah, erected in Isfahan in the early seventeenth century, for example, are mainly in Arabic. They include not only the lengthy foundation inscription over the entrance portal, the signature of the calligrapher and the date but also many bands and panels with Koranic texts, hadith and poems on the interior. One of the few Persian inscriptions is the signature of the builder on the tile mosaic mihrab in the interior, which is a rather stilted Persian form of a standard Arabic phrase (Hunarfar 1977: 401–15). In contrast to Persian, Turkish remained a lingua franca for a longer time. The first example of a monumental inscription in Turkish to survive occurs in the madrasa built by the

M O N U M E N TA L I N S C R I P T I O N S — Germayanid YaÆkub Çelebi at Kutayha in 814/1411 (Uzunçar∞ılı 1932). Turkish appears in the monumental inscriptions at Edirne in the mid-sixteenth century (Dijkema 1977, nos 28 and 29), and by the end of the century had become the standard language for foundation inscriptions in Turkey. The change may have been the result of the increased prestige of the Turkish language following the Ottoman conquests of Arabia and elsewhere in the early sixteenth century. Arabic was still used occasionally for monumental inscriptions in Turkey, particularly those in mosques, as in the foundation inscription on the portal of the Selimiye mosque at Edirne (see Figure 2.8). Persian was also used on public taps and tanks (∞ådirvån), although most are in Turkish (Aynur and Karateke 1995). The Tophane fountain (see Figure 2.9) built by Mahmud I (r. 1730–54), for example, has Turkish verses on the sides over the basins. Many of these Ottoman inscriptions, particularly the Turkish but also some of the Arabic ones, were written in verse. The verses were set in cartouches, and the shape of the cartouche can be used as a clue to dating. In these poetic inscriptions, the date is often given in a chronogram at the end. In the foundation inscription on the Selimiye Mosque, for example (Dijkema 1977, no. 37), the final couplet reads ßåra fa∂l allåh taærÈkh al-asås fa∂l yazdån kåna taærÈkh al-tamåm (‘[The phrase] “The grace of God” came to be the date/chronogram of foundation, and [the phrase] “The Grace of the Good Lord” was the date/ chronogram of completion’). The dates are written in ciphers at the upper end of each hemistich, ‘the year 976’ (corresponding to 1568–9) and ‘the year 982’ (corresponding to 1574–5). Many of the Turkish inscriptions are extremely difficult to read. Not only is the language intricate, but words are occasionally written phonetically or in an unusual order. Calligraphic flourishes and decorative arrangements also hinder legibility. Unusual interlacing and ligatures, for example, could be added between letters, and words could be written in mirror-script. Arabic was also the predominant language in east Africa. With the exception of a scant handful

— 25 —

— ISLAMIC INSCRIPTIONS —

2.8 Foundation inscription in Arabic over the portal to the Selimiye mosque at Edirne, giving chronograms with the dates of foundation (976/1568–9) and completion (982/1574–5).

— 26 —

— THE LANGUAGES USED

IN

M O N U M E N TA L I N S C R I P T I O N S —

2.9 Detail of the foundation inscription in Turkish on the fountain near Tophane in Istanbul, early eighteenth century.

— 27 —

— ISLAMIC INSCRIPTIONS — of inscriptions in Swahili or Portuguese, almost all others are in Arabic. They include a finely carved dedicatory inscription in the congregational mosque at Kizimkazi, Zanzibar (Flury 1922) and sixteen dedicatory inscriptions commemorating the foundation of mosques in Lamu, Kenya between the fourteenth and the nineteenth centuries (Freeman-Grenville and Martin 1973). The sanctity of Arabic script is clear from inscriptions in other areas, and inscriptions written

in other languages but using Arabic script are also recorded in South-east Asia during the later centuries of Islam. In Java and Sumatra, for example, inscriptions were usually written in the Javanese language, sometimes in Javanese characters but sometimes in Arabic characters. In eastern Indonesia, some nineenth-century texts were written in Arabic script but in the language of Ternate, from the sultanate of the same name.

— 28 —

— CHAPTER 3 —

Foundation Inscriptions

B

y far the most common type of monumental inscription is the foundation or restoration inscription. This type of inscription contains five basic elements in the following order: (1) the basmala or invocation to God, (2) a verb indicating what was done, (3) the object of the work, (4) the name of the patron and (5) the date of construction. These elements were already standard in the first monumental inscriptions set up under the Umayyads at the end of the seventh century and the beginning of the eighth, and continued to be so over time as inscriptions became longer and more complicated. Let us look briefly at each of the five elements in turn, using three examples of foundation inscriptions from different periods and places. The earliest (see Figures 3.10 and 3.11) is the mosaic band running around the outer face of the ambulatory around the Dome of the Rock (RCEA 9). It now contains the name of the ÆAbbasid caliph al-Maæmun, who repaired the building in 216/831 according to several other inscriptions on copper plates nailed to the wooden lintels of the interior doors (RCEA 209–10), but van Berchem (MCIA Jerusalem, no. 215) showed that the original foundation inscription had contained the name of the Umayyad caliph ÆAbd al-Malik, who ordered the building in 72/692. We can contrast this foundation and restoration inscription from the caliphal period to two later ones: a medieval example (see Figures 3.12 and 3.13) running along the east façade of the funerary complex erected by the Mamluk sultan Qalaæun on the main street of Cairo in the late thirteenth

century (RCEA 4852) and a later example (see Figures 3.14 and 3.15) running around the portal of the mosque of Shaykh Lutfallah erected by the Safavid shah ÆAbbas I on the east side of the maidan in Isfahan in the early seventeenth century (Hunarfar 1977: 401–15). As these are all major monuments erected by important rulers, they illustrate some of the changes that occurred in foundation inscriptions on royal buildings erected across the Islamic lands from earliest times to the later period. Most monumental inscriptions, like most actions carried out by devout Muslims, open by invoking God’s name. The most common form is the full basmala, ‘In the name of God the Merciful, the Compassionate’ (bism allåh alra˙mån al-ra˙Èm), but the invocation can be shortened to the first two words or lengthened with other, sometimes rhyming phrases. The form used often depends on the space available, and the short form was common in tight spaces. There were also regional variations or variations according to the type of building on which the basmala was used. For example, the phrase aÆËdhu bæillåh min al-shay†ån al-rajÈm (‘I seek refuge with God from Satan the cursed’) was often put before the basmala in foundation inscriptions in North Africa. This phrase probably reflects the conservative leanings of patrons there, who usually belonged to the Maliki school of law. This form was also common on madrasas throughout the Islamic lands and underscored the role of these buildings in disseminating the faith. In reading a foundation inscription, it is important to find the basmala, for it usually

— 29 —

— ISLAMIC INSCRIPTIONS —

3.10 Detail of the inscription band around the outer face of the arcade in the Dome of the Rock, Jerusalem, with the name of the patron.

3.11 Copy of van Berchem’s drawing of the patron’s name in the outer face of the arcade in the Dome of the Rock. The area between the dots is al-Maæmun’s restoration.

marks the beginning of the text. This is obvious on a frame band that runs along a façade or around a portal. In the case of the inscription on Qalaæun’s funerary complex in Cairo, for example, the text begins on the right corner of the façade beneath the minaret. In the mosque of Shaykh Lutfallah, the basmala and the foundation inscription begin on the right edge of the portal. The beginning of a text on a circular building is less obvious, but finding it shows how the building was planned and oriented. On the Dome of the Rock, for example, the basmala shows that both inner and outer bands of inscription begin on the south, or qibla, side. Identifying the beginning

of a circular text also means that one can then examine the words to the right of the basmala for the date, the piece of information usually put at the end of the foundation inscription and the one most sought by historians. The second element of a standard foundation text is the verb. There are many different verbs in Arabic for building. The simplest is banå and this is the one used in the inscription on the Dome of the Rock. Most foundation inscriptions, however, do not talk about constructing but about ordering construction, and hence the most common verb is amara (‘to order’). It could be used alone in this form (‘he ordered’), but more

— 30 —

— F O U N D AT I O N I N S C R I P T I O N S —

3.12 Foundation inscription across the façade of Qalaæun’s tomb complex in Cairo, 682–4/1283–5.

— 31 —

— ISLAMIC INSCRIPTIONS —

3.13 Detail of the foundation inscription across the façade of Qalaæun’s tomb complex.

often was compounded with words like bi-iqåma (‘the erection’), bi-binåæ (‘the construction’), bibunyån (‘the building’), bi-Æamal (‘the making’), bi-Æimåra (‘the building’) or bi-inshåæ (‘the establishment’). This later form became increasingly common in medieval times and is the one used on both Qalaæun’s complex and the Mosque of Shaykh Lutfallah. Other common forms were mimmå amara [bihi] (literally, ‘part [of what] he ordered’) or hadhå må amara (‘this is what he ordered’). Gaube gives a brief survey of the different forms of amara used in different places, but much more work needs to be done compiling lists of the standard forms used at different times in different areas, particularly by different classes of patron or on different kinds of buildings. Restoration texts generally used the same form as foundation texts. They can ape foundation texts. For example, the inscriptions on copper plates marking al-Maæmun’s restorations to the Dome of the Rock use the form mimmå amara bihi. Restoration inscriptions can also replace the verb amara with jaddada (‘renew’). Sometimes the specific work that was done is indicated, as in a restoration text recording the repanelling of two pillars in the Great Mosque of Damascus in 575/ 1179–80 by the Ayyubid ruler al-Malik al-Nasir Yusuf (RCEA 3343), where the inscription specifies that the ruler ordered the two pillars to be revetted in marble (jaddada tarkhÈm hadhah [sic] al-ruknayn). The verb in a foundation inscription is usually followed by the object constructed. This can be simply the pronoun huwa (‘it’), leaving it to the reader to interpret what ‘it’ is. Al-Maæmun’s restoration inscriptions use this generic pronoun. More often, however, the type of work was specified. Shah ÆAbbas’s foundation, for example, is called a mosque (masjid). Other well-known types of buildings include a congregational mosque (jåmiÆa), tower (burj), bath (hammåm), fountain (sabÈl) and the like. Sometimes, the foundation inscription could enumerate the various parts of a building complex. The foundation inscription on Qalaæun’s complex, for example, tells us that the complex includes not only the tomb but also a theological school (madrasa) and a hospital (bimaristån).

— 32 —

— F O U N D AT I O N I N S C R I P T I O N S —

3.14 Foundation inscription around the portal of the Mosque of Shaykh Lutfallah in Isfahan, dated 1012/1603–4 (after restoration).

— 33 —

— ISLAMIC INSCRIPTIONS —

3.15 Detail of the end of the foundation inscription around the portal of the Mosque of Shaykh Lutfallah, showing the date 1012/1603–4 and the signature of the calligrapher ÆAli Riza ÆAbbasi.

— 34 —

— F O U N D AT I O N I N S C R I P T I O N S — The nouns used in a foundation text can indicate either a form or a function, and the modern reader needs to interpret these words, for many of them changed in meaning over time. A particularly wide range of different terms was used for funerary monuments, including maqåm (‘resting place’), turba (‘tomb’), qabr (‘grave’), qubba (‘dome’), mashhad (‘martyrium’), qaßr (‘palace’) and even the Persian khwåbgåh (‘resting place’) as well as many others. Any definition of a particular term applicable at a particular time or place does not necessarily pertain at other times or places. This is true, for example, of the word qubba, used in both ÆAbd al-Malik’s inscription on the Dome of the Rock and Qalaæun’s inscription on his funerary complex. On the Dome of the Rock, qubba seems to denote a form, a domed building, while on Qalaæun’s funerary complex it denotes the function that took place in that form, a domed tomb. Similarly, several different words could also be used for what we today consider the same thing. The words miædhana (literally, ‘place of the call-to-prayer’), ßawmaÆa (also ‘hermitage’ or ‘monk’s cell’) and manåra can all designate what we call minaret, and the word manåra may have designated a lighthouse as well as a tower. In order to understand correctly what was meant, the reader must look at how these words were used in the specific time and place that the foundation inscription was set up. At first nouns were used alone in foundation inscriptions, but, by the eleventh and twelfth centuries, adjectives were commonly used to qualify what was built. Thus, in the foundation inscription on the Dome of the Rock, the noun qubba is used alone, whereas in the two later foundation inscriptions, each noun is qualified by an adjective. Qalaæun’s tomb is ‘noble’ (sharÈfa), his madrasa ‘great’ (muÆaΩΩama) and his hospital ‘blessed’ (mubårak). Shah ÆAbbas’s mosque is also ‘blessed’ (mubårak). These adjectives were not physical descriptors but rhetorical ones. The most common term is mubårak (‘blessed’), followed by other similar phrases such as sharÈf (‘noble’), ÆaΩÈm (‘great’) and ÆålÈ (‘elevated’). Their function was to made the building acceptable as waqf, technically the alienation in perpetuity of

property and the revenues from it for the benefit and endowment of a pious foundation, for according to Islamic tradition all religious endowments had to be pleasing to God. In a typical foundation inscription, far more space was given over to the patron than to what he built, and from the earliest inscriptions, the patron’s name was accompanied by titles and eulogies. Over the centuries, this part of the foundation inscription grew by leaps and bounds and was composed of numerous rhyming phrases that intoned the glory of the patron in ringing verse. Our three examples show this change clearly. As reconstructed by van Berchem, the foundation inscription on the Dome of the Rock records that the building was erected by God’s servant ÆAbd al-Malik, the Commander of the Faithful, and a short eulogy follows his name. The explosion in names and titles is clear when we compare this modest text on the Dome of the Rock to that on the funerary complex erected by the Mamluk sultan Qalaæun six centuries later. The bold inscription band runs along the east façade of the complex at midheight overlooking the Qasaba, the main street of medieval Cairo. The text, which begins in the north-east corner under the minaret, runs along the qibla wall of the tomb, continues across the entrance portal to the corridor linking the various parts of the complex, and then finishes along the main façade of the madrasa. The Mamluk ruler is called our lord and master, the greatest sultan, al-Malik al-Mansur, the wise, the just, the one who is aided by God, the victorious, the fighter, the triumphant, Sword of the World and the Faith, Sultan of Islam and the Muslims, lord of kings and sultans, sultan of the earth in longitude and latitude, king of the world, sultan of the two ÆIraqs and the two Egypts, king of the two continents and the two seas, inheritor of royalty, king of kings of the Arabs and Persians, possessor of the two qiblas, keeper of the two noble sanctuaries, Qalaæun al-Salihi, partner of the Commander of the Faithful. A five-part eulogy following his name asks God to prolong his greatness and glory. Shah ÆAbbas’s name and titles on the Mosque of Shaykh Lutfallah are different but almost as

— 35 —

— ISLAMIC INSCRIPTIONS — grand. He is the greatest sultan and noblest khåqån, reviver of the virtues of his pure fathers, propagator of the religious sect of the infallible imams, Abuæl-Muzaffar ÆAbbas al-Husayni alMusavi al-Safavi Bahadur Khan, and a benediction asks God to make his kingdom endure and his ships sail in the seas of eternity. Both Islamic nomenclature and Islamic titulature are complicated topics, and in order to compare the different titles and eulogies used in these inscriptions, a few short notes on the composition of Islamic names are useful. Islamic names generally have five elements in a standard order: (1) the laqab, an honorific title; (2) the kunya, a patronymic; (3) the ism, a given or personal name; (4) the nasab, a genealogy; and (5) the nisba, an epithet of origin or affiliation. In addition to these standard parts of the name, rulers and other important people were often identified by various kinds of titles. The central element is the person’s given or personal name (ism), which is included in almost all foundation inscriptions. It was traditionally drawn from a relatively small repertory of personal names sanctioned by Islamic tradition. Examples include names and epithets of the Prophet, his companions, Biblical figures who appear in the Koran, and compounds using the name of God. The Safavid shah ÆAbbas, for example, took his name from the Prophet’s uncle ÆAbbas b. ÆAbd al-Muttalib, as did the ÆAbbasid dynasty, who were descendants of his son ÆAbdallah. In the inscription on the Dome of the Rock, the Umayyad caliph’s given name is ÆAbd al-Malik, literally meaning ‘slave of the king’ and signifying that the caliph was God’s servant. The ÆAbbasid caliphs adopted regnal names of a theocratic nature, expressing dependence on God, reliance on Him, or participation with Him in the work of ruling. Thus, the repairs to the Dome of the Rock carried out in 216/831 were done by the caliph al-Maæmun (literally, ‘he who trusts in God’). With the increasing importance of Turks and Persians, given names in those languages also appear. The name Qalaæun is Turkish, although no-one is sure what, if anything, it means.

Two parts of the name follow the ism. In standard form, the given name is followed by the genealogy (nasab), with each forebear introduced by the word ibn (‘son’) or bint (‘daughter’). In transcription, these two words are often abbreviated as b. and bt. When the patron derived authority or legitimacy from his family, then the genealogy could be quite long. This was typically the case in funerary texts. In foundation inscriptions, by contrast, the genealogy is quite short, usually no more than one or two generations. None of our three examples of foundation inscriptions contains a genealogy, for none of these rulers derived much authority from his family ties. Qalaæun was a manumitted slave, and hence his paternity was irrelevant, although his descendants derived legitimacy from his. In the case of ÆAbbas, his genealogy is subsumed in his nisba. The nisba, the final element in a name, was an epithet denoting the origin or place of residence derived from the name of a tribe, town or country. It could also indicate affiliation to a legal school or religious group or sometimes a profession. Qalaæun carries a single epithet, al-Salihi, meaning that he had received training under and been manumitted by the Ayyubid sultan al-Salih Najm al-din Ayyub. ÆAbbas, by contrast, carries three. He is al-Husayni, signifying that he claimed descent from the Prophet’s grandson Husayn; al-Musavi, signifying that he was also descended from the Musavi line of the seventh ShiÆite imam Musa alKazim (d. 183/799); and al-Safavi, signifying that he belonged to the Safavid line of Sufis descended from Safi al-Din Ardabili (d. 735/1334). Two parts of the name precede the ism. The one just before it is usually the kunya, a patronymic usually comprising a compound with Abu (‘father’). It was originally composed with the name of the eldest son, but was then constructed with an attribute or quality of the bearer and assumed a metaphoric meaning. Thus ÆAbbas is Abuæl-Muzaffar (‘Father of the Victorious’). In foundation inscriptions, these four parts of the ruler’s name (kunya, ism, nasab and nisba) are often quite succinct; much more space is taken up by titles, which became increasingly flowery. In the inscription on the Dome of the

— 36 —

— F O U N D AT I O N I N S C R I P T I O N S — Rock and on coins issued in his name, ÆAbd alMalik is entitled Æabd allåh, literally meaning ‘God’s servant’. This was the standard title used by Umayyad rulers. The ÆAbbasids continued to use it, but al-Maæmun added another title, imåm (literally, ‘leader’), as in the inscriptions signalling his repairs to the Dome of the Rock. With the break-up of power in the tenth century, many new titles were awarded or assumed. One of the most important was sultan. The word sul†ån originally had the abstract sense of power or authority, but it was adopted as a title by the Saljuqs and Ghaznavids in the eleventh century. It became the most important title of later rulers. It is the one used by Qalaæun, who is not only the greatest sultan (al-sul†ån al-aÆΩam) but also the king (al-malik). It was similarly used by ÆAbbas, who is also the greatest sultan as well as the most noble khåqån. Khåqån, a word of Persian origin denoting the same idea, was used by the Qarakhanids and then adopted by various Mongol sovereigns of Persia, Mesopotamia and Asia Minor. From there, it passed into the Egyptian sphere through Mamluk relations with their eastern and northern neighbours. Part of its popularity was probably due to the fact that it rhymed with sul†ån. In the inscription on the Dome of the Rock, ÆAbd al-Malik is also identified by his caliphal title, amÈr al-muæminÈn. Variously translated as ‘Commander of the Faithful’ or ‘Prince of Believers’, it signified his authority as leader (amÈr) of the community of believers (al-muæminÈn). Both Umayyad and ÆAbbasid caliphs continued to use this title, and, beginning in the mid-eighth century, they rewarded functionaries whom they wished to honour with the title ‘Client of the Commander of the Faithful’ (mawlå amÈr al-muæminÈn). Alternate forms were soon adopted, such as ‘Sincere Friend’ or ‘Glorifier of the Commander of the Faithful’ (ßafÈ/muÆizz amÈr al-muæminÈn). As this kind of title proliferated, it decreased in importance, but it still indicated a particular ruler’s relationship to the caliph. Part of Qalaæun’s authority, for example, is derived from his nominal position as ‘Associate of the Commander of the Faithful’ (qasÈm amÈr al-muæminÈn).

ÆAbbas, of course, did not have such a caliphal title since he had no connection to the caliphate. His title, instead, is Bahadur Khan (‘distinguished khan’). It derived from the Mongol bagatur, meaning ‘hero’, and had been assumed by the Ilkhanid ruler Abu SaÆid in 1319 after he had quelled a revolt by several amirs. It continued to be used by later Timurid and Safavid rulers in Iran, who wished to trace their lineage back to the Mongols. What really distinguishes the two later inscriptions in Cairo and Isfahan from those on the Dome of the Rock in Jerusalem is the extraordinary number of honorifics (laqab, plural alqåb) that the later rulers bear. The laqab was originally a nickname, such as al-Tawil (‘the tall one’) or al-Jahiz (‘the goggle-eyed’), but soon became an honorific title. These honorifics derived from the regnal titles assumed by the early ÆAbbasid rulers. Beginning in the tenth century, these honorific titles were frequently compound constructions, often with the second element containing dÈn (‘faith’) and dawla/mulk (secular power) and sometimes umma (‘religious community’) and milla (‘religion’). Qalaæun’s main honorific, for example, is ‘Sword of the World and the Faith’ (sayf al-dunyå waæl-dÈn). These compound titles with dÈn/dawla and the like became increasingly popular, and inscriptions are some of the major sources for identifying when different variants appeared and how the nature of political authority changed. Van Berchem was one of the first to raise this topic, in an 1893 article on an inscription dated 610/1213 from the Khan al-Aqaba near Tiberias (RCEA 3720). Later scholars have followed his lead and traced almost 200 variants of these titles over the last millennium (see details in the article ‘La˚ab’ in EI 2). We can use the inscription on Qalaæun’s complex to understand some general principles about the composition and arrangement of these compound titles as they proliferated and became more laudatory. The titles were often arranged in rhyming pairs. Thus, Qalaæun is ‘sultan of Islam and Muslims’ (sul†ån al-islåm waæl-muslimÈn) and ‘lord of kings and sultans’ (sayyid al-mulËk waæl-salå†Èn), a pair contrasting the same elements

— 37 —

— ISLAMIC INSCRIPTIONS — of religion and secular power that were juxtaposed in the dÈn/dawla pair. This idea of contrast appears in several of his other compound titles. He is, for example, sultan of the earth in longitude and latitude, king of the world (sul†ån al-ar∂ dhåt al†Ël waæl-Æar∂ malik al-basȆa). The titles are also plays on words. Qalaæun’s titles ‘inheritor of royalty, king of kings of the Arabs and Persians’ (wårith al-mulk malik mulËk al-Æarab waæl-Æajam) contains three words derived from the same root m-l-k (‘to rule’) set right next to each other. Many of these compound titles also contain duals. This tradition can be traced as far back as the third caliph ÆUthman, who was known as Dhuæl-Nurayn (‘possessor of two lights’). The duality often to alluded to the two branches of government, the sword and the pen, representing two sources of authority, the military and the legal system. Thus, the vizier of the Buyid ruler Rukn al-Dawla was called Dhuæl-Kifatayn (‘possessor of the two capabilities’, that is, the sword and the pen). These titles with dual attributes were undoubtedly popular became they made ringing eulogies when composed in pairs that rhymed with the dual ending ayn. Qalaæun bears four such titles with dual attributes, arranged in two sets of rhyming pairs. In the first set, each dual is also compounded: he is sultan of the two ÆIraqs and the two Egypts (sul†ån alÆiråqayn waæl-mißrayn) and king of the two continents and the two seas (malik al-barrayn waæl-bahrayn). The first phrase refers to the division of the province of ÆIraq into Arab and Persian sections and to the parallel division of Egypt into Upper and Lower sections. The second refers to the two continents, Africa and Asia, and the Mediterranean and Red seas. In the second set, Qalaæun is possessor of the two qiblas and keeper of the two noble sanctuaries (ßå˙ib alqiblatayn khådim al-˙aramayn al-sharÈfayn). The two qiblas refer to the regular direction of prayer toward Mecca and the original qibla of Jerusalem, the two noble sanctuaries to the sanctuaries at Mecca and Medina. ÆAbbas’s titles are quite different. The Safavid shah derived his authority not from his relationship to the caliph nor from his authority over the

shrines in Mecca and Medina, but rather from his descent from the Prophet. Thus ÆAbbas’s titles underscore his affiliation to ShiÆism. He is reviver of the virtues of his pure fathers (mu˙yÈ maråsim åbåæihi al-†åhirÈn). The phrase about revival goes back to medieval times when it was associated with the revival of orthodoxy and jihåd as a propaganda weapon. Nur al-Din, for example, was ‘reviver of justice in this world and the next’ (mu˙yÈ al-Æadl fiæl-ÆålamÈn) in works which he undertook after 551/1156 (Elisséeff 1952–4). Similarly, the Anatolian Saljuq vizier Fakhr al-Din ÆAli is called the ‘reviver of orthodoxy’ (mu˙yÈ al-sunna) in the endowment deed to the Gok Madrasa in Sivas (Rogers 1976: 70). ÆAbbas has switched the focus, however, and instead of reviving orthodoxy he is reviving ShiÆism, and the phrase ‘his pure ancestors’ denotes the Prophet’s family. Much the same message is conveyed by ÆAbbas’s second title, ‘propagator of the religious sect of the infallible imams’ (murawwij madhdhab al-aæimma al-maÆßËmÈn). In addition to the various parts of his name, a foundation inscription usually includes eulogies and benedictions for the ruler. These phrases grew apace with the titles. The benediction for ÆAbd al-Malik is fairly short and asks simply God to accept [the work] from him and be pleased with him (taqabbala allåh minhu wa ra∂iya Æanhu). The emphasis in these eulogies soon changed from God’s acceptance to particular benefits for the person blessed. The benediction for al-Maæmun asks God to prolong his existence (a†åla allåh baqåæhu); it was the standard form reserved for caliphs in the ÆAbbasid period. Like titles, benedictions grew longer and more elaborate over time. The benediction following Qalaæun’s name and caliphal title asks God for five things: to make the ruler’s glory endure, to glorify his victories, to elevate his sign, to double his power and to make his cities flourish (adåma allåh naßrahu wa aÆazza inßårahu wa aÆalå manårahu wa ∂åÆifa iqtadårahu wa Æammara amßårahu). The eulogy following ÆAbbas’s name asks God the Great to make the ruler’s kingdom eternal and let his ships sail in the seas of eternity through [the grace of] Muhammad and his good pure

— 38 —

— F O U N D AT I O N I N S C R I P T I O N S — infallible family (khalada allåh taÆålå mulkahu wa ijrå fÈ bi˙år al-taæyÈd fulkahu bi-mu˙ammad wa ålihi al-†ayyibÈn al-†åhirÈn al-maÆßËmÈn). These later and longer eulogies continue the theme of extending the ruler’s life taken up by the ÆAbbasids and add that of glorifying his name. Like titles, these long eulogies were often composed as rhymes and plays on words. Foundation inscriptions also had to be laid out to fit the shape of the building, and the text could be manipulated to fit the space available. In the case of the inscription around the outer arcade in the Dome of the Rock, the words ‘Amen, Lord of the Worlds’ (amÈn rabb al-ÆalamÈn) finish the text and fill out the band along the south-east face. Most foundation inscriptions, however, are written along the main façade of a building, and as the patron’s titles and eulogies got longer, there was more scope for adjustment to fit the available space. In Qalaæun’s funerary complex, the foundation inscription is a long band that runs along the street side of the building. The Mamluk historian al-Maqrizi called such a band a †iråz, the term used for an inscribed textile or the inscription on a textile. These bands were deliberately laid out to enhance the ruler’s prestige and authority. Thus, the foundation inscription for Qalaæun’s funerary complex was composed so that the introductory phrase about ordering the work takes up the space under the minaret, while the ruler’s lengthy titles fill the space on the qibla wall of his tomb. His name Qalaæun and caliphal title, Associate of the Commander of the Faithful, fall on the projecting jamb to the right of the portal and the eulogies fill the bay over the entrance corridor connnecting the various parts of the complex. Since the madrasa in the complex also fronted on the main street, the standard form was altered and extended so that the foundation inscription continues on the left side of the entrance corridor with another long band of titles lauding the patron and grammatically qualifying the ‘him’ in the eulogies. Qalaæun receives another eight or ten compound titles referring to his role subduing enemies, infidels and rebels, before his main honorific ‘Sword of the World and the Faith’,

his name Qalaæun, his caliphal title ‘Associate of the Commander of the Faithful’ and a short benediction asking God to extend his days. Qalaæun’s funerary complex shows how a foundation text could be manipulated to unify and tie together the various parts of a complex. The foundation text on a single building could also be adapted to underscore the ruler’s sovereignty and authority. Many foundation texts are found on the portal, which is usually vaulted. In Iran and the eastern Islamic lands, the portal assumed the form of a pÈsh†åq, a flat masonry or brick structure framing three sides of an arched opening. The foundation inscription could be written along the curved sides of the arched opening, as at Ribat-i Malik, the late-eleventh-century caravanserai on the road between Samarqand and Bukhara (Survey, plate 272). The text could also be set in a U-shaped band along the sides of the framing pÈsh†åq, as on the twelfth-century caravanserai known as Ribat-i Sharaf north of Mashhad on the road to Merv (Ettinghausen and Grabar, fig. 296). A third alternative was to set the foundation inscription around the three sides of the pÈsh†åq, as was done on the mosque of Shaykh Lutfallah. Like that of Qalaæun, ÆAbbas’s foundation inscription is set at mid-height, in this case above the doorway but below the muqarnas semidome. The text is set out so that ÆAbbas’s name and titles sit exactly over the doorway, and thus anyone who entered the building had to walk, literally and figuratively, under ÆAbbas’s sway. Style enhances the message. The tail of the letter yåæ in the word taÆålå at the left side of the door returns to the right and extends all the way across the doorway to the beginning of the shah’s titles on the right side. This horizontal stroke in the middle of the band emphasises the division of the bands into two registers and distinguishes the text above the portal with the name and titles. To emphasise his role, the ruler’s name was sometimes highlighted in a different colour, particularly in Iran where inscriptions were often executed in tile mosaic. This was probably the case in the Mosque of Shaykh Lutfallah, for old photographs showing the portal before restora-

— 39 —

— ISLAMIC INSCRIPTIONS — tion (for example, Survey, plate 482) show the central section below the screened window with ÆAbbas’s titles ‘al-Husayni al-Musavi al-Safavi Bahadur’ set on a lighter ground than the dark blue used in the rest of the inscription. Colour was certainly manipulated to highlight the ruler’s name on the portal of the nearby Shah Mosque, begun some two decades later in 1025/1616, where ÆAbbas’s name and titles, Abuæl-Muzaffar ÆAbbas al-Husayni al-Musavi al-Safavi, are set in light blue letters that contrast with the rest of the inscription written in white against a dark blue ground (Blair and Bloom 1994, fig. 238). As at the mosque of Shaykh Lutfallah, the highlighted section is set not only right over the doorway, but also right below an open window. In addition to the ruler and patron, a royal foundation text sometimes mentions various other people, particularly other important government officials. In the inscription recording al-Maæmun’s restorations to the Dome of the Rock, for example, the text tells us that work was done during the administration (fÈ wilåya) of his brother Abu Ishaq, son of the caliph al-Rashid. Work was carried out under Salih b. Yahya, client of the Commander of the Faithful, one of the many clients to whom the ÆAbbasid caliphs entrusted their constructions. The phrase Æalå yad[ay], literally ‘under the hand[s] of’, usually indicates the master-of-works who was in charge of payment and accounts. If the patron was not the ruler, but one of his viziers, amirs or other subordinates, then the ruler’s name is often cited at the beginning of the foundation text following the verb. His name is introduced by a phrase such as fÈ ayyåm (‘during the days of’) or fÈ dawla (‘during the reign of’). The ruler usually receives a long string of titles, while the patron, cited later, is described in exactly the opposite way, with a few modest epithets such as al-Æabd al-∂aÆÈf (‘the weak slave’). These epithets are to be understood not literally but figuratively, as contrasts to the glorious royal figure during whose reign the work took place. Another common phrase had to do with financing: less important patrons often said that work was done out of their own money (min målih) to show that this work was a personal act of piety.

The final element in a standard foundation inscription is the date. Typically this is expressed as ‘in the year’ (fÈ sana) or ‘in the months of the year’ (fÈ shuhËr sana), sometimes preceded by the additional phrase bi-taærÈkh (‘dated’), with the number of the year written out in units, decades and centuries. Thus the foundation on the Dome of the Rock gives the date ‘in the year seventy-two’, which ran from 4 June 691 to 22 May 692. Sometimes the date is more precise, with a specific month given. Thus, the restorations to the Dome of the Rock carried out under al-Maæmun took place in the month of RabiÆ II of the year 216, corresponding to May–June 831. Historians are usually glad to find a date and often satisfied with foundation inscriptions because they are dated. We must, however, also interpret what the date means, for it usually took more than a single month or even a year to construct a major building. It is possible to interpret the Dome of the Rock in radically different ways by considering the year 72 as either the date when the building was ordered or the date when it was completed. Most foundation inscriptions mark the former (Rogers 1976), and this is why they are written using the verb amara (‘to order’). I believe this to be the correct interpretation for the Dome of the Rock as well (Blair 1992b). Usually only one date is given, but occasionally a foundation inscription is more specific, giving both the beginning and the end of construction. The typical form uses faragha (‘began’) and tamma (‘finished’). Thus the foundation inscription on the façade of Qalaæun’s funerary complex says that the work began in the months of the year 683 (which ran from 20 March 1284 to 8 March 1285) and finished in the months of the year 684 (which ran from 9 March 1285 to 26 February 1286). Such specificity in date is unusual and demands explanation. In the case of Qalaæun’s complex, we know from textual evidence that the patron was extremely proud of the speed with which his order was executed, and texts confirm that the whole complex was constructed within two years. Combining inscriptions and chronicles, Creswell (1952–9 [1978], vol. 2, p. 210) was able to establish an exact

— 40 —

— F O U N D AT I O N I N S C R I P T I O N S — chronology for the building programme. On 12 Ramadan 682/4 December 1283, the sultan used monies from his private purse to buy the land and buildings from the occupants of the site, once part of the western, or smaller, Fatimid palace. Construction on the hospital was begun in RabiÆ II 683/June–July 1284 and finished in Ramadan 683/November–December 1284. Builders then turned to the mausoleum, and dated inscriptions tell us that it was begun the next month (Shawwal 683/ December 1284–January 1285) and finished four months later (Safar 684/April–May 1285). The madrasa was begun that month and finished four months later in Jumada I 684/July–August 1285. Before citing a date as indisputable evidence for the foundation of a building, it is imperative to verify the text. Since foundation texts in Iran were often inscribed in a frame around the projecting portal or pÈsh†åq of a building, the date often falls at the lower left, near ground level where the brick band is susceptible to damage from rising ground water, salination and the like. Part of the text may be illegible, and it is often a prime spot for repair, sometimes with an incorrect date. The foundation inscription on the portal of the khanaqah at Natanz in central Iran, for example, now reads ‘in the year seven hundred and twenty-five’ (1324–5). This restoration is clearly wrong, for earlier photographs show that the first letter after the word sana (‘year’) began with a sÈn/shÈn, dictating a date that begins with six/sixty or seven/seventy. Based on a combination of historical and constructional evidence, the correct date for the khanaqah should be restored as 707/ 1307–8 (Blair 1986). Sometimes, particularly when space is at a premium, the year is squeezed in at the end of the inscription in numerals. They are often written above the word sana or slotted in above other vertical strokes. At the end of the foundation inscription on the mosque of Shaykh Lutfallah, a small band written vertically at the left edge of the main foundation inscription gives the signature of the calligrapher and the date in numerals, 1012, which ran from 11 June 1603 to 29 May 1604. One of the main reasons to erect a monumental inscription was propaganda and advertising, to broadcast a ruler’s good name and works or to

mark his sovereignty. This is already clear in the earliest monumental inscriptions, such as the milestones erected by the Umayyad caliph ÆAbd al-Malik along routes leading to his capital Damascus (RCEA 14–17) and along a difficult pass near Lake Tiberias (Sharon 1966). All of the texts are similar. Following the basmala and blessings on the Prophet, the inscriptions record that the work was ordered by God’s servant, ÆAbd al-Malik, Commander of the Faithful. The milestone found at Khan al-Hathrura (RCEA 14), a ruined caravanserai between Jerusalem and Jericho, is a good example (see Figure 3.16). The top line or two with the opening basmala and most of the blessing on the Prophet has been destroyed, but the last word is preserved. The text then records that ÆAbd al-Malik ordered the construction of this route (†arÈq) and the making of these milestones (amyål). Only the last of the original nine or ten lines actually states the distance from this milestone (mÈl) to Damascus, 109 miles. The milestones were erected not as a convenient sign for weary travellers but, like the caliph’s new epigraphic coins, as a sign of his presence and power (Bloom 1989: 45; Blair 1992b). One of the most famous examples showing how an inscription marks sovereignty was discovered by van Berchem in the foundation text on the Dome of the Rock (MCIA Jerusalem, no. 215). The text now reads that this dome was built by God’s servant, ÆAbdallah, the imam al-Maæmun, Commander of Believers, in the year 72 (Kessler 1970; Blair 1992b) but van Berchem showed that the ÆAbbasid caliph al-Maæmun had simply had his name inserted in place of that of the Umayyad caliph ÆAbd al-Malik. This change was evident on stylistic grounds, for the letters in the replaced area are more cramped and written on a darker blue ground. The date at the end of the text, however, was not changed to accord with that of al-Maæmun’s reign more than a century after the building was constructed, and van Berchem argued that the intention in altering the inscription was not to change the origin of building, which everyone knew had been built by ÆAbd al-Malik, but to substitute the name of reigning caliph for that of a deceased caliph who represented a rival and

— 41 —

— ISLAMIC INSCRIPTIONS —

3.16 Umayyad milestone in the name of ÆAbd al-Malik recording the distance of 109 miles to Damascus. Istanbul, Museum of Turkish and Islamic Art 2511.

hated dynasty. The changed inscription, like many others, was not meant to commemorate material work but as a prise de possession. The symbolic importance of many inscriptions is underscored by the fact that some are nearly unreadable and were meant to affirm symbolically the presence of the ruler (Ettinghausen 1974). Phrases are often omitted, letters misconnected or words misspelled. Sometimes these mistakes may have been the result of blind copying, and the person who drew up the inscription may not have been the same one who executed it. Sometimes, legibility did not seem to matter and was apparently readily sacrificed to decorative concerns. This is the case, for example, in a large poetic inscription in the University of Pennsylvania Museum in Philadelphia and considered the finest piece of monumental Arabic calligraphy in the United States (Faris and Miles 1940). The inscription is carved in relief on nine slabs of

black tuff stone fitted together to make a large rectangular panel measuring about a metre high and almost three metres long. The Arabic poem records the construction of a canal or reservoir and an inner gate to the grounds of the palace of Gaur, the old capital of Bengal, in 871/1466–7 by Barbak Shah, the second Ilyas Shahi ruler of Bengal. The text was obviously composed for the specific site and occasion, and the amount of work involved in producing this large inscription shows that it was important, yet it is almost impossible to read. The inscription is written in two rows of sixteen frames alternating between a sort of tughra script and open thuluth so that the letters in every other frame are inscribed in open thuluth contrasting with and varying the rhythm of the repeated verticals in the alternate panel. This stylisation is so severe and the text so difficult to read that several lines of the poem remain undeciphered.

— 42 —

— CHAPTER 4 —

Other Types of Monumental Inscriptions

F

oundation and restoration inscriptions, although by far the most common inscriptions on buildings and the ones that set the standard, are not the only type of monumental inscriptions. There are several other types that use variant forms and give different information. Endowment texts are the ones closest to foundation inscriptions. They contain extracts from waqfiyyas, legally attested charters or constitutions and guarantees of their endowments. These deeds could be drawn up only when the building was complete, and after legal attestation a copy was registered at the judiciary or appropriate state office. There are two types of endowment inscriptions. In the first type, a religious building such as a mosque, madrasa, zawiya, tomb or similar kind of building is endowed to a well-known group or religious sect. In the second, the inscription enumerates the property, buildings or objects to be used as support for the pious foundation. Two different verbs are used: waqafa (‘to endow’) and ˙abbasa (‘to tie up inalienably’). The two verbs have different shades of meaning, and the former was more common in the first type of endowment inscription and the latter in the second. There were also regional distinctions: waqafa was used in both types of endowment inscription in the eastern Islamic lands, while ˙abbasa was used in both types in Spain and North Africa. In both types, the verb is followed by the name and titles of the giver and the type of gift, which can be a single object or a long string of properties. The two types of endowment inscription often come together and follow a foundation or restoration text. A good example is the endowment text on

the mosque at Sidi Bu Medene, the shrine complex for the Sufi saint Abu Madyan ShuÆayb (d. 1197) in the village of al-ÆUbbad, 2 km east of Tlemcen in Algeria (RCEA 5764). The thirty-six lines of text are inscribed on a marble tablet encased in the column to the left of the mihrab (van Berchem 1907, plate 1). The inscription opens with a standard foundation text saying that the congregational mosque and adjacent madrasa were ordered by the Marinid sultan Abuæl-Hasan ÆAli, who receives the usual titles and eulogies. The text continues with the first type of endowment text, saying that the sultan inalienably endowed the madrasa for the pursuit of science and instruction. The text then continues with the second type of endowment text, enumerating a long list of gardens, orchards, houses, windmills, baths, and land that the sultan bought and inalienably endowed to the mosque and madrasa. In both cases, the verb is ˙abbasa. Endowment inscriptions can be quite close to foundation texts, and sometimes a foundation text can even include the information about endowment. This is the case with the foundation inscription on the Salahiya Madrasa in Jerusalem (RCEA 3453). The text is written in five lines on a marble slab above the entrance door on the west façade. It opens with a brief identification of the building as ‘the blessed madrasa’ (al-madrasa almubåraka) and then says that it was endowed (waqafa) by the Ayyubid ruler Salah al-Din (Saladin) as a madrasa for the ShafiÆite sect in 588/1192. Most of the space is given to the patron’s name and titles, including several compound honorifics (laqab), his patronymic Abuæl-Muzaffar, his personal

— 43 —

— ISLAMIC INSCRIPTIONS — name Yusuf, his genealogy (b. Ayyub b. Shadhi), his caliphal title Reviver of the Empire of the Commander of the Faithful, and eulogies asking God to glorify his victories and amass for him the benefits of this world and the next. The only difference from a standard foundation inscription is the verb waqafa and the inclusion of the group to which the building is given. In Saljuq Anatolia, it was rare to inscribe extracts of endowments on buildings, and Rogers (1976) has suggested that the foundation inscription served as a substitute to bridge the gap between the time the building was finished and the endowment deed drawn up. In other cases, the endowment inscription was distinct from the foundation inscription. The distinction could be made clear by a different form, with the foundation inscription on a large band over the portal and the endowment text written on a marble plaque. The distinction could also be made clear by different styles of script. This is the case in the lengthy inscriptions on the domed sanctuary of the congregational mosque in Qazvin, Iran (Sourdel-Thomine 1974). A long and elegant foundation inscription in cursive runs around the arches of the zone of transition. It says that the mosque was ordered during the reign of the Saljuq sultan Muhammad b. Malikshah by the governor of the city, Khumartash, and that construction took nine years, from 500/1106–7 to 508/1114–15. A second and larger inscription in floriated Kufic runs around the top of the walls of the square chamber. Clearly an abridgement of the endowment deed, it enumerates the properties endowed to the mosque and attached madrasa as well as other endowments for indigents in Medina, the upkeep of a new underground channel (qanåt) dug to supply water to the residents of Qazvin, and other purposes that are unclear because of lacunae in the text. The endowment text continues on several panels on three walls of the sanctuary, enumerating Khumartash’s endowments for the upkeep of several other pious foundations, including a Sufi convent in Mecca, a caravanserai in the village or quarter of Damghan, a well and oratory in the village of Ghirawan, and a small Sufi convent near the door of the mosque in Qazvin. A fourth text on

the west wall at the same height as the previous ones adds new stipulations about how the water in Khumartash’s underground channel is to be apportioned. This text ends with the date 509/ 1116–17, a year after the date mentioned in the foundation inscription, perhaps indicating the time needed to complete the interior decoration. This is one of the few endowment texts to survive from medieval Iran, partly because the brick used there for construction meant that such texts did not survive. Far more are known from Syria and Egypt, particularly from Mamluk times, because the stone used there ensured better preservation. Like foundation inscriptions, endowment inscriptions have a propagandistic aim. They were meant to show who endowed what, when and why, and much of the space is given over to the titles and generosity of the patron. Endowment inscriptions are also important economic and social documents, for they give details about villages, upkeep, nourishment, water supplies and living conditions. The same sort of details about daily life are given in decrees, which also supply information about administrative procedures, events of local history, and economic and fiscal policy. These types of monumental inscriptions are especially important in filling in the lives of the common people, a subject often ignored in traditional chronicles. Most decrees known date from Mamluk times. Many are found in Syria, and the major scholar associated with their publication is Jean Sauvaget (1932b, 1933, 1947–8). Decrees were clearly set up before this period, but very few have survived. Decrees were also set up in other areas. The longest Mamluk decree inscribed on stone, for example, is a fifteen-line fiscal decree abolishing taxes on objects imported to and used in Mecca. Issued by al-Ashraf ShaÆban on Monday 3 Jumada II 766/26 January 1365, it was set up on three or four pairs of columns in the Haram at Mecca (MCIA Arabie, nos 23–5; RCEA 766 007). Many similar decrees on stone plaques were also set up in Iran in the later period. The congregational mosque in Nishapur, for example, has three decrees (usually called farmån in Iran)

— 44 —

— OTHER TYPES

OF

M O N U M E N TA L I N S C R I P T I O N S —

mounted in the wall. Dated 1021/1612, they give tax relief to the citizens of Nishapur. Few of these later Iranian decrees have been published systematically. Like endowment inscriptions, decrees were often written on marble plaques. The Mamluk examples were typically set up near the market or on the walls or columns of porticos in congregational mosques. As official proclamations, they were thus installed right under the eyes of those who would benefit. Yet at the same time, as Sauvaget pointed out, many of them were poorly written. Mamluk examples are generally done in low relief, with defective writing, omitted letters or even words, orthographic and grammatical mistakes, unknown vocalisations and errors of editing. The fact that they were set up seems to have been more important than their readability to passers-by. Typical decrees from the early period open with the basmala followed by some form of amara, either bi-amr or kharaja al-amr. By the fourteenth century, the basmala was replaced by the phrase al-˙amd lillåh (‘Praise to God’). The text usually opened with the exact date of day, month and year written out in words following the phrase lamå kåna bi-ta’rÈkh (‘on’). Next came the phrase baraza al-marsËm (‘emerged the decree of’), which was followed by the name and titles of the patron. The details of the decree took up most of the text. At the end was the date, if it had not been used at the beginning, and various maledictions to anyone who violated the conditions of the decree. The maledictions invoke God’s anger and punishment on Judgment Day. They often include Koran 26:228, saying that the unjust will soon know the vicissitudes that their affairs will take. Koran 3:173, saying that God is sufficient and disposes of affairs best (˙asbunå allåh wa niæma al-wakÈl), indicates the end. A third type of inscription, funerary inscriptions, is intended to show for whom a tomb was constructed and always contains the name of the person buried there. The simplest way to introduce his name was the phrase ‘this is the grave of’ (hadhå qabr ...). This phrase was often used on tombstones (see Chapter 14). The word qabr could

be replaced by the variant, feminine form maqbara or by different words. Similar texts on cenotaphs, for example, use marqad (‘screen’) or ∂arÈkh (‘cenotaph’). Other synonyms include turba and qubba. The word qaßr was common on the tomb towers in Iran, as on the Gunbad-i Qabus, the extraordinary tomb tower near the Caspian Sea built by the amir Qabus b. Washmgir in 397/ 1006–7 and the tomb tower at Radkan built for the Bawandid amir Abu JaÆfar in 407/1016–17 (Blair 1992a, nos 19 and 31). Sometimes the name could be introduced by a more elaborate sentence with two nouns, saying that ‘this thing/place is the tomb of so-and-so’, as on the funerary inscription around the tomb tower known as the Pir-i ÆAlmadar, built in Damghan in 416/1026–7, which uses the phrase ‘this qubba is the qaßr of ...’ (Blair 1992a, no. 34). Given the importance of family, the part of the funerary inscription naming the deceased includes a long genealogy (nasab) and details about his death. On an imamzada, or shrine built particularly in Iran to honour members of the Prophet’s family, the genealogy often goes back to the Prophet or his immediate family. A good example (see Figure 4.17) is the funerary inscription on the Imamzada Yahya b. Zayd at Sar-i Pul in northern Afghanistan, built to honour an early ÆAlid martyr, Yahya b. Zayd. The large band around the top of the walls below the zone of transition gives Yahya’s genealogy four generations back to the Prophet’s nephew ÆAli b. Abi Talib and relates that Yahya was martyred at Arghuy on a Friday in the month of ShaÆban of the year 125 (June 743) during the government of Nasr b. Sayyar during the reign of the Umayyad caliph al-Walid, who is cursed (Blair 1992a, no. 75). Similarly, the funerary inscription on the splendid wooden cenotaph for the Imam al-ShafiÆi gives his genealogy ten generations back to the Prophet’s grandfather as well as the dates of Imam alShafiÆi’s birth (150/767–8), death (Friday, the last of Rajab 204/20 January 820) and burial the same day (RCEA 3332). The same information is repeated on a nearby marble column (RCEA 3333). The funerary text was often followed or accompanied by the foundation inscription saying

— 45 —

— ISLAMIC INSCRIPTIONS —

4.17 Funerary inscriptions around the interior of the Imamzada Yahya b. Zayd at Sar-i Pul in northern Afghanistan.

who ordered the tomb. It could be the person who was buried there, as with the tomb towers at Gurgan and Radkan. It could be a close relative, often a son, as in the case of the Pir-i ÆAlamdar. Female relatives were also patrons. The tomb tower at Lajim was built for a Bawandid prince by his daughter (Blair 1992a, no. 32), and Muæmina Khatun, widow of the Saljuq sultan Tughril b. Muhammad, had a tomb tower built in Nakhchivan for her son Muhammad Jahan Pahlavan, who died in 582/1186 (RCEA 3410–14). For the grave of a revered person or saint, the patron was usually a different person, who may have ordered the work years, decades or even centuries after the saint died. The Imamzada at Sar-i Pul, for example, was built c. 1100, some 350 years after Yahya b. Zayd was martyred. The patron’s name, Abu ÆAbdallah Muhammad b. Shadhan al-Farisi, is recorded in the smaller inscription around the arches. Similarly, the cenotaph for Imam al-ShafiÆi was ordered by Salah al-Din in 574/1178–9. He also built an accompanying madrasa and tomb,

which were destroyed and replaced by the present structure, built by Salah al-din’s nephew al-Kamil in 608/1211 in his honour of his mother. Thus, it is important to note that the date of death recorded in a funerary inscription is not necessarily the same as the date of foundation of the tomb. A fourth type of monumental inscription comprises commemorative texts. They are found all over the Near East, but particularly in the Arabian peninsula, where there was a strong preIslamic tradition of commemorative texts. Similarly, many are found in Iran, where monumental rock reliefs had been carved since Achaemenid times. Technique often distinguishes commemorative texts from other types of monumental inscriptions: commemorative texts are often incised or painted, not carved in relief. Incising is significantly faster and hence cheaper than relief carving, where the background around the letters has to be cut away. Commemorative inscriptions were often carved near ruins. In Arabia, many were carved in early Islamic times along the Darb Zubayda,

— 46 —

— OTHER TYPES

OF

M O N U M E N TA L I N S C R I P T I O N S —

the pilgrimage route connecting ÆIraq with the holy shrines (al-Rashid 1980, pp. 241–51). Several others are carved in the rock wall of a dam near Taæif around the foundation text recording MuÆawiya’s order for its construction in 58/ 677–8 (Miles 1948). In Iran, it was a popular practice to carve one’s name at major pre-Islamic sites in Fars province. Several commemorative texts are found, for example, at Naqsh-i Rustam, the Sasanian rock relief showing the investiture of Ardashir I, and at Bahmandiz, a rocky bluff near the pool known as Khusraw and Shirin, after the famous story, later popularised by Nizami, in which Khusraw watched Shirin bathing (Blair 1992a, no 2). A series of commemorative inscriptions was also carved from the tenth century onwards at the Achaemenid site of Persepolis. Known to the Persians as Takht-i Jamshid (‘the throne of Jamshid’), after the legendary king of Iran who created the crafts, the site served as a sort of ‘Muslim memorial’ (Melikian-Chirvani 1971b). One of the first is an eight-line text (see Figure 4.18) in the name of the Buyid ruler ÆAdud alDawla (Blair 1992a, no. 7). Commemorative inscriptions from early Islamic times usually request God’s mercy or forgiveness for a particular person. They often contain a religious formula, such as the invocation to God (basmala) or the profession of faith (shahåda), or a common Koranic verse, such as 33:56 saying that God and his angels bless the Prophet and that believers should bless him and salute him as well. Some texts also have the name of the person and a verb, typically ha∂ara (‘was present’), nazala (‘descended’) or kataba (‘wrote’). ÆAdud al-Dawla’s text at Persepolis, for example, uses ha∂ara. Texts from later times, particularly in Iran, are often in verse. Many of the people named in these commemorative texts are unknown figures, who are not recorded in major chronicles. The people mentioned in the commemorative texts took prestige from being connected with the sites where their names were inscribed. Thus, Buyid rulers carved their commemorative texts at Persepolis in the palace built by Darius.

Both the placement and the text of ÆAdud alDawla’s inscription were meant to underscore the connections between the struggling Buyid amir and his illustrious predecessors. ÆAdud alDawla’s inscription is carved on the door jamb between the portico and the main hall of the palace. The stone palace also has a trilingual inscription in Old Persian, Elamite and Babylonian identifying Darius, the great king, the king of kings, the son of Hystaspes, as its builder, and the entrance jambs are decorated with reliefs showing the Achaemenid king and two attendants proceeding from the hall. Almost a millennium later, two inscriptions in Parsik praising the Sasanian monarch Shapur II (r. 309–79) were added beside the east jamb. ÆAdud al-Dawla’s text, added just to the left of the Parsik inscriptions, was obviously meant to associate him not only with Darius, founder of the Achaemenid line, but also with Shapur II, the victorious Sasanian king who had the longest reign. Similarly, the text of ÆAdud al-Dawla’s inscription was planned to reinforce the connection between the Buyid amir and his pre-Islamic predecessors. It says that the Buyid amir was present at the site in 344/955–6 on his return, victorious, from the conquest of Isfahan, his capture of the Daylamite Ibn Makan, and his rout of the army from Khurasan. The references to these victories are meant to evoke those of Shapur II, who subdued the Romans on the west and the Kushans on the east. The Buyid amir carries not only his Arabic title, ÆAdud al-Dawla (literally, ‘Forearm of the State’), which the ÆAbbasid caliph had awarded him in 349/951, but also his Persian name Fannakhusraw (literally, ‘the refuge of Khusraw’), a popular Daylamite name that evoked the memory of the Sasanian king Khusraw I (r. 531–79), also known as Anushirwan the Just. ÆAdud al-Dawla’s ambitions, evident in the commemorative text at Persepolis, culminated in his assumption of the old Persian title shåhanshåh (‘king of kings’), used on his coins from 370/981. Sometimes we do not know why a commemorative text was inscribed. This is the case with one carved in 433/1041 on the Waruh Gorge

— 47 —

— ISLAMIC INSCRIPTIONS —

4.18 Commemorative text carved at Persepolis in the name of the Buyid ruler ÆAdud al-Dawla in 344/955–6.

— 48 —

— OTHER TYPES

OF

M O N U M E N TA L I N S C R I P T I O N S —

connecting Isfara and Waruh in the Farghana Valley of Uzbekistan (RCEA 2489; Blair 1992a, no. 42). The text says that it was ordered by the local governor to honour (tashrÈfan) the Qarakhanid ruler Arslan Tikin, who receives a long list of titles and his full name, with laqab MuÆizz alDawla, kunya Abuæl-Fadl, ism ÆAbbas, and nasab (b. Muæayyad al-ÆAdl Ilik b. Nasr b. ÆAli), a genealogy that traced his descent four generations back to his great-grandfather who had founded the ÆAlid or western branch of the Qarakhanid line. This was the time when Arslan Tikin’s father was struggling to consolidate power (the Qarakhanids had a particularly fractious political history), and the inscription may indicate that he had assigned this area to his son. Whatever the reason, carving such an inscription was a significant event, and thus these texts are useful in supplementing official written sources, which are often composed post facto, after events had been resolved. Shorter commemorative texts can be dated only on stylistic grounds, but longer texts include the date, which is sometimes given not according to the Muslim hegira calendar but in other ways. A commemorative inscription carved by the Buyid Abu Kalijar at Persepolis in 438/1046, for example, gives the day and the month in the Persian solar calendar (Blair 1992a, no. 43). The Qarakhanid inscription at the Waruh Gorge uses three calendars: the Muslim lunar calendar, the Persian solar calendar and the Syrian Christian calendar, presumably because there were many Nestorian Christians in the region. Foundation inscriptions, particularly on religious buildings, and chronicles are usually dated in the Muslim year to show the piety and official nature of the record, but commemorative texts show us that several calendars were used concurrently during medieval times, probably for agricultural activities, religious festivals and other seasonal events. A final type of monumental inscription discussed here comprises signatures of architects, builders and other workmen. Inscriptions are particularly important in preserving the names of these people and establishing their careers, for craftsmen are rarely mentioned in traditional histories and chronicles, which usually ignore

architecture and builders. The pioneering work in collecting these names was done by L. A. Mayer, whose list of Islamic architects and their works (1956a) contains 318 names compiled from signatures as well as texts. Wilber (1976) added a similar list of more than 100 builders and craftsmen known to have worked in Iran before the Timurid period. Craftsmen’s signatures do not occur regularly on buildings throughout the Islamic lands. They were much more common in Iran and adjacent regions than elsewhere. Only a handful, for example, are known from Mamluk Egypt and Syria, despite the hundreds of buildings that have survived (Meinecke 1992) A craftsman’s signature can be found in various places. It can be included within the foundation text before the date, but more often it follows the date at the end of the foundation text. The calligrapher ÆAli Riza ÆAbbasi signed the Mosque of Shaykh Lutfallah in a small vertical band at the end of the foundation inscription on the portal (see Figure 3.15) and again at the ends of the upper and lower bands around the zone of transition in the interior of the building. A craftman’s signature can also be written in another inconspicuous place, usually within a panel or cartouche. The builder’s signature on the Mosque of Shaykh Lutfallah, for example, is found on a small panel to the left of the tile mosaic mihrab in the interior. Written in two lines of white nastaÆlÈq letters on a blue ground, it gives the name of the builder, Muhammad Riza b. Ustad Husayn, the builder of Isfahan (bannå-yi ißfahånÈ), and the date in numerals 1028/1618–19. The builder Baqir also signed the large band framing the arches around the interior (Hunarfar 1977: 401–15). The standard verb used for these signatures is Æamal (‘work of’). Occasionally the variants Æumila (‘made by’) or mimmå Æumila (‘part of what was made by’) are used. Other verbs include ßanaÆ (‘made’ or ‘crafted’) and kataba (‘wrote’). Over time, as building trades became increasingly specialised, a single building might bear several signatures of different types of craftsmen. The Mosque of Shaykh Lutfallah, for example, is signed by the calligrapher and two builders.

— 49 —

— ISLAMIC INSCRIPTIONS —

4.19 Interior of the Mosque of Shaykh Lutfallah showing the mihrab.

— 50 —

— OTHER TYPES

OF

M O N U M E N TA L I N S C R I P T I O N S —

The craftsman’s signature with Æamal should be distinguished from the supervisor’s name with Æalå yad[ay] (‘under the hand[s] of’). The latter is often included in a foundation text and indicates the master-of-works who was in charge of payment and accounts. The foundation inscription marking ÆAbd al-Malik’s levelling of a difficult pass near Lake Tiberias, for example, says that it was carried out under the hands of (Æumilat Æalå yaday) Yahya b. al-Hakam, who was the caliph’s paternal uncle and the master-ofworks (Sharon 1966). The phrase Æala yad was particularly common in caliphal inscriptions from the early Islamic period, whereas craftsmen’s signatures appear more frequently from medieval times onwards. The verb in a signature is followed by the name of the craftsman, who is sometimes identified by a nisba. This epithet can indicate his profession. The most frequent is al-bannåæ (‘builder’), but other examples from the construction trade include almiÆmår (‘architect’), al-muhandis (‘engineer’), and al-ustådh and al-muÆallim (both ‘master’). The epithet indicated the relative esteem or importance of the different trades, and we have signatures of several smiths (al-˙addåd) and carpenters (alnajjår). The epithets also indicate the increasing specialisation of craftsmen in the later periods. Like the different words for types of building, these epithets designating trades must be interpreted within context. Their meaning varied over time and space. Furthermore, the same person could bear different epithets on the same building. Muhammad b. Husayn al-Damghani, a craftsman who worked on the tomb complex for Bayazid Bastami in the opening decades of the fourteenth century, for example, is called builder (al-bannåæ), engineer (al-muhandis) and stucco worker (al-jißßåß) in different signatures (EIr, s.v. ‘DåmÌånÈæ). Some epithets designating a trade were used to show the humility of the worker, particularly in face of the lofty titles of the ruler, and should be read not literally but metaphorically. A good example is the Timurid architect Qavam al-Din Shirazi, who worked for Shah Rukh and his wife Gawhar Shad and built the Friday Mosque within the shrine of Imam Riza at Mashhad.

Qavam al-Din is probably the most famous Iranian architect, yet in his signature on a panel below the main inscription ringing the sanctuary iwan there, he carries the epithet al-tayyån (‘a worker in mud brick’), although no mud brick was used in the construction of this elegant building. Qavam al-Din was not always so modest: in his signature on the madrasa at Khargird, he is called uståd (‘master’). A craftsman’s epithet can also indicate a geographical location, specifically the craftsman’s home town. The al-Damghani family, comprising Muhamad, his brother Hajji and their father Husayn b. Abi Talib, all signed work on the shrine of Bayazid Bastami. Since Hajji also signed a mosque in Damghan, it is likely that they were, as their nisba suggests, a family from Damghan that was recruited to work on the shrine at Bastam some 85 km away. Sometimes, however, even a geographical nisba must be interpreted metaphorically. This is the case with the nisba al-Shirazi carried by three architects in the Timurid period, Qavam al-Din, Ghiyath al-Din and Hajji Mahmud (Golombek and Wilber 1988: 193–4). They all built in a similar style at about the same time in the mid-fifteenth century in the same region of eastern Iran and Transoxiana, and their work can be interpreted as a school of architects. But it is historically improbable that they all had come from Shiraz. Rather, the nisba probably indicates that they had been trained in the tradition of Shiraz or were descendants of the original Shirazi builders who had been recruited by Timur several generations earlier. Craftsmen’s signatures are important social documents. In addition to identifying the work of individual builders, they help us to establish the careers of craftsmen. Like other high-status professions, the building trades were apparently local specialities, often passed down in families over a long period. The tenth-century geographer al-Maqdisi, for example, mentions Biyar, a village to the east of Damghan, as a town of architects where the men were skilled in building and planning. He tells of a man from Biyar who went to Bukhara and built a fine castle in the second

— 51 —

— ISLAMIC INSCRIPTIONS — quarter of the tenth century. The town was still producing architects a century later, for Ibrahim b. Idris, the architect who signed the tomb tower built in 490/1096–7 at Mihmandust, some 15 km east of Damghan, carries the nisba al-Biyari (Blair 1992a, no. 66). The frequent coincidence of surviving signatures with the rare textual references suggests that other undocumented notices in textual sources may also be accurate and are worthy of note. Similarly, signatures help us trace the careers of the al-Munif family of architects from Sfax, Tunisia in the seventeenth and eighteenth centuries (Mayer 1956a). ÆAtiyya al-Munif and Ahmad al-

Munif, both sons of Ibrahim, rebuilt a tower in the southern wall in 1041/1630–1, and Ahmad, ÆUmar and ÆAbd al-Latif reconstructed a gate in 1056/1646. Their professional esteem is clear, since they are all designated master (muÆallim). The family’s reputation was widespread, for Ahmad along with a compatriot named al-Darif al-Sifaqsi (‘from Sfax’) restored the qasba of the nearby city of Gafsa in the mid-seventeenth century. The family continued to practise as architects for several generations, for Tahir, son of the late Ahmad, along with his brothers and cousins, restored another gate in the city walls of Sfax in 1748.

— 52 —

— CHAPTER 5

—

Regional Studies

A

nother way of looking at inscriptions in the Islamic lands is to follow regional developments. Artists and patrons drew inspiration from the things around them, and local examples and traditions often served as the main inspirations and models for later work. Furthermore, scholars studying inscriptions have traditionally grouped them geographically, by country, city or region. This was van Berchem’s original idea in setting out his magnum opus, the Matériaux pour un Corpus Inscriptionum Arabicarum (MCIA), and much of the gathering and recording of inscriptions is still done on a geographical basis. The main bibliographic sources on inscriptions, the bibliographies in the article ‘Kitåbåt’ (‘Inscriptions’) in EI 2 and the second supplement to the Creswell bibliography, are also grouped geographically, by country. They are the essential reference tools to find comprehensive bibliographies of the inscriptions in various areas of the Islamic lands, and they should be consulted to find the broadest range of citations about inscriptions in a particular area. This chapter has a more critical aim. It picks out a few of the most important articles, volumes or series in order to show the different kinds of information and problems raised by the inscriptions from different regions – here subdivided as the Maghrib; the central Islamic lands; Iran, Afghanistan and Central Asia; and the Indian subcontinent – and the different kinds of analysis that have been attempted. It also points to areas where more work is clearly needed.

Maghrib Inscriptions from the western Islamic world,

although less frequent than those from the East and often detached from their original context, were some of the first to be studied. One of the pioneers in the field was Michele Amari. He compiled an inventory of inscriptions on buildings in Sicily. It first appeared as a series of articles in Rivista Sicula in the years 1869–70 and was then printed as a separate volume in 1875. Its fundamental importance is clear from the fact that it was reprinted almost a century later (1971). Another early work was Lévi-Provençal’s twovolume monograph on Spanish inscriptions (1931). It was conceived in a spirit close to that of the MCIA, using material collected since the early nineteenth century and additional inscriptions that the author had gathered during three trips to Spain in the 1920s. It is a good example of the value of regional studies and the kinds of social conclusions that can be drawn from them. The inscriptions are arranged geographically by province, then chronologically within the province, but the introductory essay cuts across these categories and summarises Lévi-Provençal’s findings. Most of the texts come from the major cities of the Iberian peninsula (Almería, Toledo, Cordova, Granada, Murcia, Badajoz and Seville), although some of the capital cities of the Party Kings preserve no inscriptions, and a few inscriptions, such as foundations of castillos or epitaphs of minor figures, come from isolated sites. The 227 inscriptions are mainly foundation texts and epitaphs; at the end is a summary list of inscriptions on portable objects, particularly ivories and marble. Following van Berchem, who omitted many of the Ottoman inscriptions from Cairo as poetry,

— 53 —

— ISLAMIC INSCRIPTIONS — Lévi-Provençal omitted the poetic inscriptions from the Alhambra palace in Granada, as he considered them less interesting for history, but this omission has been filled by García Gómez’s recent study (1985). Lévi-Provençal’s 227 inscriptions range in date from the ninth century to just before the Reconquista in 1492. Despite their small number, the inscriptions illuminate the brilliant period of Muslim Spain – from the reign of the Umayyad ÆAbd al-Rahman II (822–52) through the Party Kings and the Almoravids. Two-thirds of the texts are epitaphs; the remaining one-third is divided between foundation inscriptions and inscriptions on portable objects, mainly ivories and marble. The balance changed over time. In the tenth century under the Umayyads, foundation texts outnumbered funerary texts and those on portable objects. This is not surprising, for this was a time of cultural florescence when new mosques and other buildings had to be erected to meet the demands of the new faith. During the eleventh and twelfth centuries, the proportion of epitaphs increased until almost all the texts were funerary. Lévi-Provençal’s work is the foundation for continuing work on inscriptions in Spain and adjacent areas. Those from individual provinces are treated in volumes by Ocaña Jimenez (1964) on Almería, Rosseló Bordoy (1975) on Balear, and Valderrama Martínez (1975) on Tetuán. Cleaning and restoration has brought to light a great deal of new information about the inscriptions of the Alhambra, much of which is published in the Cuadernos de la Alhambra and García Gómez’s 1985 volume on the Arabic poems there. Many of the relevant articles and books are cited in the bibliography to Fernández-Puertas’ useful overview of Spanish Islamic epigraphy (1992) The material from North Africa is less well published, and most of the work concentrates on charting the development of inscriptions in a particular province or city. A survey of the inscriptions of Algeria, Corpus des inscriptions arabes et turques d’Algerie, was begun at the turn of the twentieth century, with volumes on the provinces of Algiers (Colin 1901) and Constantine (Mercier 1902), but the work was

never continued. There are useful notes in van Berchem’s old but still valuable article on Islamic epigraphy in Algeria (1905). A similar series for Tunisia, Corpus des inscriptions arabes de Tunisie, was undertaken in the 1950s, with different volumes on individual cities. Roy and Poinssot compiled a two-volume work on the inscriptions of Qayrawan (1950–8), and Zbiss published three volumes on the inscriptions of Tunis (1955), Gorjani (1962a) and Monastir (1962b). Similar works on the inscriptions from individual cities in Morocco include a series of five articles (the fifth is an index) by Alfred Bel on inscriptions in Fez published in several issues of Journal Asiatique between 1917 and 1919. Brosselard collected the inscriptions from Tlemcen, which were published in a series of articles in Revue Africaine in the 1850s and 1860s. More recently, Deverdun (1956) compiled a volume on the inscriptions of Marrakesh. There is little synthetic material on the inscriptions of North Africa, and many of the scholars working in one country are unaware of work done elsewhere. This is a pity, for many of the inscriptions are handsome. Good photographs and details of some are published in Hill and Golvin’s (1976) photographic survey of the region. Even the inscriptions on the best-known and well-photographed monuments remain outside the general domain. This is the case, for example, with the inscription on the Gate of the Wayadas/ Ouadias in Rabat (see Figure 5.20) generally considered the finest of the Almohad gates. More work on inscriptions in North Africa would also lend evidence and insights into local history, a clear boon, since, unlike the eastern Islamic lands, there are few local chronicles from the west.

Central Islamic Lands Of all regions, the rich epigraphic material from the central Islamic lands has been the subject of the most wide-ranging investigations. This is true partly because many inscriptions from the central area were well documented early in the twentieth century, not only in the first volumes of the MCIA but also in other seminal studies.

— 54 —

— REGIONAL STUDIES —

5.20 Wadayas Gate, Rabat, late twelfth century, with an unread inscription around the frame.

For example, van Berchem in his classic study (1910) with Josef Strzygowski and Gertrude Bell on Amida (modern Diyarbekır) used the inscriptions to reconstruct the history of the city (see Figure 5.21). This landmark work was done entirely on the basis of the photographs taken by General de Beylié without van Berchem ever visiting the city. The intensive study of the inscriptions from the central Islamic lands is also due to their material: many are finely carved in limestone, and their well-preserved state

makes it easier to study them and to appreciate their elegant epigraphic style. Even a well-published city such as Jerusalem shows how epigraphic study has advanced over the course of the twentieth century. During two long visits to the city, van Berchem recorded 300 Arabic inscriptions, which were prepared by Gaston Wiet for publication in the MCIA after van Berchem’s death in 1920. Since then, many new inscriptions have been found, including twenty-four published by Burgoyne and Abul-

— 55 —

— ISLAMIC INSCRIPTIONS —

5.21 Foundation text on the Kharput Gate at Diyarbekır dated 297/910.

Hajj (1979) as well as other individual texts (for example, Sharon 1977; Burgoyne 1982). These have been included in the handlist and maps of Arabic inscriptions from Jerusalem prepared by Walls and Abul-Hajj (1980). The subject has grown so vast that Sharon (1974) announced preparation of a new corpus of inscriptions from Palestine in three volumes, one on Jerusalem and two on other regions. Similarly, Gaube (1978) compiled a list of some 250 inscriptions from Syria in the later periods. The material compiled by van Berchem at the beginning of the twentieth century was used by Samuel Flury to conduct his own series of landmark studies on the evolution of epigraphic ornament in these superb inscriptions. Flury’s work on the eleventh-century inscriptions of Diyarbekır appeared both as a book in German (1920) and as a series of three articles in French published in the journal Syria in 1920 and 1921. In another series of articles (1912, 1925b, 1936) Flury used the information made available in

van Berchem’s volumes of the MCIA on Egypt to trace the extraordinary evolution of floriated Kufic under the Fatimids there (979–1171). Flury’s work is notable for its visual documentation. His fine drawings of the inscriptions bring out their artistic merits, and his charts showing the different forms of the same letter (see Figure 5.23 below for an example from the eastern Islamic lands) enable the reader to distinguish the imaginative decorative motifs. Grohman (1957) moved to a more theoretical level, tackling the problem of the origin and early development of floriated Kufic, which he considered the culmination of Arabic script. By comparing examples of lapidary styles on tombstones and buildings as well as manuscript hands, he concluded that there was a gradual development from simple Kufic to foliated script (in which the ends of letters were decorated with half-palmettes or leaves) to floriated script (in which flowers grow from the ends, or even the middles, of letters). He showed how the upright

— 56 —

— REGIONAL STUDIES — alif was successively decorated with a hook, forked top, half-palmette and trilobed palmette, decorative motifs that then moved to other letters, and argued that the transformation from foliated to floriated script had already taken place in Egypt by the mid-ninth century, based on a tombstone in the Cairo Museum dated 243/848. The information presented in these stylistic analyses has been used in turn to address the question of content. Several scholars have studied the inscriptions from Fatimid Cairo from the point of view of heterodoxy and used the inscriptions to set the monuments in their political and religious contexts. Williams (1983, 1985), for example, used inscriptions along with architectural analysis to argue that the Aqmar mosque and the myriad small tombs built under the Fatimids were part of an officially sponsored cult of ÆAlid martyrs and saints used to generate support and loyalty for the IsmaÆili imam/caliph. Bloom (1983) suggested that the inscriptions on the exterior of the Hakim mosque were chosen as part of an ideological programme that evolved during the course of construction from 989 to 1013. In another article (1987b), Bloom argued that the special form of the tasliya or eulogy that appeared on Egyptian tombstones in the tenth century was evidence for the conversion of the population to ShiÆism. Taylor (1992) criticised these approaches as overemphasising the role of ShiÆism in the development of Islamic funerary architecture. His critique contains some useful suggestions, such as the multivalent nature of Koranic texts, but it fails to give other explanations for the changes in inscriptions, unjustly privileging textual evidence over epigraphic. Other scholars have examined the content of inscriptions from Syria and the Levant to trace the development of the concept of jihåd as a propaganda weapon against the Crusaders. This line of enquiry was begun by Elisséeff in a seminal article published in the Bulletin des Études Orientales (1952–4) showing how the titulature of the Zangid ruler Nur al-Din (r. 1146–74) reflected his role in the jihåd. Tabbaa (1986) extended the argument to show how the buildings themselves, along with their inscriptions, were part of Nur

al-Din’s campaign. Carole Hillenbrand (1994) showed how the inscriptions, both historical and Koranic, on monuments and buildings erected in Syria in the first half of the twelfth century already reflect the preoccupation with jihåd there. One of the key monuments from this region, already published by Sauvaget in 1938, is the cenotaph of the Artuqid prince of Aleppo, Nur al-Dawla Balak, one of the most formidable adversaries of the Crusaders, who was killed outside Manbij in 518/1124. More recently, other scholars have opened a new line of enquiry, trying to connect the stylistic changes themselves with political ones. Bierman (1989, 1997), for example, argued that the Fatimids were the first to create a public text by making writing in Arabic the dominant visual programme on buildings, by suppressing other detracting decoration, and by placing the texts in focal areas. She connected the elaboration of script to IsmaÆili symbolism and other popular beliefs, suggesting, for example, that the interlaced låm-alif reflects the interior (båtin) hidden or esoteric meaning behind the outside (Ωåhir) appearance. Tabbaa (1994), following on her work and that of JumÆah (1969), argued that the development of floriated Kufic was not a gradual transformation from simple to foliated and then floriated script but a sudden transformation under the Fatimids, who created the script to reaffirm dynastic claims to legitimacy and distinguish their line from competitors. In the same vein, he argued that cursive script was introduced to Syria by the Zangid ruler Nur al-Din for political reasons as part of the Sunni revival in opposition to the heterodox Fatimids, who were associated with floriated Kufic. These attempts to impute political motivations to stylistic changes, however enticing, are wrong, for they are not grounded in an accurate knowledge of the epigraphic material. They overlook significant specimens, omit material that does not fit with preconceived hypotheses, and disregard events that occurred elsewhere. There are, for example, several ‘public texts’ that pre-date the Fatimid period, such as the one on the Aghlabid Mosque of the Three Doors in Qayrawan, erected

— 57 —

— ISLAMIC INSCRIPTIONS — in 866 by the Andalusian Muhammad b. Khayrun al-MaÆafiri. Its façade is decorated with a bold three-line inscription in simple Kufic (RCEA 549; Kircher 1970). Even in Cairo there are also fine examples of floriated Kufic that pre-date the arrival of the Fatimids (Bloom 1997). One is a stone plaque recording the foundation of the Biær al-Watawit by the Ikhshidid vizier Ibn al-Furat in 355/966 (see Figure 5.22). Van Berchem included the plaque in his volume of the MCIA on Egypt (no. 48) because of its elegant floriated script, but, since he was able to read only a few words, he assigned the inscription on stylistic grounds to the eleventh or perhaps the late tenth century. Wiet, however, recognised the inscription from a passage in alMaqrizi’s history, and in his supplement to van Berchem’s volume Wiet placed the inscription in its correct historical setting (MCIA Egypte 2, no. 570). Public texts written in floriated Kufic also exist on buildings outside the Fatimid

domains: the Arab-Ata mausoleum at Tim in the Farghana Valley of Central Asia, for example, has a medium-sized (35 cm) and very long (21-metre) inscription band in floriated Kufic framing the portal and ending with the date 367/977 (Blair 1992a, no. 11). Both Bierman and Tabbaa also ignore the evidence presented by grave markers, which document a gradual development of floriated and interlaced Kufic in the late tenth and eleventh centuries. This evidence includes not only the well-known lengthy list of Egyptian tombstones mined by Grohman but also the more recently discovered series of crested grave covers found at Siraf on the Persian Gulf. A particularly fine example (see Figure 14.84) made for Ibrahim b. ÆAli, who died in 383/993, shows superb floriated Kufic with intricately laced låm-alifs (Lowick 1985, no. 3). Bierman’s reason for dismissing the grave markers is that they were not official texts, but such an approach distorts the evidence.

5.22 Stone plaque recording the foundation of the Biær al-Watawit in Cairo by the Ikhshidid vizier Ibn al-Furat in 355/966.

— 58 —

— REGIONAL STUDIES — This type of reasoning also points up the dangers of arguing from a few prominent examples that have survived, such as the Azhar and Hakim mosques in Cairo, although we know that many more inscriptions were made. Virtually no monumental inscriptions are left from the ÆAbbasid capitals in ÆIraq, but we cannot take this to mean that the ÆAbbasids did not use inscriptions on their buildings. The ÆAbbasid use of inscriptions is well known from contemporary textual sources, such as the detailed account of the rules and regulations of the ÆAbbasid court entitled RusËm dår al-khilåfa written by the secretary and director of the chancellery at Baghdad, Hilal al-Sabiæ. He described, for example, how the ÆAbbasids meticulously inscribed all sorts of textiles that made up the setting for their investiture – everything from banners and the poles that supported them to the leather cushions for the seat of honour – with what the ÆAbbasids considered appropriate slogans, including Koranic verses and the names and titles of the caliph (1977: text pp. 129–30, trans. p. 76). The case of Diyarbekır shows that styles of script developed gradually and that stylistic changes did not always coincide with dynastic ones. The inscriptions in floriated Kufic there – considered by Flury the finest examples of the script, some of the most beautiful inscriptions of their time, and more sophisticated than those done in contemporary Egypt under the Fatimids – were added under the Marwanids and Saljuqs (neither an overtly ShiÆite, let alone IsmaÆili dynasty). Cursive was introduced there at the end of the twelfth century. The first extant inscription in cursive script at Diyarbekır records the opening of a postern in the rampart near Tower LVII by the Nisanid vizier Bahaæ al-Din in 578/1182. This type of sober cursive became a hallmark of inscriptions added under the Ayyubids, who took the city in 1183, but the script was already used at least a year before the Ayyubids conquered the city. Thus, it is clear that epigraphic changes were not the result of dynastic ones. Rather, politicians in the twelfth century were quick to seize upon stylistic changes and could turn epigraphic style to their advantage, transforming

a new but already developed script into a dynastic hallmark, as with floriated Kufic under the Fatimids in Egypt or cursive under the Zangids and Ayyubids in Syria. As with Egypt, van Berchem was one of the pioneers of epigraphic study in Anatolia, and the tradition begun by van Berchem and Edhem in their volume of the MCIA on Asia Minor of recording the inscriptions from Anatolia city by city has continued. Several Turkish scholars, such as Uzunçar∞ılı and Konyalı, have compiled the inscriptions from smaller towns, but their studies are often difficult to find outside Turkey and somewhat uncritical in their approach. The French have continued the same urban compilations. The architect-archaeologist Albert Gabriel, for example, gathered much valuable documentation of inscriptions in his journeys throughout Anatolia in the 1930s, 1940s and 1950s. Sometimes this work was included within his studies of Anatolian buildings, such as his two-volume work on Kayseri, Ni©de, Amasya and Tokat (1931–4) or his book on Bursa (1958). Sometimes the material was presented separately, as in Sauvaget’s appendix on Arabic inscriptions in Gabriel’s study (1940) of monuments in eastern Anatolia. The French historian Robert Mantran (1952–4, 1959) catalogued both the Arabic and the Turkish inscriptions from Bursa. One recent example of this genre is the catalogue of the Ottoman inscriptions of Edirne edited by F. Th. Dijkema (1977). The volume presents a critical edition with translations and commentaries of the 176 monumental inscriptions in Arabic script that survive in the town (see Figure 2.8). Only inscriptions with historical value are included; pious texts, poetry, dates, signatures, inscriptions on portable objects and tombstones are not. The critical readings were verified by onsite surveys in 1969 and 1972. The inscriptions are arranged not by monument, as in the volumes of the MCIA, but by date, with an index of monuments for cross-reference. Such an arrangement emphasises the epigraphic and palaeographic value of the inscriptions over their importance for the history of architecture. Dijkema’s volume is most useful for its record of individual inscriptions, and the brief introduction

— 59 —

— ISLAMIC INSCRIPTIONS — only hints at possible broader conclusions. The author notes, for example, that almost all the inscriptions are enclosed in cartouches of ten different shapes, each particular to a certain period. Hence, it is sometimes possible to determine the date of an undated inscription or to demonstrate that an inscription is a later copy of an earlier text. The evolving medallion shapes might well be applied for dating objects in other media, as in Raby’s 1986 study on the forms of medallions used in U∞ak carpets. The rich data compiled by Dijkema could be exploited for many other analyses. One could investigate, for example, urban developments and the changing nature of patronage. Most of the inscriptions from the fifteenth century, for example, record the construction of mosques, whereas public taps (çe∞me) and drinking-fountains (sabÈl) were more common in later times. Another subject is linguistic change. The first inscriptions are written in Arabic. Turkish was introduced in the mid-sixteenth century (nos 28 and 29) and quickly became standard, although Arabic was still used occasionally, particularly for mosques, and Persian was used exceptionally on several public taps and tanks (∞ådirvån). This fits the overall picture known from Anatolia, where Turkish remained a lingua franca longer than did other languages elsewhere. One of the few essays that attempts to exploit the scattered data from individual cities in Anatolia is Rogers’ 1976 article, in which he used Saljuq foundation inscriptions from thirteenth-century Anatolia as a source to illuminate architectural practices in medieval Islamic times. In it, he analysed the four standard components of foundation inscriptions – the initial statement of foundation, the titles of the reigning sultan, the name of the individual patron, and the date. He used the epigraphic evidence to argue that between any founder and the craftsmen who put up his building, there was a staff of officials headed by a Clerk of the Works who took the major architectural responsibility out of the founder’s hands. Another area of recent interest is the Gulf, and Italian scholars have been active in publishing the inscriptions from Muscat and Oman. Baldissera

has published several books and articles in Quaderni di studi arabi, a journal produced by the University of Venice, on the monuments there, most of which belong to the seventeenth century or later, and is preparing a fuller monograph on the subject.

Iran, Afghanistan and Central Asia Works about the epigraphy of Iran, Afghanistan and Central Asia are scattered and diverse. Following the type of work begun by the MCIA, Blair (1992a) catalogued the seventy-nine earliest inscriptions from the region, up to the year 500/1106, and the introductory essay is one of the few overviews of the epigraphic tradition there. The work, however, was done mainly from published sources; on-site verification was not always possible, and some of the transcriptions are inaccurate. Most works address Iranian inscriptions by area. The series published by the Iranian government on various cities and provinces, Åthår-i millÈ, often includes inscriptions within the descriptions of individual monuments, although the texts are not analysed and the readings sometimes bizarre. A more rigorous study of the inscriptions from the city of Isfahan and its environs is the fat and informative volume by Hunarfar (1977) which not only describes and records the inscriptions building by building but also sets the particular building in its historical context. Drawings and translations of the inscriptions from several sites and objects published in the Survey of Persian Art are published by Ghulam (n.d.) , although the text is not critical. A (very) few inscriptions from Islamic Iran have also been published in volume IV of the Corpus Inscriptionum Iranicarum (CII) dealing with Persian inscriptions down to the early Safavid period. Only two publications have appeared to date. Portfolio I of volume II on the province of Khurasan, edited by William Hanaway, contains twenty-four plates from three buildings in Mashhad and Nishapur. Portfolio I of volume VI on Mazandaran, edited by A. D. H. Bivar and E. Yarshater, contains seventy-two plates from several sites in the eastern part of the province. These

— 60 —

— REGIONAL STUDIES — hard-to-find and expensive volumes offer superb plates but only limited transcriptions, translations or commentaries of isolated examples. Inscriptions from several sites in Central Asia are treated in articles in Epigrafika Vostoka, a journal begun in 1947 under the editorship of the noted Russian epigrapher V. A. Krachkovskaya and devoted to inscriptions found in the former Soviet Union. After a hiatus in the 1970s, publication resumed, reaching volume 24 in 1988. A summary and analysis of the most important articles dealing with Islamic subjects in the first eight volumes was made by Grabar (1957) in the journal Ars Orientalis, and, beginning with volume 24, English summaries of the Russian articles are included in each volume. Many of the articles deal with coins and other objects in museums, but there are also articles on individual buildings or their furnishings, such as Semenov and Masson (1948) on the Gur-i Mir, Timur’s tomb in Samarqand. There are also several important surveys, such as Krachkovskaya’s 1949 article on the evolution of Kufic script in Central Asia. The articles present information that is not published elsewhere and is often inaccessible to outsiders. Many are informed by the authors’ superb knowledge of Central Asian history, but, like many works of traditional Soviet scholarship, they often suffer from unfamiliarity with work done outside the traditional borders of the Soviet Union and from an unwillingness to make general conclusions or put the inscriptions in a broader context. Many monumental inscriptions have been recorded in Afghanistan during the second half of the twentieth century, and our knowledge of epigraphic developments during medieval times there is rapidly changing. Probably the most exciting discovery was the minaret of Jam (see Figure 15.86), hidden in a remote valley in central Afghanistan (Maricq and Wiet 1959). The intact, three-storey minaret is covered with inscriptions, including all 976 words of Chapter 19 of the Koran, Surat Maryam, and a bold band in blue tile gives the name of the patron, the Ghurid sultan Ghiyath al-Din Muhammad b. Sam (r. 1163–1203). Such an elaborate programme must have been consciously chosen, and analysis of it

should shed light on the function of this enigmatic building. Even such a famous and well-published building still maintains its epigraphic secrets: the minaret is also said to contain a dated inscription over the doorway, but the inscription has never been published and the date has been read variously as 570/1174–5 or 590/1193–4. Other inscribed buildings recently discovered in the region include the small brick tomb known as that of Salar Khalil or of Baba Hatim 60 km west of Mazar-i Sharif in northern Afghanistan (Melikian-Chirvani 1968; Sourdel-Thomine 1971; Bivar 1977; Schneider 1984), a large madrasa dated 571/1176 at Shah-i Mashhad in Gargistan in north-western Afghanistan (Casimir and Glatzer 1971), and the imamzada of Yahya b. Zayd at Sar-i Pul (see Figure 4.17) in the province of Tukharistan in northern Afghanistan (Bivar 1966; Blair 1992a, no. 75). Elsewhere in Afghanistan, inscriptions have been uncovered beneath later work in wellknown buildings. This happened at the Friday Mosque at Herat, where much of the Ghurid work has recently been published, including an elaborate inscription dated 597/1201 uncovered under a Timurid portal in the north-east corner of the mosque (Melikian-Chirvani 1970). Sometimes new discoveries have been made by rereading well-known inscriptions from remote sites. Such is the case with the minaret at Ghazna popularly attributed to Mahmud (r. 998–1030). In one of her earliest and most important articles (1953), Sourdel-Thomine redated the minaret to a century later by correctly reading the name of Mahmud’s successor Bahram Shah (r. 1118–52) spelled out in relief bricks at the top of the lower storey. Similarly, the tomb at Safid Buland in the Farghana Valley of Central Asia (see Figure 2.7) was dated by Cohn-Wiener (1939) to the twelfth century on stylistic grounds, but when Nastich and Kochnev (1988) read the foundation inscription, they showed that the tomb could be dated a century earlier (c. 450/1055) (Re)reading the inscriptions on the building at Zawzan, Blair (1985b) confirmed its date in the early thirteenth century and established its function as a madrasa. Correctly reading the patron’s name or the date on a building is particularly important for

— 61 —

— ISLAMIC INSCRIPTIONS — establishing fixed points in the development of epigraphic styles, a subject that has received considerable attention in the study of epigraphy from the eastern Islamic lands. The choice of this topic is not surprising, as this area was a major centre of artistic creativity in medieval times. As in other regions, the pioneering scholar in studying the varied style of scripts used in the medieval Iranian world was Samuel Flury. He published superb drawings of the fine foliated Kufic inscriptions on the congregational mosque at Naæin (see Figure 5.23), attributed to the mid-tenth century (Viollet and Flury 1921; Flury 1930; Blair 1992a, no. 9), and, working from photographs brought back by Godard, analysed many of the monuments from Ghazna (Flury 1925a). As in the central Islamic lands, in the eastern Islamic lands floral elements continued to develop until foliated Kufic turned into floriated Kufic in which floral motifs, tendrils and scrolls grow from the ends and even the middle of letters. Fine floriated Kufic is already attested on the portal at Tim dated 367/977 (Blair 1992a, no. 11), and the style blossomed under the Saljuqs. Good examples are found on the early twelfth-century Friday mosque and Masjid-i Haydariyya at Qazvin. Sourdel-Thomine (1974) recorded and analysed the text of the extraordinary inscriptions there, and her students Bergeret and Kalus (1977) analysed the epigraphic and floral decoration of these inscriptions. While foliated and floriated decoration are found throughout the Islamic lands, one type of script – interlaced Kufic – is tied more closely to Iran and Central Asia. Again Flury was a pioneer in bringing the script to public attention, with his superb drawing of the interlaced Kufic used on the tomb tower built in the Radkan valley south of Sari for the Bawandid amir Abu JaÆfar Muhammad and completed in 411/1021. His drawing highlights the exuberant use of interlacing in, above and between single letters and groups, and the tomb tower is often cited as the classic example of interlaced script. Interlaced script was not confined to buildings in north-eastern Iran. The tomb covers recently

discovered at Siraf (see Figure 14.84) present further examples of how this script was used in monumental inscriptions in the late tenth and eleventh centuries (Lowick 1985) Interlaced script was also used on slip-painted ceramics (see Figure 1.3) attributed to the Samanid dynasty, who ruled Transoxiana from 819 to 1005. Based on her analysis of the script on these wares, Volov [Golombek] (1966) traced a formal development, dividing the script into three series based on successive transformations to the letters through natural distortion, internal modification and superimposed ornament. In addition to formal considerations, the content of the inscription also affected the style, for interlacing interfered with legibility (Blair 1989a). Interlaced Kufic was therefore more popular for pious or Koranic inscriptions than for foundation inscriptions, where it is essential to be able to read the patron’s name. The tomb tower known as the Pir-i ÆAlamdar at Damghan, dated 417/1026–7, shows how various styles of scripts were combined on the same building depending on the content of the text and the technique: the foundation inscription around the exterior in cut brick is much simpler than the Koranic text painted around the interior (Adle and Melikian-Chirvani 1972; Blair 1992a, no. 34). In order to make interlaced script more readable, the interlacing was often removed from the body of the letter and restricted to the upper zone. This type of Kufic with interlaced stems can already be seen in the inscriptions around the minaret at Tirmidh dated 423/1031–2 (Blair 1992a, no. 38). Eventually the interlacing was removed altogether from the letters below into an independent upper band, as in the foundation inscription across the south wall of the dome chamber in the mosque at Barujird datable c. 533/ 1139 (Blair 1994) or on the top of the minaret at Dawlatabad in Afghanistan (see Figure 6.27). Various types of interlaced Kufic continued to be used in the Mongol period, but were somewhat of an anachronism, as on the tomb tower erected in the first decade of the fourteenth century for the Ilkhanid prince behind the congregational mosque at Bastam (Blair 1982).

— 62 —

— REGIONAL STUDIES —

5.23 Flury’s chart of the letter shapes used in the inscription on the mosque at Naæin, mid-tenth century.

— 63 —

— ISLAMIC INSCRIPTIONS — Assigning dates to particular epigraphic styles used in the eastern Islamic lands is tricky. The problem is particularly acute when dealing with scripts used under the Ghaznavids and Ghurids, and the same monument can be assigned to different centuries on stylistic grounds. One example is the tomb recently discovered 60 km west of Mazar-i Sharif in northern Afghanistan; Melikian-Chirvani (1968) and Bivar (1977) considered it Ghaznavid; Sourdel-Thomine (1971) Ghurid. Sourdel-Thomine was also one of the principal investigators of the palace at Lashkari Bazar/Bust (Schlumberger and Sourdel-Thomine 1978), assigning the decoration of the main south palace there on stylistic grounds to the late eleventh or twelfth century in opposition to Schlumberger’s dating of the palace on archaeological grounds to the reign of Mahmud. Based on her analysis, Sourdel-Thomine also reassigned various other buildings and monuments in the area, such as the tomb at Sangbast (Sourdel and Sourdel-Thomine 1979) and the cenotaph of Mahmud (Flury 1918; Sourdel-Thomine 1981) to the twelfth century, arguing that much Ghurid architectural decoration, including the inscriptions, was a conscious revival of Ghaznavid work. Sourdel-Thomine’s stylistic analyses, however, are not universally accepted. Allen (1989), for example, challenged her twelfth-century dating of the south palace at Lashkari Bazar, arguing that the archaeology of the site supported Schlumberger’s earlier dating and that Sourdel-Thomine weighted her argument by unfairly choosing all her comparative material from the later date. Revising the date of the south palace at Lashkari Bazar would weaken Sourdel-Thomine’s arguments for dating other monuments to the twelfth century. Obviously a review of all the material working from clearly dated examples is needed to establish a relative chronology of epigraphic styles in the region.

Indian Subcontinent There is a long tradition of publishing the rich epigraphic material from the Indian subcontinent. At the end of the nineeteenth century, the Archaeological Survey of India established a separate

section for epigraphy and began to publish a specialised series, Epigraphia India. This was soon converted into an even more specialised work, Epigraphia Indo-Moslemica (often abbreviated EIM), which was published from 1907 until the partition of India and Pakistan, when its name was changed to Epigraphia Indica, Arabic & Persian Supplement (EIAPS). EIM along with its successor EIAPS is the fundamental work for the study of Islamic inscriptions from the Indian subcontinent. Anything published earlier can be found in the list compiled by Horovitz in 1909–10 containing 1,249 pre-EIM inscriptions which date between 589/1193, the year in which Qutb al-Din Aybak conquered Delhi, and 1274/1857. Another major source for the inscriptions of the subcontinent is the Annual Report on Indian Epigraphy (ARIE), published since the late nineteenth century. All of these works are mainly descriptive rather than analytical, but are extraodinarily rich in historical data. Bendrey (1944), for example, used the epigraphs published in EIM to set the monuments of India in chronological order, but, as he knew neither Arabic nor Persian, his survey missed some significant points. The key role played by the Archaeological Survey of India in recording the buildings there means that its members are some of the preeminent figures in the field of Indian Islamic epigraphy. The first was Ghulam Yazdani, who was superseded by Ziaud-din A. Desai, formerly head of Persian and Arabic epigraphy at the Archaeological Survey of India. The second supplement to the Creswell bibliography (Pearson 1984), for example, records thirty-three articles by him, mainly recording and transcribing individual inscriptions from the subcontinent. Desai has also written several overviews on the problems of Islamic epigraphy in India, such as his essay in the volume on Indian epigraphy edited by Asher and Gai (1985). This volume also contains studies of the inscriptions on individual buildings by Anthony Welch M. C. Joshi, Catherine Asher and Wayne Begley. The vast scope of the subcontinent, not to mention the current religious, ethnic and political divisions there, has given rise to several problems

— 64 —

— REGIONAL STUDIES — within the overall study of Indian Islamic epigraphy. One is the differing amounts of publication accorded to different areas of the subcontinent, depending in part on accessibility. A second problem is the lack of availability of the publications, for not all materials are always available to all scholars working on a particular area. Works published in India, for example, are not always available in Pakistan, let alone in the West. This means that subsequent studies do not always build on the back of previous work. A further problem is the isolation of Indian Islamic studies: most of the scholars working in the field know the subcontinent (or parts of it) very well, but are less familar with other areas of the Islamic lands, and they tend to draw conclusions on the basis of only the Indian material without considering the use of similar texts elsewhere. Inscriptions from several cities and areas in the subcontinent have been published, such as Tirmizi (1968) on Ajmer, or Desai (1971) on Rajasthan, but the area whose inscriptions are best recorded is Bengal. Dani (1957) published a bibliography of the inscriptions there as an appendix to volume 2 of the Journal of the Asiatic Society of Bengal, and Shamsud-din Ahmed’s Inscriptions of Bengal (1960) included texts and translations of the 214 inscriptions issued by the Muslim rulers of Bengal between 1233 and 1855. In 1992, Abdul Karim published a corpus of 310 inscriptions from Bangladesh and west Bengal, with texts, translations and analysis of historical context as well as supporting photographs and rubbings. Even Karim’s corpus is not complete, and, in reviewing the book, Mehrdad Shokoohy (1995) noted that Karim did not have access to the ARIE and listed some sixty other inscriptions, mostly from the Mughal and post-Mughal periods, published there. Shokoohy further pointed out that by covering only historical texts, Karim omitted Koranic inscriptions, which are important not only for dating purposes but also for the development of style. Siddiq’s essay (1990) surveys the distinctive epigraphic style that developed in the region, but he was apparently unaware of (or at least did not cite) the previous works by Dani, Ahmad and Karim.

Another well-researched area of epigraphic study is central India, particularly the imperial Mughal cities of Agra, Delhi and Fatehpur Sikri. Hussain’s volume (1936) on the Koranic inscriptions of Delhi has been used by Anthony Welch for several essays analysing the meaning of the Koranic inscriptions on buildings there, such as two sultanate tombs (1985). The survey which he made with Howard Crane on Tughluq architecture in the Delhi sultanate (Welch and Crane 1983) also contains a section on religious epigraphy. Nath (1979) has taken up the broader subject of the role of calligraphy in Mughal architecture, although his analysis concentrates more on interpreting the symbolic meaning of the texts rather than on describing their artistic merits. The sourcebook on Fatehpur Sikri compiled by Brand and Lowry (1985) also contains a section on the inscriptions by Z. A. Desai. The building whose incriptions have received the most work is, not surprisingly, the Taj Mahal (see Figure 5.24) Begley and Desai (1989) compiled and translated the inscriptions on the building as part of an anthology of seventeenth-century Mughal and European documentary sources on the building. In a straightforward historical article (1978–9), Begley used the inscriptions as a source for the life of the calligrapher Amanat Khan. In another, more speculative article, Begley (1979) used the Koranic inscriptions and other evidence to argue that the Taj Mahal was a symbolic replica of the heavenly Throne of God. One of the most imaginative articles about interpreting Koranic texts, Begley’s work has generated a certain amount of criticism. Some conservative Muslims find his idea distasteful or even blasphemous; some Western scholars might criticise its lack of subtlety and dogmaticism, for it overlooks the multivalent meanings of Koranic texts and the different ways in which they could be interpreted by different audiences. Epigraphy is particularly useful in documenting the coming of Islam to the Indian subcontinent, a subject of continuing controversy, and inscriptions from some of the oldest Islamic monuments there have elicited ready publication. Studying them also shows some of pitfalls of interpreting

— 65 —

— ISLAMIC INSCRIPTIONS —

5.24 Beginning of the Koranic band with Chapter 89 (Surat al-Fajr) on the south side of the gateway to the Taj Mahal at Agra, completed 1057/1647.

— 66 —

— REGIONAL STUDIES — Indian material without considering the broader Islamic context. A good example is provided by the inscriptions on two stone slabs found during the excavations of the mosque at Banbhore near Thatta in Sind province, now in Pakistan. The excavators (Khan 1963) dated one of the slabs to 109/727–8, making the building the earliest mosque in the region and important evidence for the presence of Islam there. The text, however, uses the phrase mawlå amÈr al-muæminÈn, a title that did not appear until the mid-eighth century, and the final word in the inscription should probably be read not miæa (‘hundred’) but miæat[ayn] (‘two hundred’), making the date 209/824–5. This would put the construction of the mosque closer to that of another mosque recorded in an Arabic– Sanskrit inscription dated 243/857 from the Tochi valley of Waziristan (Quraishi 1925–6; Bosworth 1965) Another group of inscriptions from medieval Islamic times that have received significant publication are those found at Badreªhvar on the Gulf of Kachh in the coastal region of Gujarat. First published by Desai (1965), the inscriptions are also analysed in Shokoohy’s monograph

(1988). Blair (1989) noted that al-Sirafi, the epithet of several people named on the shrine dated Dhuæl-Hijja 554/December 1159–January 1160 and a headstone dated 569/1174, connected the site with the active port on the Persian Gulf, where carvers produced elaborate crested grave covers decorated with fine interlaced Kufic inscriptions (see Figure 14.84) from the late tenth to the twelfth century (Lowick 1985). Inscriptions allow us to track the movements of these Sirafi merchants and craftsmen: some of the best Sirafi craftsmen, for example, migrated to Zanzibar where they carved the inscriptions in superb interlaced Kufic in the Kisimkazi Mosque dated 500/1106–7 (Flury 1922) The late twelfth century was the time of the Ghurid expansion into the subcontinent, and Holly Edwards’ analysis (1991) of the inscriptions on the structure built by ÆAli b. Karmakh near Multan in present-day Pakistan, and often considered the first Muslim tomb in the subcontinent, sheds light not only on the building’s function but also on the more general historical setting during the expansion of Islam into the area.

— 67 —

— CHAPTER 6 —

Inscriptions on Various Building Types

A

third way of dividing the vast corpus of monumental inscriptions that survive from the Islamic lands is to consider the inscriptions by the type of building or construction on which they were written. Whereas a functional typology brought out linguistic differences and a regional division brought out stylistic ones, looking at the inscriptions by building type brings out societal and institutional changes. Mosques, madrasas and minarets are considered the distinctive types of Islamic buildings, and studying the inscriptions on them can help us follow the development of Islamic society and its institutions. The fourth type of building analysed here, city walls and gates, is not distinct to Islamic civilisation, but is intended as a case study. A similar study might be made of other types of buildings, such as hospitals or fountains, as studying the inscriptions on them can help trace the changing nature of piety and communal life. Inscriptions on mosques have been treated most thoroughly. Inscriptions were a prominent part of mosque decoration from earliest times, and the Umayyad mosque at Damascus built by the Umayyad caliph al-Walid I c. 715 had an extensive epigraphic programme in blue and gold mosaic. Although destroyed, the text can be reconstructed from medieval descriptions. It included pious phrases, Koranic citations and the foundation inscription with the caliph’s name (Finster 1970). Mosque inscriptions were almost invariably written in Arabic, presumably because of the sanctity accorded to Arabic as the language in which the Koran was revealed. The mosque of Shaykh Lutfallah erected by Shah ÆAbbas at Isfahan

in the early seventeenth century (see Figure 4.19), for example, has an extensive programme of Arabic texts (Hunarfar 1977: 401–26). The band below the dome, signed by the calligrapher ÆAli Riza ÆAbbasi in 1025/1616–17, contains a Prophetic hadith about the rewards for pious Muslims and ÆAlid traditions about activities in mosques. The band below the windows, also signed by ÆAli Riza ÆAbbasi, is Koranic (Chapters 62, Surat al JumuÆa, and 110, Surat al-Nasr). The band around the walls, signed by Baqir the builder, contains Arabic poems asking the Fourteen Innocent Ones to intercede for Shaykh Lutfallah in the hereafter. There is no single survey covering the historical and religious inscriptions used in mosques in different places and times throughout the Islamic lands, but Blair compiled a brief survey of mosque inscriptions up to the eleventh century in The Oxford Encyclopaedia of Archaeology in the Near East (1996, vol. 4, pp. 58–60). It shows that foundation inscriptions in mosques are useful historical documents to measure political allegiance, for just as the khu†ba or sermon in the congregational mosque was a sign of authority, so the foundation inscription there was a testament to official policy. The inscriptions in the earliest mosques seem to illustrate general or pan-Islamic themes, such as God’s glory and greatness, but with the splintering of the Islamic community from the tenth century onwards, the inscriptions became more sectarian in nature. The Fatimids, for example, seem to have chosen verses with key phrases about the ShiÆite daÆwa to decorate the mosques in their capital Cairo (Bloom 1987b).

— 68 —

— INSCRIPTIONS

ON

VA R I O U S B U I L D I N G T Y P E S —

Using the corpus of Koranic inscriptions compiled by Dodd and Khairallah (1981), Robert Hillenbrand (1986) and Thackston (1994) investigated the Koranic inscriptions used on mosques. The most common is Koran 9:18, undoubtedly chosen because it is one of only three Koranic references to God’s mosques (masåjid allåh), a special term distinct from any masjid or place of prayer. Other common Koranic citations on mosques include the so-called Throne Verse (2:255) extolling God’s majesty, the Light Verse (24:35) in which God is eloquently described as the light of the heavens and the earth, and verses from the Chapter of Victory, Surat al-Fath (48), about God’s granting a manifest victory. Some Koranic verses became associated with parts of mosques because of individual words referring to specific practices. Verses 17:78–9 about prayer and vigil, for example, are often found on mihrabs. Another popular verse used to decorate mosques, 2:181, refers to the inviolability of endowments. From the eleventh century, hadith were inscribed alongside Koranic quotations on mosque furnishings and furniture. Most of the hadith used are not recorded in Wensinck’s extensive Concordance et indices de la tradition musulmane (1936–88), which lists traditions from canonical sources. These other hadith may well have been coined for the occasion to present specific viewpoints and are therefore important evidence for popular belief. They deserve more systematic study. One of the earliest examples is found on a wooden mihrab removed from the mosque at Iskodar in the Farghana Valley of Central Asia and datable on stylistic grounds to the early eleventh century (Blair 1992a, no. 27). The hadith describes the inestimable advantages of guarding the front line, a text suitable for the site on the eastern frontier of Islam. The earliest dated example of a hadith occurs on the minbar that Badr al-Jamali donated in 484/1081 to the sanctuary built to hold the miraculously discovered head of Husayn in Ascalon and later moved to the haram in Hebron (van Berchem 1915; Wiet 1924; RCEA 2791). The traditions inscribed there, in which the Prophet declares his two legacies to be the

Koran and his family, are a vindication of ShiÆite claims to legitimacy and must have been consciously used as a justification for the shrine. Hadith soon became more popular in decorating mosques and their furnishings. The mihrab added to the congregational mosque at Isfahan in Safar 710/July 1310 following Uljaytu’s conversion to ShiÆism (See Figure 6.25), for example, cites a hadith of ÆAli b. Abi Talib that whoever frequents a mosque will receive one of eight benedictions (van Berchem 1909a; Hunarfar 1977: 116–20). Such a hadith must have been cited specifically to win over the population in this troublesome sectarian city. An even more pointed example is found in the Shah-i Zinda, the cemetery that grew up in old Samarqand around the purported grave of the Prophet’s cousin, Qutham b. ÆAbbas. The door leading from the domed crossing at the north end of the corridor to the mosque adjoining the tomb of Qutham b. ÆAbbas is inscribed with a hadith stating that according to the Prophet, the person most like him in appearance and character is Qutham (Shishkin 1970, no. XVII; Golombek and Wilber 1988, no. 21). Inscriptions on madrasas are less well known than those on mosques, although it is often only the word used in the foundation inscription that distinguishes the building as a madrasa rather than a mosque. This was the case, for example, with the Ghurid building discovered in 1970 on the banks of the Murghab River in north-west Afghanistan (Casimir and Glatzer 1971). A large building measuring some 40 metres on a side with a central court and iwans, it has a foundation inscription around the portal identifying it as a madrasa endowed by a woman in Ramadan 571/ March–April 1176 (date corrected from the published 561/1166 in a personal note from the authors). The same situation holds for the large madrasa at Zawzan, located some 60 km south of Khwaf in eastern Iran. It was originally thought to be a mosque, but closer examination of its inscriptions yielded the date RabiÆ I 616/June 1218 and the identification as a Hanafi madrasa (Blair 1985b). The inscriptions used on madrasas can also help show us what role the founders envisaged they

— 69 —

— ISLAMIC INSCRIPTIONS —

6.25 Stucco mihrab added to the Friday Mosque in Isfahan in Safar 710/July 1310.

— 70 —

— INSCRIPTIONS

ON

VA R I O U S B U I L D I N G T Y P E S —

would play. These institutions were instrumental in the promulgation of Islam, particularly as part of the jihåd against the Crusaders in Syria, and many of the inscriptions on examples from Syria and the Levant deal with this subject (see Elisséeff 1952–4; Tabbaa 1986; C. Hillenbrand 1994). Notes and references to the inscriptions on many other madrasas are included in R. Hillenbrand’s article on madrasa architecture in the second edition of the Encyclopaedia of Islam. He points out, for example, that many madrasas in Syria included tombs and are often designated as funerary madrasas. The inscriptions referring to Koran recitation underscore the funerary nature of these ensembles, but the foundation texts distinguish the madrasa and the tomb (turba) as distinct parts.

Inscriptions show that the two parts of the funerary madrasa continued to be distinct in Mamluk times. This was the case with Qalaæun’s funerary complex (see Chapter 3) and also with the largest and finest tomb complex in Cairo, the huge one erected by Sultan Hasan in the midfourteenth century. The large foundation inscription running around the courtyard uses the word madrasa, while the band around the tomb calls the structure a qubba (MCIA Egypte 1, nos 168, 170; RCEA 764 002–3). Inscriptions also distinguish other parts of the complex, such as the four madrasas for the four schools of law (RCEA 764 004–10). Sultan Hasan’s madrasa is also decorated with several imposing Koranic inscriptions (see Figure 6.26), which have been analysed at length

6.26 Beginning of the Koranic inscription with Chapter 48:1–6 around the courtyard of the madrasa of Sultan Hasan in Cairo, completed 764/1362.

— 71 —

— ISLAMIC INSCRIPTIONS —

6.27 Minaret erected at Dawlatabad, Afghanistan in 502/1108–9.

— 72 —

— INSCRIPTIONS

ON

VA R I O U S B U I L D I N G T Y P E S —

by Dodd (Dodd 1969, substantially incorporated as Chapter 4 of Dodd and Khairallah 1981; see also the criticisms by Blair 1984). Many minarets bear inscriptions, often bands ringing the circular shafts. Van Berchem (MCIA Egypte 1, p. 481 and further in Diez 1918: 109–16) first addressed the philological problem of the minaret, and Bloom (1989) analysed how the different words such as miædhana, ßawmaÆa, manår or manåra reflect the emerging role of the minaret as a symbol of Islam. Erecting minarets became a popular form of piety in medieval Iran, where a group of almost seventy towers were built. Prominent bands encircling the shaft often contain the foundation text, for the patron was anxious to proclaim his work and establish his dominion. This is the case, for example, with the minarets erected by Abu Harb Bakhtiyar in Damghan and Simnan upon his accession as governor in the early eleventh century (Blair 1992a, nos 35, 37; Adle and Melikian-Chirvani 1972; Adle 1975). It was also the case with the minaret erected in 502/1108–9 at Dawlatabad near Balkh in Afghanistan (see Figure 6.27). The broad band at mid-height with tall cursive letters on an arabesque scroll ground gives the name and titles of the patron, Abu JaÆfar Muhammad b. ÆAli, who can be identified as the grandson of Nizam al-Mulk and vizier to the Saljuq sultan Sanjar (Sourdel-Thomine 1953). Minarets erected in Egypt, particularly in Mamluk times, were also inscribed (BehrensAbouseif 1985: 27–30, 191–7). While a few texts contain the sponsor’s name, Koranic verses became more popular. Although the verses cited may suit the construction which they adorn (for example, Koran 22:28, a text about pilgrimage, used by a sponsor who was in charge of supplying the kiswa to the KaÆba), most come from a limited repertory of Koranic texts with generalised meanings (Koran 2:255, 3:190–1, 24:35–8), referring to God as Creator of Heaven and Earth, summoning believers to prayer, or recalling the passage of hours, night to day, darkness to light, or times of prayer. These inscription bands were often set at the top of the minaret shaft, but despite their distance they were meant to be read or at least

recognised. This was particularly true in earlier periods, when legibility was enhanced by colour. The first method used was glazed tiles, first seen on the Iranian minaret erected at the end of the eleventh century in the congregational mosque at Damghan (Blair 1992a, no. 73; Adle 1982) and in 526/1131 at Sin (Smith 1939). Glazed tile was later replaced by tile mosaic. Minarets of fourteenth-century Iran often have inscriptions bands in tile mosaic set against an overall ground with repeating words executed in bannåæÈ (‘builder’s’) technique in which glazed bricks repeatedly spell out names or phrases against a ground of plain buff brick. The Mamluks considered the Iranian technique of tile mosaic so effective that at the beginning of the fourteenth century the Mamluk sultan al-Nasir Muhammad had a craftsman brought from Tabriz to execute inscriptions in tile mosaic (see Figure 6.28) to decorate the stone minarets on his mosque on the citadel (Meinecke 1976–7). The desire for legibility is also clear from the position of the inscription. Those on Cairene minarets, for example, usually have the basmala visible from the street below so that a literate person could read the first few words and identify the Koranic text. This was true even in the fifteenth century when the inscriptions on Cairene minarets became somewhat routinised and assumed more of a talismanic or ritualised function. Long bands of beautifully carved inscriptions also adorned the walls and gates of many cites in the Islamic lands, particularly in early and medieval times. The inscriptions are best preserved in areas where stone was the main material of construction, as for example, at Diyarbekır (see Figure 5.21) and Cairo (see Figure 6.29). Texts suggest that cities with brick walls, such as Isfahan or Yazd in Iran or Bukhara in Central Asia, also had inscriptions, although the iron gates are often the only parts to survive (Blair 1992a, nos 41, 49). The construction or restoration is often mentioned only in general terms such as ‘this is what was ordered by’ (mimmå amara bihi), and the inscription alone is insufficient to figure out exactly what work was done. The reason is that the inscription is meant not to record details of construction but

— 73 —

— ISLAMIC INSCRIPTIONS —

6.28 Detail of the tile mosaic inscription with the profession of faith on the north-west minaret erected by al-Nasir Muhammad in his mosque on the

rather to mark political sovereignty. The texts usually include the name of the ruler, the vizier in office at the time, the overseer of accounts, who could be a government official such as a finance officer (Æåmil) or local judge (qå∂È), and the master builder in charge of the work. The walls of Diyarbekır present a good example of the typical inscriptions from ÆAbbasid times. The one erected to the right of the northern or Kharput Gate (see Figure 5.21) says that the work, simply designated as ‘it’, was part of what was ordered by the ÆAbbasid caliph al-Muqtadir billah in 297/910 (RCEA 892). The text continues on the left side of the gate (RCEA 895) that it was carried out under the supervision of Yahya b. Ishaq al-Jarjaraæi, the Æåmil who was in charge of expenditure (al-nafaqa), and Ahmad b. Jamil alMuqtadiri the engineer (muhandis) who was in charge of the work (al-wakÈl Æalå thalik). Another longer inscription over the gate (RCEA 891 corrected in V, p. 191) adds that the vizier AbuælHasan ÆAli b. Muhammad, better known as Ibn al-Furat, supervised the project. These ÆAbbasid inscriptions set the standard, but over time the texts became became longer, as the titles of the people involved grew. We can thus use the inscriptions on the walls of Diyarbekır to trace the evolving concept of the city. Van Berchem’s 1910 monograph showed how the inscriptions could be used to document the changing political circumstances. Sauvaget’s additions in Gabriel’s travelogue through eastern Anatolia (1940) recorded more inscriptions, making it possible to trace forty-four consecutive reconstructions to the same set of city walls and showing the evolution of the city over seven centuries. With the increasing divisions of the Islamic polity in the medieval period, inscriptions on city walls came to reflect sectarian differences. The Fatimid inscriptions set up on the walls of Cairo by the Armenian general Badr al-Jamali between 480/1087 and 484/1092 are a good example (MCIA Egypte 1, nos 33, 36, 37, 520; Wiet 1941–2, 1961). The inscriptions begin with the invocation to God, the ShiÆite profession of faith, and the Throne Verse (Koran 2:255) and then report that

citadel in Cairo, 735/1335.

— 74 —

— INSCRIPTIONS

ON

VA R I O U S B U I L D I N G T Y P E S —

the walls were restored by a servant of the caliph al-Mustansir, the commander Badr al-Jamali. The texts set forth the Fatimid adherence to ShiÆism and allude to contemporary events, such as the famine and crises that brought Badr to power. Wittily composed with frequent plays on words and allusions, the texts must have been drawn up in the chancellery. The inscriptions are also carefully laid out, so that key phrases lie

in prominent positions. The name of the caliph, for example, falls over the top of the arch at the centre of the Bab al-Nasr (see Figure 6.29), so that anyone entering the gate would ride beneath the caliph’s name. Exquisite carving enhanced the effect, and the inscription on the Bab al-Futuh is carved in relief on a long 59metre band of marble blocks held in place by gilded bronze nails.

6.29 Bab al-Nasr in Cairo, begun Muharram 489/April–May 1087 (photo S. R. Peterson).

— 75 —

— CHAPTER 7 —

Stylistic Development

M

onumental inscriptions, designed for rhetorical and decorative purposes and executed in a clear sober script, had long been used in the classical world, and the first Islamic dynasty, the Umayyads of Syria, continued the epigraphic traditions of the region by ordering inscriptions in stone and mosaic, using many of the same formulas and styles but substituting Arabic for Latin. This angular monumental script used in early Islamic times is often known as Kufic, as it was once thought to have originated in the ÆIraqi city of Kufa. Although the derivation from Kufa is unwarranted, the name Kufic is useful in designating this script which is characterised by simple geometric shapes, harmonious proportions and wide spacing. The juxtaposition of horizontal and vertical strokes lends the inscription a strong internal rhythm, which is subtly enhanced by the elongation of such horizontal letters as dål and kåf. Good examples of Kufic script from the early Islamic period are the inscriptions in the name of the Umayyad caliph ÆAbd al-Malik on milestones (see Figure 3.16) and in the Dome of the Rock (see Figures 1.4, 3.10, 3.11 and 7.30). Although these early Kufic inscriptions give the impression of being unpointed, they actually use diacritical marks sparingly. Some ten words are pointed in the rock-cut inscription recording the construction of a dam near Taæif by the Umayyad caliph MuÆawiya in 58/677–8 (Miles 1948). The dots were added to clarify the reading of names and dates and to distinguish letters in words with two or even three consecutive teeth, as in banÈhu (‘built it’) and thabbatahu (‘strengthen him’). Diacritical strokes are also used on the

milestone found at Bab al-Wad (RCEA 15) to distinguish the tooth of the letter thåæ in the word thåmåniya (‘eight’) in the phrase recording that the milestone was set up to mark the distance of eight miles from Jerusalem. In these cases, then, diacritical marks were added to enhance readability and avoid confusion in significant words. The inscriptions on the Dome of the Rock suggest that already in early Islamic times diacritical marks were used for other purposes as well. Kessler (1970) recorded ninety-two separate examples of diacritical marks in the mosaic inscriptions there. None of them occurred in the historical part of the text, and only three diacritical marks were used in a single word in the rest of the band running around the outer face of the arcade (see Figures 3.10–11). The eighty-nine others occurred in the band on the inner face of the arcade (see Figure 1.4), which contains a single continuous text with Koranic excerpts and paraphrases and pious phrases extolling God, Muhammad and Islam, denouncing the Trinity, and expounding the view of Jesus as God’s servant and messenger. Kessler suggested that diacritical marks were used in the inscription on the inner face to enhance the polemic message of the text, but much of the text is similar in content to that on the outer face. Another, and probably more important, reason for adding diacritical marks to the inscription around the inner face of the arcade in the Dome of the Rock was aesthetic: to highlight the writing. The gold strokes glitter against the blue ground and call attention to the band in the same way that glittering pieces of mother-of-pearl enhance the mosaics below, which depict diadems, pectorals

— 76 —

— STYLISTIC DEVELOPMENT —

7.30 Detail of the mosaic inscription in simple Kufic on the inner face of the octagonal arcade in the Dome of the Rock, Jerusalem.

and other jewellery. When the spotlights on the dome are extinguished, the inner inscription sparkles noticeably and would have done so even more before the sixteenth century when Sulayman the Magnificent introduced the present window grilles with their double walls. Diacritical marks were thus added at the Dome of the Rock to increase the aesthetic impact of the text. This example shows that from the earliest Islamic times, the artists who drew up inscriptions were well aware of the artistic possibilities of writing. The inscriptions on the Dome of the Rock are also important for palaeographic reasons. Kessler noted that both the diacritical marks and the letter shapes used in the inscriptions on the Dome of the Rock follow those used in the earliest manuscripts of the Koran. She cited such examples as the pointed beginnings of the letters alif and Æayn, the pointed endings of dål and kåf, and the long backwardturning ending of the letter yåæ (see Figure 7.30).

The letter qåf was also marked with one stroke below the letter, an old form that has been retained in the Maghrib. Kessler concluded that scribes practised in transcribing manuscripts of the Koran must have drawn up the inscriptions on the Dome of the Rock. Since no manuscript of the Koran has survived intact from this early period, monumental epigraphy here serves as a source for manuscript palaeography. It is also clear that monumental epigraphy should be considered along with manuscripts, for the same people may have been involved in both practices. In early Kufic script, most of the letter shape is concentrated in the lower part of the inscription band, leaving the upper zone relatively empty. This imbalance bothered some scribes and artists, who gradually developed new devices to fill the upper zone. See, for example, the simple geometric filler motifs and foliated tails added to the Kufic letters in the inscription carved on the Kharput

— 77 —

— ISLAMIC INSCRIPTIONS —

7.31 End of the foundation inscription in floriated Kufic on the Aqmar Mosque in Cairo dated 519/1125–6.

— 78 —

— STYLISTIC DEVELOPMENT — Gate at Diyarbekır (see Figure 5.21). It is difficult to follow this development over the next few centuries in inscriptions on buildings, since few have survived in Mesopotamia, the capital province of the ÆAbbasid caliphate. Luckily we have another source of monumental inscriptions at hand: tombstones from Egypt. Beginning with the year 174/790, a series of more than 4,000 limestone and marble stelae survive from cemeteries in Aswan and Cairo. The tombstones provide fixed dates for the introduction of such decorative devices as the bevelled stem, arc, barb, palmette and rising tail and show that as these ornaments filled the upper zone, the script became more cramped. In the ninth and tenth centuries, artists in Egypt, the Hijaz and the western Islamic lands continued to embellish the upper ends of the letters. Foliated script evolved into a floriated one in which flowers, tendrils and scrolls seem to grow from the terminal or medial forms of the letters. Floriated Kufic was in full bloom by the mid-tenth century, and magnificent bands of floriated Kufic sculpted in

stone became a hallmark of the Fatimids, the wealthy and sophisticated rulers of Egypt from 969 to 1171. The band running across the façade of the Aqmar Mosque in Cairo (see Figure 7.31), constructed in 519/1125–6, shows how skilfully Fatimid artists sculpted these texts (RCEA 3012). Sometimes the trend towards decorating the ends of the letters with palmettes, floriation and other devices obscured the readability of the text, which devolved into a decorative pattern. This type of script is called pseudo-Kufic or Kufesque, a term coined by Miles (1964a) from Kufic, the angular script typical of early Islamic times, and arabesque, the unending geometricised scroll typical of most forms of Islamic art from the tenth century. While individual letters in pseudo-Kufic may be recognisable, words are not, and the texts are unreadable. This style of script was popular from the tenth century to the twelfth, and Miles collected many examples from the Byzantine period in Greece. One of the most famous is the Theotokos church (see Figure 7.32) built c. 950 at the monastery of Hosias Loukas in Phocis (A. Grabar 1971). The

7.32 Kufesque inscription on the façade of the Theotokos church at the monastery of Hosias Loukas in Phocis, Greece, built c. 950.

— 79 —

— ISLAMIC INSCRIPTIONS — script was also used on the portable arts, such as a ruby glass bowl with mythological scenes in the treasury of San Marco at Venice (Cutler 1974). Ettinghausen (1976) noticed that one group of inscriptions in pseudo-Kufic is composed of two framing uprights set to the right and left of an arched unit. He suggested that this tall–short– tall pattern had evolved from the word allåh, which was often embellished with a bump between the two låms on tombstones from ninth-century Egypt. The meaning of the design had been lost, and this simple and pleasing pattern was repeated on a number of buildings and objects where the use of the name of God would have been particularly inappropriate. These include not only Christian monuments, such as the marble cornice of the church at Hosias Loukas, but also other types of Islamic objects, notably ceramics, textiles and glassware. In the eastern Islamic lands, artists developed another decorative device, interlacing, to meet the same need of filling the upper zone of the inscription. Whereas in the western Islamic lands elaboration of the stems of the letters had led from bevelling to foliation and then to floriation, in the east the tendency toward elongation and distortion of horizontal letters led to internal modifications and superimposed ornament. The use of interlacing developed over the course of the ninth and tenth centuries and was already quite sophisticated by the early eleventh. The most famous example is found on the tomb tower built by the Bawandid ruler Abu JaÆfar Muhammad at Radkan in the mountains of northern Iran and completed in 411/1021. Flury’s chart of the letter shapes shows a complex use of knots and plaiting in and around the letter shapes. Such plaiting was particularly suitable for inscriptions in brick. By the eleventh century, this plethora of decorative devices threatened to obscure the readability of Kufic inscriptions. Symbolic affirmation may have superseded communication, and inscriptions were probably recognised visually rather than read literally. The growing complexity and elaboration of the script took place at the same time as the proliferation of small semi-independent dynasties and the rapid expansion of rulers’ titulature so

that even petty princes had long strings of flowery epithets attached to their names. Designers of monumental inscriptions faced a dilemma: how to reconcile the conflicting demands of the basic unreadability of many highly decorated but unpointed Kufic scripts and the increasing number of titles in any historical inscription. One solution was to move the decorative devices to the upper zone, so that only the stems of the letters, not their bodies, were knotted. This type of Kufic with interlaced stems developed in the eastern Islamic lands in the eleventh century. One of the earliest examples to survive occurs on the minaret erected at Tirmidh in 423/1031–2, probably to mark the appointment of a new Muhtajid governor for the Ghaznavids (Blair 1992a, no. 38). The minaret shaft is encircled by four bands of Kufic with interlaced stems. In the one at the top (see Figure 7.33) with the Throne Verse (Koran 2:255), the designers have not only plaited the bodies of the individual letters but also added a complementary group of knots in the stems of the upper zone to balance the weight of the letters’ bodies in the lower zone. Designers soon recognised that they did not have to depend upon the stems of the letters for the interlacing, but could add knots whenever they wanted to create a regular, if arbitrary, knotted pattern in the upper zone. This is the case in the foundation inscription in the dome chamber in the mosque at Barujird, built by the Saljuq vizier ÆIzz al-Mulk Barujirdi c. 533/1139 (Blair 1994). The stems of the letters are knotted at regular intervals, and whenever a suitable vertical was lacking, an extra knot was inserted. Painting allowed the designer more freedom in creating complex knotted patterns, and the inscription painted around the interior of the mausoleum of Safwat al-Mulk in Damascus, built in 504/ 1110–11, has three virtually separate zones: a lower zone with the bodies of the letters, an intermediate zone in which the stems cross to form stars, and an upper zone in which the stems of the letters ends in pairs of half-palmettes (illustrated in Sourdel and Sourdel 1968: 569). This type of bordered Kufic, in which an upper zone of decoration balances the lower zone

— 80 —

— STYLISTIC DEVELOPMENT —

7.33 Drawing of the upper band of Kufic with interlaced stems from the minaret erected at Tirmidh in 423/1031–2.

7.34 Bordered Kufic inscription from the Nizamiyya madrasa at Khargird in north-eastern Iran, constructed 465–70/1072–7.

containing the letter bodies, was particularly popular in the eleventh and twelfth centuries when the Ghaznavids and Saljuqs were the major powers in the eastern Islamic lands. The uppermost band with the signature on the minaret erected at Dawlatabad by the Saljuq vizier Sadr al-Din in 502/1108–9 (see Figure 6.27), for example, is done in Kufic with an interlaced or plaited border.

Designers in the eastern Islamic lands also recognised that in addition to knots other decorative motifs could be added to the upper part of the inscription band to balance the bodies below. One superb example of such a type of bordered Kufic with floral elements (see Figure 7.34) is the foundation inscription that once graced the madrasa founded by Nizam al-Mulk

— 81 —

— ISLAMIC INSCRIPTIONS — at Khargird in eastern Iran and now in the Iran Bastan Museum, Tehran (Blair 1992a, no. 57). Herzfeld called the inscription the masterpiece of Kufic epigraphy from Iran. The band was made in three sections. The lower third contains the bodies of the letters outlined with a beaded edge. This is balanced by an elaborately carved floral ornament in the upper third. The two are connected by a medial zone in which the stems of the letters are set against a background of floral arabesques. Bordered Kufic thus solves the problem engendered by the increasingly elaborate plaited Kufic: it removes the ornament to the upper zone, leaving the letter bodies in the lower zone clear and readable and providing a balanced composition. In the superb inscription from Khargird, the designer has ingeniously varied the height of the letters to enhance the aesthetic effect. The letters themselves are 11 cm thick, but project only 7.5 cm from the surface as they are embedded 3.5 cm into the wall. This allows the arabesque to swirl along the background of the middle zone, disappearing behind the stems of the letters and adding an organic element to the stately stems which march rigidly across the band. The inscription is at the same time legible and lively. The eleventh and twelfth centuries were a time of extraordinary epigraphic creativity in the eastern Islamic lands, as designers vied to create new and ingenious methods of writing out their messages in fancy varieties of Kufic script. One script that survived the longest, probably because it is so effective in brick construction, is popularly known as square Kufic. Just as Kufic is a convenient if erroneous term for the common angular script used in early Islamic times, so square Kufic is a common and convenient term for this script. Herzfeld felt that the script was derived from naskh, so he proposed that it be called square naskh, although he never articulated his exact reasoning. Following him, Myron Bement Smith (1936) adopted the form ‘rectangular’ naskh. Persian authors use the term maÆqilÈ (‘square’) or bannåæÈ (‘builder’s [technique]’). The earliest example of square Kufic to survive (see Figure 7.35) occurs on the minaret erected at

Ghazna by the Ghaznavid ruler MasÆud III (r. 1098 –1115). The cylindrical upper storey of the minaret has disappeared, leaving the lower section with a stellate octagonal plan (Pinder-Wilson 1985). An inscription in bordered Kufic in the uppermost panels (RCEA 2961) gives the patron’s name, titles and genealogy: (1) bism allåh al-rahman al-rahÈm (2) al-sul†ån al-aÆΩam (3) malik al-islåm Æalåæ aldawla (4) abË saÆd masÆËd (5) ibn ΩahÈr al-dawla (6) abÈ muΩaffar ibrahÈm (7) naßÈr khålifat [sic] allåh amÈr and (8) al-muæminÈn khallada allåh mulkahu (‘basmala, the greatest sultan, king of Islam, ÆAlaæ al-Dawla Abu SaÆd MasÆud b. Zahir al-Dawla Abi Muzaffar Ibrahim, the assistant of God’s caliph, the Commander of the Faithful, may God extend his power’). The second tier of panels above the ground contains another inscription written in square Kufic set on a diagonal. The texts repeat the sultan’s name and titles. The ones illustrated in this detail (see Figure 7.36), for example, contain the words Æalaæ al-dawla (on the right) and abÈ saÆd masÆËd (on the left). The inscription on the minaret of MasÆud III was probably one of the first inscriptions in square Kufic to have been designed. The text is difficult to read, as the letters are formed by small pieces of terracotta sandwiched between larger bricks laid vertically in stepped bond. It must have been labourintensive to design and set out the inscription. The text is also unusual as one of the only examples known of an inscription in square Kufic giving historical information. Most contain sacred names or short pious phrases of a generalised nature. MasÆud III’s minaret was designed at a time when builders were experimenting with various methods of spelling out inscriptions in square Kufic on baked brick constructions. Another technique is to set the bricks in relief so that the letters of the inscription project from the brick bonding of the structure. This technique was used on the minaret built beside the congregational mosque at Sava in 504/1110–11 (Miles 1965). Like many of its contemporaries in Iran, the minaret is decorated with bands in which the builders delighted in setting up different bonding patterns that complement and contrast with each other and catch the light in different ways.

— 82 —

— STYLISTIC DEVELOPMENT —

7.35 Minaret of MasÆud III at Ghazna c. 1100 (photo Catherine B. Asher).

— 83 —

— ISLAMIC INSCRIPTIONS —

7.36 Panels in square Kufic on the minaret of MasÆud III at Ghazna.

Several of the bands on the minaret at Sava are epigraphic. Two bands in bordered Kufic name the rulers of the time, the ÆAbbasid caliph al-Mustazhir biællah and the Saljuq sultan Muhammad b. Malikshah. A large band at the bottom contains bricks set in relief to outline letters in square Kufic spelling out the phrase lå nabÈ baÆd mu˙ammad (‘there is no prophet after Muhammad’), which is repeated three times around the minaret. This is the first extant example of square Kufic written out as a single line of text, and this experiment was rarely repeated. A far more popular method of writing on brick buildings in square Kufic was to exploit the spaces between the bricks. The earliest dated example to survive occurs on the minaret built in 515/1121–2 at Gar, a few kilometres east of Isfahan. The minaret shaft (see Figure 7.37) is decorated with a diagonal diaper of square Kufic letters that spell out the phrase al-mulk lillåh (‘power [belongs] to God’). The pattern is created by the spaces between the rising brick joints.

It was a short step for designers to fill the spaces between bricks with glazed elements so that the words were spelled out not by the dark shadows created by the recessed interstices between the bricks but by glittering surfaces that were flush with the brick bonds and contrasted with the matte surface around them. The shaft of the thirteenth-century minaret of the congregational mosque at Nushabad (R. Hillenbrand 1976) for example, is decorated with the phrase ‘knowledge is with God’ (al-Æilm Æind allåh) spelled out in square Kufic letters made of bricks glazed light blue. This became the most popular technique for covering large wall areas in the later period in Iran and adjacent areas, as on the shrine ordered by Timur for Ahmad Yasavi in Turkestan (Blair and Bloom 1994, fig. 46) or on the façade of the Madrasa al-Ghiyathiyya at Khargird, completed in 846/1442–3 (O’Kane 1976). Square Kufic clearly developed out of brick construction and was most popular in areas like Iran where baked brick was the main medium of

— 84 —

— STYLISTIC DEVELOPMENT — construction. Ghouchani (1985) gives many examples from the region of Isfahan. The style could be adapted, however, to other materials. Inscriptions in square Kufic could be executed in the stucco revetment that increasingly came to dominate interior decoration in Iran in later times. Carved stucco, of course, allowed for longer and more complicated texts. It also allowed for more latitude in design, so that the texts could be fitted to triangular spaces, muqarnas elements and the like. Some of the most elaborate examples of square Kufic inscriptions in stucco are found in the interior of the shrine built outside Isfahan in the opening decades of the fourteenth century for the Sufi shaykh known as Pir-i Bakran (Hunarfar 1977: 252–66). The side walls (see Figure 7.38) are decorated with the names of the four Orthodox Caliphs written out in square Kufic letters created by endplugs set between the baked bricks. Some of the endplugs are stamped with geometric designs, but a few are epigraphic, with the names ÆAli and Muhammad written in square Kufic. The north piers supporting the iwan are further decorated with large stucco plaques of square Kufic. The one on the east side has the names of the Fourteen Innocent Ones (Muhammad, Fatima and the Twelve Imams); the one on the west has some of the ninety-nine names of God given in Koran 59:23. These inscriptions are juxtaposed to other horizontal bands and panels with other varieties of scripts that make this building a tour-de-force of epigraphic decoration. Stucco inscriptions in square Kufic were particularly popular in Iran during the period of Mongol rule in the thirteenth and fourteenth centuries. This is not surprising, for the script resembles the square box-like script known as Phagspa introduced in 1269 by the great Yuan ruler Qubilay for writing Mongolian. This new script was quite cumbersome and was most important for insignia and seals. Yuan seals in Phagspa script were quickly distributed to the Ilkhanids in Iran, and several farmåns, or decrees, preserved in Tehran and elsewhere, are stamped in red at the end of the document and at the junction of the individual sections with the

7.37 Minaret built at Gar, Iran in 515/1121–2 with

— 85 —

a shaft decorated in square Kufic.

— ISLAMIC INSCRIPTIONS —

7.38 Interior of the shrine constructed for Pir-i Bakran at Linjan, near Isfahan in the early fourteenth century, showing the north wall with several kinds of square Kufic inscriptions.

— 86 —

— STYLISTIC DEVELOPMENT —

7.39 Interior of the Karatay madrasa at Konya built in 649/1251–2, with sacred names in square Kufic in the Turkish triangles and a Koranic inscription (2:255) in Kufic with interlaced stems around the base of the dome.

— 87 —

— ISLAMIC INSCRIPTIONS — impression of a seal sent by Qubilay from Khanbaliq. This style of Chinese seal was also imitated in Iran, when Mongol rulers of the fourteenth and fifteenth centuries ordered similar seals containing Koranic phrases and religious invocations in Arabic written in square Kufic. Indeed, the resemblance between the two scripts is so close that square Kufic is sometimes called seal script and its origins are sometimes (wrongly) thought to derive from Chinese seals, although it was clearly used in architecture before the introduction of Phagspa for seals. The popularity of square Kufic in Iran also meant that it was adapted for use in other materials elsewhere. The minaret in the Great Mosque at Mardin built in 572/1176–7 (Creswell 1932–40 [1979], vol. 2, fig. 123), for example, has a large text in square Kufic carved on the square stone base. Square Kufic was also used to decorate the shaft of the minaret built at Sinjar in Muharram 598/ October 1201. These examples in stone and brick in the Jazira were imitations of the techniques that had evolved in Iran. Designers often played with square Kufic by repeating a sacred name four times in a circle to create a wheel-like pattern. The most common names were Muhammad and ÆAli, and the inital mÈm of Muhammad or Æayn of ÆAli was often the pivot for a fourfold repetition of the name in a swastika shape. Burgoyne and Abul-Hajj (1979) give examples from the Levant. A particularly wide, and somewhat eclectic, repertory of sacred names is written in square Kufic on the Turkish triangles supporting the dome in the Karatay Madrasa built at Konya in 649/1251–2 (see Figure 7.39). Eight names are included there: Daæud, ÆIsa, Muhammad, Abu Bakr, Musa, ÆUmar, ÆUthman and ÆAli (Meinecke 1976, no. 76). The style also passed to Egypt, where it was also adapted to other materials. The earliest example of square Kufic in Cairo occurs in the interior of the mausoleum of Qalaæun (see Figure 7.40), where a panel of differently coloured marbles spells out the name Muhammad four times at the top and four times below. The technique of inlaid marble is Syrian in origin (Meinecke 1971), and colour is used to enhance the message. The

letters of the name Muhammad are done in white. Red is used to fill the hole of the letter mÈm, which occurs twice in each repetition of the name. The rest of the ground is composed of small pieces of differently coloured marble set in a diamond pattern. Ultimately the desire for readability overcame the trend towards increasingly complex decoration, and floriated and interlaced Kufic scripts were abandoned in favour of cursive ones. Cursive had been used alongside Kufic for chancellery documents since Umayyad times. In the ÆAbbasid period, famous calligraphers such as Ibn Muqla (d. 940) and Ibn al-Bawwab (d. 1022) refined and elevated cursive script to a rank suitable for copying manuscripts of the Koran. Cursive script first appears in monumental epigraphy in the eleventh century. It was used initially for religious inscriptions, such as the Koranic verses across the façade of the south dome added to the congregational mosque in Isfahan in the winter of 1086–7 (Blair 1992a, no. 61). The band is executed in carved stucco, which lends itself readily to curved lines. Since cursive script was more readable, it was quickly adopted for historical texts. In the eastern Islamic lands this was a gradual process. The first examples to survive there are stucco panels in the tomb of al-Hakim al-Tirmidhi in the name of the Qarakhanid Ahmad (d. 1089) and stone fragments from Ghazna in the name of the Ghaznavid sultan Ibrahim (d. 1099). These inscriptions in cursive were consciously juxtaposed on the same building to other inscriptions in complex Kufic scripts. The minarets at Dawlatabad (see Figure 6.27) and Ghazna (see Figure 7.35) show good examples of this wilful combination of different styles on the same building. In other areas, the change from angular to cursive was more abrupt. In Syria, for example, the Zangid ruler Nur al-Din (r. 1146–74) ordered the adoption of cursive scripts in monumental inscriptions. The transition was effected in less than a decade, and cursive inscriptions bearing his name grace many of the buildings that he ordered there. The foundation inscription for his hospital in Damascus, for example (see Figure

— 88 —

— STYLISTIC DEVELOPMENT —

7.40 Marble mosaic panel in square Kufic from the mausoleum of Qalaæun, Cairo.

— 89 —

— ISLAMIC INSCRIPTIONS —

7.41 Marble slab with the foundation inscription in naskh for Nun al-Din’s hospital in Damascus, 549/1154–5.

— 90 —

— STYLISTIC DEVELOPMENT — 7.41), was inscribed in naskh on a slab of white marble, with the cursive letters encrusted with black stucco to enhance legibility. The outside band says that the work was completed in 549/ 1154–5, while the four lines in the center give Nur al-Din’s name and titles (RCEA 3164, corrected in RCEA IX, pp. 271–2). Kufic gradually lost its pre-eminence in monumental epigraphy and from the thirteenth century became stereotyped and repetitive. This was particularly true in the western Islamic lands. In the Alhambra palace at Granada, for example (see Figure 7.42), Kufic with interlaced stems was used for cartouches in which the words are virtually illegible. These inscriptions in stylised Kufic are often played off against other cartouches with the slogan of the Nasrid ruler Muhammad I, wa lå ghålib ilåællåh (‘there is no victor save God’), written in an equally stylised cursive. In the central Islamic lands, designers reserved their greatest artistic efforts for elaborating

cursive scripts. Most inscriptions were carved in stone, but stucco was also used. One of the most beautiful (see Figure 7.43) is this stucco band on the western wall of the courtyard in the tomb complex for the amirs Salar and Sanjar in Cairo, begun in 703/1303. The two cartouches contain Koran 3:185, a typical funerary inscription that every soul shall taste of death, that the good will achieve Paradise and that the life of this world is but goods and chattels of deception. The text is superposed on a background of arabesque scrolls typical of the Iranian world. To balance the inscription and fill the upper zone, designers dropped the rigid base line characteristic of Kufic and divided words into short groups of letters suspended on the diagonal. A good example is the construction text dated 610/1213–14 on the citadel of Jerusalem in the name of the Ayyubid al-Malik al-MuÆazzam ÆIsa (MCIA Jerusalem, no. 43 and plate XLI). The phrases became more crowded as texts became

7.42 Stucco cartouches decorating the Hall of the Ambassadors in the Alhambra palace at Granada, mid-fourteenth century.

— 91 —

— ISLAMIC INSCRIPTIONS —

7.43 Stucco band with Koran 3:185 on the courtyard wall of the tomb complex for Salar and Sanjar in Cairo, begun 703/1303.

longer and incorporated administrative decrees written in several lines within rectangular frames, as in the decree in the name of the Mamluk sultan Jaqmaq added to the mosque of Princess Asalbay in the Fayyum in 845/1441–2 (MCIA Egypte 1, no. 373, plate IX, no. 6). Stucco carvers and tile workers in Iran gradually elongated the stems of the letters and filled the upper zone with various decorative motifs. These could be flowers, as on the Saljuq minaret at Dawlatabad (see Figure 6.27). They could be arabesque scrolls, as on the bands ringing the walls of the domed sanctuary added to the mosque at Ardistan in 553/1158 (Ettinghausen and Grabar 1987, fig. 298) and the contemporary tomb at Sangbast in eastern Iran (ibid., fig. 300). A second inscription could even be inserted among the stems of the main text. One of the earliest examples to survive is found on the tomb for the Ilkhanid sultan Uljaytu at Sultaniyya, finished in 715/1315, and in succeeding centuries this two-tiered inscription band became increasingly popular. Single words or even phrases could be piled above the base inscription, and the

tail of a final yåæ could be drawn back to the right across the text to serve as a sort of dividing line. The interior of the Mosque of Shaykh Lutfallah (see Figure 4.19) contains spectacular examples in tile mosaic from the seventeenth century. The introduction of tile mosaic also made colour an important feature in inscriptions. Designers juxtaposed white letters against a dark blue ground, adding vocalisation in accent colours and setting off the name of the ruler in gold, as in the restoration text in the name of the Aq Qoyunlu ruler Uzun Hasan added to the qibla iwan of the congregational mosque at Isfahan in 880/1475–6. Ottoman tile workers adapted and refined many of these techniques. The lunette panels done by the ‘Masters of Tabriz’ for the windows of the court in the Uç „erefeli Mosque at Edirne, for example, are underglaze-painted with two contrasting texts (Dijkema 1977, nos 163–4). The larger band in cursive invokes the name of the patron, the Ottoman sultan Murad II. The pious invocation reads: ‘O Lord, accept [this pious work] from the sultan, son of the sultan, Murad

— 92 —

— STYLISTIC DEVELOPMENT —

7.44 Underglaze-painted tile lunette from the Uç „erefeli Mosque at Edirne, completed in 851/1447–8.

b. Muhammad Khan’. A smaller text in a stylised Kufic is inserted in the stems of the first. It reads al-janna dår al-askhiåæ (‘Paradise is the abode of the munificent’). It alludes to the pious nature of the endowment, for only things that are pleasing to God may be endowed. Colour was used to enhance the message, and the earlier palette of blue and white was expanded to include purple and two shades of blue. In one of the surviving lunettes, the tile-makers emphasised the distinction between the two styles of script by using different colours for the two texts (see Figure 7.44).

In the later period, particularly in Iran, Turkey and India, as poetic inscriptions became more frequent, cartouches often replaced rectangular frames. Like the Selimiye (see Figure 2.8), the Uç „erefeli Mosque has such poetic inscriptions over the doorways leading from the courtyard to the mosque. The texts are painted on white marble in rectangular cartouches with rounded corners. They contain chronograms with the years that the mosque was ordered (841/1437–8) and completed (851/1447–8), but the inscriptions have been restored and the texts are now corrupt (Dijkema 1977, nos 11–13).

— 93 —

— CHAPTER 8 —

Introduction

L

ike buildings in the Islamic lands, many of the portable objects made there were inscribed. These inscriptions share many features with monumental inscriptions, and they can be discussed under some of the same rubrics used to discuss monumental inscriptions, such as language and types of text. In addition, the demands and restrictions of the individual medium meant that certain kinds of inscriptions were common on a particular kind of object, and the general introduction will therefore be followed by a lengthy description of the inscriptions on a particular medium or type of object. Chapters on the inscriptions on four major media (metalwares, woodwork, ceramics and textiles) are followed by a chapter on the inscriptions on various other minor media (glass, ivory, rock crystal and jade) and one on various types of objects (tombstones; seals, talismans and amulets; and arms and armour). The chapters are arranged in order of importance, beginning with the medium (metalwares) whose inscriptions are best known. In each chapter, a general introduction with basic bibliography is followed by a division into broad chronological groups. The major sources for learning about inscriptions on portable objects made in the Islamic lands are exhibition catalogues. One of the most comprehensive is the catalogue of the exhibition held at the Hayward Gallery in 1976, Arts of Islam. The exhibition comprised more than 600 objects, including many of the masterpieces of Islamic art. The objects are arranged by medium and then by chronology, and the individual entries contain transcriptions of the inscriptions on most pieces.

Other useful catalogues cover the inscriptions on objects from a particular country. These include Arts de l’Islam des origines à 1700, with pieces from French national collections exhibited at the Tuileries in 1971; Masterpieces of Islamic Art in the Hermitage Museum, with pieces from the Hermitage exhibited in Kuwait in 1990; and Curatola (1993), Eredità dell’Islam, with pieces from Italian collections exhibited in Venice in 1993–4. Yet other catalogues cover objects from a particular dynasty or region. One of the most important of this group is Atıl’s catalogue of a travelling exhibition on the Mamluks, Renaissance of Islam: Art of the Mamluks (1981), for inscriptions were a major decorative feature of Mamluk art and the catalogue records the inscriptions on most of the 128 pieces exhibited. Similarly, Jerrilynn Dodds’ catalogue Al-Andalus: The Art of Islamic Spain (1992) records the inscriptions on the masterpieces from that area. Catalogues that cover a particular medium will be discussed in the relevant sections below. As with monumental inscriptions, Arabic was the major language used for inscriptions on objects made in the Islamic lands. It was the preferred language for foundation texts that give the name of the patron and the date, but, as with some buildings in Iran, some objects made there in Islamic times are inscribed in other languages. Pahlavi was used in a few rare cases, particularly in the early period when local rulers had not converted to Islam. Metalwares produced in the Caspian region during early Islamic times

— 97 —

— ISLAMIC INSCRIPTIONS — consciously aped the forms and techniques of Sasanian times, and some pieces are inscribed in Pahlavi. One example is a set of three silver bowls found in Mazandaran and now in the Iran Bastan Museum in Tehran; they are engraved on the exterior in Pahlavi with the name of a local ruler in the late eighth century, Windad Ohrmazd of the Karens, and the weight in drachmas (Ghirshman 1957). As with buildings, Persian was introduced in medieval times for use on some objects. The earliest known example of a Persian inscription on an object occurs on a hemispherical bowl with a harpy found in 1909 at Perm and now in the Hermitage (Masterpieces of Islamic Art 1990, no. 11). Persian verses inscribed in Kufic around the inner rim refer to the joyful and forbidden pleasures (of wine-drinking). The bowl is undated, and, on the style of the harpy, Marschak (1986) attributed it to the ninth or tenth century; the style of the text (Kufic with rising tails) lends weight to the latter. Persian verses become standard decoration on many other types of art made in Iran from the twelfth century onwards. Persian verses are a hallmark of luxury ceramics, particularly lustre and enamelled wares, of which dated examples survive from 575/1179 onwards. Many texts have been read and recorded by Bahrami (1937, 1949) and Ghouchani (1985, 1986, 1992). Some verses were taken from well-known poets, others were composed for the occasion. The poems generally deal with the agonies of love, and, although not necessarily of the highest literary quality, they are important evidence for popular taste in medieval times. Persian verses were also used on textiles. The intricacies of weaving made it difficult, if not impossible, to incorporate a long text while weaving on a loom, but textile weavers overcame the difficulty by using another technique, resistdyeing. A silk dated to the late twelfth century in the Boston Museum of Fine Arts (no. 15.815), for example, has a Persian quatrain (Minovi 1937). Composed for the occasion, the verse is written in the first person as though the textile were speaking. It extols the beauty of the silk,

comparing it to the sun in a garden. The final hemistich gives the name of the workshop, that of Amirak the dyer. Persian foundation inscriptions appear on a few dated metalwares made in the late twelfth and early thirteenth centuries in Khurasan, the north-eastern province of Iran. The most famous example is the Bobrinsky bucket (see Figure 9.50), which has a Persian inscription around the rim giving the details of its commission and manufacture in Muharram 559/December 1163 (Masterpieces of Islamic Art 1990, no. 30). Similarly, an elaborate aquamanile in the shape of a zebu with a suckling calf and a lion attacking her back in the Hermitage (ibid., no. 39) has a Persian foundation text around its neck and head with the date Muharram 603/August–September 1206. In both cases, the texts contain many Arabic words, phrases and even constructions. These metalwares from Khurasan are exceptional, and in most cases we can distinguish, as we could with monumental inscriptions, between foundation inscriptions in Arabic and poetic texts in Persian. A good example is the well-known lustre plate in the Freer Gallery (see Figure 8.45). The long foundation inscription around the rim giving the name and titles of the patron, the date Jumada II 607/November 1210 and the signature of the artist is in Arabic, while the verses on the interior and exterior of the sides are a combination of Persian and Arabic poetry. Persian poetry became increasingly popular on objects from the fifteenth century onwards. The best known are metalwares produced under the Timurids, and Komaroff has devoted several studies to their inscriptions. The Persian verses were drawn from a wide repertory of classical poets, such as Daqiqi, Firdawsi, SaÆdi and Hafiz, and of contemporary Khurasani poets, such as Qasim al-Anwar, Jami and Salihi. The texts, usually well written and accurate, are some of the earliest records of contemporary poets. The texts also refer to the vessels on which they are inscribed. Inlaid jugs, for example, were inscribed with verses about a mashraba (‘jug’), suggesting their use as wine vessels. The inscriptions are thus important sources for contemporary terminology

— 98 —

— INSCRIPTIONS

ON

P O RTA B L E O B J E C T S : I N T R O D U C T I O N —

8.45 Detail of the inscription around the rim of a lustre plate with the scene of a sleeping groom made by Abu Zayd in Jumada II 607/November 1210. Diameter 35.2 cm. Washington

and function, and Melikian-Chirvani (1982b, Annex I) has compiled a list of the words used for these objects. These types of Persian verses continue to decorate lampstands and many other metalwares made under the Safavids. The verses on lampstands often contain the well-known image of a moth seeking a candle, a mystical metaphor for the beloved seeking God. A lampstand in the Hermitage dated 987/1579–80 (Masterpieces of Islamic Art 1990, no. 87), for example, has verses in this vein by Katibi Turshizi, SaÆdi and Hayrati Tuni as well as a quatrain by Ahli Shirazi referring to the object’s function as a chirågh (‘lamp’). Persian love poetry of the sort used on medieval luxury ceramics was also inscribed on some ceramics made in the later period. An underglazepainted dish in the Hermitage (ibid., no. 76; Blair and Bloom 1994, plate 90) has a Persian verse around the rim about the futility of love. It is followed by a sentence in Persian with a foun-

DC,

Freer Gallery of Art 41.11.

dation inscription, saying that this plate (†abaq) was made at Mashhad in 878/1444–5. A similar verse about love is inscribed around a dish in the Victoria and Albert Museum (Ferrier 1989: 267) decorated with a pair of fish painted in black under a green glaze. Many of these verses remain unread or unpublished. A good example is a damaged blueand-white pilgrim flask in the V&A (Rogers 1983, no. 152). The date written above the foot in Arabic, ‘in the months of the year 930’ (1523–4), is well known, and the piece is often cited as the earliest dated Safavid ceramic, but the Persian verses around the rim are unpublished. Collecting and recording these verses would be useful in documenting popular taste. Such a study might also help in assigning workshops, dates and provenance, a problematic area in later Iranian ceramics. The taste for poetry in the vernacular passed to the Ottoman court, and luxury objects made there were inscribed with both Persian and

— 99 —

— ISLAMIC INSCRIPTIONS — Turkish verses. The general trends are clear from several objects made for Sultan Sulayman and included in a recent travelling exhibition. (The inscriptions are treated more fully in the catalogue from the London venue by Rogers and Ward 1988.) A metal mirror with jade handle and gilding and bejewelled decoration, for example, is decorated with Persian verses in nastaÆlÈq script comparing the mirror to the haughty beauty who ravishes the soul of whoever looks upon it. In contrast, an ivory mirror has a carved border with three couplets of Ottoman verse and a dedication to Sulayman by the craftsman Ghani and the date 950/1543–4. Arabic was retained for formal foundation inscriptions, and Sulayman’s sword has a long inscription along the blade in thuluth with the sultan’s name and titles and the date 933/1526–7. The blade’s spine is inscribed in nastaÆlÈq with Persian verses and the signature of the craftsman, Ahmed Tekelü. Different styles of script were often used on the same Ottoman piece to distinguish the various languages. Persian verses are generally written in nastaÆlÈq in cartouches, as on a tile panel with a verse recording the completion of the belvedere (shåhnishÈn) in the sultan’s bath in 982/1574–5. Square Kufic was reserved for Koranic passages and pious phrases, as on a wooden Koran box or a talismanic shirt. The inscriptions sometimes show the hand of a skilled calligrapher, but on many ceramics the inscriptions are awkwardly composed and executed. Rogers and Ward (1988) dubbed one script ‘naive naskhÈ’. Like buildings, many luxury objects made in the Islamic lands contain foundation inscriptions announcing who commissioned them. These inscriptions follow the same form as foundation inscriptions on buildings, with the same five basic elements (invocation, verb, object, name of the patron, date). Foundation inscriptions on royal or courtly objects sometimes include the name of the supervisor as well. Since space was usually at a premium, shortened forms were often used. The invocation is frequently reduced to bism allåh (‘in the name of God’), the date is usually in numerals, and only the most important titles are given. Finding the invocation is

particularly useful in dealing with inscriptions on objects. It tells which way the object was meant to be seen. On a circular text, it also helps us to find the date, which is often at the end of the text. A fair example of a foundation text on an object is the one inscribed around the base of the lid on the Gerona casket (see Figure 1.2). The inscription begins at the right side of front of the casket with the shortened basmala and then invokes blessings on the reigning caliph alHakam, the Umayyad caliph of Spain (r. 961–76). The foundation text proper begins on the back. As with buildings, the verb is often some form of amara (‘ordered’). In this case, the form is mimmå amara bi-Æamalihi (‘of the things that he ordered to be made’), and the object is identified only by the pronoun ‘it’. The text continues with the name of the recipient, his son Hisham, who is identified by his kunya or patronymic Abu Walid and by his caliphal title walÈ ahd amÈr almu’minÈn (‘heir-apparent to the Commander of the Faithful’). The inscription continues that work was accomplished (tamma) under the direction of Jawdar. The phrase Æalå yaday (literally ‘under the hands of’) identifies the supervisor, not the artist, and Jawdar is known from texts as the chief eunuch of the caliphal household. He must have been in charge of the royal workshop and have supervised the distribution of precious materials, such as the silver and gold used to decorate this box. In this case, the foundation inscription does not contain a date, but it does allow us to date the box precisely to eight months of the year 976, since Hisham was only declared heir-apparent in 5 February and succeeded his father on 1 October of the same year. As with buildings, artists’ signatures can accompany foundation texts, coming either in the middle of the foundation text before the date or right after the foundation text. Signatures can also occur in another inconspicuous place. On bowls, for example, they are often found on the plain outside or under the foot. On a box, they can come between the straps or under the clasp, as on the Gerona casket (see Figure 8.46). The inconspicuous location was deliberately chosen

— 100 —

— INSCRIPTIONS

ON

P O RTA B L E O B J E C T S : I N T R O D U C T I O N —

8.46 Detail of the underside of the clasp on the Gerona casket with the signature of the artisans Badr and Tarif.

to show the humility of the artist, particularly in contrast to the lofty patron or recipient, whose name is usually inscribed earlier or in a more prominent place and often written in a different script. Titles show the same juxtaposition, and the artist is typically identified as ‘a low slave’ in contrast to his lordly patron. This identification should often be taken metaphorically: these workers were not necessarily slaves and were often quite well-known individuals who worked in high-status professions. Signatures on objects are typically introduced by the word Æamal (‘work of’). The verb ßanaÆa was used for higher-status or more meticulous work. Artisans who signed their names with Æamal on metal bowls and other objects, for example, used ßanaÆ on astrolabes and other scientific instruments. From medieval times, when objects became more elaborate and crafts more hierarchically organised, different verbs could indicate different parts of the process. This was true particularly for metalwares, woodwork

and ceramics. The signatures found on these pieces are often the best source of information about people and craftsmen in medieval times, for this subject is generally not mentioned in written texts. The Gerona casket again provides a good example of a standard signature from medieval Islamic times. The names of the artisans Badr and Tarif are incised on the underside of the clasp (see Figure 8.46), following the standard verb Æamal (‘work of’). Technique reinforces the lesser status of the artisans, for the signature is incised, a faster and cheaper technique than the relief used in the broad band around the base of the lid naming the reigning caliph al-Hakam, his son Hisham and the supervisor Jawdar. Badr and Tarif are identified as ‘his [the caliph’s] servants’, meaning that they worked in the caliphal workshop. Like the placement and the words used, the styles of script point up the distinction between the foundation text and the signature: the band around the rim is done in a larger Kufic,

— 101 —

— ISLAMIC INSCRIPTIONS — often associated with monumental inscriptions, while the signature of the artisans is incised in a script that resembles handwriting. Some objects are also inscribed with endowment inscriptions. This was the case, for example, with minbars, lampstands and other large furnishings made for the holy shrines in Mecca and Medina and other smaller shrines. These pieces were expensive gifts, and the patron wanted to have his generosity recorded. In the early period, these inscriptions are generally in Arabic, but by the fourteenth century Persian had come to be used for endowment texts. Melikian-Chirvani (1987) has published several Persian texts on candlesticks and lampstands. A tall oil lamp in an American private collection, for example, is inscribed with Arabic and Persian verses alluding to God as the light of the heavens (Koran 24:35) and an endowment text in Persian stating that it was given by the Aq Qoyunlu sultan Uzun Hasan (r. 1453–78) to the tomb of a mystic known as Bayram Baba Vali. As with foundation inscriptions, the names and titles in these endowment inscriptions can be helpful in dating works of art. These large endowment inscriptions are the same size as foundation inscriptions and are well integrated into the overall decoration of the object. Clearly planned in advance, these inscriptions show that the objects were commissioned with a specific purpose in mind. These endowment inscriptions should be distinguished from other smaller inscriptions, often scratched into a plain area on the object, saying that it was endowed to a certain place. Clearly added after an object was made, this type of smaller inscription provides a secure terminus ad quem, or date by which point an object must have been made. Such dates are particularly useful in dealing with objects that cannot be readily dated on stylistic or typological grounds. This is the case, for example, with the large corpus of European swords with Arabic inscriptions that are preserved in the arsenal of Alexandria and later in the Ottoman arsenal of St Irene in Istanbul. The Arabic inscriptions are a cornerstone in establishing a chronological typology. Even if these endowment texts were added at a much later date and cannot be taken to show

the provenance or date of an object, nevertheless these small texts can be used for other purposes. This is the case, for example, with the Chinese porcelains that the Safavid shah ÆAbbas donated to the shrine at Ardabil in 1608. The 1,162 pieces comprise celadons, white wares and blue-andwhite wares and almost all of them have an endowment text engraved into the glaze in a rectangular cartouche. The text says that ÆAbbas, slave of the shah of Sovereign power (ÆAli, the first ShiÆite imam) endowed (this) to the threshold of Shah Safi (banda-yi shåh-i wilåya ÆAbbås waqf bar åståna-yi shåh ßafÈ namËd). These endowment texts are not helpful in establishing the provenance or date of the porcelains, which were all made in China but range in date from the fourteenth century to the seventeenth. The texts are, however, useful historical documents. In the case of the porcelains, for example, they are important evidence for royal Iranian taste in the early seventeenth century, showing what imported wares were available and what Persian royalty of the time deemed fancy. The words used in the endowment inscription also provide clues as to why ÆAbbas made the expensive donation, for the text associates Shah ÆAbbas with the dynasty’s eponym Shah Safi and with ÆAli, who is lauded as the shåh-i wilåya. The foundation or endowment inscriptions on the finest wares show that they were made for or endowed by a specific patron, but such inscriptions are not found on objects made for the open market. Instead, these more ordinary wares were often decorated with short repeating texts of the type known in Arabic as duÆåæ, a prayer formula invoking God’s blessing. These prayers may have evolved from the pious invocations consistently appended to the names of caliphs, high officials and governors on glass weights and measure stamps made in Egypt (Balog 1977). Examples include aßla˙ahu allåh (‘may God set him in the way of righteousness’), abqahu allåh (‘may God grant him long life’) and akramahu allåh (‘may God honour him’). These invocations functioned like honorifics (alqåb), and a particular formula remained with a person until he was awarded a new one with higher prestige.

— 102 —

— INSCRIPTIONS

ON

P O RTA B L E O B J E C T S : I N T R O D U C T I O N —

The act of offering supplicatory or petitionary prayers is one of the main manifestations of Muslim piety. Although these prayers can be in any language, those in Arabic, being hallowed by practice, are usually preferred. The importance of these prayer formulas is already attested in the Koran. The first sura (al-Fatiha) counts as a duÆåæ, and many other verses contain supplicatory prayers, often beginning ‘Our Lord’. These prayers were also included in the traditions ascribed to the Prophet and the Imams and became particularly popular with Sufis. In folk religion, these

formulas were often invoked to secure protection against various evils, especially the evil eye and the maleficent jinn, and the act of writing such charms on amulets has been common for centuries in Iran. Such charms, spells and magical incantations were already used in ancient Iran, and with the Muslim conquest the Koran replaced Zoroastrian texts as the chief source of the charms and spells. Metalwares preserve our earliest evidence for the growing popularity of duÆåæs in medieval times. One of the earliest dated examples is an inlaid

8.47 Detail of the signature on an inlaid brass ewer in the shape of a bird made by Sulayman in 180/796–7. Ht 38 cm. St Petersburg, Hermitage IR-1567.

— 103 —

— ISLAMIC INSCRIPTIONS — bronze aquamanile in the form of a bird in the collection of the Hermitage (see Figure 8.47). A band in simple Kufic around the bird’s neck opens with the invocation to God (‘In the name of God the Merciful, the Compassionate’) and the supplicatory prayer, ‘Blessing from God’ (baraka min allåh). The text then records that the piece was made by Sulayman at a certain city, unfortunately unreadable, in the year 180/796–7. Another aquamanile in the shape of a bird in the Monastery of St Catherine at Sinai (Baer 1983, fig. 167) has a similar-style inscription across its breast invoking God’s blessings to its owner (baraka min allåh li-ßå˙ibihi). The same types of prayer formulas occur on other undated bronzes ascribed to Iran in the first centuries of the Islamic period. A bronze oil lamp attributed to the ninth-century (ibid., fig. 168), for example, is engraved in simple Kufic with the same blessing to its owner. The number of words in these supplicatory prayers was soon multiplied. Several examples on medieval Iranian metalwares are recorded and discussed in Melikian-Chirvani’s catalogue (1982b) of the bronzes in the V&A. A bronze lobed cup attributed to tenth-century Khurasan in the Herat Museum, for example, is inscribed around the rim in floriated Kufic with five or six words: baraka wa yumn wa surËr wa saÆåda wa niÆma wa d[awla?] (‘blessing and good fortune and joy and felicity and beneficence and ?good luck’). A bronze tripod stand in the Kabul Museum, recovered from a cache at Maymana east of Herat in Khurasan and attributed to the tenth or eleventh century, is inscribed with similar blessings in plaited and floriated Kufic in several places, including the ribs of domed base, the cylindrical element that tops the base, the baluster and the tray. The longest and most complete text on the tray contains eight requests (‘blessing, good fortune, joy, felicity, well-being, beneficence, victory and mercy’) for its owner (baraka wa yumn wa surËr wa saÆåda wa salåma wa niÆma wa naßr wa ra˙ma li-ßå[˙ibihi]). These words of petition are often split, with extra letters, particularly låms and alifs, inserted between or among words. Melikian-Chirvani

(1982b) has read mystical esoteric significance into these extra letters, but his interpretation is not generally accepted (see, for example, the devastating comments by Terry Allen 1985), and the explanation of an illiterate or careless craftsman truncating or blindly repeating a text is far more plausible. These supplicatory prayers were often invoked for an anonymous owner (li-ßå˙ibihi), but the name of a specific person could be included after the preposition li (‘for’). A silver cup in the Hermitage (Masterpieces of Islamic Art 1990, no. 21), attributed to eastern Iran in the eleventh century, for example, invokes good fortune, blessing, joy, felicity, wellbeing, generosity, beneficence and happiness on its owner (al-yumn waæl-baraka waæl-surËr waælsaÆåda wa’l-salåma wa’l-karåma waæl-niÆma waæl-ghib†a li-ßå˙ibihi). A comparable silver bottle, found near Tobolsk on the east side of the Urals and also in the Hermitage (van Berchem 1909b), is incised with similar blessings around its sides. A larger inscription around the flat shoulders continues that the blessings are for Abu ÆAli Ahmad b. Muhammad b. Shadhan. He served as vizier for the Saljuq rulers Chaghri Beg and Alp Arslan and preceded the well-known vizier Nizam al-Mulk. Over time, the supplicatory prayers inscribed on metalwares became increasingly elaborate. The script became more highly decorated, and the prayers were often written in plaited and floriated Kufic set against a scroll ground. The list of requests grew steadily until a single inscription could include as many as twenty-five or thirty nouns, often arranged in rhyming pairs. Adjectives were also added. A good example is found on the well-known silver wine service comprising three bowls, two small dishes, a large dish, a ewer, two jars, a bottle and a cup, inscribed with blessings for the amir AbuælÆAbbas Valgin (RCEA 2154–60; Arts of Islam 1976, nos 158–9; Melikian-Chirvani 1982b). Discovered in Iran between the First and Second World Wars, the set can be attributed to Iran in the early eleventh century based on the style of script and the amir’s caliphal title, ‘client of the Commander of the Faithful’. The inscription on

— 104 —

— INSCRIPTIONS

ON

P O RTA B L E O B J E C T S : I N T R O D U C T I O N —

the large dish invokes at least eleven requests, many with corresponding adjectives: God’s entire blessing, abundant felicity, complete well-being, ?unimpaired health, enduring beneficence, growing good luck, waxing influence, happiness, power, good fortune, joy … to the amir (baraka min allåh tåmma wa saÆåda såbigha wa salåma shåm[il]a wa Æåfiya så…a wa niÆma båqiyya wa dawla nåmiyya wa imra tåliÆa wa ghib†a wa Æizz wa yumn wa surËr … lil-amÈr …). These prayer formulas with repeated requests, often with mistakes or extra letters, became the standard text inscribed on metalwares made in the greater Iranian world in the eleventh and

twelfth centuries. Their short, repetitive texts also made them suitable for decorating textiles (Blair 1997b). Despite their profusion, the formulas have never been studied systematically. There was an evolution from shorter, simpler texts to longer, more complicated ones, and the words used also varied over time and space. Thus they offer important evidence for the evolution of popular taste. These texts may may also reflect regional variations, and thus they may be significant indicators of regional styles and a helpful tool for establishing attributions. A survey of the prayer formulas inscribed on objects made in medieval Islamic times is clearly needed.

— 105 —

— CHAPTER 9 —

Metalwares

T

here is no overall survey of the inscriptions on metalwares. The best introduction to the subject is the section on epigraphic decoration in Baer’s survey of early Islamic metalwork (1983: 187–217). It is a brief but balanced discussion of style and content on pieces manufactured up to the mid-fourteenth century. Some eighty of the finest Islamic metalwares, again mainly from the earlier periods, were published in the Arts of Islam exhibition held at the Hayward Gallery in 1976. The entries in the catalogue include a transcription and translation of the inscriptions on each piece, although there is no overall essay in which to fit the individual entries. Most studies are devoted to the inscriptions on metalwork from a particular period or region. By far the area best covered is Iran, due in part to the pioneering studies by Ivanov since the 1960s. His 1967 article on an Iranian brass tray of the fourteenth century in Epigrafika Vostoka is a landmark in using non-historical inscriptions (in this case, the Persian words used to invoke blessings on the owner) to date and attribute an object, and his 1966 article in Soobshchenie Gosudarstvennogo Ermitazha attempts to localise the production of a type of object (in this case, two candlestick bases) based on the identification of the verses in the inscriptions (in this case, by the poet Salihi). These two articles establish a methodology which has been followed by many others in the field, including Melikian-Chirvani and Komaroff. Since Ivanov publishes in Russian, however, his work is sometimes overlooked, and more people are familiar with the prolific contributions of

Melikian-Chirvani, including his short essay about inscriptions on Iranian metalwares (1976) and his catalogue of inscribed bronzes and brasses from the Persian-speaking world in the Victoria and Albert Museum (1982b). The catalogue contains extensive work on inscriptions, not only on the pieces there but on comparative material as well. Since Melikian-Chirvani uses a broad definition for medieval Iran and attributes an unusually large number of pieces there, his work should also be consulted by those interested in metalwares from other parts of the Islamic lands. Komaroff has written extensively on the inscriptions on Timurid wares, and Allan’s catalogue (1982b) of the wares excavated at Nishapur presents inscriptions from simpler, more quotidian wares from an earlier period. Inscriptions on metalwares from other areas, particularly the central Arab lands, were studied by D. S. Rice. Unfortunately, he died before completing his book on the subject, so the reader must delve among his many articles, especially the series entitled ‘Studies in Islamic Metalwork’ that appeared in the Bulletin of the School of Oriental (and African) Studies in the 1950s. He also wrote a group of articles and monographs on metalwares associated with Mosul (1949, 1951, 1957a) and another group on Mamluk metalwares, particularly the so-called Baptistère de St Louis (see Figure 9.48), the large inlaid basin in the Louvre that is perhaps the most famous piece of Islamic metalware (1950, 1953a, 1957b). In addition to the information in these studies, Rice’s work is notable for his superb drawings of the pieces and their inscriptions.

— 106 —

— M E TA LWA R E S —

9.48 Inlaid brass basin known as the Baptistère de St Louis made by Muhammad b. al-Zayn. Diameter of rim 50 cm. Paris, Louvre, LP 16.

Other places to seek information on inscribed metalwares are catalogues that record the texts on particular pieces. One of the first was Wiet’s volume Objets en Cuivre (1932), which lists 566 copper and bronze objects with historical or other significant inscriptions in the Islamic Museum in Cairo. A pioneering work in the field, it is somewhat outdated and the inscriptions should be treated cautiously. Other good catalogues of major public collections include the 1985 catalogue of metalwork in the Freer Gallery by Atıl, Chase and Jett and the 1990 catalogue of the travelling exhibition from the Hermitage, Masterpieces of Islamic Art in the Hermitage Museum. Good catalogues of private collections include Fehérvári (1976) on the Keir collection and Allan (1982a, 1986) on the Nuhad al-Said and Aron collections. Metalwork was one of the major arts under the Mamluk dynasty of Egypt and Syria, and Atıl’s 1981 catalogue of the travelling exhibition on Mamluk art, Renaissance of Islam, includes

the inscriptions on the thirty-four masterpieces exhibited there. Most catalogues illustrate the finest wares, and so learning about the inscriptions on metalwares used by ordinary folk in daily life is difficult. Several recent essays deal with the inscriptions on a particular type of metalware. SourdelThomine’s 1971 essay on keys and locks to the KaÆba is a good example of the kinds of information that can be derived from such a study. She divided the twenty-one examples from the Topkapı Museum in Istanbul into three chronological periods, dating from the late ÆAbbasid, early Mamluk and high Ottoman periods. By looking at the objects as a group, she was able to trace the evolution of ritual, politics and imperial aspirations. She also found that certain Koranic verses referring to Mecca, especially 3:90–1 about God’s house, were standard on these keys, thereby showing that other keys with similar verses might well have been made for the same purpose.

— 107 —

— ISLAMIC INSCRIPTIONS — Other significant essays on a particular type of metalware include Melikian-Chirvani’s (1986b) article on state inkwells from Iran and his 1990–1 article on beggar’s bowls (Persian kashkËl). Using inscriptions on the pieces and evidence from poetry and historical texts, he traced the evolution of these types of object over the centuries. Like much of his work on metalwares, these articles are extremely informative for the reading of inscriptions and texts, even if some of his conclusions remain unsubstantiated. The most frequent type of text written on metalwares from the Islamic lands contains blessings, good wishes and prayers, usually invoked on an anonymous owner (li-ßå˙ibihi). There are relatively few surveys of these texts, although analysing the changing words might well help us to distinguish metalwares from different regions or periods. Eugen Mittwoch compiled a preliminary list based on pieces in Sarre’s collection (Sarre 1906: 67–82), and Baer (1983: 209–12) gives a general summary. The most common word was baraka (‘blessing’), which appears on the earliest pieces. Other popular terms introduced in the middle period were al-yumn (‘good fortune’) and Æizz wa iqbål (‘glory and prosperity’). The list of attributes grew steadily over time, until a single benedictory inscription could include as many as twenty-five or thirty attributes, often arranged in rhyming pairs. Metal objects, particularly those made of precious metals or with inlay, were expensive and prized possessions, and the owner’s name was often added after the piece was made. One of the earliest examples is a set of three silver bowls and two forks found in Mazandaran and now in the Iran Bastan Museum in Tehran (Ghirshman 1957). Inscriptions roughly engraved in Pahlavi on the crosses decorating the exterior of the bowls give the name of a local ruler in the late eighth century, Windad Ohrmazd of the Karens, and the weight in drachms. Similarly, a large hemispheric silver bowl decorated with a harpy found in Perm in 1909 and now in the Hermitage (Masterpieces of Islamic Art 1990, no. 11; Marschak 1986, figs 115–18) has a short inscription in simple Kufic on the exterior saying that the bowl was for (li) the treasurer (al-khåzan) Khumartikin, a name

borne by several Turkish military slaves serving in various medieval courts. In both these cases, the owner’s name is independent of the rest of the decoration, but, beginning in the late tenth century, patrons’ names were included as part of the benedictory inscription so that the blessings inscribed on metalwares were invoked upon specific individuals. Identifying these people can help to date groups of metalwares, so scholars have paid a great deal of attention to these pieces and the names on them. They are often the fanciest objects, made of precious metals or having extensive inlay, and were specific commissions, as distinct from the majority of anonymous pieces which may have been made for the market. The people named on the earliest pieces are usually rulers or court officials, but from the twelfth century the range of patrons who ordered fine metalwares expanded to include merchants, religious leaders and tradesmen. The earlier form of li (‘for’) was also replaced by birasm (‘intended for’ or ‘on the order of’), used on a bronze lion in Berlin attributed to twelfthcentury Egypt (Museum für Islamische Kunst 1971, no. 313). The desire to have one’s name inscribed on a metal object reached its apogee under the Mamluks, rulers of Egypt and Syria from 1250 to 1517, when the inscription giving the patron’s names and titles, usually arranged in alternating cartouches and roundels or radiating bands, was the major and sometimes the sole form of decoration. Having one’s name inlaid in glittering letters on elaborately decorated metalwares became a significant sign of rank, and these fancy objects were often presented as gifts to the court. Given the importance of having one name in lights, as it were, it is all the more curious that we are unable to identify some of the patrons named on the most splendid wares, despite the plethora of texts about the Mamluk period. A good example is a stunning inlaid mirror in the Topkapı Museum (see Figure 9.49), made by the master Muhammad al-Waziri and decorated with signs of the zodiac (RCEA 6105; Köseo©lu 1987, no. 109). The band in the centre gives the patron’s name, ÆAlaæ al-Din, and the large radial inscription identifies him as

— 108 —

— M E TA LWA R E S —

9.49 Inlaid steel mirror made by the master Muhammad al-Waziri. Diameter 24 cm. Istanbul, Topkapı Palace Treasury.

a high Mamluk official (al-janåb al-ÆålÈ). Despite his lengthy titles, he has not yet been identified with any known historical figure. Patrons’ names on these magnificent Mamluk metalwares were not only written in words in dedicatory inscriptions but also represented by emblems or signs of office. The first emblems were pictorial, with polo-sticks designating a polomaster, a pen-case a secretary, and the like. The first example to survive is a cup on a candlestick made c. 1290 for the amir Kitbugha, who was cup-bearer to the sultan al-Nasir Muhammad

(Arts of Islam, 1976 no. 211; Atıl 1981, no. 16). A five-petalled rosette served as the emblem of the Rasulids, rulers of the Yemen contemporary with the Mamluks. From the mid-fourteenth century, several different charges were combined in a composite emblem, as in the eagle over the cup used by the amir Tuquztimur (d. 1345) on an inlaid bronze basin in the Islamic Museum, Cairo. These pictorial emblems were eventually replaced by epigraphic ones. The Mamluk sultan al-Nasir Muhammad (r. 1294–1340 with interruptions) used a roundel with the phrase ‘glory

— 109 —

— ISLAMIC INSCRIPTIONS — to our master, the sultan’ inscribed in the central band, and this anonymous phrase was soon replaced by the name of the specific ruler or amir. The text grew in size and length, so that the names of the sultan Qaæitbay (r. 1468–96) fill all three registers of the emblem. From the 12th century, inscriptions on metalwares name not only the person who ordered the piece but also the person for whom it was intended or the place to which it was endowed. A key piece, as for so many other innovations, is the Bobrinsky bucket, a stunning inlaid bronze bucket in the Hermitage (see Figure 9.50). The text around the rim (Ettinghausen 1943; Arts of Islam 1976, no. 180; Masterpieces of Islamic Art 1990, no. 30) tells us that it was ordered by ÆAbd al-Rahman b. ÆAbdallah for the merchant Rashid al-Din ÆAzizi b. Abuæl-Husayn al-Zanjani. The lengthy inscriptions allow us to make further deductions about the people involved and the object’s function. The person who ordered the bucket was apparently in service to the person to whom he gave it, since the commissioner, ÆAbd al-Rahman, is identified as al-Rashidi, and the person to whom he gave it is called Rashid al-Din. The commissioner may well have been a manumitted slave, since no personal name is given for him or his father. The recipient was clearly a successful merchant, for he is identified as an exalted khwåja and pride of merchants (fakhr al-tujjår). He was also a religious man, for he is identified as pillar of the faith (rukn al-dÈn), most trustworthy of Muslims (amÈn almuslimÈn) and ornament of the pilgrimage and the two shrines (that is, Mecca and Medina) (zayn al-˙åjj wa’l-˙aråmayn). The bucket was ordered to commemorate a specific event, for it is precisely dated to the month of Muharram 559/December 1163, the same month mentioned on five enamelled bowls of the same period (see below, Chapter 11). The event may have been the recipient’s successful completion of the pilgrimage to Mecca, which would have taken place at the beginning of the previous month, Dhuæl-Hijja. Inscriptions tell us that from the fourteenth century onwards one of the standard acts of piety by rulers and courtiers was to donate large metalwares to major mosques and shrines. These patrons

wanted to have their munificence remembered, and the inscriptions naming the donor are the most significant decoration. Many of these objects were light fixtures. A large candlestick base in the Museum of Fine Arts, Boston (Blair and Bloom 1994, fig. 26), for example, is decorated with four cartouches inlaid with the dedicatory inscription recording that it was endowed to the shrine of the mystic Bayazid Bastami by a vizier of the Ilkhanid sultan Uljaytu in 708/1308–9 (MelikianChirvani 1987). The largest candlestick to survive from Islamic Iran (it measures almost half a metre in diameter), it was a splendid (and expensive) present from the vizier to honour his favourite Sufi master. Similarly, the Mamluk sultan Qaæitbay had his name inscribed on several candlesticks that he endowed to the mosque of Medina in 887/1482–3 (Atıl 1981, no. 34). Another example comprises a pair of large gilt brass candlesticks in the Museum of Turkish and Islamic Art, Istanbul (no. 139) inscribed with Persian or Ottoman Turkish poems praising the light emitted from the candlesticks, apparently a paraphrase of a poem by Jami. The inscriptions further state that the candlesticks were made for the Ottoman sultan Bayazid II at Edirne and were probably intended for the mosque complex which he had built there beginning in 1484. Other large metalwares donated to mosques and shrines were large cauldrons for water. A huge cast-bronze example standing in the courtyard of the congregational mosque at Herat (MelikianChirvani 1969a) was commissioned in 767/1374–5 under the Kart rulers of the city by a dervish (qalandar). An even larger one (see Figure 15.87) was commissioned by Timur for the shrine of Ahmad Yasavi at Turkestan City on 20 Shawwal 801/25 June 1399 (Komaroff 1992a, appendix I). The inscriptions naming the patron are usually incorporated into the overall decorative schema; sometimes, however, an owner’s name was inscribed long after the metalware was made, and the name identifies a collector. The most famous example from Iran is Muhammad Muqim, whose name is incised in nastaÆlÈq script within a shield shape on the base of several fourteenthcentury bowls in the Georgian State Museum in

— 110 —

— M E TA LWA R E S —

9.50 Bobrinsky bucket made by Muhammed b. ÆAbd al-Wahid and Hajib MasÆud b. Ahmad in Muharram 559/December 1163. Bronze inlaid with copper and silver. Ht 18.5 cm. St Petersburg, Hermitage IR-2268.

— 111 —

— ISLAMIC INSCRIPTIONS — Tbilisi, the Los Angeles County Museum of Art and elsewhere (Melikian-Chirvani 1982: 228). Judging from the style of script, the inscriptions seem to have been added in the Safavid period, two or three centuries after the pieces were made, but so far it has been impossible to identify this particular individual among the many with the same name mentioned in Safavid sources. Collecting the objects that bear his name tells us about taste in Safavid times and establishes that these objects were together at one time, adding indirect weight to the suggestion of a common provenance in the same area. Beginning in the eleventh century, maxims were introduced into the repertory of inscriptions on metalwares. Given the unpointed texts, they are exceedingly difficult to decipher, and only a few have been read. Melikian-Chirvani (1986a) deciphered one that is found on several pieces attributed to the patronage of the Ghaznavids, including a bowl in the Metropolitan made by Abu Nasr the designer (al-naqqåsh): Keep your tongue by saying little. Verily calamity is linked with discourse. For every time there is a writing, For every deed a retribution, For every deed appropriate men. Careless mistakes in writing the inscriptions on metalwares are common. A small box in the V&A decorated with the ordination service of a Christian church (no. 320–1866; illustrated in Rice 1951, plate 15) is inscribed with a corrupt rendering of two verses by the pre-Islamic poet al-Nabigha al-Dhubyani from an apology and panegyric which he composed for NuÆman b. Mundhir, King of Hira. The text is a rare if not unique example of a pre-Islamic poem inscribed on Islamic metalwork, but mistakes in the metre and reading show that the artist was careless or ignorant when inscribing the text. Similarly, the inscriptions on some of the fine metalwares made for Badr al-Din Luæluæ, atabeg of Mosul from 1210 to 1259, contain mistakes and misspellings in different parts of the patron’s name, including his titles, patronymic (kunya) and even personal name (ism), yet these objects were

still accepted for use in his household. Many mistakes in the inscriptions on later metalwares seem to be akin to scribal errors. The poems inscribed on metalwares in Arabic or Persian often refer to the object’s function. One example is a gold bowl belonging to a hoard found at Nihavand near Hamadan and now in the British Museum (Ward 1993, no. 38). The exterior rim bears verses by the tenth-century poet Ibn al-Tammar from Wasit: Wine is a sun in a garment of red Chinese silk. It flows; its source is the flask. Drink, then, in the pleasance of time, since our day Is a day of delight which has brought dew. The reference to wine-drinking shows that the bowl was part of a wine service, probably for the amir Abu ShujaÆ Injutikin, whose name is inscribed on a buckle found in the hoard. Similar verses about the forbidden pleasures of drinking, but in Persian rather than Arabic, are inscribed around the inside rim of the hemispheric silver bowl decorated with a harpy in the Hermitage (Masterpieces of Islamic Art 1990, no. 11). The tradition of inscribing verses on metalwares became particularly popular in Iran. Several verses with good wishes from SaÆdi’s BËstån were used from the fourteenth century, and the range broadened considerably in the Timurid period. Developing the methodology established by Ivanov, Komaroff (1992a, 1992b) showed how fruitful it can be to identify these verses. The repetition of the same verse on different pieces can help us to identify the production of a single artisan or workshop, and, more generally, identifying and analysing these poems sheds light on the cultural milieu in which these pieces were produced. The verses on Timurid metalwares, for example, are some of the earliest or most accurate versions of poems by such local poets as Qasim al-Anwar, Salihi and Jami, and metalwares are thus an unexpected resource for the history of Persian literature. Although less common, Arabic verses were also inscribed on metalwares made elsewhere in the Islamic lands during the later period. A good example is the enormous bronze chandelier

— 112 —

— M E TA LWA R E S — suspended from the central bay of the great mosque in Taza (Blair and Bloom 1994, fig. 151). Composed of nine circular tiers of diminishing size arranged like a wedding-cake and measuring 2.5 metres in diameter, the chandelier could hold 514 glass oil lamps and is the largest extant example of a type found throughout North Africa. It is decorated with delicate arabesques and inscriptions, including verses in basȆ metre inscribed on the interior and written as though the lamp were speaking. Obviously composed for the occasion, they encourage the viewer to behold the chandelier’s magnificence and record that this wondrous lamp was offered to the mosque by the Marinid ruler Abu YaÆqub in 694/1294 (Terrasse 1943: 12–13). Many of these metalwares with inscriptions naming the patron were specific commissions, intended as a gift or as part of a personal set, and the inscriptions often include the date. It is usually written out in words, either at the end of the text naming the patron or in an inconspicuous place, such as under the clasp on a box or on the handle of a jug or bucket. Although they are rare, dated pieces are particularly important to art historians, since the date is a key to attribution. Many people look first for the date, but it is important to remember that dated pieces are the exception rather than the rule, and that adding the date implies that the piece is exceptional rather than the standard anonymous issue made for the market. Finding the date on several ewers from the early period, for example, implies that they too were specific commissions. Another type of information written on Islamic metalwares is the artist’s signature. The first place to investigate the signature of a particular metalworker is L. A. Mayer’s handbook, Islamic Metalworkers and Their Works (1959), which, although outdated, gives the largest repertory of signatures. Signed pieces outnumber dated ones, and the fairly large number of signed pieces shows that metalworking was a high-status craft in the Islamic lands. Artists’ names were among the first pieces of information inscribed on metalwares. They are found on several ewers from the early period (see below), and the smiths

who made the iron gates at Yazd and Ganga in the mid-eleventh century (Blair 1992a, nos 41, 49) are some of the first craftsmen known from Islamic Iran. The high status of metalworkers is further indicated by the epithets which they bear from the medieval period onwards. Some are called muÆallim (‘teacher’) or uståd (‘master’); others pupil tilmÈdh or servant/hireling (ghulåm). These ranks indicate a hierarchical organisation of labour, and the different verbs used in signatures confirm the increasingly elaborate organisation of the metalworking crafts in medieval times. The most common word to introduce the craftsman’s signature is Æamal (‘work of’). It also has the broadest meaning. Another term ßanÆa (‘made by’) was more limited. Rice (1953c) speculated that it was reserved for vessels whose dimensions were based on mathematical or astronomical calculations and whose execution required care and precision. His supposition is borne out by the recently discovered signatures of the metalworker Muhammad b. Hamid b. Mahmud al-Isfahani, who signed his name with ßanÆa on astrolabes, but with Æamal on a box with combination locks (Maddison 1985). With the increasing elaboration of designs and techniques of metalworking, a single piece could be signed by several craftsmen using different verbs, including naqqasha (‘engrave’ or ‘decorate’) and †aÆÆama (‘inlay’). The splendid inlaid bronze Koran box in Berlin (Arts of Islam 1976, no. 214), for example, was made (Æamal) by Muhammad b. Sunqur and inlaid (ta†ÆÈm) by Yusuf b. al-Ghawabi. Sometimes two artists are named, but it is difficult to distinguish their different jobs from the words used. The commissioning inscription around the rim of the Bobrinsky bucket (see Figure 9.50), for example, names two artists: Muhammad b. ÆAbd al-Wahid, who formed or inlaid it (the word used is ‰arb), and Hajib MasÆud b. Ahmad, the designer from Herat (al-naqqåsh-i haråt), who made (Æamal) it. An artisan’s nisba or epithet is often used to suggest the provenance of a particular metalware or group. Thus, the epithet ‘of Herat’, which occurs on several of the finest pieces made in the medieval period, including the Bobrinsky bucket and the Tbilisi ewer, is taken to show that the pieces were

— 113 —

— ISLAMIC INSCRIPTIONS — made in Herat. This is certainly possible, but without other confirmation it is not necessarily true. In the case of Herat, texts tell us that Herat was the centre of a metalworking industry, and so the assumption is logical. In other cases, it is patently wrong to assume that the artisan’s epithet shows the place of production. The epithet al-Mawsili (‘from Mosul’), for example, is found on at least thirty metalwares dating from the thirteenth to the early fourteenth century. According to the inscriptions on the pieces, at least one, the Blacas ewer in the British Museum (see Figure 9.53) was made in Mosul (RCEA 4046; Arts of Islam 1976, no. 196), but others were made in Damascus or Cairo. This nisba does refer not to a geographical site, but rather to a product or technique associated with that city or to a group of artisans who traced their lineage back to metalworkers in Mosul. Metalworkers with the nisba al-Mawsili are just like builders with the nisba al-Shirazi (see Chapter 4). There are also different kinds of signatures on metalwares. Formal signatures are more prominent and written in bolder script, sometimes at the end of the commissioning or dating text or in an isolated place, often with the date. They can be contrasted with informal signatures that are generally smaller and often contain visual puns. The difference between the two types can be seen on the so-called Baptistère de St Louis in the Louvre. The artist was justly proud of his masterpiece and signed the large brass basin in six different places. The formal signature ‘work of the master (muÆallim) Muhammad b. al-Zayn, may [God] forgive him’ is engraved on the outside rim below the rim and centred above a roundel with a rider spearing a bear (see Figure 9.48). The position, prominent size and inclusion of a benediction show that this is a formal signature. A shorter signature, ‘work of Muhammad b. al-Zayn‘, is found on five other representations of metal objects and thrones within the scenes. These informal signature are visual puns, referring to Muhammad b. al-Zayn’s talent as a metalworker who made thrones, cups and other objects. The punning nature of these signatures is even clearer on another of his pieces, the Vasselot bowl in the Louvre (Atıl

1981, no. 20), where a short signature ‘work of Ibn al-Zayn’ is found on the bowl held by one of the courtiers and similar to the object itself. Inscriptions on metalwares are written in a variety of ways using a variety of scripts, and the method and style of text affects the content. The easiest is to incise the text. This technique allows the most latitude in style and content. Inlaying the inscription is more difficult, but it has the advantage that the coloured inlay can be set off against the ground, thereby highlighting the words. A few, repetitive inscriptions are cast in relief, as on a group of medieval bronze ewers with a high spout (al-ÆUsh 1972). Since these objects were cast from moulds, the texts are usually confined to brief blessings to an anonymous owner. The same sorts of short texts could also be hammered in relief, as on a group of silver bottles, jugs and other objects from medieval Iran, or pierced, as on lamps, incense burners and other objects that required an openwork surface. Very often, different techniques were used on the same metal object for different types of text. One of the highspouted ewers in the Metropolitan Museum, for example, has other inscriptions done in the champlevé technique, including the signature of the artisan. Within the limitations imposed by the techniques of manufacture, the inscriptions on metalwares follow the same general stylistic developments as do scripts on buildings and other objects. There is the same general progression from angular scripts to cursive ones and from simple to elaborate, with foliation, floriation and knotting. One style of script, however, is specific to metalware: animated inscriptions, in which the letters or parts of them assume animal or human forms. With one exception (the so-called Baylov Stones, a stone frieze retrieved from a fortress on an island in the bay of Baku and now exhibited in the palace of the Shirvanshahs there), animated script is found only on metalwares. Animated script seems to have evolved from ornithomorphic writing in which letters are transformed into birds or end in birds’ heads. This type of bird-writing is found not only on metalwares (for example, a bronze bowl in the Metropolitan

— 114 —

— M E TA LWA R E S — Museum, illustrated in Baer 1983, fig. 177), but also on slip-painted ceramics attributed to the patronage of the Samanids, rulers of eastern Iran in the tenth century. The repertory of creatures used in animated script gradually expanded, and the ends of the letters were transformed into the heads of other types of animals, both real and imaginary, or humans. The key object for dating the introduction of zoomorphic and human-headed scripts is the Bobrinsky bucket (see Figure 9.50). In the upper band, the letters are zoomorphic, either ending in cow, dragon or bird heads or shaped like birds. Atop these zoomorphic letters are human figures. In the bottom band, the letters are human-headed. The artist’s delight in playing off different scripts is clear, and he separated these two types of animated script with a middle band of interlaced

script. These hard-to-read inscriptions contain blessings to an anonymous owner. When the artist wanted to convey historical information, including the names of the patron, the artists and the recipient and date, he used a much simpler and more readable script for the texts along the handle and rim. An inlaid brass pen-box in the Freer Gallery (see Figure 9.51) shows how metalworkers in eastern Iran in the late twelfth and early thirteenth centuries were able to combine legibility and decoration in different types of plain and animated script (Herzfeld 1936; RCEA 3671). The inscription around the lid, written in a legible naskh, conveys the historical information, with the names and titles of the owner, Majd al-Mulk, grand vizier to the penultimate Khwarazmshah ÆAlaæ al-Din Muhammad. The large inscription

9.51 Inlaid brass pen-box made by Shadhi in 607/1210–11. Length 31.4 cm. Washington

DC,

Freer Gallery of Art 36.7.

— 115 —

— ISLAMIC INSCRIPTIONS — around the base, also in naskh but with human heads, contains lengthy blessings to an anonymous owner. The inscription on the back between the hinges, although smaller and thinner, is a dramatically balanced and refined human-headed Kufic: it records the name of the artist, Shadhi, and the date 607/1210–11. Artists and objects using these zoomorphic and human-headed scripts apparently moved westwards with the Mongol invasions, for these scripts appear on wares made in the central Arab lands in the second quarter of the thirteenth century. The first dated example is a ewer in Cleveland made by Ahmad b. ÆUmar al-Mawsili in 620/1223 (Rice 1957a; Arts of Islam 1976, no. 195). These scripts were used on many wares made in the mid-thirteenth century, but became increasingly repetitive and stylised until they were finally abandoned at the end of the thirteenth century or the beginning of the fourteenth. Zoomorphic and human-headed scripts are still readable, albeit with difficulty, for only the ends of the letters are transformed into animals or humans, and the bodies of the letters below the heads are still decipherable as letters. The next step, in which the letters themselves (and not just the ends) are transformed into animals or birds, renders the inscription virtually unreadable. This script was restricted to the most generalised texts, with repeating blessings, but even these are hard to decipher. The animated inscription with good wishes on a large inlaid canteen in the Freer (Atıl et al. 1985, no. 17) was only deciphered recently, and the inscriptions on the Blacas ewer in the British Museum (see Figure 9.53) have never been deciphered. Animated script was probably not meant to be read, but rather represented the good life characterised by the hunters, revellers and other figures depicted in it (Bloom 1987a). Animated script was used on metalwares produced throughout the thirteenth century, first in Khurasan and then later in Mesopotamia. One of the key pieces decorated with animated script is the Wade Cup in Cleveland, superbly illustrated by Rice (1955) and further analysed by Ettinghausen (1957). Animated script was briefly adopted by the

Mamluks. The first datable example to survive occurs on the socket of a candlestick in Cairo made for the amir Kitbugha c. 1290 (Atıl 1981, no. 15). The script was soon abandoned by the Mamluks, for whom legible names and titles on metalwares were particularly important. Scientific instruments, such as brass astrolabes, celestial globes, divination tables and the like, form a special category of metalwares, which are particularly important in documenting the history of metalworking since many of the finer objects are signed and dated. The artisans who made scientific instruments were particularly skilful, often signing their work with ßanaÆa (‘crafted’) rather than Æamal (‘made’). The most prolific were astrolabists (a߆urlåbÈ). Mayer (1956b) compiled the basic list of their names. There is no overlap between Mayer’s volumes on Islamic astrolabists and metalworkers (1959), but as the names of more metalworkers are recorded, it is clear that the skilful artisans who made scientific instruments sometimes made other fine objects. Maddison (1985) has documented, for example, a twelfth-century Isfahani family of astrolabists comprising Hamid b. Mahmud al-Isfahani and his two sons, Muhammad (who made four astrolabes and two combination locks) and MasÆud (who signed a pen-box dated 581/1185–6). Similarly, Muhammad b. Khutlukh al-Mawsili, the artisan who made a splendid divination table in 639/ 1241–2 in the British Museum (Savage-Smith and Smith 1980), also made and signed an inlaid brass incense burner in the Aron Collection (Allan 1986, no. 1). As with other metalwares, the craft of making scientific instruments became increasingly specialised, and different artists added the decoration. This was already true by the thirteenth century, as shown by a fine astrolabe in the Deniz Museum, Istanbul, made in Damascus in 614/ 1217–18 for the Ayyubid sultan, al-Malik al-Muzaffar Sharaf al-Din. It was made by ÆAbd al-Rahman b. Sinan al-BaÆlbakki al-najjår and inlaid by Siraj al-Dimishqi, for the calculator (of scientific scales and lines), ÆAbd al-Rahman b. Abi Bakr, the muqawwam of Tabriz. This same division of labour occurs on many Safavid astrolabes as well.

— 116 —

— M E TA LWA R E S — The lengthy inscriptions on astrolabes are important evidence for centres of metalware production. A series of Andalusian astrolabes, for example, were made at Seville, Murcia (one dated 650/1252–3), Granada (three dated 664/ 1265–6, 704/1304–5 and 709/1309–10) and Guadix (one dated 720/1320–1). Astrolabes are also important in documenting the use of abjad, the alphanumeric system of dating (see Chapter 15) which is used for the numbers around the rete.

Early Period The earliest metalwares made in the Islamic lands did not include inscriptions as a part of their decoration; only gradually were inscriptions added. The novelty of inscribing pieces is evident from the inscription on a bronze vessel in the form of a bird in the Hermitage (see Figure 8.47). The all-over decoration in silver and copper inlay includes a band in simple Kufic around the neck. The text (Masterpieces of Islamic Art 1990, no. 1, with references to earlier publications ) begins with the invocation and blessings to God and then records that the piece was made by Sulayman at a certain city in the year 180/796–7. The craftsman’s unfamiliarity with writing inscriptions is clear from the awkward layout and style. He ran out of space at the end of the text and had to write the numbers of the date below the name of the city. Unfortunately, the name of the city where he made the vessel is unreadable. Suggestions have ranged from Kasan in Central Asia to the Fazz quarter in Nishapur and Kashan in Iran, but none fits the letters, which seem to read alif, låm, Æayn, sÈn, råæ. Since this ewer is a key piece in localising other pieces, such as another bird in the Monastery of St Catherine at Sinai with a similar-style inscription invoking blessings to its owner (Baer 1983, fig. 167), the question of the city inscribed on the Hermitage bird is all the more important. The difficulty in reading the unpointed inscriptions on these early ewers is clear from another ewer in the Georgian State Museum, Tbilisi, which is inscribed in a single line around the rim (see Figure 9.52). Since the work of Orbeli

and Trever 1935) and Diakonov (1947), scholars have recognised the importance of the inscription for dating and localising other ewers of a similar shape, but just about every word in the text has been debated, even where to begin reading, for the choice of beginning affects the grammar of what follows. Assuming that the text is a single sentence, and by analogy with the inscription on the bird ewer, which begins with blessings and ends with the date, then we should start reading on the right side of the rim just beyond the thumb rest. The text begins with blessings to its maker (literally, ‘he who fashioned it’), Ibn Yazid. Next is a phrase usually read as ‘part of what was made at Basra’ (mimmå Æumila bi’l-baßra). Such a reading forms the basis for assigning this group of ewers to Iraq, but Melikian-Chirvani (1976, n. 4) suggested alternative (and somewhat less plausible) readings of the last word as al-Badra, biæl-nudra (‘cast in one’) or even biæl-bukra (‘for enjoyment’). The inscription of the Tbilisi ewer ends with the date written out in words in units and tens, first read as ‘the year seven (or nine) and sixty’, meaning 67/686–7 or 69/688–9. Reading the ‘units’ digit is a minor problem. Either seven or nine is possible, although the slight slant of the

9.52 Drawing of the inscription around the rim of a bronze ewer made by Ibn Yazid at Basra? in [1]67/783–4. Georgian State Museum, Tbilisi.

— 117 —

— ISLAMIC INSCRIPTIONS — teeth might add weight to the former. Interpreting the rest of the date is more problematic. Marshak (1972) suggested that the engraver might have dropped the ‘centuries’ digit due to insufficient space and that the ewer should therefore be dated to [1]67 or [1]69 (corresponding to 783–4 or 785–6) or even [2]67 or [2]69 (corresponding to 880–1 or 882–3). Given the analogies with the text on the bird ewer, Marshak’s suggestion makes sense and shows how the inscription on a single piece should be read in conjunction with others on similar pieces. By the tenth and eleventh centuries, inscriptions played a more important role in the decoration of Islamic metalwork. Inscriptions invoking blessings and good wishes on a specific patron allow us to differentiate several geographical groups of metalwares from these centuries. A key piece for localising early Spanish metalwares is the Gerona casket (see Figures 1.2 and 8.46), whose foundation inscription dates it to eight months of the year 976. Silver caskets for storing precious objects continued to be made in Spain until the twelfth century. They are usually inscribed with blessings to an unspecified owner. Similarly, the inscriptions on silver and gold wares made in the eastern Islamic lands in the tenth and eleventh centuries name several princes and high-ranking officials. In addition to the wine service of Abuæl-ÆAbbas Valgin b. Harun (see Chapter 8, pp. 104–5), several other silver cups, jugs and bottles in the Hermitage are inscribed with the names of rulers or officals from Transoxiana. Many of the texts were first read by van Berchem (1909b) from photographs published in Smirnov’s 1909 catalogue of metalwares in the Hermitage; some are revised and summarised in MelikianChirvani’s essay (1982a) on medieval Iranian silver. Like the inscriptions on the Gerona casket, the inscriptions on these silver wares are written in a variety of scripts, often with several styles played off against each other on the same object, and are well integrated into the overall decorative scheme. One style of elegant thin Kufic script with paired verticals is engraved on a rectangular tray (CA-8255; Marschak 1986, no. 40; Masterpieces of Islamic Art 1990, no. 20) with the name and titles

of Abu Ibrahim, a Khwarazmshah who reigned in the eleventh century. The same style is used for a list of blessings around a silver flask (CA-8256; Marschak 1986, nos 138/139; Ward 1993, no. 39). The thin letters contrast with a broader band around the shoulder with thick Kufic letters set against a floral scroll ground. The inscription names the patron, Abu ÆAli Ahmad b. Muhammad b. Shadhan, who served as vizier at Balkh for the early Saljuqs at the beginning of the eleventh century. Other inscriptions in a similar thick Kufic on a scroll ground on a silver flask (V3-800; Masterpieces of Islamic Art 1990, no. 27) offer blessings to the shaykh, Abu Fadl ?Sahat b. ÆAli. Inscriptions on other precious metal vessels, including a handful of small gold jugs and silver dishes and candlesticks, name contemporary rulers and officials in western Iran, particularly the Buyids. The authenticity of many of these pieces is contested, and oddities in the wording and style of the inscriptions underscore the uncertainties about these pieces. The most (in)famous example is the so-called Alp Arslan salver in the Boston Museum of Fine Arts. A large Kufic inscription in the centre contains the words, the sultan ÆAdud al-Din, and a Kufic band around the rim states that it was offered to the illustrious sultan Alp Arslan, ordered by a queen of the age, and made by Hasan of Kashan in 459/1066–7 (RCEA 2661; Survey, plates 1347–8). The unusually lengthy inscription, its jumbled titulature and its peculiar style cast doubt on the authenticity of the piece, for inscriptions on objects from the tenth and eleventh centuries usually follow standard titulature and forms. The laqab of the Saljuq sultan Alp Arslan, for example, is ÆAdud al-Dawla, not the ÆAdud al-Din inscribed prominently on the piece. The form of the text, indicating that it was made as a gift, is unprecedented, and the style of the Kufic with curvaceous ascenders is distinct. The unusual inscription adds weight to the doubts raised about this piece and connects it to many other objects purportedly made in medieval Iran that appeared on the art market at this time (Kühnel 1956; Blair et al. 1992). Gold and silver medallions with figural decoration are also inscribed with the names of Buyid

— 118 —

— M E TA LWA R E S — rulers. According to the historian Ibn al-Athir, on 1 Muharram 378/21 April 988 (New Year’s Day) the vizier Ibn ÆAbbad presented Fakhr al-Dawla, the Buyid ruler of Rayy, with a gold dinar weighing 4.25 kg. It was inscribed on one side with seven lines of Arabic poetry extolling the piece as a sun and the ruler as king of kings and on the other with Chapter 112 of the Koran (Surat al-Ikhlas), the names of the ÆAbbasid caliph and the Buyid ruler, and the place where it was struck (Gurgan). No such enormous piece survives, but smaller ones in both silver and gold are known (Miles 1964b, 1975; Ettinghausen and Grabar 1987, fig. 253). Like many of the other objects naming the Buyids, the authenticity of some of these pieces has been questioned. Inscribed names are also the key to attributing a group of cast-bronze lions, probably used as aquamaniles or fountain heads, to Fatimid Egypt. One in Berlin (Museum für Islamische Kunst 1971, no. 313) bears the name of the person who ordered it, the governor of Egypt Shams al-Din. Another in Kassel (Arts of Islam 1976, no. 168) is inscribed with the name of the maker, ÆAbdallah the sculptor (al-maththål), and a cast-bronze lampstand (ibid., no. 171) is the work of Ibn al-Makki. The dependence on a recognisable name in an inscription for attributing a metalware from the early period to a particular locale is clear from one of the largest pieces of Islamic metalwares to survive: the Pisa griffin, so called because it was installed atop the cathedral in Pisa during the late eleventh or early twelfth century and remained there until 1828 (Dodds 1992, no. 15). A huge statue of cast bronze, it is incised over almost all the surface with textile-like patterns. A band of floriated Kufic around the body invokes blessings to an unspecified owner. While there is general agreement about an eleventh-century date, attributions range from Iran to Sicily, Egypt, North Africa and Spain. The generalised blessing and integration of the inscription into the overall decoration accord well with the dating, but investigation of the style of script and the particular words used in the blessing might cast further light on the provenance of this enigmatic piece.

Middle Period Metalwares made between the twelfth and fifteenth centuries are often packed with inscriptions, which not only identify periods and sites of production but also give us important insights into broader social concerns, such as the status of artisans, the types of patrons and the functions and uses of the objects. They also chart changes in language and culture and show the increasing artistic sophistication of metalworkers. Inscriptions are particularly important on a group of inlaid bronzes made in the late twelfth and early thirteenth centuries in Khurasan, the province of north-eastern Iran. Five dated pieces provide the chronological framework. The earliest is a pen-box in the Hermitage (Masterpieces of Islamic Art 1990, no. 28) made on 20 Dhuæl-QaÆda 542/2 April 1148. The most famous is the Bobrinsky bucket (see Figure 9.50) made in Muharram 559/ December 1163. A faceted ewer in the State Museum of Georgia, Tbilisi (Gyuzal‘yan 1938), was made in ShaÆban 577/December 1181–January 1182. The most complex piece, an aquamanile in the shape of a zebu with a suckling calf and a lion attacking her back (Giuzalian 1968; RCEA 3627; Masterpieces of Islamic Art 1990, no. 39), is dated Muharram 603/August–September 1206. Finally, a pen-case in the Freer (see Figure 9.51) was made in 607/1210–11. The inscriptions on these inlaid bronzes show the increased complexities of metalware production. All five pieces are signed, and the craftsman often bears a specialised epithet. The most common is naqqåsh (‘decorator’) found on the Bobrinsky bucket, the zebu aquamanile and the Freer penbox. The person who made the Hermitage penbox is called ÆUmar b. al-Fadl b. Yusuf al-bayyåÆ (‘the tradesman’ or ‘middleman’). The inscriptions on the Bobrinsky bucket name two artisans: the decorator and the person who formed or made it (the word used is ‰arb), Muhammad b. ÆAbd alWahid. The increasing importance of metalworkers is also clear from the Persian verses inscribed on the Tbilisi ewer: they are composed as though the metalworker were speaking and laud the beauties

— 119 —

— ISLAMIC INSCRIPTIONS — of his creation (Gyuzal’yan 1938; English translation in Ettinghausen 1970). The inscriptions also show us the wide market for these inlaid bronzes. No patron is named on the Tbilisi ewer, so it might have been made for the market, but the other pieces were specific commissions. The most important patron was the man who owned the Freer pen-box, Majd al-Mulk, grand vizier to the Khwarazmshah ruler ÆAlaæ al-Din Muhammad. The inscriptions on the other three objects mention two people, presumably the person who ordered it and the recipient. None of them seems to be related to the court. The zebu was ordered by Ruzba b. Afridun [b.] Barzin and owned by his brother, Shah Barzin. Similarly, the Hermitage pen-box was made for ÆAli b. Yusuf b. ÆUthman, and another damaged inscription apparently mentions his brother, Shaykh [ÆUth?]man. Both people mentioned on the pen-box had completed the pilgimage to Mecca (they are designated al-˙åjj), as is the recipient of the Bobrinsky bucket, the merchant Rashid al-Din ÆAzizi. It was commissioned for him by his servant, ÆAbd al-Rahman b. ÆAbdallah al-Rashidi. The inscriptions suggest that these three splendid objects may have been pilgrimage presents: they were all specific commissions, the Hermitage pen-box just before the month of pilgrimage and the Bobrinsky bucket and the zebu the month after. The inscriptions on these five dated examples of Khurasani bronzes also track the shifting linguistic and cultural milieu in the eastern Islamic world at this time, as inscriptions in Arabic give way to those in Persian. The Hermitage pen-box, like the Tbilisi ewer, is inscribed with Persian verses, one set composed in the first person as though the artisan were speaking and two other sets about the inkpot and pen. The historical texts on the Bobrinsky bucket and the zebu are also in Persian, although Arabic is maintained for the date. The personal names of patrons, recipients and artisans on these and other pieces of the same group are often Persian. Finally, the inscriptions on these inlaid bronzes show the increasing artistic sophistication of

scribes and artists in the eastern Islamic lands. The texts are written in a variety of scripts, which are often deliberately juxtaposed to heighten the aesthetic impact. Despite their complexity, the inscriptions are still legible, with simpler scripts used for historical texts and more complicated animated ones for good wishes. On the Freer pen-box, for example, the inscription around the top with the patron’s name and flowing titles is written in naskh, the good wishes around the base in human-headed naskh, and the artist’s signature between the clasps on the back in animated Kufic. Even without reading the inscriptions, someone looking at the pen-box could distinguish three types of information from the three styles of script. Shadhi’s artistic talents are clear from his signature: the words ending his name and beginning the date are wittily arranged as two birds’ heads confronting each other over the central knot. Fine metalwares were also made in western Iran during this period, and the inscription on one of the rare pieces to survive, a parcel-gilt cup in the Keir Collection (Fehérvári 1976, no. 127; corrected in Melikian-Chirvani 1982a), is the key to localising production there. The text engraved around the rim invokes good wishes on the amir Badr al-Din Qaragöz. He can be identified from textual sources as an equerry who served as governor of the Hamadan area in the late twelfth and early thirteenth centuries. The Keir cup exemplifies a common methodology used in studying Islamic metalwork: by reading the patron’s name inscribed on a piece and then identifying him in the written sources, we can localise production to a place and a time period, in this case north-western Iran on the eve of the Mongol invasions. Another group of metalwares which have been attributed to twelfth-century Iran comprises silver dishes or shallow bowls with a band of good wishes engraved around the rim. Most of these are anonymous, but one dish in the Los Angeles County Museum (M.73.5.149) invokes good wishes on a noble lady, who is identified as queen of the earth (malikat al-ar∂) and mother of Shams al-MaÆali. She is one of the few women

— 120 —

— M E TA LWA R E S — named on an Islamic metalware, and identifying her would shed light on the group. Inscriptions also help us to identify another major centre of metalworking in the Jazira (northern Mesopotamia) and Syria in the late twelfth and early thirteenth centuries. The inlaid brasses produced there are notable for their superb surface decoration which includes a wide variety of figural subjects and bold inscriptions. As with the Khurasani wares, the numerous inscriptions are the key to localising the group and tracing its evolution and working methods. These wares are often called Mosul metalwares, for twenty-three artisans used the epithet alMawsili (‘from Mosul’). The key piece is the Blacas ewer in the British Museum (see Figure 9.53). The inscription around the neck (RCEA 4046) says that ShujaÆ b. ManÆa al-Mawsili engraved (naqqasha) it in Rajab 629 (May 1232) in Mosul. Mosul was probably the main centre of production and the starting point, but some pieces were made elsewhere. Several artisans bear the epithet al-IsÆirdi (‘from Siirt’), and Damascus and Cairo are mentioned on several pieces made in the second half of the thirteenth century. The earliest dated piece is a key to the KaÆba made in Rajab 576/November– December 1180 (SourdelThomine 1971, no. 2). Production increased steadily in the first half of the thirteenth century, as shown by a large group of dated objects, and continued in the second half of the century under the patronage of the Mamluk rulers of Syria and Egypt. Most pieces of Mosul metalware are anonymous, so we can assume that they were made for the market. The suggestion of a flourishing open market is confirmed by the layout: on many pieces the inscriptions are fitted into the decoration, rather than having the decoration designed around them. These anonymous pieces include some of the finest and most famous objects, such as the Blacas ewer and the Freer canteen. The lack of named patrons has led to varying attributions. The Wade Cup in Cleveland, for example, was first attributed by Rice (1955) to Mosul in the 1220s, but more plausibly reattributed by Ettinghausen (1957) to Khurasan c. 1200.

Despite the anonymity of many wares, inscriptions show us that local rulers played an important role in fostering the metalworking industry in Mesopotamia and the Jazira. Local rulers are named on several pieces, usually with the phrase Æizz li-mawlånå (‘glory to our lord’). Five pieces name Badr al-Din, atabeg of Mosul from 1210 to 1259, and sixteen extant pieces bear the names of Ayyubid sultans. Other local rulers named on inlaid brasses include Artuq Arslan, Artuqid ruler of Mardin and Mayyafariqin from 1200 to 1239, whose name is inscribed on a candlestick base in the Haram al-Sharif Museum, Jerusalem, and Abuæl-Qasim Mahmud b. Sanjar Shah, atabeg of the town Jazirat ibn ÆUmar (now Cizre, Turkey) in the first half of the thirteenth century, who ordered a ewer and basin (Allan 1982a, no 6). Only a handful of objects name lesser officials, such as a ewer in the Freer Gallery (no. 55.22) made in 629/1232 for Shihab al-Din Tughril, gun-bearer and Turkish regent for the Ayyubid ruler of Aleppo, al-Malik al-ÆAziz, and a bowl in the Museo Civile, Bologna (Arts of Islam 1976, no. 192) made for Najm al-Din ÆUmar, an officer of Badr al-Din Luæluæ. In contrast to Khurasani wares, very few objects were commissioned by members of other social groups. Inscriptions show that metalworkers enjoyed high status during this period, when signatures are more prominent than patrons’ names. As in Iran, the metalworking tradition was becoming increasingly complex. It was organised hierarchically, with some artisans identified as master craftsmen and others as pupils or hirelings. Some pieces are signed by two artisans, and some metalworkers, such as Muhammad b. Khutlukh, made several types of objects, including scientific instruments. The prominence of inscriptions, already evident on wares made for the Ayyubids, becomes even more marked under the Mamluks. Almost all Mamluk metalwares are decorated with inscriptions, and after the 1320s inscriptions become the main decorative theme. Metalworkers continue to enjoy high status, and signatures allow us to identify seventeen metalworkers who worked under the Bahri line (1250–1390) and five or six

— 121 —

— ISLAMIC INSCRIPTIONS —

9.53 Blacas ewer made by ShujaÆ b. ManÆa al-Mawsili in Rajab 629/May 1232. Ht 30 cm. London, British Museum 66.12–69.61.

— 122 —

— M E TA LWA R E S — who worked under the Burjis (1382–1517). Most craftsmen sign themselves al-mawßilÈ (‘from Mosul’), indicating not that they worked in Mosul but that they wanted to link themselves to that tradition. One craftsman who did not was Muhammad b. al-Zayn, probably the finest metalworker from the Mamluk period and the person who made the Baptistère de St Louis (see Figure 9.48). Signatures also allow us to suggest that metalworking, like such other fine crafts as lustre pottery, was a family profession. Muhammad b. Sunqur al-Baghdadi, the craftsman who made the splendid inlaid Koran box in Berlin (Arts of Islam, 1976, no. 214) and a hexagonal table for the sultan al-Nasir Muhammad (Islamic Art in Egypt 1969, no. 61) may have been the brother of Mahmud b. Sunqur who made a pen-box now in the British Museum dated 680/1281 (Arts of Islam 1976, no. 202; Atıl 1981, no. 13). In contrast to earlier wares, however, signatures on pieces from the Mamluk period are less important than the names and titles of the amirs and other patrons. The major decoration on many pieces is a large epigraphic band in thuluth script bearing the owner’s name and titles. The earliest dated piece with such a bold calligraphic band is a ewer in the Louvre (no. 7428), made in Damascus in 659/1257 by Husayn b. Muhammad al-Mawsili for the last Ayyubid sultan, Salah al-Din Yusuf, and the style can be traced through a series of dated pieces to its florescence in the early fourteenth century. The large band sometimes contained another text, as with the Koranic verses encircling Koran boxes in Berlin and Cairo (Atıl 1981, no. 25), but pieces made for the royal or an amiral household usually had the name and titles of the patron. Objects made for the Mamluk sultan typically begin with the same phrase invoking glory to our lord (Æizz li-mawlånå) used by earlier rulers in Mesopotamia and Syria. Pieces made for other members of the court typically begin with the commissioning phrase mimmå Æumila birasm (‘one of the things made on the order of’). These large epigraphic bands are often interrupted by roundels with the owner’s emblem, either pictorial or epigraphic. The most striking are the epigraphic roundels arranged like a sun-burst,

with the tall stems of the elongated letters pointing towards the centre. Initially the prerogative of the sultan, as on an incense burner made for the sultan al-Nasir Muhammad (Allan 1982a, no. 15), this design was gradually taken over by other members of his court, as on the mirror in the Topkapı Palace collection made for an unidentified officer (see Figure 9.49). Inscriptions tell us that some Mamluk metalwares were made for export. A bowl (sold at Christie’s in 1966) and a basin in the Louvre bear the name and titles of Hugh IV of Lusignan (Rice 1956a). The commissioning inscriptions beginning ‘part of what was ordered by’ show that this Christian king of Cyprus and Jerusalem (r. 1324– 59) was considered on the rank of an amir. Other objects were made for Muslim rulers. A large group of objects (for example, Atıl 1981, nos 14 and 22), including two specifically made in Cairo, bears the names of four Rasulid sultans, rulers of the Yemen from 1250 to 1377, or their officers of state. Unlike the inscriptions on the pieces for Hugh of Lusignan, the inscriptions on the pieces for the Rasulid sultans invoke ‘glory to our lord’, implying that the Rasulid ruler was considered the equal of the Mamluk sultan. The Mamluk style of metalwork decorated with a broad band of thuluth interrupted by roundels and text invoking glory to our master was also used on a large number of hemispheric bowls, but several distinguishing features, including the inscriptions, allow us to attribute these bowls to Iran, particularly Fars province. The roundels on the Iranian pieces usually contain figures rather than emblems, and the thuluth inscriptions are often in the name of an anonymous sultan. The key to the Fars attribution lies in the inscriptions on two inlaid vessels in the Hermitage: a bucket (Masterpieces of Islamic Art 1990, no. 51) ordered by the amir Siyavush and made by Muhammad Shah al-Shirazi, who is identified as a servant of the Inju ruler of Fars, Mahmud Shah, and a bowl made for Mahmud’s son Abu Ishaq, who became ruler of Fars in 1343. Both pieces contain the phrase warÈth mulk-i sulaymån (‘heir to the kingdom of Solomon’), a phrase that MelikianChirvani (1969b, 1971a, 1971b, summarised in

— 123 —

— ISLAMIC INSCRIPTIONS — 1982b, chapter 3) identified as part of the protocol of several dynasties ruling Fars. Another epigraphic feature that distinguishes these fourteenth-century bowls from Fars is a two-tiered inscription, with a small repeating text in Kufic inserted in the stems of the tall thuluth letters. This bowl in the Freer Gallery (see Figure 9.54), for example, has a repeated phrase that seems to read al-naÆma al-kåmila (‘perfect favour’) inserted in the stems of the thuluth text offering blessings to an anonymous sultan. These pieces were clearly mass-produced, for the texts contain generalised titles and often break off in the middle of words or phrases, and the Kufic texts are extremely difficult, if not impossible, to read. Inscriptions delineate the chronological range of the Fars school of metalwork. It was in operation from the beginning of the fourteenth century, to judge from a bowl in Modena made in 705/1305 by ÆAbd al-Qadir Shirazi. The school continued to flourish under the patronage of the Muzaffarids,

for the cover of the so-called Nisan Tasi, a huge basin in the mausoleum of Jalal al-din Rumi at Konya, mentions Shah Sultan, who was appointed governor of Shiraz in 755/1354 by his uncle, the Muzaffarid ruler of Fars (Baer 1973–4, corrected in Melikian-Chirvani 1982: 155 and n. 56). Inscriptions also tell us that the clientele for these pieces was quite broad. A candlestick in the Louvre, for example, was made for and probably also designed by the noted calligrapher Ahmad Shah (Blair 1985a). The rulers named on two extremely large pieces suggest that Azarbayjan was another centre of metalworking production in fourteenth-century Iran. The inscriptions on the inlaid basin of the Nisan Tasi invoke glory on the Ilkhanid ruler Abu SaÆid (r. 1317–35), and those on a basin endowed to the shrine of Shaykh Safi at Ardabil (Y. Godard 1936) say that it was part of what was ordered by the Jalyairid ruler Shaykh Uways (r. 1356–74). The two basins use the formulas of commissioning

9.54 Inlaid brass bowl made in Fars province in the fourteenth century. Diameter 23 cm. Washington

DC,

Freer Gallery of Art 80.25.

— 124 —

— M E TA LWA R E S — standard under the Mamluks, suggesting that Mamluk protocol set the standard for Iranian metalwares. Other inscriptions on fourteenth-century Iranian metalwares contain Persian poems alluding to the object’s function. Verses in praise of winedrinking occur on several stem cups (for example, Arts of Islam 1976, no. 204), and allegorical verses about the relationship between the pen and the inkwell appear on pen-boxes. A quatrain about the moth seeking the candle flame, an allegory for the human soul seeking God, is used on candlesticks (for example, Museum für Islamische Kunst Berlin 1971, no. I 3577). Metalwares made in the Maghrib during this period also feature inscriptions as a major element of decoration. Pieces made for the Nasrids, rulers of Spain from 1230 to 1492, were often inscribed with the dynastic slogan wa lå ghålib ilåællah (‘there is no conqueror but God’). This is found, for example, on the huge bronze chandelier made for the Nasrid sultan Muhammad III in 705/1305 (Arts of Islam 1976, no. 175; Dodds 1992, no. 57). The pierced inscription there is complete and legible, but over time the slogan was distorted and shortened into a stylised motif, as on a magnificent gilt bronze bucket and several swords made in the late fourteenth and fifteenth centuries (Arts of Islam, no. 176; Dodds 1992, nos 59–64). Many of these objects were probably made at the capital, Granada, but the only Spanish metalwares on which the site of production is mentioned are astrolabes. The finest pieces of metalwork produced in North Africa during this period are chandeliers. In addition to the huge one given to the Great Mosque in Taza by the Marinid ruler Abu YaÆqub in 694/1294, there are several other examples in mosques and madrasas in Fez. The earliest known was given to the Mosque of the Andalusians in Fez by the Almohad caliph Muhammad al-Nasir (r. 1199–1214). Others were donated by the Marinids in the fourteenth century. The inscriptions show that the important point was the munificence of the ruler in donating these large chandeliers, for the inscriptions contain the ruler’s name and titles and sometimes the date,

but never the artisan’s name or place of production. Other metalwares made in North Africa during this period are cast-bronze alms bowls; in contrast to the chandeliers, they are decorated exclusively with inscriptions detailing their function (Vicaire 1944).

Late Period Broad epigraphic bands naming the patron continued to characterise Mamluk metalwares in the fifteenth century, but a new form of calligraphy was introduced, in which the vertical shafts of the thick thuluth letters rise up in pairs and cross to form pincers. This style of script can be seen on inlaid brass bowls and other objects commissioned by sultan Qaæitbay (see Figure 9.55). After Qaæitbay’s reign (1468–96), the quality of metalwork declined, and the inscriptions become less legible. Some objects of copper, tinned copper and tinned bronze have composite emblems with circular shields divided horizontally into three fields filled with assorted cups, napkins, pen-boxes and powder horns, and inscriptions indicating that they were made for anonymous amirs. Other pieces bear anonymous benedictory inscriptions that are almost illegible. The popularity and profusion of inscribed Mamluk metalwares meant that the style was revived for a generation or so at the turn of the twentieth century to meet the demand of European tourists. Like the models, these ‘Mamluk-revival’ wares are usually decorated with Arabic inscriptions containing aphorisms and polite phrases, but some pieces have engraved or gold-inlaid inscriptions in Hebrew. Late Mamluk metalwares were also popular in other contemporary circles, both Christian and Muslim. Many Mamluk metalwares from the end of the fourteenth and the fifteenth century have European coats of arms or shields left empty to receive such arms on arrival at their European destinations. One curious inlaid bowl in the Hermitage (Petsopoulos 1982, fig. 14) bears the name and titles of the Ottoman sultan Murad II (r. 1421–51 with interruption). It was intended for the Ottoman court, but it is unclear whether it was made by Mamluk craftsmen working in a

— 125 —

— ISLAMIC INSCRIPTIONS —

9.55 Inlaid brass bowl made for the Mamluk sultan Qaæitbay (r. 1468–96) Diameter of rim 32 cm. New York, The Metropolitan Museum of Art, Edward C. Moore Collection, Bequest of Edward C. Moore, 1891 (91.1.565).

Mamluk or an Ottoman city. The bowl stands out as an exception to the vast majority of Ottoman metalwares, which are anepigraphic, as the Ottomans preferred metalwares with plain surfaces or all-over decoration encrusted with jewels. Although many Ottoman metalworkers are recorded in court registers, only one is known from signed works: the goldsmith jeweller Mehmed b. ÆImad of Bosnia. He signed three objects made for Murad III and preserved in the treasury of Topkapı Palace: a binding for the sultan’s poems dated 996/1587–8, a box for the Prophet’s mantle dated 1001/ 1592–3, and a lock and key for the KaÆba dated 1002/1593–4. Inscriptions play a much more important role on metalwares made in Iran, and dates and signatures on some twenty-five metalwares provide a framework for localising a Timurid style. Eleven objects dated between 861/1456 and 910/1505 outline the heyday of production. Fourteen pieces are signed. The most common signature begins

with Æamal (‘work of’), but sometimes a more specialised terminology is used, indicating that the artisan wrote, decorated and inlaid the object (kåtibuhu wa nåqishuhu wa qåriÆuhu). None of the pieces mentions a place of production, but eight of the artisans’ names include nisbas, all relating to places in Khurasan such as Baharjan, Birjan and Quhistan. These inlaid wares were expensive and obviously made for the finest households. Most pieces are anonymous, but a few bear princely names. One jug in Berlin (Komaroff 1992a, no. 3) was made in 861/1456–7 for an unidentified Timurid prince, and two others dated 900/1495–6 and ShaÆban 903/April 1498 (ibid., no. 12) are inscribed with the name of the Timurid sultan Husayn Bayqara. Tinned wares had a wider clientele: a bowl in the Hermitage, for example (ibid., no. 29), was made for Imam Quli Kayani, perhaps a member of the Kayani clan that emigrated from Sistan to Herat during the reign of Ulugh Beg.

— 126 —

— M E TA LWA R E S — The dated and signed pieces can be used to define a distinctive style of Timurid metalware in which the figural style of earlier times is replaced by floral and vegetal motifs and inscriptions. Instead of the good wishes and anonymous titulature in Arabic characteristic of earlier metalwares, the inscriptions contain Persian verses drawn from a wide repertory of classical poets, such as Daqiqi, Firdawsi, SaÆdi and Hafiz, and contemporary Khurasani poets, such as Qasim al-Anwar, Jami and Salihi. The text usually refers to the vessel on which it is inscribed. Inlaid jugs, for example, were inscribed with verses about a mashraba (‘jug’), suggesting their use as wine vessels. The texts are usually well written and accurate and are some of the earliest records of contemporary writings. The inscriptions are written in a wide variety of cursive scripts, including nastaÆlÈq, typical of contemporary calligraphy, and point to the increasingly important role of the calligrapher in contemporary epigraphy. Signatures also help us to distinguish some sets within the problematic group of contemporary metalwares known as Veneto-Saracenic. Made of brass (or bronze), engraved with arabesques and inlaid with silver and a black organic compound, these objects were traditionally ascribed to Muslim craftsmen working in Venice in the fifteenth and early sixteenth centuries, but it is unlikely that foreign craftsmen would have been tolerated within the Venetian guild system. Ten pieces are signed by the master Mahmud al-Kurdi, including a covered bowl in the Courtauld Institute with a bilingual signature around the upper rim: the first inscription in Arabic script says ‘the work of the master Mahmud al-Kurdi, who hopes for forgiveness from his Lord’, while the second gives a transliteration in Latin script AMALEIMALENMAMUD. Mahmud al-Kurdi’s work contains features common to the Mamluk and Timurid styles, and he may have worked for the Aq Qoyunlu ruler Uzun Hasan (r. 1453–78), who had the support of the semi-autonomous Kurdish principalities and maintained links with Venice from 1463 to 1475. Seven other objects can be assigned to another master, who signed himself Zayn al-Din, Zayn al-Din ÆUmar and Zayn al-Din b. al-muÆallim (‘the teacher’). One

of them, a round-bottomed box and cover in the Bargello, belonged to the collection of Ferdinand I de Medici from 1589, so the piece must date from before that time. Zayn al-Din may have been among the metalworkers removed from Tabriz to Ottoman Turkey in 1514 when Selim I conquered the Aq Qoyunlu capital in north-west Iran. Inscriptions help to distinguish metalwares made in eastern Iran in the early sixteenth century from those of similar shape and technique made in the fifteenth century. Persian verses give way to Arabic prayers invoking ÆAli and the twelve imams, as on an inkwell in the V&A dated RabiÆ II 919/May 1513 (Melikian-Chirvani 1982b, no. 118). First inscribed on coins struck in the name of the Safavid shah IsmaÆil (r. 1501–24), this prayer to ÆAli became popular on sixteenth- and seventeenth-century metalwares, and Komaroff (1979–80) has suggested that it may be a way of distinguishing Safavid pieces from their Timurid prototypes. Decoration continues to evolve towards a dense, overall patterning in which the epigraphic cartouches are barely discernible, as on a bowl dated 916/1510–11 in the V&A (Melikian-Chirvani 1982b, no. 117). The verses by Hilali written in nastaÆlÈq script are signed by Sultan Muhammad with the type of signature used on calligraphic exercises, mashaqah (‘exercise’), identifying the writer as one of the calligraphers who had been attached to the Timurid court. Lozenge-shaped cartouches with poetic inscriptions in nastaÆlÈq continue to characterise metalwares made under the Safavids, but the inscriptions stand out more clearly as they are typically written on a scrolling leaf ground. Some poems were apparently composed for the occasion, as on a trio of cylindrical dome-covered inkwells signed by Mirak Husayn Yazdi, two in the V&A (Melikian-Chirvani 1982b, nos. 119–120) and one in the Benaki Museum, Athens (Survey, plate 1387A). Shaped like the tomb tower designed for Shaykh Safi al-Din at Ardabil, they are inscribed with Persian verses exhorting the sultan to open the inkwell and conquer the black-hearted enemy. Based on the zealous tenor of the distichs, Melikian-Chirvani suggested that these qi†Æas were composed during the reign of Shah IsmaÆil

— 127 —

— ISLAMIC INSCRIPTIONS — when the Safavid state was struggling aginst the Uzbeks in the east and the Ottomans in the west. Other objects bear poems related to their function. The typical Safavid pillar-shaped torchstand, for example, is often inscribed with verses about the moth and the candle. Dated examples show that the form became popular in the second half of the sixteenth century. The earliest to survive, in the shrine at Mashhad, was made by the founder Daæud in Lahore on 1 Jumada II 946/14 October 1539. Melikian-Chirvani (1982b: 263) suggested that he was an Iranian working for Humayun, who may have given the torch-stand to the shrine when he passed through Mashhad en route to Qazvin in 1544. Another example in the Iraq Museum, Baghdad (Melikian-Chirvani 1982b, fig. 65), was endowed to the shrine at Samarraæ by an otherwise unknown poet, Shams al-Din Katibi Kashi, in 969/1561–2. Other important dated torch-stands include one in the Metropolitan dated 986/1578–9, another in the Hermitage dated the following year (Ivanov 1960), and a third in the Musée des Arts Décoratifs, Paris, dated 996/1587–8. In addition to the Persian verses and date, the inscriptions sometimes give the name of the owner. It is usually written at the end, introduced by the word ßå˙ibih (‘owned by’). On many pieces, the cartouche with the owner’s name is blank, showing that such inscriptions were added after the object was purchased. A range of people is named in these inscriptions. For example, a bowl dated Rajab 945/November–December 1538 in the V&A (Melikian-Chirvani 1982b, no. 124) was made for a ShiÆite dervish, Mardanshah b. Ghawathshah. The Hermitage torch-stand dated 987/1579–80 was owned by Hajji Chalabi. He may be the same Hajji Husayn b. Hajji Hasan Chalabi who ordered a wine bowl which is in the V&A (Melikian-Chirvani 1982b, no. 133), and who may be identified with the Safavid envoy Husayn Chalabi. Several engraved bowls and candlesticks also bear inscriptions in Armenian naming the owner (Melikian-Chirvani 1982b, annex II). These people were probably prominent members of the Armenian community established by Shah ÆAbbas in New Julfa, south of the capital Isfahan.

These owners’ names are often the only clues to provenance, for most pieces from the late sixteenth and seventeenth centuries do not contain artists’ signatures. One exception is a group of cut steel begging bowls signed by Hajji ÆAbbas. One in the Nuhad al-Said collection (Allan 1982a, no. 26) has thuluth inscriptions in the cartouches, including the date 1015/1606–7 and the signature of the artisan, Hajji ÆAbbas, who is identified as the son of Aqa Rahim the armourer. Some begging bowls signed by Hajji ÆAbbas, however, are eighteenth- or nineteenth-century copies. One in the Hermitage, for example (Masterpieces 1990, no. 118), has inscriptions in nastaÆlÈq and the date 1207/1792–3. Another type of inscribed metalware produced under the Safavids comprises pierced steel plaques and medallions. The inscriptions often contain pious verses and phrases (for example, Atıl et al. 1985, nos 28 and 30). Tinned copper and brass wares in the Safavid style were also produced in the Indian subcontinent in the seventeenth and eighteenth centuries, and inscriptions help to differentiate the two groups. The Indian pieces are also decorated with Persian verses, but the texts sometimes contain variant readings, words are sometimes split between cartouches, and the relationship between text and object is missing. Kashmir is one possible site of production, as a wine bowl dated 1026/ 1617 in the V&A (Scarce and Elwell-Sutton 1971; Digby 1974; Melikian-Chirvani 1982b, no. 165) was made for Khwaja Muhammad Bajiv Kashmiri. Lahore is another possible site Historical inscriptions are occasionally found on Mughal weapons, as on a knife with a blade of forged steel and meteorite iron in the Freer (Atıl et al. 1985, no. 36). A poem inscribed on the side of the handle recounts how Jahangir ordered two swords, this knife and a dagger made from a meteorite that descended like lightning in 1030/1620–1. In general, however, most precious metalwares made in India under the Mughals did not include inscriptions as a major part of the decoration, for the Mughals, like their counterparts the Ottomans, preferred objects of precious metal encrusted with jewels.

— 128 —

— CHAPTER 10 —

Woodwork

I

n many areas of the Islamic lands, timber was scarce, and wood was not used for frame construction but was reserved for decorative building elements such as ceilings, lintels, doors and screens where its inherent structural advantages would be put to best use. Domestic wooden furniture was limited to small low tables and chests, as cushions and rugs replaced chairs and benches, and the finest wood was saved for furnishings in mosques and shrines, including mihrabs, minbars, maqßËras (screened enclosures for the ruler), bookcases, Koran-stands and Koran boxes. These expensive wooden objects designed for religious settings were boldly inscribed with Koranic verses and historical texts giving the donor’s name and date, and the dates on them are often the key to studying the chronological evolution of woodworking styles. A dry climate, which counters the inherent perishability of wood, also aids preservation, and many of the finest examples of Islamic woodcarving are preserved in Egypt, some in situ and others in the Museum of Islamic Art in Cairo. Scholars quickly recognised the importance of the Cairo collection, and several catalogues were prepared in the 1930s. Pauty (1931) catalogued carved examples up to the Ayyubid period, and David-Weill (1931) prepared two volumes on the inscribed pieces, the first volume on inscriptions up to the Mamluk period and the second on those of the Mamluk and Ottoman periods. The entries are arranged according to the museum accession number, but the plates are arranged chronologi-

cally so that one can quickly follow the evolution of epigraphic style, which is summarised in a brief introductory essay. The vast store of information presented in these catalogues allowed further studies of the general stylistic evolution of woodwork, as with Lamm’s 1936 study on Fatimid woodwork. French connections with Egypt have meant that the Louvre has also amassed a fine collection of Islamic, particularly Egyptian, woodwork, and the pieces there and their inscriptions are available in the 1988 catalogue by Elise Anglade. Mayer’s 1958 list of woodcarvers and their works is based on the signed pieces known at that time, but many have been added since then. These are the major works for studying the inscriptions on Islamic woodwork, for most other articles treat individual masterpieces or periods. Inscriptions on wood can be carved in relief or incised, and paint was often added to enhance legibility. At the end of the eleventh century, the technique of marquetry, in which small pieces of wood and other precious substances are closely fitted in interlocking patterns, was introduced, and long bands with inscriptions carved in relief were used to frame and enclose marquetry fields on minbars and mihrabs. These fancy objects were expensive gifts, and the inscriptions on them show that they were often made as part of polemic and pointed campaigns. These pieces of inscribed wood are therefore useful for studying the schisms that developed within the Islamic lands during medieval times.

— 129 —

— ISLAMIC INSCRIPTIONS —

Early Period Inscriptions were not used on the earliest pieces of woodwork carved in the Islamic lands, to judge from several dozen carved panels from the Aqsa Mosque in Jerusalem or the many pieces carved with a bevelled or slanted cut associated with ÆIraq. Inscribed woodwork became popular under the Tulunids, governors of Egypt from 868 to 905. The largest and most famous example is the Koranic frieze that ran around the interior arcade just below the roof of the mosque that Ahmad b. Tulun built in Cairo between 876 and 879. The Kufic letters measure some 15–19 cm high but an astounding 1,988 metres long. They are meticulously carved in high relief with a rounded upper surface and sober curves. The stems of the letters rise only a little about the line of writing to create squat letters that are deliberately stretched out horizontally. Other similarly inscribed pieces can be ascribed to Tulunid Egypt on stylistic grounds. One is a rectangular panel, probably teak, from a frieze found at the ÆAyn al-Sira cemetery near Cairo (Arts of Islam 1976, no. 434). Narrow bands with a Koranic inscription in a squat Kufic script border a central field with compartments enclosing lobed arches and rams’ horns. Another piece is a panel of thin teak veneer (ibid., no. 435) with blessings to its owner in a bevelled style. The care with which these friezes and similar pieces have been carved is evident when they are compared to an unusual group of contemporary Egyptian panels inscribed with the titles to houses and other property (David-Weill 1931, plates IV–VI; Anglade 1988, no. 20). These property deeds are carved in much flatter relief, and the letters are often crowded. They typically contain several lines of Kufic text, with the invocation and blessings to God followed by a description of the house with shops and dependencies and the owner’s name and profession. Dated examples range between 258/872 and 318/930, and these wooden property deeds are important evidence for domestic life in Egypt in the late ninth and early tenth centuries. They are apparently an Egyptian speciality, although similar examples carved on limestone are known from Jerusalem.

From the mid-tenth century in Egypt, new types of wooden fittings for tombs and shrines, such as small portable mihrabs and doors, were carved in a new style (David-Weill 1931, plates IX–X). Many of these pieces were clearly intended for ShiÆite buildings, for the mihrabs contain the names of the twelve imams and the door belonged to a tomb for a member of the ÆAlid Tabatabaæi family who died in 348/949. In style, the inscriptions show the embryonic development towards floriated script: the letters are taller and more elongated, the ends thicken and divide into bilobes, and a few tails curve like swans’ necks. The foliated style turns into full-blown floriated script on a pair of doors (David-Weill 1931, plate XI) added to the Azhar Mosque by the caliph al-Hakim in 400/1010. From the end of the eleventh century to the beginning of the thirteenth, a splendid group of large wooden furnishings were endowed to mosques and shrines in Egypt and Syria. They constitute some of the finest examples of the marquetry technique found in the Islamic lands, and the many inscriptions on them not only provide fixed dates for these wooden pieces but also help us to understand why these pieces were commissioned. The earliest extant example is the minbar that the Fatimid vizier Badr al-Jamali ordered for the shrine of Husayn at Ascalon in 484/1091–2 and later transferred to Hebron (RCEA 2790–1). Other fine pieces in Egypt include two portable mihrabs and a cenotaph added to the shrines of Sayyida Nafisa and Sayyida Ruqayya by the caliph al-Hafiz (RCEA 3092 and 3188) and a minbar ordered by the vizier al-Salih TalaæiÆ for the ÆAmri mosque at Qus in 550/1155–6 (RCEA 3189). Even finer pieces were made for the Zangid ruler Nur al-Din in Syria, including the mihrab which he added to the Maqam Ibrahim in the citadel of Aleppo in 563/1167–8 (MCIA Alep, plates XLVI–VII) and the minbar which he commissioned the following year for the Aqsa Mosque in Jerusalem (RCEA 3281). Perhaps the finest example of this school of woodcarving (see Figure 10.56), it was destroyed by arson in 1969. Rich wooden furnishings continued to be commissioned under the Ayyubids, as attested by the wealth of woodwork in the mausoleum

— 130 —

— WOODWORK —

10.56 Detail of the wooden minbar ordered by Nur al-Din Zangi in 564/1168-9 for the Aqsa Mosque in Jerusalem (after van Berchem, MCIA Jerusalem, plate XXIX).

— 131 —

— ISLAMIC INSCRIPTIONS — that the Ayyubid sultan al-Kamil built in 608/ 1211 over the grave of the Imam al-ShafiÆi in the southern cemetery of Cairo (Wiet 1933). Lintels and doors were inscribed in naskh with the name of the patron, and the unique system of wooden beams and brackets from which lamps were suspended has Kufic inscriptions on a vegetal ground. The fine Kufic inscription continues in a band around the top of the walls with Koran 36 (see Figure 10.57). The tomb for the Imam al-ShafiÆi also contains several large wooden cenotaphs. The finest is the one commissioned by Saladin in 574/1178 (RCEA 3331–2). Another one was ordered in 608/1211 by Saladin’s nephew al-Kamil in honour of his mother (RCEA 3680). Large cenotaphs were made to mark many other graves in the region, as, for example, the one made for the amir IsmaÆil b. ThaÆlab in 613/1216 (RCEA 3788). These wooden furnishings were costly gifts. They were made of small pieces of wood intricately inlaid in strapwork patterns, and Nur al-Din’s

minbar adds ivory and mother-of-pearl to the repertory of materials used. They were also important symbolic gestures. Nur al-Din’s minbar was ordered in anticipation of his taking Jerusalem from the Crusaders and was moved there by his nephew Saladin when he regained the city two decades later. The inscriptions underscore the polemic use of these expensive wooden furnishings. The long historical texts include the names and titles of the donors, which are composed with lengthy honorifics, epithets and eulogies suited to the occasion. The foundation inscription on Nur al-Din’s minbar, for example, asks God to glorify Nur al-Din’s victories, extend his power, heighten his signs, spread his emblems and standards to the two ends of the world, glorify the friends of his empire, and disgrace those who disregard his grace. The inscription on the cenotaph ordered by Saladin for al-ShafiÆi’s tomb lists the imam’s genealogy twelve generations back to the Prophet in imitation of the lengthy genealogies given on ShiÆite pieces.

10.57 Detail with the basmala and Koran 36:1–3 from the wooden frieze around the top of the walls in the tomb built for the Imam al-ShafiÆi in Cairo in 608/1211.

— 132 —

— WOODWORK — The non-historical texts on these wooden furnishings were also chosen to suit the occasion. The text on Badr al-Jamali’s minbar for the shrine of Husayn is a careful composed message that ends with a hadith in which the Prophet declares his two legacies to be the Koran and his family. One of the earliest examples of an inscribed hadith, it was clearly a vindication of ShiÆite claims to legitimacy and cited consciously as a justification for refurbishing the shrine (Wiet 1924). Koranic verses were similarly polemic. The text around al-ShafiÆi’s cenotaph begins with Koran 53:39–41, saying that man achieves nothing without striving but that should he do so, he will achieve a complete reward. The citation was surely chosen to refer to Saladin’s successful campaigns in Egypt. A more extensive programme of Koranic verses occurs on a splendid wooden cenotaph from the Mosque of al-Husayn in Cairo. The inscriptions on it, all Koranic, account for more than half the surface decoration and present a coherent programme of ShiÆite esoteric dogma (Williams 1987). The inscriptions on these wooden furnishings were crafted to enhance legibility. Many of the texts are carved with thick rounded letters that rise above the flat background plane, which is decorated with a thin undulating scroll. Slightly different styles of script were used to differentiate different types of text. On the mihrab added to the shrine for Sayyida Ruqayya, for example, the Koranic frame bands are done in Kufic on a scroll ground, while the two-line historical text is in floriated Kufic. Cursive, introduced in the mid-twelfth century, made historical texts easier to read. The best example is the foundation text on Nur alDin’s minbar, which is not only written in thick cursive letters that project above the delicate arabesque scroll in the background, but also painted in white to enhance visiblity from afar. Colour was also used to pick out letters, as on a Koranic frieze from the cenotaph for al-Kamil’s mother in which the letters were painted red. Different styles, techniques and texts were deliberately juxtaposed. An impost from the tomb of Sayidda Nafisa, for example, contains a fiveline historical text in typical Ayyubid cursive, a

frame band with Koranic verses in cursive set against a floral ground, and two lines from another Koranic verse in large square Kufic inlaid with thin bands of ivory and set on a fine ground of mashrabiyya, the lattice-like screen of turned wood that became typical of later woodwork in the Islamic lands. In the tomb for the Imam al-ShafiÆi, different scripts (cursive and Kufic on a floral ground) contained different types of text (historical and Koranic). The numerous signatures on these wooden furnishings also show us the high status of woodcarvers and the pre-eminence of an Aleppan family in this craft. The mihrab ordered by Nur al-Din for the Maqam Ibrahim is signed by MaÆali b. Salam (RCEA 3276), and the minbar which he ordered for the Aqsa mosque is signed by four craftsmen from Aleppo (RCEA 3282). One of them, Salman b. MaÆali, was probably the son of the craftsman who signed the mihrab, and the three others, two of whom were brothers and sons of a certain Yahya, were also from Aleppo. The cenotaph ordered by Saladin for the tomb of Imam al-ShafiÆi is signed by the woodcarver (al-najjår) ÆUbayd known as Ibn MaÆali, clearly another son of the craftsman who had made the other mihrab in Aleppo a decade before. The Maghrib had a particularly strong tradition of endowing fine wooden furnishings to mosques, in part because of abundant timber available in forests there. These fine pieces, especially minbars, were signs of sovereignty and were continually repaired and refurbished, and the inscriptions on these pieces are important evidence in tracing the changing sectarian schisms and political struggles in the region. The earliest minbars are known only through texts, such as the one ordered by the Rustamid ruler Idris I in 174/790 for the mosque at Tlemcen (RCEA 54) and reinscribed by his son Idris II in 199/814 (RCEA 97). The oldest extant example, the minbar added to the congregational mosque of Qayrawan, Tunisia in 248/862–3, was imported from ÆIraq and contains no inscriptions, but one from a century later, the minbar for the Andalusiyyin Mosque in Fez (Dodds 1992, no. 41), shows how these minbars became potent symbols of sovereignty.

— 133 —

— ISLAMIC INSCRIPTIONS — Terrasse (1942) and Bloom (in Dodds 1992: 249–51) untangled the complicated history of this minbar, which charts the changing fate of Fez, the key city in the struggles between the ShiÆite Fatimids of Tunisia and the Sunni Umayyads of Spain. Two rectangular cedar panels with foliated Kufic are inscribed with the date 369/980 and Koran 24:36, an allusion to the minbar’s function as the site not only of the weekly sermon (khu†ba) but also of the second call to prayer (iqåma). These panels were part of the original minbar ordered by the local ruler, Buluggin b. Ziri, who served as client of the Fatimids at this time. Buluggin’s name, however, must have been removed when the backrest of the minbar was redone, for panels from the backrest are inscribed with the name of al-Mansur, chamberlain to the Spanish Umayyad caliph Hisham, and the date Jumada II 3[7]5/9[8]5. The new historical text, carved in a similar style to that on the rectangular panels, was added just after the Umayyads retook

the city. The Umayyads were content to leave the rectangular panels with the earlier date and Koranic verses, but the original historical text on the backrest, which probably contained not only Buluggin’s name and titles but also those of his Fatimid suzerain as well as ShiÆite formulas and benedictions, was anathema. The original backrest may have been sent back to Cordova as a trophy, as was done six years earlier with the backrest from another minbar in Asila. Prominently displaying the ruler’s name on fine wooden furnishings in the major congregational mosque continued to be an important symbol of sovereignty in this disputed region in the eleventh century. This is clear from the mosque at Qayrawan, where the Zirid ruler al-MuÆizz b. Badis (r. 1016–62) erected a splendid wooden maqßËra (see Figure 10.58) to the right of the earlier minbar, which may have been restored at this time. A bold inscription below the cresting at the top of the maqßËra names the patron and

10.58 Detail of the wooden maqßËra donated to the Great Mosque at Qayrawan in 413–14/1022–3, with the name and genealogy of the Zirid ruler al-MuÆizz b. Badis in Kufic with interlaced stems.

— 134 —

— WOODWORK — says that he paid for the work from the alms tax (zakawåt) since God will not neglect to reward those who do such good things. Work was carried out under the vizier Abuæl-Qasim. The band closes with a long Koranic text (62:1–5), saying that everything in the heavens and the earth declares the praises of God, the sovereign, who bestows goodness on whomever he wishes. The work can be dated to 413–14/ 1022–23, for the vizier mentioned in the inscription was only in office for one year (RCEA 2557; date corrected in vol. 16, p. 271). No mention is made of the Fatimids, and the inscription does not contain the titles bestowed on al-MuÆizz by the Fatimid caliph, who sent an embassy in 414/1022–3 rewarding al-MuÆizz with sumptuous presents and the high-falutin title, Sharaf al-Dawla wa ÆAduduha (‘Nobility of the State and its Right Arm’). Rather, the maqßËra was probably part of al-MuÆizz’s efforts to distance himself from the ShiÆite Fatimids and placate the population who were staunch Maliki Sunnis. The style of the carving underscores the message on the Qayrawan maqßËra. To enhance legibility, the woodcarver beaded the edges of the Kufic letters and rounded their upper surfaces so that the bold inscription is set off from the richly floriated ground and the name of the patron is visible from afar. Unlike contemporary inscriptions by alMuÆizz’s overlords, the Fatimids in Egypt, which were done in floriated Kufic, this inscription is done in Kufic with knotted stems. As Flury (1920a, 1920b) noted, it is a rare and early example of this style of script which is usually associated with the eastern Islamic lands. In both text and style, then, the inscription separates al-MuÆizz from the Fatimids, a trend that would culminate two decades later when al-MuÆizz’s replaced the name of the Fatimids with that of the ÆAbbasids in the khu†ba. The tradition of fine inscribed minbars continued in the Maghrib under the Almoravids, strict and conservative followers of the Maliki sect who governed the Maghrib from 1056 to 1147. Yusuf b. Tashufin, the real founder of the empire (r. 1061–1106), donated one to the mosque at Nedroma sometime before 479/1086 (RCEA 2908). Another in the mosque of Algiers, signed

by a certain Muhammad, is dated 1 Rajab 490/14 June 1097 (RCEA 2863). The finest is the one ordered from Cordova by Yusuf’s son ÆAli (r. 1106–42). Perhaps modelled on one in Cordova itself made for the Umayyads, it was designated for the mosque which ÆAli built in Marrakesh c. 1120 but was later transferred by the Almohads to the Kutubiyya mosque there (Dodds 1992, no. 115). It is decorated with strapwork bands executed in marquetry of precious woods and ivory, and is comparable to the contemporary one made in 538/1144 for the Qarawiyyin Mosque in Fez (Terrasse 1968). A fifth Almoravid minbar was made for the congregational mosque at Tlemcen in Ramadan 533/May 1139 (RCEA 3090). The fine quality of the five surviving examples and their elaborate inscriptions with the patron’s name and date show that these minbars continued to be important signs of sovereignty. One type of inscribed wooden furnishing is distinct to congregational mosques in the Maghrib: the Æanaza, an auxiliary wooden mihrab erected in the centre of the court façade of the prayer hall. Set in the doorway of the axial nave leading to the mihrab, it was used by worshippers in the court. The front side facing the court was traditionally the most elaborately decorated, and the earliest ones were inscribed with the date. According to the Raw∂ al-Qirtås, for example, one in the Qarawiyyin Mosque in Fez bore an inscription saying that this Æanaza was made (ßuniÆat) in ShaÆban 524/July 1130 (RCEA 3031), thus under the Almoravids, and the Æanaza in the Andalusiyyin Mosque there bears an inscription at the top saying that it was finished in Muharram 606/July 1209 (Terrasse 1942), thus under the Almohads. The outer face of the Æanaza, the one exposed to rain and the elements, has often deteriorated so that the Æanaza had to be replaced. The one in the Qarawiyyin Mosque, for example, was replaced in 1289 by the Marinids, and although preserved, the outer face of the replacement has also deteriorated. Inscribed wooden furnishings for mosques were also made in the eastern Islamic lands, but they do not seem to have formed such a concerted programme, and most of the examples that

— 135 —

— ISLAMIC INSCRIPTIONS — survive were made for small villages or endowed by local patrons. The Maydan Mosque in Abiyana, a village west of Natanz in central Iran, for example, preserves a wooden minbar dated 466/ 1073 and a wooden mihrab dated 497/1103 (Ettinghausen 1952). Four columns from the congregational mosque of Khiva in the Khwarazm oasis are inscribed in simple Kufic with the name of the patron, the jurisprudent (al-faqÈh) Abuæl-Fadl al-Muhallabi, who states that he paid for them himself, and a wooden mihrab from the village mosque at Iskodar in the upper Zerafshan Valley is similarly inscribed with a hadith about guarding the frontiers (Blair 1992a, nos 26 and 27). A minbar ordered in 548/1153 for the ÆAmadiya mosque in Mosul (Arts of Islam 1976, no. 452) is inscribed along the handrail with the name of the patron, the amir Hisham al-Din, and along the rail parallel to it with the name of the three Georgian craftsmen, ÆAli b. al-Nahi, Ibrahim b. JamiÆ and ÆAli b. Salama.

Wood was more important in the eastern Islamic lands for furnishing tombs and shrines. Presenting a fancy cenotaph or other fittings to the tomb of a saint or local hero was a way of annexing his popularity. Five wooden plaques (Blair 1992a, no. 10) bear blessings to the Twelve Imams and a short text recording the restoration of a shrine (al-buqÆa) by the Buyid amir ÆAdud al-Dawla in 363/973–4. They may have been part of his restoration of one of the ShiÆite shrines near Kufa. A bier in the Israel Museum is inscribed in simple Kufic with eschatological texts about the deceased’s testimony and faith in the Twelve Imams. This rectangular piece (see Figure 10.59) from the centre of one of the long sides, for example, contains the end of the text about the deceased’s testimony that death and the grave are real; that the interrogation by Munkir and Nakir in the grave ‘is real, that the resurrection, the arising, the assembling, the accounting, the weighing, the way, Paradise and hellfire are real.

10.59 Wooden panel from a bier in the Israel Museum, made in Iran in the tenth century, with a Kufic text about the deceased’s testimony.

— 136 —

— WOODWORK — This is my belief; in it I died and in it will I be resurrected’ (˙aqq waæl-baÆth waæl-nusËr / waæl˙ashr waæl-˙isåb waæl-m[È]zån waæ-[ßi] / rå† waæl-janna waæl-når / ˙aqq hådhå iÆtiqådÈ Æalayhi muttu wa Æalayhi abÆath). The bier is one of the problematic Buyid objects reportedly found at Rayy but whose authenticity has been doubted. Shepherd had the bier tested in several ways, and it seems plausible. It is made of Pinus brutus, a tree no longer native to Iran, and radiocarbon dating produced a result in the tenth century (Shepherd 1974, Appendix A-6). This makes it an important piece for the history of Iranian epigraphy. Glidden (ibid.) noticed, for example, that it used an unnecessary support for the hamza, an orthographic feature which he often found in early Arabic texts written by Persians. One would want to compare this text with that on a grave cover in the Royal Ontario Museum, Toronto, with a similar invocation to the Twelve Imams in foliated Kufic script. Several cenotaphs were also endowed to the tomb of the seventh imam, Musa al-Kazim, at Baghdad. One large example of mulberry wood (RCEA 3976; Arts of Islam 1976, no. 453) was given by the ÆAbbasid caliph al-Mustansir in 624/1227. The lid is inscribed with Koran 33:33, the verse vindicating the rights of the ShiÆites as ‘people of the house’ (ahl al-båyt), followed by the patron’s name and date. The main decoration, a large inscription around the sides in Kufic with knotted stems, gives the imam’s name and genealogy back to the Prophet. These tombs and shrines were not always dedicated to ShiÆite saints. The finest ensemble surviving from the period is a set of folding doors from the tomb of Mahmud of Ghazna in Afghanistan (Flury 1918; Survey, plate 1462). Transferred to the Fort of Agra in 1842, they are now in the Agra Museum. The inscription (RCEA 2377) invokes God’s blessing on Mahmud, who died on 22 RabiÆ II 421/29 April 1030, but the doors are not necessarily contemporary with his death, and the style of the script, Kufic on a scroll ground, suggests that the doors might have been added in the twelfth century during a revival of interest in his cult.

Wood was also plentiful in Anatolia, and with the opening-up of the region to Islam following the battle of Manzikert in 463/1071, wood was commonly used to construct and furnish the many new mosques and shrines needed there. Doors, shutters, Koran-stands, boxes and other pieces were often inscribed with the names of the patron and woodcarver. As elsewhere, minbars were particularly elaborate, and many are preserved either in their original mosques or in the Ethnography Museum, Ankara. Oral (1962) catalogued a series of twenty-four examples dating from the early twelfth century to the sixteenth. The typical Anatolian example was heavily inscribed, with inscription bands along the rails bordering the stairs, large panels on the doors and frame at the bottom, and smaller panels with signatures of craftsmen on the sides. Texts include popular Koranic verses, hadith and long historical inscriptions with the names and titles of the patron. One of the earliest is the marquetry minbar ordered by MasÆud I for the ÆAlaæ al-Din Mosque at Konya in Rajab 550/September 1155 and carved by the master Makki, the pilgrim from Akhlat (RCEA 3200–1). The tradition continued under the amirates. The mosque that the E∞refoÌlu amir Sulayman built at Bey∞ehir at the end of the thirteenth century, for example, contains fine carved wooden doors dated 699/1299–1300 (RCEA 5082), fortyeight wooden columns with muqarnas capitals, and a fine minbar (see Figure 10.60). A line of cursive directly over the doorway gives the foundation inscription in the name of the amir Sulayman (RCEA 5083). The names of God and the four orthodox caliphs (Muhammad, Abu Bakr, ÆUmar and ÆUthman) are written in blond wood in a panel of square Kufic set on a screen at the top, and the whole is framed by a cursive band with the Throne Verse (Koran 2:255). Folding wooden stands for the Koran (Turkish rahle) were also popular. They were inscribed with Koranic texts, hadith or foundation inscriptions naming the donor. One of the most exotic is a lacquered wooden lectern given to the shrine of Jalal al-Din Rumi at Konya in 678/1279–80 (RCEA 4780; Ettinghausen and Grabar 1987, fig.

— 137 —

— ISLAMIC INSCRIPTIONS —

10.60 Wooden minbar in the Eski Jami at Bey∞ehir ordered by the E∞refo©lu amir Sulayman c. 699/1300 (photo Walter Denny).

— 138 —

— WOODWORK — 400). The interior face where the Koran was set is painted with a double-headed eagle and a lion on a background of arabesques. The exterior bears an elegant cursive inscription with the foundation text naming the donor, Jamal al-Din, who was in the service of the Saljuq vizier Sahib Ata. Wooden cenotaphs were often covered with inscriptions including prayers, Koranic verses, hadith and historical texts naming the deceased, the donor and/or the artisan. Several fine examples in the shrine museum at Konya were made for the mystic and his disciples. Rumi’s cenotaph (RCEA 4681) bears the typical decoration of a long text with his flowery titles, name, genealogy and date of his death (5 Jumada II 672/17 December 1273), and the signature of the artisan, ÆAbd al-Wahid b. Salim, who is identified as the architect (miÆmår). Other examples, such as one from the mausoleum of Sayyid Mahmud Khayrani in Ak∞ehir (Anatolian Civilisations 1983, no. 168), are inscribed with poems by Rumi. There has also been a long tradition of woodworking in the Yemen, although few examples survive from the early period. Ceilings, in particular, were elaborate coffered affairs with rich painted decoration, but to judge from the tenth-century one preserved in the Great Mosque at SanÆa, they were generally anepigraphic. Cenotaphs furnish better examples of inscribed woodwork. The most notable is the one made for al-Sayyida Arwa, consort of the Sulayhid ruler al-Mukarram ÆAli b. Ashgar (r. 1084–91) and his successor. Under her auspices, the capital was moved from SanÆaæ to Dhuæl-Jibla, north of TaÆizz, and the old palace there was converted into the city’s congregational mosque in 481/1088–9. The queen was buried in the north-west corner of the mosque, and her large wooden cenotaph is inscribed with a lower band of richly foliated Kufic using the same style of script that is found on the plaster mihrab in the mosque (Lewcock and Smith 1974, plate 11).

Late Period Wood continued to be the favoured material for fixed and portable furnishings, minbars and Koranstands in the many mosques and tombs built in

Egypt and Syria under the Mamluks (1250–1517). The precedent was clearly the tomb that the Ayyubid sultan al-Kamil had built over the grave of the Imam al-ShafiÆi in 608/1211 (Wiet 1933). It remained an important shrine in Mamluk times, and the magnificent wooden dome was restored in 885/1480 by the Mamluk sultan Qaæitbay, as noted by several small wooden and marble plaques, and again in 1186/ 1772 by the Ottoman ÆAli Bey, later pasha of St Jean of Acre, as commemorated by the long inscription painted in gold around the base of the dome. The earliest and finest woodwork ensemble preserved from Mamluk Egypt is that in the mausoleum of Sultan Qalaæun on the main street of Cairo. Wood was used for the ceilings, windows, doors, screens and cenotaph there (see Figure 10.61), and many wooden plaques and lintels are boldly inscribed with the sultan’s name and titles (MCIA Egypte 1, nos 82ff.; RCEA 4844–53). The centrepiece of the tomb is the sultan’s wooden cenotaph. The historical text (MCIA Egypte 1, no. 87), partly masked by later marble plaques, is inscribed with the Throne verse (Koran 2:255) and Qalaæun’s name and titles. The cenotaph is surrounded by a superb screen of turned wood (mashrabiyya) inscribed with the name of Qalaæun’s son, al-Nasir Muhammad (MCIA Egypte 1, no. 91). It was probably added in 703/1303–4 when al-Nasir rebuilt the minaret of the complex (MCIA Egypte 1, nos 88–90; RCEA 5160–2). Many other funerary complexes built under the Mamluk sultans of the Bahri line (1250–1389) were similarly furnished with fine ensembles of inscribed woodwork. Thanks to David-Weill’s (1931) catalogue of the woodwork in the Cairo Museum, we can follow the stylistic evolution of wooden inscriptions under the Bahri Mamluks. The naskh script is similar to that used under the Ayyubids, and the closely set, thick letters lend an impression of massiveness and solemnity. One of the finest examples is the carved frieze from the congregational mosque built in 739–40/1339–40 on the Darb al-Ahmar by Altinbugha al-Maridani, cup-bearer and son-in-law of al-Nasir Muhammad and later governor of Aleppo. It contains Koranic verses set in cartouches

— 139 —

— ISLAMIC INSCRIPTIONS —

10.61 Wooden screen added around the cenotaph in the tomb of Qalaæun in Cairo in 703/1303–4.

— 140 —

— WOODWORK — separated by roundels and is comparable in grandeur to the Kufic frieze made to decorate the mosque of Ibn Tulun some four and a half centuries earlier. Painted inscriptions were also introduced in the mid-fourteenth century. Another Koranic frieze from al-Maridani’s mosque, for example, has white letters outlined in black against a red ground. Fine woodwork continued to be made for mosques and shrines under the Burji or Circassian Mamluks (1389–1517). Most Mamluk buildings to survive date from that period, and so do most pieces of inscribed woodwork. They include Koranic and other religious texts, foundation inscriptions, and cartouches with the sultan’s name. Most are carved in relief, but painted inscriptions, which are cheaper, became increasingly popular. The mosque complex built by al-Muæayyad (r. 1412–21) presents a good example of the range of inscribed woodwork in a later Mamluk building (MCIA Egypte 1, nos 230–9). It had a lavishly decorated wooden ceiling, a fine wooden minbar inlaid with ivory, a wooden grille over the fountain, and a dikka or lectern for Koran readers. The cenotaphs in the tomb have splendid inscriptions in foliated Kufic, a deliberate revival of an earlier style. Even finer woodwork was used in the tomb complex completed by Qaæitbay in 879/1474 in the northern cemetery (MCIA Egypte 1, nos 295–304). Painted and gilded friezes ring the walls (Blair and Bloom 1994, fig. 118), and wooden plaques reiterate the sultan’s name. Fine wooden furnishings include a Koran-stand dated RabiÆ I 878/ August 1473 and the sultan’s cenotaph within a wooden screen. The script used on these later pieces became increasingly crowded, and on many pieces made for Qaæitbay the letters are so tightly packed that the background is almost obscured. Of the various inscribed wooden furnishings, the most elaborate were minbars. Donating minbars to mosques and shrines continued to be an important act of piety under the Mamluks, and the donor’s name was usually inscribed on the doors at the bottom of the steps and on the back beneath the bulbous dome. The minbar ordered by the first Mamluk sultan Baybars I for the Mosque of the Prophet in Medina has been destroyed, and the earliest Mamluk minbar to survive is the

one ordered by Sultan Lajin for the Mosque of Ibn Tulun on 10 Safar 696/8 December 1296 (RCEA 5019–20). Another contemporary minbar was ordered by the amir Qarasunqur, governor of Aleppo, for the Great Mosque there (RCEA 5292–3) at the same time that he restored the maqßËra in the mosque (RCEA 4860 and repeated as 5294). Sending a minbar and other fancy woodwork to Arabia was a sign of sovereignty over the holy shrines. The Mamluk sultans Khushqadam and Qaæitbay, for example, sent minbars there in 867/ 1462–3 and 879/1474–5. In the later fifteenth century, the quality of these minbars declined, as seen by the one ordered by Qaæitbay for his complex inside the city walls (MCIA Egypte 1, no. 310). Many of these minbars are signed, and the signatures show that woodworking continued to be a high-status craft in the Mamluk period. The minbar ordered by Baybars for Medina was made by Abu Bakr b. Yusuf. Qarasungur’s minbar in Aleppo was made by Muhammad b. ÆAli, a craftsman from Mosul who introduced ivory inlay, a feature that was slowly incorporated into Cairene examples. Teams of workers sometimes executed these large and intricate minbars. The minbar ordered by the amir Kitbugha for the congregational mosque of Hama and completed in mid-ShaÆban 701/15 April 1302 (RCEA 5136) was made by ÆAli b. Makki and ÆAbd Allah Ahmad, inlaid by Abu Bakr b. Muhammad and decorated by ÆAli b. ÆUthman. The inscriptions also show us that these woodcarvers often had long careers. Ahmad b. ÆIsa b. Ahmad al-Damyati, for example, made minbars over a twenty-five-year period in the mid-fifteenth century: he made one, later transferred to the complex of al-Ashraf Barsbay, for the mosque of al-Ghamri when it was restored by 860/1456 and another for the madrasa of Abu Bakr b. Muzhir in 885/1480–1. The names of Mamluk woodcarvers, however, do not indicate a family relationship between craftsmen, as had happened occasionally in earlier times. The abundant timber available in the Maghrib meant that wood continued to be an important decorative element in the later period. The typical building from the Marinid period in the region

— 141 —

— ISLAMIC INSCRIPTIONS —

10.62 Wooden revetment in the ÆAttarin Madrasa, Fez, 725/1325–6.

— 142 —

— WOODWORK — had a tiled dado, carved plaster along the top of the walls, and extensive decoration in carved wood for cornices, lintels, eaves and ceilings. In addition, there were turned wooden screens, grilles and balustrades. Almost all of these wooden pieces contain carved and painted inscriptions, but despite their profusion, the inscriptions play a relatively minor role (see Figure 10.62). The texts are usually set in cartouches and barely stand out from the dense background of floral and vegetal ornament. The scripts are heavily stylised, and the phrases often repeat or read in mirror reverse. The inscriptions seem to have been used more for aesthetic or decorative reasons, to play off different scripts or to fill the ground, than to communicate information. The one exception is the main foundation inscription which often rings the courtyard beneath the awning. It is usually written in a thick naskh script and is distinguished from the repeating cartouches as a band containing a continuous running text. The most spectacular wooden pieces to survive from the later period in the Maghrib are elaborately joined ceilings done in the technique known today as artesonado, in which supplementary laths are interlaced into the rafters supporting the roof to form decorative geometric patterns. These ceilings, which had developed under the Almohads, became extremely popular in the region from the thirteenth century to the fifteenth. Many of them contain small inscriptions, but the texts are repetitive and the words are almost swallowed up by the intricate interlaced patterns. A square ceiling that once covered the Torre de las Damas of the Partal palace at the Alhambra (Dodds 1992, no. 116), for example, bears small Kufic inscriptions in the octagonal zone and cursive inscriptions with the Nasrid motto in the sixteen-sided zone. Similar small inscriptions in different scripts decorate many of the other wooden ceilings still in situ. Doors, shutters and other portable furnishings were often done in marquetry, and here too inscriptions play a relatively minor role. Fine cabinet doors for the Alhambra (ibid., no. 118), for example, are anepigraphic. As in Morocco, wooden beams, consoles and eaves in Spanish buildings were decorated with

cartouches crowded with thick letters set on a densely carved floral ground. One of the finest ensembles to survive decorates the magnificent Cuarto Dorado, the façade of the Comares Palace in the Alhambra (Torres Balbás 1949, fig. 201). These wooden inscriptions resemble those carved in stucco for the Nasrids. Wooden beams were often decorated with carved and painted inscriptions, including Koranic verses. The typical Marinid madrasas in Morocco had been decorated with inscribed wooden lintels, eaves and awnings, and the same style was repeated two centuries later at the Ben Yusuf Madrasa in Marrakesh, the finest example of SaÆdian architecture, the largest madrasa in the Maghrib, and the only surviving example of one built under SaÆdian patronage. The foundation text around the court tells us that it is not, as commonly thought, a restoration of a Marinid madrasa, but a new building ordered by ÆAbd Allah al-Ghalib in 972/1564–5. Good details of these wooden inscriptions and their stucco counterparts are published in Hill and Golvin’s (1976) photographic survey of Islamic architecture in North Africa. In addition to the standard repeating inscriptions in cartouches, patrons in the Maghrib wanted to record their major donations to mosques, and fine wooden furnishings were often inscribed with the patron’s name and the date. The minbar added to the mosque at Taza by the Marinids, for example, is inlaid with ivory and ebony and bears traces of a dedicatory inscription. The Æanaza continued to be inscribed in the centre of the front side, as on the one in the mosque at Fez alJadid, founded by the Marinid sultan Abu Yusuf in 1276. Abu Hafs, son of the Marinid sultan Abu SaÆid, had a large cedar maqßËra placed in front of the mihrab in the Qarawiyyin Mosque in Fez in the early fourteenth century. Measuring nine spans high by thirty spans across by twenty-five spans deep, it was decorated with carvings, presumably including an inscription bearing the patron’s name. Pious believers, however, considered the new maqßËra an embarrassment, and the prince had to have it moved near the doorway. No trace of it survives (Terrasse 1968).

— 143 —

— ISLAMIC INSCRIPTIONS — Wooden inscriptions are important evidence in reconstructing the history of venerated congregational mosques in the region. Several wooden inscriptions, for example, record restorations to the Andalusiyyin Mosque in Fez (Terrasse 1942). The Marinid Abu SaÆid ÆUthman had a plaque with his name inscribed on it set over the door to the library (makhzåna) which he added in RabiÆ II 816/September 1413, and the Sharifan Mawlay IsmaÆil had his name and the date 1093/1692 inscribed in a poem on a new wooden awning which he had erected over the Marinid fountain in the corner of the courtyard. In contrast to Egypt, Syria or the Maghrib, where much of the inscribed woodwork is associated with court patronage, the furniture and fittings for mosques and tombs in Iran and Central Asia reflect a broader range of patrons from a cross-section of society. Extant pieces from this period are scattered in a variety of sites and collections, and little work has been done on the inscriptions beyond Leo Bronstein’s survey of decorative woodwork from the Islamic period in the Survey of Persian Art. Most of the pieces are furnishings, such as minbars, cenotaphs and doors, but a unique sandalwood box inlaid with polychrome marquetry, gold fittings and silk lining is preserved in the treasury at Topkapı Palace (Lentz and Lowry 1989, no. 49; Grube 1989). Its cover is inscribed with cartouches bearing the name of the Timurid prince Ulugh Beg and his titles gËrgån (‘son-in-law’) and sul†ån. The box is exquisitely carved with arabesques and a writhing dragon set between the cartouches, and the full use of the prince’s name and titles on the cover shows that such a fine wooden object was a prized possession of the imperial household, perhaps made, like the famous jade ewer in Lisbon (see Figure 13.83), to mark his assumption of the sultanate in 850/1447. Only a handful of wooden minbars are preserved from the later period in Iran, and they reflect a lower rank of patronage than those from other regions. A minister and merchant, Jamal al-Din Husayn b. ÆUmar b. ÆAfif, for example, donated a new minbar to the mosque at Naæin in central Iran in Rajab 711/November–December 1311 (RCEA

5300; Survey, plate 1464). It was made by the master Mahmud Shah b. Muhammad, the designer (naqqåsh) from Kirman. A comparable minbar endowed to the Friday Mosque at Isfahan in the early fourteenth century reflects contemporary innovations in calligraphy and epigraphy, such as inscriptions in square Kufic beside naturalistic leaves in high relief (Noci 1982). A rare Timurid example is the small but fine minbar made for the congregational mosque in the village of Khwaf Rud near Khargird in north-eastern Iran (O’Kane 1987, no. 28; Golombek and Wilber 1988, no. 86). The panel at the back (now in the Shrine museum in Mashhad) says that the minbar and an elegant kursÈ were donated to the mosque in 908/1502–3 by Majd al-Din Muhammad alKhwafi. He can be identified as a vizier of the Timurid sultan Husayn Bayqara who had been removed from office in the 1490s, retired to his home town, and presented this minbar as a memento to the congregational mosque there. By contrast, a large number of wooden cenotaphs are preserved in the eastern Islamic lands, and the inscriptions show that they marked the graves of a wide range of people. A group from the area around Sultaniyya, the capital of the Ilkhanids in north-western Iran, date from the late thirteenth and early fourteenth centuries (Curatola 1987). Many were made for the tombs of local imams or shaykhs, but an extraordinary example is the cenotaph of Esther from the so-called mausoleum of Esther and Mordecai at Hamadan. Although typically Ilkhanid in style, it is inscribed in Hebrew. Large cenotaphs were made for the graves of noted Sufis and their descendants. One in the Bukhara Museum, for example, was made for Yahya b. Ahmad, a grandson of the noted mystic Sayf al-Din Bakharzi who died in 1336. Smaller cenotaphs with incised inscriptions were made for local patrons: two examples, one dated 754/ 1353–4, are preserved in the isolated shrine of Tashar near Radkan in eastern Mazandaran. More examples survive from the fifteenth century, particularly from the Caspian region. Bivar and Yarshater catalogued several still in situ in remote villages in eastern Mazandaran (CII, 1978), and others have been looted and are now in museums

— 144 —

— WOODWORK — abroad. A large one in the Rhode Island School of Design (Survey, plate 1472; Lentz and Lowry 1989, no. 101), for example, was ordered for the shrine of Abuæl-Qasim in Ramadan 977/February 1473 by Gushtaham, a ruler of the local Baduspanid line. Finely carved wooden doors were also considered suitable furnishings for shrines and tombs. One pair was donated in the early fourteenth century to the mosque in the shrine of the noted mystic Bayazid at Bastam (Survey, plate 1463). Two pairs of doors, one at the main portal (Survey, plate 1467; Blair and Bloom 1994, fig. 72) and another at the entrance to the mausoleum, decorate the shrine that Timur built for the mystic Ahmad Yasavi at Turkestan at the end of the fourteenth century. A set of marquetry doors inlaid with ivory was made for Timur’s tomb, the Gur-i Mir, at the beginning of the fifteenth century (Survey, plate 1470; Lentz and Lowry 1989, fig. 15), and a contemporary set was made for the entrance to the shrine of Qutham b. ÆAbbas at the Shah-i Zinda in the same city in 807/1405–6 by Yusuf al-Shirazi. The band at the top has a large text in thuluth with a hadith: the doors of Paradise are open for the poor; mercy alights upon those who are merciful (abwåb al-janna maftË˙a Æalå fuqiråæ waæl-ra˙ma nåzila Æalå al-ru˙amåæ). Fine wooden doors were also made for local buildings: two sets of doors were donated to an unidentified shrine in Afushta near Natanz in central Iran by a local notable Sayyid Hasan al-Husayni in Ramadan 831/June–July 1428 (A. Godard 1936). Many other examples from the fifteenth century survive in shrines in Mazandaran. The layout of these wooden doors from fourteenth- and fifteenth-century Iran and Central Asia derives from contemporary bookbinding, with a larger central panel sandwiched between two smaller panels. The top panel was almost always inscribed. Sometimes the lower one was too, and occasionally there were other inscriptions running along the central edge. The main texts were usually hadith or pious phrases in Arabic, Persian or a mixture, with plays on words about doors. One common text, often repeated in miniature paintings from the same period, refers to God, the opener

of Gates. The doors to Ahmad Yasavi’s tomb are inscribed ‘The gate of sayyids is a mine of felicity’ (båb al-sådat maÆdin al-saÆådat). The bronze knockers (Lentz and Lowry 1989, fig. 68) are inlaid with a well-known hadith that the world is only an hour, so use it obediently, the same text found on the cenotaph in Rhode Island. The doors to the Gur-i Mir are inscribed with a similar hadith that the world is transitory. The doors from Afushta are inscribed with Persian verses about God, the opener, and the well-known Arabic phrase invoking happiness and long life as long as the pigeon coos. These inscriptions might well repay systematic study, for they are important evidence for the development of popular religion and the transition from Arabic to Persian. The woodcarvers of these expensive wooden doors were important craftsmen, who often signed their work, sometimes in prominent places. They were often designated master craftsmen (uståd) and had sometimes completed the pilgrimage to Mecca (˙åjjÈ). They are also called woodworker or carpenter (najjår) and carver (naqqår). The craft was handed down from father to son through several generations. Ahmad of Sari, his sons and his grandson, for example, signed several works surviving in shrines in Sari and Babol. The woodcarvers in these small shrines were often local experts. The Afushta doors, for example, are signed by the master Husayn b. ÆAli, a carpenter and carver from the nearby village of ÆUbbad. The names of the woodcarvers of the most elaborate pieces, however, suggest that they may have been recruited from afar. The cenotaph in the Imamzada at Qaydar is signed by Muhammad b. ÆAbdallah b. Abuæl-Qasim from Isfahan, and the doors to the tomb of Qutham b. ÆAbbas at the Shah-i Zinda are signed by Sayyid Yusuf of Shiraz. These wooden inscriptions also show the latest in epigraphic and calligraphic styles, juxtaposing several types of text in several styles of script. Square Kufic is used already on the doors from Bastam and appears on several other wood pieces. A second inscription in Kufic was sometimes inserted in stems of the main text in thuluth.

— 145 —

— ISLAMIC INSCRIPTIONS —

10.63 Wooden Koran-stand made by Hasan b. Sulayman Isfahani in 761/1360. Ht 130 cm. New York, The Metropolitan Museum of Art, Rogers Fund, 1910 (10.218)

The interplay of various texts and scripts is well illustrated (see Figure 10.63) by a fine Koran-stand in the Metropolitan Museum (RCEA 6337). On the exterior, the bottom is decorated with a polylobed arch enclosing a floral arabesque ground. Carved in relief above the densely packed ground is a thuluth inscription with invocations to the Twelve Imams. A small diamond at the bottom bears a word or two in square Kufic. The top is decorated with a square filled with the same arabesque ground and the word ‘God’ repeated four times. Another inscription carved in naskh at the bottom of one side gives the name of the craftsman, Hasan b. Sulayman Isfahani. Finally, an inscription incised around the top of the interior says that the stand was endowed to the so-far unidentified Sadriyya Madrasa in Dhuæl-Hijja 761/October–November 1360. Much woodwork survives from the later period in Iran and Central Asia, as palaces and pavilions typically had wooden porticos (Persian †ålår) with wooden columns and capitals, ceilings and screens (Orazi 1976). Some of these pieces bear inscriptions, which are the keys to dating, but in general the role of inscriptions diminishes in this period as inscriptions are abandoned in favour of all-over decoration. From the sixteenth century onwards in Iran, wooden doors were frequently painted and varnished in a technique commonly but erroneously called Islamic lacquer, and the epigraphic panels typical of earlier examples are replaced by figural scenes derived from book-painting. Inscribed wooden minbars continued to be popular in Anatolia in the fourteenth and fifteenth centuries. They used the same form as earlier models, but the techniques became increasingly simplified. The finest minbars were still made using the tongue-and-groove technique (Turkish kundekari), as on the minbars donated to the congregational mosques at Birgi in 722/1322–3 (RCEA 5474), Manisa in 778/1376–7 (RCEA 778 014) and Bursa in 802/1399–1400. Sometimes this laborious technique was imitated by using individual boards carved in relief and mounted on the minbar frame, and in a few later pieces this technique is further simplified so that both

— 146 —

— WOODWORK — the frame bands and the strapwork panels have been glued or pinned to boards, as on the minbar from the Ahi Ehvan mosque in Ankara (816/ 1413–14). The persistence of patrons’ names, dates and signatures shows that these were still expensive pious donations. Under the Ottomans, however, most minbars were usually made of other luxury materials such as marble, while wood was used for chests, thrones and arms made for the imperial court. These fancy objects were inlaid with ivory, precious metals, mother-of-pearl and tortoiseshell and encrusted with semi-precious stones. One of the earliest is a hexagonal walnut chest for a thirty-part Koran in the Museum of Turkish and Islamic Art in Istanbul (Anatolian Civilisations 1983, no. E19; Blair and Bloom 1994, fig. 297). The exterior is veneered with ebony encrusted with ivory panels so that the white letters of the inscriptions stand out from the dark ground. The texts include well-known Koranic verses (for example, 2:255, 3:18–19, 48:28–9), and a foundation text saying that the box to protect the Koran was ordered by sultan Bayazid II in 911/1505–6 and made by Ahmad b. Hasan al-qålibÈ al-fånÈ (‘the ephemeral frame-maker’). These fine furnishings continued to be made and endowed to tomb complexes in the sixteenth and early seventeenth centuries during the heyday of Ottoman rule (Anatolian Civilisations 1983, nos E75–8, E146–51; Atıl 1987, nos 107–11). Some pieces have generic inscriptions, with prayers, Koranic verses and hadith but without the names of specific patrons. Others include the names of court artisans, and these signatures help us to flesh out their careers. One such is Ahmed Dalgic. Recruited as a Janissary, he was trained in the imperial palace in Istanbul, studying mother-of-

pearl inlay under Ahmed Usta and architecture under Sinan. In September 1598, Ahmed was promoted to imperial court architect, serving seven years in that post before being awarded the rank of pasha. During his tenure as court architect, he designed the mausoleum of Mehmed III (r. 1595–1603), and for it he made and signed a magnificent pair of doors and a tall Koran box, both inlaid with tortoiseshell and mother-ofpearl (Anatolian Civilisations 1983, no. 151). The larger quality of inscribed woodwork surviving from the later period in the Yemen allows us to document the tradition there. Grillework, for example, may have been used since pre-Islamic times, but one of the earliest pieces to survive is the wooden screen in the tomb chamber of the Ashrafiyya madrasa built at TaÆizz in 1397–1401 (Lewcock and Smith 1974, plate 7). Wooden cenotaphs become more elaborate, with multistoreys and domed edicules, and the inscriptions show that these imposing objects were erected by the Zaydi imams in the eighteenth and nineteenth centuries to bolster their family line. The imam al-Husayn al-Mansur (r. 1726–47 with interruption) had one erected for himself in the Abhar mosque at SanÆaæ and another erected for his father al-Qasim al-Mutawakkil (r. 1716–26) in the corner of the mosque which he had built there. Al-Husayn alMansur’s son al-ÆAbbas al-Mahdi (r. 1747–76) continued the tradition. Another undated cenotaph was also installed in the mosque at SaÆda for Yahya al-Hadi, the imam who had introduced Zaydism to the Yemen in the ninth century. In addition to the foundation text, most of these pieces include short inscriptions among the arabesque and geometric decoration. The cenotaph for Yahya al-Hadi is distinguished by horizontal cartouches carved with inscriptions in the Ottoman style.

— 147 —

— CHAPTER 11 —

Ceramics

I

nscriptions on ceramics made in the Islamic lands contain a range of texts, and the length of the text depended in part on the technique used. The simplest method is to mould or carve the text in the soft clay, but this technique limits the text to single words or simple phrases. Potters could paint inscriptions on glazed objects, but the colour might run during firing and the overglaze decoration was fragile and subject to wear or even to flaking. The piece could be fired a second time, but this was an added expense used only for the finest wares. Potters in the Islamic lands also developed the technique of underglaze painting, in which the glaze protected the inscription and other decoration from wear. The inscriptions on ceramics include good wishes, Koranic quotations and poetry, but the ones that have attracted the most attention are historical texts, including the names of potters and patrons and dates of production. Historical inscriptions are particularly prominent on medieval Persian ceramics, and the importance of dates, written out in words or in numbers at the end of an inscribed band, was recognised at an early date. Sarre’s Denkmäler persischer Baukunst, published in Berlin in 1901–10, already contained a chapter on dated lustre-painted tiles. Kühnel (1924–1931) and Wiet (1933 Appendix II) expanded the number of dated examples. Ettinghausen (1935) underscored the point that dated pieces often provide the only means of localising a whole group or type, and by the time A Survey of Persian Art was published in 1939, he had compiled a list of more than 200 examples of dated Persian ceramics, including 137 lustre-painted

tiles and vessels. The number of dated pieces has continued to grow. Day (1941) added several pieces to Ettinghausen’s list, and Watson (1985) was able to list 220 examples of dated lustreware alone. Another sought-after piece of information written on ceramics is the name of the potter or workshop (Sauvaget 1965: 41–50). These signatures are particularly important since they are almost the only information that we have about potters and the only way that we can reconstruct their careers. In most cases the signature is hidden in the foot ring or on the bottom of a vessel, but sometimes it was written in a more prominent place. On earthenware bowls from ÆAbbasid times, the signature was often painted in cobalt blue on the opaque white glaze in the centre of the bowl. On medieval Persian wares, the signature is part of a long inscribed band that often ends with the name of the potter and date. Signatures predominate on expensive wares, particularly lustre-painted and enamelled wares, but are also found on cheaper, unglazed wares (Volov [Golombek] 1966; Baer 1989). Signatures are often preceded by a verb or word denoting action. The most common form is Æamal/Æamila (‘work of’/’made’). The term ßanaÆa (‘designed’) is also used, but the typical signature on lustrewares has kataba (‘wrote’) or better ‘decorated’, as inscriptions are only a part of the design. The use of the verb kataba links these ceramics with the arts of the book and implies that lustrewares were considered appropriate surfaces for painting and thus precursors to the flourishing tradition of book illustration that emerged in Iran in the fourteenth century. Sometimes it is not clear whether the name denotes a potter or

— 148 —

— CERAMICS — a workshop, for the same name may be found on a variety of wares and written in a variety of styles. A good example is Muslim (Jenkins 1968). Royal or court signatures and epigrams were also inscribed on ceramics. Bivar (1979) identified the sign-manual (Æalåma) of a Saljuq sultan, probably Arslan (r. 1161–76), incised on the bottom of a sgraffito sherd excavated at Takht-i Sulayman. It reads iætaßamtu biællåh (‘I have held fast to God’) and was apparently countersigned below, perhaps by the atabeg who held effective power. Blazons of Mamluk amirs and protocols of sultans are found on numerous potsherds (Mayer 1933), and the slogan of the Nasrid ruler Muhammad I (r. 1230 –72), wa-lå ghålib ilåæ-llåh (‘there is no victor save God’), was inscribed all over the plaster walls of the Alhambra palace at Granada (see Figure 7.42). The most common texts on Islamic ceramics are good wishes, with single words or bands invoking blessing, happiness, prosperity and the like. Although the most common, these texts are the least studied. Most studies concentrate on unique ceramics seemingly created for a specific moment and bearing long historical inscriptions, and there are very few studies of common types with repetitive texts, often with good wishes. Notable exceptions which give the complete text on each piece are Atıl’s catalogue of ceramics in the Freer Gallery, Washington DC, Ceramics from the World of Islam (1973) Philon’s catalogue of ceramics in the Benaki Museum, Athens (1980), and Grube’s catalogue of the early ceramics from the Khalili Collection in London (1996). Lack of study means that we do not have even basic information about these expressions of good wishes. We still do not know, for example, how the form of good wishes changed over time. One reason for the lack of study of these expressions of good wishes may be the difficulty in reading these texts, for many are unclear, even unreadable. Several explanations can be proposed for this illegibility. Due to the repetitive nature of ceramics, which were normally produced in multiple examples, the text may have become increasingly illegible as it was repeatedly written. This would have been the case when

the inscription, composed by a designer, was written by an illiterate craftsman. Legibility may also have been obscured by decorative concerns, and we do not know whether legibility or aesthetics was more important. At certain times, potters seem to have deliberately exploited the decorative effects, for example by painting or staining in the glaze so that the letters would run. The arbitrary division of words and phrases could also heighten the aesthetic effect (Flury 1939). A good example is the phrase ‘for its owner’ which is frequently divided li-ßå/˙ibihi. This division not only resulted in two equal groups of three letters, but also provided a well balanced, symmetrical arrangement in the first group with a low ßåd in the middle flanked by the two uprights of the låm and alif. The second group could then be squeezed in over the ßåd between the uprights. Lack of information about these expressions of good wishes on ceramics makes it difficult to explain the nature of some texts, particularly single words or names. When a word or name is not preceded by a verb, it can be unclear whether it is a signature or a general evocation of good wishes. The word saÆd, for example, may be read as either ‘happiness’ or the proper name of a potter or workshop (Jenkins1988). The word a˙mad, found on several Samanid slip-painted bowls, has been read both as the name of the potter and the word ‘more praiseworthy’ (Grube 1996: 56). The interpretation may depend on the location of the inscription and the historical context, and sometimes such ambiguity may have been intentional. Another type of inscription on Islamic ceramics which is related to the general category of good wishes comprises moralising aphorisms or proverbs. These can be taken from Koranic verses, as in fasayakfÈkahum allåh (‘God will suffice you against them’, Koran 2:137). The longest word in the Koran, it was a popular ÆAbbasid slogan that was inscribed on a variety of objects from the ninth century onwards. Other aphorisms were more popular in nature, as in the collection (1971) amassed by the twelfth-century Arab philologist al-Maydani (d. 518/1124). Moralising aphorisms,

— 149 —

— ISLAMIC INSCRIPTIONS — written in Arabic, were particularly popular on Samanid slipwares (Bol’shakov 1958, 1963a, b, 1966, 1969; Volov [Golombek] 1966; Ventrone 1974). Ghouchani (1986) as tracked down the origin of many of them, some of which are traditional sayings attributed to the Prophet’s nephew and son-in-law ÆAli. They show that the reverence for the Prophet’s family was already strong in eastern Iran by the tenth century, even if they do not show an overt dedication to ShiÆism. Poetry was also inscribed on Islamic ceramics. Some texts are phrases or poems composed as if the vessel were speaking and addressing the user. For example, an imperative asking the user to eat or drink to his health (kul/ishrab ˙aniyyan mariyyan) is inscribed on an unglazed bottle attributed to eleventh- or twelfth-century Gurgan and a contemporary series of dishes (Baer 1989). An Arabic poem written as though the jar itself were speaking and stating that ‘I quench the thirst of mankind’ is inscribed on a large water jar (˙abb), probably from the fourteenth century (Reitlinger 1951). A similar text occurs on a fragmentary Alhambra Vase in the Freer Gallery (Nykl 1957). Persian verses from both epic and lyric poetry were inscribed on medieval luxury vessels and tiles, and these have been extensively studied by Bahrami, Gyuzal’yan, Melikian-Chirvani, Ghouchani and others. By contrast, the verses on later Persian ceramics are rarely recorded. Almost no attention has been paid to the styles of writing used on ceramics and its implications. Notable exceptions are Flury, who enumerated six types of ornamental Kufic inscriptions on pottery in his essay in the Survey of Persian Art (1939: 1743–69) and illustrated them with superb drawings, and Volov [Golombek], who studied the development of interlaced Kufic on Samanid slipwares.

Early Period Inscriptions have been used to decorate Islamic ceramics at least since the early eighth century when the first distinctively Islamic ceramics appeared, and indeed it is the appearance of Arabic inscriptions that is often used to distinguish

Islamic pieces from pre-Islamic ones. The finest wares, known as Umayyad Palace Wares, have simple Arabic inscriptions and other designs painted in red on an earthenware body covered with a buff slip. Unglazed oil lamps from several sites in the eastern Mediterranean also have moulded inscriptions containing simple pious phrases and historical texts in Arabic (for example, RCEA 33–5; Grube 1996, no. 1). They give the name of the patron or artisan (who can be Muslim or Christian), the site (the most common is Jerash) and the date (published examples range from 99/717–18 to 125/742–3). Inscriptions became more prominent on wares made in the eighth and ninth centuries under the ÆAbbasids. Many inscribed fragments were uncovered during German excavations at Samarraæ, and Sarre’s publication of the ceramics from the site (Sarre 1925: 81–92) contains one of the first studies of inscriptions on Islamic ceramics. Sarre assumed that all pieces found at Samarraæ had to date from the brief period when the site served as the ÆAbbasid capital (836–92), but this is not necessarily true and one must use a wider horizon. More recently, the inscribed fragments in the extensive collection in the Benaki Museum were discussed in the catalogue by Philon (1980), in both the text and the chapter on inscriptions by Manijeh Bayani-Wolpert (pp. 293–302). Much information is amassed, although the discussions are not always clear. A similar arrangement of text and epigraphic appendix can be found in Grube’s (1996) catalogue of the early ceramics from the Khalili collection. A wide variety of polychrome glazes was one of the innovations introduced under the ÆAbbasids, and glazed examples of moulded relief wares, like their unglazed counterparts, often had short simple inscriptions in Arabic. There are, for example, several pieces with the name Abu Nasr (A. Lane 1939) His most complete signature occurs on a square condiment dish in the British Museum. The inscription is usually read Æamal abË naßr al-baßrÈ bi-mißr (‘the work of Abu Nasr al-Basri at Misr’ [that is, Egypt]), but the inscription is undotted so that al-baßrÈ has also been read alnaßrÈ, and the word bi-mißr open to question. On

— 150 —

— CERAMICS — other pieces, the text occurs in other forms (abÈ instead of abË) and other styles (the mÈm, for example, is round, not triangular as on the condiment dish), so the text cannot be always be a signature, but perhaps the name of a workshop or type of ware, and the varied find-spots of this type of ware (from Egypt to Iran) suggest that it had a wide provenance. Another signature occurs on a hemispheric bowl found at Raqqa in Syria and now in the Damascus Museum (Arts of Islam 1976, no. 250). A band around the rim is inscribed the work of Ibrahim the Christian [made] at Hira for the amir Sulayman, son of the Commander of the Faithful. The patron is believed to be Sulayman, son of the ÆAbbasid caliph al-Mansur (r. 754–75). It is extremely rare to find a specific patron, particularly such an important one, named on this utilitarian ware, and its occurrence here may perhaps be explained by the almost eggshell thinness of the piece. Another type of inscribed ÆAbbasid ceramic comprises earthenware bowls with an inscription painted in blue or green on an opaque white glaze, sometimes tin and sometimes alkaline. Since the inscription was painted with a brush, the artist had more latitude to write a longer text, and these bowls often have several lines of text written across the open surface on the interior. Texts include good wishes, particularly baraka li-ßå˙ibihi (‘blessing to its owner’), tawakkul takfå (‘trust [in God] is sufficient’), and the Koranic phrase from 2:137 fasayakfÈkahum allåh (‘God will suffice you against them’). Sometimes the pious phrase is followed by the name of the artisan, such as Æamal mu˙ammad al-[…] (‘the work of Muhammad the …’), as on a bowl in the Staatliches Museum für Völkerkunde in Munich (see Figure 11.64). Writing was used on these dishes not only to convey information, but also for decorative purposes, as the letters are often transformed ornamentally. Consequently, it is easy to read only the wellknown parts of the text, but words such as the second part of the artisan’s name are sometimes difficult or even obscure. ÆAbbasid lustrewares, both polychrome and monochrome, were also decorated with a few

11.64 Drawing of the inscription on a earthenware bowl painted in blue on an opaque white glaze with blessings to the owner and the signature of the artisan Muhammad al-? Diameter 23.5 cm. Munich, Staatliches Museum für Völkerkunde no. 28-8-82.

letters or words of text. The most common text is a stylised version of al-mulk (‘dominion [belongs to God]’). The word is sometimes repeated along a horizontal line which serves as the base for all letters, with no gap between words. This form is found both on fragments and on the monochrome lustre tiles used to decorate the mihrab at the Great Mosque at Qayrawan in Tunisia (Marçais 1928). On other polychrome wares, a single word or phrase, often repeated, is written in thick well-spaced letters with triangular endings on the letters alif and låm. These inscribed ÆAbbasid ceramics apparently served as models for inscribed wares produced in other regions of the Islamic world. For example, a bowl found at Madinat al-Zahraæ and dating from the caliphal period in the tenth century (Dodds 1992, no. 25) has the word al-mulk painted across the open face with thick letters with triangular tops like the inscriptions on ÆAbbasid polychrome wares. Inscriptions become more prominent on the slipwares associated with eastern Iran in the ninth and tenth centuries under the patronage of the Samanid dynasty (819–1005), and large numbers were excavated at Nishapur (Wilkinson 1973) Afrasiyab (old Samarqand) and other sites in

— 151 —

— ISLAMIC INSCRIPTIONS — Central Asia. These earthenware ceramics, mainly plates and bowls with a red or buff body covered with a white slip made of diluted fine white clay, presented an ideal surface for painted decoration in various metallic oxides mixed with diluted clay. One of the most common and distinctive types is a black-on-white ware, in which inscriptions and other decoration were painted in manganese over the white slip. The type may have derived from ÆAbbasid white wares, which present a similar aesthetic and have many of the same texts and calligraphic features, including extra-long verticals, triangular endings on letters, and a miniature script that gives the impression of bristles springing from a horizontal line. The texts painted on these slipwares range from single words or phrases such as baraka lißå˙ibihi (‘blessing to its owner’) to longer Arabic proverbs, ‘Patience is the key to felicity’ (al-ßabr miftå˙ al-far˙) or ‘Planning before work protects you from regret’ (al-tadbÈr qabla al-Æamal yËmminuka min al-nadm). Many examples are given in Bol’shakov, Volov [Golombek], Ventrone and Ghouchani. The inscriptions are usually written around the rim of the vessel, sometimes with another word written like a pivot in the centre, although occasionally the texts are written as a single band stretching across the surface. A decorative device often marks the beginning of the phrase. On a plate in the Freer Gallery, for example (see Figure 11.65), a palmette, set at about seven o’clock on the photo, marks the beginning of the aphorism al-jawd min akhlåq ahl aljanna (‘Generosity is a quality of the people of Paradise’). From the point of view of style, the remarkable feature of these inscriptions is the plaiting and twisting of both the bodies and the stems of the letters, a feature which often threatens to obscure the legibility of the text, yet at the same time transforms it into a vehicle of artistic achievement. Bulliet (1992) has attempted to connect this epigraphic type of slip-painted earthenwares with early converts to Islam who were more conservative, suggesting that the buff wares painted with figural scenes found at the same sites were favoured by later converts, who belonged to a more populist

sector of society. His essay, while stimulating, is not entirely convincing, for there is no clear evidence that the different types of earthenwares were used exclusively by one sector of society or another, but his essay deserves attention as a rare attempt to set the objects in context and to explain the meaning of the epigraphic ornament.

Middle Period Fine ceramics made in Egypt under the Fatimids (969–1171), particularly lustrewares, were also decorated with inscriptions, which gradually become longer and more elaborate in style as the letters sprouted leaves and flowers. The texts are usually benedictions such as naßr (‘victory’), yumn (‘good fortune’), surËr (‘joy’) and the like, written in a band across the surface or around the rim or flaring side of a plate and bowl. Fatimid lustrewares were imported into Spain and provided the model for the local style of lustreware produced there, some of which have similar bands of benedictory inscriptions around the rim. The same sorts of good wishes are also inscribed on fine ceramics made in medieval Syria, both the types known as Tell Minis and Raqqa wares (Porter 1981; Grube and Tonghini, 1988–9). These good wishes were repeated so frequently that on some pieces the texts have became illegible patterns. Names of patrons and signatures of potters or workshops also occur on a few pieces of Fatimid lustreware, and these names are particularly helpful in assigning dates to the ceramics as a group (ÆAbd al-RaæËf 1958). The most famous is a fragmentary plate in the Islamic Museum, Cairo, dedicated to Ghaban, who is called qåæid al-qawwåd (‘commander-in-chief’), a title he bore between 402/1011 and 404/1013 (al-Båshå 1956). Jenkins (1988) has identified twenty-one potters who signed seventy-two pieces from the Fatimid period, usually with the word Æamal (‘work of’). They include ÆAli al-Baytar, Ahmad and JaÆfar al-Basri, but the most profilic potter was Muslim b. al-Dahhan, whose name appears on some forty complete or fragmentary lustreware vessels (Jenkins 1968; Philon 1980: 168). A fragmentary

— 152 —

— CERAMICS —

11.65 Earthenware plate underglaze painted in brown and red on a white slip with the aphorism al-jawd min akhlåq ahl al-janna ‘Generosity is a quality of the people of Paradise’. Eastern Iran, tenth century. Diameter 21 cm. Washington

plate inscribed by him in the Benaki Museum, Athens (see Figure 11.66) is dedicated to someone whose name ends in al-Hakimi. The two-line inscription reads [Æama]l muslim bin al-dahhån wifq / …È al-˙asan iqbål al-˙akÈmÈ (‘[the wor]k of Muslim b. Dahhan to please’ / ... -i al-Hasan Iqbal al-Hakimi’). The inscription presumably refers to the Fatimid caliph al-Hakim (r. 996–1021) and

DC,

Freer Gallery of Art 65.27.

suggests that Muslim was active in the late tenth or early eleventh century. Other pieces by Muslim are inscribed bi-mißr (‘in Egypt’). Inscriptions that include the word Æamal (‘work of’) clearly indicate the signature of the potter Muslim, but the varied styles of script and forms of signature suggest that Muslim may have also been the name of the workshop. A few rare pieces of

— 153 —

— ISLAMIC INSCRIPTIONS —

11.66 Fragmentary lustre plate signed by Muslim. Egypt, c. 1000. Diameter 38 cm. Athens, Benaki Museum no. 11122.

Syrian ware are inscribed with the patron’s name, such as a vase with the name of al-Malik al-Mujahid Asad al-Dunya waæl-Din Shirkuh, an Ayyubid prince from Homs who died in 637/1240 (RCEA 4178). The question of signatures on Fatimid wares is complicated by the varying location of the writing. Most pieces are signed on the outside, in the base of the foot ring, but some are signed on the inside and others on both sides. The example of the word saÆd, meaning either happiness or a proper name, shows how interpretations differ. Although he was once thought to be the most versatile potter of the Fatimid period, Jenkins (1988) found the word in a variety of scripts on forty-six fragments of varying types and decoration and concluded that it was not a signature but a benediction. In the mid-twelfth century, a major revolution in ceramic technology took place when potters adopted an artificial body made from ground quartz mixed with a small amount of white clay

and ground glaze. This material, usually known as ‘frit’ or ‘stone-paste’, could be decorated in an extraordinarily wide range of decorative techniques. With the revolution in technology came an explosion in dated and signed pieces. By contrast, only a very few earthenwares are dated, such as two unglazed relief wares signed by alMaÆari(?) in 612/1215, a ewer in Cairo and a bottle in the Louvre (RCEA 3781 and 3782). The scarcity of dated earthenwares makes it difficult to intrepret the dates. One of the very few dated earthenwares is a high-stem bowl with sgraffito decoration in Chicago (Survey, fig. 532) with several words and ciphers crudely scratched on the bottom. The words seem to be a signature, ‘the work of Yahya the potter’ (Æamal ya˙yå kåsh[È]gar). The ciphers ‘83’ are clear, but their meaning is not. Pope interpreted them as the last two digits of a date and suggested [3]83 or [4]83 (993 or 1090), whereas Day interpreted them as the first two digits of the date and read another line incised to the right as a final 6, making the date

— 154 —

— CERAMICS — 836/1432. Without any other indication that these numerals indicate a date, they could also be a potter’s mark. The sloppiness of the line shows that the potter was no calligrapher and that the inscription was not part of the decoration. Since this type of ware is probably the most common type of glazed ceramic in the Islamic period, it is difficult to use this cipher for dating sgraffito wares. In contrast to earthenwares, many kinds of Iranian fritwares are dated, and the prominence of dates and signatures distinguishes these medieval Persian ceramics from those made elsewhere. The dates and signatures are quite conspicuous, often written at the same height and in the same style as the Koranic text or poetry that precedes them in a band on the inside or outside of a vessel or the edge of a tile. The dates and signatures usually occur on the finest vessels, suggesting that these pieces were unique commissions and setting them apart from typical ceramic wares made in multiple examples for sale on the open market. The relative quantity of dated fritwares also allows us to put the pieces in order and draw conclusions about the dates of production. The most famous examples of moulded wares with a monochrome glaze are two large vases dated 681/1282–3 and 683/1284–5 in the Metropolitan Museum (RCEA 4819; Survey, plate 759) and the Freer Gallery (Atıl 1973, no. 77). Grube (1996, no. 148) has recently published a fragmentary bottle in the Khalili Collection signed by ÆAli Buhuni and dated 534/1139–40. Lack of provenance for the piece and its unusual date, a century and a half before the two other dated pieces and twenty-five years before any other dated piece of fritware, suggest, however, that this date should be treated with caution until technical analysis can confirm the authenticity of the piece. Several pieces of underglaze-painted fritwares are also dated, but the dates are often written in ciphers, which are difficult to read. A good example is an openwork underglaze-painted ewer in the Museum of Islamic Art, Cairo. Wiet read the ciphers as 562/1166, making the ewer the earliest dated fritware, but Bahrami read the same ciphers as 657/1259 (evidence summarised in Grube 1996:

152). Other dated examples of underglazed wares painted in black and blue range from 601/1204 (RCEA 3601; Survey, plate 734A) to 612/1215 (the Macy jug in the Metropolitan Museum; RCEA 3786), and a polychrome example of underglaze-painted ware is dated 672/1274 (RCEA 4682quater.; Survey plate 734A), so the later date for the Cairo ewer seems more probable, but without publication of the text it is impossible to be definitive. The most common types of signed and dated fritwares are the two most expensive types of medieval Iranian ceramics: lustre-painted and enamelled wares. Although expensive, these techniques gave the potter great freedom of expression, enabling him to paint elaborate figural scenes and long inscriptions. Many inscriptions on lustrewares were recorded by Bahrami (1949), and the basic surveys by Watson (1985) and Ghouchani (1992) include extensive epigraphic information, as the inscriptions provide much of the evidence for the localisation and organisation of production. These inscriptions are written in two scripts: a stylised Kufic used for good wishes that sometimes degenerates into repetitive combinations of meaningless letters, and a scrawling cursive naskh with many misconnected letters used for poetry and prose, including signatures and dates. Potters signed and dated their wares in various ways. Watson’s Appendix I includes a short discussion of the words used, and his Appendix III includes a paragraph on the ciphers used for the numbers. These surveys show the value of reading many texts together, as it is possible to decipher the hastily written scrawl on some pieces because it is written more clearly on others. The high number of signed and dated pieces also allows us to draw broader conclusions about the nature of production, interpretations that are not possible when working with a rare or unusual piece. Inscriptions show that the centre of production was Kashan. A lustre bowl found at Gurgan and now in the V&A (Watson 1985, colour plate G), for example, is signed by Muhammad b. Muhammad of Nishapur ‘living in Kashan’ (al-muqÈm bi qåshån). Similarly, star tiles made in 738 and 739/

— 155 —

— ISLAMIC INSCRIPTIONS — 1337–9 (RCEA 5733–58; 5787–8) are inscribed with the name of the site Kashan. The text on several tiles asks God to protect the city from the accidents of time, suggesting perhaps that Kashan was suffering economic decline in the mid-fourteenth century. Work was carried out in a workshop, for the text on several other tiles says that the master Jamal al-naqqåsh worked in the workshop (kårkhåna) of the sayyid of sayyids Rukn al-Din Muhammad b. Sayyid Zayn al-Din ÆAli the potter (al-gha∂årÈ). The three styles of lustrewares produced at Kashan had varying amounts of epigraphy. Pieces in the monumental style, none of which is dated, have large figures painted in reserve but relatively few inscriptions. With the introduction of the miniature style, artists were free to paint small-scale designs in bands or panels on a variety of vessel shapes, and the inscriptions on these pieces are longer and more serious. In addition to stylised or pseudo-inscriptions and conventional blessings to the owner, they contain quatrains in Persian as well as four dates ranging from the third day of 575/10 June 1179 (the well-known vase in the British Museum; Survey, plate 636b) to Jumada I 590/May 1193 (a dish in the Khalili collection; Watson 1985, fig. 39) and two signatures by Abu Zayd (a fragmentary vase dated 587/1191; Watson 1985, plate 53 and a dish in a private collection in Tehran dated Rajab 598/March 1202). Inscriptions play their fullest role on lustrewares in the Kashan style, in which figures are drawn in reserve on a lustred ground relieved by a series of small spirals or commas scratched through the lustre to give it a lighter texture. Concentric bands written in different scripts often surround a central motif. Cursive inscriptions are both scratched through the lustre and painted on the white glaze; angular inscriptions are often set into ornate friezes on a scrollwork ground. The texts are mainly verses, usually Persian quatrains, although a few Arabic poems have also been identified. Generalised good wishes to the owner are common, but references to individual patrons are rare. Two notable exceptions are the vizier al-Hasan b. Salman, to whom a lustre jar in the Bahrami collection was dedicated, and a

large plate in the Freer decorated with a complex scene including a sleeping groom (see Figure 8.45). The text reserved in white from the brown ground of the rim (RCEA 3672; Guest and Ettinghausen 1961; Ghouchani 1992) contains blessings to an unidentified amir and commander (isfahsalår), the signature of the potter, and the date Jumada II 607/November 1210. Further inscriptions painted in brown on a white ground on the inside and outside walls contain a combination of various odes, a quatrain and good wishes in a mixture of Arabic and Persian. The Freer lustre plate is one of the nearly sixty dated vessels in the Kashan style known from the pre-Mongol period, and the large number of dated pieces allows us to draw various conclusions about production. We can bracket the approximate beginning and end of production from a sherd of a dish in the British Institute of Persian Studies made in Dhu’l-Hijja 595/ September 1199 (Watson 1985, fig. 55) and a bowl in the Kelekian collection (Survey, plate 773b) made in 624/1226. The dated vessels also suggest that the main period of production took place in the first two decades of the thirteenth century. The number of dated vessels in the Kashan style also allows us to eliminate other lustrewares with conflicting dates. A bowl in the Khalili collection (Grube 1996, no. 281), for example, is dated in words ‘Shawwal five hundred and four’ (March 1111), but on the basis of its shape and style of decoration, the bowl clearly belongs to the early thirteenth century. Inscriptions also play a major role in the decoration of lustre tiles (Bahrami 1937; Kiani et al. 1983; Ghouchani 1992; Carboni and Masuya 1993). Star and cross tiles usually have an inscribed border; frieze tiles have a moulded inscription in the main field; and mihrab or tombstone tiles, which can comprise up to forty individual tiles in a single ensemble, often have numerous inscriptions in a variety of scripts. This panel of three tiles decorated with a hanging lamp (see Figure 11.67), for example, has several inscriptions written in relief. The Throne Verse (Koran 2:255) begins in the rectangular frame and continues in the field below the lamp inside the arch. The signature of Hasan b. ÆAli b. Ahmad Babuya the

— 156 —

— CERAMICS — builder (al-bannåæ) fills the spandrels. The most unusual inscription is the arch of the niche, which is composed of the word fasayakfÈkahum written in interlaced Kufic. The rest of the Koranic phrase from 2:137 fasayakfÈkahum allåh wa huwa al-samÈÆ al-ÆalÈm (‘God will suffice you against them, for He is the all-hearing, the allknowing’) continues around the lamp. In addition to the verses found on vessels, tiles are also inscribed with Koranic passages and hadith. They are drawn from a limited repertory, and certain favourite passages predominate. Star tiles often have Koran 1 and 112. The Throne Verse (Koran 2:255) is also popular. Tiles from tombs are often inscribed with chapters from the end of the Koran, texts that deal with the Day of Judgment, while those from mosques are taken from the middle of the book and stress man’s duty to God. Tiles were produced for a longer span of time than vessels. The earliest dated tile in the Islamic Museum, Cairo, was made by Abu Zayd on Wednesday, the end of Safar 600/7 November 1203 (RCEA 3587; Ghouchani 1992, fig. 1). Tile production continued after the Mongol conquest of Kashan in 1224, the time when the production of lustreware vessels ceased, and a continuous series of tiles was produced into the middle of the fourteenth century. One of the latest is a star tile from the shrine of Imam Rida at Mashhad made by Haydar on 9 Ramadan 760/4 August 1359 (Ghouchani 1992, fig. 5). Production continued sporadically into the twentieth century (Watson 1975). In comparison to lustrewares, enamelled wares contain less documentary information, and even that which does exist is sometimes suspect, as it is relatively easy to fake or add to these texts which are painted over the glaze. For example, the date 640/1242 on a bowl in the V&A, which is written in an irregular form with the hundreds before the tens (fÈ sittamiæa wa arbaÆÈn) in a restored area of the bowl, should be disregarded. As with lustrewares, the texts on enamelled wares can best be read in groups. In addition to highly stylised Kufic inscriptions with good wishes and longer cursive texts with poetry, the

11.67 Panel of three lustre tiles decorated with a hanging lamp made by ÆAli b. Muhammad in 709/1309-10. New York, The Metropolitan Museum of Art, Rogers Fund, 1909 (09.87).

inscriptions on ten enamelled bowls contain dates, ranging from 576/1180 to Safar 616/April 1219 (Watson 1994). Five of the bowls were made in the month of Muharram, the first month of the lunar calendar, in the years 582 or 583 (March 1186 or 1187), and it has been suggested that the bowls were made as New Year’s presents to mark No Ruz, the beginning of the Persian solar year that occurs in the springtime. Curiously, the dates on these bowls and on other

— 157 —

— ISLAMIC INSCRIPTIONS — lustrewares are qualified as ‘the Arab hegira year’ (sana hijriyya Æarabiyya), suggesting that another calendar, such as the Persian solar one, was used in pre-Mongol Iran alongside the Muslim lunar one. The verses inscribed on these expensive fritwares are either quatrains or passages from epic texts. Ghouchani has identified hundreds of quatrains from dozens of poets. In some cases the potter was also the poet, for he signs his work qåæiluhu wa kåtibuhu baÆda må Æamilahu wa ßanaÆhu (‘the narrator [of the preceding verses] is its writer after he had made and decorated it’). In other cases, the quatrains were taken from the work of another poet, whose name is cited with the verb qåla (‘so-and-so said’), and the potter simply indicates that he wrote the verses ‘in his own hand’ (bi-kha††ihi). This phrase is often written very hastily and was at first mistaken for part of the potter’s name and misread as ‘Baza’ or ‘Abu Rufaza’. Mistakes in the verses show that the potters transcribed the poems from oral recitations rather than from written texts. The poems on fritwares also record different versions which are earlier than those found in printed editions and are thus important evidence in documenting literary history. Most of the poems deal with the agonies of love, and some are remarkable for their low literary merit. The renowned literary scholar Mojtaba Minovi, for example, dubbed the quatrains on the Freer plate ‘doggerelish and stupid’ (in Guest and Ettinghausen 1961: 29, n. 16). The poems seem to have little to do with the figural scenes painted on these wares or their ostensible function. Epic verses, common on star tiles but also found on frieze tiles and vessels, come mainly from the Shåhnåma (Gyuzal’yan 1951). From a small fragment found in a heap at the site of the Ilkhanid palace at Takht-i Sulayman, MelikianChirvani (1984, 1988, 1991) reconstructed a frieze of some twenty large moulded lustre tiles with verses from the Shåhnåma in arched niches. In two cases, the text was changed from the third to the second person singular so that the resident of the Ilkhanid palace was compared metaphorically to the kings of ancient Iran. The

verses on the tiles are taken from various parts of the epic and do not present a continuous narrative, but rather suggest general themes so that the text would have served as an aide-mémoire to evoke the glorious past and metaphorically relate the current ruler to the heroes of myth and antiquity. Many of these fritwares are signed, and the names show that fritware pottery was a high-status profession carried out by influential members of society. The names of some twenty potters are recorded. The most famous is the master Abu Zayd, who worked in both lustre and enamel. He left more signatures than any other potter, and his career can be traced over thirty-four years from 582/1186 to 616/1219 (Watson 1994). Ghouchani (1992: 5) has suggested that the two short words at the end of the name of the potter Sayyid Shams al-Din al-Hasani on the Freer plate (see Figure 8.45) should be read not as a mistaken and extra fÈ shahr but as Abi Zayd, thereby making the potter’s full name Sayyid Shams al-Din alHasani Abu Zayd. The stylised way of writing zayd corresponds with the potter’s signature elsewhere. Ghouchani’s work shows the importance of not accepting standard readings, but looking afresh at well-known pieces. It also points out the high status of these potters, who were Hasani sayyids (descendants of the prophet through his grandson Hasan) and had titles such as Shams al-Din (‘sun of religion’). These affiliations continued into the fourteenth century, when the potter ÆAli b. Jahan was also a Hasani, and another potter signed himself Jamal al-Din (‘beauty of religion’). Signatures allow us to trace several families of potters over several generations, and their careers show the waxing and waning status of the potting profession. The best-known family is the Abu Tahirs, known over three or possibly four generations (Blair 1986). Dated pieces begin with the large cenotaph for Fatima in her shrine at Qumm made by Muhammad b. Abi Tahir in 602/1206. His work, dating over the next decade, represents the artistic and technical peak achieved by the Kashan lustre potters, with calligraphy carefully balanced against background decoration. Muhammad’s son ÆAli, who produced works dated from 640/1242 to ShaÆban 663/May 1265, was the only lustre

— 158 —

— CERAMICS — potter to sign his work during the mid-thirteenth century, and although not as sophisticated as the masterpieces done earlier in the thirteenth century, it is still technically quite fine. Potters’ status and fritware quality declined in the next generation. The work by ÆAli’s son Yusuf in the first third of the fourteenth century fell off in quality, and he also worked in cheaper techniques such as underglaze-painting in blue and black. His two brothers also turned to other professions. One brother, Abuæl-Qasim, became a scribe and accountant in the Ilkhanid bureaucracy, writing a biography of sultan Uljaytu and a treatise on minerals and precious substances. The appendix describing the techniques of pottery manufacture is perhaps the only document of its sort to survive from the pre-modern Islamic world (English translation by Allan 1973). Another brother ÆIzz al-Din Mahmud became a Sufi and entered the Suhrawardi khanaqah at Natanz where he wrote a spiritual guide Mißbåh al-Hidåya wa Miftå˙ al-Kifåya (‘Light of Divine Guidance and Key to Completeness’). Fritwares were also produced in Saljuq Anatolia. They follow the general style used in Iran, but the inscriptions have often been reduced to illegibility or pseudo-inscriptions. These texts stand in marked contrast to the elaborate tile revetments used in the region, particularly for religious buildings, which are distinguished by long inscriptions (Meinecke 1976). Some texts are religious, including popular Koranic verses, hadith, pious sayings and religious formulas. Sometimes the Koranic verses were chosen to suit the building’s function (for example, 55:26–7 on a mausoleum) or as a play on words (for example, 69:28–9, including the word sul†ån (‘power’) on the Shifaiye Madrasa erected in Sivas by sultan Kaykaæus). There are also historical inscriptions in Arabic recording construction, restoration and names of artisans as well as Persian verses, including a hemistich by Firdawsi written on the tomb of Sahib Ata in Konya. Most are done in Kufic, but, with the adoption of cursive naskh in the early thirteenth century, Kufic assumed a decorative role, with formulas repeated in square Kufic or bands in Kufic with interlaced stems (see Figure 7.39).

Late period As in other media, the role of inscriptions in ceramic decoration diminished in the later period, although the occasional piece with a name or date is still one of the fundamental ways of localising a type. Inscriptions show that the production of lustreware petered out under the Mongols: there are some vessels dated in the 660s/1260s, 670s/1270s and 680s/1280s, but the latest lustre vessel in the series is dated 683/1284, although regular production of lustre tiles continued until 740/1339–40. Several moulded mihrabs and tiles glazed in a single colour, usually turquoise, have dates in the early fourteenth century, establishing a range for the production of those wares. After enamelled wares went out of fashion, they were replaced by the type of enamelled and gilded ceramics known as lajvardina ware, from låjvard, the Persian word for lapis lazuli, because of the deep blue glaze that characterises many of them. A fragmentary star tile dated 715/1315 once in the collection of Richard Ettinghausen, and a bowl dated 776/1374–5 in the Museum für Islamische Kunst, Berlin, provide documentary evidence for the span of production. The major type of wares produced in Iran in the fourteenth century, the underglaze-painted Sultanabad wares, however, are generally anepigraphic. The lustrewares produced at this time at the other end of the Islamic lands, in Spain under the Nasrids (r. 1230–1492), were also inscribed, and the large surfaces of these big pieces allowed for lengthy inscriptions, usually repeated phrases of good wishes. The best-known lustrewares from Nasrid Spain are the Alhambra Vases, so called because some were found in the Alhambra Palace at Granada (Ettinghausen 1954; Kenesson 1992). They are traditionally divided into two chronological groups based on shape, technique and epigraphic style. The earlier type, usually dated to the late thirteenth or early fourteenth century, has a bulbous shape and short neck (for example, vases in Palermo and St Petersburg, Dodds 1992: nos. 110–11) and is decorated in monochrome lustre with a wide central register inscribed in bold Kufic. The later type, usually dated

— 159 —

— ISLAMIC INSCRIPTIONS —

11.68 Fortuny Tablet. Lustre-painted terracotta. Ht 108 cm. Madrid, Instituto de Valencia de Don Juan.

to the late fourteenth or early fifteenth century (for example, a vase in the Museo Hispanomusulmán, Granada), has a more elegant shape and additional decoration in cobalt blue or gilding, with a narrow band of cursive script. The texts are usually repeated phrases of good wishes, but a fragmentary jar in the Freer Gallery also has an Arabic poem written as though the vase were speaking and asking the onlooker to contemplate its excellence (Nykl 1957). The use of repeating expressions of good wishes was typical of Nasrid work in many media,

including tiles, and the largest example of lustreware known, the Fortuny Tablet (see Figure 11.68) named after the painter Mariano Fortuny y Madrazo who purchased it in the Albaicín quarter of Granada in 1871, has a carpet-like design with an arabesque field surrounded by a border with cartouches inscribed in typical Nasrid cursive with a paean to Yusuf III (r. 1407–17): Æizz li-mawlanå al-sul†ån abÈæl-˙ajjåj al-nåßir li-dÈn allåh. Inscriptions also provide guideposts for localising and dating various types of Mamluk ceramics. Lustre production can be localised in Syria, thanks to the inscription on a large blue jar in the Museum of Islamic Art in Kuwait (Arts of Islam 1976, no. 311) that says it was ‘made for Asad al-Iskandarani; the work of Yusuf in Damascus’. Dates and signatures on several underglaze-painted wares provide a framework for production of that type. An albarello in the Museo di Capodimonte in Naples is dated 717/1317, and several pieces and fragments are dated 745/1344–5. Signatures inside the foot ring of several pieces identify potters and workshops active in the fifteenth century. Names include ÆAjami and Ustad al-Misri, but the most famous is that of Ghaybi (Abel 1930), and this name illustrates some of the problems of distinguishing workshop and potter and tracing a career. Ghaybi is best known from his signature on the extensive tile revetment in the mosque and tomb of Ghars alDin al-Tawrizi (d. 826/1423) at Damascus. The signature, Æamal ghaybÈ tawrÈzÈ (‘the work of Ghaybi of Tabriz’), suggests that Ghaybi was associated with Tabriz, a noted ceramic centre in northwestern Iran. Meinecke (1988) has identified other tilework by this workshop in Syria, Egypt and Anatolia, showing the extensive peregrinations of this workshop, and Jenkins (1984) has identified many vessels and fragments with the name Ghaybi or Ibn al-Ghaybi, suggesting that the potter, his family or his workshop also produced vessels. The most common type of Mamluk ceramic, sgraffito or incised ware, was typically decorated with bold inscriptions and blazons, but many of the epigraphic designs are poorly rendered and state that the object was ‘one of the things made for the amir …’ (mimmå Æumila bi-rasm al-amÈr …)

— 160 —

— CERAMICS — followed by the honorific titles of unidentified officials. These ceramics were probably massproduced to be sold to the households of various amirs. An occasional piece, however, was made on specific commission. A bowl found at Tod in upper Egypt, for example, is decorated in the centre with an eagle and inscribed around the rim with the name of a specific judge (Décorbert and Gayraud 1982). Some stemmed cups and fragments bear potters’ signatures, especially that of Sharaf al-Abawani (Atıl 1981). Dates and signed pieces also provide guideposts for the evolution of different types of ceramics produced in Iran in later times (EIr, ‘Ceramics xiv–xv’). Lustreware, including tiles and vessels, was produced sporadically until the twentieth century (Watson 1975, 1985), but a spurt of production occurred in the second half of the seventeenth century. Only one example is dated – a long-necked bottle once in the Wallis collection (Watson 1985, fig. 136), and even that date is unclear – but several vessels are signed ‘the work of Khatim’. Another lustreware revival occurred in the late nineteenth century and is associated with the work of ÆAli Muhammad (Scarce 1976). The sporadic production of lustreware can be documented through a few other dated lustre tiles used to mark graves or building foundations, and five mosques in central Iran have tiled minbars dating from the mid-fifteenth to the mid-sixteenth century (O’Kane 1986). One of the few sites of production that can be identified in inscriptions on Timurid ceramics is Mashhad. It is mentioned on two underglazepainted vessels: a spittoon in the Royal Scottish Museum, Edinburgh, whose foot bears a scratched inscription saying that it was made for our lord Husam al-Din Shirazi at Mashhad in 838/1444–5 (Grube 1974), and a plate in the Hermitage, whose cavetto bears an inscription that it was made there in 878/1473–4 (Ivanov 1980; Blair and Bloom 1994, fig. 90). Other rare examples of dated Timurid ceramics include a series of four pieces of black-under-green painted wares, dating between 873/1468–9 and 900/ 1494–5. Later Iranian ceramics also include a handful of unique dated specimens and several series of

dated examples. Unique pieces include a damaged blue-and-white pilgrim flask in the V&A, with an inscription around the edge in Persian verse ending with the date 930/1523–4, and a dish depicting the zodiac in the Museum für Islamische Kunst, Berlin, signed by ÆAbd al-Vahid and dated 971/1563–4. Dated series include the handful of dated pieces of Kraak ware, ranging from 1025/1616 (a teapot in the British Museum) to 1051/1641 (a base for a waterpipe in the V&A), and the signed and dated examples of blue-andwhite ranging from 1281/1864–5 to 1318/1900–1 and assigned to Naæin (Scarce 1991: 930–9). The Kraak wares also have Persian imitations of Chinese marks in the foot ring. The square or tassel marks used on earlier examples were gradually replaced by sketchy marks resembling a capital A or by multiple crosses painted in black. In general, the dated pieces from the later period are more mediocre in quality than the superb dated examples of medieval times. This may indicate a change of taste and suggest that these wares were not signed and dated by the potters as a mark of esteem, but rather were made as mementos or souvenirs for the market. Few of the texts, other than the dates, have been published. The Qajars (r. 1779–1924) had a strong taste for wares decorated with their names and other signs of rank, and even ceramics ordered from abroad had inscriptions. The dinner services that Fath ÆAli Shah (r. 1797–1834) ordered from the factories of Crown Derby and Wedgwood for the Gulistan Palace in Tehran, for example, were inscribed with his name and the date 1234/1820. Inscribed pieces also provide the basis for localising various types of Ottoman ceramics, but only recently have Atasoy and Raby (1989) integrated the very few inscribed pieces with technical and stylistic criteria to sketch an overall survey of the development of Iznik wares. The basic dated pieces of Iznik ceramics are tiles and furnishings, especially ceramic mosque lamps, made for building complexes with established dates. The extensive tile revetment at Mehmed’s complex at Bursa, known therefore as the Green (Yesil) complex and built between 822/1419 and 827/1424, shows that the impetus for the

— 161 —

— ISLAMIC INSCRIPTIONS —

11.69 Underglaze-painted mosque lamp from the Dome of the Rock in Jerusalem made by Musli in 956/1549. Ht 38.5 cm. London, British Museum 87.5-16.1.

— 162 —

— CERAMICS — extraordinary revival of glazed ceramics, after a hiatus of more than a century, came from Iran, as the mihrab is signed ‘the work of the masters of Tabriz’. Judging from tiles in other dated buildings, the atelier continued to work for the Ottoman sultans for the next half-century. Dated vessels are much rarer. One of the most famous is a small ewer in the British Museum with an inscription in Armenian mentioning the donor, a certain Abraham of Kütahya, and a date equivalent to 1510. Its rarity is clear from the fact that it is constantly cited in the literature and that its inscription has given rise to the common name of the type, Abraham of Kütayha ware. Another famous piece in the British Museum is a mosque lamp found at the Dome of the Rock in Jerusalem (see Figure 11.69). It has a fragmentary inscription around the foot ring with the date 956/1549, the signature of the artisan (read variously as Musli, Musalli or even Must[af]a), and the name of the Iznik saint E∞refzade Rumi. The only dated piece from the early seventeenth century is a dish with a prunus tree in the Musée de la Renaissance, Château d’Ecouen, inscribed and dated under the foot [10]15/1606–7 (Ennès and Kalus 1979). There are more dated examples of Iznik ceramics from the later seventeeth century, and they document the widening of the market and the declining quality of the type. Several dated pieces from the 1640s include a dish with a colt inscribed on the rim in cursive Greek with the equivalent of 1640 and a second dish with a Greek inscription mentioning Nicea and the Church of 318 Nicene Fathers and the date April 1646 (Atasoy and Raby 1989, fig. 639). A series of fifteen dishes with uncial Greek inscriptions is dated between 1666 and 1678. The texts are extracts from the Psalms or religious invocations, but the

inscriptions on them are similar in style and orthography to a tile dated 1667 designed as a setting for a fountain and inscribed with a donor or artist’s dedication and the name [La]skaris. The same epigraphic style is used on tiles in the Great Lavra inscribed with the date 1678 and the name of the maker, a Christian monk Zosimas. Inscriptions continue on wares made at Kütahya in the eighteenth century. A polylobed censor with polychrome decoration in the Cincinnati Art Museum, for example, bears an Armenian inscription stating that it was made in 1176 (1726–7 CE). As in the rest of Ottoman art, inscriptions play only a minor role on Iznik ceramics, and mosque lamps are the only group to make use of large-scale inscriptions. They have Koranic quotations or pious texts, particularly the profession of faith and the names of the Four Orthodox caliphs, as the pious nature of these texts was appropriate for objects destined for a religious setting. The inscriptions are of mediocre quality and were probably the work of the potters themselves, who were sometimes blindly copying a model. The Hebrew inscription on one lamp, for example, is written upside down and makes no sense. So few other pieces have inscriptions on them that the ones that do are considered unusual. A tankard painted in blue and red in the ‘Rhodian’ style, for example, is inscribed with an aphorism in Ottoman Turkish that the house of this world is an inn and that whoever does not see it must be mad (Arts of Islam 1976, no. 419). Ceramics were not a major part of the Indian artistic tradition, but the Mughals did collect Chinese porcelains, and those in the imperial collection were often inscribed on the underside with the emperor’s name, date of acquisition and weight of the vessel (Markel 1990).

— 163 —

— CHAPTER 12 —

Textiles

I

nscribed textiles have always played an important role in Islam. Undoubtedly the most famous is the kiswa, the veil that has covered the KaÆba in Mecca since pre-Islamic times. Today the veil is a black cloth with bands of inscriptions from the Koran woven in gold thread, but in the past it could be of virtually any colour, including white, green or even red. Supplying the veil was a great honour, and rulers fought for the right to do so as it was a sign of sovereignty. Inscribing one’s name on the kiswa became an important prerogative of the ruler from early Islamic times. The Mamluk historian al-Maqrizi, for example, preserves an account of a kiswa woven at Tinnis under the ÆAbbasid caliph al-Mahdi in 159/776 (RCEA 44). The kiswa is a unique piece made for a specific situation and now changed annually. It stands in contrast to most textiles, which were made in multiples. Setting up the loom was a time-consuming process, and once the pattern was established, it could easily be repeated. The constant repetition of patterns makes it particularly difficult to date textiles, and the inscriptions on them provide an important method for doing so. Inscriptions could be incorporated in or on textiles in several ways, depending on the fibres and looms used. The easiest was to add the text to the finished piece after weaving. Inscriptions and other decoration could be painted or blockprinted on the fabric, but the most elaborate method was to embroider the inscription on the woven textile. The embroidery technique, indigenous to Asia where silk and weft-faced weaving were typical, was only introduced to Egypt along with silk at the beginning of the ÆAbbasid period.

Inscriptions could also be added during weaving, particularly with tapestry weaving, in which mosaic-like patterns are created by using discontinuous wefts, that is, wefts that go back and forth around only a few warps to create a small coloured area. This technique was suitable for cotton, linen and woollen fabrics woven on tapestry looms. Plain linens, for example, could be decorated with bands and medallions with inscriptions (as well as figural, floral and geometric motifs) tapestry-woven in brightly coloured wool or silk. This technique, already common in pre-Islamic times, was suitable for the loose linen robes worn in Egypt. The robe was woven in the form of a cross with a slit for the head in the centre and tapestry-woven bands on either side of the neck opening and across the top of the sleeves. The garment could be made up without cutting the loom length, and the inscribed or decorated bands would fall in the appropriate places. There are few general studies of the inscriptions on textiles, in part because few scholars have expertise in both textile studies and terminology and in Arabic and epigraphy. The most successful publications are joint endeavours in which a textile specialist teams up with an epigrapher. A good example of this type of work is Otavsky and ÆAbbas’s (1995) catalogue of the medieval Islamic textiles in the Abegg Foundation in Riggisberg outside Berne, Switzerland. The foundation has a wide-ranging collection of textiles, and this catalogue contains linens, cotton and silks dating from the ninth century to the fifteenth. The entry for each piece gives a transcription and

— 164 —

— TEXTILES — translation of the text, and fine colour plates and drawings allow the reader to see the styles of script.

Early Period The two techniques by which inscriptons could be added during weaving – embroidery and tapestry – were used in early Islamic times for the large group of textiles known as tiraz (†iråz). The term is derived from the Persian word ‘to embroider’ (†aråzÈdan) and originally designated any embroidered ornament. In the early Islamic period, textiles were often decorated with inscriptions containing the caliph’s name and titles, and these fabrics were made up into robes of honour worn by the caliph or bestowed by him as official gifts. The term ‘tiraz’ came to refer either to the embroidered or tapestry-woven inscription band or to the fabrics or garments on which these inscription bands are found. To confuse the matter further, the term also referred to the workshops in which these various fabrics were made, in synecdoche for dår al-†iråz (‘factory for tiraz’). In later times, the word ‘tiraz’ was also used in the Arabic-speaking lands for the long bands inscribed with the ruler’s name and titles decorating the façades of important buildings. The long band across the front of the tomb complex of Sultan Qalaæun along the main street in Cairo is a good example of such an inscribed band (see Figures 3.12 and 3.13), and alMaqrizi called the inscription a tiraz. The institution of the tiraz, in which textiles were produced in factories under state supervision, was modelled on Byzantine and Sasanian precedents, although it is not clear whether pre-Islamic workshops were taken over and kept in operation or whether entirely new factories were founded. The institution was already in operation in early Islamic times when pieces inscribed with the caliph’s name were produced under the Umayyads of Syria (r. 661–750). A wool tapestry in the Textile Museum, Washington DC, and a fragmentary silk twill, found in Egypt and now cut up and dispersed in several museums (Day 1952; Arts of Islam 1976, no. 1), both mention Marwan, probably referring to the Umayyad caliph Marwan II

(r. 744–50). The inscription embroidered in yellow along the edge of the silk further states that it was woven in the tiraz of Ifriqiya, suggesting that a factory existed in what is now Tunisia, perhaps at Qabis (modern Gabes). The institution of the tiraz flourished in the Islamic lands under the patronage of the ÆAbbasid and Fatimid caliphs from the late ninth century until the thirteenth when a new taste evolved for patterned silks and all-over embroidered or printed cottons. Most surviving examples of tiraz come from the burial grounds at Fustat (Old Cairo), for these prized and expensive textiles were often used for shrouds, and their inscriptions, offering God’s blessing to the caliph, would impart baraka to the deceased (Sokoly 1997a). A few examples have also been preserved in European church treasuries. Major collections of tiraz fabrics are found in the Museum of Islamic Art in Cairo, which has nearly 1,000 inscribed pieces; the Victoria and Albert Museum, London; and the Textile Museum, Washington DC. The inscriptions on many tiraz textiles are listed in the RCEA. More details are given in Pfister’s publications on some of the pieces in Cairo (1936, 1945–6), Kendrick (1924) on some of the ones in the V&A, and Kühnel [and Bellinger] (1952) on those in the Textile Museum. Studies of tiraz in other collections include Britton (1938) on the pieces in the Boston Museum of Fine Arts, Day (1937) on those in the University of Michigan, Golombek and Gervers (1977) on those in the Royal Ontario Museum, Toronto, Glidden and Thompson (1988, 1989) on those in Dumbarton Oaks, and Otavsky and ÆAbbas (1995) on those in the Abegg Foundation, Riggisberg. General information about the institution is contained in the article ‘Êiråz’ in the 1st edition of the Encyclopaedia of Islam, and Micklewright (1991) raises some general questions about the role of tiraz in medieval times. A 1996 colloquium held at the Abegg Foundation in Riggisberg and published in 1997 as a volume of the Riggisberger Berichte entitled Islamische Textilkunst des Mittelalters: Aktuelle Probleme contains several articles by ÆAbbas, Blair, Sokoly and Schorta about tiraz and the inscriptions on other textiles from early Islamic times.

— 165 —

— ISLAMIC INSCRIPTIONS — The basic work to consult on the subject of tiraz is the fine catalogue of the dated tiraz fabrics in the Textile Museum published in 1952 by Kühnel [and Bellinger]. In addition to the wool tapestry made for the Umayyad caliph Marwan, the catalogue contains pieces made for thirteen ÆAbbasid and seven Fatimid caliphs. The sixteen appendices catalogue many other types of inscribed fabrics, including rush mats, Yemeni ikat cottons, painted and gilded pieces of mulham (a fabric of half-cotton, half-silk), Indian dyed cottons, and so on. Each entry begins with a brief description of the piece with technical information supplied by Louisa Bellinger. This is followed by a description of the inscription, with text given in both Arabic and translation, and several paragraphs of commentary. There is also a long discussion of the technical aspects by Bellinger, describing how the inscriptions were done. Bellinger’s essay on technique is important for epigraphy, for the choice of fabric affected the type of stitch used and the way in which the inscription could be worked (see also the essay by Schorta (1997) in the Riggisberger Berichte). It is difficult, for example, to mark out inscriptions on linen, the traditional fabric woven in Egypt: a wheel with teeth does not penetrate, and ink runs. In order to form the letters of the inscription on a linen fabric, the designer needed to count the threads, and hence he could not use the types of stitches that predominated in the eastern Islamic lands, chain and blanket stitches, since they cover the threads, rendering them impossible to count. Instead, Egyptian weavers used a variety of other stitches, including back and stem stitch, which put more strain on the ground material or embroidery thread and often produced more irregular letters. The plates, charts and notes at the end of Kühnel [and Bellinger]’s catalogue (1952) make the volume all the more useful. More than fifty plates show not only the overall inscriptions but also details of the techniques used. One chart lists the pieces in chronological order so that the reader can follow the evolution of material and spinning, stitches, factory, supervisor and date. Two other charts discuss factories and supervisors.

Studying the inscriptions on tiraz fabrics shows that over the course of several centuries both the titles and the style used for the inscription became more elaborate. A typical piece made for an ÆAbbasid caliph in the late ninth or early tenth century (see Figure 12.70) has one long line of text in a simple form of angular Kufic script. The text begins with the basmala and blessings for the caliph under whose authority the textile was made and continues with the name of the vizier who ordered the piece, the place of production, the name of the supervisor and the date. Variations in wording may indicate different places of manufacture or different rulers. The phrase Æalå yaday (literally ‘under the hands of’), for example, was used on Egyptian pieces before the name of the factory supervisor, while in the eastern Islamic world it preceded the name of the government official who ordered the textile to be made. Tiraz made for the Fatimid caliphs, who were IsmaÆili ShiÆites, often included such sectarian phrases as ‘People of the House’ (ahl albayt) or ‘his pure ancestors’ (åbaæihi al-†åhirÈn). In the tenth century, the scripts used in tiraz inscriptions became more mannered in style, as the shafts of the letters became taller and the tails were floriated. The text was sometimes written in two facing bands. A linen textile made for the Fatimid caliph al-ÆAziz in the Textile Museum (73.432), for example, has a tapestry-woven inscription of large angular script with arabesques forming discs at the heads of the high (85 mm) letters. The text invokes God’s blessing on his servant and friend, the imam al-ÆAziz; it was ordered by the vizier Ibn Kilis in 370/980–1 and probably made at Tinnis. In the eleventh century, tiraz inscriptions became shorter and more decorative with the introduction of non-historical texts that repeated a single word or phrase such as ‘blessing’ (baraka) or ‘good fortune’ (al-yumn). The crescent or shovel-shaped tails of the letters were extended below the base line. In pieces made for the caliph al-Zahir (r. 1020–35), the terminals are still attached correctly to descending letters (see Figure 12.71), but gradually the large endings

— 166 —

— TEXTILES —

12.70 Fragment of a tiraz made for the ÆAbbasid caliph al-Muktafi at Tinnis in 290/902–3. Linen embroidered with silk. Washington

were applied at regular intervals and attached indiscriminately to any letter, thereby obscuring the text and making it more difficult to read. The text on a linen piece woven a century later (Textile Museum 73.461), probably for al-Mustansir (r. 1035–94), is almost impossible to read because of the capriciously disposed shovel-shaped ends of the letters. Readable texts were eventually replaced by a decorative ‘Kufesque’ script (for example, Boston Museum of Fine Arts 30.676). The increasing illegibility of the text was accompanied by an increased use of decorative bands, either smallscale geometric patterns or designs of animals in cartouches. These narrow bands were executed in a variety of colours and generally occur in pairs, with one decorative band framed by inscribed bands and the other decorative band isolated on the ground some distance away (for example, Textile Museum 73.543 made for al-Mustansir).

DC,

Textile Museum 73.639.

In the 12th century the stylised angular scripts were replaced by cursive ones, as on a linen textile woven c. 1100 under the supervision of the vizier al-Afdal (see Figure 12.72). The decorative bands were wider and more elaborate, often containing medallions with animals or geometric motifs, and are themselves framed by narrower bands with inscriptions or pseudo-inscriptions (for example, Royal Ontario Museum 961.107.3). A wide variety of colours was used, but yellowgold often predominates, perhaps in imitation of luxury textiles that used gold-wrapped thread. Literary texts describing court vestments and furnishings speak primarily of silks, but very few pieces of this type survive. One exception (Ragib 1980) is a silk in the Islamic Museum, Cairo, of the type known as muhraq, a white silk soaked in gum, smoothed, and polished with a shell called muhra. Originally measuring some 110 by 70 cm, the bright white piece is embroidered in black

— 167 —

— ISLAMIC INSCRIPTIONS —

12.71 Fragment of a tiraz made for the Fatimid caliph al-Zahir (r. 1020–36). Linen tapestry-woven with silk and gold thread. 16.7 x 19.6 cm. The Cleveland Museum of Art, Andrew R. and Martha Holden Jennings Fund, 1982.109.

with seventeen lines of text. A magnificent band of simple Kufic at the top has Koranic texts (a paraphrase of 11:88 and 9:129) and good wishes. The sixteen lines of so-called ‘eastern’ or ‘broken’ Kufic below contain a marriage contract between two Fatimid aristocrats, an aged shaykh and the middle-aged daughter of an amir. The piece can be dated to the mid-eleventh century as the text mentions the Fatimid caliph al-Mustansir. Most surviving tiraz pieces, by contrast, are linens and other lightweight fabrics, mainly fragments from mantles, summer outfits, undergarments, turbans, shawls, sashes, napkins, presentation towels, and furnishing fabrics like curtains. The only complete garment to survive is known as the Veil of Ste Anne, so-called because

it has been preserved as a relic in the Church of Ste Anne at Apt in France (RCEA 2864; Marçais and Wiet 1934; Arts of Islam 1976, no. 8). A length of bleached linen measuring 3.1 by 1.5 metres and woven in tabby, it is decorated with three parallel bands of ornament tapestry-woven in coloured silk and gold thread. The centre band has three medallions with sphinxes encircled by an inscription in Kufic script naming the patron, the Fatimid caliph al-MustaÆli, and the supervisor of the work, his vizier al-Afdal (the same person for whom the tiraz illustrated in Figure 12.72 was made). The bands along either side are decorated with birds, animals and another inscription stating that the textile was woven in 489/1096 or 490/1097 (the last digit is not clear) at the royal factory of

— 168 —

— TEXTILES —

12.72 Fragment of a tiraz made under the supervision of the Fatimid vizier al-Afdal c. 500/1106. Linen tapestry-woven with silk. Washington

Damietta, a well-known textile manufactory in the Nile Delta. The piece was probably worn as an overgarment or mantle, with the small slits at the sides for the hands and the central band containing the inscribed medallions falling down the back. Since both the lord and bishop of Apt took part in the First Crusade in 1099, the cloth was probably brought from Egypt or Syria as plunder and preserved as a relic. In addition to these tiraz pieces made for the ÆAbbasids and Fatimids, another well-known group of inscribed fabrics from the early Islamic period comprises ikat cottons made in the Yemen. In the ikat technique, the warp and sometimes the weft threads are tie-dyed to produce patterns when woven. Ikat weaving was a speciality in Central Asia and the Yemen, where it had been introduced across the Indian Ocean from the Indian subcontinent, and a series of spectacular cotton ikats was produced at SanÆaæ and other cities there in the ninth and tenth centuries. These fabrics have characteristic variegated stripes produced by dyeing parts of the natural-colour cotton warps blue and brown before weaving and then weaving these dyed warps with natural cotton wefts. Yemeni ikats can be divided into two groups, based on the types of inscriptions on them (Golombek and Gervers 1977; Blair 1992a: 68–9). One group is decorated with inscriptions

DC,

Textile Museum 73.680.

embroidered in an interlaced Kufic script, with numerous arcs inserted for decorative purposes in the base line of the inscription and in the vertical staffs. Dated pieces range from 270/883 to 310/923. The other group of inscribed Yemeni ikats have inscriptions painted and gilded in elaborately plaited and foliated Kufic. Three pieces, dating from the late tenth century, are inscribed with the names of imams amd amirs of the Yemen. Legibility was important, and on the pieces with historical inscriptions, interlacing was restricted to single letters and decoration was restrained. In contrast, the inscriptions on other ikats with pious texts are much more elaborate, with an extraordinary number of arcs, bumps, triangles, curls and other decorative devices inserted between and around the letters Extremely laborious to read, these pious texts are legible only because of their limited repertory of phrases. This piece in the Textile Museum (see Figure 12.73), for example, is inscribed with God’s blessing on the Prophet ([ßalå a]llåh Æala mu˙ammad). Most of the painted and gilded decoration comprises triangles inserted between the two låms of allåh and the interlacing of the dål in mu˙ammad. Tiraz were also woven for the Umayyads of Spain (see, most recently, the entries in Dodds 1992 with bibliographies and colour illustrations). The earliest is the tapestry-woven fragment known

— 169 —

— ISLAMIC INSCRIPTIONS —

12.73 Fragment of a cotton ikat with painted and gilded inscription offering God’s blessing on Muhammad. Yemen, tenth century. Washington

as the Veil of Hisham II (r. 976–1013), which was found in 1853 wrapped around a reliquary in the church of Santa María del Rivero of San Estaban de Gormaz in Soria (RCEA 2124; Dodds 1992, no. 21). It is thought to have been one of the ends of the caliph’s veil or headdress which was wrapped like a turban with the ends falling loosely to the arms. Its technique and epigraphy are typically Spanish: the gold wefts are gilded strips of membrane Z-wrapped around a silk core, and the half-palmettes on the Kufic letters resemble those found on contemporary ivories and mosaics. A similar technique and epigraphic style are used on another eleventh-century textile discovered in 1978 in the Colls church in Puente de Montañana in Huesca (Dodds, 1992, no. 22).

Medieval Silks In addition to embroidery and tapestry weaving, it is also possible to incorporate an inscription while weaving a fabric, but this is a much more difficult process, especially when using a drawloom. The drawloom, which has a complicated system of drawstrings that raise and lower the pattern

DC,

Textile Museum 73.377.

warps, had been developed in ancient China in the first centuries CE. It spread to Iran in preIslamic times and was eventually adopted in all the Islamic lands where silk was woven. Since the drawloom had fine strings instead of the older cumbersome system of shafts, it allowed larger quantities of finer fabrics to be woven with less effort. It was, however, extremely laborious to set up and was used for sumptuous patterned fabrics with compound weaves, often worked with gold thread, and repeated in different combinations. Most early Persian silks did not have inscriptions, as it was difficult for the weaver to coordinate the running text with a repeating pattern. Pardoxically, it was an inscription inked on the back of one piece that allowed Shepherd and Henning (1959) to identify and localise one group of these brightly coloured silks with roundels enclosing paired birds and other motifs. The silk textile was kept in the church treasury at Huy in Belgium among the relics of St Domitian, who died in the sixth century and was canonised in the eighth and whose relics were assembled in the twelfth century. A note in Soghdian script states that the cloth was made at Zandana, a

— 170 —

— TEXTILES — town near Bukhara, and the group is therefore known as Zandaniji silks. The writing is similar to other inscriptions discovered at the Soghdian citadel of Mt Mug in Central Asia, destroyed by the invading Arabs in 722, and the group of silk textiles is therefore dated to the late seventh or the eighth century, when Islam was first brought to the area. By the tenth century, Persian weavers had figured out how to incorporate long inscription bands on their elaborately patterned drawloom silks. One of the finest examples to be preserved is the shroud of St Josse (see Figure 1.1),woven in the mid-tenth century probably in eastern Iran. Silk textiles woven on a drawloom using a similar technique but different designs were also produced at this time in western Iran. They are known, somewhat infamously, as the ‘Buyid silks’. The saga began when several silk fragments and other objects were clandestinely excavated in

late 1924 and early 1925 at a site known as Bibi Shahr Banu near Rayy, south of Tehran in Iran. Other similar textiles appeared on the market in the 1920s and 1930s, and many were acquired by major museums and private collectors. The extraordinary artistic and technical importance of these medieval silks was quickly realised, and they were the subject of many important exhibitions, publications and symposia. Following the 1931 International Exhibition of Persian Art at the Royal Academy in London, Wiet (1933, no. 3; RCEA 2177) identified one piece, a silk that had been made up into a tunic (see Figure 12.74), as inscribed with the name of the Buyid ruler Bahaæ al-Dawla (r. 998–1012), and the group became known as the ‘Buyid silks’. More and larger pieces, even complete palls woven to shape, appeared in the 1940s and 1950s, and many were acquired by the Cleveland Museum of Art and the Abegg Foundation in

12.74 Silk cloth inscribed with the name and titles of the Buyid ruler Bahaæ al-Dawla. Width 2.9 m. Mesopotamia or western Iran, c. 1000. Washington

— 171 —

DC,

Textile Museum 3.116.

— ISLAMIC INSCRIPTIONS — Riggisberg. At the same time, more and more doubts began to surface about the authenticity of these textiles. After Wiet (1948) had published eighteen more silks from an anonymous collection, Florence Day (1951) wrote a particularly vitriolic review of Wiet’s book, challenging virtually everything that he had written about the silks. The controversy continued into the 1970s. The Abegg Foundation published the so-called Riggisberg Report, nos 37–8 of the Bulletin de Liaison du Centre International d’Étude des Textiles Anciens (often abbreviated as CIETA), in which doubts were expressed about many of the thirty-nine ‘Buyid’ silks in their collection. In the next issue of the Bulletin (nos 39–40, 1974), Dorothy Shepherd countered with a spirited defence of the Buyid silks. Inscriptions played a key role in determining the authenticity of these problematic pieces in the 1990s (Blair et al. 1992). The Cleveland Museum requested a report on the inscriptions on the ‘Buyid’ textiles, including thirty-seven inscribed pieces in their collection and eighty-four in other collections. Individually, the inscriptions did not seem exceptional, but reviewing the inscriptions as a whole raised serious doubts, for only a handful, mainly the pieces associated with the original discoveries at Rayy in 1924–5, conformed to titulature and epigraphic forms standard in the period. A first step was to compare the names and titles used in the textile inscriptions with those used elsewhere, mainly in architecture since so few other inscribed objects from the period have unimpeachable credentials. Such a comparison showed that the eight textiles naming Buyid amirs and other known historical figures, including the shroud of St Josse naming Abu Mansur Bakhtikin (see Figure 1.1) and the silk naming Bahaæ al-Dawla (see Figure 12.74), followed the same protocol used in monumental inscriptions. Most of these pieces had long-established provenance. In contrast to these well-known individuals on textiles with an established provenance, none of the other people named on the ‘Buyid’ silks could be identified, although they bore significant titles such as exalted amir (al-amÈr al-ajall). Their

titles were unusual, sometimes unprecedented at such an early date. Moreover, the same or very similar names occurred on several pieces. For example, palls in Cleveland and the Textile Museum were made for people with the same name, father’s name and grandfather’s name (ÆAli b. Muhammad b. Shahrazad) but with a different great-grandfather (Rustam and Rastawayh). It was also possible to distinguish the traditional types of texts used on well-attested pieces (and on contemporary objects) – generalised formulas of good wishes to the owner – from the unusual types of texts on the other ‘Buyid’ silks, especially poems in Arabic lamenting death and asking God’s forgiveness drawn from a range of sources. Three other textiles (see Figure 12.75) inscribed with the same, very early date (203/818–19) stuck out by the unusual wealth of information inscribed on them, such as the names of the patron and weaver and place of manufacture. This wealth of information, moreover, always had some peculiarity when compared to monumental texts of the same period: it included terms or names (for example, Baghdad instead of Madinat al-Salam) not attested until much later. Studying the inscriptions on the ‘Buyid’ silks raised such grave doubts about their provenance that it was decided to test these textiles further by the method of carbon-14 analysis developed in the 1970s using an accelerator mass spectrometer. Radiocarbon dating confirmed the doubts raised by epigraphic analysis and showed that only a handful of textiles, mainly but not exclusively those associated with the 1924–5 finds at Rayy and with inscriptions naming known historical figures, are medieval in date (between the tenth and the twelfth centuries). Most are not, and are forgeries produced in the early twentieth century just after the initial finds. Some, including those with the most aberrant and over-informative inscriptions, such as one of the three supposedly made at Baghdad in 203/ 818–19, were even made after 1950 since they contain radioactive isotopes produced by atomic bombs and tests. The study of the Buyid silks has several important implications for epigraphic study. It

— 172 —

— TEXTILES —

12.75 Front of a lampas-woven silk tunic inscribed with good wishes to the amir Abu’l-Raæis Khalaf b. Mansur al-Tawaghi, the date 203/818–19, and the name of the city, Baghdad. 113.7 x 86.7 cm. Post-1950. The Cleveland Museum of Art, Gift of the Junior Council of the Cleveland Museum of Art, 1982.23.

— 173 —

— ISLAMIC INSCRIPTIONS — confirmed that inscriptions on textiles and other portable objects from this early period followed the standard forms used on architecture. It also showed that establishing a corpus of inscriptions was necessary to show up significant variants. Finally, the study made it clear that while epigraphic analysis was suggestive, it needed to be combined with other analytical techniques, including technical and stylistic analysis, to produce definitive conclusions. Copying inscribed textiles is not only a modern phenomenon; it also occurred in medieval times, when the prestige and beauty of these drawloom silks woven in the eastern Islamic lands led to their being copied in the western Islamic lands, even down to the inscriptions. The most famous copies are known as the ‘Baghdad’ silks, and a similar combination of epigraphic and technical analysis was used successfully to distinguish medieval copies from their models and assign the copies to the correct place of manufacture. One silk made into the shroud of San Pedro de Osma (d. 1109) in the cathedral of Burgo de Osma in Spain (see Figure 12.76) has large roundels with human-headed lions and small roundels with mirror inscriptions repeating the sentence ‘This is among the things made at

Baghdad, may God protect it’ (hådhå mimmå Æumila bi-mad/Ènat al-baghdåd ˙arasahå allåh). Day (1954) showed that the spelling of hådhå with two long alifs is peculiar to the western Islamic lands and confirmed the Spanish attribution for the whole group of these textiles woven in a distinctive type of lampas weave. The inscription on another textile (Shepherd 1957; Ettinghausen and Grabar 1987, fig. 138) made up into a chasuble, or ecclesiastical mantle, for the Spanish saint San Juan de Ortega (d. 1163) names the Almoravid ruler ÆAli b. Yusuf (r. 1106–42). Using the inscriptions on these pieces, it is possible to attribute some fifty examples of these sumptuous silk and gold textiles to Almoravid Spain (1056–1147). The fact that they were immediately used in burials for Christian saints shows how valuable they were in medieval times. Inscriptions are a reliable method of identifying textiles made for the Almoravids’ successors, the Almohads (r. 1130–1269), although far fewer pieces have survived, probably because of the cultural and religious conservatism of the dynasty. Like earlier Spanish textiles, these too were cherished in medieval times and were preserved in Christian burials. They are almost all tapestrywoven and generally show less figural decoration,

12.76 Drawing of the inscription on the ‘Baghdad silk’ made into the shroud of San Pedro de Osma. Spain, late eleventh century. Boston Museum of Fine Arts 33.371.

— 174 —

— TEXTILES — with interlacing replacing the figural roundels used on earlier pieces. Some have a large band with a single word or phrase, such as al-yumn (‘good fortune’) or al-baqåæ lillåh (‘everlastingness [belongs] to God’), repeated in a stylised Kufic script with paired verticals ending in palmettes. Another type of script common on Almohad textiles is a fluid cursive that shows many of the special features typical of the distinctive script known as maghribÈ and used in the western Islamic lands. These characteristics include a hook or curve at the top of shafts, a hook or spur at the bottom of the final alif, sloping shafts in the letters †åæ and kåf, and a single dot below or above to distinguish the letters fåæ and qåf. These cursive inscriptions often repeat pious phrases. The band on a silk-and-gold tapestry made into a cap for the infante Fernando of Castile, son of Alfonso VIII, founder of the Monastery of Santa María le Real de Huelgas (Dodd 1992: 110, fig. 4), for example, bears the phrase that Gómez-Moreno translated as ‘In the lord is our solace’. The same fluid script was used on larger pieces to form a border around a field decorated with a strapwork pattern. Occasionally the cursive inscription bands on Almohad textiles contain a longer text. The most famous example is the splendid banner preserved, like many other Spanish textiles, in the monastery of Las Huelgas (Dodds 1992, no. 92). It is popularly known as the Las Navas de Tolosa Banner because it was thought to have been captured in 1212 at the decisive battle when Alfonso VIII defeated the Almohad sultan al-Nasir, but the banner was probably a trophy won on campaign by king Ferdinand III (d. 1252), who gave it to the monastery during restoration. A band across the top invokes God’s blessing, and bands around the field contain Koran 61:10–12. Such banners were often carried on parade; banners inscribed with pious invocations and Koranic texts are depicted in one such parade celebrating the end of Ramadan shown in a painting from a celebrated copy of alHariri’s Maqåmåt in Paris (Bibliothèque nationale, MS arabe 5847, fol. 19r; colour illustration in Ettinghausen 1962: 118) painted by Yahya b. Mahmud al-Wasiti in ÆIraq in 634/1237. The

tenor of the Koranic inscriptions on the Las Navas banner, which mention the paradisal rewards for those who struggle on God’s path, however, suggests that it was meant to be carried on campaign, and many identical pieces could have been woven in the same pattern. Inscriptions also provide the starting point to identify the only group of textiles definitely made in medieval Sicily, the set of coronation robes used by the Holy Roman Emperor and Empress and now in the Schatzkammer in Vienna. The most famous piece is the stupendous mantle made for the Norman ruler of Sicily, Roger II Hauteville (r. 1130–54). A great semicircle of red silk, it is embroidered in gold thread and pearls with a central tree flanked by addorsed lions attacking camels. An inscription in Arabic embroidered in pearls and gold along the hem states that it was made in the royal workshop at the capital of Sicily [Palermo] in 528/1133–4 (RCEA 3058; Curatola 1993, no. 95). The set also includes an alb made for William II Hauteville (r. 1166–89), with a bilingual inscription in Latin and Arabic also embroidered in pearls and gold stating that it was made in 577/1181 at Palermo (mentioned in RCEA 3364ter.).

Later Period As the taste developed for all-over patterns, inscriptions played a diminishing role in textiles woven in later Islamic times, but a few rare pieces with historical inscriptions have been used to establish the provenance of a piece or group of textiles. For example, a silk with roundels enclosing paired lions made up into a chasuble (RCEA 4135; Arts of Islam 1976, no. 13; Ettinghausen and Grabar 1987, fig. 401) can be attributed to Saljuq Anatolia since a large inscription band along the edge bears the name and titles of Kayqubad I of Konya (r. 1219–37). Similarly, a silk made up into the burial robe of Duke Rudolf IV of Austria, who died in 1365 (RCEA 5701; Blair and Bloom 1994, fig. 23), can be attributed to Ilkhanid Iran, probably Tabriz, in the early fourteenth century, since it bears the name and titles that the Ilkhanid ruler Abu SaÆid (r. 1317–35) assumed

— 175 —

— ISLAMIC INSCRIPTIONS — after 1319. Its clear provenance allows a group of related silk and gold lampas textiles to be attributed to Iran in the late thirteenth and fourteenth centuries (Wardwell 1988–9). The earliest textile attributable to the Ottomans (Ettinghausen 1961) is a striped silk in the monastery of Studenica, Bosnia, with the name Bayazid Khan, probably referring to Bayazid I (r. 1389–1402). Reading the name on the textile is, however, not always sufficient grounds for attributing provenance, for the person named on the textile may be the recipient, and the piece woven elsewhere and intended as a gift. For example, a lampas fabric of silk and gold with paired falcons made into a cope, or long ecclesiastical cape, and preserved in the Marienkirche in Danzig (RCEA 5872; Arts of Islam 1976, no. 15), shows duodecagons enclosing paired birds whose wings are inscribed ‘Glory to our lord the just and sagacious sultan, Nasir …’. The inscription is thought to refer to the Mamluk sultan al-Nasir Muhammad b. Qalaæun (r. 1293–1341 with interruptions), but the piece may have been woven in Central Asia, since it shows technical characteristics of silks woven there. The problem of using titles to attribute textiles is particularly vexing with the Mamluks, for they drew their titles from a limited range, and some were honorific or generic. Recent studies (for example, Mackie 1984) found only a handful of pieces with the names of Mamluk sultans who could be identified precisely. They all belonged to early members of the sequence (Qalaæun, alAshraf Khalil and Muhammad b. Qalaæun), and even some of these few pieces may have been gifts as they use the Persian rather than the Arabic form of the name (Muhammad Qalaæun without an ibn). Another handful of textiles is inscribed al-malik al-nåßir, a title used by four sultans in the fourteenth century, although the silks are generally attributed to the reign of alNasir Muhammad b. Qalaæun. A striped silk inscribed Æizz li-mawlånå al-sul†ån al-malik almuæayyad (‘Glory to our lord, the sultan, the king, the one who is assisted [by God]’) seems on stylistic grounds to be earlier than the reigns of the two fifteenth-century Mamluk sultans who

bore that title, and was probably intended for the Rasulid sultan of Yemen, Muæayyad Daæud (r. 1296–1322). Other textiles are inscribed al-sul†ån al-malik al-muΩaffar and al-malik al-ashraf, titles borne by five or six sultans. Hence titulature alone is insufficient to associate most Mamluk textiles with particular sultans, and one must use inscriptions alongside technical considerations and stylistic developments seen in the other arts. Similarly for textiles woven in Spain in the later period, epigraphy has to be combined with stylistic and technical comparisons, for most of the pieces are inscribed only with short phrases. One group woven in gold and silk thread comprises Spanish variations of the striped and inscribed silks produced for the Mamluks. The inscriptions repeat the phrase Æizz li-mawlånå al-sul†ån used on the Mamluk silks, but the florid style of cursive script resembles that used earlier on pieces made for the Almohads, and the technique and decorative motifs are also Spanish. Other Spanish silks woven in the fourteenth and fifteenth centuries have strapwork patterns combined with inscriptions in angular and cursive script similar to those done in tile and stucco at the Alhambra palace in Granada (see Figure 7.42). The wilful juxtaposition of several different styles of script can be seen on the finest piece from the Nasrid period (see Figure 12.77), a splendid silk curtain now in Cleveland (good details published in Wardwell 1983). Cartouches along the central band are inscribed in naskh with wa lå ghålib ilåællåh (‘There is no victor save God’), the motto that the Nasrid dynasty (r. 1230–1492) had inscribed in the same way on many of its commissions. The cartouches are set in small epigraphic arches composed of the word al-yumn (‘good fortune’) written in mirror reverse in Kufic with interlaced stems. Bands across the top and bottom of the curtain repeat the phrase al-mulk lillåh (‘dominion [belongs] to God’) in a stylised Kufic with interlaced stems. Cartouches in the side panels contain the fuller phrase al-mulk lillåh wa˙dahË (‘dominion [belongs] to God alone’) in a stylised naskh. In all cases, the desire for symmetry was more important

— 176 —

— TEXTILES —

12.77 Lampas-woven silk curtain made for the Nasrids. 4.38 x 2.72 m. Granada, fifteenth century. The Cleveland Museum of Art, Leonard C. Hanna Jr. Bequest. 1982.16.

— 177 —

— ISLAMIC INSCRIPTIONS — than legibility, and letters are misshapen or misjoined. In the side panels, for example, the person who set up the loom (who may not have been the same person who designed the inscription) sandwiched the lillåh between the elongated stems of the alif and låms in al-mulk to form a symmetrical pattern. The final kåf, which should correctly be attached to the rest of the word almulk, hovers above the tail of the låm. Epigraphic information can still be instructive in assigning provenance, even when the inscriptions contain only generalised good wishes or even pseudo-inscriptions. There is, for example, a group of sumptuous gold lampas-weave silks that has recently appeared on the art market and is said to have come from Tibetan monasteries. On stylistic and iconographic grounds, they have been attributed to various places, ranging from Saljuq Anatolia (Wenzel 1990) to Central Asia (Wardwell 1992; Folsach and Bernsted 1993), but the style of pseudo-Kufic inscriptions with animal heads and interlaced verticals on some pieces (see Figure 12.78) lends weight to an attribution to the eastern Islamic lands, where both these epigraphic features are known. Despite the generic nature of inscriptions on many textiles made in the later period, a few

were specific commissions, and the historical references in their inscriptions help to define distinctive schools of textile production. Good examples are the two tapestry-woven banners captured at the battle of Salado in 741/1340 and preserved in Toledo Cathedral: one dated Muharram [7]12/May 1312 (RCEA 5302) was made in Fez for the Marinid sultan Abu SaÆid ÆUthman (r. 1310–31), and another dated Jumada II [7]40/ December 1339 (RCEA 5789) was made for his successor Abuæl-Hasan ÆAli (r. 1331–48). Both show the same distinctive maghribÈ style of script seen on the Las Navas banner, but the pious and Koranic inscriptions have been replaced by specific historical texts with the sultan’s name for the same apotropaic function. Banners continued to be made in North Africa in later centuries, and the distinctive maghribÈ script allows them to be distinguished from their Ottoman counterparts. The inscription tells us, for example, that a large maroon banner woven in 1094/1683 (Harvard University Art Museums 1958.20; Blair and Bloom 1994, fig. 331) was made to be carried on the pilgrimage to Mecca by members of the Qadiriyya, a Sufi order which flourished in North Africa, and the distinctive epigraphic style, with low sweeping curves,

12.78 Textile with felines, eagles and pseudo-inscription in Kufic with interlaced stems. Lampas weave, silk and gold thread. 1.70 x 1.09 m. Central Asia, thirteenth century. The Cleveland Museum of Art, Purchase from the J. H. Wade Fund, 1990.2.

— 178 —

— TEXTILES — confirms that the banner was made in North Africa, probably Morocco. The Moroccan banner belongs to the standard type of Ottoman banner (Turkish sanjak) that is shield-shaped and often decorated with a depiction of ÆAli’s legendary two-bladed sword, Dhuæl-fiqar (Denny 1974). A famous example in the Topkapı Palace, also maroon-coloured, is ascribed, quite unreliably, to the Ottoman sultan Selim I (r. 1512–20). Many others fell to Polish nobles in the sack of the Ottoman camp after the breaking of the second siege of Vienna in 1683 and are now in the Wawel Museum in Krakow. Most of them are inscribed with Koranic passages about holy war and the righteousness of the cause of true believers. One other kind of Ottoman textile was commonly decorated with inscriptions: cenotaph covers (Arts of Islam 1976, nos 32–3; Denny 1972). Made of red or green satin, they often have zigzag bands of brocaded white silk inscribed with the profession of faith in large script, prayers and invocations of the names of God and the Prophet. They also contain Koran 2:144, a verse that mentions facing the qibla, and an appropriate choice since the deceased was buried facing Mecca. By contrast, most Ottoman silks and velvets used for kaftans, pillows and other dress and furnishings were anepigraphic. Inscribed tomb covers were also made for the Safavids in Iran (Arts of Islam 1976, nos 80–2). They are typically decorated with multiple bands of inscriptions, some with incantations in Arabic, and others with historical inscriptions with the names of the donor, composer and calligrapher as well as the date. Intended for donation to the major ShiÆite shrines, they often bear the names of the Twelve Imams. One tomb cover endowed to the shrine of Shaykh Safi at Ardabil has small repeating inscriptions in square Kufic with the name Ghiyath. Ackerman (1934) took this to refer to Ghiyath al-Din ÆAli, a weaver from Yazd whose biography is given in the local seventeenth-century history of Yazd, JåmiÆ-i MufÈdÈ. Grandson of the calligrapher Kamal al-Din, Ghiyath al-Din became a favourite courtier of the Safavid Shah ÆAbbas before retiring to Yazd.

The name Ghiyath appears on several Safavid textiles woven in widely varying techniques, including velvet, satin lampas and double cloth. It is unlikely that the same person wove all these disparate pieces and more probable that he was the designer or head of the workshop or even that his name became a device or trademark added to many pieces. The same is true of the word Æabdallåh, divided in half and written in Kufic on an eclectic group of three figural silk textiles, including a satin lampas, a velvet and a double cloth in a distinctly Indian style (Bier 1987, no. 22). Inscribed silk banners were probably also made in Iran, and at least one late example made for the Qajars (Arts of Islam 1976, no. 91) is decorated with repeating bands of Koranic verses (67:51–2) and a depiction of the lion and sun. Otherwise, inscriptions were rare on most later Iranian textiles. One of the virtuoso exceptions is a red and white silk double cloth dispersed among many collections (Bier 1987, no. 25). The textile is woven so that each side shows rectangular compartments, some with figural and architectural scenes, others with polylobed cartouches inscribed with poetry, in opposite colours. The text, written in mirror reverse, is a Persian quatrain about beautiful cloth spun from the soul that plays on the value of the medium. Poetic verses, historical information or dedications were also added to printed cottons (Persian qalamkår) produced in Iran from the late seventeenth or early eighteenth century by hand-painting with a brush.

Carpets and Other Woven Pieces Inscriptions could also be incorporated in carpet designs during knotting or weaving. Pseudoinscriptions in Kufic with knotted stems appear on several types of Anatolian carpets. The earliest are the so-called Konya carpets, now attributed to the fourteenth century (see, for example, Blair and Bloom 1994, fig. 188), and similar carpets with pseudo-Kufic or Kufesque borders are depicted in contemporary Persian book-paintings. The same kinds of pseudo-epigraphic borders are also found on fifteenth- and sixteenth-century Anatolian carpets, including both large- and small-patterned

— 179 —

— ISLAMIC INSCRIPTIONS — Holbeins, so called because they are depicted in the work of Hans Holbein II (1497–1543), and the type known as re-entrant, keyhole or Bellini carpets, after the one that is depicted in Gentile Bellini’s Virgin and Child Enthroned in the National Gallery in London. With the introduction of silk and fine knotting in Safavid times, it was possible to knot extraordinarily complex patterns with amazing detail, including inscriptions. Several signed and dated carpets survive from the Safavid period and provide much of the basis for dating a large group of carpets to the mid-sixteenth century. The most famous is the impressive pair known as the Ardabil carpets, which were probably made for the shrine of Shaykh Safi at Ardabil. There is a well-preserved one in the V&A, London (Blair and Bloom 1994, fig. 214), and a patched one in the Los Angeles County Museum. Although they are said to differ in knot count, texture and pile length, both are knotted in the same technique and show the same design with the same cartouche at one end. The first two lines contain a couplet from an ode by Hafiz, saying that ‘I have no refuge in the world other than thy threshold; My head has no resting place other than this doorway’ (juz åstån-i tu-am dar jahån panåh-È nÈst / sar-i marå bi-juz Èn dar ˙awåla-gåh-i nÈst). A third line contains the signature ‘Work of a servant of the court, Maqsud of Kashan’ (Æamal-i banda-yi dargåh maqsËd kåshånÈ), and the date is written below: ‘the year 946’, corresponding to 1539–40 (Ittig 1993c). It is inconceivable that Maqsud alone tied the 60,000,000 knots in these two carpets. Rather, he probably drew up the designs for the paper cartoons from which these elegant carpets were woven at different scales. On the basis of his title, King (1996) suggested that he might have been the acting superintendent of the shrine. Maqsud must have selected the couplet from Hafiz deliberately to suit the intended location of this sumptuous pair. Based on their reconstructed size (9 x 5.30 m), King (1996) has suggested that they would have been laid side by side in the large octagonal hall known as the Jannat Sarat. The inscription supports his suggestion. The reference to threshold (åstån) would

have been appropriate for the Safavid dynastic shrine, which was often called a threshold. When the carpets were laid in the hall, the cartouches would have fallen at one edge, presumably the south-west side facing the main doorway to the courtyard and the qibla. The quotation from Hafiz, with its references to refuge (Persian panåh), resting place (˙awåla-gåh) and doorway (dar), was obviously chosen as appropriate for carpets designed for the Safavid dynastic shrine and would have been especially suitable should the hall have served, as is often imagined, as Tahmasp’s intended tomb. Another signed and dated Safavid carpet is a smaller medallion carpet with hunting scenes in the Poldi Pezzoli Museum, Milan (Arts of Islam 1976, no. 58). In the centre is a small cartouche inscribed with the name Ghiyath al-Din Jami and the date 949 (1542–3), sometimes read as 929/1522–3. The centrifugal design, with mounted and turbaned figures fighting lions, deer and other animals, suggests that the royal throne was meant to be set at the centre of the carpet, thereby covering the cartouche with the signature and date and underscoring the humility of the artist. Inscriptions were also popular on the largescale pictorial carpets introduced in Iran in the mid-nineteenth century. They often had borders with graceful calligraphy surrounding scenes from such literary works as Firdawsi’s Shåhnåma and Nizami’s Khamsa. It is also possible to include an inscription when making a flat-weave, as on the distinctive type of reversible cotton flat-weave known in Persian as zÈlË. Sets of these flat-weaves were often given as endowments to mosques and shrines in the drier regions of Iran, particularly around Yazd, and the band of inscription along the edge often contains details about the manufacture and endowment. The most famous is a set of thirteen made by Shams Qutb al-din Maybudi and given by Khanesh Begum in Ramadan 963/ July 1556 to the congregational mosque in the Nuriyya khånaqåh in Taft, a village south-east of Yazd (Afshar, 1993; Ittig 1993b). Some of the flat-woven mats or fans made of rush or hemp and commonly used in the Levant

— 180 —

— TEXTILES — and North Africa are also inscribed in Kufic. They usually contain generalised good wishes, but a few have blessings for a specific person. One fragment in Stockholm has good wishes to a specific person who may have been an unidentified ÆAbbasid caliph (Lamm 1938, no. 21), and a fan in the Abegg Foundation, Riggisberg (Otavsky and ÆAbbas 1995, no. 11) invokes blessing, good fortune and joy to Abuæl-Husayn. He can be identified as the Ikhshidid ruler of Egypt (r. 961–7), and his name is found on another piece in the Islamic Museum, Cairo (no. 1485/2). Hence many contemporary pieces (for example, a plaited rush mat in the Textile Museum, Washington 73.678, and a woven hemp and straw mat once in the alSabah Collection illustrated in Jenkins 1983: 43) are similarly dated to the tenth century.

The technique of weaving the text can make it difficult to read the inscription on a flat-weave, particularly on a single surviving example. Often the piece is fragmentary and the inscription incomplete. Words are usually unpointed, and the nisba can be read in several ways. A zÈlË in the Hermitage (Masterpieces of Islamic Art 1990, no. 61), for example, is signed by Ustad ÆAli Nushabadi and the date ends with the month Ramadan without giving the year. Dates can sometimes be read in different ways by looking at the reverse side. Afshar and Ittig (1993), for example, have argued that the date on an inscribed zÈlË fragment discovered in the congregational mosque at Maybud, signed by ÆAli Baydak b. Hajji Maybudi and customarily read as RabiÆ II 808/October 1405, should be revised to 1188/1774.

— 181 —

— CHAPTER 13 —

Other Portable Arts Glass Glasswares were produced throughout the Islamic lands from earliest times onwards. Syria and Egypt, traditional centres of glassmaking in pre-Islamic times, continued to be important centres of production in Islamic times. Museums there, particularly the Damascus Museum and the Islamic Museum, Cairo, contain some of the most important pieces of inscribed glass, as does the Corning Museum of Glass in New York State. These collections have not been published in their entirety, and often the inscriptions on the pieces are unread or unreadable. Much of the work on inscribed glass is done in exhibition catalogues, such as the 1976 exhibition at the Hayward Gallery, Arts of Islam, and Atıl’s 1981 catalogue on the art of the Mamluks. Glassmakers in the Islamic lands inherited a variety of techniques to decorate glass objects, and some of them, such as incising and relief carving, could be adapted to include inscriptions. Typically, these short passages in Arabic offered good wishes to the owner. A stemmed goblet in the Metropolitan (Jenkins 1986, no. 16), for example, has a band of Kufic incised around sides with ‘Blessing from God to its owner’ (baraka min allåh li-ßå˙ibihi). A similar text is incised on a cup in the David Collection, Copenhagen, and several examples with incised inscriptions excavated at Raqqa, Syria, are in the National Museum, Damascus. The same sorts of blessings are carved in relief on several bowls (Arts of Islam 1979, nos 129–30), but the relief technique makes it more difficult to shape the letters and the texts are very short and sometimes unreadable.

Most of these objects can only be dated approximately on stylistic grounds, mainly to the ninth or tenth centuries. Only recently have fragments of glass been recovered from controlled archaeological contexts, and even then the pieces may well have been imported. Otherwise, localisation depends on stylistic criteria, and the range of attributions is broad, so that the same piece, such as a relief-cut bowl in the Islamic Museum, Cairo, with a simple text of blessing, may be attributed to tenth-century Egypt (Arts of Islam 1976, no. 130) or Iran (Ettinghausen and Grabar 1987, fig. 245). Because of the identification of Egypt as a major centre for the production of rock crystal, much early glass was traditionally assigned to there, but recent excavations and publications have uncovered more material from Iran, which is increasingly recognised as another important centre in early Islamic times (Kröger 1995). Stamping allowed for more precise texts. Glass weights used to calibrate metal coins were the most common variety of stamped glass. In a business transaction, coins were weighed in bulk against glass weights, which could not be altered. Glass weights were produced in Egypt in large numbers from the eighth century to the fifteenth and were typically inscribed with the name of governors or officials. The earliest known examples bear the name of ÆAbdallah b. ÆAbd alMalik, who was governor and finance director of Egypt from 86/705 to 90/709. The latest weights date from the very end of the fifteenth century. Most museums have large collections of weights, which are important evidence for economic history. Those in the British Museum, for example,

— 182 —

— O T H E R P O RTA B L E A RT S — were catalogued by A. H. Morton (1985). Similar stamped appliqués bearing the names of governors or officials and the capacity that they guaranteed were attached to vessels used for measuring. In a related technique, Islamic glassmakers pinched the hot glass with tongs whose jaws were decorated with simple inscriptions and other patterns. These short texts were repeated several times on the same vessel. A cup in the Metropolitan (Jenkins 1986, no. 17), for example, is inscribed in Kufic eight times vertically along its sides with the phrase ‘blessing to its owner’ (baraka li-ßå˙ibihi). Painting on glass allowed far more latitude for inscriptions. In particular, early Islamic glassmakers exploited the expensive lustre technique, in which the surface was painted with copper or silver oxide and then fired at a controlled temperature high enough to fuse the metal to the surface but insufficient to soften the glass and cause the object to collapse. These lustred glasswares were expensive to produce, and the inscriptions on them show that they were luxury wares meant to be enjoyed by rich patrons. A cup in the Damascus Museum (no. 16021), similar in shape to the one with incised decoration in the Metropolitan, is inscribed vertically in lustre five times along its sides with the admonition ‘Drink and rejoice to the sound of music’. In contrast to early Islamic glasswares inscribed with other techniques, several examples of lustrepainted glass from Egypt and Syria bear historical inscriptions, including names of specific patrons, sites of production, and dates in the eighth and ninth centuries. The earliest datable example is a broken goblet found at Fustat in 1965 and now in the Islamic Museum, Cairo (Arts of Islam 1976, no. 119). The band of Kufic inscribed around the rim says that it was among the things ordered by ÆAbd al-Samad b. ÆAli. He was governor of Egypt for one month in 773 under the ÆAbbasid caliph al-Mansur. Another lustred-glass fragment in the Cairo Museum (no. 12739/6) says that it was among the things made in the al-fÈla style in Egypt in the year 163 of the hegira. The date, corresponding to 779–80 CE, is written in Coptic numerals, and the meaning of the phrase ‘the al-fÈla style’

is not clear (Youssef 1972). A bowl in the Corning Museum (see Figure 13.79) is inscribed ‘Basmala. Blessing to him who drinks from this cup made in Damascus under the supervision of Sunbat in the year …’. The text breaks off after the word sana (‘year’) with an initial alif, but on stylistic grounds the bowl is usually attributed to the eighth century. Damascus is also mentioned on a goblet found at Raqqa and now in the National Museum. The Kufic inscription is divided by rosettes into five parts which mix good wishes and historical information. The three that have been deciphered say: ‘Drink and be gay; made in Damascus; join in the cares of [your] mate’. Making these special kinds of glasswares was a high-status profession in the early Islamic period, and several pieces of lustred glass in the Cairo Museum are signed. These include several fragments (nos 8167, 12638/2 and 15579) signed by ÆAbbas b. Nusayr. He may possibly be connected with the Nusayr b. Ahmad b. ?Haytham who made six pear-shaped glass carafes for the amir RabiÆ (Rice 1958). Youssef (1972) identified RabiÆ as the high-living son of Ahmad b. Tulun who rebelled against his nephew Harun b. Khumarawayh and was assassinated in 896. The detailed inscriptions show that these luxury glasswares were specific commissions for the most important members of society. The texts are readable and precise, and the specialised craftsmen, particularly those who made lustred glass, took pains to make their texts legible. These texts stand in direct contrast to one on an intact cup in the Metropolitan (Jenkins 1986, no. 20), with a complete but unreadable Kufic inscription around the rim. The cup is remarkable for its fine state of preservation and for its unusual epigraphy. The text is punctuated by the letter wå’ (‘and’) surmounted by three dots and the tail of a letter that returns to the right. The inscription differs in style from the squat dotless letters used on the goblet found at Fustat and other pieces, and the cup in the Metropolitan must have been made at a different (possibly modern?) time. Inscriptions are also found on another rare type of luxury metallic glassware known as gold glass. It was made using a sandwich technique in

— 183 —

— ISLAMIC INSCRIPTIONS —

13.79 Blown glass cup with lustre decoration made in Damascus under the supervision of Sunbat. Eighth century. Ht 10.3 cm. Corning Museum of Glass 69.1.1.

which a bowl was first decorated on the interior with gold and enamel, after which another bowl was blown into the first. Several pieces have gold inscriptions enlivened with spots of blue enamel. Unlike lustre, the technique of gold glass limited the text to short words. The text around the rim of a fragmentary bowl in the David Collection, Copenhagen, for example, is written in a stylised Kufic with letters ending in four lobes like the palmettes decorating the main band around the bowl. It is apparently a pseudo-inscription that was not meant to be read.

In the late twelfth century under the Ayyubids, lustre-painted and gold glasswares were replaced by a new and simpler technique of enamelling and gilding, in which coloured glass pastes and gold, either in the form of leaf or powdered dust, were applied to the surface of the glass vessel and then fused to it by firing. The Islamic Museum, Cairo, has the largest collection of these wares. Wiet’s catalogue of the enamelled lamps and bottles there, first published in 1929 and reissued in 1982, contains entries for 118 objects, arranged by museum accession number. Eighty-seven of them

— 184 —

— O T H E R P O RTA B L E A RT S — can be dated precisely because of the people named in their inscriptions. They range from a bottle bearing the name of the last Ayyubid ruler of Aleppo, Salah al-Din Yusuf (r. 1237–60), to a lamp bearing the name of the Mamluk sultan Qaæitbay (r. 1468–96). These datable pieces includes two important series, one from each branch of the Mamluk line: thirty-five lamps in the name of the Bahri sultan Hasan (r. 1347–61 with interruption), and twenty-two in the name of the Burji sultan Barquq (r. 1382–99 with interruption). Each entry in Wiet’s catalogue contains a detailed description of the object as well as a transcription and translation of the inscriptions on it. At the end of the volume, Wiet compiled a chronological list of all dated or datable enamelled wares known to him at that time. Although both Egypt and Syria have been cited as the place where the technique of enamelling glass developed, Wenzel (1985) has worked out a line of chronological development based on the patrons named on these luxury wares. The technique may have developed out of a prototype using only gold, which can be seen on a flask in the British Museum inscribed with the name of ÆImad al-Din Zangi, atabeg of Mosul and Aleppo from 1127 to 1146. In Ayyubid times, the gold was applied to glass vessels in two different ways, either in silhouette or raised over a thick enamel ground. The bottle in Cairo mentioning the last Ayyubid ruler of Aleppo, Salah al-Din Yusuf (Arts of Islam 1976, no. 135), provides a localisation for the first type with red enamel drawing over the silhouetted gold. Pieces in the second, raised style, such as a beaker once in the Landesmuseum in Kassel, or a flask in the Hermitage, contain the characteristic Ayyubid eulogy ‘glory to our lord, the sultan’ (Æizz li-mawlånå al-sul†ån). Most of the decoration on pieces in this early style typical of the early thirteenth century is figural, but it generally includes one or more inscription bands, almost always in cursive script. Gradually in the late thirteenth century under the Mamluks, blazons were introduced. A tiny perfume bottle in the Corning Museum (Atıl 1981, no. 46), for example, is decorated with a composite blazon of a lion and a striped field.

The lion was the blazon of Baybars and his son Baraka Khan, and the striped field was used by the Hama branch of the Ayyubids who served the Mamluks. The composite blazon on the perfume bottle probably belongs to an official from Hama who was in the service of Baybars or his son. By the fourteenth century, figural motifs had declined in popularity, and inscription bands became the main subject of decoration on enamelled glass. The texts were generally set on a floral ground and interrupted by roundels with emblems, areas of contrasting pattern, or suspension handles on mosque lamps. These bold bands are ingeniously set out so that both text and technique underscore their meaning. This can be seen most clearly on mosque lamps, the most characteristic form of enamelled glass. The typical lamp (see Figure 13.80) is about 30 cm high, with a wide and flaring neck, sloping shoulders with six applied handles, bulbous body, and prominent foot or foot ring. A small glass container for water and oil with a floating wick was inserted inside the lamp, and the lamp itself was suspended by chains from the ceiling. Sometimes the chains were held together above the lamp by an egg-shaped object, also of enamelled glass. Thousands of such lamps must have been commissioned to illuminate the mosques and charitable foundations established by the Mamluks. The decoration of these lamps varies, but most have bold inscriptions in thuluth encircling both the body and neck. The inscription on the bulbous body generally names the patron and is written in reserve against a blue ground, with the letters piled on top of each other and outlined in red. Most surviving lamps were made for sultans, and the inscriptions use the royal form ‘glory to’ (Æizz li-). Some ten surviving examples were made for al-Nasir Muhammad (r. 1294–1340 with interruptions), some fifty for Hasan, and nearly forty for Barquq. Other mosque lamps were made for prominent citizens and use the form mimmå Æumila bi-rasm (‘one of the things made for’), followed by the patron’s name and sometimes his specific foundation. The inscription on the one in the Museum of Fine Arts, Boston, for example, says that it

— 185 —

— ISLAMIC INSCRIPTIONS —

13.80 Glass lamp made for the hospice of Karim al-Din. Egypt, 1310–20. Ht 27.5 cm. Boston Museum of Fine Arts 37.614.

— 186 —

— O T H E R P O RTA B L E A RT S — was made for the blessed hospice (ribå†) of Karim al-Din. He was a civil official of the class known as ‘men of the pen’. Those made for Mamluk amirs are often decorated with blazons as well as names. One made for the amir Aydakin (Jenkins 1986, no. 47), for example, has roundels with two addorsed bows on a circular red field, indicating that he was bowman to the sultan. Another made for the cup-bearer Shaykhu (Arts of Islam 1976, no. 138) has roundels with a red cup between two bars, the upper black and the lower red. The inscription around the neck is written in blue enamel outlined in red and typically contains the opening part of the Light Verse (Koran 24:35): God is the Light of the heavens and the earth; the likeness of His Light is as a wick-holder wherein is a light, the light in a glass, the glass as it were a glittering star, Lit from a blessed tree. Different lamps contain slightly different amounts of the verse. The most complete, on a lamp made for Barquq (Wiet 1929, no. 282), goes as far as the word yuqadu (‘lit’), but in most cases the Koranic text stops after the word ‘glass’ (almisbå˙). The verse describes a floating lamp in a glass and refers here to the glass lamp on which it is inscribed. When the enamelled glass lamp was lit, the patron’s name and titles written in reserve around the bulbous body would have glowed with divine light, a stunning visual realisation of the beautiful Koranic metaphor inscribed around the neck in opaque blue. Production of enamelled glass seems to have declined sharply in the fifteenth century, and only a handful of lamps date after the reign of Barquq. The inscriptions on pieces from the late fifteenth century are more hastily written, and Wiet characterised the writing on the one made for Qaæitbay at the end of the fifteenth century as ‘bastardised’. As the lamp also differs from earlier lamps in layout, style and tonality of the enamel, Wiet suggested that the inelegant letters had been laid out by a foreign workman who misunderstood the original model. These enamelled lamps were also imitated in late nineteenth-

century Europe by such masters as Emile Gallé. Although the revival wares copy the shape of the Mamluk prototypes, they are generally anepigraphic, as are most glasswares made elsewhere in the Islamic lands in later periods.

Ivory Luxury objects made of or decorated with ivory were produced throughout the Islamic world whenever the material was available. Most ivory came from Africa, so the Mediterranean Islamic lands, particularly Syria, Egypt, Spain and Sicily, were the primary centres of ivory working, but India also supplied enough to meet its own needs in medieval times. The range of ivory objects was limited by the shape and size of the elephant’s tusk: small objects such as combs, boxes and chessmen were carved whole from the block, while larger boxes were made of separate pieces usually joined with ivory pegs. Inscriptions could be carved in relief or incised, and the decoration was sometimes enhanced by painting and gilding. The major study of early Islamic ivories is Kühnel’s magisterial work, Die islamischen Elfenbeinskulpturen VIII–XIII Jahrhundert (Berlin, 1971), which contains general essays and a catalogue of 140 pieces, with superb plates and the text and a translation of the inscriptions on each piece. The earliest ivory objects generally did not have inscriptions as part of their decoration. One of the few that does is a pyxis in St Gereon in Cologne. An inscription incised with dots around the base of the conical lid invokes blessings on the caliph and says that the piece was made in Aden for ÆAbdallah the Commander of the Faithful, referring to one of two first ÆAbbasid caliphs. Another inscribed object from early Islamic times is an elaborate chessman in the Cabinet des Médailles, Paris, in the shape of a nobleman riding in a howdah on the back of an elephant and accompanied by warriors and horsemen. The underside of the base is incised in simple and undotted Kufic with the signature of the craftsman: the work of Yusuf al-Bahili. Although the carver’s nisba is an Arab tribal name well attested in the early Islamic period,

— 187 —

— ISLAMIC INSCRIPTIONS — the piece is carved in the sculptural style of India, and attributions have ranged from the ninth century (Kühnel 1971, no. 17) to the twelfth (S. C. Welch 1988, no. 72). Ivory became more readily available by the ninth and tenth centuries, and the Aghlabids of Tunisia (r. 800–909) carried on a flourishing trade in ivory from sub-Saharan Africa. One of the few pieces known to have been made in Tunisia (see Figure 13.81) is a rectangular wooden casket with ivory plaques in the Archaeological Museum in Madrid (Arts of Islam 1976, no. 145). A Kufic inscription on the lid records that the box was made for the Fatimid caliph al-MuÆizz (r. 953–75) at al-Mansuriyya, the city near Qayrawan that served as his capital until 972. The end of the inscription, now missing, once recorded the name of the person who made (ßanaÆa) it, a name read as Ahmad (or Muhammad) al-Khurasani.

This ready supply of ivory was the basis for the best-known group of inscribed ivories produced in the Islamic lands, the thirty-three objects made in Spain between the mid-tenth and the eleventh century (in addition to Kühnel 1971, see also Ferrandis 1935–40). They are all boxes. The larger caskets may have been for regalia, but the smaller ones were designed to hold jewellery and other precious commodities, for the verses inscribed on a cylindrical pyxis in the Hispanic Society of America record that it was a container for musk, camphor and ambergris. They are all carved in high relief on a smooth ground, and most have Arabic inscriptions carved in a plain or foliated Kufic script on the lower edge of the lid. The texts begin on the right front of the rectangular boxes or to the left of the clasp on the cylindrical pyxides. Since the Spanish ivories form such a closely knit group, they offer a wonderful oppor-

13.81 Wooden casket with ivory inlay made for the Fatimid caliph al-MuÆizz at al-Mansuriyya c. 970. Length 42 cm. Madrid, National Archaeological Museum 887.

— 188 —

— O T H E R P O RTA B L E A RT S — tunity to study epigraphic developments in both content and style. Over the course of two centuries, the text on Spanish ivories generally becomes longer and the script more elaborate. The first pieces began only with ‘this is what was made for’ (hadhå må Æumila li-), but later pieces offer more blessings and good wishes to the patron or owner, whose name is often recorded. The owners were members of the caliph’s family, particularly women, important officials at court and, later, members of the Dhuæl-Nunids, the Berber family of party kings (Arabic mulËk altawåæif) who took control of Toledo in the early eleventh century after the disintegration of the Spanish Umayyads. Two places of manufacture are recorded: Madinat al-Zahraæ, the Umayyad palace city east of Cordova, is mentioned on two pieces made in 355/966, and Cuenca, the Castilian city near Toledo that was under the control of the DhuælNunids, is mentioned on three pieces associated with their patronage in the mid-eleventh century. Eight of the ivories are dated between 353/964 and 441/1049–50, and several others can be dated by the titles used in the inscriptions. The high proportion of dated pieces suggests that many of the ivories were made for specific events. The inscriptions on the Spanish ivories also record the name of the person who supervised the work. Introduced by the phrase Æalå yaday (literally, ‘under the hands of’), he was generally an important slave in the caliphal household. The supervisor is distinct from the carver, who is introduced by the word Æamal (‘work of’). The craftsman’s name is sometimes sculpted in relief at the end of the main inscription, sometimes incised in a inconspicuous place elsewhere. Carvers seem to have worked in teams, for a particularly large and complex casket in Pamplona (Dodds 1992, no. 4) is incised on the interior of the lid ‘the work of Faraj and his apprentices’ (Æamal faraj wa talåmidhah), and five other names are incised in other places on the casket. The craft may have become hereditary, for the two carvers who made ivories for the Dhuæl-Nunids seem to belong to the same family. The Spanish ivories also document the increasing elaboration of Kufic script. The earliest pieces

made in the 960s, such as a cylindrical box and a casket made for the daughter of the caliph ÆAbd al-Rahman III (Dodds 1992, nos 1 and 2) and the caskets and pyxides made for Subh, concubine of al-Hakam, show relatively simple Kufic, decorated only with an occasional trilobed ending or floral motif above a letter. These decorative details were quickly multiplied and enlarged, and pieces made in the 960s, such as the famous pyxis in the Louvre made for al-Mughira (ibid., no. 3), have floral scrolls that are often suspended like boughs from the upper border of the inscription band. By the early eleventh century, as on the Pamplona casket (ibid., no. 4), the flowers float above the letters and fill the upper zone of the inscription. On this particularly elaborate piece, the letters of the inscription are also beaded. By the mid-eleventh century, as on the Palencia casket (ibid., no. 7), the simple Kufic of the earlier pieces has evolved into a floriated script. Equally fine relief-carved ivories were also produced for the Fatimids in Egypt, but in contrast to the Spanish caskets and pyxides, the Fatimid ivories are small plaques decorated with superbly carved figural scenes and do not have inscriptions. The small group of Spanish ivories made for a limited elite also stand in contrast to a much larger group of painted and gilded ivories attributed to Sicily in the twelfth and early thirteenth centuries (Pinder-Wilson and Brooke 1973). More than 200 pieces are known, including caskets and pyxides made of thin sheets of ivory secured by ivory pegs often supplemented with metal clasps and gilt metal braces. Like the Spanish ivories, the Sicilian caskets and pyxides are inscribed at the base of the lid, but the technique, style and content of the text differ from those on the earlier Spanish boxes. The inscriptions on the Sicilian pieces are painted or incised, usually in a cursive script, but occasionally in an elaborate floriated Kufic. The texts include good wishes to the owner, sometimes with the phrase Æizz yadËm [li-ßå˙ibihi] (‘glory [to its owner] endures’), a phrase not known on other portable objects. Another text comprises two well-known distichs in the metre kåmil that are repeated several times in the Thousand and One Nights and mention lovers

— 189 —

— ISLAMIC INSCRIPTIONS — beneath a coverlet. Painted and gilded caskets were also made in Spain from the twelfth to the fourteenth century, and one casket in the Institut de San Juan (Arts of Islam 1976, no. 153; Dodds 1992, no. 50) has a poem composed as though the object itself were speaking. In the later period, ivory seems to have been scarcer. A few pierced cylindrical boxes made in the fourteenth century are assigned to either Egypt or Spain. One box formerly in the Rothschild collection bears the name and titles of the Mamluk sultan Salih (r. 1351–54). Other boxes similar in style are inscribed with good wishes suggesting that the owner will enjoy good luck (Atıl 1981, no. 106) or with a poem written as though the object were speaking (Dodds 1992, no. 51). Ivory plaques were also used to decorate marquetry or inlaid doors, minbars and other furnishings. Many plaques show figural or geometric decoration, but a few are inscribed, such as the polygonal ones in the Islamic Museum, Cairo, from a table made in 1369 for Khwand Baraka, mother of Sultan ShaÆban, or the rectangular ones (Arts of Islam 1976, no. 155) with the name of the Mamluk sultan Qaæitbay (r. 1468–96). Small ivory objects and ivory inlays were also popular in Ottoman court workshops, and several fine examples were published in Atıl (1987) on the age of Sulayman. A carved oval mirror-back in the Topkapı Palace (no. 73) has an inscription carved around the rim with three couplets of Ottoman verse, a dedication to Sultan Sulayman by the craftsman Ghani, and the date 950/1543–4. Wooden boxes for the Koran were often decorated with ivory inscriptions around the base of the lid. One sixteenth-century example (no. 109) has carved plaques with Koranic verses and prayers; a dodecagonal top found in the mausoleum of Mehmed III, who died in 1603 (no. 110), has a nastaÆlÈq inscription in ivory inlaid in ebony with a hadith on the virtues of the Koran. Ivory was rarer in Iran, but was apparently used in many of the same ways. Five panels in the Iran Bastan Museum (Arts of Islam 1976, no. 156) from the tomb of Shah IsmaÆil (d. 1524) at Ardabil are carved in openwork with Koranic inscriptions in naskh against a scrolling floral ground.

Rock Crystal Rock crystal, clear colourless quartz in hexagonal crystals (Arabic billawr), was imported from East Africa, the Laccadive and Maldive Islands in the Indian Ocean, and North Africa to many of the Islamic lands, where it was carved into such small objects as phials, flasks, cruets and lamps. Some 180 medieval objects are known, most of them preserved in European royal and church treasuries and a few excavated, notably at Fustat (Old Cairo) and Susa in Iran. Many were published by C. J. Lamm in his compendium, Mittelalterliche Gläser und Steinschnittarbeiten aus dem Nahen Osten, published in Berlin in 1930. The only technique of decoration on medieval pieces is carving. Given the broad spread of decorative motifs and the extensive trade in precious objects in medieval times, it is difficult to assign provenance on the basis of motifs used. Inscriptions are therefore one of the primary ways of localising objects of rock crystal. Three rock crystal objects bear names associated with the Fatimid dynasty of Egypt (r. 909–1171). The earliest (see Figure 13.82) is a pear-shaped cruet decorated with confronted seated lions in the treasury of San Marco in Venice (Curatola 1992, no. 61; Ettinghausen and Grabar 1987, fig. 177). The Kufic inscription carved around the neck invokes blessing from God on the imam al-ÆAziz billah, the Fatimid caliph who ruled from 975 to 996. The inscription was read more than a century and a half ago by Michelangelo Lanci (1845–6), and the Kufic with rising tails is a masterpiece of the genre. One significant but so far unexplained detail is the interlaced kåf in baraka, for interlacing is characteristic of the eastern Islamic lands in the tenth century and coincides with other Persianate features of the cruet such as the shape, elongated lip, and crouching ibex on the handle. A similar cruet with confronted birds (possibly ostriches), formerly in the Medici collection and now in the Pitti Palace, Florence (Curatola 1993, no. 62; Ettinghausen and Grabar 1987, fig. 179), is inscribed over the palmette between the birds ‘for the Commander of Commanders

— 190 —

— O T H E R P O RTA B L E A RT S —

13.82 Rock-crystal cruet made for the Fatimid caliph al-ÆAziz (r. 977–96). Venice, Treasury of San Marco 80.

— 191 —

— ISLAMIC INSCRIPTIONS — personally’ (li-qåæid al-quwwåd khåßßatan). Rice (1956b) noted that qåæid al-quwwåd was a title created for the Fatimid generalissimo Husayn b. Jawhar on 17 Jumada II 390/25 May 1000. Since Husayn was deposed eight years later on 7 ShaÆban 398/18 April 1008 and resumed office briefly between 4 Muharram 401/18 August 1010 and 12 Jumada II 401/30 January 1011, the cruet can be dated precisely between 1000 and 1011, most likely between 1000 and 1008. Indeed, on a parallel with the Spanish ivory boxes, it seems likely that this expensive object might have been made specifically to commemorate the bestowal of Husayn b. Jawhar’s new title. One wonders also whether the cruet was part of a set, the other one of which was inscribed with the good wishes that were bestowed on the commander of commanders mentioned on this piece. A small jar with confronted birds in the National Archaeological Museum, Madrid (Arts of Islam 1976, no. 110), is inscribed at the neck ‘blessing from God’ (baraka min allåh), the type of text that would fit neatly with that on the Florence cruet. The third object inscribed with a Fatimid name is a simple rock-crystal crescent in the Germanische Nationalmuseum in Nuremberg (Ettinghausen and Grabar 1987, fig. 178), inscribed with the name of the caliph al-Zahir (r. 1021–36). Entirely different in style from the other two cruets, it may have been a horse ornament. It confirms, nonetheless, that Egypt continued to be a major centre of rock-crystal carving in the eleventh century. These three objects naming Fatimid caliphs or their courtiers are the basis for ascribing almost all pieces of Islamic rock crystal to Egypt in the tenth or early eleventh century. Most of the other inscribed pieces have blessings to an anonymous owner, yet even these anonymous blessings suggest that the range of patrons and the provenance of the objects were wider than imagined. This is true even for the group of six cruets similar in shape to those in Florence and Venice. One with confronted hawks in the Louvre that had belonged to the treasury of St Denis offers blessing and joy to its owner (baraka wa surËr wa [sic] li-ßå˙ibihi).

Another with confronted hawks in the treasury of Fermo Cathedral (Curatola 1993, no. 58) bears a similar inscription that has been read as baraka wa surËr biæl-sayyid al-malik al-manßËr. The title malik was only assumed by Fatimid viziers at the end of the eleventh century, and reading the inscription in this way would place this piece a century later than its stylistic cousins. To get around this difficulty, Rice suggested reading the inscription as baraka wa surËr alsayyid al-mu[æa]yyad al-manßËr to make it a generalised blessing written slightly incorrectly (the carver would have omitted the wåæ in almu[æa]yyad), but this explanation is convoluted and such a string of epithets does not conform to the standard pattern. The form of anonymous blessings varies on other rock crystals as well. A tall vase in the treasury of San Marco (Curatola 1993, no. 55) offers blessings to an anonymous lord (dawla dåæima wa niÆma kåmila wa salåma li-mawlånå) in an elaborate Kufic with rising tails. A mace head in the Islamic Museum in Cairo (Arts of Islam 1976, no. 105) invokes good wishes on an anonymous owner (iqbål wa baraka li-ßå˙ibihi) and the names of both Muhammad and ÆAli (mu˙ammad wa ÆalÈ kilåhumma), perhaps alluding to the ShiÆite affiliation of the Fatimids. A small bottle in the V&A (Pinder-Wilson 1954) offers blessings to an anonymous owner (baraka li-ßå˙ibihi), and another bottle in the Islamic Museum, Cairo (Arts of Islam 1976, no. 107), probably intended to hold kohl, offers the blessing to an anonymous female owner (baraka lißå˙ibatihi). The gender of its owner shows that some rock-crystal objects, like some ivory boxes from Spain, were made for women. Since the range of objects is so small, it would be productive to study the varying forms of these anonymous blessings and the style of script in order to delineate diffrent groups. Rock crystal was fashionable in later times, particularly under the Ottomans and Mughals, but the later taste was entirely different. The objects are generally anepigraphic and instead are encrusted with precious gems.

— 192 —

— O T H E R P O RTA B L E A RT S —

Jade and Other Hardstones Nephrite jade (Persian yashm) has been known in the Islamic lands at least since the tenth century, but the first surviving objects are connected with the patronage of the Timurids in the fifteenth century when the jade mines in Khotan passed under the nominal control of this Central Asian dynasty. Many examples were included in the exhibition of Timurid art held in 1989, and the accompanying catalogue by Lentz and Lowry, Timur and the Princely Vision, contains good photographs and summaries of the inscriptions, although no texts are given. The largest pieces of Timurid jade are two huge blocks of greenish-black nephrite, one of which was used as Timur’s cenotaph in the Gur-i Mir (Semenov 1948; Masson 1948; English summary and emendations in O. Grabar 1957: 553–5). The top of the large rectangular cenotaph is engraved with a niche bordered by a thuluth inscription that identifies it as the grave of Timur, enumerates his titles and traces his genealogy back to Genghis Khan and ÆAli b. Abi Talib. Another thuluth inscription engraved on the bottom side recounts how Timur’s grandson Ulugh Beg acquired the stone as booty after he had defeated the ruler of Moghulistan on the Ili River in 828/1424 and took it back to Samarqand, where it was carved and set over Timur’s grave. Smaller objects made for the Timurids, particularly tankards and cups, were also inscribed. Five of them bear the names of Ulugh Beg and his immediate relatives. The most famous (see Figure 13.83) is a white nephrite tankard in the Gulbenkian Foundation in Lisbon whose neck bears a relief inscription in thuluth (Arts of Islam 1976, no. 114) similar to that used on Timur’s cenotaph. Two green cups with dragon handles in the British Museum (Pinder-Wilson and Watson 1960) and the Bharat Kala Bhavan in Benares are engraved on the side with the name of Ulugh Beg Gurgan. A dark green almond-shaped seal in the Hermitage (Ivanov 1971; Lentz and Lowry 1989, no. 128) is inscribed with the name of Ulugh Beg’s mother, Gawharshad bt. Ghiyath al-Din Tarkhan (d. 1457).

The inscriptions on these four jades give only the names and titles of Timurid family members, but the inscription on the fifth Timurid jade, a tiny white casket for gems in the Asian Art Museum, San Francisco (Lentz and Lowry 1989, no. 50), is different. The inscription around the walls records that the sultan, son of the sultan, ÆAlaæ al-Dawla Bahadur Khan, may his kingdom endure, ordered the completion of this casket (durj). Grandson of Shah Rukh through Baysunghur, ÆAlaæ al-Dawla briefly proclaimed himself sultan at his father’s death in 850/1447, but was crushed a year later by his uncle Ulugh Beg. Since the five objects differ in style, it has been argued that they were gifts to the Timurids, but the inscription on the casket shows that jades were commissioned by the Timurids themselves. These inscribed objects are the basis for attributing many others of similar shape to the Timurids. Some are inscribed with benedictory good wishes in Persian verse that allude to the function of these jades. A black plaque in the Chester Beatty Library, Dublin, for example, has verses by SaÆdi and Daqiqi asking God’s protection from evil, suggesting that the plaque was an amulet. These verses are often inscribed in cartouches linked by rosettes, as on a dark green cup in the Louvre and a striped agate cup made for the Timurid sultan Husayn Bayqara in 874/1470–1 (Lentz and Lowry 1989, no. 150). The verses on the agate cup compare it to Jamshid’s cup and describe its appearance when filled with rosy wine, clearly showing the function of these objects as wine cups. The prerogative of Timurid royalty, these hardstone objects were appreciated by later rulers, particularly the Mughals, who added their own inscriptions to mark the acquisition of these precious objects belonging to their illustrious ancestors. Many Mughal jades were included in the 1982 exhibition of Indian art in London, and the accompanying catalogue, The Indian Heritage, gives translations of the texts, which are surveyed in Markel’s 1990 article on the inscribed wares made for the Mughals. Like the inscriptions done by the Timurids, the Mughal inscriptions emphasise the royal ownership of these pieces and their

— 193 —

— ISLAMIC INSCRIPTIONS —

13.83 White jade tankard made for the Timurid prince Ulugh Beg and reinscribed for the Mughal emperors Jahangir in 1022/1613 and Shah Jahan in 1056/1646. Ht 14.5 cm. Lisbon, Calouse Gulbenkian Foundation 328.

function as wine cups and tankards, but nastaÆlÈq replaces the thuluth used on the earlier pieces, and the dates are given in both hegira and regnal years. The white tankard made for Ulugh Beg, for example, was engraved twice by the Mughals: first on the upper edge of the rim in 1022/ 1613–14 with Jahangir’s name and titles, and then

below the handle in 1056/1646–7 with Shah Jahan’s title ßå˙ib qirån (‘Lord of the [Auspicious] Conjunction’). Similarly, an anepigraphic dark green tankard made for the Timurids and now in the British Museum was engraved around the neck with Persian verses alluding to wine-drinking and the date of its acquisition by

— 194 —

— O T H E R P O RTA B L E A RT S — Jahangir in 1028/1618–19, the fourteenth year of his reign, at Fatehpur Sikri. The Mughals also commissioned new jade objects. Inscribed pieces date from the reign of Jahangir onwards, and the inscriptions are similar to those added to older pieces. The earliest known is a grey-green wine cup in the Brooklyn Museum (Indian Heritage 1982, no. 350) inscribed with Persian verses in praise of wine, Jahangir’s name and titles, and the date 1607–8 given in regnal and hegira years. The most exquisite is a white cup in the shape of a halved gourd in the V&A (Indian Heritage 1982, no. 356; Blair and Bloom 1994, fig. 377) inscribed with Shah Jahan’s title and the date 1067/ 1656–7. Similar objects were made of other precious stones. A wine cup for Jahangir, recently acquired by the Rhode Island School of Design (Indian Heritage 1982, no. 372), is made of quartz and chromian muscovite and has incised decoration, perhaps in imitation of chasing on metalwares. Other types of jade objects made for the Mughal emperors were inscribed in similar ways, sometimes with additional information. For example, an inkpot in the Metropolitan (Indian Heritage 1982, no. 352) has cartouches on the body enclosing Jahangir’s name and the date 1028/ 1618–19. A second inscription on the underside of the foot ring names Muæmin Jahangiri, one of the only jade carvers to be mentioned. The location of his signature, like that of other Safavid artists,

shows his subservience to his royal patron. A tiny petalled cup in Benares (Markel 1990, fig. 8) is inscribed that it was an antidote to poison (pådzahr). The many small hardstone objects commissioned by the Mughals, including perfume phials, jewellery and other personal adornments, archer’s rings and weapons, were also inscribed, but the small surface area meant that the inscriptions are generally limited to concise statements of royal proprietorship. Daggers, swords and shields are often inscribed with exalted appellations such as ‘World Astonisher’ or ‘Tyrant Slayer’. Jade seems to have been rarer under the Safavids, but the handful of inscribed pieces shows that the inscriptions generally follow Timurid models in layout, content and style, if not technique. A black tankard in the Topkapı Palace, for example, is inscribed around the neck with the name of the first Safavid shah IsmaÆil I (r. 1501–24), but the inscription, like the other arabesque decoration, is inlaid with gold wire. A rather plain bowl in the V&A is inscribed with verses and a dedication by ÆAbbas I in 1611 to the shrine at Ardabil. A seal in the Musuem für Islamische Kunst in Berlin (Survey, plate 1458A) is inscribed with the name of ÆAbbas II (r. 1642–66). Jade was much more popular under the Ottomans, but they preferred encrusted and bejewelled decoration and the pieces are generally anepigraphic, although occasionally the Ottomans added inscriptions to earlier pieces.

— 195 —

— CHAPTER 14 —

Types of Objects

Tombstones and Other Grave Markers Despite orthodox prohibition against venerating the dead, the graves of Muslims are usually marked by upright slabs (stelae or steles), longitudinal or prismatic blocks, empty rectangular boxes (cenotaphs) or other forms of grave marker. These grave markers are made of different materials, depending on what was available locally. The most exotic was the huge block of nephrite that the Timurid prince Ulugh Beg dragged from Moghulistan in 828/1424 to mark the grave of his father Timur at Samarqand (Semenov 1948; Masson 1948). Virtually all of the fancier pieces are inscribed. As the epitaphs include the date of the deceased’s death, they provide important social and historical information about conversion, classes, titulature and sectarianism as well as the development of writing styles. The speed with which tombstones were introduced depended in part on the strength of local customs of burial. Only a few rare examples date to the first two centuries of Islam, but the year 174/790 marks the start of a continuous series of more than 4,000 limestone and marble stelae from the Egyptian cemeteries at Aswan and Cairo. Most of them were published in ten massive volumes as part of the Catalogue du Musée arabe du Caire (vols I and III by Hawary and Rached 1932–8; vols II, IV–X by Wiet 1936–42). A few other examples taken to other museums have been published separately by Walker (1951), Miles (1957), Sourdel-Thomine (1964) and others. Tombstones were gradually introduced to other areas in the western Islamic lands. The

first surviving example of a grave marker from Qayrawan dates from 235/850; that from Spain, 239/854. The series of tombstones from Dahlak, the coral archipelago in the Red Sea opposite Eritrea, begins around the same time (Schneider 1983). Tombstones were introduced more slowly to the eastern Islamic lands. In Syria, while texts mention cemeteries, the first tombstones survive only from the Fatimid period. Iran had no preIslamic tradition of erecting grave markers, and the first surviving tombstone, recently discovered in the Imamzada JaÆfar at Damghan, dates from around the year 900 (Adle 1984). It is a rare specimen, and the first series begin later. Recent excavations at Siraf on the Persian Gulf have brought to light sixty inscribed fragments from grave covers (see Figure 14.84) dated between 363/ 975 and 735/1334 (Lowick 1985), and a series of forty-two stelae from Yazd date from the mideleventh to the late twelfth century (Afshar 1969–75, 1973). From medieval times, the number of surviving grave markers increases dramatically, and most studies divide this vast range of material geographically. Some, like the pioneering works by Amari (1875) and Lévi-Provençal (1931) on Sicilian and Spanish epitaphs or the more recent one by Zbiss (1955) on Tunisia, include chapters on epitaphs in compilations of Arabic inscriptions encompassing whole countries. The majority of studies cover homogeneous collections of tombstones from individual cemeteries, towns or regions – such as the works by Bourrilly and Laoust (1927) on Sale and Rabat, Deverdun (1956) on Marrakesh, Roy and Poinssot (1950–8) on Qayrawan, Sourdel-

— 196 —

— TYPES

OF

OBJECTS —

14.84 Crested grave cover found at Siraf and made for Ibrahim b. ÆAli, who died in 383/993. Length 1.51 m (after Lowick 1985, plate XIII).

Thomine on Bust (1956b) and on the dead cities of northern Syria (1956a), Karama©arali (1972) and Rogers (1988) on Ahlat, Ory (1989) on the area around Busra in southern Syria, and Schneider (1983) on the islands of Dahlak in the Red Sea and on several sites in the Yemen (1985, 1992). Many examples from the Indian subcontinent were published in the nineteenth century in the Journal of the Asiatic Society of Bengal. One of the most complete publications is that by Moaz and Ory (1977) on the epitaphs from Bab al-Saghir, the cemetery outside Damascus. The work contains a chronological catalogue of the eighty-five tombstones with copious measure-

ments, descriptions, drawings and photographs. The authors summarise their conclusions in several tables that cut across the chronological arrangement, classifying the tombstones by material conditions (shape and material), decoration, formulas inscribed, people commemorated, and palaeography or graphics. The work could well stand as a model for future studies. Given both the range of material and its heterogeneity, it is no surprise that little attempt has been made attempt to survey and synthesise this vast corpus of epitaphs. One rare exception is Sourdel-Thomine’s article ‘abr’ (‘Grave’) in the second edition of the Encyclopaedia of Islam.

— 197 —

— ISLAMIC INSCRIPTIONS — Grave markers from some areas of the Islamic world, especially North Africa and Egypt, but now also Turkey, are more heavily documented, while those from elsewhere are treated more summarily. Most of the work on grave markers from Iran or Central Asia is hidden in hard-to-find journals (especially Epigrafika Vostoka) or relatively obscure works in Persian or Russian, and there are few syntheses or overviews of it. The epitaph, known in Arabic as kitåb (‘inscription’), is usually in Arabic, although in later periods Arabic was replaced by or mixed with indigenous languages. Nineteenth-century stelae from Central Asia, for example, are bilingual, Russian and Arabic. The text is usually in prose, but Arabic poetry occurs on a group of ninth-century stelae from Sicily, and cartouches with Persian or Turkish verses are characteristic of later examples from Iran, India and Turkey. At first, the text was carved in lines, often enclosed by a rectangular frame. Later the text was written in panels, cartouches or bands. Over time, both the script and the framing decoration become more elaborate. The earliest inscriptions were done in a simple Kufic script, but in the eighth and ninth centuries the script was embellished with such decorative devices as the barb, hook, arc and palmette. The bevelled stems of the letters evolved into floral ornament, often set on an undulating scroll ground. The Egyptian tombstones are particularly important in documenting these changes, as few monumental inscriptions survive from this period. Cursive scripts were introduced slowly from the tenth century, and bands of angular and cursive were often juxtaposed. In the fourteenth century when the Mongols ruled Iran, inscriptions in square Kufic, resembling Chinese seal script, became popular. The inscription sometimes took on such a decorative shape that it was virtually unreadable, as when Koran 2:137, the verse containing fasayakfÈkahum, the longest word in the Koran, was sculpted in the shape of a polylobed arch (Miles 1939). The epitaphs had a dual purpose – to record the name of the deceased and to bear witness to his faith. From the earliest known examples, they seem to follow a standardised formula. They begin with

the invocation or basmala. This is followed by several introductory phrases, either set phrases such as familiar Koranic verses, the benediction on the Prophet (taßliya), or the ˙amdala, or sometimes freely composed verses. The texts deal mainly with judgment, resurrection or the juxtaposition of life’s transience with God’s everlastingness. Next comes the name of the deceased with his genealogy or titles and the date of his death. The deceased’s name is followed by several eulogies, either in his honour or ‘transferred eulogies’, which offer blessings to the person who reads the epitaph. Sometimes the inscription closes with a formula of resurrection, saying that with this testimony the deceased has lived, with it he has died, and with it he shall be resurrected to life. As the function of these tombstones was to attest or witness, the most common verb to introduce the deceased was shahida (‘testified’). Later this was replaced by tuwuffiya (‘deceased’). Both of these were ultimately supplanted by a third variant, hadhå qabr (‘this is the grave of’). This form occurs on about one-third of the examples from Egypt and Aswan and then becomes the ubiquitous formula for later epitaphs from all the Islamic lands. Supplementary texts were often added to this basic statement of testimony. One of the most common is the profession of faith (shahåda), which was often put in the mouth of the deceased. Certain well-known Koranic verses were also popular. They include Koran 112, the final chapter about God’s uniqueness and the denial of the Trinity, and a clear rebuke to Christianity. Also popular in all places and periods was the socalled Throne Verse (Koran 2:255), the eloquent statement about the God’s dominion over heaven and earth. A similar sentiment is evoked by Koran 3:18, saying that God, His angels, and wise men attest to the one God and that the true religion before God is Islam (‘submission’). With its emphasis on attestation, the quotation is particularly apt on tombstones whose inscriptions used the same verb, saying that the deceased attested. Other Koranic texts became popular at different times. Koran 55:26–7, which says that all is perishable except God’s face in majesty and

— 198 —

— TYPES

OF

magnificence, would seem to be a natural quotation for tombstones, but in fact the first extant example dates only from 240/854, after which time it became increasingly popular. It is the main inscription, for example, on the crested grave cover from Siraf made for Ibrahim b. ÆAli in 383/993 (see Figure 14.84). Similarly, the phrase that every soul shall taste of death (taken from either Koran 3:185 or 29:7) became widespread from the twelfth century onwards. Other Koranic verses enjoyed local popularity: Koran 35:5 about the vanity of earthly life, for example, is typical of Andalusian tombstones. Tracing the changing formulas and Koranic quotations in epitaphs from a given area can show us changes in faith and doctrine, as has been done for example on the large corpus of material from Egypt (Wiet 1952; Bloom 1987b) and nearby areas (Schneider 1983). For example, a phrase that the worst of the misfortunes for Muslims was the loss of the Prophet occurs more than 300 times in the ninth century but was replaced under Fatimids by the Throne Verse (2:255) followed by a phrase evoking God’s mercy on the deceased through Muhammad and his pure family. A few studies discuss the style of carving on grave markers. Ocaña Jimenez (1964), for example, treated a specific local style, that of the ‘Almerian’ tombstones from Spain. One of the pioneers in the tracing of the stylistic developments of epitaphs is Janine Sourdel-Thomine. Following the method of epigraphic analysis developed by Samuel Flury (1920b, 1921) for the inscriptions from Amida/ Diyarbekır, she designed alphabetic tables so that one can quickly compare the forms of individual letters and grasp stylistic evolution. For example, in analysing the first 400 tombstones from Egypt, dating from the beginning of the ÆAbbasid period before the move to Samarraæ in the late eighth century to the establishment of Tulunid power in Egypt in the mid-ninth century, SourdelThomine (1972) sketched six concurrent styles. She has also offered penetrating analyses of later, more elaborate Kufic inscriptions, particularly those from Syria (1957). In her work on the cenotaphs from Bab al-Saghir, she treated Ayyubid

OBJECTS — Kufic, the period of the last flowering of Kufic in Syria (1950). In other cases, she analysed not only palaeographic development but also the evolution of decoration, as with the geometric medallions and borders on epitaphs from northern Syria (1956a). Careful palaeographic analysis can help trace the movement of artisans, their status, and the introduction of new styles of script. Schneider (1986), for example, showed how the stonecarver Mubarak al-Makki, who signed several ninthcentury tombstones in Cairo, was probably a native of the Hijaz.

Arms and Armour Perhaps surprisingly, much of the arms and armour produced in the Islamic lands, particularly the finest pieces, is inscribed. Inscriptions occur on eveything from helmets, body armour, belts and shirts to swords. Texts include a variety of marks of ownership, with inscriptions showing who commissioned the piece, who owned it, or to whom it was endowed. In addition, many pieces bear Koranic inscriptions and poems related to their function, both practical and symbolic. Most surviving pieces date after 1400, and inscriptions provide one of the few means of localisation as styles of weaponry, particularly those used by successful warriors, spread quickly to become international. The largest amount of material is now in Istanbul. The Ottomans maintained their royal treasury in the Topkapı Palace, and their royal pieces, including a few early examples of Islamic arms and armour, are preserved as part of the relics collected by the Ottoman sultans and stored in the Pavilion of the Holy Mantle. The Topkapı Collection is now being catalogued, and a forthcoming publication by David Alexander, Ludvik Kalus and Hulya Tezcan will include information on the inscriptions. In addition to the royal collection in the palace, the Ottomans also amassed a large collection of utilitarian weaponry, including booty from campaigns against the Aq Qoyunlu, the Safavids, the Mamluks, and the Europeans in the Balkans. Much of this was stored in the palace arsenal in the converted Byzantine church

— 199 —

— ISLAMIC INSCRIPTIONS — of St Irene. In this way, Istanbul became the major repository for Mamluk weapons, many inscribed with the names of the last sultans, as well as European weapons that had been kept in the arsenal at Alexandria. Kalus (1982) has published the inscriptions on many of these pieces. Other Islamic weapons were captured in campaigns in Europe and are scattered in military and art museums and private collections in Europe and America. Catalogues provide the best introduction to the inscriptions on this material. One of the most recent publications that includes detailed information on the inscriptions on individual pieces is David Alexander’s 1992 catalogue of some of the arms and armour in the Khalili collection. This luxury volume include not only glossy colour photographs and drawings in which the inscriptions are usually legible, but also a separate appendix giving the inscriptions and translations on each piece and a lengthy bibliography. Kalus (1982) has also published many of the pieces in European collections, such as those in the Tower of London. Many Nasrid weapons were captured at the battle of Lucena in 888/1483 when the troops of Diego Fernández de Cordóba defeated Muhammad XII, known as Boadil. These spoils remained in the hands of Fernández’s family until the beginning of the twentieth century when they were divided between the Museo del Ejército and the Real Armería in Madrid. Some of the fanciest were included in the 1992 exhibition Al-Andalus (Dodds 1992). Similarly, the Hermitage possesses many of the weapons captured in the Russian campaigns in Central Asia in the nineteenth and twentieth centuries, and several of the fanciest examples were exhibited in the 1990 show, Masterpieces of Islamic Art in the Hermitage Museum. Elgood’s (1979) book on Islamic arms and armour contains several important essays showing different ways that inscriptions can be used in dating and localising pieces, sets or groups of pieces. Anatol Ivanov, for example, discussed the Persian inscriptions on a group of Iranian daggers dating from the late fifteenth to the seventeenth century. Their steel blades were

inscribed with poems in various metres describing or glorifying the dagger and using the word khanjar (‘dagger’). Ivanov was able to use epigraphic style to suggest a chronological development for the daggers. The inscriptions are all written in nastaÆlÈq script set in cartouches, but the plain ground typical of the fifteenth century and the first half of the sixteenth became increasingly filled with vegetal designs, and a scroll ground was introduced in the seventeenth century. Yuri Miller discused the Russian inscriptions on a group of seventeenthcentury Iranian swords in the Hermitage. All made in Isfahan, the swords were apparently brought to Russia where they were inscribed by local craftsmen, probably at the armoury, with details about their provenance and date of acquisition. Helmut Nickel used the blazon on a Mamluk axe in the Metropolitan to discuss the evolution of the type. Melikian-Chirvani contributed several articles on metal arms. In one, he discussed the supplicatory prayers known as duÆåæ inscribed on a group of medieval Iranian brass discs with central bosses that he identified as bucklers, or small shields. In another, he discussed a large steel axe (tabar) inscribed with Persian and Turkish verses. In a third and longer article, he used signatures on saddle-axes to identify the work of LutfÆali, an Iranian craftsman active during the reign of the Qajar monarch Nadir Shah (1736–47). Done in a distinctive style and of the highest quality, these saddle-axes could be distinguished from modern imitations also signed with LuftÆali’s name. The basic work on signatures is Mayer’s 1962 volume on Islamic armourers and their works, but, as Melikian-Chirvani’s essay shows, signatures alone are not a guarantee of authenticity, for the work of these famous craftsmen was often copied or purposefully faked. Inscriptions in Arabic can also be used to date other types of arms and armour from outside the Islamic lands. This is the case with the large corpus of European swords with Arabic inscriptions kept in the arsenal of Alexandria and later transferred to the Ottoman arsenal of St Irene in Istanbul. Some were captured in battle; others were gifts from European princes. The blades were

— 200 —

— TYPES

OF

engraved with the names of the Mamluk sultans, and the dedicatory inscriptions are important evidence for dating these European swords. As with inscriptions on monuments and other objects, the primary language used to inscribe Islamic arms and armour was Arabic. It was used for Koranic texts, which were chosen to suit the garment and occasion. Body armour, for example, was often covered with Koranic inscriptions that reflected the wearer’s piety while engaged in jihåd. Talismanic shirts worn under the armour were covered with Koranic verses, prayers and sequences of numbers or letters meant to protect the wearer through spiritual efficacy. These texts also had apotropaic power. Both armour and shirts, for example, are often inscribed with the phrase from Koran 61:13 asking for help from God and a speedy victory (naßr min allåh wa fat˙ gharÈb) or with the opening verses of Koran 48 (Surat al-Fath), the chapter of Victory. Ottoman talismanic shirts preserved in the Topkapı Palace collection (see Figure 14.85) are almost completely covered with Koranic inscriptions, prayers and magic squares and numbers. Several shirts attributed to Safavid Iran, by contrast, are inscribed with the names of the Twelve Imams. By donning such a shirt, the wearer symbolically clothed himself in the ‘raiment of righteousness’ (libås al-taqwå) mentioned in Koran 7:26. Arms and armour made for kings and courts were often decorated with the owner’s name or his dynastic mark. A belt buckle in the L. A. Mayer collection, for example, is inscribed with the name of the Ayyubid prince Abuæl-Fida IsmaÆil (b. 1273). One of the earliest identifiable pieces of mail and plate armour is a shirt inscribed with the name of the Timurid prince Ibrahim Sultan (Lentz and Lowry 1989, no. 31). Pieces made for the Nasrids (many illustrated in Dodds 1992) often have the Nasrid slogan or stylised patterns derived from it. Ottoman pieces often had the tughra, or calligraphic emblem, of a particular sultan. Pieces belonging to the royal collection were engraved with a circular mark with several bars or strokes to show that they belonged to a particular arsenal. Drawings of several examples are illustrated in Alexander (1992).

OBJECTS — Sometimes the name or titles naming the owner ran from one piece to another in a set of armour. This is the case, for example, with a pair of legguards in the Metropolitan. The inscription on a Timurid face mask (Alexander 1992, no. 25) may have been designed to be read together with letters and words on a now-lost helmet that would have overlapped the mask in several places. In some cases, particularly of royal armour such as the shirt made for Ibrahim Sultan, the inscription was probably drawn up by a master calligrapher. In other cases, however, the fragmentary words or individual letters may be the result of copying by illiterate armourers, suggesting that the sanctity of script was more important in warding off injury than the actual message. In some cases, the script used for these inscriptions naming the owner is archaic, perhaps used to invoke the noble lineage of the wearer and associate him with glorious warriors of the past. Several helmets and other objects datable to fifteenth-century Anatolia, for example, are inscribed in a stylised Kufic of the type that had been common two centuries earlier. In later times, Arabic texts were accompanied or replaced by poems in vernacular languages. Language alone is not a key to provenance, for Persian verses were inscribed not only on weapons made in Iran but also on those made for the Ottomans and the Mughals. SaÆdi’s verses were popular with Iranian armourers, and many pieces have survived from the Qajar period. Other Persian verses invoke renowned heroes from literature, such as Rustam’s father Sam, Faramarz or Isfandiyar. Turkish was also used on arms made for the Ottomans, as on a dagger in Edinburgh (Arts of Islam 1976, no. 232) dated to the turn of the sixteenth century which is inscribed with the first couplet of a ghazal by the Ottoman poet Mecati (d. 1509). In addition to verses by famous poets, other poems were written for the occasion. Some name the type of weapon, as with the daggers published by Ivanov or the axes by Melikian-Chirvani. Other verses refer to the object’s function. Inscriptions on Ottoman guns, for example, contain generalised verses referring to the power of the weapon and sometimes to the specific enemy.

— 201 —

— ISLAMIC INSCRIPTIONS —

14.85 Detail of a talismanic shirt painted with Koranic verses. Length 122.5 cm. Istanbul, mid-sixteenth century. Istanbul, Topkapı Palace 13/1150.

— 202 —

— TYPES

OF

The Turkish verse on one gun in the Khalili collection (Alexander 1992, no. 76) says that ‘the new type of weapon that I hold in my hands is an enemy of the German [that is, Hapsburg] court’. These poems often contain puns and plays on words. The Persian verse on a seventeenth-century Ottoman pistol in the Khalili collection (Alexander 1992, no. 68), for example, contrasts the tiny size of the gun to its ability to bring down even the greatest of heroes. The play on words revolves around the word qundåq, which means both ‘swaddling-clothes’ and ‘stock of a gun’. The texts are sometimes written in the first person, as though the weapon were speaking. A fluted steel helmet decorated with gold wire made for the Ottoman sultan Bayazid II and now in the Musée de l’Armée, Paris (Arts of Islam 1976, no. 229), is inscribed with an Arabic text saying that ‘I [that is, the helmet] am the head-piece (zåjil) for the head of the imam’ Sultan Bayazid.

Seals, Talismans, Amulets and Other Small Ornaments Sigilography, the study of seals, is an important branch of Islamic epigraphy, for in contrast to pre-Islamic examples, seals made in Islamic times are almost exclusively decorated with inscriptions, especially during the first eight or nine centuries of Islam. Many Islamic seals are made of gemstones, which are also used for talismans and other ornaments, but seals are usually carved in intaglio, whereas talismans and other ornaments are carved in relief. The seals were usually used to record an owner’s mark and hence were designed to be readable yet distinct. They are therefore useful for studying the general evolution of palaeography. Most seals made before the sixteenth century, however, lack dates or identifiable names, and as the seals are often dated by the style of inscription, the argument can become somewhat circular. To date these early examples, it is often useful to compare the inscriptions to those found on tombstones and coins. Serious study of Islamic seals began in the early nineteenth century with Reinaud’s publication (1828) of the seals and talismans in the collection

OBJECTS — of le Duc de Blacas. The two-volume work included many examples as well as a general introduction to engraved stones, their inscriptions and their uses. Hammer-Purgstall’s handbook (1849) on seals appeared shortly thereafter. Given the large number of seals and their wide geographic currency, later publications concentrated on seals from a particular time or place, as with Rabino di Borgomale’s book (1945) on Iranian seals from the Safavid period onwards and Uzunçar∞ılı’s book (1959) on Ottoman seals in the Topkapı Palace. In the late twentieth century, museums have begun to publish their collections of seals, as for example the catalogue (1975) by al-Naqshbandi and al-Hurri on the seals in the Baghdad Museum or Kalus’s catalogues on seals in the Cabinet des Médailles in the Bibliothèque nationale (1981) and on seals and rings in the Ashmolean Museum (1986). Seals, rings and gems are a favourite of rich private collectors, and the recent glossy publications of the pieces in the hands of such private collectors as Benjamin Zucker (Content 1982) and David Khalili (Wenzel 1993) often contain material on the inscriptions. Kalus’s article on seals in the Dictionary of Art (vol. 16, pp. 542–3) gives a good summary of the texts and scripts used on Islamic seals and is the basis for this article. Seals from the pre-Mongol period generally contain brief, sober texts with three kinds of information. Some are religious in content, with Koranic quotations, prayers to God, Muhammad or ÆAli, or moral aphorisms. Others have the owner’s name accompanied by a set religious expression indicating his relationship to God and his humility and resignation. Still others have only the owner’s proper name, usually in the traditional Islamic pattern of ‘A son of B’, making it almost impossible to identify the individual concerned. These seals from the pre-Mongol period are usually engraved in Kufic. The earliest seals made between the seventh and ninth centuries show a horizontal base line, and this style was consciously revived in later periods when owners wished to imitate the earlier style in order to add magical power to the usual, signatory function of the

— 203 —

— ISLAMIC INSCRIPTIONS — seals. During the tenth century, some letters became more rounded. This change can be traced in inscriptions containing the name of the owner and his father, as the bowl of the letter nËn in the word ibn (‘son’) became deeper and rounded. Later, probably during the tenth or eleventh century, the bowl was transformed into a long oblique stroke that descends below the base line. At roughly the same time, certain tails and loops climb towards the top of the engraved space. The rising tail is a distinctive feature for dating these objects. Small decorative elements, such as schematised leaves or barbs, were also added. They appear most often on the lower end of the letter nËn, but are sometimes found on other characters that extend below the base line or on the tips of the long upstrokes. The letters gradually lost their rigidity, and the script became more curvilinear, either in the twelfth century or possibly earlier in the eastern Islamic world as suggested by coins from the late tenth century.

Seals dating from the thirteenth century show a new style that becomes more pronounced by the fifteenth century. Inscriptions became considerably longer. Decoration, hitherto limited to a few schematised motifs such as a six-pointed star formed by three short crossed strokes, became more naturalistic and filled the area more completely, including the spaces between the characters. The variety of gemstones used was restricted, with carnelian becoming by far the most popular. The range of languages widened, and in addition to Arabic, the inscriptions, usually short poems, could be written in Persian, and later in Turkish. From the sixteenth century, many inscriptions include the year in which the seal was made, but these dates are sometimes difficult to read. The summarily executed figures can be confused with the overall decoration, and the cipher for ‘thousand’ is often omitted as obvious and has to be supplied where appropriate.

— 204 —

— CHAPTER 15 —

Sources, Methods and Conventions

T

he previous parts of this book survey and summarise what is known about inscriptions on buildings in and objects from the Islamic lands. This part is more practical: it aims to show the reader how to find information about a particular inscription and how to record a new text. A first section describes and evaluates general reference works about Islamic epigraphy. These include volumes dealing with historical information, notably the MCIA (Matériaux pour un Corpus Inscriptionum Arabicarum), the series initiated by van Berchem in the early twentieth century, and the RCEA (Répertoire chronologique d’épigraphie arabe), the chronological list of Arabic inscriptions begun under the editorship of Étienne Combe, Jean Sauvaget and Gaston Wiet in 1931. This opening section is followed by one explaining how to identify and interpret Koranic and other religious texts. A subsequent section treats the different calendars used for dates in inscriptions. Finally, a short section gives the different symbols used in recording an inscription.

General Reference Works The massive tomes of the MCIA are the core material for studying Islamic epigraphy, and, to do much work on the subject, one must be able to exploit them fully. The first problem is to find them in the library. Sometimes they are catalogued by the names of the author(s), sometimes as a separate series (MCIA), but most often as volumes in a series of Mémoires published by the Mission archéologique française au Caire, which later became the Institut français d’archéologie orientale

du Caire. Van Berchem’s opus on Cairo (MCIA Egypte 1), for example, is volume XIX of the Mémoires de la Mission archéologique française au Caire, while Wiet’s continuation (MCIA Egypte 2) is volume 52 in the same series, but now published as Mémoires de l’Institut français d’archéologie orientale du Caire (often abbreviated MIFAO). Thus, the two works do not appear next to each other on the library shelf, and finding one volume in the MCIA does not ensure that one can quickly find the others. (For complete information on the volumes as given in their tables of contents, see ‘Volumes in the MCIA’ at the beginning of this book.) Next is the question of how the individual volumes are organised. They concentrate on architectural inscriptions, which are arranged in chronological order according to the date that the building was constructed. Repairs and renovations to a building are listed along with the initial entry on the building. Thus, Mamluk repairs to a Fatimid building such as the Azhar Mosque in Cairo (MCIA Egypte 2, nos 21–7) will be found not with other fourteenth- or fifteenth-century inscriptions but following the tenth-century entry on the mosque (MCIA Egypte 2, no. 20). Each entry is divided in two parts, one on epigraphy, the other on the archaeology of the building. A bold heading at the beginning gives the name of the building and the date, generally of completion. This is followed by a brief indication of the building’s location within the city. Each inscription within the building is listed separately. The entry begins with summary remarks about where the inscription is found within the building;

— 207 —

— ISLAMIC INSCRIPTIONS — the dimensions; the size (small, medium or large) and type (for example, simple, floriated and decorative Kufic; Ayyubid, Mamluk and Ottoman naskh) of the letters and the material on which the inscription is carved in relief (unless noted otherwise); the state of preservation; the number of lines and divisions; and the presence or absence of diacritical marks and other orthographic signs. This heading is followed by a transcription of the text in Arabic characters. Next comes a summary translation and critical analysis of the text, including any important historical, philological or archaeological deductions that can be made about it. Plates with photographs of rubbings or squeezes of the inscriptions at the end of each fascicule allow the reader to judge the style. The first volumes of the MCIA to be published, van Berchem’s work on Egypt (MCIA Egypte 1), give a good example of how these publications are organised. Following an introductory essay on the need for a corpus and a manual of Arab archaeology, van Berchem’s opus contains 545 inscriptions, all verified by the author in situ. The inscriptions are listed chronologically within dynastic headings, and the work is divided into four fascicules published between 1894 and 1903. The first fascicule covers buildings founded by the Umayyads, Tulunids, Fatimids and Ayyubids, from the foundation of the Nilometer in 97/ 715–6 to the madrasa and tomb founded by Salih Najm al-Din Ayyub in 641/1243. The second fascicule covers those buildings founded by the Bahri Mamluks, and the third, those founded by the Circassian Mamluks. The fourth fascicule treats a handful of Ottoman buildings as well as sixty objects, mainly architectural furnishings, from the Cairo Museum. Wiet’s supplementary volume on Cairo, MCIA Egypte 2, adds fifty-four inscriptions from the earliest Islamic times through the Fatimid period. One point already clear from these first volumes of the MCIA is the interest shown by scholars and epigraphists of this time in the earlier periods of Islamic history. Ottoman inscriptions, for example, are treated only summarily. There were several reasons for this emphasis on early and medieval Islam. In part, it reflects the

nineteenth-century viewpoint when scholars looked back to the eighth and ninth centuries as the golden age of a unified Islam, and many scholars considered Egypt in the later centuries merely a province of the Ottoman empire. In part, this emphasis also reflects the value of the early inscriptions, which are usually distinct and discrete and often the only contemporary evidence for certain events, as few written sources survive from the early Islamic period. By contrast, inscriptions from later times are far more plentiful and often supplemented or even replaced by contemporary accounts written from a number of viewpoints. While this emphasis in the MCIA on early or medieval Islamic times may be understandable, its impact is aggravated by the chronological organisation from earliest times onwards, with the result that the entries for the later period were often never completed. Wiet’s additions and corrections to van Berchem’s opus, for example, never reached even the Mamluk period. The MCIA is most useful, then, for the earliest times, but one needs to look elsewhere for later material. Despite this limitation, the volumes of the MCIA are of extraordinary importance for historical studies, as information from the inscriptions themselves is combined and coordinated with that supplied by texts. Wiet’s long entry on the earliest building in Cairo, the mosque of ÆAmr b. al-ÆAs (MCIA Egypte 2, pp. 1–16), for example, contains no preserved inscriptions, which are all modern, but rather those recorded in medieval texts, of which he had an extraordinarily wide and deep grasp. A second advantage of the MCIA is its superb indices, which allow the reader to find a wealth of information without reading through the entire volume. Each complete work, such as van Berchem’s four volumes on Cairo (MCIA Egypte 1), contains a single index, and a perusal of the index to van Berchem’s opus gives the reader some idea of its vast scope. The index, which took van Berchem two years to complete, is 110 pages long and contains more than 50,000 references. It includes all proper names of people and places mentioned in Arabic or French in the

— 208 —

— SOURCES, METHODS inscriptions and commentaries; the surnames and titles mentioned there; eulogies and formulas of benediction and malediction; all the Arabic words in the inscriptions; and all words in the commentaries that relate on the one hand to epigraphy, archaeology, architecture and construction and on the other to religious, political, administrative and judicial institutions, especially diplomatics and titulature, in other words to all the terms that relate to history in its broadest sense. All of the words are transcribed into French, so that one does not even need to know the Arabic alphabet to use the index. To see how useful van Berchem’s index is, we can use the example of the word Æabd. Following the main entry Æabd (‘slave’, ‘servant’), one finds a list of references first to nine entries in which the word appears in the Arabic text (several marked with an asterisk to indicate that the word appears several times in the same entry) and then references to three separate pages on which the word is discussed in the commentaries. This list for Æabd is followed by sub-entries for three different plurals (Æabadah, Æabåd and Æibåd), each with references to entries and pages. The main entry is followed by eight entries in which Æabd occurs in a phrase (for example, the title Æabd allah or the phrases alÆabd al-faqÈr and Æal-Æabd al-faqÈr ilå Allåh taÆålå), twenty-four entries of proper names beginning with Æabd (for example, ÆAbd al-ÆAziz, ÆAbd al-Baqi), and finally a separate entry on the proper name ÆAbdallah. Entries in bold point the reader to significant discussions of the term, and cross references within the indices lead to further discussions. Lest the reader miss something or forget how to use the index, a nine-page introduction lays out the categories and forms. For the novice, a few pointers make it easier to get around this extraordinary research tool. It is important to remember that slightly different systems of transcription were used in the different volumes. Some indices, for example, transcribe qåf with a q, others with a ˚; some transcribe shÈn with a ch, following the French system; others use sh following the system used in the Encyclopaedia of Islam. These are minor differ-

AND

CONVENTIONS —

ences, but should one not find any entries under an appropriate letter (for example, no words such as shaykh beginning with sh in van Berchem’s index), then one should use a modicum of sense and root around elsewhere. This is worth doing, since it is impossible to overestimate the treasures contained in these volumes. They are the fundamental works for the study of Islamic epigraphy, and their major limitation is that they cover only a few sites in the central Islamic lands. The RCEA is a more comprehensive list that sets out to amass a corpus of datable historical inscriptions in Arabic. It thus includes inscriptions on portable objects as well as those on buildings, although coins, glass weights, papyrus, manuscripts and non-dated signatures on ceramic shards and manufactured objects are deliberately omitted. Each of the original sixteen volumes of the RCEA contains 400 entries, arranged chronologically year by year up to 762/1361. Within the year, the inscriptions are generally arranged from west to east in the following order: Sicily, Spain, Morocco, Algeria, Tunisia, Egypt, other parts of Africa, Palestine, Syria, Arabia, Asia Minor, European Turkey, Armenia, Mesopotamia, the Caucasus, Persia, India and China. Inscriptions containing a ruler’s name but no date are grouped together in the last year of his reign. Other undated inscriptions are put at the beginning, middle or quarter of a century according to the hegira calendar. The entries in the RCEA are much briefer than those in the MCIA, basically a simple list without commentaries. The typical entry has a number and identification of type (for example, construction text, restoration text, funerary text, epitaph, graffito, signature, royal inscription or portable inscription). Below in capital letters is the name of the site or place where the object was made. This is followed by a heading with several brief phrases about location, size, material and script. Two lines in italics list citations where the text was published and illustrated. The main entry gives the Arabic text, followed by a translation into French. Optional last lines list citations where the text is discussed and illustrated.

— 209 —

— ISLAMIC INSCRIPTIONS — One of the major gaps in the original sixteen volumes of the RCEA was the lack of any indices. To find a particular inscription, the reader had to flip through the volumes, and the geographical indications were so vague that even the experienced scholar might overlook a text. To remedy this problem, a geographical index with an alphabetical list of sites was published in 1975. In it, each site is identified in several authoritative sources, ranging from the Times Atlas of the World and the Encyclopaedia of Islam to the classical manuals of geography such as Le Strange’s 1905 volume on the lands of the eastern caliphate. For each site in the RCEA, the inscriptions are grouped by volume, with the number of the inscription, its date and its type. Inscriptions in museums are listed separately after those in the city itself. In addition to making the inscriptions in the RCEA more readily available, the geographical index is useful in itself. It contains precise geographical identifications of obscure sites and useful cross-references among the various names of a single site. Crac des Chevaliers, for example, is cross-referenced to the medieval name of the site, given with full vocalisation as Óißn al-Akråd. In looking up Dzhambul, the reader is referred to Awliyaæ-Ata, where the site is identified as a city in the north of Farghana province, some 275 km as the crow flies north-north-west of Uzgand, with references to the Times Atlas and Le Strange. Looking over all the entries to a single city, one can also get a sense of historical change through its inscriptions. For Cordova, for example, there are almost no inscriptions from the first three volumes. With volumes IV and V, one can trace the expansion of the great mosque there, and hence the growth of the Muslim community, which receded in importance in succeeding centuries, with only a few epitaphs found in the later volumes. Not even a good geographical index, however, could solve all the problems of the original sixteen volumes of the RCEA. It was often impossible to check the accuracy of the entries, which were submitted by scholars throughout the field. The entries represent the best that was available at the time, but they are often inconsistent and the

texts unverified. Sometimes, for example, the same text was included twice or even three times because different scholars had read the date differently (for example, nos 5210bis, 5224 and 5373bis, all recording the same inscription on the portal of a khanaqah at Natanz). Hence, it was decided to adopt a more critical attitude when publication of the RCEA resumed in 1982 with volume XVII. A new numbering system was introduced, with the hegira year followed by a numerical list. Thus, inscription 762 001 is the first inscription under the year 762. Headings were expanded to include more information about location, state of preservation, material and script. A geographical index was included at the end of each volume. The biggest changes occurred in the text and translation. The text is reproduced more accurately line by line, and Koranic texts are given in full when their exact length is known. The text is also more critical. The source used is given, along with variants from other sources or suggestions by the editor. Translations are more consistent, with an attempt made to use one French word for one Arabic one. Brief commentaries are also included to explain how the date was established for an undated inscription. The scope of this monumental enterprise is clear from the ongoing history of publication of the RCEA. Volume XVII, containing approximately 400 inscriptions from the years 762/1361 to 783/1381, covers only twenty years of Islamic history. Volume XVIII, published almost a decade later, carries the work to the year 800/1398. At the rate of a new volume that covers twenty years of Islamic history appearing every decade, it will be three centuries before this enterprise reaches modern times!

Koranic and Other Religious Texts Although historical inscriptions were the most interesting to scholars compiling the MCIA and RCEA, they are outnumbered by religious inscriptions, of which the most common are citations from the Koran. Koranic inscriptions appear on all types of Islamic architecture, from

— 210 —

— SOURCES, METHODS

AND

the earliest buildings, such as the Dome of the Rock built in Jerusalem in 72/692 (see Figure 1.4), to modern ones, such as the mosque built by the Saudi King Khalid at Riyadh airport in 1984 (see Figure 1.5). Koranic inscriptions also occur on many objects made at all periods and in all areas of the Islamic lands. In early scholarship, these Koranic texts were sometimes passed over as banal or irrelevant. Several kinds of evidence, however, show that this idea is wrong and that Koranic texts were considered meaningful. Studying them, just like studying historical inscriptions, can shed light on the meaning and function of a building or object. Wiet (MCIA Egypte 2, p. 20) cited an unusual example of written justification for the selection of a Koranic text on a building: the engineer in charge of constructing the Nilometer in Egypt received a message from the caliph telling him to have Koranic verses with texts appropriate to the Mikyas inscribed on it. The engineer chose specific Koranic verses, noting that it was impossible to find better or more appropriate texts to inscribe. Size also suggests the importance of Koranic inscriptions, for sometimes the Koranic inscription is longer than the historical text on a building, and we can assume that the Koranic text is not merely decorative but meaningful. A good example is the 60-metre minaret at Jam in central Afghanistan (see Figure 15.86). Its lower shaft is decorated with an interlacing band containing chapter 19 of the Koran, the sura of Maryam. Any explanation of why this extraordinary minaret was built in the middle of this remote valley must justify why the patron went to the trouble of making an artisan write out all 976 words of this particular sura in such a prominent place on the building. In order to analyse Koranic inscriptions on buildings and objects, the researcher needs to understand the structure and arrangement of the Koran, particularly some of the peculiarities in the numbering systems used. Using a concordance, he can identify the text contained in the inscription, and he can then interpret the text and explain why it was chosen for that particular place.

— 211 —

CONVENTIONS —

15.86 Minaret of Jam, Afghanistan, late twelfth century.

— ISLAMIC INSCRIPTIONS — The Koran is accepted by Muslims as God’s word revealed to the Prophet Muhammad in Arabic. It was first transmitted orally, but, according to traditional accounts, under the third caliph ÆUthman (r. 644–56) the revelations were collected and assembled in a book. The text comprises 114 chapters or suras (Arabic sËra), which are divided into a number of verses (åyåt). The suras are arranged roughly in order of length from longest to shortest, except for the first sura, which is a petitionary prayer (duÆåæ). Variant readings of the text are possible, and by the tenth century seven readings were generally accepted. The verses can also be counted in different ways. Until recently, the text of the Koran most widely used in the West was that published by Gustav Flügel in 1834. It does not correspond to any one traditional reading, for in about half the suras Flügel changed the verse divisions in order to establish an improved text. Flügel’s edition was the basis for the English translations by R. Bell (Edinburgh, 1937–9) and A. J. Arberry (London, 1955). In 1924, a standard Egyptian edition of the Koran, known as al-Muß˙af alsharÈf or al-Quræån al-karÈm, was adopted. This is the version used in most regions of the Islamic lands today, except for parts of North Africa. The German translation by R. Paret and the French translation by R. Blachère give the numbers in both the Flügel and the Standard Egyptian editions. The translation by Marmaduke Pickthall, The Meaning of the Glorious Koran (London, 1930, reprinted 1957), follows an Indian tradition that deviates from the standard Egyptian text in a few places. Bell’s Introduction to the Quræån, revised by William Montgomery Watt (1970: 202–3), includes a table for converting verse numbers in Flügel’s edition and the Standard Egyptian text. Further details on the history of the Koranic text and the various readings and systems are given in the article Æal-uræån’ in EI 2. In addition to the various readings and numbering systems, there are two ways to refer to the suras. Muslims traditionally refer to them by name. The name is usually derived from a key word or idea in the sura. For example, the Surat al-Fath (Sura of Victory) takes its name from the

first line that promises a clear victory (fat˙an mubÈnan). Since the names were not established in Muhammad’s lifetime and did not come to be regarded as part of the text, many suras have more than one name. With the acceptance of the Standard Egyptian edition of the Koran in the 20th century, uniform names for the suras have become more prevalent, although a few suras are still known by two names (for example, Banu IsraÆil and al-Israæ for the seventeenth sura). By contrast, Western scholars usually cite the suras by number. Thus, the Surat al-Fath or Sura of Victory is the forty-eighth sura. Traditionally, scholars used Roman numerals (either capital or small) to indicate the sura, followed by Arabic numerals for the verse. Thus, the first verse of the Surat al-Fath is cited as XLVIII:1 or xlviii:1. This system leads to frequent errors, since Roman numerals are cumbersome and a digit is often dropped. A better way to record the sura and verse is by using Arabic numerals separated by a colon or a full stop, thus Koran 48:1 or Koran 48.1. A pious believer or learned scholar can often identify a Koranic inscription from a single word or phrase. Sometimes, however, it is difficult to read the text, particularly if it is written in unpointed angular script, and sometimes the researcher does not recognise the text. The simplest way then to identify a Koranic text is to use a concordance, which lists the words in the Koranic text under the appropriate Arabic root. The concordance is arranged like an Arabic dictionary, alphabetically by root and then by parts of speech. Thus to identify a text, the researcher might recognise the word fat˙ at the beginning of the inscription from the distinctive shape of the word. He or she can then look up the word under the root f-t-˙ in a concordance. In order to use a concordance, the researcher must consider the different numbering systems, particularly those used by Flügel and in the Standard Egyptian edition. Each has its own concordance: Flügel’s Concordantiae corani arabicae uses the numbering system which he established in his Corani textus arabicus, while ÆAbd al-Baki’s al-MuÆjam al-mufahras li-alfåΩ al-quræån al-karÈm

— 212 —

— SOURCES, METHODS uses the numbering system of the Standard Egyptian edition. ÆAbd al-Baki’s concordance is far easier to use, as it does not merely give the appropriate numbers of the sura and verse as Flügel’s concordance had, but also gives the phrase in which the word appears. Therefore the researcher can find the word he or she recognises and compare the surrounding words in the inscription to those on the printed page without flipping through the Arabic text of the Koran. Flügel’s edition has one minor advantage: it lists every word in the Koran, including particles such as inna, and so it is possible to look up even the most minor of words. When recording or discussing a Koranic inscription, it is essential to document which edition and which concordance are being used. Today it is easier and more common to use the numbering of the Standard Egyptian edition and ÆAbd al-Baki’s concordance, but it is important to remember that most Western scholarship done in the late nineteenth and early 20th century (including the volumes of the MCIA) is based on the Flügel numbering. Therefore, if one finds a reference in van Berchem’s MCIA Egypte 1 to Koran 2:256, the researcher must recognise that in the Standard Egyptian text this is 2:255, the famous Throne Verse. Sometimes the difference is insignificant; sometimes it is crucial. The fullest way of citing this verse would be 2:256/ 255, but as the Standard Egyptian edition is increasingly replacing the Flügel edition, 2:255 is sufficient. It is also worthwhile to record whether the entire verse is present or only part, as arguments about the meaning of a particular inscription can depend upon the inclusion of a particular word or phrase. Having identified the sura and verse in an inscription, the researcher may then want to interpret why a particular text is there. To do so, it is necessary to determine whether the citation is common or rare in inscriptions and in which locales and situations. For architectural inscriptions, one should begin by consulting Dodd and Khairallah’s two-volume collection of Koranic citations on buildings, The Image of the Word. The first volume contains several case studies

AND

CONVENTIONS —

on the iconography of Koranic epigraphy (the Dome of the Rock in Jerusalem, the Nilometer in Cairo, the hospital of Nur al-Din in Damascus, and the madrasa complex of sultan Hasan in Cairo) and essays on the use of particular Koranic texts and their placement within the architectural setting. The second volume contains the documentation: three lists of Koranic inscriptions, the first arranged by number of sura and verse, the second arranged by geographical site of the monument, and the third arranged by location within the building. To collect their citations, the authors culled published sources. Their work therefore reflects the state of the field, in which material from Egypt, the Levant and India is better recorded than that from other areas. Inscriptions from North Africa, Turkey, Iran and Central Asia are woefully under-represented in their corpus. Sometimes the authors did not know whether the citation came from the Flügel or the Standard Egyptian edition, as they could not verify all 4,000 citations included. Before generalising from their indices, the researcher is advised to verify the particular verses cited in the individual inscriptions. Documentation for Koranic verses inscribed on other media is available only piecemeal in studies of particular media or periods. Koranic verses were frequently inscribed on tombstones; Dodd and Khairallah’s work includes those published in the MCIA and the RCEA, but one might also want to consider those from elsewhere, looking, for example, at the publications by Afshar (1969–75, 1973) and Lowick (1985) on Iranian tombstones from Yazd and Siraf or by Karama©arali (1972) and Rogers (1988) on the Turkish ones from Ahlat. Koranic verses were popular as borders on lustre star tiles, and Watson (1975: 64–5) gives a brief list and analysis in his survey of lustrewares from Iran. Koranic inscriptions are rarer on metalwork; Baer (1983: 212–13) makes some general points, with references, for the metalwares up to the mid-fourteenth century. See also the comments by Melikian-Chirvani (1976). Sourdel-Thomine (1971) analyses the Koranic verses used on keys to the KaÆba.

— 213 —

— ISLAMIC INSCRIPTIONS — A few Koranic citations that epitomise important aspects of the faith are often known by epithets and have always been popular in inscriptions. Koran 2:255, the clearest statement of God’s majesty, is known as the Throne Verse (åyat al-kursÈ) and is the most frequent Koranic text cited in inscriptions. Koran 9:33, which says that God has sent his messenger with guidance and the religion of truth, is known as the Prophetic Mission. Koran 24:35, in which God is described as the light of the Heavens and the earth, is called the Light Verse (åyat al-nËr). Other popular texts include Koran 3:18, which includes a paraphrase of the first part of the profession of faith (shahåda) that there is no diety but God; sura 48 (al-Fath, Victory), which promises victory to believers; and sura 112 (al-Ikhlas, the Purifying), which describes God’s oneness. These texts were undoubtedly popular in inscriptions because they sum up important aspects of the faith.

Other verses were chosen for inscriptions because they underscored the purpose of the inscription or the function of the building or form. Thus, Koran 2:181, about the inviolability of waqf, is often included in endowment texts. Koran 9:18, which says that the person to maintain God’s mosques is he who believes in God, prays and gives alms, is one of the most frequent inscriptions on mosques (R. Hillenbrand 1986). The verse is one of the few Koranic texts referring specifically to God’s mosques (masåjid allåh), and of these, it is the most relevant to patrons. Indeed, its presence in an architectural inscription can be said to identify a building as a mosque. Similarly, Koran 17:78, about prayer and vigil, is a common text on mihrabs. In a few cases, a Koranic verse was chosen because it contained a specific word that referred to the function of a particular object. Thus, the foundation inscription around the huge cauldron from the shrine of

15.87 Bronze cauldron ordered by Timur on 20 Shawwl 801/25 June 1399 for the shrine of Ahmad Yasavi in Turkestan City. Ht 1.58 m.

— 214 —

— SOURCES, METHODS Ahmad Yasavi at Turkestan (see Figure 15.87), called a drinking vessel (al-siqåya), begins with Koran 9:19 about giving water (al-siqåya) to pilgrims. Keys to the KaÆba were often inscribed with 3:90–1, referring to God’s house. Certain verses were considered prophylactic. The last two suras are charms or incantations. According to Edward Lane’s account of contemporary life in Cairo (1836: 247–8), a book or scroll with seven suras – particularly 6, 18, 36, 44, 55, 67 and 78 – was an esteemed charm in nineteenthcentury Egypt. Another charm, usually worn in the cap and believed to ward off the devil and all evil genii, was a piece of paper inscribed with the ‘verses of protection or preservation’ (åyåt al-˙ifΩ). They comprised Koran 2:255, 12:64, 13:11, 15:17, 37:7, 41:12 and 85:20–2. The ‘verses of restoration’ (åyåt al-shifåæ), Koran 9:14, 10:57, 16:69, 17:82, 26:80 and 41:44, all of which refer to a cure (al-shifåæ), were used to charm away sickness. They were written on the inner surface of an earthenware cup or bowl. Water was added and stirred until the writing washed off, and the water infused with the sacred words was drunk by the patient (Lane 1836: 253–4). In medieval Spain, Koran 3:18 and the first words of 3:85 were considered a prophylactic formula for seeking protection from evil (taÆawwudh) and hence were used as a frame on the tombstone of an Almoravid princess who died in 496/1103 (Lévi-Provençal 1931: 31). The interpretation of a Koranic text (and hence its suitability for a given situation) could change over time, and legal scholars and others constantly re-examined and reinterpreted specific verses to justify sectarian, legal or doctrinal disputes and divisions. Several texts were popular with ShiÆites as they were often cited to show their connection to the Prophet Muhammad and to defend sectarian positions. Koran 33:33, with the phrase ‘People of the House’ (ahl al-bayt), was used to vindicate ShiÆite claims to the imamate of ÆAli and his descendants and is found on Fatimid tombs in Cairo and cenotaphs and tombstones from ShiÆite shrines in ÆIraq and Iran. Koran 5:55, which includes the word walÈkum (‘your friend’), also has special meaning to ShiÆites and

AND

CONVENTIONS —

occurs on the same types of buildings. Koran 42:23, which also mentions close kin, was another text appropriate to ShiÆites. Plays on words, so easy to make in Arabic because of the root structure, were another reason for the selection of a particular Koranic citation. Thus, the choice of 17:79, which says that God will raise one to a praiseworthy station (maqåman ma˙mËdan), is a natural one for patrons named Mahmud. Similarly, the succeeding verse, Koran 17:80, which refers to a ‘just ingoing’ (mudkhal ßidiq) and ‘just outgoing’ (mukhraj ßidiq), is appropriate for doorways. Another principle for the selection of a particular text was bracketing or synecdoche, using part to substitute for the whole. Writing the whole Koran was considered a worthy goal, and the carved wooden frieze below the ceiling in the Mosque of Ibn Tulun in Cairo, which is nearly 2 km long, is commonly said to include all of the Koran, although study has shown that there was only space for some one-fifteenth of the book (Creswell 1932 [1979]: 245). By choosing Koran 1 (al-Fatiha) and Koran 114 (al-Nas), one could be said to have written the whole Koran. This may be the reason that the first and last verses were popular on lustre star tiles. Space available was another criterion. For example, it was possible to write out the full text of a short sura from the end of the Koran on a single lustre tile. The longest word in the Koran, fasayakfÈkahum [Allah] ‘[God] will suffice you against them’ (from Koran 2:137), was popular in early Islamic times as a slogan of the ÆAbbasids. According to the rules of the ÆAbbasid court written down by the secretary Hilal al-Sabiæ, RusËm dår al-khilåfa, it decorated the pole supporting the ÆAbbasid banner (1977: text p. 130; trans. p. 77). The verse also occurs on a variety of media in the early Islamic period, ranging from a stucco fragment from Susa to wood from Fustat, an inscribed textile and a tin-glazed cobalt and cream bowl excavated at Samarraæ. The word shows up again on lustre tiles, coins and tombstones of the Mongol period in Iran where it is elaborated arranged as an arch (Miles 1939). The single word could thus be fitted on a

— 215 —

— ISLAMIC INSCRIPTIONS — small object and, with its meaningful text about God’s omnipotence, could be said to encapsulate the Koran. Often a particular verse had overlapping layers of meaning and could be chosen for different reasons in different situations. A good example is Koran 3:18. In addition to its reference to the profession of faith, it also has a funerary association through the word shahÈd (‘witness’, also ‘martyr’), and was popular on tombstones. MuÆtazilite theologians and commentators also cited the verse for the word qis† (justice) as a justification for their doctrine of rationality, and Bulliet (1969) argued that this interpretation was the reason for its inclusion on a large coin struck for Mahmud of Ghazna. Clearly Koranic texts could be interpreted in different ways in different situations, and there were different reasons for choosing a particular verse. Different texts could also be chosen for the same situation. One example of the variety of texts considered appropriate for a given situation is the Koranic texts chosen for tombstones. One expects selected verses to mention death, the Day of Judgment, resurrection and the like, and indeed many of these themes are found. Common verses include Koran 3:185 and 21:35, which say that every soul shall taste of death; Koran 22:7, that the hour will come and God will raise up all who are in their graves; and Koran 55:26–7, that everything is transitory except God’s face. Other verses (for example, Koran 23:114–16 and 39:53) invoke God’s mercy and compassion. What is interesting is how specific verses came into and went out of popularity at different times. Wiet (1952), for example, noted that both Koran 22:7 and 9:33, the Prophetic Mission, were popular on early Egyptian tombstones, but were replaced from 240/855 onwards by Koran 55:26–7. Some scholars (for example, Bloom 1987b) have connected these epigraphic changes with societal ones, notably the advent of ShiÆism under the Fatimids, and seen the choices as reflecting a popular level of taste, one that is not necessarily recorded in written documents. This analysis was criticised by Taylor (1992), but he does not offer any other explanation for

the changes other than a simple shift in epigraphic formulas. These divergent opinions show how difficult and risky it can be to interpret the choice of Koranic inscriptions in any given situation. In many cases, the choice of Koranic verses for inscriptions was quite rote, and the verses were selected for their general popularity. In a few cases, it is also possible to suggest that the choice was deliberate and that the verses were assembled to compose a programme with a specific message. The best-known examples are the mosaic inscriptions in the Dome of the Rock in Jerusalem (O. Grabar 1959; Dodd 1969; Blair 1992b) and the Great Mosque at Damascus (Finster 1970). In order to pursue this line of interpretation, it is essential to show that the verses chosen are different from what one would expect. Only then can one attempt to explain that they are meaningful. There are few reference works available to identify other kinds of religious texts inscribed on buildings and objects from the Islamic lands. The hadith, for example, are usually not found in Wensinck’s extensive Concordance et indices de la tradition musulmane (1936–88), which lists traditions from the canonical sources. The inscribed hadith were clearly chosen to suit the particular site, and some may have been coined for the occasion. The range of authorities cited is quite broad, ranging from traditions ascribed to Muhammad, ÆAli and other ShiÆite imams to those by famous literary or historical figures. Several traditions attributed to Socrates, for example, are cited on the façade of the tomb of Shirin Beka Aga erected in 787/1385–6 at the Shah-i Zinda, the necropolis outside medieval Samarqand. Similarly, the prayer formulas invoking God’s name (Arabic duÆåæ/adÆiya) are not systematically published. The inscriptions on metalwares provide some of the best evidence for these prayers in medieval times, and many are published in Melikian-Chirvani’s catalogue (1982b) of the Iranian metalwork in the V&A. The index to the volume, however, is inconsistent, and the only way to find when or if a particular word or

— 216 —

— SOURCES, METHODS phrase is used is to flip through the text. Recording these hadith, prayer formulas and other texts would be a profitable enterprise. Most seem to reflect popular taste, and they provide a counterbalance to the official history of religious taste provided by standard written authorities.

Dates and Calendars The date, often the piece of information most sought by modern scholars reading an inscription, usually comes at the end of the foundation text. Sometimes it is introduced by a few words such as ‘and this is’ (wa hadhå), ‘this took place in’ (wa kåna dhalika) or ‘with the date of’ (bitårÈkh), but often the date simply runs on after the rest of the historical text. The simplest form reads ‘in the year’ (fÈ sana, occasionally bi-sana). Learning to recognise the distinctive shape of the word sana – either the four teeth and final tåæ marbˆa in an angular script or the swooping curve, tooth and final tåæ marbˆa in a cursive one – can be an easy and effective way to find the date in an inscription. This lustre tile from a frieze (see Figure 15.88), for example, clearly contains a date. The word sana (‘year’) is followed by sabÆ wa sabÆamiæa (‘seven and seven hundred’). The word before sana ends with a låm. It is not part of the word ‘in’ (fÈ or bi-) and belongs to the month. It could be either RabiÆ I (al-awwal) or Shawwal, and by locating other tiles from the same frieze (Watson 1985: 195, no. 103) it is possible to reconstruct the entire text as Koran 76:1–7 and the date Shawwal 707/March–April 1308. The word sana (‘year’) could be replaced by Æåm. This was a common form in North Africa. According to al-Qalqashandi, one used the word Æåm in the Maghrib for apotropaic reasons, because it evoked the idea of fertility, while the word sana was applied to a year of scarcity (MCIA Egypte 2, p. 121). The word Æåm appeared occasionally in the Levant (for example, RCEA 5545), and its use in Mecca and Jerusalem was probably the result of pilgrims. It is exceptional to find it in the eastern Islamic lands, and its startling use

AND

CONVENTIONS —

at a place such as the tomb tower at Bastam in northern Iran demands a singular explanation (Blair 1982: 266–8). When space was at a premium, the ‘centuries’ digit could be dropped. Marshak (1972) proposed that this happened on the ‘Basra ewer’, the inlaid brass ewer in the Georgian State Museum in Tbilisi (see Figure 9.52), and that the date should be read as [1]67 or [1]69 (corresponding to 783–4 or 785–6) or even [2]67 or [2]69 (corresponding to 880–1 or 882–3). Dates on Persian buildings erected in the second millennium of the hegira often have only three numbers, and it is sometimes difficult to tell which digit, usually a zero, was omitted. One must always ask, therefore, whether the date is complete or whether this is only the part of the date for which there was sufficient space available. Although it was most common to give simply the year, some texts name the month. The name can be followed by one of several adjectives (see MCIA Egypte 2, pp. 35–7 for a list of these adjectives). It can also be preceded by the word shahr (‘month’). This was not a random procedure. According to the Mamluk chronicler alQalqashandi, the word shahr could be joined to all months, but was especially useful with RabiÆ and Ramadan. Epigraphy confirms his statement, for Ramadan is rarely written alone and the others are more often written without shahr. Occasionally, an inscription is even more specific, giving the day of the month and even the day of the week. The Muslim lunar calendar, which begins with the year that Muhammad emigrated from Mecca to Medina, is by far the most common one used in inscriptions throughout the Muslim lands. Known as anno hegirae, it is commonly abbrievated AH. Since it is a lunar year, it is some ten and a half days shorter than a solar year. To calculate the correspondence with Common Era (CE) dates, one should consult a concordance. Freeman-Grenville’s little handbook, The Muslim and Christian Calendars (1963) shows various ways to do so. This and other calendars are described in an article by Taqizadeh (1937–9). As most inscriptions were written in Arabic, the hegira calendar was usually cited in Arabic,

— 217 —

— ISLAMIC INSCRIPTIONS —

15.88 Lustre tile from a frieze with the date [Shawwa]l 707/March–April 1308. Kashan. 38.1 x 38.1 cm. New York, The Metropolitan Museum of Art, Gift of Emile Rey, 1912 (12.44).

but in outlying lands the hegira date was sometimes translated into the vernacular language. The small Ghurid mausoleum at Chisht in central Afghanistan, for example, has a Koranic band followed by the hegira date translated into Persian words (Blair 1985): bi-tårÈkh-i dahum jamÈd [sic] al-awwal sål-i qamar pånßad shast dË az hijrat-i payghambar mu˙ammad ßalå allåh Æalayhi (‘dated the tenth of Jamid I of the lunar year five hundred and sixty-two after the

hegira of the Prophet Muhammad, God’s blessing upon him’). Several other calendars were also used in Iran. After the Muslim conquest of Iran, Persians continued to use a solar calendar alongside the official Arabic one, and the solar calendar was used occasionally in inscriptions. The last Sasanian solar calendar went into effect during the year that Yazdigird III acceded to the throne (16 June 632–15 June 633). After his death, it was sup-

— 218 —

— SOURCES, METHODS posed to be replaced by another regnal calendar, but as he was the last Sasanian monarch, this never occurred and the Yazdigird calendar continued. To arrive at the corresponding Common Era date, one should add 631 (for dates between the vernal equinox in March and the end of December) or 632 (for dates between January 1 and the vernal equinox in March) to the Sasanian solar year. There must have been some sliding of dates in medieval times, for this process does not always work. Several tomb towers erected by minor rulers in the tenth and eleventh centuries in out-of-theway sites along the Caspian Sea have inscriptions in Arabic and Pahlavi, with the Arabic date given in the hegira calendar and the Pahlavi one in the Yazdigird era (Blair 1992a, nos 31–2). In one case (the tomb tower at Radkan), the corresponding hegira and Yazdigird calendars are 633 years apart, but in another (the tomb tower at Lajim) the Pahlavi and Arabic texts did not agree. The solar year could also be given in Arabic. The most famous of these tomb towers, the Gunbad-i Qabus erected by the Ziyarid prince Qabus b. Washmgir (Blair 1992a, no. 19), has an Arabic inscription giving the date in the lunar year 397 (which ran from 27 September 1006 to 16 September 1007) and in the solar year 375 (which ran from 15 March 1006 to 14 March 1007). Putting the two together allows us to fix the date of the tomb tower between late September 1006 and mid-March 1007. The Persian solar months were obviously in common use in Iran alongside the official lunar calendar. This is clear from the epigraphic record. An inscription carved for the Buyid Abu Kalijar at Persepolis, for example, records his presence on rËz-i bahman min måh-i åbån (the second of Aban) in the Arabic year 438/1046 (Blair 1992a, no. 43). Similarly, a tombstone from Turanpusht uses a Persian month (måh-i day) with an Arabic year 457/1065 (Afshar 1969–75, vol. 1, no. 174/10). The hegira and Yazdigird calendars were not the only ones used in medieval Iran; the Julian calendar was also used. A inscription engraved in 1041 to honour a Qarakhanid prince in the Waruh Gorge in the Farghana Valley of Uzbekistan uses three calendars: the Arabic hegira, the Persian

AND

CONVENTIONS —

Yazdigird and the Syrian Christian. The inscription gives a precise date in the Arabic lunar calendar of Tuesday 3 Jumada I 433, as well as the Persian month of Day in the year 410, and the Greek month of Kanun I, the Syriac version of the Julian calendar (Blair 1992a, no. 42). The Julian calendar was probably used by Nestorian Christians in the region. The Yazdigird era, which kept slipping backwards with respect to the solar year and needed intercalation, was reformed by the Saljuq sultan Malikshah. The new Jalali or Maliki era was introduced with the vernal equinox of the year 471–2 (15 March 1079), but to my knowledge, it was never used in epigraphy. Another new calendar was introduced under the Ilkhanid Ghazan in 701/1302. Known as the Khani era, it was popularised in the 1330s by his nephew Abu SaÆid who used it on many coins. This popularity may explain why a graffito penned on the mosque at Ashtarjan in 733/1333 gives the date in a mixture of Arabic and Persian (Miles 1974), but the Khani calendar was not used in many other inscriptions that survive. Neither was the Turko-Mongol animal calendar which appears in texts of the same period (Melville 1994). Although not common in epigraphy, these calendars must have continued in daily use. A glazed tombstone made for Bibi Malik Khatun in 1481 (Musée de Sèvres, Paris; see Watson 1985, fig. 131) is dated ‘the beginning of the month of Tir Jalali of the year 180 khåniyya, corresponding to the year 886 of the hegira of the Prophet’. The crude moulding and unidentified patron show that the tombstone was not on a royal commission, and most formal inscriptions from the later period in Iran continue to use the regular hegira calendar. In India, a new calendar was instituted by the emperor Akbar in 992/1584, the twenty-ninth year of his reign. Known as the Ilahi calendar, it was described by Abuæl-Fadl ÆAllami in his chronicle of Akbar’s reign, the ÅæÈn-i akbarÈ (vol. 2, p. 30). The Ilahi era was made to begin not from the date of its institution but retrospectively from the first New Year (the vernal equinox) twenty days after the emperor’s accession in 1556. It was reportedly used until the accession of Shah Jahan in 1037/1628.

— 219 —

— ISLAMIC INSCRIPTIONS — on his accession, but his inscription was added twenty years later, perhaps when an official inventory of the treasury was made. Local calendars were also used in other outlying areas. In Java, for example, a purely lunar calendar was introduced in 1633, and dates in inscriptions composed after this time are often given in this typically Javanese calendar. Most of these systems are limited in their use. In addition to different calendars, there are also different systems of numeration. The first and probably the best known is the abjad system, the alphanumeric system in which each letter of the Arabic alphabet was used to represent a different numerical value. The letters were arranged in the sequence of older Semitic alphabets, and the name itself was formed from the first four letters. For the sake of pronunciation and memorisation, the letters are grouped into pronounceable but meaningless words: æabjad hawwaz ˙u††iy kalaman saÆfaß qarashat thakhadh ∂aΩagh. The article ‘Abjad’ in the Encyclopaedia Iranica gives a convenient table. In the Maghrib, the fifth, sixth and eighth groups were arranged differently, and the letters are grouped æabajid hawazin ˙u†iyin kalamanin ßaÆfa∂in qurisat thakhudh Ωaghshin. The abjad system of dating was commonly used by scientists, divinators and others for astronomical tables, astrological horoscopes and death, compositional and regnal chronograms. It also occurs on magic squares, talismans and other forms of letter magic. It was standard on astrolabes, and two contemporary examples show that the eastern and western systems were already in use in medieval times. A tenth-century astrolabe made in ÆIraq by Nastulus/Bastulus (see Figure 15.89), for example, uses the eastern system with sÈn for 60 and ßåd for 90, whereas one in the Germanische Nationalmuseum, Nuremberg, made in 472/ 1079, uses the Maghribi system with ßåd for 60 and ∂åd for 90. The ÆIraqi astrolabe is also dated in abjad on the back of the kursÈ: sana shiya (‘the year of the mark’), and the value of the three letters (shÈn or 300 + yåæ or 10 + håæ or 5) adds up to 315, corresponding to 927–8 CE.

— 220 —

ˆ

Despite the introduction of the Ilahi calendar, however, the hegira calendar remained the most common in monumental inscriptions in India, particularly for mosques and other religious buildings. In the Taj Mahal, for example, the inscriptions on the cenotaphs of Mumtaz Mahal and Shah Jahan give the dates of their death in the hegira years 1040 (1631) and 1076 (1666) respectively. Similarly, the signature of the calligrapher Amanat Khan in cartouches on the interior of the tomb uses the hegira date 1045 (1635–6). Sometimes, however, a dual system that mentioned both the hegira and the regnal years was used. Thus, Amanat Khan’s full signature at the end of the Koranic band around the interior of the tomb gives the hegira date 1048 and the twelfth regnal year of Shah Jahan’s reign (1638–9). This system of dual dates giving the emperor’s regnal year along with the hegira year is generally used on jades and other precious objects made for or appropriated by the Mughal emperors (many published in Markel 1990). For example, the jade inkpot made for Jahangir and now in the Metropolitan Museum has a foundation inscription and signature on the underside of the foot ring saying that it was made for the emperor in the fourteenth year of his reign corresponding to the hegira year 1028 (1618–19) by the artist Muæmin. Another example is the white jade tankard made for Ulugh Beg (see Figure 13.83). It was reinscribed for Jahangir in 1613–14 with a lengthy inscription running around the upper edge of the rim giving the emperor’s name and titles and the date of the eighth regnal year corresponding to the hegira year 1022 (sana 8 julËs mu†åbiq sana 1022 hijrÈ). Shah Jahan added his own, shorter inscription in taÆlÈq below the handle, 1056 ßå˙ib qirån thånÈ 20, with his sobriquet ‘Lord of the Second Conjunction’ flanked by two dates, 1056 of the hegira calendar (1646–7) and the twentieth year of his reign. Although the dates of these reinscriptions are clear, the reasons for making them are not. According to the emperor’s memoirs, Jahangir acquired the jade cup in 1608, some five years before his name was inscribed on it. The cup undoubtedly passed to Shah Jahan

ˆ

— SOURCES, METHODS This difference between eastern and western systems of abjad has important ramifications for localising other objects. It can be applied, for example, to early manuscripts of the Koran, in which the verses are traditionally numbered according to the abjad system. Bloom (1986, 1991) has suggested that the so-called Blue Koran should be attributed to North Africa on the basis of the abjad system used. His method might be profitably be extended to other types of objects.

AND

CONVENTIONS —

The use of abjad numbering can also be used to show what types of people were involved in making an object. Rogers and Ward (1988, no. 110) suggested, for example, that the absence of abjad numbering on a talismanic shirt made for the Ottomans (see Figure 14.85) shows that the artist was a calligrapher/illuminator rather than a specialist in talismanic calculations. The cursive style of script certainly supports this suggestion. The abjad system was also used for chrono-

15.89 Bronze astrolabe made by Bastulus or Nastulus in ÆIraq in 315/927–8. Diameter 17.5 cm. Kuwait, Dar al-Athar al-Islamiyya LNS 36M.

— 221 —

— ISLAMIC INSCRIPTIONS — grams at the end of poetic inscriptions, which were particularly popular on Ottoman buildings (see Chapter 2). Another system of numeration used in the early Islamic period is called Coptic (zimåm). Used in early Arab administration records, it is not common in epigraphy. One rare example is a very early piece of lustred glass from eighth-century Egypt and now in the Cairo Museum (Youssef 1972). These systems of numeration were eventually replaced by the so-called Arabic numerals. Lemay gives an extensive discussion of the subject in his article ‘Arabic numerals’ in the Dictionary of the Middle Ages (vol. 1, pp. 2382–98). More properly called the Hindu-Arabic system, this is a simple and ingenious system in which any quantity can be represented by nine figures using a decimal point. This system of number representation had arisen in India at least by the fifth century and was diffused by Muslim civilisation. Although known, it was not used in early Islamic times, when numbers in manuscripts and on objects were written out in words or transcribed in abjad or occasionally in Coptic numerals. The use of numerals in manuscripts was only adopted slowly, and astrologers and mathematicians remained firmly attached to older systems of number representation well into late medieval times. Even in the sixteenth and seventeenth centuries, scribal rendering of the Hindu numerals is clumsy and inconsistent. A survey of the practice on objects might be of interest to historians of science in charting the gradual adoption of this system in the Islamic lands. Arabic numerals were used, for example, in dates on Persian lustre ceramics, particularly tiles, during the thirteenth century. The forms of the Arabic numerals also changed over time. Irani (1955–6) has compiled various forms of numerals used in manuscripts dating from the eleventh to the eighteenth century. Zero was originally written as a small circle and then gradually reduced to a point. Both forms are used on medieval Persian lustrewares. The biggest change occurred in the numeral for five. In the traditional form, adopted from the Hindu numeral, five was written something like a figure-eight

with a flat right side. This is the form used on medieval lustrewares. Gradually, the cusp on the left became less significant and moved downward. In the modern form for five in most of the Arab world, the cusp is suppressed so that five is written as a circle. By contrast, in Iran, the cusp is at the bottom of the numeral so that five is written as an upside-down heart. The forms of the Arabic numerals are written in different ways in different places. In India, for example, seven was often written on its side. In Spain and the Maghrib, during the ninth or more probably the tenth century, the Arabic numerals developed a distinct form known as ‘dust numerals’ (al-arqåm al-ghubårÈ). The name is probably derived from the dust board on which calculations were done in medieval times. These ‘dust’ forms were adopted for the modern European numerals. This system is not widespread in epigraphy.

Recording an Inscription The first step in studying an inscription is to record it as written. Various systems are used to record an inscription from the Islamic lands. One of the most popular and convenient, the Leiden Bracket system, uses these conventions when adapted for Arabic epigraphy (Burgoyne and Abul-Hajj 1979): [ ] Lacuna. Square brackets enclose text that is lost, where the surface is broken away or so worn and eroded that the inscribed text does not survive. The square brackets can be filled in in several ways: [ ] Conjectural restoration of missing text proposed by editor. […] Text missing, no conjectural restoration proposed by editor. Each dot represents approximately one missing letter. This can 3 ca. 3 also be written as […], […], and so on. [---] The same, but the length of missing text is unknown to the editor.

— 222 —

— SOURCES, METHODS

AND

< > Conjectural addition. Angle brackets enclose corrections and additions proposed by the editor. These are usually letters which properly belong to the text but have been erroneously omitted when the text was written. { } Conjectural deletion. Braces enclose letters which do not properly belong to the text but were erroneously added when the text was written. They should be ignored in reading. ° ° ° Conjectural interpolation of doubtful letters. The superscript circle ° above a letter indicates that only incomplete traces of the letter survive, and that these traces,

CONVENTIONS — if taken out of context, could be interpreted in more than one way. Thus the ° does not indicate damage as such but rather palaeographic ambiguity resulting from damage and indicates that the reading of the individual letter is an editorial conjecture.

…

Undeciphered letter traces. Traces of letters who palaeographic interpretation is not apparent to the editor. Each dot represents traces of approximately one letter.

Dijkema (1977) gives further marks showing how this system can be adapted to accommodate Ottoman poetic texts.

— 223 —

Bibliography ÆAbd al-RaæËf, ÆAlÈ YËsuf, ‘KhazzåfËn min al-Æaßr al-få†imÈ wa asålÈbuhum al-fannÈya’, Bulletin of the Faculty of Arts, Cairo University 20 (1958), 173–279. Abel, A., Gaibi et les grands faïenciers égyptiens d’époque Mamelouke (Cairo, 1930). AbËæl-Fa∂l ÆAllåmÈ, ÅæÈn-i akbarÈ, trans. H. S. Jarett, 3 vols (reprinted Delhi, 1989). Ackerman, Phyllis, ‘A biography of Ghiyath the Weaver’, Bulletin of the American Institute for Persian Art and Archaeology 7 (December 1934), 9–13. Adle, Chahryar, ‘Le minaret du Masjed-i JåmeÆ de Semnån circa 412–25/1030–34’, Studia Iranica 4 (1975), 177–86. — ‘KatÈba-yi kåshÈ-yi manår-i masjid-i jåmiÆ-i Dåmghån’, Athar 7–9 (1361/1982), 297–300. — ‘Recherches archéologiques en Iran sur le Kumes médieval. Rapport préliminaire pour 1982–1983’, Académie des Inscriptions et Belles-Lettres, Comptes Rendues des Séances de l’Année 1984 (April–June 1984), 292–7. Adle, Chahryar and Assadullah Suren Melikian-Chirvani, . ‘Les monuments du XIe siècle du Dåmqån’, Studia Iranica 1 (1972), 229–97. Afshar, Iraj, Yådgårhå-yi Yazd, 3 vols (Tehran, 1348–54/ 1969–75). — ‘Two 12th century gravestones of Yazd in Ma∑had and Washington’, Studia Iranica 2 (1973), 203–11. — ‘ZÈlË’, in Annette Ittig (ed.), The Carpets and Textiles of Iran: New Perspectives in Research = Iranian Studies 25:1–2 (1993), 31–7. Ahmed, Shamsud-din, Inscriptions of Bengal, vol 4 (Being a Corpus of Inscriptions of the Muslim Rulers of Bengal from 1233 to 1855 A.C.) (Rajshahi, 1960). Alexander, David, The Arts of War: Arms and Armour of the 7th to 19th Centuries, The Nasser D. Khalili Collection of Islamic Art (London, 1992). Allan, J. W., ‘Abu’l-Qasim’s treatise on ceramics’, Iran 11 (1973), 111–20. — Islamic Metalwork: The Nuhad Es-Said Collection (London, 1982a). — Nishapur: Metalwork of the Early Islamic Period (New York, 1982b). — Metalwork of the Islamic World: The Aron Collection (London, 1986).

— ‘The Nishapur metalwork: cultural interaction in early Islamic Iran’, in Priscilla P. Soucek (ed.), Content and Context of Visual Arts in the Islamic World, College Art Association of America (University Park and London, 1988), pp. 1–11. Allen, Terry, ‘Review of books on Islamic metalwork, architecture, and archaeology’, Ars Orientalis 15 (1985), 149–54. — ‘Notes on Bust-II’, Iran 27 (1989), 57–66. Amari, M., Le epigrafi arabiche di Sicilia, iscrizione edili. (Palermo, 1875; reprinted Palermo, 1971). The Anatolian Civilisations III: Seljuk/Ottoman (Istanbul, 1983). Anglade, Elise, Catalogue des boiseries de la section islamique, Musée du Louvre (Paris, 1988). Arts de l’Islam des origines à 1700 dans les collections publiques françaises (Paris, 1971). The Arts of Islam, exhibition catalogue, Hayward Gallery (London, 1976). Asher, Catherine B., ‘The tomb of IbråhÈm SËr: epigraphs and implications’, in Frederick M. Asher and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985), pp. 275–82. Asher, Frederick M. and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985). Atasoy, Nurhan and Julian Raby, Iznik: The Pottery of Ottoman Turkey (London, 1989). Atıl, Esin, Ceramics from the World of Islam [in the Freer Gallery of Art] (Washington DC, 1973). — Renaissance of Islam: Art of the Mamluks (Washington DC, 1981). — The Age of Sultan Süleyman the Magnificent, (Washington DC and New York, 1987). Atıl, Esin, W. T. Chase and Paul Jett, Islamic Metalwork in the Freer Gallery of Art (Washington DC, 1985). Aynur, Hatice and Hakan T. Karateke, III. Ahmed Devri Òstanbul Çe∞meleri (1703–1730) (Istanbul, 1995). Baer, Eva, ‘The Nisan Tasi: a study in Persian-Mongol metalware’, Kunst des Orients 9 (1973–4), 1–46. — Metalwork in Medieval Islamic Art (Albany, 1983). — ‘Jeweled ceramics from medieval Islam: a note on the ambiguity of Islamic ornament’, Muqarnas 6 (1989), 83–97. Bahrami, Mehdi, Recherches sur les carreaux de revêtement lustré dans la céramique persane du XIIIe au

— 224 —

— BIBLIOGRAPHY — XVe siècles (étoiles et croix) (Paris, 1937). — Gurgan Faiences (Cairo, 1949; reprinted Costa Mesa, 1988). Baldissera, E., ‘Le inscrizioni nel palazzo di ∆abrÈn (Sultanato di Omàn)’, Quaderni di studi arabi 9 (1991), 29–91. — ‘Scritta lignea del palazzo di al-ManßËr (Sultanato Di Omàn)’, Quaderni di studi arabi 11 (1994a [1993]), 27–36. — Al-Kitåbåt fÈæl-masåjid al-umåniya al-qadÈma (Muscat, 1994b). — Qaßr Jabrayn wa-kitåbåtuh (Muscat, 1994c). Balog, Paul, ‘Pious invocations probably used as titles of office or as honorific titles in Umayyad and ÆAbbåsid times’, in Miriam Rosen-Ayalon (ed.), Studies in Memory of Gaston Wiet (Jerusalem, 1977), pp. 61–8. al-Båshå, Óasan Ma˙mËd, ‘Êabaq min al-khazaf bi-ism Ghaban Mawlå al-Óåkim Biæamrillåh’, Bulletin of the Faculty of Arts, Cairo University 18 (1956), 71–85. Begley, Wayne E., ‘Amånat Khån and the calligraphy on the Tåj Ma˙al’, Kunst des Orients 12 (1978–9), 5–60. — ‘The myth of the Taj Mahal and a new theory of its symbolic meaning’, Art Bulletin 56 (1979), 7–37. — ‘A Mughal caravanserai built and inscribed by Amanat Khan, calligrapher of the Taj Mahal’, in Frederick M. Asher and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985), pp. 283–90. Begley, W. E. and Z. A. Desai, Taj Mahal: The Illumined Tomb (Cambridge MA, 1989). Behrens-Abouseif, Doris, The Minarets of Cairo (Cairo, 1985). Bel, Alfred, ‘Inscriptions arabes de Fès’, Journal Asiatique, 11th series, 9 (1917), 303–29; 10 (1917), 81–170, 215–67; 12 (1918), 189–276, 337–99; 13 (1919), 5–96; 15 (1919), 467–79. Bendrey, V. S., Studies in Muslim Inscriptions (Bombay, 1944). Berchem, Max van, ‘Eine arabische Inschrift aus dem Ostjordanlande mit historischen Erläuterungen’, Zeitschrift des Deutschen Palaestina-Vereins 16 (1893), 84–105. — Matériaux pour un Corpus Inscriptionum Arabicarum I: Egypte 1, Mémoires de la Mission archéologique française au Caire XIX (Paris, 1894–1903). — ‘L’épigraphie musulmane en Algérie: étude sur le corpus’, Revue Africaine 49 (1905), 160–91. — ‘Titres califiens d’occident à propos de quelques monnaies Mérinides et Ziyanides’, Journal Asiatique, 10th series, 9 (1907), 245–335.

— ‘Une inscription du sultan mongol Uldjaitu’, in Mélanges Hartwig Derenbour (Paris, 1909a), pp. 367–78. — ‘Inscriptions mobilières arabes en Russie’, Journal Asiatique, 10th series, 14 (1909b), 401–13. — Amida (Heidelberg, 1910). — ‘La chaire de la mosquée d’Hébron et le martyrion de la tête de Husain à Ascalon’, in Festschrift Eduard Sachau (Berlin, 1915), pp. 298–310. — Matériaux pour un Corpus Inscriptionum Arabicarum II: Syrie du Sud, Mémoires de l’Institut français archéologique du Caire XLIII–XLV (Cairo, 1920–2). Berchem, Max van and Halil Edhem, Matériaux pour un Corpus Inscriptionum Arabicarum III: Asie Mineure 1, Mémoires de l’Institut français archéologique du Caire XXIX (Cairo, 1910–17). Bergeret, Jean and Ludvik Kalus, ‘Analyse de décors épigraphiques et floraux à Qazwin au début du VIe/XIIe siècle’, Revue des Études Islamiques 45:1 (1977), 89–130. Bernus, M., H. Marchal and G. Vial, ‘Le suaire de SaintJosse’, Bulletin de Liaison du Centre International d’Etudes des Textiles Anciens 33 (1971), 1–57. Bier, Carol (ed.), Woven from the Soul, Spun from the Heart: Textile Arts of Safavid and Qajar Iran 16th– 19th Centuries (Washington DC, 1987). Bierman, Irene A., ‘The art of the public text: medieval Islamic rule’, in I. Lavin (ed.), World Art: Themes of Unity in Diversity: Acts of the XXVIth International Congress of the History of Art (University Park and London, 1989), vol. 2, pp. 283–90. — ‘Inscribing the city: Fatimid Cairo’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997), 105–14. Bisheh, Ghazi, ‘Óammåm al-Íarå˙ in the light of recent excavations’, Damaszener Mitteilungen 4 (1989), 225–30. Bivar, A. D. H., ‘SeljËqid Ziyårats of Sar-i Pul (Afghanistan)’, Bulletin of the School of Oriental and African Studies 19 (1966), 57–63. — ‘The inscription of Sålår KhalÈl in Afghanistan’, Journal of the Royal Asiatic Society (1977), 145–9. — ‘The SaljËq sign-manual represented on a sgraffito potsherd’, Journal of the Royal Asiatic Society (1979), 9–15. Bivar, A. D. H. and Ehsan Yarshater (eds), Corpus Inscriptionum Iranicarum, Part IV: Persian Inscriptions down to the Early Safavid Period, vol. 6: Måzandarån Province, Portfolio I: Eastern Måzandarån I (London, 1978).

— 225 —

— ISLAMIC INSCRIPTIONS — Blair, Sheila S., ‘The inscription from the tomb tower at Baståm: an analysis of Ilkhanid epigraphy’, in C. Adle (ed.), Art et Société dans le Monde Iranien, (Paris, 1982), pp. 263–86. — ‘Review of Erica Cruikshank Dodd and Shereen Khairallah, The Image of the Word’, Arabica 31 (November 1984), 337–42. — ‘Artists and patronage in late fourteenth-century Iran in light of two catalogues of Islamic metalwork’, Bulletin of the School of Oriental and African Studies 48:1 (1985a), 53–9. — ‘The madrasa at Zuzan: Islamic architecture in eastern Iran on the eve of the Mongol invasions’, Muqarnas 3 (1985b), 75–91. — ‘A medieval Persian builder’, Journal of the Society of Architectural Historians 45 (1986), 389–95. — ‘Legibility versus decoration in Islamic epigraphy: the case of interlacing’, in I. Lavin (ed.), World Art: Themes of Unity in Diversity: Acts of the XXVIth International Congress of the History of Art (University Park and London, 1989a), vol. 2, pp. 329–34. — ‘Review of Mehrdad Shokoohy and Natalie H. Shokoohy, Óißår-i FÈrËza: Sultanate and Early Mughal Architecture in the District of Hisar, India and Mehrdad Shokoohy, Bhadreªvar: The Oldest Islamic Monuments in India’, Journal of the Society of Architectural Historians 48:4 (1989b), 390–91. — The Monumental Inscriptions from Early Islamic Iran and Transoxiana, Supplements to Muqarnas (Leiden, 1992a). — ‘What is the date of the Dome of the Rock?’, in Julian Raby and Jeremy Johns (eds), Bayt al-Maqdis: ÆAbd al-Malik’s Jerusalem, Part One, Oxford Studies in Islamic Art (Oxford, 1992b), pp. 59–88. — ‘An inscription from BarËjird: new data on domed pavilions in SaljËq mosques’, in Robert Hillenbrand (ed.), The Art of the Seljuqs in Iran and Anatolia, Proceedings of a symposium held in Edinburgh in 1982 (Costa Mesa CA, 1994), pp. 4–11. — ‘Inscriptions on medieval Islamic textiles’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997a), 95–104. — ‘A note on the prayers inscribed on several medieval silk textiles in the Abegg Foundation’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997b), 129–38. Blair, Sheila S. and Jonathan M. Bloom, The Art and Architecture of Islam, 1250–1800, The Pelican History of Art (London and New Haven, 1994).

Blair, Sheila S., Jonathan M. Bloom and Anne E. Wardwell, ‘Re-evaluating the date of the “Buyid” silks by epigraphic and radiocarbon analysis’, Ars Orientalis 22 (1992), 1–43. Bloom, Jonathan M., ‘The mosque of al-Hakim in Cairo’, Muqarnas 1 (1983), 15–36. — ‘Al-Maæmun’s blue Koran?’, Revue des Études Islamiques 54 (1986), 61–5. — ‘A Mamluk basin in the L. A. Mayer Memorial Institute’, Islamic Art 2 (1987a), 15–26. — ‘The mosque of the Qarafa in Cairo’, Muqarnas 4 (1987), 7–20. — Minaret: Symbol of Islam, Oxford Studies in Islamic Art (Oxford, 1989). — ‘The early Fatimid blue Koran manuscript’, GraecoArabica 4 (1991), 171–8. — The Fatimids (909–1171), their ideology and their art’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997), 15–26. Bol’shakov, O. G., ‘Arabskie nadpisi na polivnoy keramike sredney azii IX–XII vv’, Epigrafika Vostoka 12 (1958), 23–38; 15 (1963), 73–87; 16 (1963), 35–55; 17 (1966), 54–62; 19 (1969), 42–50. Bombaci, Alessio, The KËfic Inscription in Persian Verses in the Court of the Royal Palace of MasÆËd III at Ghazni (Rome, 1966). Bosworth, C. E., ‘Notes on the pre-Ghaznavid history of eastern Afghanistan’, Islamic Quarterly 9 (1965), 22–3. Bourrilly, J. and E. Laoust, Stèles funeraires marocaines (Paris, 1927). Brand, Michael and Glenn D. Lowry (eds), FatehpurSikri: A Sourcebook (Cambridge MA, 1985). Britton, N. P., A Study of Some Early Islamic Textiles in the Museum of Fine Arts, Boston (Boston MA, 1938) Brosselard, Charles, ‘Les inscriptions arabes de Tlemcen’, Revue Africaine 3–6 (1856–62): 3: 81–94, 161–72, 241–8, 321–40, 401–19; 4: 1–17, 81–93, 161–74, 241–58, 321–31; 5: 14–30, 161–73, 241–60, 321–36, 401–21; 6: 11–21, 161–72. Bulliet, Richard, ‘A MuÆtazilite coin of Ma˙mËd of Ghazna’, American Numismatic Society Museum Notes 15 (1969), 119–29. — ‘Pottery styles and social status in medieval Khurasan’, in A. Bernard Knapp (ed.), Archaeology, Annales, and Ethnohistory (Cambridge, 1992), pp. 75–82. Burgoyne, Michael H., ‘A recently discovered Marwånid inscription in Jerusalem’, Levant 14 (1982), 118–21.

— 226 —

— BIBLIOGRAPHY — Burgoyne, Michael Hamilton and Amal Abul-Hajj, ‘Twenty-four medieval Arabic inscriptions from Jerusalem’, Levant 11 (1979), 112–37. Carboni, Stefano and Tomoko Masuya, Persian Tiles (New York, 1993). Casimir, Michael J. and Bernt Glatzer, ‘„åh-i Ma∑had, a recently discovered madrasah of the ≈hurid period in GarÌistån (Afghanistan)’, East and West new series, 21 (March–June 1971), 53–68. Cohn-Wiener, Ernst, ‘A Turanic monument of the twelfth century A.D’, Ars Islamica 6 (1939), 145–9. Colin, G., Corpus des inscriptions arabes et turques de l’Algerie: Département d’Alger (Paris, 1901). Combe, Étienne, Jean Sauvaget and Gaston Wiet (eds), Répertoire chronologique d’épigraphie arabe, 19 vols (Cairo, 1931–). Content, Derek J., Islamic Rings and Gems: The Benjamin Zucker Collection (London, 1982). Creswell, K. A. C., Early Muslim Architecture, 2 vols (Oxford: Clarendon Press, 1932–40; 2nd edn of vol. 1 in two vols, 1969; reprinted New York, 1979). — The Muslim Architecture of Egypt, 2 vols (Oxford, 1952–9; reprinted New York, 1978). — A Bibliography of the Architecture, Arts and Crafts of Islam to 1st Jan. 1960 (Cairo and Vaduz, 1961; reprinted 1978). — A Bibliography of the Architecture, Arts and Crafts of Islam; Supplement Jan. 1960 to Jan. 1972 (Cairo, 1973). [For second supplement see Pearson.] Curatola, Giovanni, ‘Some Ilkhanid woodwork from the area of Sultaniyya’, Islamic Art 2 (1987), 97–116. — Eredità dell’Islam ([Venice], 1993). Curatola, Giovanni and Marco Spallanzani, Mattonelle Islamiche/Islamic Tiles, Lo specchio del Bargello (Florence, 1981). Cutler, Anthony, ‘The mythological bowl in the treasury of San Marco at Venice’, in Dikran Kouymjian (ed.), Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles (Beirut, 1974), pp. 235–54. Dani, Ahmad Hasan, Bibliography of the Muslim Inscriptions of Bengal (down to A.D. 1538), Appendix to Journal of the Asiatic Society of Bengal, 2 (1957) (Dacca, 1957). David-Weill, J., Les Bois à épigraphes jusqu’à l’époque mamlouke (Cairo, 1931). Day, Florence, ‘Dated †iråz in the collection of the University of Michigan’, Ars Islamica 4 (1937), 421–48.

— ‘A review of “The Ceramic Art in Islamic Times. B. Dated Faience”’, Ars Islamica 8 (1941), 49–58. — ‘Review of Soieries Persanes by Gaston Wiet’, Ars Islamica 15–16 (1951), 231–44. — ‘The †iråz silk of Marwån’, in G. C. Miles (ed.), Archaeologica Orientalia in Memoriam Ernst Herzfeld (Locust Valley NY, 1952), pp. 39–61. — ‘The inscription of the Boston “Baghdad” silk: a note on method in epigraphy’, Ars Orientalis 1 (1954), 191–4. Décorbert, Christian and Roland-Pierre Gayraud, ‘Une céramique d’époque mamelouke trouvée à Êπd’, Annales Islamologiques 18 (1982), 95–104. Denny, Walter B., ‘Ottoman Turkish textiles’, Textile Museum Journal 3 (December 1972), 55–66. — ‘A group of silk Islamic banners’, Textile Museum Journal 4:1 (1974), 67–81. Desai, Z. A., ‘KËfÈ epitaphs from Bhadreswar in Gujarat’, Epigraphia Indica, Arabic and Persian Supplement (1965), 1–8. — Published Muslim Inscriptions of Rajasthan (Jaipur, 1971). — ‘Islamic inscriptions: their bearing on monuments’, in Frederick M. Asher and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985), pp. 251–6. Description de l’Egypte, publiée par les ordres de Napoléon Bonaparte (reprinted Cologne, 1994). Deverdun, G., Inscriptions arabes de Marrakech (Rabat, 1956). Diakonov, M. M., ‘Ob odnoy rasmey arabskoy nadpisi’, Epigrafika Vostoka I (1947), 5–8. Diez, Ernst, Churasanische Baudenkmäler (Berlin, 1918). Digby, Simon, ‘More historic Kashmir metalwork?’, Iran 12 (1974), 181–5. Dijkema, F. Th., The Ottoman Historical Monumental Inscriptions in Edirne (Leiden, 1977). Dodd, Erica Cruikshank, ‘The image of the word (notes on the religious iconography of Islam)’, Berytus 18 (1969), 35–62. Dodd, Erica Cruikshank and Shereen Khairallah, The Image of the Word: A Study of Quranic Verses in Islamic Architecture, 2 vols (Beirut, 1981). Dodds, Jerrilynn D. (ed.), Al-Andalus: The Art of Islamic Spain (New York, 1992). [Edgüer] Rıdvån Nåfiz and Ismåæil Óa˚˚È [Uzunçar∞ılı], Sivas ∞ehri (Istanbul, 1346/1928). Edwards, Holly, ‘The Ribat of ÆAli B. Karmakh’, Iran 29 (1991), 85–94.

— 227 —

— ISLAMIC INSCRIPTIONS — Elgood, Robert (ed.), Islamic Arms and Armour (London, 1979). Elisséeff, Nikita, ‘La titulature de NËr ad-DÈn d’après ses inscriptions’, Bulletin des Études Orientales 14 (1952–4), 155–96. Ennès, P. and L. Kalus, ‘Un plat d’ıznik à inscription turque’, Revue du Louvre 29 (1979), 258–60. Ettinghausen, Richard, ‘Important pieces of Persian pottery in London collections’, Ars Islamica 2 (1935), 45–64. — ‘Dated faience’, in A. U. Pope and P. Ackerman (eds), A Survey of Persian Art (London and New York, 1939), pp. 1,667–96. — ‘The Bobrinski “kettle”: patron and style of an Islamic bronze’, Gazette des Beaux-Arts, 6th series, 24 (1943), 193–208. — ‘The “beveled” style in the post-Samarra period’, in G. C. Miles (ed.), Archaeologica Orientalia in Memoriam Ernst Herzfeld (Locust Valley NY, 1952), pp. 72–82. — ‘Notes on the lusterware of Spain’, Ars Orientalis 1 (1954), 145–8. — ‘The “Wade Cup” in the Cleveland Museum of Art: its origins and decorations’, Ars Orientalis 2 (1957), 327–66. — ‘An early Ottoman textile’, in First International Congress of Turkish Arts (Ankara, 1961), pp. 134–40. — Arab Painting (Geneva, 1962). — ‘The flowering of Saljuk art’, Metropolitan Museum Journal 3 (1970), 111–31. — ‘Arabic epigraphy: communication or symbolic affirmation’, in Dikran Kouymjian (ed.), Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles (Beirut, 1974), pp. 297–317. — ‘Kufesque in Byzantine Greece, the Latin West and the Muslim world’, in A Colloquium in Memory of George Carpenter Miles (1904–1975) (New York, 1976), pp. 28–47. Ettinghausen, Richard and Oleg Grabar, The Art and Architecture of Islam: 650–1250, The Pelican History of Art (Harmondsworth, 1987). Faris, Nabih A. and George C. Miles, ‘An inscription of Bårbak Shah of Bengal’, Ars Islamica 7 (1940), 141–7. Fehérvári, Géza, Islamic Metalwork of the Eighth Century to the Fifteenth Century in the Keir Collection. (London, 1976). Fernández-Puertas, Antonio, ‘Calligraphy in al-Andalus’, in Salma Khadra Jayyusi (ed.), The Legacy of Muslim Spain (Leiden, 1992), pp. 639–76.

Ferrandis, José, Marfiles árabes de Occidente, 2 vols (Madrid, 1935–40). Ferrier, R. W., The Arts of Persia (New Haven CT and London, 1989). Finster, Barbara, ‘Die Mosaiken der Umayyadenmoschee’, Kunst des Orients 7 (1970), 83–141. Flury, S., Die Ornamente der Hakim- und Ashar-Moschee (Heidelberg, 1912). — ‘Das Schriftband an der Türe des Ma˙mËd von Ghazna (998–1030)’, Der Islam 8 (1918), 214–27. — Islamische Schriftbänder: Amida-Diarbekr XI. Jahrhundert (Paris, 1920a) — ‘Bandeaux ornementés à inscriptions arabes, AmidaDiarbekr IXe [sic] siècle’, Syria 1 (1920b), 235–49, 318–28; 2 (1921), 54–62. — ‘The Kufic inscriptions of Kisimkazi mosque, Zanzibar, 500 H. (A.D. 1107)’, Journal of the Royal Asiatic Society (April 1922), 257–64. — ‘Le décor épigraphique des monuments de Ghazna’, Syria 6 (1925a), 61–90. — ‘Ein Stuckmi˙råb des IV. (X.) Jahrhunderts’, Jahrbuch der Asiatischen Kunst 1 (1925b), 106–9. — ‘La mosquée de Nåyin’, Syria 11 (1930), 43–58. — ‘Le décor épigraphique des monuments fatimides du Caire’, Syria 17 (1936), 365–76. — ‘Ornamental KËfic inscriptions on pottery’, in A. U. Pope and P. Ackerman (eds), A Survey of Persian Art (London and New York, 1939), pp. 1,743–69. Folsach, Kjeld von and Anne-Marie Keblow Bernsted, Woven Treasures – Textiles from the World of Islam, exhibition catalogue (Copenhagen, 1993). Freeman-Grenville, G. S. P., The Muslim and Christian Calendars (London, 1963; reprinted 1977). Freeman-Grenville, G. S. P. and B. G. Martin, ‘A preliminary handlist of the Arabic inscriptions of the eastern African coast’, Journal of the Royal Asiatic Society (1973), 98–122. Frishman, Martin and Khan Hasan-Uddin, The Mosque: History, Architectural Development and Regional Diversity (London, 1994). Gabriel, Albert, Monuments turcs d’Anatolie, 2 vols (Paris, 1931–4). — Voyages archéologiques dans la Turquie orientale (Paris, 1940). — Une capitale turque, Brousse (Bursa) (Paris, 1958). García Gómez, E., Poemas árabes en los Muros y Fuentes de la Alhambra (Madrid, 1985). Gaube, Heinz, ‘An examination of the ruins of Qasr BurquÆ’, Annual of the Department of Antiquities of Jordan 19 (1974), 93–100.

— 228 —

— BIBLIOGRAPHY — — Arabische Inschriften aus Syrie, Beiruter Texte und Studien 17 (Beirut, 1978). — ‘Epigraphik’, in Wolfdietrich Fischer (ed.), Grundriß der Arabischen Philologie, I: Sprachwissenschaft (Wiesbaden, 1982), pp. 210–25. Ghirshman, R., ‘Argenterie d’un seigneur sassanide’, Ars Orientalis 2 (1957), 77–82. Ghouchani, Abdallah [Quchani], Kha††-i KËfÈ-i MaqilÈ dar Masåjid-i Baståni-i Ißfahån (Tehran, 1985). — KatÈbahå-yi sufal-i NayshåbËr (Tehran, 1986). — Persian Poetry on the Tiles of Takht-i Sulaymån (13th Century) (Tehran, 1992). Ghulam, Yousif, Introduction to the Art of Arabic Calligraphy in Iran (Tehran, n.d.). Giuzalian [Gyuzal’yan], L. T., ‘The bronze qalamdan (pencase) 542/1148 from the Hermitage Collection (1936–1965)’, Ars Orientalis 7 (1968), 95–119. Glidden, Harold W. and Deborah Thompson, ‘Êiråz fabrics in the Byzantine Collection, Dumbarton Oaks, Part One: Êiråz from Egypt’, Bulletin of the Asia Institute 2 (1988), 119–39. — ‘Êiråz fabrics in the Byzantine Collection, Dumbarton Oaks, Parts Two and Three: Êiråz from the Yemen, Iraq, Iran, and an unknown place’, Bulletin of the Asia Institute 3 (1989), 89–105. Godard, André, ‘Na†anz (province de Kåshån)’, Åthår-é Ïrån 1 (1936), 75–106. Godard, Yedda A., ‘Bassin de cuivre au nom de Shaikh Uwais’, Åthår-é Ïrån 1 (1936), 371–3. Golombek, L. and V. Gervers, ‘Êiråz fabrics in the Royal Ontario Museum‘, in V. Gervers (ed.), Studies in Textile History in Memory of Harold B. Burnham (Toronto, 1977), pp. 82–125. Golombek, Lisa and Donald Wilber, The Timurid Architecture of Iran and Turan, 2 vols (Princeton NJ, 1988). Grabar, André, ‘La décoration architecturale de l’église de la Vierge à Saint-Luc en Phocide et le début des influences islamiques sur l’art byzantin de Grèce’, Comptes Rendus de l’Académie des Inscriptions et Belles-Lettres (1971), 15–37. Grabar, Oleg, ‘Review of Epigrafika Vostoka I–VIII’, Ars Orientalis 1 (1957), 547–60. — ‘The Umayyad Dome of the Rock in Jerusalem’, Ars Orientalis 3 (1959), 33–62. Grohmann, Adolf, ‘The origin and early development of floriated KËfic’, Ars Orientalis 2 (1957), 183–214. — Arabische Paläographie, 2 vols (Vienna, 1971). Grube, Ernst J., ‘Notes on the decorative arts of the Timurid period’, in Gururåjamañjarikå: Studi in

onore di Giuseppe Tucci (Naples, 1974), vol. 1, pp. 233–79. — ‘Notes on the decorative arts of the Timurid period, II’, Islamic Art 3 (1989), 175–208. — Cobalt and Lustre, The Nasser D. Khalili Collection of Islamic Art (London, 1996). Grube, Ernst J. and Cristina Tonghini, ‘Towards a history of Syrian Islamic ceramics before 1500’, Islamic Art 3 (1988–9), 59–94. Guest, Grace D. and Richard Ettinghausen, ‘The iconography of a Kashan luster plate’, Ars Orientalis 4 (1961), 25–64. Gyuzal’yan, L. T., ‘Bronzvoi Kuvshin 1182 G.’, in Pamiatniki Epochi Rusavelli (Leningrad, 1938), pp. 227–36. — ‘Otryvok iz shakhname na glinyanykh izdeliyakh XIII–XIV vv’, Epigrafika Vostoka 4 (1951), 40–55. — ‘Nadpis na lyustrovom izraztse 624/1227 G. iz Kievskogo Muzeya’, Epigrafika Vostoka 11 (1956), 33–43. — ‘Nadpisi na dyukh sferokonicheskikh sosudakh iz Baylakana’, Palestinskiy Sbornik 13 (1965), 166–70. — ‘Nadpis na lustrovom kuvshine 1179 G. iz Britanskogo Muzeya’, Epigrafika Vostoka 17 (1966), 63–4. Hammer-Purgstall, Freiherr, Abhandlung über die Siegel der Araber, Perser und Türken (Vienna, 1849). Hanaway, William L. Jr (ed.), Corpus Inscriptionum Iranicarum, Part IV: Persian Inscriptions down to the Early Safavid Period, vol. 2: Khorasan Province, Portfolio I: Khorasan I (London, 1977). Hawary, H. and H. Rached, Les stèles funéraires, vols 1 and 3 (Cairo, 1932–8). el-Hawary, Hassan M. and Gaston Wiet, Matériaux pour un Corpus Inscriptionum Arabicarum IV: Arabie, Mémoires de l’Institut français archéologique du Caire CIX/1 (Cairo, 1985). Herzfeld, Ernst, ‘Postsasanidische Inschriften’, Archaeologische Mitteilungen aus Iran 4:3 (1933), 140–56. — ‘A bronze pen-case’, Ars Islamica 3 (1936), 35–43. — ‘Damascus: studies in architecture’, Ars Islamica 9 (1942), 1–53. — Matériaux pour un Corpus Inscriptionum Arabicarum II: Inscriptions et Monuments d’Alep, Mémoires de l’Institut français d’archéologie orientale du Caire LXXVI–LXXVIII (Cairo, 1954–5). Hilål al-ÍåbiÆ, RusËm dår al-khilåfah (The Rules and Regulations of the ÆAbbåsid Court), trans. Elie A. Salem (Beirut, 1977). Hill, Derek and Lucien Golvin, Islamic Architecture in North Africa, intro. Robert Hillenbrand (London, 1976).

— 229 —

— ISLAMIC INSCRIPTIONS — Hillenbrand, Carole, ‘Jihåd Propaganda in Syria from the time of the first Crusade until the death of Zengi: the evidence of monumental inscriptions’, in K. Athamina and R. Heacock (eds), The Frankish Wars and Their Influence in Palestine (Jerusalem, 1994), pp. 60–9. Hillenbrand, Robert, ‘SaljËq monuments in Iran: the mosques of NËshåbåd’, Oriental Art 22 (Autumn 1976), 265–77. — ‘Qur’anic epigraphy in medieval Islamic architecture’, Revue des Etudes Islamiques [Mélanges Dominique Sourdel] 54 (1986), 171–87. Horovitz, J., ‘A list of the published Mohamedan inscriptions of India’, Epigraphia Indo-Moslemica (1909–10), 30–144. Hunarfar, Lu†fallah, GanjÈna-yi åthår-i TårÈkhÈ-yi Ißfahån (Tehran, 1350/1977). Hussain, M. A., A Record of All the Quræanic and NonHistorical Epigraphs on the Protected Monuments in the Delhi Province, Memoirs of the Archaeological Survey of India 47 (Calcutta, 1936). The Indian Heritage: Court Life and Arts Under Mughal Rule, exhibition catalogue (London, 1982). Irani, Rida A. K., ‘Arabic numeral forms’, Centaurus 4:1 (1955–6), 1–12. Islamic Art in Egypt 969–1517, catalogue of an exhibition held in April 1969 at Semiramis Hotel, Cairo (Cairo, 1969). Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997). Ittig, Annette (ed.), The Carpets and Textiles of Iran: New Perspectives in Research = Iranian Studies 25:1–2 (1993a). — ‘Notes on a ZÈlË fragment dated 963/1556 in the Islamic Museum, Cairo’, in idem (ed.), The Carpets and Textiles of Iran: New Perspectives in Research = Iranian Studies 25:1–2 (1993b), 37–42. — ‘Historian’s choice: the Victoria & Albert Museum’s “Ardabil” Carpet’, Hali 69 (1993c), 81–3. Ivanov, A. A., ‘O pervonachal’nom naznachenii tak nazuivaemykh Iranskikh “podsvechnikov” XVI–XVII vv. [On the original purpose of the so-called Iranian “candlesticks” of the 16th–17th centuries’, in Issledovaniya Po Istorii Kul’tury Narodov Vostoka. Sbornik V. Chest’ Akademika I. A. Orbeli (1960), pp. 337–45. — ‘Osnovanie podsvechnika 880 g. Kh. (1475–1476 Gg) so stikhami poeta Salikhi’, Soobshchenie Gosudarvennogo Ermitazha 27 (1966), 67–70. — ‘Iranskiy mednyy podnas XIV v’, Epigrafika Vostoka 18 (1967), 94–104.

— ‘Pechat’ Gaykhar-shad’, Strani I Narodi Vostoka 10 (1971), 199–201. — ‘Fayansovoe blyudo XV veka iz Mashkhada’, Soobscheniya Gosudarstvennogo Ordena Lenina Ermitazha 45 (1980), 64–6, 79–80. Jaussen, J. A., ‘Inscriptions arabes de la ville d’Hébron’, Bulletin de l’Institut d’Archéologie Orientale 25 (1925), 1–45. Jayyusi, Salma Khadra (ed.), The Legacy of Muslim Spain, Handbuch der Orientalistik (Leiden, 1992). Jenkins, Marilyn, ‘Muslim: an early Fatimid ceramicist’, Bulletin of the Metropolitan Museum of Art 26 (1968), 359–69. — (ed.), Islamic Art in the Kuwait National Museum: The al-Sabah Collection (London, 1983). — ‘Mamluk underglaze painted pottery: foundations for further study’, Muqarnas 2 (1984), 95–114. — ‘Islamic glass: a brief history’, The Metropolitan Museum of Art Bulletin 44:2 (1986), entire issue. — ‘SaÆd: content and context’, technical note by Gary W. Carriveau, in Priscilla P. Soucek (ed.), Content and Context of Visual Arts in the Islamic World, College Art Association of America (University Park and London, 1988), pp. 67–89. Joshi, M. C., ‘The authorship of Purånå Qilæa and its buildings’, in Frederick M. Asher and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985), pp. 269–74. JumÆah, Ibrahim, Diråsa fÈ Ta†awwur al-kitåbåt alkËfiyya Æalå A˙jår fÈ Mißr fÈæl-qurËn al-khamsa alËla lil-hijra (Cairo, 1969). Kalus, Ludvik, Catalogue des Cachets, Bulles et Talismans Islamiques (Paris, 1981). — ‘Donations pieuses d’épées médiévales à l’arsenal d’Alexandrie’, Revue des Études Islamiques 50 (1982), 1–174. — Ashmolean Museum: Catalogue of Islamic Seals and Talismans (Oxford, 1986). Karama©arali, Beyhan, Ahlat Mezartaslari (Ankara, 1972). Karim, Abdul, Corpus of the Arabic and Persian Inscriptions of Bengal, Asiatic Society of Bangladesh (Dhaka, 1992). Kendrick, A. F., Catalogue of Muhammadan Textiles of the Medieval Period (London, 1924). Kenesson, Summer S., ‘Nasrid luster pottery: the Alhambra vases’, Muqarnas 9 (1992), 93–115. Kessler, Christel, ‘ÆAbd al-Malik’s inscription in the Dome of the Rock: a reconsideration’, Journal of the Royal Asiatic Society 3 (1970), 2–14. Khan, F. A., Banbhore (Karachi, [1963]).

— 230 —

— BIBLIOGRAPHY — Kiani, M. Y., F. Karimi and A. Gouchani, Introduction to the Art of Iranian Tilework in the Islamic Period (Tehran, 1983). King, Donald, ‘The Ardabil puzzle unravelled’, Hali 88 (1996), 88–92. Kircher, G., ‘Die Moschee des Mu˙ammad B. HairËn (“Drei-Tore-Moschee”) in Qairawån/Tunesien. Erster Bericht’, Mitteilungen des Deutschen Archäologischen Instituts Abteilung Kairo 26 (1970), 141–68. Komaroff, Linda, ‘Timurid to Safavid Iran: continuity and change’, Marsyas 20 (1979–80), 11–16. — ‘Pen-case and candlestick: two sources for the development of Persian inlaid metalwork’, Metropolitan Museum Journal 23 (1988), 89–102. — The Golden Disk of Heaven: Metalwork of Timurid Iran, Persian Art Series (Costa Mesa CA, 1992a). — ‘Persian verses of gold and silver: the inscriptions on Timurid metalwork’, in Lisa Golombek and Maria Subtelny (eds), Timurid Art and Culture: Iran and Central Asia in the Fifteenth Century, supplement to Muqarnas (Leiden, 1992b), pp. 144–57. — ‘Paintings in silver and gold: the decoration of Persian metalwork and its relationship to manuscript illustration’, Studies in the Decorative Arts 2 (Fall 1994), 2–34. Köseo©lu, Cengiz, The Topkapı Saray Museum: The Treasury, trans., expanded and ed. J. M. Rogers (Boston, 1987). Krachkovskaya, V. A., ‘Evolyutsiya Kuficheskogo pis’ma v sredney azii’, Epigrafika Vostoka 3 (1949), 3–27. Kröger, Jens, Nishapur: Glass of the Early Islamic Period (New York, 1995). Kühnel, Ernst, ‘Datierte persische Faÿencen’, Jahrbuch der Asiatischen Kunst 1 (1924), 42–52. — ‘Dated Persian lustred pottery’, Eastern Art 3 (1931), 230–6. — Catalogue of Dated Tiraz Fabrics: Umayyad, Abbasid, Fatimid, Technical analysis by Louisa Bellinger (The Textile Museum, Washington, DC, 1952). — ‘Die Kunst Persiens unter den Buyiden’, Zeitschrift der Deutschen Morgenländischen Gesellschaft 106 (1956), 78–92. — Die islamischen Elfenbeinskulpturen VIII–XIII Jahrhundert (Berlin, 1971). Lamm, C. J., Mittelalterliche Gläser und Steinschnittarbeiten aus dem Nahen Osten (Berlin, 1930). — ‘Fatimid woodwork, its style and chronology’, Bulletin de l’Institut d’Égypte 18 (1936), 59–91. — ‘Dated or datable Êiråz in Sweden’, Monde Oriental 32 (1938), 103–25.

Lanci, Michelangelo, Tratto delle simboliche rappresentanze arabiche (Paris, 1845–6). Lane, Arthur, ‘Glazed relief ware of the ninth century A.D.’, Ars Islamica 6 (1939), 56–65. Lane, Edward William, An Account of the Manners and Customs of the Modern Egyptians (London, 1836; reprinted New York, 1973). Le Strange, Guy, The Lands of the Eastern Caliphate, Cambridge Geographical Series (Cambridge, 1905). Lentz, Thomas W. and Glenn D. Lowry, Timur and the Princely Vision (Los Angeles, 1989). Lévi-Provençal, E., Inscriptions arabes d’Espagne (Leiden and Paris, 1931). Lewcock, R. B. and G. R. Smith, ‘Three medieval mosques in the Yemen’, Oriental Art 20 (1974), 75–86, 192–203. Lowick, Nicholas M., Siraf XV: The Coins and Monumental Inscriptions (London, 1985). Mackie, Louise W., ‘Toward an understanding of Mamluk silks: national and international considerations’, Muqarnas 2 (1984), 127–46. Maddison, Francis, ‘Al-JazarÈ’s combination lock: two contemporary examples’, in Julian Raby (ed.), The Art of Syria and the JazÈra 1100–1250, Oxford Studies in Islamic Art (Oxford, 1985), pp. 141–58. Mantran, Robert, ‘Les inscriptions arabes de Brousse’, Bulletin d’études Orientales 14 (1952–4), 87–114. — ‘Les inscriptions turques de Brousse’, Oriens 11 (1959), 115–70. Marçais, Georges, Les Faïences à Reflets Métalliques de la Grande Mosquée de Kairouan (Paris, 1928). Marçais, Georges and Gaston Wiet, ‘Le “voile de Sainte Anne” d’Apt’, Academie des Inscriptions et BellesLettres, Fondation Eugène Piot, Monuments et Mémoires 34 (1934), 1–18. Maricq, André and Gaston Wiet, Le Minaret de Djam: la découverte de la capitale des Sultans Ghorides (XIIe–XIIIe siècles) = Mémoires de la Délégation archéologique française en Afghanistan, vol. 16 (Paris, 1959). Markel, Stephen, ‘Fit for an emperor: inscribed works of decorative art acquired by the Great Mughals’, Orientations 21 (August 1990), 22–36. Marshak, B. I., ‘Bronzoviy kuvshin iz Samarkanda’, in A. A. Ivanov (ed.), Srednaya Aziya i Iran (Leningrad, 1972), 61–90. Mars[c]hak, Boris, Silberschätze des Orients: Metallkunst des 3.–13. Jahrhunderts und ihre Kontinuität, trans. L. Schirmer (Leipzig, 1986). Masson, M. E., ‘Tretii kusok nefritovogo namogil’nika TÈmËra’, Epigrafika Vostaka 2 (1948), 63–75.

— 231 —

— ISLAMIC INSCRIPTIONS — Masterpieces of Islamic Art in the Hermitage Museum (Kuwait, 1990). al-MaydånÈ, Abuæl Fa∂l A˙mad b. Mu˙ammad alNaysåbËrÈ, MajmaÆ al-amthål (Beirut, 1391/1971). Mayer, L. A., Saracenic Heraldry (Oxford, 1933). — Islamic Architects and Their Works (Geneva, 1956a). — Islamic Astrolabists and Their Works (Geneva, 1956b). — Islamic Woodcarvers and Their Works (Geneva, 1958). — Islamic Metalworkers and Their Works (Geneva, 1959). — Islamic Armourers and Their Works (Geneva, 1962). Meinecke, Michael, ‘Mamlukische Marmordekorationen in der Osmanischen Turkei’, Mitteilungen des Deutschen Archäologischen Instituts: Abteilung Kairo 27:2 (1971), 207–20. — Fayencedekorationen Seldschukischer Sakralbauten in Kleinasien, 2 vols (Tübingen, 1976). — ‘Die Mamlukischen Faiencemosaikdekorationen: eine Werkstätte aus Tabriz in Kairo (1330–1350)’, Kunst des Orients 11 (1976–7), 85–144. — ‘Syrian blue-and-white tiles of the 9th/15th century’, Damaszener Mitteilungen 3 (1988), 203–14. — Die Mamlukische Architektur in Ägypten und Syrien, 2 vols, Abhandlungen des Deutschen Archäologischen Instituts Kairo, Islamische Reihe (Glückstadt, 1992). Melikian-Chirvani, A. S., ‘Remarques préliminaires sur un Mausolée Ghaznévide’, Arts Asiatiques 17 (1968), 59–86. — ‘Un bassin iranien de l’an 1375’, Gazette des Beauxarts 78 (1969a), 5–18. — ‘Bronzes et cuivres iraniens du Louvre: I. L’École du Fars au XIVe siècle’, Journal Asiatique 257 (1969b), 20–36. — ‘Eastern Iranian architecture: à propos of the GhËrid parts of the Great Mosque of Haråt’, Bulletin of the School of Oriental and African Studies 33:2 (1970), 322–6. — ‘Nouvelles remarques sur l’école du Fars à propos des bassins iraniens du XIVe s. au Musée des BeauxArts’, Bulletin des Musées et Monuments Lyonnais 4:3 (1971a), 361–91. — ‘Le royaume de Salomon: les inscriptions persanes de sites achéménides’, Le monde iranien et l’Islam 1 (1971b), 1–41. — ‘Iranian metal-work and the written word’, Apollo (April 1976), 286–91. — ‘Essais sur la sociologie de l’art islamique – I. Argenterie et féodalité dans l’Iran médieval’, in C. Adle (ed.), Art et Société dans le Monde Iranien (Paris, 1982a), pp. 143–75. — Islamic Metalwork from the Iranian World 8–18th Centuries (London, 1982b).

— ‘Le Shåh-nåme, la gnose soufie et le pouvoir mongol’, Journal Asiatique 272 (1984), 249–338. — ‘Silver in Islamic Iran: the evidence from literature and epigraphy’, in Michael Vickers (ed.), Pots and Pans: A Colloquium on Precious Metals and Ceramics, Oxford Studies in Islamic Art (Oxford, 1986a), pp. 89–106. — ‘State inkwells in Islamic Iran’, Journal of the Walters Art Gallery 44 (1986b), 70–94. — ‘The lights of Sufi shrines’, Islamic Art 2 (1987), 117–48. — ‘Le livre des rois, miroir du destin, I’, Studia Iranica 17:1 (1988), 7–46. — ‘From the royal boat to the beggar’s bowl’, Islamic Art 4 (1990–1), 3–112. — ‘Le livre des rois, miroir du destin, II: Takht-e Soleymån et la symbolique du Shåh-nåme’, Studia Iranica 20:1 (1991), 33–148. Melville, Charles, ‘The Chinese Uighur animal calendar in Persian historiography of the Mongol period’, Iran 32 (1994), 83–98. Mercier, G., Corpus des inscriptions arabes et turques de l’Algerie: département de Constantine (Paris, 1902). Micklewright, Nancy, ‘Êiråz fragments: unanswered questions about medieval Islamic textiles’, in C. G. Fisher (ed.), Brocade of the Pen: The Art of Islamic Writing (East Lansing MI, 1991), pp. 31–46. Miles, George C., ‘Epitaphs from an Isfahan graveyard’, Ars Islamica 6 (1939), 151–7. — ‘Early Islamic inscriptions near Tåæif in the Óijåz’, Journal of Near Eastern Studies 7 (1948), 236–42. — ‘Early Islamic tombstones from Egypt in the Museum of Fine Arts, Boston’, Ars Orientalis 2 (1957), 215–26. — Byzantium and the Arabs: relations in Crete and the Aegean area’, Dumbarton Oaks Papers 18 (1964a), 1–32. — ‘A portrait of the Buyid prince Rukn al-Dawlah’, American Numismatic Society Museum Notes 11 (1964b), 283–93. — ‘Inscriptions on the minarets of Saveh, Iran’, in Studies in Islamic Art and Architecture in Honour of Professor K. A. C. Creswell (Cairo, 1965), pp. 163–78. — ‘The inscriptions of the Masjid-i JamiÆ at Ashtarjan’, Iran 12 (1974), 89–98. — ‘Numismatics’, in R. N. Frye (ed.), From the Arab Invasion to the Saljuqs, Cambridge History of Iran, vol. 4 (Cambridge, 1975), pp. 364–77. Minovi, M., ‘A Persian quatrain on a dyed silk’, Bulletin of the American Institute for Iranian Art and Archaeology 5 (December 1937), 170–1.

— 232 —

— BIBLIOGRAPHY — Moaz, K. and S. Ory, Inscriptions Arabes de Damas: Les Stèles Funéraires. I. Cimetière d’al-Bâb al-Sagîr (Damascus, 1977). Morton, A. H., Glass Weights in the British Museum (London, 1985). Museum für Islamische Kunst Berlin, Katalog 1971 (Berlin, 1971). al-NaqshbandÈ, ÆUsåma Nåßir and Óayåt ÆAbÈd ÆAlÈ alÓurrÈ, Al-akhtåm al-islåmiyya fÈæl-mat˙af al-ÆiråqÈ (Baghdad, 1975). Nastich, V. N. and B. D. Kochnev, ‘K atributsii mavzoleya Shakh-Fazil’, Epigrafika Vostoka 24 (1988), 69–76. Nath, R., Calligraphic Art in Mughal Architecture (Calcutta, 1979). Noci, Francesco, ‘Un antico minbar in legna della Masjid-i Jamæa di Isfahan’, Quaderni di Studi Arabi 2 (1982), 1–12. Nykl, A. R., ‘The inscription on the Freer vase’, Ars Orientalis 2 (1957), 496–7. Ocaña Jimenez, M., Repertorio de inscripciones árabes de Almería (Madrid, 1964). O’Kane, Bernard, ‘The Madrasa al-Ghiyasiyya at Khargird’, Iran 14 (1976), 79–92. — ‘The tiled minbars of Iran’, Annales Islamologiques 22 (1986), 133–55. — Timurid Architecture in Khurasan (Costa Mesa CA, 1987). Oral, M. Zeki, ‘Anadolu’da San’at Degeri Olan Ahsap Minberler, Kitabeleri Ve Tarihçeleri’, Vakiflar Dergisi 5 (1962), 23–77. Orazi, Roberto, The Wooden Gratings in Safavid Architecture (Rome, 1976). Orbeli, I. A. and K. V. Trever, Sasanidskiy Metall (Leningrad, 1935). Ory, Solange, Cimetières et inscriptions du Óawrån et du ∆abal al-DurËz, Éditions Recherche sur les Civilisations (Paris, 1989). Otavsky, Karel and Mu˙ammad ÆAbbås Mu˙ammad SalÈm, Mittelalterliche Textilien I: Ägypten, Persien und Mesopotamien, Spanien und Nordafrika, in collaboration with Cordula M. Kessler (Riggisberg, 1995). Pauty, E., Les bois sculptés jusqu’à l’époque ayyoubide (Cairo, 1931). Pearson, J. D. assisted by Michael Meinecke and George T. Scanlon, A Bibliography of the Architecture, Arts and Crafts of Islam; Second Supplement Jan. 1972 to Dec. 1980 (with Omissions from Previous Years (Cairo, 1984). Petsopoulos, Yanni (ed.), Tulips, Arabesques and Turbans: Decorative Arts from the Ottoman Empire (New York, 1982).

Pfister, R., ‘Matériaux pour servir au classement des textiles égyptiens postérieurs à la Conquête Arabe’, Revue des Arts Asiatiques 10 (1936), 1–16, 73–85. — ‘Toiles à inscriptions Abbasides et Fatimides’, Bulletin d’Études Orientales 9 (1945–6), 46–90. Philon, Helen, Early Islamic Ceramics: Ninth to Late Twelfth Centuries, Benaki Museum Athens catalogue (London, 1980). Pinder-Wilson, Ralph, ‘Some rock crystals of the Islamic period’, British Museum Quarterly 19 (1954), 84–7. — ‘The minaret of MasÆËd III at Ghazni’, in Studies in Islamic Art (London, 1985), pp. 89–102. Pinder-Wilson, R. H. and C. N. L. Brooke, ‘The reliquary of St. Petroc and the ivories of Norman Sicily’, Archaeologia 104 (1973), 261–306. Pinder-Wilson, Ralph and William Watson, ‘An inscribed jade cup from Samarqand’, British Museum Quarterly 23 (1960), 19–21. Porter, Venetia, Medieval Syrian Pottery (Oxford, 1981). Quraishi, Muhammad Hamid, ‘A Kufic Sarada inscription from the Peshawar Museum’, Epigraphia Indo-Moslemica (1925–6), 27–8. Rabino di Borgomale, H. L., Coins, Medals and Seals of the Shâhs of Irân (Cambridge, 1945). Raby, Julian, ‘Court and export: Part 2. The U∞ak carpets’, in Robert Pinner and Walter B. Denny (eds), Oriental Carpet and Textile Studies 2: Carpets of the Mediterranean Countries 1400–1600 (London, 1986), pp. 177–88. . Rå™ib, Yusuf, ‘Un contrat de mariage sur soie d’Égypte Fatimide’, Annales Islamologiques 16 (1980), 31–7. al-Rashid, Saad A., Darb Zubaydah: The Pilgrimage Road from Kufa to Mecca (Riyadh, 1980). Reinaud, J.-T., Monumens arabes, persans et turcs, du cabinet de M. le Duc de Blacas et d’autres cabinets, 2 vols (Paris, 1828). Reitlinger, Gerald, ‘Unglazed relief pottery from northern Mesopotamia’, Ars Islamica 15–16 (1951), 11–22. Rice, D. S., ‘The oldest dated “Mosul” candlestick’, Burlington Magazine 91 (1949), 334–40. — ‘The blazons of the “Baptistère de Saint Louis’’’, Bulletin of the School of Oriental and African Studies 23 (1950), 367–80. — ‘The brasses of Badr al-DÈn Luæluæ’, Bulletin of the School of Oriental and African Studies 13:3 (1951), 627–34. — ‘Studies in Islamic metalwork’, Bulletin of the School of Oriental and African Studies 14 (1952), 564–78. — The Baptistère de Saint-Louis (Paris, 1953a).

— 233 —

— ISLAMIC INSCRIPTIONS — — ‘Studies in Islamic metalwork – II’, Bulletin of the School of Oriental and African Studies 15 (1953b), 61–79. — ‘Studies in Islamic metalwork – III’, Bulletin of the School of Oriental and African Studies 15 (1953c), 229–45. — ‘Studies in Islamic metalwork – IV’, Bulletin of the School of Oriental and African Studies 15 (1953d), 489–503. — ‘Studies in Islamic metalwork – V’, Bulletin of the School of Oriental and African Studies 17 (1955), 206–31. — ‘Arabic inscriptions on a brass basin made for Hugh IV de Lusignan’, in Studi orientalistici in onore di Giorgio Levi Della Vida (Rome: Istituto Per L’Oriente, 1956a), vol. 2, pp. 390–402. — ‘A datable Islamic rock crystal’, Oriental Art, new series, 2 (Autumn 1956b), 85–93. — ‘Inlaid brasses from the workshop of A˙mad alDhakÈ al-MawßilÈ’, Ars Orientalis 2 (1957a), 283–326. — ‘Two unusual Mamluk metal works’, Bulletin of the School of Oriental and African Studies 20 (1957b), 487–500. — ‘Studies in Islamic metalwork – VI’, Bulletin of the School of Oriental and African Studies 21 (1958), 225–53. Rogers, J. M., ‘Waqf and patronage in Seljuk Anatolia: the epigraphic evidence’, Anatolian Studies 26 (1976), 69–103. — Islamic Art and Design: 1500–1700 (London, 1983). — ‘Calligraphy and common script: epitaphs from Aswan and Akhlat’, in Priscilla P. Soucek (ed.), Content and Context of Visual Arts in the Islamic World, College Art Association of America (University Park and London, 1988), pp. 105–37. Rogers, J. M. and R. M. Ward, Süleyman the Magnificent (London, 1988). Rosseló Bordoy, Guillem, Corpus Balear de epigrafía árabe (Mallorca, 1975). Roy, B. and P. Poinssot, Inscriptions arabes de Kairouan (Paris, 1950–8) SalÈm, Mu˙ammad ÆAbbås Mu˙ammad, ‘The Function of some woven fabrics in Riggisberg’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997a), 65–70. — ‘The inscriptions on Islamic fabrics in the Abegg Foundation’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997b), 123–8. Sarre, Friedrich, Denkmäler persischer Baukunst (Berlin,

1901–10). — Sammlung F. Sarre. Teil I – Metall (Berlin, 1906). — Die Keramik von Samarra (Berlin, 1925). Sauvaget, Jean, Les monuments historiques de Damas (Beirut, 1932a). — ‘Décrets mamlouks de Syrie’, Bulletin des Études Orientales 2 (1932b), 1–52; 3 (1933), 1–29; 12 (1947–8), 5–60. — ‘La tombe de l’Orto˚Ède Balak’, Ars Islamica 5 (1938), 207–15. — ‘Glanes épigraphiques’, Revue des Études Islamiques 6 (1941–6), 17–28. — ‘Introduction à l’étude de la céramique musulmane’, Revue des Études Islamiques 33 (1965), 1–64. Savage-Smith, Emilie and Marion B. Smith, Islamic Geomancy and a Thirteenth-Century Divinatory Device (Malibu CA, 1980). Scarce, Jennifer, ‘ÆAli Mohammad Isfahani, tile maker of Tehran’, Oriental Art 22 (1976), 278–88. — ‘The arts of the eighteenth to twentieth centuries: architecture, ceramics, metalwork, textiles’, in Peter Avery, Gavin Hambly and Charles Melville (eds), From Nadir Shah to the Islamic Republic, The Cambridge History of Iran (Cambridge, 1991), pp. 890–958. Scarce, Jennifer and L. P. Elwell-Sutton, ‘A problem piece of Kashmiri metalwork’, Iran 9 (1971), 71–85. Schlumberger, Daniel and Janine Sourdel-Thomine, Lashkari Bazar: une résidence royale Ghaznévide et Ghoride, 3 vols, Mémoires de la délégation archéologique française en Afghanistan, vol. XVIII (Paris, 1978). Schneider, Madeleine, Stèles funéraires musulmanes des îles Dahlak (Mer Rouge), 2 vols, Textes arabes et études islamiques 19 (Cairo, 1983). — ‘Remarques au sujet d’une inscription du mausolée dit de Baba Hatim’, Studia Iranica 13:1 (1984), 165–8. — ‘Les inscriptions arabes de l’ensemble architectural de Ûafår-∫È BÈn (Yémen du nord)’, Journal Asiatique 273 (1985), 61–137, 293–369. — Mubårak al-MakkÈ: An Arabic Lapicide of the Third/Ninth Century, Journal of Semitic Studies Monograph No. 9 (Manchester, 1986). — ‘Deux études d’épigraphie arabe’, Journal Asiatique 275 (1987), 223–52. — ‘Mußallå DåwËd (Yémen du nord) et ses stèles funéraires musulmanes’, Journal Asiatique 280 (1992), 1–79.

— 234 —

— BIBLIOGRAPHY — Schorta, Regula, ‘Technische Aspekte gewebter Inschriften’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997), 139–43. Semenov, A. A., ‘Nadpisi na nadgroviyax TÈmËra e ego Potomkov v Gur-i Emire’, Epigrafika Vostoka 2 (1948), 49–62. Sharon, M., ‘An Arabic inscription from the time of the Caliph ÆAbd al-Malik’, Bulletin of the School of Oriental and African Studies 29 (1966), 369–72. — ‘Un nouveau corpus des inscriptions arabes de Palestine’, Revue des Études Islamiques 42 (1974), 185–91. — ‘The AyyËbid walls of Jerusalem: a new inscription from the time of al-MuÆaΩΩam ÆÏså’, in Myriam Rosen-Ayalon (ed.), Studies in Memory of Gaston Wiet (Jerusalem, 1977), pp. 179–94. Shepherd, Dorothy G., ‘A dated Hispano-Islamic silk’, Ars Orientalis 2 (1957), 373–82. — ‘Medieval Persian silks in fact and fancy (a refutation of the Riggisberg Report)’, Bulletin de liaison du centre international d’étude des textiles anciens 39–40 (1974). Shepherd, Dorothy and W. Henning, ‘ZandanÈjÈ identified?’ in Aus der Welt der Islamischen Kunst: Festschrift für Ernst Kühnel, (Berlin, 1959), pp. 15–40. Shishkin, V. A., ‘Nadpisi v ansamble Shakhi-Zinda’, Zodchestvo Uzbekistana 2 (1970), 7–71. Shokoohy, Mehrdad, Bhadresvar: The Oldest Islamic Monuments in India (Leiden, 1988). — ‘Review of Abdul Karim, Corpus of the Arabic and Persian Inscriptions of Bengal’, Bulletin of the School of Oriental and African Studies 58:1 (1995), 161–7. Siddiq, Mohammad Yusuf, ‘An epigraphical journey to an eastern Islamic land’, Muqarnas 7 (1990), 83–109. Smirnov, V. I., Argenterie Orientale (St. Petersburg, 1909). Smith, Myron Bement, ‘The Manårs of Ißfahån’, Åthåré Ïrån 1 (1936), 313–58. — ‘Two dated Seljuk monuments at Sin’, Ars Islamica 6 (1939), 1–10. Smith, M. B. and Paul Wittek, ‘The wood mimbar in the Masdjid-i DjamiÆ, Nain’, Ars Islamica 5 (1938), 21–35. Sobernheim, M., Matériaux pour un Corpus Inscriptionum Arabicarum II: Syrie du Nord, Mémoires de l’Institut français archéologique du Caire XXV (Cairo, 1909). Sokoly, Jochen A., ‘Between life and death: the funerary context of tiraz textiles’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997a), 71–8.

— ‘Towards a model of early Islamic textile institutions in Egypt’, in Islamische Textilkunst des Mittelalters: Aktuelle Probleme = Riggisberger Berichte 5 (1997b), 115–22. Sourdel, D. and J. Sourdel, La civilisation de l’Islam classique (Paris, 1968). Sourdel, Dominique and Janine Sourdel-Thomine, ‘À propos des monuments des Sangbast’, Iran 17 (1979), 109–14. Sourdel-Thomine, Janine, Epitaphes Coufiques de Båb SaghÈr vol. 4 of Les Monuments Ayyoubides de Damas (Paris, 1950). — ‘Deux minarets d’époque Seljoukide en Afghanistan’, Syria 30:1–2 (1953), 110–21. — ‘Stèles arabes anciennes de Syrie du nord’, Annales Archéologiques de Syrie 6 (1956a), 11–38. — ‘Stèles arabes de Bust (Afghanistan)’, Arabica 3 (1956b), 285–306. — ‘Le Coufique Alépin de l’époque Seljoukide’, in Mélanges Louis Massignon (Damascus, 1957), vol. 3, pp. 301–18. — ‘Deux épitaphs arabes anciennes’, Eretz-Israel 7 (1964), 112–15. — ‘Clefs et serrures de la KaÆba’, Revue des Études Islamiques 39:1 (1971), 29–86. — ‘Quelques réflexions sur l’écriture des premières stèles arabes du Caire’, Annales Islamologiques 11 (1972), 23–35. — ‘Inscriptions Seljoukides et salles à coupoles de Qazwin en Iran’, Revue des Études Islamiques 42 (1974), 3–43. — ‘A propos du cénotaphe de MahmËd à Ghazna (Afghanistan)’, in Abbas Daneshvari (ed.), Essays in Islamic Art and Architecture in Honor of Katharina Otto-Dorn (Malibu CA, 1981), pp. 127–35. Tabbaa, Yasser, ‘Monuments with a message: propagation of jihåd under NËr al-DÈn (1146–1174)’, in Vladimir P. Goss and Christine Verzár Bornstein (eds), The Meeting of Two Worlds: Cultural Exchange Between East and West During the Period of the Crusades, Studies in Medieval Culture (Kalamazoo MI, 1986), pp. 223–41. — ‘The transformation of Arabic writing: Part 1, Quræånic calligraphy’, Ars Orientalis 21 (1991), 119–48. — ‘The transformation of Arabic writing: Part 2, The public text’, Ars Orientalis 24 (1994), 119–47. Taqizadeh, S. H., ‘Various eras and calendars used in the countries of Islam’, Bulletin of the School of Oriental and African Studies 9 (1937), 903–22; 10 (1939), 107–32.

— 235 —

— ISLAMIC INSCRIPTIONS — Taylor, Christopher S., ‘Reevaluating the ShiÆi role in the development of monumental funerary architecture: the case of Egypt’, Muqarnas 9 (1992), 1–10. Terrasse, Henri, La Mosquée des Andalous à Fès (Paris, 1942). — La Grande Mosquée de Taza (Paris, 1943). — La Mosquée al-Qaraouiyin à Fès (Paris, 1968). Thackston, Wheeler M., ‘The role of calligraphy’, in Martin Frishman and Khan Hasan-Uddin (eds), The Mosque: History, Architectural Development & Regional Diversity (London, 1994), pp. 43–53. Tirmizi, S. A. I., Ajmer through Inscriptions (1532–1852 A.D.) (New Delhi, 1968). Torres Balbás, L., Arte almohade – Arte nazari – Arte mudéjar [= Ars Hispaniae 4] (Madrid, 1949). al-ÆUsh, M. Abuæl-Faraj, ‘A bronze ewer with a high spout in the Metropolitan Museum of Art and analagous pieces’, in Richard Ettinghausen (ed.), Islamic Art in the Metropolitan Museum of Art (New York, 1972), pp. 187–98. Uzunçar∞ılı, I. H., Tokat, Niksar, Zile, Turhal, Pazar, Amasya Vilåyet (Istanbul, 1345/1927). — Afyon Karahisar, Sadıklı, Bolvadin (Istanbul, 1347/ 1929). — Kütahya ∞ehri (Istanbul, 1932). — Topkapı Sarayı Müzesi Mühürler Seksiyonı Rehberi (Istanbul, 1959). Valderrama Martínez, F., Inscriptiones árabes de Tetuán (Madrid, 1975). Ventrone, Giovanna, ‘Iscrizioni inedite su ceramica samanide in collezioni italiane’, in Gururåjamañjarikå: Studi in onore di Giuseppe Tucci (Naples, 1974), vol. 1, pp. 221–32. Vicaire, Marcel, ‘Notes sur quatre mesures d’aumône inédits’, Hespéris 31 (1944), 1–14. Viollet, Henri and Samuel Flury, ‘Un monument des premiers siècles de l’hégire en Perse’, Syria 2 (1921), 226–34, 305–16. Volov [Golombek], Lisa, ‘Plaited Kufic on Samanid epigraphic pottery’, Ars Orientalis 6 (1966), 107–34. Walker, J., ‘Six Kufic inscriptions’, Le Muséon 54 (1951), 151–6. Walls, Archibald G. and Amal Abul-Hajj, Arabic Inscriptions in Jerusalem: A Handlist and Maps (London, 1980). Ward, Rachel, Islamic Metalwork (New York, 1993). Wardwell, Anne E., ‘A fifteenth-century silk curtain from Muslim Spain’, Bulletin of the Cleveland Museum of Art 70 (1983), 58–72. — ‘Panni tartarici: eastern Islamic silks woven with

gold and silver (13th and 14th centuries)’, Islamic Art 3 (1988–9), 95–173. — ‘Two silk and gold textiles of the early Mongol period’, Bulletin of the Cleveland Museum of Art 79 (December 1992), 354–79. Watson, Oliver, ‘Persian lustre ware from the 14th to the 19th centuries’, Le monde iranien et l’Islam 3 (1975), 63–80. — Persian Lustre Ware (London, 1985). — ‘Documentary mÈnåæÈ and AbË Zaid’s bowls’, in Robert Hillenbrand (ed.), The Art of the Seljuqs in Iran and Anatoli, Proceedings of a symposium held in Edinburgh in 1982 (Costa Mesa CA, 1994), pp. 170–80. Watt, W. Montgomery, Bell’s Introduction to the QurÆån, Islamic Surveys (Edinburgh, 1970). Welch, Anthony, ‘Quræån and tomb: the religious epigraphs of two early Sultanate tombs in Delhi’, in Frederick M. Asher and G. S. Gai (eds), Indian Epigraphy: Its Bearing on the History of Art (New Delhi, 1985), pp. 257–68. Welch, Anthony and Howard Crane, ‘The Tughluqs: master builders of the Delhi Sultanate’, Muqarnas 1 (1983), 123–66. Welch, Stuart Cary, India: Art and Culture 1300–1900 (New York and Munich, 1988; reprinted 1993). Wensinck, A. J., Concordance et indices de la tradition musulmane, 8 vols (Leiden, 1936–88; reprinted 1992). Wenzel, Marian, ‘Towards an assessment of Ayyubid glass style’, in Julian Raby (ed.), The Art of Syria and the JazÈra 1100–1250, Oxford Studies in Islamic Art (Oxford, 1985), pp. 99–112. — ‘Textiles and the Seljuks of Anatolia’, Hali, 53 (1990), 136–43. — Ornament and Amulet: Rings of the Islamic Lands, ed. Julian Raby, The Nasser D. Khalili Collection of Islamic Art (London, 1993). Wiet, Gaston, ‘Notes d’épigraphie Syro-musulmane’, Syria 5 (1924), 216–53. — Lampes et bouteilles en verre émaillé, catalogue générale du Musée Arabe du Caire (Cairo, 1929). — Matériaux pour un Corpus Inscriptionum Arabicarum I: Égypte 2, Mémoires de l’Institut français archéologique du Caire LIII (Cairo, 1929–30). — Objects en cuivre, catalogue générale du Musée Arabe du Caire (Cairo, 1932). — L’exposition persane de 1931, Musée national de l’art arabe (Cairo, 1933). — Les stèles funéraires, vols II and IV–X, Musée Arabe (Cairo, 1936–42).

— 236 —

— BIBLIOGRAPHY — — ‘Nouvelles inscriptions fatimides’, Bulletin de l’Institut d’Égypte 24 (1941–2), 145–55. — Soieries Persanes (Cairo, 1948). — ‘Stèles coufiques d’Égypte et du Soudan’, Journal Asiatique 240 (1952), 273–97. — ‘Une nouvelle inscription fatimide au Caire’, Journal Asiatique 249 (1961), 13–20. Wilber, Donald N., ‘Builders and craftsmen of Islamic Iran: the earlier periods’, Art and Archaeology Research Papers 10 (1976), 31–9. Wilkinson, Charles K., Nishapur: Pottery of the Early Islamic Period (New York, 1973). Williams, C., ‘The cult of ÆAlid saints in the Fatimid monuments of Cairo. Part I: The mosque of alAqmar’, Muqarnas 1 (1983), 37–52. — ‘The cult of ÆAlid saints in the Fatimid monuments of Cairo. Part II: The mausolea’, Muqarnas 3 (1985), 39–60.

— ‘The Quræanic inscriptions on the Tabut of alHusayn in Cairo’, Islamic Art 2 (1987), 3–14. — ‘The mosque of Sitt Hadaq’, Muqarnas 11 (1994), 55–64. Yakubovsky, A. Yu., ‘Dve nadpisi na severnom mavzolee 1152 g. v Uzgende’, Epigrafika Vostoka 1 (1947), 27–32. Youssef, Abdel Raouf, ‘Etude sur le verre égyptien (résumé)’, in Colloque international sur l’histoire du Caire (Gräfenhainichen, 1972), pp. 467–8. Zbiss, S. M., Inscriptions de Tunis et de sa banlieue, Corpus des inscriptions arabes de Tunisie (Tunis, 1955). — Inscriptions de Gorjani, Corpus des inscriptions arabes de Tunisie (Tunis, 1962a). — Inscriptions de Monastir, Corpus des inscriptions arabes de Tunisie (Tunis, 1962b).

— 237 —

Index ÆAbbas I (Safavid), 29–40, 68, 102, 128, 179 ÆAbbasid(s), 8, 29, 36–41, 47, 59, 74, 79, 84, 88, 107, 119, 137, 164, 165, 166, 181, 183, 199, 215 ÆAbd al-Malik (Umayyad), 9, 16, 21, 29–41, 51, 76 abjad, 117, 220–2 Abu SaÆid (Ilkhanid), 37, 124, 175, 219 Achaemenid(s), 3, 46, 47 adjective(s), 35, 104–5, 217 Afghanistan, 45, 60–4, 69, 73, 137, 218 agate, 193 Aghlabid(s), 57, 188 Agra, 65, 137, 220 ahl al-bayt, 137, 166, 215 Æalå yad[ay], 40, 51, 100, 166, 189 Aleppo, 21, 57, 121, 130, 133, 139, 141, 185 Alexandria, 102, 200 Algeria, 43, 54, 135 ÆAli b. Abi Talib, 45, 69, 85, 88, 102, 127, 179, 193, 215, 216 ÆAlid(s), 45, 57, 68, 130 Almohad(s), 54, 125, 135, 143, 174–6 Almoravid(s), 54, 135, 215 Æamal, 49–51, 101–2, 113, 116, 117, 126, 189 amara, 30–2, 40, 45, 73, 100 amulet(s), 103, 193, 203–4 Anatolia, 13, 24, 37, 59–60, 74, 137, 146, 175, 178, 179, 201 aphorism(s), 125, 149–50, 152, 203; see also proverb(s) Aq Qoyunlu, 92, 102, 127, 199 Arabia, 10–11, 13, 25, 46 Arabic, 21, 26–8, 68, 97–9, 102, 103, 112, 119, 120, 127, 145 Armenian(s), 128, 163 arms, 147, 199–203 Artuqid(s), 57, 121 astrolabe(s), 101, 113, 116, 125, 220–1 Ayyubid(s), 32, 36, 43, 59, 91, 116, 121, 123, 130, 133, 139, 184–85, 201, 208 al-ÆAziz (Fatimid), 166, 190 Babylonian, 3, 47 Badr al-Din Luæluæ (vizier and ruler of Mosul), 112, 121 Badr al-Jamali (Fatimid), 69, 74, 130, 133 Baghdad, 8, 128, 137, 172, 174 Balkh, 16, 73, 118

bannåæÈ see Kufic, square banner(s), 59, 175, 178, 215 Baptistère de St Louis, 106, 114, 123 baraka, 8, 104, 108, 151, 152, 165, 166, 182, 183, 190, 192 basmala, 11, 29–30, 41, 45, 47, 73, 74, 100, 166, 183, 198 Bastam, 51, 62, 110, 145, 217 Bawandid(s), 45, 46, 62, 80 benediction(s), 36, 38–9, 152, 154, 209 Berlin, 108, 113, 119, 123, 126, 159, 161, 195 Blacas Ewer, 114, 116, 121 blazon(s), 149, 185, 187, 190 blessing(s), 7–8, 19, 41, 100, 104, 108, 114–19, 124, 130, 149, 165–9, 175, 181, 189, 192 Bobrinsky bucket, 98, 110–19 Boston, 98, 110, 118, 165, 167, 185 brick(s), 80, 82, 84 builder(s), 5, 24, 49–52 Bukhara, 24, 39, 51, 73, 144, 171 Bursa, 59, 146, 161 Buyid(s), 23, 38, 47, 49, 118, 119, 136, 171, 172, 219 Byzantine(s), 79, 165 Cairo, 4, 12, 13, 29–40, 45–6, 53, 57, 58, 59, 68, 71, 73–5, 79, 88, 91, 109, 114, 116, 121, 123, 129, 130, 133, 139, 141, 152, 154, 155, 157, 165, 167, 181–4, 190, 192, 207, 208, 211, 213, 215, 222 calendar(s), 23, 49, 158, 217–20 calligraphy, 8, 15–16, 25, 41–2, 49, 65, 88, 123, 124, 125, 127, 144–5, 152, 155, 158, 179–80, 201, 221–22 candlestick(s), 109–10, 116, 125 caravanserai(s), 39, 41, 44 carpet(s), 60, 179–81 cartouche(s), 11, 25, 49, 60, 91, 100, 108, 110, 127, 128, 139, 141, 143, 176, 193, 196 carving, 42, 46, 49, 143, 148 casket(s), 6, 100, 118, 188, 193 Caspian Sea, 21, 45, 97, 144, 219 cenotaph(s), 45, 46, 57, 62, 64, 67, 130, 132, 133, 137, 139, 141, 144, 158, 179, 193, 220 Central Asia, 16, 58, 60–4, 69, 73, 117, 144–6, 152, 169, 171, 176, 178, 198, 200, 213 chancellery, 21, 75, 88

— 238 —

— INDEX — chandelier(s), 112–3, 125 chessman, 187 China, 170 Chinese, 102 Christian(s), 4, 49, 80, 112, 123, 125, 150, 163, 174, 219 chronogram(s), 25, 93, 220, 221–2 Cleveland, 116, 121, 171, 172, 176 coats of arms, 125 coins, 12–13, 15–16, 21, 37, 41, 61, 209, 215, 219 colour(s), 39–40, 73, 88, 92–3, 114, 133 Commander of the Faithful, 35, 37, 39, 40, 41, 44, 100, 104, 151 Cordova, 53, 134, 135, 189, 210 Crusader(s), 57, 71, 132, 169 cursive, 16, 44, 59, 88–91, 114, 127, 133, 137, 157, 163, 167, 175, 198, 221 Damascus, 21, 32, 41, 68, 80, 88, 114, 116, 121, 123, 151, 160, 182, 183, 197, 213, 216 Damghan, 44, 45, 46, 51, 62, 73, 196 date(s), 23, 25, 40–1, 45, 49, 67, 99, 100, 102, 113, 116, 117–18, 120, 130, 148, 150, 154–7, 180, 217–22 Dawlatabad, 62, 73, 81, 88, 92 day(s), 40, 45, 49 decrees, 20, 44–5, 85, 92 Dhu’l-Nunid(s), 189 diacritical marks, 24, 76–7, 208 Diyarbekır, 55, 56, 59, 73, 74, 79, 199 duÆåæ, 102–5, 108, 200, 212, 216 Edirne, 25, 59, 92, 93, 110 Egypt(ian), 3, 4, 12, 13, 37, 38, 44, 49, 56–9, 73, 79, 80, 88, 102, 107, 108, 119, 129, 130, 139, 151, 152, 160, 164, 166, 169, 182, 187, 208, 213, 216 Elamite, 3, 47 Encyclopaedia of Islam, 14–15, 71 endowment(s), 16, 20, 35, 43–4, 93, 102 eulogy/eulogies, 35, 38–9, 43, 44, 57 Farghana, 23, 49, 58, 61, 69, 210, 219 farmån, 43–4, 85 Fars, 47, 123, 124 fasayakfÈkahum, 149, 151, 157, 198, 215 Fatimid(s), 4, 41, 56–9, 68, 69, 74, 75, 79, 119, 129, 130, 134, 135, 152, 153, 154, 165, 166, 168, 188–90, 192, 207, 208, 215, 216 Fez, 54, 125 133–5, 143–4

Firdawsi, 24, 98, 127, 159, 180 foundation texts, 24, 25, 29–42, 43–6, 54, 60, 62, 68, 69, 71, 88, 98–100, 137 fountain(s), 25, 27, 32, 60, 119, 141, 144, 163 Four Orthodox Caliphs, 85, 137, 163 Fourteen Innocent Ones, 68, 85 gate(s), 42, 52, 59, 73–5, 79, 113 Gawhar Shad (Timurid), 51, 193 genealogy, 36, 44, 45, 49, 82, 132, 198 Ghazna, 24, 61, 62, 82, 88, 137 Ghaznavid(s), 24, 37, 64, 80–2, 88, 112 Ghurid(s), 61, 64, 67, 69, 218 gift(s), 11, 132, 165, 176 glass, 182–7 glass weights, 182, 209 good wishes, 102–5, 108–9, 148, 149, 155, 156, 168, 178, 181, 189, 190, 192 Gothic, 4 Granada, 53, 54, 91, 117, 125, 143, 149, 159, 176 Greek(s), 3, 4, 21, 163, 219 Gurgan, 45, 46, 119, 150, 155, 193, 219 hadith, 19, 21, 25, 68, 69, 133, 137, 139, 145, 147, 157, 159, 216 Hafiz, 98, 127, 180 al-Hakam (Umayyad of Spain), 6, 100–1, 189 Hebrew, 125, 144, 163 Herat, 61, 104, 110, 114, 126 Hijaz, 4, 21, 79 historiography, 12–15 honorifics, 37, 102, 132 hospital(s), 32, 35, 41 Idrisid(s), 133 ikat, 166, 169 Ikhshidid(s), 58, 181 Ilkhanid(s), 37, 62, 85, 92, 110, 124, 144, 158, 175, 219 illegibility, 41, 159, 167 imam(s), 36, 37, 41, 103, 127, 130, 144, 216 Imam al-ShafiÆi, 45, 46, 132 incising, 46, 101, 114 India(n), 14, 64–7, 93, 128, 169, 179, 187, 188, 197, 213, 222 Inju, 123 inkwell(s), 108, 127 interlacing, 25, 62, 80 invocation, 100, 117, 130; see also basmala Iran(ian), 3, 14, 16, 21, 23–5, 37, 39, 41, 44–7, 49, 51,

— 239 —

— ISLAMIC INSCRIPTIONS — 60–4, 69, 73, 80, 82, 84, 92, 93, 97, 102, 104, 106, 108, 110, 112, 119, 123, 126, 144, 145, 146, 151, 182, 196, 213 ÆIraq, 38, 47, 59, 130, 133, 220 Isfahan, 5, 25, 29–32, 35, 39–41, 47, 49, 60, 68, 69, 73, 80, 84, 85, 88, 92, 116, 128, 144 ivory/ivories, 53, 54, 100, 132, 135, 141, 147, 187–90 jade, 144, 193–5, 199, 220 Java, 28, 220 Javanese, 28 Jazira, 88, 121 Jerusalem, 9, 13, 16, 19, 29–41, 43, 55, 76–7, 91, 121, 123, 130–3, 136, 211–17 Jesus, 4, 10, 76 jihåd, 6, 38, 57, 71, 175, 179, 201 KaÆba, 73, 107, 121, 126, 164, 213, 215 Kashan, 117, 155, 156, 157 key(s), 107, 121, 126 Khargird, 51, 82, 84, 144 Khurasan(i), 24, 25, 47, 60, 98, 104, 116, 119–21, 126–7, 199, 213, 215 Konya, 24, 88, 124, 137, 139, 159, 179 Koran(ic), 10–12, 16, 19, 21, 25, 36, 57, 59, 62, 65, 68, 69, 71, 73, 76, 77, 88, 100, 103, 123, 129, 130, 133, 137, 139, 141, 143, 147, 148, 155, 157, 159, 163, 164, 168, 178, 179, 190, 198, 199, 201, 203, 207, 210–17 Light Verse see Sura 24:36 Sura 1, 103, 157, 215 Sura 2:137, 149, 157, 151, 198, 215 Sura 2:144, 179 Sura 2:181, 214 Sura 2:255, 69, 73, 74, 80, 137, 139, 147, 156–7, 195–6, 213–15 Sura 3:18, 147, 198, 214–16 Sura 3:85, 215 Sura 3:90–1, 107, 215 Sura 3:167, 45 Sura 3:185, 91, 199, 216 Sura 3:190–1, 73 Sura 5:55, 215 Sura 6, 215 Sura 7:26, 201 Sura 9:14, 215 Sura 9:18, 69, 214 Sura 9:19, 215 Sura 9:33, 214, 216 Sura 9:129, 168

Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura Sura

— 240 —

10:57, 215 11:88, 168 12:64, 215 13:11, 215 15:17, 215 16:69, 215 17:78, 69, 214 17:79, 69, 215 17:80, 215 17:82, 215 17, 212 19, 61, 211 21:35, 216 22:7, 216 22:28, 73 23:114–16, 216 24:35, 11, 69, 73, 102, 187, 214 24:36, 11, 73, 134 26:80, 215 26:228, 45 29:7, 199 33:33, 137, 215 33:56, 47 35:5, 199 36, 132, 215 37:7, 215 39:53, 216 41:12, 215 41:44, 215 42:23, 215 48, 69, 201, 212, 214 48:28–9, 147 50:17–18, 11 53:39–41, 133 55:26–7, 159, 198, 216 57:1–7, 11 59:23, 85 61:10–12, 175 61:13, 201 62:1–5, 135 62, 68 67, 215 67:51–2, 179 68, 11 69:18–19, 12 69:28–9, 159 76:1–7, 217 85:20–2, 215 96:1–5, 11 110, 68

— INDEX — Sura 112, 119, 157, 198, 214, 215 Throne Verse see Sura 2:255 Koran box(es), 100, 113, 123, 147, 190 Kufesque see Kufic, pseudoKufic, 12, 13, 21, 24, 25, 56, 61, 76, 79, 80, 93, 98, 101, 108, 118, 124, 130, 132, 133, 141, 155, 157, 159, 166, 170, 175, 181, 182, 183, 184, 188, 189, 190, 198, 208 animated, 114, 120 bordered, 80–4 ‘broken’, 168 floriated, 24, 44, 56–9, 62, 78–9, 104, 119, 130, 133, 135 foliated, 56–7, 62, 77, 79, 130, 134, 137, 139, 141, 169 interlaced, 16, 58, 62, 67, 80, 104, 150, 152, 169, 190 pseudo-, 79–80, 167, 179 simple, 57, 58, 77, 117, 136 square, 73, 82–9, 100, 133, 137, 144, 145, 146, 159, 179, 198 with interlaced stems, 62, 80, 91, 135, 137, 159, 176, 179 with rising tails, 98 language(s), 21–8, 60, 68, 97, 119; see also Arabic, Babylonian, Elamite, Greek, Hebrew, Javanese, Nabatean, Pahlavi, Persian, Syriac, Ternate, Turkish Latin, 76, 127, 175 legibility, 24–5, 73, 80, 91, 115, 120, 125, 133, 135, 169, 178, 183 Levant, 57, 71, 88, 213, 217 London, 13, 99, 106, 112, 114, 116, 123, 127, 128, 149, 150, 155, 156, 161, 163, 165, 180, 182, 185, 193, 194, 195, 200, 203, 216 Los Angeles, 112, 120, 180 love(rs), 98, 99, 158, 189 madrasa(s), 29, 32, 35, 38, 39, 41, 43, 44, 46, 51, 61, 69–73, 81, 125, 143 Madrid, 190, 192, 200 Maghrib, 14, 29, 43, 53–4, 77, 79, 80, 91, 97, 113, 119, 125, 133, 135, 141, 143, 144, 181, 195, 198, 212, 213, 217, 220, 221, 222 maghribÈ, 175, 178 Mamluk(s), 29, 35, 37, 39, 44, 45, 49, 71, 73, 92, 97, 106, 107, 108, 109, 110, 116, 121, 123, 125, 127, 129, 139, 141, 160, 176, 182, 185, 187, 190, 199, 201, 207, 208 Mamluk-revival, 125, 187

al-Maæmun (ÆAbbasid), 29, 32, 36, 37, 38, 40, 41 al-Maqrizi, 39, 58, 164, 165 marble(s), 24, 32, 43, 44, 45, 53, 54, 75, 79, 88, 91, 93, 147 Marinid(s), 43, 113, 125, 141, 143, 144, 178 Marrakesh, 54, 135, 143, 196 Mashhad, 25, 51, 60, 99, 128, 144, 157, 161 Mazandaran, 23, 60, 98, 108, 144, 145 MCIA, 13–14, 16, 19, 53–60, 207–13 Mecca, 11, 38, 44, 102, 107, 110, 120, 164, 178, 179, 217 Medina, 38, 44, 102, 110, 141, 217 Mesopotamia, 37, 79, 116, 121 mihrab(s), 25, 43, 69, 130, 135, 136, 139, 214 milestone(s), 16, 21, 41, 76 minaret(s), 35, 61–2, 73–4, 80–2, 84, 88, 92, 139, 211 minbar(s), 69, 102, 130–46, 161 mistake(s), 42, 45, 105, 112, 158 Moghulistan, 193, 196 Mongol(s), 24, 37, 62, 85, 116, 157, 159, 198, 215 Mongolian, 85 month(s), 40, 45, 49, 54, 217, 219 Dhu'l-Hijja, 67, 110, 146, 156 Dhu'l-QaÆda, 119 Jumada I, 41, 139 Jumada II, 98, 128, 134, 156, 178, 192 Muharram, 88, 98, 110, 119, 135, 157, 178 RabiÆ I, 69 RabiÆ II, 40, 41, 127 Rajab, 45, 121, 128 Ramadan, 41, 69 Safar, 41, 69, 157 ShaÆban, 45, 119, 126, 158, 190 Shawwal, 41, 110, 156 Morocco, 54, 143, 179 mosaic(s), 9, 29, 39, 49, 68, 73, 76, 92, 170, 216 mosque(s), 32, 43, 51, 60–2, 67–70, 130, 214 Mosul, 13, 106, 112, 114, 121, 123, 136, 141, 185 al-muÆallim, 51, 52, 113, 114, 127 Mughal(s), 65, 128, 163, 192, 193, 195, 201, 220 Muhammad (Prophet), 10, 11, 19, 36, 38, 41, 45, 47, 69, 76, 85, 88, 103, 126, 132, 133, 137, 150, 169, 179, 203, 209, 212, 214, 215, 216, 217, 218 al-mulk [lillåh], 84, 151, 176–8 al-Mustansir (Fatimid), 75, 137, 167–8 Nabatean, 21 al-najjår, 51, 116, 133, 145 name(s), 36, 45, 49, 100, 108, 110, 112 al-naqqåsh, 112, 113, 119, 144

— 241 —

— ISLAMIC INSCRIPTIONS — al-Nasir Muhammad (Mamluk), 73, 109, 123, 139, 176, 185 naskh, 24, 82, 99–100, 116, 120, 132, 139, 143, 146, 155, 159, 176, 190, 208 Nasrid(s), 91, 125, 143, 149, 159, 176, 200, 201 nastaÆlÈq, 49, 100, 100, 110, 127, 194, 200 Natanz, 41, 136, 145, 210 New York, 112, 114, 128, 146, 155, 182, 183, 188, 195, 200, 201, 220 nisba, 36, 51, 52, 113, 114, 126, 181, 187 Nishapur, 16, 25, 44, 60, 106, 117, 151 Nizam al-Mulk (Saljuq), 73, 81, 104 North Africa see Maghrib numerals, 25, 41, 49, 154–5, 203, 220–2 Nur al-Din (Zangid), 38, 57, 88, 130, 132, 133 Ottoman(s), 25, 53, 59, 77, 92, 99, 100, 102, 107, 125, 127, 128, 129, 139, 147, 161, 163, 176, 178, 179, 190, 192, 195, 221–2 owner(s), 110, 128 Pahlavi, 21–3, 97, 108, 219 paint(ing), 46, 93, 129, 133, 139, 141, 143, 146, 148, 169 palace(s), 24, 35, 41, 42, 47, 54, 64, 91, 139, 158 Paris, 4, 12, 106, 114, 123, 128, 129, 154, 187, 192, 198, 203 Party Kings, 53, 54, 189 patron(s), 35, 36, 40, 44, 45, 46, 61, 62, 73, 100, 108, 113, 120, 121 pen-box(es), 115, 116, 119, 120, 123, 125 Persepolis, 23, 47, 49, 219 Persian, 3, 8–9, 21, 23–5, 35, 37, 47, 98–100, 102, 106, 110–12, 119, 120, 125, 127, 128, 145, 150, 156, 159, 161, 165, 179, 193, 195, 198, 219 Persian Gulf, 58, 60, 196 pious phrase(s), 10, 19, 21, 25, 59, 62, 68, 82, 84, 93, 100, 102, 145, 150, 159 plays on words, 38, 39, 75, 114, 203, 215 poetry, 24–5, 42, 53, 59, 93, 98–9, 106, 112, 119, 125, 128, 139, 148, 150, 155, 156, 157, 172, 179, 198, 199, 200, 201, 204, 222 prayer(s), 102–5, 108, 127, 139, 147, 200, 201, 203, 216, 217; see also duÆåæ pre-Islamic inscriptions, 3–4, 47, 112, 150, 170 proverb(s), 7–8, 149–50, 152; see also aphorism(s) pseudo-inscriptions, 156, 159, 167, 178, 179 Qaæitbay (Mamluk), 110, 125, 185, 187, 190 Qajar(s), 161, 200, 201

Qalaæun (Mamluk), 29, 35–9 Qarakhanid(s), 23, 24, 37, 49, 88 Qayrawan, 54, 57, 151, 188, 196 Qazvin, 44, 62, 128 Radkan, 45, 46, 62, 80 Raqqa, 151, 152, 182, 183 Rasulid(s), 109, 123 RCEA, 14, 16, 20, 209–10 readability, 76, 79, 80, 82, 88, 116, 149, 201 relief(s), 3, 8, 45, 46, 47, 61, 82, 101, 114, 156 rock crystal, 190–2 Roman(s), 3, 47 Romanesque, 3–4 SaÆdi, 98, 99, 112, 127, 193, 198 Safavid(s), 25, 29, 36, 37, 38, 60, 99, 102, 112, 116, 127, 128, 195, 199, 201, 203 Salah al-Din (Saladin), 43, 46, 123, 195 Saljuq(s), 37, 44, 59, 62, 73, 80, 81, 84, 92, 104, 118 Saljuqs of Rum, 38, 44, 46, 60, 149, 159 Samanid(s), 4, 6–8, 16, 24, 115, 149, 150, 151 Samarqand, 24, 39, 61, 69, 193, 196 SamarraÆ, 4, 128, 150 Sar-i Pul, 45–6, 61 Sasanian(s), 23, 47, 98 scripts see cursive, Kufic, maghribÈ, naskh, nastaÆlÈq, seal script, thuluth seal script, 88 seals, 12–13, 85, 203–4 ShiÆism, 36, 38, 57, 59, 68, 69, 74, 75, 128, 150, 192 Shiraz, 51, 114, 124 Sicily/Sicilian, 53, 119, 187, 189, 196, 198 signature(s), 20, 24, 25, 49–52, 59, 100–1, 113–14, 121, 128, 148, 150, 151, 152, 154, 155, 158, 183 silk(s), 4–5, 98, 164–79 Siraf, 58, 62, 67, 196, 199 Sivas, 24, 38, 159 smith(s), 51, 113 Spain, 6, 14, 43, 53, 54, 97, 117, 119, 125, 159, 187, 188, 196, 199 St Petersburg, 98, 99, 104, 108, 110, 117, 118, 119, 123, 125, 128, 159, 161,181, 185, 193 stone, 42, 44, 55, 58, 67, 73, 88, 91 stucco, 51, 85, 88, 91, 92 Sufi(s), 44, 103, 110, 159 Syria(n), 4, 13, 21, 44, 49, 56, 57, 59, 71, 88, 107, 108, 121, 151, 154, 160, 187, 195, 196, 199 Syriac, 21

— 242 —

— INDEX — Tabriz, 73, 92, 127, 160 Taæif, 21, 47, 76 talisman(s)/talismanic, 13, 100, 103, 201, 203–4, 221 Tehran, 82, 98, 108, 161, 190 Ternate, 28 thuluth, 11, 42, 100, 123, 124, 125, 128, 185, 193, 194 tile(s), 73, 92–3, 100, 156, 157, 159 Timur, 51, 61, 84, 110, 193, 196 Timurid(s), 25, 37, 49, 51, 61, 98, 106, 112, 126, 127, 161, 193, 196, 201 †iråz, 39, 165–70 titles/titulature, 35–42, 43–5, 51, 59, 60, 80, 82, 91, 100, 112 tomb(s), 21, 24, 32, 35, 39, 41, 43, 45–6, 51, 57, 61, 64, 65, 67, 71, 88, 91, 92, 145 tomb tower(s), 23, 45, 46, 52, 62, 80 tombstone(s), 13, 35, 45–9, 56–9, 59, 79, 80, 195–9 Transoxiana, 14, 24, 51, 62, 118 Tunisia, 52, 54, 151, 188, 196 Turkey, 3, 14, 93 Turkish, 25, 27, 36, 60, 100, 110, 163, 198, 200, 201, 203, 204 Uljaytu (Ilkhanid), 69, 92, 110, 159 Ulugh Beg (Timurid), 126, 193, 194, 195 Umayyad(s), 9, 16, 21, 29, 36, 37, 41, 45, 47, 68, 76, 88 Umayyad(s) of Spain, 6, 54, 100, 189 ÆUmar (Orthodox Caliph), 38, 88, 137 ÆUthman (Orthodox Caliph), 88, 137

Uzun Hasan (Aq Qoyunlu), 92, 102, 127 Venice, 80, 127, 190, 192 verb(s), 30, 32, 40, 43–4, 45, 49, 100; see also Æamal, amara verse(s), 8, 47, 47, 98, 99, 100, 112, 119, 127, 128, 149, 156, 158, 159, 161, 189, 190, 193, 195 vizier(s), 58, 59, 73, 74, 80, 104, 110, 115, 118, 120, 140 al-Walid (Umayyad), 45, 68 walls, 73–5 waqf see endowment Washington DC, 11, 98, 107, 115, 116, 119, 121, 124, 128, 149, 150, 152, 155, 156, 158, 160 wine, 98, 104, 112, 118, 125, 128, 193, 194, 195 women, 36, 46, 69, 120, 133, 139, 180, 189, 190, 192, 193 workshop(s), 99, 148, 156, 160 Yazd, 1, 73, 113 year(s), 40, 41, 45, 49, 113, 116, 117, 135, 183, 188, 217–21 Yemen, 2, 109, 123, 187 Zangid(s), 57, 88 Zanzibar, 28, 67 Zaydi(s), 147 zodiac, 108, 161

— 243 —