Viewing Inscriptions in the Late Antique and Medieval World 9781107092419, 1107092418

Table of contents :
Cover
Half title
Title
Copyright
Contents
Illustrations
Contributors
Acknowledgements
Introduction: Viewing Inscriptions
One. Inscriptions, Royal Spaces and Iranian Identity: Epigraphic Practices in Persia and the Ancient Iranian World
Two. Prayers on Site: The Materiality of Devotional Graffiti and the Production of Early Christian Sacred Space
Three. Erasure and Memory: Aghlabid and Fatimid Inscriptions in North Africa
Four. Textual Icons: Viewing Inscriptions in Medieval Georgia
Five. Pseudo-Arabic 'Inscriptions' and the Pilgrim's Path at Hosios Loukas
Six. Arabic Inscriptions in the Cappella Palatina: Performativity, Audience, Legibility and Illegibility
Seven. Intercession and Succession, Enlightenment and Reflection: The Inscriptional and Decorative Programme of the Qaratay Madrasa, Konya
Eight. Remembering Fernando: Multilingualism in Medieval Iberia
Nine. Displaying the Word: Words as Visual Signs in the Armenian Architectural Decoration of the Monastery of Noravank` (14th century)
Ten. Written in Stone: Civic Memory and Monumental Writing in the Cathedral of San Lorenzo in Genoa
Eleven. Place, Space and Style: Craftsmen's Signatures in Medieval Islamic Art
Afterword: Re-Viewing Inscriptions
Index

Citation preview

VIEWING INSCRIPTIONS IN THE LATE ANTIQUE AND MEDIEVAL WORLD Inscriptions convey meaning not just by their contents but also by other means, such as choice of script, location, scale, spatial organisation, letterform, legibility and clarity. The essays in this book consider these visual qualities of inscriptions, ranging across the Mediterranean and the Near East from Spain to Iran and beyond, including Norman Sicily, Islamic North Africa, Byzantium, medieval Italy, Georgia and Armenia. Although most essays focus on late antiquity and the Middle Ages, they also look back at Achaemenid Iran and forward to Mughal India. Topics discussed include real and pseudo-writing, multilingual inscriptions, graffiti, writing disguised as images and images disguised as words. From public texts set up on mountainsides or on church and madrasa walls to intimate craftsmen’s signatures barely visible on the undersides of precious objects, the inscriptions discussed in this volume reveal their meanings as textual and visual devices. Antony Eastmond is AG Leventis Reader in the History of Byzantine Art at the Courtauld Institute of Art, University of London. He has written extensively on the art and culture of medieval Georgia and its relations with Byzantium. He also works on Byzantine ivories. He is the author of The Glory of Byzantium and Early Christendom (2013), as well as Art and Identity in Thirteenth-Century Byzantium: Hagia Sophia and the Empire of Trebizond (2008) and Royal Imagery in Medieval Georgia (1998). He has published articles in the Art Bulletin, Art History, Dumbarton Oaks Papers and Speculum. He recently held a Leverhulme Major Research Fellowship during which he worked on a study of cultural interaction in eastern Anatolia on the eve of the Mongol invasions.

Viewing Inscriptions in the Late Antique and Medieval World Edited by

Antony Eastmond Courtauld Institute of Art, University of London

32 Avenue of the Americas, New York, ny 10013-2473, usa Cambridge University Press is part of the University of Cambridge. It furthers the University’s mission by disseminating knowledge in the pursuit of education, learning and research at the highest international levels of excellence. www.cambridge.org Information on this title: www.cambridge.org/9781107092419  C Cambridge University Press 2015

This publication is in copyright. Subject to statutory exception and to the provisions of relevant collective licensing agreements, no reproduction of any part may take place without the written permission of Cambridge University Press. First published 2015 Printed in the United States of America A catalog record for this publication is available from the British Library. Library of Congress Cataloging in Publication Data Viewing inscriptions in the late antique and medieval world / edited by Antony Eastmond (Courtauld Institute of Art, University of London). pages cm Includes bibliographical references and index. isbn 978-1-107-09241-9 (hardback) 1. Inscriptions – History – To 1500. 2. Visual communication – History – To 1500. 3. Civilization, Ancient – Sources. 4. Civilization, Medieval – Sources. 5. Mediterranean Region – Antiquities. 6. Middle East – Antiquities. I. Eastmond, Antony, 1966– cn77.v54 2015 929 .5–dc23 2014032617 isbn 978-1-107-09241-9 Hardback Cambridge University Press has no responsibility for the persistence or accuracy of urls for external or third-party Internet Web sites referred to in this publication and does not guarantee that any content on such Web sites is, or will remain, accurate or appropriate.

CONTENTS 鵼

Illustrations

page vii

Contributors

xiii

Acknowledgements

xvii

Introduction: Viewing Inscriptions Antony Eastmond

one. Inscriptions, Royal Spaces and Iranian Identity: Epigraphic Practices in Persia and the Ancient Iranian World Matthew P. Canepa

1

10

two. Prayers on Site: The Materiality of Devotional Graffiti and the Production of Early Christian Sacred Space Ann Marie Yasin

36

three. Erasure and Memory: Aghlabid and Fatimid Inscriptions in North Africa Jonathan M. Bloom

61

four. Textual Icons: Viewing Inscriptions in Medieval Georgia Antony Eastmond

five. Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas Alicia Walker

76

99

six. Arabic Inscriptions in the Cappella Palatina: Performativity, Audience, Legibility and Illegibility Jeremy Johns

124

seven. Intercession and Succession, Enlightenment and Reflection: The Inscriptional and Decorative Programme of the Qaratay Madrasa, Konya Scott Redford

148

v

vi

Contents

eight. Remembering Fernando: Multilingualism in Medieval Iberia Tom Nickson

nine. Displaying the Word: Words as Visual Signs in the Armenian Architectural Decoration of the Monastery of Noravank‘ (14th century) Ioanna Rapti ten. Written in Stone: Civic Memory and Monumental Writing in the Cathedral of San Lorenzo in Genoa Stefania Gerevini

170

187

205

eleven. Place, Space and Style: Craftsmen’s Signatures in Medieval Islamic Art Sheila S. Blair

230

Afterword: Re-Viewing Inscriptions Antony Eastmond

249

Index

257

ILLUSTRATIONS 鵼

1. 2. 3.

4. 5.

6.

7. 8. 9. 10. 11.

12. 13. 14.

Inscriptions and rock relief of Darius I (522–486 BCE), Bisotun, Iran. View of the Gate of All Lands, Persepolis, Iran. Created by Xerxes I (486–465 BCE). Detail of the Old Persian inscription (center, above the wing) flanked by the Elamite and Babylonian versions on the interior of the northern wall of the Gate of All Lands, Persepolis, Iran. Cult Foundation of Antiochus III, Laodicea-in-Media. The Kaʿba-ye Zardosht (foreground) with a partial view of the sites of Shabuhr I’s Parthian and Greek inscriptions (lower courses of masonry) with an Achaemenid tomb and Sasanian relief in the background. Naqsh-e Rostam, Iran. Sasanian stucco panel from Umm az-Zaʿatir (near Ctesiphon) with Pahlavi letters possibly forming a monogram of Middle Persian abzud (‘increased’). Museum fur ¨ Islamische Kunst, Berlin. Jubilee Doors, St. John Lateran, Rome. In situ section of graffiti at the Memoria Apostolorum triclia. Philae, Temple of Isis/Church of St. Stephen, south wall of hypostyle hall, west side of entrance, with insc. nos. 205–14. Graffiti-covered plaster fragments from the Memoria Apostolorum triclia. a. Graffiti wall at the shrine of St. Felix, Cimitile-Nola. b. Drawing of graffiti on left portion of graffiti wall at the shrine of St. Felix, Cimitile-Nola. Resafa, Basilica of the Holy Cross, detail of graffiti wall fragment B in situ. Reconstruction drawing of the interior of the Crypt of the Popes in the catacomb of Callixtus, Rome. Reconstruction drawing of the entrance of the Crypt of the Popes in the catacomb of Callixtus, Rome.

page 15 18

19 21

27

29 37 42 43 45

47 49 50 51 vii

viii

Illustrations

15. 16.

17. 18. 19. 20. 21. 22. 23. 24. 25.

26. 27. 28. 29. 30.

31.

32.

33.

34.

Resafa, Basilica of the Holy Cross, plan with location of graffiti wall and martyrium indicated. Carved stone plaque set in the east fac¸ade of the Great Mosque of Sfax, recording the restoration of the mosque in 378 H/988 CE. Carved marble slab to the left of the mihrab in the Great Mosque of Kairouan. Kairouan, Great Mosque, Qubbat al-Bahu, interior. Sousse, Great Mosque, replacement inscription over portal. Mahdia, Great Mosque, portal showing empty entablature. Inscribed marble plaque found in the Zawiyat al-Gharyaniyya at Kairouan. Raqqada, Museum of Islamic Art. Kumurdo cathedral ( Javakheti, Georgia). East fac¸ade, 964. Kumurdo cathedral ( Javakheti, Georgia). South porch, main tympanum. Inscriptions nos. 1 & 2, 964. Kumurdo cathedral ( Javakheti, Georgia). South porch, main tympanum. Inscriptions nos. 1 & 2, 964. Ishkhani cathedral (Tao-Klarjeti, now in Turkey). Oratory chapel, north door. Inscription of Gurgen, king of Kartli, 994–1008. Parkhali cathedral (Tao-Klarjeti; now in Turkey). South fac¸ade, inscription at arcade level, c.970. Kumurdo cathedral ( Javakheti, Georgia). South porch, Festival inscriptions, 964 to early eleventh century. Kumurdo cathedral ( Javakheti, Georgia). South porch, Festival inscriptions, 964 to early eleventh century. Icon of St Symeon the younger commissioned by Antoni, bishop of Ishkhani, c.1015. View of the south wall of the east end of the north church showing cloisonn´e bricks with pseudo-Arabic patterns, Monastery of Hosios Loukas, Byzantine, ca. 961, Phokis, Greece. Detail of the stringcourse in the exterior apse wall of the north church showing carved pseudo-Arabic patterns, Monastery of Hosios Loukas, Byzantine, marble, ca. 961, Phokis, Greece. Reconstruction drawing of the templon screen of the tenth-century north church, Monastery of Hosios Loukas, showing pseudo-Arabic marble carving. General view of the bema and apse of the crypt in the katholikon (south church), Monastery of Hosios Loukas, first half of the eleventh century. Detail of the pseudo-Arabic pattern in the north impost block flanking the bema of the crypt in the katholikon (south church), Monastery of Hosios Loukas, fresco, 1040s.

53

63 65 67 69 70 71 77 80 81

83 86 90 91 93

103

105

107

111

113

Illustrations

35.

36.

37. 38.

39.

40.

41.

42.

43.

44.

45. 46. 47. 48. 49.

Presentation of Christ at the Temple, southwest squinch of the katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. Hagios Demetrios, soffit of the arch flanking the Presentation scene to the south, katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. Detail of Fig. 36 showing the pseudo-Arabic pattern on the shield held by Hagios Demetrios. Hagios Prokopios, soffit of the arch flanking the Presentation scene to the north, katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. Palermo, Palazzo Reale, Cappella Palatina. The painted wooden ceilings of the nave and the two aisles, seen from below. Palermo, Galleria Nazionale della Sicilia, inventory nos. 5104 & 5105. The two longest of the three fragments of Arabic verse inscriptions in opus sectile from the Cappella Palatina. Palermo, Palazzo Reale, Cappella Palatina. Detail of a stellate coffer (south side, third from west) from the central zone of the painted wooden ceiling of the nave. Palermo, Palazzo Reale, Cappella Palatina. Detail of panel at the base of the first large unit of the muqarnas zone from the west end on the north side of the ceiling of the nave, showing a male half-figure holding two crosses, above an Arabic inscription. Palermo, Palazzo Reale, Cappella Palatina. Details of the Arabic inscription hidden in the intrados of the deep little niche in the centre of the fourth small unit from the west end of the north side of the muqarnas zone: wa-l-yumn wa-l-kif¯aya wa-l-‘izz wa-l- . . . (‘bliss and capability and power and . . . ’). Palermo, Palazzo Reale, Cappella Palatina. Detail of a stellate coffer (south side, third from west) from the central zone of the painted wooden ceiling of the nave, showing (below) a pseudo-inscription composed of the letter s¯ın attached to an intertwined l¯am-alif (or alif-l¯am), followed by a reversed letter s¯ın. The portal of the Qaratay Madrasa at the beginning of the twentieth century. Qaratay Madrasa ı¯w¯an, general view. Qaratay Madrasa, general view of courtyard wall. Qaratay Madrasa, southwestern ‘Turkish triangle’. Qaratay Madrasa, diagram showing the arrangement of the prophetic and caliphal names in the segments of the Turkish triangles.

ix

115

116 117

118

126

127

127

133

139

141 151 155 157 158

159

x

Illustrations

50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.

63.

64.

65. 66. 67. 68.

69.

70.

Qaratay Madrasa detail of the inscription at the base of the dome. Arabic and Hebrew epitaphs from the tomb of Fernando III in Seville cathedral, between 1252 and 1284. Latin and Castilian epitaphs from the tomb of Fernando III in Seville cathedral, between 1252 and 1284. Reconstruction of the converted mosque-cathedral at Seville. Seal of the city of Seville, 1311 (left). Eighteenth-century drawing in Seville cathedral archive, MS. 57-3-40 (fol. 63r.). Marmaˇse¯ n, south fac¸ade, detail, founder’s inscription. Noravank‘, western fac¸ade of the gawit‘. Noravank‘, western fac¸ade, lower tympanum, with the Virgin and the Child. Noravank‘, western fac¸ade, lower tympanum, with the Virgin and the Child. Noravank‘, burial church of Prince Bułtel, inscription. Noravank‘, western fac¸ade, upper tympanum as seen at the door of the gawit‘. Genoa, Cathedral of San Lorenzo. View of the central nave. Genoa, San Lorenzo. View of north wall with dedicatory and foundation inscriptions and bust portrait of Janus, c. 1307. Genoa, Cathedral of San Lorenzo. North wall, gallery, detail of the foundation inscription and sculpted effigy of Janus, c. 1307. Genoa, Cathedral of San Lorenzo. South wall, detail of the junction between dedicatory and foundation inscription, c. 1312. Genoa, Cathedral of San Lorenzo. South wall, central section of the foundation inscription, c. 1312. Genoa, Cathedral of San Lorenzo. South wall, western section of the foundation inscription, c. 1312. Signature of Khalaf between the hinges of the cylindrical box in the Hispanic Society of America D7532, c. 965. Signature of Badr and Tarif on the underside of the clasp on the Gerona Casket, 976. Treasury of the Cathedral of Gerona. Signature of Misbah under the throne on the front of the Pamplona Casket, 1004–5. Museo de Navarra, inv. no. 1360-B. Signature of al-ʿAziz incised on the rear jamb of the lower left frame of the minbar made in Cordoba in 1137 for the Almoravid Mosque in Marrakesh.

161 171 172 173 175 189 191 193 195 197 199 207

208

208

209 210 211 232

233

235

237

Illustrations

71.

72.

Signature of Bichitr on the footstool beneath the enthroned Jahangir presenting a book to a Sufi, c. 1615–20. Freer Gallery 1942.15a. Signature of ʿAli ibn Muhammad al-Zammaki on a glass lamp made c. 1330 for Sayf al-Din Qawsun. Metropolitan Museum of Art 17.190.991.

xi

240

241

CONTRIBUTORS 鵼

Sheila S. Blair Norma Jean Calderwood University Professor of Islamic and Asian Art, Boston College; Hamad bin Khalifa Endowed Chair of Islamic Art, Virginia Commonwealth University, Richmond, VA. Sheila S. Blair is a historian of Islamic art who investigates both writing and illustration, especially the interaction between the two. Her latest book is Text and Image in Medieval Persian Art (Edinburgh, 2014). Among many other projects, she is currently preparing a chapter on the arts of the Mongol period and a paper on the reasons for illustrating historical manuscripts made during that period. Jonathan M. Bloom Norma Jean Calderwood University Professor of Islamic and Asian Art, Boston College; Hamad bin Khalifa Endowed Chair of Islamic Art, Virginia Commonwealth University, Richmond, VA. Jonathan M. Bloom has written on many aspects of Islamic art and architecture, but is particularly interested in the Mediterranean region, especially Islamic North Africa and Spain, in the medieval period. His most recent books include The Minaret (Edinburgh, 2013) and, with Sheila Blair as co-editor, God Is Beautiful and Loves Beauty (New Haven and London, 2013). He is currently working on a book about Islamic architecture in North Africa and Spain for Yale University Press. Matthew P. Canepa Associate Professor of Art History at the University of Minnesota–Twin Cities where he is a faculty member in the departments of Art History and Classical and Near Eastern Studies. Matthew P. Canepa earned his Ph.D. from the University of Chicago. His research focuses on the intersection of art, ritual, landscape and power in the eastern Mediterranean, Persia and the wider Iranian world. His publications include The Two Eyes of the Earth: Art and Ritual of Kingship between Rome and Sasanian Iran (Berkeley, 2009), which was awarded the James Henry Breasted Prize from the American Historical Association for the best book in English in any field of history prior to 1000 CE, and the edited volume, Theorizing Cross-Cultural Interaction among the Ancient and Early Medieval Mediterranean, Near East and Asia (Ars Orientalis 38, 2010). Antony Eastmond AG Leventis Reader in the History of Byzantine Art, Courtauld Institute of Art, University of London. Antony Eastmond has written xiii

xiv

Contributors

extensively on the art and culture of medieval Georgia and its relations with Byzantium. He also works on Byzantine ivories. He is the author of The Glory of Byzantium and Early Christendom (2013), as well as Art and Identity in ThirteenthCentury Byzantium: Hagia Sophia and the Empire of Trebizond (2008) and Royal Imagery in Medieval Georgia (1998). He has published articles in the Art Bulletin, Art History, Dumbarton Oaks Papers and Speculum. He recently held a Leverhulme Major Research Fellowship during which he worked on a study of cultural interaction in eastern Anatolia on the eve of the Mongol invasions. Stefania Gerevini Assistant Director at the British School at Rome. Stefania Gerevini’s current research focuses on the appropriation and meaning of Byzantine visual language in late medieval Italy, particularly Genoa and Venice, and on the conceptualizations and artistic applications of light and transparency in medieval art. Her recent publications include ‘The Grotto of the Virgin: Artistic Reuse and Cultural Identity in Medieval Venice’ (Gesta, 2014) and Christus Crystallus: Rock Crystal, Theology and Materiality in the Medieval West (British Museum Press, 2014). Jeremy Johns Professor of the Art and Archaeology of the Islamic Mediterranean; Director, Khalili Research Centre for the Art and Material Culture of the Middle East, University of Oxford. Jeremy Johns is principally interested in relations between Muslim and Christian societies in the medieval Mediterranean as manifested in material and visual culture. His research has focused upon the archaeology of the transition from late antiquity to early Islam in the Levant and, especially, upon the archaeology, history and art history of Sicily under Islamic and Norman rule, from the Muslims’ conquest of the island in the ninth century to the destruction of the Islamic community of Sicily by Frederick II. His recent and forthcoming publications include the first comprehensive study of the Islamic painted ceilings of the Capella Palatina in Palermo, as well as editions and studies of Arabic and bilingual documents from Norman Sicily. He is currently engaged in a multidisciplinary study of the medieval Islamic rock crystal industry. Tom Nickson Lecturer in Medieval Art and Architecture, Courtauld Institute of Art, University of London. Tom Nickson’s recent research has focused on encounters of Christian and Muslim texts and traditions in medieval Iberia, and their historiography. His book, Toledo Cathedral: Building Histories in Medieval Castile will be published in 2016. Ioanna Rapti Director of Studies in the Department of Religious Studies in the History of Art and Architecture of the Byzantine World and the Christian Orient, ´ ´ Ecole pratique des Hautes Etudes, Paris. Ioanna Rapti’s research interests include Armenian iconography and court culture, historical geography of Armenian Cilicia, text and image, history of Byzantine art history and photoarchives.

Contributors

Scott Redford Nasser D. Khalili Chair in the Art and Archaeology of Islam, Department of the History of Art & Archaeology, School of Oriental and African Studies, University of London. Scott Redford’s most recent book Legends of Authority: The 1215 Seljuk Inscriptions of Sinop Citadel, Turkey was published in 2014 by Koc¸ University Press in Istanbul. Alicia Walker Assistant Professor, Bryn Mawr College. Alicia Walker’s research focuses on cross-cultural artistic interaction between the Byzantine and medieval Islamic worlds and gender issues in Byzantine art and material culture. Her work has appeared in Art Bulletin, Gesta, Medieval History Journal, Studies in Iconography, Ars Orientalis and Muqarnas. Her first book, The Emperor and the World: Exotic Elements and the Imaging of Middle Byzantine Imperial Power, was published by Cambridge University Press in 2012. Ann Marie Yasin Associate Professor of Classics and Art History, University of Southern California. Ann Marie Yasin’s first book, Saints and Church Spaces in the Late Antique Mediterranean: Architecture, Cult, and Community, appeared in 2009. Her current research examines evidence for the adaptation and renovation of early Christian churches to understand how alterations to the material fabrics of sacred structures shaped communal history and devotional practices over time.

xv

ACKNOWLEDGEMENTS 鵼

This volume has evolved from a series of meetings of a research network entitled Viewing Texts: Word as Image and Ornament in Medieval Inscriptions. The meetings, which were held between 2008 and 2010, were organised under the auspices of the Beyond Text programme that was funded by the Arts and Humanities Research Council, under the direction of Evelyn Welch.1 The project was jointly organised with Prof Liz James of the University of Sussex, whose aid and support throughout were invaluable. The network meetings, the website and bibliography (http://projects.beyondtext.ac.uk/wordasimage/index.php) were set up and administered by Stefania Gerevini with help from Eleni Dimitriadou and Laura Veneskey, to whom I am very grateful. The meetings were facilitated by Cynthia de Souza and Ingrid Guiot, staff members of the Research Forum at the Courtauld Institute. The other members of the network included Timothy Greenwood, Bernard O’Kane, Arietta Papaconstantinou, Charlotte Rouech´e and Avinoam Shalem. Although they are not represented in this volume by essays, they are represented in spirit: their contributions to the discussions at the various meetings were invaluable in the formation of many of the ideas presented in this volume. I greatly appreciate all that they gave to the meetings, and hope that when they see this book, they will understand my gratitude to them.

1

www.beyondtext.ac.uk

xvii

鵻 INTRODUCTION

VIEWING INSCRIPTIONS 鵼 Antony Eastmond

WRITING AS ART

Inscriptions communicate. Whether they are royal proclamations, pious prayers, wise sayings, historical accounts or simple records of names, inscriptions were considered important by those who made and read them, and they have become crucial historical tools to modern historians. Yet, like all means of communication between humans, the information they contain is coded not just in the words that constitute the text but also in a wide variety of nonverbal forms.1 In speaking, these nonverbal forms include the kinesic messages conveyed by gesture, posture, facial expression and movement. Inscriptions, whether inscribed on a monumental scale on the side of a mountain or carved in letters just millimetres high on an ivory casket, similarly rely on nonverbal elements – choices of script, scale, location, spatial organisation, letter style, clarity and legibility – for much of their meaning. The chapters in this volume all address this nonverbal visual evidence, the other information embedded in inscriptions. They consider writing as art, not simply as an art form (the more traditional and well-established study of calligraphy, literally ‘beautiful writing’ from the Greek κάλλος and γραφή). Rather, they see inscriptions as important constituents of wider visual environments. All focus on inscriptions, whether painted, carved or formed from tiles and bricks: texts inscribed on the floors, walls and ceilings of buildings; chiselled on a monumental scale into the bare rock of cliff faces; or placed, barely visibly, on ivory boxes and glass lamps. They vary from royal proclamations, set up sometimes in three or four languages simultaneously and inscribed with great care and precision or with great effort in difficult-to-reach places, to graffiti hastily and illicitly 1

2

Antony Eastmond

scratched into walls. These words are not simply representations of disembodied utterances. Once created, they become physical objects whose materiality is an essential element in the means by which they convey meaning. This volume deliberately brings together scholars who work on a broad range of periods and fields that stretch all the way around and across the Mediterranean, from Spain to the Caucasus on the north side, from Tunisia to Egypt and Syria on the south, and with Sicily at its heart. Beyond the Mediterranean, they look to the Iranian world, Central Asia and the Indian subcontinent. The chapters concentrate on the late antique and medieval worlds, but they look back to the fifth century BCE and forward to the seventeenth century CE. In the Christian world they move from late antique Rome to Byzantine Greece and Georgia around the year 1000. They examine Sicily in the twelfth century, Spain in the thirteenth, and Italy and Armenia in the fourteenth century. In the Islamic world the chapters range across the Umayyad, Aghlabid, Fatimid, Seljuk, Ottoman, Safavid and Mughal realms; in Chapter 1 Matthew Canepa traces Sasanian and early Islamic practices in Iran to the epigraphic cultures of earlier Persian societies, the Achaemenids and Seleucids. Despite this extraordinary diversity in period and place, all the chapters are united by their interest in the ways in which groups in societies exploited the presence of writing to convey additional meanings beyond their verbal content. The aim of this chronological and geographic breadth is to encourage conversations across modern disciplinary, regional and period boundaries in academia and to explore the common uses of writing, as well as the particular differences employed by different cultures at different times. Inscriptions have tended to be treated as collections of words, whose materiality is incidental. Such assumptions underlie the origins of the great corpora of inscriptions, which were often motivated by positivist concerns about the factual content that could be gleaned by reading such texts. The R´epertoire chronologique d’´epigraphie arabe (RCEA), the great eighteen-volume corpus of Arabic historical inscriptions, for example, is completely unillustrated and only provides modern transcriptions and translations of all its Arabic texts.2 With no interest in the appearance of any text, it is impossible to glean any of the kinds of information that the chapters in this volume seek to exploit. Even those corpora that do include photographs tend to use them primarily as a means to corroborate the editors’ reading of the text itself, rather than as a source of additional information.3 The essential premise of this book is that inscriptions are not just disembodied words that can be studied in isolation. Instead they must be considered as material entities, whose meaning is determined as much by their physical qualities as by their contents. None of the chapters seeks to deny the importance of reading inscriptions. Indeed the contents remain important and are central to understanding the ways in which they have been set up and used. However, in addition to their contents, the ways in which words were presented to onlookers is a key source of information and a generator of meaning that should not be ignored.

Introduction

Meaning can be generated simply by the formal qualities of inscriptions: the shape and arrangement of the script used, the size of letters, the legibility and readability of the inscription. Scripts can deliberately look to the distant past, as with the use of ancient kufic in the fourteenth-century Sultan Hasan Complex in Cairo,4 or the revival of Roman script in the fac¸ade inscription of Salerno cathedral in the eleventh century.5 Equally, as Stefania Gerevini shows in Chapter 10, self-consciously modern scripts can be employed, even when trying to invent a long and venerable history. Meaning is further developed by the relationship between the texts and their physical contexts. The layout and sequencing of texts can affect how viewers interact with the buildings or landscapes in which they are located. The visual qualities of texts, the ways in which they wrap around buildings or cluster in particular places, can give them agency to encourage ritual or other interactions between readers/viewers and the texts or monuments. As Chapter 1 shows, readability is a central feature of Achaemenid Persian inscriptions. Placed high on cliff faces, they could never have been read without some form of crib or accompanying oral performance. They were clearly meant only to be seen. Their illegibility and their permanence in the landscape made them expressions of power far more potent than their often banal contents would suggest. Similar arguments have been proposed for other inscriptions in other cultures that were also placed beyond the sight or reach of most viewers.6 The apparent legibility of a text could be undercut in other ways. Inscriptions in Christian churches in Byzantium,7 Syria8 and Armenia9 are all recorded as being written in clear, legible scripts, but the words are represented in cipher. These inscriptions thus present themselves as script, but can only be read by those initiated into the code. Equally, the monograms employed in many of the same churches clearly relate to writing, but the ways in which letters were combined meant that they were (and often remain) impossible to decode. The writing conveys meaning, but at the same time its presentation is deliberately unhelpful to the reader.10 The visual qualities of inscriptions allow them to play with their onlookers, who can be manipulated to act as the viewer of an inscription rather than its reader. Texts can be presented both to encourage and discourage reading, shifting emphasis between the textual content and the formal appearance of the inscription. These strategies of presentation dictate the ease and the speed of reading and can require onlookers to dedicate long periods of study in order to access their contents. They can both facilitate and obfuscate access to their contents and verbal meaning. The different strategies employed all help us better understand the reason the writing was put there in the first place.11

READING AND VIEWING

The question of who read inscriptions is an important one.12 The ability to read is a multifaceted problem that concerns not just the literacy and education of the

3

4

Antony Eastmond

reader but also many features of the text.13 Is the inscription legible? Is it readable? The first question concerns the form of the writing, the ways in which the letters are presented and how they fit into the text as a whole: are the individual letters distinct? Are individual words presented clearly, with or without abbreviations and other sigla? How do the words form into sentences: in a linear form with word breaks or in a more complex order with words run together or awkwardly split across lines? Readability is affected not only by these same issues but also by other matters such as the size of the lettering and its placement: can it actually be seen by the naked eye? These points all affect the ways in which the text can be read. As Paul Saenger has noted, texts that do not leave spaces between words force their readers to approach them differently.14 They require greater concentration and can only be read more slowly. This changes the readers’ relationship both to the inscription and to the object or monument on which it is inscribed, because it now must be studied and considered over a longer period of time in order to read the text fully. Indeed we might wonder whether the words written on objects were always meant to be read and, if so, by whom. Did patrons expect that the texts that they commissioned to be placed on works of art would (or could) be read by everyone who saw them? How were the words perceived by illiterate viewers? In what ways might they have been seen when they are considered simply in terms of their visual appearance and effect, with no thought to their verbal meaning, when a formalist reading is applied to them? How do they function as ornament? The ability to exclude readers becomes a way of restricting access to their contents, which confers a new status on the text as a bearer of secret and therefore potentially more powerful meanings.15 For those inscriptions that cannot easily be seen, the question arises of the relationship between the text and other means by which its meaning might be transmitted. How do the inscriptions relate to the spoken word, to oral recitations of the text from memory, or to other written versions kept elsewhere?16 Does the inaccessibility of the text change its meaning? Were inscriptions, like Victorian children, to be seen but not heard? Or should we assume that there were alternative means to access these texts, in which case the inscription becomes merely a visual cue or mnemonic? The importance of memory in the premodern age and the repetitious training by which monks and mullahs (among others) learned the Bible or Qur‘an by heart would allow words or phrases to cue such viewers to recite much longer passages. When considering the relationship between the inscription and the reader, literacy is clearly a central issue. However, literacy is not an absolute: it is a sliding scale that runs between full literacy – the ability to read, write and compose – and various forms of functional literacy.17 Functional literacy can vary between signature literacy (the ability just to read and sign your name); differing degrees of ability to recognise particular letters, words or phrases; or the ability to read and write in one particular language.18 Each person, therefore, brings a different

Introduction

degree of knowledge and experience to bear on inscriptions and, as a result, will take different things from them. And the further we move down the scale of literacy the greater the importance of viewing over reading. The less you can read, the greater the shift from reading to viewing and the more the writing becomes a visual symbol rather than a means of verbal communication. In these instances the context of the inscription takes on more importance than the contents. But as some of the chapters in this volume show, artists could play on the blurred line between reading and viewing. Inscriptions can be placed in such obscure places or made on such a small scale that reading must be preceded by an act of intense, detailed viewing. Equally, multilingual inscriptions, as discussed by Tom Nickson (Chapter 8) and Matthew Canepa (Chapter 1), were rarely intended to be comprehensible to all: rather each community had access to different versions of the text, and meanings arose from the varying nature of the contents of each language’s inscription – from the order and juxtaposition of the languages and from the ways in which the placement of the different texts could evoke notions of community or hierarchy and authority. In other cases artists exploited the overlap between writing and ornament. In Chapter 7 Scott Redford discusses the different registers of writing and ornament in the Qaratay Madrasa in Konya, which are predicated on a literate viewer, but it is clear that not all the inscriptions are meant simply to be read. Inscriptions are disguised as ornament, and ornament is presented as inscription. The aesthetics of writing are very much to the fore here. A parallel case is presented by Ioanna Rapti in Chapter 9 of a fourteenth-century Armenian inscription in which the letters are barely distinguishable from the ornament that surrounds them. All require detailed scrutiny to distinguish one from the other. In Chapter 6 Jeremy Johns finds a more perplexing overlap of real and pseudo-writing in the twelfthcentury Cappella Palatina in Palermo. The Cappella Palatina inscriptions demonstrate also the importance of placement. This is also a central concern of Chapter 11, in which Sheila Blair considers the placement of craftsmen’s signatures on a range of Islamic works of art that could be examined in the most intimate of circumstances. Here the location of the signature conveys more meaning than the words themselves, particularly about the relationship between the craftsman and the patron. It shows the playfulness of medieval writing.

INTERACTIONS AND CONVERSATIONS

In recent years the study of art and text has become the focus of increasing attention. Volumes devoted to ‘Art and Text’ have explored the relationship between words and images in Greek, Roman and Byzantine culture; analyses of writing and inscriptions have long been central to the study of Islamic art.19 These studies consider not just the relationship of words to images but also the idea of text

5

6

Antony Eastmond

as art.20 This book does not seek to reproduce that work; rather it is interested in the legacy of those Greco-Roman epigraphic habits on the successor states around the Mediterranean and their impact on their neighbours in the larger Iranian world.21 Much of the existing work in this area has been conducted within disciplinary boundaries, divided between scholars of the Latin West or of the Greek East, or beyond that of the Christian Caucasus; between Islamic scholars of the western or eastern shores of the Mediterranean; between those working on antiquity and late antiquity and those working on the later middle ages. The individual case studies presented here are designed to draw out some of the shared ways in which all these different societies used scripts, the degree of interchange and overlap between them, as well as the distinctive traditions and uses of words in each society. Inscriptions engage in conversations with other inscriptions across space and time. Later texts react both visually and verbally to earlier inscriptions, and they can evoke or echo inscriptions being set up at the same time but in different places or by different cultures. This volume is intended to provoke and further that conversation, which takes many forms. Matthew Canepa in Chapter 1 shows the importance of Hellenistic traditions in the Persian world in the aftermath of the conquests of Alexander the Great and the ways in which these traditions then interacted with the indigenous epigraphic culture over the next millennium. Equally, Jonathan Bloom in Chapter 3 proposes that the Aghlabid inscriptions of North Africa can be traced to the lasting legacy of the surviving inscriptions of the Roman Empire in the region.22 These conversations took place over time; but there is equally much evidence of similar conversations taking place between contemporaneous neighbouring cultures. Ioanna Rapti in Chapter 9 argues that the inscriptions at Amał u Noravank‘ in Armenia are heavily indebted to the Islamic culture of inscriptions, now transposed to a Christian monument. The recurrence of pseudo-writing in Byzantium, Sicily and Seljuk Anatolia points to another common use of the appearance of writing in the medieval world – although in this case the three chapters that examine the phenomenon, by Alicia Walker, Jeremy Johns and Scott Redford (Chapters 5, 6 and 7 respectively), propose different meanings and functions for this imitation of writing. It is clear that inscriptions, whether formal or informal, do not exist in splendid isolation, but engage in conversations with other inscriptions. Few of the inscriptions discussed in this book stand alone. Rather they are parts of more extensive arrangements of writing, and they must be seen as such. Whether in monasteries or madrasas the layout of texts on walls requires readers to look from one text to another, and the arrangement produces meanings. The conversations take place both over space, between walls, between mountains, between sides of an object – but also across time. Cumulative inscriptions respond to what was already there – answering, developing, altering the context and the meaning of the older text. They mark changes in power. This is true both for formal and informal texts. Ann Marie Yasin (Chapter 2) and Antony Eastmond (Chapter 4)

Introduction

both demonstrate the ways in which inscriptions attract others and the way in which the chorus of voices that emerge gains resonance from the multitude of writing. Conversations also happen between languages, as the examples of multilingual inscriptions in this volume show.23

MEMORY AND PERFORMATIVITY

When read, inscriptions serve as prompts to memory, never less so than when modern scholars use them to reconstruct people or events otherwise forgotten by history. This function was clearly always central to their existence. However, memories could be encoded in other aspects of an inscription’s appearance. Alicia Walker, in Chapter 5 on the pseudo-writing in the Byzantine monastery at Hosios Loukas, argues that writing could embed and evoke memories through form. This is an idea that recurs in Stefania Gerevini’s Chapter 10 about the fourteenthcentury inscriptions in Genoa cathedral. She shows the way in which the past (however fictitious) could become a tangible and apparently permanent part of the present. However, as Jonathan Bloom shows in Chapter 3, memory could be eradicated, and the consequences of this loss should not be ignored by contemporary historians. Whether inscriptions are totally eradicated or only selectively erased (as at Sfax), the idea that we are meant to remember to forget what is destroyed is apparent from the empty spaces left behind. The absence of a text is as important as its presence. This form of damnatio memoriae has a long history that looks back to the Roman and before that the Egyptian world.24 The contents and design of texts also raise questions about their performative aspects: the layout, arrangement and re-arrangement or reuse of texts affect how buildings are viewed, as well as the ways in which texts institutionalise and affect communities’ memories in public spaces. The linear nature of most Christian and Islamic writing requires readers to approach it in particular ways, especially when presented in monumental form. These texts thus lead viewers through monuments and objects in particular ways. In Chapter 7 Scott Redford evokes this successive layering of texts in the Qaratay Madrasa – the ways in which different types of text, presented in different scripts, led students from the portal to the interior, from the ground up to heaven, from consideration of moral behaviours on earth to the celebration of divine mysteries in heaven. There is thus a logic, and an order, to looking at these monuments and objects, determined by their inscriptions. However, it can easily be misunderstood by those who do not understand the writing system involved. This is evident in Ernst Kuhnel’s cata¨ logue of Islamic ivories published posthumously in 1971.25 His editors understood that the inscriptions helped organise the appearance of the objects, and so they laid out multiple photographs of each object to lead the viewer round them, starting from the beginning of the main inscription. However, they rotated each object clockwise, whereas the Arabic inscriptions actually require the objects to

7

8

Antony Eastmond

be rotated in the opposite direction; each ivory is thus effectively presented in reverse. In Chapter 5 Alicia Walker argues that even pseudo-writing could have a performative element as viewers’ expectations of the linearity of writing encouraged them to circumambulate the church at Hosios Loukas. The eleven essays gathered in this book bring together all these themes, but their case studies are divided across a wide chronological and geographical range. They show the common ways in which writing could be used to inform but also manipulate viewers, and the ways in which the appearance of writing created its own meanings above and beyond the content of the texts themselves. The essays also reveal the culturally specific ways in which different societies exploited inscriptions, depending on the form of alphabets and writing systems employed and their possibilities and limitations. The expectations of viewers and readers about inscriptions also varied. The types of text they believed that they would encounter varied, as did the degree of authority, power or spirituality with which they imbued those texts. These essays show the possibilities for new ways of thinking about how inscriptions can be studied, revealing much about the ways in which different cultures appropriated, adapted and developed the epigraphic habits of their neighbours and predecessors. They show the power of the nonverbal qualities of the written word. NOTES

1. For an introduction see Peter A. Andersen, Nonverbal communication: Forms and functions, 2nd edition (Long Grove, IL, 2007). 2. R´epertoire chronologique d’´epigraphie arabe 18 ´ vols, eds. Etienne Combe, Jean Sauvaget, Gaston Wiet (Cairo, 1931–91). The new online version of the corpus now includes photographs where available: www.epigraphieislamique.org. 3. For example, A. A. Khachatrian, Korpus arabskikh nadpisei Armenii (Erevan, 1987); I. A. Orbeli, Corpus Inscriptionum Armenicarum, 8 vols. (Erevan, 1966–). 4. Doris Behrens-Abouseif, Cairo of the Mamluks: A history of the architecture and its culture (London, 2007), 211; Abdallah Kahil, The Sultan Hasan Complex in Cairo, 1357– 1364: A case study in the formation of Mamluk style, Beiruter Texte und Studien 98 (Beirut, Wurzburg, 2008). ¨ 5. Armando Petrucci, Public lettering: Script, power, and culture (Chicago and London, 1993), originally published as La Scrittura: Ideologia e rappresentazione (Turin, 1980). 6. Liz James, ‘“And shall these mute stones speak?” Text as art’, in Art and text in Byzantine culture, ed. Liz James (Cambridge, 2007), 188–206.

7. Catherine Jolivet-L´evy and Nicole Lemaigre Demesnil, ‘Saint-Serge de Matian`e, son d´ecor sculpt´e et ses inscriptions’, Travaux et Memoires 15 (2005), 67–84. 8. William K. Prentice, “Magical formulae on lintels of the Christian period in Syria,” American Journal of Archaeology 10.2 (1906), 137–50. 9. Tim Greenwood, “A corpus of early medieval armenian inscriptions,” Dumbarton Oaks Papers 58 (2004), 27–91, at 82–3, no.6. 10. On these ideas see Antony Eastmond, “Monograms and the art of unhelpful writing in Late Antiquity,” in Sign and design, eds. Jeffrey Hamburger and Brigitte M. BedosRezak (Washington DC, forthcoming). 11. Oleg Grabar, ‘Graffiti or proclamations: Why write on buildings?’, in The Cairo heritage, ed. D. Behrens-Abouseif (Cairo, 2006), 69–75 reprinted in his Islamic Art and Beyond [Constructing the Study of Islamic Art: 3] (Ashgate: Variorum reprints, 2006), Study XVII, 239–44. 12. Liz James, ‘“And shall these mute stones speak?” Text as art’. 13. Michael Camille, ‘Seeing and reading: Some visual implications of medieval literacy and illiteracy’, Art History 8 (1985), 26–49.

Introduction

14. Paul Saenger, Space between words: The origins of silent reading (Stanford, 1997). 15. John Baines, ‘The earliest Egyptian writing: Development, context, purpose,’ in The first writing. Script invention as history and process, ed. S. Houston (Cambridge, 2004). 16. Mary Beard, ‘The function of the written word in Roman religion’, in Literacy in the Roman world, Journal of Roman Archaeology. Supplementary Series 3, eds. Mary Beard, A. K. Bowman, M. Corbier, et al. (Ann Arbor, MI, 1991), 35–58. 17. William V. Harris, Ancient literacy (Cambridge, MA and London, 1989); and the discussion of it in ‘Literacy in the Roman world’, Journal of Roman Archaeology. Supplementary Series 3, eds. Mary Beard, A. K. Bowman, M. Corbier, et al. (Ann Arbor, MI, 1991). 18. Herbert C. Youtie, ‘Because they do not know ¨ Papyrologie und Epiletters’, Zeitschrift fur graphik 19 (1975), 101–8. 19. Simon Goldhill and Robin Osborne eds., Art and text in ancient Greek culture (Cambridge, 1994); Jas Elsner, ed. Art and text in Roman culture (Cambridge, 1996); Liz James ed. Art and text in Byzantine culture (Cambridge, 2007); Sheila S. Blair, Text and image in medieval Persian art (Edinburgh, 2013). Also see Sheila S. Blair, Islamic inscriptions (Edinburgh, 1998); Irene A. Bierman, Writing signs: The Fatimid public text (Berkeley CA, 1998).

9

20. In Byzantine studies, this is most evident in Liz James, ‘“And shall these mute stones speak?” Text as art’; Amy Papalexandrou, ‘Text in context: Eloquent monuments and the Byzantine beholder’, Word & Image 17.3 (2001), 259–83. 21. Ramsay Macmullen, ‘The epigraphic habit in the Roman Empire’, American Journal of Philology 103.3 (1982), 233–46; Elizabeth A. Meyer, ‘Explaining the epigraphic habit in the Roman Empire: The evidence of epitaphs’, Journal of Roman Studies 80 (1990), 74– 97. 22. For the equivalent legacy in medieval Italy see Petrucci, Public lettering: Script, power, and culture. 23. See also Antony Eastmond, ‘Inscriptions and authority in Ani’, in Der Doppeladler – Byzanz und die Seldschuken in Anatolien vom sp¨aten 11. bis 13. Jahrhundert, eds. Neslihan AsutayEffenberger, F. Daim (Mainz, in press). 24. Greg Woolf, ‘Monumental writing and the expansion of Roman society in the early empire’, Journal of Roman Studies 86 (1996), 22–39. For the erasure of Akhenaten, see B. G. Trigger, B. J. Kemp, D. O’Connor and A. B. Lloyd, Ancient Egypt: A social history (Cambridge, 1983), 186–7. Die islamischen Elfenbein25. Ernst Kuhnel, ¨ skulpturen, VIII.-XIII. Jahrhundert (Berlin, 1971). I owe this observation to Sheila Blair.

鵻 CHAPTER ONE

INSCRIPTIONS, ROYAL SPACES AND IRANIAN IDENTITY: EPIGRAPHIC PRACTICES IN PERSIA AND THE ANCIENT IRANIAN WORLD 鵼 Matthew P. Canepa

I am the Mazda-worshipping lord Shabuhr, king of kings of Iran and NonIran, of divine nature, son of the Mazda-worshipping lord Ardaxshir, king of kings of Iran, of divine nature, grandson of the lord Pabag the king. I am the sovereign of the Kingdom of Iran. I hold the following lands: Persia . . . Shabuhr I (239/40–270/2 CE), Kaʿba-ye Zardosht inscription from Naqsh-e Rostam, Iran, §1-2. I am Darius the great king, king of kings, king in Persia, king of lands of all races, son of Vishtaspa, of the line of Achaemenes, a Persian, son of a Persian, an Iranian of Iranian lineage. Darius, king of kings, proclaims: by the will of Ahura Mazda these are the lands I held outside of Persia . . . . Darius I (522–486 BCE), tomb inscription from Naqsh-e Rostam, Iran, (DB §2-3).

From the height of the Achaemenid Empire in the early fifth century BCE to the fall of the Sasanian Empire in the mid-seventh century CE, inscriptions played an important role in the development and expression of kingship in Persia and the ancient Iranian world. As with many aspects of the long history of Iranian kingship, stunning continuities and deep ruptures mark Iranian epigraphic practices. Invasions of new peoples and the growth of new empires introduced new scripts and languages, which often displaced those of the previous regime. New visual and architectural traditions modified the ways in which patrons deployed texts and viewers experienced them. Although such cultural discontinuities often rendered the texts of fallen empires’ inscriptions incomprehensible or even alien, the inscriptions themselves continued to be powerful visual and topographical 10

Inscriptions, Royal Spaces and Iranian Identity

features of the landscapes of Persia and the wider Iranian world. Their very presence made the sites where they were carved popular and powerful, with some accumulating inscriptions from multiple dynasties as new regimes responded to their presence, if not their content. Inscriptions played an important role in articulating Iranian culture’s sense of space and became a focal point for struggles over cultural memory and Iranian royal identity. As illustrated by the parallels between the Middle and Old Persian royal inscriptions that open this chapter, the content of late antique inscriptions attests to stunning cultural continuities in Persian kingship. Both of these inscriptions were carved at the same site: the Persian necropolis and ritual centre at Naqsh-e Rostam near Persepolis. However, by late antiquity, the Old Persian script and language had long fallen out of use, and Shabuhr I and his subjects could not read and compare the content of their late antique inscriptions with the ancient predecessors, even if their contents might attest to great continuities. As explored later in this chapter, what made Shabuhr I’s statement significant was the fact that it was carved onto an ancient Persian structure located at the necropolis of the Achaemenid kings, capitalizing on the visual presence of the Achaemenid inscription nearby. The inscribed and sculpted landscape of the Iranian plateau presented a reservoir of royal precedent that transcended centuries and was arguably more important than its content for establishing connections with the past. With or without cultural and linguistic continuity, an inscription’s very visual and spatial presence presented a constant inspiration and precedent for aspiring rulers to take control of the symbolic landscape of Iran and negotiate a relationship with their predecessors and competitors.

INSCRIBING IRANIAN ROYAL IDENTITY

Along with rock reliefs and coins, inscriptions are one of the few unquestionably authentic primary sources in pre-Islamic Iranian historiography.1 Although the field has depended on them to provide the most reliable data for studies of Iranian cultural, political and art history, most scholarship on the inscriptions themselves has centred on their internal content, decipherment and dating.2 The internal textual contents of inscriptions were certainly important. However, the fact that inscriptions functioned for the majority of the individuals who saw them as visual and spatial features of the natural or built environment alludes to a more complex set of interactions among their contents and contexts. This chapter offers an approach to Iranian inscriptions that expands the focus beyond their basic vocations of preserving and disseminating a text. It presents a history of epigraphic practices in Iranian Western Asia from the rise of the Achaemenids to the coming of Islam, noting both ruptures and continuities, and examines the impact of inscriptions on their spatial, architectonic and ritual contexts. Their potential effect extended far beyond their internal contents: their very presence could

11

12

Matthew P. Canepa

shape the culture’s experience of its environment, both natural and built, as well as the activities that went on around them. Inscriptions were visual actors much like the figural rock reliefs alongside which they were often carved. In fact, many inscriptions in ancient Iran were meaningful only as graphic rather than textual signifiers, especially those looming high above on cliffs or adorning nearly inaccessible architectural features in palaces. Their physical presence and visual impact as linear, rectangular or simply textured features on the surfaces of walls, architectural members, rock faces and objects continually conveyed meaning beyond their internal content. Although the significances of inscriptions’ content (if understood) were quite closely determined, their visual meanings were not, nor were those meanings closely linked to their internal features. In fact, many of the examples considered in this chapter were not actually legible because of their physical distance from the viewer, the difficulty of their writing systems (even for the literate), or, with the passage of time, the loss of any knowledge of the inscriptions’ scripts and languages. Such episodic legibility foregrounds three important facts. First, the textual content of many inscriptions could only be understood through some process of mediation. For example, the text of the immense, Achaemenid rock-cut inscription at Bisotun was carved at such a great height on the rock face that it could not be read from below. Its royal patron, Darius I, states that he disseminated copies and translations of the inscription throughout the empire, ‘in clay and parchment’.3 An Aramaic translation survives on papyrus fragments discovered in Egypt while pieces of a basalt stele from Babylon preserve a Babylonian translation along with a reduced version of Bisotun’s accompanying sculptural relief.4 In addition, when Darius I had the Old Persian inscription prepared, he commanded that it be performatively read in his presence after it was carved into the rock.5 Although such singular royal efforts to communicate and fix the cultural understanding of the visual experience of an inscription were rare, ephemeral courtly discourse, such as acclamations, epic poetry and royal proclamations, continually enlivened and reinvigorated the significance of inscriptions in the mind of viewers (though not necessarily ‘readers’). Secondly, inscriptions were among those interrelated techniques that converted undifferentiated places into politically and socially potent space. ‘As individuals identify with an environment so their identity comes to be constructed through that environment.’6 Inscriptions were one of several techniques that could articulate and meaningfully knit the features of natural and architectonic spaces into a larger semantically and politically powerful whole. Their content or simply their formal features could link their surroundings to a wider network of royal inscriptions as experienced by the viewer. These texts were thus tangible visual links that integrated the natural and human-made features of sites, the activities that took place there and the identities of individual patrons into a wider network of cognitive associations and social agency as perceived by the

Inscriptions, Royal Spaces and Iranian Identity

patron and viewers. Unauthorized texts (graffiti) or a new imperial regime’s addition to, or disfigurement of, preexisting texts, images or structures had the power to impose on these features a reinforcing or subversive commentary. This commentary could alienate the viewer from previous understandings and experiences of spaces or structures or the activities that took place there. Depending on its location, a new inscription in the environment could divide up a viewer’s experience of the landscape by imposing new borders. Conversely, an inscription could dissolve old frontiers and impose on previously disconnected or independent landscapes or cities the marks of their subjection and integration into an empire. Taking into account inscriptions’ wider semantic network, such changes could therefore reshape the cultural understanding of the natural or built environment and individuals’ understandings of themselves in relation to it. Performativity and repeated movement through space played a central role in the process of generating and maintaining such spaces.7 Not coincidentally, inscriptions in Iranian natural or architectural contexts were closely bound up with the performance of ritualized acts. They proliferate at spaces that hosted regular ritual performances, but also appear in places that had hosted a unique event. For example, the Sasanian king of kings Shabuhr I’s monumental and smaller scale rock inscriptions were closely connected with ritual performances related to his funerary cult at Naqsh-e Rostam. In contrast, at a relatively obscure site, the king left a single inscription simply to commemorate a particularly impressive feat of archery. When subjects viewed an inscription, they reacted not only to it and its spatial contexts but also to the dynamics of the web of social agency imputed to them. Inscriptions acted as an extension of royal will and a distributed royal self.8 Even if their contents were not exactly known, inscriptions’ tangible presence extended the power and presence of the royal patron beyond the palace into the landscape. A few legible lines of an inscription at the threshold of an audience hall, common throughout the palaces of Persepolis – or even its darkened, textured surface looming overheard – could become linked with that emotional experience of joy or dread of the viewers as they moved through the space. An inscription viewed on a cliff side could have the same effect as viewing the king of kings from afar. This experience could occur only one time or be reinforced over repeated encounters. As long as the regime held power, to be caught disturbing a royal inscription would be as seditious and dangerous an act as attempting to attack the king of kings himself and could result in equally dire consequences. Taken altogether, inscriptions’ internal contents (understood immediately, through mediation, or later reinventions), visual features, and spatial, ritual and social contexts and associations all contributed to complex, overlapping networks that extended beyond an individual inscription. Thirdly, inscriptions, especially those carved into the living rock, intervened not in space, but in perceptions of time.9 They could translate ephemeral ritual performance, or even a brief movement through a site, into a permanent presence in culturally significant space. Inscriptions could tangibly collapse gulfs

13

14

Matthew P. Canepa

of time that might separate two patrons, as is most vividly illustrated by Shabuhr I’s great trilingual inscription at Naqsh-e Rostam, which the Sasanian king carved on the tower of the Kaʿba-ye Zardosht, built by Darius I more than five hundred years before. They could create an experience of coherence and continuity among vastly disparate monuments or historical actors. By making subsequent inscriptional or visual additions to the same site, patrons, in effect, imposed new meanings and contexts on old inscriptions. When new regimes from outside the Iranian plateau and the Iranian cultural sphere encroached on or conquered Iranian territory, creating or destroying inscriptions was a major technique by which they forged a meaningful relationship with Iranian royal identity and Iran itself. As they accumulated at sites, inscriptions began to articulate Iranian culture’s sense of the past; even if their contents were not understood, their silent presence attested to the long succession of kings and empires.

THE CREATION OF AN IRANIAN ROYAL EPIGRAPHIC PRACTICE IN ACHAEMENID PERSIA

Darius I (r. 522–486 BCE) of the Achaemenid dynasty was the first king of the Persian Empire (550–330 BCE) to carve monumental inscriptions in palaces and on the landscape of the empire.10 He sponsored the invention of the Old Persian cuneiform syllabary and produced the largest volume of inscriptions of the Achaemenid era. His successors cultivated and followed the repertoire of Achaemenid royal epigraphic practices that he established with little deviation. Achaemenid inscriptions appear mainly in two contexts: as monumental reliefs carved into the living rock, often in close proximity to figural relief sculpture, or incorporated into the architectural fabric of palaces, again near extensive relief decoration.11 Although hieroglyphic inscriptions featured prominently in Egyptian tombs, temples and palaces, Darius I drew most heavily on the precedents of the Achaemenids’ more geographically proximate predecessors, such as the Babylonians and Assyrians in Mesopotamia, and the Lullubi and Elamites on the Iranian plateau.12 Rock-cut inscriptions accompanied a number of figural rock reliefs carved by regional kings of non-Iranian peoples, such as the Lullubi and Elamites, as well as by Assyrian conquerors.13 The development of Achaemenid royal epigraphic practice parallels the development of their official royal art and architecture. The Achaemenids appropriated aspects of the visual and palatial traditions of empires they conquered, but did not simply replicate them unchanged. In all cases the Achaemenids innovated new traditions and modified the ones they appropriated, integrating them into a harmonious whole and subordinating them to Persian imperial ideology and tastes. Most Achaemenid inscriptions were trilingual, with Old Persian, Elamite and Babylonian Akkadian versions arranged in parallel columns or panels. Old Persian was the supreme language of power and was associated with the king

Inscriptions, Royal Spaces and Iranian Identity

1. Inscriptions and rock relief of Darius I (522–486 BCE), Bisotun, Iran. (Photo: Matthew Canepa)

of kings and the Persian aristocracy. As such, the Old Persian version was always largest and appeared in the dominant position, either at the centre or, if the three inscriptions were combined into a single panel, on the left and thus first in the order that it was read. Although the syllabary was loosely based on Akkadian cuneiform shapes, which Elamite also adapted, Old Persian appears visually distinct from both. Old Persian inscriptions could stand alone, although this was not normally the case for Elamite or Babylonian inscriptions. This hierarchy persisted even in inscriptions produced in the provinces using writing systems and languages not normally used in the heartland.14 Akkadian was a Mesopotamian prestige language, and although certain individuals in the court understood it, it was not widely used. In contrast, from very early in the empire, the Persians adopted the Elamites’ traditions of record keeping. As attested by the Persepolis Fortification Archive, the Achaemenids used the Elamite language and Elamite scribal and bureaucratic conventions as the basis for their own extensive bureaucracies that managed the various, overlapped regional economies of their empire.15 Although they might have spoken Old Persian, Iranian officials in the homeland of Parsa conducted most of their internal official written business in Elamite. Considering it was the script and language that officials used on a daily basis, the Elamite versions of inscriptions would have been just as legible as – or even more legible than – the Old Persian inscriptions to literate people. Darius I created the largest and longest rock-cut inscription of both the Achaemenid Empire and the ancient Iranian world at the site of Bisotun in ancient Media (Fig. 1).16 He carved a monumental trilingual inscription celebrating his

15

16

Matthew P. Canepa

victories over rivals for the throne and over rebellious provinces at the foothills of a craggy mountain, which was an important landmark on the route between the Mesopotamian lowlands and the Iranian plateau. The inscription was integrated with a monumental relief sculpture that reflected many of the same themes. The king carved the relief sculpture and Elamite inscription first in 520 BCE, followed by the Babylonian version.17 Darius I commanded a new writing system be created for the ‘Aryan’ language (ariya) and subsequently added the Old Persian version in 519 BCE.18 Under Darius I the monument remained dynamic. Darius I updated the relief and inscriptions with subsequent events, even effacing the Elamite inscription and recarving it in a different location to provide room for an additional sculpted figure of a new rebel. Locally, the inscription and the relief marked and claimed a site that appears to have had preexisting religious and military significance and it is likely that Darius defeated one of his major rivals for the throne nearby. In a broader sense, the relief literally and symbolically inscribed Darius I’s monumental presence onto this artery of the trade routes. Although the Bisotun inscriptions would not have been legible from the ground, Darius I disseminated the text throughout the empire in a variety of languages and in portable media. Darius set up an inscribed black basalt stele on Babylon’s Processional Way leading to the Ishtar Gate.19 It contained both a Babylonian translation of the Bisotun inscription and a condensed version of Bisotun’s great figural rock relief, suggesting that the visual and textual message formed part of a unified propaganda program.20 Babylon was among the provinces whose revolt Darius I had bloodily put down. The presence of his new inscribed monument transformed the significance of the famous thoroughfare and gate by visually and spatially altering one of the focal points of the rituals of the Babylonian New Year. Like the language of the inscription, the sculptural and epigraphic conventions of the stele as well as its placement engaged with Babylonian tradition.21 Other royal steles had stood in this area, including one created by Babylon’s last king, Nabonidus (556–539 BCE), who was overthrown by Cyrus.22 The inscribed stele’s presence on the Sacred Way continuously reminded the viewer that it was god’s will that Darius I be king and Babylon remain subject to the Persian Empire. Darius I and his son and successor, Xerxes I, carved rock inscriptions at several other sites throughout the empire, marking and shaping the natural environment in the process. At Darius I’s rock-cut tomb at Naqsh-e Rostam near Persepolis, the inscription was subordinate to the tomb’s figural sculpture. Much like at Bisotun, Darius I’s tomb inscription was carved after the sculptural relief was completed; it spreads over its surface, at places commenting on the sculpture and identifying the figures. At other sites, the inscriptions themselves are the only visual feature. The site of Ganj-nama, located in Media twelve kilometres from Ecbatana, hosted two small rectangular inscriptions carved into the rock near a waterfall on a route through the Alvand mountains. The contents of the inscriptions, which are identical but for the name of the kings, relate

Inscriptions, Royal Spaces and Iranian Identity

specifically to the site itself, but extol the Great God, the ‘Wise Lord’ Ahura Mazda, for bringing the king to the throne and lists the king’s lineage and conquests. Much like at Bisotun, Ganj-nama takes advantage of a prominent natural feature (a waterfall) on the east–west route to inscribe the presence of the kings into the landscape. Similarly, the only rock inscription that the dynasty carved outside of Iran consists of a panel on the towering cliffs of the Citadel of Van in present-day Turkey (ancient Tuˇspa). The citadel had been the core of the Urartian Empire, and its cliffs contained several rock-cut Urartian tombs. Darius started the inscription, carving away its frame, and his son, Xerxes I, finished it, adding the inscription itself.23 The inscription contains a standard preamble and then simply states that Darius I began the inscription and Xerxes I brought to completion what his father had begun. The text acknowledges the power of monumental inscriptions to mark and claim sites and surrounding space, even if their contents were rather mundane. The inscription at Van implies that the act of carving it and its visual presence were more important than its contents. This gets to the heart of the ultimate impact and purpose of Achaemenid monumental rock-cut inscriptions. In fact, the majority of the rock-cut inscriptions, with the possible exception of Ganj-nama, would have been illegible because of their location, high up on cliff faces. Their contents, however, were no doubt known and likely circulated in other monumental media, judging from evidence of the multiple copies of Darius I’s Bisotun inscription. Papyri fragments of the Bisotun inscription, which emerged in Elephantine, Egypt, corroborate Darius I’s claim that its contents were circulated in widely read languages and were possibly read out loud in public settings.24 Indeed, the king mentions that, after the Old Persian version was inscribed at Bisotun, the inscription was read out loud to him in its entirety, a public performance that likely occurred for the inscription on his tomb at Naqsh-e Rostam too.25 This ritualized act linked the visual impact of row upon row of inscriptions spreading across the rock face with their intended meaning and with the royal presence. Nevertheless, their indelible visual presence, if not the entire content of the inscription, was continually indexed to the royal presence, marking both the landscape and viewer’s minds. A visitor could not walk far without encountering an inscription in an Achaemenid palace. Departing from several Western Asian palatial traditions, the stone or glazed brick revetment of the walls of Persian palaces carried rich figural reliefs and intricate geometric and vegetal ornament and inscriptions. The palaces at Ecbatana, Susa, Babylon and especially Persepolis featured inscriptions in many prominent spaces, as well as hidden on their baked bricks and deep in their foundations.26 Architectural members such as column bases even received inscriptions commemorating their patron’s work. Even luxury items like silver or gold plate used at banquets carried inscriptions. Taking advantage of inscriptions’ ability to manipulate perceptions of time and authority, Darius I carved patronal inscriptions that he attributed to the founder of the empire, Cyrus the Great, at the

17

18

Matthew P. Canepa

2. View of the Gate of All Lands, Persepolis, Iran. Created by Xerxes I (486–465 BCE). (Photo: Matthew Canepa)

palace of Cyrus at Pasargadae. These inscriptions essentially put words in the mouth of Cyrus that portrayed his ancestry according to a spurious genealogy that Darius used to link himself to the founder of the empire. These inscriptions, in a sense, allowed Darius to provide empirical proof for his ideological fiction and claim Cyrus’s palace at Pasargadae as a monument to the newly invented ‘Achaemenid’ imperial dynasty.27 Extensive inscriptions joined figural sculpture in adorning the palace walls’ limestone orthostats, or in the case of Susa, baked brick revetment. Inscriptions were especially prominent at ritually and politically potent sites within the palace. At Persepolis, where these features are better preserved, inscriptions were placed in spaces that hosted the visits of envoys, subject rulers and governors, either in mass gatherings or individual processions. The palace’s main gatehouse, Xerxes I’s Gate of All Lands (Figs. 2 & 3), and the courtyard of the great audience hall, the apad¯ana, featured prominent inscriptions ensuring they were integral to the visitor’s experience of the site. The Old Persian version (XPa) of the inscriptions of Xerxes I’s Gate of All Lands provides a good example of the content of these inscriptions. After a typical preamble Xerxes I proclaims, ‘By the will of Wise Lord, I built this Gate of All Lands (imam duvarθim visadahyum adam akunavam). Abundant other Good was wrought in this [palace of ] Persepolis, which I built and my father built. Whatever work shows itself to be Good, we built it all by the will of the Wise Lord.’28 In the interior of the palace, inscriptions marked the facades of adorsed stairways facing courtyards and pilasters at entrance porches in the private palaces of Darius I and Xerxes I at Persepolis (the tacara and hadiˇs,

Inscriptions, Royal Spaces and Iranian Identity

3. Detail of the Old Persian inscription (center, above the wing) flanked by the Elamite and Babylonian versions on the interior of the northern wall of the Gate of All Lands, Persepolis, Iran. (Photo: Matthew Canepa)

respectively). Public or private, it is quite likely that these inscriptions were frequently read. Given their rather stereotyped content, they stood as indexes to their content, spurring the memory of the viewer and not requiring a long time to be spent in front of them. Considered this way, inscriptions could articulate and shape the elite viewers’ inner cognitive narrative and experience of the architectural space that they were led through.

AFTER ALEXANDER: RUPTURE AND REINVENTION

Epigraphic practices in the Iranian world changed markedly after the conquests of Alexander (d. 323 BCE) and the establishment of the Seleucid Empire (ca. 312 BCE). As with the history of art and architecture in the Iranian world, deep ruptures, unexpected continuities and constant influences from conquerors and neighbouring peoples define the history of monumental inscriptions in Iran after Alexander. In Babylonia, whose religious traditions the Seleucids cultivated, the Seleucid kings even continued the royal prerogative of laying down their own royal foundation deposits.29 However, this continuity and royal engagement with Babylonian religion did not extend beyond the old cities of Babylonia. The use of cuneiform scripts, be they Old Persian, Elamite or Babylonian, fell out of use in the majority of the lands of the former Achaemenid Empire, including the Iranian-speaking provinces of Persia, Media and Bactria.30 The Greek language

19

20

Matthew P. Canepa

and, just as importantly, Greek epigraphic practices quickly displaced Persian traditions.31 The majority of public inscriptions known from Seleucid Mesopotamia and Iran and the breakaway Greek kingdom of Bactria follow Greek traditions. These largely consist of inscriptions carved into stone steles or statue bases erected near official buildings, including sanctuaries, a fortress and a gymnasium. In addition, for the first time, imperial Aramaic, which had been used in record keeping and official communications but never for monumental inscriptions, begins to take this new role. It survived into the Macedonian and Parthian eras as an administrative language alongside Greek and soon became a language deemed worthy of monumental inscriptions itself. Affecting later Middle Iranian practices, under the Seleucids inscriptions no longer appear high up on cliff sides but more often than not, near eye level with a new emphasis on legibility drawn from the Greek traditions they continue. The Seleucid inscriptions can still claim important sites by their very presence, but they now more readily reach out to a wider constituency in a way that the Achaemenid inscriptions tended to do only in intimate, controlled palace settings. A series of inscribed steles connected with Antiochus III’s expansion of the official Seleucid ruler cult provides the best example of this adaptation of a Greek epigraphic practice to a new Iranian context. Three copies of his edict of 193 have been discovered at sites spread across the Seleucid Empire: Dodurga, Turkey (Eriza, Phrygia); Nehavand in Iran (ancient Laodicea-in-Media); and a fortress in the region of Kermanshah, Iran (Media).32 In this edict, Antiochus III introduced the official cult of the queen, Laodice, alongside the cult of the royal ancestors (progonoi) and his own cult.33 In the text the king ordered the edict displayed in the ‘most illustrious of sanctuaries’ or, in the case of the Kermanshah edict, an important fortress, though the sanctuaries themselves were apparently located in or near the cities.34 In the version from Laodicea-in-Media, which was carved onto a pedimented stele, a letter of an official, a certain Menedemos, precedes the king’s edict; it reiterates the king’s command that the edict be publicly displayed (Fig. 4). The steles and their distribution show that the Iranian plateau was integrated with the rest of the Seleucid Empire and experienced a royal inscriptional environment closely tied with a built and cultic environment similar to that cultivated in the Seleucid Mediterranean, Syria and Mesopotamia. This prestigious setting and function had a deep impression on Iran and Iranian practices of kingship, making monumental inscriptions an expected visual component of later Iranian dynastic cults.35 Although they were found out of context in a later Parthian house, fragments of an official decree from a sanctuary dedicated to the royal cult at Seleucia-on-the-Tigris indicate that this epigraphic practice occurred in the metropolis too.36 Only a handful of rock-cut Greek inscriptions survive from the period of Seleucid control of Mesopotamia and Iran (ca. 312–141 BCE); however, they demonstrate how Mediterranean epigraphic practices swept over Western Asia and permanently modified the ways in which inscriptions were deployed.37 Below Darius

Inscriptions, Royal Spaces and Iranian Identity

4. Cult Foundation of Antiochus III, Laodicea-in-Media. (Photo: Matthew Canepa)

21

22

Matthew P. Canepa

I’s relief, on the level of the plain, a bilingual Greek and Aramaic inscription accompanied a high relief sculpture of a reclining Herakles marking the entrance to the Hellenistic sanctuary precinct at Bisotun. The inscription, which is placed on a relief portrayal of a pedimented stele, dedicates the statue to the ‘preservation’ of the governor of the Upper Satrapies.38 The stele shape and content allude to a cultic context, whereas the feasting Herakles puts it into a wider ritual context of Seleucid royal ‘tryphic’ display.39 Although the nature of its royal statement and its epigraphic and iconographic languages differ from Darius I’s, the Seleucid relief nevertheless engages the Achaemenid precedent through its medium and the very site itself. The other inscription known from the Seleucid era, also related to Herakles, comes from a rock-cut Hellenistic fortress at the site of Karafto, which guarded the northern Median frontier of the Seleucid Empire.40 Typical of apotropaic domestic inscriptions found in the Mediterranean, a Greek inscription carved on the exterior of a rock-cut lintel of one of the site’s more finished, upper residential apartments reads: “Herakles dwells here, let nothing evil enter.” This inscription relates to private domestic inscriptions, but in what likely was an official context of the garrison commander’s quarters, it advertised Herakles’ protection of the Seleucid official in a manner parallel to the Bisotun inscription and relief. The presence of the inscription was just as important as its content: its durable physical presence was the key element that converted the rock-cut chamber into a divinely protected space. The major official buildings that have been excavated at the Seleucid capital of Seleucia-on-the-Tigris, such as the stoa, archives and theatre, have not yielded examples of monumental dedicatory inscriptions of the sort one would expect from such structures in the Mediterranean.41 However, this could be a result of the perishable materials used in their architectural decoration. At Ai Khanum, a Seleucid foundation and later capital of the independent GrecoBactrian kingdom (ca. 250–145 BCE), the surviving inscriptions appear in public places overtly connected with the city’s Greek identity, including the gymnasium and a cult site dedicated to the city’s founder.42 Patrons and gods who do not appear to be Greek could, however, be involved in Greek epigraphic practices. Located about 100 km west of Ai Khanum at the confluence of the Oxos River (modern Amu Darya) and a tributary, the Temple of the Oxos (Takht-e Sangin) was the region’s major cult site during the Seleucid Empire and, later, the independent kingdom of Bactria.43 Among the many cult dedications excavated at the site was an inscribed altar dedicated to the river Oxos, which carried a small bronze figure of a silenus playing the aulos. The language of the inscription is Greek, as is the religious custom of such a dedication and most of the other archaeological evidence of dedications at the site.44 The name of the patron contained in the inscription is Iranian, which indicates that Greek epigraphic practices had become an important component of the religious life of this culturally diverse region. Demonstrating the extent to which Seleucid epigraphic practices influenced Iran and served as a means of communication between Iran and the wider world,

Inscriptions, Royal Spaces and Iranian Identity

new and competing empires directly engaged Seleucid epigraphic practices. The later Mauryan emperor Ashoka (r. 269–232 BCE) created a series of inscriptions carved onto monumental columns or into the living rock across his empire as part of his policy of propagating Buddhism as an imperial religion.45 The majority of these edicts were in Prakrit written in Brahmi script, but a number of rock-cut inscriptions in regions that bordered the Greco-Bactrian kingdom incorporated Seleucid and Achaemenid traditions, thereby establishing a mixed precedent that the later Kushan rulers of these regions engaged.46 With new and neighbouring kings and dynasties eager to take control of this royal technology of power or communicate in prestigious idiom, Seleucid epigraphic practices created a common cultural continuum that connected Mediterranean, Western and South Asian epigraphic practices. After the fall of the Seleucids and the Greco-Bactrian kingdom, new Iranian dynasties engaged the art and architectural traditions of Macedonian charismatic kingship as practiced by the Seleucids and Greco-Bactrian kings. Oftentimes employing Greek artisans, these dynasties took the classical tradition in new directions. Their epigraphic practices follow a similar path of departure and creative reinvention. The Arsacid Empire (ca. 238 BCE–224 CE) initially used Greek as their administrative language and Seleucid conventions as a starting point for their own traditions, even as they adopted the Iranian language of their home province, Parthia. They used Greek in their early monumental inscriptions, a vocation that the other Seleucid administrative language, Aramaic, later replaced. The majority of monumental rock reliefs sponsored by the Arsacids were carved at the site of Bisotun, still dominated by Darius I’s great relief towering above. Among them the site hosted, on the lower rock face, a relief of courtiers paying allegiance to the Arsacid king of kings Mithradates II, over which ran a Greek inscription identifying the figures. Although the composition of the figural relief responds to that of Darius I’s sculpture on the cliff above, the accompanying inscription, running overhead in capital Greek letters, evokes a Greek entablature inscription. Aramaic inscriptions appear as rather crude additions in later Parthian reliefs. A freestanding boulder at Bisotun (the ‘Parthian stone’) carried a relief depicting scenes of sacrifice by a king of kings named Vologases, identified by an Aramaic inscription. This is not so much a reflection of the status of languages within the court itself. The Parthian language functioned as a medium of prestigious oral communication among the Iranian elite, particularly epic poetry or verse performed in a courtly setting. The Arsacids continued the tradition of Greek and Aramaic, which presented a visual and linguistic rhetoric of continuity with previous royal traditions, be they on stele, cliffs or coins. The Kushan dynasty (ca. 40–360 CE), which ruled Bactria and northern India, originally adopted Greek as an administrative language. Many of their epigraphic practices depart from Greco-Bactrian precedents though they eventually engaged South Asian traditions as well. The dynasty adapted the Greek alphabet to write the Iranian language of Bactria and continued using Seleucid dating

23

24

Matthew P. Canepa

conventions. The Kushans built a number of dynastic sanctuaries (Bactrian: bagolaggo) in Bactria that selectively engaged Greco-Bactrian and Iranian traditions.47 These sanctuaries adapted several features of Hellenistic architecture and traditions of dynastic cult, including a close connection between a sanctuary’s monumental inscriptions, cult regulations and dynastic memory. The early Kushan kings Vima Kadphises, Kanishka I and Huvishka were responsible for building or restoring a number of dynastic sanctuaries in which inscriptions featured prominently. Two Bactrian dynastic sanctuaries known by their modern toponyms Surkh Kotal and Rabatak and associated with Kanishka I (r. ca. 127–150 CE), both feature large inscribed slabs outlining the patronage and cult practiced at the site, paralleling the Seleucid inscribed royal edicts.48 In addition, the exterior fac¸ade of the main gateway of Surkh Kotal, whose architect was Greek, integrates a stone stringcourse carrying a monumental inscription.49 In contrast to the long detailed inscriptions on slabs, this inscription, integrated into the wall, was not fully legible unless the viewer/reader walked along the length of the exterior fac¸ade before entering into the sanctuary.50 The early Kushans also engaged with the royal tradition of carving trilingual rock-cut inscriptions in the natural environment. One example has survived, and given the wilful damage inflicted on it by the region’s inhabitants in both ancient and modern times, it is not unlikely that there were more throughout the empire.51 All told, Kushan approaches to monumental inscriptions were continuous with wider developments of post-Hellenistic Iranian kingship, but also integrated South Asian traditions as well, reflecting the reach of their empire. The last Persian dynasty of Iran before Islam, the Sasanians (224–642 CE), introduced a new, innovatory style of kingship. The founder of the dynasty, Ardaxshir I (224–239/40 CE), and his son Shabuhr I (239/40–270/2 CE) projected carefully crafted and unified visual and discursive messages that rooted their rule in the long history of Iranian kingship and marked them as visually and ideologically distinctive from the regimes they conquered. The early Sasanians cultivated the arts and produced finely crafted images in a variety of media that were demonstrably more refined than that of the Arsacids.52 Paralleling this, the early Sasanians developed epigraphic practices that engaged the long tradition of ancient Western Asian monumental inscriptions, but marked a break with Arsacid practices. We have only a few poor examples of monumental inscriptions in Parthian written with Aramaic chancellery script and conventions (Pahlavi); the Sasanians, however, self-consciously elevated their vernacular, Middle Persian, to be the premier language of monumental inscriptions. They adapted a Pahlavi alphabet to provide a visually impressive script. The alphabet was derived from the one used by scribes, and court scribes appear to have carved several major inscriptions.53 A handful of private aristocratic inscriptions survive, but the most prominent were those produced by the royal court. Although they had no knowledge of the languages and scripts, like the Seleucids and Arsacids, the Sasanians were drawn to many of the Achaemenid

Inscriptions, Royal Spaces and Iranian Identity

inscriptions they observed in their homeland. Compared to previous regimes, their relationship with them was much more intense and intimate. Along with sculptural reliefs, inscriptions were a key technique that the Sasanians used in forging enduring connections with Achaemenid sites and structures.54 At the Achaemenid necropolis of Naqsh-e Rostam, several Sasanian kings of kings carved monumental rock reliefs below the rock-cut tombs of the Achaemenid kings of kings. In addition to carving a rock relief at the site, Shabuhr I carved a trilingual inscription into the lower face of three sides of an Achaemenid tower constructed by Darius I, the Kaʿba-ye Zardosht (Fig. 5). Shabuhr I’s Kaʿba-ye Zardosht inscription, which is the longest and most important Sasanian inscription known, performed three main vocations.55 First of all, the inscription recorded the king of kings’ deeds and the extent of his empire in Middle Persian, Parthian and Greek, the languages of the two empires that the Sasanians conquered either in fact or symbolically. The content of the inscription is remarkable in and of itself. Although separated from the Achaemenids by centuries, composed in a much changed Persian language, and written in an entirely unrelated script, the structure, themes and even lexical choices of the Middle Persian version of Shabuhr I’s inscription are incredibly close to features that are common across many of the Old Persian inscriptions. At this stage we can only speculate whether these close correspondences arose from oral tradition or some other origin, but they reflect a wider cultural and political urge to collapse the gulf of centuries separating the Sasanians from their ancient and half-understood predecessors.56 Secondly, its very physical presence claimed the Achaemenid structure for the Sasanian dynasty, forging a permanent link between the two dynasties and providing, like the rock reliefs, visual ‘evidence’ of the implied relationship between them. Finally, the content of the inscriptions established a series of fires (¯adur) honouring the soul and memory of Shabuhr I, his queen and several of his sons. The inscriptions thus offered an anchor and constant commentary for cult that likely took place at Naqsh-e Rostam, if not around the tower itself. Unlike Darius I’s inscription, which remained inaccessible high on the cliff above, Shabuhr I’s inscriptions were accessible and legible to whomever was allowed to draw near the tower. In fact, internal clues suggest that Sasanian inscriptions were intended to be read and even read out loud, constantly reenlivening the memory of their creator in the process.57 The close connection between inscriptions and ritual activity extended beyond purely religious ritual. Shabuhr I carved inscriptions marking the landscape with the memory of his performance of a ritualized feat of strength and prowess. At two sites in Pars, Shabuhr I carved bilingual Middle Persian and Parthian inscriptions marking and celebrating the distance he could shoot an arrow, challenging ‘whoever may be strong of arm’, to attempt to match him.58 Although their content is more playful, these inscriptions share a similar sensibility with the Kaʿba inscription and indeed the older Iranian rock-cut inscriptions. They act as an interface between ritual performance, the viewer and the

25

26

Matthew P. Canepa

built or natural environment of the realm, permanently marking it with the king’s identity. Shabuhr I’s royal epigraphic practices inspired further activities at Naqsh-e Rostam and elsewhere in the empire. After the death of Wahram II (276–93 CE), Kerdir, the empire’s chief cleric, added a long Middle Persian inscription to the Kaʿba, engaging Shabuhr I’s precedent of discursive, visual, spatial engagement. Kerdir chose to carve his inscription directly below Shabuhr I’s Middle Persian inscription, indicating that this northwest face of the Kaʿba, which was the only one that did not face the fortification walls or was not encumbered by the stairway, was the dominant focus of the monument during the Sasanian era. In the inscription Kerdir claimed that Shabuhr I gave him control of all the cults at the site, thus using the inscription to graft his presence onto the site’s most prestigious ritual activities too. Kerdir carved two other inscriptions – a short inscription with a bust portrait contiguous with Shabuhr I’s rock relief and a similar bust relief and inscription at the cluster of Sasanian rock reliefs at Naqsh-e Rajab. Underscoring the fact that Sasanian inscriptions were meant to be legible and even audible, Kerdir’s Kaʿba inscription commands the viewer, ‘to read it out loud’.59 Inspired by the architectural, inscriptional and, possibly, ritual precedent set by Shabuhr I’s inscriptional intervention at the Kaʿba-ye Zardosht, Narseh (293– 302 CE) built an ashlar masonry tower on which he carved a bilingual inscription, in Middle Persian and Parthian, commemorating and legitimating his seizure of the throne.60 The tower marked the site where the grandees of the realm met and recognized his claim to the throne. In the reign of Shabuhr II (309–79 CE), a client king and brother of the king of kings, Shabuhr Saganshah (‘king of the Sakas’), carved two related inscriptions at Persepolis in the tacara of Darius I.61 The first memorialized his visit to the site and the rituals he performed for the gods, and the second, carved eighteen years later, commemorated this earlier visit to the palace of his ancestors. Although Shabuhr Saganshah’s inscriptions are in Middle Persian and the Pahlavi alphabet, they respond visually to the tacara’s Achaemenid inscriptions, which were located on the limestone piers at the entrance. Shabuhr’s inscriptions, located on the door frame, are arranged in a rectangle as were the Achaemenid inscriptions, though unlike the Achaemenid inscriptions, they do not have a linear frame surrounding them. Establishing an important precedent for Islamic art, in the late Sasanian Empire inscriptions became increasingly popular in a new context and medium: architectural ornament and stucco carving. Already used in the dynasty’s early palaces, stucco applied to rough stone masonry or mud brick walls became the favoured method of finishing and ornamenting prestigious structures in the Sasanian Empire.62 The ornamental motifs of early Sasanian palaces derived from a number of sources. The two palaces of Ardaxshir I at Ardaxshir-Xwarrah incorporated Achaemenid-inspired ornament, which later Shabuhr I juxtaposed with Roman-inspired ornament at his palace Bishapur.63 In the late empire,

Inscriptions, Royal Spaces and Iranian Identity

5. The Kaʿba-ye Zardosht (foreground) with a partial view of the sites of Shabuhr I’s Parthian and Greek inscriptions (lower courses of masonry) with an Achaemenid tomb and Sasanian relief in the background. Naqsh-e Rostam, Iran. (Photo: Matthew Canepa)

27

28

Matthew P. Canepa

Sasanian stucco carving developed a more regular repertoire of vegetal and geometric motifs that tended to cover vast areas with repeating, symmetrical motifs – often incorporating symbolic figural elements, such as boars, royal hunters, birds or rams, within ornamental frames. Treated in much the same way as these figural motifs, late Sasanian stucco incorporated ‘heraldic’ devices (n¯ısˇa¯ nih¯a, sing. n¯ısˇa¯ n) consisting of geometrically arranged Pahlavi letters set among ornamental patterns or, commonly, the symbolic figural motif of symmetrical wings64 (Fig. 6). Like the wings, the Pahlavi monograms often mimicked elements of the Sasanian royal crown, with shapes evoking the crescent moon or solar disk. Whereas earlier Iranian inscriptions were often finely and carefully wrought, Sasanian stucco carving integrated inscriptions into vegetal and geometric ornamental compositions. Halfway between monograms and nomadic tribal symbols (tamgas), such devices were associated with aristocratic families or royal institutions and appear widely on sealing stones, rock reliefs and, to a lesser extent, silver plate. All of the examples in stucco come from aristocratic dwellings, including urban palaces and country villas, and prestigious religious complexes. Although they served some of the same vocations as inscriptions in other media, such as linking a site or structure with a patron, they became an independent vehicle for aesthetic enjoyment in and of themselves. This merger between architectural ornament and writing in Sasanian stucco carving had no small influence on the origins and development of early Islamic calligraphy and its place within architectural ornament.

CONCLUSIONS

A variety of influences shaped Iranian royal epigraphic practices – from the silent, yet constant, presence of monuments of long-forgotten regimes to the prestige of new, non-Iranian royal traditions. Achaemenid inscriptions departed from the practices of the Western Asian empires they conquered and subsumed. Although Achaemenid conventions fell out of use when Macedonian kingship effaced Persian royal traditions, the sites the Achaemenids had favoured continued to draw patrons. The inscriptions from Bisotun, Takht-e Sangin and Ai Khanum all point to the impact of Hellenistic inscriptional practices on the Iranian world. Whereas Bisotun showcases the work of a Macedonian patron at a site associated with Persian kingship, Takht-e Sangin is particularly indicative of the other aspect of this process. Although both the name of the god (Oxos < ∗ Vaxˇsu) and the patron ¯ (Atrosokes < ∗ Atrǝ.sauka) of the altar at Takht-e Sangin are Iranian, the medium of worship and dedication came from the Greek world. By late antiquity a new repertoire of Iranian epigraphic practices had emerged that incorporated aspects of Hellenistic practices while responding anew to the older sites and traditions. The later reinventions of the Parthians and Sasanians responded to the inscriptions at sites and monuments of earlier dynasties and the growing competition among contemporary Middle Iranian sovereigns.

Inscriptions, Royal Spaces and Iranian Identity

6. Sasanian stucco panel from Umm az-Zaʿatir (near Ctesiphon) with Pahlavi letters possibly forming a monogram of Middle Persian abzud (‘increased’). Museum fur ¨ Islamische Kunst, Berlin. (Photo: Matthew Canepa)

Beyond merely preserving the contents of a text, inscriptions indelibly altered the visual impact and cultural meaning of a site or structure. They interpreted and commented on natural features, such as the mountain of Bisotun, or cultural features, such as the Kaʿba-ye Zardosht or the accumulated activities of succeeding regimes at sites like Bisotun. Inscriptions, along with figural reliefs, could establish a site with no previous activity as culturally important or manipulate the connotations of a site that already had accumulated cultural significance.65 Given their durable nature, rock-cut inscriptions could shape a region’s experience of its landscape and its populations’ experience of the past. By establishing tangible links between a patron and a natural or human-made feature, inscriptions imposed a visual and textual commentary on a site and could shape viewers’ experience of it and reactions to it. Done throughout a region or even empire, they could articulate a wider topography of power that testified to a dynasty’s

29

30

Matthew P. Canepa

control over a wide swath of territory and affected how viewers and subjects related to the landscape as they moved through it. Monumental rock-cut inscriptions served as crucial technologies of memory and power in ancient Iran. The visual presence of inscriptions implied that whatever their content expressed, even if it was the claim of a new dynasty, was as ancient and immutable as the natural feature or ancient structure into which it was carved. Inscriptions added to preexisting structures, reliefs or even to other inscriptions formed a permanent link between their cultural significance and the contemporary patron. This process could be rather subtle, such as in the case of Darius I’s inscriptions at Pasargadae, which linked Cyrus to the Achaemenid Dynasty, or Shabuhr Saganshah’s inscriptions in Darius I’s tacara at Persepolis, which established a simple connection between the patron and the general grandeur and illustrious (though half-remembered) patron of the site. Because of their volume and detail, Shabuhr I’s Kaʿba-ye Zardosht inscriptions radically reconfigured the significance of the site of Naqsh-e Rostam, permanently binding it to his identity and memory. Kerdir’s additions to Shabuhr I’s reliefs and inscriptions aggrandised his standing and connection to the great king of kings and permanently fixed it in stone. These inscriptions, in a sense, provided a permanent visual commentary by which later viewers would view the site and inscription’s patron. Sovereigns often attempted to claim or efface what was left of the topography of their royal predecessors. In some cases, such as in the Persian material, this could occur with the accumulation of multiple inscriptions and reliefs at a single site. The very act of creating an inscription could be a royal or, depending on its placement, imperial act. Even if the language, script or patron of an inscription might be non-Iranian, a site such as Bisotun and its adjacent features connected the patron to a deeper Iranian tradition. Depending on their ambitions and the history and cultural dynamics of their empire, patrons engaged multiple overlapping spheres and traditions, from regional precedents. For example, Darius I responded to the millennium-old Lullubi inscriptions and reliefs near Bisotun as well as the remnants of more recent empires, such as the Assyrian reliefs and inscriptions. Shabuhr I’s sponsorship of fine Greek epigraphy at the Kaʿbaye Zardosht sought to appropriate and subsume the traditions of the Roman Empire, as did the iconography and Roman sculptural forms of his later reliefs, the classical architectural and ornamental features of his palace at Bishapur, and Roman engineering in his hydraulic works in Khuzestan.66 Royal patrons often carved inscriptions at centres of ritual activity, and their content could comment on and shape the performance of the rituals, even to the point of their content being read out loud in the ritual space. Achaemenid palaces hosted a variety of royal rituals, and due to their location at palace gateways, entranceways and central courtyards, inscriptions served as visual backdrops to grand reception rituals and even articulated vessels used in royal banquets or weights used in courtly transactions. These inscriptions repeatedly voice an intimate connection between the Great God, Great King, the empire, and the

Inscriptions, Royal Spaces and Iranian Identity

structure and happiness for humankind. Their visual presence kept the meaning of the texts on the mind of the ritual participants, be they literate members of the court or visitors to whom the inscriptions’ contents were mediated. This reinforced the polemic of Achaemenid ideology that orchestrated the visual, architectural and ritual displays taking place at the court: the Achaemenid Empire and Achaemenid palace were divinely ordained creations that prefigure the postapocalyptic perfection of the world.67 Inscriptions could articulate ritual movement around a natural feature or structure or preserve the memory of a single ritual performance, making that punctiliar event a permanent topographic and cultural feature of the environment. Their placement drew attention to certain features and invited (or even forced) the viewer or reader to interact with the site and move around it in the way the patron envisioned. The inscriptions marking the distance of Shabuhr I’s bowshot and the site of Shabuhr Saganshah’s ritual meal at Persepolis both serve this purpose. Inscriptions could also perform ritual work in and of themselves. The dedicatory inscriptions at the Temple of the Oxos, in the gymnasium of Ai Khanum or on the Bisotun Herakles document celebrate a patron’s actions and its efficacy. The foundation tablets at Susa and Persepolis and the apotropaic inscription at the Karafto caves ‘worked’ simply because of their continued presence. In the most complex examples inscriptions established and regulated ritual activity and provided visual authority for their continued and proper performance. The Seleucid tradition of using inscriptions to establish and regulate cults, especially dynastic cult, had an enduring impact on several Iranian dynasties. Sasanian inscriptions forged a particularly close connection between ritual practice, place and inscriptions in the Middle Iranian era. Inscriptions were central in Shabuhr’s renovation of Naqsh-e Rostam, serving to claim the Kaʿba and establish the fires for his and his family’s soul. Their visual and spatial presence provided a pivotal link between the topography of a site and its ritual life. Despite cultural and political discontinuities, the constant presence of inscriptions from earlier dynasties inspired later practices, even though the memories of their original patrons and dynasties had long since dissipated. Although their cuneiform characters were mute and unintelligible to the late antique viewer, the very enduring presence of the Achaemenid rock-cut inscriptions at Naqshe Rostam or Persepolis inspired later dynasties to come to terms with the site using similar means. Rather than merely a product of Iranian tradition, inscriptions generated and replicated practices that formed and maintained Iranian royal tradition. This points to one of the most important and enduring functions of inscriptions in Iran: the presence of inscriptions at sites drew later interventions, from the Achaemenid through the Qajar era. Creating inscriptions in an ancient manner and engaging with ancient sites by carving new inscriptions were two of the most important activities of later Islamic kings and became a royal act in of itself. After Iranian culture recovered from the Arab invasions, local Iranian rulers established their connections with the grand pre-Islamic heritage by reprising Pahlavi inscriptions in their tomb towers.68 The Buyids, Safavids, and

31

32

Matthew P. Canepa

Qajars carved inscriptions at sites such as Persepolis, Naqsh-e Rostam, Bisotun and Taq-e Bostan.69 Crucial for the development of Islamic art, the late Sasanian dynasty’s experimentations with inscription as architectural ornament provided a departure point for this major vocation of inscriptions in Islamic ornament. NOTES

1. This chapter uses the abbreviations for Achaemenid inscriptions in Rudiger Schmitt, ¨ Die altpersischen Inschriften der Achaimeniden (Wiesbaden, 2009). Other abbreviations collected and explained in, Matthew P. Canepa, The Two Eyes of the Earth: Art and Ritual of Kingship between Rome and Sasanian Iran (Berkeley, 2009), xvii–xx. 2. For an entry into the literature on Iranian epigraphy, see Helmut Humbach, ‘Epigraphy i. Old Persian and Middle Iranian Epigraphy’, Encyclopaedia Iranica online, 1998, www.iranicaonline.org/articles/epigraphy-i; Philip Huyse, ‘Epigraphy ii. Greek Inscriptions from Ancient Iran’, Encyclopaedia Iranica online, 1998, www.iranica.com. 3. DB §70. 4. Ursula Seidl, ‘Ein Monument Darius’ I. ¨ Assyriologie und aus Babylon’, Zeitschrift fur vorderasiatische Arch¨aologie 89 (1999), 101– 14. Rudiger Schmitt, ‘Bisotun iii. Darius’s ¨ Inscriptions,’ Encyclopaedia Iranica online, 1990 [updated 2013], www.iranicaonline.org/ articles/bisotun-iii. 5. DB §70. Schmitt, ‘Bisotun iii. Darius’s Inscriptions’. 6. Neil Leach, ‘Towards a Theory of Identification with Place’, Perspecta 33 (2002), 126–33. 7. Ibid., 129. 8. However, in using this phrase I should stress that I do not advocate attributing independent agency to the inscription, but approach it a tool for effecting the king’s agency. Cf. Alfred Gell, Art and Agency: An Anthropological Theory (Oxford, 1998), 222–3. 9. Matthew Canepa, ‘Technologies of Memory in Early Sasanian Iran: Achaemenid Sites and Sasanian Identity’, American Journal of Archaeology 114.4 (2010), 563–96. 10. Humbach, ‘Epigraphy. i. Old and Middle Persian Epigraphy’. 11. For similar issues in Iranian figural rock reliefs, see Matthew P. Canepa, ‘Topographies of Power: Theorizing the Visual, Spatial and Ritual Contexts of Rock Reliefs in Ancient Iran’, in Of Rocks and Water: ¨ ur Towards an Archaeology of Place, ed. Om ¨ Harmans¸ah (Oxford and Philadelphia, 2014),

12.

13. 14.

15.

16.

55–92 (though ‘glazed brick’ should be corrected to ‘basalt’ on p. 77). ¨ ur Om ¨ Harmans¸ah, ‘Source of the Tigris: Event, Place and Performance in the Assyrian Landscapes of the Early Iron Age’, Archaeological Dialogues 14.2 (2007), 179–204. Ann Shafer, ‘Assyrian Royal Monuments on the Periphery: Ritual and the Making of Imperial Space’, in Near Eastern Art in Context, ed. J. Cheng and M. H. Feldman (Leiden, 2007), 133–60. Louis Vanden Berghe, Reliefs rupestres de l’Ir¯an ancien (Brussels, 1983), 19–32. In Egypt, Darius I set up inscribed steles at Suez (DZa-f ), which commemorated the reopening of a canal from the Red Sea to the Nile. A statue of Darius I was found at the palace of Susa, one of two that originally flanked its official entrance. It was also carved with Egyptian craftsmanship, though with substantial Persian influence. Although they incorporated Egyptian hieroglyphs and were produced with Egyptian craftsmanship, the Egyptian inscriptions were subordinate to the object’s cuneiform inscriptions and, in the case of the statues, their Persian iconography. As the content of the inscriptions makes clear, they functioned primarily as tangible evidence of Egypt’s subjection: ‘This is the stone image Darius the king caused to be made in Egypt, so that whoever sees it thereafter knows that the Persian man holds Egypt.’ DSab, Schmitt, Altpersische Inschriften, 146; R. Boucharlat, ‘Susa iii. The Achaemenid Period’, Encyclopaedia Iranica online (2009), www.iranica .com. Jean Yoyotte, ‘The Egyptian Statue of Darius’, in The Palace of Darius at Susa: The Great Royal Residence of Achaemenid Persia, ed. Jean Perrot (London and New York, 2013), 241–71. Wouter Henkelman, The Other Gods Who Are: Studies in Elamite-Iranian Acculturation Based on the Persepolis Fortification Texts (Leiden, 2008), 86. Heinz Luschey, ‘Bisotun ii. Archaeology’, Encyclopaedia Iranica online (1989), www .iranica.com.

Inscriptions, Royal Spaces and Iranian Identity

17. Hinz Luschey, ‘Ausgrabungen in Bisutun’, ¯ an-e B¯ast¯an 2.1 (1965), Anjoman-e Farhang-e Ir¯ 19–41. Hinz Luschey, ‘Studien zu dem Darius-Relief in Bisutun’, Arch¨aologische Mitteilungen aus Iran n.s. 1 (1968), 63–94. 18. DB §70. 19. Ursula Seidl, ‘Ein Relief Dareios’ I. in Babylon’, Arch¨aologische Mitteilungen aus Iran 9 (1976), 125–30; Ursula Seidl, ‘Eine Triumphstele Darius’ I. aus Babylon’, in Babylon: ¨ Focus mesopotamischer Geschichte, Wiege fruher Gelehrsamkeit, Mythos in der Moderne, ed. J. Renger (Saarbrucken, 1999), 297–306. ¨ Ursula Seidl, ‘Ein Monument Darius’ I. aus Babylon’, 113. 20. Some had argued for a version of the Bisotun relief in fragments of glazed brick from Susa, although this has since been rejected. Bruno Jacobs, ‘Eine Weitere Kopie des Bistun- Reliefs?- Zu einem Reliefziegel aus Susa’, Arch¨aologische Mitteilungen aus Iran n.s. 29 (1997), 303–08. 21. Seidl, ‘Eine Triumphstele Darius’ I. aus Babylon’, 303. 22. Hanspeter Schaudig, Die Inschriften Nabonids von Babylon und Kyros’ des Grossen samt den in ihrem Umfeld entstandenen Tendenzschriften. Textausgabe und Grammatik Alter Orient und Altes Testament 256 (Munster, 2001), ¨ 514. 23. G. D. Summers, ‘Archaeological Evidence for the Achemenid Period in Eastern Turkey’, Anatolian Studies 43 (1993), 85–108. 24. J. C. Greenfield and B. Porten, The Bisitun Inscription of Darius the Great. Aramaic Version (London, 1982). 25. DB §70. 26. The Achaemenid kings of kings incorporated texts into the foundations of several of their palatial structures, modifying a royal practice with a long history ancient Mesopotamia. Whereas Sumerian, Akkadian and Babylonian examples almost always include curses against anyone who should harm the structure, the Persian examples extol the king of kings and his empire in a manner similar to his other inscriptions, sometimes offer some sort of detail about the creation of the structure, and simply ask for Ahura Mazda’s protection for the king and his works. Richard S. Ellis, Foundation Deposits in Ancient Mesopotamia (New Haven, 1968). Darius I’s Susa inscription, which describes the construction of the palace, was inscribed on a clay tablet 22.5 by 26.5 cm and placed in the Palace of Darius under a threshold in the southern area of the place. Paris, Louvre: Sb 2787.

33

27.

28.

29. 30.

31.

32.

33.

Darius I placed stone boxes containing precious metal plaques ca. 32.5 by 33 cm by .2 cm in the southeastern and northeastern corners of the main walls of Perspolis’s apad¯ana. Tehran: 4116. Erich Schmidt, Persepolis I: Structures, Reliefs, Inscriptions (Chicago, 1953), 70 and 79, figs. 42 a-b, 43. Eight examples of stone foundation slabs from Xerxes I’s construction efforts were found collected in the Garrison Quarters but never used. Persepolis: Room 16, Schmidt, Persepolis I, fig. 87 D-E; Erich Schmidt, Persepolis II: Contents of the Treasury and Other Discoveries (Chicago, 1957), 52–3, pl. 21a. Margaret Cool Root, ‘Temple to Palace-King to Cosmos: Achaemenid Foundation Texts in Iran,’ in From the Foundations to the Crenellations: Essays on Temple Building in the Ancient Near East and Hebrew Bible, ed. Mark J. Boda and Jamie Novotny (Munster, 2010), ¨ 165–219. For the wider context, see Bruce Lincoln, Happiness for Mankind: Achaemenid Religion and the Imperial Project (Leuven, Paris, Walpol, 2012), 375–92. Bruce Lincoln, Religion, Empire, and Torture: The Case of Achaemenid Persia, with a Postscript on Abu Ghraib (Chicago, 2007), 45. XPa §3; Schmitt, Altpersische Inschriften, 153– 54. On the ‘theology of vision’ that manifests ‘the Good’ in Persian royal imperial discourse, see Lincoln, Happiness for Mankind, 180–83, 191–93. Ellis, Foundation Deposits, 161–3, 184 nos. 42 and 43. Although cuneiform fell out of use, cuneiform literacy likely persisted for many more years, which has important implications for the legibility of inscriptions. Mark Geller, ¨ Assyriologie ‘The Last Wedge’, Zeitschrift fur und Vorderasiatische Arch¨aologie 87 (1997), 43– 95. For a corpus of texts and entry into the literature, see Georges Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, Corpus Inscriptionum Iranicarum Pt. II vol. I (London, 2012). Matthew P. Canepa, ‘Dynastic Sanctuaries and Iranian Kingship between Alexander and Islam’, in Persian Kingship and Architecture: Strategies of Power in Iran from the Achaemenids to the Pahlavis, eds. Sussan Babaie and T. Grigor (London and New York, 2014), 65–117. For further bibliography, see note 32. Debord, ‘Le culte royal chez les S´eleucides’, 291–300. Virgilio, Diadema, Lancia, Porpora,

34

34.

35. 36.

37. 38.

39.

40.

41.

42.

Matthew P. Canepa

123. Van Nuffelen, ‘Le culte royale de l’empire des s´eleucides’, 278–85. Mehdi Rahbar and Sajjad Alibaigi, ‘The Hunt for Laodicea: A Greek Temple in Nahavand, Iran’, Antiquity 83, no. 322 (December 2009), http://antiquity.ac.uk/projgall/alibaigi322/. Canepa, ‘Dynastic Sanctuaries’. Leroy Waterman, Preliminary Report upon the Excavations at Tel Umar, Iraq (Ann Arbor, 1931), 1:9–17. Robert Harbold McDowell, Stamped and Inscribed Objects from Seleucia on the Tigris (Ann Arbor, 1935), 258. Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, nos. 70 and 75. Louis Robert, ‘Review of R.M. Frazer, Samothrace 2.1: Inscriptions on Stone’, Gnomon 35 (1963), 50–79. Heinz Luschey, ‘Bisotun ii. Archaeology’, Encyclopedia Iranica online, (1989), www.iranica.com. Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, no. 70, 145–48. For its context in the construction of the Seleucid Empire, see Paul Kosmin, Land of the Elephant Kings: Space, Territory, and Ideology in the Seleucid Empire (Cambridge, MA and London, 2014), 162–64. Paul Bernard, ‘Heracles, Les Grottes de Karafto et le sanctuaire du Mont Sambulos en Iran’, Studia Iranica 9 (1980), 301– 24. J. Tubach, ‘Herakles vom Berge Sanbulos’, Ancient Society (1995), 241–71. Von Gall’s argument that the site is a sanctuary has not gained much credence: H. von Gall, ‘Karafto Caves’, Encyclopaedia Iranica online, 2010, www.iranica.com. Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, no. 75, 154–56. Vito Messina, Seleucia al Tigri l’Edificio degli Archivi: Lo scavo e le fasi architettoniche (Florence, 2006), 27–69. For the larger context, Rachel Mairs, The Hellenistic Far East: Archaeology, Language, Identity in Greek Central Asia (Berkeley, 2014). For a review of the literature on Ai Khanum see, Rachel Mairs, The Archaeology of the Hellenistic Far East: A Survey (Oxford, 2011), 26– 29 and 38–39. Georges Rougemont, ‘Hellenism in Central Asia and the North-West of the Indo-Pakistan Sub-Continent: The Epigraphic Evidence’, Ancient Civilizations from Scythia to Siberia 18, no. 1 (2012), 175–182. Rachel Mairs, ‘The “Temple with Indented Niches” at Ai Khanoum: Ethnic and Civic Identity in Hellenistic Bactria’, in Cults, Creeds and Contests in the Post-Classical City, eds. Richard Alston and Onno van Nijf (Leiden, 2009).

43. Mairs, Archaeology of the Hellenistic Far East, 25–6 and 30. ¨ 44. Gunvor Lindstrom, ‘Heiligtumer und Kulte ¨ im hellenistischen Baktrien und Babylonien – ein Vergleich’, in Alexander der Grosse und die ¨ Offnung der Welt. Aisens Kulturen im Wandel, eds. Svend Hansen, Alfried Wieczorek and Michael Tellenbach (Mannheim, 2009), 127– 34. 45. Harry Falk, A´sokan Sites and Artefacts: A Source-Book with Bibliography (Mainz, 2006). 46. In Alexandria-Arachosia (Kandahar), a Greek city founded by Alexander, Ashoka sponsored edicts in Greek, (Kandahar I), Greek and Aramaic (Kandahar I) or Aramaic and Maghadi Prakrit transliterated with Aramaic script (Kandahar III). The region around Jalalabad to the north and Taxila to the southeast hosted a number of minor inscriptions in Aramaic that mention Ashoka or use his reign for dating: Falk, A´sokan Sites, 242–53. For the Greek inscriptions, see Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, nos. 82– 83, 157–73. 47. Canepa, ‘Dynastic Sanctuaries’. 48. Daniel Schlumberger et al., Surkh Kotal en Bactriane, 2 vols. (Paris, 1983–90), 2:132–8. The inscription of Rabatak was discovered out of context, with only a few architectural fragments marking the existence of the sanctuary it belonged to. Nicholas Sims-Willians, ‘The Bactrian Inscription of Rabatak: A New Reading’, Bulletin of the Asia Institute 18 (2004 [2008]), 53–68. 49. SK1: Schlumberger, Surkh Kotal, 2: 53, 134, and plate 35.94. Rougemont, Inscriptions grecques d’Iran et d’Asie centrale, no. 94, 194. 50. In contrast to the sanctuaries built in Bactria, a Kushan dynastic sanctuary built on the Gangetic plain near the winter capital, Mathura, was built using Indian architectural and cultic traditions and epigraphic practices. G´erard Fussman, ‘The M¯at. devakula: A New Approach to its Understanding’, in Mathur¯a: The Cultural Heritage New Delhi, ed. Doris Meth Srinivasan (New Delhi, 1989), 193–9. 51. About 50 km west from Ghazni, the Kushan king Vima Takto carved a series of inscriptions on a mountain, Mt. Qarabayu, which contributed to the western part of a ring of mountains surrounding a volcanic caldera, the Dasht-e Nawur. These inscriptions were in Bactrian with several Greek conventions, mixed Sanskrit using Kharoshti script and in another, currently undeciphered script. This inscription commemorates the passage

Inscriptions, Royal Spaces and Iranian Identity

52.

53. 54. 55.

56.

of the Kushan king of kings through the mountains, his conquest of the region and establishment of a cult at the site to commemorate his victory. Gerard Fussman, ‘Documents e´ pigraphiques kouchans. I. Inscriptions et antiquit´es du Daˇst-e N¯awur’, ¯ ´ Bulletin de l’Ecole Franc¸aise d’Extrˆeme-Orient 61 (1974), 1–76. For an overview of Sasanian rock reliefs, see Matthew P. Canepa, ‘Sasanian Rock Reliefs’, in The Oxford Handbook of Ancient Iran, ed. D. T. Potts (Oxford, 2013), 856–77. For an overview of the other arts, see the catalogue, Les Perses sassanides: Fastes d’un empire oubli´e (224–642) (Paris, 2006). As indicated in the colophon to the inscription of Kerdir, Naqsh-e Rajab. Canepa, ‘Technologies of Memory’, 580–4. Philip Huyse, Die dreisprachige Inschrift ˇ ˇ Sabuhrs I. an der Kaba-i Zardust (SKZ), Vols. Corpus Inscriptionum Iranicarum, Pt. III, 2 vols. (London, 1999). P.O. Skjærvø, ‘Thematic and Linguistic Parallels in the Achaemenian and Sassanian Inscriptions’, Papers in Honour of Professor Mary Boyce Acta Iranica 24 (Leiden, 1985), 593–603. Touraj Daryaee, ‘Sasanians and their Ancestors’, Proceedings of the 5th Conference of the Societas Iranologica Europæa, 2 vols., eds. A. Panaino and A. Piras (Milan: Mimesis, 2006) 1:287–93. Parallels explored more fully in M. Rahim Shayegan, Arsacids amd Sasanians: Political Ideology in Post-Hellenistic and Late Antique Persia (Cambridge, 2011), 1–29.

35

57. Kerdir, Kaʿba-ye Zardosht, 21. 58. Canepa, The Two Eyes of the Earth, 313 n.85. 59. KKZ 21. Canepa, ‘Technologies of Memory’, 582 n.87. 60. H. Humbach and P.O. Skjærvø, The Sassanian Inscription of Paikuli, 3 vols. (Weisbaden, 1983). Canepa, ‘Technologies of Memory’, 589. 61. Back, Sassanidischen Staatinschriften, 589–90. ¨ 62. Jens Kroger, Sasanidischer Stuckdekor (Mainz ¨ am Rhein, 1982); Jens Kroger, ‘Stucco Decoration in Iranian Architecture’, Encyclopaedia Iranica online, 2005, www.iranica.com. 63. Canepa, ‘Technologies of Memory’, 572–4, 584–5. ¨ 64. Catalogued in Kroger, Sasanidischer Stuckdekor: from Umm az-Zaʿt¯ır (Ctesiphon), 52– 4; Maʿa¯ rid. I (Ctesiphon), 84–5; Maʿa¯ rid. IV (Ctesiphon), 127; Mesopotamia (specific provenance unknown), 141; Takt-e ¯ Soleym¯an, 145; K¯ısˇ (east of Babylon), 189; ¯ an (Khurasan), 193–4. D¯amg¯ 65. Canepa, ‘Topographies of Power’. 66. Canepa, The Two Eyes of the Earth, 53–78. 67. Lincoln, Happiness for Mankind, 357–74. 68. Melanie Dawn Michailidis, ‘Landmarks of the Persian Renaissane: Monumental Funerary Architecture in Iran and Central Asia in the Tenth and Eleventh Centuries’ (Ph.D. Dissertation, Massachusetts Institute of Technology, 2007). 69. Shiela S. Blair, ‘Epigraphy iii. Arabic inscriptions in Persia’, Encyclopaedia Iranica online, 1998, www.iranica.com.

鵻 CHAPTER TWO

PRAYERS ON SITE: THE MATERIALITY OF DEVOTIONAL GRAFFITI AND THE PRODUCTION OF EARLY CHRISTIAN SACRED SPACE 鵼 Ann Marie Yasin

When the celebration of the Jubilee Holy Year of 2000 ended, the massive set of sculpted bronze doors that Pope John Paul II commissioned for the east porch of the church of St. John Lateran in Rome were ceremonially sealed.1 Pilgrims and tourists cannot now pass through the closed entranceway, yet when they visit the church many also come to see the doors, and especially to touch them (Fig. 7). When doing so, visitors take part in a pattern of mimetic action by repeating the gestures modelled by those before them: they touch the sculpted foot of the infant Christ and then kiss their contact-hand or make the sign of the cross with it (indeed, an informal queue frequently forms as individuals watch others approach the doors while waiting a turn for their own personal encounter). Although it has been only a little more than a decade since they were cast, the hands of thousands of visitors have irrevocably altered the appearance of the bronze doors. With each touch, visitors contribute to the polishing of the infant’s sculpted bronze limb. In the personal, bodily contact of pilgrims’ flesh to holy image, the once uniformly textured, compositionally coherent and officially commissioned door panel is transformed into a site of perceptible and inviting popular piety. A small fraction of the visitors to the Lateran leave a more individualized mark on the site by contributing to the scores of finely scratched graffiti that cover the jambs and surrounding mouldings of the Holy Doors. Some are anonymous

I am extremely grateful to Antony Eastmond, David Frankfurter, and the two anonymous readers for their insights and feedback on the text and to fellow workshop participants at the Courtauld Institute of Art and audiences at the University of Southern California and at the University of California, Riverside where subsequent of versions of this chapter were presented.

36

Prayers on Site

7. Jubilee Doors, St. John Lateran, Rome. (Photo: Ann Marie Yasin)

crosses and others now illegible marks, but many preserve names and, occasionally, dates of earlier visitors to the site. These scribbles, like the shiny bronze foot of the baby Jesus on the door, are physical manifestations of personal, ephemeral encounters with the site. They are material traces of the presence and religiously inspired gestures of previous visitors to the place; they signal past actions and individuals, but are made present each time the gesture is renewed. They are humble yet evocative, and we might imagine that a great deal of their affective

37

38

Ann Marie Yasin

power, their power to inspire others to copy and repeat the action with their own bodies, comes from the perception of the traces as acts of popular piety: ecclesiastical authorities neither prescribe nor regulate the gestures, but nevertheless permit and tacitly approve them (the graffiti are not covered over or buffed out; groups of pilgrims touching the Holy Doors do not attract the attention of guards or clerics). These kinds of pious gestures and the material marks they leave on holy places have a long history. This chapter examines the earliest surviving traces of graffiti at Christian holy sites as both evidence of and impetus for personal, physical interaction with a holy place. Like the modern rubbings and scratches in the Lateran porch, the traces are modest. Yet, then as now, graffiti provide rare and direct physical evidence of individual users’ interaction with buildings and urban spaces. In contrast to idealized or prescriptive textual descriptions of ritual, graffiti offer hard evidence of actual devotional gestures carried out at holy sites. The contents of the handwritten messages are telling in their patterns and simplicity. The meaning and impact of the texts, however, are inherently linked to their materiality. I suggest that it is through their status as site-specific artefacts that the inscribed names of holy figures and their devotees evoke the presence of both. Moreover, the cumulative nature of graffiti invites active participation and renews and amplifies this message of divine and devotee presence over time. Equally significant is what the physical characteristics and placement of the texts reveal about the graffiti writers’ and viewers’ perception of and interaction with sacred space. Graffiti declare the efficacious sanctity of a place by attesting to multitudes of previous devotions enacted there. Their irregular, handmade traces evoke the very bodily experiences of their diverse writers and likewise engage viewers on a similarly personal and somatic level.

GRAFFITI IN CULTURAL AND SPATIAL CONTEXT

Before proceeding further, some clarification of terms will be useful. In contemporary usage, the word ‘graffiti’ is applied to a wide range of writing in both modern and ancient contexts. In today’s world, the term usually conjures up associations with defacement, anti-establishment expression, or, increasingly since the 1970s and 1980s, with pop culture and edgy contemporary art that critique and play off of mainstream characterizations of the genre.2 It is important to note, however, that the overtones of vandalism or illicit behaviour that accompany graffiti in the modern world do not appear to have been universal or even central to opinions about informal wall writing in the Greco-Roman sphere.3 This is not to say that graffiti were always and everywhere deemed desirable or acceptable – scattered ancient literary sources allow us to sniff out traces of elite condescension towards the medium – but it is clear that we are dealing with a culture of unofficial public writing very different from our own.4 Ancient graffiti

Prayers on Site

communicated a wide range of verbal content, from scatological jokes and love letters, to political commentary, literary riffs, accounting notations and prayers. It was also apparently tolerated, indeed sometimes even written, by authorities and/or property owners in a wide range of contexts and situations, including not only in latrines and bars but also in public squares, on and in houses, by military personnel, and on statues, tombs and shrines.5 It is this last type – graffiti scrawled on the surfaces of sites considered holy – that primarily concerns us here. As sources of evidence, these graffiti present numerous challenges. Some ancient Christian graffiti were carved into durable stone supports, but many were scratched into or drawn on the thin layers of plaster that coated walls and other surfaces. They are therefore inherently fragile, subject to destruction when buildings are damaged or renovated, and rarely survive in anything close to complete and fully legible form.6 The sites on which this chapter focuses have been selected because they house graffiti both well enough preserved at the time of their discovery to have been recorded in at least some degree of detail and dateable to the late antique period (third–seventh century).7 Within this time span, although the inclusion of absolute dates in the texts themselves is exceptionally rare, some graffiti can be more firmly dated than others because of their archaeological contexts.8 At Rome’s Memoria Apostolorum and the memorial of St. Peter on the Vatican Hill, for example, the older architectural features on which graffiti were written were replaced by grander buildings in the first decades of the fourth century, burying and thus preserving the earlier graffiti-covered surfaces until their discovery in the twentieth century.9 At the early Christian church under the Liebfrauenkirche at Trier, graffiti cut into the plaster-coated choir screen were found amid the excavated debris of the church that had been destroyed in the fifth century.10 In situ sections of plaster walls at both the Church of the Holy Cross at Resafa (ancient Sergiopolis) in Syria and the shrine of St. Felix at Cimitile-Nola preserve graffiti left by late antique writers thanks to subsequent installations that sealed off the inscribed surfaces until they were once again revealed by modern excavations.11 Elsewhere, however, the exposed surfaces of long-lived holy sites, such as the Roman catacombs, the Church of the Theotokos in the Parthenon in Athens, the Church of St. Stephen at Philae, and the Monastery of Apa Apollo in Bawit, Egypt, preserve writing that can span generations, even centuries.12 It is important to bear in mind, therefore, that the archaeology of each site presents its own window onto graffiti activity, either in a relatively concentrated time frame or over a longer dur´ee. Most analyses of early Christian graffiti have used them as a source of historical evidence about the development of cult – the graffiti attest to visitation to a particular holy place, the veneration of a particular martyr or pilgrimage to a particular shrine. Indeed, from the earliest chapter of Christian archaeology, graffiti have been central to the project of charting an early Christian sacred topography. For example, in his monumental nineteenth-century Roma

39

40

Ann Marie Yasin

sotteranea cristiana, Giovanni Battista de Rossi’s careful attention to epigraphy allowed him and subsequent scholars to map out entire catacomb systems by using graffiti to distinguish sacred ‘hotspots’ where especially venerated individuals were buried.13 As the authors of one later handbook of Christian archaeology wrote, ‘Almost always, in the Roman catacombs, they [= graffiti] indicated the proximity of a historical crypt; at times they enable us to follow, step by step, the same way as the ancient visitors used when they went from the upper basilicas to the subterranean chapels which contained the most venerated tombs’.14 Like centuries-old breadcrumbs, in other words, catacomb graffiti allow modern investigators to trace their way back through the labyrinthine subterranean corridors to the historically venerated places. Although indebted to this valuable work, this chapter focuses less on plotting a map of historical cult sites than on examining what Christian graffiti can tell us about the lived experience of sacred spaces.15 In doing so, I examine early Christian graffiti from two related angles. On the one hand I consider graffiti as material manifestations of devotional practice. Graffiti present one type of response to a religious site and contribute to our understanding of late antique behaviour and ritual at and around holy sites more broadly, along with pilgrimage, prayers, offerings, and other devotional actions such as kneeling, kissing and lighting candles. On the other hand, I consider how graffiti, through their form and placement, communicated to viewers, conditioned subsequent visitors’ interaction with the site, and transformed sacred landscapes over time. As Colin Adams has written, ‘graffiti are dynamic in the sense that they form a body of evidence at particular locations that grows over time, and thus create and develop the history of a location – they are deeply linked to place’.16 Essentially these two aspects constitute opposing faces of the same coin, one focusing on the producers of the inscribed texts and their actions at a sacred place, and the other on the inscriptions’ reception, their effects on subsequent visitors to the site. Given the fragile nature of the evidence, its relatively serendipitous rate of survival and the inconsistent conventions of publication, our picture is necessarily tenuous and incomplete. Nevertheless, I argue that examination of the diverse body of evidence from across the Mediterranean allows us to identify three key spatial and material aspects of early Christian graffiti. First, graffiti do not merely indicate the contours of a preexisting sacred topography, but are rather an active means through which to interact with and shape a natural or manmade landscape. Not planned by the patrons or builders of a site, but incorporated into architectural fabrics, the inscribed names, prayers and acclamations speak very directly to the physical action of writing as a popular devotional response to particular places. Second, the graffiti attracted other graffiti. The greatest proof of this is the fact that the handwritten messages regularly occur in groups – not as isolated texts, but in clusters that comprise a larger graffiti-laden field. Carving or drawing graffiti was a means of both joining and perpetuating a devotional community, for the writing often mimics earlier writers’ gestures at a site while

Prayers on Site

simultaneously inspiring and inviting the production of future graffiti. Such texts adhered to collective conventions while also remaining vehicles for individual expression. Finally, the graffiti worked collectively to transform the supports on which they were inscribed and thereby constructed new meanings for the space bound by those physical structures. They palpably signalled to viewers that the place was an active and effective venue for Christian prayer. In this sense, graffiti both established a kind of space for communal action and provided a means of inserting oneself into that community.

MATERIALIZING PRESENCE AND PRAYERS

Even if graffiti are relatively uncommon at major Greco-Roman temples,17 substantial corpora of hand-written inscriptions have been recovered from a number of cult sites, such as the shrine of Hercules Curinus outside Sulmona in Italy, the mithraeum at Dura Europos, the Gallo-Roman sanctuary at Chˆateauneuf, and Egyptian shrines at Abydos and Philae.18 These places give us a glimpse at a site-specific, interactive and user-generated facet of the material culture of sacred space. The content of the texts conveys the writer’s religious intent while the writing’s form and placement transform the site itself. At Chˆateauneuf, for example, fragments of painted plaster reveal that the exterior walls of the cella were covered with scratched messages – records of donations made to the shrine and acclamations of the gods. Some call out to divinities broadly (e.g. no. 3: ‘Gods! Goddesses!’), but many of the plaster fragments preserve the names of specific gods: Mercury, Maia, the local god Limetus, Roma, and the divine Roman emperor each appear multiple times.19 The divinities’ names literally covered the walls of the fanum. The προσκύνημα (‘act of worship’) type inscriptions common to Greek graffiti found at Egyptian shrines from the end of the second century BCE ‘speak’ their writers’ messages with even more syntactical directness.20 More than one hundred examples of these prayers of adoration adorn the walls of Isis’s sanctuary at Philae, for example, and are written in the first person by named devotees who direct their acts of worship to the goddess, who is also directly named.21 The formulae used in graffiti at the mithraeum at Dura Europos are different, but many seek to achieve similar results: memorializing the acclamation of the named god by a particular named devotee.22 These records of donations to the gods and of vows made in their names heaped up evidence of the divinities’ power upon the very fabric of their sacred houses. The role of naming is thus key to many of the surviving graffiti from Greco-Roman shrines. The handwritten inscriptions provide a tangible proof at the site of the shrine of the presence of the gods named – Hercules Curinus/Quirinus in his shrine outside Sulmona, Mithras in his Durene sanctuary, Serapis and Bes at the Memnonion at Abydos, and so on.23 Although much of the ancient graffiti carved or painted elsewhere on an ancient city’s surfaces

41

42

Ann Marie Yasin

¨ 8. In situ section of graffiti at the Memoria Apostolorum triclia. (After: Styger, Romische Martyrergrufte, vol. 2, pl. 18) ¨

were uncredited (anonymity being a central component of their function, for example, as political commentary or social expression), this is clearly not the case for shrines.24 Graffiti found at cult sites regularly declare, one after another, the names of the individuals responsible for the scrawl. Thus, much like votive offerings, the graffiti attest both to the relationship between specific devotee and deity and to the special status of the sanctuary space. They are also, like votives, both performative and commemorative, materializing a gesture of communication while at the same time remaining on site after the fact as a record of the exchange. The graffiti ensure that, even after the visitor leaves the shrine bodily, his or her name persists, physically inscribed on the holy site. Graffiti at early Christian sacred places too consisted largely of names and slightly more expanded texts that regularly adopted standard expressions for communication with the non-earthly realm. The walls of the so-called triclia area at the Memoria Apostolorum, a kind of above-ground trapezoidal portico with masonry benches on the Via Appia outside of Rome, were awash with scratched and scribbled marks of visitors (Fig. 8): the recovered plaster fragments attest to more than 400 individuals whose names were inscribed in the time between the area’s installation in the mid-third century and its burial as part of the construction of the Basilica Apostolorum (today’s San Sebastiano) in the early fourth century.25 Inscribing one’s name is the most direct and individualized way of recording one’s personal presence at the site. Some graffiti emphasize the immediacy of

Prayers on Site

9. Philae, Temple of Isis/Church of St. Stephen, south wall of hypostyle hall, west side of entrance, with insc. nos. 205–14. (After: Bernand, Inscriptions grecques et latins, vol. 2, pl. 46)

the encounter with the place even more emphatically. ‘I, Dioskoros’ and ‘I, Josaias,’ for example, are but two of the Greek Christian graffiti gouged into the south wall of the hypostyle hall of the temple of Isis at Philae, Egypt, after it was converted to a church of St. Stephen in the sixth century, that heighten the sense of personal presence by preceding the visitor’s name with a resounding first-person ‘ἐγὼ’ (Fig. 9).26 Others on the same stretch of wall declare the subject’s humility in relation to the divine through the addition of the common epithet ‘servant of God’ or ‘servant of Jesus’.27 At the same time, graffiti also offered a way to materialize prayers for those not able to be there in person. Indeed, a rare late antique literary reference to graffiti writing at an early Christian holy place attests to the practice of inscribing not one’s own name but those of others. In an account by a pilgrim to the Holy Land from Piacenza in the late sixth century, the author writes of visiting Cana and of reclining on the very wedding couch where Christ did. There, he says, ‘I, although unworthy, wrote the names of my parents’.28 Archaeological evidence bears out this practice. It is not infrequent, for example, for a single graffito to list many individuals. At Resafa, for instance, one Sergios left a graffito asking God to remember him as well as a number of other relatives, including his father, daughter and sister.29 In graffiti like this, although it is possible that all of those listed were present at the time of writing, it is probably more likely that some of the names included were written by the family member(s) who had ventured to the holy place and prayed there on behalf of them all. This is undoubtedly the case for graffiti such as that a certain Leo wrote to St. Felix at Cimitile-Nola for himself and ‘all his’, using the blanket phrase ‘cum suis omnibus’, which covered

43

44

Ann Marie Yasin

household members broadly, as well as for some handwritten inscriptions at the Memoria Apostolorum that record prayers for those no longer living.30 Through the material inscription of their names and the prayers made in their honour, therefore, graffiti could make it possible even for those not present bodily to have a physically inscribed presence at the holy site. Many of the texts spell out material prayers very explicitly. Early Christian graffiti writers regularly drew on a common storehouse of formulae such as the Latin ‘pete pro X’ (‘pray for X’) or the Greek ‘Βοήθε X’ (‘help X’) to articulate their message.31 The addressee was sometimes God, but also often a saint or saints, as was particularly common in the graffiti scratched into the tufa or plaster coating of Roman catacomb walls.32 The repetition and conventionality of the texts are striking. Some take the form of a benediction on behalf of the named individual. At least twenty-four of the graffiti gouged into the plaster face of the choir screen of the fourth-century church under the present Liebfrauenkirche at Trier, for example, include variations on the common acclamation ‘May you live in God’ (e.g. vivas in deo, vivas in XP, vivas in Domino, vivas in deo XP semper, etc.).33 On the triclia walls of the Memoria Apostolorum, approximately half of the 189 more or less complete graffiti name Peter and Paul directly (see Figs. 8 & 10).34 Here the most common prayer appealed to the apostles to ‘remember so-and-so’ (‘in mente habete’), written at least sixty different times at the site.35 These texts adopt the grammatical form of a direct address to the holy figures (with imperatives and vocatives). The result is an extremely densely packed field of handwritten inscriptions, as if the walls themselves speak on behalf of or in the voice of the texts’ authors to call out to the apostles by name. At the most fundamental level, the graffiti are a mode of communication between their writers on earth and God, or saintly intercessors, in the divine realm. The communication may have once been conveyed orally (through uttered prayers) in a passing moment during one’s visit to a shrine, but the graffito presents a lasting material vestige of the exchange at the site. For the hundreds of individuals whose names were accompanied by these formulae, the written text was a physical trace that concretized their presence, a sign of having been to the site or having been remembered by others there, and the act of writing the graffito was also an overt, if humble, gesture of devotion, an active appeal to the heavenly realm.

AUDIENCE AND ACTION

As material acts of devotion, the graffiti thus both communicated to the heavens and commemorated that exchange. As we have seen, they attested to the piety of the writers and the power of heavenly figures, and for both groups they ensured a form of physical presence through the lasting material trace of their proper names on the fabric of the site. Graffiti also, however, would have been encountered by

Prayers on Site

10. Graffiti-covered plaster fragments from the Memoria Apostolorum triclia. (After: Styger, ¨ R¨omische Martyrergrufte, vol. 2, pl. 21)

other visitors to the place, and the fact that the inscriptions cluster on specific sections of walls illustrates that graffiti invited action – specifically, the response of mirroring the gesture of earlier writers by adding one’s own mark to the site.36 The result is not the kind of ‘dialogue’ between writers prompted by many earlier Roman graffiti such as the texts scratched on Pompeian walls (or modern bathroom stalls!) that frequently find scrawled rejoinders, whether playful, clever, satirical or derogatory.37 Rather, graffiti at sacred places encouraged viewers to copy the gesture of prayer writing, of adding their own names, of joining the ranks of those who had gone before in seeking divine assistance at that place. In some cases the form of previously written graffiti seems directly to influence others produced in the same place. For instance, graffiti often adopt formulae that are common at that particular site, but are found less frequently in other regions of the early Christian world. The ‘᾿Εγώ + name’ inscriptions from Philae discussed in the previous section (Fig. 9), for example, are not found amongst

45

46

Ann Marie Yasin

the roughly contemporary Greek graffiti from Resafa.38 It seems reasonable to suggest that, for a would-be graffiti writer at Philae, the choice to word one’s own inscription in this way could very well have been influenced by the form of other devotional inscriptions at the site. At the Memoria Apostolorum, as we have seen, a good half of the preserved graffiti write out the names of the two apostles with the result that the words ‘Petre et Paule’ are repeated at least 100 times on the same wall (see Figs. 8 & 10).39 Yet in the dense palimpsest of graffiti on the so-called wall g of the memorial under St. Peter’s, the apostle’s full name is conspicuously absent, and the patch of plaster is instead replete with scratched chi-rho monograms, symbols that are not common in the graffiti at the Memoria Apostolorum triclia across town.40 At Trier, approximately two-thirds of the more or less complete graffiti open or close with a chi-rho monogram, creating a striking type of visual conformity not found at contemporary sites.41 At the so-called Grotto of St. Paul on the outskirts of Ephesos, nearly all of the graffiti drawn or scratched into the fourth- to sixth-century plaster layers adopt a variation of ‘Paul [or Christ, or Lord], help your servant’ (βοήθη/βοήθι τοῦ δούλου σοῦ).42 We find therefore, a certain, limited degree of internal cohesion among graffiti texts from some sites, and where this is the case, the graffiti themselves would have served as an important means of communicating and transmitting those local preferences and conventions. Indeed, Rachel Mair’s observation about Hellenistic and Roman graffiti to Pan at El Karnais in the Egyptian desert is apposite for us as well: ‘writing dedicatory graffiti . . . was a repetitive self-reproducing practice, reinforced by constant performance’.43 Adding one’s own scrawl, whether simply one’s name or a longer text, to the cluster of graffiti at a sacred site was then both a means to communicate with the divine and to join the ranks of others who have done so at the same place before. Yet through the words they wrote, certain individuals proclaimed an even more specific place within the community of Christian devotees to the shrine. For example, some graffiti testify to foreign names and foreign languages or explicitly indicate the writer’s distant homeland.44 In the panel we have examined from the Hypostyle Hall at Philae, for example, four Christian inscriptions identify their writers as Nubians (Fig. 9).45 In addition to what they indicate about the history of pilgrimage, such inscriptions also reveal the importance of the holy sites as epigraphic spaces. The graffito could serve as a key medium for the representation, performance and registration of one’s status as a pious Christian from near or far. When indications of ecclesiastical titles are present, as they are increasingly from the seventh to eighth centuries in the graffiti from Italy, they also indicate a growing distinction between spaces accessed, venerated – and graffitied – by lay people and those apparently reserved for clergy.46 Registering your devotion through the physical act of inscribing your name amidst the others written on the walls of a holy site was therefore not only a devotional but also a social gesture. Mary Beard’s important discussion of the role of names inscribed at Roman sanctuaries as a means of what she calls

Prayers on Site

47

(a)

(b)

11. a. Graffiti wall at the shrine of St. Felix, Cimitile-Nola. (Photo: Ann Marie Yasin). b. Drawing of graffiti on left portion of graffiti wall at the shrine of St. Felix, Cimitile-Nola. (After: Ferrua, ‘Graffiti di pellegrini’, 19, fig. 2)

‘signing up for paganism’ is relevant for us here.47 She sees inscribed votive texts as playing a ‘central role in defining the place of the individual within traditional paganism’, solidifying and making permanent the relationship between devotee and god that is otherwise expressed through the occasional, fleeting ritual of sacrifice.48 I would argue that the Christian graffiti we have been examining also fulfil these important roles. The inscription of one’s name at a shrine constitutes a physical trace of the appeal to God or a saint. It registers one’s affiliation with Christianity and even to a specific saint’s cult – a record of ‘membership’ in the religious community, in the Beardian sense. I would press further, however, the importance of the location and form of the inscriptions. Writing one’s name can be seen as a declaration of joining, an assertion of membership in the group, but it can also be used as a means to stand out. Through their appearance and scale many graffiti adhere closely to norms, as we have seen. Others, however, clearly strive to distinguish themselves from the surrounding texts. On the red wall at St. Felix’s shrine at Cimitile, for example, the 4- to 5-cm high letters of the fragmentary text at the top (‘BI’ on one line followed by ‘ERIO’ on a second line) dwarf the rest of the texts below them (Fig. 11).49 We find the same disparity of scale in other clusters of graffiti, for example in the case of the large ‘κύριε βοέ[θησον]’ (‘Lord, help!’) text engraved

48

Ann Marie Yasin

into the Resafa wall (Fig. 12)50 and the deeply gauged, 3.5- to 4-cm high letters of a certain Victor’s prayer to Paul and Peter at the Memoria Apostolorum (Fig. 8).51 Frames and other graphic arrangements can similarly distinguish individual texts. Slightly below the large, attention-grabbing prayer to the Lord on the Resafa wall, the roughly rectangular tabula ansata frame around the text of one Symeon’s plea employs an archaizing epigraphic technique to segregate and elevate that writer’s message (Fig. 12).52 Other writers instead chose to capitalize on a wall’s original painted decoration. Certain graffiti at the Memoria Aposotolum, such as Victor’s pronounced inscription, for example, are aligned with the painted borders of the preexisting garden scene, thus taking advantage of the polychrome effects and attracting visual attention by the way they are inserted within the larger compositional programme (Fig. 8).53 Likewise, at the Monastery of Apa Apollo at Bawit, Egypt, the painted plaster walls of the chapels of the sixth or seventh to eighth centuries include many that explicitly appropriate the elaborate geometrical patterning of the painted decoration in their design and orientation.54 The surviving evidence of graffiti-encrusted walls thus reveals that the scribbled texts were not only passively witnessed: from some viewers they also elicited a parallel gestural response. Making one’s own mark at a holy place was a means of materializing one’s wishes and prayers to the divine along with the others who had patently done so before. The wall created a virtual community of devotees that transcended a given time, and adding one’s own graffito provided a means of joining and literally carving out one’s place in the group.

PERCEPTION AND CREATION OF SACRED SPACE

Graffiti could affect ancient viewers on other registers as well. When a visitor encountered an inscribed graffito, its status as a text that was written by a fellow devotee rather than a professional stone cutter was clear. A lettered visitor might decipher the words of a specific graffito and find a connection to its message, but even the illiterate would find themselves occupying the very same physical position of the text’s writer and thereby understand that they were not alone in the experiences and motivations that brought them to the place. Collectively, one of the messages a mass of graffiti would have conveyed to visitors most forcefully was about the nature of the place itself – the popular recognition of its special sanctity and accessibility. The material properties of graffiti speak to the physicality of the devotional act of writing the text, of cutting or drawing letters on the wall. Unlike the clean products of professionally tooled lettering with which formal inscriptions deny the impression of the craftsmen’s physical labour, such as funerary epitaphs or the finely wrought lettering produced by the calligrapher Filocalus for Damasus, Bishop of Rome, in the second half of the fourth century

Prayers on Site

12. Resafa, Basilica of the Holy Cross, detail of graffiti wall fragment B in situ. (Photo after: Resafaarchiv DAI Berlin, courtesy T. Ulbert)

(Fig. 13), the graffiti’s uneven scratches and chips fervently reveal the hands of the different writers who scraped pointed instruments, or even fingernails, through the surface of the plaster (Fig. 14).55 The visual mash-up or juxtaposition of different hands, scales and scripts thus reveals far more than is said by its textual content; its materiality renders it a discernible record of multitudes of individual physical gestures of piety.56 In contrast to a formal inscription that resonates with a steadfast voice of institutional authority (conveyed by such aspects as the placement, material, alignment and letterforms), the clusters of graffiti seem to murmur a steady hum of separate and unorchestrated yet harmonious prayers. The disparity in the formal properties of the writing found on a single patch of wall drives home the personal and unpremeditated nature of the texts. More than the content of the prayers, their varied handwriting presents viewers with hard evidence for a particular form of spontaneous popular devotion performed over and over again at the site.57 In addition, the appearance and location of the graffiti allow viewers to relate to the texts bodily, regardless of their degree of literacy and ability to access the textual content.58 Philae’s graffiti-covered monumental walls notwithstanding, most of the graffiti at Christian holy sites are very intimate affairs both in terms of the personal nature of their content and their scale.59 They differ markedly from formal, professionally carved inscriptions not only in their diverse and irregular scripts, as we have seen, but also in their challenging legibility. In contrast to formal inscriptions that use visibility-enhancing graphic conventions, such as

49

50

Ann Marie Yasin

13. Reconstruction drawing of the interior of the Crypt of the Popes in the catacomb of Callixtus, Rome. (De Rossi, Roma sotterranea, vol. 2, pl. 1A)

regular lines and text blocks, the form and spacing of letters, the scale and depth of carving, and the possible addition of pigment, most devotional graffiti with their irregular scripts, shallow carving and letter heights, often no more than 1–2 cm, are quite difficult to see from a distance. At the same time, the relative illegibility of graffiti in comparison to official and professionally produced inscriptions ensures that, when it is noticed, it has the potential to engender a profoundly different reception and reaction in its viewers. In general, to see a

Prayers on Site

14. Reconstruction drawing of the entrance of the Crypt of the Popes in the catacomb of Callixtus, Rome. (After: De Rossi, Roma sotterranea, vol. 2, pl. 29.1)

particular graffito, one needs to stand close to the inscribed surface. Notice, for example, the distribution of graffiti at the entrance to the Crypt of the Popes (Fig. 14), which in real life are more difficult to see than they appear in de Rossi’s drawing because they are simply scratched into the white plaster. The texts do not span the entire surface of the wall, but congregate at roughly eye level at the threshold of the burial chamber. This puts the viewer in approximately the same position as the writer had been at the time he or she inscribed the text. These graffiti, in other words, work on a personal, somatic level. The human scale of the texts encourages a sense of personal identification with writers, not necessarily as specific individuals, but as fellow devout visitors who engaged in similar rituals and experiences at the same exact spot. Considered from the perspective of an ancient Christian visitor to the site, therefore, devotional graffiti offered traces of others who had previously, in that very place, made verbal and physical gestures of communication with the divine. In this way, graffiti turned the place into a particular type of epigraphic environment, one that spoke of the masses of devotional gestures carried out there.60 Borrowing a phrase from V´eronique Plesch’s work on later medieval church graffiti, a gallery of graffiti prayers turn the place into a perceptible ‘contact zone’ between earthly and heavenly realms.61 The texts render the site a recognizable space of prayer and communication. By creating such a contact zone the introduction of graffiti could transform a site both physically and symbolically. For example, the petitions and invocations

51

52

Ann Marie Yasin

engraved on either side of the entrance to the Crypt of the Popes in the Catacomb of Callixtus or at the shrine of St. Felix in the cemetery zone of Cimitile north of Nola conveyed that these burial grounds were no longer merely places for interring and remembering the dead (Figs. 11 and 14). They had instead become spaces for communicating with heavenly martyrs and seeking divine assistance. In these cases, a cemetery zone has become a more affective sacred place – it attracts a broader audience who come not only to honour and commemorate the deceased but also to seek help from the ‘special dead’ for themselves or for their loved ones. At other sites, graffiti participate in an even more dramatic type of transformation. Historically pagan shrines that were made into churches in the late antique period, such as the Parthenon in Athens or the sanctuary of Isis at Philae, underwent certain alterations, such as the addition of apses, Christian dedication inscriptions and monumental crosses.62 These architectural changes mark an official liturgical and administrative conversion of the structure, but inscribed Christian graffiti enact the transformation at a palpable level of individual religious experience for they testify to the devotions of specific Christians and to the effectiveness of Christian prayers at the once pagan site. The columns of the Parthenon in Athens, for example, support more than 230 Christian inscriptions, prayers and epitaphs spanning the late sixth or seventh to the fifteenth century.63 At Philae, more than thirty Christian devotional inscriptions, written after the temple’s conversion to a Christian church in the sixth century, upstage the earlier Greco-Roman graffiti directed to Isis.64 Furthermore, at both Athens and Philae, the Christian identity of the graffiti prayers is graphically underscored by the frequent inclusion of engraved crosses or chi-rho symbols. At these sites, the Christian graffiti are therefore an important element in both enacting and perceiving the ritual and cultural conversion of traditional pagan religious landscapes and structures. The specific location of such graffiti-formed ‘contact zones’ within the larger sacred topography of each site is also noteworthy. As discussed earlier, graffiti often come down to us only by chance, and rarely do we have surfaces from an entire sacred complex with which to map relative densities of texts. Yet it is striking that evidence from numerous sites indicates that entrances or thresholds were often key locations for inscribing devotions. In addition to what this might indicate about social conventions or authorities’ control over access to other, more holy areas, it also reveals a claim to the liminal space as a site of popular devotion.65 The passageway walls in the Catacomb of Callixtus at the entrance to the Crypt of the Popes, for example are crowded with inscribed names and individual devotional prayers: the writing fields directly line both sides of the entrance to the chamber (Fig. 14). The cluster of graffiti we have examined from Philae was located on the wall of the hypostyle wall, directly adjacent to the entranceway (Fig. 9).66 Interestingly too, at the Parthenon church the Christian inscriptions are not evenly distributed around the building. Most columns on

Prayers on Site

15. Resafa, Basilica of the Holy Cross, plan with location of graffiti wall and martyrium indicated. (After: Resafaarchiv DAI Berlin, courtesy T. Ulbert)

the long flanks have no or at most one or two inscribed texts, but at the entrance end the number explodes: 188 graffiti have been documented on the fourteen columns that form the inner and outer colonnades at the church’s western face.67 Moreover, within the general pattern of preference for the entrance side, certain columns flanking heavily used steps and doorways were clearly more popular than others, with some columns having attracted more than twenty or thirty graffiti each.68 The situation was perhaps somewhat analogous at the Church of the Holy Cross, built in 559 at Resafa in the eastern Syrian desert. Here, in a subsidiary chamber the chance survival of a section of wall plaster preserved under a later interior dividing wall revealed an exceptionally dense field of eightyseven Greek inscribed graffiti (Figs. 12 and 15).69 The architectural location of the room is critical. This thickly graffitied chamber connected the courtyard on the north side of the basilica to the martyrium room where the relics (likely of the True Cross) were displayed in a monumental reliquary, as indicated by the steps,

53

54

Ann Marie Yasin

columns and pavement patterns.70 Which other plastered surfaces of the complex supported inscribed prayers we do not know, but it is striking that this heavily written wall would have been encountered by visitors as they entered from the courtyard before they reached the shrine proper. Indeed, like the texts flanking the doorway to the Crypt of the Popes in Rome and those carved on the columns at the west end of the Parthenon church, the Resafa graffiti spatially anticipated the primary focal point of devotion. They materially defined the threshold as a place of prayer and conditioned the visitors’ movement into increasingly sacred space. For the pious visitor, carving a graffito was one of a range of devotional gestures that could have accompanied a visit to a holy place. But unlike bowing in supplication, intoning a prayer, lighting a candle or partaking of the Eucharist, inscribing a graffito offered a tangible interaction with the very substance of the site. Through it one’s presence at the shrine and one’s appeal to higher powers there found fixed form and a modicum of endurance. What is more, although some scholars have argued that Christian graffiti were not meant for a human audience, that the inscribed prayers were rather fully about communicating with the divine,71 the form and placement of the inscriptions suggest otherwise. Even when words are not easily discernible, the graffiti communicate to and elicit responses from their human viewers through their location and material properties: their position on architectural supports; their juxtaposition to other features (including earlier inscribed texts); their use of conventional or unique formulae, symbols or framing devices; and the shallowness, scale and regularity of their strokes. To subsequent viewers, graffiti present lasting, visible testimony of past devotional actions performed at the site. They operate on personal and physical levels, inviting viewers to occupy the bodily position of the writers, and demonstrating a means for performing and displaying participation within the social group of devotees. The inscriptions sanction, even encourage, those moved to mirror the gesture and scratch their own mark on the shrine’s surface. Graffiti also operate collectively and spatially. Galleries of graffiti conjure religious communities that extend beyond those worshippers present at any given moment. The traces, neither officially designed nor preordained, testify to parallel actions carried out in the same area by multitudes of individuals over time. Graffiti thus transform larger sacred topographies by signalling the use of a preexisting space now explicitly for communicating and memorializing individual petitions to the divine. It is clear then that, for us, graffiti’s value as material artefacts of devotional and social behaviour far exceeds that of their textual content alone. NOTES

1. Each of Rome’s four main basilicas has a set of ‘Holy Doors’ that are opened ritually by the pope or his representative at the beginning of the Jubilee year and then walled up

at its conclusion (for a summary, see Herbert Thurston, s.v. ‘Holy Year of Jubilee’. The Catholic Encyclopedia (New York, 1910): www.newadvent.org/cathen/08531c.htm

Prayers on Site

2.

3.

4.

5.

6.

7.

(15 Aug. 2011). For the Vatican’s description of the ritual and significance of the opening of the Holy doors for the Jubilee year of 2000, see www.vatican.va/news services/ liturgy/documents/ns lit doc 14121999 porta-santa en.html. Graffiti as art is more popular than ever: the Museum of Contemporary Art in Los Angeles reports that more than 200,000 people visited its recent exhibit of graffiti art, The Art in the Streets, during its less than four-month run in 2011, the highest attended exhibition in the museum’s history to date (‘MOCA Announces Record-Breaking Exhibition Attendance’, Museum of Contemporary Art, Press Release, August 10, 2011: www.moca.org/pressroom/index.php). On the history of modern graffiti and the art world, see Susan Stewart, ‘Ceci Tuera Cela: Graffiti as Crime and Art’, in Life after Postmodernism: Essays on Value and Culture, ed. J. Fekete (New York, 1987), 161– 80; and Jeffery Deitch, Roger Gastman, and Aaron Rose, eds. Art in the Streets (New York, 2011). See J. A. Baird and Claire Taylor, ‘Ancient Graffiti in Context: Introduction’, in Ancient Graffiti in Context, eds. J. A. Baird and Claire Taylor (New York, 2011), 3–4; though cf. Roger S. Bagnall, Everyday Writing in the Graeco-Roman East (Berkeley, 2011), 24. Alexei V. Zadorojnyi, ‘Transcripts of Dissent? Political Graffiti and Elite Ideology under the Principate’, in Ancient Graffiti in Context. eds. J. A. Baird and Claire Taylor (New York, 2011), 110–33. See the essays and bibliography in Baird and Taylor, Ancient Graffiti in Context. Cf. Juliet Fleming, Graffiti and the Writing Arts of Early Modern England (Philadelphia, 2001), 29–72. Werner Eck, ‘Graffiti an Pilgerorten im ¨ sp¨atromischen Reich’, in Akten des XII. inter¨ christliche Arch¨aologie. nationalen Kongresses fur Bonn 22–28. September 1991, Jahrbuch fur ¨ Antike und Christentum Erg¨anzungsband 20,1 (Munster, 1995), 208–10. Most work on early Christian graffiti is limited to study of a particular site. For an important synthetic account, see Eck, ‘Graffiti an Pilgerorten’. Also still useful is Henri Leclercq, s.v. ‘Graffites’, in Dictionnaire d’archéologie chrétienne et de liturgie, ed. F. Cabrol (Paris, 1925), VI.2: cols. 1453–1542. Andrea Binsfeld’s recent study of the early Christian graffiti from Trier also includes a valuable summary of comparanda from other sites: Vivas in Deo. Die Graffiti der

55

8. 9.

10.

11.

¨ fruhchristlichen Kirchenanlage in Trier (Trier, 2006), 43–146. On dating graffiti, see Eck, ‘Graffiti an Pilgerorten’, 211–12. Memoria Apostolorum: Inscriptiones christianae urbis Romae septimo saeculo antiquiores, nova series. 10 vols, eds. Angelo Silvagni, Antonio Ferrua, D. Mazzoleni and C. Carletti (Rome and Vatican City, 1922–1992) [hereafter ICUR], V.12907–13096; Antonio Ferrua, La basilica e la catacomba di S. Sebastiano, 2nd ed. (Vatican City, 1990); Antonio Ferrua, ‘Rileggendo i graffiti di S. Sebastiano’, La civilt`a cattolica 116 (1965), 428–37; Antonio Ferrua, ‘Memorie dei SS. Pietro e Paolo nel epigrafia’, in Saecularia Petri et Pauli, eds. B. M. Apollonj Getti, et al. (Vatican City, 1969), 131–48; F. Grossi-Gondi, ‘Il “refrigerium” celebrato in onore dei SS. Apostoli Pietro ¨ christe Paolo’, R¨omische Quartalschrift fur liche Altertumskunde und Kirchengeschichte 29 (1995), 222–49; St Peter’s: Margherita Guarducci, I graffiti sotto la Confessione di San Pietro in Vaticano, 3 vols. (Vatican City, 1958). The graffiti fragments were found in 1949– 50 in the area of the presbyterium of the south basilica of the cathedral complex: Nancy Gauthier’s catalogue of the Trier graffiti (Recueil des inscriptions chr´etiennes de la Gaul ant´erieures a` la Renaissance carolingienne. Vol. 1: Premi`ere Belgique [Paris, 1975], 544– 54) has been updated by Binsfeld, Vivas in Deo, 172–90. See also Andrea Binsfeld, ‘Die Graffiti der fruhchristlichen Kirchenanlage ¨ in Trier’, in Neue Forschungen zu den Anf¨angen des Christentums im Rheinland, ed. S. Ristow (Munster, 2004), 235–52; Mark A. Handley, Death, Society and Culture: Inscriptions and Epitaphs in Gaul and Spain, AD 300–750, BAR International Series 1135 (Oxford, 2003), 163–64; Mark A. Handley, ‘Beyond Hagiography: Epigraphic Commemoration and the Cult of Saints in Late Antique Trier’, in Society and Culture in Late Antique Gaul: Revisiting the Sources, eds. R. W. Mathisen and D. Shanzer (London, 2001), 187–200. For a summary of the building phases, see Sebastian Ristow, ¨ Fruhes Christentum im Rheinland. Die Zeugnisse der arch¨aologischen und historischen Quellen ¨ an Rhein, Maas und Mosel (Koln, 2007), 193– 203. See also n. 33. Resafa: Thilo Ulbert, Resafa II: Die Basilica des Heiligen Kreuzes in Resafa-Sergiupolis (Mainz am Rhein, 1986). Within the same volume, the Greek graffiti are catalogued by Cor¨ nelia Romer (‘Die griechischen Graffiti’, 171–7) and the Arabic by Raif Georges

56

Ann Marie Yasin

Khoury (‘Die arabischen Inschriften’, 179– 80); Binsfeld, Vivas in Deo, 106–13. CimitileNola: Antonio Ferrua, ‘Graffiti di pellegrini alla tomba di San Felice’, Palladio n.s. 13 (1963), 17–19; Binsfeld, Vivas in Deo, 85–93. 12. In addition to the sources listed in nn. 6–7, key recent publications addressing graffiti in Roman catacombs include: Carlo Carletti, ‘Viatores ad martyres: Testimonianze scritte altomedievali nelle catacombe romane’, in Epigrafia medieval greca e latina: ideologia e funzione. Atti del Seminario di Erice, 12– 18 settembre 1991 (Spoleto, 1995), 197–225; ibid., ‘ “Scrivere i santi”: epigrafia del pellegrinaggio a Roma nei secoli VII-IX’, in Roma fra Oriente e Occidente, 19–24 aprile 2001 (Spoleto, 2002), 323–60; and the overview with select examples in Binsfeld, Vivas in Deo, 70–85. On the Parthenon church graffiti, see A. K. Orlandos and L. Vranousis, Τα χαράγματα τοῦ Παρθενῶνος. Les graffiti du Parth´enon. Inscriptions grav´ees sur les colonnes du Parth´enon a` l’´epoque pal´eochr´etienne et byzantine (Athens, 1973); and Anthony Kaldellis, The Christian Parthenon: Classicism and Pilgrimage in Byzantine Athens (Cambridge, 2009), 74–80. For Philae see: Andr´e Bernand, Les inscriptions grecques de Philae, 2 vols. (Paris, 1969); P. Nautin, ‘La conversion du temple de Philae en e´ glise chr´etienne’, Cahiers arch´eologiques 17 (1967), 1–43; and Jitse H. F. Dijkstra, Philae and the End of Ancient Egyptian Religion: A Regional Study of Religious Transformation (298–642 CE), (Leuven, 2008). On Bawit: Jean Maspero, Fouilles ex´ecut´ees a` Baouˆıt, M´emoires publi´es par les membres de l’Institut franc¸ais d’arch´eologie orientale du Caire 59 (Cairo, 1931). For a recent overview of the history and evidence from the monastery, see Alain Delattre, Papyrus coptes et grecs du monast`ere d’apa Apollˆo de Baouˆıt conserve´es aux Mus´ees Royaux d’Art et d’Histoire de Bruxelles (Louvain-la-Neuve, 2004), 29–109, and the brief summary in Roger S. Bagnall and Dominic W. Rathbone, Egypt from Alexander to the Copts: An Archaeological and Historical Guide (London, 2004), 175–8. 13. Giovanni Battista De Rossi, La Roma sotterranea cristiana. 3 vols. (Rome, 1864–77). See also the important recent study by Vincenzo Fiocchi Nicolai: ‘Itinera ad sanctos. Testimonianze monumentali del passaggio dei pellegrini nei santuari del suburbia romano’, in ¨ Akten des XII. Internationalen Kongresses fur christliche Arch¨aologie (Munster, 1995) 2: 763– ¨ 75.

14. Orazio Marucchi and Hubert Vecchierello, Manual of Christian Archaeology, 4th ed. (Paterson, New Jersey, 1935), 265. 15. For a historiographical overview of late antique spatial studies and a call to focus on ‘lived space’, see Luke Lavan, ‘Late Antique Urban Topography: From Architecture to Human Space’, in Theory and Practice in Late Antique Archaeology (Leiden, 2003), 171–95. 16. Colin Adams, ‘Travel and the Perception of Space in the Eastern Desert’, in Wahrnehmung und Erfassung geographischer R¨aume in der Antike, ed. M. Rathmann (Mainz am Rhein, 2007), 215. 17. Martin Langner, Antike Graffitizeichnungen. Motive, Gestaltung und Bedeutung, Pallia 11 (Wiesbaden, 2001), 21. 18. Sulmona: Margherita Guarducci, ‘Graffiti parietali nel santuario di Ercole Curino presso Sulmona’, in Scritti sul mondo antico in memoria di Fulvio Grosso, ed. L. Gasparini (Rome, 1981), 225–40. Dura Europos: F. Cumont and M. I. Rostovtzeff, ‘Dipinti and Graffiti’, in The Excavations at Dura-Europos conducted by Yale University and the French Academy of Inscriptions and Letters. Preliminary Report of the Seventh and Eighth Seasons of Work 1933–1934 and 1934–1935. eds. M. I. Rostovtzeff, F. E. Brown, and C. B. Welles (New Haven, 1939), 116–28; E. D. Francis, ‘Mithraic graffiti from Dura-Europos’, in Mithraic Studies: Proceedings of the First International Congress of Mithraic Studies, ed. John R. Hinnells (Manchester, 1971), 2: 424– 45; Franz Cumont, ‘The Dura Mithraeum,’ in Mithraic Studies: Proceedings of the First International Congress of Mithraic Studies, ed. John R. Hinnells (Manchester, 1971), 1:150– 214, esp. 194–205. Chˆateauneuf: Christian Mermet, ‘Le sanctuaire gallo-romain de Chˆateauneuf (Savoie)’, Gallia 50 (1993), 95– 138. Abydos: Ian Rutherford, ‘Pilgrimage in Greco-Roman Egypt: New Perspectives on Graffiti from the Memnonion at Abydos’, in Ancient Perspectives on Egypt, eds. R. Matthews and C. Roemer (London, 2003), 171–90. Philae: Bernand, Inscriptions grecques de Philae; Ian Rutherford, ‘Island of the Extremity: Space, Language and Power in the Pilgrimage Traditions of Philae’, in Pilgrimage and Holy Space in Late Antique Egypt, ed. David Frankfurter (Leiden, 1998), 229–56; Nautin, ‘Conversion du temple’, and Dijkstra, Philae and the End of Ancient Egyptian Religion. 19. Mermet, ‘Sanctuaire gallo-romain’. See also Alix Barbet and Michel Fuchs, Les murs murmurent. Graffiti gallo-romains. Catalogue de

Prayers on Site

20.

21. 22.

23. 24.

25.

l’esposition cr´ee´ e au Mus´ee romain de LausanneVidy, 2008 (Gollion, 2008), 152–63 with color reproductions of Chˆateauneuf examples from the mid first century and a plaster fragment from the exterior of the courtyard wall of the temple at Jublains, dated after the mid second century, with proper names of gods and temple visitors. Giovanni Geraci, ‘Ricerche sul Proskynema’, Aegyptus 51 (1971), 3–211; Rutherford, ‘Island of the Extremity’, 237; and Rutherford, ‘Pilgrimage in Greco-Roman Egypt’, 179. Bernand, Inscriptions grecques de Philae; Geraci, ‘Ricerche sul Proskynema’, 115–36. Most of the sixty-some catalogued graffiti and dipinti from the Dura Europos mithraeum date to the second phase of the structure, from ca. 210–40 CE, and more than half of the Dura mithraeum graffiti use the ‘nama’ formula, a kind of cultic acclamation or greeting appropriated from Old Persian (Francis, ‘Mithraic graffiti’, 438; Cumont, ‘Dura Mithraeum’, 195–6; Cumont and Rostovtzeff, ‘Dipinti and Graffiti’, 117–22). E.g.: ῾Υπὲρ Νίκης τοῦ Κυρί/ου ἡμῶν Αὐτοκράτορος. / Νάμα θεῷ Μίθρᾳ, / νάμα πατράσι Λιβει/ανῷ καὶ Θεοδώρῳ, / νάμα καὶ Μαρείνῳ πε/τίτορι, νάμα πᾶσι τοῖς / συνδεξίοις παρὰ τῷ θε[ῷ]. (Francis, ‘Mithraic graffiti’, 438). See sources in n. 18. E.g. see Zadorojnyi, ‘Transcripts of Dissent?’ and cf. the religious graffiti discussed by J. A. Baird, ‘The Graffiti of Dura-Europos: A Contextual Approach’, in Ancient Graffiti in Context, eds. J. A. Baird and Claire Taylor (New York, 2011), 56. Monica Smith offers important insights on the role of anonymity in modern graffiti: ‘Walls Have Ears: A Contextual Approach to Graffiti’, International Folklore Review 4 (1986), 100–5. ICUR V.12907–13096; Carlo Carletti, ‘Nascita e sviluppo del formulario epigrafico cristiano: prassi e ideologia’, in Les iscrizioni dei cristiani in Vaticano. Materiali e contribute scientifici per una mostra epigrafica, Inscriptiones sanctae sedis 2. ed. Ivan Di Stefano Manzella (Vatican City, 1997), 148. On the Memoria Apostolorum more broadly, with a description of the triclia, see: F. Tolotti, Memorie degli Apostoli in Catacumbas (Vatican City, 1953); R. Krautheimer, and S. Corbett, ‘S. Sebastiano’, in Corpus Basilicarum Christianarum Romae. The Early Christian Basilicas of Rome (IVIX Cent.) (Vatican City, 1970), 4: 99–147, esp. 114–18; and Paul Styger, R¨omische Mar¨ tyrergrufte, 2 vols. (Berlin, 1935), 1:15–36.

57

26. Bernand, Inscriptions grecques et latins, 2:268– 77, insc. 205–15; Nautin, ‘Conversion du temple’, 29–31; Dijkstra, Philae, 335. 27. E.g. ‘᾿Εγὼ ᾿Ιωάνν|ης δοῦλος| [Θεο]ῦ’ and ‘᾿Εγὼ ᾿Αάρων Νουβὰ δοῦλος [᾿Ι](ησοῦ)ς’ (Nautin, ‘Conversion du temple’, 31; on the selfidentification of the subject of the second inscription as Nubian, see the section, ‘Audience and Action’). 28. Full passage of Antonini Placentini Itinerarium 4, in Itinera hierosolymita saeculi IIIIVIII, (CSEL 39) ed. P. Geyer (Vienna, 1898), 161: ‘Deinde milia tria venimus in Cana, ubi ad nuptias fuit Dominus, et accumsimus in ipso accubitu, ubi ego indignus nomina parentum meorum scripsi’. An English translation of the full text is available in John Wilkinson, Jerusalem Pilgrims before the Crusades (Warminster, England, 1977), 79–89. See also Eck, ‘Graffiti an Pilgerorten’, 206, and Gary Vikan, Early Byzantine Pilgrimage Art, rev. ed. (Washington, D.C., 2010), 71. ¨ 29. Romer, ‘Griechischen Graffiti’, 173, insc. 14. 30. Ferrua, ‘Graffiti di pellegrini’, 18; Ferrua, ‘Rileggendo i graffiti’, 432–3. 31. Carletti, ‘Nascita e sviluppo’, 149 on prayer formulae at the Memoria Apostolorum. See also Eck, ‘Graffiti an Pilgerorten’, 217, and H. Leclercq, s.v. ‘Boetheia’, in Dictionnaire d’archéologie chrétienne et de liturgie, ed. F. Cabrol (Paris, 1910) 2.1: cols. 962–6. 32. Carletti, ‘Nascita e sviluppo’. 33. Binsfeld, Vivas in Deo, 24–6 and 35–7, and id., ‘Graffiti der fruhchristlichen Kirchen¨ anlage’, 241–3. Recent review of the evidence has determined that the graffiti-covered plaster fragments come from two (previously thought to be three) phases of the choir screen, the first apparently from some time between the 330s and the middle, or just after the middle, of the century, and the second from the reign of Valentinian I (364–75) until the church’s destruction in the fifth century (Binsfeld, Vivas in Deo, 15–22). 34. Ferrua, ‘Rileggendo i graffiti’, 431. 35. Ibid., 431–2. 36. Cf. Jean Guyon, Le cimeti`ere aux deux lauriers. Recherches sur les catacombs romaines (Rome, 1987), 471. 37. Rebecca Benefiel, ‘Dialogues of Ancient Graffiti in the House of Maius Castricius in Pompeii’, American Journal of Archaeology 114:1 (2010), 67–9; id. ‘Dialogues of Graffiti in the House of the Four Styles at Pompeii (Casa Dei Quattro Stili, I.8.17,11)’, in Ancient Graffiti in Context, eds. J. A. Baird and Claire

58

38. 39.

40.

41.

42.

43.

44.

45.

Ann Marie Yasin

Taylor (New York, 2011), 20–48; Peter Keegan, ‘Blogging Rome: Graffiti as Speech-Act and Cultural Discourse’, in Ancient Graffiti in Context, eds. J. A. Baird and Claire Taylor (New York, 2011), 165–90; Smith, ‘Walls Have Ears’. ¨ See Romer, ‘Griechischen Graffiti’. Sometimes in Greek and in reverse order (e.g. ICUR V.13052: ‘Παῦλε καὶ Πέτρε’). See n. 31. Carletti, ‘Nascita e sviluppo’, 149. Guarducci’s study suggested that Peter was nevertheless indicated crypographically in the ‘wall g’ graffiti by abbreviations and monograms (Graffiti sotto la confessione, v. 2). In addition, she sees the fragmentary ‘ΠΕΤ’ graffito from the nearby ‘red wall’ as clear indication of Peter’s name, although many others have been more circumspect, e.g. Ferrua, ‘Memorie dei SS. Pietro e Paolo’, 133–4. On the christograms, see Binsfeld, Vivas in Deo, 39–41. Mark Handley suggests that text and script similarities among the Trier graffiti might mean that they were carved by designated cathedral personnel or that they ‘reflect a desire to have one’s graffito conform to a perceived norm’ (‘Beyond Hagiography: Epigraphic Commemoration and the Cult of Saints in Late Antique Trier’, in Society and Culture in Late Antique Gaul: Revisiting the Sources. eds. Ralph W. Mathisen and Danuta Shanzer [Ashgate, 2001], 194, n. 37; reprised in Mark A. Handley, Death, Society and Culture: Inscriptions and Epitaphs in Gaul and Spain, AD 300–750 [Oxford, 2003], 171). Renate Pillinger, ‘Neue Entdeckungen in der sogenannten Paulusgrotte von Ephesos’, Mitteilungen zur christliche Arch¨aologie 6 (2000), 16–29; Binsfeld, Vivas in Deo, 93–5. Rachel Mairs, ‘Egyptian “Inscriptions” and Greek “Graffiti” at El Kanais in the Egyptian Eastern Desert’, in Ancient Graffiti in Context, eds. J. A. Baird and Claire Taylor (New York, 2011), 158. Such evidence has been examined for what it can tell us of patterns of long-distance travel, e.g. Eck, ‘Graffiti an Pilgerorten, 213–18; Carletti, ‘Scrivere i santi’,’ 351–5. From the south wall of the Hypostyle hall near the entrance, no. 205: ᾿Εγὼ Θεωδόσιος Νουβα; no. 208 (directly below inscription of Dioscoros above offering vessels held by king): ᾿Εγὼ Ε . . . Ν[ου]βα; no. 210 (immediately below graffiti of Sophronios = no. 209): Νουβα; 213 (before hands of king carrying offerings):+ ᾿Εγὼ ᾿Αάρων Νου/βα /

46.

47.

48. 49. 50.

51. 52.

53. 54.

δοῦλος / [᾿Ι](ησοῦ)ς (Bernand, Inscriptions grecques, 2:268–77). See Rutherford, ‘Island of the Extremity’. Carlo Carletti, ‘Testimonianze scritte del pellegrinaggio altomedievale in occidente Roma e l’Italia’, in Los muros tienen la palabra: materiales para una historia de los graffiti, eds. M. Gimeno Blay and Mar´ıa Luz Mandingorra Llavata (Valencia, 1997), 81–4; Carletti, ‘Viatores ad martyres’, 212–17. Mary Beard, ‘Writing and Religion: Ancient Literacy and the Function of the Written Word in Roman Religion, Question: What ˆ of Writing in Graeco-Roman Was the Role Paganism’, in Literacy in the Roman World, JRA Supplement 3, eds. Mary Beard, A. K. Bowman, M. Corbier, et al. (Ann Arbor, 1991), 35–58. Ibid., 47–8. Ferrua, ‘Graffiti di pellegrini’, 18. ¨ Romer, ‘Griechischen Graffiti’, 172, insc. 3 on wall fragment B with βοέ[θησον] for βοήθησον. ICUR V.12989. ¨ Romer, ‘Griechischen Graffiti’, 173, insc. 18: Συμεώ / Μάρας. On the tabula ansata frame used for divine invocations both on the Memoria Apostolorum triclia wall and in the nearly contemporary graffiti recording votives from the station of the VII Cohort of the Vigili, see Anna Holst Blennow, ‘The Graffiti in the Cryptoporticus of the Horti Sallustiani in the Area of the Embassy of the United States of America in Rome’, in Unexpected Voices: The Graffiti in the Cryptoporticus of the Horti Sallustiani and Papers from a Conference on Graffiti at the Swedish Institute in Rome, 7 March 2003, ed. Olof Brandt, Skrifter utgivna av Svenska Istitutet I Rom, 4o , 59 (Stockholm, 2008), 59. On tabulae ansatae in general, see Langner, Antike Graffitizeichnungen, 27. Cf. Blennow, ‘Graffiti in the cryptoporticus’, 60. Most of the site does not survive postexcavation. These panels, from the walls of hall 6, are preserved in the Coptic Museum in Cairo (reproduced in Gawdat Gabra and Mariane Eaton-Krauss, The Treasures of Coptic Art in the Coptic Museum and Churches of Old Cairo [Cairo, 2006], 88–9); others from the same room are in the collection of the Louvre. The room is described with excavation photos of the paintings in situ in Maspero, Fouilles execute´es a` Baouˆıt, 20–3, pls. 15– 25. It was a long (ca. 6 m × 30 m) rectangular chamber with a bench built along the south,

Prayers on Site

55.

56.

57.

58.

west and east walls; Delattre and Walters suggest that the dimensions indicate that it could have been a refectory or other kind of reception place for pilgrims (Delattre, Papyrus coptes et grecs du monast`ere, 50; C. C. Walters, Monastic Archaeology in Egypt [Warminster, 1974], 109). The tomb of the monastic founders, Apollo and his companion Phib, was presumably the goal of the pilgrimage (Delattre, Papyrus coptes et grecs du monast`ere, 53); 337 graffiti from the room, mostly in Coptic, have been published (Maspero, Fouilles execute´es a` Baouˆıt, 63–120). A plan of the complex can be found in Jean Cl´edat, Le monast`ere et la n´ecropole de Baouit, M´emoires publi´es par les members de l’Institut franc¸ais d’arch´eologie orientale du Caire, 111 (Cairo, 1999), plans I–IV. On Damasus’s epigraphy, see Epigrammata damasiana, ed. A. Ferrua (Vatican City, 1942); N. Gray, ‘The Filocalian Letter’, Papers of the British School at Rome 24 (1965), 5–13; Dennis Trout, ‘Damasus and the Invention of Early Christian Rome’, Journal of Medieval and Early Christian Studies 33:3 (2003), 517– 36. Cf. V´eronique Plesch, ‘Memory on the Wall: Graffiti on Religious Wall Paintings’, Journal of Medieval and Early Modern Studies 32:1 (2002), 182. In the interesting case of the tomb of a priest named Eulalios in the Domatilla catacomb at Rome, devotional graffiti provide evidence of popular recognition of a saint not officially recognized by the church: Philippe Pergola, ‘Le “saint” prˆetre Eulalios: un cas singulier de veneration a` la fin du IVe si`ecle’, in Quaeritur inventus colitur: miscellanea in onore di padre Umberto Maria Fasola (Vatican City, 1989), 2: 543–60. Following William Harris’s fundamental text (Ancient Literacy [Cambridge, Mass., 1989]), much important work has sought to refine our understanding of literacy in terms of a broader spectrum (e.g. the essays collected in Literacy in the Roman World, Journal of Roman Archaeology Supplement 3, eds. Mary Beard et al. (Ann Arbor, 1991); Rosalind Thomas, ‘Writing, Reading, Public and Private “Literacies”: Functional Literacy and Democratic Literacy in Greece’, in Ancient Literacies: The Culture of Reading in Greece and Rome, eds. W. A. Johnson and H. N. Parker (Oxford, 2009), 13–45; M. C. A. Macdonald, ‘Literacy in an Oral Environment’, in Writing and Near Eastern Society. Papers in Honour of Alan R. Millard, eds. P. Bienkowski, C. B. Mee, and

59

59.

60.

61. 62.

63.

64.

E. A. Slater (New York, 2005), 49–118. For more on the relationship between graffiti and literacy, see also Bagnall, Everyday Writing, 25–6; Baird and Taylor, ‘Ancient Graffiti in Context’, 9–11; Mairs, ‘Egyptian “Inscriptions”’, 162; Keegan, ‘Blogging Rome’; Handley, Death, Society and Culture, 170–1. On the question of whether the proskynema inscriptions at Philae were carved by cult personnel or the pilgrims themselves, see ´ Etienne Bernard, ‘R´eflexions sur les proscyn`emes’, in M´elanges Franc¸ois Kerlou´egan, eds. D. Conso, N. Fick, and B. Poulle (Paris, 1994), 51–3. I am grateful to David Frankfurter for this reference. On the concept of epigraphic environments, see Greg Woolf, ‘Monumental Writing and the Expansion of Roman Society in the Early Empire’, Journal of Roman Studies 86 (1996), 22–39, and Ann Marie Yasin, Saints and Church Spaces in the Late Antique Mediterranean: Architecture, Cult, and Community (Cambridge, 2009), esp. 101–2. Plesch, ‘Memory on the Wall’, 168. On the conversion of the Parthenon, the dating of which is disputed (at least by the end of the sixth century based on archaeological evidence but with literary sources possibly pointing to a date as early as the end of the fifth century), see Bryan Ward-Perkins, ‘ReUsing the Architectural Legacy of the Past, entre id´eologie et pragmatisme’, in The Idea and the Ideal of the Town between Late Antiquity and the Early Middle Ages, eds. G. P. Brogiolo and Bryan Ward-Perkins (Leiden, 1999), 225–44. Orlandos and Vranousis, Graffiti du Parth´enon; Kaldellis, Christian Parthenon, 74– 80. Sixty-five of the catalogued inscriptions include date formulae (Orlandos and Vranousis, Graffiti du Parth´enon, esp. 28–30). Kaldellis calculates that 104 of the inscriptions are prayers, generally addressed to God of the Theotokos, and sixty-four are epitaphs. There are also at least eighty incompletely published Christian graffiti from the walls of the Propylaia (Christian Parthenon, 75–7). Bernand, Inscriptions grecques de Philae, 2: 268– 77, 305–14. The conversion of the sanctuary of Philae to the Church of St. Stephen, dated between 535–7 based on Procopius (Bel. Pers. I.19.37), is commemorated in a heavily damaged foundation inscription engraved on the south face of the North Pylon directly east of the central door leading to the hypostyle hall (insc. no. 200, Bernand, Inscriptions grecques de Philae, 2: 251–6; see also Nautin, ‘Conversion

60

Ann Marie Yasin

du temple’, and Dijkstra, Philae and the End of Ancient Egyptian Religion. 65. Cf. Eck, ‘Graffiti an Pilgerorten’, 221–2. 66. Bernand, Inscriptions greques, 2: 268–74, plus the diagram at pl. 110. 67. Orlandos and Vranousis, Graffiti du Parth´enon, esp. 15, with plan indicating graffiti placement. Recall that the ‘back’ of the ancient temple on the west side turned into the church’s ‘front’ when the addition of the church’s eastern apse reversed the building’s orientation. It is also important to note the wide chronological range of the graffiti and that the Parthenon church presents a somewhat special case of official scribes: although 32 inscriptions provide only names and were most likely carved by pilgrims or other visitors, and more than 100 of them are prayers (see n. 63), as Kaldellis notes, ‘the

68. 69.

70. 71.

plurality of the inscriptions were carved by – or on behalf of – the temple staff’ (Christian Parthenon, 78). Orlandos and Vranousis, Graffiti du Parth´enon, 15. For a description of the chamber, see Ulbert, Resafa II, 62–8 and the useful overview of the site in Pauline Donceel-Voute, ˆ Les pavements des e´ glises byzantines de Syrie et du Liban. D´ecor, arch´eologie et liturgie, Publications d’histoire de l’art et d’arch´eologie de l’Universit´e catholique de Louvain 69 (Louvain-la-neuve, ¨ 1988), 1: 273–79. For the graffiti, see Romer ‘Griechischen Graffiti’. Ulbert, Resafa II, 43–60; Donceel-Voute, ˆ Pavements des e´ glises, 279. E.g. Carletti’s ‘messaggi unidirezionali’ (‘Scrivere i santi’, 333 and ‘Testimonianze scritte’, 73).

鵻 CHAPTER THREE

ERASURE AND MEMORY: AGHLABID AND FATIMID INSCRIPTIONS IN NORTH AFRICA 鵼 Jonathan M. Bloom

According to the thirteenth-century North African historian Ibn ʿIdhari, in 909 when the Fatimid imam publicly revealed himself as the caliph al-Mahdi at Kairouan (now in Tunisia), he ordered his name mentioned in the Friday sermon and inscribed on coins, the two standard vehicles for the expression of sovereignty in the Islamic lands. He also ordered that the names of all previous patrons be removed from ‘mosques, cisterns, forts, and bridges’ and replaced with his own.1 His erasure and subsequent relabelling of monumental inscriptions recall the case of the Abbasid caliph al-Maʾmun (r. 813–33) who tampered with the foundation inscription of the Dome of the Rock in Jerusalem, adroitly replacing the name of the original patron, the Umayyad caliph ʿAbd al-Malik (r. 685–705), with his own.2 It is difficult to imagine that either ruler believed that such a change would cause the public to forget the name of the founder.3 Rather, as the Swiss Arabist Max van Berchem suggested many years ago, the act was a prise de possession, ensuring to the new claimant the symbolic advantages of citation, that is, the benefits from any blessing the building might possess. Nearly a century and a half after the Fatimid ruler al-Mahdi, the scene would be replayed on the same stage when the Fatimid client, the Zirid governor al-Mu‘izz ibn Badis (r. 1016–62), returned to the Abbasid fold and anathematised his erstwhile overlords, removing their names from inscriptions on coins, flags, standards and buildings.4 Although no examples of the outright substitution of names on the model of al-Maʾmun’s action at the Dome of the Rock have survived in Tunisia, a wellknown example of simple erasure is a carved stone plaque set in the east fac¸ade of

61

62

Jonathan M. Bloom

the Great Mosque of Sfax (Fig. 16).5 Originally written in six lines of an elegant floriated and foliated kufic typical of the Fatimid or Zirid period, the inscription is missing several phrases and even one complete line, which have been carefully chiselled away. The date of the inscription, 378 spelled out in words at the bottom left, has not been defaced, so we know it records the restoration of the mosque carried out in 988 CE. Because this date falls during the reign of al-Mansur b. Buluggin b. Ziri (r. 984–95), the Zirid governor for the Egyptian Fatimid caliph al-ʿAziz (r. 975–96), the missing phrases were surely the names and titles of the patron and his overlord, as well as praises and blessings on them and their works. Once the Zirids had renounced their Fatimid suzerains, these particularly Shiʿi sentiments would have been deemed offensive and were removed. At the same time the obvious erasure surely called attention to the missing words, which most readers would in any case have found difficult to decipher because the inscription is written in a hard-to-read monumental script. But otherwise evidence of the erasures and substitutions mentioned in the sources has been difficult to find. A visit to Tunisia in May 2010 with a group of scholars from the Kunsthistorisches Institut, Florence, and Antony Eastmond’s invitation to participate in the Viewing Texts workshop at the Courtauld Institute in London the following month prompted this reconsideration not only of the evidence for the erasure of inscriptions in the Aghlabid and Fatimid monuments of Tunisia but also of the consequences of these visible and invisible erasures. We should always remember the fallacy of arguing from ignorance: absence of evidence is not evidence of absence. Of all the regions in the Islamic lands from which buildings survive from the ninth and tenth centuries, Tunisia presents some of the most varied structures in both number and quality.6 At least seven major buildings – four congregational mosques, two neighbourhood mosques, and a ribat (a sort of fortified monastery) – survive from the ninth century in the cities of Kairouan, Sousse, Tunis and Sfax; all these cities were ruled by the Aghlabid dynasty (r. 800– 909), who governed for the Abbasid caliphs resident in distant Baghdad. The most important is the Congregational Mosque of Kairouan, often known as the Mosque of Sidi ʿUqba, a companion of the Prophet Muhammad who supposedly founded it in the seventh century. The present building, however, dates from the ninth century when it was entirely rebuilt by the Aghlabid governors. Several stages of construction have been discerned, but most scholars date the major portion of the present building to a campaign by Ziyadat Allah (r. 817–38) in 836.7 The smaller Congregational Mosque of Sousse was built in 859, presumably after the mosque in the nearby ribat was deemed insufficient for the growing Muslim population of the city.8 The Congregational Mosque of Tunis, known as the Zaytuna, or Olive-Tree Mosque, is dated by inscription to the equivalent of 864–5.9 The style and techniques of construction at the Congregational Mosque of Sfax indicate that it was certainly begun in the ninth century, although it bears no foundation inscription with an exact date.10

Erasure and Memory

16. Carved stone plaque set in the east fac¸ade of the Great Mosque of Sfax, recording the restoration of the mosque in 378 H/988 CE. (Photo: Jonathan Bloom)

In the tenth and eleventh centuries these buildings were repeatedly modified and new ones constructed in such cities as Mahdia, founded on the coast in 917 by the Fatimid caliph al-Mahdi, and Sabra-Mansuriya, a palace city founded three decades later by his grandson al-Mansur in the outskirts of Kairouan, although the site has long been a ruin and the mosque no longer exists.11 Consideration of these Aghlabid and Fatimid buildings as an ensemble reveals where inscriptions are normally found and where they are apparently missing, suggesting that the rulers in medieval Ifriqiya, the region roughly corresponding to modern Tunisia, regularly removed or defaced the inscriptions of their predecessors. The surviving record, therefore, does not accurately reflect what had existed originally. Although no single building exhibits all the features of the ‘typical’ Aghlabid congregational mosque, the model can be reconstructed as a somewhat irregular quadrilateral stone building containing a large open court surrounded by an arcaded portico, with a deep hypostyle prayer hall on the qibla or Mecca-facing side. Minarets of varying form, ranging from massive towers to short bastions, are placed somewhere along the wall opposite the qibla.12 The arcades around the court and in the prayer hall are normally supported by a forest of reused late antique and early Christian columns retrieved from local buildings, although piers are used to support the arcades in Sousse and Sfax. The arcades in turn support walls on which rest flat roofs constructed of transverse wooden beams laid relatively close together, themselves supporting smaller wooden planks, and the whole covered with matting and a water-resistant plaster. The width of the

63

64

Jonathan M. Bloom

porticoes and aisles in the hypostyle halls is normally about 3–4.5 m (10–15 feet), necessitated by the limited length of the timbers locally available for the ceilings and roofs. The most important area of the mosque – the bay directly in front of the mihrab at the centre of the qibla wall where the wider central aisle leading from the courtyard intersects the wider aisle running transversely along the qibla wall – is typically covered by a elegant masonry dome that rises above the surrounding roof. The dome is often decorated on the interior with carved (and painted) designs and bands of inscriptions quoting the Qur’an and naming the patron. In addition, similar inscriptions are often found running across the fac¸ade of the prayer hall. In the tenth century, the Fatimids adapted this model for their mosque at Mahdia and presumably at Sabra-Mansuriya as well.13 Although they eschewed the Kairouan prototype of a massive minaret opposite the mihrab, they and the Zirids often added a monumental entrance in its place (e.g. Mahdia), as well as a ‘narthex-portico’ to the qibla fac¸ade of the courtyard of the congregational mosques they remodelled (e.g. Kairouan, Tunis, Sfax); the central bay of this element was marked by another dome, decorated in much the same fashion as the ante-mihrab dome. At the Congregational Mosque in Kairouan, the earliest in the series, the antemihrab dome bears an inscription painted in white and red on four wooden planks 15 cm wide, which are nailed to structural beams inserted in the wall above the great arches that support the zone of transition.14 The text, comprising verses 2:255 and 256 from the Qur’an (the ‘Throne’ and the following verse, celebrating God’s majesty), is written in a kufic script typical of the Aghlabid period. Because the inscription bears no historical information, the date of this dome has been a matter of some dispute: whereas K. A. C. Creswell attributed it to a restoration under Abu Ibrahim Ahmad in 862, Georges Marc¸ais believed more sensibly that it was part of the original work of his father Ziyadat Allah in 836 and that only the arcades lining the central aisle were added at the later date, presumably to strengthen the somewhat shaky structure. The niche of the mihrab is composed of twenty-eight beautifully carved marble panels set in a rectangular grid of four tiers of seven panels. The horizontal band between the first and second tiers is carved in relief with the entire short text of Qur’an 112 (Ikhlas, dealing with God’s oneness), followed by blessings on Muhammad to fill up the available space.15 A recently discovered inscription above the upper central panel that was hidden by the line of tiles along the lower edge of the mihrab’s semidome states that this was the work of Abu al-ʿAfiya, the Andalusi servant or slave (ghulam al-andalus¯ı ). Hiding this inscription seems not to have been a conscious attempt to erase the artist’s name, but rather a judgement that an artist’s signature was inappropriate in such a prominent and religiously significant place.16 In contrast, a large white marble plaque carved in low relief with a mihrab design has been inserted in the wall to the left of the mihrab (Fig. 17). It is inscribed in the upper right spandrel with a circular motif containing four words

Erasure and Memory

17. Carved marble slab to the left of the mihrab in the Great Mosque of Kairouan. (Photo: Jonathan Bloom)

written radially that share the letter mim: al-h.amd li’l-h.am¯ıd al-mabd¯ı al-ʿam¯ıd, ‘Praise to the Praised, the Creator, the Resurrector’.17 The phrase does not make much sense and seems to have been chosen for its epigraphic potential to make a radial design that share a common letter. Several scholars have noted the similarity between the radial inscription on the plaque and those on some Fatimid coins minted in Sicily, and Roy and Poinssot suggested many years ago that this magnificent slab of marble might have been brought to the Kairouan mosque from the long-destroyed Fatimid mosque of Sabra-Mansuriya, presumably as a

65

66

Jonathan M. Bloom

trophy after the Fatimids left the region. Nobody, however, has remarked that logic demands that the upper left spandrel of the plaque should have had a corresponding radial inscription. Access to the mosque is virtually impossible for non-Muslims, so it has been impossible for me to examine the plaque closely, but from a distance it appears that the slab has been gently abraded at the upper left to remove all traces of the corresponding inscription, which presumably would have offered praises of a Shiʿi nature on Muhammad and his family that later Sunni worshipers might have found objectionable. One may imagine that after the initial subtle alteration few worshippers would have noticed its absence, particularly because the surviving inscription is not only difficult to read but makes little sense. In this case, in contrast to the Sfax inscription, the low relief of the carving allowed the inscription (assuming it existed) to be abraded so gently that nobody noticed the loss. To judge from the example of the Aghlabid congregational mosques at Sousse and Tunis, inscriptions should also have run across the fac¸ade of the prayer hall at Kairouan, but that portion of the Aghlabid building is no longer visible because at some point in the late ninth or tenth century (the date is disputed) a ‘gallery-narthex’ – an arcade two bays deep – was added across the entire width of the building.18 Perhaps the remains of the inscription – presumably praising an Aghlabid patron – exists behind the masonry. If it was – or is – there, one may imagine that the builders of the narthex were not upset that their addition conveniently obscured an inscription presumably in praise of the Aghlabid ruler. This scenario would argue for a tenth-century date for the narthex under Shiʿi patronage. In turn, the central element of the narthex is the elegant and impressive Qubbat al-Bahu, a dome chamber resting on four arches. On its interior (Fig. 18), the walls above the arches support an octagonal zone of transition of eight arches; in the spandrels between them, brackets support small columns that ‘support’ a sixteen-sided zone on which rests the dome. Just above the columns and below the sixteen windows of the dome’s drum, there is an empty horizontal band, a sort of entablature, which would have been ideally suited for an inscription, although none exists there now. If this dome, along with the entire gallery-narthex along the court, was added by Fatimids or Zirids, did the now empty entablature once support the carved plaques of a foundation inscription commemorating their work that was deliberately removed either after the Zirids broke with the Fatimids or when the region returned definitively to Maliki law? Although no other evidence for such an inscription survives in Kairouan, the comparable dome at the Zatuna mosque in Tunis has a perfectly legible but mutilated foundation inscription in the equivalent place. The Zaytuna mosque in Tunis, begun in 864–5, has a much fuller complement of inscriptions than survive from the mosque at Kairouan. An inscription around the base of the dome in front of the mihrab is written in letters about 20 cm high and states, ‘In the name of God the Merciful the Compassionate. This is among the works that the imam [al-Must]aʿin bi’llah, Commander of the

Erasure and Memory

18. Kairouan, Great Mosque, Qubbat al-Bahu, interior. (Photo: Jonathan Bloom)

Believers, the Abbasi, seeking God’s reward in the desire that He will be satisfied with him, ordered by the hands of Nusayr his mawla, in the year 250 (864–65). O ye who believe, may justice rule your witnessing of God! [Qur’an 4:134] The work of Fath’.19 A similar inscription was written across the fac¸ade of the prayer hall under the present gallery-narthex, but some of the panels have disappeared, including the name of the founder, making the reading difficult.20 When the portico and dome were added, presumably under Zirid patronage, this Aghlabid inscription was mutilated and underwent transformations that inverted some of the panels. Nevertheless, the name of Nusayr and the date of 250 remain. It appears as if this frieze, which must have run from one end of the prayer hall fac¸ade to the other, was cut and remounted so it could fit into the dimensions of the gallery-narthex when it was added to the mosque.21 As at Kairouan, the central element of the portico is an elegant domed chamber known as the Qubbat al-Bahu. An inscription dated 991–2 runs around the interior of its base, but the name of the sovereign has been deleted, leaving a visible gap in the inscription – although we know from historical sources that in that year al-Mansur b. Buluggin b. Ziri governed the region for the Fatimid caliph al-ʿAziz in Cairo.22 Presumably the name was effaced after the Zirids fell from power. Another inscription on the impost blocks above two groups of columns supporting the dome states that four workmen – Ahmad al-Burjini, Abu al-Thana, Abdallah b. al-Qaffas, and Bishr b. al-Burhini – supervised the project. This text also indicates that the work

67

68

Jonathan M. Bloom

included the construction of the vaulted basement, the galleries and the dome; it also states the duration of the work, from Rabi I 380 (May–June 990) to Jumada I 385 ( June 995), when the Zirid was still in power. Considering the formulaic nature of Islamic inscriptions (discussed in Chapter 11 by Sheila Blair), it is curious that the patron who paid for the work is nowhere mentioned – suggesting once again that later patrons might have erased inconvenient reminders of past acts that might have been inscribed on the other impost blocks supporting the dome.23 Virtually every other Aghlabid building bears an inscription. The earliest, the Ribat at Sousse, has a small marble plaque over the door to the tower engraved with a text in kufic: ‘In the name of God the Merciful the Compassionate. Blessings from God. Among the things that was ordered by the amir Ziyadat Allah b. Ibrahim, may God prolong his life, by the hands of Masrur, the servant, his mawla, in the year 206 (821)’.24 The small neighbourhood mosque at Sousse known as the Bu Fatata Mosque, one of the earliest surviving examples of a nine-bay mosque plan, has a large inscription across the fac¸ade written in kufic letters in relief: ‘In the name of God the Merciful the Compassionate. Blessing from God and mercy on . . . . his mawla. Al-Aghlab b. Ibrahim built this mosque to proclaim the name of God, by the hands of . . . . h.dr [?] mawla of the amir’. Although the mawla‘s name is unclear, the amir al-Aghlab ruled from 838–41, so the mosque can be dated to those years. In Sousse, the Congregational Mosque has a kufic inscription carved in relatively high relief running around the courtyard, beginning under the narthexportico that was added to the mosque in the seventeenth century. The inscription states that the amir Abu’l-Abbas b. al-Aghlab built it in the year 236 (850–1) by the care of his servant (khadim) Mudam.25 The inscription continues around the courtyard with several Qur’anic quotations, including 7:43, 15:49, 9:18, and 112. Over the main doorway to the prayer hall (Fig. 19), however, the text is interrupted by a later addition in lower relief, which presumably replaces an earlier patron’s name that would have occupied the most prominent position as one entered the mosque. Finally, the Mosque of the Three Doors at Kairouan, dated to 866, has several lines of inscription across the fac¸ade carved in relief: the text comprises the basmala (invocation) and several Qur’anic verses (33:70–71 and 30:3) followed by a foundation text: ‘The construction of this mosque was ordered by Muhammad b. Khayrun al-Maʿafiri, the Andalusian, in order to get closer to God and in the hope of His forgiveness and mercy’.26 Although the inscription has been somewhat garbled when the blocks on which it was inscribed were reset during some earlier restoration, it has not been defaced, presumably because it was a private foundation and did not mention any sovereign. Given this extensive evidence for the presence and importance of inscriptions in the Aghlabid buildings of Tunisia, it is all the more surprising that the early Fatimid mosque at al-Mahdia, built in 917 generally following the plan of the

Erasure and Memory

19. Sousse, Great Mosque, replacement inscription over portal. (Photo: Jonathan Bloom)

Congregational Mosque of Kairouan, bears no inscriptions at all. The absence of inscriptions is striking because decoration with inscriptions increased in popularity throughout the Islamic lands in the tenth century and subsequent Fatimid mosques in Egypt had lavish epigraphic decoration.27 The Mahdia mosque was entirely reconstructed in the 1960s, but absolutely no evidence of inscriptions was found during the archaeological investigation of the site then, suggesting that all the Fatimid-era inscriptions had been deliberately removed at some time during the building’s long and complicated history.28 The court fac¸ades have been entirely reconstructed several times over the centuries, so it is difficult to say whether there might have once have been inscriptions across the prayer hall, as at Tunis or Sousse. The blank entablature running around three sides of the projecting portal, the only portion of the mosque not rebuilt, would have been particularly appropriate for an inscription, as would the now-blank lunette over the main entrance (Fig. 20). A beautiful white marble plaque inscribed ‘Enter it in peace, forebearing [from fear; Qur’an 15:46]’ in an elegant foliated kufic script, and now in the Museum of Islamic Art at Raqqada (Fig. 21), is said to have come from the fourteenth-century Zawiyat al-Gharyaniyya in Kairouan, but the style of script surely indicates that it must date to the Fatimid or Zirid period.29 The text is perfectly appropriate for use on a gateway or portal, and it may have been removed from one of the gates of Sabra-Mansuriya. Whatever its origin, the plaque gives us some idea of the style of inscription the Fatimids might have used on the portal to the Mahdia mosque; the text would undoubtedly have followed standard protocol, invoking blessings on the Fatimid ruler and the Prophet’s family and concluding with the date.30

69

70

Jonathan M. Bloom

20. Mahdia, Great Mosque, portal showing empty entablature. (Photo: Jonathan Bloom)

The repeated erasure and replacement of inscriptions from the buildings of ninth- and tenth-century North Africa are unusual, if not quite unprecedented, in the history of Islamic architecture, demonstrating that, far from being merely symbolic affirmations of the faith, these communications were ‘read’ and interpreted by contemporary viewers.31 The strong partisan positions demonstrated by these repeated mutilations and substitutions undoubtedly had their origins in the complex political and religious history of the region. During the Aghlabid period Kairouan was a centre of Sunni and particularly Maliki learning; however, once the Fatimids replaced the Aghlabids in the tenth century, the new rulers deemed the Aghlabid inscriptions praising their Sunni Abbasid overlords to be deeply objectionable and ordered them defaced, because the Shiʿi Fatimids denied the Abbasids’ claim as legitimate successors to the Prophet. After the Sunnis regained power in Ifriqiya, they in turn deemed Fatimid and Zirid inscriptions praising the Fatimid caliphs and the family of the Prophet to be equally objectionable and ordered them defaced. In contrast, in Egypt, where the Fatimid

Erasure and Memory

21. Inscribed marble plaque found in the Zawiyat al-Gharyaniyya at Kairouan. Raqqada, Museum of Islamic Art. (Photo: Jonathan Bloom)

rulers moved in 973 never to return to North Africa and where they were similarly replaced by Sunnis after Saladin extinguished the dynasty in 1171, surviving Fatimid buildings and inscriptions in Cairo show no obvious signs of the kind of mutilation that appears to have been relatively common in Ifriqiya. For example, al-Azhar, the Fatimid mosque built in 970 that became the centre of the dynasty’s intellectual life, is today revered as the epicentre of orthodox Sunni learning, and many of its Fatimid-era inscriptions remain – although it must be admitted that the original exterior of the building, where other inscriptions might have existed, is entirely hidden behind later additions. Nevertheless, a Fatimid building such as the al-Aqmar mosque, erected on the main street of Fatimid Cairo, still preserves its kufic inscriptions that heap praise on the Fatimid caliph and his vizier, Qur’anic verses mentioning ‘the People of the House’ (which the Shiʿa interpret to be a unequivocal reference to the family of ʿAli), and epigraphic designs worked around the names of Muhammad and ʿAli.32 It is, however, possible that a circular space above the blind arch on the right side of the fac¸ade (which has now been restored with an innocuous rosette) once contained an epigraphic device that later viewers might have found objectionable. Moreover, the puzzling absence of contemporary mihrabs from the major Fatimid mosques in Cairo (Fatimid-era mihrabs still survive in out-of-the-way shrines) might suggest that they too had to be replaced because of the objectionable sentiments their inscriptions expressed.33 To return to North Africa, the examples discussed reveal four ways in which inscriptions appear to have been erased and replaced during the tenth and eleventh centuries. In a case like the Kairouan mihrab signature, the inscription was simply hidden by subsequent additions, perhaps because the signed panel had been reused in some way. In the cases of the fac¸ades at Kairouan and Tunis, inscriptions were obliterated or defaced ‘accidentally on purpose’ by constructing something to obscure them from view; in other instances, such as the Kairouan dome or the Mahdia portal, presumed inscriptions on plaques were simply removed, leaving blank fields that only comparison with other buildings reveals what should once have been there. Finally, the cases of the plaque at Sfax,

71

72

Jonathan M. Bloom

the mihrab at Kairouan and the Qubbat al-Bahu at Tunis show more or less obvious erasures of offending words and phrases by subsequent viewers. For the first three types of erasures, later viewers hardly knew that the original inscriptions existed, whereas the fourth type must have drawn attention to the spaces where words or phrases are missing and led the viewer to puzzle out what might once have been there. If that is true, the mutilators’ desire to erase the memory of the past and take possession of it might have been better served by ‘leaving well enough alone’.34 Unlike Ayyubid Egypt, which seems to have sought to forget the past by simply ignoring it, the erasures and gaps in this last group of examples acted as a visible and a permanent reminder of the unacceptability of earlier regimes and their beliefs. The absence of words and inscriptions, therefore, could provide as much meaning as their presence. These inscriptions also reveal the extraordinary power and importance of writing in this region. The Roman province of Africa Proconsularis, corresponding roughly to modern Tunisia, had been heavily Romanized, and with the arrival of Christianity, Latin continued to be the language of epigraphy, although the formal quality of the writing declined. The Arabs who conquered the region in the seventh century certainly found Latin inscriptions everywhere and probably a large number of Latin speakers. In the twelfth century, for example, the Arab geographer al-Idrisi reported that Ifriqiyan Latin was still spoken at Gafsa.35 The early Arab coinage minted at Kairouan is inscribed in Latin, not Arabic, and gives the profession of faith in an abbreviated Latin curiously reminiscent of Arabic, in which only the consonants and long vowels are written.36 Arab builders carefully reused blocks inscribed in Latin in their buildings, and it is no accident that the lower courses of the Kairouan minaret are composed of massive blocks inscribed in Latin but carefully set upside down, as if to proclaim that the old order had been reversed. The fortress at Gafsa, in the south of Tunisia, has so many reused Latin inscriptions that a nineteenth-century French traveller called it a virtual ‘mus´ee e´ pigraphique.’37 Aghlabid Kairouan is also believed to have been an important centre of manuscript production, although the first manuscripts that can be ascribed with certainty to the city or region date from the eleventh century. Already in the thirteenth century the Kairouan mosque preserved an important collection of early manuscripts and bindings, and many scholars believe that the magnificent ‘Blue Qur’an’ was copied there in the mid tenth century under Fatimid patronage.38 All of this suggests that the cities of Ifriqiya, whether under Aghlabid or Fatimid rule, were the most important centres of literacy and writing between the Nile and the Atlantic in the early medieval period. In the wider Islamic lands, this was the period when the majority of the population converted to Islam, and truly Muslim societies – whatever their sectarian affiliations – emerged.39 And in medieval Kairouan religious scholars devoted their energies to studying the variant readings of the Qur’an and other such matters.40

Erasure and Memory

73

The widespread use of public writing in Ifriqiya already in the ninth and tenth century casts doubt on Irene Bierman’s hypothesis about the development of a distinctly Egyptian ‘Fatimid public text’,41 because these North African inscriptions predate the advent of the Fatimids by decades and have nothing at all to do with Egypt. Bierman argued that the Fatimid rulers of Egypt were the first to use ‘public texts’ on buildings and textiles to present their own distinct ideology to the diverse members of Cairene society. Because the Fatimids were in hiding from their Abbasid rivals before they revealed themselves in early tenthcentury North Africa, they must have learned about the use of ‘public texts’ from their Aghlabid predecessors. Furthermore, monumental epigraphy was used not only in ninth-century North Africa but also in many Muslim lands from the Iberian peninsula to Iran to decorate mosque portals and fac¸ades.42 In the case of North Africa, a far more intriguing connection might be to the revival of public writing in early medieval Italy, where the presence of Latin inscriptions inspired medieval writers to create their own monumental inscriptions on walls and over gateways.43 Could it be that the Arabophone peoples of Ifriqiya were similarly inspired by the region’s rich heritage of Latin monumental epigraphy – whether or not they could actually read the texts – not only to decorate their buildings with monumental inscriptions in Arabic but also to care very much about what the inscriptions actually said? The repeated erasure and replacement of monumental inscriptions in medieval North Africa also underscore the importance of appreciating the impermanence of the ‘permanent’ and realizing that much is missing from the historical record. Although Ifriqiya preserves an unusually rich corpus of buildings and inscriptions from the ninth and tenth centuries, particularly in comparison to other regions of the Islamic lands, the historical record is always incomplete. We can, however, use our knowledge of what should exist, rather than simply studying what does exist, to begin to understand what has vanished from the historical record and why.

NOTES

1. Ibn ‘Idh¯ar¯ı al-Marr¯akush¯ı, Kit¯ab al-Bay¯an alMaghrib (Histoire de l’Afrique du nord et de l’Espagne musulmane), eds. G. S. Colin and ´ L´evi-Provenc¸al (Leiden, 1948), 1:159; E. Jonathan M. Bloom, ‘The Origins of Fatimid Art,’ Muqarnas 3 (1985), 21. 2. Max van Berchem, Mat´eriaux pour un Corpus Inscriptionum Arabicarum II: Syrie du sud: Jerusalem (Cairo, 1920–27), 2:228–42. 3. For example, in 985, nearly two centuries after al-Ma’mun tampered with the inscription on the Dome of the Rock, the Palestinian geographer al-Muqaddasi correctly identified the original patron of

the building as the Umayyad caliph ʿcor al-Malik. al-Muqaddas¯ı, Ah.san al-taq¯as¯ım f¯ı maʿrifat al-aq¯al¯ım, ed. M. J. de Goeje, Bibliotheca Geographorum Arabicorum (Leiden, 1967), 159; al-Muqaddas¯ı, The Best Divisions for Knowledge of the Regions [Ah.san alTaq¯as¯ım f¯ı Maʿrifat al-Aq¯alim], trans. Basil Collins, reviewed by Mohammad Hamid Altaʾi (Reading, 1994), 135. 4. Ibn ‘Idhari, Bay¯an, 1:277. 5. Georges Marc¸ais and Lucien Golvin, La Grande Mosqu´ee de Sfax, Institut National d’Arch´eologie et Arts de Tunis, Notes et Documents, vol. iii (Tunis, 1960); Lucien Golvin,

74

Jonathan M. Bloom

6.

7.

8. 9.

10. 11.

12.

13. 14.

15. 16.

17. 18.

19.

Essai sur l’architecture religieuse musulmane (Paris, 1970–9), 164; Jonathan M. Bloom, Arts of the City Victorious: Islamic Art and Architecture in Fatimid North Africa and Egypt (London, 2007), 8. The most convenient introductions to the architecture of this period remain K. A. C. Creswell, Early Muslim Architecture, Vol. II (Oxford, 1940) and Georges Marc¸ais, Architecture musulmane d’occident (Paris, 1954). Additional information on Aghlabid architecture is to be found in Golvin, Essai, vol. III. Paul Sebag, The Great Mosque of Kairouan, trans. Richard Howard, Andr´e Martin, photographer (New York, 1965). Alexandre L´ezine, Sousse: Les Monuments musulmans (Tunis, 1968). ¯ ‘Abd al-Az¯ız al-Dawl¯atl¯ı, Al-Zaytuna: ‘Ashra ¯ min al-fann al-mi‘m¯ar¯ı al-Tunis¯ ¯ ı (Tunis, qurun 1999). Marc¸ais and Golvin, Mosqu´ee de Sfax. Alexandre L´ezine, Mahdiya: recherches d’arch´eologie islamique, Arch´eologie M´editerran´eenne (Paris, 1965). Jonathan Bloom, Minaret: Symbol of Islam, Oxford Studies in Islamic Art, VII (Oxford, 1989). Jonathan M. Bloom, ‘Origins.’ Creswell, Early Muslim Architecture, 2:315; B. Roy and P. Poinssot, Inscriptions arabes de Kairouan (Paris, 1950–58), 9–11. Roy and Poinssot, Inscriptions, 15; Golvin, Essai, III: fig. 90. The presence and subsequent concealment of the signature raise the possibility, first mentioned by medieval chroniclers, that the panels were originally intended for some other purpose. See Golvin, Essai, III: 228– 43. I presented a study of these panels, ‘The Marble Mihrab Panels of the Great Mosque of Kairouan’ at ‘The Aghlabids and Their Neighbours’, held at the UNC Centre in London on May 23–24, 2014; the proceedings are to be edited and published by Glaire Anderson, Corisande Fenwick and Mariam Rosser-Owen. Roy and Poinssot, Inscriptions Arabes de Kairouan, 16. See the inconclusive discussion in Faouzi Mahfoudh, Architecture et urbanisme en Ifriqiya m´edi´evale: proposition pour une nouvelle approche (Manouba, 2003), 84–5. ´ Etienne Combe, Jean Sauvaget and Gaston Wiet, R´epertoire chronologique d’´epigraphie arabe (Cairo, 1931) [hereafter RCEA], II: 75, no. 505; S. M. Zbiss, Corpus des Inscriptions

20. 21.

22.

23. 24. 25. 26.

27.

28. 29. 30. 31.

32. 33.

34.

Arabes de Tunisie (Tunis, 1955), 27; Marc¸ais, Architecture musulmane d’occident, 7 n. 5. Zbiss, Corpus, 27–8; Golvin, Essai, III: 153. Golvin, Essai, III: 153. Photos of the inscription can be found at www.epigraphieislamique.org/epi/picture.php?pho_refext= 8415 (accessed 10 July, 2013). Zbiss, Corpus (Tunis, 1955), 38–9; Golvin, Essai, III: 154. Because non-Muslims are not allowed to enter the mosque beyond one of the side porticos, I have not seen the inscription myself, but its placement is shown in al¯ Dawl¯atl¯ı, Al-Zaytuna, fig. 77. These inscriptions are illustrated in al¯ Dawl¯atl¯ı, Al-Zaytuna, 91–6. Combe, Sauvaget and Wiet, RCEA, no. 143. Golvin, Essai, III: 212. Combe, Sauvaget and Wiet, RCEA, II, no. 549; G. Kircher, ‘Die Moschee des Muh.ammad b. Hairun ¯ (‘Drei-Tore-Moschee’) in Qairaw¯an/Tunesien. Erster Bericht,’ Mitteilungen des Deutschen Arch¨aooglischen Instituts Abteilung Kairo 26 (1970), 141–68. Bloom, Arts, 29–30 and chapter 3; Jonathan M. Bloom, ‘The Mosque of al-Hakim in Cairo,’ Muqarnas 1 (1983), 15–36. L´ezine, Mahdiya. Zbiss, Corpus, 78. Bloom, Arts, 39; Zbiss, Corpus Des Inscriptions Arabes de Tunisie, 77–8, no. 37. Richard Ettinghausen, ‘Arabic Epigraphy: Communication or Symbolic Affirmation?’ in Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles, ed. Dikran Kouymjian (Beirut, 1974), 297–317. Bloom, Arts, 139–45. Jonathan M. Bloom, ‘Fatimid Mosques,’ in The Cambridge World History of Religious Architecture, ed. Richard Etlin (Cambridge, forthcoming). The erasure of inscriptions from Tunisian buildings in the tenth century finds a remarkable parallel in a spectacular rectangular box and lid carved out of a single piece of ivory that is now in the David Collection, Copenhagen. The lid of the box, like many other ivory boxes made in the second half of the tenth century at Madinat al-Zahra and ´ Cordoba, once bore an Arabic inscription that was effaced, presumably when the box was repurposed as a Christian reliquary. For the David box and the inscriptions on related pieces, see The Ivories of Muslim Spain, Papers from a symposium held in Copenhagen from the 18th to the 20th of November 2003, in Journal

Erasure and Memory

of the David Collection, vol. 2/1–2 (2005), especially Sheila S. Blair, ‘What the Inscriptions Tell Us: Text and Message on the Ivories from al-Andalus,’ 75–99. 35. The Encyclopaedia of Islam, New Edition, ed. H. A. R. Gibb et al. (Leiden, 1960), s.v. ‘K . afsa’. See also Walter E. Kaegi, Muslim Expansion and Byzantine Collapse in North Africa (Cambridge, 2010). 36. J. Walker, Arab-Byzantine and Post-Reform Umaiyad Coins, Catalogue of the Muhammadan Coins in the British Museum (London, 1956), 54, no. 143 See also www .britishmuseum.org/research/search_the_ collection_database/search_object_details .aspx?objectid=902982&partid=1&search Text=gold+North+Africa+coin&numpages= 10&orig=%2fresearch%2fsearch_the_ collection_database.aspx¤tPage=4 37. Norman Douglas, ‘Stones of Gafsa,’ North American Review 194 (November 1911), 51.

75

38. For a recent summary, see Jonathan M. Bloom, ‘The Blue Koran Revisited,’ conference presentation, Codicolog´ıa e Historia del ´ Libro Manuscrito en Carateres Arabes (Madrid, 2010) For an entirely different interpretation of the date and place of production, see Alain George, ‘Calligraphy, Colour and Light in the Blue Qur’an,’ Journal of Qur’anic Studies 11 (2009), 75–125. 39. Richard W. Bulliet, Conversion to Islam in the Medieval Period: An Essay in Quantitative History (Cambridge, MA, 1979). 40. Sheila S. Blair, Islamic Calligraphy (Edinburgh, 2006), 173ff. 41. Irene A. Bierman, Writing Signs: The Fatimid Public Text (Berkeley, 1998). 42. A convenient introduction to the subject is Sheila S. Blair, Islamic Inscriptions (Edinburgh, 1998). 43. Armando Petrucci, Public Lettering: Script, Power, and Culture, trans. Linda Lappin (Chicago and London, 1993).

鵻 CHAPTER FOUR

TEXTUAL ICONS: VIEWING INSCRIPTIONS IN MEDIEVAL GEORGIA 鵼 Antony Eastmond

High on the east fac¸ade of the cathedral of Kumurdo in Georgia is a short and apparently simple inscription: ‘Christ, have mercy on the bishop Iovane, builder of this [church], on that day [of judgement]’ (Fig. 22).1 Inscribed in 964 CE in the Georgian asomtavruli (majuscule) alphabet, the lettering has been laid out with particular care. The carving is in relief and the letters are neatly blocked out, carefully separated and distinguished one from another. They are stately and upright; there is no sense of hurry or attempt at vigour or animation among the letters. The inscription is nearly 5 m long, and the letters are approximately 40 cm high. Even though the relief is shallow, the letters are crisp and stand out particularly well in the strong light of the Javakheti plateau in southern Georgia. The impact of the inscription is heightened by the severity of the barely articulated fac¸ade, which accentuates the few areas of decoration.2 The tall, thin form of the letters and the absence of serifs match the austerity of the architecture. Even from a distance every letter can be made out; this was clearly an inscription that was designed to be legible. The question is: in what ways was it read? The visibility of the inscription at Kumurdo and the prominence, care and beauty with which it was executed lie at the heart of this chapter. This text and another twenty-seven texts that were inscribed on the monument in the first fifty years of its existence have long been known and studied by historians. However, their interest in these inscriptions has focused entirely on their contents. They have mined them for their positivist value, their ‘facts’ about the social, political and economic history of the cathedral. These inscriptions have also been linked to outside referents: historians have argued that as they seem to be transcriptions of texts in other media they are too partial to study on their own, now those

76

Textual Icons

22. Kumurdo cathedral ( Javakheti, Georgia). East fac¸ade, 964. (Photo: Antony Eastmond)

77

78

Antony Eastmond

‘originals’ are lost.3 The texts have not been studied as material objects, as images of words as much as words themselves. This chapter considers the visual context of monumental texts in Georgia in the decades around the year 1000 CE. It raises questions about the ways in which such texts may be viewed as much as read, and it suggests that, in many ways, their form superseded their contents. There has been a growth of interest in recent years in the ways in which inscriptions interacted both with the monuments on which they were placed and with the people who encountered them. Work on Byzantine inscriptions has stressed their performative nature: as readers walk around the exterior and interior of the churches of H. Polyeuktos (524 CE) and SS. Sergios and Bakchos (530 CE) in Constantinople, or encircle the apse of the church of the Virgin of Skripou in Boeotia, Greece (874 CE), they are made to enact the prayers the inscriptions contain.4 These arguments have become entwined in questions about whether and how such texts may have been read and by whom.5 The presence of monumental inscriptions in Fatimid Cairo has been explained in terms of the use of language to demarcate space,6 and those of early Armenia have been presented as assertions of permanence, trust and legitimacy in an era of conflict.7 The inscriptions of Romanesque France have been described simply as ‘stone charters’.8 I argue that in Georgia around the year 1000 inscriptions held a particular function as visual objects: designed as much to be seen as to be read, they carried meanings beyond those conveyed by their contents. These meanings were determined by the ways in which inscriptions were presented as blocks of text and as carefully defined ornamental elements in the decoration of churches, and by the way in which they were located in specific areas of churches in this period. I propose that as visual objects, what I refer to as ‘textual icons’, their overall form carries the same visual power and figural correspondence as icons in the Orthodox world. The blocks of text acted as mnemonic devices to imprint an icon of prayer on the mind of the viewer, and images were long regarded as the most effective way of framing words and memories in the medieval mind.9 The visual presence and consequent function of inscriptions also had an impact on the types of text that were chosen to be inscribed in monumental form.10 It is these varieties that the rest of this chapter presents. The idea of a text as a visual prayer is neatly encapsulated in Iovane’s foundation text on the east fac¸ade of Kumurdo cathedral. Although its individual letters are legible, the presentation of the prayer itself presents obstacles that impede attempts to read it. There is no punctuation or separation between words, and there is an extensive use of abbreviations. This is evident in the difference in length between the text as inscribed: ddddddddddddddddddddddddddddddd

and in its fully expanded form: dddddd, dddddddd dddddd dddddddddd, ddddd dddddddddd ddddd ddd11

Textual Icons

However, the complexity of expanding the text to separate words, supply missing elements and so make a grammatical sentence is not necessarily a hindrance to comprehending it as a prayer. The overall text follows a stock formula for an intercessory prayer, and the abbreviations employed are standard, not least for frequently used words such as dd for d(dddd)d (Christe) and ddd for d(dd)d(ddd)d (ˇseicqale ‘have mercy on’). Indeed, the phrase ‘Christ have mercy on’ is commonplace. It was often inscribed on churches and finds exact parallels (including the same abbreviations) on other Georgian monuments.12 The phrase appears in all media. In the magnificent three-volume manuscript of the Old Testament (now Mount Athos, Iviron MS Georg. 1) commissioned by the Georgian monk Tornik in 978, a decade after Kumurdo was built, the phrase appears no less than ten times with minor variations, interspersed throughout the 960 folios of the text, occurring each time at the end of a book of the Old Testament (along with other comments that underline the spiritual and physical effort involved in copying out such a huge text).13 Similar intercessory texts also appear on icons, such as that added to the repouss´e gold cover for an icon of the Mother of God from Khobi. This was commissioned by the king of Abkhazeti, Leon III Anchabadze (957–67), in whose territory Kumurdo stood. The phrase appears on the upper edge of the lower part of the frame, immediately below the figure of the Virgin: ‘Ever-pure Mother of God, be intercessor before Christ for the soul of king Leon’.14 As with Kumurdo cathedral and the Old Testament manuscript, the efficacy of the prayer lies in the combination of text and monument/object. In all cases, the inscriptions work as part of a larger ensemble, and the value of their prayers is demonstrated by the magnificence of the material object to which they are attached. Because they are inseparable they work together to establish the form of prayer and the name of the principal intended recipient of it. However, whilst readers had to search through the many folios of the manuscript to find Tornik’s prayers, and viewers must study the icon closely to separate the ornate, serifed letters of the inscription from the dense repouss´e interlace that subsumes it on the golden frame, the apse inscription at Kumurdo is all too visible. Even to someone with no knowledge of Georgian, the different letters are clear and easily differentiated. These features all suggest that clarity and visibility were paramount. Even if you cannot read the inscription, you are meant to see it, and to know that it could be read. The brevity of the text and the common nature of its intercessory formula suggest that, even if it could not be deciphered as words (i.e. read by a literate viewer), it could be recognised as a phrase, a form of contextual reading: the shapes of the first letters and the general length of the text enable it to be memorised or recognised as a particular form of intercessory prayer.15 Although nonliterate readers may not have been able to decipher the name of the donor, they would have recognised the text as the prayer of particular individuals, whose names would be known to God. The common use across the Orthodox world of

79

80

Antony Eastmond

23. Kumurdo cathedral ( Javakheti, Georgia). South porch, main tympanum. Inscriptions nos. 1 & 2, 964. (Photo: Antony Eastmond)

deliberately anonymous inscriptions, such as ‘As a prayer for one whose name God knows’, indicates that the presence of a prayer was often more important than knowledge of its donors’ identities.16 The prayer is the block of text, as much as the individual words themselves. This is evident from the way in which the text at Kumurdo had to compromise its symmetry to ensure its recognition as a prayer. It was clearly carved in situ (note the way that the bottoms of the letters d, d and d at the end of the inscription are cut into the stone that forms the head of the right-hand niche) and overran its originally allotted space in the centre of the fac¸ade. The imbalance of the text suggests that even this recognisable formula could not be made any shorter without compromising its comprehensibility. Alongside the east fac¸ade inscription at Kumurdo, we must consider two more foundation texts. Both are found around the main entrance, on the south side of the cathedral, where they were carved on a single vast monolith, set over the door (Figs. 23 & 24). The inscriptions are carved in relief in the same way as the east fac¸ade inscription, although using less elongated, serifed letter forms. If one of the oddities about the east fac¸ade inscription was the way its original text clearly overran, so that the neat symmetry of the letters between the two niches was lost, here the two texts were clearly designed together to fill the space. They are both frame and text, with the majority of the tympanum given over to a basket-like latticework of stars. The abbreviations across both inscriptions are reasonably even, suggesting that they were planned to fill the spaces allotted to them.

Textual Icons

24. Kumurdo cathedral ( Javakheti, Georgia). South porch, main tympanum. Inscriptions nos. 1 & 2, 964. (Drawing: Antony Eastmond)

The first, which fills the upper part of the semi-circle and the left side, gives full details of the foundation of the cathedral: With God’s help, the bishop Iovane laid the foundation of this church by my hand – of the sinful Sakostari, in the time of king Leon – may he be glorified by God – in koronikon 184 [= 964 CE] the first of May, Saturday, at the new moon, when Zvia was eristavi; this foundation was laid by him. Christ be a fellow fighter to your slave, amen.17

The accuracy of the timing seems to be reflected in the precision of the lettering. The other inscription, to the right of the central grid, is equal in grandeur but more humble in tone, seeking worshippers’ prayers for the builders of the cathedral: Holy church, protect and pardon your servants and all your builders on that Day of Judgement; mention the wretched Giorgi in your prayers.18

The clarity and simplicity that are visible in the text of the apse inscription are less evident in those around the main door, muddied by the need to acknowledge others, as the king, Leon III, and the local nobleman, Zvia, both receive an honourable mention, along with the architect/designer, Sakostari. Even the chief mason, Giorgi, is awarded a memorial. Nevertheless, the inscriptions bind the church to these individuals. There is no ambiguity about ownership and power and their relationship to piety. The foundation texts reveal a tension between legibility (clarity of letter forms) and readability (the ability actually to read and make sense of the texts). The contents of the two inscriptions in the porch are more particularised than the formulaic apse inscription, and they are full of abbreviations, contractions and

81

82

Antony Eastmond

elisions that are not marked. These features compress the texts considerably: the main inscription has lost nearly one third of its letters (down to 167 from 237). It is not easy to separate or decipher the various words: there is no punctuation or division between words, and words are even subdivided and overspill from one line to another. This is most evident in the first inscription. The indication of the date is divided between the end of the first (upper) part of the inscription and the first two lines of the second (left side) part. The first block ends with the word koronikonsa [in the year . . . ], and the second block starts with the alphanumeric date, itself split over two lines: d [rae = 100] on one line and then d [par = 80] and d [don = 4] on the next (i.e. 184 + 780 = 964). Viewers reading the inscription starting with the left-hand block are faced with a meaningless jumble of incomprehensible letters rpd. Given the importance of this carving as the authenticating inscription of the church, entrenching its foundation in history and memory, this is a curious but crucial place to break the text. (The final line of the first section of the inscription could easily have been changed to make space for the date by altering the design of the two stars on either side.) It is evident that the overall appearance of the text took priority over its comprehensibility, and that the tympanum works most successfully as a cohesive whole rather than as a compilation of elements. The length and detail of the texts seem to replicate that of the colophons of manuscripts in this period: the main colophon in Tornik’s Old Testament fills four folios and is supplemented by a second page-long colophon.19 Three points emerge from the viewing of these main foundation texts. The first is their aura of readability, which is more a visual fiction than a practical application: they are designed to look as if they can be read. The second is the visual nature of their design. Rather than giving the whole of the tympanum over to the inscription (as happens elsewhere, as we see later), the texts were arranged in conjunction with other decorative elements, which forced them to take on a distinctively ornamental character. They were designed to be part of a larger artistic ensemble that is structurally integral to the cathedral. The third is their location: over the principal entrance into the cathedral. All three characteristics can be found on other monuments in Georgia erected at the same time as Kumurdo. They are evident, for example on two major inscriptions on the cathedral of Ishkhani. In the tenth century this was one of the major ecclesiastical centres of the neighbouring region of Tao-Klarjeti, and it received much patronage from its rulers, the rival Bagrationi dynasty.20 The first inscription is found not on the cathedral itself, but on its neighbouring oratory, about 30 meters south-west of the cathedral. It was erected by Leon III’s brother-in-law, Gurgen Bagrationi (king of Kartli, 994–1008). It is found over the north door, where it is surrounded by an outer band of animals that lend themselves to a paradisiacal interpretation, an idea supported by the text itself (Fig. 25):21

Textual Icons

25. Ishkhani cathedral (Tao-Klarjeti, now in Turkey). Oratory chapel, north door. Inscription of Gurgen, king of Kartli, 994–1008. (Photo: Antony Eastmond)

In the name of God, I, Gurgen, king of kings, built this holy church in the name of the Holy Mother of God in prayer for my soul. Holy Mother of God protect me on the day of the last judgement. This is the gate of the Lord and into which the righteous shall enter. [I beseech] all who will enter to commemorate me in your prayers. Koronikon was SKV (i.e. 226 + 780 = 1006).

This text is perfomative, asking viewers to seek salvation through the acts of reading and entering the building. The inscription portrays a conflict between clarity and legibility on the one hand, and a desire for ornament on the other. Its form begins to blur the boundaries between text, decoration and image. The letterforms here are marked by their deeply drilled serifs and the care to make them fit the space. The inscription starts beautifully – separating words with colons and explicitly marking contractions; there are even lines between the rows to help guide the eye. But this marking and spacing had to be dropped after the fourth line, and by the end, the final word is split over three lines to fit the ever reducing radius of the semicircle; the date in both Georgian and Armenian appears at the end on the horizontals.22 This tympanum presents itself simultaneously as a text and as a work of art. The inscription is little more than a frame for the extraordinary rinceau that dominates the tympanum and the frieze of animals that appears on the hood moulding. The text is part of the ornament here. An identical arrangement is found in a later inscription over the south door into the cathedral that was added by Gurgen’s grandson Giorgi I (king of united Georgia, 1014–27) and his bishop

83

84

Antony Eastmond

Antoni Tsagereli. The second inscription is dependent in form and meaning on the earlier one. It is also located on a semi-circular tympanum and has eight concentric lines of ever-decreasing length. The final five letters form a cross at the centre of the composition:23 Jesus Christ exalt in both lives king Giorgi and his children, amen. In the name of God, I, the miserable Antoni, bishop of Ishkhani, renovated this porch of the holy catholic church for the glorification of kings: king Giorgi and his children, in prayer for the soul of Bagrat Kuropalates, [and] for forgiveness of our sins. Holy Catholic church aid and protect us before God and be our abode during the awful day of retribution before the Universal Judge [so that] we can be worthy of forgiveness for our sins and of [the] benevolence of God. I beseech all entering [through] this gate to commemorate me in prayers.

It is clear that the layout of these texts was a central concern. They are carefully contrived ensembles, whose importance lies in their overall appearance rather than their individual words. Texts are subordinated to the general decoration of the portal, and their design and meaning are linked to the function of the space they inhabit. There are parallels here with the ways in which the aesthetic qualities of words were exploited in the Islamic world. Oleg Grabar has written extensively about the shifting balance between the visual composition of inscriptions and their meanings.24 In their contents and their locations, the two Ishkhani inscriptions refer to the Christian theology of doors. They stand at the threshold between outside and inside, between secular space and sacred, and between damnation and salvation. This liminal location adds to the spiritual power of the texts. Their placement over entrances also allows the inscriptions to draw authority by referring to another visual cue, based on the expectation of Georgians approaching churches. Monolithic tympana had for centuries been an integral feature of Georgian church design, presenting an elaborate and self-contained location for major sculptural works. The most significant, and frequent, iconographic programme to be placed here was the Glorification of the Cross, the central symbol of Georgian Christianity that looked back to the country’s conversion myth in the early fourth century. This imagery can be found on numerous portals from the seventh century, as at Jvari (586–605) overlooking Mtskheta, the medieval capital of Georgia,25 or at contemporary monuments such as Savane (1046).26 Although only the inscription on the south door to Iskhani consciously evoked the cross, it is reasonable to suppose that all these inscriptions sought to take on the aura of authority that came with the cross and with the idea of the door as entrance to the holy. This is acknowledged in the inscriptions, confirmed by their location and sanctified by scripture. It is certainly the case that other inscriptions, such as that at Zarzma, which are examined in more detail later, do make use of the cross. Just as the inscriptions sought to create memories, so too they relied

Textual Icons

on existing memories and patterns of expectations to be enacted and understood. A similar analogy works for the inscriptions around apses, which were also often adorned with crosses, acting, like the inscriptions, as a seal of piety, authenticity and authority. At Kumurdo, a cross was structurally embedded in the masonry of the apse in red tufa. The importance of legibility in these inscriptions is emphasised by their choice of alphabet itself. The texts were all written in asomtavruli, the original Georgian majuscule alphabet first recorded in the fifth century CE.27 The role played by the choice of letterforms in the establishment of these blocks of text as visual as much as literary objects was an active issue in the tenth century. At that time the status of the asomtavruli alphabet was changing, because it was beginning to be replaced by a new cursive alphabet, known as nuskhuri.28 When Kumurdo was built both alphabets were used alongside each other, and there is no clear sense of hierarchy between them: Tornik’s Old Testament of 978 (Mount Athos, Iviron MS Georg. 1) was written in the new nuskhuri alphabet, whereas his Samotkhe (meaning Paradise, a collection of the writings of Sts John Chrysostom and Ephrem the Syrian; Mount Athos, Iviron, MS Georg. 9), which was copied out at the same time and at the same monastery (Oshki, in the province of Tao-Klarjeti in south-western Georgia), but finished a year earlier in 977, was entirely written in asomtavruli.29 By the eleventh century, a more stable relationship between the two alphabets had emerged, with asomtavruli becoming the predominant alphabet to use for chapter titles and also for monumental inscriptions, whether painted or incised, but nuskhuri becoming the standard scribal hand. The impact of the choice of alphabet is evident by looking at another church erected in Bagrationi-controlled Tao-Klarjeti. The monastery of Parkhali, built at the same time as Kumurdo by Davit Kuropalates (961–1001), prince of TaoKlarjeti, is an extraordinary revival of basilican architecture (one of two examples from the 960s–70s). Shortly after it was completed, it was adorned with an inscription written in an early variant of nuskhuri: nuskha-khutsuri (ecclesiastical cursive).30 The text appears on the south fac¸ade of the church, between arches six and seven of the clerestory arcade (Fig. 26). In terms of content this inscription is directly comparable with those at Kumurdo and specifically concerns embellishments and adornments to the church: † In the name of God, and through the intercession of the Holy Mother of God and St John the Baptist, God glorify Iovane, patriarch of Kartli and the entire east; Kurapalati has built the chasdudara [arcade] of Parkhali church; but no one adorned it. And he [Iovane] did adorn it; May he be glorified for ever! God bless the soul of father superior Giorgi Ghaghueli, amen, amen. Iovane Superior.31

Yet, in visual impact the contrast with the inscriptions at Kumurdo could not be more marked. The text looks more like modern graffiti than a commemoration of the enlargement and improvement of a major monastery. It is an

85

86

Antony Eastmond

26. Parkhali cathedral (Tao-Klarjeti; now in Turkey). South fac¸ade, inscription at arcade level, c.970. (Photo: Antony Eastmond)

altogether more hastily executed piece of work: lines rise and fall along their uneven lengths, and the letterforms are variable, not to mention of differing thicknesses of line and clarity. It looks as if it were painted hastily in a matter of minutes, smudged and fading towards the end. It certainly shows none of the care with which the foundation and embellishment inscriptions were carved by their contemporaries at Kumurdo and Ishkhani. It appears impermanent and contingent, but its contents suggest that it was designed to be understood as the same form of permanent marker as those other carvings. It is perhaps ironic that an inscription that talks about adornment should, on the face of it, be so crude and ugly. The contrast between the forms of inscription suggests that a different intent lay behind each: at Kumurdo and Ishkhani the texts are both icons and prayers, whereas at Parkhali the text is meant primarily to document and record, rather than act as a form of prayer. The exceptional nature of the inscription at Parkhali suggests that this was not the dominant concern of inscriptions at this time. The discussion of alphabets also raises questions about expectations of literacy: the Parkhali text is in the new and evolving ‘working’ alphabet of Georgian monks and includes aids such as punctuation, capitals and marks to denote contractions and abbreviations. All inscriptions, of course, presume and imply a literate audience. Although it is impossible to estimate the degree of literacy in tenth-century Georgia,32 it is possible to gauge interest in texts at that time. This interest is most eloquently established by the monk Iovane Mtatsmindeli (known

Textual Icons

in the west as John the Iberian or John the Hagiorite, died c. 1002); he started life as a monk at the neighbouring monastery of Otkhta Eklesia in Tao-Klarjeti, before heading west to become the founding abbot of the Iviron monastery on Mount Athos (where Tornik was eventually to retire with his collection of manuscripts).33 Iovane (and later his son) was a prolific translator from Greek into Georgian, and he recorded that he translated thirty-three books, as well as the complete Old and New Testaments.34 His work was in response to what he saw as the terrible state of education in Georgia: However, in our Georgian language, nobody anywhere could be found who could make these holy books of the interpretation of the holy Gospel accessible: while the Churches of Greece and Rome were full [of such people], those of our country were in destitution. And not only these books, but many others were lacking in our language. Seeing this, I, the poor Iovane, the least of monks, I was grief stricken by such scarcity of books in the Georgian lands. I therefore placed on myself many sacrifices and efforts, and I gave a complete Greek education to my son Ekvtime, so that he should be able to translate from Greek into Georgian the books which were [then] copied by us.

Within a monastic context, therefore, we can assume a voracious appetite for texts in the late tenth century, even if the degree to which that extended to the laity is much more uncertain. However, my argument is that the form of the monumental inscriptions that are this chapter’s focus obviates the need for the level of literacy that Iovane laments. Functional literacy adds precision to the prayers, but even to view them is to enact their contents. The concerns with legibility, ornament and location that the texts at Kumurdo and Ishkhani reveal have important consequences for our understanding of the other kinds of text inscribed on Georgian monuments in this period. These texts are very different in content, but, because they draw on the same locational and visual cues, they are also able to be enacted as prayers and icons. The first comes from the church Oshki in Tao-Klarjeti, built by Davit Kuropalates and his brother Bagrat between 963–73. Oshki was the most ambitious and ornate Georgian building of the tenth century. The inscription (one of many) is over the principal, south door. The surviving portion stretches over twelve densely written lines. Although it is a foundation text, its extraordinary length and bureaucratic detail give more prominence to economics than to piety, yet the scale of the former is a demonstration of the latter. Its location and use of imperial purple ink also help confer spiritual value on the text: By the grace of God who protects those who honour him and who benevolently leads to success all good deeds of those who love his name. With belief in the Holy Trinity and the intercession of our Holy Queen – Mother of God – with the grace of the Holy Wood of Life with the help and intercession of the glorious Baptist and all the saints who devoted themselves to God from time immemorial [it is] for the glorification and praise of all

87

88

Antony Eastmond

these saints that our divinely crowned kings began the construction of this holy church so that in it could be celebrated holy days and the saints splendidly commemorated, as it is written: “partake in the commemoration of the saints.” By the desire of all those saints those glorious kings of ours did not spare transient riches for that which is permanent, in order to merit the aid of all those saints, and may God help in both lives the sons of the divinely blessed Adarnase kuropalates, Bagrat eristavt eristavi, Davit magistros. The Holy Trinity shall protect all three on his right side and glorify them in both lives. I, Grigol, was made worthy by God to serve the Lord through them. I was the supervisor of the labours and may God convince them of my devotion to them, their servant. The yearly expenditures and wages for the masons and hired labourers and for the oxen wagons carrying the sand was k [20] thousand drama, the wine c [5,000] phisos, the iron n [50] litres, the grain smi [250] grivi, the number of steadily labouring masons, carpenters and blacksmiths [was] ji [70], the number of oxen carrying the stone l [30] the mules and other pack animals which were transporting the spondiki from the environs of the monastery of Holy Grigol l [30] the other carriers which were gathering . . . . . . . . . j [60] . . . . . . . . . . . . . . . j [60] and . . . . . . [the total number of ] men and those labouring [in the construction of the church] p [80].35

A second example from Ishkhani, dated 1017, moves even further from the foundational concerns of the inscriptions we have seen so far. It has no direct relationship to the fabric of the cathedral, but rather presents a partial history of the bishopric. It was painted in white on a lilac ground beside the door from the north atrium into the naos of the cathedral. Only eighteen of its twenty-seven lines have been read by scholars: In koronikon rlz [137 = 917] in the month of September z [7] with the help of Christ and the blessing of the holy catholic church by the orders of Adarnase, king of the Georgians, our blessed father Basil was appointed as bishop, and he served this holy church it [19] years with outstanding dedication, and in complete faith, and he entrusted his soul to God in the month of December kv [26] on Friday. May Christ rest his soul, amen. After him by the orders of our glorious and worthy kings – may they be exalted by God – David, king of the Georgians, Ashot kuropalates, Bagrat magistros [and Sumbat antipatrikios] dispatched to Greece-Trebizond our honourable, worthy father Stepane, and he was consecrated as archbishop [of Ishkhani] by the hand of the honourable and God-imbued Greek patriarch Basil in the month March a [1], Sunday. May God endow him benevolently with success, amen. Because at that time [the catholicos] of Kartli blessed . . . 36

In 980 an inscription was set up at the monastery of Zarzma by one of Davit Kuropalates’ subjects, the monk Tornik. Although it, like the inscription at Oshki, is ostensibly a foundation text, it is quickly apparent that its real concerns lie

Textual Icons

elsewhere: recounting Tornik’s role in the suppression of the revolt of Bardas Skleros against the Byzantine emperor Basil II in 979, a defining moment both in the reign of Basil II and in Georgian-Byzantine relations. The text, the second half of which is now lost, was clearly originally very extensive, and its prominence was ensured by the monumental scale of the lintel on which it was carved (it originally measured about 300 × 100 cm, but was subsequently reduced in size to about 200 × 100 cm, probably when it was reused in its current location, as the lintel over the door of a fourteenth-century chapel adjoining the monastery’s bell tower). The inscription has been treated by historians simply as a primary textual source, but the presence of the central cross is a reminder that it also belongs to this other, visual context:37 In the name of God, and with the intercession of the holy mother of God, I, Ivane, son of Suli, built this holy chapel. At the time when in Greece Skliaros rebelled, Davit kuropalates – may God exalt him – sent us all to go to help the holy emperors. We forced Skliaros to flee. I, in that country, which is called Kharsanani, to the district called Sarvenisi, [where] sat on the throne . . . 38

All three inscriptions use the same visual cues – elegance of script and location around a door – as the foundation inscriptions from Kumurdo and Ishkhani to give them spiritual authority. However, they differ in one key way: all three are considerably longer. It is doubtful that they were intended to be read in full very often (if at all). Their length argues against this as an everyday practice, unless as a form of veneration or meditation. However, their contents do not concern everyday commemorations, and their relevance to the forms of the buildings in which they are located is limited. Moreover, although their locations may give them authority, they also militate against them being read. The doorways are narrow and in confined spaces; anyone lingering to read the texts would block the entrance to other worshippers seeking to walk into the churches. However, part of the meaning and power of these inscriptions derives from their enormity. As has been argued for the scale of the inscriptions of the Res gestae of Augustus in the first century CE, length in and of itself conveys authority and truth.39 Even for the literate, the importance of the texts is conveyed by the relentless march of words over so many lines and at such length, and their location confirms their pious intent. They are the lapidary equivalents of Tornik’s extended colophons. The quantity of words again suggests that what was most significant was the block of text: it is the sum (the visible paragraph) rather than the parts (the words and their meanings) that counts. With these examples in mind we can return to one final group of inscriptions at Kumurdo. The contents of the texts are very unusual: they are legal transactions rather than commemorations of donations or memorials. All concern the establishment of feast days to commemorate members of the local nobility. They are all much shorter texts (see the Appendix), but the way they cluster

89

90

Antony Eastmond

27. Kumurdo cathedral ( Javakheti, Georgia). South porch, Festival inscriptions, 964- early eleventh century. (Photo: Antony Eastmond)

around the main door and repeat many of the same textual formulae gives them a visual coherence, despite the fact that they were carved by different craftsmen at different times using different letterforms (Figs. 27 & 28). This makes them, as a group, analogous to the extensive inscriptions that we have just examined. The first was set up by the founding bishop, Iovane, immediately to the right of the main entrance to the church: In the name of God I, the bishop Iovane, fixed the feast day of the eristavi Vache on Easter day. May whoever alters this be cursed by this icon, my cross. [no. 3]40

This was followed by a series of similar texts that were added piecemeal between 964 and the early eleventh century (see the Appendix). Historians have been concerned by the problem of what these texts represent. They are not commemorative in the same way as the apse and foundation inscriptions. They are much closer to legal documents that concern economics and practice, not piety. They establish and eternalise commemorations rather than perform them: to read them out is to confirm a transaction, not to commend

Textual Icons

28. Kumurdo cathedral ( Javakheti, Georgia). South porch, Festival inscriptions, 964- early eleventh century. (Drawing: Antony Eastmond)

someone’s soul to God. They cannot even be seen as the equivalent of synodika, manuscripts that list the names to be prayed for day by day through the year (but that do not record the paying for it).41 The inscriptions are remarkable as documents. Their format suggests that they replicate a text or proclamation that would surely have been written and stored elsewhere in the cathedral archives (although no such manuscript is known). It is unclear how the inscriptions would relate to such documents. Do these monumental examples replace those documents or simply publicise them? Which document would hold the legal authority in the case of a dispute? Although these questions remain pertinent, they ignore the visual context of the inscriptions. I argue that even though the contents of the inscriptions may be peculiarly legal, their form ties them into the ideas about textual icons outlined earlier. Although the first inscription may have been commissioned for a specific reason (now unknown), its meaning was soon altered by the gradual accumulation of other texts around it. In turn, these gained meaning by being placed in juxtaposition to those already in place, and once there, they began to change the meaning of the earlier texts and the location. What began as the commemoration of the builders and their lords was gradually transformed into a site of communal memory. It was a shift from the individual to the corporate. All

91

92

Antony Eastmond

undermined notions of time and historical moment. Visually, none of the later inscriptions has the same level of visual authority as the original. They were all incised, rather than cut in relief, and used a quicker and less precise script. Instead, they derived their power from quantity – through their huddling together and their accumulation. Whatever appearance the south portal had in the years immediately after the consecration of the church had been transformed a century later. The entrance wall had become a solid block of texts, each of which was related to the others around it to form a single entity. The apparently random placement of the texts around the door offers no order in which to read them. Indeed, this surely mattered little as their repetitive contents need no sequence. It is their visual coherence as replicated blocks of text that matters. One final clue about the way in which this group of inscriptions worked comes from the references in them to outside referents: ‘May whoever alters this be cursed by this icon, my cross’. Only three texts include crosses at the start of their inscriptions (nos. 4, 7, 10). It therefore seems unlikely that these are internal textual references. Do they refer to a cross (signature) made on some ‘original’ document, or are we meant to assume here that there was some other icon or cross in front of which such oaths were sworn? Icons and crosses are referred to in inscription no. 10, and of course icons and crosses were commissioned to stand inside the great Georgian churches; many examples survive from the tenth century.42 Certainly, by the fifteenth century Kumurdo was home to a famous, miracle-working icon, but its history before that time is unknown. Three niches exist in which icons could have been placed, one to the right of the main door and the other two in the east walls of the south-west and north-west apses. However, if an icon was placed to the right of the main door, in close proximity to the inscriptions, it was clearly already gone by the time Molozon’s commemoration was fixed and commemorated by an inscription in the niche itself (no. 6). It therefore seems unlikely that these permanent texts defer to a moveable object elsewhere. I propose that we should see the text itself as ‘this icon’. It is the blocks that decorate the exterior that have become the pious image, the textual icon. They do not need to be read: each block is transformed from a legal transaction into a visual encapsulation of a prayer, and their power grows through their accumulation and massing around the church’s entrance. They act in a similar way to the amassing of icons on the templon screen in the Byzantine world, or the massed presence of saints on menologion icons.43 In all these cases, the individual elements had a power and a purpose, but they became subsumed in the larger whole. A final parallel for the importance of words in Georgian art can be found on contemporary icons, which give great prominence to texts. Georgia has a strong tradition of placing texts on icons, even allowing them to dominate the composition as a whole. Antoni, the bishop of Ishkhani who remodelled the south door of Ishkhani, also commissioned an icon of St Symeon the younger in c. 1015 (Fig. 29). He depicted himself below the stylite, but gave equal prominence to a text – the visual echo of the bishop on the right side of the saint’s column:

Textual Icons

29. Icon of St Symeon the younger commissioned by Antoni, bishop of Ishkhani, c.1015. (National Museums of Georgia, State Museum of Fine Art)

93

94

Antony Eastmond

I, the wretched Antoni Tsagereli, during my time at Ishkhani, cultivated the desolate region of Mkhatvareuli, and giving thanks to St Symeon, made this icon and set it up in the church for the salvation of my soul. Whosoever alters this small offering of mine, may St Symeon punish him before Christ on the day of judgement. Amen.44

Just as the commemorative inscriptions at Kumurdo stand in for individuals, so too the text here duplicates the individual. They are presented as equivalents: two ways of representing the same act and prayer of donation. The importance of language was central to Christianity, thanks to the opening of St John’s Gospel: ‘In the beginning was the Word, and the Word was with God, and the Word was God’. This verse gave words a physical presence and an ability to act. However, the power of language was elaborated in Georgia in the tenth century. Ioane Zosime, a monk active at the Georgian monasteries at Mar Saba in Jordan and at Mount Sinai from 949–87, composed a mystical poem The praise and glorification of the Georgian language.45 In it, Ioane gave the Georgian tongue superiority over all languages. He promoted Georgian as older than any other language (by ninety-four years, in his complex numerological symbolism) and suggested that it will be the language of the last judgement: Buried is the Georgian language As a witness until the day of the second coming, So that God may examine every language Through this language. And so the language Is sleeping to this day And in the Gospels this language Is called Lazarus . . . And this language, Beautified and blessed by the name of the Lord, Humble and afflicted, Awaits the day of the second coming of the Lord.46

This mystical poem is an extraordinary expression of belief in the power and importance of a language. With such emphasis being placed on language, we should expect that monumental inscriptions were more than simply bearers of information. They were also proclamations of theological truth that had to bear the weight of performing the last judgement. Georgian inscriptions must be considered both as artistic works, with consideration of their beauty, location, and the care with which they were made, and also as art: textual icons that held the same theological value as images, serving as representations of truth with access to the divine.

Textual Icons

95

APPENDIX: THE FESTIVAL INSCRIPTIONS AT KUMURDO Location

text

no. 3

to the right of the door

no. 4

below no. 3

no. 5

to the right of the niche over the door into the small chapel in the niche to the right of the door

In the name of God I, the bishop Iovane, fixed the feast day of the eristavi Vache on Easter day. May whoever alters this be cursed by this icon, my cross.a In the name of God I, the bishop Grigol, fixed the feast day of Goliat, on the day of Ignatius. May whoever alters this be cursed by this icon, my cross.b On Great Sunday [be] the feast day of Mirdat, on the eve of Christmas, that of Goliat. May whoever alters this be cursed by God’s mouth.c In the name of God, I, the bishop Grigol, have fixed the feast day of Molozon on the Friday of Palm week, on the feast day of the holy bishops. May whoever does not carry this out in full be cursed by this icon and our cross. Full fee has been taken.d Christ have mercy on Molozon’s soul, amen.e

no. 6

no. 7 no. 8

around the arch of the niche to the right of the door recorded on the west wall

no. 9

around the arch of the main door and in the spandrel

no. 10

below no. 4

no. 11

recorded to the left of the main door

a b c d e f g h i

Silogava, Kumurdo, no. 4. Ibid., no. 13. Ibid., no. 14. Ibid., no. 15. Ibid., no. 16. Ibid., no. 19. Ibid., no. 20. Ibid., no. 18. Ibid., no. 21.

[In the name of God, I] Eprem the bishop, [fixed for the feast day . . . ] . . . the charge has been taken in full.f In the name of God, I, the bishop Eprem, fixed the feast day of Gabriel Kumurdoeli, the fifth bishop, on the Friday of Shrovetide, at the festival of the holy archangels. And the full charge has been received from him. May whoever alters this be cursed by this icon and my cross. Christ have mercy on the soul of Gabriel Kumurdoeli, amen.g In the name of God, I, Zosime Kumurdoeli, invested the altar, chalice, dish, the cross, and embellished the icon of the Saviour by the hand of Davit. Fixed his feast day with kyrie eleison on the fifth [day] of Palm Week; let the archpriest celebrate the feast on the [day] of the Holy Trinity by singing a hymn. May he who alters this be cursed by this icon.h On . . . of January, the feast day of father Pavle Mangleli. God, let the divine service and prayer of that day be of benefit to him. Whoever does not celebrate this, let God demand from him, amen.i

96

Antony Eastmond

NOTES

1. Valeri Silogava, Kumurdo tadzris epigrap’ika [The Epigraphy of the church of K’umurdo] (Tbilisi, 1994), no. 3; Nodar Shoshiashvili, Aghmosavlet da samkhret sakartvelo V-X ss. kartuli tsartserebis korpusi: lapidaruli tsartserebi [A Corpus of Georgian Inscriptions in East and South Georgia, V-X centuries], vol. 1 (Tbilisi, 1980), no. 146. 2. On the architecture see N. P. Severov and Giorgi N. Chubinashvili, Kumurdo i Nikorcminda [Kumurdo and Nikortsminda] (Moscow, 1947), 5–16. 3. Linda Seidel, Legends in Limestone: Lazarus, Gislebertus, and the Cathedral of Autun (Chicago, London, 1999), 15. 4. Amy Papalexandrou, ‘Text in context: eloquent monuments and the Byzantine beholder,’ Word & Image 17.3 (2001), 259– 83. 5. Papalexandrou, ‘Text in context: eloquent monuments and the Byzantine beholder’; Liz James, “And shall these mute stones speak” Text as art,’ in Art and Text in Byzantine Culture, ed. L. James (Cambridge, 2007), 188–206; Amy Papalexandrou, ‘Echoes of Orality in the Monumental Inscriptions of Byzantium,’ in Art and Text in Byzantine Culture, ed. L. James (Cambridge, 2007), 139–60. 6. Irene A. Bierman, Writing Signs. The Fatimid Public Text (Berkeley, CA, 1998). 7. Christina Maranci, ‘Trusting architecture: inscribed churches and their function on the eastern frontier,’ in Between Paris and Fresno: Armenian studies in honor of Dickran Kouymjian, ed. B. Der Mughrdechian, Armenian Studies Series, 13 (Costa Mesa, CA, 2007), 39–49. 8. Seidel, Legends in Limestone, 15. 9. Mary J. Carruthers, The Book of Memory: A Study of Memory in Medieval Culture (Cambridge, 2008). 10. It should be noted that a similar interest in the presentation of monumental, extensive inscriptions can equally be found in Armenia in this period; but this paper is limited to Georgian material for the sake of simplicity and cohesiveness. Compare Chapter 9 in this volume. 11. This transcribes into the modern Georgian mkhedruli alphabet as d[dddd]d, d[dd]d[ddd]d ddd[d]dd dd[dd]d[ddd]dd, dd[d]dd d[d]d[d]d [d]d[d]dd, dd[d]dd d[d]d. 12. There are numerous examples in: Shoshiashvili, Aghmosavlet da samkhret sakartvelo,

13.

14.

15. 16.

17.

18. 19.

20.

21.

22. 23.

eg. nos. 131, 142, 143; Wachtang Djobadze, Early Medieval Georgian Monasteries in Historic ˇ set’i (Stuttgart, 1992), Tao, Klarjet’i and Savˇ 213 and pl. 313; 184 and pl. 254 (this time written in nuskhuri script and accompanied by an image of the figure, Theodore, named in the prayer) from the monastic church at Parkhali. Robert P. Blake, ‘The Athos Codex of the Georgian Old Testament,’ Harvard Theological Review 22.1 (1929), 33–56, with supplement in Robert P. Blake, ‘Catalogue des manuscrits g´eorgiens dans la biblioth`eque de la Laura d’Iviron du Mont Athos,’ Revue de l’Orient Chr´etien 28 (1931), 289–361, 301– 4. The manuscript is now bound in two volumes, and the prayers appear on vol. 1, fols. 143r, 148r, 220v, 399r; vol. 2, fols. 117v, 171r, 219v, 258r, 277v, 302v. Giorgi N. Chubinashvili, Gruzinskoe chekannoe iskusstvo [Georgian repouss´e art] (Tbilisi, 1959), 573–576. Bierman, Writing Signs, 25–6 with further references. For examples see Robin Cormack, The Church of Saint Demetrios. The Watercolours and drawings of W. S. George (Thessaloniki, 1985), nos. 31, 38. An equivalent appears in Georgia at Mijna: ‘Christ have mercy’: Shoshiashvili, Aghmosavlet da samkhret sakartvelo, no. 132. Shoshiashvili, Aghmosavlet da samkhret sakartvelo, no. 145; Silogava, Kumurdo, no. 1. The koronikon is a paschal calendar: a 532-year cycle starting in 780, see Venance Grumel, Trait´e d’´etudes byzantines: I: La Chronologie (Paris, 1958), 151–3. Shoshiashvili, Aghmosavlet da samkhret sakartvelo, no. 147; Silogava, Kumurdo, no. 2. Blake, ‘The Athos Codex of the Georgian Old Testament,’ 50–3: vol. 2, fols. 219v, 427r– 428v. The most recent study of the complex architectural history of Ishkhani is Markus R. P. Bogisch, The Appropriation of Imperial Splendour. Ecclesiastical Architecture and Monumental Sculpture in Medieval Tao-Klarjeti around 1000 AD (Copenhagen, 2009), 179–206; see also Djobadze, Early Medieval Georgian Monasteries, 191–217. Djobadze, Early Medieval Georgian Monasteries, 216–17. The inscription borrows from Psalm 118.20. Ibid., 216 n.690. Ibid., 211–13.

Textual Icons

24. Oleg Grabar, The Mediation of Ornament, Bollingen series, 35 (Princeton, 1993), 88, 99–101. 25. Demetre Tumanishvili, Tamar Khundadze, and Davit Khoshtaria, Jvari, Church of the Holy Cross at Mtskheta (Tbilisi, 2008), 67– 69. 26. Vakhtang Beridze, ‘Savane (XI saukunis kartuli khurotmodzghvrebis dzegli)’ [‘Savane (Georgian artistic monument of the eleventh century)’], Ars Georgica 1 (1942), 77–132. Other contemporary examples include the south porches at Khakhuli (Djobadze, Early Medieval Georgian Monasteries, fig. 49, pl. 202); and Nikortsminda (Natela A. Aladashvili, Monumental’naia skul’ptura Gruzii [Monumental Sculpture of Georgia] (Moscow, 1977), pls. 167–9). 27. The earliest known inscription is that of Abbot Antonios from floor mosaic of the monastery of St Theodore at Bir el-Qutt near Bethlehem, early fifth century: M. Tarchnishvili, ‘Les r´ecentes d´ecouvertes e´ pigraphiques et litt´eraires en g´eorgien,’ Le Mus´eon 63 (1950), 248–51; image in Adriano Alpago-Novello, Vakhtang Beridze, and Jacqueline Lafontaine-Dosogne, Art and Architecture of Medieval Georgia (Louvain La Neuve, Milan, 1980), 44–5, figs. 17 & 18. 28. For palaeography see Ilia Abuladze, Karthuli tseris nimushebi: Paleographiuli albomi [Samples of Georgian Writing: Paleographic Album] (Tbilisi, 1973). 29. Treasures of Mount Athos, ed. A. A. Karakatsanis (Thessaloniki, 1997), 20.3 (Old Testament), 20.2 (samotkhe). 30. On the architecture of Parkhali see Djobadze, Early Medieval Georgian Monasteries, 178–90. 31. Valeri Silogava, Oshki: Tenth Century Memorial Church (Tbilisi, 2006), 195–6; Djobadze, Early Medieval Georgian Monasteries, 184. The title ‘patriarch of Kartli and all the east’ indicates that it was erected under catholicos Iovane I Okropiri (980–1001). 32. Compare the attempt to consider literacy in Byzantium: Margaret Mullett, ‘Writing in Early Mediaeval Byzantium,’ in The Uses of Literacy in Early Medieval Europe, ed. Rosamond McKitterick (Cambridge, 1990), 156–85. 33. For the life of Iovane see Bernadette MartinHisard, ‘La Vie de Jean et Euthyme et le statut du monast`ere des Ib`eres sur l’Athos,’ Revue des e´ tudes byzantines 49 (1991), 67– 142. 34. Mount Athos, Iviron MS Georg. 10, fols 332v–336r; Blake, ‘Catalogue des manuscrits

97

35.

36. 37.

38.

39.

40. 41.

42.

43.

44.

g´eorgiens dans la biblioth`eque de la Laura d’Iviron du Mont Athos,’ 339–44. Djobadze, Early Medieval Georgian Monasteries, 132–4. The reference in the inscription to Davit magistros refers to the earlier Byzantine court title held by Davit kuropalates prior to 979. Ibid., 209–10. N. Adontz, ‘Tornik le Moine,’ Byzantion 13 (1938), 143–64; C. Holmes, Basil II and the Governance of Empire (976–1025) (Oxford, 2006), 242–3. E. Taq’aishvili, ‘Zarzmskii monastir,’ ego restavratsiia i freski’ [‘Zarzma monastery, its restoration and frescoes’], Sbornik materialov dlia opisanie mestnostei i plemen Kavkaza [A Collection of Materials for the Description of the Places and Tribes of the Caucasus] 35 (1905), 16–22. Shoshiashvili, Aghmosavlet da samkhret sakartvelo, no. 156. Most recently published in Valeri Silogava, ‘Zarzmis lapidaruli epigrapika’ [‘Lapidary Epigraphy at Zarzma’], in Zarzma, ed. N. Z. Vachnadze (Tbilisi, 2007), 105–9. Ja´s Elsner, ‘Inventing Imperium: Texts and the Propaganda of Monuments in Augustan Rome,’ in Art and Text in Roman culture, ed. J. Elsner (Cambridge, 1996), 32–53. Shoshiashvili, Aghmosavlet da samkhret sakartvelo, no. 148; Silogava, Kumurdo, no. 4. Darejan K’ldiashvili, Sinas tsm. ekaterines monastris kartvelta eklesiis sulta matiane [Synodikon of the Georgian Church at the Monastery of St. Catherine on Sinai] (Tbilisi, 2008). Guaram Abramishvili and Alexander Javakhishvili, Jewellery and Metalwork in the Museums of Georgia (Leningrad, 1986), figs. 102, 116, 126. See, for example, the hundreds of saints on the bilingual Georgian-Greek hexaptych from Sinai: George Galavaris, An Eleventh-Century Hexaptych of the Saint Catherine’s Monastery at Mount Sinai (Venice, Athens, 2009); although this misses the crucial evidence about the Georgian material: Zaza Skhirtladze, ‘Ioane Tokhabi – sinaze moghvatse kartveli mkhatvari’ [‘Ioane Tokhabi, a Georgian Artist Working at Sinai’], Literatura da Khelovneba [Literature and Art] no. 3 (1998), 61–72. Chubinashvili, Gruzinskoe chekannoe iskusstvo, 302–8; M. Bartholomaei, ‘Lettres de M. Bartholomaei, relatives aux antiquit´es g´eorgiennes,’ M´elanges Asiatiques tir´es du Bulletin Historico-Philologique de l’Acad´emie Imp´eriale des Sciences de Saint-P´etersbourg 2 (1856), 281–2.

98

Antony Eastmond

45. M. Tarchnishvili, Geschichte der kirchlichen georgischen Literatur (Vatican, 1955), 109– 14. 46. Aleksandr A. Tsagareli, Pamiatniki gruzinskoi stariny v sviatoi zeml’ i na Sinia, [Monuments

of Georgian antiquity in the Holy Land and at Sinai] (St Petersburg, 1888), 203; trans. Donald Rayfield, The Literature of Georgia. A History (Oxford, 2000), second edition, 33–4.

鵻 CHAPTER FIVE

PSEUDO-ARABIC ‘INSCRIPTIONS’ AND THE PILGRIM’S PATH AT HOSIOS LOUKAS 鵼 Alicia Walker

A topic of long-standing interest in Byzantine architectural history has been the intriguing presence of pseudo-Arabic ‘inscriptions’ in the decorative programmes of some middle Byzantine churches.1 The earliest example is attested in the painted ceramic tiles that once adorned the interior of the Monastery of Constantine Lips in Constantinople (Istanbul) (ca. 907).2 The motifs become more common, however, in tenth- to twelfth-century churches in Greece, where they are found predominantly as cloisonn´e brick designs in exterior walls and adorning objects and buildings depicted in wall paintings and mosaics.3 Although in certain instances these script-like forms closely imitate specific Arabic letters or letter groups, they typically do not compose coherent words or phrases; hence, their identification as pseudo-epigraphy.4 In art-historical scholarship, pseudoArabic is commonly categorized as ornament – that is to say, as decorative motifs that did not carry specific semantic value.5 There is no doubt that some medieval viewers perceived pseudo-Arabic as only an aesthetic embellishment. Yet the frequency and consistency of these motifs in middle Byzantine churches raise the question of whether pseudo-Arabic was intended – in some cases – to convey particular meanings or associations. In proposing a connotative function for pseudo-Arabic, this chapter explores what Lothar Von Falkenhausen describes as the ‘iconic use of writing, in which characters assume a role akin to that of iconographic motifs rather than transmit the words of a specific message’.6 I argue

I thank Antony Eastmond, Elena Boeck, Heather Grossman, Amanda Luyster and Vasileios Marinis for their insightful comments on earlier versions of this chapter. Of course, any errors of fact or interpretation remain mine alone. I am grateful to Carolyn Connor for generously sharing her photographs of the crypt at Hosios Loukas.

99

100

Alicia Walker

that, although pseudo-Arabic in Byzantine monumental programmes could not be read in a literal sense, the proximity of these motifs to Arabic script would have allowed them to retain status as language by ‘symbolically affirming’ the identity and alterity of Islam and the regions over which medieval Muslims held dominion.7 I take as a case study the decorative programme of the middle Byzantine monastery of Hosios Loukas in Phokis, Greece, which was a renowned medieval healing shrine and pilgrimage destination. The site includes two conjoined churches positioned north-south of one another. The north church retains the earliest example of pseudo-Arabic cloisonn´e brick, and additional pseudo-Arabic motifs in carved marble, wall painting and mosaic are found in the external and internal decorative programmes of both buildings. I argue that pseudo-Arabic played a key role in shaping the pilgrim’s physical and conceptual experience of the site by connoting Muslim groups and the geographic territories under their control.8 I concur with scholars who posit that the north building can be understood as a ‘victory church’, and I extend this message to encompass pseudo-Arabic as both a sign of Byzantine triumph over Muslim foes in that period and as an apotropaic motif that demarcated boundaries and protected the sacred spaces of the church. I differ from earlier interpreters, however, by arguing that over time this meaning was partially displaced by another set of associations that recognised Arabic as the dominant ‘public text’ of the Holy Land, thereby imbuing pseudo-Arabic with a geographic association that carried a different range of religious and political implications.9 I suggest that, in this later phase, pseudo-Arabic connoted the sacred authority and spiritual efficacy of Christian holy sites in the East, while ultimately calling the viewer to reclaim these territories from Muslim dominion. My reading depends on a consideration of pseudo-Arabic in its architectural, iconographic and ritual contexts; I attend to its location within the structures at Hosios Loukas and the implications that this placement had for pilgrims’ reception of these ‘inscriptions’.10 I propose that pseudo-Arabic served to direct the visitor’s performance and experience of pilgrimage by demarcating sacred space and contributing to a liminoid phenomenon, which would have enhanced the perceived efficacy of the site.11 In this respect, I follow Irene Bierman’s analysis of the communicative potential of public text, in which the meaning of an inscription was ‘not completely contained in the writing itself but, rather, grows in the web of contextual relationships woven between the official writing, the patrons, the range of beholders, and the established contexts in which that writing was placed’.12

THE MONASTERY OF HOSIOS LOUKAS

Hosios Loukas is positioned along the western slope of Mount Helicon in the region of Boeotia in central Greece. During the middle Byzantine period,

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

Thebes, the closest major town and the capital of the theme (administrative territory) of Hellas, was located about 30 miles (45 km) to the southwest of the site. There survive no foundation documents for the monastery and only minimal historical references to its first years. For this reason Luke’s vita, written shortly after his death in the third quarter of the tenth century, is an important source for the initial history of the site.13 According to the vita, Luke settled in the area around 946.14 During his lifetime, the complex was elaborated with a newly built church dedicated to St. Barbara and funded by the regional governor, who had become a devotee of the saint.15 After Luke’s death in 953 and at his request, his cell was converted to a tomb; soon after the burial site was enclosed, and later a shrine was erected over it.16 The buildings mentioned in the vita were subsequently replaced by the two churches that stand at the site today.17 Although their exact chronology remains unresolved, the smaller, north church, dedicated to the Panagia or Theotokos (Virgin Mary), was without doubt built first, as confirmed by archaeological investigations completed in the 1960s.18 It follows a cross-in-square plan, with a compact naos capped by a steeply rising central dome. The interior decoration is now almost completely lost, but fragmentary remains of the original sculptural programme and pavements suggest its former grandeur. I find most persuasive Laskarina Boura’s proposed dating, which is based on close scrutiny of the building’s structure, comparative analysis of its sculptural ensemble and assessment of the available historical evidence. She places its foundation soon after 961 and suggests an association of the monument with the patronage of the Byzantine emperor Romanos II (r. 959–63).19 The eleventh-century south church follows a modified cross-in-square plan and consists of an upper church and a lower crypt. The church takes the form of a central octagon encased in a square, which is in turn surrounded by a larger rectangle; four cross-arms connect the inner square and outer rectangle. The interior is dominated by a vast dome (measuring almost 30 feet [ca. 9 m] in diameter) and is famous for its well-preserved programme of figural mosaics and marble revetment. Equally renowned is the crypt, which is outfitted with one of the best preserved wall painting programmes of middle Byzantine art. Almost all scholars agree that the katholikon (the main, south church) dates to the first half of the eleventh century and postdates the vita.20 Opinion diverges, however, regarding the precise period of its foundation. I endorse the interpretation of Manolis Chatzidakis, who places the dedication in the year 1011 (when the relics of Luke were translated to a shrine in the new, south church) and ascribes patronage of the katholikon to the abbot Philotheos.21 Controversy also surrounds the chronology of the decorative programmes. I find persuasive the argument of Nicolas Oikonomides, who dates the mosaics of the katholikon to the 1040s, associating them with the patronage of the abbot Theodosios (fl. 1048). Oikonomides dates the fresco programme of the crypt to shortly after 1048 based on the presence of Theodosios’s portraits.22

101

102

Alicia Walker

Although the exact circumstances surrounding the patronage of the monastery’s buildings are obscure, both churches show innovative architectural forms and extravagant decorative programmes that represent the most sophisticated level of artistic achievement of the time, especially in Greece. Scholars have long proposed a connection between these monuments and the artistic and architectural production of Constantinople; however the buildings also demonstrate novel adaptations of models from the capital. It is likely that some form of support for the projects came from elite echelons of Byzantine society, whether highranking members of the imperial administration stationed locally, regional magnate families or emperors themselves. At the same time, it is clear that the leaders of the monastic community were also important agents of these undertakings. The churches must, therefore, be seen as the product of collaborations among a number of interested parties, operating both locally and in the broader field of middle Byzantine society.

PSEUDO-ARABIC AND THE PILGRIM’S EXPERIENCE OF THE TENTH-CENTURY CHURCH

Pseudo-Arabic is prominently displayed on the exterior of the tenth-century church at Hosios Loukas, dispersed across its surface in the form of cloisonn´e brickwork (Fig. 30). The individual blocks are substantial, measuring about eleven inches in height. Their relatively large scale allows them to be seen clearly from the ground, even when a viewer stands some distance away. Although the motifs proliferate around the eastern fac¸ade, they are found on all four walls of the structure. One course of pseudo-Arabic cloisonn´e brick motifs runs in a continuous frieze just below the roof of the central apse; a second frieze runs to either side of the arches of the upper window of the central apse; and a third frieze runs just below the roofs of the side apses and continues onto the eastern ends of the north and south walls.23 The next nine courses include isolated pseudo-Arabic blocks, placed at regular intervals horizontally and vertically in an almost checkerboard pattern. These blocks are similarly concentrated around the east fac¸ade and the east ends of the north and south walls; however, individual pseudo-Arabic cloisonn´e blocks originally appeared on the western surfaces of the north and south walls as well as on the west fac¸ade, as revealed during archaeological investigations in the 1960s that exposed exterior surfaces of the north church obscured by later decorations and the construction of the katholikon.24 The ‘letter’ forms of these cloisonn´e brick motifs show the distinctive leaflike finials and angled forms of foliated kufic, a script type that emerged in the ninth century and is attested in Islamic monuments and objects across an expansive geographic range from Spain to Iran.25 Although the motifs at Hosios Loukas do not form readable script, they do resemble Arabic letters, including a medial s¯ın, a medial m¯ım and combinations that include various forms

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

30. View of the south wall of the east end of the north church showing cloisonn´e bricks with pseudo-Arabic patterns, Monastery of Hosios Loukas, Byzantine, ca. 961, Phokis, Greece. (Photo: Alicia Walker)

of l¯am, ‘alif, k¯af and ‘ayn.26 The source from which Byzantine designers and craftsmen drew these motifs is unknown. Foliated and floriated kufic inscriptions were particularly popular in objects and monuments produced during the reign of the Fatimid Dynasty (909–1171), with its capital first at Mahdia (along the east coast of modern-day Tunisia) and from 969 at Cairo (Egypt). Kufic inscriptions appear prominently, however, on North African buildings that predate the Fatimids, employed already by the Aghlabids (with their capital at Kairouan in modern-day Tunisia) (r. 800–909) and in Egypt before the arrival of the Fatimids.27 Indeed before the construction of the north church at Hosios Loukas in the 960s, the Byzantines maintained diplomatic and trade relations with Islamic groups in North Africa, including a delegation sent by Romanos I Lekapenos (r. 920–44) in 937–8 to the Ikshidid court in Egypt (935–69).28 Given the relative geographic proximity of Byzantium and North Africa, as well as the longstanding contacts between these regions, it is possible that the model for these pseudoArabic motifs came from Egypt or a nearby location.29 For example, the emirate of Crete, which maintained close commercial and diplomatic ties with Egypt, especially under the Tulunid dynasty (868–905), was conquered and sacked by the Byzantines in 961 and may have provided a source for examples of foliated kufic.30 It is likely that the original entrance to the monastery of Hosios Loukas was located in the southeast corner of the complex. As a result, the tenth-century visitor approached the church from the east and would have first encountered

103

104

Alicia Walker

the apse fac¸ade with its concentration of pseudo-Arabic motifs. In addition, two chambers that originally flanked the sides of the west fac¸ade were open at ground level, allowing uninterrupted access to all four sides of the building.31 This plan would have facilitated the circulation of pilgrims to Luke’s tomb, the focal point of his cult, which in the tenth century was housed in an oratory that was likely located immediately to the south of the north church.32 It is reasonable to speculate that some effort was made to unify the decorative programmes of the church and oratory, perhaps employing pseudo-Arabic as a common motif. The appearance of the oratory is, however, unknown. The possibility of an intentional, dynamic use of pseudo-inscriptions at Hosios Loukas finds support in the way that actual inscriptions functioned at other medieval pilgrimage sites and churches, in particular at the nearby ninth-century church of Skripou in Orchomenos, located approximately 15 miles (24 km) northeast of Hosios Loukas.33 As Amy Papalexandrou posits, the extensive Greek inscriptions that wrap around the exterior of that building could have cued spectators to move around the church in ritual fashion as they vocalized the epigraphic texts inscribed on the monument’s walls.34 Such practices were not unique to Skripou. Linear, encircling inscriptions on churches in seventh-century Armenia seem similarly related to ritual circumambulation, raising the possibility of a wider medieval practice of dynamic interaction with inscribed structures.35 Comprehension of the content of actual inscriptions was not necessary to these ritual performances. As Christina Maranci notes, not all the participants in ceremonial processions around inscribed Armenian buildings would have been able to read the epigraphy.36 The majority of viewers thus recognised the script as a ritual sign, and its strategic disposition cued a particular physical interaction with the monument. Although the pseudo-inscriptions at Hosios Loukas cannot be read, their evocation of letters and words in sequence and their disposition around the exterior of the church still invite the viewer to circumambulate the structure. This effect is particularly evident at the eastern end of the church, where continuous friezes of pseudo-Arabic wrap around the apses and draw the viewer to the north and south walls. Furthermore, both the friezes and the isolated blocks of pseudo-Arabic observe a regular ground line and consist largely of forms with sharp angles projecting horizontally. Their ductus creates a strong lateral pull. In this way, the pseudo-inscriptions on the exterior of the north church could have served a role similar to actual inscriptions on some medieval monuments by engaging viewers both visually and physically in order to direct their pathways around the building. Pseudo-Arabic is also found in segments of a marble stringcourse that adorn the east fac¸ade and bind together the three apsidal projections at the spring-point of the window arches of the southern and central apses (Fig. 31). The forms of these pseudo-Arabic motifs occur in three types, and each is more complicated than those in cloisonn´e brick. In some sections, the marble carving displays pseudo-script composed of stacked horizontal lines and densely packed vertical

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

31. Detail of the stringcourse in the exterior apse wall of the north church showing carved pseudo-Arabic patterns, Monastery of Hosios Loukas, Byzantine, marble, ca. 961, Phokis, Greece. (Photo: Alicia Walker)

extensions, which end in thickly foliated finials; these vertical elements are regularly knotted together just below the terminal leaves. Although specific letterlike forms are difficult to isolate, the same key forms that appear in the cloisonn´e brick (recalling the letters l¯am, ‘alif, k¯af and ‘ayn) are discernible. The carvings possess a consistent rhythm of repeated motifs and a horizontal uniformity that evoke writing and guide the viewer in a lateral direction. This visual and physical engagement is strengthened by regular, deep drill holes that punctuate the designs and highlight the foliate embellishments. The stringcourse recalls epigraphic marble bands found in other middle Byzantine churches; for example, a Greek inscription around the exterior of the apsidal wall in the early tenth-century monastery church of Constantine Lips in Constantinople records the founder’s dedication and requests for divine favour. The concentrated decoration around the apsidal area of both churches recognises the particularly sacred – and therefore venerable and powerful – nature of this part of the building. The marble inscription at the monastery of Constantine Lips was originally inlaid with metal. This enhancement would have made it relatively legible from the ground, although to follow the text and read the inscription the viewer had to move along the fac¸ade. The question remains, however, whether an average viewer would have possessed a level of literacy elevated enough to understand it. Indeed, as Liz James notes, the inscriptions at Constantine Lips

105

106

Alicia Walker

could have communicated with the viewer less through their literal content and more as signs of class, status, apotropaic force or ornament.37 Again, the literal content of a medieval monumental text was neither its only nor necessarily its most important aspect. On entering the north church at Hosios Loukas, one is struck today by the empty surfaces of the interior. Only fragments of the original decorative programme are preserved. However, extensive dislocated marble carving has been recovered from throughout the site, and a significant proportion of this material is decorated with pseudo-Arabic in a form that closely resembles the marble stringcourse on the exterior of the apse. These carvings include fragments of the original templon screen (the barrier between the naos, or public area of the church, and the bema, or the altar area, which was restricted to the clergy) (Fig. 32) and of the original proskynetarion (a frame in which the main icon of the church was placed).38 In addition a carved marble slab decorated around its edges with pseudo-Arabic (now used as a tomb cover in the crypt) is thought to have been part of the tenth-century sculptural ensemble. Square cuts at each of the four corners indicate that the structure for which the slab was initially carved was equipped to support a canopy. Boura suggests that it was part of Luke’s tomb.39 Although a tomb structure could have included a canopy, this feature raises the possibility that the slab instead functioned as an altar table covered by a ciborium, in which case it would have displayed ornament consistent with that of the templon, thereby creating a unified decorative programme for the liturgical furnishings of the church.40 With these reconstructions of the interior marble reliefs in mind, we can appreciate how pseudo-Arabic originally served to direct the spectator’s interaction with the north church and to highlight the most sacred areas of the complex. As pilgrims approached the building from any side, the pseudo-Arabic motifs would have guided them around the structure, whereas the concentration of cloisonn´e brick and marble pseudo-Arabic decoration on the exterior surface of the east end would have highlighted this area as particularly important. On passing into the church, pseudo-Arabic ornaments would draw the spectator towards the focal point of the interior, the bema, and its most important liturgical and symbolic features, the templon and altar (or possibly tomb). This directed movement created an interactive relationship between the building and the viewer, such that the pilgrim found inscribed on the structure directions for the pathway he or she was to follow, as well as an indication of the most privileged, sacred areas of the building. The similar medium and style of pseudoArabic marble carving on the east fac¸ade and the templon created a subtly proleptic relationship between the outside and inside decorations of the apsidal area. This material and stylistic paralleling established not only a visual and processional connection but also an experience of anticipation and culmination that punctuated qualitative shifts in the sacred nature of space: the pseudo-Arabic motifs encircling the church and concentrated around the apse at the east fac¸ade

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

32. Reconstruction drawing of the templon screen of the tenth-century north church, Monastery of Hosios Loukas, showing pseudo-Arabic marble carving. (After L. Bouras, Ho glyptos diakosmos tou naou t¯es Panagias sto monast¯eri tou Hosiou Louka (Athens, 1980), 101, drawing 3)

were again encountered, densely clustered, at the templon, the point of transition between lay and clerical space. In this way, the pseudo-Arabic ‘inscriptions’ could have heightened the spectator’s awareness of his or her progress through the spiritual and social spheres demarcated by the architecture and decoration of the church.41 By delineating boundaries between sacred and profane spaces, the use of pseudo-Arabic in the tenth-century church calls to mind the late antique and medieval practices of marking liminal zones with protective signs.42 This potentially apotropaic function is apparent in at least two characteristics of the motifs at Hosios Loukas: the way that the pseudo-Arabic friezes wrapped around the exterior surface of the building, marking a sacred circuit around it, and once inside the structure, the concentration of pseudo-Arabic on the templon screen at the point of transition into the altar area, the most sacred zone of the church. Scholars have long raised the possibility of an apotropaic function for pseudo-Arabic motifs found in medieval Christian monuments and objects.43 Most notably, Richard Ettinghausen proposed that a common unit of pseudo-Arabic consisting of two vertical elements framing a shorter central element (which he labelled the ‘tallshort-tall syndrome’) represents a debased form of the Arabic word for God (All¯ah) and was employed as a protective sign on medieval Islamic objects and monuments.44 Although some of the pseudo-inscriptions at Hosios Loukas conform to the tall-short-tall-syndrome, no evidence suggests that knowledge of this

107

108

Alicia Walker

Islamic talisman was transferred to the Byzantine context. Still, Byzantine viewers may have possessed more general familiarity with the apotropaic properties of Arabic. Holly Edwards argues in relation to the talismanic use of Qur’anic inscriptions in Islamic art and architecture that ‘[w]earing a coat of armor covered with tiny and illegible Qur’anic passages is comparable to being inside a building enhanced with an extensive epigraphic programme; in both cases, one is protected by the immanence of the Word’.45 Powerful ‘words’ were not limited to Qur’anic phrases. Pseudo-Arabic may have functioned in Byzantine objects and monuments as a protective sign because of its foreign character and indecipherable nature; late antique and medieval magical languages were considered potent in part because they were incomprehensible to all except the otherworldly forces to whom they were directed and the few practitioners who were initiated in the mysteries of sacred and mystical knowledge.46 At Hosios Loukas, interspersed among the pseudo-Arabic motifs appear several crosses – the Christian apotropaic sign par excellence – further contributing to the possibility that the cloisonn´e brick designs collaborated to fulfil a protective function, harnessing and controlling the force of a powerful foreign script through the sign of the cross.47 Beyond a possibly apotropaic significance for pseudo-Arabic, the motifs as they appeared in this particular moment and monument may have added to the message that Hosios Loukas conveyed through the associations that Byzantine viewers held for Arabic script as the public text of Islam.48 The pilgrim’s understanding of the pseudo-Arabic on the north church may have been shaped by historical events of the late tenth century, particularly confrontations between Byzantine and Islamic military forces in the eastern Mediterranean. Luke’s posthumous miracles involved healing the sick and possessed; however his vita records that during his lifetime he was also recognised as a holy seer.49 Most famously, he is said to have foretold the conquest of Arab Crete by the Byzantines, which was achieved in 961, less than ten years after the saint’s death.50 This prophecy would have been particularly salient in the region of central Greece where Luke’s cult developed because the area suffered regularly from Arab raids in the tenth century. As discussed earlier, I endorse a chronology for the north church that places its foundation soon after the Byzantine conquest of Crete, presumably in recognition of Luke’s prophecy of this event and the fame that fulfilment of his prediction brought to the monastery. As a result of regional military encounters, Arabic script may have taken on particular meaning as a mark of Christian triumph over Muslim adversaries. After the conquest of Crete, the Byzantines claimed a large amount of spoils. According to the tenth-century history of Leo the Deacon, a wondrous diversity of objects – including ‘full sets of armor, helmets, swords, and breastplates, all gilded, and countless spears, shields, and back-bent bows’ – were publicly displayed during the triumphal procession celebrated in Constantinople.51 Although Leo does not specify that the arms and armour paraded through Constantinople were

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

inscribed, medieval Islamic armaments were commonly decorated with writing.52 Furthermore the capital would not have been the only destination for these trophies. The Byzantine regiments of the area of Hosios Loukas contributed large contingents of soldiers to other expeditions in the region, and it is likely that they participated in the Cretan campaign and returned with spoils.53 If the pseudo-Arabic motifs on the walls of the tenth-century church are interpreted in relation to trophies garnered in the Cretan victory, they can be understood as symbols of Christian triumph over Islam.54 As Papalexandrou notes, the use of foreign scripts on Byzantine buildings established a relationship with ‘other’ cultures, the meaning of which varied depending on context, but could include the ‘inferiority of a distinct cultural identity as demonstrated by the enslavement and reification of its script’.55 In this respect, the interspersing of crosses among the pseudo-Arabic cloisonn´e brick motifs on the exterior walls at Hosios Loukas may have further expressed symbolic dominance over Islam through the submission of its public text to this most eminent Christian sign.56 Support for the idea of viewing pseudo-Arabic at Hosios Loukas as a sign of triumph is found in a fragment of fresco dating to the tenth century that originally decorated the exterior wall of the narthex of the north church.57 The wall was later incorporated into the south church, and the painting was sealed behind eleventh-century marble revetment.58 This well-known image depicts the Old Testament general Joshua, who wears a helmet and head cloth that are both inscribed with a highly abstract form of pseudo-Arabic. This detail may have evoked the foreign spoils claimed by triumphant soldiers, like Joshua and like the local military men who returned from Crete. We can speculate that the latter group might have made offerings in thanks for their victory, thereby helping to establish the new north church and its triumphal programme.59 Scholars also raise the possibility that the Byzantine emperor Romanos II, during whose reign Crete was defeated, contributed to the construction of a new church at Hosios Loukas as a votive in recognition of Luke’s prophecy.60 Pseudo-Arabic inscriptions on the exterior of the Panagia would have immediately reminded the pilgrim who entered the complex that this building was made possible by Byzantine triumph over the Muslims of Crete, an accomplishment foretold by Luke.

PSEUDO-ARABIC AT HOSIOS LOUKAS IN THE ELEVENTH CENTURY

A number of changes took place at Hosios Loukas and in the surrounding region between the second half of the tenth century, when the north church was constructed, and the first half of the eleventh century, when the south church is thought to have been built and decorated. One development was the increased prominence of the monastery, which emerged as an important regional healing cult.61 This growth in popularity both facilitated and necessitated construction

109

110

Alicia Walker

of the katholikon, which was likely founded in 1011 and decorated over the subsequent centuries. Given the more than fifty-year time span between the conquest of Crete and the completion of the south church, it stands to reason that pseudo-Arabic likewise underwent a transformation in significance and purpose. I suggest that its reference changed from one of prophecy and local military triumph in the tenth century to an association in the eleventh century with the Holy Land and the sacred sites frequented by Christian pilgrims in that region, as well as the need to liberate these sites from Muslim oppression. Support for this reading comes from several features of the crypt and upper church that create a programme of loca sancta parallelisms and that project pseudo-Arabic as both the public text of the Holy Land and an attribute of Christian military saints’ battle for the faith. Hosios Loukas’s geographic position placed it at an active crossroads of tenthcentury east–west communication. Located four miles (ca. 6.5 km) from the bay at Steiri (Antikyra) along the northern shore of the Gulf of Corinth, the site offered a relatively convenient stopover for those travelling by sea. The vita of the Saint notes that people in transit to and from Europe and North Africa to points east passed through Boeotia, mooring at the coast of the Gulf of Corinth and then crossing to the ports on the Aegean Sea. The region was visited by monks journeying between Rome and Jerusalem; the bishop of Corinth, who was en route to Constantinople; envoys of an imperial delegate to Egypt, who was waylaid in Corinth; and other travellers moving between the eastern and western Mediterranean.62 These routes were determined in large part by the difficulty of travel around the Peloponnese imposed by Arab control of Crete, a situation that presumably changed after the Byzantine conquest of the island in 961. Yet by the third quarter of the tenth century, the reputation of the monastery was well established, and the site may have served as an attraction in and of itself. Easily accessible from major ports, the area presumably continued as a thoroughfare for east–west traffic. In particular, western European pilgrimage to the Holy Land surged during 1020s, and both land travel via the Balkans and sea travel from southern Italy were among the preferred routes.63 In the first decades of the eleventh century, persecutions of Jews and Christians and limitations on Christian pilgrimage to the Holy Land instigated by the Fatimid caliph al-Hakim (r. 996–1021) heightened Muslim-Christian tensions.64 The proposed date for the consecration of the southern church in 1011 positions it in the aftermath of al-Hakim’s infamous destruction of the Holy Sepulchre in 1009.65 In 1016 Basil II (r. 976–1025) banned trade with the Fatimids, an embargo that lasted until 1027, but appears not to have been rigorously enforced.66 In the 1030s an agreement was reached for the Byzantines to renovate the Holy Sepulchre, an undertaking realized by Constantine IX Monomachos (r. 1042–55) in 1042.67 During the early eleventh century, and especially in the 1020s and 1030s, Christian pilgrimage to the Holy Land saw a significant upswing as a result of millennialism, Fatimid-Byzantine commercial and political d´etente and the

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

33. General view of the bema and apse of the crypt in the katholikon (south church), Monastery of Hosios Loukas, first half of the eleventh century. (Photo courtesy Carolyn Connor)

consolidation of Byzantine military strength in the eastern Byzantine Empire.68 The decorative programmes of the south church and the crypt at Hosios Loukas – which were likely completed in the 1040s – can be interpreted in relation to these phenomena of increased pilgrimage to Palestine and imperial renovation of the Holy Sepulchre on the one hand, and continuing Muslim subjugation of the Holy Land on the other. In the eleventh century, pseudo-Arabic motifs were still visible on the eastern end of the north church, but construction of the new building to the south resulted in the concealment of these motifs on the south wall of the north church. On entering the crypt in the south church, however, the viewer encountered pseudoArabic again, at the east end in the impost blocks to either side of the bema (Figs. 33 & 34) and on the cover of the tomb to the north of the altar (the same marble slab that Boura associates with the original sculptural programme of the north church). I propose that these references to Islamic public text participated in an iconographic programme that sought to compare Hosios Loukas to the revered monasteries of the Holy Land and their pilgrimage cults. Subterranean funerary chapels are relatively rare in middle Byzantine churches. The presence of one at Hosios Loukas was likely in part the result of practical necessity: the uneven terrain of the site required a substructure to bring

111

112

Alicia Walker

the new south church to the same level as the preexisting north church. Yet the resulting architectural form recalls the crypts found in early Byzantine shrines dedicated to holy men of Palestine, such as the renowned St Sabas.69 Although less prominent in contemporary notions of Holy Land pilgrimage, these monasteries were a major attraction in the medieval period. Many of the pilgrims going to Palestine in the eighth to fourteenth centuries were monks (or individuals preparing to take their vows), and their destinations were just as often the monasteries of ascetic holy men as the traditional loca sancta of Christ’s life and deeds.70 The presence of Luke’s tomb as the focus of pilgrimage at Hosios Loukas created another parallel to the Palestinian monasteries, which were often designed with the founder’s tomb at the centre of the complex and burials of other community members encircling it.71 Indeed, the crypt at Hosios Loukas may have served as the burial chamber for the distinguished abbots of the community, who were venerated in conjunction with Luke.72 The fresco programme of the crypt further promotes a connection between the monastic community of Hosios Loukas and the great leaders and centres of Orthodox monasticism. Portraits of eleventh-century abbots of the monastery are painted in the southeast vault (including a second Luke [d. ca. 1005]; Philotheos, the likely founder of the katholikon [fl. ca. 1011]; Athanasios; and Theodosios, the possible patron of the katholikon mosaics [fl. 1048]). They are compared with four great eponymous sanctified monks in the northeast vault: Luke himself; the great Egyptian early church father Athanasios of Alexandria (fourth century); Theodosios the koinobiarches (fourth century), who founded a revered monastery in Palestine along the River Jordan just north of the Dead Sea; and the somewhat obscure Philotheos the Confessor (tenth century [?]), a monk from Asia Minor known for his miracles, especially his posthumous healing power.73 Eight additional portraits of monks in the north-west and southwest vaults include fourth- to ninth-century hermits and abbots who founded important monasteries or penned significant monastic writings; these foundations were the key centres of early monastic life in Egypt, Syria, Palestine and Bithynia.74 Other features of the programme include scenes from the life of Christ, but these familiar narrative vignettes are depicted in an unusual fashion – on the arched spaces of the walls at eye level and in a scale comparable to the viewer’s own body. As a result they depict these most hallowed events in immediate and intimate terms, functioning more like windows onto a surrounding landscape than iconic records of distant history. Pilgrims who came to Hosios Loukas after having toured the tombs of monks in the Holy Land might have felt transported back to the sacred sites of the East, whereas those who were on their way to Jerusalem and its environs would have had a first taste of the sanctified, otherworldly environments that awaited them. In either case, they would have perceived Hosios Loukas as an extension of this sacred terrain, and its monks as a community parallel to the venerable monasteries of the East. The pseudo-Arabic

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

34. Detail of the pseudo-Arabic pattern in the north impost block flanking the bema of the crypt in the katholikon (south church), Monastery of Hosios Loukas, fresco, 1040s. (Photo courtesy Carolyn Connor)

motifs prominently positioned at the entrance to the crypt’s bema further contributed to making present these distant hallowed territories under Muslim control by introducing Arabic, the public text of Fatimid Palestine, into the decorative programme. Visiting the subterranean crypt would have also contributed to the pilgrim’s spiritual experience by facilitating a sense of separation from the normal world and an ensuing state of liminality. Descent from the ground level of the churches instigated a physical transition enhanced by the shift from the bright, warm and open environment of the exterior to the dark, cool and confined space of the funerary chapel. The ‘otherness’ of the crypt may have aided the supplicant’s transformation from illness to health by creating a place seemingly outside normal time and space, a sanctified environment in which miraculous healings could more readily occur. By imitating the shrines of holy monks of the East, the crypt of Hosios Loukas would have claimed equivalent sacred status for its own saint and community, as well as comparable spiritual efficacy as a site of pilgrimage and healing. Much as the physical environment of the crypt recalled or anticipated distant holy sites of the East, pseudo-Arabic inscriptions might have been employed to enhance the alterity of the space and its affiliation with the loca sancta, specifically by recognising the status of the Christian Holy Land as a territory currently under Muslim control. The comparisons forged in the crypt could have inflected the pilgrims’ understanding of the site as they exited the space and reencountered the pseudo-Arabic motifs decorating the exterior of the north church,

113

114

Alicia Walker

which now expressed a claim to sacred otherness and authority for the site as a whole. The katholikon at Hosios Loukas includes one final concentration of pseudoArabic features, which appear in the upper church and contribute an additional dimension to our understanding of how pseudo-Arabic may have operated at Hosios Loukas across the history of the monastery and across the spaces of its shrines. Within the katholikon, pseudo-Arabic appears in three mosaics located in the south-west corner of the naos. The squinch shows the Presentation of Christ at the Temple (Luke 2: 22–40) (Fig. 35). Mary and Joseph pass the infant to the Jewish priest, Symeon, while at the left stands the Prophetess Anna. A ciborium in the background represents the Jewish Temple and is decorated with a simple motif that evokes Arabic script, recalling the letters aliph, l¯am and nun. ¯ 75 Support for the idea that the Byzantines intended this motif to be understood not just as decoration but also as inscription is found in its relationship to the Greek identifying label that runs to either side of the ciborium and reads ‘Η ΥΠΑΠΑΝΤΗ’ (The Presentation). The inscription strategically intersects with the pseudo-Arabic motif, creating a visual alignment that frames it as language, but a language contrasted explicitly with Greek. Although the motif on the dome of the ciborium does not represent actual words, its juxtaposition with Greek instructs the viewer to think of the pseudo-Arabic as a sign of linguistic – and therefore cultural – otherness. Ciboria were of course prominent liturgical furnishings in Byzantine churches, and an analogy with contemporary Byzantine types is implied in the Hosios Loukas mosaic by the presence of a cross at the summit of the canopy and on the textile hanging beneath it. The cross conveys a Byzantine typology that parallels the Old Temple of Jewish tradition with the New Temple, the Christian church. But at Hosios Loukas, the Temple is simultaneously allied with a cultural tradition distinct from that of Christian Byzantium through the inclusion of a pseudo-Arabic motif. Christian pilgrims identified the early Islamic buildings in Jerusalem, which were visible and visitable on the Temple Mount, as ancient Jewish monuments. By at least 870, the Dome of the Rock (an Islamic shrine built by the Umayyad caliph Abd al-Malik [r. 685–705] in 691) was equated with the Temple of Solomon.76 Indeed, some pilgrims explicitly stated that the Dome of the Rock was the building in which the Virgin had presented Christ to Symeon.77 Among its many noteworthy features, the Dome of the Rock is perhaps best known for its inscriptions rendered in kufic script. The ornate mosaic friezes, which encircle the inner and outer surfaces of the octagonal arcade, include an extensive epigraphic programme that runs just below the cornice.78 In the 1020s these mosaics were renewed under the Fatimids, who are credited with making significant renovations throughout Jerusalem in the first half of the eleventh century. Little evidence of these efforts remains, in large part because the Crusaders drastically transformed the city in the twelfth century, obliterating evidence of Islam from its landscape.79 Still given the prominence of inscriptions

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

35. Presentation of Christ at the Temple, southwest squinch of the katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. (Photo Credit: Gianni Dagli Orti / The Art Archive at Art Resource, NY)

in the Fatimid monuments of Cairo, it is likely that public text played a similarly important role in their Jerusalem building projects.80 A rare fragment of a Fatimid inscription, possibly dating to the tenth century and originally located in the atrium of the Church of the Holy Sepulchre, provides a sense of what must have been the pervasive presence of Arabic script throughout eleventh-century Jerusalem.81 Located in the most revered of Christian buildings, it represents both the real authority that Muslims exercised in the Holy Land and the symbolic dominion that Arabic conveyed to the Christian viewer. In the mosaic of the Presentation, Byzantine conceptions of Jerusalem, both present and past, merged to create a system of associations within which Arabic script became an ambivalent sign for the geographic distance of the loca sancta and Muslim dominance in this region. To either side of the squinch depicting the Presentation at the Temple, on the soffits of the abutting arches, the military saints Demetrios (Figs. 36 & 37) and Prokopios (Fig. 38) are depicted holding shields that show similar pseudo-Arabic motifs at their centres. Shields, greaves and swords inscribed with pseudo-Arabic were relatively common attributes of middle Byzantine military saints, appearing in tenth- to twelfth-century ivories, panel paintings, manuscripts and wall paintings.82 Yet despite the large number of military saints depicted in the mosaics at

115

116

Alicia Walker

36. Hagios Demetrios, soffit of the arch flanking the Presentation scene to the south, katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. (Photo Credit: Erich Lessing / Art Resource, NY)

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

37. Detail of Fig. 36 showing the pseudo-Arabic pattern on the shield held by Hagios Demetrios. (Photo Credit: Erich Lessing / Art Resource, NY)

Hosios Loukas, Prokopios and Demetrios are the only soldiers who hold shields inscribed with pseudo-Arabic.83 The concentration of such armaments to either side of the Presentation suggests that the pseudo-Arabic motifs on the ciborium of the Jewish Temple and the shields of the saints were intended to be understood in relation to one another. I propose that in the upper building, the themes of military triumph and the Holy Land are combined, achieving a synthesis of connotations found in the north church and in the crypt. In the tenth-century north church, the pilgrim was called to recognise pseudo-Arabic as a sign of victory over the Arabs of Crete and more broadly as a sign of Christian dominance over Islam; in the eleventh-century crypt, pseudo-Arabic contributed to a broader decorative programme that evoked distant Christian holy sites under the subjugation of another Muslim adversary, the Fatimids. In the eleventh-century upper church, Arabic is inscribed on a specific locus sanctus, the Jewish Temple, which has yet to be liberated from Islamic control, while the flanking saints appeal to the viewer to join their ranks and claim triumph over the Muslims who dominate Jerusalem. In this way, we might understand pseudo-Arabic as a means not only to direct and enhance the pilgrims’ spiritual experience at Hosios Loukas but also to call on them to continue their pious actions beyond the ritual space of the monastery.

117

118

Alicia Walker

38. Hagios Prokopios, soffit of the arch flanking the Presentation scene to the north, katholikon (south church), Monastery of Hosios Loukas, mosaic, 1040s. (Photo Credit: Erich Lessing / Art Resource, NY)

CONCLUSIONS

In sum, I propose that pseudo-Arabic at Hosios Loukas can be understood as an integral part of the visitor’s experience and understanding of this space, but that these motifs operated as mutable signs, the meaning of which changed in response to new historical circumstances of the monastic community and the pilgrims who visited it. Although most middle Byzantine viewers were unable to read Arabic, they would have encountered the language inscribed on objects from Islamic lands (such as spoils taken from the Arabs of Crete) or monuments located in the Holy Land, thereby equipping them to recognise pseudo-Arabic as a marker of alterity, whether religious, geographic or cultural. The conquest of Crete in the tenth century could have inflected pseudo-Arabic to serve as a mark of military and religious triumph, whereas its placement encircling the exterior

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

119

of the north church and concentrated at the templon argues for a role in demarcating and possibly protecting sacred space. Over the course of the tenth and eleventh centuries, a shift in the meaning of pseudo-Arabic took place as memories of the Arabs of Crete were replaced with new attitudes toward Islam forged in response to Muslim hegemony in the East. The crypt of the eleventh-century south church employs pseudo-Arabic motifs as one of several features that create a liminoid environment for the pilgrim, conjuring a sense of sacred places in the ‘other’ domain of the Holy Land and thereby enhancing the sense of Luke’s own spiritual status through comparison with the great fathers of Orthodox monasticism and the sites of their cults. Finally, in the south-west corner of the katholikon naos, the concentration of pseudo-Arabic motifs on the shields of military saints and on the dome of Jewish Temple simultaneously positions the loca sancta of the Holy Land under the sign of Muslim dominion while reminding the viewer of the Christian duty and ability to challenge this subjugation. In these ways, the meaning of pseudo-Arabic at Hosios Loukas shifted over time, responding to changes in the relationship of the monastery – and Christianity more broadly – to the Islamic world. NOTES

1. The conventional terms found in scholarship on Byzantine art and architectural history for these motifs (i.e. ‘pseudo-kufic [Cufic]’, ‘kufesque [cufesque]’) employ the name of a specific type of early medieval Arabic script, kufic, which is characterised by an emphatically horizontal base line and angular letter forms. I prefer the more general term ‘pseudo-Arabic’ because not all these Byzantine motifs coincide with the characteristics of kufic script, employing in some instances rounded or stacked (as opposed to linear, horizontal) ‘letter’ forms. Furthermore it is unclear if middle Byzantine craftsmen or patrons were aware of the distinctions between or even the existence of different types of Arabic script. The term ‘pseudo-Arabic’ avoids the implication that these motifs were understood by Byzantine makers and viewers to evoke kufic script specifically. 2. Sharon E. J. Gerstel and Julia A. Lauffenburger, A Lost Art Rediscovered: The Architectural Ceramics of Byzantium (University Park, Penn., 2001), cat. no. VI.16, 194–5. 3. See esp. George Miles, ‘Byzantium and the Arabs: Relations in Crete and the Aegean Area,’ Dumbarton Oaks Papers 18 (1964), 20–32; Nikos Nikonanos, ‘Keramoplastikes Koufikes Diakosm¯eseis sta Mn¯emeia t¯es Peri¯ Athenon,’ ¯ ox¯es ton in Afier¯oma st¯e mn¯em¯e Stylianou Pelekanid¯e (Thessalonike, 1983),

330–51. The approach to pseudo-Arabic most proximate to my own interpretation is found in Sarah A. Taft, ‘Pseudo-Kufic Decoration on Byzantine Sculpture,’ Abstracts of Papers, 14th Byzantine Studies Conference (N.p., 1998), n. pag. For additional discussion of pseudo-Arabic in medieval architecture and the role of real and pseudo-inscriptions in relation to architectural space, see Chapters 7 and 9 in this volume. 4. Regarding parallels between pseudo-Arabic inscriptions and actual Arabic script, see S. D. T. Spittle, ‘Cufic Lettering in Christian Art,’ Archaeological Journal 111 (1954), 138–52, esp. 138–42; Richard Ettinghausen, ‘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), 28–47; Chrysanthos Kanellopoulos and Lara Tohme, ‘A True Kufic ¯ Inscription on the Kapnikarea Church in Athens?’ Al-Mas¯aq 20/2 (2008), 133–9. Even in cases where actual words or phrases can be discerned, however, there is no evidence that their meaning would have been intelligible to the patrons, designers and craftsmen responsible for these Byzantine monuments or the medieval people who used the spaces in which the ‘inscriptions’ appear. For discussion of issues relating to the legibility of actual (and pseudo-) Arabic inscriptions

120

5.

6.

7.

8.

Alicia Walker

and their potential to be recognized ‘as texts’ even when unreadable, see Chapter 6 in this volume. See H. Megaw, ‘The Chronology of Some Middle-Byzantine Churches,’ Annual of the British School at Athens 32 (1931/1932), 90– 130 at 110; Nikolaos Gkioles, ‘The Church of Kapnikarea in Athens: Remarks on Its History, Typology, and Form,’ Zograf 31 (2006– 7), 20. For an interpretation of the Islamicizing decorations at Hosios Loukas as a sign of luxury and exoticism, see Andr´e Grabar, ‘La d´ecoration architecturale de l’´eglise de la Vierge a` Saint-Luc en Phocide, et les d´ebuts des influences islamiques sur l’art byzantin de Gr`ece,’ Acad´emie des Inscriptions et Belle-Lettres comptes rendus des s´eances de l’ann´ee 1971 (Paris, 1971), 15–37. Lothar Von Falkenhausen, ‘Inconsequential Incomprehensions: Some Instances of Chinese Writing in Alien Contexts,’ Res 35 (1999), 45. A similar line of reasoning is followed in Liz James, ‘“And Shall These Mute Stones Speak?” Text as Art,’ in Art and Text in Byzantine Culture, ed. L. James (Cambridge, 2007), 194–7 and 199–200; and Chapter 4 in this volume. I draw here from Richard Ettinghausen’s notion of Arabic epigraphy’s ability to ‘symbolically affirm’ the sacredness and authority of the Qur’an even in instances when an inscription is incorrectly or incompletely rendered. ‘Arabic Epigraphy: Communication or Symbolic Affirmation,’ in Near Eastern Numismatics, Iconography, Epigraphy and History: Studies in Honor of George C. Miles, ed. Dickran K. Kouymijian (Beirut, 1974), 297–318. I expand, however, beyond sacred contexts of use to consider how pseudo-Arabic could evoke Islamic culture and identity more broadly. In this respect, particularly relevant to my discussion is Irene Bierman’s notion of a ‘territorial’ function for ‘public texts’ (see n.9), defined as ‘the many ways in which officially sponsored writing reinforced and perpetuated to beholders both the solidarity of their group, binding it, and its exclusivity from other groups, bounding it from another or others.’ Writing Signs. The Fatimid Public Text (Berkeley, 1998), 16. In this respect, this chapter is allied with an effort to foreground the material context of pilgrimage as an essential factor in the participant’s experience. On this point, see Simon Coleman and John Elsner, ‘The Pilgrim’s Progress: Art, Architecture, and Ritual

9.

10.

11.

12. 13.

14. 15. 16. 17.

18.

19.

20.

Movement at Sinai,’ World Archaeology 26/1 (1994), 73–89. The term ‘public text,’ as coined by Bierman, refers to writing inscribed in the built environment by a dominant sociopolitical power in a way that is visually accessible to the population at large and rendered with the intent of communicating and creating social identity and difference. ‘The Art of the Public Text: Medieval Islamic Rule,’ in ed. Irving Lavin, World Art. Themes of Unity in Diversity, 2 vols. (University Park, Penn., 1989), 2:283– 90; Bierman, Writing Signs, 1–27, esp. xi–xii, 1, 4. In this volume, Chapter 1 makes a related argument for the ability of inscriptions to transform ‘places into politically and socially potent spaces.’ In this regard, my approach resonates with those of other authors in this volume; see esp. Chapters 2, 7, 9 and 10. On the power of liminal states in the experience of pilgrimage, see Victor Turner and Edith Turner, Image and Pilgrimage in Christian Culture (New York, 1978), 1– 39. Bierman, Writing Signs, 15. W. Robert Connor and Carolyn L. Connor, The Life and Miracles of Saint Luke of Steiris: Text, Translation and Commentary (Brookline, Mass., 1994). Also see Nicolas Oikonomides, ‘The First Century of the Monastery of Hosios Loukas,’ Dumbarton Oaks Papers 46 (1992), 245–55. Connor and Connor, The Life, ch. 54–5. Ibid., ch. 59. Ibid., ch. 64–67. Paul Mylonas, ‘Nouvelles remarques sur le complexe de Saint-Luc en Phocide,’ Cahiers arch´eologiques 40 (1992), 115–22. Stikas, To Oikodomikon Chronikon tēs Monēs Hosiou Louka Ph¯okidos (Athens, 1970). Regarding the architecture of the site, see now Paulos Mylonas, Mon¯e tou Hosiou Louka tou Steiri¯ot¯e. H¯e architektonik¯e t¯on tessar¯on na¯on (Athens, 2005). L. Boura, Ho glyptos diakosmos tou naou t¯es Panagias sto monast¯eri tou Hosiou Louka (Athens, 1980), 123–4, 134. The diverging voice is that of Carolyn Connor, who posits that the current buildings correspond to those described in the Saint’s vita. She therefore dates both churches to the tenth century. This interpretation has not received wide acceptance. Art and Miracles in Medieval Byzantium: The Crypt of Hosios Loukas and Its Frescos (Princeton, 1991), 82–3, 122.

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

21. M. Chatzidakis, ‘A propos de la date et du fondateur de Saint-Luc,’ Cahiers arch´eologiques 19 (1969), 127–50; idem, ‘Pr´ecisions sur le fondateur de Saint-Luc,’ Cahiers arch´eologiques 22 (1972), 87–88; Oikonomides, ‘The First Century,’ 249–50. Paul Mylonas, who places the construction of the katholikon between 1011 and 1048 and possibly in 1022, is refuted by Oikonomides. See Paul M. Mylonas, ‘Gavits arm´eniens et Litae byzantines. Observations nouvelles sur le complexe de Saint-Luc en Phocide,’ Cahiers arch´eologiques 38 (1990), 99–122; idem, ‘Nouvelles remarques,’ esp. 120–21; Oikonomides, ‘The First Century,’ 250. 22. Oikonomides, ‘The First Century,’ 251. Based on stylistic and technical features, Doula Mouriki also dates the mosaics and frescoes at Hosios Loukas to the 1040s; see ‘Stylistic Trends in Monumental Painting of Greece during the Eleventh and Twelfth Centuries,’ Dumbarton Oaks Papers 34/35 (1980/1), 77–124, esp. 86. 23. See the drawings in Robert W. Schultz and Sidney H. Barnsley, The Monastery of St. Luke of Stiris in Phocis (London, 1901), pls. 9, 10 and 11. 24. Stikas, To Oikodomikon Chronikon, 146–60. 25. Sheila S. Blair, Islamic Inscriptions (New York, 1998), 63, fig. 5.23; also see 56–9; 77– 80. 26. See Schultz and Barnsley, The Monastery, pl. 11. 27. Jonathan M. Bloom, Arts of the City Victorious. Islamic Art and Architecture in Fatimid North Africa and Egypt (New Haven, 2007), 29, 54; and his Chapter 3 in this volume. 28. This delegation might be the one that Luke encountered at Corinth, as reported in his vita. Connor and Connor, The Life, ch. 44. On Byzantine-Egyptian trade before and during the Fatimid period, see David Jacoby, ‘Byzantine Trade with Egypt from the Mid-Tenth Century to the Fourth Crusade,’ Thesaurismata 30 (2000), 25–77, esp. 30–47; Jonathan Shepherd, ‘Holy Land, Lost Lands, Realpolitik: Imperial Byzantine Thinking about Syria and Palestine in the Later 10th and 11th Centuries,’ Al-Qant.ara 33.2 (2012), 505–45, esp. 520–2. 29. It is unknown if the model was a monument or object copied by Byzantine craftsmen or if the motifs were transmitted by Muslim itinerant artisans working at Byzantine sites, although all these possibilities have been suggested. See Miles, ‘Byzantium and the Arabs,’ 20, 29–32; Boura, Ho glyptos diakosmos, 126;

30.

31.

32.

33.

34.

35.

36. 37. 38.

121

Anthony Cutler, ‘The Parallel Universes of Arab and Byzantine Art (with Special Reference to the Fatimid Era),’ in L’Egypte fatimide: son art et son histoire, ed. Marianne Barrucand (Paris, 1999), 638–42. The material culture of Islamic Crete is almost completely lost, but historical evidence suggests the island had a strong economy and active commercial relations, indicating the conditions for vibrant artistic and architectural production. See Vassilios Christides, The Conquest of Crete by the Arabs (ca. 824): A Turning Point in the Struggle between Byzantium and Islam (Athens, 1984), 116–22. ´ ci´c, Architecture in the Balkans Slobodan Curˇ ¨ from Diocletian to Suleyman the Magnificent (New Haven, 2010), 298. Also see Mylonas, ‘Gavitas arm´eniens et Litae byzantines,’ esp. 107–9. Here I endorse the construction sequence and analysis of architectural relationships proposed by Mylonas, but not his dating. See ‘Nouvelles remarques,’ 116–20. Regarding the role of actual inscriptions in articulating the sacred topography of Byzantine pilgrimage sites through strategic repetition and placement, see Coleman and Elsner, ‘The Pilgrim’s Progress,’ esp. 78– 81. Amy Papalexandrou, ‘Text in Context: Eloquent Monuments and the Byzantine Beholder,’ Word and Image 17.3 (2001), 259– 83. Christina Maranci, ‘Performance and Church Exterior in Medieval Armenia,’ in Visualizing Medieval Performance. Perspectives, Histories, Contexts, ed. Elina Gertsman (Aldershot, U.K., 2008), 16–32, esp. 18–21. Ibid., 20. James, ‘“And Shall These Mute Stones Speak?”’ 197–203. Boura, Ho glyptos diakosmos, 98–109, 130– 32; 101 and 103, drawings 3 and 4; figs. 165–72. Also see Taft, ‘Pseudo-Kufic Decoration on Byzantine Sculpture,’ n. pag.; Sophia Kalopissi-Verti, ‘The Proskynetaria of the Templon and Narthex: Form, Imagery, Spatial Connections, and Reception,’ in Thresholds of the Sacred: Architectural, Art Historical, Liturgical, and Theological Perspectives on Religious Screens, East and West, ed. Sharon Gerstel (Washington, D.C., 2006), esp. 108– 9. Boura has identified a pair of marble parapet reliefs with pseudo-Arabic borders that appear below the windows in the exterior east wall of the katholikon (south church) as

122

39. 40.

41.

42.

43. 44. 45.

46.

47. 48. 49.

Alicia Walker

remains from the original embellishments of the interior of the north church. She argues that materials yielded from alterations to the north church at the time of the construction of the south church were incorporated into the south church to give the two buildings a unified appearance. Ho glyptos diakosmos, 99– 100, figs. 171–2. Boura, Ho glyptos diakosmos, 112–14, 133. The altar area of the north church can accommodate the marble slab: the bema has a width of ca. 3 m, whereas the sarcophagus lid measures 1.83 m in length. A similar argument regarding the function of graffiti in early Christian sacred spaces is made in Chapter 2 in this volume. For the medieval practice of protecting doorways and walls with talismanic markings, including inscriptions, see James, ‘“And Shall These Mute Stones Speak?”’, 197–200; Julia Gonnella, ‘Columns and Hieroglyphs: Magic Spolia in Medieval Islamic Architecture of Northern Syria,’ Muqarnas 27 (2010), 103– 20; and Chapters 4 and 9 in this volume. For instance, see Miles, ‘Byzantium and the Arabs,’ 27; Boura, Ho glyptos diakosmos, 126. Ettinghausen, ‘Kufesque in Byzantine Greece,’ 35–9, 41–5. Holly Edwards, ‘Text, Context, Architext: The Qur’an as Architectural Inscription,’ in Fisher, Brocade of the Pen. The Art of Islamic Writing, ed. Carol Garrett (East Lansing, Mich., 1991), 67. On the magical, talismanic properties of arcane languages and scripts in Byzantium, see Alicia Walker, ‘Meaningful Mingling: Classicizing Imagery and Islamicizing Script in a Byzantine Bowl,’ The Art Bulletin 90.1 (2008), 32–53; James, ‘“And Shall These Mute Stones Speak?”’, 197–200; Gonnella, ‘Columns and Hieroglyphs,’ 104–7. Images of the crosses are found in Schultz and Barnsley, The Monastery, pls. 9–11. For the definition of ‘public text,’ see n.9. Connor and Connor, The Life, ch. 23–4, 30, 44, 59, 60, 62, 64. Hosios Loukas is located ca. 37 miles (60 km) from the ancient site of Delphi, the home of the renowned Pythian oracle of Apollo. The region had a longstanding association with prophecy from GrecoRoman times, which may have had bearing on Luke’s promotion as a seer. Indeed, Luke’s vita compares him to the ancient oracle, denouncing the pagan source as ‘oblique and ambiguous,’ while Luke’s counsel ‘was beneficial, clearly, and unambiguously expressed.’ Connor and Connor, The Life, ch. 58.

50. Connor and Connor, The Life, ch. 60. The vita also notes he regularly warned his followers of other ‘invasions of the heathens’ as well as an imminent Bulgarian (‘Scythian’) attack. Connor and Connor, The Life, ch. 24, 62. 51. Alice-Mary Talbot and Denis F. Sullivan, trans., The History of Leo the Deacon: Byzantine Military Expansion in the Tenth Century (Washington, D.C., 2005), 81; cited by Carolyn L. Connor, ‘Hosios Loukas as a Victory Church,’ Greek, Roman, and Byzantine Studies 33.3 (1992), 293–308 at 300. 52. An idea of their possible appearance is provided by inscribed ninth- and tenthcentury Islamic armaments. See Bashir Mohamed, ed., The Arts of the Muslim Knight: The Furusiyya Art Foundation Collection (Milan, 2008), cat. nos. 7, 68–70, 87, 90–91, 104, 232. The chronicler known as the continuator of Theophanes reports that Muslim soldiers who fought against the Byzantines at Amorion in 838 wrote on their shields. Theophanes Continuatus, Chronographia, ed. E. Bekker, Corpus Scriptorum historiae byzantinae 22 (Bonn, 1838), 125, ch. 30, ll. 16–21; cited by Piotr Grotowski, Arms and Armour of the Warrior Saints: Tradition and Innovation in Byzantine Iconography (843–1261) (Leiden, 2010), 242. 53. Connor, ‘Hosios Loukas as a Victory Church,’ 298–9. 54. This reading complements that of Connor, who, however, does not expressly cite pseudoArabic as a symbol of triumph. I disagree with Connor’s identification of the eleventhcentury katholikon as the victory church (ibid., 305–6) and instead associate this status with the tenth-century church. 55. Amy Papalexandrou, ‘Memory Tattered and Torn: Spolia in the Heartland of Byzantine Hellenism,’ in Archaeologies of Memory, eds. Ruth M. Van Dyke and Susan E. Alcock, (Oxford, 2003), 72. 56. Regarding the use of multilingual inscriptions as a mark of cultural-political conquest and control, see Chapter 6 and 8 in this volume. 57. Connor, ‘Hosios Loukas as a Victory Church,’ 304–5. 58. It was revealed during archaeological explorations in the 1960s. See Stikas, To Oikodomikon Chronikon, 174–8. 59. Connor, ‘Hosios Loukas as a Victory Church,’ 299–303. 60. Boura, Ho glyptos diakosmos, 134. Connor instead suggests that the general responsible for the victory, Nikephoros Phokas, could

Pseudo-Arabic ‘Inscriptions’ and the Pilgrim’s Path at Hosios Loukas

61.

62.

63.

64.

65.

66. 67.

have been the donor. ‘Hosios Loukas as a Victory Church,’ 301. Connor and Connor, The Life, ch. 68–84. Alice-Mary Talbot notes that Luke’s cult was not only the third most active site for middle Byzantine miraculous healings but also one of the few shrines whose prominence endured beyond a few decades. ‘Pilgrimage to Healing Shrines: The Evidence of Miracle Accounts,’ Dumbarton Oaks Papers 56 (2002), 153–73, esp. 156, 163, 165, Appendix 3. Connor and Connor, The Life, ch. 9, 21, 41–2, 44, 52, 66; Oikonomides, ‘The First Century,’ 253. Although the area of Boeotia is not specified as a stopover for these pilgrims. See Shepherd, ‘Holy Land,’ 523; Franc¸oise Micheau, ‘Les itin´eraires maritimes et continentaux des p`elerinages vers J´erusalem,’ in Occident et Orient au Xe si`ecle. Actes du IXe congr`es de la soci´et´e des historiens m´edi´evistes de l’enseignement sup´erieur public (Paris, 1978), 89–91. Alice-Mary Talbot, ‘Byzantine Pilgrimage to the Holy Land from the Eighth to the Fifteenth Century,’ in The Sabaite Heritage in the Orthodox Church from the Fifth Century to the Present, ed. J. Patrich (Leuven, 2001), 100– 101; David Jacoby, ‘Bishop Gunther of Bamberg, Byzantium and Christian Pilgrimage to the Holy Land in the Eleventh Century,’ in Zwischen Polis, Provinz und Peripherie. Beitr¨age zur byzantinischen Geschichte und Kultur, ed. Lars M. Hoffmann (Wiesbaden, 2005), 279– 84. Shepherd emphasises, however, that Antioch was of greater commercial and strategic significance to Byzantine interests in SyriaPalestine than was Jerusalem, and the vandalism against the Holy Sepulchre would not have been a major factor in shaping Byzantine policy in the area. Rather it was the growing interest in pilgrimage to the site that would have motivated imperial intervention. Shepherd, ‘Holy Land,’ 512, 519–20, 527–35. Ibid., 519–20. Jacoby, ‘Bishop Gunther,’ 271–4; Shepherd, ‘Holy Land,’ 530–5.

123

68. Talbot, ‘Byzantine Pilgrimage,’ 101; Jacoby, ‘Bishop Gunther,’ 270–2; Shepherd, ‘Holy Land,’ 515–19, 522–6. 69. Connor, Art and Miracles, 72–4. 70. Talbot, ‘Byzantine Pilgrimage,’ 101–2; Shepherd, ‘Holy Land,’ 524–6. 71. John Binns, ‘The Concept of Sacred Space in the Monasteries of Byzantine Palestine,’ in Case Studies in Archaeology and World Religion, ed. T. Insoll, BAR International Series 755 (Oxford, 1999), 30; Connor, Art and Miracles, 49. 72. Oikonomides, ‘The First Century,’ 251. 73. Connor, Art and Miracles, 48–53, fig. 10, although see revisions to Connor’s identification of the vault portraits in Oikonomides, ‘The First Century,’ 250–1. 74. Connor, Art and Miracles, 47–48. 75. Compare, for example, the inscriptions on tenth-century Fatimid textiles in Claus-Peter Haase, ‘Some Aspects of Fatimid Calligraphy on Textiles,’ L’Egypte fatimide: son art et son histoire, in ed. Marianne Barrucand (Paris, 1999), 342–43. 76. Pamela Berger, The Crescent on the Temple: The Dome of the Rock as Image of the Ancient Jewish Sanctuary (Leiden, 2012), 55–73, 55. 77. Carol Herselle Krinsky, ‘Representations of the Temple of Solomon before 1500,’ Journal of the Warburg and Courtauld Institutes 33 (1970), 1–19, 6. 78. Oleg Grabar, The Shape of the Holy: Early Islamic Jerusalem (Princeton, 1996), esp. 56– 104. 79. Bloom, Arts of the City Victorious, 81–3. 80. See Bierman, Writing Signs. 81. The inscription was removed in the nineteenth century. See Heribert Busse, ‘Die ¨ ‘Umar-Moschee im ostlichen Atrium der Grabeskirche,’ Zeitschrift des Deutschen Pal¨astrina-Vereins 109.1 (1993), 73–82, esp. 75–6, 79–80. 82. Grotowski, Arms and Armour, 190, 241–3. 83. In addition, the military saint Merkurios, positioned on the surface of the soffit opposite Procopios, wears greaves inscribed with pseudo-Arabic. See Stikas, To Oikodomikon Chronikon, pl. 40, 47a.

鵻 CHAPTER SIX

ARABIC INSCRIPTIONS IN THE CAPPELLA PALATINA: PERFORMATIVITY, AUDIENCE, LEGIBILITY AND ILLEGIBILITY 鵼 Jeremy Johns

INTRODUCTION: ARABIC INSCRIPTIONS IN NORMAN SICILY

The twelve sons of Tancred, baron of Hauteville in Normandy, who crossed the Alps in the mid eleventh century and became leaders of the Normans in Southern Italy and ancestors of the kings of Sicily, were men of war, not of letters. Nonetheless, the younger of them soon acquired the epigraphic habit, and by the 1080s, Latin, Greek and Arabic texts were regularly issued in their names. For example, Robert Guiscard appeared as a duke of imperial Rome in the Latin text on the fac¸ade of Salerno Cathedral,1 as a Byzantine duke in the Greek inscription from San Pietro de Balnearis in Palermo dated to the year 6589 in the era of ¯ al-ajall malik .siqill¯ıya (‘Robert, the mighty Constantinople,2 and as ab¯art al-duqa duke, the ruler of Sicily’) in the Arabic legend in the field of a tar`ı minted at Palermo in the year 464 of the hijra (1072 CE).3 It is highly unlikely that Robert would himself have been able to read any of these texts. Indeed, the persistence of the exclusively Islamic profession of faith on the obverse of the Palermo tar`ı of 1072 seems to indicate that neither Robert nor his Christian officials could have been aware, even at second hand, of the content of the Arabic legends.4 Thus it seems clear that, from the outset, such inscriptions were less important for their content than as demonstrations that the Norman conquerors had appropriated the languages, scripts and styles of the previous rulers of the South.5 Such assertions of the legitimacy of their rule over all their new subjects and territories served primarily as nonliterary signifiers of the triumphant appropriation of what are arguably the most immediately distinctive visual cultural identifiers of the communities conquered and now ruled by the Normans. The distinctively

124

Arabic Inscriptions in the Cappella Palatina

different scripts of Arabic, Greek and Latin were used in this way throughout the brief reign of the De Hautevilles to symbolise the three dominant cultures of the South – Arabic for the Muslims, Greek for the Greek Orthodox Christians, and Latin for the Christians who followed the rite of Rome. Even after the collapse of the dynasty and the succession of the Hohenstaufen emperor Henry VI, his panegyrist, Peter of Eboli, would still describe Palermo as ‘a happy city, blessed with a three-tongued people’, and would visualise that populace as the trilingual royal chancery with its Greek, Saracen and Latin scribes side by side, each writing his own distinctive script.6 The earliest Norman inscriptions had essentially continued or revived the epigraphic practices of the earlier rulers of the South, but with the coronation of Robert Guiscard’s nephew, Roger, as the first king of the new kingdom of Sicily on Christmas Day 1130, a new deliberately heterogeneous royal art and architecture began to be assembled and developed. This new style did not perpetuate an indigenous Sicilian tradition inherited from the previous rulers of the island, but rather depended upon the importation of art and artists from the contemporary Mediterranean – Byzantium, Rome and, most relevant here, the Islamic Mediterranean, especially Fatimid Cairo but also the Berber emirates that succeeded the Fatimids in the central Maghrib, the Zirids and the Hammadids, and the Almoravids of Morocco. From this time, new styles of Arabic epigraphy were introduced and began to be developed that reflected contemporary usage and did not hark back to the Kalbid past in Sicily. Under King Roger and his successors, such new Arabic texts came to be used systematically to project, enhance and manage the royal image both on royal buildings and on the ceremonial vestments worn by the king in the churches, palaces and processions where he was displayed to his subjects, and on his coins and documents that reached even the most remote corners of his kingdom.7 The Norman kings and their ministers systematically employed bilingual, trilingual and, in one famous surviving example,8 quadrilingual texts to convey the message that the cohesive power of the Norman ruler had bound together the different cultures and peoples of the kingdom into a single Sicilian people, the populus trilinguis. Eugenius of Palermo, the Greek minister and panegyrist of William II, expressed that claim in these words: ‘Do not you [the king] harmonise the inharmonious, and mix together the unmixable . . . with wise foresight blending and uniting into a single race disparate and incongruent peoples’.9 Arabic, Greek and Latin texts were systematically contraposed both on single artefacts – such as bilingual coins from the royal mint, bilingual documents from the royal chancery, bilingual inscriptions on regalia from the royal wardrobe, and trilingual inscriptions on royal monuments – and in single architectural settings, such as the great royal churches of which Cefalu, ` the Cappella Palatina and Monreale survive largely intact. By far the richest and most interesting example of this purposeful contraposition of text is the Cappella Palatina, where the Greek and

125

126

Jeremy Johns

39. Palermo, Palazzo Reale, Cappella Palatina. The painted wooden ceilings of the nave and the two aisles, seen from below. (After Brenk, ed., 2010, 4, unnumbered figure before p. 384. C 2010 Franco Cosimo Panini Editore Spa) Photo Ghigo Roli 

Latin inscriptions of the mosaics are set against the Arabic texts of the painted ceilings of the nave and two aisles (Fig. 39), of the doors and of the moveable objects used within the chapel.10 Although in what follows discussion focuses upon the Arabic inscriptions of the Cappella Palatina,11 the multilingual background against which they are set in the chapel, as well as in the royal art and architecture of Norman Sicily as a whole, is not ignored. A very few words here suffice to introduce the content of the Arabic inscriptions of the Norman kings, which may be divided into three distinct groups. One commemorates acts of construction, including the trilingual inscription from a public building in Termini Imerese overseen by King Roger’s eunuch PeterBarrun, ¯ 12 and the trilingual inscription from King Roger’s water-clock now displayed outside the entrance to the Cappella Palatina.13 A second group comprises panegyric verses that once decorated the royal palaces, including the opus sectile inscriptions from the Cappella Palatina (Fig. 40),14 and from Roger’s lost palace in Messina.15 The third group consists of supplications (adʿiya) made to God that he may bestow upon the king a variety of blessings, qualities and virtues. This last group is the largest to survive and is represented by the inscription embroidered in gold on the hem of King Roger’s mantle,16 by the inscription from the bell tower of the lost royal church of San Giacomo la Mazara,17 and – most importantly – by the Arabic inscriptions from the painted ceilings of the Cappella Palatina (Fig. 41).18

Arabic Inscriptions in the Cappella Palatina

40. Palermo, Galleria Nazionale della Sicilia, inventory nos. 5104 & 5105. The two longest of the three fragments of Arabic verse inscriptions in opus sectile from the Cappella Palatina. C 2010 Jeremy Johns) (Photo Jeremy Johns 

41. Palermo, Palazzo Reale, Cappella Palatina. Detail of a stellate coffer (south side, third from west) from the central zone of the painted wooden ceiling of the nave. (Khalili Research Centre C 2010 University of Edinburgh & The Barakat Trust) Archive, Slide no. ISL15777. 

127

128

Jeremy Johns

ARABIC INSCRIPTIONS IN THE CAPPELLA PALATINA

The principal palace of the Norman kings in Palermo, built on the highest point within the walls on the site of the acropolis of the ancient city and replacing a palace of the Kalbid emirs, was extensively rebuilt and enlarged by King Roger and his successors. The structure now known as the Cappella Palatina was built on the first floor of an inner court and rested directly upon the vaulted roof of an earlier and smaller chapel of St. Peter. The structural history of the Cappella Palatina is complex and still highly problematic and controversial. However, it is now generally accepted that it was originally intended to combine a sanctuary in the east end with a three-aisled basilical hall extending westwards from the sanctuary, with a platform for the royal throne in the centre of its western wall. The manner in which these two elements – a sanctuary for the celebration of the divine liturgy oriented from west to east, and a royal hall focused upon the king enthroned in the west end – actually functioned is far from clear and may have changed and developed over time as King Roger and his successors adapted the original plan, altered the structure and added to its decoration. At the time of its consecration, probably in June 1143, the walls, vaults and central cupola of the sanctuary were extensively decorated with mosaics, while the walls of the western hall seem to have been undecorated below the elaborately painted wooden ceilings (Fig. 39) and above the polychrome opus sectile pavements and dadoes. Although earlier scholarship tended to make a clear distinction between the ‘Byzantine’ mosaics of the sanctuary and the ‘Islamic’ ceilings and pavements of the western hall, recent studies have tended to speculate that there was from the outset a programme that unified the two elements; these studies have attempted to reconstruct detailed examples of the ways in which the mosaics of the sanctuary were intended to collaborate with the ceilings and pavements of the western hall in the projection of a single royal programme.19 Today, the Arabic inscriptions that survive in the Cappella Palatina can scarcely be seen in the upper reaches of the painted ceilings of the western hall (Figs. 39 & 41), but it is likely that, in the early 1140s before its mosaics were set, Arabic epigraphy would have been the most immediately conspicuous decorative element of the royal hall. The doors leading into the hall were furnished with metal handles decorated with Arabic supplications (adʿiya).20 Inside, the doors may have been framed by white marble panels inlaid with porphyry and serpentine opus sectile inscriptions reciting Arabic panegyric verses in praise of King Roger.21 After the nave mosaics were completed,22 the gap between their upper border and the lower border of the wooden ceilings was covered by the tall wooden cornices, now overpainted with Latin texts commemorating restorations commissioned by the kings of Aragon, but in all likelihood originally bearing Arabic inscriptions.23 Above them, hundreds of panels in the wooden ceilings were painted with Arabic inscriptions and pseudo-epigraphic designs.24 Even the furnishings of the chapel would have been decorated with Arabic texts: the ivory chests and

Arabic Inscriptions in the Cappella Palatina

reliquaries, such as the famous incrusted casket still conserved in the treasury with its long inscriptions;25 the textile hangings on the walls and the curtains in the arches of the arcades;26 the liturgical cloths, vestments, vessels and plate; and the glass lamps suspended from the ceiling27 – all would have borne Arabic inscriptions. So, too, would the vestments of the king and his courtiers and servants, and the cushions and carpets upon which they sat.28 The original location of the water-clock that bore the trilingual inscription that is now displayed on the wall outside the Cappella Palatina is unknown, but if it was originally associated with the royal chapel, then it proclaimed King Roger’s mastery of not just the three cultures of his kingdom but also of time itself.29 And, if the famous trilingual psalter in the British Library (MS Harley 5786) was indeed used in the royal chapel, then the liturgy was sung there in Arabic, just as it undoubtedly was in Greek and in Latin, and that psalter together with other Arabic liturgical texts may also have been kept in the chapel.30 Such a proliferation of Arabic texts in the Cappella Palatina may today be difficult to imagine, and it may help to recall that, for centuries before the rise of Islam and thereafter until today, Arabic was one of the languages of Christianity. After the Islamic conquests and the spread of Arabic as the language of government, learning and commerce throughout much of the Mediterranean, many indigenous Christian communities from Spain to the Tigris adopted that language for common usage and even came to worship in Arabic. In Norman Sicily, many Greek Orthodox Christians were Arabic speakers, and many leaders of their community were prominent in and around the royal court. Amongst them were both immigrants from the Islamic Mediterranean and the descendants of South Italians and Sicilians who had adopted Arabic under Islamic rule. Chief amongst the former was George of Antioch, King Roger’s first minister and the main architect of his multicultural monarchy. When George built Santa Maria dell’Ammiraglio at about the same time as the Cappella Palatina or perhaps slightly before it,31 he placed an Arabic inscription in the drum of the dome: painted on sixteen wooden boards, it recited the Epinikion Hymn from the Anaphora and the Great Doxology, corresponding to the Sanctus of the Roman liturgy. The Archangels, who are depicted in the mosaics of the dome prostrating themselves before the throne of Christ Pantokrator, are thus seen to worship Him by singing the hymn of victory that is written in Arabic immediately beneath their feet.32 The status of Arabic as a Christian language may perhaps have facilitated the appropriation of Arabic by a Christian king. However, the content of most royal inscriptions was not explicitly Christian, but rather deliberately appropriated Islamic texts and employed them to serve the Norman ruler. In the Arabic supplications (adʿiya), a series of royal qualities and virtues that were typically invoked mina ll¯ah (‘from God’) were here summoned on behalf of the Christian king. In the opus sectile inscriptions in the Cappella Palatina (Fig. 40), the royal palace is figured as the Kaʿba at Mecca, whereas the reader is imagined as a Muslim pilgrim and his acts of homage and submission to King Roger are

129

130

Jeremy Johns

described in terms proper to the sacred rituals of the Islamic h.ajj.33 This was no syncretistic attempt to reconcile Christianity and Islam, but rather a masterful act of cultural appropriation intended to enhance the image of the king, of precisely the kind that Ibn Jubayr was to see as evidence of the Norman king’s ‘ability’, one that he perceived as a dangerous threat to Islam.34 PERFORMATIVITY

No primary source independent of the Arabic inscriptions that decorate the Cappella Palatina suggests that they prescribed the encounter between the viewer and the building. Nor does the location and display of those Arabic inscriptions that survive in situ suggest that they could have guided or led the viewer around the building. Nonetheless, the content of one group of inscriptions and a figural scene incorporating part of another group do suggest that the Arabic inscriptions referred to ceremonies and rituals that were appropriated by the Norman king and that, if they were not actually performed in the royal hall in the western end of the chapel, were at least explicitly evoked by the texts. The three fragments of inscriptions in opus sectile from the Cappella Palatina are all that survive from what was clearly once an extensive programme of Arabic verse inscriptions that decorated King Roger’s palace in Palermo.35 It is almost inconceivable that these verses would not have been recited in front of the king, at least on the occasion of their installation or of the inauguration of the hall, although no witness has left a record of such a public performance of their content. It is not known for certain where the inscriptions were originally located, but their provenance, form and content all suggest that they originally framed doorways leading to the royal hall in the western end of the Cappella Palatina. The smallest of the three is too fragmentary to read and translate with confidence, but the other two fragments may be transcribed and rendered as follows: Fig. 40A (metre ramal)

[ . . . wa-]lthim ruknahu¯ baʿda ltiz¯amin ∗ wa-taʾammal m¯a h.aw¯ahu¯ min jam¯al¯ı wa [ . . . ] [ . . . and] kiss its corner after pressing yourself [to it] ∗ and contemplate the beauties that it contains. And . . . . . . . . . .. Fig. 40B (metre k¯amil)

Arabic Inscriptions in the Cappella Palatina

1. [ . . . ] munyatan ∗ wa-tuʿa¯ jilu l-taqb¯ıla wa-l-tasl¯ım¯a s¯am¯a Ruj¯aru [ . . . ] 2. d¯ar a[l- . . . ] 1. [ . . . ] an object of desire ∗ and you hurry to give the kiss and to make the submission. Roger has sought to surpass [ . . . ] 2. The house of [ . . . ]

Both contain a series of explicit references to the ceremonies of the .taw¯af that form part of the h.ajj or Islamic pilgrimage to Mecca. The pilgrim must circumambulate the Kaʿba seven times, keeping it always on his right and beginning and ending at the Black Stone that is set into its eastern corner (rukn; cf. A, rukn). The first three circuits must be performed at the rapid pace called ramal – the name of the ‘running’ metre in which the verses of A were composed – and the last four, at a normal pace. Each time he passes it, the pilgrim should make a special effort to kiss (taqb¯ıl; cf. B, al-taqb¯ıl) the Black Stone or at least to touch it (istil¯am; cf. B, al-tasl¯ım). After he has completed seven circuits, in order to acquire a special blessing, the pilgrim presses himself against al-multazam (cf. A, iltiz¯am), a stretch of the wall of the Kaʿba between the door and the Black Stone, drinks from the holy well of Zamzam, and finally kisses the Black Stone in farewell (note the sequence; cf. A, “kiss its corner after pressing yourself to it”). In these verses, King Roger’s hall becomes the Kaʿba, and the reader is figured as a pilgrim whose acts of homage and submission to the king and his palace are represented as the sacred ceremonies of the h.ajj. This depiction seems to go beyond the common literary trope in which a visitor is compared to a pilgrim and his destination to the Kaʿba: it seems highly likely that these verses were inspired by a panegyric addressed to the Fatimid caliph in which his palace was similarly figured as an analogue of the Kaʿba. Indeed, the hypothesis that best accounts for the presence of such inscriptions in the Cappella Palatina is that they imitated similar inscriptions that once decorated the lost Fatimid palaces of Cairo. There is some reason to think that visitors to the Fatimid court may have been obliged to embrace and to kiss the gates of the palace in a ritual intended to act out the idea that, just as the im¯am was God’s representative on earth (khal¯ıfat All¯ah), so was his palace the analogue of the House of God (Bayt All¯ah) in Mecca. Could it be that the inscriptions in the Cappella Palatina similarly prescribed the rituals that visitors to King Roger’s palace were expected to perform? This is precisely what the texts claim to do. In fragment A, the verbs ilthim (‘kiss’) and taʾammal (‘contemplate’) are both in the imperative mood addressed to a singular masculine subject, whereas the indicative verb (tuʿa¯ jilu ‘you hurry’) in fragment B also has a masculine second-person singular subject, so that the verses address the reader directly and instruct him to perform these rites. What is more, some of the surviving fragments of the Arabic verses in the opus sectile inscriptions from Messina direct a specific group of visitors to King Roger’s palace: y¯a maʿshara

131

132

Jeremy Johns

¯ ¯ (‘Enter, o nobles of the kingdom, for l-mulki dkhulu-hu fa inna-hu d¯aru l-khulud it [the palace] is the abode of eternal life’).36 Difficult as it is to imagine that King Roger’s Norman barons would have performed such rites, it is conceivable that his crypto-Muslim servants and Greek Christian ministers, who had been educated and trained within the palace and would have been familiar with comparable ceremonies from Byzantine and Islamic courts, did actually perform them. The question remains open, but another piece of evidence suggests that other ceremonies, which the Norman barons would have found no less outlandish, may have been performed in the royal hall. The largest corpus of Arabic inscriptions in the Cappella Palatina comprises the adʿiya or supplications made to God that he may bestow upon the king a variety of blessings, qualities and virtues. A flavour of these may be had from one of the kufic inscriptions painted in white that form the borders of the star-shaped coffers of the central part of the ceiling of the nave (Fig. 41): alnas.r wa-l-iqb¯al wa-l-yumn wa-l-z.afar wa-l-sal¯ama wa-l-h.ifz. wa-l-h.im¯aya wa-l-yumn al-ʿizz al-z.afar wa-l-nas.r al-ʿizz al-kam¯al wa-l-saʿd wa-l-sal¯ama wa-l-yumn wal-ʿizz wa-l-nas.r wa-l-saʿd wa-l-kam¯al wa-l-i[jl¯al/fd.a¯ l/qb¯al ?] (‘victory and propitious-fate and bliss and attainment and security and vigilance and protection and bliss and power, attainment and victory, power, perfection and good-fortune and security and bliss and power and victory and good-fortune and perfection and [magnificence/generosity/propitious-fate ?]’).37 Eighteen of the twenty-six adʿiya (not counting the adjectives) used in the Cappella Palatina appear ubiquitously throughout the medieval Islamic world in supplicatory inscriptions, especially on portable objects such as ceramics, ivories, metalwork and textiles, but also on buildings.38 Hjalmar Torp was therefore almost certainly mistaken in asserting that the adʿiya in the Cappella Palatina are merely translations into Arabic of Byzantine acclamations.39 Nonetheless, the possibility remains that these adʿiya were indeed recited in the royal hall of the Norman kings. Although no contemporary source describes any such ceremonial, a unique panel in the painted ceiling of the nave of the Cappella Palatina may depict just such a scene of acclamation (Fig. 42). It is located immediately to the north of the throne platform in the west end of the hall at the base of the first large unit of the muqarnas zone of the north side, beneath a pair of panels depicting seated rulers.40 In good light, a viewer blessed with perfect eyesight could have read the scene (but not, I dare say, the inscription) from the pavement of the hall and from the throne platform itself. The panel is occupied by a male half-figure, fully frontal, his eyes turned upwards, bearded and with long dark hair falling over his shoulders. His two arms are outstretched and in each of his closed fists he holds a cross. He wears a robe with decorated golden armbands and cuffs, and with an unusually elaborate golden band around the neck that falls vertically down the chest. The folds of the fabric of his robe are indicated by fine white lines that form concentric oblong swirls at the bottom of his torso, raising

Arabic Inscriptions in the Cappella Palatina

42. Palermo, Palazzo Reale, Cappella Palatina. Detail of panel at the base of the first large unit of the muqarnas zone from the west end on the north side of the ceiling of the nave, showing a male half-figure holding two crosses, above an Arabic inscription. (Khalili Research C 2010 University of Edinburgh & The Barakat Centre Archive, Slide no. ISL15197.  Trust)

the possibility that the artist may have intended to represent the figure on his knees; if not, then the figure is truncated at the waist in a peculiarly arbitrary fashion. Below the figure, a single line of Arabic in cursive script runs the full width of the panel, reciting the rhyming couplet of al-ʿizz al-d¯aʾim al-saʿd al-q¯aʾim alʿizz . . . (‘lasting power, established good fortune, . . . power . . . ’). The artist was unable to fit the adjectival complement of the final noun (al-ʿizz, ‘power’) into the panel, but there can be no doubt that it was either al-d¯aʾim (‘enduring’) or al-q¯aʾim (‘established’); this couplet is the most frequently repeated phrase in the Arabic inscriptions of the ceilings. This is the only instance in which a duʿa¯ is associated with a figurative scene, and it is the only figurative scene to incorporate a formal text.41 The pair of hand-crosses displayed by the figure suggest that the panel may depict a ceremony that was related to Byzantine benedictory rituals featuring a pair of handheld blessing-candles (dik¯erion), which developed from first imperial and then patriarchal ceremonies in the tenth to twelfth centuries.42 The unique association of this scene with written adʿiya suggests that the ceremony evoked is likely to have been a supplicatory ritual, in which God was beseeched to bestow ‘enduring power [and] established good fortune’ upon the king. If so, then the half-figure might represent a priest or bishop, and the decorated golden band around the neck of his robe his liturgical stole. Alternatively, it may be significant

133

134

Jeremy Johns

that the figure’s beard and long hair falling over his shoulders are also to be found in the representations of rulers in the ceiling – thus raising the possibility that the scene may evoke a ceremony in which the king either prayed to God for lasting power and good fortune, or blessed the assembly on the model of Byzantine imperial practice. In that case, the decorated golden band might represent the royal loros. The location of the panel in the immediate vicinity of the throne platform strengthens the possibility that it may refer to a royal ceremony. Were this to be so, then this panel would belong to a group of approximately ten genre scenes that depict life in King Roger’s palace,43 but would be the only one to refer to a ceremony that was actually performed within the western hall of the Cappella Palatina – and the only one in which image and text were employed in self-reference. No contemporary witness confirms that the lists of adʿiya or supplications that dominate the surviving Arabic inscriptions from Norman Sicily were ever recited aloud, but the texts themselves suggest that they may have been intended for public performance. The inscription from King Roger’s mantle is composed in rhymed prose or sajʿ, which breaks up the text into a series of short, rhyming couplets that would have been suitable for recitation aloud as a litany of royal blessings: mimm¯a ʿumila bi-l-khiz¯anati ¯ l-malak¯ıyati l-maʿmurati bi-l-saʿdi wa-l-ijl¯ali wa-l-majdi wa-l-kam¯ali wa-l-t.awli wa-l-ifd.a¯ li ¯ wa-l-iqb¯ali wa-l-qabuli wa-l-sam¯ah.ati wa-l-jal¯ali wa-l-fakhri wa-l-jam¯ali ¯ wa-bulughi l-am¯an¯ı wa-l-¯am¯ali wa-t.ı¯bi l-ayy¯ami wa-l-lay¯al[¯ı] bi-l zaw¯ali wa-l¯a ntiq¯ali bi-l-ʿizzi wa-l-diʿa¯ yati wa-l-h.ifz.i wa-l-h.im¯ayati wa-l-saʿdi wa-l-sal¯amati wa-l-nas.ri wa-l-kif¯ayati bi-mad¯ınati S.iqill¯ıyata sanata tham¯anin wa-ʿishr¯ına wa-khamsimiʾatin

(“This was made in the most royal, flourishing wardrobe, with good fortune, magnificence, splendour, perfection, might, superiority, generosity, prosperity, propitious fate, dignity, glory, beauty, attainment of desires and hopes, pleasure of days and nights without end or removal, with power, declaration of faith, vigilance, protection, good fortune, security, victory and capability, in the city of Sicily, in the year five-hundred and twenty-eight [1133–4 CE]”)44

Arabic Inscriptions in the Cappella Palatina

The supplications inscribed on the ceilings of the Cappella Palatina are not composed in sajʿ, and the individual texts are too short to be considered to be litanies, but the lexical repertoire is much the same as that used for the inscription on Roger’s mantle.45 Were all the adʿiya in the ceilings to be read together as a single continuous text, they would assume the cadence and shape of a litany. The case is not proven, but the evidence is sufficient to keep it open. No contemporary source helps us identify who might have recited such a litany but, if anyone, it would have been the royal courtiers, whether those standing bodily before the king or those imagined in the paintings of the ceiling. Such a litany might have been understood to have been spoken by the building itself, rather in the way in which the palace addressed the verses in the opus sectile inscriptions to an individual courtier. In this case, of course, the adʿiya are supplications addressed to God – prayers of praise that were intended to be overheard by the king and his courtiers.

AUDIENCE AND LEGIBILITY

So far, the legibility of the Arabic inscriptions in the Cappella Palatina has not been discussed, although the majority of visitors to the chapel would have been unable to read them as texts. This apparently straightforward claim, however, is complicated both by variety in the content, form and location of the inscriptions themselves and by the range of readers and viewers. A first distinction may be made between those able and unable to read Arabic. In the twelfth century, very few Latin Christians, even educated and highly trained officials of the royal court and administration, would have been able to read Arabic.46 The only significant exceptions – Arab and Berber Muslims who had converted or been converted to the Latin rite – seem to have been crypto-Muslims. A much larger number of Greek Christians native to Sicily and Calabria would have been able to read Arabic, and many of the leading officers of the court and administration were drawn from this group. A few Greek Christians from outside the island, such as King Roger’s chief minister, George of Antioch, one of the chief architects of the multicultural Sicilian monarchy, were also able to read Arabic. Many Jews would have been able to read Arabic, but I dare say that none were royal courtiers. But by far and away the largest group who would have been able to read the inscriptions in the Cappella Palatina were educated Muslims, amongst whom five groups may be distinguished. First were the scholars and men of letters, such as the geographer al-Idr¯ıs¯ı and the poets ʿAbd al-Rah.m¯an al-Buth¯ır¯ı and ʿAbd al-Rah.m¯an al-It.r¯abanish¯ı, who were attracted to the Norman court by the patronage of the king and of his leading officials. The author or authors of the verses in the opus sectile inscriptions may well have belonged to this group. Second were other elite visitors to the Norman court, including ambassadors from Fatimid Cairo and from other

135

136

Jeremy Johns

Islamic rulers. A letter borne by one such ambassador from the caliph al-H . a¯ fiz. li-d¯ın All¯ah (1131–49) to King Roger demonstrates how closely Cairo monitored the use of Arabic by the Norman court.47 Third were the leaders of the Muslim community of the island, such as the civil leader al-q¯aʾid Abu¯ ʿAbd All¯ah H ¯ himself a patron of men of letters, and the Banu¯ Raj¯aʾ who provided . ammud, the q¯ad.ı¯ of Palermo in three successive generations.48 Fourth were the ‘men of the pen’, the administrators, secretaries and scribes employed in the royal d¯ıw¯an, a mixed group that could include both leaders of the Muslim community and crypto-Muslim eunuch slaves who had been educated and trained within the royal court and administration.49 Finally, and perhaps most problematically, are those who made the inscriptions – metalworkers, painters and stonemasons. All of the inscriptions, except perhaps the pseudo-inscriptions, were designed by more or less highly skilled calligraphers who are perhaps unlikely to have been the craftsmen themselves, but rather either the distant artists of traditional ‘kufic’ designs preserved in the painters’ portfolios or the contemporary artists of new epigraphic designs in cursive script – there is reason to conclude that both groups were responsible for the epigraphic designs painted in the Cappella Palatina.50 Contemporary calligraphers may very well have come from the first group of visiting men of letters or from the fourth group of ‘men of the pen’ employed within the palace. Although members of all five of these groups would have been more or less able to read the Arabic texts inscribed in the Cappella Palatina, that does not mean that they all actually did so. For example, the opus sectile inscriptions in the Cappella Palatina (Fig. 40) and in the royal palace in Messina are written in large cursive script, with the dark green inlaid characters contrasting strongly with the white marble matrix. They seem to have been placed around the doorways giving entrance to the palace and were thus both highly visible and perfectly legible.51 Nonetheless, the manner in which they originally combined horizontal lines of text with vertical ones that might run from top to bottom or from bottom to top meant that even a man of Arabic letters would first have had to puzzle out their arrangement before he could have begun to read their content. Even when he had done so, the juxtaposition of verses in different metres created a series of intratextual obstacles that he would have had to overcome before he could have accessed their meaning. Today, this is difficult enough to do seated undisturbed at a desk, with the texts spread out in front of one. How much more difficult would it have been for an ambassador or a courtier passing nervously through a doorway into the presence of the king? In much the same way, the text of the long list of adʿiya embroidered in gold on the hem of King Roger’s mantle could only have been read when it was displayed fully open, as it is today in the Weltliche Schatzkammer in the Schweizerhof in Vienna.52 In the twelfth century, perhaps only its designer and the embroiderers who made the mantle, as well as the servants who cared for it in the royal wardrobe, would have had the opportunity to read its inscription – and, of course, only those of them who could read Arabic.

Arabic Inscriptions in the Cappella Palatina

One is drawn towards the conclusion that only the poet who composed the verses, and perhaps the officials who commissioned these inscriptions and oversaw their installation, would ever have actually read them on the mantle itself. We cannot even be sure that the embroiderers could read the words created by their own fingers. However, as we have already seen, the rhymed prose inscription on the border of Roger’s mantle seems to be composed for recitation, whereas the verses inscribed in opus sectile are likely to have been recited at least on their inauguration. Their content may thus have reached an audience far greater than those who actually read the inscribed texts. Let us turn now to consider the impact of the Arabic inscriptions in the Cappella Palatina upon those unable to read or understand Arabic. Although nearly all viewers would have recognised them as assertions of royal authority, their reactions to such assertions would have varied greatly according to their ethnic and linguistic community, their political allegiance and their education. William Malconvenant, the younger son of a Norman baron who was struggling to make a career in the trilingual administration and signed his name in a crude kufic script written from left to right, may have looked on his royal master’s Arabic inscriptions with approbation and wistful envy.53 However, his contemporary, Matthew Bonellus, another Norman baron from Sicily, was one of the leaders of a series of violent revolts against the multicultural ministers of the royal court, which sparked off pogroms against the Muslim citizens, and would have seen the display of Arabic as symptomatic of all that was rotten in the court of Sicily.54 Whereas the royal priest Grisandus usurped the royal monopoly to set up a trilingual memorial to his Norman father, and even a quadrilingual monument to his mother,55 other Latin churchmen, such as Stephen du Perche, archbishopelect of Palermo, disapproved so strongly of the Saracen ways tolerated in Sicily that he wrote to Pope Alexander III urging him to intervene.56 Not only would the Arabic inscriptions in the Cappella Palatina have been viewed in very different ways by such diverse individual viewers but those viewers are likely to have displayed their reactions in very different ways according to whether they were in the presence of the king and his ministers or alone in the company of their most trusted friends. Given that no contemporary source discusses the use of Arabic inscriptions by the Norman kings of Sicily – except for Al-Idr¯ıs¯ı and Ibn Jubayr in the brief passages cited later – there is perhaps little more to be said about audience and legibility, but the following brief discussion addresses the illegibility of most of the surviving Arabic inscriptions in the Cappella Palatina.

ILLEGIBILITY

The Greek and Latin texts in the mosaics of the Cappella Palatina are easily legible to the extent that they are written clearly and more or less correctly, and

137

138

Jeremy Johns

they can be read from the pavement of the chapel by those with good sight in the natural light of a sunny day. These texts, for the most part the names of saints and verses from scripture, enter into a dialogue with their images, which enhances the legibility of both. In the nave and the two aisles, where individual images and their texts are arranged sequentially to retell the familiar stories of Genesis and the lives of St Peter and St Paul, a familiar sequential narrative further supports legibility. Seventy-three panels in the painted ceilings now preserve Arabic inscriptions, not counting the four that are part of figural scenes and the many pseudoepigraphic designs, but none is legible in the manner of the Greek and Latin inscriptions. They are located in four areas of the ceilings: two in the borders of the coffers in the aisle ceilings; twenty-three (not counting one pseudoepigraphic design) in the intrados of the deep little niches in the centre of the small units of the muqarnas zone (Fig. 43); thirty in the semi-circular lunettes in the upper tier of the muqarnas zone; and eighteen (including two repainted with pseudo-Arabic in the eighteenth century) in the borders of the star-shaped coffers in the central zone of the nave ceiling (Fig. 41).57 Before electric light and optical lenses, at a time when approximately half of the population in middle age was afflicted by myopia and other impairments to sight, even the largest and most conspicuous inscriptions painted in the ceiling of the nave would have been extremely difficult if not impossible to read from the pavement below. Only at the end of the eighteenth century was it first recorded that the ceiling was decorated with Arabic inscriptions.58 When Michele Amari made the first reading and transcription of the texts of the central part of the ceiling of the nave, he did so by using optical lenses, photographs and mirrors to reflect sunlight into the dark recesses of the ceiling; even with such tools, he was unable to read the inscriptions in the niches of the muqarnas zone.59 One sunny day in February 2010, soon after the recent restoration had cleaned them and made them more easily visible than they had been at any time since they were first painted, I was able to read by natural light and with the naked eye the inscriptions bordering the star-shaped coffers in the central section of the ceiling of the nave (Fig. 41). But I was able to do so only because I am naturally long-sighted and because I already knew from long study of detailed photographs the precise content of each inscription. On the same occasion, I was unable to read the inscriptions in the south aisle even though they are much closer to the pavement than those of the nave; that is because the gold letters do not stand out clearly from the dark blue background and the foliate scrollwork. Nor was I able to read most of the inscriptions in the semicircular lunettes in the upper tier of the muqarnas zone. And I could read none of the inscriptions that decorate the intrados of the deep, little niches in the small units of the muqarnas (Fig. 43). But even though I could not read most of these inscriptions, I could see virtually all of them and recognise them as texts, just as I could see and mistake for texts many more pseudo-epigraphic designs composed of Arabic letterforms without

Arabic Inscriptions in the Cappella Palatina

43. Palermo, Palazzo Reale, Cappella Palatina. Details of the Arabic inscription hidden in the intrados of the deep little niche in the centre of the fourth small unit from the west end of the north side of the muqarnas zone: wa-l-yumn wa-l-kif¯aya wa-l-ʿizz wa-l- . . . (‘bliss and capability and power and . . . ’) (Khalili Research Centre Archive, Slides nos. ISL15108, 15110, 15111, C 2010 University of Edinburgh & The Barakat Trust) 15112. 

literary content. However, I was unable to appreciate the full variety of scripts and the elegance and ingenuity of the calligraphic compositions, in part because the inscriptions are too high to permit the close study of epigraphic and decorative detail, but also because the different forms are too widely dispersed throughout the ceiling to convey any immediate impression of their range and variety. From the ground, it is nearly impossible for one not already familiar with the inscriptions to distinguish the literary texts from the pseudo-inscriptions. The latter are designs composed of Arabic characters arranged in a pattern, which has no literal meaning and which often breaks the rules of Arabic orthography. The use of pseudo-inscriptions is common to Islamic art in general and owes nothing to the particular context of the Cappella Palatina. The manner in which here, as elsewhere in Islamic art, pseudo-inscriptions tend to be most commonly used to fill the borders of primary decorative elements, including true inscriptions, suggests that they offered an economical solution for the decoration of such marginal panels – allowing the artists to reproduce easily adaptable and infinitely repeatable patterns from their copybooks or memories, rather than more complicated and less adaptable calligraphic compositions with literary content. With this in mind, let us consider more closely the inscriptions and pseudoinscriptions of one of the eight-pointed stars that fill the octagonal coffers of

139

140

Jeremy Johns

the central zone of the ceiling of the nave (Fig. 41).60 At the centre of the coffer is a eight-lobed dome painted with an intricate, curvilinear strapwork design that radiates outwards from a central eight-pointed star, creating six concentric layers of interstitial spaces, each of a different shape – with the larger ones filled by palmettes, trefoil sprigs and other foliate designs. The central dome is enclosed within two borders that both sit within the cavity of the coffer. The inner border is a rounded moulding decorated as if with a winding rope of two contrasting designs. The outer border is painted with a kufic inscription, with white letters outlined in red, set on a bed of foliated scrollwork on a dark blue ground. The inscription is interrupted at the eight corners by large palmette buds, which effectively address the problem of how the inscription can be made to turn the acute corners. (A transliteration and translation of the text of the kufic inscription have already been given.) The central dome and its two borders together form a composite unit that fits over the stellate cavity at the centre of the octagonal coffer. The transformation of the eight-sided coffer into the eight-pointed cavity is effected by the six bases of stalactite pendants and by the two capitals of the adjoining unit of the muqarnas zone of transition, which form eight obtuseangled projections pointing inwards towards the centre. Four of these projections are painted with pseudo-inscriptions, and four with geometric designs, forming in effect a border along the edge of the stellate cavity.61 In all twenty of the octagonal coffers of the central zone of the ceiling, the equivalent obtuse-angled projections are decorated with an apparently random mix of pseudo-inscriptions and geometric or foliate designs, reinforcing the conclusion that pseudoinscriptions were used in these marginal panels for essentially practical reasons of economy. If we focus more closely upon one interstice between two points of the star (Fig. 44), we can see how an Arabic inscription in which the textual meaning is perfectly legible – al-saʿd wa-l-kam¯al wa-l-i . . . (here seen upside down at the top of the figure) – is juxtaposed with a pseudo-inscription composed of the letter s¯ın attached to an intertwined l¯am-alif (or alif-l¯am), which is in turn attached to a letter s¯ın reversed as in a mirror. Not only is the resulting pattern wholly without textual content but also it defies the rules of Arabic orthography. However, neither the legible Arabic text nor the unreadable pseudo-inscription could have been read from the pavement of the chapel. In effect, without reopening the question of the ability of the audience to read Arabic, both inscription and pseudo-inscription would have been illegible to the twelfth-century viewer. It is tempting to argue that the primary function of the Arabic inscriptions in the borders of the stellate coffers was not to convey textual meaning, but rather to pick out in white the shape of the eight-pointed stars in the central section, thereby making them particularly conspicuous within the multifaceted and heavily decorated ceiling. That these stellate coffers were recognised as stars in the twelfth century is confirmed by Philagathos Kerameos who remarks that the ceiling ‘imitates the heavens when, through the clear air, the host of stars shines

Arabic Inscriptions in the Cappella Palatina

44. Palermo, Palazzo Reale, Cappella Palatina. Detail of a stellate coffer (south side, third from west) from the central zone of the painted wooden ceiling of the nave, showing (below) a pseudo-inscription composed of the letter s¯ın attached to an intertwined l¯am-alif (or alif-l¯am), C 2010 followed by a reversed letter s¯ın. (Khalili Research Centre Archive, Slide no. ISL15784.  University of Edinburgh & The Barakat Trust)

everywhere’.62 But the difficulty with this line of argument is that whereas the same visual effect could have been obtained by using a pseudo-inscription or, for that matter, a geometric or foliate pattern, or any other design, so long as it were painted in white or in any pale contrasting colour, the artists went to great pains to outline the stars with a perfectly legible text carefully composed in elegant calligraphy. Moreover, similar Arabic texts decorate the recesses of the muqarnas zone that are almost invisible from the ground, so that these perfectly legible inscriptions cannot be read today even with a powerful torch and binoculars. What is more, the most elaborate and inventive calligraphic compositions in the ceiling are reserved for these invisible recesses, where no one but the artists themselves would have noticed had they used pseudo-inscriptions or even non-epigraphic designs. The ceiling of the nave of the Cappella Palatina was almost certainly built and painted by a team of itinerant craftsmen according to a model that they had tried and tested many times elsewhere on Islamic palaces and pavilions. It may be that such structures, none of which survive, were significantly lower in height than the nave, so that by transplanting such a ceiling from the hall of an Islamic palace to the nave of the Christian chapel most of its decoration – including its inscriptions – were thereby made illegible. Be that as it may, perfectly legible Arabic inscriptions are frequently located so as to be illegible or invisible even

141

142

Jeremy Johns

in Islamic buildings, so that it would be rash to attribute the illegibility of the Arabic inscriptions in the Cappella Palatina solely to that act of transplantation. It may be instructive to approach the matter of illegibility from another perspective. We have already seen that the Arabic inscription embroidered in gold on the hem of King Roger’s mantle combines the date and place of production with supplicatory formulae similar and in some cases identical to those painted on the ceiling of the Cappella Palatina. Displayed as it is today in the Vienna Schatzkammer, the Arabic text can be read by anyone familiar with the script and the slightly arcane vocabulary, and accustomed to reading sajʿ63 – skills, incidentally, that only a well-educated man of Arabic letters would have possessed in twelfth-century Palermo. There is no reason, however, to assume that the mantle was ever displayed in this manner at the Norman court, and we know from the coronation ordines of the Norman kings and from other sources that the royal vestments were kept in the wardrobe, under the charge of a special chamberlain.64 The mantle and the other Norman royal vestments later became part of the ceremonial regalia of the Holy Roman Empire, and contemporary illustrations survive of rulers actually wearing the Sicilian robes. However, Jean Adam Delsenbach’s imaginative engravings of various emperors wearing the full regalia give an altogether misleading impression of the legibility of the inscriptions on the regalia and, indeed, of the portability of the vestments.65 The artist’s primary concern was to document the regalia themselves, not to give a realistic account of how they looked when worn. In fact, the text on Roger’s mantle could never have been read, as it were, in situ. Delsenbach pretends that the mantle fell straight from shoulder to floor without a single fold, exhibiting the text as if it were inscribed on a rigid canopy, when of course it would have been hidden within folds that would have moved continuously with the wearer. What is more, the mantle weighs a knee-trembling fifty kilograms, and few men could have stood still long enough for would-be readers on their hands and knees to crawl about trying to decipher the inscription. Albrecht Durer’s far more impression¨ istic reconstruction of Charlemagne wearing the Sicilian vestments may be closer to reality; it may even be significant that Durer seems not to have recognised the ¨ inscription as meaningful text.66 If the relatively clear and large inscription on the mantle could not have been read while it was being worn by the king, still less could the much smaller and poorly executed inscriptions in Latin and cursive Arabic script on his alba have been read in situ, hidden as they were not just by the folds of the garment but also by the mantle and the stole.67 And, of course, no observer could have read the inscription on the leggings while they were being worn, for not only does the text begin on one leg and continue on the other but also the leggings would have been completely hidden by the successive layers of the mantle, alba and tunic.68 Thus the regalia of the Norman kings bore perfectly legible Arabic inscriptions that, like those in their palace chapel, would ordinarily have been

Arabic Inscriptions in the Cappella Palatina

completely illegible even to those few of their courtiers who possessed the linguistic skills that would have enabled them to read them. These inscriptions principally operated not as texts but as signifiers that conveyed not through content but by form a complicated nexus of messages about the nature of the Norman monarchy to a diverse audience. The use of Arabic in these royal media proclaimed first and foremost that the king was master of the Arabic language and, by extension, of all other aspects of Arab and Islamic culture. Although Arabic inscriptions on royal coins, documents and inscriptions successfully created the impression that the kings themselves had mastered the language, there is little doubt that they were in fact unable to read and write Arabic.69 Contemporaries nonetheless regarded the kings’ mastery of Arabic as a marvellous and potentially threatening demonstration of their authority and power. Ibn Jubayr, a Spanish Muslim visitor to the court of William II, wrote, ‘May God protect the Muslims from [the king’s] enmity and ability (ʿa¯ diyata-hu wa-bast.ata-hu); amongst the marvellous things told about him is that he reads and writes in Arabic’.70 Similarly, the variety of Arabic scripts employed by King Roger on a variety of media had been noted by al-Idr¯ıs¯ı, the author of the geographical work sometime referred to as the Kit¯ab Ruj¯ar (‘Book of Roger’). Of Palermo cathedral, he remarks that ‘it is difficult to comprehend the marvellous workmanship, the artful choice and inventive wonders of the range of images, and the varieties of ornaments and inscriptions’. And, although he does not mention the Cappella Palatina itself, he describes how the royal palace was ‘constructed and adorned with most wonderful curiosities (bi-aʿjabi l-mughtarib¯at) and filled with things of extraordinary character (bad¯aʾiʿ al-s.if¯at)’, phrases that, as Michele Amari assumed, may refer respectively to calligraphy and representational images.71

TOWARDS SOME CONCLUSIONS

The variety of different forms of the Arabic script used in the Cappella Palatina indicated the king’s mastery over all things Arab and Islamic, and the juxtaposition of Arabic with Greek and Latin scripts reminded the viewer that it was not just Arabic that the king claimed to have mastered. In the royal chapel, as in bilingual coins from the royal mint, bilingual documents from the royal chancery, and multilingual inscriptions on royal buildings and erected by royal servants, the Norman king proclaimed not only his mastery of the three individual cultures of the island but also that Arab, Greek and Latin had been merged under his beneficent rule into a single populus trilinguis. But other forces also governed the use of Arabic inscriptions in the Cappella Palatina. The strongest were the artistic traditions respected by the painters of the ceilings. The inscriptions and pseudoinscriptions belonged to their stock repertoire. The content, location and form of the inscriptions were largely determined by what the painters had done time

143

144

Jeremy Johns

and time before, presumably for Muslim patrons, and had brought with them to Norman Palermo. Such traditions incorporated tried and tested solutions to practical artistic challenges: visual – how to emphasise the stellate outline of the coffers of the central zone; economic – how to save time and skilled labour in the decoration of marginal zones; and physical – how to decorate the intricately multifaceted surfaces of the ceiling while precariously balanced on a rickety scaffolding more than ten metres above the ground. In contrast, the polychrome inlaid stone inscriptions from the west end of the royal hall, supplemented by the similar fragments from the royal palace in Messina, illustrate the most extreme departure from tradition by King Roger’s epigraphers. Both the technique of opus sectile and the outlandish cursive ductus of these inscriptions are unique to Norman Sicily; the former borrows from South Italian ecclesiastical decoration, the latter perhaps mirrors the administrative script used in the royal d¯ıw¯an. Visuality again predominates: here, close to the royal throne, a conspicuous display of porphyry and serpentine is strongly contrasted against a white marble background. The verses themselves, except for the inclusion of the barbaric name of Ruj¯ar, demonstrate an unimaginative, if wholly competent, adherence to the traditions of Arabic panegyric. But, by addressing visitors to the palace in the imperative voice, they seek to prescribe their approach to the palace, as if it were the Meccan sanctuary or a paradise on earth. Even the repetitive lists of adʿiya painted in the ceilings may evoke litany-like acclamations performed as part of the public presentation of the king. All of these motive forces – the proclamation of royal policy, practical artistic traditions, and what may be regarded as some sort of performative utterance – simultaneously drove the Arabic inscriptional programme in the Cappella Palatina. NOTES

1. Antonio Braca, Il Duomo di Salerno: Architettura e culture artistiche del Medioevo e dell’Et`a Moderna (Salerno, 2003), 21– 5: . . . ROBBERTUS DUX R[omani] IMP[erii] MAXIM[us] TRIUMPHATOR DE ERARIO PECULIARI. 2. Andr´e Guillou, Recueil des inscriptions grecques ´ m´edi´evales d’Italie, Collection de l’Ecole franc¸aise de Rome 22 (Rome, 1996), 210– 11, no. 195: . . . ἐv ἡμέραις τοῦ λαμπροτάτου δουκὸς ʿΡουμβέρτου . . . 3. Lucia Travaini, La monetazione nell’Italia normanna, Istituto Storico Italiano per il Medio Evo Nuovi Studi Storici 28 (Rome, 1995), 109–10, 398–9, no. 61. 4. Travaini, Monetazione, 109–110, 398–399, no. ¯ all¯ah 61: l¯a illah ill¯a all¯ah muh.ammad rasul ¯ all¯ah arsalahu bi-l-hud¯a / muh.ammad rasul wa-d¯ın l-h.aqq li-yuz.hirahu ʿal¯a l-d¯ın kullihi (‘There is no god except God. Muh.ammad

is the messenger of God. / He sent him with guidance and the religion of truth to cause it to prevail over all religions’). 5. For this assertion, see the following, all with extensive further bibliography: Jeremy Johns, Arabic Administration in Norman Sicily: The Royal D¯ıw¯an, (Cambridge, 2002), especially pp. 284–300; Alex Metcalfe, Muslims and Christians in Norman Sicily. Arabic speakers and the end of Islam (London, 2003), especially 99– 113; J. Johns, “L’iscrizioni arabe dei re normanni di Sicilia: una rilettura,” in Nobiles officinae: perle, filigrane e trame di seta dal Palazzo Reale di Palermo, 2 vols., ed. Maria Andaloro (Catania, 2006), 2.47–67; Eng. trans., 2.324–37. See also Karla Malette, “Translating Sicily,” Medieval Encounters 9 (2003), 140– 63. 6. Peter of Eboli, Liber ad honorem Augusti sive ¨ de rebus Siculis, f.97v., ed. Theo Kolzer and

Arabic Inscriptions in the Cappella Palatina

7.

8.

9.

10.

11.

12.

13.

14.

Marlis St¨ahli (Sigmaringen, 1994), 44–5: urbs felix, populo dotata trilingui; illustration of the trilingual chancery, 59 (f. 101v.). For a general discussion of the Arabic inscriptions of the Norman rulers, with up to date bibliography, see Johns, ‘L’iscrizioni arabe dei re normanni,’ in Nobiles officinae, ed. Andaloro, 2.47–67; Eng. trans., 2.324–337. For the quadrilingual inscription in which Arabic, Greek, Latin and Hebrew scripts are used as symbols for the four religious communities of Sicily – Muslim, Greek Orthodox, Latin Catholic and Jewish, see J. Johns, ‘Lapidi sepolcrali in memoria di Anna e Drogo, genitori di Grisanto, chierico del re Ruggero,’ in Nobiles officinae, ed. Andaloro, 1.519–23; Eng. trans., 1.775–8. Eugenius of Palermo, Versus iambici XXIV. 65–9, ed. Marcello Gigante, Istituto siciliano di studi bizantini e neoellenici, Testi e Monumenti, Testi 10 (Palermo, 1964), 130; It. trans. 163–4. For the most up-to-date and comprehensive accounts of the Cappella Palatina, see the superb photographs accompanied by descriptive schede and analytical essays in La Cappella Palatina a Palermo, Mirabilia Italiae 15, 4 vols., ed. Beat Brenk (Modena, 2010). See also the studies collected in Die Cappella Palatina in Palermo – Geschichte, Kunst, Funktionen. Forschungsergebnisse der Restauri¨ erung Hg. im Auftrag der Stiftung Wurth, ed. Thomas Dittelbach (Kunzelsau, 2011). ¨ For a comprehensive catalogue and discusssion, see J. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ in Cappella Palatina, ed. Brenk, 1.353–86. J. Johns, ‘Lastra con iscrizione trilingue di ¯ il Gaito, eunuco alla corte di Pietro (Barrun) Ruggero II,’ in Nobiles officinae, ed. Andaloro, 1.510–11; Eng. trans., 1.771–2. See also the objections of A. Nef, Conqu´erir et gouverner la Sicile islamique aux XIe et XIIe si`ecles, Bib´ lioth`eque des Ecoles franc¸aises d’Ath´enes et de Rome, fasc. 346, Rome, 2011, pp. 335–36, that are refuted in V. von Falkenhausen, N. Jamil and J. Johns, ‘The twelfth-century doc` uments of St. George’s of Troccoli (Sicily)’, forthcoming, note 82. J. Johns, ‘Lastra con iscrizione trilingue dalla clessidra di re Ruggero,’ in Nobiles officinae, ed. Andaloro, 1.512–13; Eng. trans. 1.772– 773. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.382–3. J. Johns, ‘Tre lastre frammentarie con iscrizioni arabe in lode di Ruggero II dal palazzo di Palermo,’ in Nobiles officinae, ed. Andaloro,

145

15.

16.

17. 18. 19.

20. 21. 22.

23.

24. 25.

26.

1.499–501; Eng. trans. 1.765–6. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.379–82. Annliese Nef, ‘Venti blocchi frammentari con iscrizioni arabe in lode di Ruggero II dal palazzo di Messina,’ in Nobiles officinae, ed. Andaloro, 1.502–9; Eng. trans., 1.766–70. For the mantle, see: William Tronzo, ‘The Mantle of Roger II of Sicily,’ in Stewart Gordon, Robes and Honor. The medieval world of investiture (New York, 2001), 241–253; Oleg Grabar, ‘The experience of Islamic art,’ in Irene A. Bierman, The Experience of Islamic Art on the Margins of Islam (Los Angeles, 2005), 11–59; Rotraud Bauer, ‘Manto di Ruggero II,’ in Nobiles officinae, ed. Andaloro, 1.45–9; Eng. trans. 1.577–9. See also notes 65–8. Johns, ‘L’iscrizioni arabe dei re normanni,’ 47–8; Eng. trans. 324. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.355–79. See the collections of recent studies (and the secondary works listed therein) cited in note 10. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.383–6. See note 14. For this relative date, see J. Johns, ‘The Date of the Ceiling of the Cappella Palatina in Palermo,’ in Ernst J. Grube and J. Johns, The Painted Ceilings of the Cappella Palatina, Islamic Art Supplement I (Genova and New York, 2005), 6–7. See also the work cited in note 23. J. Johns and Stefano Riccioni, Schede nos. 161, 371 and 475, in Cappella Palatina, ed. Brenk, 2.440–1, 490–1, 551–5. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.355–79. Maria Pia Demma, ‘Cofano con figure intarsiate e iscrizione nasḫ,’ in Nobiles officinae, ed. Andaloro, 1.136–9. For a less than wholly satisfactory reading and Italian translation of the inscriptions: Michele Amari, Le epigrafi arabiche di Sicilia, trascritte, tradotte e illustrate, 2nd ed., rev. Francesco Gabrieli (Palermo, 1971), 290–309. ‘A great many curtains are hung, the fabric of which is threads of silk, woven with gold and various dyes, that the Phoenicians have embroidered with wonderful skill and elaborate artistry’: from the ekphrasis of the Cappella Palatina by Philagathos Kerameos, trans. Johns, ‘The Date of the Ceiling,’ 13– 14; see also Johns’ discussion of this passage, 4 and 6. Hangings are shown in the

146

27.

28.

29. 30.

31.

32.

33. 34.

35.

36.

37. 38.

Jeremy Johns

illustration of the Cappella Palatina by Peter ¨ of Eboli, Liber ad honorem Augusti, eds. Kolzer and St¨ahli, 43 (f. 97r.). Compare the illustrations of the Cappella Palatina by Peter of Eboli, Liber ad honorem ¨ Augusti, eds. Kolzer and St¨ahli, 43 (f. 97r.) and 47 (f. 98r.), and the painting of the Cappella Palatina in the north muqarnas zone of the ceiling of the nave discussed by Johns, Fig. 591 and Scheda no. 591, in Cappella Palatina, ed. Brenk, 2.581–2 and 4.459, fig. 591. Andaloro, ed., Nobiles officinae, 1.44–9, 55– 9, 60–1. Tarif Al Samman, ‘Arabische Insch¨ riften auf den Kronungsgew¨ andern des Hei¨ ligen Romischen Reiches,’ Jahrbuch der kunsthistorischen Sammlungen in Wien 78 (1982), 7–34, with discussion in Johns, ‘Iscrizioni arabe dei re normanni,’ 2.47–67; Eng. trans., 2.324–36. See also the works cited in note 16, and notes 65–8. See note 13. J. Johns, ‘The Greek Church and the Conversion of Muslims in Norman Sicily?,’ in Stefanos Efthymiadis, Claudia Rapp and Dimitris Tsougarakis, eds., Bosphorus. Essays in Honour of Cyril Mango, Byzantinische Forschungen 21 (Amsterdam, 1995), 141–2. Augusta Acconcia Longo, ‘Considerazioni sulla chiesa di S. Maria dell’Ammiraglio e sulla Cappella Palatina di Palermo’, in ᾿Αμπελοκήπιον. Studi di amici e colleghi in onore di Vera von Falkenhausen IV, Nea Rhome 4 (Rome, 2007), 267–94. Ernst Kitzinger, The Mosaics of St. Mary’s of the Admiral in Palermo (Washington, D.C., 1990), 38–9, 131. Johns, ‘Tre lastre’. Ibn Jubayr al-Kin¯an¯ı, Abu¯ l-H . usayn Muh.ammad ibn Ah.mad, Rih.lat al-Kin¯an¯ı, ed. Michael Jan de Goeje and William Wright (Oxford, 1907), 325: kaf¯a ll¯ahu l-muslim¯ına ʿa¯ diyata-hu wa-bast.ata-hu wa-min ʿaj¯ıbi sha ʾni-hi l-mutah.addathu bi-hi anna-hu yaqraʾu wa-yaktubu bi-l-ʿarab¯ıya. See also p. 26 and note 70. For the most recent discussion of these, with earlier bibliography, see the works cited in note 14. Nef, ‘Venti blocchi,’ 505, no. 3. The verb ¯ is the masculine plural of the imper(udkhulu) ative mood. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.377, and 4.730, fig. 1045. Johns, ‘Iscrizioni arabe dei re normanni,’ 2.56, and 2.67, tav. III; Eng. trans. 2.328, and 2.337, Table III. See also Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.359–62.

39. Hjalmar Torp, ‘Normannerkongen Roger II av Sicilia i lys av bysantinsk og senantikk herskerideologi,’ Kongelige Norske Videnskabers Selskabs Forhandlinger (1991), 101–120. In a similar vein, see also Staale Sinding-Larsen, ‘Plura ordinantur ad unum. Some Perspectives regarding the “Arab-Islamic” ceiling of the Cappella Palatina at Palermo (1132– 1143),’ Institutum Romanum Norvegiae, Acta ad archaeologiam et artium historiam pertinentia 7 (1989) 55–96. 40. Johns, Schede nos. 505–507 (muqarnas unit), 511 (male half-figure with two crosses) and 514–15 (seated rulers), in Cappella Palatina, ed. Brenk, 2.562, 564– 65, 565. 41. Three images of drummers have a single Arabic word inscribed on the heads of their drums: see Johns, Scheda no. 381, 2.508, and figs. 655, 704 and 745, in Cappella Palatina, ed. Brenk, 4.497, 527 and 553. 42. Vassa Larin, ‘The Dikerion and Trikerion of the Byzantine Pontifical Rite: Origins and Significance,’ Orientalia Christiana Periodica 74 (2008), 417–30. 43. J. Johns, ‘Le pitture del soffitto della Cappella Palatina,’ in Cappella Palatina, ed. Brenk, 1.400–1. 44. Samman, ‘Arabische Inschriften,’ 31–4. Johns, ‘Iscrizioni arabe dei re normanni’ 2.53–5; Eng. trans. 2.327. 45. Johns, ‘Iscrizioni arabe dei re normanni’ 2.67, tav. III; Eng. trans. 2.337, tav. III (the attentive reader will note that I have since revised the English translation of some of the Arabic adʿiya: see Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.361–2). 46. Metcalfe, Muslims and Christians, 99–113. Nadia Jamil and J. Johns, ‘Signs of the Times: Arabic Signatures as a Measure of Acculturation in Norman Sicily,’ Muqarnas 21 (2004), 181–92. 47. Johns, Arabic Administration, 264. 48. Ibid., 88, 89. 49. Ibid., 212–56. Metcalfe, Muslims and Christians, 99–113. 50. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.356–8. 51. Johns, ‘Tre lastre’. Nef, ‘Venti blocchi’. 52. Samman, ‘Arabische Inschriften,’ 31–4 and figs. 15–19. 53. Jamil and Johns, ‘Signs of the Times,’ 184–6. 54. For Bonellus, see Graham A. Loud, ed. and trans., The History of the Tyrants of Sicily by ‘Hugo Falcandus’. Manchester Medieval Sources (Manchester, 1998), 21, 86, 92–104, 107–8, 110–12, 115–20, 122–4, 197, 229–32. 55. Johns, ‘Lapidi sepolcrali’.

Arabic Inscriptions in the Cappella Palatina

56. Johns, Arabic Administration, 229 and note 68. 57. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.355–79. 58. Johns, ‘Le pitture del soffitto della Cappella Palatina,’ 1.389. 59. Ibid., 1.389–90. 60. Johns, ‘Iscrizioni arabe nella Cappella Palatina,’ 1.377, and 4.730, fig. 1045. 61. Note, however, that the hastae of the kufic inscription radiate outwards, obeying the norm for the arrangement of inscriptions that circumscribe a central space, whereas those of the pseudo-inscriptions point inwards towards the centre of the star. This may indicate that the artists or designers conceived of the pseudo-inscriptions as belonging to the decorative scheme of the stalactite pendants rather than of the stellate domes. 62. Johns, ‘The Date of the Ceiling,’ 13. 63. See the works cited in note 16. 64. R. Elze, ‘Tre ordines per l’incoronazione di un re e di una regina del regno normanno di Sicilia,’ in Atti del congresso internazionale di studi sulla Sicilia normanna (Palermo 4–8 dicembre 1972), Istituto di Storia Medievale, Universit`a di Palermo, 1973, B, sections 4–5, 21. 65. Beatrix Kriller, ‘Delineation exacte des Ornem´ens Imp´eriaux du Sainte Empire Romain et Allemand. Gard´es dans la ville libre et imp´eriale de Nuremberg. Dessin´es et grav´es aux d´epens de Feu Monsieur Le S´enateur Jerˆome Guillaume Ebner d’Eschenbach, par Jean Adam Delsenbach. Avec les Saintes Reliques grav´ees d’apr`es les dessins de Fr´ederic Iuvenell,

147

66.

67.

68.

69.

70. 71.

Nuremberg, Gottlieb Schneider 1790,’ in Nobiles officinae, ed. Andaloro, 1.84–87; Eng. trans., 1.592–3. Mathias F. Muller, ‘Carlo Magno con le ¨ vesti dell’incoronazione del Sacro Romano Impero,’ in Nobiles officinae, ed. Andaloro, 1.80–1; Eng. trans., 1.590–1. Rotraud Bauer, ‘Alba di Guglielmo II,’ in Nobiles officinae, ed. Andaloro, 1.55–9; Eng. trans., 1.580–2. Samman, ‘Arabische Inschriften,’ 19–24. Rotraud Bauer, ‘Calze di Guglielmo II,’ in Nobiles officinae, ed. Andaloro, 1.61; Eng. trans., 1.582–3. Samman, ‘Arabische Inschriften,’ 25–31. Jamil and Johns, ‘Signs of the Times,’ 182– 4. Vera von Falkenhausen, ‘Una Babele di lingue: a chi l’ultima parola? Plurilinguismo sacro e profano nel regno normanno-svevo,’ Archivio storico per la Calabria e la Lucania 76 (2010), 23. For a rather different reading of the same evidence, see Metcalfe, Muslims and Christians, 102–4. Ibn Jubayr, Rih.la, 325. See also note 34. Abu¯ ʿAbd All¯ah Muh.ammad ibn ʿAbd All¯ah ibn Idr¯ıs al-H ¯ ı al-H . ammud¯ . asan¯ı Al-Idr¯ıs¯ı, Opus Geographicum sive ‘Liber ad eorum delectationem qui terras peragrare studeant’ (Kitab nuzhat al-mushtaq), ed. Antonio Bombaci et al. (Naples, 1970–6), p. 591; Michele Amari, Biblioteca arabo-sicula, ossia raccolta di testi Arabici che toccano la geografia, la storia, la biografia e la bibliografia della Sicilia. Raccolti e tradotti in Italiano, 2nd ed., rev. Umberto Rizzitano et al., 3 vols. (Palermo, 1997–8), 1.59.

鵻 CHAPTER SEVEN

INTERCESSION AND SUCCESSION, ENLIGHTENMENT AND REFLECTION: THE INSCRIPTIONAL AND DECORATIVE PROGRAMME OF THE QARATAY MADRASA, KONYA 鵼 Scott Redford

This chapter addresses a traditional art-historical subject: a patron and a building. The patron in question is Jalal al-Din Qaratay, court emir and regent of the Seljuk sultanate between the years ca. 1249 and 1254, a period when the Seljuk state was weak and the three sons for whom he served as regent were children. For all intents and purposes, during this time, Qaratay ruled the Seljuk sultanate. The building he commissioned during the period of his regency served both as a school for religious and legal sciences, or madrasa, and as his mausoleum. He built it in the Seljuk capital of Konya around the year 1251. Known as the Qaratay Madrasa, this building is located at the base of the citadel hill in the centre of the city. Throughout the building, inscriptions, presented in a variety of forms, play a significant role.1 Contemporaneous historical and religious sources report that Qaratay was a very pious Muslim, with his piety taking a mystic and ascetic bent. They record that he was well known to another Jalal al-Din living in Konya at the time: his younger contemporary Jalal al-Din Rumi, Sufi founder of the Mawlawiyya/Mevlevi dervish order, popularly called the Whirling Dervishes. As an index of his asceticism, two contemporaneous historical sources note that Qaratay forswore eating meat, drinking wine and engaging in sexual intercourse. One source also notes that literate and illiterate, Muslim and non-Muslim, the orphaned and the widowed, but especially Sufis and all manner of religious men from far and wide, sought him out and benefitted from his generosity.2 In the foundation document of his madrasa, as well as that of the large caravanserai that he endowed in the previous decade, Qaratay prescribes that in place of oil lamps, which were the normal means of lighting in these (and other

148

Intercession and Succession, Enlightenment and Reflection

contemporaneous) buildings, candles be lit in certain places and at certain times. According to the instructions in these documents, candles were to burn in the mosque of the caravanserai all year long – before, during and after prayer times at dusk and dawn and in the madrasa on three holy nights: those celebrating the revelation of the Qur’an, the conception of the Prophet and the descent of the Qur’an. The foundation document of his caravanserai also prescribes that all comers, Muslim or non-Muslim, free or slave, man or woman, be treated equally.3 The Seljuk historian Ibn Bibi states that Qaratay was himself of slave origin and from Anatolia; likely he was a Christian captured and converted to Islam at an early age. We know that Qaratay had two brothers, both of whom also had assumed Turco-Islamic names and served as high-ranking officials of the Seljuk sultanate. One of these, Kamal al-Din Rumtash, built his own madrasa, today in ruins, directly across the street from that of his brother.4 In this chapter, I use inscriptional and other data from the Qaratay Madrasa to argue for the creation of spaces, constructed architecturally and socially, that recognised a continuum of religious practice between Christianity and Islam as valid. The placement of writing within these spaces also served to emphasise continuity.5 In addition to a continuum in practice, in Islam there is a natural sense of succession or completion between Islam and the religions of other peoples of the book, one that is explored later in this chapter. Whether or not it related to the veneration of saints in Christianity, medieval Islamic mysticism had a particular interest in holy men, their sanctity and their relation to other divine powers. The rise in importance of the veneration of the Prophet Muhammad as al-ins¯an al-k¯amil, the perfect man, a model for Muslims, coincided with the spread of the practice of night-time celebrations of such signal events as the beginning of the revelation of the Qur’an, the conception of Muhammad, the beginning of the dictation of the Qur’an to him and his night voyage into the heavens. The celebration of these sacred nights, a development of medieval Islamic practice relating in part to the self-identification of Sufi holy men with Muhammad, gained in importance at this time. Ibn Bibi reports that Qaratay’s piety caused him to be granted the privilege of writing the title Vali All¯ah f¯ı’l-ard. next to his signature on official Seljuk state documents. This term can be translated as ‘God’s friend on earth’, but the term Vali also means holy man or saint in Sufism, as can be seen by its use in the foundation document of the madrasa. The prescription of celebration of these holy nights in the madrasa can be tied not only to Sunni Muslim practice at this time but also to a possible self-identification of the ascetic/mystic emir with the Prophet Muhammad himself. In this respect it is interesting to note that the name of the artist who copied and illustrated the eleventh-century Persian language romance of Varqeh and Gulshah at this time can be found among ¨ the list of the witnesses to the foundation document of the Qaratay Madrasa. This romance invokes the Prophet Muhammad on many occasions and ends with the revivification of the lovers by God as the result of prayers made by the Prophet Muhammad himself and taken to the heavens by the Angel Gabriel.6

149

150

Scott Redford

THE DECORATIVE AND EPIGRAPHIC PROGRAMME OF THE QARATAY MADRASA I: THE PORTAL

This much biographical information is apposite to a chapter that argues for an intimate relationship between this patron and the artisans decorating the portal and the interior of the madrasa. Michael Meinecke has mustered stylistic evidence for the presence of a workshop of tile mosaic artisans working in Konya from the early 1230s to the 1250s, linking the only signed work, the Sırc¸alı Madrasa (1242–3), with the mosaic tilework of the dome and mih.r¯ab of the Alaeddin Mosque (which he dated to the early 1230s); this signed work was also linked to later buildings such as the madrasa of Qaratay’s brother Kamal al-Din Rumtash (1248–9) and the Qaratay Madrasa itself (1251–3), all built by members of the Seljuk elite in the capital city of the sultanate. A decades-long period of residency and acquaintanceship between artists and members of the Seljuk ruling class as the result of high-level patronage accords not only with the increasing technical sophistication of the tilework in these buildings but also the integration of tile technique with the decorative and epigraphic programme exhibited in the Qaratay Madrasa.7 The masons who worked on the marble inlay and inscriptions of the fac¸ade of the building (Fig. 45) are unknown, and there are no good candidates like that of Muhammad al-Tusi, who was proposed by Meinecke as the leader of the tile workshop. The majority of the scholarship dealing with the fac¸ade of the Qaratay Madrasa has been concerned principally with its relationship to the nearby Alaeddin Mosque, whose main entrance was furnished with a similar doorway and interlace marble decoration by an architect from Damascus in the early 1220s; researchers have also examined problems with the foundation inscription of the building, which is located on the fac¸ade.8 Let us look instead at the religious inscriptions of the facade. An unusual Qur’anic quotation runs on either side of the entrance. It is executed in an elegant cursive script that does not aim for the monumentality of the foundation inscription, with its drawn-out hastae: rather, it is written in a style more reminiscent of a Qur’anic chapter heading. The quotation is from the Sura of the Ant, verse 19, which recounts the deeds of Solomon, a prophet in Islam. Solomon also held a special place in mystical Islam because of his esoteric knowledge, in this chapter the knowledge of the language of birds and insects. This verse introduces us to the concept of God granting prophets esoteric knowledge and, without identifying him by name, introduces Solomon (and indirectly, his father David, who is named in the previous verse). The door frame of the Qaratay Madrasa copies the format of the door frame of the nearby central mosque of the Seljuk dynasty, the Alaeddin Mosque, but with modifications. Here, instead of the opening sura of the Qur’an, which is around the doorway of the Alaeddin Mosque, twenty-two short, simple h.ad¯ıth (sayings of the Prophet Muhammad), written in easily legible cursive script, are carved into

Intercession and Succession, Enlightenment and Reflection

45. The portal of the Qaratay Madrasa at the beginning of the twentieth century. (After J. H. ¨ Loytved, Konia. Inschriften der seldschukischen Bauten (Berlin: J. Springer, 1907), pl. 47)

the marble door frame, the sole entrance and exit of the building. These h.ad¯ıth form the first part of the epigraphic programme of the madrasa that calls to mind the educational function of the building. The h.ad¯ıth chosen for inclusion here are all structured in the same very short, simple declarative sentences, appropriate

151

152

Scott Redford

for the level of understanding of students studying Islamic law and theology in Arabic, a language that was not likely to have been native to any of them. They are written in the individual cells, or leaves, of the vegetal design carved in the marble door surrounds. These h.ad¯ıth, constituting pithy homilies with pedagogic intent, can be read separately or together as a text, with connections between individual h.ad¯ıth and, especially toward the end of the series, as a commentary connected with the founder. In this way, they, like other elements of the decorative programme of the madrasa, are capable of being understood separately, but also together, with connections and levels of meaning apprehended only after the kind of repeated exposure that goes with daily ingress and egress. The twenty-two h.ad¯ıth are written within the thirty-seven cells of the design; several then span two cells, often eliminating the wa, or ‘and’ that links the two parts of the h.ad¯ıth as written in had¯ıth compendia, or other words. Following the direction of writing of the Arabic script, the h.ad¯ıth start at the lower right-hand corner and end at the lower left-hand corner. The h.ad¯ıth on the right-hand side of the doorway largely concern moral precepts, with a certain emphasis on finance: they warn against indebtedness and emphasise group values. The h.ad¯ıth above the doorway continue in this vein, but also address matters of praxis – again aspects important to students. For instance, a sequence of three h.ad¯ıth in the middle of this part of the frame can be thought of as having being grouped together for a purpose. The first of the three states that one should offer greetings before beginning conversation, al-sal¯am qabl (al-)kal¯am. This is followed by a h.ad¯ıth proclaiming that questioning is half of knowledge, al-su’¯al nis.f al-‘ilm, and one emphasising the importance of prayer for worship (inna) al-du’¯a huwa al-‘ib¯ada. In this way the initiation of speech, the content of (educational) speech and the importance of speech (as part of prayer) for faith are all addressed, with a progression from the initiation of speech through its use to attain different levels of knowledge. On the left-hand side, moral precepts dominate, with a return to the ‘x is half of y’ formulation quoted earlier. The longest h.ad¯ıth found here introduces what I consider to be a biographical element relating to the founder, an element that continues until the end. This h.ad¯ıth states that precaution is half of life, love is half of intelligence, and worry is half of old age: al-tadb¯ır nis.f al-‘aysh (wa) al-tawaddud nis.f al-‘aql (wa) al-hamm nis.f al-haram. After this come h.ad¯ıth that describe good and bad actions in opposition. The first of these uses the word khayr, which means good, but also denotes good work(s): good is customary (and) evil is troublesome, al-khayr ‘¯ada (wa) al-sharr laj¯aja. Coming after a h.ad¯ıth mentioning old age, this h.ad¯ıth can be seen as containing a personal reference to Qaratay and the good work of his madrasa. The thirty-second and thirty-third cells contain another h.ad¯ıth with a subject that, in the context of the other h.ad¯ıth, is so unusual as to demand a biographical explanation. This h.ad¯ıth states that a

Intercession and Succession, Enlightenment and Reflection

child is a cause of greed and cowardice, (inna) al-walad mabkhala majbana. Coming as it does after h.ad¯ıth mentioning old age and pious works, this h.ad¯ıth may be connected to the political context of Qaratay’s regency: the factions surrounding each of the three young brothers of the ruling dynasty in general and the debauched behaviour of the teenaged sultan ‘Izz al-Din in particular. The string of h.ad¯ıth ends with two that rhyme. The first of these once again contains ominous portents: foul language comes from harshness, al-badh¯a’ min al-jaf¯a’, whereas the final h.ad¯ıth fittingly proclaims the superiority of the Muslim holy book: the Qur’an is the remedy, al-Qur’¯an huwa al-daw¯a’.9 To summarize, the portal of the Qaratay Madrasa is a complex composition that reflects the close partnership of patron and craftsmen. A foundation inscription, something usually found over the door, lies near the top of the entire portal. Below the inscriptional band is a seemingly abstract composition of black-andwhite marble, but one whose shapes recall the angular writing conventionally called kufic.10 A shallow muqarnas vault rises above the doorway. At the level of the base of this vault, the muqarnas is flanked by an unusual Qur’anic quotation, which itself runs across the fac¸ade just above the level of the top of the had¯ıth surrounding the entranceway. The primary function of the had¯ıth is pedagogical, but they also introduce the concept of an inscriptional text that can be read and understood separately, as well as a whole, and whose meaning reveals itself over time; they add a biographical element to religious inscriptions. The position of the had¯ıth around a doorway also introduces the idea of passage, something also found in the interior of the building.

THE DECORATIVE AND EPIGRAPHIC PROGRAMME OF THE QARATAY MADRASA II: THE DOMED COURTYARD

Once through the portal doorway, a visitor would have immediately entered a domed, tiled vestibule. Unfortunately, the dome that once rose atop this space, one that connects the portal to the domed courtyard of the madrasa, has long since collapsed. At one time this dome was decorated with tile mosaic, several pieces of which are still in situ above the interior of the doorway and around one side of the entrance into the main domed space of the interior. From here, a doorway leads unobtrusively into a corner of the main space of the building, a courtyard topped by a large hemispheric dome.11 This dome once terminated in an oculus, now covered, that provided the principal source of light to the interior. There is a square pool in the middle of the courtyard, directly underneath the oculus. This central covered courtyard is surrounded by subsidiary spaces. To three sides were rooms that have collapsed over the centuries and been rebuilt. Judging from the number of doorways on these three sides, they must have led to the rooms that housed at least some of the

153

154

Scott Redford

students studying here. To the far side, opposite the entrance to the courtyard, is a raised vaulted hall, or ı¯w¯an (Fig. 46). It is flanked by two doorways, both leading into domed rooms. This centralized building, with a dome hovering over the main space, is given direction by the ı¯w¯an at the back opposite the entrance, the main teaching space of the madrasa. At the back of this ı¯w¯an is a window, the second major source of lighting for the main space of the building. The domed tomb chamber of Qaratay himself is entered through a doorway to the left of this ı¯w¯an. The raised ı¯w¯an opposite the entrance gives the building two foci, two directional axes. The eye is led to these two axes – the horizontal one leading to from east to the ı¯w¯an to the west, and the vertical one leading to the dome – both because of the concentration and variety of the tile mosaic patterns and inscriptions and because both axes culminate in the two main light sources for the building: the window at the back of the ı¯w¯an and the oculus at the top of the dome. The inscriptional and decorative programmes of the Qaratay Madrasa combine the pedagogical function of the space with the spiritual inclinations of the patron. This madrasa is justly famous for its tilework, which, in its opulence, abundance and variety, recalls a palace rather than a school. The contrasting modesty of the decoration of the tomb chamber, with its untiled dome, can be seen as an indirect acknowledgement of the funerary function of this space. To the abundance of tile and variety of tile technique can be added skill: the surviving mosaic tilework displays, despite occasional lapses, an advanced capability in the planning and disposition of complex patterns and long inscriptions in and around complex spaces. And then there are the patterns and inscriptions, which themselves are not unique to this space. It is in integration of architectural form and space and tile decoration, rather than in novelty of motif and quotation, that they proclaim their mastery.12 In what follows, I use images and description to portray the way in which writing is employed not only to draw attention to the dome and the ı¯w¯an but also, through its interplay with the decoration and the architectural space, to advance different kinds and levels of understanding. As I proposed for the had¯ıth around the doorway of the portal, I also argue that the writing’s intended meaning unfolds not only in relationship to architectural space but also over the time spent by a frequenter of this building, another pedagogical and spiritual concern. The focus is on the decoration of the walls and dome of the central courtyard; however, the major ‘eye-catching’ device of the ı¯w¯an – the inscription around its entrance arch and the accompanying raised band of reticulated ornament – also mediates the transition between the ı¯w¯an and the central space of the domed courtyard. Over the centuries, the lower walls of this building have lost most of their tile decoration. However, from surviving examples, we can surmise that they were covered in their entirety with hexagonal tiles with monochrome turquoise glaze.

Intercession and Succession, Enlightenment and Reflection

46. Qaratay Madrasa ı¯w¯an, general view. (Photo: O. Eravs¸ar)

The glazed surface of all surviving tiles is covered with stencilled gilded geometric and floral decoration, often featuring the six-pointed star known in Islamic cultures as the seal of Solomon. The only writing consists of the repetition of two phrases: al-‘Izz lill¯ah, Glory to God, and al-Mulk lill¯ah, Sovereignty (alone) belongs to God. Both are commonly found on Seljuk and other medieval Islamic coins and inscriptions. The second phrase is used so often in Seljuk Anatolia that it is not outlandish to propose it as the religious motto, or one of the mottos or devices, used by the Seljuk sultans as a signature. The luxury of using gold-stencilled designs was confined to this lower level. All other tile decoration employs the tile mosaic technique. Two inscriptions run above the level of this gilt-turquoise tile dado. Both are Qur’anic, and both were ‘written’ and designed in order to be read, with only slightly monumentalized Seljuk cursive writing in black standing out, plain and legible against a background of turquoise ornament scrolling against a white ground (Fig. 47). The inscription framing the entrance to the ı¯w¯an from the courtyard is perhaps the most frequently found Qur’anic inscription in Islamic architecture, the Throne Verse (Sura of the Cow, verse 255), a central statement of Islamic faith. This inscription frames the ı¯w¯an, at its top reaching almost to the base of the dome (Fig. 46). The introduction of the basmala at the beginning of the inscription allows the phrase about interceding with God (the verb is shaf ’a) to hang at the very top of the arch. The placement of this phrase at the top of the arch constitutes one piece of evidence for intercession as one of the major themes of the epigraphic programme. Typical to the epigraphic programme of this building, the

155

156

Scott Redford

message is imported not by the choice of unusual and more pertinent Qur’anic passages, but by the visually, spatially and stylistically unusual and innovative ways in which common passages are displayed, as this example demonstrates. Equally legible is another Qur’anic passage that tops the arches above the entrances to the rooms flanking the three sides of the courtyard. It reproduces other parts of the Sura of the Cow, verses 284–6 (Fig. 47). In this architectural context, these verses may be interpreted in two ways. The first interpretation is a pedagogic one. Like the Throne Verse around the ı¯w¯an, they contain succinct and powerful statements about God, His omnipotence and His relationship with believers. On the walls of the Qaratay Madrasa courtyard, they are written in a clear, legible cursive hand, appropriate for apprehension by the students, and the writing stands out by virtue of its contrasting colour. Ornament scrolls as if behind the writing, but does not combine with it. Continuing as it does from door to door, this inscription visually unites the courtyard. A second possible interpretation emphasizes verse 285, ties this inscription to those in the pendentives, and relates the mortal realm to that of the divine through God’s messengers, prophets. This verse specifically mentions the validity of all prophets. Between the Qur’anic quotation and the pendentives runs a band of decoration, one that at first looks like the style of writing – interlaced kufic – encountered in the dome (Fig. 47). In this band, the interlace is prominent, but below it lie only three letter shapes that repeat themselves over and over. With the exception of the ı¯w¯an, this pseudo-epigraphic band runs all the way around the courtyard. The transition from the easily read cursive script of the gilded tiles and the Qur’anic verses to the knotted, angular kufic-style decoration is striking. Even though this band does not contain writing, its decoration (1) looks like writing, (2) is easily distinguishable as pseudo-writing and (3) marks a transition from cursive script to the use of kufic in all inscriptions above it. Looking up, the next tile patterns encountered are massive triangular bands spreading between the four corners of the courtyard and the dome; one is pictured here (Fig. 48). This is a kind of fan pendentive known in art and architectural historical literature as a ‘Turkish triangle’ because of its use in Seljuk buildings in this era. Each of these triangular areas is divided into five thin triangular wedges separated one from the next by double bands of vegetal decoration. Like the interlaced pseudo-kufic band below it, the field of each triangle also contains angular patterns that resemble writing. On closer inspection, it is possible to puzzle out the repetition of single names within each triangular segment. With one exception, a single name occurs in each of the twenty segments (five to each triangle, four triangles, one at each corner of the courtyard) (Fig. 49). Just as the Throne Verse running around the arch of the ı¯w¯an contributes to both the horizontal and vertical axes of movement of the building, so too these triangles contribute both a vertical and horizontal movement. The vertical axis is much more evident: starting from a point touching the top of the

Intercession and Succession, Enlightenment and Reflection

47. Qaratay Madrasa, general view of courtyard wall. (Photo: O. Eravs¸ar)

pseudo-epigraphic frieze, these triangles fan out to the base of the dome to such an extent that they barely leave any stretch of wall uncovered there. As such, they contribute visually to the dominance of the dome, lifting the eye upwards toward it. It is in their content, and not in their form, that these triangular fans impart a horizontal movement to the building. This axis of movement also relates to the temporal issue raised earlier, because it presents prophetic history, quite naturally in such a setting, from an Islamic point of view. This point of view is expressed within each single triangle, as well as in the overall scheme governing all four of the triangles. As mentioned earlier, each of the triangles of the four fan pendentives has five segments. The name of the Prophet Muhammad is found in the central segment of each triangle. The four segments flanking this central segment seem to have been conceived temporally, with earlier prophets on the outermost pair, and later prophets, or Abu Bakr, the first of the Rightly Guided Caliphs (the four that succeeded Muhammad after his death) filling the inner pair. As the diagram reproduced in Fig. 49 shows, on the southeast triangle (that above the entrance), Muhammad’s name is flanked by those of Jesus and Abu Bakr, whereas on the outermost segments, Moses is to one side and, likely due to error, Jesus and Abu Bakr, exceptionally, share the same segment. The northeastern triangle is clearer: the name of Muhammad is flanked by those of Jesus and Abu Bakr, with the names David and Moses on the other segments to either side of them. As we have seen, the temporal succession (or completion) of Abrahamic prophets by the Prophet Muhammad then continues with caliphs. On the two triangles

157

158

Scott Redford

48. Qaratay Madrasa, southwestern ‘Turkish triangle’. (Photo: O. Eravs¸ar)

to the east, or back of the courtyard, only the name of the first of the Rightly Guided Caliphs, Abu Bakr, is present. On the western triangles, he is ‘succeeded’ by the other three Rightly Guided Caliphs: ‘Umar, ‘Ali and ‘Uthman. In both the north-west and south-west triangles, the name of Muhammad is flanked by

Intercession and Succession, Enlightenment and Reflection

49. Qaratay Madrasa, diagram showing the arrangement of the prophetic and caliphal names in the segments of the Turkish triangles. (After Michael Meinecke, Fayencedekorationen seldschukischer Sakralbauten in Kleinasien [Tubingen: Ernst Wasmuth, 1976], Volume 2, Illustration ¨ 43.) [1 = David; 2 = Jesus; 3 = Muhammad; 4 = Abu¯ Bakr; 5 = Moses; 6 = ‘Umar; 7 = ‘Uthm¯an; 8 = ‘Al¯ı; 9 = Muhammad]

segments bearing the names of ‘Ali and ‘Uthman. However, in these two triangles, the succession seems to be intended to be read both centrally and sequentially from right to left, with the second Caliph ‘Umar followed by the ‘Uthman, and then, after Muhammad in the central segment, ‘Ali. This leaves one final segment with no more Rightly Guided Caliphs, so it is filled in once again with the name of Muhammad. Earlier I referred to the association of the thirteenth-century Konya-based copyist/illustrator of the eleventh-century Persian romance of Varqeh and Golshah with the social milieu of Jalal al-Din Qaratay, as testified by his serving as a witness to the foundation document of his madrasa. Given the prominence and placement of the name of the Prophet Muhammad in this building, it is interesting to note the prominent intercessory role played by Muhammad in the poem. Moreover, although they are not mentioned in the text, in this romance the four men who were to become the Rightly Guided Caliphs are illustrated seated together with Muhammad himself. Priscilla Soucek has noted that this is the first surviving instance of figural representation of the Prophet Muhammad together with these of his companions, a subject that she notes finds increased favour later in the thirteenth century and into the fourteenth century in Islamic manuscript painting. The particular miniature in the manuscript of Varqeh and Golsh¯ah that represents the Prophet together with these four of his companions introduces the king of Syria into the scene, necessitating the removal of all four of

159

160

Scott Redford

these figures to one side of the centrally located figure of Muhammad. However, later representations portray, as is represented onomastically in the fan pendentives of the Qaratay Madrasa, Muhammad in the middle, flanked by ‘Ali, ‘Uthman, Abu Bakr and ‘Umar.13 The temporal succession of prophets and caliphs has its own internal order within each triangle, even though that order was not strictly followed. What is clear, however, from the absence of the names of David, Moses and Jesus from the two triangles flanking the ı¯w¯an to the west of the building is that the succession of prophetic and caliphal names proposes a horizontal movement parallel to the movement of the visitor/student/teacher to the school, from entrance to ı¯w¯an. To the temporal element of prophetic and caliph succession across the space of the courtyard can be compared the time it would take a student living and studying in the building to puzzle out the intricate, interlocking kufic names of each segment, put them together in individual sequences within each triangle and arrive at the overall idea underlying this major element of the decorative programme. The world of scholarship continues to use the key first published by Mehmet ¨ Onder in 1962 (Fig. 49). What if a student studying here had no such crib? The decorative programme is meant to be revealed through mental application over time, as the eye strays from the ı¯w¯an and the mind strays from the sound of the lesson passing from the ı¯w¯an and reverberating through the domed space of the courtyard. The tile mosaic decoration of the interior of the hemispherical dome of the Qaratay Madrasa is one of the best-known works of Anatolian Seljuk art. The reticulated patterns found in subsidiary parts of the building, on the fac¸ade, and flanking doorways and inscriptions resolve here into a web more complex than the hexagon-centred patterns below, spinning a series of webs that contain four rows of sixteen centriform, segmented stellate forms (Fig. 50). These forms, which emerge from and disappear into the edges of the design field, become smaller and get closer to one another as they approach the oculus at the top of the dome. Bordering this extraordinary field, which has been likened to everything from a field of daisies to whirling dervishes, are two more panels of writing to its top and bottom. Like the writing below, they both play on the border between pattern and writing. The reward behind the dare implicit in the complexity of the knotted patterning and distorted letterforms of the knotted interlace kufic of the lower inscription lies in the recognition of a Qur’anic passage reproduced below in a much more legible form: the Throne Verse (Fig. 50). Once again, a temporal element is built into the decorative programme – perhaps the understanding gained by extended exposure to this inscription would give more profound insight to the student than the verity that may have accompanied the easier reading of the same famous Qur’anic passage given closer to the level of human use of the space below in legible cursive script.

Intercession and Succession, Enlightenment and Reflection

50. Qaratay Madrasa detail of the inscription at the base of the dome. (Photo: S. Redford)

161

162

Scott Redford

Just as the stellate forms in the tile mosaic decoration of the dome grow smaller as they approach the oculus, so, too, does the writing on the upper band, immediately below the oculus, diminish in size, thereby increasing the difficulty of decipherment. However, the shapes of the letters are less distorted, and there is less knotting and interlace, leading to a quicker understanding that this band, too, reproduces the Throne Verse. There is one addition to both of these bands that is not found in its cursive iteration below: the subsequent verse, no. 256: There is no compulsion in religion. The right direction is henceforth distinct from error. And he who rejecteth false deities and believeth in God hath grasped a firm handhold which will never break.14

I leave it to the reader to decide whether this addition can be related to the unusual inclusion of a Qur’anic inscription giving voice to Solomon on the portal and of the names of David, Moses and Jesus on the interior, or whether it simply fills the added space available in the dome.

DISCUSSION AND CONCLUSION

The attentive reader may have surmised a connection between the tilework, the spiritual predilections of the patron who paid lavishly for it and the general outlines of the structure of the attainment of spiritual knowledge. The thesis proposed here is based on a correspondence between mystic ideas of divinity, knowledge and truth; both literal and figural enlightenment and reflection; and levels of knowledge. The meaning that the tile programme aimed to impart was based both on the content of the inscriptions (especially the Qur’anic inscriptions, which are the most common) and on their placement and style. To summarize: at the bottom, at eye level, we begin with gilded inscriptions – themselves declaring the power of God, but recognisable as mottoes used in imperial/sultanic contexts. They are mixed, however, with decoration that implies some sort of esoteric knowledge, a clue, if you will. The next level contains inscriptions (both around the arch of the ı¯w¯an and above the entrances to the rooms around the other three sides of the courtyard) that are at the same time forceful, clear and well-known Qur’anic statements of the power of God. Immediately above this level, in the courtyard, is a layer of obfuscation: writing that is not writing, but is easy to recognise as such. It introduces the angular form of writing known as kufic, which was not usually used at the time; however, when used, it was often associated with esoteric knowledge. Therefore it could also be thought of as demarcating a boundary, in the way pseudo-kufic decoration would have been used on a contemporaneous carpet. Above this layer is the opposite: kufic writing that looks like decoration, but is not. This writing consists of names of prophets and caliphs. Ascending, as they are, they can be viewed as being in an

Intercession and Succession, Enlightenment and Reflection

intercessory position between the human and the divine, but the arrangement of names also argues for a temporal progression from Abrahamic prophets through Muhammad to the Rightly Guided Caliphs. These names literally fan out to the base of the dome, on which the Throne Verse is written twice, at its base and its top, at the opening of the oculus. In both instances it is written in kufic, first in an almost impossibly beautiful and difficult way and then in a more comprehensible one. Above this is the open sky, main source of light for the building. Passing through the oculus, light circles through the courtyard as the day progresses, literally enlightening parts of the decoration, the writing and all that is in between (Fig. 50). At the same time, the light reflects in the pool directly below the oculus, diffusing in a more indirect way and subtly distorted by the gradual flow of water through it. Enlightenment and reflection – two processes that take time – are presented together to those studying in the ı¯w¯an of the madrasa as daily facts, but the increasing complexity of the decoration/writing as the tile programme ascends invites, even demands, the passage of time (and the lifting of one’s head from books, or one’s gaze from the teacher). The passage of time implicit in the movement from the doorway into the covered courtyard to the main place of instruction, the ı¯w¯an, is paralleled by the movement of progression of prophetic time, from the other Abrahamic religions to Islam (even as Muhammad, the prophet who ushered in the return to the original faith of Abraham, is omnipresent). The levels of inscriptional and pseudo-inscriptional tilework that culminate in the oculus of the dome are structured in a way analogous to the levels (manzil, pl. man¯azil) of knowledge of God attainable through study. Muhammad had a prominent role in Islam, mystical or not, as an intercessor, so his prominence in the inscriptional programme of the fan pendentives comes as no surprise. Muhammed can also be related to the idea of enlightenment through ¯ althe widespread belief in his luminous quality, the light of Muhammad, nur Muhammad; this concept was found in the writings of mystics present in Anatolia – Rumi, Ibn ‘Arabi and Yunus Emre – at about this time. Thus, the inscriptional programme invites us to think of him as both intercessor for Muslims and as the possessor of a divine light that can lead the Muslim student toward enlightenment. Both aspects of Muhammad also link him to the celebration here of events connected with his life at night, which are illuminated with candles that are prescribed in the building’s foundation document. Even though the placement of Muhammad’s name encourages such an interpretation, it must be noted that the formulators of the epigraphic programme did not use h.ad¯ıth or different Qur’anic passages associating Muhammad with both light and with intercession. Yet this absence of such writings is congruent with the general avoidance of explicit statements here and the use of placement, programme and epigraphic style as clues to meaning. According to Islamic belief, Muhammad intercedes with God on behalf of believers on Judgement Day. In this sense, the prominence of intercession as a theme of the inscriptional programme of the Qaratay Madrasa can

163

164

Scott Redford

be thought of as chosen for a structure that houses the tomb of its patron, who was advanced in age when it was being built, and died soon after its completion.15 Earlier in the chapter I used the term ‘axis’ in a standard art-historical way to describe the organization of space, movement, decoration and light in the interior of the Qaratay Madrasa. This word can be used in another way, as in the term axis mundi, which is one way to translate the word qut.b. This word is a common one in medieval Sufi writings used to describe a saint, a holy man, teacher or one with special spiritual insight. This word, which can also be a title, is not used anywhere in the inscriptional programme of the Qaratay Madrasa. Rather, it is used twice in the foundation document of the building. On the fac¸ade of the Qaratay Madrasa, the foundation inscription gives the reigning sultan only two titles, both standard issue: ‘the Greatest Sultan’ and ‘Shadow of God in the World’. The same inscription mentions Jalal al-Din Qaratay without giving him a title, although his high status is implied by the absence of any version of the phrase, ‘the humble servant in need of the mercy of God’, which usually precedes the names of emirs in Seljuk inscriptions. In a manner similar to Seljuk foundation inscriptions, the foundation document of the Qaratay Madrasa is organised in descending order. It begins with praise for God; followed by praise for the caliph in Baghdad; then praise, names and titles for the Seljuk Sultan (‘Izz al-Din); and finally for Qaratay himself. In the way that Sufism at the time used temporal terminology and titles for religious people, in the last section of this document, Jalal al-Din Qaratay refers to himself, among other ‘titles’, as ‘Lord of the Ascetics, Axis (qut.b) of the Believers, Master of the Emirs and the Saints, Helper of the People, Source of Blessings, Cream of the Times, Exemplar of the Saints (awliy¯a – plural of wal¯ı ) Crown of the Sufis’.16 In the foundation document of his caravanserai, Qaratay refers to himself as, among other things, ‘King of the Sufis (abd¯al)’ and ‘Axis (qut.b) of the Ascetics’.17 These and (many) other titles and names given in the foundation documents are congruent with the depiction of Jalal al-Din Qaratay as a pious Muslim with mystical inclinations. The title qut.b is of particular interest: Franklin Lewis informs us that, in the writings of Jalal al-Din Rumi, the prophets, saints and caliphs who were qut.b (pl. aqt.a¯ b) included David, Solomon, Moses, Jesus and Muhammad, followed by the Rightly Guided Caliphs: in short all of those named or referred to in the inscriptional programme of the madrasa itself. Ibn Bibi writes that Jalal al-Din Qaratay was possessed of the spirit of S.uhayb (a companion of the Prophet known for his abstemiousness) and with knowledge of the esoteric (asr¯ar-e ghayb). Still, despite shared nomenclature, a spiritual gap exists between saints and prophets.18 In a manner similar to the Sunni orthodoxy of the inscriptional programme, the foundation document does not mention mystical instruction, but rather prescribes that Sunni orthodox Islamic theology and law of the Hanafi school were to be taught here. The main teacher (mudarris) was to be from this legal school,

Intercession and Succession, Enlightenment and Reflection

although imams from the three other Sunni schools of law were also permitted. ¯ and furu’ ¯ as found in madrasas across the Islamic world at Instruction was in us.ul this time. This prescription is one reason that I have emphasized the patron and not the curriculum of instruction as a driving force behind the decorative and epigraphic programme.19 A cottage industry of writers and publishers has regularly imparted religious, especially mystical, meaning to works of Islamic art and architecture. The subject of mystical interpretation of a work of Islamic architecture, therefore, is a sensitive one, not only because it is hard to prove but also because maintaining that artistic traditions exist mainly to embody eternal religious verities opens one up to charges of cultural essentialism.20 In finding analogies between the neoplatonic structures of Sufi thought, with their levels of meaning, and the levels of ornament and writing in the domed courtyard of the Qaratay madrasa, I hope to have grounded my argument securely enough in the cultural circumstances of mid-thirteenth century Konya and the person of Jalal al-Din Qaratay to avoid these charges. In closing, I return to an issue raised at the beginning of this chapter, and that is the relationship of Christianity to Islam. This is an important issue, personally for a convert such as Jalal al-Din Qaratay and generally in medieval Anatolia, a land that had been predominantly Christian for centuries and one that still, at this time, must have had a population that was majority Christian. The 1202 foundation document of the Altun Aba Madrasa in Konya prescribes the use of ¨ Medrese candles on the three holy nights mentioned earlier. The 1280 Sivas Gok foundation document stipulates that candles be lighted during prayers as well as mealtimes on these nights. Together with the foundation documents of Qaratay, these documents cover almost the entire thirteenth century and therefore can be seen as representative of Islamic religious practice in Anatolia then: not a cryptoChristian practice prescribed by one man, but an association of candlelight with faith common with Christian practice in this place and time.21 The prescription of the use of candles in two buildings built and endowed by Jalal al-Din Qaratay participates in this trend; wittingly or unwittingly there is a continuity of praxis between these two religions. Is there anything else here that bridges these religious and cultural traditions? The domed madrasa is a feature of medieval Anatolian Islamic architecture and as such is not found elsewhere in the Islamic world. The number of domed madrasas in Anatolia dating to the twelfth and thirteenth centuries is small in comparison to those with open courtyards. In the context of traditions of religious architecture, it is, on the surface, easy to think about the domed madrasa as a form influenced by Byzantine ecclesiastical architecture until one sees its single hemispheric dome, so different in shape and size from Byzantine, Armenian or Georgian domed churches of the time.22 The Qaratay Madrasa is the first of a series of domed madrasas with such domes and without columnar support, a series that begins in the 1250s and continues

165

166

Scott Redford

for more than thirty years. Therefore, it can be considered an important index not only of a regional tradition but also of its revival and modification, and the question of intent lies behind the choice of this building type. The placement of the names of prophets and caliphs in the pendentives and the sense of the progression between them, although not paralleling Byzantine or other Christian practice, seem to be beholden to it in some measure. This is also the case with the twin foci on the ı¯w¯an and the dome and with the hierarchical presentation of sacred space. However, rather than an instance of cultural borrowing, perhaps one can see this practice as another instance of the creation of a distinctly Islamic space that contains Christianity within it, in the way that Islam acknowledges Jesus as a prophet.23 This sense of progression from Christianity to Islam can also be seen in the tomb chamber of Qaratay within his madrasa. Its dome is devoid of decoration, in stark contrast to the dome of the courtyard. Almost nothing remains of its inscriptional programme except part of the same tile dado with a gilded inscription and part of a Qur’anic inscription in the same legible cursive hand, part of a sura (The Troops, verse 9) found on the wall. This sura is concerned with death and the afterlife, and therefore appropriate to the funereal nature of this space. The verse that originally was reproduced in full on the walls of the tomb chamber, however, is concerned not only with an interest in the afterlife but also with the difference in knowing (the truth) between (Muslim) believers and unbelievers, a condition that also bears on Qaratay’s own passage from Christianity to Islam. Remarkably, the door frame of the entrance into the tomb chamber from the courtyard is made up of architectural sculpture that has been recognisably reused from a Byzantine church, the only piece of spolia obviously employed in this building. This could be seen as another, different means of referencing Christianity as prefatory to the truths of Islam: to enter the tomb chamber, one has to pass through a doorway recalling a Christian church. In a similar fashion, the Qur’anic inscription on the portal anticipates the progression of prophetic and historical time elaborated on the Turkish triangles of the interior. The Qaratay Madrasa appears to be the first surviving Islamic building whose epigraphic programme places the names of important caliphs and prophets in high places, especially around the base of the dome – a practice that became widespread in religious architecture in later centuries. However, there are other buildings in Anatolia, both earlier and later than this building, that also have inscriptions placing the name of Muhammad in high places. These inscriptions are more prosaic (and more legible, too!). The earliest surviving example seems to be the placement of the name of Muhammad in star form at the top of a dome of the Ulu Cami at Malatya, which has been dated to 1224. The names of Allah and Muhammad are placed at the base of the minarets of the C ¸ ifte Minareli Medrese in Erzurum, undated but from about this time period, and in the Sahib Ata complex in Konya, dated 1258. The Tahir and Zuhre mosque in Konya, ¨

Intercession and Succession, Enlightenment and Reflection

167

also undated by inscription, but belonging to the late thirteenth century, bears, on the crown of the dome, tile mosaic decoration, which is also in star form, with the names of Muhammad and the Rightly Guided Caliphs. These examples illustrate not only similar concerns at the time but also a less developed decorative programme.24 The Qaratay Madrasa uses changes in level and in script style and difficulty to invoke the secrecy and mystification of divine knowledge. Doing so did exclude from understanding both the ‘literate and illiterate’ who benefitted from Jalal alDin Qaratay’s generosity, but not the teachers and students who lived and worked in the Qaratay Madrasa and had its secrets revealed to them gradually, through literal and metaphorical reflection and illumination. Jalal al-Din Qaratay can be seen as playing a dual intercessory role: as a patron in building the building, but also as a saintly man, as God’s friend on earth revealing through his actions knowledge of God to others, just as did the prophets and caliphs in the past. In turn, he relied on the Prophet Muhammad to intercede with God on his behalf on Judgement Day. NOTES

1. The fourth chapter of Sara Nur Yıldız, Mongol Rule in Seljuk Anatolia: the Politics of Conquest and History Writing 1243–1282 (Brill: Leiden, in press) gives a lucid account of the politics of the period of regency. 2. Bar Hebraeus, E. A. Wallis Budge trans., The Chronography of Gregory Abu’l Faraj (Piscataway, NJ, 2003 (reprint of 1932 Oxford University Press edition)), 1: 413 ‘And there was in Kanya (sic) a certain noble, an old slave of Sultan Ala ad-Din, whose name was Jalal alDin Karatai (sic), and he was an ascetic who abstained from the eating of flesh, and from the drinking of wine, and from women, and he was a good and merciful man.’ See also ¨ Ibn Bibi, El-Evamiru’l-‘Ala’iyye fi’l-Umuri’l‘Ala’iyye (Ankara, 1956), 593–4, and Franklin Lewis, Rumi Past and Present, East and West (Oxford, 2000), 278–9. 3. See Osman Turan, ‘Celaleddin Karatay, Vakıfları ve Vakfiyeleri,’ Belleten 47 (1948), 132 for the relevant passage concerning the madrasa; 95–6 for that in the waqfiyya (endowment/foundation document) of the caravanserai. The same prescription is given in the waqfiyya of his brother’s madrasa across the street and in the later (1280) waqfiyya ¨ Medrese, built in Sivas in 1271 of the Gok by Sahib Ata, an emir of nonslave origin; see Sadi Bayram and Ahmet Hamdi Karabacak, ‘Sahib Ata Fahru’d-Din Ali’nin Konya, ¨ ˙ ¨ Imaret ve Sivas Gokmedrese Vakfiyeleri,’

Vakıflar Dergisi 13 (1981), 56. In all of these documents, the sacred nights are referred to only by their dates. ¨ 4. Ibn Bibi, El-Evamiru’l-‘Ala’iyye, 593; Turan, ‘Celaleddin Karatay,’ 47–9 on the broth˙ ers mentioned in the waqfiyyas; Ibrahim Hakkı Konyalı, Konya Tarihi (Konya, 1964), 875 for the theory that he was from Konya, and Michael Meinecke, Fayencedekorationen seldschukischer Sakralbauten in Kleinasien (Tubingen, 1976), Vol. 2, 278–83, for the ¨ remains of the building, which interestingly was not domed. 5. See, for instance, Robert Jordan (trans.), ‘Pantokrator. Typikon of Emperor John II Komnenos for the Monastery of Christ Pantokrator in Constantinople’ in eds. John Thomas and Angela Constantinides Hero, Byzantine Monastic Foundation Documents (Washington, D.C., 2000), 740–3 for the burning of oil lamps in the church, including all night, although it is specified that candles burn in the most important parts of the church, in front of important icons and during important feast days. ¨ 6. See Ibn Bibi, El-Evamiru’l-‘Ala’iyye, 569; also Marion Holmes Katz, The Birth of the Prophet Muh.ammad. Devotional Piety in Sunni Islam (London: Routledge, 2007), 50–62 for the emergence of a genre of texts relating to the Prophet’s birth in the twelfth century in Iran and Iraq and the emergence of Sunni

168

Scott Redford

celebrations of it, and 147–68 for notions of homogeneous and nonhomogeneous time in Islam in relation to the merits of certain days such as the birth of the Prophet. See Assadullah Souren Melikian-Chirvani, ‘Le roman de Varqe et Golsah: Essai sur les rapports de l’esth´etique litt´eraire et de l’esth´etique plastique dans l’Iran pr´e-mongol, suivi de la traduction du po`eme,’ Arts asiatiques 22 (1970), which gives a French translation of the text of Ayyuq¯ ¯ ı’s Varqeh o Golsh¯ah, with the climactic episode of prayer and resurrection found in verses 2168–2224. On page 13, Melikian-Chirvani concludes, following Ahmed Ates¸, ‘Un vieux po`eme romanesque persan: r´ecit de Warqah et Gulshah,’ Ars Orientalis 4 (1961), 146–7, that the composition of this work likely dates to the reign of Ghaznavid Sultan Mahmud ¯ (d. 1030), based on the language of the text and the mention of titles congruent with those of this sul¨ tan. M. Kemal Ozergin, ‘Selc¸uklu Sanatc¸ısı Nakkas¸ Abdul-Mu’min el-Hoyi Hakkında,’ ¨ Belleten 34 (1970), 227, made the connection between the artist who signed this manuscript and the name on the Qaratay endowment document. 7. Meinecke, Fayencedekorationen, 2: 298 and ˘ MehMeinecke, ‘Tuslu Mimar Osman Oglu ˘ Mehmed ve Konya’da 13.unc med Oglu ¨ u¨ ¨ Etnografya ¨ Yuzyılda bir C ¸ ini Atolyesi,’ Turk ¨ Dergisi 11 (1968), 81–93. 8. The inscription dates to the period of Qaratay’s regency for three brothers. Although quite naturally the inscription should have mentioned ‘Izz al-Din, the only one of the brothers then holding the title of sultan, it does not. Rather, it bears the name of his brother ‘Ala’ al-Din. The block of stone that should mention this brother is the only one of the ten blocks of marble that form this inscription not to have a double, but a single line of writing. Doubts about its originality are not limited to its format: the next panel refers to the latter part of ‘Ala’ al-Din’s name as Kayka’us, which is the latter part of the name of his brother ‘Izz al-Din, and not Kayqubadh as it rightly should be. The only brother mentioned in the waqfiyya as sultan is ‘Izz al-Din, so it is obvious that this inscription has been tampered with and (partially) changed to the name of his youngest brother. Other problems concern the choice of the content: the genealogy of the inscription is much longer than usual, but there is no caliphal title. Qaratay’s name is mentioned, but without title, whereas the inscription,

9.

10.

11.

12.

13.

14.

15.

16. 17. 18. 19.

unusually, has a benediction for him, but not for the sultan. ˘ Mehmet Eminoglu, Karatay Medresesi Yazı ˙ Incileri (Konya, n.d.), 16–41. See Konyalı, ˘ Karatay Konya Tarihi 847 and Eminoglu, Medresesi 15–41 for h.adith collections that could have been consulted for these wellknown sayings of the Prophet. ˘ Eminoglu, Karatay Medresesi 10, without demonstrating how he arrives at this conclusion, writes that this abstract geometric patterning actually is a symmetric pairing of the words ‘Allah’ and ‘Muh.ammad’. Aptullah Kuran, Anadolu Medreseleri (Ankara, 1969), 51–2, considers this to have been a domed space, whereas Meinecke, Fayencedekorationen Vol. 2, 287, proposes a segmented dome. Earlier, I discussed the relationship between the fac¸ade of this building and that of the Alaeddin mosque. Because of this connection, and because it is likely that the same workshop executed the tile mosaic here and, twenty years earlier, of the mihrab and dome of the Alaeddin mosque, this tile program could be seen as an elaboration or reworking of the program there. Unfortunately, little remains of the earlier programme – the heart of the mihrab was replaced in the nineteenth century, and the dome has lost its tile decoration, so any possible connection between the two tile programs remains impossible to explore in any sustained way. Priscilla Soucek, ‘The Life of the Prophet: Illustrated Versions,’ in Content and Context of Visual Arts in the Islamic World, ed. Priscilla Soucek (University Park, 1988), 194. Mohammed Marmaduke Pickthall (trans.), The Meaning of the Glorious Koran (New York: Mentor, n.d.). I modified this translation only by substituting ‘God’ for ‘Allah’. Annemarie Schimmel, And Muhammad Is His Messenger. The Veneration of the Prophet in Islamic Piety (Chapel Hill, 1985), 83–4 for his role as intercessor, 123–30 for the light of Muh.ammad, 127 for Ibn ‘Arabi’s promotion of the idea of the light of Muhammad, and 134 for Ibn ‘Arabi’s idea of Muhammad as ‘the suture between the Divine and the created world’ and ‘the isthmus between the Necessary and contingent existence’. Turan, ‘Celaleddin Karatay,’ 130. Ibid., 91. ¨ Ibn Bibi, El-Evamiru’l-‘Ala’iyye, 593–4; Lewis, Rumi, 410. Turan ‘Celaleddin Karatay,’ 122–3.

Intercession and Succession, Enlightenment and Reflection

20. For a critique of a strand of Islamic art history that finds unchanging religious and mystical ˘ meaning in decoration, see Gulru Necipoglu, ¨ The Topkapı Scroll – Geometry and Ornament in Islamic Architecture (Los Angeles, 1995), 73–83. 21. Osman Turan, ‘S¸emseddin Altun-Aba, Vakfiyyesi ve Hayatı,’ Belleten 11 (1947), 229; Sadi Bayram and Ahmet Hamdi Karabacak, ‘Sahib Ata Fahru’d-Din Ali’nin Konya, ¨ ˙ ¨ Imaret ve Sivas Gokmedrese Vakfiyeleri,’ Vakıflar Dergisi 13 (1981), 56. 22. Robert Hillenbrand, Islamic Architecture. Form, Function, and Meaning (New York, 1994), 210, notes that of fifty-three madrasas surviving from medieval Anatolia, only fifteen are domed. On pp. 210–12, he discusses possible reasons – climatic, compositional and cultural – for the phenomenon of the domed madrasa. In considering the matter of overlap with Christian traditions he has the following to say: ‘First, the compact madrasa with large central dome and smaller domed or vaulted areas surrounding it cannot fail to recall the standard type of mid-Byzantine church, which was widespread in Anatolia at the time of the Turkish conquest. Nor is this

169

resemblance simply a matter of external silhouette; the rear ı¯w¯an flanked by dome chambers in the domical madrasas of Konya brings to mind a Byzantine church apse flanked by diaconicon and prothesis.’ Kuran, Anadolu Medreseleri, 8–9, finds the origin of the domed courtyard madrasa in pre-Islamic Central Asian domestic architecture and the origin of the open-courtyard madrasa in Buddhist monasteries. 23. See Kuran, Anadolu Medreseleri, 62–3, for the subgroup of domed madrasas erected beginning with this building. 24. Meinecke, Fayencedekorationen, 2: 141, for the C ¸ ifte Minareli Medrese in Erzurum, 308 for the Sahib Ata complex in Konya (1258). There are similar roundels at the ¨ Medrese bases of the minarets of the Gok (1271) and C ¸ ifte Minareli Medrese (1271– 2) in Sivas that also likely once contained these names; unfortunately these roundels are now empty. For the Tahir and Zuhre mosque ¨ in Konya see Ruc ¨ ¸ han Arık and Olus¸ Arık, Tiles. Treasures of Anatolian Soil. Tiles of the ˙ Seljuk and Beylik Periods (Istanbul, 2008), 151, fig. 109 and Meinecke, Fayencedekorationen, 2: 360.

鵻 CHAPTER EIGHT

REMEMBERING FERNANDO: MULTILINGUALISM IN MEDIEVAL IBERIA 鵼 Tom Nickson

Scholars of medieval Iberia have invariably understood multilingual inscriptions as signs of multiconfessional and multiethnic communities, ‘relic[s] of Judaeo-Spanish harmony’, the written deposits of a culture of convivencia (living together).1 Indeed it was in the 1962 edition of his seminal formulation of ˜ that Am´erico Castro the concept of convivencia, La realidad hist´orica de Espana, first drew the attention of modern scholars to the multilingual epitaphs of ´ (d. 1252).2 Written in Latin, Castilian, King Fernando III of Castile-Leon Arabic and Hebrew, these epitaphs originally belonged to the tomb erected by Fernando’s son, Alfonso X, in the mosque that served as Seville’s cathedral following Fernando’s capture of the city in 1248. In a recent assessment of convivencia in the visual arts, the epitaphs were described as Alfonso’s elegy to his conqueror father’s plural kingdom, and to a Castilian universe in which the public presence of Jews and Muslims was a matter of course, peoples of the realm with their own monumental languages in which the king himself could inscribe their versions of the life of Ferdinand. No less, though, this was Alfonso’s proclamation of his own imperial vision and ambitions, especially those that arose from the culture of translation his father’s tomb immortalizes.3

Although somewhat qualified, this beguiling account betrays notions of plurality that are as much indebted to the authors’ experiences of the ‘melting pot’ ´ cultures of Manhattan as to the historical reality, the ‘realidad historica’, of medieval Iberia.4 In this chapter I question such pluralistic interpretations of multilingual texts: the careful juxtaposition of different scripts and languages was never neutral, I argue, but resulted from a complex process of multiple translation that 170

Remembering Fernando

51. Arabic and Hebrew epitaphs from the tomb of Fernando III in Seville cathedral, between 1252 and 1284. (Photo: Luis Arenas, after an anonymous watercolour)

produced a fiction of universal, sapiential kingship and multiconfessional consensus. On one level the epitaphs worked visually, especially for those – the great majority – who could not read them in their entirety (or chose not to), but saw them next to emblems of Seville’s new Christian rulers and could probably at least recognise different scripts. Yet the majority of contemporary visitors to Seville’s converted mosque would not have even seen them. Therefore I also consider the epitaphs’ oral and aural performance and reception.5 How did their meaning and content repeat and resound through space and time? How can a fresh examination of these epitaphs shed further light on the material and written artefacts typically associated with convivencia and the plural cultures of the Mediterranean? The epitaphs survive on two rectangular marble panels, now found embedded in the altar of Seville cathedral’s sixteenth-century royal chapel (Figs. 51 & 52). Each panel contains two epitaphs divided by a band of castles and rampant ´ – the two kingdoms united lions, the heraldic emblems of Castile and Leon by Fernando III in 1230. On the altar’s north face are the Castilian and Latin epitaphs, carved in a legible Lombardic script that bears traces of paint and gilding. To the south is the Arabic epitaph, carved in elaborate floriated kufic, and paired with the Hebrew epitaph. In all four epitaphs the writing is raised rather than inscribed, a time-consuming and expensive technique rarely used in Latin or Castilian inscriptions of the period.6 Notoriously difficult to access or photograph, the inscriptions nonetheless received considerable scholarly attention in the early modern period, largely as documents of Fernando’s saintly life and death, or of different calendar systems (the date of Fernando’s death is recorded according to the Gregorian, Hispanic era, Hajiri or biblical calendars, respectively).7 THE MONUMENT TO FERNANDO III

The epitaphs were installed in their current location as part of the reorganisation of Seville’s royal chapel in 1579, but must have been conceived soon after

171

172

Tom Nickson

52. Latin and Castilian epitaphs from the tomb of Fernando III in Seville cathedral, between 1252 and 1284. (Photo: Luis Arenas)

Fernando’s death in May 1252 and well before the 1290s, when Fernando’s tomb monument achieved its final form.8 The tomb was dismantled when the converted mosque-cathedral was destroyed and replaced by the current gothic structure in the fifteenth century, but fortunately recent scholarship has clarified the original appearance of both tomb and mosque-cathedral and can be summarised here.9 The plan of Seville’s Almohad mosque survives below the foundations of the current gothic cathedral, and its minaret still survives in the cathedral’s famous bell tower, the Giralda.10 When consecrated as a cathedral – some time after Seville’s capture in 1248 – the mosque was enriched with Christian paintings and ´ furnishings. As in Cordoba’s Great Mosque, which was converted in 1236, new inscriptions in Latin, Castilian and even Arabic were gradually added alongside the Arabic epigraphy retained from the old mosque: a fittingly multilingual context for Fernando’s epitaphs.11 The building was also reoriented towards the east, although the high altar was located not at the east end but at the centre of the old mosque, roughly on an axis with the old mihrab.12 At his death in 1252 Fernando was buried near this high altar, but in 1279 Alfonso X moved the body of Fernando’s wife, Beatrice of Swabia, to a new monument erected in a vast royal chapel that occupied most of the eastern half of the mosque-cathedral (Fig. 53).13

Remembering Fernando

53. Reconstruction of the converted mosque-cathedral at Seville, after A. Jimenez. A = royal chapel, B = high altar, C = former mihrab.

Fernando may have been moved at the same time, but probably earlier; Alfonso was also buried in the chapel after his death in 1284. The final appearance of the extraordinary funerary monument before its partial destruction in the fifteenth century can be pieced together from visual evidence, inventories, surviving fragments and written descriptions. Raised up on a vaulted platform were three sarcophagi, silver-plated and covered with heraldic emblems. Behind them, in some kind of tabernacle, were life-sized seated effigies of Fernando flanked by Alfonso and Beatrice, the whole monument brightly lit by lamps.14 The effigy of Fernando carried the sword with which, somewhat anachronistically, he had supposedly slain Mohammed; this was removed and carried each year in the annual procession to celebrate Seville’s

173

174

Tom Nickson

conquest (St Clement’s Day), and was occasionally taken on military campaigns.15 Raised above these effigies and encased in a silver tabernacle was the Virgen de los Reyes, which still survives in the royal chapel, albeit much transformed.16 An early fourteenth-century seal from Seville gives the clearest indication of this arrangement (Fig. 54). As with the effigies, the Virgen de los Reyes had moveable parts and wore real clothes and real jewellery. The combination of these figures must have been an extraordinarily potent statement of the presence, authority and beliefs of Seville’s new political and religious rulers given that there had been no figural sculpture under Almohad rule.

READING THE EPITAPHS

The epitaphs were almost certainly surrounded by other inscriptions in Latin, Castilian and Arabic, but it is not clear exactly where in the royal chapel Fernando’s epitaphs were originally placed.17 The opening of the Latin and Castilian epitaphs, ‘here lies’, implies proximity to the sarcophagi, though at 142 cm long × 57 cm high the epitaphs would not easily fit either on the sarcophagi or on nearby piers.18 The whole chapel was enclosed by an iron screen with gates: various contemporary descriptions indicate that these gates were usually locked and well guarded, and even though the epitaphs preserve fragments of their original paint, it is unlikely that they could have been seen or read by those outside the large chapel.19 Nor is it certain that lay visitors regularly entered the chapel to visit Fernando’s tomb: Fernando was not formally canonised until 1671, and there is only limited evidence of a saintly cult in the medieval period.20 Yet a detailed description of the chapel from 1345 implies that access was not entirely restricted to its chaplains, and that on occasions visitors could admire the tomb – and its epitaphs – from a close distance.21 This is confirmed by a mid fourteenthcentury chronicle of Alfonso X’s rule, which suggests that large crowds could visit the chapel on certain days:22 Every year king Alfonso commemorated the anniversary of the death of his father, king Fernando, in this fashion. A great many men would come from all over Andalucia to do Fernando honour, and brought with them all the banners and standards of each place, and with each banner they brought many wax candles [ . . . ] king Alhamar of Granada [Ibn al-Ahmar] [ . . . ] sent king Alfonso important men of his court and with them one hundred foot soldiers; each one carried a lighted candle of white wax, and these hundred candles they would place around the tomb where King Fernando lay buried. This, Aben Alhamar did in honour of the king.23

This account of Muslim ambassadors and soldiers in Seville cathedral corroborates evidence from other sources that Jews and Muslims regularly entered church

Remembering Fernando

54. Seal of the city of Seville, 1311 (left). Eighteenth-century drawing in Seville cathedral ´ Colombina, Seville) archive, MS. 57-3-40 (fol. 63r.). (Photo: Institucion

spaces in medieval Iberia; it describes one occasion when the Arabic epitaph might have been read, perhaps aloud.24 Despite the flight of most of Seville’s native Arabic-speaking community in 1248 and in the following decades, a small number of converts to Christianity (conversos) remained in the city and could also have read the Arabic epitaph.25 And although there seems to have been no sizeable Jewish community in Seville before its conquest, Alfonso X encouraged the settlement there of Arabic-speaking Jews and Christians (Mozarabs) from Toledo who would make capable multilingual administrators – and potential readers of the epitaphs.26 But to what extent were these epitaphs actually read, even by those with the necessary linguistic skills? Castro observed that Fernando’s Latin epitaph is the only one to include an inflammatory passage describing how Fernando seized Seville ‘from the hands of the pagans and restored it to the Christian cult’, a phrase indebted to contemporary chronicles and chancery documents.27 This

175

176

Tom Nickson

suggests that there was an expectation that the other three epitaphs might be read by non-Christians; that the epitaph was first composed in the preferred language of the Church, Latin, and then discreetly bowdlerized in translation, in deference either to the translator or to a notional community of readers. This suggestion is corroborated by what is known of contemporary translation practices: fluency in both Hebrew and Arabic was largely confined to Jews, who could also understand Castilian but were seldom literate in Latin.28 Thus one person (perhaps Seville’s first archbishop, Raimundo de Losana, d. 1286?) must have been responsible for the Latin and Castilian epitaphs, and then the latter – either in oral or written form – was converted into Hebrew and Arabic by a Jewish or converso translator.29 This would explain, for instance, why the Hebrew epitaph has little literary merit: its awkward grammar and lack of traditional biblical or Hebrew idiom betray the fact that it derives from the Castilian epitaph.30 Once composed, the texts were probably then written directly onto the marble, to be carved by specialists in inscriptions who may have understood their content only partially or not at all.31 But the Castilian epitaph not only attests to an intermediate stage in the translation process. It also represents the endorsement and elevation of Castilian as a new national language, spoken by people of all confessions – a project closely associated with Alfonso X, though already begun under Fernando III.32 Here, the new status of Castilian is underlined by its parity with the Latin, Hebrew and Arabic epitaphs in terms of size and position. Brief consideration of multilingual texts elsewhere in the Peninsula sheds further light on the significance of Fernando’s epitaphs. The epitaph of Miqail ibn Simino/Miguel Jim´enez (d. 1156), originally in the Mozarabic parish church of Santas Justa y Rufina in Toledo, is written in the two languages used regularly by Toledo’s Mozarabs, Arabic and Latin. But it was not absolutely necessary to read the epitaphs to understand them: their visual appearance was equally significant. Many Toledans would have recognised the Arabic and Latin scripts even if they could not decipher them, and thus they would have understood that this was the epitaph of a Mozarab, descended from Christians living in Arabic domains.33 Elsewhere, different languages are rendered together in the same script. For instance, the Latin epitaph of Petrus de la Cera (d. 1265), now in Seville’s Giralda, ends with three lines in Castilian that repeat the date of his death, not according to the Roman system but following the Hispanic era system, which dates events from the birth of Augustus (38 BCE).34 The reasons for this are not entirely clear, especially because the reverse is found in contemporary chancery documents. Fernando III’s Privilegio Rodado to Seville cathedral of 20 March 1252, for instance, is largely written in Castilian, but is dated in Latin and according to the Gregorian calendar. Echoing (or anticipating) the epitaphs, the scribe also added that this was ‘four years after that same most victorious king Fernando captured the most noble city of Seville and restored it to the Christian cult’.35

Remembering Fernando

Further examples attest to the close relationship between diplomacy, commemoration and multilingualism in postconquest Seville. In the 1250s, Alfonso X’s Muslim vassals were frequently named as witnesses to Seville’s royal privileges, their names helpfully transliterated by notaries in a clear gothic hand alongside other members of the court. Thus ‘don Aboabdille auenac¸ar, Rey de Granada et vassallo del Rey don Alffonsso’ is named as a witness to a document of August 1252.36 But a number of thirteenth-century peninsular surrender treatises also survive where Latin and Arabic content is inter-lineated.37 These clearly have a different function from the chancery documents, but do not depend on bilingual readers. Whether one reads the Arabic or the Latin, the close juxtaposition of familiar and unfamiliar languages implies shared content and thus the consent of those communities associated with the unfamiliar language: readers of one script could simply assume that the other script had precisely the same meaning. Fernando’s Hebrew and Arabic epitaphs can be interpreted in the same way. Like the keys with Hebrew, Arabic and Castilian epigraphy that were supposedly given to Fernando at Seville’s conquest, the epitaphs claim to speak to Seville’s Jewish and Muslim communities and at the same time for them, implying a willing transfer of power.38 Their elegance and permanence conceal discord. They are sealed and ratified by their juxtaposition with one another and with the royal arms that divide them. Their physical similarity intimates identical content, disguising subtle differences: Christians, Jews and Muslims are united in their celebration of Fernando’s virtues, they hint seductively. It was not necessary to read each epitaph in order to identify them with a notional confessional community. Their mere presence suggests the consent of each notional community, providing reassurance, in particular, for new Christian settlers in this frontier city – especially necessary after the mud´ejar revolts of 1264, when Muslims living in Christian territories rose up against Alfonso X.

THE WIDER DISCOURSE

Those who did read the Latin or Castilian epitaphs, at least in part, may have recognised wider affinities.39 The epitaphs’ opening lines mirror the conventional listing of Fernando’s territories from contemporary chancery formulae, endowing the epitaphs with quasi-official documentary status.40 The description of Seville as ‘metropolis totius Hispaniae’ asserts Seville’s pretensions to ecclesi˜ implies astical primacy, whereas the city’s designation as ‘cabeza de toda Espana’ political precedence over other cities in the realm.41 The account of Seville’s con´ quest chimes with other monumental inscriptions, such as those above Cordoba ´ cathedral’s original high altar that commemorate Cordoba’s conquest.42 The events recorded in the epitaphs were also remembered and reenacted liturgically:

177

178

Tom Nickson

in 1254 Alfonso X endowed the feast of St Clement (the day of Seville’s conquest) in Seville cathedral in honour of ‘the very noble king don Fernando who lies buried there’, and the anniversary of Fernando’s death was supported by a series of royal endowments and papal indulgences.43 The epitaphs’ precise dating of Fernando’s death thus serves as a monumental (and hence authoritative) equivalent to the cathedral’s books of anniversaries, privileges and obits – parts of which were typically read aloud during the offices read during the hours of Prime.44 To some extent Fernando’s tomb and epitaphs may thus have served as prompts for oral commemoration, so that the content of the epitaphs resounded throughout Seville’s converted mosque and beyond. In thirteenth-century Saint-Denis the vitae of Louis VI and Dagobert were read on the anniversaries of their deaths, in the manner of saints.45 No such practice is explicitly recorded in medieval Castile, although commemoration of Alfonso VIII at Las Huelgas, orchestrated by Alfonso X and culminating in the consecration of the Holy Cross altar in 1279, was enriched with a variety of processions, musical pieces and feasts.46 A number of poems were likewise composed in honour of Seville’s capture and Fernando’s death by the jongleur, Pero da Ponte, and presumably were performed at court.47 An early fourteenth-century life of Fernando, seemingly conceived in response to contemporary lives of St Louis, also derives from oral sources – perhaps the sermon preached at Fernando’s death or brief vitae read on his anniversary.48 In this way, the epitaphs can be properly understood as part of a much wider practice of commemoration in Seville cathedral and beyond. Events from the lives of Fernando and Alfonso were also included in the collection of Marian miracle stories known as the Cantigas de Santa Mar´ıa, which were to be sung in the cathedral on feast days, according to Alfonso’s will.49 Indeed, Cantiga 292 actually refers to a miracle that took place at Fernando’s tomb.50 Fernando and Alfonso were thus surreptitiously commemorated, their memory performed through song and images, even during liturgical celebrations that were not explicitly royal. It is therefore apt that Francisco Prado-Vilar has identified in the Cantigas a strategy of presenting narratives of trans-confessional consent: they stage, in his words, a ‘national utopia’.51 The same is true of the epitaphs. Connections with a wider discourse of commemoration are indeed profound. In the prologue to the Setenario, Fernando’s very name is laden with moral significance. This law code, a kind of first draft of the famous Siete Partidas, was conceived under Fernando, but received a new prologue in the last years of Alfonso’s life, probably after 1279 when he rearranged the royal tombs in Seville.52 In the opening section – often described as an elegy to Fernando – the letters making up Fernando’s name become an acrostic recording his virtues, with the wording similar in parts to Fernando’s Castilian epitaph.53 Such widely used word plays imply that words and letters carried a special moral resonance that reached beyond their content. Just as each letter of Fernando’s name prompted contemplation of his virtues, so too could short texts such as the epitaphs prompt

Remembering Fernando

far more extensive acts of reflection and commemoration.54 In the words of Seville’s very own saintly scholar, St Isidore, written letters were invented ‘in order to remember things. For lest they fly into oblivion, they are bound in letters [litteris alligantur]. For so great is the variety of things that all cannot be learned by hearing, nor contained only in memory’.55 Support for these ideas is found in an epilogue added around 1255 to the Libro de los doce sabios, a kind of mirror of princes compiled under Fernando III and Alfonso X.56 This epilogue – written roughly at the same time as the epitaphs were made – describes twelve wise men who approach Alfonso and recommend that Fernando’s tomb [sepoltura] ‘be titled [titulado] with the sayings [dichos] of each of us, so that his saintly and good memory endures in the world forever’. Alfonso then ‘asked them that they put these things into writing so that he could then put them on his [Fernando’s] tomb in golden letters, very richly made, as befitted it’. These short sayings include such phrases as ‘Your end is better than your beginning’; ‘Your memory will be greater than the time that you lived’; ‘In your lifetime you were handsome in body, in death you prove handsome in soul’; and ‘Till now those who knew you praised you, now those that didn’t know you will also praise you’.57 The story is more moral than historical, and manuscripts of the Libro were not widely diffused.58 Yet such trite apothegms not only suggest that the performance of remembering was more important than the content of commemorative texts but also that epitaphs were considered the products of oral and collective practices, that they gave voice to communities of mourners. This tale of royal counsel further configures the tomb as the site where sapiential kingship is inherited, and indeed ideas of sapientialism are key to a proper understanding of the epitaphs. Alfonso X’s self-fashioning as el Sabio, the Learned, is well known, but Fernando III was also framed in Solomonic terms by his contemporaries.59 In Castile and across the medieval Mediterranean multilingual abilities were a topos, an index of learning especially suited for those with imperial ambitions.60 Jeremy Johns has recently shown, for example, that the use of Arabic signatures or multilingual inscriptions by the Norman kings of Sicily does not prove their fluency in these languages, but rather the desire to appropriate the cultures with which they were associated and to demonstrate the king’s superior knowledge.61 The same was true in thirteenth-century Castile. Alfonso X’s sponsorship of multiple translations has occasionally seduced scholars into believing that Alfonso X knew Arabic, despite evidence to the contrary.62 The use of different scripts and languages on Fernando’s epitaphs is equally seductive, but should be understood in the same way as those translations, or as the choice to be buried in Andalusi textiles with Arabic epigraphy.63 The multilingual epitaphs hint at access to arcane sources of knowledge and proximity to prestigious systems of luxury and display; they imply that Fernando and Alfonso ruled all the peoples that made up their kingdom(s). Yet they are no mere reflection of a polyglot society, nor even an elegy to Fernando’s plural kingdom. Even for those who could not read them, the epitaphs fashion an image of sapiential kingship,

179

180

Tom Nickson

effacing dissent and staging a utopian vision of a multiconfessional society united in its recognition of crown and church. TRANSLATIONS OF THE EPITAPHS Latin

´ Here lies the most illustrious king Ferrandus of Castile, and of Toledo, Leon, ´ Galicia, Seville, Cordoba, Murcia, and Ja´en, who conquered all Hispania, the most loyal, the most veracious, the most constant, the most just, the most energetic, the most tenacious, the most liberal, the most patient, the most pious, the most humble and the most effective in fear and in the service of God. He destroyed and utterly exterminated the impudence of his enemies. He raised up and exalted all who were his friends; he seized the city of Seville, capital and metropolis of all Spain, from the hands of the pagans and restored it to the Christian cult, and that is the city where he paid his debt to nature and passed to the Lord on the last day of May in the year of the Incarnation of the Lord, 1252.

Castilian

Here lies the very honourable king don Ferrando, lord of Castile and of Toledo, ´ of Galicia, of Seville, of Cordoba, ´ of Leon, of Murcia, and of Ja´en, he who ˜ the most loyal and the truest and most forthright and most conquered all Espana, courageous and most comely and most illustrious and most forbearing and most humble and he who most feared God, and he who most served Him, and he who broke and destroyed all his enemies and who exalted and honoured all his friends, and conquered the city of Seville, which is the head of all Spain, and he died there on the last day of May in the era of 1290.

Arabic

´ This is the tomb of the great king don Ferando, lord of Castile, Toledo, Leon, ´ Galicia, Seville, Cordoba, Murcia, and Ja´en, may God be pleased with him. Who ruled all al-Andalus, [who is] the most faithful, the most veracious, the most enduring, the most just, the most valiant, the most propitious, the most noble, the most forbearing, the most visionary, the greatest in modesty, most suitable to God and His greatest servant. He died (God had mercy on him) on the Friday night and God raised him. He ennobled and honoured his loved ones and took possession of the city of Seville, which is the head of all al-Andalus, and in which he broke and destroyed all his enemies. [He] died on the twentieth of the month of Rabi’ of the year 650 al-hijra.

Remembering Fernando

181

Hebrew In this place is the tomb of the great king don Ferando ´ and Galicia and Seville Lord of Castile, Toledo and Leon ´ And Cordoba and Murcia and Ja´en. May his soul [rest] in Paradise. He who Conquered all Sefarad, the upright, righteous, the stronghold and tower, The mighty, pious, humble, [a] G’d-fearing [man] who served Him all His days. Who broke and destroyed all his enemies and raised up and honoured all His friends, and who conquered the city of Seville, which is the head [= capital city] of all Sefarad and who died in it on Friday night the twenty-second day of the month of Sivan in the year five thousand and twelve since the Creation of the world.

NOTES

I am extremely grateful to the following for their advice on these epitaphs: Raquel Alonso, Jessica Berenbeim, Susan Einbinder, Laura Fern´andez Fern´andez, Eva Frojmovic, Francisco Hern´andez, Kirstin Kennedy, Marcia Kupfer and Mariam Rosser-Owen. Susan Einbinder kindly provided a new translation of the Hebrew epitaph, but all other translations are based on those by Sarah Pearce in J. Dodds, M. R. Menocal, and A. Krasner, The Arts of Intimacy. Christians, Jews, and Muslims in the Making of Castilian Culture (Yale, 2008), 200–1, with ´ slight amendments. The transcriptions reproduced in The Arts of Intimacy, taken from Enrique Florez, are inaccurate. 1. E. Kedourie, “Introduction”, in Spain and the Jews: the Sephardi experience. 1492 and after, ed. Elie Kedourie (London, 1992), 31. 2. Am´erico Castro, La realidad histórica de España, 2nd ed. (Mexico City, 1962), 38–9. 3. See Jerrilynn Dodds, Maria Menocal, and Abigail Krasner, The Arts of Intimacy, 196– 202. 4. See the interview with Jerrilynn Dodds in Education Update 14.10 ( June 2009): www.slc .edu/media/news-events/pdf/Jerrilyn Dodds .pdf (accessed 7 July 2010). 5. The classic study of partial literacy is Michael Camille, ‘Seeing and Reading: Some Visual Implications of Medieval Literacy and Illiteracy,’ Art History 8.1 (1985), 24–49. 6. Raised epigraphy was used in Arabic and Hebrew contexts, however. See Dodds, Menocal and Krasner, The Arts of Intimacy, 160; Santiago Palomero Plaza, ‘Capitel bilingue ¨ de Toledo,’ in Memoria de Sefarad, ed. I. G. Bango Torviso (Toledo, 2003), 245. I am grateful to Peter Draper for this observation. ˜ 7. See Alonso Nunez de Castro, Vida de San Fernando, el tercer, rey de Castilla, y Le´on (Madrid, 1673), fols. 140r–41r; Juan Jos´e Antequera Luengo, Memorias sepulcrales de la Catedral de Sevilla. Los manuscritos de Loaysa

y Gonz´alez de Le´on (Seville, n.d.), 56–7; Jordi ´ ed. Epigraf´ıa hebrea (Madrid, Casanovas Miro, ´ 2005), 78–9, and especially Henrique Florez, Elogios del santo rey Dn. Fernando . . . etc. (Madrid, 1754). 8. See Bartolom´e Pozuelo Calero, ‘El tumulo ´ erigido en Sevilla a la apertura de la nueva Capilla Real (1579), obra emblem´atica del Licenciado Francisco Pacheco,’ Excerpta philologica: Revista de filolog´ıa griega y latina de la Universidad de C´adiz 3 (1993), 349– ˜ 72; Fernando Guti´errez Banos, Las empresas art´ısticas de Sancho IV El Bravo (Valladolid, 1997), 151. The tomb must have been largely finished when Juan Gil de Zamora described it c. 1270: Roc´ıo S´anchez Ameijeiras, ‘La ´ fortuna Sevillana del Codice Florentino de las Cantigas: tumbas, textos e im´agenes,’ Quintana 1 (2002), 260; Laura Fern´andez Fern´andez, ‘“Muy noble, et mucho alto et ´ de la mucho honrado”. La construccion imagen de Fernando III,’ in Fernando III. Tiempo de Cruzada, eds. Carlos de Ayala Mart´ınez and Mart´ın Federico R´ıos Saloma (Madrid, 2012), 138–41. Payments of 15,000 maravedies were still being paid ‘A los monumentos de los reyes’ in 1294: Francisco Hern´andez, Las Rentas del Rey. Sociedad y Fisco

182

Tom Nickson

en el Reino Castellano del siglo XIII, 2 vols. (Madrid, 1993), 1: 498. 9. Tomb: Javier Mart´ınez de Aguirre, ‘La ´ primera escultura funeraria gotica en Sevilla: la Capilla Real y el sepulcro de Guzm´an ˜ de el Bueno (1248–1320),’ Archivo Espanol Arte 68 (1995), 113–22; Mar´ıa Jesus ´ Sanz Serrano, ‘Imagen del antiguo tabern´aculo de plata, de la Capilla Real de Sevilla, a trav´es de dos sellos medievales,’ Laboratorio de Arte: Revista del Departamento de Historia del Arte 11 (1998), 51–68; Cynthia Chamberlin, ‘Unless the Pen Writes as It Should: the proto-cult of Saint Fernando III in Seville in the thirteenth and fourteenth centuries,’ in Sevilla 1248. Congreso Internacional commemorativo de la conquista de la ciudad de Sevilla por Fernando III, Rey de Castilla y Le´on, ed. Manuel Gonz´alez Jim´enez (Madrid, 2000), 389–418; Mar´ıa Jesus ´ Sanz Serrano, ‘Ajuares funerarios de Fernando III, Beatriz de Suabia y Alfonso X,’ in Sevilla 1248, ed. Gonz´alez Jim´enez, 419–47; Teresa Laguna Paul, ´ ‘La capilla de los reyes de la primitiva catedral de Santa Mar´ıa de Sevilla y las relaciones de la Corona Castellana con el cabildo hispalense en su etapa fundacional,’ in Mara˜ Medieval. Tesoro Sagrada y villas de la Espana Monarqu´ıa, ed. Isidro Bango Torviso (Madrid, 2001), 235–49; S´anchez Ameijeiras, ‘La ´ fortuna Sevillana’; Raquel Alonso Alvarez, ‘De Carlomagno al Cid: la memoria de Fernando III en la Capilla Real de Sevilla,’ in Fernando III y su tiempo (1201–1252). VIII ´ 2003), Congreso de Estudios Medievales (Leon, 471–88; Amy Remensynder, ‘Marian Monarchy in Thirteenth-Century Castile,’ in The Experience of Power in Medieval Europe, 950– 1350, eds. Robert Berkhofer III, Alan Cooper and Adam Kosto (Aldershot, 2005), 253–70; ´ Alonso Alvarez, ‘Los enterramientos de los ´ y Castilla hasta Sancho IV. reyes de Leon Continuidad din´astica y memoria regia’, eSpania 3 (2007); Laguna Paul, ´ ‘El Imperio y la Corona de Castilla: la visita a la capilla de los Reyes de Sevilla en 1500,’ in El intercambio art´ıstico entre los reinos hispanos y las cortes europeas en la Baja Edad Media, eds. Mar´ıa C. Cosmen, Mar´ıa Victoria Herr´aez Ortega ´ Gomez-Calcerrada ´ ´ and Mar´ıa Pellon (Leon, 2009), 217–38; idem, ‘Si el nuestro cuerpo fuere enterrado en Sevilla’. Alfonso X y la Capilla de los Reyes,’ in Alfonso X el Sabio, ed. I. Bango Torviso (Murcia, 2009), 116– 29; Fern´andez Fern´andez, ‘“Muy noble, et mucho alto”. Mosque conversion’: Laguna Paul, ´ ‘La aljama cristianizada. Memoria de

10.

11.

12.

13. 14.

15.

16.

la catedral de Santa Mar´ıa de Sevilla,’ in Metropolos totius hispaniae. 750 aniversario de la incorporaci´on de Sevilla a la corona castellana, ed. A. J. Morales (Seville, 1998), 41–71; Amy Remensynder, ‘The Colonization of Sacred Architecture: the Virgin Mary, mosques, and temples in medieval Spain and early sixteenth-century Mexico,’ in Monks & Nuns, Saints & Outcasts. Religion in Medieval Society. Essays in Honor of Lester K. Little, eds. Sharon Farmer and Barbara Rosenwein (London, 2000), 189–219; Alfonso Jim´enez Mart´ın, ‘Las Fechas de las Formas,’ in La Cated´ ´ y fabrica de ral Gotica de Sevilla. Fundacion la obra nueva, ed. Alfonso Jim´enez Mart´ın (Seville, 2006), 15–113; Antonio Almagro, ´ ‘De Mezquita a Catedral. Una adaptacion imposible,’ in La piedra postrera. V centenario de la conclusi´on de la Catedral de Sevilla, ed. Alfonso Jim´enez Mart´ın (Seville, 2007), 13–45. See Alfonso Jim´enez Mart´ın, ‘Notas sobre la mezquita mayor de la Sevilla almohade,’ Artigrama: Revista del Departamento de Historia del Arte de la Universidad de Zaragoza 22 (2007), 131–54. See, for example, the bronze doors preserved from the old mosque, or the new sacristy door carved after 1248 with fragments of Arabic: ´ en ʿAbd al-Az¯ız S¯alem, ‘La Puerta del Perdon la gran mezquita de la alcazaba almohade de Sevilla,’ Al-Andalus 43 (1978); Teresa Laguna Paul, ´ ‘Door of Ancient Tabernacle,’ in Ibn Khaldun: The Mediterranean in the 14th cen´ tury. The rise and fall of empires, ed. Jeronimo ´ P´aez Lopez (Seville, 2006). Laguna Paul, ´ ‘Alfonso X y la Capilla de los ´ Reyes,’ 117. The same occurred at Cordoba: Manuel Nieto Cumplido, La catedral de ´ C´ordoba (Cordoba, 1998), 449–53. Manuel Gonz´alez Jim´enez, Alfonso X El Sabio (Barcelona, 2004), 45. Fidel Fita, ‘Biograf´ıas de San Fernando y Alfonso el Sabio por Gil de Zamora,’ Bolet´ın de la Real Academia de la Historia 5 (1884), 308–28, 321; Hern´andez, Las Rentas, vol. 1, 391. See Teofilo F. Ruiz, ‘The Symbolic Meaning of Sword and Palio in Late Medieval and Early Modern Ritual Entries: the case of Seville,’ in Localizaci´on: Memoria y civilizaci´on: anuario de historia de la Universidad de Navarra 12 (2009), 13–48; Laguna Paul, ´ ‘Espada del rey Fernando III, el Santo,’ in Metropolis totius hispaniae, ed. Morales, 130. Laguna Paul, ´ ‘Virgen de los Reyes,’ in Maravillas, ed. Bango Torviso, 435.

Remembering Fernando

17. See Laguna Paul, ´ ‘El Imperio,’ 235, for the Andalusi fabrics that belonged to the chapel. 18. Sarcophagi dimensions: Sanz, ‘Ajuares funerarios,’ 443. For other royal epitaphs in medieval Iberia, see Xavier Dectot, Les tombeaux des familles royales de la p´eninsule ib´erique au ˆ (Turnhout, 2008), 76–80. Moyen Age 19. See cantigas 292 & 324: Kathleen Kulp-Hill, Songs of Holy Mary of Alfonso X, The Wise. A translation of the Cantigas de Santa Maria (Temple, 2000), 353, 393; Alvar Garc´ıa de Santa Mar´ıa, Cronica de Juan II de Castilla, ed. Juan de Mata Carriazo (Madrid, 1982), 190; Hern´andez, Las Rentas, vol. 1, 392. ˜ entitled 20. A chapter in the Estoria de Espana, ‘Miracles that God performed for the sake of holy King Fernando, who lies in Seville, after he had died’, remains blank: Diego ´ Men´endez Pidal, eds., Catal´an and Ramon ˜ 3rd ed. Primera cr´onica general de Espana, (Madrid, 1977), ch. 1135. See also Luis ´ Fern´andez Gallardo, ‘La Cronica particular de San Fernando: sobre los or´ıgenes de la ´ cronica real castellana, II. Los contenidos,’ Cahiers d’´etudes hispaniques m´edi´evales (2010), 215–46, 224, n.34. 21. See Chamberlin, ‘The proto-cult,’ 397, n.30. The description does not refer to the epitaphs. 22. Such crowds may explain the great size of the chapel, which is especially striking compared with the small scale of other royal chapels in medieval Iberia. See Juan Carlos Ruiz Souza, ‘Capillas Reales funerarias catedralicias de ´ Nuevas hipotesis ´ Castilla y Leon: interpret´ ativas de las catedrals de Sevilla, Cordoba y Toledo,’ Anuario del Departamento de Historia y Teor´ıa del Arte 18 (2006), 9–30, 11–12. 23. Gonz´alez Jim´enez, Manuel, ed. Cr´onica de ´ el Ms. II/2777 de la Biblioteca Alfonso X: segun del Palacio Real, Madrid (Murcia, 1998), 27. Translation adapted from Fern´an S´anchez de Valladolid, The Chronicle of Alfonso X, trans. Joseph O’Callaghan (Lexington, Ky., 2002), 46–7. 24. See Jos´e S´anchez Herrero, Concilios provinciales y s´ınodos toledanos de los siglos XIV y XV: la religiosidad cristiana del clero y pueblo (La Laguna, 1976), 37; 182; Mark D. Meyerson, The Muslims of Valencia in the Age of Fernando and Isabel: between coexistence and crusade (Berkeley, 1991), 45; Hieronymous ˜ y Portugal (1494– Munzer, Viaje por Espana ¨ 1495) (Madrid, 1991), 291; Nieto Cumplido, La catedral de C´ordoba, 321; KulpHill, Songs, 202 (cantiga 167). For the performance of inscriptions see the later

183

25.

26.

27.

28.

discussion and Chapters 1, 6 and 7 in this volume. See Isabel Montes Romero-Camacho, ‘Mud´ejares y jud´ıos en la Sevilla del siglo XIII,’ in Sevilla 1248, ed. Gonz´alez Jim´enez, 467–98; Heather Ecker, ‘How to Administer a Conquered City in al-Andalus: mosques, parish churches and parishes,’ in Under the Influence: questioning the comparative in medieval Castile, eds. Cynthia Robinson and Leyla Rouhi (Boston, 2005), 45–65, 54–5. Conversos: Klaus Wagner, Regesto de documentos del Archivo de Protocolos de Sevilla referentes a jud´ıos y moros (Sevilla, 1978), 10–11; Heather Ecker, ‘Administradores moz´arabes en Sevilla despu´es de la conquista,’ in Sevilla 1248, ed. Gonz´alez Jim´enez, 821–38, 828–30. For multilingual epitaphs and conversos in Palermo, see Barbara Zeitler, ‘“Urbs felix dotata popula trilinguii”: some thoughts about a twelfthcentury funerary memorial from Palermo,’ Medieval Encounters 2 (1996), 132–5; Jeremy Johns, ‘Three Funerary Memorials to Anna and Drogo, Parents of the Royal Priest Grisandus,’ in Nobiles Officinae: Perle, filigrane e trame di seta dal Palazzo Reale di Palermo, ed. Maria Andaloro, 2 vols. (Catania, 2006), 1: 775–8. Alfonso’s foundation of an ‘estudio y escuelas generales’ of Latin and Arabic in Seville in 1254 may be seen as part of this attempt, though it is not clear if anything came of this: Gonz´alez Jim´enez, Alfonso X, 424; Ecker, ‘Administradores moz´arabes’. Castro, La realidad, 38–9. See In´es Fern´andez ˜ ´ por etapas de la Ordonez, ‘La composicion Chronica latina regum Castellae (1223–1237) de Juan de Soria,’ e-Spania 2 (2006), §57–64. I am grateful to Daniel Rico for this observation. Elaine Miller, Jewish Multiglossia: Hebrew, Arabic, and Castilian in medieval Spain (Newark, Del., 2000), 85–92; Gerold Hilty, ‘El plurilinguismo en la corte de Alfonso X ¨ el Sabio,’ in Actas del V Congreso Internacional ˜ de Historia de la Lengua Espanola, eds. Mar´ıa Teresa Echenique Elizondo, Juan P. S´anchez M´endez and F. Javier Satorre Grau (Madrid, 2002), 207–20, 208; 218; One well-known exception is Yehuda Mosca, who was fluent in Hebrew, Latin and Arabic: Eleazor Gutwirth, ‘“Entendudos”: Translation and representation in the Castile of Alfonso the Learned,’ The Modern Language Review 93, no. 2 (1998), 384–99, 389. See also Pilar ´ Tello, Jud´ıos de Toledo, 2 vols. (MadLeon rid, 1979), vol. 2, doc. 431; Roger Wright,

184

29. 30.

31.

32.

33.

34.

35.

36.

37.

Tom Nickson

Late Latin and Early Romance in Spain and Carolingian France (Liverpool, 1982), 241–4; Francisco J. Hern´andez, ‘The Jews and the Origins of Romance Script in Castile: a new paradigm,’ Medieval Encounters 15 (2009), 259–306. On Raimundo, see Chamberlin, ‘The protocult,’ 408. I am grateful to Susan Einbinder and Eva Frojmovic for their advice on the Hebrew epitaph. Although the Hebrew epitaph is carved rather elegantly, Sheila Blair suggested to me that the Arabic epigraphy is fairly clumsy and formulaic. On carving inscriptions, see Robert Favreau, Les inscriptions m´edi´evales (Turnhout, 1979), 21–31; 50–2. For template makers in Timurid Iran, see Lisa Golombek and Donald Wilber, The Timurid Architecture of Iran and Turan, 2 vols. (Princeton, N.J., 1988), vol. 1, 117–18. See especially Jos´e Manuel Nieto Soria, ‘La monarqu´ıa fundacional de Fernando III’, in Fernando III y su tiempo, 31–66; Francisco M´arquez Villanueva, El concepto cultural alfons´ı (Barcelona, 2004), 41–8; Ray Harris-Northall, ‘Aspects of Official Lan´ Latin and guage Usage in Castile and Leon: the vernacular in the early thirteenth century,’ in Medieval Iberia: changing societies and cultures in contact and transition, eds. Ivy Corfis and Ray Harris-Northall (Woodbridge, 2007), 165–74, 167. ´ Evariste L´evi-Provenc¸al, Inscriptions arabes d’Espagne, 2 vols. (Leiden, 1931), doc. 81; Francisco J. Hern´andez, ‘Language and Cultural Identity: the Mozarabs of Toledo,’ Bolet´ın Burriel 1 (1989), 29–48. Luis F. Mart´ınez Montiel, ‘L´apida de Pedro de la Cera,’ in Metropolis totius hispaniae, ed. Morales, 262. Manuel Gonz´alez Jim´enez, ed. Diplomatario andaluz de Alfonso X (Seville, 1991), 7. See also Pilar Ostos Salcedo, ‘Canciller´ıa castellana y lengua vern´acula: su proceso de consol´ idacion,’ Espacio, tiempo y forma. Serie III, Historia medieval 17 (2004), 471–84. Mercedes Borrero Fern´andez, Seville, Ciudad de Privilegios (Seville, 1995), 90–4. A few contemporary Castilian documents from Seville are signed by Mozarabs in Arabic, however: Ecker, ‘Administradores moz´arabes,’ 836–8. Robert Burns, Paul Chevedden, and Miguel de Epalza, Negotiating Cultures: bilingual surrender treaties in Muslim-Crusader Spain under James the Conqueror (Leiden, 1999).

38. Keys: Teresa Laguna Paul, ´ ‘Llaves de la ciudad de Sevilla,’ in Memoria de Sefarad, ed. Isidro Bango Torviso (Toledo, 2003), 266. 39. The epitaphs still await detailed philological analysis, but they omit, for instance, the epithets of ‘noble’ and ‘bienaventurado’ that are preferred in contemporary descriptions of Fernando. See Chamberlin, ‘The proto-cult,’ 390. Christopher Norton has pointed out to me that the section of the Latin epitaph from ‘Qui contrivit’ to ‘amicos suos’ is reminiscent of the Magnificat – which is highly appropriate given Fernando’s Marian devotion. The string of laudatory epiphets is also common in both Hebrew and Arabic contexts: Ibn Khaldun: The Mediterranean in the 14th Cen´ tury: The rise and fall of empires, ed. Jeronimo ´ P´aez Lopez (Seville, 2006), 110; 238; Francisco Cantera Burgos and Jos´e Mar´ıa Mill´as ˜ Vallicrosa, Las inscripciones hebraicas de Espana (Madrid, 1956), 62–4; 144–5; 168–70. The formula radi allahu anhu in the Arabic epitaph, ‘may God be pleased with him’, is usually reserved for companions of the Prophet: Dodds, Menocal and Krasner, The Arts of Intimacy, 199. See Nieto Cumplido, La catedral de C´ordoba, 450. 40. See, for instance, Fernando’s privilege to Seville cathedral: Gonz´alez Jim´enez, Diplomatario, 6. 41. I am grateful to Francisco Hern´andez for this observation. It may even claim to compete with Granada as head (ras) of all Andalus. For ˜ employed Alfonso’s title of ‘rey de Espana’, just three times between 1257 and 1259, see Laura Fern´andez Fern´andez, Arte y ciencia en el ‘scriptorium’ de Alfonso X el Sabio (Seville, 2013), 48–9. 42. See Nieto Cumplido, La catedral de C´ordoba, 450. 43. Antonio Ballesteros Beretta, Sevilla en el siglo XIII (Madrid, 1913), docs 68 & 96; Jos´e ´ S´anchez Herrero and Mar´ıa Carmen Alvarez M´arquez, ‘Fiestas y devociones en la Catedral de Sevilla a trav´es de las concesiones medi˜ evales de indulgencias,’ Revista Espanola de Derecho Canonico 46 (1989), 129–78, docs 2 & 3. 44. For instance, the record of the mensual commemoration of Fernando’s death in the cathedral’s ‘Curso de los Aniversarios y dotaciones que ha la eglesia de Seville,’ (Archivo ´ V, de la Catedral de Sevilla [ACS], Seccion Patronatos, libro 1, fol. 42v), or the copy of his privileges in ‘Tumbo A’ (ACS, Sec. II, Lib. 1474, fols. 1r–3r). For discussion of sim´ see Mauricio Herrero ilar parallels in Leon,

Remembering Fernando

45.

46.

47.

48.

49.

Jim´enez, Colecci´on documental del archivo de la ´ Catedral de Le´on: Obituarios medievales (Leon, 1994), 47–8; 55–6; Ana Su´arez Gonz´alez, ‘Del pergamino a la piedra? De la piedra al pergamino? (Entre diplomas, obituarios y epi´ tafios medievales de San Isidoro de Leon),’ Anuario de Estudios Medievales 33, no. 1 (2003), 365–415, 380–1; 412. Niels Rasmussen, ‘The Liturgy at SaintDenis: A preliminary study,’ in Abbot Suger and Saint-Denis, ed. Paula Gerson (New York, 1987), 41–7, 44. See Roc´ıo S´anchez Ameijeiras, ‘El “cementerio real,” de Alfonso VIII en Las Huelgas de Burgos,’ in Cultura, poder y mecenazgo, ed. Alfredo Vigo Trasancos (Santiago de Com´ postela, 1998); Raquel Alonso Alvarez, ‘La cabecera de las iglesias cistercienses femeninas en la Corona de Castilla: clausura, cura ´ aristocr´atica y monialum y representacion regia,’ Hortus Artium Medievalium 15, no. 2 (2009), 341–54. Nicasio Salvador Miguel, ‘La actividad literaria en la corte de Fernando III,’ in Sevilla 1248, ed. Gonz´alez Jim´enez, 685–99, 697–8. For earlier versions of royal funereal laments, see Higinio Angl`es, El C´odex Musical de Las ´ Huelgas. Musica a veus dels segles xiii–xiv, 3 vols. (Barcelona, 1931), vol. 1, nos 169 & 172. ´ Luis Fern´andez Gallardo, ‘La Cronica particular de San Fernando: sobre los or´ıgenes ´ de la cronica real castellana, I. Aspectos formales,’ Cahiers d’´etudes hispaniques m´edi´evales (2009), 245–65, 254–64. For the sermon, see Catal´an and Men´endez Pidal, Primera cr´onica general, ch. 1134. See Robert Folger, ‘A Genealogy of Castilian Historiography: from ´ Nomina regum to Semblanzas,’ La coronica 32, no. 3 (2004), 49–68, for discussion of analogies between later biographies and funerary monuments. Gonz´alez Jim´enez, Diplomatario, 559. See Martha Schaffer, ‘Marginal Notes in the Toledo Manuscript of Alfonso el Sabio’s Cantigas de Santa Maria: observations on composition, compilation, and performance,’ Bulletin of the Cantigueiros de Santa Maria 7 (1995), 65–84, 69; Kirstin Kennedy, ‘Alfonso’s Miraculous Book: patronage, politics, and performance in the cantigas de Santa Maria,’ in The Appearances of Medieval Rituals: the play of construction and modification, ed. Nils Holger Petersen (Turnhout, 2004), 199– 212; Laura Fern´andez Fern´andez, ‘Cantigas de Santa Mar´ıa: fortuna de sus manuscritos,’ Alcanate. Revista de estudios alfons´ıes 6 (2008– 9), 323–48, esp. 344–5.

185

50. See S´anchez Ameijeiras, ‘La fortuna Sevillana’. 51. Francisco Prado-Vilar, ‘The Gothic Anamorphic Gaze: regarding the worth of others,’ in Under the Influence, eds. Robinson and Rouhi, 67–100, 74. 52. See Georges Martin, ‘Alphonse X ou la science politique (Sept´enaire, 1–11),’ Cahiers de linguistique hispanique m´edi´evale (1993), 79– 100. 53. For instance, ‘Reciedumbre de voluntad y de obras para quebrantar a los enemigos de Dios y a los malhechores’: Alfonso X, Setenario, ed. Kenneth H. Vanderford (Barcelona, 1984), 8–9; Georges Martin, ‘Datation du Sept´enaire: rappels et nouvelles consid´erations,’ Cahiers de linguistique hispanique m´edi´evale (2001), 325–42, 329. 54. Compare cantiga 56: Kulp-Hill, Songs, 73. 55. Stephen Barney, ed., The Etymologies of Isidore of Seville (Cambridge, 2006), 1.3.2, cited in Mary Carruthers, The Book of Memory: a study of memory in medieval culture, 2nd ed. (Cambridge, 2008), 139. For thirteenthcentury echoes of Isidore’s formulation, see Luis F. Gallardo, ‘De Lucas de Tuy a Alfonso el Sabio: idea de la historia y proyecto historiogr´afico,’ Revista de Po´etica Medieval 12 (2004), 53–119, 67-69; Catal´an and Men´endez Pidal, Primera cr´onica general, vol. 1, 3–4. 56. Hugo. Bizzarri, ‘Sermones y espejos de pr´ıncipes castellanos’, Anuario de Estudios Medievales, 42/1 (2012), 168; John Walsh, El libro de los doze sabios, o tractado de la nobleza y lealtad, ca. 1237. Estudio y edici´on (Madrid, 1975), 23–33. 57. Walsh, El libro, 117–18. 58. See Franc¸ois Foronda, ‘Sociedad pol´ıtica, propaganda mon´arquica y ‘regimen’, en la Castilla del siglo XIII. En torno al “Libro de los doze sabios,”’ Edad Media: revista de historia 7 (2005–6), 13–36; Barry Taylor, ‘Libro de los doce sabios,’ in Diccionario ˜ filol´ogico de literatura medieval espanola: textos y transmisi´on, eds. Carlos Alvar and Jos´e Manuel Luc´ıa Meg´ıas (Madrid, 2002), 812– 14. 59. Evelyn Procter, Alfonso X of Castile: patron of literature and learning (Oxford, 1951); Kirstin Kennedy, ‘The Sabio-Topos: prologues of Alfonso X in the context of his thirteenthcentury royal contemporaries,’ in Proceedings of the Ninth Colloquium, eds. Andy Beresford and Alan D. Deyermond (London, 2000), 175–90; Manuel Rodr´ıguez de la Pisa, ‘El

186

Tom Nickson

paradigma de los reyes sabios en el De rebus hispaniae de Rodrigo Jim´enez de Rada,’ in Sevilla 1248, ed. Gonz´alez Jim´enez, 757–65; Maribel Fierro, ‘Alfonso X “The Wise”: The last Almohad caliph?,’ Medieval Encounters 15 (2009), 175–98. 60. See Peter Linehan, History and the Historians of Medieval Spain (Oxford, 1993), 328, n.50, for the renowned multilingualism of Rodrigo Jim´enez de Rada, archbishop of Toledo (d. 1247). 61. Jeremy Johns, Arabic Administration in Norman Sicily: the royal d¯ıw¯an (Cambridge, 2002); Jeremy Johns and Nadia Jamil, ‘Signs of the Times: Arabic signatures as a measure of acculturation in Norman Sicily,’ Muqarnas

21 (2004), 181–92; and Chapter 6 in this volume. 62. See Cynthia L. Robinson, ‘Review of “The Arts of Intimacy,’” Art Bulletin 91.3 (2009), 369. 63. Compare Kirstin Kennedy, ‘Influence and Power: Arabic iconography in Alfonso X’s Book of Chess,’ in Under the Influence. The concept of influence and the study of illuminated manuscripts, eds. John Lowden and Alixe Bovey (Turnhout, 2007), 89–96, 91; Mar´ıa J. Feliciano, ‘Muslim Shrouds for Christian Kings? A reassessment of andalusi textiles in thirteenth-century Castilian life and ritual,’ in Under the Influence: questioning the comparative, eds. Robinson and Rouhi, 101–32.

鵻 CHAPTER NINE

DISPLAYING THE WORD: WORDS AS VISUAL SIGNS IN THE ARMENIAN ARCHITECTURAL DECORATION OF THE MONASTERY OF NORAVANK‘ (14TH CENTURY) 鵼 Ioanna Rapti

From its creation in the fifth century CE the Armenian alphabet was sanctified. The vita of Mesrop Maˇstoc‘, the inventor of the alphabet, makes of him a new Moses, receiving the signs of writing as Moses received the tablets of the law. Consequently the alphabet appears as a gift of divine providence.1 The Armenian Church celebrates the memory of Mesrop Maˇstoc‘ along with St Sahak, the Patriarch who commissioned the alphabet, and the Holy Translators, the disciples of Maˇstoc‘ who, as soon the alphabet was set, launched the translation of the Bible and liturgical texts. In contrast with the high level of literacy in the Roman world, writing erupts into Armenia as a definition of linguistic and religious identity with the script being itself a system of sacred signs. Similar developments occurred at the same time in neighbouring Georgia and Caucasian Albania and, eventually, in the Balkans and Russia following the conversion of the Slavs.2 In addition to the magical character of its letters, the Armenian script became a component of artistic expression, and inscriptions constitute the safest criterion for the ascription to the Armenian realm of heterogeneous works of art in various media. As first-hand documents, Armenian inscriptions have been carefully collected,3 classified according to geographical and chronological criteria and studied as texts and sources; in contrast, the interest in scribal forms has been limited to the traditional methodology of epigraphy, seeking to date writings and objects.4 Armenian inscriptions did not enter art-historical discourse until Timothy Greenwood published a corpus of early Armenian inscriptions.5 His introductory remarks raised the question of the meaning of these architectural inscriptions beyond their content and pointed out the significance of their placement and display, which often matched epigraphic practices of the contemporary early Byzantine world. A few years earlier, Jean-Pierre Mah´e’s reediting of the 187

188

Ioanna Rapti

eleventh-century inscription on the western door of the cathedral of Ani shed new light on the administration of the city under the Byzantine emperor Constantine X (1059–67) and showed the specific function of the inscription as a public legal text.6 Thus the cathedral had a role as the civic centre, the ‘forum’ of the city, in addition to its primary religious function. Along with the slightly earlier inscription by the governor Aaron (1055–6) on the same wall, this document, posted at a focal point of public space, belongs to a rarely extant urban context where inscriptions seem to function as indicators of space and as a means to shape it. The question is still open as to why and how this practice of publication rooted in the Roman world was revived in Ani and what its reception was. Although publication of legal documents remains exceptional in urban public contexts, inscriptions combining legal arrangements with pious offerings abound in the wall archives of Armenian monastic churches. They are laid out with a similar sense of space: all but haphazard, their setting seems to be governed by content, while place, layout, legibility and visibility establish further relations between these inscribed statements and contribute to the organisation of the space. Such a sense of organisation can be better and clearly observed in gawit‘s, the equivalent of Byzantine narthexes, which may be seen as the public or social space of the monastic environment, alongside the sacred space of the church and chapels and the working rooms of libraries, kitchens or workshops.7 The insights these gawit‘s provide into the visual value of public and monastic inscriptions rely solely upon internal evidence, as historical accounts are silent about them, and calligraphy seems mainly relevant to the art of the book without an elaborated theoretical framework. The simplicity of the architectural school of Bagratid Ani is distinguished by elegant clarity in the setting and the writing of the inscriptions in harmonious and clear lettering with almost no ligatures. The text is sometimes set in columns under arched frames that, unlike earlier frieze inscriptions, recall manuscript decorative pages such as those of the Eusebian letter and the canon tables in Gospel books. Now expanded onto a monumental scale, the layout of decoratively written pages underlines the form of the inscriptions, stressing the height of the building. In the cathedral of Ani, the inscription of Queen Katranid¯e celebrating the completion of the building in 1001 stands at eye level framed by arches; the building lacks further exterior decoration, apart from a band of floral ornament around the windows and doors, and two lions flanking the porch. The inscription dominated the wall and conveyed the memory of the patron, embellishing the building in the way of a donor portrait.8 A similar display is found in the nearby monastery of Marmaˇse¯ n (dated 1029), built by the prince Vahram Paxlawuni, with the founder’s inscription in the middle of the south fac¸ade of the main church where its careful and deep carving catches the abundant rising light (Fig. 55).9 Established as a characteristic of Bagratid aristocratic architecture, this display of inscriptions was interestingly emulated in architecture in the area in the thirteenth century, but did not spread further.

Displaying the Word

55. Marmaˇse¯ n, south fac¸ade, detail, founder’s inscription. (Photo: Ioanna Rapti)

Apart from Bagratid architecture and the aforementioned inscriptions of archival character, the foundation or dedicatory Armenian inscriptions are associated with the main doors of churches. Inscriptions, which often run around the buildings of the early period, seem to be transferred in medieval monuments to the focal point of the door. They circle the door frame or fill in the tympanum above. This shift from the earlier circular inscriptions was probably a simpler solution in terms of time and work while conveying similar apotropaic

189

190

Ioanna Rapti

and protective functions, the door being a synecdoche for the whole church. Inscriptions become an essential decorative element of the buildings, and sometimes the only one. Architectural inscriptions counterbalance the scarcity of exterior decoration; their prominent presence was perhaps a response to the specific reservation of the Armenian Church and audience towards religious images. Without being formally formulated and doctrinally established this attitude was forged during a period of severe ordeals when the Armenian Church had to redefine itself against the radical heresy of aniconic trends and Arab political dominion. Catholicos ¯ John of Oǰun emphasized the sanctification of the materials while the wider cultural context favoured a visual language based on script and ornament.10 The sacred character of the script, assumed without interruption since the invention of the alphabet, may have found a significant parallel and possible model in the primary role awarded to inscriptions in the aesthetics and architecture of Islamic art. Religious and even devotional images did not disappear, but seem to be concealed in the intimacy of the enclosed space of manuscripts and particularly the Gospels. Despite some notable exceptions in the tenth and eleventh centuries, images reappear in buildings and cross-stones (khatchkars) from the thirteenth century on and then are in constant but moderate use. They do not replace inscribed decoration, but most often interact with it within the same architectural space or decorative compositions. This chapter focuses on an outstanding monument of that period, the monastery of Noravank‘, and particularly on the decoration of the fac¸ade of its gawit‘ (Fig. 56). Its two sculpted tympana set in two superimposed levels display a sophisticated decorative program with unique iconographic and epigraphic features. In this monument, the layout and rendering of the inscriptions present unparalleled variety and originality, which, I argue, seem to be a deliberate choice as part of the overall composition rather than experimentation in the management of the decorative space. Noravank‘ is among the highlights of late medieval Armenian architecture, included in surveys of Armenian art and cultural itineraries across Armenia.11 However, the aesthetic function of its inscriptions has been overlooked, apart from cursory references to the irregularity of those in the upper tympanum, which has been charged to the sculptor’s inability to balance the composition.12 Through the analysis of the iconographic and epigraphic elements of this decoration, this chapter addresses the inscriptions as visual, artistic and semantic components of the tympana, investigating the reasons and the effects of their display, as well as their possible reception. The Noravank‘ tympana offer a compelling case study of the interplay between text and image in terms not only of content but also, and particularly, of form. Thus they are of particular relevance to a perpetual art-historical issue, the relation between text and image, and they invite a challenge to the verbal approach of medieval imagery through the visual apprehension of written words that they seem to imply.13

Displaying the Word

56. Noravank‘, western fac¸ade of the gawit‘. (Photo: Ioanna Rapti)

Standing on a ledge on the side of a canyon and overlooking the plain of Ełeknajor, the monastery of Amał u Noravank‘ was erected over the course of the thirteenth century in the lands of the princely family of Orbelian.14 It is a harsh location providing natural protection, in addition to the thick enclosure around the estate. Several members of the family were patrons for the building and the decoration of the complex with its churches and chapels, and many were buried there.15 Step‘anos Orbelian, bishop of Siwnik‘ and renowned historian,

191

192

Ioanna Rapti

documents in detail the growth of the monastery, which was to become the see of the bishopric, in the late thirteenth century.16 In addition to its paramount importance in ecclesiastical administration, the monastery seems to be amongst the major architectural achievements of its time and one of the most accomplished representatives of monastic architecture in that area. The ecclesiastical core of the complex comprises three adjacent churches, and in 1339 the doublefloored funerary church was built in the eastern part of the site. The main church, the one in the middle, erected in 1216–23, was expanded by a large rectangular hall, the gawit‘ or ǰamatun, a hall typical in Armenian monastic architecture with multiple possible functions, the main of them being to house commemorative monuments and prestigious burials. The construction of the gawit‘ was certainly part of the initial architectural project, but it seems to have been accomplished in two stages as revealed by the inconsistency between the upper and lower part in the interior.17 The date of the completion and subsequently that of the sculpted tympana is debated. The dating of the inscription on the lower tympanum as 1321 that Sedrak Barxudaryan assumed in the Corpus of Armenian Inscriptions18 has convincingly been shown to be wrong,19 raising again the questions of the date of the sculptures and their attribution. Whether the tympana date to the last quarter of the thirteenth century20 or to the first half of the fourteenth century,21 there is no doubt that they belong to a wellestablished artistic trend that stems from the area of the Vayoc‘ Jor and spreads across the entire province of Siwnik‘ in a series of aristocratic monastic churches, patronized by local princes. The tympana of the gawit‘ of Noravank‘ present a series of features shared in common with works sculpted elsewhere in the region, the most characteristic being the theme of the Virgin with the Child treated with variations on several fac¸ades, which all refer to the Incarnation and exploit various models of devotional images. However, both tympana are distinguished by the unique complexity of their organisation. None of the related works gather so many compositional elements and iconographic components in such accumulative ensembles. Although inscriptions with elegant rendering and theological sensitivity are used in most of the sculptures produced in Siwnik‘, they all lack the sophisticated effects of the two tympana of the gawit‘ at Noravank‘ and their variety of formal epigraphic devices. The lower relief shows the Virgin frontal (Fig. 57), heels together but knees apart, holding in her left arm the Christ Child. He sits on her knee, which is covered by a richly ornate textile. Following the type of the Hodigitria, she points to the child with her right hand while a dove spreads its wings on the top of her head. The image finds its closest parallel in the tympanum of the church in the neighbouring village of Areni, dated to 1321, with which it shares an almost identical flattened throne, hidden by a thick decorative textile. As at Noravank‘ these throne covers are filled with an ornamental pattern of angled lines that form polygonal motifs on the field and an undulating chain on the border.22 The lower

Displaying the Word

57. Noravank‘, western fac¸ade, lower tympanum, with the Virgin and the Child. (Photo: Ioanna Rapti)

fringe is adorned with a crocheted selvage of alternating pomegranates and fleurde-lis.23 Such decorative textiles are among the numerous themes that Armenian art shares with the Islamic world, particularly after the Mongol conquest.24 The distinctive features of the Noravank‘ panel are the higher relief of the main figures and their more lively attitudes, which prefigure the corporeality of the Virgin of Tenderness on the tympanum of Spitakawor, carved in 1320. More characteristic is the foliate background in which appear two figures of prophets and related inscriptions. The prophets, unlike the traditional iconography of full- or halfstanding figures facing the viewer or turning in three-quarter position, are here depicted in bust and strict profile.25 The one on the right grips a scroll firmly in his clenched fist and unfolds it; the lower part of the scroll bears his name ԵՍԱՅԻ (Esayi, Isaiah). The middle part of the phylactery bears the word ԿՈՅՍ (koys, Virgin) (Fig. 58) in a very legible script, but surprisingly running from right to left. Thanks to the clarity of the script, the text is so easy to read that the peculiar letter order of the word may not even be noticed. The inversion of the word might be the effect of an attempt to render the transparency of the scroll’s verso as it twists. Inverted script is rather exceptional in Armenian art, and twisting scrolls are rare too.26 In some examples of the latter encountered in Cilician manuscripts, the text is displayed horizontally, but the reading does not require any special effort because the excerpt is only rotated.27

193

194

Ioanna Rapti

The left part of the background is inhabited by another prophetic bust also in profile. The rounded edge of the letter Ձ touches his lips and seems to escape from his mouth.28 The reading of the word it introduces is uncertain. Barxudaryan ignored it in his edition,29 whereas Avagyan read two letters ՁԵ and proposed to rectify the reading as ՉԵ, which according to him could be the date of the sculpture, 1251 (a date obviously too early and one that he did not defend further).30 Donabedian tentatively suggested the reading of the word ՁԱՅՆ (jayn, ‘voice’).31 Indeed the shape of the second letter clearly seen against the foliage is closer to that of the letter as seen in the incised inscription on the upper frame, with a clear example just above: a shorter second stroke on which the shaft is attached by a loop. The third letter can be identified behind the prophet’s back. It is composed of two loose curls, looking like a 3. It may be an inverted Յ, which fits better with the setting of the ornament because the roundels of the letter follow harmoniously the frame of the tympanum, thereby enabling a satisfactory reading for the letters above. In this case the word ‘voice’ (ՁԱՅՆ, jayn) ends with the centrally placed letter Ն (N). The remaining part of the background is domin˙ each written progressively smaller. One ated by three other letters ԲԱՌ (BAR), reads from the lower left edge to the middle of the field where the letter Ն (N) common to the previous word completes them as an article and allows the reading ‘ba˙rn’ meaning ‘the word’, ‘the logos’, possibly to be translated as ‘this is the Logos’. The letters on the left part of the composition are not supposed to transcribe a specific verse or excerpt and do not form a sentence together.32 The peculiar display of the inscriptions in the field has not really been discussed. Barxudaryan saw two inscriptions in the tympanum: one running round the semicircular edge, the other combining the words in the field of the tympanum with those along its lower edge. He edited these to read ԲԱՌՆ [ԽՐ]ԱՏ ԵՒ ՈԶ ՎՃԻՌ. ԵՍԱՅ[Ի] (ba˙rn [xr]at ew oˇc‘ vˇci˙r Esayi, ‘The word of prediction and not decision of Isaiah’). However, this transcription ignores the place of the words and the variety of the scripts, forging a continuous reading, which seems unlikely.33 Barxudaryan’s reading was challenged in 1975 by Suren Avagyan, who transcribed the complete text running around the tympanum as one inscription and rightly read the two last letters on the semicircle separately as parts of two different words: ԱՅՍ Է ԱՌ ԻՄ ԱՒՐՀՆԵԱԼ ԵՒ ԱՀԵՂ ԱՆՈԻՆՆ ԱԾ Ի ԾԱԳԱՑ ՄԻՆՋ Ի ԾԱԳՍ ԾԱԳԻՆ ՈՐ ՈՉ ՀԱՏ ՈՉ ՎՃԱՐ (Ays e¯ a˙r im awrhneal ew aheł anownn ac i cagac‘ minǰ i cags cagin

or oˇc‘ hat oˇc‘ vˇcar: ‘This is my blessed and fearful name of God from the ends to the ends of the end who (is) without seed and without compensation’).34 He did not notice the punctuation, which appears clearly after the third word on the left part of the arched frame. Thus the reading should be amended as: ԱՅՍ Է ԱՌ ԻՄ։ ԱՒՐՀՆԵԱԼ Է ԱՀԵՂ ԱՆՈԻՆՆ . . . ‘This is in my stead. Blessed is the fearful name of God from the ends to the ends . . . ’. Avagyan rightly disconnected the reading of the floriated word ԲԱՌՆ (ba˙rn, word, logos)35 in the field from the rest of the inscription and pointed out the inverted spelling of the word ‘Virgin’,

Displaying the Word

58. Noravank‘, western fac¸ade, lower tympanum, with the Virgin and the Child. (Photo: Ioanna Rapti)

deciphering, less convincingly, the previous letters as ԱՅ: this does not make sense and only the first letter is clearly visible here. A possible reading might be ԱՅՍ (ays, this) or, more likely, reading the letter Հ at the folding of the scroll, [Ա]ՀԱ (aha, behold), which matches Isaiah’s prophecy of the virginal birth (Isa. 7:14). The difference of scale, along with the shape and the curving of the letters, makes it much easier to decipher the word ‘the Word’ than the smaller word ‘voice’ above. The vegetal ornament throughout and the letters are cut out in a plane surface and recall champlev´e sculpture and metalwork. In contrast, the design of the letters with a slight thickening of the upright strokes, the floriated terminal at the curving of the Բ and the waving hook of the Ն betray conspicuous inspiration from the art of the pen and moreover from the Islamic realm. Unlike kuficesque decorative scripts used in non-Muslim art, the script is here perfectly legible. Another inscription of a quite similar aspect, mingling ornate letters with abundant ligatures and vegetation, is encountered elsewhere in Noravank‘, in the funerary church erected in 1339 by Bułt‘el (Fig. 59). The inscription on a block, now in the interior above the door of the upper floor,36 records the name and the title of the founder mingled in abundant foliage within a cartouche: ԲՈՒՂԹԵԼ ԻՇԽԱՆԱՑ ԻՇԽԱՆ (Bułt‘el Iˇsxanac‘ Iˇsxan: ‘Bułt‘el, Prince of Princes’). In this case the choice of this style of written form may be easily explained as an expression of secular aristocratic identity, as in the adoption of the Mongol attire and iconography in princely images.37 However, the similarity between these two

195

196

Ioanna Rapti

kuficesque inscriptions is limited to their common inspiration; their execution does not support a closer connection and, consequently, a later date for the reliefs of the gawit‘. As an aesthetic device, the use of kuficesque script is not surprising in the context of the monument: the architecture is based upon a tradition that has appropriated muqarnas vaults and used abundantly interlaced friezes and arabesques. In the centre of the composition, the throne of the Virgin has a cover with just such a diamond-grid ornament. The integration of Armenia into the Mongol empire where the local elites quickly ensured their position and status prompted the development of the Armenian decorative language and informed the stylistic choices of patrons and artists. The province of Siwnik‘ provides a significant number of Arabic and bilingual Armenian-Arabic inscriptions from this period, although most are encountered in secular contexts.38 However, in spite of the close connections between the Arabic-writing Muslim world and Armenia, which are mirrored in visual arts from the early tenth century, the use of Arabic-looking lettering as ornament is notably rare. Kuficesque inscriptions are almost entirely absent from Armenian architecture. In manuscripts, this type of epigraphic decoration is seldom integrated into either iconographic or stylistic contexts: the eleventh-century Mułni Gospels displays a pseudo-kufic ornament along with Nilotic iconography within a programme combining archaic elements and medieval innovations.39 In 1232, the painter Grigor abundantly exploited an Arabicinspired script for both legible and nonlegible inscriptions in the canon tables and frames of other images in the so-called Gospels of the Translators.40 Nevertheless these works do not represent a constant trend in epigraphic decoration as is seen in the series of tenth- and eleventh-century churches of the Greek peninsula in Byzantium.41 The sparing use of this specific type of decoration, which may have been more commonly adopted in secular objects and objects of fine art now lost, sharply contrasts with the numerous borrowings from Muslim dress and ornamentation that were deeply embedded in Armenian material culture. At the time the Noravank‘ tympanum was sculpted, oriental-looking epigraphic ornamentation was in vogue in Western Europe, particularly in Italy. From the late thirteenth century on, its use was promoted through actual and depicted luxury textiles, which often bore woven or embroidered inscriptions.42 When disconnected from dress, and especially when integrated into religious images and figured in books, these inscriptions may have recalled implicitly the linguistic reality of biblical times or pointed roughly to the linguistic otherness of the Holy Land.43 It is intriguing to wonder whether the sculptor of Noravank‘ was aware of this trend, because he was aware – perhaps through intermediary models – of other Western formulae, such as the feet of Christ nailed together, employed on the upper tympanum. But even if his inventive interpretation of the image of the Mother of God also responded to a European orientalist fashion, the use of oriental script is

Displaying the Word

59. Noravank‘, burial church of prince Bułtel, inscription. (Photo: Ioanna Rapti)

animated by a specific sensitivity and comprehension of its Islamic sources, which doubtlessly relied upon the artist’s own experience. The simultaneous use of different scripts and images in the Virgin tympanum of the gawit‘ at Noravank‘ seems to be an aesthetic and semantic choice. The prophets in bust are a usual means to stress the theological meaning of the image of the Virgin with the Child, likely inspired by painted representations. The difference in the writing used in the two parts of the composition may be itself of some significance: on the right, the medium, the unfolded scroll, alludes to the orality of the prophetic words. It also points to the fulfilment of Old Testament, but its central part ‘Virgin’ also works as a label for the enthroned Mother of God. The inverted direction of the word koys (Virgin) remains intriguing. It may indicate the hand of a sculptor whose drawing was primarily motivated by aesthetic concerns, or perhaps he was a sculptor trained in an Islamic milieu using readily available patterns from Islamic textiles or decorative arts. In either case, the inverted direction of the word testifies to a familiarity with the Islamic visual approach to both script and ornament.44 The idea of the fulfilled prophecy is also conveyed on the left by the barely legible word ‘voice’. The kuficesque writing of ‘the Word’ refers precisely to the image, and it works, in its turn, as a surprising and interpretive label for the Christ Child. At the same time, the choice of the word ‘Word’/‘Logos’ may evoke an intriguing parallel to the prophetic announcement: Logos is not a traditional

197

198

Ioanna Rapti

label in the Armenian iconography of the Mother of God or Christ. The largescale dominant transcription of the word ‘the Word’ conveys the fundamental sense of the Logos, but at the same time its kuficesque script may point to the Islamic counterpart of this concept, the revelation through the ‘Word’ the representation of which ‘offered a meaningful substitute for the traditional human figure that represents divinity in pagan and Christian religions’.45 The specific feature of the left part of the composition is that, in spite of their Arabic calligraphic style, the lettering does not follow the regular distribution of Arabic inscriptions – and their imitations – through well-defined friezes or cartouches in architecture and various objects.46 The letters of ‘the voice’ are difficult to distinguish against the floriated background, whereas those of ‘the Word’ expand upwards in a graceful movement responding to the unfolding of the scroll. The letters’ unusually free display within the field contributes to a unique dramatization of the whole tympanum. The unusual choice of these words and their unparalleled display complicate their legibility and meet, more than other inscriptions in Armenian, the Islamic use of partial legibility or nonlegibility for the nomina sacra or theological formulae, intended to be seen rather than read.47 The aesthetic of the script appears to be a means to enhance the mystery of the Incarnation concealed in the established and standardized iconography of the Hodigitria. Thus, the tympanum offers a rare and perhaps unique example of the integration of kuficesque script – visible, readable and meaningful as a component of a religious image. The upper tympanum hosts a composition unparalleled in medieval art, perhaps one of its most compelling and inventive creations (Fig. 60). The Ancient of Days, a figure with a thick beard and curly hair, dominates the surface by its large scale and high relief; his right arm and gaze are raised. In his left hand he holds a male head seen in profile, eye wide open and risen toward the face of God. A dove descending from the divine lips reaches his mouth. Another dove with folded wings, set symmetrically to the flying one, fills the space between the two parts of the beard, with its beak pointed towards God’s mouth. Without exclusive links to visual or textual narratives, the image recalls both the insufflation of life in the Creation and the hand of God holding the human soul. The identification of the Ancient of Days (ՀԻՆ ԱՒՈՒՐՑ, hin awurc‘) set on the upper frame of the tympanum under the salient outer arch is obscure even for a literate viewer. The lack of any other label for the components of the central image enhances its emblematic character and its semantic richness encompassing theophany and trinity, paternity and eschatology. It is worth pointing out the analogy with the concept of the circularity of the process of salvation, consisting of emanation, exemplarity and return.48 How far the exploitation of this complex idea, which was promoted by the Franciscans, may be indebted to the friars who were crossing this area on their way to Persia, or to the revival in the monastic schools of the Siwnik‘ of neoplatonic philosophy in which the concept originated is, however, beyond the scope of this chapter.

Displaying the Word

60. Noravank‘, western fac¸ade, upper tympanum as seen at the door of the gawit‘. (Photo: Ioanna Rapti)

The left part of the composition is filled with an image of the Crucifixion in surprisingly reduced proportions, which make it look like a miniature next to the central figure. In spite of its condensed nature, the Crucifixion is distinguished by the exceptional addition of the prophet Daniel. This identification for the figure lying on the ground below Golgotha, in the place usually occupied by the skull of Adam, is suggested by the inscription on the lower frame of the tympanum.49 Apart from Daniel, the only other figure of the Crucifixion who is identified is the Virgin, whose usual abbreviation ‘Mother of God’ (ԱԾԱԾԻՆ, Acacin) runs vertically down behind her figure on the frame of the composition. Christ’s head turns up towards the blessing hand of the Ancient of Days. He is crucified before the eyes of personifications of the moon and the sun, depicted here as male profiles instead of the more usual standardized images inscribed in medallions. The left arm of the cross encroaches upon the right arm of the Ancient of Days, cut out of the same block, showing that the juxtaposition of these components was part of the initial plan with the Crucifixion designed to fit into the available space. This necessitated the considerably reduced scale of John on the right. The dense relief of the left part of the panel is not counterbalanced on the other side, where only a bird is perched on the edge of the tympanum and by the text that fills the rest of the ground. The uneven distribution of figures and script leaves the text at the most shadowy place because of the exposure and of the salient outer arch.

199

200

Ioanna Rapti

The inscription breaks with all the unwritten standards of architectural epigraphy. The surprising irregularity of the lettering, presenting different sizes and inclinations as well as many letters that hesitate between capital and minuscule, makes it look rather like graffiti: ԸՆԴ ՍՏԵՂԾԻԼՆ ԱԴԱՄ ՆՈՐՈԳԵՑԱՒ ԵՐԿԻՆՔ ԵՒ ԵՐԿԻՐ ՈՐ ՄԻՇՏ ԱՒՐՀՆԵՆ ԶԱԾ.50 (Ac. Ǝnd stełciln Adam norogecaw erkink‘ ew erkir or miˇst awrhnen zac. ‘God. Through the shaping of Adam were restored heaven and earth so as eternally to bless God’).51 Such a translation requires that Adam – here in nominative followed by the article – be read as a genitive. Conversely, given the absence of punctuation, a different reading may also be possible: ‘God. According to the Creation Adam was renewed (restored). [These are] Heaven and earth, which eternally bless God’. The inscription does not seem to quote a specific text; its lack of syntactical unity is reminiscent of the formulation of captions accompanying images in various works and particularly in manuscripts. In any case the content is elaborate and deals with the fundamental idea of the salvation of the human race, fallen with Adam, through the sacrifice of the incarnate God, a new Adam. Analysing the inscription in terms of its layout allows a better understanding of what Jean-Michel Thierry called ‘savant d´esordre’ and Sirarpie Der Nersessian criticised as awkwardness.52 Instead of calibrating the text in ruled lines within the available space – as inscriptions in several lines are usually set – the carver or writer seems to have managed the epigraphic field so as to privilege the beginning and the end of the text, namely the abbreviation ԱԾ (God). This catches the eye thanks to its upright design and the large size and the deeper incision of the letters. When standing before the entrance of the gawit‘, before analysing the visual information, the viewer first encounters the hands of God and his name. Any viewers on the ground could hardly notice the open eye of Adam’s head and could not face God’s gaze staring far above, as indicated by the incisions of the pupils in his eyes. Some could also notice the beginning of the word ‘bless’, written above the second abbreviation of the word God, in smaller size but upright. Like the abbreviation for God, it was perhaps carved separately and before the rest of the inscription, which filled the space between with an effort to parallel the outline of the tympanum. Unlike the unfamiliar kuficesque words in the lower tympanum, the inscription of the upper tympanum was apparently not intended to be read further than the abbreviated name of God. The complexity of the lettering diverts the viewer from seeking to read and apprehend the image through the accompanying text. It only guides the mind to contemplate with fear the promise of salvation as evoked suggestively by the entire composition. The deliberate anarchy of the script translates and emphasizes the unintelligibility of the divine essence and of the economy of salvation. Set higher than six metres from the ground the tympanum was to be seen from a distance and from below. The disproportionate figure and size of the letters should be better understood as parallel visual devices to stress the complex meaning of the image and its sense of mystery.

Displaying the Word

201

The Armenian tradition adopted and developed the gnostic concept of Adam as ‘archetypal man’, which was interestingly transposed in the mystical approach of the alphabet. The parallel between the creation of the first human and the alphabet is mirrored in characteristic enigmas encountered in manuscripts and in oral tradition, such as ‘The hand of the Highest created me, the Highest released me in the high spheres’ or ‘I speak among the dead and among the inanimate I raise a voice’.53 The sensitive approach to the script as a visual element is not exclusive to the Armenian realm, but it is a major component of its artistic expression. The western fac¸ade of the gawit‘ of Noravank‘ offers the most characteristic example of how perennial and dynamic this element is. Alongside the traditional pattern of inscriptions in careful elegant script framing and overlooking the entrance to the sacred, the – most likely – anonymous sculptor of the tympana capitalizes upon the visual power of the script through its innovative and astute integration in these two images of unparalleled mastery. His surprising borrowing from the aesthetic of Islamic script for most sacred words in the lower tympanum and his audacious breaking of the unspoken epigraphic rules of order and symmetry in the upper one do not contrast with the fine theological content of images and texts; on the contrary, they emphasise it with particular effectiveness. Their contemplative and symbolic function, beyond text, conveys a special insight into the culture and mentality of the Armenian elites flourishing in the Mongol empire.

NOTES

1. James R. Russel, ‘On the Origins and Invention of the Armenian Script,’ Le Mus´eon 107 (1994), 317–33. Gabriele Winkler, Koriwns ¨ Biographie des Mesrop Maˇstoc‘. Ubersetzung und Kommentar, Orientalia Christiana Analecta (Rome, 1994) and Jean-Pierre Mah´e, ‘La Vie de Maˇstoc‘, traduction annot´ee,’ Revue des ´ Etudes Arm´eniennes 30 (2005–2007), 57–97. See also Karen Yuzbaˇsyan, ‘L’invention de l’alphabet arm´enien. De la langue parl´ee a` la ´ langue e´ crite,’ Revue des Etudes Arm´eniennes 33 (2011), 67–129. 2. Chrisrian Hannick, ‘Le d´eveloppement des langues regionales et l’introduction d’alphabets dans des communaut´es illettr´ees,’ in East and West: Modes of Communication. Proceedings of the First Plenary Conference at Merida, eds. Evangelos Chrysos and Ian Wood (Leiden, 1999), 93–110. 3. Before the corpora compiled in the twentieth century in modern Armenia and in the nineteenth century by antiquarians, the thirteenth-century historian Step‘anos Orbelian was perhaps the earliest Armenian

4.

5.

6.

7.

author to integrate inscriptions in his chronicle and use them as sources. Corpus Inscriptionum Armenicarum / Divan Hay Vimagrut‘ean, 7 vols (Erevan, 1960–). ‘Evgenia Muˇsełian, Hayer¯en arjanagrut‘yamb a˙rarkanerǝ [Objects with Armenian Inscriptions] (Erevan 1964). For an overview of the history of the Armenian script (from the cuneiform alphabets used in Armenia to the era of the computer) see the exhibition catalogue Arm´enie, la magie de l’´ecrit, ed. Claude Mutafian (Paris, 2007) with further bibliography. Timothy Greenwood, ‘A Corpus of Early Medieval Armenian Inscriptions,’ Dumbarton Oaks Papers 58 (2004), 27–91. Jean-Pierre Mah´e, ‘Ani sous Constantin X, d’apr`es une inscription de 1060,’ in M´elanges Gilbert Dagron, Travaux et M´emoires 14 (2002), 403–14. Unfortunately there is not yet any comprehensive study on the development of Armenian monastic institutions as centres of economic and social life in the period

202

8.

9.

10.

11.

12.

Ioanna Rapti

after the Arab conquest. For an informative case study see A. T. Baladian, JeanMichel Thierry and Jean-Pierre Mah´e, Le couvent de Ho˙romos d’apr`es les archives de Toros Toramanian (Paris, 2002), in particular the chapter on the inscriptions, 147–68. For the emergence of monasteries as a major development of the period after the Arab conquest of Armenia, see Jean-Pierre Mah´e, ‘Le christianisme arm´enien avant et apr`es l’Islam: rupture et continuit´e,’ in Armenia Sacra. Memoire sacr´ee des Arm´eniens, eds. Jannic Durand, Ioanna Rapti and Dorota Giovannoni (Paris, 2007), 117–19. Paolo Cuneo, Ani, Documents of Armenian architecture 12 (Milan, 1984). Nikolai Marr, Ani, rˆeve d’Arm´enie (Paris, 2001) [partial translation from Marr’s monograph in Russian] and Iosif. A. Orbeli, Divan hay vimagrut‘yan, vol. 1: Ani k‘ałak‘ (Erevan, 1960). A second contemporary inscription on the left completes the record in a similar script. For Marmaˇse¯ n see Paolo Cuneo, ‘Le Couvent de Marmaˇse¯ n et l’´ecole architec´ turale d’Ani,’ Revue des Etudes Arm´eniennes 23 (1992), 419–71 and Paolo Cuneo, Marmashen Documents of Armenian architecture 16 (Milan, 1986). Cf. also Jean-Michel Thierry and Patrick Donab´edian, Les arts arm´eniens (Paris, 1987), 167–8, 123–4, 553. Mah´e, ‘Le christianisme arm´enien avant et apr`es l’Islam,’ 116. On the debates on the use of images in Armenia see Sirarpie Der Nersessian, ‘Image Worship in Armenia and Its Opponents,’ Armenian Quarterly 1, 67– ´ 81 reprinted in her Etudes Byzantines et Arm´eniennes (Leuven, 1973), 405–15. Ioanna Rapti, ‘Le statut des images dans l’art et le culte arm´eniens,’ Actes du colloque sur l’aniconisme a` Byzance (Mus´ee d’art et d’histoire de Gen`eve, 2008), in press. Sirarpie Der Nersessian, Armenian Art (New York, 1978), figs 142–3; O. Khalpakchian, Architectural Ensembles of Armenia (Moscow, 1980), 441–5; Thierry and Donab´edian, Arts arm´eniens, 478–9. Jasmine Dum-Tragut, Armenien: 3000 Jahre Kultur zwischen West und Ost (Berlin, 2008), 345–9. In the postSoviet period, all the buildings of the Noravank‘ complex underwent heavy restorations, including rebuilding of the cupolas in the main and the burial church. Nevertheless it seems that the fac¸ade of the gawit‘ was not altered by the twentieth-century restorations. Sirarpie Der Nersessian, ‘Deux tympans sculpt´es armeniens datant de 1321’, Cahiers arch´eologiques 25 (1976), 139–22. At the

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

beginning of her study, p. 110, Der Nersessian observes without further commentary ‘l’irr´egularit´e de l’´ecriture et l’´echelle des lettres’, while later Thierry and Donab´edian, Les arts arm´eniens, 479, note ‘un savant d´esordre’. Elizabeth Sears, ‘“Reading” Images’, in Reading Medieval Images. The Art Historian and the Object, eds. Elizabeth Sears and Thelma K. Thomas (Michigan, 2002), 1–7. Adriano Alpago-Novello and Giulio Ieni, Amaghu Noravank‘, Documents of Armenian Architecture 14 (Milan, 1985), 6. Thierry and Donab´edian, Les arts arm´eniens, 478–9; Lilith Zakarian, ‘Les arts en Grande Arm´enie (XIIIe -XVe si`ecle)’, in Armenia Sacra, 326–7. Step‘anos Orbelian, Histoire de la Siounie, trans. Marie-F´elicit´e Brosset (St Petersburg, 1864), 196–9, 202–4 and 265–72. This is indicated by the fact that the ambitious main church was not provided with any specific decoration and arrangements on the fac¸ade that were to be hidden by the gawit‘, which, moreover, was the only way of accessing the church. Sedrak Barxudaryan, Divan Hay Vimagrut‘yan, vol. 3, Vayoc‘ Jor, Ełegnacor ew Azizbekovi sˇrǰanner (Erevan, 1967), 222, no. 705. The date was used in one of the arguments to ascribe both tympana to the famous Momik, an elusive artist, sculptor and painter, whose signature is encountered in some carved cross-stones (khatchkars) in the monastery and its vicinity. Thierry and Donab´edian, Les arts arm´eniens, 479. Patrick Donab´edian ‘Le khatchkar du XIIIe si`ecle aux temps modernes,’ in Armenia Sacra, 313 and Zakarian, ‘Les arts en Grande Arm´enie,’ 328. Two khatchkars dated to 1303 and 1308 have been transferred from Noravank‘ to Etchmiadzin. Suren A. Avagyan, ‘Noravank‘i gavt‘i baravori arjanagrowt‘yownǝ’ [The inscription of the lintel of the gawit‘ of Noravank‘], Lłaber fasc. 8 (1975), 106–16. This early date has been suggested by Patrick Donab´edian, ‘Les particularit´es stylistiques d’un monument sculpt´e de Noravank‘ et sa ´ datation,’ Revue des Etudes Arm´eniennes 17 (1983), 395–405. This date has been maintained by Alpago– Novello and Ieni, Noravank‘, 18. Der Nersessian, ‘Deux tympans sculpt´es,’ 122. This motif abounds in tiles and in manuscripts from the early fourteenth century: Linda Komaroff and Stefan Carboni, eds.,

Displaying the Word

23.

24. 25.

26.

27. 28.

29. 30. 31. 32.

The Legacy of Genghis Khan, Courtly Art and Culture in Western Asia, 1256–1353 (New York, 2002), cat. 63 and fig. 158, and p. 206, figs. 249 and 252. Donab´edian, ‘Les particularit´es stylistiques,’ 401–2 and 405, considers that the frame of the throne cover is rare in Armenian art and should be seen as a characteristic of the school of Vayoc‘ Jor of the late thirteenth century. Zak‘arian, ‘Les arts en Grande Arm´enie,’ 324–5. Ioanna Rapti, ‘Gloses proph´etiques sur l’´evangile: A` propos de quelques manuscrits arm´eniens enlumin´es en Cilicie dans les ann´ees 1260,’ Dumbarton Oaks Papers 58 (2004), 123–4. The only Armenian example to my knowledge consists in two (out of more than fifty) illustrations of the so-called Vehap‘a˙r Gospels, an early eleventh-century Gospelbook with abundant captions and transcriptions of the conversations between the figures. In these two instances, the petition of the Canaanite and the healing of the epileptic, the inverted legends transcribe direct addresses to Christ: not only do they run leftwards but all the letters are inverted as seen through a mirror. For the manuscript see: Thomas F. Mathews, Armenian Gospel Iconography. The Tradition of the Glajor Gospels, Dumbarton Oaks Studies 29 (Washington DC, 1993), passim and fig. 99a. Cf. also the facsimile edition of the manuscript: Artaˇse¯ s Matevossian and Tatiana Izma¨ılova, The Gospels of Vasgen Catholicos (Facsimil´e). Codicological, Paleographic and Miniature Studies (Erevan, 2000) [in Armenian with English title and summary] and Ioanna Rapti, Mots et images au service du Verbe: l´egendes, annotations et paroles dans les e´ vangiles de Vehap‘ar (XIe si`ecle), Travaux et M´emoires 18 (2014), M´elanges Jean-Pierre Mah´e, A. Mardirossian, A. Ouzounian, C. Zuckerman (eds), 547–73. Rapti, ‘Gloses,’ 123–4, figs. 4–5. Donab´edian, ‘Les particularit´es stylistiques,’ 404 and figs. 2 and 14, observes that this rare device occurs also in a category of khatchkar called amenap‘rkiˇc‘, which comprises a few examples but of great artistic quality. Barxudaryan, Divan Hay Vimagrut‘yan, vol. 3, no. 705, p. 222. Avagyan, ‘Noravank‘i gavt‘i,’ 107. Donab´edian, ‘Les particularit´es stylistiques,’ 404. Der Nersessian, ‘Deux tympans sculpt´es armeniens,’ 116, did not comment on the

203

33.

34.

35.

36. 37.

38.

39.

40. 41. 42.

43. 44.

45.

inscriptions except to note that they do not allow us to identify the prophet who, she suggested, should be Micah because of his relation to the Incarnation. There is no room on the lintel for the two letters restored by Barxudaryan, who did not justify his reading, and the discontinuity of the letters precludes a continuous reading. He also ignored the first part of the scroll and discarded ՁԱՅՆ (jayn, voice) on the left. Avagyan, ‘Noravank‘i gavt‘i,’ 106–7. Avagyan distinguished the last two letters at the right part of the lower frame of the tympanum that are today scratched. It should be noticed that the syntaxis a˙r + genitive is extremely unusual: A. Meillet Etudes de linguistique et de philologie compar´ee, 65, notes a single example in the text of the Gospels. Barxudaryan, Divan Hay Vimagrut‘yan, vol. 3, no. 783, p. 240. ¯ c‘i See for example the depiction of Prince Eaˇ Proˇsean in the reliquary he commissioned in 1300 or the hunting relief of prince Amir Hasan made for the church of Spitakawor: Armenia Sacra, cat. 143 and 130. Seta Dadoyan, The Armenians in the Medieval Islamic World (New Brunswick, 2011), 47 and A. Xaˇcatryan, Korpus arabskikh nadpisei Armenii [Arabic inscriptions in Armenia] (Erevan, 1987). T. Izmailova, ‘Le cycle des fˆetes du t´etrae´ vangile de Mugna (Matenadaran no 6201),’ ´ Revue des Etudes arm´eniennes 6 (1969), 105–39. Matenadaran 2743, in Armenia Sacra, 183, no. 70. See Chapter 5 in this volume. Inscriptions depicted on textiles were merely ornamental, but they could also be significant in conveying the role and the status of the figures clothed with those garments: Rosamond E. Mack, Bazaar to Piazza: Islamic Trade and Italian Art, 1300–1600 (Berkeley and Los Angeles, 2002), 52–4. Oleg Grabar, The Mediation of Ornament (Princeton, 1992), 117 and 127, points out the variety in the use of Arabic script beyond Islamic art that makes generalization inadequate. Mack, Bazaar to Piazza, 71. The juxtaposition of various complex kufic scripts and cursive inscriptions is common in Islamic decorative arts: Sheila Blair, Islamic inscriptions (Edinburgh, 1998), 88. Erica C. Dodd and Shereen Khairallah, The Image of the Word, A Study of Quranic Verses in Islamic Architecture, vol. 1 (Beirut, 1981), 3. There is an interesting parallel between the

204

Ioanna Rapti

Qur’an’s qualification for the prophet ummi (unlettered) and the virginity of Mary as conditions to become the messengers of divine revelation. Annemarie Schimmel, Calligraphy and Islamic culture (New York, 1984), 77. 46. Eva Baer, Islamic Ornament (New York, 1998), 60–2. 47. Grabar, The Mediation of Ornament, 117. Richard Ettinghausen, ‘Arabic Epigraphy: Communication or Symbolic Affirmation’, in Near Eastern Numismatics, Iconography, Epigraphy and History. Studies in Honour of George C. Miles, ed. Dikran Kouymjian (Beirut, 1974), 304, 307–9. 48. Amy Neff, ‘Palma dabit palmam: Franciscan Themes in a Late 13th Century Italian Devotional Manuscript’, Journal of the Warburg and Courtauld Institutes 65 (2002), 46– 51.

49. The addition of the figure of Daniel is most likely related to the commentary of Vardan A˙rewelc‘i on the Book of Daniel, which seems to have had considerable influence on contemporary eschatological imagery in Greater Armenia, cf. Lilith Zakarian, Aghdjots. Saint Stepanos (Erevan, 2007), 186–94. 50. Barxudaryan, Divan Hay Vimagrut‘yan, vol. 3, no. 705, p. 222. The abbreviations are in bold to indicate the important size of these letters. 51. Der Nersessian, ‘Deux tympans sculpt´es,’ 114. 52. See n.12. 53. Jean-Pierre Mah´e, ‘Six e´ nigmes arm´eniennes anciennes sur le mythe de l’homme primor´ dial,’ Revue des Etudes Arm´eniennes 15 (1981), 45–57, in part. p. 53 and Nikoghayos Adontz, Denys de Thrace et les commentateurs arm´eniens (Leuven 1970), LX–LXI.

鵻 CHAPTER TEN

WRITTEN IN STONE: CIVIC MEMORY AND MONUMENTAL WRITING IN THE CATHEDRAL OF SAN LORENZO IN GENOA 鵼 Stefania Gerevini

In 1296, a devastating fire started by civic dissenters irreversibly damaged the cathedral of San Lorenzo in Genoa. Shortly afterwards, in the years 1307–12, two sets of monumental inscriptions were set up along the central nave of the church, marking the completion of the works of repair on this part of the building. These epigraphs, elegantly written in the most accessible space of the church, form the subject of this chapter, which addresses the interactions between the textual and visual functioning of the inscriptions and explores the ways in which their appearance contributed to give their contents meaning and increased their significance and efficacy at a time of acute political and social tensions in Genoa. The epigraphs in San Lorenzo were primarily intended to celebrate the officers who supervised the building enterprise following the fire of 1296, as well as to praise the Genoese state and civic community, which had financed the reconstruction. However, between 1307 and 1312 the city witnessed fierce conflicts and frequent turnovers in its government, until its political representatives, exhausted by the unending civil tensions, spontaneously consigned the city to the emperorelect Henry VII in 1311. This chapter proposes that the inscriptions in San Lorenzo contributed to the normalization of these dramatic political events and fostered civic cohesion by advertising the names of the mythical founders of Genoa, publicizing the illustrious origins of the city-state and celebrating its ancient legacies. Far from being exclusively defined by their textual contents, the ways in which the epigraphs operated and conveyed meaning in medieval Genoa were significantly determined by their specific location, form and visual properties. Written in large-scale, elegant Gothic capital letters and displayed in the most accessible space of the cathedral, these inscriptions were highly visible, clearly legible 205

206

Stefania Gerevini

and undoubtedly intended to be experienced by as wide an audience as possible. Examining the placement, extension, form and pagination of these inscriptions, as well as the interactions between their textual and visual meanings, this chapter investigates how public writing could signify and engage medieval audiences beyond (or in addition to) the act of reading. The very decision of the Genoese government to record the ancient origins and recent history of the city on the walls of the cathedral is meaningful, and it testifies to the authority that inscribed text enjoyed in Genoa and to its significance in processes of historical interpretation and verification. As this chapter suggests, writing in stone was employed in San Lorenzo to (literally) consolidate the unsteady historical memory and civic identity of the city. The durable nature of engraved text and the combination of historical and mythical events in these inscriptions also raise intriguing questions regarding the relationship between real and imagined memories in Genoa and the different artistic media available for their preservation and promotion. Finally, the epigraphs called into question issues of temporality and historicity. They extended along the length of the nave, engaging beholders in a visual and temporal perambulation. They emphasised the continuity and unity between the time of the mythical foundation of Genoa, the recent restoration of the cathedral, and the instant of viewing, implicitly presenting the latter as the latest episodes – and the culmination– of an illustrious chronological and historical progression. The cathedral of San Lorenzo in Genoa has a rectangular ground plan with three semicircular apses. It is divided into a central nave and two lateral aisles preceded by an atrium. The arcade comprises nine bays and is surmounted by a false gallery, also divided into nine arches and sustained by alternated pillars and columns.1 The north and south walls of the nave carry lengthy Latin inscriptions in Gothic majuscule.2 The north wall features two texts (Figs. 61 & 62). The first is a dedicatory inscription, engraved on the banded surface that separates the lower and upper storeys of the nave (Fig. 62). It consists of a single line of text, and it extends over three bays, starting above the fifth column and ending above the eighth. The text runs just below the gallery cornice and is neatly carved along a band of white marble. Each letter is filled with black pigment. The epigraph is preceded and followed by the Genoese coat of arms, a red cross on white ground, and reads: MCCCVII ♦♦PASTONUS DE NIGRO ♦ ET NICOLAUS DE GOANO ♦♦ FECERUNT RENOVARI HOC OPUS ♦ DE DECENO LEGATORUM ♦ 1307. Pastonus de Nigro and Nicholas de Goano had this structure renovated with the legates’ tithe.3

The north wall also carries another inscription (Figs. 62 & 63). It occupies the lower half of the banded pillar that separates the fifth and sixth bays of the false

Written in Stone

61. Genoa, Cathedral of San Lorenzo. View of the central nave. (Photo: Stefania Gerevini)

gallery. Set above the dedicatory epigraph of 1307, this inscription is composed of six lines. Each couplet is arranged over one of three horizontal bands: the first and last two lines are engraved on black slate, whereas the central couplet is carved on a slab of white marble. The characters are significantly smaller

207

62. Genoa, San Lorenzo. View of north wall with dedicatory and foundation inscriptions and bust portrait of Janus, c. 1307. (Photo: Stefania Gerevini)

63. Genoa, Cathedral of San Lorenzo. North wall, gallery, detail of the foundation inscription and sculpted effigy of Janus, c. 1307. (Photo: Stefania Gerevini)

208

Written in Stone

64. Genoa, Cathedral of San Lorenzo. South wall, detail of the junction between dedicatory and foundation inscription, c. 1312. (Photo: Stefania Gerevini)

(about half the size) than those of the dedicatory epigraph described earlier, but the script is identical in the two cases (only the shape of the letter ‘E’ differs slightly). The inscription, which features several abbreviations, runs below the sculpted bust-effigy of a king placed on a foliated corbel and indicates the identity of the figure: JAN[US] P[RI]M[US] REX ITALIE ♦ DE P[RO] GENIE GIGAN TIU[M] ♦ Q[UI] FU[N]DA VIT IAN[UAM] ♦ T[EM]P[O]R[E] ABR[A]HE ♦4 [This is] Janus, first king of Italy, of the offspring of Giants, who founded Genoa at the time of Abraham

The foundation text works in conjunction with the sculpted portrait located above it. Together, they evoke the myth of foundation of the town of Genoa and its legendary founder, Janus, recently reclaimed by thirteenth-century Genoese writers.5 The inscription, although undated, was almost certainly engraved at the same time as the dedicatory epigraph on the cornice, because this area of the wall would only be accessible with the aid of the high scaffolding put in place for the restoration of the arcade and gallery. The south wall of the nave of San Lorenzo also carries extensive monumental writing. Two consecutive inscriptions cover the entire length of the nave (Figs. 61, 64–66). Engraved in Gothic capital letters that match those on the

209

210

Stefania Gerevini

65. Genoa, Cathedral of San Lorenzo. South wall, central section of the foundation inscription, c. 1312. (Photo: Stefania Gerevini)

north wall, they read from east to west. The text fills a band of white marble below the cornice that divides the gallery from the lower arcade, thus occupying the same position as the dedicatory inscription on the opposite wall. The inscription at the east end of the nave echoes the phrasing and meaning of the dedication of 1307. It commemorates the two officers who supervised the works of restoration of the south arcade, completed in 1312:

Written in Stone

66. Genoa, Cathedral of San Lorenzo. South wall, western section of the foundation inscription, c. 1312. (Photo: Stefania Gerevini)

MCCCXII FILIPP[US] D[E] NIGRO [ET] NICOLAU[S] D[E] GOANO REPARATORES HUI[US] ECCL[ES]IE FECER[UN]T RENOVA[R]I HOC OP[US] D[E] D[E]CENO LEGATO[RUM] 1312. Philip de Nigro and Nicholas de Goano, repairers of this church, had this structure renovated with the legates’ tithe.6

211

212

Stefania Gerevini

After a punctuation mark, the inscription continues along the same line with a summary of the origins and early history of Genoa, longer and more informative than that on the north wall: JAN[US] PRI[N]CEPS TROIAN[US] ASTROLOGIA PERIT[US] NAVIGA[N]DO AD HABITA[N]DU[M] LOCU[M] QUERE[N]S SANU[M] D[OMI]NABILE[M] [ET] SECURU[M] IANUA[M] IA[M] FU[N]DATA[M] A IANO REGE YTALIE P[RO]NEPOTE NOE VENIT ET EAM CERNE[N]S MARE ET MO[N]TIB[US] TUTISSIMA[M] A[M]PLIAVIT NO[M]I[N]E ET POSSE. Janus, prince of Troy, experienced in astrology, sailing in search of a place to live that would be healthy, safe and easy to rule, came to Genoa, which had already been founded by Janus king of Italy and great grandchild of Noah, and, judging that it was very well protected by the sea and the mountains, he increased its fame and power.7

The epigraphs along the nave of San Lorenzo were evidently intended to perform a public function. They were engraved in large, well-spaced and clearly legible capital characters; contained few abbreviations; and were displayed in the most accessible space of the most important religious and civic building of the city. Furthermore, they were written in Latin, which, as has been suggested, was extremely tenacious in Genoa. Unlike other Italian city-states, where vernacular began to replace Latin in private and public writings at the end of the Duecento, Genoa made exclusive use of Latin for the recording of contracts and wills, as well as of all public acts, confining the vernacular to poetry and private letters.8 We do not possess precise estimates of the degree of literacy of the local population. However, the pervasive occurrence of Latin in the city, the association of this language with the local legal and notarial tradition, and the widespread use of Latin as the international merchants’ language suggest that the monumental inscriptions in San Lorenzo – commissioned by the public authorities of Genoa, written in Latin and elegantly engraved in Gothic capital script – participated in the local tradition of official and administrative writing and were devised as authoritative means of public communication. The monumental inscriptions carved in San Lorenzo in 1307–12 were not an isolated phenomenon. An extraordinary number of inscriptions of different contents, size, length and purpose were set up between the twelfth and the fourteenth century by the political and ecclesiastical authorities of Genoa as well as by private patrons on the inner and outer surfaces of city walls, religious edifices and secular buildings to fulfil a wide range of communicative and celebratory functions.9 Furthermore, the sheer quantity of epigraphic material that has been excavated in Crimea (controlled by the Genoese in the later middle ages) and on the Genoese walls of Galata in Constantinople, further testifies to the Genoese proclivity for public writing.10 A monumental inscription was even set up in 1106–9 by King Baldwin I in the church of the Holy Sepulchre

Written in Stone

in Jerusalem, presumably upon request of the Genoese, to record the commercial privileges granted to Genoa as a reward for its support in the conquest of Acre, Caesarea and Arsuf.11 These examples powerfully demonstrate that monumental writing was a favoured means of communication and certification in medieval Genoa. More specifically, they confirm that the epigraphs carved along the nave of San Lorenzo – the most impressive extant epigraphic endeavour of medieval Genoa and the only surviving engraved texts commemorating the origins and early history of the city – were conceived as exceptionally significant public announcements.12 Just what messages did they convey, and how did they operate visually? Initially, the inscriptions along the nave of San Lorenzo may be understood as memorials of a remarkable public enterprise. The expression hoc opus (lit. ‘this work’) that occurs in both dedicatory inscriptions refers specifically to the works of repair carried out on the left and right arcades and galleries. As mentioned earlier, the columns, burnt by a fire in 1296, were gradually replaced with new shafts, without dismantling the surrounding Romanesque wall.13 Pastonus de Nigro, Philippus de Nigro and Nicholas De Goano, the reparatores (repairers) honoured in the dedicatory inscriptions on either side, were public officers appointed by the Commune to supervise and manage the restoration of the cathedral. Their office was established in 1297, following the great fire that had devastated San Lorenzo, and is attested until 1317, testifying to the long duration of the reconstruction. The restorers of the cathedral were appointed in pairs, received no salary for their work, and remained on duty for six months, at the end of which they could either be replaced or reappointed.14 The fact that Nicholas of Goano, commemorated in 1307, was still on duty in 1312, when the epigraph along the right wall of the cathedral nave was carved, suggests that some officers were reappointed for years, perhaps to ensure the continuity of the building project. The inscriptions meticulously indicate that the works of repair were funded with public money, namely with the deceno legatorum – the legates’ tithe. This was a consolidated tradition of the Genoese Commune: according to a law issued by the Commune in 1174, 10 per cent of all pious legates bequeathed by Genoese citizens to any institution should be destined to the cathedral of San Lorenzo.15 In the years 1270–85, when building work in the cathedral had reached a halt, the Commune diverted those funds to the construction of the harbour, the other major public enterprise of those years.16 However, after the fire of 1296, the tithe was in large part redirected towards the cathedral, and, as the dedicatory inscriptions suggests, became a major financial source for its repairs.17 The two texts were engraved in large, neat characters along the nave, between the two rows of newly installed columns. Their very existence on the recently restored walls was clearly intended as a ‘signature’: they revealed the site of public intervention, advertised the successful completion of the ambitious architectural endeavour and publicly praised the officers who had managed it on behalf of the

213

214

Stefania Gerevini

Commune, performing what Oleg Grabar labelled as an indicative function.18 Furthermore, the two texts reminded the audience that the repair was the result of efficient public spending, implicitly acknowledging the collective generosity of the Genoese community, whose pious bequests had generated the financial resources necessary to carry out the works. Thus, the dedicatory inscriptions ultimately functioned as public memorials of the collective efforts of the Genoese state and civic community, incidentally conveying a message of social cohesion that, as illustrated later, was far from the everyday reality of late medieval Genoa. The textual meanings of the inscriptions of 1307 and 1312 were significantly amplified by their physical setting and visual properties, as was their capacity to aesthetically engage viewers. To begin with, the engraved texts performed an important ornamental role. Unlike the side aisles and the west wall of the church, the central nave of San Lorenzo carried no paintings or large-scale sculptural programme. With the exception of a few sculpted capitals and the bust of Janus situated above the foundation inscription on the north wall, the inscriptions represented the only decorative element of the banded walls of the nave. More experienced readers may have been able to appreciate the choice of Gothic majuscule script (rather than the more archaizing Roman capital lettering), which identified the texts as contemporary creations and further advertised the arcades of the nave as the outcome of a recent (and challenging) building campaign. However, their relative isolation, length and the large size of their characters increased the visual and symbolic prominence of the engraved texts, whose commanding presence and aesthetic qualities could be experienced and appreciated not only by literate readers but also by viewers with limited or no literacy. In addition, the commemorative inscriptions were written in clearly defined and well-spaced characters; they were formulaic and syntactically very straightforward; and the choice of words was almost identical in the two cases. Therefore, although the complete decoding of the inscriptions would presumably only be accessible to fully literate readers, viewers with at least a rudimentary familiarity with text would probably be able to identify the two inscriptions as a pair, and perhaps could recognise portions of the sentences and reconstruct their meaning.19 Yet decorating the central nave and paying tribute to two teams of zealous officers were only two of the aims of the cathedral inscriptions. More importantly, the epigraphs carved along the nave of San Lorenzo praised the city of Genoa by recapitulating its illustrious past, namely its foundation and enlargement by two eponymous heroes. The association between the Ligurian city and Janus was formulated in Genoa relatively late, in the thirteenth century. The connection was prompted by the homophony between the medieval denomination of the city, Ianua or Janua, and the name of the mythical character Ianus/Janus. The meaning of the Latin term ianua (door), exploited by Genoese writers to acclaim Genoa’s role as a two-way point of access to Lombardy and to the sea, further facilitated the association, because the Roman bifrontal deity Janus supervised thresholds.20 Ursone da Sestri was the first author to refer to

Written in Stone

Genoa as the city of Janus in his municipal chronicle, written in the early thirteenth century.21 In the late Duecento, Jacopo Doria further investigated and expanded the legend in his official annals (1280–94), attributing the foundation of the city to a fictitious Trojan nobleman named Janus.22 Finally, the most elaborate version of the myth, obtained through the reorganisation and (sometimes inventive) assemblage of a variety of classical and medieval sources and oral traditions, was compiled between 1295–8 by the archbishop of Genoa Jacopo da Varagine. In the first part of his Chronica Civitatis Ianuensis, Jacopo da Varagine identified three different characters, all known under the name of Janus, who played a prominent role in the foundation and subsequent aggrandisement of the city of Genoa. According to this version, Genoa was first founded as a small settlement named Ianicula (little Ianua) by Janus, the ancient king of Italy, who came from the East at the time of Moses, and founded Genoa seven hundred and seven years prior to the foundation of Rome, in 1546 BCE.23 The enlargement of ancient Ianicula to Ianua/Janua was due to a later Janus, the Trojan hero already mentioned by Jacopo Doria. After the war of Troy, Janus allegedly fled the city and, navigating westward, ‘joyfully got ashore’ in Liguria, where he set to aggrandise ancient Ianicula by building a fortress in the (medieval) district of Castellum, near the (medieval) palace of the archbishop, and strong fortifications around the inhabited centre. Jacopo da Varagine proudly calculated that this enlargement had taken place in 1019 BCE, 420 years prior to the foundation of Rome.24 Finally, the chronicler explained that a third Janus, king of Epirus, had been venerated as a god by the Romans and that the Genoese, allies of Rome, had also worshipped the same divinity prior to their conversion to Christianity.25 The epigraph engraved on the pillar of the north gallery of San Lorenzo evokes the earliest founder of the city. The lengthy inscription carved on the south wall of the nave, instead, primarily commemorates the re-founder of Genoa, the survivor of the war of Troy first mentioned by Jacopo Doria. These inscriptions indicate the widespread popularity of the myth of foundation of Genoa in the early Trecento. Also, as public commissions located inside the cathedral of the city, they demonstrate the keen interest that both the political and ecclesiastical authorities of Genoa took in issues of history and civic memory at this time and reveal the strong interconnections between political and religious identities in medieval city-states. Most importantly, the decision taken by the Genoese government to inscribe the mythical origins of the city on the cathedral walls is instructive of the authority and dependability attributed to monumental writing in late medieval Genoa. The myth of the foundation of Genoa lacked any historical evidence, and the silence of ancient written sources over this event had much troubled thirteenth-century Genoese writers. Jacopo Doria had hesitantly presented the story of Janus as vulgaris opinio (word of mouth) in the early 1290s.26 And Jacopo da Varagine, in spite of the numerous classical and medieval works cited in his version of the early history of Genoa, had also admitted that he had

215

216

Stefania Gerevini

relied on oral traditions more than on historical or literary sources in his recollection of the origins of the city. His statement on this subject is worth reporting in full, because it implicitly indicates that epigraphy may have been employed in San Lorenzo to certify and bestow an aura of truth onto the legendary origins of the city: Although these facts [the stories of Janus of Troy] cannot be found in any ancient history-book, they are testified as true by public and consolidated fame. Where the authority [of written sources] lacks, it is replaced by the celebrity of fame. The kind of fame that does not originate from small, but from grave matters, that is not particular but general, that is not new but ancient, by its very nature of fame does not only induce presuppositions and imply a certain probability, but also induces and produces evidence.27

The epigraphs in San Lorenzo provided the Janus myth precisely with the authority (auctoritas) that it lacked, and whose importance Jacopo da Varagine had uneasily attempted to minimize. The epigraphic commemoration of the deeds of the eponymous heroes, engraved in large characters and publicly visible inside the most important religious and civic building of Genoa, confirmed the widespread but volatile fame upon which the legend had thus far rested with the confident, durable assertion and the sound evidence of words written in stone.28 The role of inscriptions as repositories of civic memory and as signals of historical truth is likely to have been particularly prominent in early fourteenth-century Genoa. Civic memory had traditionally been recorded in the city through stately commissioned annals, continuatively produced and archived since 1099.29 After 1294, however, when Jacopo Doria presented the Commune with his history of Genoa, the production of municipal annals reached a halt, possibly because of the turbulent political situation of the following years; no municipal history of Genoa survives from the early Trecento.30 In this context, the role of the monumental epigraphs as public memorials is likely to have been especially significant, and their role in the construction and advertisement of the city’s official history may hardly be overestimated. The use of inscriptions as public certifications and as tokens of truth in Genoa interestingly compares with two near-contemporary cases of historical authentication in late medieval Italy. The public authorities of Padua also resorted to public writing to relaunch the image of the mythical founder of the city, the Trojan prince Antenor, and to celebrate the classical past of the town. In 1283, Antenor’s alleged funerary monument, already on display in the city centre, was renovated and aggrandised. On this occasion the tomb was equipped with a four-verse celebratory inscription commissioned from the leading local humanist Lovato Lovati. The text was written in Latin and modelled after classical sources, but was engraved in Gothic capital script. The choice of Gothic would identify the inscription as a medieval rather than an antique creation and suggests that the monument was engineered not only to celebrate Padua’s classical past but also to

Written in Stone

signal the continuity between the city’s contemporary history, its renowned tradition of learning, and its ancient origins. In the same town, an inscribed Roman tombstone was found near the local convent of Santa Giustina in the early fourteenth century. Mistakenly identified as the funerary epigraph of the Roman historian Livy, native of Padua, the inscribed tombstone was immediately set up on a wall at Santa Giustina and also promoted as the target of civic cult and as a material token of the classical past of the city.31 A rather different take on practices of historical authentication is offered by medieval Venice, where images rather than engraved texts appear to have functioned as bearers of truth. Venice was, like Genoa, an international commercial power, and similarly to the Ligurian town, it undertook a quest for illustrious origins that could illuminate and amplify its late medieval achievements. Even more than Genoa, however, Venice struggled with the complete lack of evidence of its existence in antiquity. Although medieval Venetian sources also made substantial attempts to link the origins of the city with the Trojan saga, the city chiefly resorted to later but equally significant events to craft its historical identity: it primarily associated its foundation and destiny with early Christianity and with the evangelist Mark.32 Interestingly enough, in concluding his account of the adventurous translation of the body of the evangelist from Alexandria to Venice in the ninth century, the late thirteenth-century Venetian chronicler Martino da Canale recommended that readers verify the truthfulness of his account by looking at the fac¸ade of the church of San Marco. There, he added, ‘This story is written in the same way as I retold it’.33 Curiously, Martino da Canale’s passage did not refer to a set of monumental inscriptions, but to the large-scale figurative mosaics that adorned the fac¸ade of San Marco and represented the most salient episodes of the translation of the evangelist to Venice. The different functions attributed to words and images in Padua, Venice and Genoa stimulate reflections about the mutable relationship between writing and visual representation and their relative authority in medieval city-states. However, they also collectively demonstrate that both images and public writing could be regarded as bearers of truth and that both could play an important part in the production and validation of historical claims. Incidentally, these examples also indicate that the realms of engraved words and images overlapped. Martino da Canale explicitly referred to the mosaics of San Marco as a form of writing. He indicated that they should be understood as significant public announcements and committed the dependability of his own historical narration to the authority of those images.34 Conversely, the Paduan and Genoese cases suggest that inscribed words shared in the function of images, engaging their viewers and conveying meaning not only through their textual contents but also by means of their physical presence and aesthetic qualities. The dedicatory inscriptions and the Janus epigraphs could convey meaning independently. However, one epigraph of each type was engraved on either wall, and in both cases the dedication and foundation texts were displayed near each

217

218

Stefania Gerevini

other and written in identical script. Their physical proximity and visual similarity would encourage approaching them as integrated textual ensembles, and as a consequence, the events that they narrated were also symbolically brought together, engaging viewers in issues of temporality and historicity at multiple levels. By means of visual juxtaposition, the boundaries between mythical event and contemporary history of Genoa were blurred, as were those between its real and imagined memories, presented together and given equal status on the walls of the cathedral. As a result, the renewal of the church was inserted within a long and illustrious history of rebuildings that went back to biblical times, and the restoration of the cathedral was implicitly related to the foundation of Genoa. Likewise, the reparatores Pastonus, Philip and Nicholas, whose names were displayed side by side with those of the illustrious founders of Genoa, were tacitly presented as epigones of the two Januses, whom Jacopo da Varagine had meaningfully praised in his chronicle as the actual builders of the city.35 The layout of the inscriptions, which stretched along the two sides of the nave and visually embraced the Genoese community gathered under its vaults, would further emphasise the continuity between myth, history and contemporary life in Genoa. Also, it would suggest interpreting the campaign of restoration of San Lorenzo as a symbolic re-foundation, not only of the church but also of the entire city of Genoa, whose violent civil war had caused the destruction of the cathedral a few years earlier. The spatial progression of the inscriptions, which developed towards the altar area and then retroceded on the south wall towards the atrium of the cathedral and its exit, implicated viewers in a visual perambulation along the nave. Thus, the experience and duration of viewing were made to converge with the mythical and historical developments that the inscriptions evoked; beholders, in turn, were encouraged to apprehend those events as the framework of their individual and collective histories. Finally (and much like the monumental inscriptions engraved on the walls of Georgian churches that Antony Eastmond discusses in Chapter 4), the physical arrangement of the epigraphs symbolically joined the secular time of history, which the texts evoked, with the cyclical time of liturgy, celebrated in the cathedral. The renewal of the cathedral of Genoa, its present history and its remote origins were thus integrated in the broader story of human salvation, reenacted in the ceremonials of San Lorenzo. Whether or not they were originally conceived as a unified textual programme, the overall contents of the inscriptions on the north and south walls of the nave ultimately mirrored each other, and the epigraphs functioned as a visual ensemble. However, the specific information that each wall provides and the visual layout of text differ in the two cases. On the north wall, the inscriptions extend over a limited portion of the nave. Also, the dedication is engraved in large letters and without abbreviations, whereas the Janus epigraph is written in smaller script, includes abbreviations and is confined to a pillar at gallery level (Fig. 62). The different size and arrangement of the two epigraphs suggest that, although visually complementary, the two inscriptions did not have the same significance

Written in Stone

and would not be experienced in the same way by viewers standing in the nave. The dedicatory inscription occupied a more conspicuous position and would be seen and read more easily: hence, the celebration of the reparatores of the church (and of the Genoese state) is likely to have been conceived as the centrepiece of the epigraphic programme on this side of the nave. The arrangement and visual impact of the inscriptions on the south wall are significantly dissimilar. The dedicatory text and the foundation epigraph are neatly carved one after the other, on a single line, and occupy the entire nave length. Their script is identical, and the spacing between letters and between words is the same, as is the height of individual letters, which also matches the lettering on the north wall. However, the Janus inscription is much longer than the dedication (6.5 against 2.5 bays), and the characters that compose it are remarkably wider (0.3 to 0.5 times larger) than those forming the dedicatory epigraph. This phenomenon is particularly apparent in the incipit of the second phrase, where the letters of the title ‘Janus prince of Troy’ are noticeably larger and visually more prominent than the characters forming the preceding words (Fig. 64). This variation is likely to have been deliberate. The Janus epigraph begins in front of the sculpted effigy of the founder of Genoa placed in the north gallery, and this symmetry, which is unlikely to be accidental, may have influenced the layout of all text on this side of the nave. If the Janus epigraph was planned to face its counterpart on the north wall, then the dedicatory inscription would have to be squeezed in the two and a half bays available to the east of the sixth column, leaving the remaining six and a half bays free for the (longer) Janus text (Fig. 61). This solution imposed that a higher number of abbreviations be employed in the dedicatory text on the south wall than in its equivalent on the north side, but was otherwise convenient in various respects. The officers in charge of the restoration were commemorated at the east end of the nave, nearer the altar and closer to the relics of St Syros, the first bishop of Genoa. Instead, the Janus text, which referred to the pre-Christian past of the city and expounded its political heritage, aptly covered the western section of the nave, occupied by the laity.36 The varying width of letters also contributed to visually distinguish the two (physically contiguous) components of the epigraph. The identical form, height and spacing of letters, in turn, preserved their visual consistency, as well as the aesthetic cohesion of the whole nave. By virtue of its length and location, and of the greater width of its characters, the Janus epigraph would be visually more imposing and more easily readable for viewers in the nave than the dedicatory inscription that accompanied it. Assigning more space, and thus greater symbolic importance, to the myth of foundation of the city, this layout reversed the visual arrangement of the north wall, where the dedicatory text, larger and more visible, was preeminent. A succinct examination of the troubled political history of Genoa in the early fourteenth century may shed some light on the possible reasons behind such different pagination.

219

220

Stefania Gerevini

As was the case with many other central and northern Italian Communes, the political life of late medieval Genoa was marked by violent social frictions between aristocratic family groups and non-noble wealthy merchants and artisans, as well as by relentless political confrontations between Guelph and Ghibelline parties and, in the early fourteenth century, amongst their inner factions.37 The years 1306–12, when the restoration of the cathedral was under way, were a period of unprecedented civic violence and hectic political turnovers.38 Following the expulsion of the Guelph party from Genoa after the riots and fire of 1296, the city was intermittently ruled by the two leading Ghibelline families Doria and the Spinola through the office of the captains of the people, and by a foreign podest`a formally in charge of the government between 1299 and 1306. By this time, however, the once-strong alliance between the Doria and the Spinola had deteriorated, due to partisan interests and to divergences over international matters – particularly, Genoa’s prospective conduct towards the crown of Anjou. When Opizzino Spinola and Bernabo` Doria were proclaimed captains of the people in January 1306, after a vehement outbreak of urban combat, the disagreements inside the Ghibelline party were irreparable and soon produced further armed conflicts.39 In 1307, the Doria joined forces with the Guelph Grimaldi, their ancient enemies, to attempt a coup d’´etat against Opizzino Spinola, who had meanwhile revealed explicit monocratic tendencies, repeatedly sidestepping his colleague. Following the failure of the coup, Bernabo` Doria, although formally uninvolved in the incident, was deposed and imprisoned, and Opizzino had himself proclaimed solo ruler (capitaneus perpetuus et generalis comunis et populi Janue) in the summer of 1308. The autocratic turn provoked an immediate reaction of the Genoese nobility, whose dissent against Opizzino temporarily superseded the deep-rooted antagonisms between Guelphs and Ghibellines. The Ghibelline enclaves of the Doria and Spinola de platea joined forces with the Guelph leaders Grimaldi and Fieschi, and defeated Opizzino at Sestri Ponente, outside Genoa, in June 1309. Opizzino continued his armed resistance on the Ligurian coast until the following year. Meanwhile, the government of the city was temporarily entrusted to a council of twelve governors, presided by an abbot and a podest`a.40 Such was the political situation in Genoa when, in January 1311, Henry VII of Luxembourg, emperor-elect of the Holy Roman Empire, was crowned king of Italy in Milan and received the oath of fealty from the Communes of Lombardy – uniting the region and bringing imperial rule into an area that had thus far been divided into independent and belligerent city-states. The advent of the emperor-elect and the new political configuration of Lombardy provoked conflicting responses in the neighbouring states of northern and central Italy. While Guelph Florence, allied with Bologna, set to organise military resistance against Henry VII, Genoa, torn apart by decades of civic unrest and factionalism, welcomed the emperor-elect as the forbearer of a new age of peace.41 Representatives of the Genoese commune attended Henry VII’s coronation in Milan,

Written in Stone

and once they were reassured that Genoa’s ancient imperial privileges, including the Commune’s fiscal and judicial autonomy, would be respected by the new emperor-elect, they swore an oath of fealty on behalf of the Commune, recognising Henry VII as the legitimate lord of Genoa.42 Genoa’s pledge of allegiance to Henry VII was ratified during the following months. The emperor-elect arrived in Genoa on 21 October 1311, where he was solemnly greeted and escorted on a solemn procession by the political authorities of the Commune – the podest`a and abbot of the people accompanied by representatives of the aristocracy and of the popolani.43 As soon as the solemn reception was over, Henry set to bring order to the city and define the specifics of his dominion over Genoa, a task that would engage him for several months, until mid-February 1312. He undertook elaborate investigations, questioning leaders of the aristocratic families as well as representatives of the trade guilds as to the best solution to adopt in Genoa. On 21 November 1311 the Genoese political representatives revised the terms of their fealty to Henry VII, requesting that he accept full and direct sovereignty over their city and over its contado for twenty years and publicly declaring that this would be the only way to guarantee peace.44 Henry accepted their request, and the following day his sovereignty over Genoa was solemnly ratified before the Genoese people in front of the cathedral of San Lorenzo.45 Unfortunately, the new political order was short-lived. Henry VII remained in Genoa until 16 February 1312, withheld by a wave of plague that struck the city in December 1311 and caused the death of his wife Margaret of Brabant.46 Meanwhile, the opposition against the emperor-elect in central Italy increased dramatically and soon resulted in open military confrontation.47 The untimely death of Henry VII, which occurred in 1313 shortly after his coronation as Holy Roman Emperor in Rome 1312, put an end to his attempts to bring imperial rule and political concord in Italy, and following his death, Genoa underwent a new phase of political and civic unrest.48 Scholarship has already suggested that the keen interest taken by the political and ecclesiastical authorities of Genoa in the mythical origins and early history of the city, as well as their decision to advertise them on the walls of the cathedral, may have been connected with the dramatic civic struggles of the first decade of the Trecento. Clario di Fabio has argued that the monumental inscriptions, the sculptures executed in the cathedral in the same years (and particularly the bust of Janus in the north gallery) and the frescoes that were set up in the church shortly after 1312 should be regarded as a unified programme intended to advertise the political and religious origins of the city at a time of intense political unrest.49 Carrie Benes has more specifically suggested that the monumental epigraphs in San Lorenzo were meant to provide contemporary Genoese with powerful counterexamples of civic unity and cohesion and to encourage them to abandon personal rivalries and live up to the standards set by their ancestry.50

221

222

Stefania Gerevini

The epigraphic apparatus of San Lorenzo was undoubtedly related to the dramatic political events of the years 1306–12. The inscriptions commemorated in a durable form an ambitious public enterprise, accomplished at a difficult time, when the very survival of the Commune of Genoa was at peril. They reminded the fragmented civic community and its petty rulers that they shared common origins and a common past, encouraging their collaboration in the present. Finally, they advertised and certified the (imagined) history of the city both as an antidote against its domestic social and political decline and as a proud recognition of its contemporary commercial successes in the Mediterranean. In addition to fulfilling these broad civic functions, it is also possible that the inscriptions in San Lorenzo, and particularly the Janus text on the south wall, may have more directly reflected the conclusive – and most critical – event of that decade, the advent of Henry VII and his short-lived dominion over Genoa. Although its pacifying effects were not enduring, Henry VII’s coming to Genoa was a political and historical event of unprecedented importance for the Ligurian town. At one level, it signified a tragic political failure: Genoa renounced its political autonomy, which it had strenuously defended against imperial claims for more than a century. At another level, though, the advent of the emperor represented an attractive opportunity for political renewal of the city, which could finally look forward to a government capable of placating its civic unrest and of bypassing the rivalries and animosities among its noble families. The stopover of the emperor-elect in Genoa coincided with the final phase of restoration of the south arcade of the nave of San Lorenzo, which appears to have lasted five years, from 1307 to 1312. The inscriptions were presumably engraved after the structural works on the arcade were completed, and it is therefore improbable that they were set up before 1311, when Henry VII undertook his journey through Italy. Thus, the emperor-elect’s sojourn in Genoa is likely to have overlapped with (or shortly preceded) the engraving of the monumental epigraphs on the south wall. Furthermore, the Genoese formally transferred the dominion of their city to the emperor-elect in front of the cathedral of San Lorenzo on 22 November 1311, and Henry VII undoubtedly visited the cathedral during his protracted stay in Genoa. In this context, it seems plausible that the inscriptions on the south wall, sponsored by the Genoese state, may have alluded to the recent political events. As already discussed, the Janus inscription on the south wall recapitulates the mythical origins of Genoa, founded by Janus, the first king of Italy, and later re-founded by a survivor of the war of Troy known with the same name. As was the case with the inscription on the north wall, the founder of Genoa is identified here as Iano rege Ytalie, ‘Janus, king of Italy’. The title Rex Italie, which in the middle ages implied sovereignty over the northern regions of Italy and upper Tuscany, enjoyed an enduring association with the western empire. Charlemagne had conquered this area from the Lombard king Desiderius in 774, and his successors, starting with his grandson Lothar I (795–855), were designated in official

Written in Stone

documents with the title Rex Italie, synonymous with Rex Longobardorum (king of the Lombards). 51 Interestingly, just before he sojourned in Genoa and took over the government of the city, Henry VII, following the ancient Carolingian tradition, had been crowned king of Italy in the Basilica of Saint Ambrose in Milan, on 6 January 1311.52 Contemporaries would undoubtedly be sensitive to the heraldic analogy: the epigraphic apparatus of San Lorenzo would thus implicitly encourage them to associate the mythical king of Italy Janus with his successor Henry VII, whose legitimate rulership and status as a novel founder of the city the texts indirectly advertised. This hypothesis is further reinforced by the incipit of the Janus inscription on the south wall. The text begins with a solemn commemoration of the second founder of Genoa, ‘Janus prince of Troy’ (Jan[us] pri[n]ceps Trojan[us] ), whose name and title are visually emphasised by means of wider characters. Interestingly enough, neither Jacopo Doria nor Jacopo da Varagine had referred to the Trojan Janus as princeps. The former had simply described him as ‘a nobleman from Troy, named Janus’ (nobilis trojanus nomine Janus), and Jacopo da Varagine had identified him as a Trojan magnate (magnas) and citizen (cives), acknowledging his aristocratic status but never designating him as a ruler.53 Thus the conferral of a regal title on the mythical hero Janus represented a departure from available textual sources, and the great visual prominence given to this portion of text seems to indicate that such deviation was deliberate and that the beholders’ attention was intended to focus on this detail. The authorities of Genoa may simply have attributed a royal title to their mythical founder in order to enhance the legendary lineage of medieval Genoa. However, the concomitant handover of the government of the city to Henry VII may have provided additional reasons for this change. The emphasis placed by the inscription on the royal status and on the foreign provenance of Genoa’s ancient founders may have been intended as an implicit tribute to Henry VII, who shared both these qualities and was habitually designated as princeps in official documents.54 At the same time, the epigraph addressed a reassuring message to the local community. Albeit greeted with optimism, the advent of the emperor-elect marked a significant loss of political autonomy for Genoa. Revealing that the early history of the city was marked by the interventions of foreign rulers, the inscription set a positive precedent for the advent of Henry VII, and thus normalised and publicised the recent political turn, projecting onto him the city’s expectations of aggrandisement and renewal. Finally, the inscriptions may have conveyed an indirect message of political submissiveness. Jacopo Doria provided no description of ancient Ianicula (the small settlement enlarged by the Trojan hero Janus) in his account of the foundations of the city, and Jacopo da Varagine concisely explained that Janus ‘carefully considering the site, liked it very much’ (multum sibi placuit).55 The Janus inscription on the south wall creatively supplemented its literary antecedents, describing the early settlement as ‘healthy, easy to rule, and secure’. As explained earlier, the

223

224

Stefania Gerevini

surrender of Genoa to the emperor-elect in 1311 was not immediate or unconditional. The city carefully negotiated the duration of Henry’s direct control and defended its long-lived administrative and fiscal privileges, eventually submitting to the emperor for a limited period of twenty years and obtaining that the contado be administered by Genoese citizens. In the light of this delicate political balance and of the overall fragility of Henry’s rule over Italy, the adjective dominabilis (easy to rule or easy to defend) applied to the ancient settlement of Ianicula seems particularly meaningful, because it may have propagandised – against the manifest unruliness of the city – Genoa’s current docility and loyalty to the emperor-elect. Thus the epigraphs that cover the north and south walls of the nave of the cathedral of San Lorenzo were an ambitious architectural and artistic endeavour, and they represent an exceptional opportunity to investigate the textual and visual functioning of monumental writing in late medieval Italy. Firstly, they acted as the material celebrations of the officers who had supervised the reconstruction of the nave of the cathedral, greatly damaged by the fire of 1296. As such, they signal the complex dynamics of social progression and political recognition of individuals in the medieval commune, as well as the relevant part that public writing could play in fostering (or sanctioning) social ascent.56 Secondly, and more importantly, the physical durability and public accessibility of the inscriptions in San Lorenzo enhanced the official character of their statements, and thus the credibility of the events that they narrated, allowing them to function as public deeds that performed a variety of tasks. At a first level, they complemented – and, for some time at least, replaced – municipal annals in registering the collective memory of Genoa in the early Trecento, testifying to the authority of the engraved word in Genoa and to its capacity to produce and promote shared memories. The cathedral inscriptions, however, did not only certify and advertise the past of the city. They also functioned, to use Matthew Canepa’s effective wording, as veritable technologies of memory. Occasionally departing from available literary sources, they also marginally overwrote and redefined the past of Genoa in order to address the pressing political and civic agenda of the early Trecento. Thus, they pose intriguing questions about the relative hierarchy between archival and epigraphic writing in the middle ages and about the different roles that they played in fixing, preserving and communicating official civic memory. In the context of the turbulent political life of the early Trecento, the inscriptions may finally have functioned as implicit political commentaries, not unlike the epigraphic erasures that Jonathan Bloom comments upon in Chapter 3. Compelled by its own undisciplined ruling class to give up its autonomy, the city reinterpreted the latest political events in the light of its (recently crafted) legendary history and inscribed the walls of its civic temple to normalise its recent past and to celebrate the grandeur of the city, in spite of all. By means of their textual contents, visual properties and spatial layout, the inscriptions in San Lorenzo engaged their audience aesthetically and

Written in Stone

225

intellectually, implicating viewers in a visual journey along the nave and across Genoese history. As discussed earlier, their style and extension drew the attention of beholders to the novelty and exceptionality of the building campaign. Also, the visual juxtaposition of historical facts and mythical events encouraged them to apprehend the legendary and historical past of the city as a continuous and unified narrative. Finally, the epigraphs required that viewers move along the nave, either physically or visually, and integrated the act and moment of viewing in the events that the epigraphs evoked, as well as in the liturgical space of the cathedral. Ultimately, they integrated individual viewers and the Genoese community as a whole in a unified narrative that stretched from biblical and mythical times to the Trecento; and they joined the linear narrative of history with the cyclical progression of liturgy, implicitly reminding the faithful of the ultimate goal of Christian communities: human salvation. NOTES

1. The most comprehensive survey of the cathedral of Genoa in the middle ages is La Cattedrale di Genova nel Medioevo. Secoli VIXIV, ed. Clario di Fabio (Genoa, 1998). On the modern campaigns of restoration of the cathedral, see Clario Di Fabio, ‘San Lorenzo,’ in Medioevo Restaurato. Genova 1860–1940, ed. Colette Dufour Bozzo (Genoa, 1984), 193– 248. Other studies on San Lorenzo include Giuseppe Banchero, Il Duomo di Genova (Genoa, 1855), with extensive documentary appendixes; Camillo Boito, Il Duomo di Genova ed i Nuovi Lavori (Milan, 1910); Tipografia della Gioventu, ` La Cattedrale di Genova 1118– 1918 (Genoa, 1918); Guglielmo Salvi, ‘La Cattedrale di Genova. San Lorenzo,’ Italia Sacra, N.S. 2 (1931), 845–1003; Bartolomeo Pesce and Antonio Camposanto, La Cattedrale di Genova (Genoa, 1959). See also Fulvio Cervini, I Portali della Cattedrale di Genova e il Gotico Europeo (Florence, 1993), with extensive bibliography. 2. I am very grateful to Prof. Flavia De Rubeis for her help with the palaeography of the inscription. For an introduction to the development of Gothic capital script in medieval Italy and for an examination of its regional variants, see Flavia De Rubeis, ‘La Capitale Romanica e la Gotica Epigrafica: una Relazione Difficile,’ Scripta 1 (2008), 33– 44. 3. The inscription was first edited (with minor inaccuracies) by Augusta Silva, Corpus Inscriptionum Medii Aevii Liguriae, 3 vols. (Genoa, 1987), vol. 3: Genova: Centro Storico, n.25, 16– 17. Also edited and discussed by Clario Di

4.

5.

6.

7.

Fabio, ‘L’Incendio del 1296 e la ‘Reparatio Ecclesie’ fra 1297 e 1317,’ in La Cattedrale di Genova, ed. Di Fabio, n.17, 252; and by Carrie E. Benes, Urban Legends. Civic Identity and the Classical Past in Northern Italy, 1250–1350 (University Park, 2011), n.79, 214, (translation into English at page 78). Silva, Corpus Inscriptionum, n.53, 32; Clario di Fabio, ‘Sculture, Affreschi ed Epigrafi. La Citt`a e i suoi Miti delle Origini. Fonti, Committenti, Esecutori,’ in La Cattedrale di Genova, ed. Di Fabio, 260 and n.10, 278; Benes, Urban Legends, 79 and n.81, 214. The legend and its literary fortune in medieval Genoa are discussed in detail later. The significance of the Janus inscriptions in the cathedral of San Lorenzo has primarily been discussed in Di Fabio, ‘Sculture, Affreschi ed Epigrafi,’ 258–79; and in Benes, Urban Legends, 63–87. On the sculpted effigy of Janus, see Clario Di Fabio, ‘La Scultura Bronzea a Genova nel Medioevo e il Programma Decorativo della Cattedrale nel Primo Trecento,’ Bollettino d’Arte, 55 (Series VI), May-June (1989), 1–44; and Clario di Fabio, ‘Cantieri, Scultori ed Episodi di Committenza nel Trecento,’ in Niveo de Marmore. L’uso Artistico del Marmo di Carrara dall’ XI al XV secolo, ed. Enrico Castelnuovo (Genoa, 1992), 223–33. Silva, Corpus Inscriptionum, n.26, 17; Di Fabio, ‘L’incendio del 1296,’ n.17, 252; and Benes, Urban Legends, 81 and n.84, 214. Di Fabio, ‘Sculture, Affreschi ed Epigrafi,’ 260 and n.10, 278; and Benes, Urban Legends, 81 and n.84, 214.

226

Stefania Gerevini

8. Stephen Epstein, Genoa and the Genoese, 958– 1528 (Chapel Hill, NC, 1996), 161–2. The reference work on education and literacy in medieval Genoa is Giovanna Petti Balbi, L’insegnamento nella Liguria Medievale: Scuole, Maestri, Libri (Genoa, 1979); see also Giovanna Petti Balbi, Una Citt`a e il suo Mare. Genova nel Medioevo (Bologna, 1991), 264–85. 9. Public writing in Genoa still awaits a comprehensive investigation. On the civic meaning of two inscribed panels installed in 1155 on the eastern city gate (Porta Soprana), see Colette Dufour Bozzo, La Porta Urbana nel Medioevo. Porta Soprana di Sant’Andrea in Genova; Immagine di una Citt`a (Rome, 1989). The elaborate political messages conveyed by the multi-line inscriptions that cover the fac¸ade of San Matteo have been examined by Rebecca Muller, ‘Constructing Fame in ¨ a Town: The Case of Medieval Genoa,’ Medium Aevum Quotidianum, 47 (2003), 8– 32, and Sic Hostes Ianua Frangit. Spolien und Trophaen im mittelalterlichen Genua (Marburg, 2002), 108–56 (126–33 texts on the fac¸ade of San Matteo). The same volume also briefly discusses two kufic inscriptions with Qur’anic verses installed above the northern arcade of the church of Santa Maria di Castello, near the city port. Muller, Sic Hostes Ianua ¨ Frangit, 65 and cat. 8, 207–9 with further bibliography. Secular buildings could also carry extensive epigraphic apparatuses. Palazzo San Giorgio, built in 1257 as the new communal palace, and soon converted into the headquarters of the customs and tax department of the city, features a significant number of inscriptions, including a commemorative epigraph dating from 1260 (Silva, Corpus Inscriptionum, n.190, 109), a spoiled inscription stolen from Pisa on the occasion of the conquest of the castle of Lerici in 1251 (Silva, Corpus Inscriptionum, n.191, 109–10) and a number of epigraphic charters (fifteenthsixteenth centuries) that recorded the mortgaging of public securities to private Genoese benefactors. The function of these inscribed charters has not been systematically assessed. However, they were edited by Silva, Corpus Inscriptionum, n.192–209, 110–27. Some photographs may be found in Orlando Grosso, Il Palazzo di San Giorgio (Genoa, 1984), 64–79. 10. Luigi Volpicella, Elena Skrzinska, and Ettore Rossi, Iscrizioni Genovesi in Crimea ed in Costantinopoli. Atti della Societ`a Ligure di Storia Patria, vol. 56 (Genoa, 1928). 11. The inscription may have been located in the gallery of the Rotunda, in the proximity of the

holy sepulchre of Christ. It was destroyed by King Amaury in 1160. Jaroslav Folda, The Art of the Crusaders in the Holy Land. 1098–1187 (Cambridge, 1995), 62–4, with further bibliography. 12. The significant visual impact exerted by the inscriptions in San Lorenzo is also indirectly confirmed by a reference made to them by Petrarch in his brief description of Genoa written in 1358. Francesco Petrarca, ‘Itinerarium ad Sepulcrum Domini,’ 12, in Opera Omnia, ed. Pasquale Stoppelli (Roma, 2000, CD-ROM version); also edited and translated in Italian in Petrarca il Viaggiatore. Guida ad un Viaggio in Terra Santa, ed. Raffaella Cavalieri (Rome, 2007), n.12, 120–1. 13. For a painstaking account of the works of restoration in the cathedral, and for their chronology, see Di Fabio, ‘L’incendio del 1296,’ 223–53, esp. 230–1 for details of the replacement of the columns. On the frescoes painted in the cathedral during the final phase of the restoration, see Robert Nelson, ‘A Byzantine Painter in Trecento Genoa: The Last Judgement at S. Lorenzo,’ The Art Bulletin, 67.4 (1985), 548–66; and Robert Nelson, ‘Byzantine Icons in Genoa before the “Mandylion”,’ in Intorno al Sacro Volto. Genova, Bisanzio e il Mediterraneo (secoli XI–XIV), eds. Anna Rosa Calderoni Masetti, Colette Dufour Bozzo, and Gerhard Wolf (Venice, 2007), 79–92. 14. Leges Genuenses, Historiae Patriae Monumenta, vol. 18, eds. Vittorio Poggi, Luigi Tommaso Belgrano, and Cornelio Desimoni (Turin, 1901), c. 121–2. For a biographic profile and chronology of these officers between 1297 and 1317, see Di Fabio, ‘L’incendio del 1296,’ 223–30, with further references. Di Fabio indicates in the same essay (n.14, 252) that an enlarged group of four officers was commemorated in a dedicatory inscription in 1317. This epigraph (Silva, Corpus Inscriptionum, n.28, 18) was originally attached to the altar of the chapel of Saint James, but is currently visible on the intrados of the door leading into the sacristy of San Lorenzo. On the office of the constituti, and more generally on the administration of the cathedral during the middle ages, see Valeria Polonio Felloni, ‘Da ‘opere’ a Pubblica Magistratura. La Cura della Cattedrale e del Porto nella Genova Medievale,’ in Opera. Carattere e Ruolo delle Fabbriche Cittadine fino all’Inizio dell’Et`a Moderna, eds. Margaret Haines and Lucio Riccetti (Florence, 1990), 117– 36.

Written in Stone

15. I Libri Iurium della Repubblica di Genova, vol. 1/1, ed. Antonella Rovere (Genoa, 1992), n.230, cited in Polonio Felloni, ‘Da ‘opera’ a Pubblica Magistratura,’ 123. 16. Statuto dei Padri del Comune della Repubblica Genovese, ed. Cornelio De Simoni (Genoa, 1885), 44, 180, cited in Polonio Felloni, ‘Da ‘opera’ a Pubblica Magistratura,’ 129. 17. Leges Genuenses, n.204, c. 121–2; and Statuto dei Padri del Comune, 180, also cited in Di Fabio, ‘L’incendio del 1296,’ 223 and n.11, 252 and in Polonio Felloni, ‘Da ‘opera’ a Pubblica Magistratura,’ 130. 18. Oleg Grabar, ‘Graffiti or Proclamations: Why Write on Buildings?’ In Islamic Art and Beyond, vol. 3, Constructing the Study of Islamic Art (Farnham, 2006), 239. The essay was first published in The Cairo Heritage, ed. Doris Behrens-Abouseif (Cairo, 2000), 69–75. 19. For a discussion of the different audiences of medieval epigraphy, see Vincent Debiais, Message de Pierre. La Lecture des Inscriptions dans la Communication M´edi´evale (XIII–XIV si`ecles) (Turnhout, 2009), 247–92, and further bibliographic references, 403–7. This volume represents an excellent introduction to many other issues discussed in the present chapter. 20. Jacopo Doria (c. 1233- d. before 1305), ‘Annales Ianuenses’, in Annali Genovesi di Caffaro e dei Suoi Continuatori. Dal 1099 al 1293, 5 vols. (Genoa, 1890–1929), vol. 5, ed. Cesare Imperiale di Sant’Angelo (1929), 8; Giovanni Balbi (d. 1298?), ‘Ianua’ and ‘Ianus’, in Catholicon seu Summa Prosodiae (Genoa, 1286; edited and printed in Venice, 1506), Part V, lacks page number; Jacopo da Varagine (c. 1230– 98), ‘Chronica Civitatis Ianuensis,’ in Cronaca della Citt`a di Genova dalle Origini al 1297, ed. Stefania Bertini Guidetti (Genoa, 1995), 359, also associates Genoa with the Latin ianua. On the significance of onomastics in medieval Genoa, see Clario di Fabio, ‘Il ‘Mito delle Origini’ e il nome di Genova nel Medioevo,’ Bollettino Ligustico per la Storia e la Cultura Regionale, 31 (1979), 37–44. 21. The reference to Janus occurs in the annals under the year 1227. Annali Genovesi, vol. 3, ed. Cesare Imperiale di Sant’Angelo (1923), 26; discussed by Giovanna Petti Balbi, Caffaro e la Cronachistica Genovese (Genoa, 1982), 58. 22. Jacopo Doria, ‘Annales Ianuenses’, 4. 23. Jacopo da Varagine, ‘Chronica,’ 342–4, and 349 for the alleged date of foundation. 24. Ibid., 344–7, and 351 for the alleged date of aggrandisement.

227

25. Ibid., 348–9. 26. Jacopo Doria, ‘Annales Ianuenses’, 4. 27. ‘Ista igitur, etsi non inveniantur in aliqua antiqua ystoria, inveniuntur vera esse ex fama publica et antiqua. ubi deest auctoritas, suplet/fame celebritas; fama, que non [habet] ortum a levibus, sed a gravibus, que non est particularis sed generalis, que non est nova sed antiqua, talis in quantum fama non solum inducit presumptionem, non solum facit robabilitatem, sed etiam inducit e facit probationem.’ Jacopo da Varagine, ‘Chronica,’ 347–8. 28. The employment of monumental writing (particularly on cathedral walls) as a means to sanction and publicise legal documents, laws or other messages of public importance is widely attested in medieval Italy. For a general introduction, see Nicoletta Giov`e Marchioli, ‘L’Epigrafia Comunale Cittadina,’ in Le Forme della Propaganda Politica nel Due e nel Trecento, ed. Paolo Cammarosano (Rome, 1994), 263–86. More specifically, in 1214 a law regulating the alienability of Genoese public debt appears to have been engraved on the walls of the cathedral of San Lorenzo, as in Banchero, Il Duomo di Genova, 35, after Michele Giuseppe Canale, Nuova Istoria della Repubblica di Genova, del suo Commercio e della sua Letteratura (Florence, 1858), 231 (wrong page in Banchero). However, Canale does not provide the details of his primary source, and no traces of these inscriptions are currently visible in the cathedral. 29. See Annali Genovesi. The Genoese annalistic tradition has been analysed by Giovanna Petti Balbi, Caffaro e la Cronachistica Genovese (Genoa, 1982), partly reprinted in Giovanna Petti Balbi, Una Citt`a e il suo Mare, 249–63. For a brief examination of Italian historiographical trends in the twelfth century (including Genoa), see Chris Wickham, ‘The Sense of the Past in Italian Communal Narratives,’ in The Perception of the Past in TwelfthCentury Europe, ed. Paul Magdalino (London, 1992), 173–89. On Caffaro, also see Richard D. Face, ‘Secular History in Twelfth-Century Italy: Caffaro of Genoa,’ Journal of Medieval History 6 (1980), 169–84. For a comparison between historical discourse in Genoa and in Venice, see Elizabeth Crouzet-Pavan, ‘Genes et Venise: Discours Historiques et Imaginaires de la Cit´e,’ in Le Forme della Propaganda Politica, ed. Paolo Cammarosano, 427–53. 30. Petti Balbi, Caffaro e la Cronachistica Genovese, 81–2, where it is also specified that the tradition of municipal chronicles in Genoa

228

31.

32.

33.

34.

35.

36.

Stefania Gerevini

was only resumed in 1396, with the annals by Giorgio Stella; and Petti Balbi, Una Citt`a e il suo Mare, 258–9, where it is clarified that Jacopo da Varagine’s ‘Chronica’, although rooted in the annalistic tradition, was not a stately commissioned work and significantly differed from municipal annals in terms of its structure and aims. On Antenor’s monument, see Padova per Antenore. Atti della giornata di studio tenutasi il 14 dicembre 1989 presso il Museo Civico Archeologico agli Eremitani e altri interventi, ed. Girolamo Zampieri (Padua, 1990). Antenor’s tomb inscription and Livy’s tombstone are discussed in the context of the promotion of the mythical and classical past of medieval Padua in Benes, Urban Legends, 39–60, with extensive bibliography. Literature on the medieval history of Venice and on its quest for origins is vast. The reference reading on this subject is still Patricia Fortini Brown, Venice and Antiquity. The Venetian Sense of the Past (New Haven and London, 1996). By the same author, see also ‘History as Myth. Medieval Perceptions of Venice’s Roman and Byzantine Past,’ in The Making of Byzantine History: Studies Dedicated to Donald M. Nicol, eds. Roderick Beaton and Charlotte Rouech´e (London, 1993), 145–57. For an introduction to the civic cult of St Mark in Venice, see Otto Demus, The Church of San Marco in Venice. History, Architecture, Sculpture (Washington, DC, 1960); and Thomas Dale, ‘Inventing a Sacred Past. Pictorial Narratives of St. Mark the Evangelist in Aquileia and Venice, ca. 1000–1300,’ Dumbarton Oaks Papers, 48 (1994), 53–104. Martino Da Canal, La cronique des V´eniciens. Cronaca veneta de maistre Martin da Canal, dall’origine della citt`a sino all’anno 1275, eds. Filippo Luigi Polidori, transl. Giovanni Galvani, Archivio Storico Italiano, ser. I, 8 (1845), 290. This passage is similarly discussed in Fortini Brown, Venice and Antiquity, 33 and in Patricia Fortini Brown, Venetian Narrative Painting in the Age of Carpaccio (New Haven, 1988), 79–80, with further remarks on the role of painted images as tokens of truth in Venice. Jacopo da Varagine explicitly indicated that the earlier Janus ‘built’ Genoa (Ianuam construxit), and carefully described what areas of the city were erected by the Trojan Janus. Jacopo da Varagine, ‘Chronica,’ 343 and 347. The body of St Syros had been reinvented inside the altar of San Lorenzo and offered

37.

38.

39. 40. 41.

42.

43.

to public veneration by Archbishop Jacopo da Varagine in 1293. Jacopo da Varagine, ‘Chronica,’ 500–1. For a critical appraisal of this event, see Stefania Bertini Guidetti, ‘Contrastare la crisi della Chiesa Cattedrale: Iacopo da Varagine e la Costruzione di un’Ideologia Propagandistica,’ in Le Vie del Mediterraneo. Idee, Uomini, Oggetti (secoli XI– XVI), ed. Gabriella Airaldi (Genoa, 1997), esp. 158–62. The reference work for the late medieval history of Genoa is still Georg Caro, Genua und die Machte am Mittelmeer, 1257–1311, 2 vols. (Halle, 1895–9), transl. in Italian as Georg Caro, Genova e la Supremazia sul Mediterraneo. Atti della Societ`a Ligure di Storia Patria, New Series, 14–15 (Genoa, 1974–5). See also Epstein, Genoa and the Genoese; and Storia di Genova: Mediterraneo, Europa, Atlantico, ed. Dino Puncuh (Genoa, 2003). On the acute political strife of those years, see Caro, Genova e la Supremazia, vol. 2 (15), 303–81; and Axel Goria, ‘Le Lotte Intestine in Genova tra il 1305 e il 1309,’ in Miscellanea di Storia Ligure in onore di Giorgio Falco, ed. Giorgio Falco (Milan, 1962), 251–80. Caro, Genova e la Supremazia, vol.2 (15), 318– 27; and Goria, ‘Le Lotte Intestine,’ 253–6. Caro, Genova e la Supremazia, vol.2 (15), 341–52. The most detailed modern account of the emperor-elect’s Italian adventure is William Bowsky, Henry VII in Italy. The Conflict of Empire and City-State, 1310–1313 (Lincoln, 1960). A very informative account of the sovereign’s visit to Genoa and of the related political events is Caro, Genova e la Supremazia, vol. 2 (15), 370–81. For a contemporary account of the events, see Nicholaus De Butrinto, Relatio de Itinere Italico Henrici VII Imperatoris ad Clementem V Papam, Rerum Italicarum Scriptores 9 (Milan, 1726), 887–934. The event occurred on 28 January 1311. Acta Henrici VII Imperatoris Romanorum et Monumenta Quaedam Alia Medii Aevi, ed. Wilhelm G. Donniges, 2 vols (Berlin, 1839), vol. 1, n.70, 37–8; Monumenta Germaniae Historica. Leges,Constitutiones et Acta Publica Imperatorum et Regum, 4.1 (Hannover, 1906), n.567, 525–6. commented in Caro, Genova e la Supremazia, vol. 2 (15), 370–2. Details of the ceremony of reception and of the procession, which was stationed in front of the cathedral of San Lorenzo, are provided by Caro, Genova e la Supremazia, vol. 2 (15), 373–4, with references to primary

Written in Stone

44.

45.

46.

47. 48. 49. 50. 51.

sources; an informative – albeit later – account of the event was also provided by the Genoese annalist Giorgio Stella (c. 1365– 1420), Annales Genuenses, ed. Giovanna Petti Balbi, Rerum Italicarum Scriptores, vol.17.2 (Bologna, 1975), 77. MGH, Const., vol. 4.1, n.706, 682, lines 28– 32; the request had already been formally advanced to the king during a civic assembly on 14 November 1311. See MGH, Const., vol. 4.1, n.705, 681–2. Acta sunt hec in civitate Ianue in platea ante ecclesiam sancti Laurencii, in generali et pleno parlamento seu arengho popullo Ianue ibidem ad vicem preconis et sono campane congregato, presente magna parte civium Ianue tam nobelium quam popularium in multitudine copiosa. MGH, Const., vol. 4.1, n.708, 687. The new administrative and political order given by Henry VII to Genoa is described in MGH, Const., vol. 4.1, n.710, 691–3. For a modern account and appraisal, see Bowsky, Henry VII, 133–5. Bowsky, Henry VII, 137, with further references; Giovanni Villani (c. 1275–1348), Nuova Cronica, ed. Giovanni Porta, 3 vols. (Parma, 1991), vol. 2, 233. Bowsky, Henry VII, 139–203. Ibid., 203–11; Epstein, Genoa and the Genoese, 194–202. Di Fabio, ‘Sculture, Affreschi ed Epigrafi,’ 258–79. Benes, Urban Legends, 79–82. Annamaria Ambrosioni, ‘La Corona Ferrea e le Incoronazioni: certezze e ipotesi,’ in La Corona Ferrea nell’Europa degli Imperi, eds.

229

52.

53. 54.

55. 56.

Graziella Buccellati and Annamaria Ambrosioni (Milan, 1995), 1. La Corona, il Regno e l’Impero: un Millennio di Storia, xx and n.2, xx, with further references. Nicholas of Butrinto, Relatio, 894–5; MGH Leges, Leges (in Folio), 2 (Hannover, 1837), 503–10. For a modern account see Bowsky, Henry VII, 82–3. On the coronation of Frankish kings (and future emperors) in Milan and nearby Monza, see Carlo Paganini, ‘Incoronazioni di Re in Lombardia,’ in La Corona Ferrea, eds. Ambrosioni and Buccellati, vol.1, 11–43, and, in the same volume, Reinhard Elze, ‘Monza und die eiserne Krone im Mittelalter,’ 45–59. Jacopo Doria, ‘Annales Ianuenses’, 4; Jacopo da Varagine, ‘Chronica,’ 344–7. See the documents from the MGH cited in n.44–45, where Henry VII is alternately referred to as rex and princeps. Jacopo da Varagine, ‘Chronica,’ 347. The function of inscriptions as a means of social publicity in medieval Genoa is also confirmed by the banded fac¸ade of San Matteo, the family church of the Dorias. The fac¸ade was covered with a set of lengthy inscriptions in Gothic capital letters in the first half of the fourteenth century. These texts, publicly visible and strategically located at the heart of the family’s urban enclave, commemorated leading members of the family and recapitulated their victories, powerfully confirming the function of monumental writing as means of social recognition in Genoa. Muller, ‘Con¨ structing Fame in a Medieval Town,’ 8–32, and other references cited in n.9.

鵻 CHAPTER ELEVEN

PLACE, SPACE AND STYLE: CRAFTSMEN’S SIGNATURES IN MEDIEVAL ISLAMIC ART 鵼 Sheila S. Blair

The extensive use of Arabic inscriptions is one of the hallmarks distinguishing works of art created under Muslim patronage.1 Monumental inscriptions had long been used in the ancient Near East and the Roman Mediterranean, and the role of such inscriptions not only continued but also increased under Muslim patronage.2 Bold inscriptions decorate the fac¸ades, courtyards and interiors of mosques, madrasas and other large structures built by Muslims.3 Objects made for Muslims are also regularly decorated with inscription bands set amidst other geometrical, floral and even figural decoration. Indeed, Arabic inscriptions became so tied to the concepts of Muslim identity and territory that they were even imitated in pseudo-Arabic script and included in multilingual inscriptions as signs of cultural appropriation by those who had conquered the Muslims, such as ´ 4 the Byzantines in Greece, the Normans in Sicily and the kings of Castile-Leon. The Arabic inscriptions that have received the most attention in the scholarly literature are historical texts giving names and dates.5 The standard inscription describing the foundation of a building or the commissioning of an object lists the participants hierarchically. First comes the name of the ruling sovereign during whose reign the work was made, often modified by elaborate titles to underscore his vaunted position. The ruler’s name is followed in order by the names of the patron, recipient and supervisor of the work, with the date concluding the inscription. When space is available, the craftsman’s name, introduced typically by the word ʿamal (‘work of ’) or sometimes by the word .sanaʿ (‘craftsmanship of ’), can be appended at the end after the date. This format became increasingly standardised in the Abbasid period (749–1258). An early example is the 20-cm foundation text dated AH 250/864–5 CE now mounted around the base of the dome in front of the mihrab in the Zaytuna Mosque in Tunis, which names the 230

Place, Space and Style

ruling caliph al-Mustaʿin bi’llah, his client Nusayr and the craftsman Fath (see Chapter 4 for this and other comparable examples). Over time, these inscriptions became longer and more complicated, but the form remained the same. This chapter addresses a quite different type of Arabic inscriptions: short, informal ones naming the craftsman. Unlike foundation inscriptions, which were planned from the beginning as part of the decoration, these informal signatures of craftsmen were added to already finished works of art, usually inserted in inconspicuous places within the decoration, and often written in a casual style. These informal signatures also differ from graffiti, mementos added by later visitors to mark their presence, such as those added by pilgrims to early Christian sites as evidence of their personal, physical interaction with a holy place (see Chapter 2).6 These craftsmen’s signatures differ too from ascriptions, typically inscriptions added by other people to assign works to particular hands, such as those added in the margins of miniature paintings. The chapter begins with such craftsmen’s signatures on a series of more than thirty ivory containers made in the Iberian peninsula from the mid tenth to the mid eleventh century.7 They form a tight group that falls directly within the purview of this volume and the workshops that engendered it: inscriptions in the arts of the late antique and medieval Mediterranean worlds. From this core set of examples on ivory containers, the chapter expands through both time and space to show how the differences between formal inscriptions and informal signatures continued in other works of art produced later in the Maghrib, or western Islamic lands, and then more broadly as far as Central Asia and India over the next seven centuries of Muslim rule there.

CRAFTSMEN’S SIGNATURES ON IVORY BOXES FROM IBERIA

The formal inscriptions on the ivory boxes made in medieval Iberia are always written at the base of the lids, whether rectangular or round. The texts follow a standardised layout that regularly begins on the front, either at the right edge of the rectangular lids or to the left of the clasp of the round ones. The inscriptions are always carved in relief and must have been set out when designing the decorative programme for the entire container because the text was planned to jump across the space where the clasp or hinge was later affixed. A few of the formal dedicatory inscriptions on the ivory containers end with the name of a craftsman. The inscription on the so-called Fitero Casket, a prismatic box made in AH 355/965–6 CE and now in Santa Maria La Real in Fitero, for example, ends on the right side with the text ʿamal khalaf (‘work of Khalaf ’).8 The same craftsman, whose name literally means successor or scion, signed another cylindrical box now in the Hispanic Society of America (D752).9 Here, however, the formal foundation inscription around the base of the lid is replaced by a poem written as though the object itself were talking:

231

232

Sheila S. Blair

67. Signature of Khalaf between the hinges of the cylindrical box in the Hispanic Society of America D7532, c. 965. (After: Kuhnel, pl. 14, no. 28d) ¨

The sight I offer is of the fairest, the firm breast of a delicate maiden. Beauty has invested me with splendid raiment that makes a display of jewels. I am a receptacle for musk, camphor and ambergris.

Khalaf ’s informal signature on the box in the Hispanic Society uses the same formula as on the Fitero Casket, ʿamal khalaf (‘work of Khalaf ’), but is not part of the formal inscription with the poem carved in relief around the base of the lid. Instead the signature is incised between the hinges on the back of the box in the Hispanic Society (Fig. 67). A contemporary metal container, the so-called Gerona Casket now in the Treasury of the Cathedral of Gerona (inv. no. 64), confirms the subtle placement of such informal craftsmen’s signatures.10 A wooden box with a pyramidal lid that is covered with hammered plaques of silver decorated with gilt and niello, the Gerona Casket imitates ivory boxes because the overall decoration of a stylized vegetal scroll includes a strap, useless on a metal box but clearly copying the metal strap used to secure the lid to the bottom of an ivory box. The Gerona Casket carries the same type of formal foundation inscription found on the ivory containers: a relief band beginning at the right corner of the front side and running around the base of the rectangular lid. The text contains the typical commissioning inscription with the names arranged hierarchically. First come blessings to the ruler and patron, the caliph al-Hakam II. Next is the name of the recipient, his son and heir apparent Abu’l-Walid Hisham. Because Hisham was declared heir apparent on February 5 and succeeded his father on October 1 of that same year, the Gerona casket can be dated precisely to 976 CE. The band

Place, Space and Style

68. Signature of Badr and Tarif on the underside of the clasp on the Gerona Casket, 976. Treasury of the Cathedral of Gerona. (After: Dodds ed., Al-Andalus, p. 208)

ends with the information that the object was completed under the supervision of Jawdhar, known from texts as an important palace slave (al-fatan al-kab¯ır; literally, ‘the great youth/slave’) who held the posts of Superintendent of Gold- and Silversmiths and Grand Falconer.11 The craftsmen’s signature on the Gerona Casket is inscribed on the underside of the clasp affixing the lid (Fig. 68): ʿamal badr wa .tar¯ıf ʿab¯ıdihi (‘work of Badr and Tarif, his servants’). The writing tries to imitate the floriated kufic used in the formal inscription on the exterior, but the inscription is more casual in both technique and content. It is incised, not carved in relief, and the Arabic is faulty. Assuming that Badr and Tarif were two servants, then the text should have the dual ʿabdayhi, but the inscription is clearly written with the plural ʿab¯ıdihi. The grammatical mistake in the signature contrasts with the formal inscriptions in relief that use correct, formal Arabic. The signature on the Gerona Casket is also an amusing verbal and visual pun. The formal inscription on the exterior says literally that the box was made under the two hands of Jawdhar, and the placement of the signature on the underside of the clasp, which could be opened only when grasped between fingers and thumb, suggests that the two craftsmen Badr (whose name literally means ‘full moon’)

233

234

Sheila S. Blair

and Tarif (literally, ‘curious’ or ‘strange’) worked under the thumb of Jawdhar, superintendent in the caliphal atelier. The use of these informally incised signatures to show the lowliness of the artisans, especially with respect to their more important overlords and overseers, is confirmed by the group of signatures on the so-called Pamplona Casket (Museo de Navarra, inv. No. 1360-B), a prismatic box made in AH 395/1004–5 CE for the ʿAmirid chamberlain and de facto ruler of Cordoba, ʿAbd al-Malik ibn Mansur.12 The sizeable box has a long but typical foundation inscription around the rectangular base of the lid. It begins with a series of blessings asking God to extend the life of the chamberlain ʿAbd al-Malik, who is lauded with several titles including Sayf al-Dawla (‘Sword of the State’). The text further states that the box was to be made under the supervision of his chief page (al-fatan al-kab¯ır), a person who carries the epithet al-ʿAmiri, showing his affiliation with the patron ʿAbd al-Malik who belonged to the ʿAmirid family. The Pamplona Casket is the largest of all the ivory containers, measuring 38.4 × 23.7 × 23.6 cm, exactly the same dimensions as the metal Gerona Casket and three times the width of the small ivory Fitero Casket. The Pamplona Casket is also the most complicated of surviving ivory boxes, constructed of nineteen plaques, seventeen of them carved. Not surprisingly, then, it also bears evidence of a team of artisans, with six inscriptions naming five separate craftsmen. The largest inscription with a craftsman’s signature is incised on the interior of the lid, where it would be seen immediately by anyone who opened the casket. The text reads ʿamal faraj maʿa tal¯amidhihi (‘work of Faraj with his pupils/apprentices’). The signatures of five individual craftsmen, each preceded by the word ʿamal (‘work of ’), are incised in inconspicuous places in various figural scenes on five different plaques on the lid and four sides. Faraj himself signed the calf of the left leg of a lion slayer on the back right side of the lid. A second craftsman, Misbah, signed the platform of the throne beneath the large figure seated on the right side of the front (Fig. 69). A third craftsman, Khayr, signed the shield carried by a hunter defending himself against two lions in the central medallion on the back. A fourth craftsman, Rashid, signed the hindquarters of the right-hand deer being devoured by a lion in the left medallion on the left side. And a fifth craftsman, Saʿada, signed the hindquarters of the left deer being devoured by a lion in the right medallion on the right side. Contrary to doubts expressed by John Beckwith,13 all of these inscriptions are craftsmen’s signatures. They are introduced by ʿamal (‘work of ’), a word that typically introduces the name of an artisan. Furthermore, their names are typical auspicious attributes for slaves: Faraj, (‘joy’), Misbah (‘light’), Khayr (‘goodness’), Rashid (‘rightly guided’) and Saʿada (‘felicity’). The formal and informal signatures on the ivory boxes combine several traditions of how craftsmen signed their works in Umayyad Iberia.14 On the one hand, the formal signatures carved in relief at the end of the foundation inscription around the rim can be compared to those on the finely carved marble

Place, Space and Style

69. Signature of Misbah under the throne on the front of the Pamplona Casket, 1004–5. Museo de Navarra, inv. no. 1360-B. (After: Kuhnel, pl. 33, no. 55c) ¨

capitals added under the caliph al-Hakam to the congregational mosque in Cordoba or made for the nearby palace city of Madinat al-Zahraʾ.15 Each of these capitals is a tour de force of the carvers’ art, deeply carved on several levels with multiple tiers of acanthus leaves and volutes with drilled decoration. The upper rim of many bears a formal foundation inscription that follows the standard formula. The text on one column now in the Dar al-Athar al-Islamiyya in Kuwait (LNS 2 S), for example, begins with the name of the ruler, al-Hakam, and then adds that the work was carried out under his chief page (al-fatan al-kab¯ır) in the year AH 362/962–3 CE.16 In addition to the formal commissioning inscriptions, some capitals have a projecting panel on one side that continues the text with the name of the artisan, preceded by the standard ʿamal (‘work of ’).17 On the other hand, the incised technique and rounded script of the informal signatures on these ivory boxes connect them to the masons’ marks found in many

235

236

Sheila S. Blair

architectural traditions, including that of the Umayyads in Iberia. For example, some symbols as well as names, occasionally introduced by the word ʿamal (‘work of ’), are carved on the columns added to the Great Mosque of Cordoba in AH 354/965 CE.18 Most of the texts are roughly incised, even illegible, and the craftsmen named on the columns may have been illiterate, at least in classical Arabic, as was the person who incised the names on the Gerona Casket. Like masons’ marks, all of these informal inscriptions were added to finished works, as opposed to the more formal signatures in the foundation inscriptions, which were planned in advance of carving the decoration. A similar combination of formal foundation inscription and informal signature continued to be used on works of art in ivory and other precious materials produced in Cordoba into the twelfth century. The best example is the fabulous minbar formerly in the Kutubiyya Mosque in Marrakesh.19 A formal foundation inscription carved in relief around the backrest in kufic script proclaims that the minbar was begun on AH 1 Muharram (New Year’s Day) 532/16 September 1137 CE, in the city of Cordoba for ‘this venerated congregational mosque’ (h¯adh¯a al-j¯amiʿ al-mukarram), referring to the mosque in Marrakesh established by ʿAli ibn Yusuf (r. 1107–42) – the first great ruler of the Almoravids, the Berber Muslim dynasty that controlled North Africa and southern Iberia. Ordered specially from Cordoba, the minbar must have been shipped eight hundred kilometres down the Guadalquivir River and across the straits of Gibraltar and then toted on the backs of pack animals over the Atlas Mountains to its final destination in the newly founded capital of Marrakesh. Restorations to the minbar carried out in 1996 with the help of the Metropolitan Museum of Art uncovered parts of a craftsman’s signature incised on the rear jamb of the lower left frame (Fig. 70): al-ʿaz¯ız (‘the dear one’). The space before the name is missing, but may have contained the word ʿamal (‘work of ’) or .sanaʿa (‘craftsmanship of ’). Both name and placement again may be puns. The craftsman’s name, which literally means ‘Dear One’, is typical of a slave, but can also be construed as an allusion to the luxury of this extraordinary minbar, decorated with a thousand exquisitely carved panels set in strapwork bands painstakingly assembled from exotic woods and bone. The signature, furthermore, may have once been hidden behind the carved ivory capital and colonette supporting the arch, a spot that would have been grasped by the preacher as he prepared to ascend the pulpit’s stairs. Just as the slaves Badr and Tarif were under the hands of Jawdhar, so al-ʿAziz would have been under the hands of the preacher, who in pronouncing his Friday sermon would have proclaimed the name of ʿAli ibn Yusuf as ruler of Marrakesh.20 In juxtaposition to these informal craftsmen’s signatures ‘under the hands’ of their overlords and overseers, at least one ivory carver signed his name under the patron’s foot: on the front panel of the Pamplona Casket, Misbah’s name is inscribed on the platform throne (see Fig. 69). Contrary to expectation, the signature is not precisely in the centre, but is set slightly to the right so that it falls directly under the foot of the enthroned figure. Holding a braided sceptre

Place, Space and Style

70. Signature of al-ʿAziz incised on the rear jamb of the lower left frame of the minbar made in Cordoba in 1137 for the Almoravid Mosque in Marrakesh. (Photo: Jonathan Bloom)

(khayzur¯an) in his right hand and wearing a seal ring (kh¯atam) on his left, both caliphal emblems, the bearded figure is flanked by attendants carrying other appurtenances of authority such as a flywhisk, straw fan and perfume bottle. He is not, however, the actual caliph, but must represent the patron, the ʿAmirid ʿAbd al-Malik ibn Mansur.21 The location of the craftsman’s signature implies that Misbah is ‘under the foot’ of the chamberlain, thereby showing that the chamberlain also controlled the caliphal workshops.

CRAFTSMEN’S SIGNATURES ON LATER WORKS OF ISLAMIC ART

Setting the craftsman’s signature under the foot of his patron became a common and effective visual pun, used for signatures on objects created across the Muslim lands in succeeding centuries. Space here allows for only a few representative examples that show how widespread this concept became, being used in many media and at many times and places.

237

238

Sheila S. Blair

One well-known example of a craftsman’s signature beneath the foot of his patron occurs on a detached page from a manuscript of the D¯ıv¯an of Hafiz, now in the Art and History Trust collection on loan to the Sackler Gallery in Washington, D.C.22 Illustrating verses added at the top of the palace about the sighting of the new moon for the New Year and hence sometimes called ‘The Celebration of ʿId,’ the painting is signed by Sultan Muhammad ʿIraqi, court painter in the workshop of the Safavid shah Tahmasp (r. 1522–76).23 The lyric poet Hafiz had originally composed the poem c. 1376 to eulogize his patron, the Muzaffarid Shah Shujaʿ, whose first name Shah is mentioned in the opening line, by suggesting that viewers behold the new moon in the face of the king. The painter Sultan Muhammad in turn turned the verbal conceit into a visual one praising his own patron, Shah Tahmasp, who is surely the youthful prince sitting on the throne in the centre of the composition. The painter, who belonged to the royal workshop, flatters his patron Tahmasp by signing the painting at the base of the throne, literally under his foot.24 In Sultan Muhammad’s painting, the throne is set in the centre of a floral carpet, and we actually have a contemporary example of a craftsman’s signature on such a carpet in the Poldi Pezzoli Museum in Milan (d. t. 1).25 A large rectangle measuring 5.7 by 3.65 meters, the carpet is knotted in wool on silk warps and cotton wefts. The design shows a centripetal composition with a central medallion surrounded by mounted riders wearing the distinctive Safavid turban (t¯aj). The use of figures in the design shows that the carpet could not have been intended for a religious setting. Rather, it was probably designed so that the owner would have sat enthroned in the centre surrounded by his courtiers. Slight oddities in layout (the medallion is slightly off centre towards the bottom), colour (a warm madder red rather than the usual cool wine red made from expensive insect dyes), and weave (at the final few inches of both ends, a silk warp has been tied onto the cotton one so that when the carpet was cut off the loom, the ends would have made a silken fringe, changing a cheaper cotton foundation into a more costly silken one) suggest that the carpet might have been intended for an important provincial governor rather than the shah himself. A large inscription in the centre of the medallion contains a Persian couplet with the craftsman’s signature: By the diligence of Ghiyath al-Din Jami was completed This renowned work that appeals to us by its beauty.

The date is inserted vertically in numbers at the end: the year 949, corresponding to 1542–3 CE. When seated in the centre facing the lower (and slightly shorter) end, the patron’s foot would have rested on the craftsman’s name.26 The idea of signing paintings and other works of art beneath the foot of the enthroned ruler became something of a clich´e in the arts produced across the Islamic lands. A manuscript of Jami’s Bah¯arist¯an made in Bukhara in AH 958/1551 CE (Minneapolis Museum of Art 52.31) shows the device becoming routine.27 The manuscript, transcribed by Mir Husayn al-Husayni, calligrapher and

Place, Space and Style

librarian in the court atelier at Bukhara, was probably intended for the Shibanid ruler himself, because in addition to the fine calligraphy, it has fine illumination and one large double-page painting. The painting illustrates the section of the moralizing text concerning Alexander’s advice to rulers. The specific episode relates how Alexander had captured a fortress by mean of a ruse and ordered it razed. While his workmen were carrying out his instructions, Alexander’s advisors told him that a learned philosopher resided in the fortress, and so Alexander had him summoned. The double-page painting depicts this moment of Alexander encountering the philosopher, with two scenes loosely connected by a river. On the right, Alexander encounters the philosopher, while on the left his workmen raze the castle. ¯ mu˙zahhib The painting is signed on the lower step of the throne ʿamal mah.mud (‘work of Mahmud, the gilder’). A well-known calligrapher, painter and gilder, Mahmud was responsible for the transfer of styles of book art from the Timurid ateliers at Herat to the Uzbek court at Bukhara.28 He himself was taken from Herat to Bukhara in 1528, where he produced many manuscripts for the Shibanids, especially for ʿUbaydallah (r. 1533–9) and his son ʿAbd al-ʿAziz (r. 1539–51). This scene of Alexander encountering the philosopher was standard in fine Bukharan copies of Jami’s text and was used in another manuscript prepared twenty-five years earlier and now in the Gulbenkian Foundation in Lisbon.29 But in the later copy Mahmud Muzahhib deliberately acknowledges his service to his new Shibanid patrons by inscribing his name on the flat stair just below Alexander’s dangling foot. The enthroned Alexander is clearly intended as a metaphor for the Shibanid ruler of Bukhara, and the painter places himself symbolically ‘under the foot’ of his patron. The same device of a painter’s signature under the foot of a figure intended to represent the reigning sovereign was transferred to India, as shown by a splendid painting dubbed ‘The Emperor’s Choice’ (Freer Gallery of Art 1942.15a), made c. 1622–7 and once mounted in the St. Petersburg Album.30 Four verses at the corners set the subject: they relate that ‘Jahangir is Emperor in form and spirit through the grace of God; although kings may stand before him, he seeks the guidance of dervishes’. The painting depicts Jahangir, set again in an enormous halo and seated on an hourglass throne, whose base bears an invocation ‘God is Great. O Shah, may your life span a thousand years’. The throne itself is an allegory representing the grains of time. Seated on it, the ruler hands a book to a white-bearded Sufi. The two figures below the shaykh represent contemporary rulers: the Ottoman sultan and King James I of England. The entire image is generally reckoned an allegory showing Jahangir preferring a Sufi to kings. The small figure in the lower left of ‘The Emperor’s Choice’ is probably a portrait of the artist, who signed his name beneath the throne on the footstool that resembles a royal seal: work of (ʿamal ) the poor Bichitr (Fig. 71). Like his predecessor Mahmud Muzahhib, Bichitr would have been metaphorically stepped on by the ruler. Also like his predecessor, Bichitr was anything but poor and unknown. Active from around 1615 to 1650, he worked in the Mughal royal

239

240

Sheila S. Blair

71. Signature of Bichitr on the footstool beneath the enthroned Jahangir presenting a book to a Sufi, c. 1615–20. Freer Gallery 1942.15a. (Photo: John Seyller)

ateliers and was celebrated for his accomplished technique, surface brilliance and formal elegance.31 His signature is thus a visual and verbal conceit. In addition to paintings with craftsmen’s signatures beneath the foot of their overlords, we also have examples of actual objects signed on the foot. Hanging lamps are the best example. The typical piece has a wide and flaring neck, bulbous body with applied handles, and prominent foot. A small glass container for water and oil with a floating wick would have been inserted inside the lamp, and the lamp itself would have been suspended by chains from the ceiling. The lamps are typically decorated with bold inscriptions that are replete with visual and verbal puns. A key piece is the glass lamp made c. 1330 for Sayf al-Din Qawsun, cupbearer to the Mamluk sultan al-Nasir Muhammad (Metropolitan Museum of Art 17.190.991).32 The three medallions on the neck contain a cup, the emblem indicating the patron’s rank as cupbearer. The medallions are enclosed within an inscription painted in reserve against a blue ground. The text contains the so-called Light Verse (Qur’an 24: 35), saying that

Place, Space and Style

72. Signature of ʿAli ibn Muhammad al-Zammaki on a glass lamp made c. 1330 for Sayf alDin Qawsun. Metropolitan Museum of Art 17.190.991. (After: Carboni and Whitehouse 2001, p. 233)

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)].

The verse literally describes God’s light through the metaphor of the wick floating in a dish of oil inside a glass lamp.33 When the lamp was lit, the Divine Word would have glowed, a stunning realization of the Qurʾanic metaphor inscribed around the neck. The image would have been even more resplendent with dozens of such lamps hanging together. Not coincidentally, the Divine Word also circumscribes, and thus illuminates, the amiral emblem of the cup, painted in thick red against an orange-yellow ground. The patron’s identity is confirmed in the name and titles painted in cobalt blue around the globular body. The inscription identifies the lamp as made for his excellency, the cupbearer (al-s¯aq¯ı) Sayf al-Din Qawsun, who carries the epithets al-malik¯ı and al-n¯a.sir¯ı, showing his affiliation with the Mamluk sultan al-Malik al-Nasir Muhammad. The lamp can be dated c. 1330, the time of Qawsun’s active patronage in Cairo. These two inscriptions with the Light Verse and the patron’s name are typical of Mamluk mosque lamps (although often written in reverse techniques, with the patron’s name in reserve and the Qurʾanic text in blue), but a third inscription around the foot of Qawsun’s lamp is more unusual (Fig. 72). Written in reserve in a rather loose style, it contains the name of an artisan: work of (ʿamal ) the poor

241

242

Sheila S. Blair

slave ʿAli ibn Muhammad. He carries an epithet that has been read in various ways. One suggestion is al-Barmaki, suggesting that he may have belonged to the Barmakid family of Iranian origin that rose to power under the Abbasids in eighth- and ninth-century Baghdad. Such a reading is unlikely, because the epithet was uncommon in Mamluk times and the reading does not follow the shape of the letters. A much more plausible reading of the craftsman’s epithet that fits both with the shape of the letters and the many puns in the lamp’s decoration is al-Zammaki. The epithet derives from the verb zammaka, a term used in the signatures of several contemporary Qurʾan manuscripts to mean outlining. For example, in a dispersed seven-volume manuscript transcribed by Muhammad ibn al-Wahid for the sultan Baybars in AH 704–5/1304–6 CE, the illuminator Aydughdi outlined the gold letters that had been calligraphed by Muhammad ibn al-Wahid.34 In a single-volume Qur’an manuscript dated AH 713/1314 CE, a person named ʿAli ibn Muhammad outlined the gold letters that had been transcribed by the Ayyubid prince and calligrapher Shadhi ibn Muhammad for the Mamluk sultan al-Nasir Muhammad.35 Such outlining was necessary, because the calligraphy in both these manuscripts was painted with adhesive and then covered with gold leaf so that the gilt letters had rough edges.36 Thus it is no surprise that the term zammaka (‘to outline’) typically occurs in Mamluk texts following dhahhaba (‘to gild’). The same craftsman ʿAli ibn Muhammad also painted his name in gold on the foot of a hanging lamp made for the mosque founded by the amir Sayf al-Din Ulmas in AH 730/1330 CE (Museum of Islamic Art, Cairo, no. 3154).37 The patron was the chief chamberlain (h.a¯ jib) to the same Mamluk sultan al-Nasir Muhammad. The craftsman’s epithet on Ulmas’s mosque lamp is written a¯ mak¯ı, but Gaston Wiet suggested sensibly that the medial l¯am-z¯al had just been omitted in al-Zammaki. ʿAli ibn Muhammad al-Zammaki’s signatures on these two lamps are the only such signatures on the many Mamluk glass lamps that survive, and he must have been an important person to have placed his name there. The signature in the single-volume Qurʾan manuscript copied by Shadhi may provide a clue to his identity. The outliner there could well be the same person, because he has the name ʿAli ibn Muhammad, but has different epithets: the draftsman (al-rass¯am) known as the left-handed (al-aʿsar). ʿAli ibn Muhammad was part of the increasingly complex and hierarchically arranged scriptorium associated with the Mamluk sultans in Cairo. Thus, Aydughdi, who had outlined the letters in the seven-volume Qur’an manuscript made for Sultan Baybars, moved up the ladder to become an illuminator in the single-volume manuscript made a decade later for Sultan al-Nasir Muhammad. In addition to these two hanging glass lamps signed by ʿAli ibn Muhammad al-Zammaki, one hanging ceramic lamp bears a signature: a striking piece painted in blue, green and black and dated AH 956/1549 CE (British Museum

Place, Space and Style

1887.0516.1).38 Discovered in the mid nineteenth century on the Haram in Jerusalem, it must have been part of the renovations to the Dome of the Rock carried out under the Ottoman sultan Sul¨eyman the Magnificent (r. 1520–66). As with other signed objects, the inscriptions on the hanging ceramic lamp are full of puns.39 The band around the neck contains a h.ad¯ıth saying that the believer in a mosque is like a fish in water, whereas the false believer in a mosque is like a bird in a cage. The larger band around the body of the lamp implores God, praised in all His works as the supporter of prosperity and possessor of hidden graces, to save us from that which we fear. The large band at the bottom of the body has another h.ad¯ıth cited on the authority of Abu Hurayra, a companion of the Prophet Muhammad, that God the Highest is good and receives none but the good. These inscriptions show that the lamp was a pious gift to a mosque or shrine. Even more intriguing is the fragmentary rhyming text on the foot of the ceramic lamp. Written in cartouches in a round naskh script, it refers to the lake of Iznik, where this type of fine pottery was made, and to a local mystic Es¸refzade Rumi, who is recorded as being the patron saint of Iznik potters in the following century. The text plays on the meaning of dary¯a as lake or ocean in the physical sense and as the Ocean of Goodness in mystical terms. The inscription also includes the date (AH Jumada I 956/June 1549 CE) and the name of the craftsman, the decorator (naqq¯ash) Musli, who is described as the poor (al-faq¯ır) and the base (al-h.aq¯ır). These rhyming epithets are once again puns, because the work is anything but poor and base. This hanging lamp signed by Musli in AH 956/1549 CE and a fragmentary companion also in the British Museum (G 1983.57)40 attest to the establishment of polychrome blue and green decoration in the so-called saz floral style in the 1540s. To judge from the fine quality of these and related works, Musli and his associates worked at the premier ceramic workshop of the time, making some of the most striking pottery ever produced in the Islamic lands.

THE SIGNIFICANCE OF CRAFTSMEN’S SIGNATURES

What then can we learn from craftsmen’s signatures on works of Islamic art? Signatures occur on the finest objects, those made for princes or their courtiers of the most precious materials, from ivory to enameled glass and polychrome pottery, as well as deluxe manuscripts and elaborate single-page paintings for albums. On these objects, informal signatures differ from more formal ones by place, space and style. They are not part of the foundation or commissioning inscription, but were added in an inconspicuous location after the rest of the work had been completed. On ivories, they are incised, rather than carved in relief; on other objects they are often set off in a cartouche and written in a different and looser script. Studying these informal texts tells us much.

243

244

Sheila S. Blair

Signatures can show how an industry developed and a workshop functioned. Whereas an individual craftsman such as Khalaf could carve a small monolithic box such as the Fitero Casket dated AH 355/965–6 CE, a team was needed to carve the large multipart Pamplona Casket dated AH 395/1004–5 CE. From the informal signatures on the latter work, we can infer that Faraj was the master craftsman, who probably laid out the entire design. He himself carved the lid, while distributing individual plaques to at least four other apprentices in his workshop (Misbah, the front; Khayr, the back; and Rashid and Saʿada, the left and right sides respectively). The informal signatures thus suggest a method of production, one that should be followed up with close examination of the box to see whether it is possible to detect minute differences in individual hands. The signatures of these ivory boxes also suggest that the ivory-carving industry, which sprang seemingly ex novo in Iberia in the mid tenth century, expanded substantially in the intervening generation or two between these two dated ivory boxes. The same type of expansion and specialization occurred in the Mamluk scriptorium in the early fourteenth century, as individual specialists became responsible for calligraphy, gilding, outlining and the like. The Mamluk scriptorium also provided for upward mobility, as a craftsman like Aydughdi could move from one specialty to another more prestigious one. Similar specialization in the scriptorium continued elsewhere in later times,41 and by the Mughal period, the book atelier was so large that responsibilities were divided among individual painters responsible for composition, faces and so forth. Signatures also show us that the crafts evolved differently from the thirteenth century onwards, when there was an increasing divergence between the artisan who made the object, whether potter, smith or the like, and the person who designed or decorated it. Thus, the Ottoman craftsman Musli is identified as naqq¯ash (‘designer’). He did not make the lamp; rather, he designed the decoration. Similarly, the Mamluk ʿAli ibn Muhammad al-Zammaki was not a glassmaker but a calligrapher/decorator. Their signatures show how designs worked out in the court scriptorium were executed in different media, whether books, ceramics, glass or even metal, as on the superb incense burner made for al-Nasir Muhammad at the same time as the glass lamps.42 One major reason for this change in the organization of craftsmanship was the increasing role of paper, which allowed designs to be transferred from one medium to another and therefore for dynastic or regional styles to develop.43 What is less clear is why some craftsmen signed their works, but others did not. The carver Khalaf is a good example. He is the only artisan from the first generation of ivory carvers in Iberia to have signed his works, but he was not very good. The carving on both of his signed boxes, the rectangular Fitero Casket and the round box in the Hispanic Society, is mediocre, especially in comparison with that on the stupendous contemporary box made for the prince al-Mughira in

Place, Space and Style

AH 357/967–8 CE (Louvre, AO 4068) or on the stunning anonymous box recently acquired by the David Collection in Copenhagen (no. 5/2002), probably by the same hand.44 The latter two are both unsigned. We can understand better why craftsmen signed their works in the way that they did. Signing a work at the end of the foundation inscription put the artisan on the lowest rung in the hierarchy of names, as did Bichitr’s position as the smallest figure in the lower left of the ‘Emperor’s Choice’. But if there was no room in the foundation text or no foundation text at all, as in the case of the round box in the Hispanic Society, incising the name in an inconspicuous place was another way of underscoring the artisan’s humility and servitude. To judge from their names, many of these artisans were non-Muslims or recent converts. This is the case with many of the ivory carvers in the Umayyad palace workshops whose names such as ‘Light’ or ‘Felicity’ are the typical attributes of slaves. The same is true for al-ʿAziz, the woodcarver of the Almoravid minbar, and for the Mughal painter Bichitr, whose name is an Arabized version of the Sanskrit Vicitra (‘splendid’). By signing the work below the thumb rest or under the foot of the royal recipient, the artisan was acknowledging his servitude; this was especially clear in the case of Khalaf, whose name derives from the same Arabic root as the title of his overlord, the Umayyad caliph (Arabic: khal¯ıfa). The inconspicuous placement of the signatures – where the object would have been grasped in the hands, beneath the foot of the patron, or on or under the foot of the object – parallels the epithets of the artisans, who are often labelled ‘poor’ or ‘humble.’45 The adjectives play on the craftsmen’s humility in the face of royalty or even of God. Their signatures are verbal puns, in the same way that their names and Bichitr’s portrait of himself were visual ones. Like the names and the adjectives that modify them, the placement of these informal signatures was a double entendre. Reading them often requires close inspection of the object, and deciphering them can be seen as a sort of reward for close study. Similarly, the signatures reflect not only the humility of the craftsman toward the patron but also function as a sign of his status as part of the caliphal or royal workshop. Such a double meaning seems to have been understood by patrons, at least in such later examples as the individual paintings and drawings made for the Safavid shah ʿAbbas in Iran in the early seventeenth century. Many of these works were signed by his court artists, who often used the word ‘poor’ (faq¯ır) or ‘least’ (kamtar¯ın) in their signatures. Before the paintings and drawings were mounted in albums, the shah stamped them with his ring seal, using a rectangular or oval seal like the one evoked by the footstool in Bichitr’s painting of the ‘Emperor’s Choice’. On the figures of the handsome youths, Shah ʿAbbas deliberately stamped his seal under the feet of the model, perhaps to express admiration not only of the craft but also of the image,46 thereby acknowledging in return his ‘indebtedness’ to them.

245

246

Sheila S. Blair

These informal signatures, even if sometimes roughly scratched and hidden in the decoration, show that the iconography of text and that of image on these works of art deserve close reading together.

NOTES

1. A basic introduction to the subject is Sheila Blair, Islamic Inscriptions (Edinburgh, 1998). 2. See, for example, Chapter 1 in this volume. 3. See Chapters 6 and 7 in this volume. 4. See Chapters 5, 6 and 8 in this volume. 5. The most readily available collection of Arabic historical inscriptions is the R´epertoire ´ chronologique d’´epigraphie arabe, eds. E. Combe, J. Sauvaget, and G. Wiet (Cairo, 1936–), now supplemented by an online ´ version, Thesaurus d’Epigraphie Islamique, available by subscription at www.epigraphieislamique.org. 6. See Chapter 2 in this volume. 7. Ernst Kuhnel, Die islamischen Elfenbeinskulp¨ turen VIII–XIII. Jahrhundert, 2 vols. (Berlin, 1972). Some are published in Jerrilynn Dodds, ed., Al-Andalus: The Art of Islamic Spain (New York, 1992), with their Arabic texts. The ivories were also the subject of a 2003 symposium in Copenhagen, the papers of which are published as The Ivories of Muslim Spain, eds. Kjeld von Folsach and J. Meyer, Journal of the David Collection 2 (2005). My article there, Sheila Blair, ‘What the Inscriptions Tell Us: Text and Message on the Ivories from al-Andalus,’ Journal of the David Collection 2 (2005), 75–100, discusses the inscriptions. See also the two-volume catalogue by Ángel Gal´an y Galindo, Marfiles Medievales del ´ Islam (Cordoba, 2005). 8. Kuhnel, Die islamischen Elfenbeinskulpturen, ¨ no. 23; Journal of the David Collection 2 (2005), cat. no. 7c and pl. 7c; Gal´an y Galindo, Marfiles Medievales, II: 28–29, no. 02005. In the commissioning inscription around the lid, the recipient is identified as al-ah.abb wall¯ada, a phrase that has engendered much confusion. Following E. L´evi-Provenc¸al, Inscription sarabes d’Espagne (Paris and Leiden, 1931), 186–7, the phrase is usually taken to refer to al-Hakam’s consort Subh, who received two other ivory boxes dated AH 353/964 CE, on which she is called al-sayyida umm ʿabd al-rah.m¯an (the lady, the mother of ʿAbd al-Rahman). See, for example, Manuela Mar´ın’s biography of Subh in the Encyclopaedia of Islam, 2nd ed.; Francisco PradoVilar, ‘Circular Visions of Fertility and

Punishment: Caliphal Ivory Caskets from alAndalus,’ Muqarnas 14 (1997), 20–2; Blair, ‘What the Inscriptions Tell Us,’ 90; Gal´an y Galindo, Marfiles Medievales, II: 27–28; and Sophie Makariou, ‘The al-Mugh¯ıra Pyxis and Spanish Umayyad Ivories. Aims and Tools of Power,’ in Umayyad Legacies: Medieval Memories from Syria to Spain, eds. Antoine Borrut and Paul M. Cobb (Leiden, 2010), 314–15. The phrase al-ah.abb wall¯ada is taken to mean ‘the most beloved, she who bears children’ or ‘the most beloved progenitrix,’ and the boxes are interpreted as presents to her in honour of her childbearing capabilities, for in addition to ʿAbd al-Rahman, born in 962, the year indicated on the first set of boxes, she gave birth to another son Hisham, born in 965, the year before the second set of boxes. Based on Arabic grammar, however, this identification is not possible, because as Wheeler Thackston confirmed, that reading would require al-ah.abb al-wall¯ada, with the definite article before both words. Instead, the phrase must be translated as ‘the most beloved Wallada’ and must refer to Wallada, one of four daughters of ʿAbd al-Rahman III by his favourite wife Maryam and sisters of al-Hakam (see Makariou, ‘al-Mughira Pyxis,’ 314 and n.3). Antonio Vallejo Triano pointed ˜ Jim´enez had out to me that Manuel Ocana already made this identification in his mono´ graph El cufico hispano y su evoluci´on (Madrid, 1970), 37, and Arabic grammar shows that he is correct. This identification of the recipient as Wallada confirms that boxes were made for many females at the Umayyad court, but undercuts the idea of their being baby presents and explains why the years on the boxes do not correspond to the years of the children’s births. 9. Kuhnel, Die islamischen Elfenbeinskulpturen, ¨ no. 28; Journal of the David Collection 2 (2005), cat. no. 14; Gal´an y Galindo, Marfiles Medievales, II: 30–1, no. 02006, 10. Al-Andalus, ed. Dodds, no. 9; Jonathan Bloom and Sheila Blair, Islamic Arts (London, 1997), 253–6; Journal of the David Collection 2 (2005), fig. 41.

Place, Space and Style

11. Makariou, ‘The al-Mugh¯ıra Pyxis.’ 12. Kuhnel, Die islamischen Elfenbeinskulpturen, ¨ no. 35; Al-Andalus, ed. Dodds, no. 4; Bloom and Blair, Islamic Arts, 256–9; Journal of the David Collection 2 (2005), cat. no. 20; Gal´an y Galindo, Marfiles Medievales, II: 48–52, no. 02018. 13. John Beckwith, Caskets from Cordoba (London, 1960), 29. 14. Sheila Blair and Jonatahn Bloom, ‘Signatures on Works of Islamic Art and Architecture,’ Damaszener Mitteilungen 11 (1999) [2001], 49–66. 15. These capitals were the subject of a paper presented by Antonio Vallejo Triano at the Fourth Biennial Hamad bin Khalifa conference in Doha in October 2011, ‘The Architectural Decoration of Capitals from the Umayyad Caliphate of al-Andalus: The Example of Madinat al-Zahraʾ’ which was published as ‘Two Capitals from the ´ Umayyad Caliphate of Cordoba’, God is Beautiful and Loves Beauty: The Object in Islamic Art and Culture, eds. Sheila Blair and Jonathan Bloom (London, 2013), 103–21. 16. Illustrated in Al-Andalus, ed. Dodds, no. 39. 17. The name on the capital in Kuwait is hard to read; the letters seems to read ʿamal fath. alasir ʿabd, but the text makes no sense. 18. Authorities in charge of the building’s restoration have made plaster stamps of these signatures and mounted them for display on the back wall there. 19. Al-Andalus, ed. Dodds, no. 115. Jonathan Bloom et al. The Minbar from the Kutubiyya Mosque (New York, 1998). 20. For other craftsmen who signed the contemporary stucco work in the Qarawiyyin mosque in Fez, see H. Terrasse, La Mosqu´ee al-Qaraouiyin a` F`es (Paris, 1968), 79–80 and pls. 47 and 51. 21. Sophie Makariou, ‘Quelques r´eflexions sur les objets au nom de ʿAbd al-Malik ibn alMansur, ˆ Arch´eologie islamique 11 (2001), 47– 60; Blair, ‘What the Inscriptions Tell Us,’ 80–1. 22. Abolala Soudavar, Art of the Persian Courts (New York, 1992), no. 59. 23. Brief biography in Grove Encyclopedia of Islamic Art and Architecture, eds. Jonathan Bloom and Sheila Blair (New York, 2009): ‘Sultan Muhammad.’ 24. Sheila Blair, Text and Image in Medieval Persian Art (Edinburgh, 2014), 240–1. 25. Hunt for Paradise: Courts Arts of Safavid Iran 1501–1576, eds. Jon Thompson and Sheila Canby (Milan, 2003), no. 12.19.

247

26. Richard Ettinghausen, ‘The Boston Hunting Carpet in Historical Perspective,’ Boston Museum Bulletin 69, nos. 355 and 356 (1971), 371–2. The entry in Hunt for Paradise, eds. Thompson and Canby, no. 12.19, contains the perceptive comments about the attribution, but suggests that Ettinghausen had suggested the shah’s posterior would have rested on the signature. Actually, it would have been his foot. 27. Eleanor G. Sims, ‘Persian Miniature Painting in the Minneapolis Institute of Arts,’ Minneapolis Institute of Arts Bulletin 71 (1975), 51– 73. 28. Brief biography in Grove Encyclopedia of Islamic Art and Architecture: ‘Mahmud Muzahhib.’ 29. Illustrated in Survey of Persian Art, eds. A. U. Pope and P. Ackerman (R Osaka, 1977), pl. 888. 30. Richard Ettinghausen, ‘The Emperor’s Choice,’ De Artibus Opuscula LX: Essays in Honor of Erwin Panofsky, ed. M. Meiss (New York, 1961), 98–120. The device was also used in the Deccan: see K. Overton, ‘A Collector and His Portrait: Book Arts and Painting for Ibrahim ʿAdil Shah II of Bijapur (r. 1580–1627),’ (PhD dissertation, University of California, Los Angeles, 2011), 224–5, 239–41. 31. Brief biography in Grove Encyclopedia of Islamic of Art and Architecture: ‘Bichitr.’ 32. Stefano Carboni and David Whitehouse, Glass of the Sultans (New Haven, 2001), 232– 4, no. 116. 33. Sheila S. Blair, ‘Written, Spoken, Envisioned: The Many Facets of the Qur’an in Art,’ The Qur’an in Art, ed. F. Suleman (London, 2007), 280–2. ¯ (Lon34. David James, Qur’ans of the Mamluks don, 1988), 66–7 and no. 1; Sheila Blair, Islamic Calligraphy (Edinburgh, 2006), 345–9 and fig. 8.13. James’s definition of the term zammaka to mean the laying on of colours is wrong. 35. James, Qur’ans of the Mamluks, no. 6; Blair, Islamic Calligraphy, 326–9 and fig. 8.5. 36. I thank Joan Cummins, conservator at the Boston Museum of Fine Arts; Laura Weinstein, curator there; and Jonathan Bloom for their help in looking at Qurʾan manuscripts and verifying the use of gold leaf in this way. 37. G. Wiet, Catalogue g´en´eral du Mus´ee arabe du Caire: Lampes et bouteilles en verre e´ maill´e (Cairo, 1982), 122–4, no. 3154. 38. J. Michael Rogers and Rachel M. Ward, ¨ Suleymanthe Magnificent (London, 1988), no. 148; Nurhan Atasoy and Julian Raby, Iznik:

248

39.

40. 41.

42. 43.

44.

Sheila S. Blair

the Pottery of Ottoman Turkey (London, 1989), 135–42 and figs. 239 and 355; Bloom and Blair, Islamic Arts, 394–5, cover and pl. 213. C. Drury Fortnum, ‘On a Lamp of “Persian Ware” Made for the Mosque of Omar at Jerusalem in 1549,’ Archaeologia 42 (1869), 394–5. Atasoy and Raby, Iznik, no. 238. M. S. Simpson, ‘The Making of Manuscripts and the Workings of the Kitab-khana in Safavid Iran,’ The Artist’s Workshop, ed. P. Lukehart (Washington, DC, 1993), 105–21. Blair, Islamic Calligraphy, fig. 8.9. Jonathan Bloom, Paper before Print: The History and Impact of Paper in the Islamic World (New Haven, 2001). For the former, see Kuhnel, Die islamischen ¨ Elfenbeinskulpturen, no. 31; Dodds, ed., AlAndalus, no. 3; Journal of the David Collection 2 (2005), cat. no. 11, and most recently,

Makariou, ‘The al-Mugh¯ıra Pyxis’. For the latter, see Journal of the David Collection 2 (2005), cat. no. 10. 45. An ivory chessman in the form of a nobleman riding in a howdah on an elephant accompanied by warrior and horsemen in the Cabinet des M´edailles in Paris is signed on the underside of the base ‘work of (ʿamal ) Yusuf al-Bahili’; Blair and Bloom, ‘Signatures,’ 52 and n.21. 46. A. Adamova, ‘On the Attribution of Persian Paintings and Drawings at the Time of Shah ʿAbbas I: Seals and Attributory Inscriptions,’ Persian Painting from the Mongols to the Qajars, ed. Robert Hillenbrand (London, 2000), 19– 38 and ‘Muhammad Q¯asim and the Isfahan School of Painting,’ Society and Culture in the Early Modern Middle East: Studies on Isfahan in the Safavid Period, ed. A. Newman (Leiden, 2003), 193–212, esp. n.1.

AFTERWORD: RE-VIEWING INSCRIPTIONS 鵼 Antony Eastmond

The inscriptions discussed in this volume were created over a period of two thousand years, by societies spread over three continents and separated by the Mediterranean and Arabian Seas. They span a distance from west to east of more than seven thousand kilometres. Most of the inscriptions discussed were made for religious settings, but those religions were extraordinarily diverse – from Babylonian cults and Zoroastrianism in the Iranian world, to the Greco-Roman pantheon of gods around the Mediterranean, and from Christianity to Islam. Other inscriptions belong to the secular realm, but this arena covers an equally broad field – on the one hand mighty texts that express the grandeur of power and demonstrate the authority of rulers within their palaces and across entire landscapes, and on the other delicate inscriptions adorning small-scale objects that evoke the intimacy of love and the delight in beauty of their owners. Yet other inscriptions were hastily scratched or incised by visitors, often illicitly, seeking to leave a physical testament to their presence in particular locations. Seen in these terms, the chapters in this volume reveal the remarkable variety of contexts in which inscriptions appear, as well as the diversity of reasons that underlay their creation as scribes, masons and carvers gave words a tangible, three-dimensional physicality that – in most cases, but by no means all – has ensured that they have survived to speak to viewers and readers across the centuries. The chapters in this volume, however, have been concerned not so much with that apparent diversity of function and subject matter as with the visual strategies that were employed to enable these texts to express much more than the semantic content of their words. Here it is possible to find many areas of overlap and convergence. In some cases the convergences are explicit within a chapter, no more so than in Chapter 1’s presentation of the ways in which the distinct epigraphic practices of the Achaemenid and Hellenistic rulers of Iran were appropriated by their Sasanian successors. However, in most cases the ties and overlaps must be read between and across the chapters in this volume.

249

250

Antony Eastmond

MULTILINGUAL INSCRIPTIONS

The need for a cross-cultural approach to the subject of viewing inscriptions is most evident in the frequent appearance of multilingual inscriptions in all the regions, periods and cultures that this volume covers. Chapters 1 and 8 focus on multilingual inscriptions, but multilingual inscriptions were produced by all the cultures represented in the volume. Bilingual Greek and Roman texts were produced in the Byzantine world;1 bi- and trilingual texts, involving Arabic, Armenian, Persian and Syriac were made in the ethnically and linguistically diverse world of eastern Anatolia and the Caucasus;2 and occasionally quadrilingual texts survive, such as the 1148 epitaph of Anna, mother of the priest Grisandus in Sicily composed in Greek, Latin and Arabic (written twice, once in Hebrew letters), or the Georgian, Armenian, Persian and Uighur graffito scratched in 1352 on the wall of the refectory at the remote Udabno monastery in the Gareja Desert in Georgia (on the current border with Azerbaijan).3 All such texts raise questions about hierarchy. The juxtaposition of languages inevitably involves power relationships, determined by which text is dominant. This feature is most easily demonstrated visually through the prominence and placement of each text. However, hierarchy can be disguised through a fiction of equivalence in which inscriptions appear to be given equal value, but this needs to be recognised as much as a visual and political ploy as the obvious assertion of the prominence of one language over another. The uniform appearance of such texts disguises whether they actually say the same thing or whether each speaks to its own audience. How many people could have recognised these differences remains a moot point, but the possible discrepancies between them highlight the different roles played by the content and the appearance of inscriptions. Ultimately the significance of all multilingual inscriptions rests on issues of translation: first how a text in one language is converted into the different languages around it, and then how that oral or written text is translated into the medium of the inscription, whether incised, cast or painted. Were the same people involved at every stage, or – in a form of written Chinese whispers – were meanings subtly altered by the pragmatic choices made by the craftsmen or the restrictions imposed on them by their own ignorance of foreign tongues or by more practical issues of space or medium? The chapters in this volume all imply the existence of a ‘designer’, someone who has conceived the message to be conveyed in each inscription and with an overall vision for how it should appear, both in terms of content and visual form. But the incursion of craftsmen into the visual field of the Andalusian ivories (Chapter 11) is a reminder of the intermediary role played by such workers in executing this vision. The level of education of these craftsmen is unclear. In the case of the ivory workers, the precision and artistry with which they carved both the principal texts and the surreptitious appearance of their own names elsewhere required a high level of literacy. It is clear that these ivory workers could

Afterword

read and understand the texts they were required to transpose, but in other cases it is less clear. For multilingual inscriptions it is much less likely that the carvers had equal command of all the texts. However, we might also doubt whether the patrons could have read all the languages they chose to be commemorated in. As a result such inscriptions work through expectation and imagination rather than through actuality: readers of one language become viewers of another, and their assumptions about the contents of those unreadable texts are largely determined by their appearance in relation to the inscriptions they can read. As Chapter 8 shows, this allows divergent messages to be conveyed whilst maintaining a fiction of convergence. Even in monolingual inscriptions the choice of language can be important: the Arabic text of the inscriptions in the Qaratay Madrasa in Konya, discussed in Chapter 7, was distinct from the Persian literary language and everyday spoken Turkish of its patron. In contemporary palaces in Seljuk Turkey, the decorative scheme often revolved around Persian inscriptions, revealing the way in which the choice of inscriptional language helped construct the different social and religious spaces of the Seljuk elite.4 THE MEDIATION OF CRAFTSMEN

The mediation of craftsmen affected not just the ways in which different languages could be presented but also had an impact on every aspect of inscriptions. The awkward breaks and unbalanced placement of the inscriptions at Kumurdo in Georgia (Chapter 4) suggest that in some cases the compromises forced on the artists and in others their desire to produce symmetrical carvings both triumphed over the content of the original texts and any hopes on the patron or designer’s part for how they might be inscribed. At Noravank’ in Armenia (Chapter 9), the opposite was the case. It was the quality of the artists that potentially undermined the text: their brilliantly inventive visual solution to the depiction of the Word of God created new visual meanings but simultaneously threatened to obscure their verbal contents (as the modern disagreements about how to read the inscription eloquently show). The graffiti discussed in Chapter 2 provide a further level of non-official intervention in the visual field. Whether tacitly approved by the church authorities or not, the marks left by pilgrims at sites around the Christian world indelibly altered the religious space and transformed the visual and spiritual experience of later visitors. INSCRIPTIONS AND ENTRANCES

A further theme that emerges from many of the chapters in this volume concerns the placement of inscriptions. In some cases the location of inscriptions

251

252

Antony Eastmond

gives them performative value, as readers are forced to move round the texts (or in the case of ivories to manipulate the objects in their hands) to read them. This movement then instilled a ritual element to the encounter, with a processional quality whose order can be determined by the placement and order of inscriptions. It allowed artists and designers to create architectural and iconographic paths that the reader had to follow. In architecture inscriptions are associated with particular places, frequently appearing in conjunction with doors and windows (see Chapters 2, 4, 6, 7, and 9). To some extent such placements are common sense: an inscription at a door must be encountered by anyone seeking to enter that building. Such inscriptions act as textual introductions, establishing a frame of mind and conditioning the reader before he or she enters the building. At the Qaratay Madrasa that frame of mind concerned the behaviour of the entrant; at Kumurdo it was to establish the memorial and commemorative nature of the building; in Achaemenid palaces it was to instil awe in the supplicant before he entered the shah of shah’s semi-divine presence. The relationship between texts and windows draws on a very different idea. Here, the placement of inscriptions is concerned not with the approach of the individual, but with the relationship between exterior and interior spaces. These inscriptions become guards, permanently keeping watch over the potentially vulnerable openings into the precious interior. The apotropaic qualities that are ascribed to such texts, usually regardless of their actual content, indicate the power embedded in monumental writing. This idea that the written word had an intrinsic power that came simply from its physical presence, rather than from any particular aspect of its contents, had both positive and negative connotations. Fragments of inscriptions believed to have been made by early Christian martyrs were carefully treasured and reused in twelfth-century Rome, even to create papal thrones.5 At the same time, Roman inscriptions of a similar age were treated more warily in Ankara where they were placed in the city walls (again near entrances), but with the writing carefully positioned upside down to deny the writing its latent power. In thirteenth-century Cairo pharaonic inscriptions were reused as the thresholds of doorways into mosques and madrasas. Worshippers were forced to tread on the hieroglyphs as they entered as a symbol of the vanquishing of the old religions by Islam in Egypt.6

INSCRIPTIONS AND AUTHORITY

Inscriptions across all cultures relate to the same texts written in other formats – whether biblical or Qur‘anic quotations, memorials inscribed in prayer lists, or transcriptions of laws or royal proclamations. There was clearly an understanding that inscriptions looked beyond their own immediate existence, to a broader world of writing, deriving ultimately from the spiritual authority of religious

Afterword

texts. They linked this world with the divine realm, providing a concrete embodiment of the word of God, serving as a reminder of the impermanence of this world and the permanence of that to come. The authority of inscriptions, embedded both in beliefs about the status of the written word but also in other physical qualities, such as the scale and perceived durability of writing, had effects that can be traced throughout this volume. For example, Genoa’s imaginary history was made more tangible by its appearance in written form on the walls of the city’s cathedral (see Chapter 10). Inscriptions could set out legal and quasi-legal texts, but the monumental nature of their appearance transformed such texts into representations of that law. There are close comparisons with the way in which, in the Byzantine world, images of the emperor enjoyed the same status as the ruler himself. This made the ruler ubiquitously present even though the man himself could only ever be in one place. Inscriptions could also enjoy a similar reduplicative function. They too made rulers and patrons, and their official pronouncements, prayers and testaments, ever present. Even if the inscriptions may actually have deferred to texts written in other media (inscribed in wax or clay tablets, written in ink on parchment or papyrus), the scale, permanence and visibility of the monumental versions surely eclipsed that of the original, even if the lawyers and archivists whose livelihoods depended on the originals may have disagreed. The placement of such legal texts had the power to transform their surroundings. Gathered together on individual walls like a permanent noticeboard they could change the status of the building on which they stood, turning the wall into a document of memory, commemoration or power. In turn, the building could begin to transform the urban or rural context around it. These inscriptions placed the presence of authority permanently in front of the population, although whether such signs acted as a reassurance about the existence of a power to maintain social order or a repressive reminder of that power’s omnipresence must have depended on the viewer. The craftsmen’s inscriptions discussed in Chapter 11 – hidden behind locks, placed beneath rulers’ feet, hidden on the legs of carved animals – play on these ideas of inscriptions and the law in a different way. They show the lure of inscriptions, deriving from their status as quasi-legal markers of patronage and ownership and from their associations with permanence and memory. These craftsmen’s desire to remember themselves in inscriptions, albeit in a more surreptitious way, displays their wish to appropriate this commemorative quality for themselves. These inscriptions reveal not only the changing status of craftsmen in society in different cultures but also their view of themselves. In the Islamic world such men were clearly considered as individuals with reputations; in the Caucasus artists’ names are also frequently inscribed on monuments and works of art. In contrast, in the Byzantine Empire and the Latin West, named artists remain much rarer before the fourteenth century. These variations show how much could change

253

254

Antony Eastmond

in one geographical area: the Andalusian ivories were produced three hundred years before anonymous workers produced the tomb of Fernando III at Seville. The makers of each lived in the same region, but the status of artists in each age was clearly very different. The world of inscriptions also reveals how fragile are these ideas of permanence. Whether pilgrims ever expected their graffitied prayers to last for so long (or even to be studied by future generations) is unlikely, but their creation was dependent on an idea that they would be read by later visitors, who would then bring their names and prayers once more before God. Chapter 3 highlights the danger of visual silence: buildings bereft of inscriptions, but with clear spaces for them to exist. Does space for an inscription imply an expectation to find one there? And if it is absent, should this be read as a significant fact of omission, rather than simply an absence? The evidence accumulated in this volume of the status of inscriptions and their importance as tools of political power supports the view that the former is the case. It reminds us to look again at all the monuments in this volume to wonder whether there are similar silences there too. Are there instances where the desire of later owners has been to forget the past, rather than – as in the case of the partially obliterated inscription at Sfax – to remember to forget?

PSEUDO-WRITING

Although most of the visual functions of inscriptions raised in individual chapters in this book find parallels in other chapters, there are cases in which visible similarities prove to be misleading. The Cappella Palatina, the Qaratay Madrasa and Hosios Loukas all employ pseudo-kufic (or pseudo-Arabic) writing. This is evidently a form of fictive writing that derives from the Islamic world, but was taken up by both Christian and Muslim cultures around the Mediterranean. However, the three examples here all use this form of writing to very different ends. What at first sight appears to be a pan-Mediterranean feature is actually revealed to be very culturally specific. It is also a feature the meaning of which, as Chapter 5 shows, could change over time.

INSCRIPTIONS AND IDENTITY

A final theme that pervades all the chapters is that of the relationship between inscriptions and identity. The addition of an inscription to a building, a mountainside, a tomb or an ivory helped create a social context for that object. It did so through inclusion and exclusion. It divided those who encountered it between viewers and readers. Before that, it divided them between those with access to the inscription and those without, between those who could read or understand

Afterword

255

it and those who could not, between those with the status, wealth and power to make the inscription (or have it made) and those who could not. The inscription thereby helped not only to establish or reinforce social and hierarchical differences within a society but also to develop wider cultural differences between one society and another. Particular forms of inscription encouraged their viewers to associate themselves more explicitly with their society, thereby consolidating its membership. The identities that emerge could be dictated from above, by makers of monumental inscriptions, but also from below. The agglomeration of graffiti in Rome is an example of a willing coming together of individuals to associate themselves with a particular ideal. The choice of language and location excluded others, thereby establishing the limits of the social group. Similarly the destruction of inscriptions could redirect the identity of a social group, imposing new limits and exclusions on its undesirable elements, and at the same time establishing a break with the past and an opportunity to redefine it afresh.

THE BEAUTY OF WRITING

One aspect of inscriptions that does not appear often in the chapters in this volume concerns the beauty of writing, the ability of artists to arrest viewers’ attention through the skill with which they could produce harmonious, handsomely laid-out, elegantly proportioned lettering. It is most evident in the carvings at Noravank’ (see Chapter 9), but in that instance the beauty comes as much from the place of the words in a larger field of ornamentation as in the care taken to form the shapes of each letter. Yet the beguiling character of writing must lie at the heart of much of the impact of inscriptions explored by all the chapters in this volume. The dense rectangular fields created by the small scale, angular forms of cuneiform, the decorative nature of the church fac¸ades at Hosios Loukas created by the brickwork of pseudo-writing, the square kufic at the Qaratay Madrasa that converts the pendentives beneath the dome from an inscriptional zone into a geometric carpet pattern – all depend on a high level of artistry. Ultimately it is the consummate skill of the artists that transformed these words into physical entities that this volume celebrates. NOTES

1. From the sixth century: Denis Feissel, ‘Un rescrit de Justinien d´ecouvert a` Didymes (1er avril 533),’ Chiron. Mitteilungen der Kommission ¨ alte Geschichte und Epigraphik des Deutschen fur Arch¨aologsichen Instituts 34 (2004), 285–365; from the twelfth century: Jaroslav Folda, Art of the Crusaders in the Holy Land, 1095–1187 (Cambridge, 1995), 347–50.

2. For the bi-lingual doors at Bethlehem (Armenian, Arabic): Zehava Jacoby, ‘The Medieval Doors of the Church of the Nativity at Bethlehem,’ in Le Porte di Bronzo dall’Antichit`a al Secolo XIII, ed. S. Salomi (Rome, 1990), 121–34; for the trilingual inscription at Hekimhan (Armenian, Arabic, Syriac): Kurt Erdmann, Das Anatolische

256

Antony Eastmond

Karavansaray des 13. Jahrhunderts, Istanbuler Forschungen 21 (Berlin, 1961), 1, no. 18; Hakkı Acun, Anadolu Selc¸uklu D¨onemi Kervansaraylari (Ankara, 2007), 105–19. See also Antony Eastmond, ‘Inscriptions and Authority in Ani,’ in Der Doppeladler – Byzanz und die Seldschuken in Anatolien vom sp¨aten 11. bis 13. Jahrhundert, eds. Neslihan AsutayEffenberger, F. Daim (Mainz, in press). 3. Barbara Zeitler, ‘“Urbs felix dotata populo trilingui”: Some Thoughts about a TwelfthCentury Funerary Memorial from Palermo,’ Medieval Encounters 2.2 (1996), 114–39;

Darejan K’ldiashvili and Zaza Skhirt’ladze, Garejis epigrap’iuli dzeglebi [The Inscriptions of Gareja] (Tbilisi, 1999), vol. 1, part 1: The Lavra of St David, Udabno Monastery (11th–18th cc.), no. 86. 4. Ruchan Arık and Olus¸ Arık, Tiles: Treasures ¨ of Anatolian Soil: Tiles of the Seljuk and Beylik Periods (Istanbul, 2008), figs. 188, 382. 5. Mary Stroll, Symbols as Power. The Papacy after the Investiture Contest (Leiden, 1991), 106–9. 6. Keppel A.C. Creswell, The Muslim architecture ¯ ¯ of Egypt 2: Ayyubids and early Bah.rite Mamluks, A.D. 1171–1326 (Oxford, 1959), 101.

INDEX 鵼

Aaron, Armenian Governor (1055–1056), 188 Abbasid caliphate (749–1258), 62, 230, 242 abbreviations, 79, 86, 200, 209, 212, 218 ʿAbd al-Aziz, Ruler of Bukarah (r. 1539–51), 239 ʿAbd al-Malik ibn Mansur, ʿAmirid Chamberlain (11th Century), 234, 237 ʿAbd al-Malik, Umayyad Caliph (r. 685–705), 61 Abu Ibrahim Ahmad, Aghlabid Emir (r. 856–863), 64 Abu¯ ʿAbd All¯ah H ¯ Norman Civil Leader, . ammud, 136 Abu’l-Abbas b. al-Aghlab, Emir, 68 Achaemenid Empire, 10, 14 Bisotun, 15, 17, 23, 28 Ganj-nama, 16 Ishtar Gate, Babylon, 16 Naqsh-e Rostam, 16, 17, 25, 31 Palace of Cyrus, Pasargadae, 18 palaces, 14, 17 Persepolis, 13, 18, 26, 31 adʿiya, 126, 128, 129, 132, 133, 135, 136, 144 Aghlabid dynasty (r. 800–909), 62, 64, 70, 103 Albrecht Durer (1471–1528), 142 ¨ Alexander III, Pope (1159–1181), 137 Alexander the Great, King of Macedonia (356–323 BCE), 19, 239 Alfonso VIII, King of Castile (r. 1158–1214), 178 ´ (r. 1252–1284), Alfonso X, King of Castile-Leon 170, 178, 179 ʿAli b Yusuf, Almoravid Ruler (r. 1107–42), 236 ʿAli ibn Muhammad al-Zammaki, Calligrapher/Decorator (14th Century), 242, 244 Almohad dynasty (1147–1269), 174 alphabets Armenian, 187, 190, 201 Georgian asomtavruli, 76, 85 nuskha-khutsuri, 85 nuskhuri, 85 Greek, 23 Pahlavi, 24, 26, 31

Antiochos III, Ruler of the Seleukid Empire (r. 222–187 BCE), 20 Antoni Tsagereli, Bishop, 84 apotropaic, 22, 31, 100, 106, 107, 189 appropriation, 14, 124 Ardaxshir I, Founder of the Sasanian Empire (r. 224–240), 24, 26 Armenia, 78, 104 Ani, Cathedral, 188 Areni Church, 192 Marmaˇse¯ n, Monastery, 188 Noravank‘, Monastery of, 191–192, 197 Funerary church, 195 gawit‘, 192, 201 upper tympanum, 198 Siwnik‘, 192, 196, 198 Arsacid dynasty (247 BCE–224 AD), 23 Ashoka, Mauryan Emperor (r. 269–232 BCE), 23 Athanasios of Alexandria, Church Father (4th Century AD), 112 al-ʿAziz, Fatimid Caliph (r. 975–996), 62, 67 Bactria, 20 Rabatak, 24 Surkh Kotal, 24 Bagrationi dynasty, 82, 188 Baldwin I, King of Jerusalem (r. 1100–1118), 212 Bardas Skleros, Byzantine General (d.991), 89 Basil II, Emperor of Byzantium (r. 976–1025), 89, 110 ´ Beatrice of Swabia, Queen Consort of Castile-Leon (1205–1235), 172 Bernabo` Doria, Genoese Captain, 220 Bichitr, Mughal Craftsman, 239 Buddhism, 23 Byzantium. See individual monuments calendars, 171, 176 Caliphs, Rightly Guided, 157, 159, 162, 164, 167 Cantigas de Santa Mar´ıa, 178 Charlemagne, Emperor of the Romans (r. 800–814), 222 commemoration, 89, 90, 177, 178

257

258

Index

Constantine IX Monomachos, Emperor of Byzantium (r. 1042–1055), 110 Constantine X, Emperor of Byzantium (r. 1059–1067), 188 Constantinople, 102 Constantine Lips, monastery of, 99, 105 Galata, 212 Hagios Polyeutkos, 78 SS. Sergios and Bakchos, 78 contractions, 81, 86 convivencia, 170 Crusaders, 114 Cyrus the Great, King of Persia (r. 559–530 BCE), 17 Damasus, Bishop of Rome (366–384), 48 Darius I, King of Persia (r. 522–486 BCE), 12, 14, 17–18, 20, 23, 25, 26 Davit Kuropalates, Prince of Tao-Klarjeti (961–1001), 85, 87 Desiderius, Lombard King (r. 756–774), 222 devotional practice, 40 dik¯erion, 133 donor portrait, 188 Egypt Apa Apollo, Monastery of, 39, 48 Cairo, 103, 131 al-Aqmar Mosque, 71 al-Azhar Mosque, 71 Church of St Stephen, Philae, 39, 41, 43, 45, 46, 52 El Karnais, 46 epigraphic practice, 14, 20, 22, 23, 28, 125, 196 Eugenius of Palermo (1130–1202), 125 Fatimid dynasty (909–1171), 103 Fatimid Egypt. See Egypt ´ (r. 1217–1252), Fernando III, King of Castile-Leon 170 effigy, 173 Libro de los doce sabios, 179 Privilegio Rodado, 176 tomb, 172 vita. See vitae foundation text, 68, 78, 81, 82, 87, 88, 153, 164, 189, 209, 217, 230, 231, 236 France, 78 Chˆateauneuf Sanctuary, 41 Franciscans, 198 Genesis, 138 George of Antioch (d.1151), 129, 135 Georgia Ishkhani Cathedral, 82, 84, 86, 88 Jvari, 84 Kumurdo Cathedral, 76, 80, 85, 86, 89 Oshki Monastery, 85, 87

Otkhta Eklesia Monastery, 87 Parkhali Monastery, 85 Savane, 84 Zarzma Monastery, 88 Germany Liebfrauenkirche, Trier, 39, 44, 46 Gerona Casket, 232 Ghibelline factions, 220 Giorgi I, King of Georgia (r. 1014–1027), 83 graffiti, 13, 36, 38, 85, 200, 231 anonymity, 42 Greco-Bactrian kingdom (ca. 250–145 BCE), 22 Greece Athens, Church of the Theotokos, 39, 52 Hosios Loukas, 100 Church of the Panagia, 101, 102, 117 funerary chapel, 111, 113, 117 katholikon, 101, 109, 111, 114, 117 military saints, 115 Presentation of Christ at the Temple, 114 templon screen, 106, 107 Skripou (Orchomenos), Church of the Virgin, 78, 104 Guelph factions, 220 Gurgen Bagrationi, King of Kartli (r. 994–1008), 82 h.a¯ dith, 150–153, 243 al-Hafiz. li-d¯ın All¯ah, Fatimid Caliph (r. 1131–49), 136 ´ al-Hakam II, Caliph of Cordoba (r. 961–976), 232 al-Hakim, Fatimid Caliph (r. 996–1021), 110 h.ajj, 130 rukn, 131 .taw¯af, 131 Henry VI, Holy Roman Emperor (r. 1191–1197), 125 Henry VII, Holy Roman Emperor (r. 1312–1313), 205, 220–221, 223 Herakles, 22, 31 ´ Hisham III, Caliph of Cordoba (r. 976–1009), 232 Holy Land, 43, 100, 110–113, 115, 118, 196 Huvishka, Kushan King (r. 140–180 AD), 24 Ibn Jubayr, 143 Ibn ʿIdhari, Historian (d.1312), 61 icon Mother of God from Khobi, Georgia, 79 St Symeon the Younger, Georgia, c.1015, 92 identity, 14 Ifriqiya. See Tunisia inscriptions, 25, 64 appearance, 17, 205 carving, 12, 14, 76, 176 coinage, 11, 61, 65, 72, 125 commemorative, 42, 94, 126, 214 content, 16, 17, 76, 106, 124, 171, 176, 187 embroidered, 126, 196, 238 erasure, 14, 61, 62, 70, 71, 72

Index

form, 78, 86 function, 86 icon frame, 79, 92 kufic, 62, 64, 102, 132, 153, 156, 162, 171, 233 location, 54, 82, 84, 87, 130, 187, 243 monumental, 14, 20, 24, 73, 87, 89, 94, 177, 205 multi-lingual, 14, 22, 24, 25, 26, 125, 170, 174, 196, 230 performance, 17, 26, 31, 44, 134, 171 performativity, 13, 42, 83, 130 signatures, 230, 231, 233, 234, 236, 237, 243, 244 textual icon, 78, 86, 87, 91, 92, 94 visibility, 79, 188, 205 Ioane Zosime, Monk (949–987), 94 Iovane Mtatsmindeli/John the Iberian, Monk (d.1002), 86, 87, 90 Islamic mysticism, 149, 162, 167 Israel Jerusalem. See Jerusalem Italy Cimitile-Nola, Shrine of St Felix, 39, 43, 47, 52 Genoa, San Lorenzo, 205, 206–212, 221 Janus epigraphs, 217, 218, 219, 222 nave, 213, 224 Hercules Curinus, Shrine of, 41 Milan, Basilica of St Ambrose, 223 Padua Santa Giustina, 217 Tomb of Antenor, 216 Rome Catacomb of Callixtus, 51, 52 Catacombs, 39 Memoria Apostolorum, 39, 42, 44, 46, 48 St John Lateran, 36 St Peter Memorial, Vatican Hill, 39 Salerno Cathedral, 124 Venice, 217 San Marco, 217 Ivories (Iberian), 231, 235 Fitero Casket, 231, 234, 244 Gerona Casket, 234 Hispanic Society of America D752, 231 Pamplona Casket, 234, 236, 244 ʿIzz al-Din, Seljuk Sultan (r. 1246–1256), 153, 164 Jacopo da Varagine, Archbishop of Genoa (1230–1298), 215, 218, 223 Chronica Civitatis Ianuensis, 215 Jacopo Doria (born 1233), 215–216, 223 Jahangir, Mughal Emperor (r. 1605–1627), 239 Jalal al-Din Qaratay, Regent of the Seljuk Sultanate (1249–1254), 148, 152, 159, 164 Jalal al-Din Rumi, Sufi Mystic (1207–1273), 148, 163 James I, King of England (r. 1603–1625), 239

259

Janus, 209, 214, 216, 222, 223 Jean Adam Delsenbach (1687–1765), 142 Jerusalem, 110, 112 Church of the Holy Sepulchre, 110, 115, 213 Dome of the Rock, 61, 114, 243 Joshua, Old Testament General, 109 Kamal al-Din Rumtash (1248–1249), 150 Kanishka I, Kushan King (r. 127–140 AD), 24 Katranid¯e, Wife of King Gagik I of Ani (r. 989–1017), 188 khatchkars, 190 Kit¯ab Ruj¯ar, 143 Kushan dynasty (1st AD–4th AD), 23 lamps hanging, 242, 243 Sayf al-Din Qawsun, 240 languages, 46 Arabic, 72, 125, 129, 135, 137, 170, 171, 176, 196, 233 Aramaic, 20, 22, 23 Armenian, 83 Babylonian Akkadian, 14 Castilian, 170, 171, 176 Elamite, 14, 16 Georgian, 83, 87, 94 Greek, 19, 22, 23, 25, 87, 125, 129, 137 Hebrew, 170, 171 Latin, 72, 125, 129, 137, 170, 171, 176, 206, 212, 216 magical, 108 Parthian, 25, 26 Persian Middle, 24, 25, 26 Old, 11, 14, 17, 18 Prakrit, 23 Laodike III, Wife of Antiochos III (d.191 BCE), 20 Last Judgement, 94 legibility, 12, 15, 16, 17, 20, 25, 49, 62, 76, 81, 87, 102, 135, 140, 188, 195, 212 Leo the Deacon, Byzantine Historian (born ca.950), 108 Leon III Anchabadze, King of Abkhazeti (r. 957–967), 79, 81 liminality, 52, 84, 107, 113 literacy, 48, 72, 79, 86, 87, 105, 171, 187, 212, 214, 236 loca sancta, 110, 112, 115 Lothar I, Emperor of the Romans (r. 817–855), 222 al-Nasir Muhammad, Mamluk Sultan of Egypt (r. 1293–1341), 240, 241, 242 Mahmud Muzahhib, Craftsman, 239 Mahmud, Calligrapher at Bukarah, 239 al-Maʾmun, Abbasid Caliph (r. 813–833), 61

260

Index

al-Mansur b. Buluggin b. Ziri, Zirid Governor (r. 984–995), 62, 67 manuscripts, 72, 239, 242 Art and History Trust Collection, D¯ıv¯an of Hafiz, 238 Erevan, Matenadaran 2743, Gospels of the Translators, 196 Erevan, Matenadaran 6201, Mułni Gospels, 196 London, British Library MS Harley 5786, Psalter, 129 Minneapolis, Minneapolis Museum of Art 52.31, Bah¯arist¯an, 238 Mount Athos, Iviron MS Georg.1, Old Testament, 79, 85 Mount Athos, Iviron MS Georg.9, Samotkhe, 85 Margaret of Brabant, Wife of Henry VII (1276–1311), 221 Mark, Evangelist, 217 Martino da Canale, Venetian Chronicler (13th Century), 217 materiality, 38, 49 Matthew Bonellus, Italo-Norman Baron, 137 memory, 84 civic, 215, 216, 224 collective, 91, 224 public, 214 Miqail ibn Simino/Miguel Jim´enez (d. 1156), 176 Mir Husayn al-Husayni, Calligrapher at Bukarah, 238 Mithradates II, King of Parthia (r. 123–88 BCE), 23 monogram, 28 chi-rho, 46, 52 Morocco Marrakesh minbar of the Kutubiyya Mosque, 236 Mount Athos, 87 Mozarab, 175, 176 Muhammad ibn Mansur al-Mahdi, Abbasid Caliph (r. 775–785), 61 Muhammad, Prophet, 62, 64, 71, 149, 157, 159, 163 al-Muʿizz ibn Badis, Zirid Governor, (r. 1016–1062), 61 Narseh, Sasanian King (r. 293–302), 26 Opizzino Spinola, Genoese Captain, 220 otherness, 100, 109, 113, 118, 196 panegyric, 126, 128, 131, 144 patronage, 10, 13, 17, 24, 31, 61, 66, 68, 82, 101, 112, 135, 148, 150, 162, 188, 230, 241 Paul, Apostle, 44, 46, 48, 138 perambulation, 104, 106, 206, 218 Persian Empire (550–330 BCE), 14 Peter of Eboli, Chronicler (fl.1196–1220), 125 Peter, Apostle, 44, 46, 48, 138 Philagathos Kerameos, 140

piety, 44, 49, 81 popular, 38, 40 pilgrimage, 36, 39, 43, 100, 104, 106, 110–112, 113, 117, 129, 131 power, 14, 81 prayer, 43, 54, 78, 79, 86, 87 prolepsis, 106 pseudo-epigraphy, 99, 104, 115, 128, 136, 138, 139, 141, 143, 156, 162, 196, 197, 230 public text, 38, 73, 100, 108, 110, 188, 212 Qur’an, 64, 149, 155, 160, 240, 242 Raimundo de Losana, Archbishop of Seville (d.1286), 176 readability, 78, 81, 82, 89, 136, 138, 193, 219 Rightly Guided Caliphs. See Caliphs, Rightly Guided ritual, 13, 25, 30, 130, 131 Robert Guiscard, Duke of Sicily (1015–1085), 124 Roger II, King of Norman Sicily (r. 1130–1154), 125 mantle of, 126, 134, 142 water-clock, 126, 129 Romanos I Lekapenos, Emperor of Byzantium (r. 920–944), 103 Romanos II, Emperor of Byzantium (r. 959–963), 101, 109 sajʿ, 135, 142 Saladin, Founder of the Ayyubid dynasty (r. 1174–1193), 71 sapientialism, 171, 179 Sasanian dynasty (224–642), 24 Sasanian Empire, 10 Ardaxshir-Xwarrah, 26 Kaʿba-ye Zardosht, 25, 26 Pars, 25 stucco carving, 28 script, 49, 170, 176 Akkadian cuneiform, 15 Brahmi, 23 Gothic capital, 206, 216 Lombardic, 171 naskh, 243 Old Persian cuneiform, 14 Seleukid Empire, 19, 20 Ai Khanum, 22, 28 Bisotun, 22 inscribed stele, 20 Karafto, 22 Seleukeia-on-the-Tigris, 22 Temple of the Oxos, 22, 31 Setenario, 178 Shabuhr I, Sasanian King (r. 240–270 AD), 11, 14, 24–26 Shabuhr II, Sasanian King (r. 309–379), 26 Shabuhr Saganshah, 26

Index

Shiʿi, 62, 66, 70 Sicily, 179 Cappella Palatina, 125, 128 nave ceiling, 132, 138, 139 opus sectile, 129, 135, 136 Cefalu` Cathedral, 125 San Giacomo la Mazara, 126 Santa Maria dell’Ammiraglio, 129 Siete Partidas, 178 signifiers, 12, 124, 143, 206 Spain ´ 171 Castile and Leon, ´ Cordoba Cathedral, 177 ´ Great Mosque of Cordoba, 172, 236 Madinat al-Zahraʾ, 235 Santas Justa y Rufina, Toledo, 176 Seville Cathedral, 170, 176 Giralda, 172 St Clement’s day, 174, 178 Step‘anos Orbelian, Armenian Historian (1250–1305), 191 Sufism, 149, 164, 165 Sul¨eyman the Magnificent, Ottoman Sultan (r. 1520–1566), 243 Sultan Muhammad ʿIraqi, Court Painter (16th Century), 238 Sunni, 66, 70, 149, 164 synodika, 91 Syria Church of the Holy Cross, 39, 43, 46, 48, 53 Dura Europos, 41 Syros, Saint and Bishop of Genoa, 219 Tahmasp I, Shah of Iran (r. 1522–1576), 238 text appearance, 82, 176 arrangement, 219 compression, 82 document, 177 layout, 84, 190, 218, 231 legal, 89, 188 mnemonic device, 78 ornamental, 82 record, 86 sequence, 92, 104 spiritual power, 84 Theodosios, Abbot (fl. 1048), 101 theology, 94, 201 cross, 84 doors, 84 topography power, 29 sacred, 39, 40, 52 symbolic, 11 Tornik, Georgian Monk, 79, 82, 88 triumph, 100, 108, 117 Tulunid dynasty (868–905), 103

261

Tunisia Gafsa Fortress, 72 Kairouan Congregational Mosque/Mosque of Sidi ʿUqba, 62, 64, 66, 71 Mosque of the Three Doors, 68, 71 Mahdia, 64, 69, 103 Sabra-Mansuriya, 64, 65, 69 Sfax, Great Mosque, 62, 66 Sousse Bu Fatata Mosque, 68 Congregational Mosque, 62, 66, 68 Ribat, 68 Zaytuna Mosque, Tunis, 62, 66, 71, 230 Turkey Ephesos, Grotto of St Paul, 46 Erzurum, C ¸ ifte Minareli Medrese, 166 Istanbul. See Constantinople Konya Alaeddin Mosque, 150 Qaratay Madrasa, 148, 150, 164, 165: courtyard, 153; dome, 160 Sahib Ata, 166 Sırc¸alı Madrasa, 150 Tahir and Zuhre Mosque, 166 ¨ Malatya, Ulu Cami, 166 Van, Citadel of, 17

ʿUbaydallah, Ruler of Bukarah (r. 1533–39), 239 Ursone da Sestri, Italian Poet, 214

Vahram Paxlawuni, Armenian Prince (11th Century), 188 Varqeh and Golsh¯ah, 149, 159 Vima Kadphises, Kushan King (r. 90–100 AD), 24 Virgen de los Reyes, 174 Virgin and Child Hodigitria, 192, 198 vitae Dagobert, 178 Fernando III, 178 Louis VI, 178 Mesrop Maˇstoc‘, 187 St Luke, 101, 108 votive offerings, 42 Wahram II (r. 276–293), 26 William II, King of Norman Sicily (r. 1166–1189), 143 William Malconvenant, Italo-Norman Baron (1183–1203), 137 word play, 178, 233, 236, 240, 243 Xerxes I, King of Persia (r. 486–465 BCE), 16, 18 Ziyadat Allah, Aghlabid Governor (r. 817–838), 62, 64