Brill Publishing The Dragon in Medieval East Christian and Islamic Art (2011).pdf - PDFCOFFEE.COM (2024)

The Dragon in Medieval East Christian and Islamic Art

Islamic History and Civilization Studies and Texts Editorial Board

Sebastian Günther Wadad Kadi

VOLUME 86

The Dragon in Medieval East Christian and Islamic Art With a Foreword by Robert Hillenbrand By

Sara Kuehn

LEIDEN • BOSTON 2011

This book is printed on acid-free paper. Kuehn, Sara. â•… The dragon in medieval East Christian and Islamic art / by Sara Kuehn ; with a foreword by Robert Hillenbrand. â•…â•… p. cm. — (Islamic history and civilization ; v. 86) â•… Includes bibliographical references and index. â•… ISBN 978-90-04-18663-7 (hardback : alk. paper) ╇ 1. Dragons in art. 2. Islamic art and symbolism. 3. Christian art and symbolism— Medieval, 500-1500. I. Title. II. Series. â•… N7745.D73K84 2011 â•… 704.9’47—dc22

2011008359

ISSN╇ 0929-2403 ISBN 978 90 04 18663 7 Copyright 2011 by by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill, Global Oriental, Hotei Publishing, IDC Publishers, Martinus Nijhoff Publishers and VSP. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill NV provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.

introduction

For my parents

v

vi

introduction

Wie sollten wir jener alten Mythen vergessen können, die am Anfange aller Völker stehen, der Mythen von den Drachen, die sich im äußersten Augenblick in Prinzessinnen verwandeln; vielleicht sind alle Drachen unseres Lebens Prinzessinnen, die nur darauf warten, uns einmal schön und mutig zu sehen. Rainer Maria Rilke “Briefe an einen jungen Dichter”

contents

vii

CONTENTS Notes on the Bibliography, Transliteration, Dating and Illustrations ix Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇╅ xi ╅ Robert Hillenbrand Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ xiii PART I

introduction Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅╇ 3 Characteristics of the (serpent-)dragon . . . . . . . . . . . . . . . . . . . . . . .╅╇ 5 Existing contributions to the study of dragon ico�no�g�raphy. . . . . . .╅ 10 1╅ The Medieval Islamic World from Central Asia to Anatolia: The Historical and Cultural Context .. . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅ 15 2╅ Dragons on Monumental Settings in Regions West of Iran╇ . . . . . .╅ 21 3╅ The Dragon Motif on Portable Objects . . . . . . . . . . . . . . . . . . . . . . . .╅ a. Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅ b. The dragon motif on accoutrements relating to the hunt or war. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅ c. The dragon motif on objects of personal adornment. . . . . . . . .╅ d. The dragon motif on vessels. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅

35 35 36 44 47

part ii

the dragon and the natural world 4â•… Dragons and the Powers of the Earth . . . . . . . . . . . . . . . . . . . . . . . . .â•… a. The dragon and the elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•… b. The serpent(-dragon) jinn. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•… c. The treasure-guarding dragon. . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•… d. The dragon and vegetal compositions. . . . . . . . . . . . . . . . . . . . . .â•…

51 51 56 59 63

5â•… Dragons and Animals of the Natural and the Mythical Realms . . .â•… a. The dragon in animal combat scenes. . . . . . . . . . . . . . . . . . . . . . .â•… b. The dragon and mythical creatures .. . . . . . . . . . . . . . . . . . . . . . .â•… c. The dragon and the bird . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•… d. The dragon and the feline. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•… e. The dragon and the elephant . . . . . . . . . . . . . . . . . . . . . . . . . . . . .â•…

73 73 74 76 78 82

part iii

the dragon and its royal and heroic associations 6╅ The Dragon in Scenes of Combat. . . . . . . . . . . . . . . . . . . . . . . . . . . . .╅ 87 a. The dragon combat in ancient myth. . . . . . . . . . . . . . . . . . . . . . .╅ 87 b. Iconography of the medieval dragon-fighter . . . . . . . . . . . . . . . .╅ 92 c. The Eastern Christian holy rider as dragon-fighter . . . . . . . . . . ╇ 102

viii

contents

7â•… The Dragon in Relation to Royal or Heroic Figures . . . . . . . . . . . . . ╇ 111 a. Royal and heroic associations of the dragon in literary accounts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 111 b.╇ Paired dragons with a central anthropomorphic figure. . . . . . . ╇ 113 c.╇ Symbolism of the dragon’s tongue. . . . . . . . . . . . . . . . . . . . . . . . . ╇ 127 part iv

the dragon in astrology, alchemy, medicine and magiC 8â•… The Dragon and Astrology. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 133 a.╇ Astrology in medieval Central Asia. . . . . . . . . . . . . . . . . . . . . . . . ╇ 133 b.╇ The dragon in medieval Islamic Â�astrology . . . . . . . . . . . . . . . . . . ╇ 136 9â•… Vestiges of Ancient Dragon Iconographies. . . . . . . . . . . . . . . . . . . . . ╇ 145 a.╇ The encircling dragon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 145 b.╇ The drinking dragon. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 155 10 The Knotted Dragon Motif . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 159 a.╇ The ancient practice of knot tying. . . . . . . . . . . . . . . . . . . . . . . . . ╇ 159 b.╇ The knotted dragon.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 161 11 The Dragon and the Magico-Medical Sphere. . . . . . . . . . . . . . . . . . . ╇ 169 a. The dragon as prophylaxis and cure . . . . . . . . . . . . . . . . . . . . . . . ╇ 169 b. The dragon and the theriaca as illustrated in the Kitāb al-diryāq. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 171 c.╇ Studies on the properties of serpent(-dragons) and the effects of their venom. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 176 d. The serpent(-dragon) stone. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 180 12 The Dragon and the Province of Magic and Divination. . . . . . . . . . ╇ 183 a. The dragon and talismans . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 183 b. The dragon staff. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 186 part v

the dragon as mystical metaphor 13 The dragon and sound . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 191 14 The dragon as symbol of transformation . . . . . . . . . . . . . . . . . . . . . ╇ 195 Concluding Remarks. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 205 Epilogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 209 a. East-west exchange and the metamorphosis of dragon imagery╇ .209 b. Syncretism and the dragon .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 228 List of Abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 237 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 239 List of Illustrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ╇ 273 General Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285 Illustrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . following 298╇

introduction

ix

Notes on the bibliography, transliteration, dating and illustrations

The bibliography is limited to works and articles that are cited in the study. All bibliographic references for books and periodicals are given in abbreviated form in the footnotes with complete citations appearing in the bibliography. Encyclopaedia articles and dictionary entries are cited only in the notes. Unless passages are quoted, editions of classical authors are not cited. The system of transliteration of Arabic, Persian and Turkish words used in this work is based on that of the Encyclopaedia of Islam, second edition, with several modifications: “q” replaces “ḳ”; “j” is used instead of “dj.” “Th,” “kh,” “dg,” “sh” and “gh” are not underlined.

With the exception of certain words such as Qurʾān, neither italics nor diacritical marks are used for those Arabic, Persian or Turkish names and terms that have entered into common English usage like large cities, geographical locations and dynasties. Where specific dates pertaining to the Islamic realm are referenced, both Islamic (hijrī) and Christian (Gregorian) dates will be given, the Islamic date appearing first. Where a century or decade is mentioned only Christian dates are cited. Photographs included in the illustrations were taken by the author unless otherwise noted.

x

introduction

foreword

xi

FOREWORD

This book is part of a much longer and comprehensive study on which Dr Kuehn has been labouring for over a decade and whose aim is to trace the iconography of the composite mythical creature known as the serpent-dragon from the mists of antiquity to the later middle ages. Her geographical focus in the study as a whole is principally Western and Central Asia but she remains continually alert to the manifestations of her theme in neighbouring cultures to the east (including India and China) and the west. The continuity of this arresting image across vast gulfs of space and time in the most diverse cultures of the Old World from the Atlantic to the Pacific is quite startling. That continuity in itself constitutes a major challenge to anyone seeking to tell a connected story that extends across continents, cultures and millennia. The volume of scholarship on the art of Western and Central Asia has grown exponentially in the last couple of generations. In the field of Islamic art alone, it is clearly no longer a reasonable ambition to produce a companion volume to Creswell’s magisterial Bibliography of the Architecture, Arts and Crafts of Islam to 1st Jan. 1960; such a work would need to be several times the size and weight of that huge tome in order to cover what has been produced in the last fifty years. But as the volume of scholarship expands, so, by a seemingly ineluctable law, does its scope contract. More and more people write about less and less. The dangers of over-specialisation and tunnel vision loom large. Artificial boundaries, whether chronological, geographical, cultural or confessional, are set and then ferociously policed. Scholarship operates in watertight compartments, to the detriment of that open-mindedness, that cross-fertilisation of disciplines and, more generally, the linking of disparate bodies of information that have traditionally been regarded as the litmus text of creative thinking in academe. Iconographical studies are especially vulnerable to this shift from the macroscopic to the microscopic mode. Images readily adapt to changes in use, in faith and context, not to mention changes in location or scale, but they do tend to guard their core meanings most tenaciously. Neverthe-

less, an altered context, especially if it involves a transfer from one faith to another – such as Isis suckling Harpocrates, often regarded as an immediate model for the Christian image of the Virgin and Child – can trigger unexpected accretions and adaptations of meaning. Thus there can develop over the centuries a pool of ideas associated with a given image, and it requires expert judgment and erudition to make the right choices from that pool in any particular case. The body of evidence and allusion that accumulates in this way becomes increasingly difficult to control and to understand. Such, then, are some of the difficulties confronting an extensive iconographical study of the kind that Dr Kuehn has produced. To overcome those difficulties calls for a special kind of scholar, one that was much more commonly encountered several generations ago. Happily Dr Kuehn fits that bill, and has the sheer erudition, the wideranging sympathies, the creative imagination and the indefatigable intellectual curiosity to match. Methodically and passionately she follows the leads of her research wherever they take her, crossing numerous disciplinary boundaries en route. The result is a many-textured study of remarkable boldness and finesse that, firmly grounded in the thought-worlds of Bronze Age Central Asia and the Hellenistic empire, explores the full flowering of the serpent-dragon motif in medieval East Christian and Islamic art, most especially in Anatolia. The range of reference is extensive – from the mythic origins of the theme to such detailed aspects as the dragon tamer, combat scenes, the significance of knotting, and the serpent-dragon as an element of personal adornment. We learn of its interaction with other animals and how it functioned as an emblem of war and of the hunt, as a guardian of treasure and as an avatar of chthonic powers; and its sinister side helps to explain its appearance in Christian contexts in association with such saints as George and Theodore. Yet it also had multiple royal and heroic associations, as shown for example by the dracontine throne with its apotropaic role. Small wonder that this fabulous creature developed an apocalyptic significance and figured largely in the Islamic sciences

xii

foreword

– whether in star lore or toxicology, magic or cosmology. These various excurses reveal a manylayered thought world shared by Arabs, Persians and Turks as by Byzantine, Armenian, Syriac and Georgian Christians. The serpent-dragon appears on mausolea and gravestones, on mosques and madrasas, on monasteries and churches, on bastions and caravansarais, on city gates and palace frescoes, on pottery galore, on coins and figured silks, on mirrors and belt buckles. Usually it carries a symbolic charge, for example as an amulet or talisman, but it is also at home in narrative contexts. Altogether this is pioneering original work, and it demonstrates an enviable capacity to move from

one culture to another – classical, Christian, Zoroastrian, Islamic – in a remarkably sure-footed way. It is packed with cogent arguments and unexpected insights. Dr Kuehn is a born explorer and has a natural affinity for cross-cultural work. She disdains the quick fix and is ready to do whatever is required to prove her point. Her list of authorities is startling in its length and completeness. But those authorities are merely a means to an end – the tale’s the thing, and it casts a potent spell. Robert Hillenbrand University of Edinburgh

introduction

xiii

Acknowledgements

The topic of dragon iconography in medieval Islamic art was suggested to me for my doctoral thesis by my Doktorvater Professor Claus-Peter Haase in 1999. I am immensely grateful to him for having given me the opportunity to work on the tremendous and awe-inspiring yet, at the same time, exceedingly challenging topic that now forms the basis for this book. I especially valued his continued support and advice while he was extremely busy as director of the Museum for Islamic Art in Berlin. During the years in which I was engaged in this research, I regularly worked on the preparation of curatorial documentation for the al-Sabāh Collection, Kuwait National Museum, in Kuwait. This allowed me to gain an insight into one of the most outstanding collections of Islamic works of art, among whose treasures are many objects bearing revealing manifestations of the dragon. I would like to express my gratitude to His Highness Shaykh Nasser Sabāh al-Ahmed al-Sabāh and his wife Her Highness Shaykha Hussah Sabāh alSalem al-Sabāh for generously granting me permission to publish several objects from their collection. I am much indebted to the curators of the collection, Manuel Keene and Sue Kaoukji, for their long-standing encouragement and help. My thanks also extend to my colleagues at the Dar al-Athar al-Islamiyyah. It took ten years to complete this vast research project and put it into writing. The text has been read with keen intelligence and an extraordinary editorial eye by Jill Tilden who moreover was a valued source of intellectual support and encouragement throughout these years. I feel profoundly grateful for the privilege of her friendship. In addition, I wish to thank Professor Eberhard König of the Free University of Berlin, second reader of my thesis, for his sympathetic reading of my interpretations and his continuous generous support. Reverend Dr Vrej Nersessian of the British Library, London, read the sections concerning Armenian art, sharing with me some of his profound knowledge on Armenian art and offering additional valuable suggestions. For this I am extremely grateful. My thanks also go to Dr Hratschja Tamrasjan, Director of the Â�Matenadaran

in Yerevan, Armenia, as well as to the curator Dr Lilith Zakarian, for their generous help. I am further indebted to Professor Dr Peter Zieme of the Berlin-Brandenburgische Akademie der Wissenschaften for reading the text and making valuable comments and suggestions. I owe particular thanks to my editors, especially Ms Kathy van Vliet-Leigh and Ms Ellen Girmscheid, for inviting me to submit my text for publication in this series and for their formative guidance, patience and support throughout the publication process. For comments and criticisms on earlier versions of this work I also thank the anonymous readers for Brill. For their kind assistance in obtaining images and line drawings, grateful thanks are moreover due to Prof. James Allan, Ms Katherine Baker, Dr Wolfgang Baum, Dr Sheila Canby, Ms Maria Teresa Fortuna Canivet, Dr Stefano Carboni, Prof. Dr Falko Daim, Dr Catherine Â�Depierraz, Mr Alfred Diwersy, Dr Joachim Gierlichs, Prof. Dr Claus-Peter Haase, Dr Navina Haider, Mr Isao Kurita, Dr Sophie Makariou, Ms Armine Melkonyan, Dr Nahla Nassar, Rev. Dr Vrej Â�Nersessian, Mr Harvey B. Plotnick, Dr Christoph Rauch, Prof. Dr Scott Redford, Prof. Dr Mehmet Öz, Prof. Dr Yasser Tabbaa, Dr Hratschja Â�Tamrasjan, Dr Nicole Thierry, Dr Jean Michel Thierry, Dr Lilith Zakarian. I am particularly indebted to Prof. Dr Piotrovsky, Ms Elena Obuhovich and Prof. Dr Alexander Nikitin for their tremendous help. Above all, my deep gratitude is due to Professor Robert Hillenbrand, Professor Emeritus of the University of Edinburgh, for his inspired vision and open-hearted generosity of spirit that I will not forget. His constructive guidance was instrumental in allowing this work to be successfully delivered from the dragon’s maw. All errors and inadequacies that remain are, of course, to be attributed to me alone. Finally, immeasurable thanks are due to my parents, Heidi and Michael Kuehn, who generously helped me emotionally and financially to complete this project. Their manifold support and tolerance have sustained me throughout. For all these reasons, the book is dedicated to them as a small token of my thanks.

xiv

introduction

introduction

part one

introduction

1

introduction

3

INTRODUCTION

The aim of this research is to contextualise and chart, as far as possible, the complex iconography of the dragon in the medieval Islamic world,1 by interrogating the many factors, contexts and contingencies that helped to shape and transform it.2 The study focuses on the identification of the dragon imagery in a medieval Central Asian 3 cultural context, in what may be described as Irano-Turkish territories, from where it was disseminated by people of predominantly Turkic and Iranian stock.4 It necessarily draws on a vast corpus of imagery of long artistic and iconographic tradition which originates from an equally vast geographic area of enormous cultural and ethnic complexity, with a primary emphasis on the transmission of the dragon iconography from Central Asia to Anatolia. Importantly, the latter comprises to a large extent parts of the region that formed part of the empire of Alexander the Great at his death in 323 bc, constituting ancient Sogdia, Bactria, the Indus Valley, Parthia, Media, the Transcaucasus and Anatolia. A common feature of these regions is therefore to have been subject for three to four centuries to intermittent waves of Hellenistic influence. Arab conquests of Central Asia began to gain momentum from 86/705 when Qutayba ibn Muslim was appointed governor of Khurasan, from where he led incursions into neighbouring regions.5 This led to a process of Islamicisation in the city states of sedentary Central Asia and the subsequent transformation of the entire region

into a centre of Islamic civilisation. It also resulted in the assimilation and subsequent Islamicisation of the steppe peoples of Turko-Mongol heritage. Islamic-period Central Asia naturally inherited artistic traditions from preceding dynasties such as the Sasanians (c. 224–651) and the Sogdians (fifth–eighth centuries). A true melting pot of peoples and cultures, the region had from earliest times served as a mediator and transmitter of artistic trends as they passed from east to west Asia and vice versa. This phenomenon was taken even further in the vast spatial entity of Islam, where economic links facilitated the transmission of knowledge as well as cultural and artistic exchange among peoples of different backgrounds and thus, in spite of the multicultural setting, conveyed a feeling of unity and a sense of belonging to a common civilisation.6 Medieval Islamic society was a mixture of several regional cultures which included Muslims and non-Muslims speaking many languages, including Arabic, Persian, Syriac, Hebrew, Armenian, Turkish, Kurdish and various local dialects. The approach in the following essays is thus necessarily broadly comparative since evidently, as Julie Scott Meisami has aptly put it, “the medieval world does not stop at, say, the border between Christian ByÂ�zantium and Islamic territories, it is also clear that valuable insights may be gained from comparing the various manifestations of what is, to a great extent, a unified tradition, which shares certain basic attitudes and assumptions despite the par-

╇ Throughout this investigation the traditional historical era, commonly referred to as the medieval period, is defined as spanning the eighth to the thirteenth century. 2 ╇ On the history of the study of iconography in Islamic art, see the recent resumé of Ernst Grube (2005, pp.╯13–33) with an extensive list of references. 3 ╇ Today “Central Asia” has acquired a narrower meaning associating it with its use in the former Soviet Union and can be said to include the territories of Uzbekistan, TurkÂ� menistan, Tajikistan, Kyrgyzstan, Kazakhstan as well as Mongolia, the Tibet Autonomous Region and the Xinjiang Uighur Autonomous Province in northwest China. However throughout the present study the term is used in its broader sense following the cultural definition of Central Asia given by UNESCO in two meetings of experts, held in 1978 and 1979, which is anchored in the multi-volume History of Civilization of Central Asia. The Final Report of 12 February 1979, Paris, reads that the spatial dimensions include “territories lying at present within the boundaries of Afghanistan, the western

part of China, northern India, northeastern Iran, Mongolia, Pakistan and the [former] Central Asian Republics of the USSR.” See Miroshnikov, 1992, repr. 1999, pp.╯259–80 (the discussion also includes a brief outline of the historical usage of the term). The problems of defining the conceptual geographies of Central Asia are revisited by Akiner, 1998, pp.╯3–62. 4 ╇ The words “Turkic” as well as “Iranian” are used as general designations to denote people whose ruler or majority spoke a Turkic or an Iranian language. Tribal confederacies in Central Asia were very heterogeneous and under various cultural influences. Cf. Frye, 2005, p.╯149, n. 1. 5 ╇ On the Muslim Arab campaign in Central Asia and subsequent consolidation of power, see the classical study of Gibb, 1923. 6 ╇ The accounts of medieval travellers show that there were, in fact, apart from sea frontiers, no clearly defined boundary lines within the Islamic empire. See Bauer, 1995, pp.╯34–6.

1

4

introduction

ticular local colouring of the individual cultures that make up the whole.” Therefore, since it pertains to more than one culture and geographical region, the study necessarily addresses the multicultural and hybrid facets of the dragon motif as it evolved in these regions and examines how the motif was accepted and incorporated into the artistic repertory. An investigation into the visual phenomenon of the dragon, which evolved from its pre-Islamic origins to manifest itself in varied but analogous and interrelated forms across this wide spatial and temporal entity, necessitates a broad overview of the entire spectrum of images as they appear on diverse media. In doing so the study, moreover, inevitably exhibits some of the difficulties arising from the necessity of crossing academic boundaries. An interdisciplinary method of analysis has been pursued, involving not only art historical but literary, epigraphical and historical evidence. During the ten years it took to compile the vast body of data the sheer scope of the material, in cultural, confessional, geographical and chronological terms, threatened to overwhelm all attempts at containment and control. Necessarily, given the vastness of the subject, only certain aspects of the multilayered and multivalent character of the topic can be treated. This study identifies and discusses specific themes pertaining to the dragon iconography which can be observed over a long period of time. The likeness of the dragon is commonly associated with Asia and more specifically with China, being a paramount Chinese emblem. Yet its iconographic expression was known and used in a Central Asian context during the Bronze Age period, i.e. from the late third to early second millennium bc,7 and was again extensively employed in the so-called “animal style” which was transmitted in the wake of the migrations of the ancient nomads of the Scytho-Siberian culture.8 The late outflows of the culture which produced this style include, for instance, the Xiongnu of Mongolia and the Yuezhi (Rouzhi), who were driven out of present-day Gansu province in China by the Xiongnu in the second century bc and migrated

to the region of Bactria that lies between the mountains of the Hindu Kush and the classical Oxus river (known as Āmū Daryā). Known as the Kushāṇas, they entered the Eurasian heartlands and the Indian subcontinent in the first or second centuries ad. Under subsequent Central Asian dynasties such as the Sasanians and Sogdians (who were closely linked with the Turkic empires and played the role of active agents of cultural interaction), the dragon motif continued to be extensively employed and was to become a prominent emblem of the Great Saljuq Turks. This so-called “Saljuq-style” dragon was a motif in common currency from Central Asia to Anatolia (Rūm, the “Roman”/Byzantine lands) long before its place was taken by a so-called “Chinese-style” dragon, introduced in the aftermath of the Mongol invasion during the rule of the Chaghatayids (624/1227–764/1363), the Batuʾids (624/1227–907/1502) and the Ilkhanids (654/1256–754/1353) when China marked one pole of the Mongol empire at its time of greatest territorial expansion. This gave rise to a Chinese and Chinese-inspired but Mongol version of the dragon that began to appear for instance on the tile revetments of the Ilkhanid summer residence at Takht-i Sulaimān, built in the 1270s in the Azerbaijan region of present-day Iran, as well as in some early fourteenth-century manuscripts. The transmission of the visual rendering of the motif was the result of an acculturation process in which it was translated into a Central Asian context. The focus of this study is precisely on the manifestations of the dragon as evinced in the cultural and artistic context of the medieval Central Asian world before the phenomenon of the “Chinese-style” dragon occurred in the arts of Islam during the latter half of the thirteenth and fourteenth century with the establishment of the Mongols in Central Asia. Examples dating to after the Mongol invasion are employed only in so far as they illustrate a particularly pertinent symbolic feature in the stylistic continuation of the “Saljuq-style” dragon (the term “Saljuq” being used throughout this study in an extended sense, geographically and chronologically). The issue of

7 ╇ The dragon (together with its smaller relative, the serpent, as will be shown) is a universally attested motif, perhaps generated simultaneously by a number of cultures. An “Eastern,” and more specifically “Central Asian,” provenance can therefore not be inferred for one of the most ancient iconographies of mankind. However extensive material evidence of the motif, so far fairly unknown, has been discovered in the Central Asia region from at the least the

late third millennium bc. This body of evidence is investigated by the present writer in a separate, forthcoming monograph which discusses the dragon iconography from 2500 bc to 650 ad. See also Kuehn, 2009, pp. 43–67. 8 ╇ The application of this term has been questioned since it seems to exclude geographically the important branch of the same culture that inhabited Mongolia. Cf. Jacobson, 1999, p.╯173.

introduction inherent parallels in artistic expression as well as the adaptation and incorporation of seemingly “Chinese” and Chinese-inspired Mongol stylistic formulae, in particular under the Ilkhanids, and their amalgamation with the visual contexts of the Central Asian region will be addressed in the first part of the Epilogue. Characteristics of the (serpent-)dragon Since earliest antiquity the dragon has been a richly multivalent symbol of complex mythical and symbolic value characterised by a coalescence of maleficence and beneficence. Owing to this inherent polyvalence and ambiguity, it has been called “one of the most complex symbolisms of the history of cultures.”9 Its iconography is a recurring and popular image in the architecture and the arts of the medieval Islamic world. Yet despite its wide diffusion, the symbolism that survives from the Central Eurasian world of the medieval period is often elusive and even cryptic. The composite mythical creatures are endowed with features or parts belonging to various animals generally recognisable across cultural-aesthetic boundaries, the reptilian, feline and raptorial motif being prevalent in the overall composition; they thus often carry chthonic, aquatic and aerial aspects. It is however the reptilian characteristics that predominate in the iconography of the medieval Central Asian dragon. This is not only displayed in visual information but also demonstrated by written sources. In a passage from the Arabian collection of tales of the Alf layla wa-layla (“Thousand and One Nights”),10 the physician asks the slave girl Tawaddud to name him a serpent that lays eggs. In response she names the dragon, in other words, a grown serpent.11 The same notion is expressed by the Ghaznawid poet Masʿūd-i Razī when he advises sulṭān Masʿūd ibn Maḥmūd of Ghazna (r. 421/1030–432/1040): Do not extend security and do not give time and opportunity [to the enemy] Given time the snake turns into a dragon.12 ╇ Le Goff, 1980, p.╯162. ╇ A collection of stories in Arabic, Thousand and One Nights, appears to have formed around a Persian framework and to have developed with many additions from various locations from the ninth and tenth centuries, taking final shape in the thirteenth century. Cf. Littmann, “Alf layla wa-layla,” EI2 I, 358b. 11 ╇ Cf. Tausendundeine Nacht (Brandenburg, 1973, p.╯70), which contains this episode. 12 ╇ Rīzā Qulī Khān Hidāyat, Majmaʿ al-fuṣaḥāʾ, ed. 9

10

5

The tenth-century compilations of the IkhÂ�wān al-Ṣafāʾ (Brethren of Purity), a sect of the Ismāʿīlis, similarly note that the dragon, king of all crawling creatures, has the viper as wazīr.13 That the dragon is in fact a large serpent is noted much earlier in the writings of the fifth-century Armenian apologist, Eznik of Koghb.14 Literary sources of various types, from practical writings (that is to say, pharmacopoeia, travelogues or books on magic), to theological and exegetical writings, poetry, fables and in particular epics,15 such as the early eleventh-century Persianlanguage masterpiece, the Shāh-nāma (“Book of Kings”), prove invaluable in the effort to establish a relationship between dragon iconography and its possible iconological content, in other words the endlessly varied contemporary cultural concepts that generated these notions. Such variations result in an apparently limitless repertoire of iconographical formulae for the dragon which, according to the Shāh-nāma: lived in the water and overland now in the river and anon in the sun [i.e., on the earth], and could pull a ferocious elephant with its tail.16

This verse describes the dragon’s ability to undergo environmental and spatial changes. In his wellknown bestiary, Ḥayāt al-ḥayawān al-kubrā (“The Life of the Biggest Animals”), the fourteenth-century scholar Kamāl al-Dīn al-Damīrī (745/808– 1344/1405) similarly observes that: Serpents are originally in their nature aquatic [creatures] and can live in the sea after having been land ones, and on land after having been marine ones.17

Allusions such as these reveal that the dragon was able to cross boundaries within its natural environment, metamorphosing from land to sea creature and vice versa. The physical changes accompanying such shape-shifting all form part of the dragon iconography in medieval Islamic art so that the creature is, for instance, portrayed variously without legs, with two forelegs or with Muṣaffa, M., 6 vols., Tehran, 1961, vol.╯2, p.╯1170, as cited in Daneshvari, 1993, p.╯17, n. 13. 13 ╇ Ikhwān al-Ṣafāʾ, tr. and ed. Dieterici, 1858, pp.╯83–4. 14 ╇ Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, pp.╯593–4, ch. 133. 15 ╇ For a discussion of Islamic mythology comprising the creation myths, the lives of the Prophets and eschatology, see Thackston, 1990, pp.╯186–201. 16 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯6, p.╯41, ll. 427–9. 17 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯636.

6

introduction

four legs. It may thus have a quadruped body, a serpentine body or a quadruped protome extending into ophidian coils. Overall however the serpentine body with or without legs enjoyed greater prominence. In its astrological manifestation, the dragon necessarily has a celestial quality which, as the sage astrologer Jāmāsp relates to Gushtāsp (Av. Wishtāspa, the Greek Hystaspes), the Kayanian king of Iranian traditional history and first Mazdaist on the throne, is all-powerful: No one can safely pass that fateful wheel. Who has by wisdom or by manliness escaped the knifesharp claws of that celestial dragon? What has to be will be. There is no doubt. The shrewdest man has not escaped his fate.18

The avian aspect of the dragon is often expressed through its portrayal with wings. The latter are assoÂ�ciated with the power of flight, a well-known vehicle for the transition from one realm into another. An early Armenian translation of the thirdcentury Christian theologian Origen’s writings underlines the fiendish nature of the dragon: And we call vishap many of the largest animals; of those on land, the elephant and the serpent, and an evil and violent man, but when the names are once applied, they do not change their nature. We call vishap also the invisible evil power, which, asking power of the Lord, struck the righteous man with grievous blows, not in one part, but in all his parts, outer and inner.19

Yet it is also interesting to note the assertion that vishaps (Av. vishāpa) can fly, as stated by the thirteenth-century Â�Armenian philosopher and historian, Wahrām Wardapet (also known as Rabuni Sevleṛntsʿi), in a letter to the Armenian king Hetʿum of Cilicia (Lesser Armenia, Armenian kingdom from 1198–1375): Many men have seen vishaps ascend from earth to heaven.20 18 ╇ The Story of Rustam and Isfandiyār, tr. and ed. Clinton, 1999, p.╯33; for Gushtāsp’s family tree, idem, pp.╯24–5. 19 ╇Yerevan, Matenadaran MS 6036, fols. 124b-125a, cited by Petrsyan, E., and Najaryan, H., eds., Nshkharner Oroginesi haykakan tʿargmanutʿyunnericʿ, vol.╯1, Ejmiatsin, 1979, pp.╯2, 22, as referred to by Russell, 1987, p.╯207. 20 ╇ Alishan, G., Hin hawatkʿ kam hetʿanosakan krōnkʿ Hayokʿ (“The Ancient Faith or Pagan Religion of the Armenians”), Venice, 1910 ed., p.╯187, as cited in Russell, 1987, pp.╯206–7. It is of note that before the Fall the Genesis serpent is described as a winged creature with legs. Such a giant winged quadruped serpent is portrayed, for instance, in the wall paintings showing events related to the book of Genesis on the drum of the dome (far right) in the Armenian pala tine church of the Holy Cross at Aghtʿamar (915–921).

However, unlike in most of Christian culture where the overall image of the serpent or dragon is predominantly associated with its portrayal in the Bible as the epitome of evil and sin,21 the position of serpents and dragons in Islamic culture is ambiguous and can have benevolent as well as malevolent connotations. The serpent as symbol of evil does not exist in the Qurʾān where it appears only once in the story of the staff of Mūsā (Moses) metamorphosed into a serpent (sūras 20, 20; and 79, 16).22 However, both serpents and dragons figure more frequently in Persian than in Arabic tradition. This ambiguity in the nature of the dragon is also mirrored in the Persian language, the word for dragon (azdahā) being used to describe “a strong and brave man,” or “passionate testy person” as well as “a tyrant.”23 A positive image of serpents or dragons as powerful, friendly and helpful beings persists in Arabic as well as Persian poetry.24 A grateful serpent is depicted in the pre-Islamic Arab writings of the sixth-century poet ʿAbid ibn al-Abraṣ who when travelling through the desert with members of his tribe, the Banū Asad, took pity on a serpent that was tormented by thirst and gave it his last drops of water to drink. During the night the camels bolted and vanished. When ʿAbid, close to despair, was searching for his mount, he heard the voice of the grateful serpent offering him a camel to ride. Because of his meritorious act ʿAbid was thus one of the surviving members of the outing to return to the clan.25 Another story of a lifesaving serpent is recorded by al-Damīrī, citing the shaykh Abu ’l-Ḥasan ʿAlī ibn Muḥammad al-Muzayyin al-Ṣagīr al-Ṣūfī (d.╯328/939–40), who fell into a well in the desert of Tabūk and was saved by a viper which, he states: …wound itself round me, whilst I remained perfectly still in my heart without any emotion; it then twisted its tail round me and took me out See Mathews, 1982, pp.╯245–57; Thierry, 1987, p.╯384, fig.╯266. The fact that the Genesis serpent is winged is also mentioned in the Jewish Apocalypse of Moses, 26; Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯5, pp.╯123–4, n. 4. 21 ╇ On the positive aspect of the serpent in western medieval symbolism, cf. Le Goff, 1979, pp.╯53–90, repr. 1978, pp.╯236–79; English tr., pp.╯159–88. 22 ╇ For the figure of the serpent in Arabic culture, cf. Ruska, “Ḥayyā,” EI² III, 334b; Kopf, “Afʿā,” EI² I, 214b. 23 ╇ Steingass, 1892, repr. 1981, p.╯45. 24 ╇ Touching stories of helpful serpents were also known in antiquity; see Pliny, Naturalis Historia VIII 61; Aelian, De Natura Animalium 6.17 and 63. 25 ╇ The story goes back to Ibn al-Kalbī and is dismissed

introduction of the well, and then untwisting its tail from my body went away.26

Farīd al-Dīn ʿAṭṭār’s (b. c. 513/1119) hagiography Tadhkirat al-awliyāʾ (“Memoirs of Saints”) contains another image of a benign serpent, described as fanning the mystical lovers from time to time “with a branch of narcissus held in its mouth.”27 In Abū Ṭāhir Ṭarsūsī’s twelfth-century compilation of prose narratives, Dārāb-nāma (an Iranian recension of the Alexander Romance), the hero, Dārāb (Darius), is confronted with a sympathetic dragon which helps him to find his abducted mother, queen Humāy.28 The story of Ardashīr in the same epic begins with the tale of the origin of a dragon that evolved out of a worm in an apple, perhaps representing a romanticised account of the introduction of sericulture into Iran,29 when the sight of silkworms transforming into spinning cocoons must not have been uncommon. Among the collection of fables entitled Marzubān-nāma (“Tales of Marzubān”) recorded by Saʿd al-Dīn Warāwīnī in 607–22/1210–25, who presented his collection to Abu ’l-Qāsim Rabīb al-Dīn, the vizier to the Ildeñizid/Eldigüzid atābeg of Azerbaijan (Ādharbyjān), Özbek ibn Muḥammad, there are five stories about serpents. One of these accounts deals with a pious, generous serpent who has the power to interpret dreams and who saves a weaver from punishment by helping him to remind the king of his forgotten dreams. It selflessly continues to help the weaver even though the latter deceives the serpent on two occasions.30 The stories thus portray the serpent-dragon’s compassion as a sign of innate benevolence, high merit or kindness, exemplifying human virtues. On the other hand, the awesome and terrifying nature of the serpent-dragon forced humans into a subordinate, defensive role, thus for instance the by Abu ’l-Faraj (XXII, 85–6) as a “manifest fabrication.” Cited after Kilpatrick, 2003, p.╯117. 26 ╇ Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯58. 27 ╇ Ed. Nicholson, R.A., Tehran, 1370/1991, pp.╯46, 184, as cited in Gohrab, 2000, p.╯86. Cf. al-Hujwīrī, Kashf al-Maḥjūb, p.╯118; Gohrab, 2003, p.╯81. 28 ╇ Gohrab, 2000, p.╯85. 29 ╇Yamamoto, 2003, p.╯75. 30 ╇ Tr. Levy, pp.╯222–7. 31 ╇ Christensen, 1931, tr. 1993, p.╯23. 32 ╇ According to Zaehner (1961, p.╯162): “...the extraordinary zest with which the Magi are alleged to have killed ‘with their own hands’ flying and creeping things, can scarcely be accounted for except on the supposition that they thought such creatures to be the handiwork of an evil power. It is they, then, who would be responsible for the cut-and-dried division of creation into two mutually antagonistic halves—the creatures of the Holy Spirit on the

7

Yasht hymn (“Songs of Praise”) collection of the surviving Avestan texts, the earliest scriptures of the ancient Persian religion, Zoroastrianism, lists not only various types of legendary or mythical “first man” or “first king,” but also dragon men and killers of dragons, transmitted mainly from the Indo-Iranian period.31 The later Zoroastrian scriptures of the Vidēvdāt (Vendidād), perhaps influenced by the customs of the Median priests, the Magi,32 contained a radically reconfigured view of the universe. The “law against the daevas” divided “creation into two mutually antagonistic halves—the creatures of the Holy Spirit on the one hand and the creatures of the Destructive Spirit on the other.”33 According to this understanding serpents or dragons (Av. azhi-, Pahl. azh-) were identified as creatures of the “hostile spirit” Ahriman. They were defined as evil, Â�noxious, harmful to man and his animals and crops (Av. khrafstra)34 and thus deserving of death.35 This inherent ambiguity is exemplified in the demon Azhi Dahāka/Azhdahāk found in the Avestan texts, the notorious dragon who tried to seize the khvarәnah- (Mid. Pers. khwarrah “glory, God-given fortune, splendour”) of Iran’s Aryan rulers of traditional history, attempting, in other words, to make himself ruler of the Aryans.36 After several great battles, he was overcome by the dragon-fighter Thraētaona/Frēdōn (the Avestic counterpart of the Vedic dragon-slayer Indra). Hence from an early time, variants of this epic seem to have attributed to usurpers some traits that seem to have been borrowed from the dragonman.37 Long familiar as a monstrous tyrant,38 he becomes in New Persian or Arabic narratives the Babylonian tyrant Ẓaḥḥāk (al-Ḍaḥḥāk)/Dahāk, who belonged to the Pīshdādian, the early mythione hand—and the creatures of the Destructive Spirit on the other. Thus they can be regarded as the true authors of that rigid dualism that was to characterize the Zoroastrianism of a later period, but which is only implicit in the Gāthās [“songs”] of Zoroaster.” 33 ╇ Zaehner, 1961, p.╯162. 34 ╇ Boyce, 1975, repr. 1996, pp.╯90–1. The special stick used by the Zoroastrians to kill noxious creatures of various kinds is called a mār-gan (“snake-killer”); Russell, 1987, p.╯461. The custom of killing of khrafstras is also mentioned by Plutarch (De Iside et Osiride 46; De Invidia et Odio 3.537B; Questiones Conviviales 4.5.2.670D). 35 ╇ Vidēvdāt 14.5; 18.73. 36 ╇ Cf. Christensen, 1931, tr. 1993, p.╯26; Gershevitch, 1959, p.╯59; Zaehner, 1961, pp.╯150–3; Sarkhosh Curtis and Stewart, eds., 2005, pp.╯102–3. 37 ╇ Christensen, 1931, tr. 1993, p.╯27. 38 ╇Yasht 5.29–30; 15.19.

introduction

8

cal Iranian kings who established civilisation, and, in turn, became the son of a king of the Arabs.39 Having overthrown the Iranian king Jamshīd (Av. Yima Khshaēta “Yima the brilliant”) with popular support, Dahāk is corrupted by Iblīs/Satan,40 and from this time snakes issue from his shoulders, his demonic human-to-hominoid-dragon transformation thus representing a form of moral retribution.41 He then imposes his tyranny on Iran for a thousand years until he in turn is overcome by the conqueror Frēdōn who imprisons him in Mount Damāwand.42 It is noteworthy that this paradigm of an evil king nevertheless also enjoyed a favourable reputation in Iranian history.43 One of the astrologers of the ʿAbbasid caliph Hārūn al-Rashīd, Abū Sahl al-Faḍl ibn Nawbakht, working at the caliph’s proverbial Treasure House of Wisdom (Khizānat al-ḥikma), the great library, translation bureau and institute for the promotion of the philosophical sciences, describes Ẓaḥḥāk as founder of palaces of science and as living in a domain governed by the beneficent planet Jupiter.44 Hence, by implication, ibn Nawbakht equals Ẓaḥḥāk, the founder of palaces of science, with the caliph, who was the founder of the celebrated Khizānat al-ḥikma.45 Furthermore, tribal confederacies, dynasties and heroes identified with

the dragon and claimed their descent from the demonic king. Rustam, the hero par excellence of the Iranian epic (in particular in Firdawsī’s Shāh-nāma), traces his descent to Ẓaḥḥāk/Dahāk,46 as did the Kushāṇas of the Yuezhi confederacy (c. first–third centuries) who ruled over the Central Asian regions which comprise present-day Tajikistan, Afghanistan, Pakistan and northern India47 as well as the Islamic dynasty of Ghūr.48 As Shapur Shahbazi has cogently argued, Dahāk is believed to have been “the eponymous father of a formidable Iranian people,”49 the Dahae/ Dahī, Sacians who formed the core of the Arsacid invaders (12–428) of Parthia, one of the five divisions of the Iranians.50 Descent from Dahāk was moreover claimed by the Armenians of the region near Lake Sevan.51 Finally, in the Turkish epic Ṣaltūq-nāma (“Book of Ṣaltuq”), the first ruler of the world, Eslem, son of Adam, becomes the father of Ẓaḥḥāk the Turk, ancestor of all Turkish sovereigns.52 However, at the same time the dragon Azhi Dahāka/Dahāk was in some cases regarded as “the incarnation of the demonic par excellence.”53 His symbolic value was drastically “historicised” and identified by various societies or groups with real or external enemies such as foreign nations or oppressive powers or rulers.54 In his Patmutʿiwn

╇ See Yarshater, 1983a, pp.╯426–9. ╇ In Qurʾānic tradition Iblīs is both an angel (sūra 20, 34) and “one among the jinn” (sūra 18, 50). An important difference between Islamic and Christian perceptions regarding Satan (from the Hebrew śāṭān, “adversary”) hence lies, according to Arent Jan Wensinck (“Iblīs,” EI² III, 668a), in the fact that: “Muslim thought remains undecided as to whether he was an angel or a jinn, and does not pronounce an opinion on the possibility of his being a “fallen angel.”↜” 41 ╇ See a detailed description of this episode in al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯19–27. Cf. Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯63–5, ll. 178–97, pp.╯69–71, ll. 14–44; vol.╯2, pp.╯45, 60, 75. 42 ╇ Cf. Boyce, 1975, repr. 1996, pp.╯67, 91, 98, 100, 103, 283, 289, 293. Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯113, ll. 518–27. 43 ╇ Shahbazi, 1993, p.╯159. 44 ╇ Ibn Nawbakht, Kitāb al-nahmaṭān, quoted in Ibn al-Nadīm, Kitāb al-Fihrist, Cairo, n.d., pp.╯345–8, as cited in Pingree, 1968, p.╯9 and ns. 2–4, p.╯10, n. 1, pp.╯11–2, 69. 45 ╇ Pingree, 1968, p.╯12. 46 ╇ With the approval of his grandfather Sām, Rustam’s father, Zāl, married Rūdāba, the daughter of Mahrāb, the king of Kābul, a descendant of Dahāk; al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯73–97. Cf. Khāleqī-Moṭlaq, 1971, pp.╯31, 35, 39–40; Monchi-Zadeh, 1975, pp.╯109–10; de Bruijn, “Sām,” EI2 VIII, 1011a. For Rustam’s family tree, see The Story of Rustam and Isfandiyār, tr. and ed. Clinton, 1999, p.╯26. In the Shāh-nāma Mahrāb is described as idolater; Monchi-Zadeh (1975, pp.╯109–11,

142–3) associates the name of the king of Kābul, Mahrāb, with the title Mahrāj (= Mahārāja), hence linking him with India. Cf. von Spiegel (1871, p.╯567) who has considered him to be Buddhist. 47 ╇ This is indicated by the story of Kūsh, the nephew of Ẓaḥḥāk and founder of the Shar-i Kūshan (= Kūshānshar), noted in an epic of Īrānshāh, the son of Abu ’l-Khayr (Safa, Ḥamāsa sarāʾī dar Iran, pp.╯296–300; cf. The Mujmal al-Tawārīkh, pp.╯89, 187, 189), as cited in Shahbazi, 1993, p.╯159. 48 ╇ Quoted by ʿUthmān ibn Muḥammad al-Jūzjānī (fl. c. 685/1260), the historian of the Ghurids; Bosworth, “Ghūrids,” EI² II, 1099a. Cf. Shahbazi, 1993, p.╯159. 49 ╇ Movses Khorenatsi, Patmutʿiwn Hayocʿ (“History of the Armenians”), p.╯127: “The one they [= the Persians, in other words some Iranians] call Biurasp [Bīwarāsp] Azhdahak was their ancestor,” cited after Shahbazi, 1993, p.╯159 and n. 123. 50 ╇ Bailey, 1959, pp.╯71–115. 51 ╇ Khorenatsi, Patmutʿiwn Hayocʿ, II.49, cited after Shahbazi, 1993, p.╯159. According to a reference by Khorenatsi (Patmutʿiwn Hayocʿ, I.30) as well as Thomas Arcuni’s Collection des historiens Arméniens, Petersburg, 1874, p.╯47, there even existed an Armenian noble family called Azhdahāk; cited after Widengren, 1969, p.╯17 and n. 35. 52 ╇ Mélikoff, 1960, vol.╯1, p.╯43 and n. 1; Dedes, 1996, p.╯29, n. 80. 53 ╇ Russell, 1987, p.╯43. 54 ╇ For a brief discussion of the Greek word drakōn as appellation of a historical person or a people in Hebrew, Aramaic and Greek literature, see Schlüter, 1982, pp.╯44–6.

39 40

introduction Hayocʿ (“History of the Armenians”), which is ostensibly written in the fifth century but probably dated to the mid-eighth century in its present form,55 the Armenian historian Moses of Chorene (Movsēs Khorenatsi) calls the neighbouring Medes, mar, which is a homonymic of the Persian word mār (“snake”). The History refers to their offspring as “progeny of the dragon” (vishap-azun), or human snakes,56 while the archetype of evil misrule, Azhi Dahāka/Azhdahāk, the dragon in man-shape (or the human in dragonshape)57 of the Sasanian epics, is identified with the historical Median king Astyages (Med. Â�Rishtivaiga) against whom the Armenian king Tigran rebelled.58 Khorenatsi also refers to the first century ad invading Alans, an Iranian people of the Caucasus, and their offspring, as descendants of Azhdahāk.59 The name of the latter continued to be used as a symbol for historical enemies; in particular, the new world power of the Saljuq Turks, whose conquests provoked a sense of the apocalyptic in medieval Christian Transcaucasia.60 The eleventh-century Armenian scholar and theologian, John (Yovhannes) of Tarawn, declared that the Antichrist – the dragon bound, at the time of the Crucifixion, for a thousand years – was now free once again and had returned with the help of the Saljuq Turks.61 In the discussion of the story of Paradise in post-Qurʾānic canonical traditions, the serpentdragon’s inherent ambivalence is also expressed. In the primordial Paradise the serpent is said to have been the most beautiful and strong of animals,62 who was: ...shaped like a camel and like the camel, could stand erect. She had a multi-coloured tail, red,

╇ See, for instance, the “Introduction” of Khorenatsi: History of the Armenians, tr. and ed. Thomson, 1978, repr. 1980. 56 ╇ Ishkol-Kerovpian, “Višap,” WdM IV, 1, pp.╯155–7; Russell, 2004, p.╯627. 57 ╇ Cf. Schwartz, 1980, pp.╯123–4. 58 ╇ Purporting to transmit a report of the earlier historian Mar Apas Catina, Khorenatsi notes this in a passage of his Patmutʿiwn Hayocʿ (tr. Langlois, 1872, p.╯39). Russell, 2004, p.╯1170. 59 ╇ Alemany, 2000, p.╯285. 60 ╇ Cf. White, 1991, p.╯193. 61 ╇ Their invasion is described by Matthew of Edessa (Mattʿēos Uá¹łhayetsi), who uses visionary apocalyptic imagery: the Turks are “winged serpents” (ojkʿ tʿewaworkʿ) or “death-breathing dragons” (vishapn€mahashuntch). Russell, 2004, p.╯883. 62 ╇ Cf. Wheeler, 2002, p.╯25. 63 ╇ Al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯38. 64 ╇ Idem, pp.╯39, 53. It is of note that of all the animals 55

9

yellow, green, white, black, a mane of pearl, hair of topaz, eyes like the planets Venus and Jupiter, and an aroma like musk blended with ambergris.63

Owing to the instigation of the bird of Paradise, the peacock (ṭāʾūs), which he saw at the gate of Paradise, Iblīs made use of the serpent and managed to trick her by speaking to her in a soft voice until she had confidence in him: she opened her mouth, ... Iblīs jumped in and sat down between her fangs (thus the fangs of snakes became poisonous until the end of the ages)64

and so he eluded the angels guarding Paradise who would not have admitted him. A narrative ascribed to Wahb ibn Munabbih (b. 34/654–5), a Yemenite descendant from a family of Persian origin, describes the Fall which led to the expulsion from Paradise: When Iblīs wanted to cause [Adam and Eve] to slip, he entered into the stomach (jawf) of the serpent; the serpent [then] had four legs and was like a Bactrian [camel] (bukhtīya), one of the most beautiful creatures God had created. When the serpent entered the garden, Iblīs came out of its stomach (jawf); he took [a fruit] from the tree [the Tree of Immortality (Qurʾān, sūra 20, 116–21)] that God had forbidden to Adam and Eve and brought it to Eve.65

As a consequence of the service rendered to Iblīs, the serpent is not only banished from Paradise, but loses her legs, which reenter her body; she will dwell in dark places and only earth will be her food;66 she is condemned to crawl on her belly becoming “malformed and deprived of the power of speech, mute and forked-tongued.”67

in Paradise the serpent and the peacock are singled out to become the pawns of Iblīs; both were severely punished (idem, pp.╯46–7) but only the peacock was rehabilitated (idem, p.╯53). 65 ╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, vol.╯1, p.╯108; see also idem, Jāmiʿ al-Bayān, I, p.╯235, cited after Katz, 2002, p.╯179. Jewish Midrashic literature similarly records that the serpent of the Garden of Eden originally had feet; Gray, 1906, p.╯186. It is worth mentioning that in the Qurʾān it is not Eve who entices Adam to disobey God; Iblīs speaks to both and in one instance only to Adam (sūras 7, 20–2; 20, 120–2). 66 ╇ Al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯53; al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, vol.╯1, pp.╯525–6. 67 ╇ Al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯46. See also the second-century bc Hebrew work, Book of Jubilees 3.28, as well as Philo of Alexandria, De Opificio Mundi 55.156.

introduction

10

Existing contributions to the study of dragon iconography As far as the symbolism of the dragon is concerned, the field for the greater part has been prepared through valuable studies in stylistic and iconographic development, of which those of Giovanni Curatola, Katharina Otto-Dorn, Gönül Öney and Abbas Daneshvari rank foremost.68 Considering the prominent position accorded to the dragon motif in the arts of the medieval Islamic period, it is however surprising that so far only Giovanni Curatola has devoted a study exclusive to this iconography and its characteristics. His study, which investigates the overall relation of the dragon to Chinese and Central Asian sources as well as its appearance in manuscripts of the post-Timurid period,69 leaves room however for the exploration and identification of the larger phenomenon, its manifestations and crosscurrents across a period of many centuries and many cultures. Some studies have concentrated on a specific medium. Otto-Dorn70 and Öney71 have focused in particular on the monumental sculpture of Anatolia, providing an extensive catalogue of known monumental stone reliefs. Individual reliefs were investigated by Jean-Paul Roux72 and Özden Süslü73 in Saljuq-period Anatolia and by Joachim Gierlichs in northern Mesopotamia (the Jazīra).74 In a more recent work Gierlichs established an important catalogue raisonné of animal reliefs on monuments of the Saljuq and Artuqid periods and their successors throughout Anatolia and the Jazīra, focusing also on the depiction of the dragon.75 The mass of data thus assembled is truly exhaustive and provides a firm foundation for further research. The present study consists of a total of 14 chapters. Chapter 1 outlines the historical and cultural context within which the dragon iconography flourished in the medieval Islamic world from Central Asia to Anatolia. Representations of the dragon on monumental sculpture, both Islamic and Christian, of the medieval period, mainly in

the form of architectural decoration, are discussed in chapter 2. The symbolic significance attributed to the dragon is closely interlinked with the perception of how the boundaries between the realm of the supernatural creature and man are negotiated. It serves as a liminal marker, constituting at the same time a powerful protective device. Chapter 3 turns to portable art, where the dragon appeared as an expression of fundamental social, moral and sociological concepts as well as a metaphor of sociopolitical authority and ideal rulership.╯The symbolic appropriation of the dragon and control over it figured also among the paraphernalia of heroism and rulership as well as appearing on objects of personal adornment and on vessels. It is moreover noteworthy that, while there is a large body of dragon depictions on portable items from the entire Central Asian region, of which a selection is examined here, their existence on architectural structures in the Western Central Asian (previously also “Eastern Iranian”)76 world has so far not been documented. This is due to the fact that no figural sculpture is associated with the brick architecture of the Iranian world from about 1000 to 1200. Moreover, comparatively little architecture of this period survives from the “crossroads of Asia,” i.e. the region of present-day Afghanistan. Hence, only the representation of dragons on monumental settings in regions west of Iran are considered. Many pre-Islamic thought systems and practices were assimilated into early Islamic culture. Beliefs in spirits or jinn and their manifestation as serpent genii have been studied for example by Joseph Henninger77 and Ernst Zbinden,78 and these are considered in chapter 4, which analyses the intricate connection of dragon iconography with a multiplicity of natural phenomena as the means through which the continuous correlation and interchange between human society and the natural world were mediated. Dragons are paramount symbols of the elements or forces present or active in the cosmic world. This chapter focuses on the dragons’ association with the four great ele-

╇ Curatola, 1979; idem, 1982; and idem, 1989; Â�Otto-Dorn, esp.╯1978–9, pp.╯25–36; Öney, 1969a, eadem, 1969b, and eadem, 1978; Daneshvari, 1993. 69 ╇ Curatola, 1979; idem, 1982; and idem, 1989. 70 ╇ Otto-Dorn, 1959, pp.╯63–5, eadem, 1963, pp.╯131–3, eadem, 1978–9, esp.╯pp.╯25–36. 71 ╇ Öney, 1969a, eadem, 1969b, and eadem, 1978. 72 ╇ Roux, 1972, and idem, 1980. 73 ╇ Süslü, 1987. 74 ╇ Gierlichs, 1995.

╇ Idem, 1996, pp.╯28–40; also idem, 1993, and idem, 1998. ╇ The term is used in a geographical sense to indicate all the regions that stretch from the Caspian Sea and the Central Iranian desert in the west to the Indus river in the east, and from the coastal strip along the Arabian Sea in the south to the banks of the Syr Daryā in the north; hence comprising part of present-day Iran, Afghanistan, Pakistan, Turkmenistan, Uzbekistan and Tajikistan. 77 ╇ Henninger, 1963. 78 ╇ Zbinden, 1953.

68

75 76

introduction

11

ments—earth, water, fire and wind. They express, in a mythical language, aspects of the natural setting and the positive or dangerous qualities of those aspects, such as rain, drought or flood. Their particular connection with the medium of water involves a nexus of ideas which also determines the dragon’s affinity with other symbols of fertility of the vegetal world. The chthonic nature of the mythical creature may be associated with the belief in dragons guarding both treasures hidden in the earth and the sources of nature’s abundance. The ability of serpents and dragons to undergo physical change underlay, moreover, their association with numinous or demonic power. The overall physical division of animals from humans and the great diversity and physical distinctiveness among animal species have made animals the preferred medium for the representation of the other, non-human, sphere, which was often characterised by the effortless passage across physical and geographical boundaries. Dragons within the “natural” animal species and the realm of the supernatural and the strange, as well as the visual fusion of two or more animals, are looked at in chapter 5. Beyond a relation with the “natural” world, the dragon is embued, on a cosmic level, with a more complex meaning as a bringer of stability or disorder, stasis or dynamism, life or death. As mentioned earlier the mythical creature has a similar meaning on a “social” or “political” level, symbolising the enemies or, as will be shown, in some cases, the champions of a society, group or class. By using the iconography of the dragon, rulers aimed to ensure symbolic control over subject dominions. This pivotal role accorded to the dragon, which figures prominently in royal and heroic symbolism and as an indicator of political power, intellectual supremacy and socio-religious dominance, is discussed in chapter 6. Among those who have worked on these concepts is Eva Baer,79 who has formulated the expression and thoroughly explored the iconography of “the ruler in the cosmic setting,” a form of imagery closely

linked to the ancient concept of the royal or heroic dragon-tamer. The dragon’s overall symbolism as well as the dragon-fighting myths in Old and Middle Iranian and Persian literature have been examined in depth, respectively, by Prods Oktor Skjærvø and Djalal Khāleqī-Moṭlaq,80 while its Armenian aspect is treated by James Russell.81 In his book, How to Kill a Dragon: Aspects of IndoEuropean Poetics, Calvert Watkins similarly looks in detail at the structure of the ancient myths on the slaying of dragons, which recur throughout the Indo-European poetic tradition.82 Many of these topics have most recently been reinvestigated and expanded by Michael Janda in his wonderful 2010 book, Die Musik nach dem Chaos: Der Schöpfungsmythos der europäischen Vorzeit.83 The dragon combat motif played a central role mainly in the chronicles of heroic combat, in tales of romance, or in allegories of mystical initiation or religious teaching.84 The slaying or subduing of a dragon by a hero or divinity in ancient myth, and its iconography and mode of transmission in the medieval Islamic world as well as in the Eastern Christian sphere, are explored in chapter 7. A further important aspect is the eschatological role fulfilled by some dragon fighters. The role of the equestrian warrior saints as dragon-slayers has been extensively researched by the prominent scholar of Byzantine hagiography and art, Christopher Walter.85 More recently Oya Pancaroğlu has examined the role of the itinerant dragonslayer in medieval Anatolia.86 Chapter 8 aims to uncover further layers of dragon iconology in the context of astrology, a highly esteemed science in the east, its association with light and its role as a vehicle to convey cosmological ideas. The dragon representation in medieval Islamic astrology has been addressed in a number of studies, foremost among which remains that of Willy Hartner (1938) who demonstrated over seventy years ago the influence of the conceptualisation of the two pseudo-planetary “lunar nodes” (al-ʿuqdatāni) on Islamic artisans.87 Guitty Azarpay seeks to explain the theme of the

╇ Baer, 1981. ╇ Skjærvø, “Aždahā I,” EIr; Khāleqī-Moṭlaq, “Aždahā II. In Persian Literature,” EIr. 81 ╇ Russell, “Aždahā IV. Armenian Aždahāk,” EIr, and idem, 2004. 82 ╇ Watkins, 1995. In the mythologies of Indo-European speaking societies (Indic, Iranian, Hittite, Greek, Roman, Germanic and Armenian) versions or traces of a type of myth have been found wherein a god or hero overcomes a mythical dragon-like creature.

83 ╇ I am indebted to Professor Almut Hintze for pointing it out to me. 84 ╇ Russell, 2004, pp.╯1032, 1285–6. 85 ╇ Walter, 1989a; idem, 1995; idem, 1999, and idem, 2003. 86 ╇ Pancaroğlu, 2004. 87 ╇ Hartner, 1938; idem, 1959, and idem, 1973–4, as well as idem, “Al-Djawzahar,” EI² II, 501b. Cf. Öney, 1969a, pp.╯193–216; Otto-Dorn, 1978–9, pp.╯125–36; Azarpay, 1978, pp.╯363–74.

79 80

introduction

12

interlaced dragons in the light of its astrological significance.88 Moreover, two more recent exhibitions on celestial imagery in Islamic art in New York and Paris also included the astrological association of the dragon.89 Chapter 9 surveys a theme that has previously attracted little attention: the survival within medieval Islamic tradition of the ouroboros, the iconography of a serpent devouring its own tail. Similarly, the motif of the drinking dragon on medieval Islamic and Christian two- and threedimensional art represents the visual remnants that survived the mutations of an ancient belief. Likewise the knotted dragon motif in the Islamic period, related to the ancient magical practice of knot tying for apotropaic ends, discussed in chapter 10, has barely been addressed other than in connection with alchemy.90 The dragon figure also plays a role in the various branches of knowledge that comprise the science of alchemy as well as the more esoteric or occult sciences which include talismanic astrology. Chapter 11 introduces aspects of the dragon as prophylaxis and cure in the province of magic and divination. It discusses the dragon’s association with the theriaca as illustrated in the famed Kitāb al-diryāq. It also looks at serpents, their venoms and other characteristics as well as the corresponding antidotes, chiefly theriaca and bezoar-type stones. The dragon’s role in esoteric or magical sciences, in particular sympathetic and talismanic astrological magic, as well as in the preparation of talismans is examined in chapter 12. The supernatural power of the serpent-staff is the subject of the second section of the chapter, which has been explored in terms of Jewish influences on Islamic magic by Alexander Fodor in the context of a thirteenth-century treatise in which the motif of Moses’ serpent-rod is discussed as a magical device.91 In chapter 13 the association of the dragon with the sound-world is considered. One of these aspects is the speaking dragon; another more rare occurrence is its link with sounds transformed into music. Chapter 14 then concludes the discussion,

focusing on the dragon as pre-eminent symbol and paradigm of change and transformation, thus addressing a key aspect of its phenomenology. The most important study of the iconography of the dragon in the cult of the saints and mystics is provided by Abbas Daneshvari,92 whose research moreover offers a significant contribution to the iconological interpretation of the multivalent symbolism of the dragon by underlining its beneficial and apotropaic aspect.93 In medieval Ṣūfīsm (taṣawwuf), the mystical dimension of Islam, the symbolism of the dragon illustrates the theme of moral transformation on the spiritual path, often mediated by the figure of a mystic. The first part of the Epilogue shows how in the wake of the Mongol invasion dragon imagery appears in different stylistic guises. The first half of the Ilkhanid period presents a transitional period in which a new “Chinese,” “Mongol” or, more broadly, “Central Asian” style appears which in the second half of their reign gradually overtakes and amalgamates with the conventional “Saljuq-style” dragon. This emergence in the Mongol period of a hitherto unknown style, termed “Islamic chinoiserie,” from the relationship between Chinese and Iranian art, has been investigated by Yuka Kadoi.94 The second part demonstrates the eminent role played by the dragon, itself the ideal image of incarnate liminality, in frontier societies in Transoxania, Khurasan and Anatolia characterised by ghāzīs fighting for the defence and victory of Islam. This is evident in the epic-chivalrous frontier narratives describing jihād against dragons, as for instance in the early Turkish Anatolian epic, the BaṭÂ�ṭāl-nāma (“Book of Baṭṭāl”), which was more recently translated by Yorgos Dedes,95 or the epic romance of the Türkmen Dānishmend ruler, the Dānishmend-nāma (“Book of Dānishmend”), also based on orally transmitted traditions and composed in the twelfth or early thirteenth century, translated by Irène Mélikoff,96 who in the romance is identified with Sayyid Baṭṭāl.97 This phenomenon was complemented by the crosscultural convergence of saintly cults prominently

╇ Azarpay, 1978. ╇ Following the Stars, 1997; L’Apparence des cieux, 1998. 90 ╇ Rogers, 1969; Moulierac, 1987; Savage-Smith, 1997, pp.╯324–33. 91 ╇ Fodor, 1978. 92 ╇ In December 2004 Abbas Daneshvari gave a series of four lectures (the Yarshater Lectures) at the School of Afri-

can and Oriental Studies in London on the iconography of the dragon in Persian art. His publication on the subject is eagerly awaited. 93 ╇ Daneshvari, 1993. 94 ╇ Kadoi, 2009. 95 ╇ Dedes, 1996. 96 ╇ Mélikoff, 1960. 97 ╇ Eadem, 1960, vol.╯1, p.╯103.

88 89

introduction

13

Since there are no previous comprehensive studies of the dragon in the Islamic world,98 the present exercise is prone to all the risks that are characteristic of such an endeavour. Depictions of dragons in the material culture of the medieval Islamic period have never been fully catalogued as a corpus, hence in what follows only a representative sampling will be considered in which the selected objects are individually discussed. Inevitably, numerous potentially relevant objects may have been overlooked or considered too briefly, and the aim of tracking down the exact symbolic significance of the serpent and its greater relative, the dragon, may not have been fully realised. Nevertheless, the approach adopted here will at least bring into focus the complexity of the semantic horizon associated with these images. This endeavour is particularly precarious since modern perception of the elements that determine different modes of interpretation will not necessarily reflect the meaning and mental associations attributed to them by the medieval populations studied. Imagery that elicited clearly identifiable connotations and allusions in the cultural milieu

of the medieval Central Asian world often remains incomprehensible for modern people. Because of the refractory nature of the evidence, some iconographic elements of the period are based upon a number of deductions while some others necessarily remain elusive, even impenetrable. It is therefore the intention of the present study to locate certain iconographic details that so far have not been understood or, perhaps, even recognised. At the same time it has to be borne in mind that what may be considered a symbol in the twentieth century may well represent an authentic transfer of a reality for man in medieval Central Asia and beyond. Even so it is hoped that the following chapters will shed some light on the perception of the great dragon beast within the overall intellectual and visual universe of the medieval Irano-Turkish world. Finally, the path that has led to the realisation of this study has been long and tortuous: the serpent is not called a dragon for nothing. Nonetheless, this study represents but a small contribution to a better comprehension of the complex multivalent symbolism of the dragon in the medieval Islamic sphere. Some propositions consequently have a provisional character that, with the help of new material, specialists will confirm or invalidate.

98 ╇ Giovanni Curatola’s thorough work on the dragon focuses on the overall relation of its iconography to Chinese and Central Asian sources as well as its appearance in Â�manu-

scripts of the post-Timurid period. Curatola, 1979; idem, 1982; and idem, 1989. Other previous studies cited above have mainly focused on particular topics.

involving the dragon in which Islamic, TurkoMongol, Jewish and Christian beliefs overlapped and amalgamated.

the medieval islamic world from central asia to anatolia

15

chapter one

The medieval Islamic world from Central Asia to Anatolia: the historical and cultural context

With its rapid expansion to the status of world power, the Islamic world became increasingly open culturally to the transmission and active appropriation of ancient learning from GraecoRoman and Indo-Iranian sources. The westward movement of the culture of Western Central Asia, in particular that of greater Khurasan, resulted not only in an outflow of savants and artists from this region but in a general tendency to “Easternise.” This coincided with the westward migration of ever-growing numbers of Turkic-speaking tribes into Western Asia which increased from the late tenth century onwards. From the end of the eleventh century until the onslaught of the Mongols in the mid-thirteenth century, the Saljuqs and their “successor states” ruled a large region from India to Egypt, perpetuating the heritage of Western Central Asian art and culture in their new homeland. Two major currents profoundly influenced the formation of the Islamic world from its inception. One was the transmission of ancient learning from the Greek, Central Asian (in particular the Iranian) and Indian cultural realms, provoking an intense intellectual ferment in the Islamic world.1 This was linked with and reflected by the second, which saw the culture of Western Central Asia flowing westwards, facilitated by the largescale migrations of Turkish-speaking people into Islamic lands from the late tenth century onwards.2 With the establishment of the Islamic polity (dār al-islām, “abode of Islam”) in the wake of the Arab conquests after the death of the Prophet Muḥammad in 10/632, the lifting of political and religious barriers from Morocco to India promoted greater movement of goods, people and ideas across a vast region.3 It united areas and

people that for over a millennium, since Alexander the Great (r. 356–323 bc), had been subject to varying degrees of Hellenisation. Throughout the Umayyad period (41/661– 132/750), and possibly beyond the mid-eighth century, Greek was widely current in greater Mesopotamia and Palestine as the native language of a significant portion of the population4 and was moreover cultivated in the many Christian monasteries and cloisters. The cultivation of Hellenistic philosophy and science at centres of learning that had flourished during the first six centuries of the Christian era was well entrenched and further developed in the regions that were part of the Roman, later the Byzantine and Sasanian empires, and finally the caliphate, throughout the Fertile Crescent, from Edessa (al-Ruhā) and Qinnasrīn in the west, through Nisibis and Mosul in northern Mesopotamia to Jundaysābūr in Khūzistān, well into western Iran.5 To these should be added at least two other major centres of Hellenistic Â�science and learning, Ḥarrān (ancient Carrhae) in northwestern Mesopotamia just south of Edessa. In 47/667 the Muslim armies crossed the river Oxus and by 95/713 Transoxania had come within the expanding fold of Islam. The Eurasian heartlands, in particular the cities of Marw and Balkh in Khurasan (covering a wide extent of land comprising regions in present-day Afghanistan, Turkmenistan and Iran),6 which were well-known for their libraries until their destruction by the Mongols in the early thirteenth century, constituted an important locus for integrating and transmitting knowledge.7 After the ʿAbbasid revolution, the transfer of the seat of the caliphate from Syria to Iraq and the building in 145/762 of a new capital, Baghdad (close to the ruins of the Sasanian capital ofÂ�

1 ╇ The most in-depth monograph on the Graeco-Arabic translation movement and the political and social factors involved in it is certainly that of Gutas, 1998. 2 ╇ While the migratory routes of peoples were mostly from east to west, there were also significant concurrent flows in the opposite direction as well as southwards.

╇ Gutas, 1998, p.╯13; Bauer, 1995, pp.╯34–6. ╇ Gutas, 1998, p.╯117. 5 ╇ Idem, p.╯14. 6 ╇ Idem, p.╯14. 7 ╇ Ruska, 1926. Cf. Needham and Wang, 1965, p.╯369; Gutas, 1998, p.╯50 and n. 39. 3 4

chapter one

16

Ctesiphon), placed the ʿAbbasids (132/750– 656/1258) in the heartlands of the former Sasanian empire (c. 224–651). During Sasanian times scholarly activity was partly motivated by a Zoroastrian imperial ideology that would see all learning ultimately derive from the Avestan texts, the Zoroastrian canonical scriptures. It is perhaps in this context that the burgeoning Graeco-Persian translation activities which culminated in the reign of Khusraw I Anūshirwān (r. 531–578), generating a culture of translation that survived into early ʿAbbasid times, are best understood.8 Political and economic support from the Western Central Asian world, in particular from greater Khurasan (in other words the wider Iranian world), was indispensable for the ʿAbbasid victory.9 The early ʿAbbasid caliphs tried to legitimise their rule by expanding their ideology to include the concerns of the “Persian” contingent, thereby bringing about the incorporation of Sasanian culture, still dominant for large parts of the population east of Iraq, into mainstream ʿAbbasid culture.10 Two components of the Sasanian culture, Zoroastrian imperial ideology and political astrology, proved to be of immense significance to the caliph al-Manṣūr, who was in many ways the actual founder of the ʿAbbasid caliphate, in helping him to consolidate the ʿAbbasid cause. Between the second half of the eighth and eleventh centuries, intense scientific activity was accompanied by a prodigious effort to garner and then translate, assimilate and cultivate scientific and pseudo-scientific treatises, for instance on theology, medicine, astrology and logic, mostly via Syriac and Persian (Pahlawī) into the Arabic sphere. Especially in the field of astronomy and astrology translations were often also made from Persian or Sanskrit.11 The translation movement was actively patronised by the ʿAbbasid rulers while at the same time representing a “social phenomenon” which was “subsidized by an enormous outlay of funds, both public and private.”12 Support for these undertakings “cut across all lines of religious, sectarian, ethnic, tribal and linguistic demarcation. Patrons were Arabs and non-Arabs, Muslims and non-Muslims, Sunnīs and Shīʿites, generals and civilians, merchants and land-owners, etc.”13 While the most widely known evidence for Muslim cultural borrowing lies in this vast ╇ Idem, 1998, pp.╯25–6. ╇ Cf. Frye, 2005, p.╯4. 10 ╇ Gutas, 1998, p.╯29; Spuler, 1976, pp.╯342–7. 11 ╇ Pingree, 1963.

corpus of Arabic translations, the simultaneous cultural appropriation was much broader, pervading all modes of life. By the twelfth and thirteenth centuries, Islamic culture had fully internalised and synthesised concepts emerging from a multitude of scientific works acquired largely from Graeco-Roman and Indo-Iranian sources. The movement of craftsmen from eastern Islamic lands, in particular from greater Khurasan, westwards was a decisive process, which accelerated between the fall of the Iranian dynasty of the Samanids (204/819–395/1005) and the late thirteenth century, a period that might broadly be described as the Turko-Mongol era. Hemmed in by deserts both to the south and the north, the Khurasan mountain range and the plains along its slopes have always represented a significant, and often the most important communication artery between east and west. The tendency to easternise was initially an internal phenomenon within Western Central Asia which began with the Iranian Samanid dynasty that ruled in Transoxania (known during the Middle Ages as mā warāʾ al-nahr, “the land which lies beyond the river [Oxus]” at the eastern margin of Khurasan facing the still pagan Turks) and then in Khurasan from the ninth century onwards. For a time, the Samanids constituted the border between the Islamicised lands and the still incompletely Islamicised Turkic dynasty of the Qarakhanids. East of the regions dominated by the Qarakhanids were the lands of the Oghuz tribes from whom the Saljuqs would emerge. Under the aegis of the Samanids and that of other petty courts of the east, Iranian literature flourished, reaching an extraordinary stage of maturity and eloquence with authors such as Rūdakī, Daqīqī and al-Kisāʾī of Marw.14 It was at this time that the ancient Iranian epic traditions were rekindled, and in 346/957 the governor of Ṭūs in Khurasan, Abū ManÂ�ṣūr Muḥammad ibn ʿAbd al-Razzāq, commissioned the translation of Pahlawī (that is, the Middle Persian of the Sasanians) texts of the national epic into New Persian. These were taken up by the great poet Abu ’l-Qāsim Firdawsī of Ṭūs (c. 329–30/ 940–1–c. 411/1020 or 416/1025), author of the monumental versified epic retelling of the history of the pre-Islamic Iranian kings and heroes ╇ Gutas, 1998, p.╯2. ╇ Idem, pp.╯2, 5, see also 134–5. 14 ╇ Cf. Bosworth, “Samanids,” EI2 VIII, 1025b.

8

12

9

13

the medieval islamic world from central asia to anatolia from mythico-legendary times until the arrival of Islam, which in the context of the period may be seen in terms of an Iranian revival characterised by an interest in national history. Known as the Shāh-nāma, it was completed in 400/1010 after about thirty years of writing. Together with Asadī Ṭūsī’s (d.╯c. 465/1072) slightly later heroic epic Garshāsp-nāma (“Epic of Garshāsp”), the oldest of the epics complementary to Firdawsī’s Shāh-nāma, it serves as one of the main sources for various descriptions of dragons comprising the following range of features that not only incorporate characteristics of various species into a single body but had the ability to cross boundaries within the natural environment it inhabits and was found in different locations: [The dragon] is sometimes described as a wolf, a tiger, shīr-e kappī, i.e., a sort of sphinx (combined lion and ape), or simply as a patyāra (maleficent creature), or a black cloud. ... it has one head and mouth, exhaling fire and smoke from its hellish mouth, and inhaling with enough force to suck in a horse and rider, or a crocodile from the water, or an eagle from the sky. … It is big as a mountain. Its head resembles a thicket of hair and its bristles stretch down to the ground like nooses. It has two horns the size of the branch of a tree, ten gaz or eighty cubits long. Its eyes are the size of wagon wheels or like two tanks of blood. They shine from afar as brightly as stars at night, as two glittering diamonds, as two blazing torches, or as two mirrors held beneath the sun. It has two tusks, each the length of the hero’s arm or of a stag’s horns. Humans and animals hang from its teeth. When it sticks its long, black tongue out of its mouth it hangs down onto the road like a black tree. Its skin has scales like a fish, each as big as a shield. It has eight feet, though most often it drags itself over the ground, and when it moves it makes the valleys and plains tremble, and a river of yellow poison as deep as a spear flows from its tail and nose. Its color is variously described, e.g., as dark yellow or gray, black, blue. ... Its lair, guarded day and night, is on a mountain (usually said to be near the sea, whence the azhdahā itself originated) or rock the same color as its body and is shunned by all living things, animals and plants. The sources variously locate it on the Kashaf-rūd near Ṭūs, on Mount Shekāwand in Kabul, India, “Māzandarān,” on ╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. ╇ The term Türkmen appears in Islamic sources from the tenth century onwards. See Bosworth, 1968, p.╯17. 15 16

17

Mount Saqīlā in the land of the Romans, Mount Zahāb in the Yaman, or in Ṭabarestān.15

At the same time there was a florescence of local arts and crafts. With this cultural background, some of the Turkic tribes, in particular the Oghuz and Qarluq (henceforth known as Türkmen/Turkoman),16 converted to Islam and became heirs to the local Muslim civilisations, Iranian and then Arabic. By the late tenth century Selchük (transcribed in Arabic as Saljūq), the son of Toqaq Temir Yaligh, commander of the Oghuz Yabghu, embraced Islam and became a ghāzī (Muslim warrior for the faith) against his still pagan fellow-tribesmen. He and his followers soon became embroiled in a power struggle with the weakening Samanids, a vacuum filled shortly after by the emerging Turkic dynasties, the Ghaznawids and Qarakhanids, thus setting the stage for the rise of the Great Saljuq Empire. Turkic tribes had long been in contact with Western Central Asian culture. The GhazÂ�naÂ�vid dynÂ�asty of Turkish origin was founded by Â�Sebüktigin, a general and governor for the Samanids and with its capital in Ghazna lasted for over two hundred years, from 367/977–8 to 583/1187. The Ghaznawids were favourably disposed towards Iranian culture, and their courts became outstanding cultural centres. They thus fulfilled what was expected of royal patrons, as the polymath of the age, Abu ’l-Rayḥān Muḥammad ibn Aḥmad al-Bīrūnī, who finished his days at Ghazna, put it: “to do this [sc. to honour learning and its representatives is] … the duty of those who rule over them, of kings and princes. For they alone can free the minds of scholars from the daily anxieties for the necessities of life, and stimulate their energies to earn more fame and favour, the yearning for which is the pith and marrow of human nature.”17 The Indian campaigns of Maḥmūd ibn Sebüktigin (r. 389/999–421/1030) brought a great influx of plundered temple treasures into the capital, Ghazna. The empire reached its zenith under Maḥmūd: no expense was spared in beautifying the capital and the sulṭān brought scholars, craftsmen and artisans from the lands he had conquered to Ghazna as well as to such provincial centres as Herat, Balkh and Lashkarī Bāzār, resulting in a flowering of the arts, and of architecture in particular.18 ╇ Kitāb fī Taḥqīq mā li-l-Hind (“Book of Inquiry into India”), tr. Sachau, 1887, p.╯152. 18 ╇ Bosworth, 1963, pp.╯139–41. Cf. Bombaci and Scerrato, 1959. 17

18

chapter one

Although Firdawsī makes the Oxus the traditional boundary between Iran and Tūrān (the Central Asian region beyond the Jayḥūn/Āmū Daryā), the land of the nomadic world of Western Central Asia, and states that there was a natural dislike between the two groups, which were like “two elements, fire and water, which rage against each other in the depths of the heart,”19 there was never, as Bosworth has pointed out, a cut and dried distinction between the two racial groups.20 They had a long history of interaction and the Turkish people were well-known to the Iranians, who had often been invaded by steppe peoples of diverse ethnic origins. The antithesis between Iran and Tūrān, emblematising a dualistic conception of the world and of history, thus appears by Firdawsī’s time to have been more a literary and archaising preconception of the Iranian national consciousness than a reflection of the actual state of affairs.21 The Qarakhanid Muslim philologist, Maḥmūd al-Kāshgharī, remarks in his lexicographic encyclopaedia Dīwān lughāt al-turk, written in 463/1071, that all of Transoxania, which was closely linked with the Eurasian steppe, was once inhabited by Turkic peoples, “but when the Iranians (al-Furs) became numerous, it became just like Persian territory (bilād al-ʿAjam).”22 Thus the two worlds had to a large part become intermingled culturally as well as ethnically. The westward migration of ever-growing numbers of Turkic-speaking tribes, the dominant force being the Saljuqs, into Western Asia increased from the late tenth century onwards. This accelerated following their decisive victory under Ṭoghrıl Beg over the Ghaznawids at Dandāndaqān (located between Marw and Sarakhs) in 431/1040, after which all of Iran lay open before them. In

447/1055 Ṭoghrıl Beg entered Baghdad and in 449/1058, when he entered for the second time, the ʿAbbasid caliph al-Qāʾim legitimised his rule by the bestowing of honorific titles. The victory of the second Saljuq sulṭān Alp Arslan against the Byzantine ruler Romanus IV Diogenes at Manzikert/Malāzgird, north of Lake Van, in Dhu ’l-Qaʿda 463/August 1071, on a day following a moonless night,23 effectively destroyed the ability of Byzantium (Rūm) to defend its eastern boundaries. This led to the gradual settlement of hellenised Asia Minor, or Anadolu/Anatolia as it was later to be known under the Turks, by successive waves of mostly nomadic Turkic tribes, and to the establishment of the sultanate of the Saljuqs of Rūm (Saljūqiyān-i Rūm)24 in the central and eastern territories. By the end of the century the entire territory from the Armenian and Georgian marches to the Aegean sea was in Saljuq hands. In their new homeland, the Turks perpetuated the heritage of Western Central Asian art and culture with a “markedly Khurasanian flavour.”25 This was facilitated by the fact that the migratory movement of Turkic peoples swept along migrant craftsmen from the Central Asian, in particular East Iranian, world, a process much intensified by the invasion of Khurasan by the Mongol army under Genghis Khān in the 1220s.26 The signature of master craftsmen on Anatolian tilework suggests that innovation on Anatolian Saljuq architecture owed much to craftsmen from Khurasan or Ghurid Herat27 and points to Eastern Iran as one of the earlier and most important centres of artistic innovation. Katharina Otto-Dorn has proposed that some of the tile revetments at the now destroyed palace-citadel at Â�Kubadabad (623/1226–634/1237), southwest of Konya, built

19 ╇ Kowalski, 1939–49, pp.╯87–9. The moon was a preeminent Turkish emblem and the sun a Iranian one; idem, pp.╯98–9. 20 ╇ Bosworth, 1963, p.╯205. 21 ╇ Idem, p.╯206. 22 ╇ Tr. Atalay, B., vol.╯3, Ankara, 1939–1943, pp.╯149–50, as cited in Bosworth, 1963, p.╯206. 23 ╇ Hillenbrand, C., “Malāzgird,” EI2 VI, 242b. 24 ╇ The use of the ethnic/dynastic term Rūm by the Anatolian Saljuqs whose principality was based on the region of Konya and southern Cappadocia reflected their conception as heirs to the Byzantines in south-central Anatolia, territories which continued to be strongly Greek in ethnos. Bosworth, “Rūm. Relations between the Islamic powers and the Byzantines,” EI2 VIII, 601a. 25 ╇ Melikian-Chirvani, 1974, pp.╯112, 114. Note however Oleg Grabar’s caveats (2006, pp.╯314–5) with regard to anÂ� Â�argumenÂ�tation for a Khurasan “style” or “mode” as being hypoÂ�-thetical on visual and historical grounds and merely

indicative of “an active industry or artisanship of metalwork in Khurasan.” 26 ╇ On the large-scale movement of metalworkers to western Iran, Anatolia, the Jazira, Syria and Egypt, see Ward, 1993, pp.╯79, 87. 27 ╇ The signature of the bannāʾ Muḥammad ibn Muḥammad ibn ʿUthmān al-Ṭūsī, who probably came from KhuÂ�rasan, found on the tile-mosaic of the Sırçalı madrasa in Konya (640/1242–3) is discussed by Meinecke, 1976, vol.╯1, pp.╯35–45, and idem, vol.╯2, no. 71. For further examples of craftsmen who, judging from their geographical epithet (nisba; generally pointing to someone’s tribal, geographical or religious affiliation) may have come from the eastern Islamic lands to the west, see, for instance, those listed in idem, vol.╯1, pp.╯187–9, addendum II (i.e., Aḥmad ibn Abī Bakr al-Marandī, c. 612/1215, or miʿmār Badr al-Dīn Tabrīzī, post-672/1273); Pickett, 1997, pp.╯37–41. See however the cautionary remarks on this subject, idem, n. 349.

the medieval islamic world from central asia to anatolia

19

at the apogee of Saljuq power by sulṭān ʿAlāʾ al-Dīn Kay Qubādh I (r. 616/1219–634/1237), may have been made by Iranian craftsmen.28 Moreover, the extent to which the Rūm Saljuqs (just like their Iranian cousins, the Great Saljuqs) embraced the Iranian tradition of kingship was particularly marked, as also evidenced by their tendency to choose pre-Islamic Iranian royal names.29 Furthermore, Persian was the official language of the court and administration in the sultanate of Rūm, which welcomed streams of poets and mystics (ṣūfis, “those who wear wool”) from the Western Central Asian world, the most celebrated of whom was certainly Jalāl al-Dīn Rūmī (Rabīʿ I 604/30 September 1207–5 Jumāda II 672/1273), who had left Balkh (ancient Bactra) at the age of twelve with his family.30 On the other hand, the sedentary population over which the Saljuqs ruled was extremely heterogeneous and included large numbers of Armenian, Georgian and Greek Christians. The transmission of art and culture from the greater Khurasan region was also apparent in the Transcaucasian region, particularly in thriving twelfth-century Georgia31 and Armenia, regions, valuable repositories of ancient oral and iconographical traditions that had been in the orbit of successive stages of Iranian culture since antiq-

uity.32 Marriage alliances between eleventh- and twelfth-century Muslim ruling families and Georgian33 and Armenian34 royal families resulted in a nexus of Christian-Muslim relations, while providing intermediaries between Christians and Muslims. More Islamic culture penetrated into Georgia during the reign of queen Tʿamar of Georgia (1184–1211/2), whose territory stretched from Azerbaijan to the borders of Cherkessia, and from Erzurum to Ganja, forming a pan-Caucasian Empire. The evidence of this can be seen in GeorÂ�gian literature35 and manuscript illustration36 based on Iranian models. This influence is also discernible in metalwork from Daghistan, located east of Georgia and west of Iran.37 In spite of internecine strife the successive and partly overlapping major dynasties of the Ghaznawids, the Ghurids, the Qarakhanids, the Great Saljuqs, the Zangids, the Ayyubids and the Saljuqs of Rūm succeeded in spite of the ethnic diversity of their subjects in creating a comparatively unified culture from India to Egypt. In particular the artistic traditions of the Western Central Asian world, the Caucasus, Transcaucasia, northern Syria, eastern Anatolia and northern Mesopotamia certainly had to varying degrees a symbiotic development. In the twelfth and thirteenth centuries, the region underwent

28 ╇ Otto-Dorn and Önder, 1969, pp.╯468–9. Cf. MeliÂ�kianChirvani, 1974, p.╯115 and n. 25. 29 ╇ At least from trom the time the Saljuqs entered Iran, the Turks came under the influence of Iranian culture. The impact of this influence is reflected, for instance, in al-Juwaynī’s account of the last Great Saljuq sulṭān of Iran and Iraq, Ṭoghrıl III ibn Arslan (r. 571/1176–590/1194), reciting verses from the Shāh-nāma (ed. Vullers, p.╯188, ll. 1060–2) while wielding his heavy mace in battle:

(Allen, 1932, pp.╯91–2). The ruler of Erzurum ordered his son, Mughīth al-Dīn Toghrıl, to convert to Christianity to marry Rusudan (1223–1247), the daughter of queen Tʿamar of Georgia, and heir to the throne after the sudden death of her brother, whose daughter was in turn married to the Saljuq sulṭān of Rūm, Giyāth al-Dīn Kay Khusraw II (634/1237–644/1246); after his death she married the Parwāna Muḥammad ibn Sulaymān Muʿīn al-Dīn (executed in 676/1277), one of the most powerful Anatolian magnates of the thirteenth century. Cf. Minorsky, 1953, p.╯135; Melikian-Chirvani, 1974, p.╯113, n. 17; Rogers, 1976, p.╯316. 34 ╇ The eleventh-century ruler of Dwīn, Abu ’l-Aswār Shawūr, of the minor Armenian dynasty of the Shaddādids, was married to a sister of the Armenian king of Tashir, David Anholin. Akhsatān ibn Manuchihr ibn Afrīdūn, ruler of Ani, and son of this princess, also married an Armenian princess of the Bagratid house (Minorsky, 1953, p.╯81). See also Melikian-Chirvani, 1974, p.╯113, n. 17. 35 ╇ For instance the close similarities between the eleventhcentury narrative poem, Wīs u Ramīn, and Shota RustaÂ�veli’s twelfth-century epic Vepkhis-tkaosani (“The Knight in the Panther’s Skin”); see idem, 1974, p.╯112 and n. 14. 36 ╇ See, for instance, a Georgian astrological treatise of 1188, illustrated under Islamic influence which MelikianChirvani (1974, pp.╯112–3 and n. 15) interpretes as Persian, detecting Khurasanian influence in the Kufic inscriptions. Amiranašvili, 1966, pls. 56–66 and pp.╯28–30, where the writer also connects the Iranian stylistic influence with the Persianising aspect in Georgian literature. 37 ╇ Melikian-Chirvani, 1974, p.╯113 and n. 16.

“When the dust arose from the countless army, the cheeks of our worthies turned pale. As for me I raised the mace that kills with a single blow and felled that host upon the spot. I uttered a yell from my saddle saying, ‘The earth has become a millstone upon them.’” Cited after Taʾrīkh-i jahān-gushāy, tr. Boyle, 1912–37, vol.╯2, p.╯302. 30 ╇ Cf. Melikian-Chirvani, 1974, pp.╯114–5. 31 ╇ The close relationship between cultures in the twelfth century appears to have been a result of the Christian-Islamic symbiosis at Tbilisi among the ruling families (Minorsky, 1953, p.╯157; attested by king Dimitri’s attendance at Friday prayers in 548/1153 (as recorded by al-Fāriqī), and p.╯135 (Christian-Islamic marriages)). Cf. Melikian-Chirvani, 1974, p.╯112 and n. 15. 32 ╇ Baltrušaitis, 1929, pp.╯43–5; Melikian-Chirvani, 1974, pp.╯113–5 and ns. 17, 18, 20. 33 ╇ After his incursions into Georgia against Bagrat IV, the Saljuq sulṭān Alp Arslan (455/1063–465/1073) strengthened his influence there by marrying one of the king’s nieces

chapter one

20

sweeping socio-political, economic and artistic changes that made “northern Mesopotamia” – a geographical entity known in medieval Islam as the Jazīra (“island”), the northern part of the territory located between the Tigris and Euphrates rivers (today divided between eastern Syria, northern Iraq and southeastern Turkey) – “one of the liveliest regions.”38 Political stability, though not hegemony, brought about relative prosperity. Following the decline of the Great Saljuq dynasty from around 512/1118, the area was divided into a number of Turkish and Kurdish principalities.39 The Saljuq “successor states” included the Artuqids of Amīda/Diyārbakr (end of the fifth/ eleventh to the beginning of the ninth/fifteenth century), the Zangids of northern Syria (521– 2/1127–8 until 579/1183 in Aleppo and until 631/1233 in Mosul), the Ayyubids of Syria and Egypt (564/1169–658/1260) and the Saljuqs of Rūm (Anatolia, c. 483/1081–c. 707/1307). The

╇ Ettinghausen and Grabar, 1987, pp.╯297–9. ╇ For an outline of the political, religious and cultural climate of the region, see Hillenbrand, C., 1985. 38 39

rapid increase of small principalities founded by atābegs/atābaks (“father-beys”) or various members of the Saljuq dynasty resulted in the establishment of numerous courts, all competing with each other for cultural prestige, which may well have provided the impetus for the proliferation of innovative images. However, this was severely disrupted under the impact of invasions from the east, first by the Turkic Khwārazm shāhs in the 1220s and shortly after by the Mongols, which brought about a cataclysm with great social upheavals, destruction and discontinuity. After the battle of Köse Dagh in the region of Sivas (Sebasteia) in 641/1243, the Mongols occupied Anatolia and Saljuq autonomy was lost forever. For a time the Saljuq sultanate continued as a Mongol province, although some Türkmen emirs maintained small principalities of their own in distant mountainous districts, but finally the Saljuq dynasty came to its end.

dragons on monumental settings in regions west of iran

21

chapter two

Dragons on monumental settings in regions west of Iran

Carved friezes or high-relief sculpture on stonework of the medieval Islamic period, very often figural in nature and including a repertory of animals and fabulous beasts, were portrayed in particular on architectural monuments. These could include the gates, archivolts and doorways of secular monuments such as city walls, palaces and caravanserais, as well as religious monuments like mosques, madrasas and funerary structures. Among the mythical creatures depicted on medieval Islamic monuments, the dragon occupies a significant place.1 However, while the iconography is entirely absent from Western Central Asian monumental art until the fifteenth century,2 it is characteristic of the area to the west of Iran. As noted earlier, this is due to the fact that no figural sculpture is associated with the brick architecture of the Iranian world from about 1000 to 1200, whereas the dragon motif is conspicuous on portable items from the greater Khurasan region, where the motif has a very ancient history. Its appearance on the architecture of eleventh- and twelfth-century northern Mesopotamia, the Jazīra, Syria and Anatolia,3 may on the one hand be due to depictions in the so-called “minor arts” whose very nature is their potential for portability, allowing for the long-distance diffusion of motifs.4 This may be compounded by another reason which may be sought locally. The principal building material employed for the architecture of these regions is cut-stone with brick playing a minor role. The preference for this material is due

to its availability as well as the local builders’ mastery of stone masonry, a skill which was naturally made use of by their new rulers, the Saljuqs. The subjugated local population was largely Christian, who only gradually converted to the faith of their rulers, and not only employed their traditions of stone-carving, woodwork, stucco and tile-mosaic but also their decorative repertoire.5 It is evident that dragon motifs were not limited to Islamic monuments or portable objects, but were used equally by Christian artists and the artists of other faiths. The iconography of the dragon clearly enjoyed cross-cultural popularity in the medieval era. In fact representations of the motif on stone-carved architectural reliefs in the predomiÂ�nantly Christian Transcaucasian realm (that is most of present-day Armenia, Georgia and Azerbaijian) precede its first known depictions in neighbouring eleventh- and twelfth-century northern Mesopotamia, Syria and Anatolia. In particular in the Armenian and Georgian regions this type of imagery can be found from about the ninth century onwards, especially the dragon combat motif (the earliest depictions of which may be datable to the seventh century), which will be examined in chapter 7. Often depicted in mirror image, dragon themes appear above or around entrances and portals which represent the boundary between the exterior and interior. The placement reflects the senÂ�sitivity of the threshold (Lat. limen) as both a metaphor for the monument it protects and the

1 ╇ For dragon imagery on Islamic architecture, see in particular the monograph by Öney, 1969a, as well as the research of Otto-Dorn, 1978–9, esp.╯pp.╯25–36. 2 ╇ Two large yellow dragons set against a blue background in mosaic faience were shown in the tympanum of the portal arch of Abu ’l-Qāsim Babur’s mosque dating from 848/1444–5 situated in the shrine complex of Jamāl al-Ḥaqq wa ’l-Dīn at Anau near Ashgabat in Â�Turkmenistan, which was destroyed when the area was struck by an earthquake in 1948. Some of the dragon mosaic has been rÂ�ecovered and is now housed at the Fine Arts Museum of Ashgabat. The portal was photographed by the German art historian Ernst Cohn-Wiener in the 1920s, whose collection of Â�photographs taken in west Turkestan is kept at the British Museum and published online in the digital library of

archnet.org; ArchNet Image ID ICW0120. For a discussion of the Anau dragon motif, see Pugachenkova, 1956, pp.╯125–9. Dragons also appear in the spandrels of a fifteenth-century mosque at the shrine-complex of Turbat-i Sheikh Jām halfway between Mashhad and Herat in Khurasan; see Daneshvari, 1993, pl.╯I, fig.╯1. 3 ╇ It is also noteworthy that the dragon motif was not introduced into Egypt and the Maghrib until Mamlūk times, when this transmission probably took place through Mosuli craftsmen. Ibrāhīm, 1976, pp.╯12, 15–6. 4 ╇ Cf. Hoffman, 2001, pp.╯17–22. 5 ╇ I would like to thank Professor Robert Hillenbrand for drawing my attention to this point. Cf. Meinecke, 1976, vol.╯1, pp.╯5–6.

chapter two

22

liminal or transitory realm within which the passage from the perilous exterior to the secure interior might be negotiated.6 The doorway, gates and further openings, or other vulnerable zones, were considered to be a favourite abode of dangerous spirits, hence the particular precautions, rites of spatial passage and apotropaic sacrifices enacted in these areas on special occasions.7 One of the primary uses of the iconography of the dragon was therefore to identify the threshold and thereby operate as liminal marker. The choice of the dragon in this capacity reflects a consistently followed principle according to which anyone entering a building is confronted by the dragon figure. While the dragon’s intermediate position and hybrid character itself is marked out as incarnate liminality, characterised by ambiguity, its fierce, menacing aspect is here found in a “helpful” context in that it is directed towards the outside world, warding off all hostile attack, and turning into a symbolism of defence of those that are “inside.” It is employed in its apotropaic capacity to ward off evil and afford protection by taking a defensive role against baleful creatures and dangerous influences such as natural catastrophes. Its semantic horizon is thereby extended to include the function of guardian of the threshold akin to a “tutelary spirit” imbued with prophylactic and talismanic power.8 The apotropaic and protective function of the dragon may further be associated with belief in the Evil Eye, the blighting glance of envy, which belongs to one of the most ancient concepts of humanity, prevalent in medieval Islamic culture and referred to in the Qurʾān (sūra 68, 51–3).9 The Evil Eye was feared and apotropaic symbolism, sometimes in the form of fixed representations of the Eye attached to architectural structures, was used to ward off its malevolent gaze, to disseminate evil and to warrant protection. It may further be noted that the Evil Eye was often considered

to be most harmful to those in the liminal stages of life, such as newborn children, the newly married or pregnant women. Moreover, in keeping with ideas relating to the protection of deceased individuals entering the liminal and transitional states, burials are sensitive sites. At funerary sites and mausolea, the dragons serve as markers of sacred spaces. The depiction of dragons may be associated with the fact that some dangerous animals, in particular vipers, are held to have a poisonous glance.10 This is illustrated by the story of the marine dragons reported by the renowned tenth-century Arab geographer and traveller, Abu ’l-Ḥasan ʿAlī ibn al-Ḥusayn ibn ʿAlī al-Masʿūdī (d.╯345/956). Marine creatures are said to have so frightened the workmen who built Alexandria as to prevent Iskandar (Alexander the Great) from constructing the city. He is therefore said to have dived into the sea in a glass box inserted into a wooden box from where he drew pictures of the sea monsters. These images were used to construct metal effigies which were then set up opposite the place where building was being carried out. When the dragons emerged from the sea and saw the images they fled, enabling Iskandar to complete the building of Alexandria.11 This ruse was hence linked to the idea that if confronted by representations of themselves, dragons would be repelled by their own noxious power. The account has a precedent in the biblical story of the plague of poisonous snakes which so devastated the people of Israel that Moses intervened on their behalf, setting a bronze image of a serpent upon a pole (Numbers 21:6, 7 and 9):

╇ Cf. Flood, 2006, p.╯149. ╇ Kitzinger, 1970, p.╯640 and n. 7; Engemann, 1975, pp.╯44–8; Henninger, 2004, pp.╯14 and n. 66, p.╯22 and ns. 109–11, pp.╯31–2; Zbinden, 1953, pp.╯36, 44. 8 ╇ Kühnel, 1950, pp.╯4–18. Cf. also Otto-Dorn, 1959, p.╯75 and n. 38, and eadem, 1978–9, p.╯130; Öney, 1969a, pp.╯214– 25; Roux, 1972, p.╯393. 9 ╇ Schimmel, 1994, p.╯91. Cf. Marçais, “ʿAyn,” EI2 I, 786a. 10 ╇ Al-Damīrī (Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯55, 633) records that the face of a certain kind of serpent, called al-Aṣalah, whose looks kill by the mere sight, “is like that of a human being, that it is of an immensely large size, and that it remains in the same condition even if a thousand years pass over it.” For other stories

about poisonous snakes whose looks alone are enough to kill, see Ullmann, 1992, p.╯111; Ruska, “Almās,” EI¹ I, p.╯313; Ruska and Plessner, “Almās,” EI² I, 419a. The poisonous serpent or basilisk whose glance could kill a man was already known in antiquity (cf. Pliny, Naturalis Historia VII 1.2 and 49.70; VIII 33). 11 ╇ Al-Masʿūdī, Kitāb murūj al-dhahab, tr. and ed. Barbier de Meynard and de Courteille, 1917, vol.╯2, pp.╯425–7. See also Pseudo-Callisthenes, Historia Alexandri Magni, ed. Kroll, 1926, p.╯32. The story is also recorded by Ibn Khaldūn who discounts the practical aspect of it and observes “the story of the many heads they have is intended to indicate ugliness and frightfulness. It is not meant to be taken literally.” Â�MuÂ�qadÂ�dima, tr. Rosenthal, 1958, vol.╯1, p.╯73.

6 7

The Lord sent seraph serpents against the people. They bit the people and many of the Israelites died. The people came to Moses and said, “we sinned, by speaking against the Lord and against you. Intercede with the Lord to take away the

dragons on monumental settings in regions west of iran serpents from us!” And Moses interceded for the people. Then the Lord said to Moses, “Make a seraph figure and mount it on a standard. And if anyone who is bitten looks at it, he shall recover.” Moses made a copper serpent and mounted it on a standard; and when anyone was bitten by a serpent, he would look at the copper serpent and recover.12

23

survived fled, and it remained empty for a while. Then, to drive these calamities away, one sultan [Byzantine emperor] ... fabricated a talisman, possibly the bronze one now presently shaped like three serpents.16

Serpents and snakes appeared in [Constantinople] and decimated men and cattle alike. Those who

ʿIzz al-Dīn Muḥammad ibn Shaddād (613/1217– 684/1285), the chronicler of the cities of Syria and the Jazīra in the thirteenth century, mentions a dragon tower (burj al-thaʿābīn) in Aleppo that was supposed not only to prevent the detrimental effects of snake-bites but also to protect the city’s inhabitants.17 The late Mamluk topographer Abu ’l-Faḍl Muḥammad Ibn al-Shiḥna (active 1400– 1450) more clearly specifies that the tower referred to as burj al-thaʿābīn serves as a talisman against serpents conferring immunity from snake bite in Aleppo.18 Such apotropaic renditions were not unique to Islamic architecture but were also found in the Christian environment. This is evidenced by the twelfth-century ascetic Abu ’l-Ḥāsan ʿAlī ibn Abī Bakr al-Harawī (d.╯611/1215), who in his description of sacred sites of Anatolia describes a talismanic design of a double-headed serpent in the church of Mart [sic; perhaps a corruption of Mār] Dāris in Mayyāfāriqīn, northeast of Diyārbakr.19 The conception of the existence of a serpent with heads at both ends of its body appears to have been widespread in the medieval Islamic world. It is discussed in the best-known book on animals of the Arab world, the Kitāb al-Ḥayawān (“Book of Animals”), written by the ninth-century litÂ�Â�térateur Abū ʿUthmān ʿAmr ibn Baḥr, known by his nickname al-Jāḥiẓ (“the google-eyed”; 159/775–6–255/868–9). He quotes the Arabic

12 ╇ Tr. after Wilson, 2001, p.╯75. The term seraph (Hebr. saraf) appears to be the general name for poisonous snakes whose poison, metaphorically, soref (“burns”) the body. See also Astour, 1965, pp.╯232–3; Hendel, “Neshutan,” DDD, pp.╯615–6. 13 ╇ The Greek cities had dedicated the serpent column to the Delphic temple to commemorate victory over the Persians at Plateia in 479 bc. During the rule of Constantine I (r. 306–337) the column, which originally was eight meter high, was brought to Constantinople to decorate the spina (central divider) of the Hippodrome. The intertwined heads of the serpents, into which the names of the victorious cities had been engraved and which once carried a golden tripod, can be seen in sixteenth-century Ottoman miniatures and remained in place until about 1700. Only the shaft of the monument remains in situ; one of the serpent heads is preserved in the Istanbul Archaeological Museum. Cf. Eder, W., “Schlangensäule,” DNP 11, 2001, p.╯184; “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, p.╯xli and n. 1. 14 ╇ The word talisman (ṭilsam) is used throughout the text to denote any type of object made to protect the owner, that is to say to avert the power of the Evil Eye, and to promote

well-being; it also relates to architectural sculptures that served an apotropaic function. 15 ╇ The text is included in Ibn Rusteh’s early tenth-century geographical work, Kitāb al-aʿlāq al-nafīsa. See El-Cheikh, 2004, p.╯148. 16 ╇ Akhbār al-duwal wa athār al-uwal fī al-tārīkh, 3 vols., eds. Saʿd, F., and Ḥaṭiṭ, A., Beirut, 1992, vol.╯3, p.╯192, cited after El-Cheikh, 2004, p.╯221. 17 ╇ Al-Aʿlāq al-khaṭīra, the part on Aleppo, tr. and ed. Sourdel, D., Damaskus, 1953, p.╯123. Cf. Herzfeld, 1955, pp.╯24–5; Meri, 2002, p.╯206, n. 360. 18 ╇ Durr, cited after Herzfeld, 1955, p.╯25. To this may be added the tradition that Balīnūs (the Pseudo-Apollonius of Tyana), known as the great master of talismans (ṣāḥib al-ṭilasmāt), is reported to have left in many towns charms for protection against such adversities as serpents, scorpions or storms. Cf. the Kitāb Ṭalāsīm Balīnās al-akbar, Paris, Bibliothèque Nationale, Ms. 2250, fols. 84–134; see Vajda, 1953, p.€696; Sezgin, 1971, pp.╯77–90. 19 ╇ Kitāb al-Ishārāt ilā maʿrifat al-ziyārāt, tr. and ed. Sourdel-Thomime, 1957, p.╯65. Cf. Meri, 2002, p.╯206, n. 360, and idem, 2006, p.╯162.

The expedient relates to the magical principle of effecting a cure for snake venom by viewing the image of the serpent, thus following the principle of homeopathic (or imitative) magic, similia similibus, according to which things are believed to act upon each other, even at a distance, if they are alike in some relevant manner. The purpose of the bronze image was therefore therapeutic: anyone bitten by a serpent could be healed by looking at it. Since the peril was identified with the demonic power within the serpent, the bronze image upon a pole constituted a counter-equivalent power which served as effective prophylaxis. The famous copper alloy serpent column, the triple-headed serpent tripod of Delphi in the Hippodrome of the city of Constantinople, is an example of apotropaic sculpture intended to afford protection from poisonous creatures, including serpents.13 The early tenth-century Arab captive, Hārūn ibn Yaḥyā, notes four copper serpents in Constantinople that served as talismans14 to render noxious creatures inoffensive.15 The sixteenth-century Ottoman historian Aḥmad ibn Yūsuf al-Qaramānī (d.╯1019/1610) similarly reports:

chapter two

24

translation or paraphrase of the Aristotelian Historia Animalium and, while duly noting his personal scepticism as an educated man, records the following anecdote:

Depictions of double-headed serpents, or amphisbaena, are a recurrent feature on architecture, most often positioned above gates, as will be shown in the following. Many city gates also carried prominent epigraphic panels commemorating a victory, invoking good fortune or deflecting evil influences.21 Qurʾānic verses in particular are considered the most powerful of all “talismans.”22 Often these protective inscriptions were complemented or replaced by images of the dragon which served a prophylactic and talismanic (in the sense of apotropaic) function. It was thus their purpose to protect, avert the power of evil, and to promote well-being. In 579/1183–4 Diyārbakr (Diyarbakır), the historic city of Āmid, was conquered by the Türkmen Artuqids who in the same year made significant additions to the striking black basalt city walls to commemorate the victory.23 A large frieze graces the West Gate (the former Rūm Gate and present Urfa Gate) and eulogises the patron Abū ʿAbdallāh Muḥammad ibn Qara Arslan ibn Dāwūd ibn Suqmān (561/1166–581/1185) with extensive titles and Qurʾānic verses (sūras 48, 18; and 61, 13) which were deemed to have specific powers relating to the attaining of victory: “Help comes from God and victory is near.”24 The protective power of the inscription is augmented by carved reliefs: a pair of symmetrical confronted dragons with forelegs is surmounted by the composition of a bird of prey, probably an eagle, wings out-

stretched, and standing atop the horns of a bull’s head, the nose of which is threaded through with a large copper alloy ring carved onto the keystone of the arched gate. The dragons are rendered in formal pose with one foreleg raised (sometimes considered to be “heraldic”), their elongated twisting serpentine tail arranged first in a socalled pretzel-like (or heart-shaped) knot and then into two loops; their supernatural properties are represented by the presence of wings and by the long sinuous tongues protruding from the open mouths (fig.╯1).25 These details which represent a “heraldic” stylisation of the creature establish not only the symbolism of the Saljuq-period dragon but indicate at the same time a ceremonial and ritual of subjugation and domestication. However in contrast to some of the later Saljuqperiod dragon sculpture, discussed below, that was intended to be highly visible from far away, the Diyārbakr dragon reliefs, which represent the earliest surviving such examples of the medieval Islamic period, are rather discreetly portrayed: they are too small to be detected from a distance and thus remain curiously invisible. It may be hypothesised that the inconspicuous nature of the depictions reflects a formative level of the conceptualisation; hence, the craftsmen were still rather unfamiliar with the dragon iconography and thus perhaps did not wish to display it too ostentatiously. Yet the fluidly rendered portrayal, which is in line with later distinctive representations when the motif had become increasingly common, betrays this assumption. Indeed it may be noted that, with exceptions such as the Konya dragon sculptures discussed below, the iconic image of the dragon was generally not conspicuously displayed throughout the medieval period. This may be due to the fact that size and associated visibility were not considered to be essential criteria for the inherently propitious and apotropaic powers of the sculptures to take effect.26 Related but much larger, plastically carved decoration was characteristic of the early thirteenth-century city walls, gates or citadel of the Saljuq capital Konya (ancient Iconium) in central Anatolia, built by sulṭān ʿAlāʾ al-Dīn Kay Qubādh I (r. 616/1219–634/1237) to protect the city from

╇ Kopf, 1953, p.╯400. ╇ Van Berchem and Strzygowski, 1910, p.╯73. 22 ╇ Porter, 2006, p.╯794. 23 ╇ Gabriel, 1931, vol.╯1, pp.╯166–8, fig.╯136; van Berchem and Strzygowski, 1910, pp.╯82–4; Kühnel, 1950, p.╯8. 24 ╇ Cf. Hauptmann von Gladiss, ed., 2006, pp.╯31–2.

25 ╇ Van Berchem, 1910, p.╯82, fig.╯30 B (drawing), pl.╯17, Öney, 1969a, figs. 30 a (photograph of the entire composition) and b (line drawing); Gierlichs, 1996, pl.╯48.5 (photograph of the dragon to the left). 26 ╇ I am most grateful to Professor Robert Hillenbrand to have raised this subject for discussion.

The Master of Logic (Aristotle) states that a serpent with two heads was observed. â•… I asked a bedouin about that, and he asserted that it be [sic] true. Whereupon I said to him: In the direction of which one of the two heads does it creep, and with which one does it eat and bite? He replied: As to its creeping it does not creep (at all), but proceeds towards its aim by rolling itself as do boys on the sand. As to its eating it takes its evening meal with one mouth and its breakfast with the other. As to its biting it does so with both its heads.20

20 21

dragons on monumental settings in regions west of iran

25

the Mongol advance. A group of double-headed dragon reliefs from these now destroyed monuments is preserved in the İnce Minare Müzesi, Konya.27 Iconographic and stylistic variations certainly exist between the dragon sculptures of this period, but the basic conception remains the same for all surviving examples. Akin to the Artuqid Diyārbakr dragon reliefs the Konya sculptures are thus similarly portrayed with their inner forelegs raised, the “Saljuq-style” heads with cusped pricked up ears projecting at the top, the wide-open jaws terminating in upward curled tips revealing sharp teeth and the tongues thrust out to reveal bifid tips. The upswept, curved wings with tightly curled tips project from the haunches; their serpentine tails are knotted at mid-section. In these fragments however the tails curve backwards and then taper to a small dragon head that snaps with its snout at the dragon-tail (fig.╯60). Although preserved as individual panels, the dragons would probably have been represented as antithetical pairs.28 In contrast to the shallowly rendered Diyārbakr dragon reliefs, these dragons are plastically sculpted, the monumentality of their size adding to their visually dramatic appearance. The large stone panels may have been placed on or near the city gates and would have made an impressive sight on the city wall, visible from afar. The chronicler of the history of the Rūm Â�Saljuqids, al-Ḥusayn ibn Muḥammad ibn ʿAlī al-Jaʿfarī al-Rughadī, known as Ibn al-Bībī (d.╯after 684/ December 1285), who was head of the chancellery of the Secretariat of State,29 describes the elaborate royal ceremonies the Saljuqs were known to have staged at the city gates where they received visitors before conducting them into the city.30 The placement of several dragon sculptures on the city walls and gates would have amplified the

impression of power and good fortune which the sculptures were probably meant to convey. Closely related dragon figures are known from the interior decoration of Saljuq palaces. Two small fragments have been discovered at the now destroyed palace-citadel at Kubadabad on the west bank of Lake Beyşehir, near Akşehir, of the sulṭān ʿAlāʾ al-Dīn Kay Qubādh I (r. 616/1219–634/1237), whose reign marked the height of Rūm Saljuq power. One of these is a fragmentary stone relief-carved with a dragon’s head,31 the other is a star tile showing paired dragons whose necks cross over and whose wideopen confronted jaws reveal rows of pointed teeth and bifid tongues.32 A third piece was discovered while excavating the ruins of the pavilion (köshk) on Konya’s citadel, which was probably built during the reign of the Rūm Saljuq sulṭān Rukn al-Dīn Qılıch Arslan IV (r. 646/1248–647/1249, 655/1257–664/1266).33 It represents a plaster Â�fragment moulded in low relief with a pair of related addorsed dragons separated by a braided band; their gaping snouts are turned backwards, the feathery wings raised, and the tails form a pretÂ�zel-like knot and a loop before tapering to a point.34 One of the most outstanding examples of the dragon iconography on city gates certainly must have been the monumental sculptures on the archivolt of the so-called Talisman Gate (Bāb al- Ṭilasm) in Baghdad (figs. 2, 139a and b) which was destroyed in 1917 during the First World War.35 As indicated by the name of the gate, built in 618/1221–2 under the great caliph Abu l-ʿAbbās Aḥmad al-Nāṣir li-Dīn Allāh (577/1181– 620/1223), this type of imagery had talismanic (as also implied by the gate’s name), or at least apotropaic connotations. It showed a seated figure

27 ╇ Konya, İnce Minare Müzesi, inv. no. 889 (Sarre, 1909, p.╯14, fig.╯16; Önder, 1961, p.╯70, fig.╯2; Öney, 1969a, p.╯195, fig.╯3; Otto-Dorn, 1959, pl.╯VIII, fig.╯36; Gierlichs, 1996, p.╯197, cat. no. 43, and pl.╯38.1), inv. no. 890 (Sarre, 1909, p.╯13, fig.╯15; Önder, 1961, p.╯70, fig.╯1; Diyarbekirli, 1968, p.╯370, fig.╯5; Öney, 1969a, p.╯194, fig.╯1; Gierlichs, 1996, p.╯198, cat. no. 44, and pl.╯38. 2), and inv. no. 1394 (Öney, 1969a, p.╯195, fig.╯2; and eadem, 1978, p.╯46, fig.╯32; Gierlichs, 1996, p.╯200, cat. no. 46, and pl.╯39.1; Grube and Johns, 2005, p.╯230, fig.╯77.2). 28 ╇ Otto-Dorn, 1978–9, p.╯125. Cf. Gierlichs, 1996, p.╯197. 29 ╇ Cf. Duda, “Ibn Bībī,” EI2 III, 737b. 30 ╇ Such as the welcome accorded to shaykh ʿUmar Suhrawardī at Konya; al-Awāmir al-ʿAlāʾiyya fī ’l-ūmur al-ʿAlāʾiyya (“History of the Rūm Saljuqs”), completed in 680/181, tr. Duda, 1959, pp.╯102–3.

╇ Öney, 1969a, p.╯196, fig.╯4; Otto-Dorn, 1978–9, p.╯125. ╇ Otto-Dorn, 1978–9, p.╯125; Arik and Arik, 2008, p.╯313, fig.╯300 (colour reproduction). 33 ╇ Rogers, “Saldjūḳids,” EI2 VIII, 936a. 34 ╇ Plaster fragment. Konya, Alaeddin Palace. Thirteenth century. Height 33 cm, width 19.5 cm. Konya Museum, inv. no. 1029. Öney, 1969a, p.╯196, fig.╯4; İnal, 1970–1, fig.╯6; The Anatolian Civilisations, 1983, p.╯36, D.39. On the dating of the köshk, see Sarre, 1936, pp.╯36–7; Meinecke, 1976, vol.╯1, pp.╯71–2. 35 ╇ Preusser, 1911, p.╯16 top; Sarre and Herzfeld, 1920, vol.╯2, pp.╯152–6, vol.╯3, pls. 10 (lower photograph) – 11; Hartner, 1938, fig.╯26; Sarre, 1936, fig.╯26 (detail); Kühnel, 1950, p.╯11, fig.╯12. Gierlichs, 1996, pl.╯66.1. Meinecke, 1989, p.╯58, fig.╯7. Hauptmann von Gladiss, ed., 2006, p.╯114, fig.╯12. 31 32

chapter two

26

that presumably represented the caliph in the act of subduing a pair of mighty confronted dragons whose expansive serpentine bodies entirely filled the rest of the archivolt. The monumental sculptures, moreover, are not only to be seen as images of power, for the added symbolism of the central figure contributes an important authoritative component as will be considered in chapter 6. A pair of monumental intertwined doubleheaded dragons guards one of the monumental entrance gates to the citadel of Aleppo, which was legendary for its impregnability.36 The large relief-carved frieze with the interlaced dragons surmounts a pointed archivolt with a raised frame at the main portal known as Serpent Gate (Bāb al-Ḥayyāt, re-built probably around 606/1209– 10)37 at the eastern tower of the citadel rebuilt under the Ayyubid ruler al-Malik al-Ẓāhir ibn Salāḥ al-Dīn (568/1173–613/1216). The two heads, one at the springing of the arch and the other at the apex, are crowned by a pair of cusped ears and punctuated with small round eyes; their pointed snouts reveal a row of prominent pointed teeth with bifid tongues thrusting out. Scaly rufflike collars from which project what appear to be tiny, upswept, cusped wings accentuate the base of the necks and delineate the bodies. Their slender, serpentine bodies are thrice knotted on either side into evenly spaced, pretzel-like shapes. The entwined necks at the apex result in an addorsed position of the dragon heads that with wide-open jaws appear to grasp or attack their bodies; this is mirrored in the lower necks and heads of the dragons at the tail tips that are twisted around roundels enclosing eight-pointed star-rosettes (figs.╯3a and b),38 which Willy Hartner has interpreted as solar symbols.39 Astrological symbolism may however be but one component of the commanding composition which in its fantastic, fierce, and awe-inspiring aspect, as Yasser Tabbaa points out, exudes above all also a symbolism of power.40 What is more, the motif serves to strengthen the belief in the impregnability and inviolability of the citadel and to function as a powerful protective device. Added

to the talismanic aspect is the potent symbolism of the mysterious interlace of the dragons41 and the knotting of their bodies, a conspicuous feature on many depictions of the dragon, as discussed below in chapter 10. Significant relations between the Islamic world and the Caucasus region, in particular Armenia, were established in early ʿAbbasid times. Later on, with the Saljuq conquests of the eastern Anatolian region, the Armenian iconographic repertoire served as a source of inspiration, and reciprocal contacts between the Saljuqs and the cultural sphere of the Caucasus were established.42 The close geographic proximity led furthermore to a natural sharing of iconographic emblems. After the Saljuq raids that led to the capture of Ani, the ancient Armenian capital, near Kars, in 456/1064, the city continued to flourish under the Kurdish Shaddādid emirs who subsequently bought the city from the Saljuqs in 464/1072. During the Shaddādid period, as a result of the Byzantine and Muslim wars which led to the interruption of direct trade, Ani became an intermediary of the trade between Iraq and the Black Sea and thus developed into an important and wealthy trading centre.43 The long stylised bodies of a pair of dragons horizontally circumscribe two round towers (nos. 46 and 62) of the northern city wall of Ani. Their bodies are rendered in the form of a thick diagonally hatched moulding, resembling a twisted rope, and end in large heads in profile with gaping mouths revealing rows of teeth and tongues with bifid tips that flank a frontally rendered bovine head which in one case holds a ring in its mouth (fig.╯130).44 The Shaddādids of Ani (c. 464/1072– 595/1198–9 with interruptions) ordered work on the walls, though the reliefs may equally well have been added when the Shaddādid Abū Shudjāʿ Minūchihr ibn Abi ’l-Aswār Shāwur (d.╯c. 512/ 1118) was governing the city.45 In the same vein as ʿIzz al-Dīn Ibn Shaddād’s thirteenth-century elucidation of serpent imagery on a tower in Aleppo which was meant not only to prevent the detrimental effects of snake-bites but also to protect the

╇ Cf. Tabbaa, 1997, pp.╯54, 76. ╇ Herzfeld, 1954–5, p.╯85, no. 36; Tabbaa, 1997, p.╯75. 38 ╇ Cf. Roux, 1980, pp.╯316–7, fig.╯10; Gierlichs, 1996, pl.╯65, 1–3; Tabbaa, 1997, figs. 25, 26. 39 ╇ Hartner, 1938, p.╯144. 40 ╇ Yasser Tabbaa (1997, p. 77) mentions yet another fragmentary dragon sculpture in the form of a stone block which was reused in the late Ottoman rebuilding of the western wall of the Damascus citadel. 41 ╇ Cf. the discussion in Herzfeld, 1954–5, pp.╯236–9. 42 ╇ Cf. Otto-Dorn, 1978–9, p.╯104.

╇ Minorsky, 1953, pp.╯104–5. ╇ Öney, 1969a, pp.╯206–7, figs. 28, 29; Gierlichs, 1996, pl.╯70.2, 3. 45 ╇ The sculptures have been variously dated to the late tenth century (977–989), the Armenian period; to the late eleventh and twelfth century when the Shaddādids carried out work on the walls and added much of the ornamentation; and to the thirteenth century when the Zakʿarids renewed the walls. Sinclair, 1987, vol.╯1, pp.╯360–2; Gierlichs, 1996, pp.╯96–8 with further references.

36 37

43 44

dragons on monumental settings in regions west of iran

27

city’s inhabitants,46 the two paired dragon sculpÂ�tures probably served as Â�powerful protection for the citizens of Ani. The added potency conveyed by the symbolism of the dragons flanking a bull’s head is examined in chapter 7. The dragons’ sculpted representations on the Ani towers however show that their symbolism was not restricted to just one religious creed, as Muslims and Christians alike made use of the iconography. One of the most striking features of Anatolian Saljuq architecture is the chain of caravanserais (khāns) that link the principal cities of the Sultanate of Rūm. Their prime function was evidently to service the north-south overland trade, in other words to provide for and protect travelling merchants. However, as Michael Rogers has pointed out, the east-west trade was much less developed and in spite of the increasingly difficult terrain, the density of distribution of caravanserais east of Sivas noticeably decreases. Nonetheless, to facilitate trade along the Araxes, the local Georgian dynasty built their own chain of caravanserais, which appears to have been modelled on the Saljuq system.47 The depiction of the dragons on the khāns afforded further protection for travellers and caravans from any evil such as raids. A pair of monumental antithetical dragons are depicted on a deeply carved relief band at the back of the entrance īwān at Karatay Han on the former trade road linking Kayseri with Malatya. According to the epigraphic frieze above the main door, the khān was built during the reign of sulṭān Ghiyāth al-Dīn Kay Khusraw II, son of ʿAlāʾ al-Dīn Kay Qubādh I, in 638/1240–1.48 The expansive serpentine bodies, entirely stylised by three parallel moulded bands, are formed of a horizontal guilloche band which extends to frame the entire arch and interlaces at the apex to form a central circular motif, presumably alluding to stellar symbolism, as will be further discussed below. Tongues with bifid tips touch the edges of the central motif, projecting from the toothed jaws of the substantial dragon heads which are finely carved in profile with slightly gaping long wrinkled snouts, the tips

terminating in a tight curl. The heads have small, almond-shaped eyes and the cheeks are enlivened by fine spiralling motifs. A pair of cusped ears crowns the heads; the mane is swept back and covers the uppermost section of the finely carved scaly neck (figs. 4a and b).49 The same khān includes a spring housed in a türbe-like building that is circumscribed by an elaborate muqarnas frieze, comprising a menagerie of fifteen animals among which is the atypical depiction of a single dragon.50 Unusually, the dragon’s body forms not only a pretzel-like knot just below the neck but the very long serpentine tail is knotted to form a maze of interlace (fig.╯5). The dragon’s hide is covered allover with a spotted pattern. Its head with wide-open mouth is turned backwards towards a bird perched in an adjacent niche that holds a round object in its beak. It may be reasonable to propose that the accumulation of knotted interlace in the dragon’s tail symbolised an increase in the protection against evil influences. The resulting maze probably denoted the ability to resist disentanglement by Evil Eyes and may have been considered as added potency. At the same time the complex tangle ensured that the innate forces of the dragon itself are also securely bound in the maze of its own making. Katharina Otto-Dorn, followed by Gönül Öney, associates the reliefs with the twelve animal cycle,51 a view which has been challenged by Jean-Paul Roux on the basis that the discrepancy between the animals depicted on the frieze and the animals associated with the twelve animal cycle is too great.52 At the now partly destroyed thirteenth-century caravanserai, Susuz Han (Susuz Khān), dated c. 644/1246,53 located about one kilometre south of Bucak just off the Burdur-Antalya road, the ogives of a pair of recessed muqarnas niches that flank the portal are each surmounted by a pair of antithetically presented dragons in profile. The heads of the mythical creatures are crowned by curved horns, they have elongated wide-open snouts ending in curled-up tips, the sinuous necks are covered with scales and from the protomes project curved wings and short forelegs. At the

46 ╇ Al-Aʿlāq al-khaṭīra fī dhikr umarāʾ al-Shaʾm wa ’l-Jazīra, the part of Aleppo, tr. and ed. Sourdel, D., Beirut, 1953, p.╯123, as cited in Meri, 2002, p.╯206, n. 360. 47 ╇ Rogers, 1976, pp.╯322–6, and idem, “Saldjūḳids,” EI2 VIII, 936a. 48 ╇ Erdmann, 1961, pp.╯123–4, no. 32. 49 ╇ Cf. Öney, 1969a, p.╯198, fig.╯10; Roux, 1972, p.╯393, figs. 16 and 17; Gierlichs, 1996, pl.╯10.1, 2; Grube and Johns, 2005, p.╯234, fig.╯79.2.

50 ╇ Roux, 1972, pp.╯386–9, figs. 13–5; Önder, 1987, p.╯595; Gierlichs, 1996, pls. 8.3 (complete view of the frieze with Â�fifteen animals); 9.2 (dragon); Hakkı, ed. 2007, p.╯362, fig.╯2 (drawing of animal frieze), p.╯370 (photograph of the frieze). 51 ╇ Otto-Dorn, 1963, p.╯143. For a brief discussion of the twelve animal cycle of years, see the Epilogue. 52 ╇ Roux, 1972, pp.╯387–92. 53 ╇ Erdmann, 1961, p.╯114.

28

chapter two

apex the confronted dragon mouths flank a small, rounded human head with clearly demarcated eyes, ears, nose and mouth (figs. 7 and 124, detail of one muqarnas niche).54 The dragons’ stylised festooned tail, which echoes the contemporary festoon on the arches of the “Kiosk Mosque” at Sultan Han (discussed below), frames the entire arch (without however ending in a second head at the tail tip as shown on the south-facing ogive arch at Sultan Han).55€The composition is further distinguished by a pair of winged figures flanking a central now destroyed motif that seem to hover protectively over the composition and can be assumed to have celestial significance.56 Their presence seems to bestow a honorific dimension upon the enigmatic iconography of the mask-like human heads tightly enclosed by the dragons’ gaping jaws, a feature examined further in chapter 7. The dragon occurs once more on the caravanserai but as a single depiction (fig.╯6). Set within the tight angular interlacing strapwork to the right side of the façade of Susuz Han is the small figure of a single dragon entwined in a pretzel-like knot, its re-curving tail end passing through the knot, which additionally contains, lower down, another depiction of a human face and rosettes.57 An interesting composition involving the draÂ�gon occurs on the façade above a window on Kesikköprü Han situated on the Kırşehir-Kayseri Road to the south of Kırşehir in central Anatolia.

It shows a frontally rendered projecting bovine head from the mouth of which springs a pair of addorsed dragons, with upward arching and once looped or knotted necks, their wide-open jaws oriented to the top.58 An inscription dates the khān to 667/1268–9 and names Jibrāʾīl ibn Chāchā, vizier and governor of Kırşehir under sulṭān Ghiyāth al-Dīn Kai Khusraw III (664/1266–680/1281), as its patron.59 As mentioned earlier, Ani, the ancient Armenian capital, had become an important and wealthy trading centre during the Shaddādid period.60 In 1124, under David II, Ani was conquered by the Georgians who built their own chain of caravanserais.61 David II laid the foundation for the power of the Georgian pan-Transcaucasian monarchy (in Peter Golden’s term)62 that reached its zenith under queen Tʿamar (1184–1211/2). The queen’s victories were chiefly due to the military successes of the Christianised Kurdish generals Zakʿare and Ivane, whose family name in Georgian is Mkhargrdzeli “Longomani.”63 The brothers took Ani in 1199 or 1201 and the queen bestowed it on them as fief.64 Under the rule of the Zakʿarids (the dynastic name of the Mkhargrdzelis), which survived for a while even after the Mongol conquest in 1239, the city experienced a renaissance and became again an important centre of internaÂ�tional trade. The route passed through Armenia to the Black Sea ports where Trebizond had become the

54 ╇ It is noteworthy that in addition to the contemporary carved decoration one also finds the reuse of late antique and Byzantine architectural elements at the khān which includes an example of a section of a lintel carved with a vegetal frieze enlivened with small human faces projecting from the reÂ�Â�cesses; documented during the author’s visit in October 2008. 55 ╇ Riefstahl, 1931, p.╯67, pl.╯125; Kühnel, 1950, p.╯8; Â�Gierlichs, 1996, pp.╯95, 162–4, pl.╯11.1–2. 56 ╇ The motif recalls the winged figures or angels, generally referred to as Nike (Victory) or Tyche (Fortuna), depicted on the sides of the now destroyed Larenda Gate of Konya (Texier, 1862; Sarre, 1910, pl.╯CIX, and idem, 1936, pp.╯8–9, figs. 3, 4; cf. the bas-relief of winged figures of c. 617/1220, now preserved in the İnce Minare Müzesi, Konya, inv. nos. 883, 884) and may also be compared to the victoriae set into the spandrels of the monumental rock-cut arch at Tāq-i Bustān built by Khusraw II Parwīz (590–628), or the early Christian motif derived from the Roman composition of the imago clipeata held by winged figures (cf. L’Orange, 1953, pp.╯90–102); for instance, the flying figures holding aloft a cross within a wreath rendered on the south façade of the seventh-century church of Ptghni (Ptghavankʿ), Ararat (Thierry, 1987, p.╯365, fig.╯199); or on the façade of the church of Dshwari of Mzcheta, built between 586–7 and 604–5 (Baltrušaitis, 1929, pl.╯LXXVI, fig.╯125). The motif can also be observed on portable objects such as the sixth-century Byzantine ivory bindings of the Ejmiatsin Gospels (Der Nersessian, 2001, p.╯155, cat. no. 77). Cf. Redford, 1993, pp.╯153–5. According to Christian church doctrine the souls of the faithful were car-

ried to heaven by angels who guarded them from dangers on the way. The need to curb the cult of angel veneration as it appeared increasingly idolatrous is reflected in the canons of the council of Laodicea (c. 363–364). Protective imagery of this type occurs at entry points of Transcaucasian churches. On each side of the central window of the sixth- to eighth-century church at Ōdzun, for instance, an angel holds what appears to be the coiling tail of a serpent and the serpents’ bodies intertwine to form a knot at the apex (however on account of the advanced surface deterioration the composition is unfortunately difficult to assess; cf. Redgate, 2000, p.╯126); a similar serpent knot surmounts the window of the south portal at the seventh-century church of Mren (c. 640). 57 ╇ Öney, 1969a, p.╯199, figs. 15 a (line drawing) and b (photograph); Gierlichs, 1996, pl.╯11.3–4. 58 ╇ Öney, 1969a, pp.╯184, 207–8, fig.╯31; İnal, p.╯160, fig.╯12; Gierlichs (1996, pp.╯171–2, pl.╯17.6) suggests that the relief carving might be a spolia. When the author visited the khān in 2008, it had just been renovated and the relief with the bovine head between two dragons, which previously had been in very weathered condition, no longer existed. 59 ╇ Erdmann, 1961, p.╯77. 60 ╇ Minorsky, 1953, pp.╯104–5. 61 ╇ Rogers, 1976, pp.╯322–6, and idem, “Saldjūḳids,” EI2 VIII, 936a. 62 ╇ Golden, 1983, p.╯66. 63 ╇ Minorsky, 1953, p.╯102. 64 ╇ Idem, p.╯103. Cf. Barthold [Minorsky], “Ānī,” EI² I, 507a.

dragons on monumental settings in regions west of iran

29

Byzantine capital (1204) after the fall of Constantinople to the crusaders.65 In Ani two caravanserais were built under Zakʿarid rule. The façade of the southern caravanserai (596/1200–633/1236)66 was richly decorated with a pair of confronted dragons with wings and what appear to be forelegs above mythical winged quadrupeds carved onto the spandrels of the ogival arch which was originally covered with a bi-coloured inlay of carved polygons (fig.╯8).67 The overall decorative scheme should however, as Rogers notes, be seen in the context of the evident taste of the Zakʿarid governors of the city for Anatolian Saljuq decoration which had resulted in the creation of a new “semi-Saljuq” Transcaucasian style that continued to flourish in Armenia and metropolitan Georgia long after the decline of the Zakʿarids.68 In the context of the confronted dragon representations in the spandrels of the early thirteenth-century caravanserai in Ani, it is worth mentioning that a near-identical location was reserved for the hybrid beasts in the upper section of architectural structures recorded in the two-dimensional medium of an Armenian manuscript of slightly earlier date, transcribed and illuminated in Cilician Armenia. The shift in geography of Armenian cultural centres from the Armenian plateau westwards occurred after Saljuq raids that led to the capture of Ani in 1064 and Kars the following year, at which time the king, Gagik-Abas, was driven into Cappadocian exile. This led to massive western migrations of Armenians which contributed to the re-Armenisation of ancient Armenia Minor as well as Cappadocia and Cilicia to the south. In the kingdom of Cilicia in Tarsos, miniature painting attained a high degree of excellence. Here in the monasteries of Mlich and Skevra the L’viv (Lemberg) Gospel was transcribed and lavishly illuminated from 1193 until 1198/99.69 The ten canon tables

are set within architectural frames consisting of two columns supporting arches. Interestingly in the first two tables the uppermost sections of the rectangular headpieces carry confronted winged dragons, with long raptor-like forelegs, in the spandrels that surmount the archivolts. The beasts are set against curling foliage, which is held in the dragons’ snouts on the headpiece over the Letter of Eusebius to Carpianus on the second canon table (fig.╯9).70 The iconography of the dragons on these miniatures is clearly identifiable as Eastern Christian, or perhaps particularly Armenian, distinguishable from the Jazīran and Anatolian “Saljuq-style” dragon by the long raptor-like legs and the shorter snouts. The Armenian predilection for representing dragons may perhaps be associated with the fact that in Armenia the dragon (vishap) belongs to the pre-Christian substrate71 and as a result is part of an ancient iconographical tradition, combined here with canonical scenes from Christian iconography. In spite of the fact that these represent two-dimensional compositions on paper, they nonetheless suggest that placing paired dragons in the spandrels of an arch was not rare in pan-Transcaucasian architecture. This is further corroborated by an analogous composition found on the arch of a thirteenth- or fourteenth-century wooden door from Godaïk in Ararat province, which is carved with confronted dragons represented without wings or legs.72 The placement of the dragons in the arch of a door may once again support the supposition that a protective function was intended. A stone relief, now no longer extant, of paired confronted dragons with a quadripartite knot at mid-section with further knotted interlaces above and below was found at the hospital (darüşşifa) of the atābeg Lālā Jamāl al-Dīn Farrukh during the reign of the Rūm Saljuq sulṭān ʿAlāʾ al-Dīn Kay Qubādh I of Konya in Çankırı (Gangra),

65 ╇ Manandian, A., O torgovle i gorodakh Armenii, Erevan, 1954, p.╯278, as cited in Barthold [Minorsky], “Ānī,” EI² I, 507a. Cf. Redgate, 2000, p.╯258. 66 ╇ Rogers, 1976, p.╯324. 67 ╇ A photograph of the dragon relief in the right spandrel is reproduced in Sakisian, 1940, pl.╯XVIII, fig.╯33. 68 ╇ Rogers, 1976, pp.╯315–26. 69 ╇ Prinzing and Schmidt, eds., 1997, pp.╯18–21. 70 ╇ Akinian, 1930, p.╯7, fig.╯1; Prinzing and Schmidt, eds., 1997, pls. I, II. Comparable dragons, likewise positioned in the spandrels of the rectangular structures that surmount the archivolts, feature also in thirteenth-century Armenian manuscripts, such as in the Gospel book commissioned in 1273 by Ter Simeon, the abbot of the monastery of Skevra (Istanbul, Topkapı Sarayı Museum, Library, Ms. 122, fol. 8,

Letter to Eusebius, see Der Nersessian and Agemian, 1993, vol.╯2, fig.╯318, page to the right). 71 ╇ Russell, 2004, p.╯453. Of note are the large carved stone steles found throughout the Transcaucasus and beyond, perhaps datable to the first two millennia bc, often referred to as vishap stones, that are generally erected near a spring or reservoir; hence, they probably are of some ritual or Â�religious significance. Cf. Marr and Smirnov, 1931; Piotrovskiy, 1939. The steles are also sometimes called vishap azhdahā on account of their prodigious size by analogy with azhdahā mard, “giant man”; azhdahā kʿar, “megalith”; and vishap kʿar, “stone fish.” See Tchukasizian, 1964, p.╯326 and n. 21 with further references. 72 ╇ Hovsépian, 1937, pp.╯164–5, fig.╯67.

30

chapter two

outside Ankara, built according to the inscription in Muḥarram 633/1235.73 The serpentine bodies were oriented to the left, their confronted heads with open jaws demarcated by almond-shaped eyes and topped by cusped ears, revealing sharp teeth and tongues.74 The heads would probably have been confronted although it is no longer possible to reconstruct this on account of the deteriorated condition of the lower dragon head (fig.╯175). Interesting in this regard is the record by Albert Gabriel of Süheyl Ünver’s suggestion that the paired dragons, joined at mid-section by a quadripartite knot, were an ancient symbol of healing used on hospitals, transmitted through the Saljuqs,75 an interpretation followed by Mehmet Önder who also associates the dragon iconography on the darüşşifa with healing.76 A second affiliation of the dragon with hosÂ� pitals in Anatolia is found among the plaster reliefs on the façade of the Kay Kāwūs Darüşşifa in Sivas, datable to 614/1217, built during the reign of Kay Kāwūs I ibn Kay Khusraw I (r. 608/1211– 616/1220). Now in very poor condition, they show traces of the body of one dragon with forelegs and spiralling tail, which can be presumed to have been complemented by a second dragon.77 Only three examples of the representation of dragons are known from Islamic sacred architecture from the period of the Saljuqs and their “successor states.” One example is the stone relief at a mausoleum (türbe) known locally as that of the Emir Saltuq after whom the Türkmen Saltuqid dynasty (c. 465/1072–598/1202), former commanders of the Saljuq army, is named. It is the largest and most unusual of a complex of three tombs (Üç Kümbetler) just south of the walled city of Erzurum, near the Tabriz Gate.78 The drum is circumscribed by eight fan-shaped arched niches formed by the gables of the octagon’s roof, inside which are carved animal, vegetal and geometric compositions.79 Among the reliefs is a pair

of upright confronted dragons whose wide-open elongated fleshy snouts with upward-curling tips reveal a row of sharp teeth and projecting tongues and touch at the tips to form a diamond-shaped enclosure. Small, rounded eyes and small, pointed ears demarcate the heads. Their bodies entwine to form a loop at mid-section and interlace again at the bottom angle of the niche whence they curve upward to frame both diagonal sides. The two creatures thus entirely fill the architectural space into which they are fitted (fig.╯10).80 A striking parallel to the dragons on the türbe of Emir Saltuq is found at the church of Saint Gregory, which belonged to a monastery, located at the edge of Ani above the cliffs of the Arpa Çay gorge. The presence of a new class of wealthy merchants that formed during the eleventh and twelfth century in Ani is attested to by the inscription of the merchant Tigran Honentsʿ on the church he erected in 1215 and dedicated to Saint Gregory the Illuminator. The Greek-Orthodox tendency favoured during Zakʿarid rule continued to predominate in the architectural design of this church.81 However the depiction of the dragons, which appears on the fan-shaped spandrels of a blind arcade, follows the well-established Saljuqperiod conventions. The recumbent confronted dragons are carved in a horizontal arrangement in the upper section. They are portrayed resting on their forelegs, their heads crowned by a pair of pointed ears. The hybrid creatures have squinting eyes and the characteristic wide-open jaws with rolled-up ends, sharp teeth and flickering tongues with bifid tips which nearly touch at the centre. The sizable upper bodies extend into tapering tails which loop twice then arch over the back (fig.╯11).82 It is interesting that the placement of the dragons in a niche is not the only feature to recall the dragons depicted on the türbe of Emir Saltuq, the date of construction of which is uncer-

73 ╇ Meinecke, 1976, vol.╯2, p.╯103. Cf. the general references in van Berchem, 1910, pp.╯82–4; Gabriel and Sauvaget, 1940, vol.╯1, p.╯166, fig.╯137. 74 ╇ The relief has been lost since 1940 and is only known from photographs and drawings done before this date. Cf. Gierlichs, 1996, p.╯156. 75 ╇ Cited after Gabriel and Sauvaget, 1940, p.╯168, n. 1. 76 ╇ Cf. Süslü, 1987, p.╯641. By the Ottoman period, the hospital became a place of spiritual healing and snake charming. Terzioğlu, A., Mittelalterliche islamische Krankenhäuser, Technische Universität, Berlin, 1968, p.╯126, as cited in Tabbaa, 2003, p.╯112. The conspicuous knotting aspect of the dragons is discussed in chapter 10. 77 ╇ Öney, 1969a, p.╯198, fig.╯9. 78 ╇ The city withstood the Saltuqid onslaught until 473/ 1080, when it became the capital of the Saltuqid principality.

It is during the period of their rule, which lasted for about thirty years, that Otto-Dorn (1964, p.╯151) suggests that the construction of the türbe took place. A later dating before the city was taken by the Mongols in 639/1242 has also been put forward. Ünal, 1968, p.╯160; Gierlichs, 1996, p.╯146, n. 6. For a mid- or late fourteenth-century dating, see Sinclair, 1998, p.╯212. 79 ╇ Cf. Sinclair, 1998, p.╯212. The compositions have been associated with the animal cycle, see Otto-Dorn, 1978–9, pp.╯126, 144. 80 ╇ Öney, 1969a, fig.╯23; Otto-Dorn, 1978–9, p.╯126, fig.╯22; Gierlichs, 1996, pp.╯145–7, pl.╯1.8. 81 ╇ Barthold [Minorsky], “Ānī,” EI² I, 507a. 82 ╇ Otto-Dorn, 1978–9, p.╯127, fig.╯23; Gierlichs, 1996, pl.╯70.6.

dragons on monumental settings in regions west of iran

31

tain. The dragons on the mausoleum are rendered upright with doubly entwined bodies and without forelegs, whereas the dragons on the church of Saint Gregory have a more horizontal orientation with individually looped tails. Nevertheless, there is a consistency in the overall iconographic programme, which is why the dragon sculptures on the so-called türbe of Emir Saltuq may probably be asigned a thirteenth-century date. The second Saljuq dragon sculpture on Islamic sacred architecture is found on the small “Kiosk Mosque” situated in the arcaded rectangular courtyard of the double-section caravanserai Sultan Han, located northeast of Kayseri, on the main road that once linked Konya, Kayseri and Sivas to the east (Iraq and Iran). It is the second largest Saljuq caravanserai in Anatolia and was built between 629/1232 and 633/1236 on the orders of ʿAlāʾ al-Dīn Kay Qubādh I of Konya, as attested by an inscription on the portal. Resting on a four-bay substructure, the south- and east-facing monumental ogival arches are both symmetrically framed with a complex serpentine festoon. The latter is composed of reciprocally arranged pretzel-like shapes, culminating at the apex in confronted dragon protomes. The heads with large, almond-shaped eyes, topped by backward-projecting pointed ears, have wide-open jaws revealing sharp teeth and tongues (fig.╯12).83 Both the south- and east-facing reliefs are closely related but while the dragon protomes on the south side do not touch each other at the apex, the bodies of the dragon protomes on the east side are joined and enlivened by dots.84 Moreover, on the south-facing arch the dragon festoons end in small, inverted dragon heads with large eyes, necks bent inwards, with the open jaws appearing to hold the tip of the outer edge of the festoonband;85 it is thus interesting to observe that they seem to bite (in other words “swallow” or “disgorge”) their own tail tip, an aspect discussed in chapter 9 (fig.╯13). While this feature is not

recognisable on the east-facing arch (possibly on account of the surface deterioration), it shows, interestingly, an additional upward oriented, small dragon head, growing out of one of the bends of the dragon festoon to the left (fig.╯14).86 Also of note is the fact that both serpentine festoons are surmounted by a further band enclosing a tightly woven knotted composition distinguished by a small eight-petalled star-rosette in the interstitial area at the apex (although on the east side, on account of the surface deterioration, this is no longer identifiable). Otto-Dorn interprets the rosette as a planetary symbol suggesting an astralmythological reading of the iconography.87 The third example is found in the relief sculpture of a pair of dragon protomes which spring from the base of a central vegetal composition topped by a double-headed eagle on the façade of the Çifte Minare madrasa at Erzurum (probably before 640/1242–3).88 The relief featuring dragons springing from a conventionalised vegetal or tree-like composition, whose important symbolism is examined in chapter 4, is set within ogival niches at either side of the main façade, but only the relief to the right was completed (fig.╯43).89 Rogers associates the Çifte Minare madrasa as well as the buildings of the Great Mosque and hospital at Divriği (626/1228–9 or later) with elements of a Caucasian building tradition (in particular with the influence of western Georgia (Tao-Klargeti)).90 A fourth, yet less conspicuous, instance of dragons on sacred architecture of the Saljuq period may be noted in passing. The end of one of the inscriptions at the façade of the Ak Mosque (617/1220–634/1237) at Anamur near Alanya shows a single double-headed dragon, knotted at mid-section and terminating at either end in an ophidian head.91 In the sacred architecture of the pan-Transcaucasian realm the dragon, as mentioned above,

83 ╇ Cf. Riefstahl, 1932, p.╯92; Kühnel, 1950, p.╯8, fig.╯15; Erdmann, 1961, pp.╯94–5, fig.╯152; Öney, 1969a, p.╯197, figs. 6, 7, 7a, and eadem, 1978, p.╯45, fig.╯31; İnal, 1970–1, p.╯163, fig.╯23; Otto-Dorn, 1978–9, p.╯127, fig.╯24; Grube and Johns, 2005, p.╯234, fig.╯79.4. For a detailed description, see also Roux, 1980, pp.╯316–7 and fig.╯10. 84 ╇ Öney, 1969a, figs. 6, 7; Gierlichs, 1996, pl.╯6.1, 2. 85 ╇ This detail is documented by Öney, 1969a, fig.╯7a. Cf. Gierlichs, 1996, pl.╯6.1, featuring the entire festoon on which however it is difficult to discern this feature. 86 ╇ See detail in Gierlichs, 1996, pl.╯7.3. 87 ╇ Otto-Dorn, 1978–9, pp.╯130–1, fig.╯24. 88 ╇ Michael Rogers (1965, pp.╯63–85 and 1974a, pp.╯77–

119, esp.╯117–9) suggests that the fall of Erzurum to the Mongols in 1242–3 represented an architectural (as well as political) terminus ad quem for the dating of the Çifte Minare madrasa. For a comprehensive list of suggested dates of the Çifte Minare madrasa, see Meinecke, 1976, pp.╯136–7. 89 ╇ Cf. Bachmann, 1913, pl.╯66; Öney, 1969a, p.╯208 and fig.╯32, and eadem, 1969b; Otto-Dorn, 1978–9, p.╯127, fig.╯25; Gierlichs, 1996, pl.╯12.1, 4, 5. 90 ╇ Rogers, 1974a, pp.╯103–6 and ns. 76, 77. Cf. idem, “Saldjūḳids,” EI2 VIII, 936a. 91 ╇ Roux, 1980, p.╯316; Öney, 1969a, p.╯176, figs. 12 a and b.

chapter two

32

is found already at a much earlier date. While its different manifestations will be examined in the following chapters, a noteworthy example, part of a frieze on the eastern façade of the Georgian church of the Virgin in Martʿvili in western Â�Georgia, founded by king George II of Aphkhasia (912–957),92 may be mentioned in this context. The frieze, which is finely carved with scrolling vines bearing clusters of grapes, also features Â�several composite animals: a winged dragon, shown in profile with forelegs, is flanked on either side by other imaginary creatures. On its left is a centaur-like depiction with Â�quadruped body in profile, extending into a frontally rendered human upper body with large, stylised heart-shaped head which appears to be crowned by a halo, while on its right is a composite mythical animal now commonly identified as Sasanian-style sēnmurv. It is rendered in profile with the protome of a canine dragon and the characteristic peacock tail.93 The dragon’s head is turned backwards towards the centaur, its long gaping snout revealing a row of teeth, the bifid tongue projecting towards its raised, unfolded wing. The creature’s very long sinuous ophidian body forms a large loop, ascending behind the body to descend and taper to a point below the body of the centaur, whose forelegs rest on the attenuated tail end. Of note is the ruff-like loop around the dragon’s neck, the ends of which curve sharply upwards, a feature which appears to be shared by the centaur (fig.╯15). The apotropaic character of these mythical animals, which are also a salient feature on the façade of the Armenian church of the Holy Cross at Aghtʿamar (915–921), may be presumed.94 Further of note is the striking difference in the depiction between this apotropaic dragon, rendered with forelegs, and the relief with a bicephalic ophidian dragon speared by two riders shown above the western entrance of the church, discussed below (fig.╯107). Another important architectural feature is found at the Armenian monastic ensemble at

Sanahin, near Alaverdi in Lori province, Armenia, which was established in 966. The monastery, which also served as Bagratid necropolis, not only functioned as a religious retreat but was an intellectual academy with a scriptorium for the copying and illuminating of manuscripts as well as a library. Significantly, it is in the latter which was added in 1063 that relief-carved interlaced dragons are prominently depicted on a column capital. After the destruction by the Mongol invasion decades earlier, the library was restored by bishop Stephen Sarkis and his disciples Herapet, David, Hesou and Karapet, in the late twelfth century (fig.╯16).95 Of particular significance is the dragon iconography in the Eastern Christian sphere of the Jazīra. Extremely interesting are its representations at the thirteenth-century monastery of Mār Behnām, also called Deir al-Khiḍr, southeast of Mosul, near the ancient Assyrian city of Nimrud situated between the river Tigris and the upper Zab.96 The large, fort-like monastery was founded in the second half of the fourth century as a memorial to the Christian martyrs Behnām and his sister Sarah.97 The lintel immediately above the southern outer door in an internal corridor of the monastery carries a relief-carved depiction of a pair of addorsed recumbent dragons resting on their forelegs. Their expansive finely scaled ophidian tails form a pretzel-like knot followed by a loop and are interlaced to form a horizontally oriented figure of eight along the central axis. Small narrow wings attached to the haunches extend towards the back. The creatures are flanked by seated lions portrayed in profile with head en face. The paws of the inner legs are slightly raised and the long tails are drawn behind the flank of one of the hind legs and ascend in a slight curve terminating in dragon-headed finials behind the arched back. A weathered frontally rendered lion head

╇ The church of the Virgin in Martʿvili is variously dated between the seventh and the tenth centuries, cf. Baltrušaitis, 1929, p.╯104 (“Martʿvili”). Mepisaschwili und Zinzadze (1987, p.╯160, fig.╯234) date it to the seventh century. 93 ╇ For a detailed discussion of the sēnmurv, see Harper, 1961–62, pp. 95–101, and Schmidt, 1980, pp.╯1–85. 94 ╇ Cf. Der Nersessian, 1965, figs. 23, 28, 29, 43. 95 ╇ Baltrušaitis, 1929, pl.╯LXIV, figs. 99, 100; Sakisian, 1939, fig.╯30; Manoukian, Agopik and Armen, eds., 1967–9; Khalʿpakhʿchian, 1971, p.╯145, fig.╯110, right side (drawing). 96 ╇ The dating of the reliefs is uncertain. Inscriptions in the church show that renovations of the choir were undertaken by the patriarch Mār Athanasius of Antioch in 559/1164 (Pognon, 1907, pp.╯134–5, no. 75). Herzfeld (Sarre and

Herzfeld, 1920, vol.╯2, p.╯266) dates the reliefs to the period of Badr al-Dīn Luʾluʾ (618/1222–657/1259). Further extensions and renovations were undertaken between 1248 and 1295. It is further of note that a 33 line Syriac inscription engraved on the walls of the church attests that the monastery was spared from destruction by the Ilkhan Baidu in 1295. Pognon, 1907, pp.╯132–42, 235, no. 76; Fiey, 1959, p.╯50, and idem, 1965, pp.╯584–5. For a detailed discussion of the dating, see idem, 1965, pp.╯590–7. 97 ╇ The monastery is named after Behnām, allegedly of the Assyrian family of Sennecherib II, the governor of Nineveh, who was killed together with his sister during the persecution of the Christians by Ardashīr (279–283), son of Shāpūr. Cf. idem, 1965, vol.╯2, pp.╯565–74.

92

dragons on monumental settings in regions west of iran

33

projects above the centre of the relief. Below, on the actual lintel, a Christian cross, from whose base also spring regardant quadruped dragons, is flanked, in turn, by pairs of confronted regardant birds whose tail feathers end in dragon heads with gaping jaws snapping at the birds’ tails (figs. 17a and b, details figs. 50 and 74).98 The symbiosis of birds or felines with the dragon will be further examined in chapter 5. The bilaterally symmetrical configuration of the dragons shows them with entwined tails tapering to a pointed tip and forming a vertically oriented figure of eight. Their scaly serpentine bodies are doubly intertwined in a pretzel-like knot and a simple loop.╯The large heads, with characteristic wide-open jaws revealing two long pointed fangs and twisted forked tongues, face stemmed cups. The almond-shaped eyes seem to squint slightly; small, cusped ears project at the top of the head. Their supernatural properties are underlined by a pair of arched narrow horns, swept towards the back, and by the slender, curved wings that spring from an ornament that winds around the dragons’ haunches, terminating in an angular curl (a detail which can also be seen in the Cizre dragon-knockers, fig.╯83). The inner front leg is slightly raised, a feature that is paralleled in the figures of the flanking lions. Significant are equally the closely related carved mouldings that provide an enlarged frame for the entire portal both at the monastery of Mār Behnām and at the thirteenth-century mausoleum of Imām Bahir in Mosul,99 which was possibly erected at the order of Badr al-Dīn Luʾluʾ (618/1222–657/1259).100 In both cases stylistically and iconographically closely related compositions

of six pairs of dragons form trilobed arch-shaped niches that accentuate a portal. At Mār Behnām the second southern portal leading to the chapel of the baptistery is framed by a moulding in the form of a Â�knotted interlace composed of the bodies of six Â�dragons that enclose a total of twenty-one niches (fig.╯18).101 The horizontal section of the portal is defined by the entwined necks of the addorsed dragons, with the exception of the central arch-shaped enclosure that is topped by a projecting lion head en face. The dragon heads are crowned by arched horns and small, cusped ears folded to the back; they are rendered with the characteristic wide-open mouths revealing a long tongue, the elongated upper jaws terminating in a rolled-up tip.╯At the base the tails form a single loop.╯The vertically interlaced ogival arch-shapes each comprise eight niches: three enclose hooded standing figures holding a book in the left hand and a cross in the right hand, probably representing monks,102 while the others contain the symbol of the cross. The horizontal interlace encloses a tall standing figure next to a small, nude figure and a rider on horseback,103 also alternating with the cross motif. The lintel carries a central Christian cross which extends at the base into a stemmed palmette. At the portal leading to the vestibule at the mausoleum of Imām Bahir the frame is similarly set off with a knotted interlace, here accentuated with a scaly pattern, enclosing twelve niches, which contain foliate arabesques surmounted in the arched section with a tiny muqarnas (fig.╯19).104 The lintel is inscribed with the words “Muḥammad, al-Malik, Allāh, ʿAlī.” The outer dragon pairs form a vertical interlace that encloses

98 ╇ Cf. Preusser, 1911, p.╯11, pls. 5, 6.2; Fiey, 1959, p.╯145, n. 1 a, fig.╯12; and idem, 1965, vol.╯2, pp.╯565–9, esp.╯pp.╯605–6; Kühnel, 1950, p.╯12, fig.╯14; Gierlichs, 1996, pl.╯59.1 (upper section of door). 99 ╇ Located near Bāb Sinjār. The entire portal frame was brought to Baghdad in 1939, cf. Fiey, 1965, p.╯595, n. 1; Gierlichs, 1996, p.╯230, n. 455. Ministry of Culture and Arts, ed., Guide to the ʿIraq Museum, 1979, fig. 49. 100 ╇ Farès, 1953, p.╯52. An inscription on the eastern wall gives the date 699/1300; however it is unclear whether this date refers to the construction or a renovation of the Â�monument. Another inscription on the southern wall names the builder, ʿAbd al-Rāḥim ibn Aḥmad. See Uluçam, 1989, p.╯141. 101 ╇ Preusser, 1911, pl.╯12; Fiey, 1965, pl.╯E; Gierlichs, 1996, pl.╯59.4 (upper section of portal). 102 ╇ Two figures are identified by inscriptions; Fiey, 1965, p.╯601. 103 ╇ The two carvings are often referred to as relating to the scene of the baptism of Sarah by Mār Matta, yet on account of the nude figure it has been suggested that they might either represent the baptism of Mār Behnām (Fiey, 1965, p.╯599 and n. 2), or the baptism of Christ (Sarre and Herzfeld, 1920,

vol.╯2, p.╯247); the other carving has been interpreted as the entry into Jerusalem (eidem, 1920, vol.╯2, p.╯247). 104 ╇ Also published in Mossoul au temps des Atabeks, ed. Saʿīd al-Daywahchī, Mosul, 1958, fig.╯36; ʿAṭṭa al-Hadīthī and Hanāʾ ʿAbd al-Khāliq, al-Qibāb al-Makhrūṭiyya fi ’l-ʿIrāq (“Conical Domes in Iraq”), Baghdad, Ministry of Information, Directorate General of Antiquities, 1974, pl.╯82; Janabi, 1982, fig.╯170 B; Uluçam, 1989, pp.╯141–3, figs. 314, 315. The tomb chamber of the mausoleum, which carried fragments of the Āyat al-Kursī, the Throne Verse (sūra 2, 255), on the four walls of the chamber, a mihrab with a hanging lamp and prayers for the holy family of Shīʿism on the qibla wall, is discussed by al-Suyūfī (d.╯1901), Majmūʿ al-kitābāt al-muḥarrarah fī abniyat al-Moṣulʾ, ed. Saʿīd al-Daywahchī, Mosul, 1956, pp.╯141–7; Ibn Faḍlallah al-ʿUmarī, Manhal al-awliyāʾ wa mashrab al-aṣfiya min sādāt al-Moṣul al-hadbāʾ, ed. Saʿīd al-Daywahchī, Mosul, 1967, pp.╯225–6; both cited after Khoury, 1992, pp.╯23–4, ns. 20, 28. Shrines with references to Shīʿī imams and their descendants are thought to have played a propagandistic role in Badr al-Dīn Luʾluʾ’s campaign to convert madrasas into Shīʿī shrines. Cf. also Khoury, 1992, p.╯14, n. 29, with reference to Janabi, 1982, pp.╯53–4.

34

chapter two

four niches each, knotted with a simple loop at the points of juncture. At the top the dragons terminate in the same manner as on the portal of Mār Behnām; however at the bottom of the niches the slightly tapered tails entwine to form a loop, whence they curl inward to terminate in a bird’s head with curved beak which pecks at the tail (comparable to the birds on the tail tips of dragons on the Cizre door-knockers (fig.╯83) and on the wing tips of the dragons at the Bāb al-Ṭilasm in Baghdad (fig.╯139a)). Just like the bottom mouldings, the central niches terminate at the top in addorsed heads and at the bottom in

tail ends that entwine to form a loop and thence taper to a pointed tip.╯Also of note are the small star-rosettes which on both portals are set into the interstices of the vertical bands. The representation of dragons situated at the approach to the most sacred part of the building, both in the case of the monastery of Mār Behnām, where they feature on the second portal leading to the chapel of the baptistery, and at the mausoleum of Imām Bahir, where they appear on the portal leading to the vestibule, indicates that their depiction was intended to serve as monumental apotropaion and to guard the entrance to a sacred space.

the dragon motif on portable objects

35

chapter three

The dragon motif on portable objects

By their very nature, portable artefacts are characterised by their potential for movement across geographical, cultural and religious boundaries.1 Most of the objects under discussion testify to an overall westward migration of the dragon iconography from the Western Central Asian regions, facilitated by the movement of ever growing numbers of Turkic-speaking tribes, the dominant force being the Saljuqs, into Western Asia. The latter took much from the artistic traditions of the Ghaznawids whose state they had destroyed.2 During the Saljuq period, the main artistic centres were located in greater Khurasan and northern Iran, with merchants and artisans becoming the principal bearers of cultural traditions.3 Able to move freely in the enormous empire, which reached from Khurasan to Syria and Asia Minor, the population aided the diffusion of prevailing styles and tastes which had a “markedly Khurasanian flavour.”4 Belonging to a wider, shared visual culture across the medieval Central Asian sphere, portable objects shared emblematic themes, reflected also on monumental representations. The dragon motif is found on a great variety of artefacts, notably in the category of personal objects: specifically on accoutrements pertaining to the hunt or war and objects of personal adornment. The semantic value of the dragon motif on the former would naturally have served to enhance the efficacy of these items. When personal in nature and worn on the body, objects with this motif were popularly believed to provide the wearer with a prophylactic or apotropaic safeguard against a

variety of real or imaginary dangers. The belief in the magical power of images meant that they could function as talismans intended, for instance, to promote well-being and to protect from the power of evil. This is related to the age-old belief in the agency of envy and jealousy and the Evil Eye, which certainly survived through the medieval Islamic period,5 and the neutralisation of the harm that was intended to the person by wearing such a piece. Hence the serpent or dragon comes to be looked upon as harbinger of good luck and bestower of prosperity. In addition, such objects were often believed to endow their owners with certain abilities or powers. One of the most widespread functions of talismans was in the form of amulets (tamāʾim, sing. tamīma, or taʿāwīdh, sing. taʿwīdh) intended to gain the assistance of unseen, supernatural forces that were believed to influence the affairs of humankind to achieve certain desired outcomes. It is of course impossible to fathom how “potent” the motif was for the wearer, it may indeed be that some did not consider the dragon motif to have such explicit powers, using it as an ornament, but with prophylactic intent. In general, though, it may reasonably be conjectured that the dragon iconography carried implicit semantics imbued with passively apotropaic, that is to say protective or actively beneficial properties, in other words empowering qualities designed to be imparted to the wearer. Other objects such as vessels with this iconography may often have functioned as portable “apotropaia.” In spite of the fact that many pieces have a very varied iconography of which the dragon is only part, it may be hypothesised that dragon motifs, too, served to magnify the

╇ Cf. Hoffman, 2001, pp.╯17–22. ╇ Marshak, 1986, p.╯358. 3 ╇ Melikian-Chirvani, 1974, pp.╯112, 114; Marshak, 1986, p.╯359. 4 ╇ Cf. Melikian-Chirvani, 1974, pp.╯112, 114. See also p. 18, n. 25. 5 ╇ In his Qaṣīdah Sāsānīyah, the tenth century poet Misʿar

ibn Muhalhil al-Khazrajī al-Yanbūʿī, known as Abū Dulaf, refers to beggars who wrote out talismanic charms and sold little clay tablets allegedly made from the earth of the tomb of al-Ḥuseyn in Karbalāʾ, undoubtedly alluding to the production of amulets. For this and further references, see Bosworth, 1976, vol.╯1, pp.╯86–8, 90, 128, vol.╯2, pp.╯192, 198–9, 221, 243.

a.╇ Overview

1 2

36

chapter three

intended effect of the vessels which were presumably meant to protect their maker6 and more often their owner, so functioning as protective devices. This is emphasised by inscriptions invoking familiar expressions of wishes for blessings, luck, health, or long life for the mostly anonymous owner, which are frequently of amuletic character in themselves; these are rendered often in combination with figural decorations of a symbolic or “magical” significance such as the dragon. A clear function of the dragon’s iconography was thus to reinforce the propitious, apotropaic, or even magical or supernatural powers of such portable objects. b.╇ The dragon motif on accoutrements relating to the hunt or war Dragon imagery is attested on weapons and banners from early Zoroastrian times. It is particularly associated with the mace and with the finials of ceremonial weapons or staffs which may carry sculpted dragons with a human or animal figure in their maw. Dragon banners, an important part of military insignia from ancient times, appear in Iranian art and literature. The dragon motif similarly occurs on weapon fittings, ritually significant belt/strap fittings and equestrian accoutrements. The use of such imagery on the paraphernalia of heroism and rulership communicated mastery over the dragon and appropriation of its formidable qualities. When featured on objects of personal adornment such as jewellery or beltornaments, the motif endowed such items with prophylactic or apotropaic powers.

6 ╇ Cf. the twelfth-century large silver-inlaid brass ewer, now in the Georgian State Museum, Tbilisi, Georgia, which bears the inscription:

... seven heavenly bodies, however proud they may be, are protection for the one who works so. Allan, 1982a, repr. 1999, p.╯49. 7 ╇Yasht 10.96; cf. for instance, also, “cudgel of bronze” (Rigveda 1.80.12). Gershevitch, 1959, p.╯121; Watkins, 1995, p.╯411. 8 ╇ Watkins, 1995, pp.╯411–3. 9 ╇ Idem, pp.╯331–2. 10 ╇ Wikander, 1938, pp.╯60, 64–6, 99; Widengren, 1969, p.╯249. 11 ╇ Asadī Ṭūsī, Garshāsp-nāma, p.╯269, l. 10, referred to by Khāleqī-Moṭlaq, “Aždahā II,” EIr; De Blois, “Â�Garshāsp-nāma (or Karshāsp-nāma),” EIr; Boyce, 1975, repr. 1996, p.╯63; Sarkhosh Curtis, 1993, p.╯26; Bivar, 2000, p.╯24. 12 ╇ Rosenfield, 1967, figs. 2, 2b (photograph on the left).

The dragon motif on weapons Cudgels seem to have been the most widespread weapon in early Mesopotamia and Iran. “Cast in yellow bronze,”7 they were the most powerful and the most victorious of all the weapons of the Vedic and Avestan gods.8 Cudgels or maces were also associated with the dragon-fighting Indo-Iranian mythical heroes,9 and hence possibly serve to characterise them as primordial warriors.10 Significantly, the mace of the legendary dragon-fighting hero Kərəsāspa (Garshāsp in New Persian poetry), celebrated already in the Zoroastrian Yasna and Vidēvdāt, is said to have been carved in the shape of a dragon head.11 Similarly the Mathura portrait statue of the Kushāṇa king Kanishka is shown with a giant club tapering to an open-mouthed head of a dragon-like creature.12 The representation of the dragon on ceremonial weapons must be evidence of the intention to endow the weapon and hence its owner with the magical powers of the dragon. The club was greatly favoured by the Parthians (250 bc–226 ad), too. The club of Herakles, the most popular of Greek heroes, even appears as architectural decoration in the early Parthian monument referred to as the “Round Hall” in Nisa, the Parthian metropolis in present-day Turkmenistan.13 Maces and battle axes were used in the Parthian and Sasanian periods.14 An important depiction is found on a bas-relief from a small house-temple in Parthian Hatra in northern Mesopotamia (an integral part of Iran in Parthian and Sasanian times), which shows the composite figure of Herakles-Nergal,15 the god of the realm of death and the underworld, who can

The heads of the mythical creatures, the Indian composite marine creature, makara, and the Central Asian dragon, were sometimes portrayed in a stylistically closely related manner. Since only the head is portrayed, it is impossible to identify it with a degree of certainty as belonging to either creature, though both the makara and the Central Asian dragon can to a large extent be considered semantically equivalent. Clubs terminating in dragon-like heads are featured in theÂ� seventh-century wall paintings at Sogdian Afrāsīyāb; see Albaum, 1975, fig.╯13. On the makara in Indian iconography, see Vogel, 1929–30, pp.╯133–47; Coomaraswamy, 1928–31, repr. Delhi, 1971, pp.╯47–56, esp.╯pp.╯47–9; Combaz, 1945, pp.╯146–55; Bosch, 1960; Rosenfield, 1967, pp.╯179–83; Boardman, 1986, pp.╯451–3. 13 ╇ Colledge, 1986, p.╯21 and pl.╯XLIa. 14 ╇ Chegini and Nikitin, 1996, repr. 1999, p.╯53. 15 ╇ Bivar, 1975a, vol.╯2, pl.╯4a; Drijvers, 1978, p.╯172; Â�Winkelmann, 2004, pp.╯248–51, fig.╯102.

the dragon motif on portable objects

37

be at once life- and death-giving,16 whose attribute is the serpent.17 Clad in Parthian garb he is shown wielding with one hand a double-bladed battle axe, the right blade of which is replaced by a serpent, and clasping the hilt of a dagger in his other hand. The god is girded with a snake-like rope to which three quadrupeds, probably dogs, are connected, the tail of one of the quadrupeds being also in the form of a serpent. Serpents spring from the god’s shoulders and rise from either side of his waist, while another serpent rests at his feet. To his right he is flanked by a large cult-standard (semeion), near the foot of which another snake and a scorpion are featured. The close association of the ophidian creature with the archaic weapon is reflected in the Persian word gurza which not only means “a large wooden club or mace,” but also “a large headed serpent.”18 Among the Iranian and Turkish tribes the gurz was a weapon of special ceremonial importance.19 Mace bearers were part of the guard of the Samanid and Ghaznawid sulṭān. As attested by the Ghaznawid historian Abu ’l-Faḍl Muḥammad ibn Ḥusayn Bayhaqī (d. 470/1077), the mace was the favourite weapon of sulṭān Masʿūd ibn Maḥmūd of Ghazna (r. 421/1030– 432/1040).20 The mace even appears on the decoration of silver-inlaid metalwork: among the planets depicted on the cover of the twelfthcentury Vaso Vescovali in the British Museum, London, is an eight-armed deity, identified as the planet Mars, holding different weapons among which is a dragon-headed club.21 The iconography of a dragon head holding a quadruped or a human being in its open jaws is found on several finials of ceremonial weapons or staffs. Dragons are often represented as fierce, devouring creatures. However, since the animal or human being appears in the dragon’s jaws it is not clear whether the act is one of ingestion or expulsion. The heads are portrayed as character-

istic “Saljuq”-type dragon heads22 which are typically represented as ophidian, the elongated lips (sometimes only the upper lip) curved upwards and rolled outwards revealing a proportionately deep cavity with large fangs. An important example of such a ceremonial copper alloy mace terminating in a “ Saljuq-type” dragon head with characteristic gaping jaw revealing teeth and tongue, perhaps made in Afghanistan and datable between the eleventh to the thirteenth centuries, is housed in the David Collection in Copenhagen (fig.╯20). Finials of such dragon staffs are preserved in several collections worldwide. An example in the Furusiyya Art Collection in Vaduz which is thought to come from twelfthor thirteenth-century Anatolia is fashioned with large almond-shaped eyes, small, rounded ears, and curved horns, and the back of the neck is embellished with elongated drop-shaped cartouches enclosing spiralling foliage. The particular feature of the mace head is that the dragon’s maw holds the body of a quadruped, possibly a feline (fig.╯21). The wide-open jaws of another twelfthor thirteenth-century copper alloy dragon-headed finial with curved horns, small pointed ears and with a small loop for attachment at the back of the ophidian neck, is filled, in a corresponding manner, with the seated figure of a human being (figs. 22a and b).23 The finial may have topped a ceremonial staff, thought to be either from Iran or the Jazīra region, and is now in the Musée du Louvre in Paris. The dragon head’s very long jaws with drawn-up lips ending in curved tips have paired fangs that frame the human figure at top and bottom. The stylised figure is rendered with a rounded mask-like head. Another closely related twelfth- or thirteenth-century finial is in the David Collection in Copenhagen (fig.╯23). Thought to come from Iran or Afghanistan, or perhaps Anatolia, it is cast in the form of a dragon

╇ Dhorme, 1949, pp.╯40–3, 51. ╇ For further related examples of chthonic deities with the serpent as attribute, see Winkelmann, 2004, pp.╯252–9, figs. 103–6. 18 ╇ Cited after Steingass, 1892, repr. 1981, p.╯1082. 19 ╇ Herzfeld, 1927, vol.╯2, pl.╯LXVI. Cf. Furūsiyya, 1996, vol.╯2, p.╯97. 20 ╇ Abu ’l-Faḍl Bayhaqī, Taʾrīkh-i Masʿūdī, ed. Ghanī and Fayyūḍ, Tehran, 1324/1945, as cited in Bosworth, “Lashkar-i Bāzār,” EI² V, 690b; idem, 1963, p.╯120. 21 ╇ Hartner, 1973–4, pp.╯119–20 and fig.╯17, no. 5. 22 ╇ Cf. L’Islam dans les collections nationales, 1977, p.╯102, cat. no. 161. In the scene of Mūsā, Aaron (Hārūn ibn ʿImrān)

and the Israelites watching the Egyptians drown in the Red Sea, depicted in the illustrated copy of the Jāmiʿ al-tawārīkh (“Compendium of Chronicles”) by Rashīd al-Dīn Ṭabīb, made in Tabriz in 714/1314, Mūsā holds an elongated staff which ends in a closely related open-jawed dragon head. Talbot-Rice, 1957, p.╯61, cat. no. 11. 23 ╇ A further eleventh- or twelfth-century silver- and copper-inlaid copper alloy dragon-headed finial, probably a sword pommel, thought to come from greater Khurasan or the Jazīra/Anatolia, shows the monster holding a stylised human being in its wide-open maw. Vaduz, Furusiyya Art Collection, inv. no. RB-94. L’art des chevaliers, 2007, p.╯107, cat. no. 71 (only depicting the profile).

16 17

38

chapter three

head holding in its wide-open mouth a bovid, whose protome peers out of the dragon’s jaws.24 Interesting in this regard is the use in medieval Iranian poetry of the metaphor “caught in the dragon’s maw” (akin to “held in the dragon’s claws”) which is conventionally used to reflect a potentially fatal calamity. In the pre-Islamic epic, Wīs u Rāmīn (“Wīs and Rāmīn”), translated (from Pahlawī into classical Persian) and versified by Fakhr al-Dīn Asʿad Gurgānī around 442/1050, the protagonist Rāmīn uses it, for instance, to describe his separation from his beloved Wīs, the daughter of the queen of Media who is the wife of his older brother, king Mūbad of Marw: ...I have left my hostage heart with you... It is as if, upon your soul I swear, I’m in a dragon’s jaws when you are not there.25

In the Shāh-nāma, in which the ancient history of Iran, from its legendary origins down to the extinction of the Sasanian dynasty in 652 was recorded, this metaphor is used to describe a political misfortune, such as the defeat of the Iranians by the Türkmen: The world thou wouldst have said, “is in the dragon’s maw, Or Heaven level with earth.26

24 ╇ The iconography of the dragon devouring or disgorging a human being or an animal such as a felid or a bovid is also known in the Christian iconography of the Caucasus. A dragon with quadruped forelegs and a looped tail, portrayed in profile and depicted in the process of swallowing or delivering a proportionally small human figure, is shown above the southern entrance of the mid-tenth-century Georgian church of Beris-Sakdari, near the village of Eredwi in the Patara Liakhvi Gorge. The depiction is probably related to the story in the book of Jonah in the Old Testament of a sea-monster or ketos who devoured and cast up the hero under divine command. For the ketos, translated vishap, which swallowed Jonas in later Christian Armenian art, see Russell, 2004, p.╯373. The imagery of the dwellingplace of sinners “in the midst of the jaws of the dragon of the outer darkness” also repeatedly appears in the GnosticChristian writings of the Pistis Sophia (c. fourth century) which are further discussed in the following chapters; Pistis Sophia, text ed. Schmidt and tr. Macdermot, 1978, bk. III, ch. 108, p.╯551, ch. 119, p.╯609, ch. 121, p.╯617. The fact that the imagery was depicted above the entrance to the church also indicates that its iconography was associated with the warding off of evil and the affording of protection. Referring to the work of Vladimir Propp, Boris Marshak has pointed out the archaism of the theme of the hero being devoured by a monster (Propp, V.Y., Istoricheskie korni volshebnoi skazki (“The Morphology of a Fairy Tale”), Leningrad, 1946, pp.╯200–23, as cited in Marshak, 2002, p.╯49, n. 39). According to Propp (1984a, pp.╯116–8, 207, 208, and idem, 1984b, p.╯96) the imitation of devouring and expectorating of a hero by an animal such as a dragon was sometimes part of

Or, the hopelessness of a political situation, as expressed by one of the last Sasanian rulers, Â�Khusraw II (Khusraw Parwīz, r. 591–628): But what can this avail now that my head is in the dragon’s maw?27

In Islamic culture dreams were considered an important means of communication with the world of the unknown. Their meaning would be explained, often as a prophetic message from the world of the unseen.28 Hence, it is interesting to consider a dream interpretation recorded in al-Damīrī’s fourteenth-century para-zoological encyclopaedia which reverses the generally negative associations of the ophidian-devouring process. He states that: He who dreams of a serpent swallowing him, will obtain power.29

In the heroic epic Garshāsp-nāma composed by Asadī Ṭūsī in 456–8/1064–6, the eponymous hero, who is the great-great uncle of the legendary warrior Rustam, is requested by Ẓaḥḥāk at the tender age of fourteen to slay a dragon that dwells on Mount Shekāwand having emerged from the sea following a storm.30 The hero accomplishes the feat by clubbing the beast to death with a cudgel carved in the form of a dragon head.31 Employan initiation process, the successful completion of this ritual entitling the initiate to start a new phase of life or existence. An analogous relief probably representing Jonah in the maw of a mythical creature, a “whale” with quadruped forelegs (only the protome is featured) is shown above the south door (east side) of the Georgian Tao-Klardjeti monastery church of Haho (Georgian Khakhuli), modern Bağlar Başi in northeastern Turkey, datable between the tenth and eleventh centuries (Baltrušaitis, 1929, pl.╯LXXI, fig.╯118; Winfield, 1968, pp.╯62–3 (line drawing, fig.╯6), pl.╯30b). However, whereas on the relief of the Georgian church of Beris-Sakdari, the human figure is depicted with its head in the dragon’s maw (as if being swallowed), the Haho relief shows the figure’s upper body and head topped by small fish projecting from the beast’s jaws (as if being spat out). 25 ╇ Tr. and ed. Davis, 2008, p.╯388; the same metaphor is employed on pp.╯143, 166 and 230. 26 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯6, p.╯538. 27 ╇ Idem, vol.╯6, p.╯88. For further examples cited in the Shāh-nāma, see, for instance, idem, vol.╯3, p.╯171, l. 377, p.╯469, l. 670; vol.╯4, p.╯13, l. 95; vol.╯5, p.╯13, l. 95; vol.╯6, p.╯233, l. 876. 28 ╇ Marzolph and van Leeuwen, 2004, p.╯542. 29 ╇ Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯655–6. 30 ╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr; Yamamoto, 2003, p.╯115. 31 ╇ Asadī Ṭūsī, Garshāsp-nāma, p.╯269, l. 10, cited after Khāleqī-Moṭlaq, “Aždahā II,” EIr. In the Shāh-nāma, the hero employs serpent-like sword (tr. and ed. Mohl, 1838– 1878, vol.╯5, p.╯341, l. 728).

the dragon motif on portable objects

39

ing the principle of sympathetic magic – “like affects like” – Garshāsp succeeds in killing the dragon by means of a weapon carved with its own likeness. This magical power appears to be contagious and can be transmitted from its source, in other words from the dragon onto different kinds of implements such as the mace32 or a staff, as evidenced, for instance, by a story recorded by al-Kisāʾī. Here Mūsā similarly employes a mimetic or “homeopathic” principle by using his serpentstaff to strike a giant serpent that has devoured all the sheep of the Prophet Shuʿayb’s flock that pass through an exceptionally fertile valley thereby cutting it in two.33 Literary accounts of the medieval Islamic period describe the bejewelled weapons that were paraded on ceremonial occasions.34 Sabres were probably introduced into the central Islamic lands by the Turkic guard of the ʿAbbasid caliph al-Muʿtaṣim (218/833–227/842), which can be seen on a representation of a ninth-century wall painting in a building at Nīshāpūr that depicts a horseman with a belt with hanging straps designed to support a sabre.35 Quillon blocks extending into downward-curving prongs that terminate in dragon heads are a common feature on twelfth- or thirteenth-century sabres or daggers, as for instance on a gilded copper alloy sword guard fragment, or a nielloed silver scabbard, both thought to be from Syria or Palestine and now preserved in the Furusiyya Art Collection, Vaduz (fig.╯111).36 Literary sources such as the Shāh-nāma similarly use the image of the (serpent-)dragon to describe swords, as for instance, those belonging to the dragon-fighting Kayanian king Gushtāsp,37 to Iskandar38 or to Fūr (Porus), king of India, who stopped Alexander’s advance in India.39 The Turks attributed magical properties to jade (nephrite) and called it the “stone of victory.” They used it extensively for fittings of weapons such as handles and quillon blocks, as well as for objects of adornment such as belt fittings and

rings, in the belief that its presence would aid in attaining victory over their opponents.40 Very informative in this regard is the entry under the heading yashf (“jade”) of al-Bīrūnī’s pharmacological work, the Kitāb al-Ṣaydala fi ’l-Ṭibb (composed in 442/1050), which states that dragon iconography was engraved on jade and used by the Turks to adorn swords:

╇ Jeffers, 1996, p.╯95. ╇ Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯223. 34 ╇ Cf. the description of a celebration at the Ghaznawid sulṭān Masʿūd’s court in 429/1038 by Abu ’l-Faḍl Bayhaqī, (Taʾrīkh-i Masʿūdī, ed. Ghanī and Fayyūḍ, Tehran, 1324/1945, pp.╯539–41). Cf. Bosworth, 1963, pp.╯135–7. 35 ╇ The wall painting is now preserved in the National Museum of Iran, Tehran. Hakimov, 2000, p. 445, fig. 30. 36 ╇ L’art des chevaliers, 2007, p.╯154, cat. no. 147 and pp.╯155–7, cat. no. 148; Chevaux et cavaliers arabes, 2002, pp.╯118–9, cat. no. 57. An Iranian jade (nephrite) quillon block from the hilt of a sword, dated to the first half of

the fourteenth century, is preserved in the Metropolitan Museum of Art, inv. no. 64.133.3. See Melikian-Chirvani, 1997a, p.╯159, fig.╯27. 37 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯4, p.╯341, l. 728. 38 ╇ Idem, vol.╯4, pp.╯153, l. 628. 39 ╇ Idem, vol.╯5, pp.╯151–5. 40 ╇ Melikian-Chirvani, 1997a, pp.╯131–3. 41 ╇ Ed. ‘Abbās Zariyāb, Tehran, 1370/1991, p.╯203, as cited in Melikian-Chirvani, 1997a, p.╯131. 42 ╇ Tr. Meisami, 1993, p.╯91. 43 ╇ Ed. Nafīsī, S., Tehran, 1960, p.╯53, cited after Daneshvari, 1993, pp.╯16–7, n. 7.

32 33

The yashf stone: this is yashb on which they engrave the radiate dragon [al-shuʿāʿ is translated as “ray of light” in Steingass]. We tested it without the engraving and it delivered [a result]. Its characteristic, they say, is to dispel stomach pains. “The stone of victory” is a variety of it and that is why the Turks adorn their swords with it.41

The affiliation of the dragon with arms was also made by the great twelfth-century poet Ilyās ibn Yūsuf Niẓāmī Ganjawī (535–40/1141–6– 575–613/1180–217) in his romance Haft Paykar (“Seven Portraits”), when he compared to dragons the blades of the idealised fifth-century Sasanian king Bahrām Gūr’s (Wāhram V, r. 420–38) army, and its arrows to the serpents of Ẓaḥḥāk, the tyrannical foreign ruler of Iran in the Shāhnāma from whose shoulders sprouted the notorious serpents.42 Similarly, the panegyrist and epistolographer Rashīd-i Waṭwāṭ (508–9/1114– 5–573/1177–8 or 578/1182–3), who was born in Balkh, and spent most of his life as poet at the court of Khwārazm-shāh Atsїz, writes in his dīwān: Thousands of lion-hearted and elephant-bodied warriors Are eaten by dogs [after being slain] by your dragon-lance.43

A thirteenth-century silver chape is completely covered with the depiction of a pair of upright dragons that are addorsed along a central vertical ridge. The chape was found together with its long knife in Herat, in present-day Afghanistan, and is now in the al-Sabāh Collection in Kuwait

40

chapter three

(fig.╯24). The so-called “Saljuq-style” heads of the fabulous beasts are turned towards the back thus confronting each other with their wide-open jaws terminating in upward curled tips and revealing the tongues, the long cusped ears projecting at the top.╯The quadruped protomes with snugly-fitted feet are oriented towards the edge of the chape, their serpentine bodies with slender, curved wings forming five loops attenuating towards its tip.╯It is interesting to note a closely related depiction found on a twelfth- or thirteenth-century silk fragment from Samangan province in presentday Afghanistan, also housed in the al-Sabāh Collection in Kuwait, which is woven with an upright pair of dragons with gaping jaws, here in confronted position and wearing beaded collars, their winding serpent-like bodies having clusters of three dots in the bends and set against a background of foliate scrolls (fig.╯25). The dragon motif on banners In the Middle Iranian period, a dragon (azhdahā) was often depicted on standards carried in battle as a symbol of martial valour, intended to frighten the enemy by its ferocious aspect and to show the ruler’s power.44 These banners are referred to several times in the Shāh-nāma, in which the ancient history of Iran, from its legendary origins down to the dissolution of the Sasanian dynasty was recorded, as azhdahā-paykar (“having a dragon’s body”).45 In his De historia conscribenda sit (XXIX) the second-century Greek writer Lucian describes the war against the Parthian king Vologesus III who defeated the Romans at Elegia in 162, destroying the Roman legion and killing the commander Severianus. Lucian notes that the Parthians used banners with different emblems to differentiate the divisions of their army, a dragon ensign (dracōn) preceding a thousand-man division. He refers to another historian’s vivid report of these dragon ensigns, which were made of light material, attached to open-jawed heads and mounted on poles, so that they would move in the breeze like enormous serpent-dragons, so much so that they appeared to the Romans to be: 44 ╇ It is of note that the Indian makara, discussed above (p. 36, n. 12), also formed the head of the battle standard (makaradhvaja) of Rudra or Shiva, and was later also carried by Siddhārtha’s son, Rāhula. It served as battle standard for the hosts of Mara, who attacked the Buddha. The makara also served as designation for specific battle formation of troops that take the form of a makara during battle manoeuvres. Cited after Beer, 2004, p.╯68. 45 ╇ Skjærvø, “Aždahā I,” EIr.

...alive and of enormous size; that they were born in Persia a little way beyond Iberia; that they are bound to long poles, and raised on high, create terror while the Parthians are coming from a distance; that in the encounter itself at close quarters they are freed and sent against the enemy; that in fact they had swallowed many of our men in this way and coiled themselves around others and suffocated and crushed them.46

Later on the emblem was introduced into the Roman army where the standard-bearer bore the title draconarius. Such a banner with a dragon ensign belonging to the Dacians and their Sarmatian allies is portrayed on the narrative reliefs of Trajan’s column which commemorates the Dacian wars of the early second century ad. The fluttering dragon ensigns were described by the contemporary Greek historian Flavius Arrian in his Tactica (XXXV 2–5) as being of “Scythian” origin: Scythian ensigns are serpents of good length, tied to staffs. They are made out of pieces of dyed material. The heads and the bodies through to the tail are made in order to appear as terrifying as possible … They swell in the wind of a ride so that they look like those serpents and even begin to whistle when the breath of air is strong enough.47

The dragon banner in the Roman army appears to have been instituted in the wake of the deployment of new tactical divisions as part of the Roman emperor Diocletian’s (r. 284–305) reorganisation of the military machinery.48 Such dragon banners are again described by the fourth-century Roman historian Ammianus Marcellinus, who served in the army of Constantine II (337–361) in Persia, in his Rerum gestarum libri (XVI 10.7 and 12.39). In his record of Constantine’s entry into Rome in 357 (XVI 10.7) Marcellinus writes that the emporer: ...was surrounded by dragons, woven out of purple thread and bound to the golden and jewelled tips of spears, with wide mouths open to the breeze and hence hissing as if roused by anger, and leaving their tails winding in the wind.49 46 ╇ Lucian, Quomodo historia conscribenda sit, tr. Kilburn, K., London, 1959, vol.╯VI, pp.╯42–3. Cf. Widengren, 1969, pp.╯17–8 and n. 38; Shahbazi, “Derafš,” EIr. 47 ╇ Arrian, Tactica (XXXV), cited after Lebedynsky, 1995, pp.╯93–4. 48 ╇ Haussig, 1992, p.╯29. 49 ╇ Cf. Widengren, 1969, p.╯18, n. 39 (with further references).

the dragon motif on portable objects According to Moses of Chorene, in commemoration of the Armenian king Tigran’s resettlement of the defeated Medes to the area of Goghtʿn and around the foot of Mount Ararat, the windsock-like silk dragon (vishap) banner, adopted from the Parthians, became the Armenian king’s heraldic sign.50 In his account of the battle of Jiraw the sight of the banners (III.37) is similarly vividly conveyed:

41

showing that different royal emblems could be used concurrently.53 A dragon banner may be depicted on a bone plaque which is incised with a battle scene and probably served as a belt element.54 It was found in a necropolis near the village of Orlat in the district of Koshrabad, west of Samarqand, and, according to Boris Marshak, does not reflect a local but a Central Asian nomadic tradition.55 The mail-clad combatants are wielding lances, bows, swords and battle axes.56 Strapped to the lancer on the lower left is a military emblem, which may have served as badge of rank, a long, flowing banner on a pole

akin to the dragon banners represented on wall paintings in Chinese Turkestan (present-day Xinjiang), discussed below. Marcus Mode tentatively ascribes the plaque to the reign of the Xiongnu king of Sogdiana (Su-te) in the 30s of the fifth century attested by The History of the Wei (Weishu).57 The Xiongnu were known to have fought under the dragon banner, a traditional ensign of the military forces of the steppes.58 Similar banners, some rendered in serpent or dragon-like form, are depicted on wall paintings in Chinese Turkestan, a region whose pivotal position at the crossroads between China and Central Asia resulted in a broad cultural synthesis which embraced the western Turks or Kök Türks (T’uchueh or Tujue). In the Cave of the Painter at Kizil (Kezier), a sixth- to seventh-century wall painting illustrates an army with a dragon banner.59 During the German Turfan expeditions at the beginning of the twentieth century, Albert von le Coq recorded several such banners. At Kizil, in the “Cave of the Dove,” a seventh- or eighthcentury wall painting features a senāpatiratna, a deity who symbolises the wrathful power to overcome enemies, carrying a banner in the form of a serpentine body terminating in a swallow-tailed pennant with projecting dragon head with gaping jaws,60 and in the “Caves with Fireplace,” a warrior riding a war elephant also holds such a dragon banner.61 A wall painting discovered in the eighthcentury “Cave of the Doves” at Kirish, the ruins of Simsim, portrays a dragon-king (nāgarāja) above whom floats a large banner with lupine head and what appears to be a serpentine body.62 The wolf was one of the prevalent emblems on the military standards of Turkish tribes; other such emblems seem to have been the moon or the dragon.63

50 ╇ Russell, 2004, pp.╯621–40, esp. 624 and 622–23 (for a discussion of the historical Tigrans that could have inÂ�spired the orally transmitted legend); also pp. 1047–8 (with a translation of the late twelfth-century Armenian Catholicos Nersēs Šnorhali’s text Interpretive Explanation, Mingled with Supplication, of the Standards of the Kings of Armenia which recalls the king and his banner). 51 ╇ Idem, p.╯627. 52 ╇ Idem, pp.╯624–5 and n.╯9. It is interesting to observe that such dragon standards are still represented in thirteenthcentury Armenian miniatures (Washington, DC, Freer Gallery of Art, Ms. 32.18, p.╯513; “Judas leading the Multitude,” see Der Nersessian and Agemian, 1993, vol.╯2, fig.╯251). 53 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯589, ll. 2107–2108. 54 ╇ Mode, 2006, p.╯420 and n. 4. 55 ╇ Marshak, B.I., “Iskusstvo Sogda,” Tsentral’naya Aziya. Novye pamyatniki pis’mennosti I iskusstva. Sbornik statey, eds., Piotrovskiy, B.B. and Bongard-Levin, G.M., Moscow, 1987, p.╯235, as cited in Mode, 2006, pp.╯421–2. Cf. Brentjes, 1989, p.╯41. 56 ╇ Bone plaque with battle scene. Orlat, Koshrabad district, west of Samarqand, present-day Uzbekistan. Ilyasov

and Rusanov, 1997–8, pl.╯IV:1; Mode, 2006, p.╯444, fig.╯1. Cf. Brentjes, 1989, p.╯40, fig.╯3. 57 ╇ Mode, 2006, p.╯433. For a comprehensive list of suggested dates of the plaques, cf. esp.╯pp.╯421–2, 433. 58 ╇ Idem, p.╯433. 59 ╇ G. 356; von Le Coq, 1925, p.╯54, fig.╯50. 60 ╇ G. 274; idem, p.╯72, fig.╯117. 61 ╇ G. 50; idem, p.╯55, fig.╯53. 62 ╇ G. 432; idem, p.╯68, fig.╯101. The lupine figures recall the fact that the T‘u-chüeh tribe, which was part of the Â�Xiongnu confederacy, is said to have depicted a wolf on their banners; Eberhard, 1979, pp.╯52–3. 63 ╇ Roux, 1979, p.╯170. The seventh-century Chinese dynastic annals, the Pei shih, as well as the Chou shu (50, 4a), explicitly state that the Turks put wolf heads on their standards; Sinor, 1982, p.╯233 and 1996, repr. 1999, p.╯329; Liu, 1958, p.╯9. It is interesting to observe that wolf-standards were sometimes referred to as dragon-standards; see Schmidt, 1980, p.╯63. The dragon, the wolf and the moon were also emblems on the respective flags of Rustam, Gurgīn and Farīburz (Shāh-nāma; tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯66).

the sinuous rippling of the dragon [banners], puffed up by the blast of air, their jaws yawning frightfully.51

An account of the anonymous fifth-century Armenian Epic History or Buzandaran Patmutʿiwnkʿ (IV.2) also describes the battle standard of the Arsacid Armenians (54–428) as a silk dragon banner.52 A description of the advance of the Iranian army is given in the grand heroic epic, Firdawsī’s Shāh-nāma, which states that: behind each banner, there followed another banner – some with dragons, others with the image of eagles

42

chapter three of the celebrated feat carried out by Farāmarz, son of Shahrbānū Irem and Rustam, in slaying a dragon called hissing serpent (mār-i juwshā).73 An extended list of heroes thus seems to have claimed the right to own a dragon banner as emblem. The visual appropriation of the dragon’s likeness on individual and dynastic banners not only communicated mastery over the mythical creature but also implied that through victory the vanquisher had been able to appropriate the formidable qualities of the dragon.

In Asadī Ṭūsī’s Garshāsp-nāma the eponymous hero’s victory over the dragon was commemorated by a flag (dirafsh) embellished with the representation of a black dragon (azhdahāy-i siyāh) and a pole surmounted by a golden lion, in turn topped by a moon.64 The banner was passed on to Garshāsp’s descendants and became his family’s coat of arms.65 Likewise, the standard of the paladin Rustam, whose ancestors are thought to have been Saka people who came to Sakastān/ Sīstān and Zābulistān in the late second century bc,66 was “blazoned with the dragon’s form and from its tip a golden lion roars.”67 His grandfather, Sām, also had a dragon banner,68 as did Rustam’s son, Farāmarz.69 The Shāh-nāma account records that when Bahrām Chōbīn received the supreme command, the Sasanian king Hurmuz IV (578– 590) handed him the purple dragon banner with the words: “You are indeed a second Rustam.”70 Not only did Bahrām Chōbīn receive the distinguishing emblem which likened him to the great hero, but, in addition, he claimed Arsacid Parthian descent and was thus heir to the traditional Arsacid dragon banner.71 Dragon banners bearing the effigy of a gold-coloured dragon without legs or wings, the open mouth revealing the tongue, set against red and black grounds alternately, are featured on a page from a dispersed Shāhnāma. Painted in Shiraz in 742/1341, the miniature, which was formerly in the collection of the prince and the princess Sadruddin Aga Khan, now in the Aga Khan Foundation, portrays the victory of the dragon-slayer Bīzhan, grandson of the Iranian commander, Gūdarz, over Hūmān, brother of the Tūrānian commander, Pīrān;72 the banners being displayed in commemoration

Girding with a belt was a rite of passage in all Central Asian, in particular Iranian, societies from ancient times.74 Adorned parade belts played an important role in the investiture of warriors in the Indo-Iranian world and were insignia of cultic or ritual significance as well as symbols of social distinction.75 In the Armeno-Parthian dynasty the belt was presented to the dihqāns (“members of the lesser feudal nobility”) along with a ring and a banner as tokens of royal service.76 The origin of the custom has been attributed to the sacred rope-girdle (kustī) of the Zoroastrians, for whom, however, such girding was an act of consecration, as in the rite of initiation. In Turkish and Islamic society the belt was of great importance; it was an integral part of male costume, often presented by the ruler, and hence regarded as an insignia of rank.77 The nomadic warrior seems to have been girt with a pair of belts78 adorned in accordance with

64 ╇ Widengren, 1969, p.╯17, n. 35; Asadī Ṭūsī, Garshāspnāma, pp.╯49–63 (63.35–6), as cited in Khāleqī-Moṭlaq, “Aždahā II,” EIr. A dragon-shaped long narrow flag floating from the mast of a boat is featured on a ninth-century lusterware plate from Nīshāpūr. Papadopoulo, 1979, fig. 420. 65 ╇ Op. cit. 66 ╇ De Bruijn, “Rustam,” EI2 VIII, 636b. Cf. the discussion in Shahbazi, 1993, pp.╯157–8. 67 ╇ Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯107, ll. 1224–1255, p.╯113, ll. 1292–1293; vol.╯5, pp.╯85, 89; The Tragedy of Sohráb and Rostám, tr. and ed. Clinton, 1987, pp.╯95, 549–50. Cf. Shahbazi, 1993, p.╯159; MelikianChirvani, 1998, p.╯179. 68 ╇ Widengren, 1969, p.╯17, n. 35, with reference to the Garshāsp-nāma, as also cited by Shahbazi, 1993, p.╯159. 69 ╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. 70 ╇ Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯6, pp.╯587–9, ll. 525–31. It is of note that like the dragon banners surrounding Constantine when he entered Rome, as described by Ammianus Marcellinus (see above), the Shāhnāma account specifies that the colour of Bahrām Chōbīn’s dragon banner was – just like that of Rustam – purple; tr. and ed. Mohl, 1838–1878, vol.╯6, pp.╯587–9, ll. 525–31. 71 ╇ Shahbazi, “Derafš,” EIr.

72 ╇ “Victory of Bīzhan over Human.” Page of a dispersed Shāh-nāma. Iran, Shiraz. 1341. Height 36.9 cm, width 30.7 cm. Collection of the Prince and the Princess Sadruddin Aga Khan, Ms. 006/E. Chevaux et cavaliers arabes, 2002, p.╯161, cat. no. 119. A further example of a dragon banner is Â�illustrated in an Ilkhanid-period Shāh-nāma manuscript; Pope and Ackerman, eds., 1938–9, repr. 1964–81, vol. 10, pl. 834. 73 ╇ Farāmarz-nāma, London, British Museum, Ms. Or. 2946, fols. 24, 25, as cited in Khāleqī-Moṭlaq, “Aždahā II,” EIr. 74 ╇ It is of note that the belt did not play an important role in Greek tradition. Cf. Brentjes, 1989, p.╯42. 75 ╇ Cf. Widengren, 1969, pp.╯21–32. 76 ╇ Brentjes, 1989, p.╯43. The Parthian king Uthal is depicted with a belt whose central belt plaques are rendered with winged and apparently bearded dragons with coiled serpentine tails resting on their forelegs. See Winkelmann, 2004, pp.╯10–3, fig.╯5b. 77 ╇ Cf. the discussion of Katharina Otto-Dorn, 1961–2, pp.╯9–13. 78 ╇ As evidenced for instance by the first-century ad finds in the Sarmatian kurgan near Porogi, nomadic warriors customarily wore two belts. Simonenko, 1991, p.╯215, fig.╯1.

The dragon motif on belt/strap fittings and equestrian accoutrements

the dragon motif on portable objects

43

rank and ancestry.79 Hence on one belt a diagonally suspended bow in a long narrow bow-case and a quiver filled with arrows could be carried, while a sword or sabre and a dagger might be suspended from the second.80 Paired belts, a ceremonial main belt and another with straps from which the sword and other weapons were suspended, were, according to the Tarikh-i Bukhārā, worn by the young attendants (bandagān) at the court of the queen of Bukhara.81 Reports on the Turks in their homeland emphasise the importance of belts, which in addition to their utilitarian function served as status symbols. The Turks are described as rich in cattle, horses and sheep and possessing “many vessels of gold and silver. They have many weapons. They have silver belts…”82 The Avars, an Ibero-Caucasian people who were subjects of the Turks, followed in the wake of the great migrations of nomadic peoples from the Eurasian steppes, fleeing to the north Caucasus region in 558 and from thence migrating westwards.83 Late Avar period iconography is preserved almost exclusively on items of personal adornment, in particular compartmentalised belt sets with multiple plaques figuring more naturalistic tamgas in the form of fabulous creatures reflecting “ancestral, totemic ideas.”84 Often these are shown in combat with ungulates or human beings perhaps symbolising “the cycle of death and rebirth.” Among the animals another hybrid creature with a long history in the Near East, the griffin, appears to have been particularly prevalent, however the dragon also made an appearance as did natural animals such as horses, eagles and wild boars.85 An elongated silver tongue-strap fitting (“Hauptriemenzunge”) with one arched end featuring a quadruped dragon was unearthed from grave 292 of an Avar necropolis in Abony,

Hungary, attributed to the second half of the seventh century (fig.╯26).86 The beast is rendered in profile with elongated open snout, the upper lip curving upwards revealing a pointed tongue, crowned by horns or ears, with a beard projecting from the chin; the elongated, undulant, serpentine body covered with a spotted pattern is demarcated by crest-like, spiky protrusions, and rests on three curved legs with pointed protrusions at the feet, probably representing unsheathed claws. Depicted at mid-section of the body is what appears to be a small version of the dragon. An association of this representation with the Chinese dragon through Byzantine mediation has been proposed by Falko Daim on the grounds that the depiction of the dragon is comparatively rare in Avar iconography.87 Samuel Szádeczky-Kardoss however relates this iconography to Hellenistic (that is Seleucid and Parthian) and Sasanian influence, while underlining that “the subject matter [of the Avars] is taken from the body of beliefs proper to the peoples of the steppe.”88 Belts also had an important ceremonial significance and were a symbol of authority for the Samanids, Ghaznawids and Saljuqs.89 Tenth- or eleventh-century Western Central Asian belt hooks were commonly S-shaped with a central cuboctahedral or spherical knob, terminated at either end by a horned dragon head, one end bent to a closed position and with a heavy rectangular strap-slot.90 The dragon heads are shown with the wide-open snouts characteristic of the “Saljuq” type, the upper lip curving upwards and revealing a deep cavity with stumpy fangs. A roughly contemporary horned dragon with open snout and sinuous body also occurs on narrow essentially rectangular copper alloy belt strap fittings from present-day Afghanistan, some of which are gilded.91

╇ Szádeczky-Kardoss, 1990, p.╯126. ╇ Cf. the belts of the sandstone statue of the Śaka satrap Caṣṭana (first half of the first century). Czuma, 1985, pp.╯112–3, cat. no. 43; Azarpay, 1981, pp.╯122–5. 81 ╇ Schefer, 1892, pp.╯7–12. 82 ╇ Gardīzī, Zayn al-akhbār, in V.V. Bartol’d (W. Barthold), Otč o poezdke v Srednyuyu Aziyu s naučnoy tsel’yu 1893–1894 gg., in Zapiski Imperatorskoy Akademii Nauk, ser. VII, t. i, 74–175. Pers. text and Russ. tr. repr. in Sočineniya, Moscow 1963–73, vol.╯8, p.╯35, cited after Golden, “Pečenengs,” EI² VIII, 289a. 83 ╇ Pohl, 1988, pp.╯28–9; Barthold and Golden, “Khazar,” EI2 IV, 1172a. 84 ╇ Szádeczky-Kardoss, 1990, pp.╯126–8. 85 ╇ Pohl, 1988, p.╯289. 86 ╇ Cf. Xi’an, 2006, p.╯356, cat. no. 271. 87 ╇ Daim, 2000, pp.╯134–6, fig.╯60; Xi’an, 2006, p.╯356,

cat. no. 271 (catalogue entry by Falko Daim). 88 ╇ Szádeczky-Kardoss, 1990, p.╯128. It is interesting to observe that a plaque in the form of a reptilian mythical creature with large gaping snout bearing the images of fish on its belly, datable between the sixth and the eighth century, can be detected among the animal copper alloys from the governorate of Perm, north of the Caspian; see Oborin and Tshagin, 1988, p.╯93, cat. no. 69. 89 ╇ Bosworth, 1963, p.╯38. 90 ╇ Several of these examples are preserved in Kuwait, al-Sabāh Collection, Kuwait National Museum, such as inv. nos. LNS 617 J, LNS 1122 J, LNS 1123 J, LNS 1125 J b, LNS 2762 J. 91 ╇ See an example in Kuwait, al-Sabāh Collection, Kuwait National Museum, inv. no. LNS 3103 J; putative origin: Herat, Afghanistan.

79 80

chapter three

44

An unusual depiction of a standing quadruped dragon appears on a richly gilded copper alloy fitting, probably for the strap of a parade belt, now in a private collection (fig.╯27). The head, turned backwards, has gaping jaws with fleshy folds, the upper lip terminating in a tight curl revealing rows of teeth, conspicuous projecting fangs and a long flickering tongue. Small, circular eyes punctuate the head, which is crowned by small, cusped ears and extends into the long sinuous neck surmounted by a crest defined by deep hatching. A split palmette projects from the haunches at the front and another extends from the tip of the long sinuous tail which curves under the flank to ascend vertically above the back. The legs with delineated four-clawed feet are firmly planted on two platforms that project to form the base of the fitting, adding to the imposing plastic effect of the creature. In spite of the uncommon appearance, its powerful and expressive monumentality must be the result of a long-established tradition. Remarkably, distant cousins of the enigmatic beast, albeit no dragons, appear to survive on the low-relief carvings that cover a large part of the exterior of a cathedral of Saint Dmitry (1193–1197) in Vladimir in the northeastern principality Vladimir-Suzdal of Rus’ situated around the river Volga. Named after the warrior saint Demetrius of Salonika, the cathedral was commissioned by prince Vsevolod III (1176–1212) for his own use. Contacts with the east on the part of this principality, by way of the Volga, were much closer than with western Europe and it is generally thought that most of the characteristic local artistic features are due to Caucasian influences.92 Vladimir-Suzdal had close contacts with Armenia and Georgia, exemplified by the fact that prince Yuri Bogolyubsky, Vsevolod’s nephew, was chosen as husband for the famed Georgian queen Tʿamar (1184–1211/2) in 1184. On all three middle pediments of the cathedral the biblical king David is portrayed as surrounded by the animals, several of which are rendered with stylistic aspects reminiscent of the dragon on the fitting, such as the elongated necks, the long legs ending in forceful delineated feet and, in particular, the elongated tails that

curve under the haunches to project vertically above the back, ending in a split-palmette.93 The stylistic and iconographic stimuli for these relief sculptures are to be sought in post-Sasanian, Islamic and Transcaucasian art which may well have been well-known through textiles and other portable items.94 As a reflection of these eastern stimuli the late twelfth-century date of creation of the Vladimir relief sculptures thus provides an approximate terminus ante quem for dating the fitting with its forceful relief featuring a standing quadruped dragon which, in contrast to the more tame appearance of the beasts in the Vladimir reliefs, is characterised by an immediacy and innate animal nature, imbued with the vigour, physical power and lively spirit of wild beasts. The dragon motif also extends to horsemanship.╯Niẓām al-Mulk, the celebrated vizier of the Saljuqid sulṭān Alp Arslan and former Â�Ghaznawid functionary, used the figurative expression “Exalted Stirrup” (riqāb-i ʿālī) in his Siyāsat Nāma when referring to the sulṭān.95 The metonym was used at Turkish as well as Iranian courts to denote “the sovereign himself or his presence, the foot of the throne.”96 It is thus fitting that a pair of standing quadruped dragons are depicted on the shoulders of a cast copper alloy horse stirrup which is probably from the Ghaznawid world, datable to the eleventh or twelfth century, now in the Nasser D. Khalili Collection of Islamic Art, London (fig.╯28).

╇ Buxton, 1934, pp.╯24–5. ╇ Allenow et al., 1992, figs. 205 and 228. 94 ╇ Eidem, p.╯35. The style has also been compared with Transcaucasian relief sculptures on the seventh-century church of Ptghni (Ptghnavankʿ) in Ararat province and the church of the Holy Cross of Aghtʿamar on Lake Van built

between 915 and 921. See Buxton, 1934, pp.╯25–6. 95 ╇ Siyāsat Nāma, p.╯133, pp.╯98–9; cf. Silverstein, 2007, p.╯131. 96 ╇ Deny, “Riqāb,” EI² VIII, p.╯528b. The use of the term is also attested in Ibn Bībī, al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, pp.╯95–6.

92 93

c.╇ The dragon motif on objects of personal adornment In the Western Central Asian world the dragon also features prominently on objects pertaining to personal adornment. Quadruped Â�dragons appear on an eleventh- or twelfth-century soft stone mould of rectangular outline used for tooling leather, reportedly from Herat in northern Afghanistan and housed in the al-Sabāh Collection in Kuwait (figs.╯29a and b). Next to the patterning for two faces of a pouch featuring astrological and hunting scenes (as well as two small flaps), the mould is intaglio-carved with a pair of large, confronted quadruped dragons portrayed

the dragon motif on portable objects

45

with elongated slender bodies standing on tall, splayed legs, with one leg raised in a formal stance. Their snarling heads with long wide-open snouts reveal protruding sinuous tongues, while the top of the head has pricked ears and prominent elegantly curved horns. The creatures are characterised by very long, narrow wings that surmount the back, and thin, elongated S-shaped tails. The dragon also makes an appearance on a range of jewellery. A roughly trapezoidal twelfth-century grey steatite jewellery mould from Maimana province (Faryab) in Afghanistan, now in the al-Sabāh Collection in Kuwait, is deeply carved for a simple loop ear wire, terminating at one end with a dragon head and a vent.97 An eleventh- or twelfth-century copper alloy ring from northern Afghanistan, preserved in the same collection, has tapering shanks that terminate on either side at the apex in doublecollared, bulging-eyed dragon heads, their lower jaws joining at the tips.98 Intertwined dragon protomes, joined by a single loop, appear as appliqués on the shoulders of a twelfth-century niello-inlaid gold finger ring, also in the al-Sabāh Collection in Kuwait. The creatures are defined by confronted wideopen snouts with upward-curling tips that hold up a rectangular bevelled-edged agate seal stone which is inscribed in reverse Kufic script: “my sufficiency is in God and [it] suffices” (fig.╯30). These pious words were probably a supplication for help to overcome difficulties.99 The efficacy of the prayer was buttressed by the entwined dragons which in turn would have helped to enhance the ring’s protective properties. The motif of the entwined dragons is Â�further shown on an unglazed earthenware press mould of tapering rectangular form from Balkh in northern Afghanistan also in the al-Sabāh Â�Collection in Kuwait. The twelfth-century mould has a reÂ�markable pictorial programme. It shows a pair of addorsed regardant quadruped dragons in rampant posture with geometrically patterned elongated bodies. Importantly, the winged creatures entwine at mid-section. Their heads are crowned by long, curved horns and the open jaws revealing long tongues touch the scrolling foliage bearing round buds or fruit, possibly

pomegranates, against which the entire composition is set and which itself again terminates in gaping dragon heads. A small feline Â�couchant, probably a lion, is framed between the legs of the dragons, which are surmounted in their turn by addorsed upright quadrupeds alternating with a pair of confronted human figures seated, legs folded back, on a horizontal platform. Three stacked round objects that closely resemble the produce of the flowering branches (which are perhaps pomegranates, the fruit associated with the concept of fertility) appear between these figures (figs. 31a and b). To this may be added a very important and more complex composition of paired entwined dragons in mirror image, joined at the centre by a quadripartite knot. This composition, considered below in chapter 10, is featured on a matrix for belt ornaments which can be connected to the Ghurid sulṭān Ghiyāth al-Dīn Muḥammad ibn Sām (r. 569/1173–599/1202–3), the builder of the minaret of Jām, or his younger brother, Muʿizz al-Dīn Muḥammad ibn Sām (r. 569/1173– 602/1206), with whom he ruled in partnership and also shared the ism (name) Muḥammad and the nasab (a person’s relation to his forefather) ibn Sām (fig.╯169).100 In this context it is interesting to note that the motif of the interlaced dragons was minted as an added emblem on the coinage of Fakhr al-Dīn Qara Arslan ibn Dāwūd (539/1144–562/1167), the Artuqid ruler of Ḥiṣn Kayfā and Khartpert, which otherwise follows late Byzantine conventions (figs. 32a and b). The dragon protomes are represented addorsed but with necks twisted backwards so that the open-mouthed heads are confronted. Their raised wings touch at the centre in a circular tip while the forelegs extend down and forward to the medallion edge. The lower ophidian body entwines once and forms a quadripartite loop which then extends to frame the lobed medallion. The addition of the dragon motif countermark, superimposed on the lower left quadrant of a reproduction of a late Â�Byzantine coin type featuring an enthroned nimbate Christ holding a book,101 may be seen not only as an example of Türkmen efforts to create a multi-cultural

97 ╇ Length 8.53 cm, width 2.94 cm. Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. no. LNS 2619 J. 98 ╇ Height 2.59 cm, width 2.4 cm. Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. no. LNS 1596. 99 ╇ The formula is a well-known supplication in modern times.

100 ╇ From a set of 77 copper alloy matrices for belt/ strap fittings, Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, LNS 2558 J a-x2. 101 ╇ This type of motif is seen on coins of Manuel I Â�Komnenos (1143–1180). See Lane Poole, p. 123, no. 329.

chapter three

46

paradigm, embracing as it does both ByÂ�zanÂ�tine Greek and Türkmen visual traditions, but also as a step towards asserting the identity of this Artuqid ruler. The choice of this particular motif as emblem by a Türkmen leader is particularly important since it gives weight to the hypothesis that the interlaced dragon figure was introduced into Islamic art from Central Asia via the Turkish dynasties.102 The use of a symbol which must have carried Iranian cultural associations may thereby represent a conscious effort to revive a visual heritage from the past which at the same time served as a means of self-identification. Dragon imagery on Central Asian objects of personal adornment was also found at the frontier town of Ūtrār, located at the confluence of the Aryss and the Syr Daryā rivers. Excavations yielded a signet ring featuring a rider on horseback killing a dragon, an iconography that will be discussed in greater depth in chapter 7, together with bracelets terminating in dragon heads.103 The incident at Ūtrār where a Mongolian caravan was massacred by Khwārazmian officials led to the invasion of Transoxania by Genghis Khān’s troops in the autumn of 616/1219 and the city’s destruction shortly thereafter. It signalled the beginning of the Mongol conquest of Western Asia. Whereas the Central Asian world yields a wide range of objects of adornment decorated with the likeness of the dragon, comparable finds from Anatolia and the Jazīra from the eleventh to the thirteenth century are generally extremely rare. However, on the back of an early- to mid-thirteenth-century Saljuq gold-inlaid steel mirror, which would have been a prime accoutrement, now in the Topkapı Sarayı Müzesi, the dragon occurs three times with different connotations. The central field is decorated with a mounted falconer on a richly caparisoned horse with a hunting dog attached by a leash to the saddle. A small, looped dragon with raised, gaping head appears just in front of the horse’s hooves, a quadruped, probably a fox, seeks cover at the back, and a

game bird flies above. The benedictory inscription in minute letters on the horse’s harness underlines the rider’s elevated position. The scene is circumscribed by a band enclosing a procession of symmetrically arranged real quadrupeds as well as mythical creatures, the latter including a centaur-archer whose tail terminates in a dragon head. The band is crowned at the top by a pair of expressive dragons with small wings and forelegs whose upper bodies cross (but do not loop) so that the gaping mouth of each appears to snap at the other dragon’s looped tail end (fig.╯33).104 Their position at the apex, as Priscilla Soucek notes, brings to mind the use of architectural dragons as guardians at gates.105 Such a mirror would have belonged to the requisites of the nobility and may well, as Oya Pancaroğlu proposes, have been used as instrument “of allegorical reflection and divination.” This mirror thus “embodies a vision of kingship that extends beyond the horizons of temporal human dominion while affirming the universality of its royal centre.”106 Mirrors have a long history of association with apotropaic properties in antiquity and the medieval period, for they have the power of turning evil back upon itself.107 They are often linked with magic and the Latin version of the magical manual Ghāyat al-ḥakīm (“The Philosopher’s Goal”), attributed to Abū Maslama Muḥammad al-Majrītī (who wrote between 443/1052 and 448/1056),108 includes instructions on how to make a magic mirror, ascribed to Jābir ibn Ḥayyān, the semi-legendary eighth- or ninth-century author of a large body of Hermetic alchemical literature. The owner of this mirror was said to have power over the winds, humankind and demons.109 The amplified depiction of the dragon on such a multi-layered object, as pair at the apex, below the horse’s hooves and as the centaur’s tail, shows that it was considered an extremely valued and compelling iconography that was deemed necessary to further increase the potency of the mirror’s inherent properties.

╇ Azarpay, 1978, p.╯366, n. 20. ╇ Baipakov, 1992, p.╯110; the date of the pieces is not mentioned. They could also date to the period after the destruction by the Mongols in 617/1220, since the city regained some of its commercial prominence by the middle of the thirteenth century, as attested by the travelling Armenian king Hetʿum II of Cilicia (Lesser Armenia, Armenian kingdom from 1198–1375) who in his Account of the Â�Eastern Kingdoms (p.╯128, as cited in Bretschneider, 1888, repr. 1967, p.╯57) called Utrār (Otrar) “the greatest city of Turkestan.”

104 ╇ Cf. Öney, 1969a, p.╯171, fig.╯21; Erginsoy, 1978, pp.╯456–7, figs. 225 a and b; Glory of Byzantium, 1997, p.╯424, cat. no. 282; Turks, 2005, cat. no. 72. 105 ╇ Glory of Byzantium, 1997, p.╯424 (catalogue entry no. 282 by Priscilla Soucek). 106 ╇ Turks, 2005, p.╯395, cat. no. 72. Cf. also Pellat, “Mirʾāt,” EI² VII, 105b. 107 ╇ Ullmann, 1992, pp.╯55–61; Maguire, 1994, p.╯267. 108 ╇ On Abū Maslama Muḥammad al-Majrītī, see Sezgin, 1971, pp.╯294–8. 109 ╇ Cf. Strohmaier, 1989, p.╯267.

102 103

the dragon motif on portable objects d.╇ The dragon motif on vessels The depiction of the dragon is frequently found sculpted as part of vessels, for instance on the arched handles of two celebrated buckets both of which are now preserved in the State Hermitage, St. Petersburg. One of these is the richly silverand copper-inlaid copper alloy bucket, named after its collector, Count Alexei Bobrinsky, the famous “Bobrinski bucket,” which was purchased in Bukhara in 1885 by N.N. Shavrov, the adjuntant of General Chernyayev, governor general of Turkestan (fig.╯34). It was probably made in Herat, one of the main cultural centres of the province of Khurasan which flourished especially under the Ghurid dynasty, during whose rule the vessel was produced as indicated by the date muḥarram 559/December 1163 inscribed in Kufic at the top band of the handle of the bucket.110 The loops of the handle are in the form of a leaping lion and on the inside a dragon protome, from whose gaping mouth issues the four-sided arched section inscribed on two sides in naskhī with benedictory inscriptions (fig.╯56). The second bucket, which in the mid-nineteenth century was in the Parisian collection of Louis Fould before coming into the possession of the St. Petersburg jeweller, A.K. Fabergé, is signed by its maker, Muḥammad ibn Nāṣir ibn Muḥammad al-Harawī. The toponymic (nisba) al-Harawī (“from Herat”) perhaps indicates the origin of the maker, and indeed the bucket is thought to have been made in late twelfth- or early thirteenth-century Khurasan, probably in Herat.111 Yet nothing is known of the maker who could also have migrated from Khurasan westwards; moreover, on the basis of its faceted body and gilt ground as well as several decorative elements, it has recently been attributed to the early thirteenth-century Jazīra, northern Syria, or possibly Anatolia.112 Its handle is closely related to that of the Bobrinski bucket and is similarly held in place by loops in the form of curved 110 ╇ Cf. Loukonine and Ivanov, eds., 2003, pp.╯114–5, cat. no. 116. Cf. Pope and Ackerman, eds., 1938–9, repr. 1964– 81, vol. 13, pl.╯1308; Rice, 1955, pls. XIX–XX; Mayer, 1959, p.╯61; Ettinghausen, 1943, pp.╯193–208; Hartner, 1973–4, p.╯122, fig.╯18. 111 ╇ See however Eva Baer’s (1983, pp.╯301–2) caveat with regard to the assumption that the nisba carried a geographic association that indicates the place of the artist’s workshop.╯Not only could the artist have left his native town but it could also have indicated a special product or specialised technique.

47

dragons topped here by projecting lion-headed knobs (fig.╯57).113 The dragon’s close connection with water has been manifested since ancient times and its ensuing depiction on vessels containing liquid is known at least from the early medieval period. Often this is expressed in dragon-headed spouts. This feature appears on an automaton depicted in the treatise written by the court engineer Abu ’l-ʿIzz Ismāʿīl ibn al-Razzāẓ al-Jazarī (fl. second half of sixth/twelfth century) which details the various automata commissioned by the Artuqid ruler Nāṣir al-Dīn Maḥmūd ibn Muḥammad al-Malik al-Ṣāliḥ (r. 597/1201– 619/1222) for the court’s amusement. The resulting work, Kitāb fī maʿrifat al-ḥiyāl al-handasiyya (“Book of Knowledge of Ingenious Mechanical Devices”), dates to c. 1200.114 Al-Jazarī’s “Handwashing machine,” which was designed for ritual ablutions, is depicted with a dragon-spouted ewer on a leaf from the earliest extant manuscript of this work dated to the end of Shaʿbān 602/about 10 April 1206, copied by Muḥammad ibn Yūsuf ibn ʿUthmān al-Ḥaṣkafī (“of Ḥiṣn Kayfā”) at Diyārbakr, and now preserved in the Topkapı Sarayı Library, Istanbul. When the machine was turned on, water flowed from the cistern in the servant’s chest into the ewer, the bird on the lid of the vessel whistled and the liquid poured out of the gaping mouth of the dragon-headed spout (fig.╯35).115 Such a spout is also portrayed on a Jazīra-type copper alloy ewer (with recently replaced silver inlay), the so-called “Homberg Ewer,” now preserved in the Keir Collection in London, which has an overall decagonal outline. The lower part of the neck is inscribed with the signature of the artist, Aḥmad al-Dhakī, the engraver, al-Mawṣilī and the date 640/1242.116 The dragon motif in manifold variations is frequently found as part of the decorative programme of vessels, as for instance on a rectangular brass tray inlaid with silver and with a central cruciform depression in which four pairs of dragons 112 ╇ Ivanov, 2004, p.╯174; see also idem, n. 19 with reference to Oktay Aslapana (1971, p.╯284), who came to the same conclusion many years ago. 113 ╇ Another thirteenth-century bucket of Anatolian provenance with handles terminating in dragon heads was sold at Sotheby’s, London, 1990; cf. Ivanov, 2004, p.╯175, fig.╯2. 114 ╇ For al-Jazarī’s sources, see Kitāb fī maʿrifat al-ḥiyāl al-handasiyya, tr. Hill, 1974, p.╯74. 115 ╇ Cf. Rogers, tr., exp.╯and ed., 1986, p.╯30, cat. no. 10. 116 ╇ Fehérvári, 1976, p.╯105, cat. no. 131, pl.╯I.

chapter three

48

are portrayed with their heads turned back and their tails intertwined. The well-wishing Arabic inscriptions are typical of Western Central Asian metalwork. In the medieval Islamic world inscriptions bestowing blessings on the owner were often combined with figural decorations of a symbolic or magical intention,117 in order to magnify the overall apotropaic function intended to benefit the maker and owner of the objects. Not least because of the epigraphic bands which are typical of Khurasani metalwork, the tray has been attributed to the first half of thirteenth-century northeastern Iran; however on the basis of the creatures’ entwined tails an Anatolian or Jazīran provenance has recently been suggested.118 A related composition also fills the inner band of three zigzag bands that circumscribe the conical base of a thirteenth-century copper alloy candlestick base, inlaid with silver, which is preserved in the Metropolitan Museum of Art, New York. It comprises bilateral compositions of conÂ�fronted pairs of quadruped dragons in profile. The tail tips

of the creatures are entwined with the contiguous addorsed dragons’ tails (fig.╯36). They are portrayed with their inner forelegs raised, their wideopen jaws revealing the tongues and the upswept, curved wings with tightly curled tip projecting from the haunches. The dragons’ hide is covered with scales, the tails knotted at mid-section to form a figure of eight. Importantly, the tapering tail tips transform into small dragon heads which appear to grasp or snap at the serpentine coils.119 An epigraphic frieze around the shoulder invokes blessings such as glory, prosperity, perfect health, good fortune, felicity, etc., upon the owner. It is notable that the paired dragons have close analogies to the dragon figures carved in low relief on three relief-carved stone panels from the now destroyed city walls, gates or citadel of Konya (618/1221) (fig.╯60). Like their monumental cousins, the miniature versions of the dragons on the candlestick base may well have conveyed an impression of might and good fortune, thus augmenting the impact of the epigraphic blessings.

╇ Cf. Kerner, 2004, pp.╯218–9 and ns. 79, 80. ╇ Ivanov, 2004, p.╯174. The tray is published in the exhibition catalogue Islam and the Medieval West, 1975, cat. no. 58. Dimensions 8 inch square. At the time of publication

the piece was in the collection of N. Anavian. 119 ╇ Cf. d’Avennes, 1877, pl.╯161–5, fig.╯3; Dimand, 1926, p.╯197, fig.╯5.

117 118

dragons and the powers of the earth

part two

the dragon and the natural world

49

dragons and the powers of the earth

51

chapter four

Dragons and the powers of the earth

a.╇ The dragon and the elements Equally at home on land and sea, the dragon is associated with remote places and phenomena of the natural world. Its aquatic nature is profoundly ambivalent: as water dweller it can be both benevolent guardian and malevolent destroyer. In its threatening manifestation the creature is linked to adverse climatological phenomena such as thunder, rain, lightning or earthquakes. The symbolic complexity of the dragon is thus expressed through its ability to cross boundaries within the natural environment it inhabits. The distinction between land- and sea-beast is often blurred. That the dragon or the large serpent can be both aquatic and terrestrial was noted in the fifth-century Armenian theological writings of Eznik of Koghb.1 According to the texts of the Ikhwān al-Ṣafāʾ (Brethren of Purity, established c. 373/983), the likeness, character and manner of the dragon is like the sea serpent.2 Both aquatic and terrestrial, the dragon, like its close cousin the amphibian serpent, is thus characterised by a wet-dry dichotomy as noted by the fourteenth-century scholar Kamāl al-Dīn al-Damīrī in his Ḥayāt al-ḥayawān al-kubrā.3 They dwell not only in springs, wells, rivers, lakes or sea water, but also in mountains, forests, caverns, caves, crevices and other subterranean enclosures, hence lending themselves to association with the underworld and chthonic forces. In tunnelling into the earth and resurfacing again above ground, they

1 ╇ Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, pp.╯593–4, ch. 133. 2 ╇ Tr. and ed. Dieterici, 1858, pp.╯114–6. 3 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯636. See p.╯5. 4 ╇ This observation is recorded by al-Bīrūnī in his Kitāb al-Āthār al-Bāqiya (“The Chronology of Ancient Nations”) (tr. and ed. Sachau, 1876–8, p.╯248) in which he states that during the cold season he himself found that:

...in Khwārizm, they gather in the interior of the earth and roll themselves up one round the other so that the greatest part of them is visible, and they look like a ball. In this condition they remain during the winter until this time.

are associated with the fertilisation of the earth. Their absence and re-emergence according to the cycle of the seasons (during the dormant season it hibernated in the ground)4 may also be seen as a metamorphosis. In the Rigvedic pantheon a primordial “serpent of the deep,” Ahi Budhnyà, is known;5 the Vedic áhi- meaning “serpent, snake,” while budhnyàis an adjectival derivative of budhnàs “bottom, base.” The origin and abode of the “dragon of the deep” is the dark bottom of heavenly waters, he is “sitting in the depth of rivers” (budhne nadīnāṃ rajaḥsu sīdan).6 In the Rigveda (dating from 1500–1000 bc) budhnàs is used of the root (in heaven) of the cosmological Nyagrodha tree (1.24.7),7 hence associating the serpent with a tree.8 In later Indian literature water is known as the abode of serpent demons.9 Apart from the aquatic monster Gandarəβa who lives in Lake Vārukasha (originally perhaps denoting a specific location such as Lake Aral or the Caspian Sea), the Iranian Zoroastrian dragons were terrestrial creatures, “inhabitants of this world,” and the connection with water is less evident in Zoroastrian literature, with the exception of some references to river-dwelling dragons.10 Yet in almost all of the stories in Iranian literature, the dragon’s lair is close to either a source of water or the sea,11 for instance, the dragon-fighter Garshāsp in the Garshāsp-nāma kills a dragon which had emerged from the sea and made its abode on Mount Shekāwand, while Sām in the Shāh-nāma12 slays

5 ╇ Oldenberg, 1894, repr. 1977, pp.╯71–2; Watkins, 1995, pp.╯460–2. 6 ╇ Grassmann, 1873, repr. 1976, cols. 909–10. Cf. Watkins, 1995, pp.╯460–2. 7 ╇ Watkins, 1995, p.╯460. 8 ╇ After he was killed by Indra, the Rigvedic hymns (1.32.5) state the following about the dragon Vá¹łtra: “As trunks of trees, what time the axe hath felled them, low on the earth so lies the prostrate dragon” (tr. Hotchkin Griffith, 1889, p.╯20). 9 ╇ Oldenberg, 1894, repr. 1977, p.╯71; Vogel, 1926, pp.╯32– 33, 115–6, 209, 244; Bosch, 1960, pp.╯33–4, 51–3, 136–7. 10 ╇ Boyce, 1975, repr. 1996, pp.╯90–1. 11 ╇ Cf. Khāleqī-Moṭlaq, “Aždahā II,” EIr. 12 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯309,

52

chapter four

a dragon from the river Kashafrūd in Khurasan.13 The dragon’s aquatic nature is clearly perceptible in the Armenian vishap (Georgian veshapi, also a fabulous serpent; Syriac wshp; a loanword from the Avestan vishāpa-, “whose saliva is poison”) was used as an epithet to azhi, “serpent,”14 which was said to dwell in the waters of a lake.15 In the ancient Semitic world the predominant belief was that both wood and water are potent generating forces,16 a notion which continued to exist in the medieval Islamic period.17 Holy trees are known either by direct assertion or by implication to be associated with spring shrines.18 Often the sacred spring or well spirit or numen was an accompaniment of a sacred tree19 or sacred place.20 Sources of life such as wood and water are also considered to be channels of a greater power; the power that is contained in them is thought to be “contagious.”21 In the Islamic period trees and twigs as part of trees were widely used for religio-magical purposes, and by extension, any magic rods or wands used in such practices were related to the idea of woody plants.22 The association of vegetation and the dragon is reflected in the description by the universal historian Abū Jaʿfar Muḥammad ibn Jarīr ibn Yazīd al-Ṭabarī (c. 224–5/839–310/923) of the rod of Mūsā (Moses) as a: ...two-pronged fork with a crook under the meeting point of the twigs and when it was turned into a serpent, the two twigs formed the mouth of the serpent with its forked tongue, while the crook took the shape of the crest.23

ll. 1175–176. In the Pahlawī Riwāyat of Garshāsp, the hero fights in the sea with the dragon Gandarw for nine days and nights; after his victory he slaughters fifteen horses and eats them. See Monchi-Zadeh, 1975, p.╯138. 13 ╇ The early thirteenth-century historian Ibn Isfandīyār (I, p.╯89), probably a native from Āmul, has similarly recorded a tale from Māzandarān in which Sām had vanquished a dragon at an otherwise unknown location called Kāva Kalāda in the same province near the sea; see KhāleqīMoṭlaq, “Aždahā II,” EIr. 14 ╇ Nirangastān 48, cited after Ananikian, “Armenia (Zoroastrian),” ERE, vol.╯5, part 2, 1914, p.╯800; Skjærvø, “Aždahā I,” EIr; Boyce, 1975, repr. 1996, p.╯91, n. 42. 15 ╇ Russell, 2004, p.╯373. 16 ╇ Robertson Smith, 1889, repr. 1927, pp.╯132–3, 135–6, 166–7; Jeffers, 1996, p.╯145. 17 ╇ Cf. Whitehouse, “Holiness (Semitic),” ERE 6, 1913, p.╯754. 18 ╇ Cf. Wood, 1916, p.╯19. 19 ╇ Robertson Smith, 1889, repr. 1927, p.╯136; Jeffers, 1996, p.╯163 and n. 107. 20 ╇ Eadem, pp.╯169–70; Whitehouse, “Holiness (Semitic),” ERE 6, 1913, p.╯754. 21 ╇ Jeffers, 1996, p.╯145.

The numinous power inherent in the serpent, discussed further below, is also reflected in the Qurʾānic story of Mūsā’s rod turning into a serpent, which is an example of the living power of the rod.24 The ancient association of the dragon with water is revealed in the names of streams, lakes, pools or springs that are also often compared with each other in poetic simile. The serpent as guardian and custodian of a water source is referred to by the ancient Greek term drakōn in the second-century ad compilations of PseudoApollodorus.25 According to the Byzantine historian Procopius of Caesarea, a river in Bithynia was called Drakōn because its shape resembled that of the fabulous monster.26 The sacred fountain of Ephca at Palmyra, which is a sulphurous spring, is associated with a demon in serpent form.27 As William Robertson Smith has pointed out, there are indications that in certain instances the original sanctuary was at a well beneath the town as was the case of the original sanctuaries of Jerusalem, such as the fountain of En-Rogel (1 Kings I, 9, 38) where Adonijah held his sacrificial feast, located near the “serpent’s stone” which may possibly be identified with the “dragon well” (Nehemiah II, 13).28 Sacrifices offered at the well of Abraham at Mamre were said to be eaten by the serpent denizen of the water.29 In Syria sacred springs were thought to be guarded by spirits in the form of giant pythons.30 The chronicler of the early Byzantine period, Ioannes Malalas (d.╯c. 570/580), mentions that the partly subterranean river Orontes in Asia Minor was called Drakon,31 ╇ Cf. Schimmel, 1994, pp.╯29–30. ╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, I, Cairo, n.d., p.╯401; al-Thaʿlabī, ʿArāʾis al-majālis fī qiṣaṣ al-anbiyāʾ, Cairo, n.d., p.╯90, cited after Fodor, 1978, p.╯4. 24 ╇ Schimmel, 1994, p.╯30. It is also of note that Ezekiel (17.3-10) likens the king to a vine and (19:11) calls the vine “fit to be carved into a royal scepter.” 25 ╇ Apollodorus, Bibliotheca III 4.1. The central episode of the myth of Kadmos is his victory over the spring-guarding drakōn, Ares’ son, whose teeth he sowed in the earth and out of which grew armed warriors (the Spartoi). On the Kadmos myth, see Fontenrose, 1959, repr. 1980, pp.╯306–20; Astour, 1965, pp.╯156–61. 26 ╇ Procopius VII, Buildings V 2.3-10. 27 ╇ Robertson Smith, 1889, repr. 1927, pp.╯168–9. 28 ╇ Idem, p.╯172, n. 3. 29 ╇ Idem, p.╯177. 30 ╇ Whitehouse, “Holiness (Semitic),” ERE 6, 1913, pp.╯751–3; Chelhod, 1955, p.╯105 and n. 6. For other examples in Syria and Palestine, cf. Robertson Smith, 1889, repr. 1927, pp.╯168, 176–7. 31 ╇ Chronographia, tr. and ed. Dindorf, 1831, p.╯38,1. Cf. Robertson Smith, 1889, repr. 1927, pp.╯171–2; 22 23

dragons and the powers of the earth

53

and was, according to Strabo, also known as Typhon and Ophites.32 The continued association of the Orontes with the mythical dragon is shown in a Turkish legend from Hatay which states that the river was created when the dragon took flight from the mysterious immortal Islamic Prophet Khiḍr by digging underground channels.33 In Persian the close affiliation between the serpent and the stream of water is reflected in the word liwāʾ which comes to mean “the winding of the river” as well as “the twisting or coiling of a serpent.”34 A serpent spring also makes an appearance in the Kitāb-i Samak ʿAyyār (“Book of Samak the Adventurer”), collected by Farāmarz ibn Khudādādh al-Arrajānī and written down by Ṣadaqa ibn Abu ’l-Qāsim Shīrāzī,35 which is thought to be rooted in the Parthian period.36 The association of the dragon with water is similarly evident in the Transcaucasian tradition. The appellation of the Armenian river Awji near Awjaberd in Gegharkʿunikʿ province contains just like the toponym the word awj (serpent).37 One of the most ancient Armenian cults, that of tree worship, is often connected with water sources and serpents. In his collection of Armenian folklore entitled Krots-Prots (ch. 9), the eighteenthcentury clergyman Garegin Servantsian records the ancient belief that aged serpents come to a certain source to shed their skins, eat a flower which only they know, bathe in the spring and are then rejuvenated. Anyone who finds the same

flower and eats it and then drinks three times from the water in which the serpent bathed will similarly become immortal.38 In pre-Islamic Central Asia, in particular in the regions of presentday Afghanistan, the connection of dragons with water was expressed by the pan-Indic serpent deities (nāgas), the serpent genii, who dwell in terrestrial water sources, and to whom were attributed tempests and floods.39 In Kashmir, the word nāg occurs in a variety of names of springs, rivers, or reservoirs, for instance, Lake Nīla Nāg in the region of Nāgām (ancient Nāgrāma).40 This close association of the dragon with the element of water also led to its depiction on waterspouts. This is evidenced in the Islamic period in the example of Karatay Han situated between Kayseri and Malatya. Two of the conduits used for drainage from the gutters of the roof (gargoyles) of the mid-thirteenth-century caravanserai have the appearance of winged dragonite protomes that hold in their wide-open mouths what appears to be the stylised upper body of a human being whose hands clutch the monsters’ forelegs (fig.╯37).41 The dichotomy of the dragons is evident in their activities either as guardians of natural sources or as ravagers in the form of destructive natural phenomena. In the popular mind dragons lent themselves naturally to functioning as symbols for the mysterious and destructive forces of the earth. The notion of dragons as guardians of water

Fontenrose, 1959, repr. 1980, pp.╯277–8. 32 ╇ Reportedly this is because the wounded drakōn Typhon had crawled into the underground channels of the source of this river to seek shelter from the thunderbolts of Zeus (Strabo, Geography XVI, 750). The name of the Greek she-dragon Pythōn (phuth- from Vedic bhudh-) who is slain by Apollo with his arrow (Fontenrose, 1959, repr. 1980, pp.╯13–45) also demonstrates the inherent designation of a serpent creature of the watery deep (Watkins, 1995, p.╯461). The Pythōn myth, as Joseph Fontenrose (1959, repr. 1980, pp.╯77–93, 193) shows, largely corresponds with that of Typhon. In the Greek tradition the outer ring of the world’s water, the ocean (Ogenos-Okeanos), is also the lair of Ophioneus (Ophiūchus), the serpent-man who tried to overcome Zeus (Janda, 2010, pp.╯71–89; Russell, 2004, p.╯718). In Phaedo Plato’s literary figure, Socrates, compares the rivers to the serpent: “There are some, then, that after having encircled the earth with one or more coils, like snakes, descend so deeply that they come out at the lowest point of Tartarus,” as cited in Mastrocinque, 2005, p.╯29. Cf. Robertson Smith, 1889, repr. 1927, p.╯176 and n. 4. 33 ╇ Franke, 2000, pp.╯100, 547–9. 34 ╇ Cited after Steingass, 1892, repr. 1981, p.╯1130. 35 ╇ See the preface of the Oxford Kitāb-i Samak ʿAyyār (Bodleian Library Ms. Ouseley 379–81); Sachau and Ethé,

1889, pp.╯422–3; Gaillard, 1987, pp.╯10–1. 36 ╇ Gaillard, 1987, pp.╯64, 155. 37 ╇ Alishan, G., Hin hawatkʿ kam hetʿ anosakan krōnkʿ Hayokʿ (“The Ancient Faith or Pagan Religion of the Armenians”), Venice, 1910 ed., pp.╯165–6, as cited in Russell, 2004, p.╯461. 38 ╇ Ishkol-Kerovpian, “Baum- und Pflanzenkult,”WdM IV, 1, pp.╯105–6. 39 ╇ The nāgas are genii of lakes and springs, and are worshipped for their beneficent as well as destructive aspects as powers of the waters. They are considered also as guardians of treasure and givers of vital forces stored up in springs and wells. The accounts of the seventh-century Chinese pilgrim Xuanzang (Hsüan-tsang) show that the nāga cult was still flourishing in parts of Western Central Asia and northern India in the seventh century. Cf. Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, pp.╯121–3. For an in-depth study of the nāgas in Indian iconography, see Vogel, 1926. 40 ╇ Idem, pp.╯227–31, with further examples. 41 ╇ See Erdmann, 1962, pl.╯IX. Roux (1972, pp.╯373–5, figs. 1 and 2), however, recognises these spouts to be sculpted in the form of a fantastic leonine animal. Gargoyles in the form of dragon heads are also found at the fourteenth-century Araboğlu Mosque in Karaman, south of Konya. See idem, p.╯373, n. 2.

chapter four

54

sources, which was well-known in ancient Greek sources,42 is recorded in the medieval period by the twelfth-century Byzantine scholar Ioannes Tzetzes, a commentator on Lycophron.43 By contrast, destructive natural phenomena attributed to dragon monsters, also well-established themes in ancient Greek sources,44 were mentioned by the tenth-century Arab encyclopaedist al-Masʿūdī who associates meteorological phenomena with the mythical creatures. He reports that the Caspian and the Mediterranean (near Tripoli and Latakia) were “prolific in sea monsters” (kathīr al-tanānīn), adding that according to tradition the sea monster (al-tannīn) was a “black wind nurtured in the depth of the sea, which ascends to the zephyrs.” He adds the account that the al-tannīn were: …black serpents existing in the plains and mountains, in which places there are floods and rainstorms, carrying them down into the sea, where they feed upon the sea-creatures so that their bodies attain great size, and their age is extended, and in the end some of them achieve the age of 500 years, and become the lords of the sea. These stories are by no means denied by the Persians, who assert that the monsters have seven heads, and are called Ajdahā [Azhidahāka].45

account of the formidable fire created by the venom which sits between his jaws and moves freely in flames within his body, the dragon endeavours to obtain relief in freezing temperatures. The text describes that as a result of the intensity of the fiery heat of his venom, the dragonking seeks as abode the peaks of high hills and mountains, above the regions of mild air, in a region where the cold is so intense that there can be neither clouds nor rain, and where neither plants nor animals can survive.47 Citing Ibn ʿAbbās, al-Masʿūdī, moreover, reports the medieval Islamic idea that “when the tail of a dragon strikes a large edifice, (like) a tree or mountain, it destroys it. Furthermore, at times when the dragon breathes it sets fire to large trees,”48 apparently describing natural manifestations such as earthquakes. This association between dragons and fire is similarly mentioned in the Kitāb-i Samak ʿAyyār.49 The fiery breath of the notoriously pelagic biblical Leviathan is also described in Job (41:11–3): Out of his mouth go burning torches, and sparks of fire leap forth. Out of his nostrils goes smoke, as out of a burning pot or cauldron. His breath kindles coals and a flame goes out of his mouth.

The celebrated thirteenth-century cosmographer and geographer Zakariyyāʾ ibn Muḥammad ibn Maḥmūd al-Qazwīnī (c. 600/1203–4–682/1283), who originated from an Arab family that had been Iranised after settling at Qazwīn, similarly relates that the Iranians believed “the sea-dragon to be either a hurricane or a black serpent dwelling on the sea bottom.”46 The Ikhwān al-Ṣafāʾ, whose Rasāʾil (Epistles) appeared in the tenth century, states that, on

Moses of Chorene’s writings also preserve preChristian Armenian religious poetry which describes the birth of Vahagn, god of strength and victory. Sudden storms or winds on Lake Van are a sign that dragons (vishap) live in the lake, growing there until they are large enough to destroy the world, at which point Vahagn (probably conflated in this story with the Hurrian weather god Teshub, the Urartean Teisheba)50 drags the Â�dragons up from the depths to take them into the sky to burn up in the sun.51 The twelfth-Â�

╇ A giant dragon (drakōn) guarded the local spring, sacred to the god Ares, at the future site of the Boeotian city Thebes (Apollonius, Argonautica 3.1178–87; cf. also Euripides, Phoenician Women 930–5 and Bacchae 1274, 1314–5). 43 ╇ Ioannis Tzetzae Historiae, ed. Leone, P.A.M., Naples, 1968, p.╯404, 399, as cited in Bouras, p.╯67 and n. 41. 44 ╇ The contumacious Greek primeval monster Typhon was defeated by Zeus but as punishment lived on under the earth (under volcanoes or in Tartaros); he personified volcanism, being thus associated with volcanic eruptions. The classical Greek lyric poet Pindar, for instance, ascribes the volcanic action of Etna in Sicily to the drakōn Typhon who was imprisoned under the mountain (Pythian Odes 1.15–28; see also Aeschylus, Prometheus Bound, 351–72). Typhon also appeared as demon of storms and whirlwinds (Hesiod, Theogony 846, 869–80). 45 ╇ Kitāb marūj al-dhahab, tr. and ed. Barbier de Meynard and de Courteille, 1917, vol.╯vol.╯1, pp.╯266–7.

╇ Kitāb ʿajāʾib al-makhlūqāt, ed. Wüstenfeld, 1849, repr. 1967, p.╯129. See also Badiee, 1978, pp.╯120–1. 47 ╇ Tr. and ed. Dieterici, 1858, pp.╯115. 48 ╇ Kitāb marūj al-dhahab, tr. and ed. Barbier de Meynard and de Courteille, 1917, vol.╯1, p.╯267. For a translation of the passage into German, see Monchi-Zadeh, 1975, p.╯159, n. 33. 49 ╇ Gaillard, 1987, p.╯63. 50 ╇ Geo Widengren (1966, p.╯444) suggests an amalgamation of Iranian beliefs held by Armenians with the older Anatolian substratum of the Hurrian song of the monster Ullikummi, whom the weather god Teshub (Urartean Teisheba) smites. Cf. Ishkol-Kerovpian, “Vahagn,”WdM IV, 1, pp.╯149– 52; Schwartz, 1975, p.╯416; Russell, 1987, p.╯29, and idem, 2004, pp.╯357–61, 373; Mahé, 1994. However, according to Mary Boyce (1975, repr. 1996, p.╯64, n. 279), the tale of Vahagn killing a monster may be a late development. 51 ╇ Patmutʿiwn Hayocʿ, tr. Langlois, 1872, p.╯41 and n. 1. See also Russell, 2004, pp.╯357, 361, 617–8 and n. 23, 631 and n. 38, 1132, 1287, n. 39; van Lint, 2009, pp.╯257–8.

42

46

dragons and the powers of the earth

55

century Armenian historian Matthew of Edessa (Mattʿēos Uá¹łhayetsi) relates the eruption of volÂ� canoes to the fiendish nature of dragons, referring to this phenomenon metaphorically as the dragons of Mount Ararat fighting those of Mount Aragac.52 An echo of the ancient association of dragons with water seems distantly to reverberate in the ancient Iranian festival of Sada, the celebration of which, according to the historian of the Ghaznawid dynasty, Abu ’l-Faḍl Bayhaqī, resumed during Ghaznawid times. For the celebration of Sada, a festival held fifty days before Nawrūz (the celebration of the return of spring), large fires are lit in commemoration of the Pīshdādian king Hūshang (Haoshyaṅha), the first lawful king who reigned over the seven climes of the world, over the demons and the sorcerers, and according to al-Bīrūnī, and later Firdawsī, was the inventor of fire. As stated by the latter, Hūshang also originated the idea of using irrigation canals. Firdawsī, moreover, records the king’s feat of valour in vanquishing a monstrous dragon that infested the country and which the king himself attacked with stones, “when one of them falling with prodigious force upon another, struck fire and set herbage and surrounding trees in a blaze, and consumed the dragon in the flames.”53 The writer adds that the legendary pre-Islamic monarch “gave orders that prayers should be said facing a fire, saying: It is the spark given by God (Īzād); worship it if you are wise.”54 Sada was held in winter “to strengthen the sun and to help bring back warmth and light to the world.”55 As was customary, sulṭān Masʿūd of Ghazna (the son and successor of Maḥmūd) chose to celebrate it beside a stream, where a fire was lit56 to aid symbolically the stream of water “in his subterranean task of protecting plants and springs from frost.”57 The affiliation between dragon, water and earth also becomes apparent in the Alexander Romance by the Pseudo-Callisthenes, thought to have originated at some time prior to the third century ad

as a collection of oral legends and other material. A good portion of the material was progressively gathered in written form, probably assembled in Alexandria, and entered the Iranian tradition no later than the Sasanian period. In this romance Iskandar (Alexander) is flying through the air on the back of the eagles when, at the highest point of the sky, he sees an enormous serpent whose coils enclose a disc representing the world surrounded by the ocean.58 In medieval Iranian poetry, the transformative power of the dragon is sometimes evoked as a portent to signal changes in the course of human events or impending alterations in the cosmic cycle. This type of metaphor is employed in the verse romance, Wīs u Rāmīn, composed by Fakhr al-Dīn Gurgānī for the governor of Iṣfahān on behalf of the Saljuqids. The story relates how Rāmīn has become unfaithful to Wīs who remonstrates with him in a long elaborate letter reminding him of her love. Rāmīn sets off to Marw in the hope of a reconciliation with his beloved, but when he arrives on horseback, a snowstorm is in progress. The meteorological phenomenon is figuratively associated with the dragon to evoke human emotions:

52 ╇ Abeghyan, M., Erkeri Zhoghovadsu (“Collected Works”), Erevan, 1966–75, vol.╯7, p.╯65, and Avandapatum, no. 11, as cited in Russell, 1987, p.╯206. 53 ╇ Cited after Steingass, 1892, repr. 1981, p.╯663. 54 ╇ Massé, “Hūshang,” EI2 III, 637b. 55 ╇ Boyce, 1983, p.╯800. 56 ╇ Bayhaqī, Taʾrīkh-i Masʿūdī, ed. Ghanī and Fayyūḍ, Tehran, 1324/1945, p.╯278, as cited in Lambton, “Marasīm: 3. In Iran,” EI² VI, 518a.

╇ Boyce, 1983, p.╯801. ╇ Pseudo-Callisthenes II, ch. 41, tr. and ed. Stoneman, 1991, p.╯123. Cf. the imagery described in Reitzenstein, 1904, p.╯31; Millet, 1923, p.╯94; Grabar, 1951, pp.╯47–8; see also the discussion on the astral ascension of the Sasanian king in L’Orange, 1953, pp.╯64–79, which argues for a primeval oriental origin of the motif of Iskandar’s heavenly ascent (p.╯69). 59 ╇ Tr. and ed. Davis, 2008, p.╯448.

... the skies became like some vast dragon breathing tongues of flame; the snow was like a poison, since within it men’s hearts would freeze and stiffen in a minute, black clouds were massing, blocking out the light, choking back breath, depriving eyes of sight; the snow blew with such force that elephants could not have stood its vehemence.59

The natural phenomenon thus described correlates with the human event by functioning as a form of inauspicious portent. A dragon in the form of a black cloud appears also in the epic poem Bahman-nāma (“Book of Bahman”), in which Ādar Barzīn, the son of the dragon-slayer Farāmarz, recognises that the cloud is a transformed dragon which came out of a mountain every spring to violate the daughter of the local ruler, Bīwarāsp/Dahāk. He kills the dragon with

57 58

56

chapter four

arrows and then washes at a spring.60 In this story the killing of the dragon is followed by contact with water. In the Shāh-nāma a similar reference is made. Both heroes, Rustam, after he had killed the dragon Babr-i bayān which came out of the sea once a week, and Borzū, after he had vanquished the dragon on Mount Zahāb, lose consciousness, and on reviving also wash themselves in a spring.61 Likewise in another episode of the Shāh-nāma, after accomplishing the third of his seven trials (haft khwān), that of slaying the dragon, Isfandiyār, one of the prince-heroes of the epic and eldest son of shāh Gushtāsp, loses consciousness. On reviving, “he called for a new garment, and then immersed himself in a nearby stream, washed the dirt from his body. Thus cleansed, he came before the Lord. Down on the ground again, contorting and writhing like a serpent, he cried out: Must it not be that the dragon-slayer is constantly sustained by the One who grips the world?

The troops invoked blessings upon their leader, and all of the company bowed low before the Just Provider.”62 Frequent reference is thus made in the legends to a loss of consciousness of the hero in the aftermath of the dragon-slaying and his subsequent contact with water that is known to have a “magical” cleansing effect and to be a transformative agent. The purification thus constitutes a rite of separation from the act that has been accomplished. That water is an agent of transformation is further evidenced in an account of the slaying of a dragon (shīr-i kappī) in Turkestan by Bahrām Chōbīn, Sasanian commander of Hurmuz IV (r. 578–590), rival to the throne of his son Khusraw

60 ╇ Bahman-nāma, BM Or. 2780, fols. 180, 181, as cited in Khāleqī-Moṭlaq, “Aždahā II,” EIr. Cf. Hanaway, “Bahmannāma,” EIr. 61 ╇ Cf. Khāleqī-Moṭlaq, “Aždahā II,” EIr. 62 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯4, p.╯503, ll. 1702– 1714. English tr. as cited in Dickson and Welch, 1981, vol.╯II, p.╯201. See also al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯309–12. 63 ╇ Khāleqī-Moṭlaq, “Aždahā II. In Persian Literature,” EIr. 64 ╇ Robertson Smith, 1889, repr. 1927, p.╯168. 65 ╇ Whitehouse, “Holiness (Semitic),” ERE 6, 1913, p.╯754; Robertson Smith, 1889, repr. 1927, pp.╯119, 168, 171–2, 176. 66 ╇ Closely related to the nature of the jinn seem to be the Armenian shahapets, serpent genii of places, and as such “supernatural protectors” of tombs, homesteads or rural sites, often also residing in vegetation, especially trees. AghatanÂ�gelos, pp.╯56–7, as cited in “461.šahap,” “462.šahapet,” Â�Hübschmann, 1895, pp.╯208–9. On the shahapets, see Ananikian, 1925, pp.╯74–6; Ishkol-Kerovpian, “Šahapet,” WdM IV, 1, p.╯136; Russell, 1987, pp.╯329–34. The fifth-centuryÂ�

II Parwīz. However, in this case it is the dragon that would become invulnerable if it went to a certain spring and wetted its hair.63 b.╇ The serpent(-dragon) jinn Sources of water such as wells or streams, maniÂ� festations of life around which vegetation spread, were thought to be endowed with properties of generation, cleansing and in some cases with medicinal or healing virtues.64 These were sometimes also regarded by the Arabs and other Semites as inhabited by the serpent genie (jinn), and hence had a sacred significance.65 Serpents as well as the jinn (pl.╯ajnān, “genii”),66 the supernatural spirits invisible to the human eye, whose existence is recognised in the Qurʾān (sūras 72, 130; 37, 158; 51, 56; and 55, 15), were considered from the oldest times as the general earthly genius loci of trees, in particular roots of trees, as well as caves, springs and wells.67 According to a tradition of the Prophet Muḥammad cited by al-Damīrī in his para-zoological encyclopaedia, chthonic creatures such as serpents represent one of three categories of jinn.68 While usually invisible, the jinn liked to manifest, according to Muslim popular belief, as creeping creatures, reptiles and amphibians, in particular serpents.69 In pre-Islamic Arabia, the jinn were regarded as semi-divinities.70 The serpents’ close association with metamorphosis was motivated by their mutant nature, manifested by behaviour such as the periodical sloughing off their skin, living in water or tunnels beneath the earth and alternating between land and water. This aptitude to metamorphose was often seen Armenian apologist, Eznik of Koghb (fl. c. 430–c. 450), Bishop of Bagrewand, notes that they: ...appeared sometimes as a man, sometimes as a serpent, because of which it was made possible for serpentworship to be introduced into the world. Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, p.╯594, ch. 138; see also pp.╯574–5, ch. 64. The theme of the serpent as genius loci occurs also in the Alexander Romance; Pseudo-Callisthenes I, ch. 32, tr. and ed. Stoneman, 1991, p.╯65. 67 ╇ Wellhausen, 1897, pp.╯106, 212, 214. Robertson Smith, 1889, repr. 1927, p.╯120, n. 1. Macdonald [Massé], “Djinn,” EI² II, 546b. Zbinden, 1953, p.╯49. 68 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯449. 69 ╇ Nöldeke, 1860, pp.╯412–4, and idem, 1913, p.╯669; Robertson Smith, 1889, repr. 1927, pp.╯120, 129; Ruska, “Ḥayyā,” EI² III, 334b; Gohrab, 2000, p.╯87. 70 ╇ Robertson Smith, 1889, repr. 1927, pp.╯121–3, 138. Cf. Macdonald [Massé], “Djinn,” EI² II, 546b.

dragons and the powers of the earth

57

as a source of numinous or demonic power. As primarily chthonic creatures, yet able to undergo so portentous a transformation, the jinn emÂ�Â�body “the undefined and innominate divine,”71 having a very close and necessary link with the serpent.72 The close association of some jinn with trees through their characterisation as the spirit that resides in vegetation,73 is exemplified in the story of Ḥarb ibn Omayya and Mirdās ibn Abī ʿĀmir, historical persons who lived a generation before Muḥammad. When the two men set fire to an inaccessible, knotty thicket, the jinn, taking the form of white serpents, flew out of the burning grove of al-Qurayya with doleful cries and the intruders died soon afterwards.74 It is believed that the jinn slew them “because they had set fire to their dwelling place” and thus violated their haunt.75 In the Islamic period the pagan gods of the so-called Jāhiliyya period, considered to be the time of ignorance and false beliefs, were broken, their sanctuaries destroyed and their guardians dispersed.76 The gods and demigods were subsequently downgraded into jinn.77 The serpent was one of the most ancient sacred symbols of the pre-Islamic cults,78 however since Islam broke with these pre-Islamic practices, Muḥammad gave orders to kill the serpent (amara bi-Â�qatlil-aym) even in the midst of prayer,79 even if the believer is in a state of sacralisation (iḥrām), or in

the sacred enclosure of the Meccan sanctuary.80 Yet these pre-Islamic cults were only gradually and hesitantly abandoned.81 The ninth-century historian of Mecca and its sanctuary, al-Azraqī (d.╯222/837), reports on the authority of Ṭalq ibn Ḥabīb that in the first century of Islam a male serpent circumambulated the Kaʿba, the most sacred building of Islam, called the House of God (bayt Allāh), located in the centre of the Great Mosque of Mecca, and when warned by the Muslims that were present, suddenly took to the skies and disappeared:

╇ Wellhausen, 1897, repr. 2007, p.╯106. ╇ Nöldeke, 1860, p.╯413. 73 ╇ The ancient conception of trees as animated beings (in ancient Greece, for instance, Aristotle, De plantis, I, p.╯815; Plutarch, De placitis philosophorum, V, 26; for Hebrew lore, see Judges, 9–10; 2 Kings, 9), also explains the particularly close association of jinn with trees; cf. Robertson Smith, 1889, repr. 1927, pp.╯132–3. 74 ╇ Abu ’l-Faraj, Aghānī VI, p.╯92 and XX, pp.╯135–7; Yāqūt, Muʿjam al-buldān III, p.╯85, as cited in Wellhausen, 1897, repr. 2007, pp.╯152–3; Robertson Smith, 1889, repr. 1927, p.╯133; Zbinden, 1953, p.╯76. Moreover, killing a serpent is said to make enemies of the spirits; see Henninger, 2004, pp.╯15–6. 75 ╇ Abu ’l-Faraj, Aghānī VI, p.╯92 and XX, 135–7, cited after Robertson Smith, 1889, repr. 1927, p.╯133. 76 ╇ Atallah, 1975, p.╯166. 77 ╇ Robertson Smith, 1889, repr. 1927, p.╯120. 78 ╇ Atallah, 1975, p.╯166. There was a custom to take sacred oaths “by the serpent between the two ḥarra [basaltic lava fields],” recorded in Ibn Hishām’s edition of the Ibn Isḥāq’s Sīra, ed. Wüstenfeld, 1857, p.╯16; Ibn al-Athīr, Nihāya, I, p.╯450; The History of al-Ṭabarī, vol.╯5, tr. and ed. Bosworth, 1999, p.╯179 and n. 460. Cf. Robertson Smith, 1889, repr. 1927, p.╯130, n. 1; Atallah, 1975, p.╯167. Wahib Atallah (pp.╯162–9) moreover conclusively deduces that the translations for the sacred phrases, Aymu-l-Lāh and Aymu-l-Kaʿba

were “by the serpent of God” and “by the serpent of the Kaʿba,” respectively, and hence meant that, just like the oath discussed above, these formulas were made “by the sacred serpent.” 79 ╇ For this as well as further traditions, see Ḥamd Allāh al-Mustawfī al-Qazwīnī, The Zoological Section of the Nuzhatu-l-qulūb, tr. and ed. Stephenson, J., London, 1928, Persian text, pp.╯55–6, tr. p.╯38, as cited in Ettinghausen, 1955, p.╯277. 80 ╇ Al-Azraqī, Kitāb Akhbār Makka, I, pp.╯377–9, as cited in Atallah, 1975, p.╯166. 81 ╇ Idem, p.╯166. 82 ╇ Cited in al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯221. The French translation of the text more precisely states: “...“Être pieux, muʿtamir, Dieu a agrée ta prière et ta visite du lieu saint. Mais, notre pays ne manque pas d’esclaves ni d’hommes incapables d’apprécier le bien, sufahāʾ. Nous craignons qu’il t’arrive malheure.” Alors, le serpent gonfla sa tête en forme de boule, baṭḥāʾ, il enroula sa queue autour de cette boule, il s’énleva dans les airs et disparu dans le ciel.” Al-Azraqī, Kitāb Akhbār Makka, I, p.╯263, as cited in Atallah, 1975, p.╯166. 83 ╇ Nöldeke, 1860, p.╯415. 84 ╇ In his commentary on Ibn Hishām (ed. Wüstenfeld, vol.╯2, pp.╯41–2) ʿAbd al-Raḥmān al-Suhaylī notes that in several cases an orthodox Muslim is said to have wrapped a dead serpent in a piece of his cloak and to have buried it.

71 72

We were seated with ʿAbd Allāh b. ʿAmr b. al-ʿĀṣ in the Ḥijr; the sun having come there (the shade having contracted) the assembly rose up, when we beheld the glistening of a serpent which had come out of the gate of the Beni-Shaibah [Banū Shayba]. The eyes of the men were raised to look at it; it went the circuit of the House seven times and prayed with two bendings of the body behind the place of Abraham, when we went and said to it, ‘O thou visitor, God has ordained thy blood to be shed, and there are in our land slaves and fools of whose mischief to thee we are afraid?’ It then went away in the direction of the sky and we did not see any more of it.82

Theodor Nöldeke has associated the behaviour of these Muslims with the belief in jinn,83 probably in order to give an explanation for such vestiges of the ancient cults.84 Even so the Prophet Muḥammad insisted on regarding the serpent as

58

chapter four

an enemy of Islam,85 as shown in several canonical traditions (the so-called ḥadīth)86 such as: “We have not made peace with serpents since the time we became their enemies;”87 “whoever leaves them [the serpents] (alone) is not one of us;”88 “whoever leaves a serpent alone from fear of its revenging itself on him, has on him the curse of God, the angels and men, – all of them;”89 “whoever kills a serpent will have as it were killed a man believing in the plurality of gods, and whoever leaves a serpent (alone) fearing retaliation from it, is not one of us.”90 Yet these dictates had to be eased with regard to domestic serpents91 and it is known that Muḥammad saw in each serpent not only a maleficent but also a benign spirit,92 thus continuing to ascribe to them a certain positive power. According to tradition, it is forbidden to kill serpents that dwell in human habitation because these are beneficial jinn.93 Before killing a domestic serpent it has to be forewarned three times, or during three days,94 of the danger it faces and of the obligation of the faithful to pursue it.95 The persistent adherence to the belief that every house has its serpent guardian that is the real owner of the place – probably linked to the belief in ancestral spirits – may in some way be connected with this tradition.96 Al-Damīrī also records the interpretation of

Post-Qurʾānic traditions thus portray the (serpent-)dragon as no more than the genius loci, fulfilling solely the function of guardian.98 Yet it appears that in spite of official strictures, the cult of the serpent was abandoned only slowly under Islam, while domestic serpents, considered as beneficial jinn in their role as genius loci, continued to be tolerated. The traditions also maintain that the Kaʿba, the most famous sanctuary of Islam, was built upon a serpent. When Abraham (Ibrāhīm) wished to build the Kaʿba, the Sakīna (the Hebrew shekhīnā, “dwelling,” or “presence,” is usually considered the source for the Arabic sakīna)99 unfolded itself like a snake on the first foundations which had already been laid by Adam or the angels, saying, “Build upon me,”100 “and so he built; hence every Bedouin in flight and every powerful person Â�inevitably circumambulates the sanctuary under the Sakīna’s protection.”101 Al-Ṭabarī expounds on this tradition and describes the serpent as

In order to justify these vestiges of the ancient cults the serpent was said to represent “a believing jinn”; cf. al-Damīrī, I, p.╯233, cited after Robertson Smith, 1889, repr. 1927, p.╯444, n. 1. 85 ╇ Al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯649. Cf. Atallah, 1975, p.╯166. 86 ╇ The term ḥadīth, or “communication,” denotes codified reports that convey the normative sayings and deeds of the Prophet Muḥammad, based, according to Muslim belief, on first-hand accounts of reliable witnesses to those utterances and events; after the Qurʾān, the ḥadīth constitute the second most important basis of Islamic law. 87 ╇ Al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯649–50. Cf. Atallah, 1975, pp.╯166–7. 88 ╇ Op. cit. 89 ╇ Op. cit. 90 ╇ Op. cit. 91 ╇ Atallah, 1975, p.╯167. Cf. Massé, 1938, vol.╯1, pp.╯201–2. 92 ╇ Nöldeke, 1860, p.╯415; Wellhausen, 1897, repr. 2007, p.╯153. 93 ╇ Nöldeke, 1860, pp.╯415–6; Wellhausen, 1897, repr. 2007, pp.╯151, 164; Zbinden, 1953, p.╯76; Henninger, 2004, p.╯31. 94 ╇ Al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯650–1; cf. Atallah, 1975, p.╯167. 95 ╇ Ibn al-Athīr, Nihāya, ed. Zāwī, Cairo, 1963, “ḥarağ,” I, p.╯362, cited after Atallah, 1975, p.╯167. 96 ╇ Donaldson, 1938, repr. 1973, p.╯168; Atallah, 1975, p.╯167. These house snakes are revered also throughout the Persian-speaking world; see Russell, 1987, p.╯461. The fifthcentury Armenian author Eznik of Koghb writes for instance

that “now if there be a heathen who may think a being evil by nature, let him be opposed by the co-practitioners of his own art, the worshippers of serpents, for they now tame serpents to such a degree that they can call them into houses of talismans (yuṙt ʿiwkʿ) and offer them food, as did the Babylonians with the dragon they worshipped, but the beloved of God killed it with the same accustomed food.” Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, p.╯575, ch. 65. The dragon-killing mentioned by Eznik refers to the apocryphal book of the Old Testament Bel and the Dragon in the book of Daniel LXX in which there was giant serpent (drakōn) worshipped at Babylon which was killed by feeding lumps of a concoction of pitch, fat and hair to the dragon, causing it to burst. Cf. Gunkel, 1895, pp.╯320–3. There is a distinct possibility that ophiolatry and ophiomancy (cf. al-Bīrūnī, Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, pp.╯217–9, see Jadwal al-ikhtiyārāt (“Table of Selections”); Panaino, 2005, p.╯73–89) was practised in the Persian-speaking world. In this connection it is of interest to note that the Yezidis venerate a serpent carved at the height of a man and painted black on the wall to one side of the entrance to their holiest shrine, Shaykh ʿAdī, near Mosul in Mesopotamia. See Russell, 1987, p.╯461; Bachmann, 1913, pls. 14–6. 97 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯656. 98 ╇ Wensinck, 1916, repr. 1978, p.╯60. 99 ╇ Fahd, “Sākīna,” EI² VIII, 888b. 100 ╇ Al-Azraqī, Kitāb Akhbār Makkah, ed. Wüstenfeld, 1858–61, p.╯30, as cited in Wensinck, 1916, repr. 1978, p.╯60. Fahd, “Sākīna,” EI² VIII, 888b. 101 ╇ Op. cit.

a dream in his Ḥayāt al-ḥayawān al-kubrā which underlines the continued importance accorded to domestic serpents: If one dreams as if the serpents out of a place have disappeared, there will be an epidemic (of plague), and the mortality in that place will increase, for serpents indicate life.97

dragons and the powers of the earth “a stormy wind with two heads. One of them followed the other till it reached Mecca; there it wound itself like a serpent on the spot of the sacred house.”102 The foundation of the Kaʿba is further described as “a wind called the wind Al-Khadjūdj which had two wings and a head like a serpent’s.”103 A similar description is given by Ḥusayn ibn Muḥammad al-Diyārbakrī in his Taʾrīkh al-khamīs in which it is said to possess “two serpents’ heads, one behind the other.”104 Such traditions endow the great serpent with a sacred as well as a mythological character. In Islamic tradition, the (serpent-)dragon is thus closely associated with the foundation of the Kaʿba. It evidently has not only a mythological or apotropaic function but a sacred character, following the ancient Semitic and Iranian traditions.105 Its supernatural qualities are manifest in the winged and double-headed appearance. It is further significant, as Arent Jan Wensinck points out, that in most of the traditions the Meccan serpent is either the Sakīna or a being sent by God, hence “not a demoniac but a divine being.”106 Finally, a madman was said to be supernaturally possessed by jinn (majnūn).107 At the same time, jinn sometimes endowed men with special knowledge as seems to be implied by the word for the mantic figure of a poet (shaʿīr), who was thought to be endowed with demonic or supernat╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, vol.╯1, tr. and ed. de Goeje, 1879–1901, p.╯275, 8–10, cited after Wensinck, 1916, repr. 1978, p.╯61 and n. 2. The pre-Islamic origins of the belief in serpents as guardians of graves and foundations in general (see Wensinck, 1916, repr. 1978, p.╯60) are supported by Gustaf Dalman’s research in Petra, where he photographed a large conical stone carved in relief with a coiled serpent (1908, vol.╯1, pp.╯218–20, figs. 141–3). It was placed upon a gigantic stone block under which is a large room with niches for the reception of the deceased. 103 ╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, vol.╯1, ed. de Goeje, 1879–1901, p.╯276, 16–7, as cited in Wensinck, 1916, repr. 1978, p.╯61. 104 ╇ Vol. 1, Cairo, 1283, p.╯98, as cited in Wensinck, 1916, repr. 1978, p.╯61. 105 ╇ Cf. Wensinck, 1916, repr. 1978, p.╯60. 106 ╇ Idem, p.╯65. 107 ╇ Robertson Smith, 1889, repr. 1927, p.╯128. 108 ╇ Cf. Gohrab, 2000, pp.╯83–95, esp.╯87–93. 109 ╇ Ed. Asgharzada, A.A., and Babayev, F., Baku, 1965, ch. 20, 51–3, cited after Gohrab, 2000, p.╯87. Closely related imagery is abundantly used in the romance Wīs u Rāmīn, tr. and ed. Davis, 2008, pp.╯29, 89, 195, 209, 236, 473. 110 ╇ In an Ugaritic poem of Baal V AB:D 35–44, the goddess ‘Anat declares that: 102

Did I not crush El’s Darling, Sea? Nor destroy River, the great god?

59

ural inspiration. The close association of the jinn with serpents is also emblematised in Iranian literature. In his romantic epic of the popular legend Laylā wa Majnūn, Niẓāmī of Ganja in Azerbaijan uses ophidian imagery to depict the lovers.108 In particular Majnūn’s serpent-like appearance and his dwelling in a cave or ruin underline the fact that he is possessed by jinn: There, in that particular ruined place, He is creeping like a serpent over a stone. He is insane, in pain and distressed, and like a demon, he is far from the eye of man. Due to his wounds, his soul is pierced; the marrow of his bones can be seen.109

c.╇ The treasure-guarding dragon Just as the dragon keeps the wealth of the waters concealed, it also keeps, by implication, guard over the wealth which is concealed in the earth. The belief concerning dragons guarding treasures hidden in the earth and the sources of nature’s abundance may be linked with their chthonic nature.110 Within the sanctuary of the Kaʿba in Mecca a deep pit or well was situated, called khizāna (“treasury”), on account of the offerings of jewellery and precious objects that were thrown Nor muzzle Tannin full well? I crushed the writhing serpent, The powerful one of seven heads, ... I fought and I inherited gold. Hence by fighting one or more dragons, which apparently guarded the gold, ‘Anat obtained the precious metal. Cited after Astour, 1965, pp.╯291–2. In ancient Greek lore the great dragon (drakōn) who guards some treasure or holy site was an important theme, and either protects its holdings as its own possession or was placed there by the owner of the site as its guardian. The subterranean golden apples located in the uttermost parts of the earth in the garden of the Hesperides were guarded by the Nymphs, daughters of Atlas, and the drakōn (Hesiod, Theogony 333–5); the oracle at Delphi was protected by the female dragon (drakaina) Pythōn that was killed by Apollo; the spring at Thebes was watched over by a drakōn killed by Kadmos, and the golden fleece was guarded by a drakōn killed by Medea or Jason. Euripides (Herakles 397, tr. after Buschor, Munich, 1952) calls the treasure-guarding serpent the “reddish drakōn, the formidable coiled gurdian.” Plato notes that owing to his sharp sight, the dragon guards hidden treasures as well as temples and oracles (Phaedrus, 4.20.3–4); cf. Merkelbach, “Drache,” RAC IV, 1959, pp.╯226–7. For post-biblical Jewish sources describing the Genesis serpent as possessing “silver, gold, gems, and pearls,” before the Fall, see Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯1, p.╯71. The treasure-guarding serpent is also known in Indian lore, such as the fable in the

60

chapter four

into it and preserved there.111 This was probably done with the intention, as Wensinck suggests, of appeasing the god of the netherworld.112 Islamic tradition relates how in the period shortly before the rise of Islam, at a time when the ancient Arab tribes of the Jurhum controlled the Kaʿba, these treasures were stolen from the sanctuary.113 Thereupon, according to the Kitāb Akhbār Makkah (“Book of Information on Mecca”) assembled by al-Azraqī, the ninth-century historian of Mecca and its sanctuary,114 God sent “a serpent which had a black back and a white belly and a head like the head of a he-goat; this serpent guarded the sacred House [and its treasure] during five hundred years.”115 The serpent is called a “large serpent” and has its dwelling place in the pit of the Kaʿba, where it guards the treasures.116 It is fed by daily food offerings thrown into the sacred pit by the faithful and basks in the sun on the stones of the Kaʿba.117 This is further elaborated in the biography of the Prophet Muḥammad by the Arab author ʿAlī ibn Burhān al-Dīn al-Ḥalabī who writes that “Allāh sent a white serpent with a black head and a black tail and its head was like the head of a he-goat. It had to dwell in this pit in order to guard the objects there.”118 Al-Azraqī notes that as a result the treasure of the Kaʿba

held all the votive offerings, including arrows used for divination, jewellery for adorning idols, and gold,119 and also that Muḥammad uncovered a large amount of gold in the well.120 During the preIslamic period, sacred serpents were also known to have guarded the sanctuaries at Nakhla where the cult of the goddess al-ʿUzzā prospered as well as at Dawmat al-Jandal where the god Wadd was served.121 These sacred serpents were also fed at the sanctuaries, often with milk.122 According to the tenth-century Arabic geographer Abū Muḥammad al-Ḥasan ibn Aḥmad al-Hamdānī (d. 334/945–6), a treasure lay hidden near the high mountains Ḥufāsh and Milḥān (Rayshān) in south Arabia. Many Arabs sought this treasure but they were always prevented from reaching it by a serpent in the shape of a high mountain which barred their approach.123 The indirect association of the serpent with wealth in the interpretation of a dream is recorded by al-Damīrī in his€Ḥayāt al-ḥayawān al-kubrā according to which:

Panchatantra of the hooded serpent in the anthill who daily bestows a gold piece on the poor Brahmin Haridatta (Vogel, 1926, pp.╯173–4) or the legend of the gold-guarding black serpent of Rājagá¹łiha who was subdued by the Buddha (Avadāna Śalākā IV, 1, ed. Spencer, vol.╯2, pp.╯289–91, as cited by Vogel, 1926, p.╯21). Al-Bīrūnī records an episode from the Saṁhitā in which Varuṇa’s son, Agastya, devoured the water of the ocean so that it disappeared, revealing the lower parts of Mount Vindhya:

pp.╯12–3, 85, as cited in Wensinck, 1916, repr. 1978, p.╯63. Cf. Atallah, 1975, p.╯164. 116 ╇ Ed. Wüstenfeld, 1858–61, p.╯170, 10, cited after Â�Wensinck, 1916, repr. 1978, p.╯63. On the account of the Â�serpent in the well of the Kaʿba, sent by God to guard its treasury in the time before the Kaʿba was secured with doors and a lock, see Canova, 1994, pp.╯421–5. 117 ╇ Atallah, 1975, p.╯164. The role of the paternal grandfather of the Prophet, ʿAbd al-Muṭṭalib, in providing security for the treasury of the Kaʿba is also of relevance with regard to the possible existence of a serpent cult at the site in pre-Islamic times. Cf. Nöldeke, 1869, p.╯416; Fahd, 1968, esp.╯pp.╯40–1; Atallah, 1975, p.╯166. 118 ╇ Al-Sīra al-Ḥalabiyya, Cairo, 1292, vol.╯1, p.╯189, 3–5, as cited in Wensinck, 1916, repr. 1978, pp.╯63–4 and n. 1. 119 ╇ Kitāb Akhbār Makkah, pp.╯119–24. 120 ╇ Idem, p.╯124; al-Qāḍī al-Rashīd ibn al-Zubayr, Kitāb al-Hādāya wa ’l-Tuḥaf (“Book of Gifts and Rarities”), tr. Ghādah al-Ḥijjāwī al-Qaddūmī, no. 175. 121 ╇ Atallah, 1975, p.╯165. 122 ╇ Idem, pp.╯164–6. The notion that serpents are drawn to milk appears also in the tale of the “Queen of the Serpents” included in Alf layla wa-layla where the Serpent-Queen is enticed into a cage by bowls of milk and wine (The Book of the Thousand Nights and a Night, tr. Burton, 1885, vol.╯2, pp.╯594–6). 123 ╇ Ṣifat Jazīrat al-ʿArab, (“Description of the Arab Peninsula”), ed. Müller, Leiden, 1884, p.╯68, as cited in Schleifer, “Ḥufāsh,” EI² III, 548b. 124 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯655.

...whilst the makara and the water animals were clinging to it. They scratched the mountain till they pierced it and dug mines in it, in which there remained gems and pearls. Kitāb fī Taḥqīq mā li-l-Hind, tr. Sachau, 1887, p.╯93. The simile of the pearl-guarding monster similarly appears inWīs u Rāmīn, tr. and ed. Davis, 2008, p.╯473. 111 ╇ Al-Azraqī, Kitāb Akhbār Makkah, ed. Wüstenfeld, 1858–61, pp.╯41, 14–6; 49, 1; 106, 14–6; 111, 11; 169; 171, 4; as cited in Wensinck, 1916, repr. 1978, pp.╯29–30. Atallah, 1975, p.╯164. 112 ╇ Wensinck, 1916, repr. 1978, p. 30; see also Wellhausen, 1897, repr. 2007, p.╯103; Atallah, 1975, p.╯164. 113 ╇ Al-Azraqī, Kitāb Akhbār Makkah, ed. Wüstenfeld, 1858–61, p.╯48, 8–10, cited after Wensinck, 1916, repr. 1978, p.╯63. 114 ╇ The book was put together before 251/865, although it also includes references from as late as 310/922–3; cf. Grabar, 1985, p.╯2 and n. 12. 115 ╇ Ed. Wüstenfeld, 1858–61, p.╯48, 8–10, German tr. vol.╯4,

He who dreams of being bitten by a serpent, in consequence of which the bitten part is swollen, will obtain great wealth, for the poison indicates wealth and the swelling indicates an increase in it.124

dragons and the powers of the earth

61

The motif of a black snake as positive augury also appears in the Shāh-nāma when the young Â�Buzurjmihr, the later minister of Khusraw Anūshirwān, was breathed on by a black snake which was interpreted by his companion who Â�witnessed the scene as a sign that Buzurjmihr would attain a position of great power.127 The association of the colour black with the concept of royalty is interesting and may perhaps be related to the black banner of the ʿAbbasids (black being supposedly the colour of the ProÂ� phet’s banner) but could also be a vestige of

conceptualisations that were introduced by the Turko-Mongol dynasties for whom the colour carried exalted associations.128 In his romance Haft Paykar, which was completed in 593/1197, the poet Niẓāmī Ganjawī refers to “dragon-like locks on treasures rested.”129 Surviving examples of such dragon locks are rare.130 However dragon knockers that have a part attached to a door, and another part linked to it by a hinge that may be lifted and used to strike a plate fitted to a door are used to gain entrance to a monument. Well-known examples are, for instance, the copper alloy knockers discussed below that adorned the doors of the Great Mosque (Ulu Cami) of Cizre (the former Jazīrat ibn ʿUmar), now in the Türk ve İslam Eserleri Müzesi in Istanbul (figs. 82 and 83). The motif of the treasure-guarding dragon appears also in the first part of Niẓāmī’s prose romance Iskandar-nāma (“Alexander Romance”), called Sharaf-nāma (“Book of Honour”). He describes how a certain Balīnūs, whose name is the Arabic version of Apollonius (of Tyana), a renowned sage and magician of the first century ad and author of several handbooks on magic, accompanied Iskandar, represented as an Islamicised Graeco-Iranian hero,131 on his conquest of

125 ╇ Idem, p.╯656. It is interesting to compare this with the Zoroastrian practice of taking snake omens as recorded in the New Persian text, the Mār-nāma (“Book of Snakes”) included in the Persian Riwāyats of Dastūr Dārāb Hormazdyār that date from 1679; one of the verse-lines stating, for instance, “If you see a snake on the day of Hormazd, your honour, property and income will increase.” See Panaino, 2005, p.╯79. 126 ╇ Idem, p.╯656. 127 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯6, pp.╯247–9, ll. 1037–1045. 128 ╇ The adjective qara, literally “black,” designates the prime compass point of the north, the place of origin of the Altaic tribes, hence acquiring the meaning “principal,” “chief” (in contrast to other compass points which are also described in terms of colour such as ak (“white”) indicating the south). Cf. Pritsak, 1950–5, pp.╯244–5, 255. Qara was used in the onomastic of the Turkish tribes (for instance, the eighth-century seasonal camp of the eastern Turks, Qara Qum; the Uighur capital from 744 to 840, Qarabalghasun; the southern Uighur capital after 840, Qara Xocho; and the first Yuan capital, Qaraqorum) and as an epithet to the names of Xiongnu and Turkic rulers. There are also the Â�Qara-Qitai (“Black Qitan”) who after the fall of the Liao empire ruled over nearly the whole of Central Asia from the Oxus to the Altai mountains until 1175 and the Qarakhanids that ruled in the lands of Central Asia straddling the T’ien-shan mountains from the tenth to the early thirteenth centuries. It also carries the meaning “strong, powerful” (Kramers, “Ḳarā,” EI² IV, 572b) as evidenced by its use in personal names of the Islamic period, such as the name of the Artuqid ruler Abū ʿAbdallāh Muḥammad ibn Qara Arslan (“black lion”) ibn Dāwūd ibn Suqmān (561/1166–581/1185). 129 ╇ Tr. Meisami, 1993, p.╯221.

130 ╇ Such a lock is featured on a gilded and nielloed silver casket, now in the Treasury of St. Mark’s, Venice. Thought to have been either in crusader possession (Marshak, 1986, pp.╯119–20 and n. 89, figs. 163–7) or the product of a southern Italian or Spanish workshop (Nobilis Officinae, 2005, pp.╯171–2, cat. no. 13; Venise et l’Orient, p.╯123, cat. no. 93), the casket has been variously dated from the twelfth to the fourteenth century: Anna Schwinger dates it to the twelfth century, or alternatively, proposes it as a thirteenth- or fourteenth-century reproduction (Nobilis Officinae, 2005, p.╯171, cat. no. 13), while Boris Marshak (1986, p.╯435) has suggested to date it to the second half of the thirteenth century. Most recently Francesca Leoni has proposed that the casket originates from southern Italy, or possibly Spain, and should be dated to the end of the thirteenth or the early fourteenth century (Venise et l’Orient, p.╯123, cat. no. 93; catalogue entry by Leoni). The flat lid is framed at the sides by epigraphic bands in Kufic, flanking two seated musicians whose depiction conforms to the usual Islamic courtly iconography. The decoration on the front of the casket includes two large interlaced ornaments each enclosing a central cross-shape, suggesting that it may have been made for the Western market. One of the sides depicts a human-headed harpy with a tail terminating in a large dragon head with gaping mouth. Of particular note are the three extremely finely worked yet prominent hinges that terminate in two confronted dragon heads, their wideopen mouths touching at the tips. The lock, which ends in closely related dragon heads, fastens just above a pair of lion heads. 131 ╇ In spite of mythologising features that go back to the Pseudo-Callisthenes, it is noteworthy that medieval Islamic authors differentiated between the legendary and

In another interpretation, the dreamer is not bitten by a serpent but merely dreams of a smooth serpent that he can take wherever he likes, which is interpreted as a sign that the dreamer will become wealthy: If one dreams of possessing a sleek (smooth) serpent which he has taken wherever he likes, [he] will obtain riches and become prosperous.125

Moreover a dream of a black serpent (instead of a smooth serpent) is said to indicate the acquisition of power and governance: He who dreams of possessing a black serpent will acquire a kingdom and government.126

62

chapter four

Iran. Through his talismanic powers Balīnūs helped the king to defeat Āẓar Humā, the powerful priestess of a fire temple, who in the form of a dragon guarded the holy fire, in other words, the treasure of the temple.132 The same motif is extensively used in Niẓāmī’s Haft Paykar, considered in chapter 14, and also in his Laylā wa Majnūn. When the news of Laylā’s death reaches Majnūn, he hastens madly to her grave and: ...rolls in the same way as a serpent, Or a worm coiling beneath the earth, With a thousand toils, he coiled himself around The grave-stone like a serpent on a treasure.133

Even in her death Majnūn continues to keep watch over the grave of his beloved, Laylā, like a serpent guarding a treasure: The cavern is always the home of a snake; O Moon, why has the cavern become your residence? I will lament in your cave for you are my beloved; you are the companion of the cave (yār-i ghār), how can I not weep? You turned out to be a treasure in the earth; If you are not a treasure, why are you then thus [in the earth]? Every treasure, which is in a cave, has a serpent upon its skirt. I am a serpent which is the sentinel of your grave watching the treasure on your grave.134

Serpents as guardians of treasures are a widely known topos in folk tales, as recorded in the Alf layla wa-layla, for instance the giant fire-spitting serpent guarding Solomon’s magic ring in the story of the adventures of Bulūqiyā inserted into the story of the Queen of the Serpents.135 The story of the Valley of the Diamonds, which recurs in many eastern sources, associates serpents with diamonds. It is found in the legend of the Hellethe real figure of the Macedonian conqueror, as evidenced by al-Bīrūnī’s Kitāb al-Āthār (tr. and ed. Sachau, 1876–8, p.╯127) in which he comments upon the catastrophic consequences of the historical Alexander’s invasion in 330 bc which led to the extinction of the Achaemenid empire. 132 ╇ Dīwān, pp.╯974–5, cited in Pseudo-Apollonius of Tyana, tr. and ed. Weisser, 1990, p.╯27. The theme of Alexander as dragon-slayer appears in the Syriac version of the Pseudo-Callisthenes, ch. 7, tr. and ed. Budge, 1889, repr. 2003, pp.╯lxxiv, 107–9, as well as in the Ethiopic version, idem, p.╯ciii. 133 ╇ Ed. Asgharzada, A.A., and Babayev, F., Baku, 1965, ch. 62, 42–3, cited after Gohrab, 2000, p.╯90. 134 ╇ Idem. Cf. idem, 2003, p.╯83. 135 ╇ Marzolph and van Leeuwen, 2004, p.╯130. 136 ╇ Laufer, 1915, p.╯20. 137 ╇ Das Steinbuch, tr. and ed. Ruska, 1912, pp.╯14–6.

nistic Orient represented by De XII gemmis, the tale of the fourth-century bishop of Constantia (Cyprus), Epiphanius,136 as well as by Pseudo-Aristotle.137 It is retold in the Alexander Romance138 and also recorded by al-Bīrūnī in his Kitāb al-Jamāhir fī maʿrifat al-jawāhir (“Comprehensive Book on the Knowledge of Precious Minerals”),139 by Niẓāmī in his Iskandar-nāma,140 by the Arabian mineralogist Aḥmad al-Tifāshī (580/1184– 651/1253),141 in the collection of travel narratives of Sindbād in the Alf layla wa-layla142 as well as later in the travelogue of the merchant explorer Marco Polo (c. 1254–c. 1324).143 It recounts that in the east on the extreme frontier of Khurasan there is a deep valley in which diamonds lie, guarded by poisonous snakes upon which no man can gaze without dying; however this power endures only so long as the serpents live. On his way to India Iskandar, the Dhu ’l-Qarnayn (“two-horned”) of the Qurʾān (sūra 18, 82–98), is said to have ventured into such a diamond pit. He outwitted the serpents by holding up mirrors as shields in which the serpents saw their own reflection and died.144 The association of diamonds with serpents may also explain the belief in the toxicity of diamonds first advocated in the Pseudo-Aristotle, which cautions its readers “against taking the diamonds into their mouths, because the saliva of the snakes adheres to it so that it deals out death.”145 The association of dragons with gems is also recorded in Armenian epic lore, as related by the eleventh-century Armenian scholar Grigor Â�Magistros (990–1058) in a letter addressed to prince Tʿornik Mamikonean: A fish called Azhdahāk (i.e., Azhi Dahāka) ... gave a concubine a huge pearl... she took it to the king, who had it set into a crown ... The king then ordered that the gods be honoured with rich offerings, and that this fish be carven, together

╇ Stoneman, 1991, p.╯2. ╇ P.╯99, 11–102, 5, as cited in Ullmann, 1994, p.╯107. Cf. Ruska, “Almās,” EI¹ I, p.╯313; Ruska [and Plessner], “Almās,” EI² I, 419a. 140 ╇ See Laufer, 1915, pp.╯10–1. 141 ╇ Raineri Biscia, A., Fior di pensieri sulle pietre preziose di Ahmed Teifascite, Bologna, 2nd ed., 1906, pp.╯21, 54. In this version the serpents are “able to swallow an entire man (“inghiottiscono un uomo intero”),” cited after Laufer, 1915, p.╯13 and n. 1. 142 ╇ Marzolph and van Leeuwen, 2004, p.╯385. 143 ╇ Laufer, 1915, pp.╯10–21, esp.╯pp.╯11, n. 2, 14, and 18, n. 1. 144 ╇ Idem, pp.╯10–21. Ullmann, 1992, p.╯111. Cf. Ruska, “Almās,” EI¹ I, p.╯313; Ruska [and Plessner], “Almās,” EI² I, 419a. 145 ╇ Laufer, 1915, pp.╯40–1. 138 139

dragons and the powers of the earth with the effigies of the other divinities, and that sacrifices be made on the banks of the river Phison, where it had appeared to the concubine.146

d.╇ The dragon and vegetal compositions An instance of the association of the serpent with a tree has already been noted in the figure of the Rigvedic primordial “serpent of the deep,” Ahi Budhnyà, which resides at the roots of the tree. This may be related to the notion of the giant serpent as custodian of a tree, often represented as stylised vegetation, which became an important iconographical motif, reflected in particular on the decoration of portable objects found throughout the Central Asian regions. The prevalence of this theme in these regions is corroborated in the depiction of a fork-shaped harness ornament, datable between the sixth and the eighth centuries, which was excavated from a tomb at Kebinai in southern Siberia (fig.╯38). The pendant is cast in openwork with paired antithetical dragons emerging from a stylised vegetal composition that springs from a horned mask-like head at the bottom. Long-necked regardant dragon protomes with deep-set eyes, lupine or canine snouts and feathery wings, springing from a central foliate stalk and emitting or biting the leaves are shown on a belt fitting serving as hinge joint, cast in openwork. The fitting was unearthed in Novocherkassk in the lower Don region and is attributed to the second half of the eighth century (fig.╯39).147 A Byzantine attribution has been suggested on the basis that Byzantine workshops are thought to have existed in the proximity.148

╇ Tchukasizian, 1964, p.╯325. ╇ Cf. Daim, 2000, p.╯128, fig.╯51. 148 ╇ Idem, p.╯130. The paucity of early medieval records of this important visual element from the wider Central Asian world warrants a brief geographical digression. It is interesting to point out that the motif also exists on eighthcentury Lombard and Frankish art such as the sarcophagus of Abbess Theodota [sic], Pavia, I. 720–30 (Atroshenko and Collins, 1985, p.╯139, fig.╯88; Daim, 2000, drawing on p.╯319, fig.╯13.3), on the Baptistry of Callisto (so-called Sigualdus), dated 762 to 776, in the Museo Archeologico Nazionale, Cividale del Friuli (Daim, 2000, p.╯128, fig.╯52, drawing on p.╯319, fig.╯13.2), and on the eighth-century silk-embroidered white linen tunic of the Frankish queen, Saint Bathilde of Chelles, which is thought to have been made at the Merovingian court under Byzantine influence (Vida, 2000, p.╯312, fig 6.2; cf. Vierck, 1978, pp.╯521–64). The iconographic expression was probably transmitted to Europe in the wake of the great 146 147

63

In the Central Asian world the iconography is further recorded on a small early medieval ancestral fire altar (chirāgh khāna) from Samarqand.149 The niche, which formed the nucleus of a house, is decorated on both end panels with a stylised tree-like vegetal composition flanked by a pair of perching birds with long florid tails, probably representing peacocks. Springing from the trunk of the tree are arched regardant Â�dragons that flank a large central crescent-shaped bracket, above which flutters a pair of small birds (figs. 40 a and b). After the destruction of the fire temples following the advent of Islam, these chirāgh khāna, miniature imitations of fire temples, were placed as small altars in private houses for the luminaries that manifest the presence of the eternal fire. A small fire was lit in the niche, the denomination chirāgh khāna translating as house of the lamp, in other words a domestic “fire temple.” Together with the dragons that spring from the central tree this altar contains images of roosters. The rooster motif is a well-known visual element harking back to Sasanian and early Islamic iconographies. The idea of the rooster (Av. parō. dərəs-, “he who foresees [dawn]”) as a sacred bird or a bird endowed with special powers was characteristic of Iranian beliefs. In the Avesta it is considered as the bird of Sraosha, the angel who defends mankind from demons.150 Generally in Zoroastrianism it is regarded as an ally of the powers of light and goodness, as an enemy to evil and demonic beings, and as a symbol of royalty and true religion; thus in the Vidēvdāt Ahura Mazdā says of the rooster: That bird lifteth up his voice at the mighty dawn (saying), ‘Arise, O men, laud Best Righteousness, contemn the demons’.151

migrations of nomadic peoples from the Eurasian steppes such as the Avars, subjects of the Turks, who fled to the north Caucasus region in 558 and then migrated westwards into the province of Pannonia (which lay to the east of northern Italy), where later waves of Avar migrants continued to join them (cf. Pohl, 1988, pp.╯28–9; Barthold and Golden, “Khazar,” EI2 IV, 1172a). The motif of dragon protomes projecting from vegetal compositions that is found on eighth-century Lombard and Frankish art may thus have been inspired by Central Asian prototypes. 149 ╇ Rempel’, 1983, p.╯127, fig.╯49.5; idem, 1987a, p.╯103, fig.╯42a; and idem, 1987b, pl.╯LI, fig.╯11, altar to the left side. 150 ╇ Vidēvdāt 18.14–5; Gershevitch, 1959, p.╯62. In Jewish post-biblical scriptures (Yoma 21a) the crowing of the cock is likewise known to drive demons away, see Ginzberg, 1909– 38, repr. 1946, and idem, 1955, vol.╯5, p.╯173, n. 16. 151 ╇ Vidēvdāt, Fargard 18.2.14 (The Zend-Avesta, tr. DarÂ� mesÂ�Â�teter, 1880); Bundahishn 19.33 (Sacred Books of the East,

64

chapter four

In antiquity, the religious significance of the cock as sacred bird of the sun whose rising he announces is described by the second-century Greek geographer Pausanias.152 In medieval Islam it was similarly associated with light and Paradise, as well as with royalty, but also with the magicianking Sulaymān ibn Dāwūd who was acquainted with the speech of birds and animals (Qurʾān 27, 16–9).153 In his Kitāb ʿajāʾib al-makhlūqāt al-Qazwīnī notes a prophetic ḥadīth according to which God placed under the divine throne a rooster whose wingspan extended from east to west. At the approach of dawn the rooster would beat his wings and cry out loud: “Glory to God, the Almighty!” At this all the roosters in the world would respond in similar fashion, praising God by beating their wings and crowing.154 A similar type of imagery as the one on the chirāgh khāna can also be found in a Buddhist context in Chinese Turkestan (present-day Xinjiang). When the near one-hundred-year steppe empire of the Uighurs in Mongolia (744–840), which had essentially been a continuation of the Turkic qaghanate, was destroyed in 840, one group fled westward into the region of Chinese Turkestan, where they ultimately founded two kingdoms. Here the Uighurs became the culture bearers of Central Asia.155 On a wall painting from Cave 19 in Uighur Bezeklik in Qocho (present-day Turfan), datable between the ninth and the eleventh centuries, a pair of dragons in “Central Asian Chinese-style” springing from a water pool are closely associated with abundant vegetation. The confronted snarling dragons ascend in upright position, their tail ends entwined. Dramatically presented with their raptor-like legs raised and extended talons unsheathed, they guard between them a treasure, shown springing from the lotus throne and floating over a flaming pearl, the Buddhist ruyi baozhu (Skr. cintāmaṇi),156 from

which sprout the formidable, profusely reticulated branches of a tree. Its luxuriant growth is nourished by the fructifying waters of the pool in which small dragons cavort (fig.╯41).157 A pair of knotted dragons features in a related depiction of the same period from a now destroyed wall painting, also from the cave monastery of Bezeklik, recorded only by a drawing made by the German archaeologist Albert Grünwedel during the third German expedition to Chinese Turkestan in 1906 to 1907 (fig.╯168).158 A similar conceptualisation of the dragon and the tree is emblematically portrayed in an Islamic context as part of a large water-clock automaton. It is described in an illustrated horological treatise attributed to Archimedes of which several medieval Arabic versions survive.159 The second section of this treatise gives an account of a tree with silver birds set between two mountains behind which serpents hide:

tr. West, E.W., Oxford, 1897); see also The Zohar, vol.╯4, p.╯369 (Wayikra 22 b) (tr. Simon, M., New York, 1949) in which the cock calls upon men to praise God and study law; cited after Bonner, 1950, p.╯125 and n. 11. Cf. Harper, 1978, p.╯65; Schwartz, 1985, p.╯661. 152 ╇ Graeciae Descriptio V 25, 9. Cf. Bonner, 1950, p.╯127. 153 ╇ Viré and Baer, “Ṭāwūs, Ṭāʾūs,” EI2 X, 396a. 154 ╇ Tr. and ed. Giese, 1986, p.╯219. 155 ╇ Golden, 2004, p.╯22. 156 ╇ The pearl (Skr. maṇi) was a symbol of the Buddha and of law. In Sino-Indian traditions it was “a wishing jewel, granting the desires of its possessors.” Moreover it was associated with the moon. The “dragon-pearl” symbolises the full moon, the Buddhist ruyi baozhu (Skr. cintāmaṇi), the wish-

fullfilling jewel of the Indian nāgas, which symbolises transcendent wisdom. See Schafer, 1963, pp.╯181, 243. For the reception of the motif consisting of three points surmounting two wavy lines, called çintamani in Persian and Islamic art, see Soucek, “Cintāmani,” EIr, and Kadoi, 2006–7, pp.╯33–49. 157 ╇ This mural painting, which was taken to Berlin by the German expedition, is presumed to have been destroyed during the bombing of World War II. 158 ╇ Grünwedel, 1912, fig.╯590. 159 ╇ Hill, 1976; cf. idem, 1984, p.╯230. 160 ╇ Idem, 1976, pp.╯30–1, figs. 14, 15. See also idem, 1984, pp.╯229–30, fig.╯12.2 for the design of the water machinery from Archimedes. 161 ╇ Idem, 1984, p.╯230.

We will now describe how a tree is made, upon whose branches are birds. This tree is to be fixed vertically between two mountains, with a fingerlength between each mountain and the stem of the tree. When an hour elapses then a snake emerges from a hole in the mountain’s foot. The hole is inside the floor of the tree or opposite the birds that are on the branches of the tree. When the two snakes emerge the birds shriek and whistle as long as the snakes are visible. Then the snakes go back inside their holes and the shrieking and whistling of the birds ceases.160

While parts of the Pseudo-Archimedes treatise are first mentioned in Islamic sources of the tenth century, this type of clock seems to belong to a much earlier tradition and was known to have been built in Sasanian and Byzantine times.161 During the medieval Islamic period, automata in the form of silver gilded trees with artificial singing birds were known as expressions of imperial

dragons and the powers of the earth

65

might, particularly in ʿAbbasid times;162 slightly later they are also recorded as having surrounded the throne of the Byzantine emperor.163 The motif of the tree with serpents is rendered on a two-dimensional medium, a polychromeenamel overglaze-painted ceramic bowl (so-called mīnāʾī ware) from Rayy, datable to the second half of the twelfth century. Here a central tree springs from a small mound at the base of the bowl. At mid-section the tree trunk is separated by a mound-like conical motif around which a dragon appears to be coiling and from the tip of which emerges the giant upward oriented gaping jaws of a dragon with bifid tongue. The trunk grows out of this creature’s maw into a treetop with lush foliage topped by a pair of perched birds. The scene is flanked on either side by a rider and in turn by a figure seated cross-legged (fig.╯42).164 An analogous motif is represented by the wellknown relief sculptures that decorate the stone minaret buttresses of the Çifte Minare madrasa at Erzurum, probably built just before the fall of the city to the Mongols in 640/1242–3.165 Set within an ogival arch-shaped niche surrounded by a triple torus frame, the composition was intended to appear at either side of the main entrance of the madrasa, but only the relief to the right was completed.166 The sculptures are carved to form a pair of dragon protomes that spring from the stem of a central palm tree with a double-headed eagle at its summit (fig.╯43).167 The scaly upward-curving serpentine necks extend into large, upward oriented heads with large, almond-shaped eyes and small, pointed ears, their open jaws with outward curled tips replete with rows of teeth and tongues with bifid tips. Just behind the heads, the slender necks are embellished with guilloche chains, a feature

they share with the dragons that flank a central bovine head on the city walls of Ani (fig.╯130) and the dragons that frame the large, composite star-rosette from Konya (fig.╯154). The dragons’ curving ophidian bodies are once looped, then join to grow upwards into the stemmed vegetation above, while an inverted stemmed palmette springs from the interstice created below the point of juncture of the bodies. Out of the large, crescent-shaped bracket which encloses the stem grow eight large “palmate” leaves that fan out symmetrically and bear what appears to be fruit, probably pomegranates; on the tips of the side fronds (second leaves from the bottom) perch small birds. The pomegranates, anomalous in the context of a palm tree, link the tree with the idea of fertility and immortality – both in itself and as a dwelling place for birds. Öney interprets the dragons as chthonic symbols (in contrast to the solar/light symbol of the double-headed bird) which guard the tree.168 Similarly, Otto-Dorn reads this form of imagery in the light of astrological lore, according to which the dragons threaten the luminaries represented by the crescent-shaped bracket and the doubleheaded eagle.169 The serpent at the foot of the tree seems to have been a well-known conceptualisation in medieval Islam, as is evidenced by a saying recorded by the fourteenth-century encyclopaedist al-Damīrī and attributed to ʿAbd Allāh ibn ʿAmr ibn al-ʿĀṣ, a contemporary of the Prophet Muḥammad, in which he compares himself to the serpent at the foot of the tree:

162 ╇ The wonders contained in the newly built caliphal palace, such as an automaton in the form of trees, are reported in al-Khātib’s account in the History of Baghdad of the arrival of Byzantine ambassadors to the ʿAbbasid capital in 305/917–18. They described seeing a silver tree in a large pond with silver or gilded whistling birds perched on its branches. Idem, p.╯205 and n. 17. 163 ╇ In the Byzantine treatise of “The Throne of Solomon” in Constantinople, the throne of the emperor is depicted as surrounded by automata which included a jewel-encrusted tree with artificial singing birds, roaring lions and moving beasts, a description which is corroborated by Liutprand, Bishop of Cremona, who visited Byzantium as envoy in 336/948 and again in 355/966. Cf. idem, p.╯205. The tree stood in the chrysotriclinium before the emperor’s throne; outside there was a fountain with a silver eagle that covered the conduits by holding a serpent in its talons. Ebersolt, 1923, pp.╯55–7. See also Soucek, 1997, p.╯405 and ns. 13–8 with further references.

164 ╇ İnal, 1970–1, fig.╯17. The bowl does not seem to have been published elsewhere; its overall condition and the extent of possible overpainting which could have a bearing on the imagery are not known to the author. However, in spite of these uncertainties, the bowl was nonetheless included as it represents a rare example of this iconoÂ�graphy. 165 ╇ See p.╯31, n. 87. 166 ╇ Above the tree in the right hand panel is a doubleheaded eagle, the equivalent part in the left hand panel is left unfinished. On the side faces of the portal are similar compositions but without dragons. See also Gierlichs, 1996, pl.╯12.1, 4, 5. 167 ╇ Cf. idem, pl.╯12.1; Öney, 1969a, p.╯208 and fig.╯32, and eadem, 1969b. 168 ╇ Öney, 1969a, p.╯208. 169 ╇ Otto-Dorn, 1978–9, pp.╯133–4. 170 ╇ Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯634.

I support [equally well] good and evil, and am â•… inexorable, Like the serpent at the foot of the trees. 170

66

chapter four

The close associtation of the tree and the serpent is also evident in the Risālat of Ibn Faḍlān, secretary of the caliphal mission to Volga Bulgaria from 309/921 to 310/922, that relates the story of a tree being cut to size which then begins to move and crawl away in the form of a giant dragon.171 Given the historical relations of Armenians with the principality of Erzurum (Karin), which at various times belonged to Armenia and under Islamic rulership also had a Christian Armenian population, it is interesting to relate the Çifte Minare madrasa motif to earlier as well as contemporary Armenian references preserved in manuscript illumination. A significant variant on the imagery is presented in a collection of sermons of 1216 illustrated by Barsegh Kʿahanay (the priest) in Skevra monastery, Cilician Armenia (fig.╯44). It portrays a tall multi-branched tree bearing large foliage, buds and fruit that are pecked at by different types of birds that perch on the branches. Yet instead of growing out of the base of the tree and rearing up from there, the interlaced slender trunks of the tree are here shown to be enclosed by an entwined pair of giant scaly serpents forming a double loop.╯The motif represents an example of the standard depiction of the tree or pole together with the serpent forming the figure of eight (both open and closed) as also repesented in the symbol of the caduceus (kerykeion) discussed below. The reptiles gnaw, with heads upreared, at the uppermost leaves. A small bird is perched on the summit of the tree while a pair of confronting human-headed birds, probably harpies, flanks the base of the trunk. The latter bears an intricate interlace terminating in pendant palmette-shaped buds. The tree-withserpents motif certainly alludes here to a complex of ideas, not least the serpent and the fruit-bearing tree in the Garden of Eden.172 Painted about three decades before the construction of the Çifte

Minare madrasa, the miniature thus documents a variation of the motif in a Christian context.173 A large marginal ornament from an Armenian Gospel of Luke illuminated in 1323 shows openmouthed dragons with projecting tongues that touch the base of a more stylised, tree-like vegetal motif. The latter contains the symbols of the four Evangelists set amidst foliage that tapers to an arched apex. It is surmounted by an ornate cross flanked by clusters of grapes instead of the birds that top the tree-like compositions on the collection of sermons of 1216 and the mid-thirteenthcentury Çifte Minare madrasa. The ornament was illustrated by the miniaturist Tʿoros Taronatsi in the important scriptorium of the monastery of Gladzor in southern Armenia (fig.╯45).174 In view of the close analogies between the dragon compositions on the façade of the Çifte Minare madrasa and contemporary Armenian stone carvings, Rogers has referred to the reliefs as “barely islamicised versions of Armenian khatchkʿars (commemorative cross-stones).”175 This is particular evident on vishap-type Armenian Christian sepulchral steles or commemorative cross-stones (khatchkʿar), such as a twelfth- or thirteenth-century example from Makravankʿ in Ararat province, on which a closely related form of imagery can be deduced. On this khatchkʿar a pair of stylised dragons springs from the base of a cross, whence their serpentine bodies bifurcate, form a loop and then curve upwards to terminate in stylised heads with wide-open jaws. Just below the heads, ornamental bands enclose collars in the form of a figure of eight which accent the ophidian necks and delineate the bodies, the latter being enlivened by parallel decorative stripes. The bases of the small crosses are shown to rest on the tips of short tongues projecting from the dragons’ mouths whose lips end in inward-curling tips (fig.╯46).176

171 ╇ Ed. Dahhān, S., Damascus, 1959, pp.╯127–8 (fol. 4 206 wāw), as cited in Montgomery, 2006, p.╯72. 172 ╇ A reference to the tree is also found for instance in the Proverbs of Solomon: “A Tree of Life is wisdom for those who acquire it” (3.18). On the meaning of the Tree of Knowledge and the Tree of Life, see Agathangelos, Teaching of St. Gregory, tr. and ed. Thomson, 1970, ch. 277; The Armenian Commentary on Genesis Attributed to Ephrem the Syrian, tr. Mathews, 1998, pp.╯21–2, 27–8, 32–3, 37 and n. 116, 39; see also Wallace, 1985, pp.╯101–32. 173 ╇ I am grateful to Professor Robert Hillenbrand for pointing this out to me. 174 ╇ For a closely related example, see British Library Ms. Add. 15,411, fol. 92a, dated 1321, which is also illustrated at the monastery of Gladzor in southern Armenia by Tʿoros

Taronatsi; see The Christian Orient, 1978, pl.╯13 (caption 119). The same conceptualisation can be observed in an eleventh- or twelfth-century Georgian illustration of the Gospel book from Ghelati near Kutaisi in northwestern Georgia which shows a tree-like composition from the base of which projects a root-like vertical extension that carries a single curved serpent with gaping mouth revealing a bifid tongue. Georgia, Tbilisi, Institute of the Academy of Â�Sciences of Georgia, Ms. A 908, fol. 16; see Amiranašvili, 1966, fig.╯35. 175 ╇ Rogers, “Saldjūḳids,” EI2 VIII, 936a. 176 ╇ For another closely related vishap-type khatchkʿar with small crosses issuing from the gaping dragon heads, see Armenië, 2001, p.╯52, photograph, lower left side, third khatchkʿar from the left.

dragons and the powers of the earth

67

It is notable that khatchkʿar decorated with the cross, the “Sign” of God or “Wood of Life,” as the main decorative motif, symbolise primarily the salvation of the souls of the departed in whose memory they were erected.177 In the case of vishap-type khatchkʿar it appears therefore less likely that the dragons are shown ingesting the crosses and it may be presumed that, conversely, the mythical creatures are represented as deliverers and as givers of fecundity and prosperity. The visual conflation of dragon bodies with knotted ornaments is perhaps indebted to the canon-tables of Armenian Gospel books.178 An important marginal ornament features a cross resting on an inverted heart-shaped interlace of split-palmettes which evolves into two confronted dragon heads, the necks enclosed in narrow ornamental collars. Just as on the vishaptype khatchkʿar discussed above, the creatures are distinctly portrayed with their tongues darting from the wide-open mouths to touch the base of the cross (fig.╯47). The ophidian heads are capped by pointed ears, the most characteristic aspect being however the wide open, curved snouts, the upper lip ending in a rolled-up tip.╯The ornament is portrayed on Mark’s first page, copied by the priest Hohannes [sic], son of the priest Manuk, in 1171 in Edessa (now known as Urfa) in southeast Anatolia.179 The rendering of the dragons on the vishaptype khatchkʿar as well as the Edessan marginal ornament reveals analogies with the arch-shaped double-headed knotted dragons with wide-open jaws, pointed ears and horn-like protuberances, the bodies covered with an interlaced palmette scroll, that are featured on some of the funerary steles at the vast cemetery of Akhlāṭ. The latter is situated at the northwest corner of Lake Van between Eski Akhlāṭ and the Ottoman qalʿa and was principally erected between 1250 and 1350 (figs. 48 and 49).180 The points of resemblance are noted by Rogers who hypothesises that the overall decoration of the richly carved tombstones, some

of which become highly stylised,181 “was a reaction to the traditional decoration of Armenian khatchkʿar.”182 Representations that associate the dragons yet again with the Christian cross are moreover found at the monastery of Deir Mār Behnām in Mosul. On the lintel of the southern outer door, just below the relief-carved representation of a pair of entwined dragons (figs. 17a and b), examined above, is a central Greek cross. From its base extends an arched cartouche which encloses a pair of stylised regardant quadruped dragons viewed from behind, whose arched bodies are crossing (fig.╯50).183 The dragons’ gaping mouths and their projecting tongues touch the tip of the cusped lower end of the cross. An interesting parallel occurs on the portal leading to the chapel of the baptistery at Mār Behnām, where the frame is carved with a knotted serpentine moulding forming ogee arch-shaped niches which also enclose crosses whose lower ends extend into stemmed palmettes. An important khatchkʿar in the church of Surb Astvatsatsin (Holy Mother of God) in Sevanavankʿ, located on the northwestÂ�ern shore of Lake Sevan in the eastern Armenian province of Gegharkʿunikʿ, depicts the scene of God expelÂ�ling Adam and Eve from Paradise flanked by a pair of dragons with knotted tails and open mouths revealing the tongues (fig.╯51). It is interesting to note the early rabbinic assessment of the association of Adam and Eve with the serpent: “the serpent is your [sc. Eve] serpent and you are Adam’s serpent.”184 Yet for its role in the fall of Adam and Eve, God condemned the serpent forever to eat “earth” (Genesis 3.14). However rather than being portrayed as evil, it is once again shown to touch the bottom tip of the long cross held by God in his right hand with its projecting tongue.185 This ambiguity inherent in the figure of the serpent(-dragon) is elucidated in a passage from the Epistle of Barnabas in which Moses says:

177 ╇ Der Nersessian, 2001, p.╯110. In the Armenian Hymnal the rod is referred to as Holy Cross and Tree of Life, a staff that gives life (kensatu), a staff of power (zawrutʿean). Russell, 2004, p.╯1148. 178 ╇ Cf. Rogers, “Saldjūḳids,” EI2 VIII, 936a. 179 ╇ Izmailova, T.A., “Edesskaia rukopis 1171 goda (M. 313),” Kultura i iskusstvo nardo vostoka 8, Trudi gosudarstvennogo Orgena Lenina Ermitazha, 19, Leningrad, 1978, pp.╯84–101 (English summary, pp.╯117–8), as cited in Der Nersessian and Agemian, 1993, p.╯30. 180 ╇ Rogers, 1988, p.╯109.

181 ╇ Four steles carved with dragons are said to be extant; Süslü, 1987, p.╯640. Cf. Karamağaralı, 1972, p.╯187, nos. 734– 234a (no. 72, epitaph dated 23 Ṣafar 700/7 November 1300). Öney, 1969a, fig.╯11. 182 ╇ Rogers, 1988, p.╯120. 183 ╇ Researched by Preusser, 1911, n. 11, pp.╯5–6, pls. 5, 6.2. Cf. Fiey, 1965, vol.╯1, pp.╯565–09, esp.╯pp.╯605–6; Kühnel, 1950 (church of Khiḍr Ilyās). 184 ╇ Genesis Rabbah 20; Wallace, 1985, p.╯148. 185 ╇ Tr. Lake, 1914, vol.╯1, p.╯385.

chapter four

68

Whenever one of you, he said, be bitten, let him come to the serpent that is placed upon the tree, and let him hope, in faith that it though dead is able to give life, and he shall straightway be saved.

In the Armenian apocryphal tradition a promise was made to Adam at the expulsion that he will be restored as ruler of Paradise after the Second Coming, an event which may be implied in the carving with the giant serpents appearing to guard the gates of Paradise. At least from the beginning of the fifth century, the cross was regarded as a powerful amulet,186 hence in the early Christian period crosses placed on or near entrances served primarily protective and apotropaic functions.187 The Christian cross together with the dragons thus undeniably has a special function as an apotropaion guarding a threshold. The association of the dragon with the cross may be related to the symbol of the serpent as it appears in the Gospel of John (3:14): And as Moses lifted up the serpent in the wilderness, even so must the Son of Man be lifted up.

In his homily on the Gospel of John, the fourthcentury archbishop of Constantinople, John Chrysostom (c. 347–407), similarly compares the “lifting up” of the serpent with the manner of Jesus’ death and parallels the Christ himself with the likeness of the serpent:

╇ Canaan, 1938, p.╯175, n. 249. ╇ Kitzinger, 1970, p.╯640. 188 ╇ Tr. Sr. Thomas Aquinas Goggin, 1957–60, 1:262–3. 189 ╇ Although he derides this reference as indulgence in “tricks and hallucinations”; Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, pp.╯292–3. 190 ╇ XII 5–6, tr. Kirsopp, L., The Apostolic Fathers, London and New York, 1914, vol.╯1, p.╯385, cited after Leisegang, 1955, repr. 1979, p.╯229. Cf. Daniélou, 1964, pp.╯92, 271; Thomson, 1970, ch. 473, n. 1. 191 ╇ Apocalypse of Moses 16–9; Justin, Dialogue 124; Origen, De principiis 3.2.1. In this connection it is interesting to consider an eighth-century Armenian text in which the church is said to be “a paradise planted by God which evil cannot enter; for from that first one we were expelled through the wiles of the serpent, but we enter this one through Christ … In that other paradise was planted a tree of life, forbidden to the firstborn; while here we have the cross of life, planted by the same planter, and which has taken root and has blossomed through His life-giving blood.” Aucher, J.B., Domini Johannis Ozniensis Philosophi Armenorum Catholici Opera, Venice, 1834, pp.╯290–2, as cited in Der Nersessian, 1965, p.╯42 and n. 96. 186 187

In the former, the uplifted serpent healed the bites of the serpents; in the latter, the crucified Jesus healed the wounds inflicted by the spiritual dragon. In the former, there was the uplifted brass fashioned in the likeness of a serpent; in the latter, the Lord’s body formed by the spirit.188

The tenth-century polymath al-Bīrūnī, one of the most eminent literary and scientific figures of the medieval Islamic period, records a Christian source which also relates the brazen serpent to the cross: God ordered the Israelites to make a serpent of brass and to hang it on a beam, which was to be erected, for the purpose of keeping off the injury done by the serpents when they had become very numerous in the desert. Now from this fact they infer and maintain that it was a prophecy and a hint indicative of the Cross (of Christ).189

Correspondingly, as Hans Leisegang asserts in his discussion of the Epistle of Barnabas,190 the serpent lifted up by Moses is characterised as a representation of Jesus, an identification that became an established motif in Christian symbolism, and the counterpart of the satanic serpent in Paradise.191 At the same time, the multivalent aspect of the symbol of the cross, employed as metaphor for both the crucifixion and the resurrection192 of Christ, has to be taken into account. In Armenian exegetical works the khatchkʿar is often elaborated as “Wood of Life” (pʿayt kenats).193 The equation of the “living” cross with vegetation194 provides an 192 ╇ With regard to the Christian Feast of the Apparition of the Flaming Cross in Heaven, al-Bīrūnī records that, according to Christian scholars, Constantine’s mother, Helena, went to Jerusalem to find the cross of Christ which when placed upon a dead body could resurrect the dead. He further notes that the wood used for the cross is referred to as “the wood of Paeonia,” which is frequently “attached to a man who suffers from epilepsy, being considered as a symbol of the resurrection of the dead” (Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, pp.╯293–3). For related narratives on the association of the Cross and the Tree of Life, see Wünsche, 1905, pp.╯15–7. 193 ╇ See€the Homily by David Anyaghtʿ, cited in Russell, 2004, pp.╯630 and 1194. In Agathangelos’ fifth-century theological teachings, Teaching of St. Gregory, the cross is praised as a “Tree of Life rooted in the earth” (tr. and ed. Thomson, 1970, pp.╯7–9, 21–3, and chs. 577–86, 618–31). On the cult of the Tree of Life in Armenia, see Russell, 1987, p.╯33. The Tree of Life as prefiguration of the Cross of Christ is mentioned in the Christian Syriac Meʿārath gazzē (Cave of Treasures, tr. and ed. Budge, 1927, p.╯34). 194 ╇ Cf. Teaching of St. Gregory, tr. and ed. Thomson, 1970, pp.╯159–62, and chs. 641–54.

dragons and the powers of the earth analogy to the dragon’s inherently close relation with vegetation and the associated beneficial qualities of fertility/fecundity. This in addition to the shared special apotropaic functions naturally accords the dragon the preeminent position of a vehicle that issues, thus perhaps wittingly emits or, by extension, one may hypothesise, delivers, and henceforth protects the “living” cross. *** Yet another manifestation of symmetrical dragon heads is found in an embroidery on cotton ground from the greater Khurasan region, now in a private collection. Here the motif is takes the form of addorsed symmetrically arranged monster heads with tall floppy ears springing from a central vegetal composition, probably representing a conventionalised tree motif visualised as stylised plant. The necks of the creatures extend into the “branches of the tree” in such as way that the bodies are fused with the scrolls. The textile has been radiocarbon dated to the thirteenth or fourteenth century (fig.╯52). In his ground-breaking article on “pseudo-planetary nodes” in Islamic art, written in 1938, Hartner has identified these monster heads with long ears growing from scrolling tendrils as “the dragon progeny threatening the luminaries or, vicariously, their domicilia and exaltations.”195 However, he qualifies the astrological interpretation by suggesting that “in all probability, various elements – astronomical, astrological and mythological – were here fused in one.”196 The iconography of the “inhabited scroll” which includes not only dragon heads, but other animal as well as human heads is ubiquitously applied specifically on late eleventh- to early thirteenth-century works of art, in particular metalwork, from Western Central Asia, especially the greater Khurasan region.197

╇ Hartner, 1959, pp.╯237–9, and idem, 1973–4, p.╯112, 118. An anachronical but perhaps not completely irrelevant depiction exists on a late Parthian-period stucco capital (the imagery is 24 cm square) from Qalʿa-i Yazdigird. On the capital a pair of interlaced dragons with slender snouts and long floppy ears, for which there appear to be no prototypes, is shown. Keall, 1967, p.╯115, fig.╯6. 196 ╇ Hartner, 1973–4, pp.╯112–3. 197 ╇ Cf. Melikian-Chirvani, 1982, p.╯145, fig.╯51; PinderWilson, 1997, p.╯343. 198 ╇ Bombaci and Scerrato, 1959, pp.╯13–5 and figs. 11, 12; Baer, 1965, pp.╯15, 66, pl.╯XLVII, figs. 82, 83. Although on the two surviving marble panels no creatures that can be identified as dragons inhabit the scrolls, it is noteworthy that the hexagonal interlace enclosed within the border of 195

69

The long-eared dragon heads are also a common feature on contemporaneously dated press-moulds, most of which have a Khurasani provenance. Ralph Pinder-Wilson hence identifies these scrolls as having their origin in Khurasan, first occurring on marble reliefs, attributed to one of the Ghaznawid palaces (perhaps of Masʿūd III (r. 492/1099–508/1115) or Bahrām Shāh (r. 511/1117–552/1153) or of a later date) excavated at Ghazna by the Italian Archaeological Mission to Afghanistan.198 Spiralling tendrils transforming into dragon- and other mythical heads are figured on a mould which bears a dedicatory inscription to a Ghurid general and governor of Herat, ʿIzz al-Dīn Ḥusayn ibn Kharmil (d.╯607/1210–1).199 Openwork gilded silver plaques found in the valley of Kotchkar in the Semirechye, now in the State Hermitage in St. Petersburg, are further examples of vegetal scrolls inhabited with the monster and animal heads, among which the dragon is prominently represented.200 Rinceaux with such heads also form the all-over decoration on the doublures of a Qurʾān from Anatolia or the Jazīra, datable to the mid-thirteenth or fourteenth century.201 This ornamenting of the doublures of the sacred book of Islam with vegetal scrolls from which spring mythical and naturalistically rendered zoomorphic heads offers an interesting parallel to the practice, evidenced in Ilkhanid Anatolia, of decorating the façades of sacred architecture, such as mosques and madrasas, with such “inhabited scrolls.” A more detailed discussion of this can be found below in the Epilogue. The wide distribution of the motif is, moreover, attested by its representation throughout the Caucasus region, such as on an archivolt fragment from Daghistan in the eastern Caucasus. It is carved with a vivid depiction of a mighty winged dragon pursuing a slender cervid, the one of the panels includes a winged and horned quadruped with a long gaping snout which may be recognised as a quadruped dragon (Bombaci and Scerrato, 1959, fig.╯12; Baer, 1965, fig.╯83). The fabulous creature is rendered just below a similar such quadruped which however is distinguished by a bird’s beak and small pointed ears and thus clearly typifies as a griffin. 199 ╇ Pinder-Wilson, 1997, pp.╯346–7, cat. no. 215. 200 ╇ See Strzygowski, 1916, p.╯214, fig.╯180 (outer vertically-oriented plaques). 201 ╇ Doublures of two sections of a 30-part Qurʾān from Anatolia or the Jazīra, dated to the mid-thirteenth or fourteenth century. The Nasser D. Khalili Collection of Islamic Art, London, inv. nos. QUR433 and QUR132. Pinder-Wilson, 1997, p.╯342, fig.╯11.

chapter four

70

inner edge being framed by a rinceau terminating in gaping dragon heads with projecting tongues (figs. 63 and 53, detail).202 The dragon’s vegetal affiliation, conveyed through its repetitive projection from the foliate bends, was presumably not merely decorative, intending rather to underline the associated, apparently positive and beneficial, semantic meaning of the visual expression. During the twelfth and thirteenth centuries, rinceaux that evolve into gaping dragon heads and protomes also figure in Armenian miniature painting, the earliest dated examples of the motif appearing in marginal ornaments of the famous so-called Homiliary of Mush executed on parchment leaves at the monastery of Avagvankʿ near Erēz/Erznga(n) (present-day Erzincan)203 in northeastern Anatolia between 1200 to 1202 (figs. 126–129, 131–133). The fusion of a winged dragon’s body with the leafy branches is also shown in a headpiece in the same manuscript, which is further discussed below. The hind parts of the mythical creatures are transformed here into vegetal scrolls, thereby highlighting the emergence of two confronted dragon protomes with forelegs, the necks are again enlivened by ornamental collars (fig.╯61). A more detailed rinceau appears in a headpiece of a later manuscript, copied in the scriptoria of Sis, the capital of the Armenian Kingdom of Cilicia, in 1274, which had a princely sponsor, Marshal Oshin, son of Kostandin of Lambron and queen Keran’s uncle. It features on the first page of the Gospel of Luke and is filled with an animated all-over foliate interlace “inhabited” by animal and human heads among which the dragon is prominently represented. For the most part the stems are issuing from animal and dragon necks, the tips of the wideopen dragon jaws, and the pointed tips of the tight-fitting headdress worn by the human heads. The headpiece is crowned at the apex by a foliate interlace in the form of an inverted heart-shape from which project two large, confronted dragon heads whose gaping jaws reveal the tongues that touch the upper edge of the headpiece and whose lower jaws extend into split-palmettes that curve along the same edge (fig.╯54).

In another version of the dragon’s association with vegetation, rather than vegetation sprouting dragon heads or dragon bodies growing into or out of vegetal stems, the mythical creature itself issues vegetation. This is represented on a Samanid-period bowl excavated near Tashkent, Uzbekistan, featuring a sketchy composition of a stylised rider with large, round head and raised hands mounted on a recumbent dragon, from whose elongated open snout and hindquarters sprout two twigs of undulant foliage (fig.╯55).204 The iconography of the dragon as producer of vegetation is further exemplified by two analogous depictions on the arched handles of the two celebrated silver- and copper-inlaid copper alloy buckets, cited above, both preserved in the State Hermitage, St. Petersburg. The loops of the handle of the so-called Bobrinski bucket (inscribed with the date muḥarram 559/December 1163) are fashioned out of dragon protomes enlived by a spotted pattern. From the open mouths springs (a devouring interpretation may be presumed to be incongruous) the four-sided arched section inscribed on two sides in naskhī with benedictory inscriptions and on the top band of the handle in Kufic with the date muḥarram 559/December 1163. On the outside the mythical creatures are flanked by leaping lions (fig.╯56). The so-called Fould bucket (signed by its maker, Muḥammad ibn Nāṣir ibn Muḥammad al-Harawī, and recently attributed to early thirteenth-century Jazīra, northern Syria, or possibly Anatolia) has a handle that is similarly held in place by loops in the form of arched dragons topped at the outside by projecting lionheaded knobs. From the dragons’ open mouths emerges the six-sided arched handle, decorated on all facets with scrolling foliage in place of the benedictory inscriptions of the Bobrinsky handle. On either side of the handle the foliage gives way to lion heads that frame a square link inlaid with a composite quadripartite interlaced knot at the apex (originally topped by a now lost finial) (fig.╯57).205 Yet another aspect of the dragon’s relation with vegetation is shown in the Eastern Â�Christian world in the eleventh-century Georgian church

╇ Cf. Bashkirov, 1931, p.╯95 and pl.╯72. ╇ When Ibn Baṭṭūṭa visited the city in 731/1331, he described the inhabitants of the town as predominantly Armenian. See Tuḥfat al-nuẓẓār fī gharāʾib al-amṣār wa-ʿajāʾib al-asfār, tr. Gibb, vol.╯2, p.╯437. 204 ╇ This as yet unpublished bowl has been handled and photographed by the author. It was excavated together with other Samanid-period earthenware bowls; in spite of

the fact that the piece has not been tested, its body and glazing are characteristic of this period, hence in spite of the rather crude and perhaps unusual depiction displayed on it, it has been deemed sufficiently interesting to be included as a reference, albeit one of a more folkloristic character. 205 ╇ Cf. Mayer, 1959, pl.╯X; Loukonine and Ivanov, eds., 2003, pp.╯123–4, cat. no. 126.

202 203

dragons and the powers of the earth

71

of Â�Çengelli, located near Kağizman in the steep mountains that soar above the Aras valley, southwest of Kars, curiously placed within a formerly Armenian community. Inside the church the northern capital is carved in relief with a recumbent looped serpent portrayed with upstretched head which reaches towards a cluster of grapes suspended just above its gaping mouth (fig.╯58).

The head is capped by a pair of small, cusped ears and the upper lip of the elongated snout terminates in the curled-up tip characteristic of the “Saljuq-type” dragon. The other face of the capital features the relief sculpture of a composite tree bearing pomegranates and bunches of grapes, both kinds of fruit generally carrying paradisical associations.206

╇ Composite vine scroll friezes bearing grapes and someÂ� times also pomegranates frequently appear on the façades of tenth-century Armenian and Georgian churches, the most prominent example being the Bagratid church of the Holy Cross of Aghtʿamar (915–921) at Lake Van where in addition to bands of pomegranate framing the exterior window arches, two continuous vine friezes circumscribe the exterior of the church. Cf. Der Nersessian, 1965, pp.╯11, 25, pl.╯31. The pomegranate is one of the symbols of the Iranian fertility goddess Ardvī Sūrā Anāhitā (Mid. Pers. Anāhīd, Grk. Anaïtis) who was venerated by the Armenians that shared the religion of the Persians and the Medes. See Soudavar, 2003, pp. 58, 74, pl. 165, fig. 71. For an example of comparable

decoration on Georgian Â�churches, see the tenth-century church of the Virgin in Martʿvili; Â�Mepisaschwili und Â�Zinzadze, 1987, p.╯160, fig.╯234. Of note is Â�further the description of the holy garment Â�decorated with pomegranates and bells which was ordered by Moses in Exodus 39. Much earlier examples of these motifs are portrayed in the carved decoration of the church of Saint Â�Polyeuktos (c. 524–527) in Constantinople (Istanbul) featuring, for instance, the combination of pomegranates and split palmettes on a modillion or a vine growing out of a vase on a column capital. Cf. McKenzie, 2006, pls. 559, 567. According to Genesis (9:20), the Tree of Life was probably the stem of a vine.

206

dragons and animals of the natural and the mythical realms

73

chapter five

Dragons and animals of the natural and the mythical realms

a.╇ The dragon in animal combat scenes Ancient Near Eastern animal motifs such as the theme of combatant animals figure prominently in the decorative repertoire of the medieval Islamic period. Animal combat scenes also appear to reflect the visual language of so-called Steppe Empire Eurasian “animal style” art.1 Dragons locked in combat with each other or with other real or imaginary creatures played a particularly significant part in the artistic vocabulary of many Eurasian pastoral peoples. This ancient iconographical theme of animal combat, in which the dragon is pitted against another animal, continues to appear in both Islamic and Christian contexts. A fluidly rendered representation, evoking the immediacy and innate animal nature intrinsic in wild beasts, is found in a twelfth- or thirteenth-century openwork decoration of a brazier from the Iranian world or Anatolia. The scene depicts the combat of two exceedingly fierce and powerful creatures, a feline quadruped with a quadruped dragon.2 The latter has an elongated ophidian body tapering to the tip of a long tail and is twisted around the entire body of its striding adversary, beginning at the right hind leg, circling the body once, and then the neck, re-emerging behind the feline’s head,

biting with an elongated snout into the feline’s neck, who in turn mauls the mythical creature’s shoulders. The densely patterned fur of the felid and scaly hide of the fabulous beast are finely contrasted (fig.╯59).3 This type of imagery has been interpreted as a classical metaphor of the dragon as eclipse symbol swallowing the lion, symbol of the Sun (the zodiacal sign of Leo is known as the house of the Sun).4 It is moreover generally assumed that the creature emerging victorious from the combat would assimilate some of the essential qualities of the vanquished party; in this case the dragon perhaps absorbing qualities attributed to the lion such as bravery, courage and magnanimity as well as ferocity, voraciousness and wildness. A comparable spirit seems to pervade the dragon reliefs preserved from the now destroyed city walls of the Saljuq capital Konya (618/1221), now in the İnce Minare Müzesi in Konya. The dragons’ expansive knotted tails curve upwards and terminate in small dragon heads with open mouths that appear to grasp or attack the tails, hence visually conflating two dragons (fig.╯60).5 The influence of the bestiaries that populated the imaginary world of the medieval period perhaps stimulated the visual fusion of two or more animals, often of a different type, a common feature

1 ╇ In his study on the theme of animal combat, Jean-Paul Roux (1981, pp.╯5–11) came to the conclusion that the cultivation of shamanism among the Eurasian peoples perhaps played a role in the development of the motif. 2 ╇ An engraved copper alloy plaque in the form of a haloed sphinx passant shown in combat with a winged dragon whose body is circling the body once and then the neck, re-emerging behind the sphinx’s head and aiming to bite the back was sold at Sotheby’s, London, 13 April, 1988, lot 252. 3 ╇ Cf. L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯97, cat. no. 64 (dated here to the twelfth or thirteenth century); Hauptmann von Gladiss and Kröger, 1985, p.╯54, cat. no. 264 (dated here to the fourteenth century). Sections probably belonging to the same artefact are found in several collections, such as Copenhagen, the David Collection, inv. no. 31a/1975; von Folsach, 1990, p.╯197, cat. no. 327, and idem, 1991, p.╯53, cat. no. 91 (catalogued with an Iranian

provenance and dated to the thirteenth or fourteenth cen tury). For a complete side of the brazier with such openwork decoration, formerly in the Harari Collection, Cairo, see Erginsoy, 1978, p.╯331, fig.╯175, and further fragments of side panels, fig.╯176 A (in the Grenoble Museum) and fig.╯177 B (in Baltimore, Walters Art Gallery). Cf. Grube and Johns, 2005, p.╯221, fig.╯72.7. 4 ╇ Cf. L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯97, cat. no. 64. As has been demonstrated by Hartner and Ettinghausen, the lion combat motif dates back to ancient Mesopotamia and Iran having been interpreted as an Â�astronomical symbol for the constellations Leo and Taurus that developed into a royal and finally a religious symbol. Hartner and Ettinghausen, 1964, pp.╯161–71. See also Kuz’mina, 1987, pp.╯729–45. 5 ╇ Sarre, 1909, p.╯13, fig.╯15; Önder, 1961, p.╯70, fig.╯1; Diyarbekirli, 1968, p.╯370, fig.╯5; Öney, 1969a, p.╯194, fig.╯1; Gierlichs, 1996, p.╯198, cat. no. 44, pl.╯38. 2.

74

chapter five

in the arts of Western Central Asia, particularly from the eleventh to the thirteenth century. Often however compositions represent a physically strong animal attacking a weaker one. The interesting depiction of an apparent contest between two dragons, hence two fabulous creatures of equal strength, perhaps symbolises the perpetual act of predation. An animal combat also appears as part of a bilaterally symmetrical foliate interlace which fills a headpiece in the Mush Homiliary illuminated at the monastery of Avagvankʿ near Erznga(n) between 1200 and 1202. It terminates at one end in the protomes of two confronted winged dragons with forelegs and at the other in two goat heads topped by long curved horns, the latter snapping at the dragon wings while the dragon snouts clinch the goat necks (fig.╯61).6 The façades of Armenian churches are similarly adorned with combatant animals, also featuring the likeness of the dragon or giant serpent. The bas-relief of a bird in combat with an ophidian dragon is shown over the window on the southern façade of the small twelfth- or thirteenthcentury chapel behind the main church of Surb Astvatsatsin, which forms part of the monastic complex Makaravankʿ in Tavush province, Armenia.7 Here the long serpentine body of the dragon is slung around the neck of the bird that in turn hacks with its large beak at the dragon’s neck (fig.╯62).8 A certain archaism of representation displayed in the vividly expressed animal nature of the creatures is preserved in the north Caucasus region – clearly featuring the dragon as antagonistic force in animal combat. The depiction echoes the theme of a lion killing a weaker animal which had been current in the Near Eastern and Central Asian world for millennia.9 The portrayal cer-

tainly reflects the immediacy of mythic animal combat. Used as architectural ornament, the dragon is portrayed as a predatory and voracious creature, mauling another quadruped, presumably a cervid, shown with backward turned head, on an archivolt fragment from Daghistan in the eastern Caucasus (fig.╯63).10 The dragon’s head is crowned with slanted cusped ears that are folded to the back, the flame-like mane is composed of four large, contiguous, cusped teardrop shapes that project from the sinuous necks; the slender wing that springs from the haunches is stiffly raised over the back ending in a small curl at the tip.╯The mythical creature is rendered with a pair of muscular forelegs, the inner leg raised as if to attack its prey, while the other balances the weight of the body; its elongated undulating serpentine tail forms a central loop and terminates in a curl.

6 ╇ This depiction may be an allusion to vishap-a-kagh, a term employed in Armenian legend (from vishap and kagh, the latter meaning “male goat,” hircus). It may be associated with the god Vahagn of the ancient Armeno-Parthian pantheon who bears the epithet vishapakʿagh, “dragonslayer.” I am grateful to Rev. Dr Vrej Nersessian (personal communication) for elucidating this point. Cf. IshkolKerovpian, “Vahagn,”WdM IV, 1, pp.╯149–52, esp.╯p.╯150. 7 ╇ See Makaravankʿ, Documents of Armenian Architecture/ Documenti Di Architettura Armena, vol.╯22, Venice, 1993. 8 ╇ The same motif is found frequently in marginal ornaÂ� mentation of Armenian manuscripts. For instance, in a Gospel book, vellum, 318 fols., dated 1290, copied in Drazark, Cilician Armenia, by the scribe Tʿoros and illuminated by Tʿoros Roslin, Ms. 5736, fol. 10b. Mnatsakanyan, 1955, p.╯532, fig.╯1040. 9 ╇ Cf. Ettinghausen, 1972, pp.╯44–5.

10 ╇ Bashkirov, 1931, pl.╯72; Baltrušaitis, 1929, pl.╯LXVI, fig.╯107. Anatoli Ivanov (1976) has revised Bashkirov’s twelfth- to fourteenth-century dating of most Daghistan sculpture preserved in the walls of the mosques of Kubachi, on the basis of tombstones from Kubachi and a neighbouring village as well as an inscribed archivolt (fig.╯133 in the publication) datable to the fourteenth or early fifteenth century. However, the archivolt fragment with the dragon and cervid was not part of the group of pieces discussed and published by Ivanov, nor do the carvings of the published tombstones (figs. 123, 126, 127, 130, 135, 136) bear any stylistic resemblance to the carving on the fragment (on the basis of the photographs), which is why a twelfth- or thirteenth-century dating still seems justified. The contention that most of the reliefs were not made in Kubachi but elsewhere in Daghistan (p.╯203), is certainly worthy of notice.

b.╇ The dragon and mythical creatures The visually conflated dragons of the Konya reliefs (fig.╯60) show that the body of the dragon was itself seen to be subject to mutations. Similarly, hybrids resulted from the dragon incorporating parts of other animals and mythical creatures and vice versa other animals and mythical creatures merging with parts of the dragon. Most commonly it was the head that was assimilated. The multiplex creatures thus juxtapose two principles into a unified being creating a duality, an ambiguity which simultaneously contrasts and fuses two metaphorical principles, as, for instance, the well-known example of the bird and the serpent or dragon representing two eternal antagonists which are discussed in the following. The nature of such hybrid beings is frequently associated with transitory or liminal states exemplified in such

dragons and animals of the natural and the mythical realms

75

tales as the Alexander Romance which recount Alexander’s journey to the outer reaches of the earth, the land of darkness, without Sun, Moon or stars, where he meets such mythical creatures. Antecedents for dragon head motifs issuing from different zoomorphic junctures, for instance, from the tips of the wings or the tail ends,11 may ultimately be found in the ancient “animal style.”12 The synthesis with the dragon is portrayed on animals from the real and the imaginary realm. Among the theriomorphic examples, the dragontailed lion was one of the most common, as will be shown in the following. Predominant among the dragon-tailed fabulous creatures from the mythical realm are those usually defined as griffins and unicorns.13 A thirteenth-century relief from the now destroyed fortress in Konya portrays a winged unicorn with prominent dragon tail in pursuit of a bovid (fig.╯64).14 Often however the creatures fusing with parts of the dragon were therianthropomorphic beings combining the form of a human with those of an animal. Human-headed mythical creatures, generally identified as harpies, sphinxes15 or centaurs, were portrayed with limbs transforming into dragon heads on diverse objects of the Eastern Islamic world. A harpy (a human-headed bird) whose florid tail terminates in a large upwardcurving dragon protome with forelegs and gaping mouth serves as finial of an eleventh-century copper alloy incense burner typical of artistic production from the greater Khurasan region.16 A sphinx (a human-headed quadruped, often represented with a female head and the body of a feline, frequently a lion(ess)) and a griffin (a hybrid of a bird and a quadruped), both with their upwardly curving tails terminating in dragon

heads with open mouths that are oriented towards the heads of the fantastic animals, are prominently rendered on a frieze which runs around the body of a twelfth-century gilded and nielloed silver jug from Iran.17 Anatolian examples include architectural manifestations such as the carved reliefs of two winged quadrupeds with dragon-tails, possibly representing a sphinx and a griffin, on the bastion (burç) in the southwest of the city wall of Diyārbakr (Ulu Bandan tower no. 31) which according to the inscription was built in 604/1208–09 by the builder (bannāʾ) Ibrāhīm ibn Jaʿfar under the patronage of the Artuqid ruler Nāṣir al-Dīn Maḥmūd ibn Muḥammad al-Malik al-Ṣāliḥ (597/1201– 619/1222).18 A star tile unearthed during the 1992 excavations of the the now destroyed small palace at Kubadabad (623/1226–634/1237) built by sulṭān ʿAlāʾ al-Dīn Kay Qubādh I shows a gracefully rendered sphinx marchant, the haloed head in profile turned backwards and the tail ending in a dragon head capped by pointed ears, a wisp of flowing hair at the back of the neck and open jaws revealing the tongue (fig.╯65). A pair of confronted perched harpies whose upstretched wing tips Â�terminate in dragon heads are found on the türbe of Hüdavend Hatun/Khudāwand Khātūn (712/1312) in Niğde, examined in more detail below (fig.╯66). An example of a portable object, representing perhaps a fountain head, is the small twelfth- or thirteenth-century pierced copper alloy figure of a recumbent lion-bodied sphinx with a tall crown whose raised curving wings as well as the upward-curling tail end in gaping dragon heads.19 To this dragon-tailed menagerie may be added a dragon-tailed sphinx portrayed to the left of the portal of a caravanserai in the Selim moun-

11 ╇ Body parts of natural and mythical animals could also transform into body parts, mostly the heads, of other animals of the natural world. For instance, an eleventh- or twelfthcentury copper alloy incense burner in openwork from present-day Afghanistan in the form of a standing lion has a tail that terminates in a bird head (David Collection, Copenhagen, inv. no. 48/1981; see von Folsach, 1991, p.╯44, cat. no. 30, ill. on p.╯15). Birds whose heads transform into those of hares are found on the cavetto of a twelfth- or thirteenth-century silver-inlaid dish with polylobed cavity from the Khurasan region, in the Musée du Louvre in Paris, inv. no. OA 6479 (Paris, Musée du Louvre, Département des Antiquités orientales, Section Islamique; Pope and Ackerman, eds., 1938–9, repr. 1964–81, p.╯1315; L’Etrange et le Merveilleux en terres d’Islam, 2001, pp.╯50–2, cat. no. 32, (see detail of cat. no. 32)). 12 ╇ Cf. Otto-Dorn, 1978–9, pp.╯130, 136. 13 ╇ For a discussion of the unicorn-elephant fight, see Ettinghausen’s monograph on the unicorn, 1950, pp.╯84–91

(with a further dragon-tailed example, pl.╯31, left). 14 ╇ Cf. Gierlichs, 1996, pl.╯37.2–3. 15 ╇ For an in-depth study on sphinxes and harpies in medieval Islamic art, see Baer, 1965. For a more general study which includes modern interpretations of the sphinx, see Regier, 2004. 16 ╇ Berlin, Museum für Islamische Kunst; Pope and Ackerman, eds., 1938–9, repr. 1964–81, vol.╯13, p.╯1278 B; Erginsoy, 1978, p.╯158, fig.╯70. 17 ╇ Berlin, Museum für Islamische Kunst, inv. no. I.2210; Gierlichs, 1993, p.╯53, cat. no. 50. 18 ╇ Cf. Gabriel and Sauvaget, 1940, p.╯98, fig.╯72, pp.╯120, 171, fig.╯14, pls. LVIII. 2, LXVIII. 1; Gierlichs, 1996, pl.╯51.5 and 6. 19 ╇ Previously Diyārbakr Müzesi, Erginsoy, 1978, pp.╯447–8, figs. 222 a and b; Baer, 1965, fig.╯29. Now in Copenhagen, the David Collection, inv. no. 5/1978; von Â�Folsach, 1990, p.╯197, cat. no. 326.

76

chapter five

tains in southern Armenia. The mythical creature has a powerful, perhaps bovine, quadruped body structure, topped by a slender neck which supports the frontally depicted human head whose features are no longer identifiable on account of the weathered condition with a prominent threepronged crown-like headdress. The creature’s tail curves upward to form a pretzel-like knot terminating above the slender wings with curled tip in a dragon head whose open jaws reveal the protruding tongue. According to the Persian inscription on the outside lintel the caravanserai was built in 727/1326–7 whereas the Armenian inscription on the right inside the portal states that it was constructed six or seven years later by Chesar Ōrbēlian and his brothers (fig.╯67); just like the türbe of Hüdavend Hatun, the caravanserai was erected during the reign of the last Ilkhanid ruler Abū Saʿīd Baḥādur Khān (r. 717/1316–736/1335). Hence in contrast to the animal combat in which the victorious antagonist would assimilate some of the essential attributes of the defeated animal, the hybrid examples referred to above amalgamated not only the form but presumably also the nature of the disparate elements, in other words their innate characteristics and attributes. This fusion was a conspicuous feature perhaps intended not only to reinforce the visual impact but also to augment the intended effect of a potent symbol. It has been shown that one of the most commonly depicted themes is that of dragon heads issuing from the tails or wing tips of real or imaginary creatures some of which may sometimes be identified as one of the constellations visualised as one of the twelve signs

of the zodiac, as discussed below. It may be surmised that these depictions, other than those in an astronomical or astrological context, do not have a special astrological connection,20 but reflect a general eastern spirit of ʿajāʾib, akin to that of the western merveille, at the mythical creatures of God’s Creation (al-makhlūqāt) that are thought to exist beyond the limits of the known world. This is expressed in visual hybridisations that were particularly pervasive during the twelfth and thirteenth century in the medieval Central Asian world and beyond.

20 ╇ Cf. Rogers, 1969, pp.╯154–5 and ns. 15 and 16; Badiee, 1978, pp.╯255–6. 21 ╇ The Babylonian myth of Etana (surviving in Old Babylonian, Middle Assyrian and Neo-Assyrian recensions) that contains the fable of the serpent and the Zu bird relates how both lived in peaceful coexistence, “in the shade of the styrax tree begets the serpent; on its crown begets the eagle” until one day the eagle treacherously devoured the young of the serpent and the serpent, in revenge, broke the eagle’s wings and threw him into a deep pit where he was doomed to perish. Etana found and healed the eagle who to show his gratitude took him to heaven to bring from there the “plant of birth” without which Etana could have no offspring. See Knipe, 1967, pp.╯340–5; also Küster, 1913, pp.╯52–3, 127–31, Astour, 1965, p.╯235, and Reiner, 1991, repr. 2003, p.╯301. It is of note that the Zu bird represents the evil protagonist against the the serpent as victim. 22 ╇ Homer (Iliad XII 202–9; see also verse 220) describes the battle between eagle and serpent with the words:

who held it between chest and neck, twisting backward. And he dropped it to the ground, hurting with pain, and it fell in the midst of the army while he gave a cry and flew off down the wind. The Trojans shivered to see the shining snake (ophis) which lay among them, a portent from Zeus of the aegis.

...and a bird came ... a high-flying eagle ... holding a huge scarlet snake (drakōn) in its talons, alive still, twitching. Nor did it give up the fight. It struck the one

c.╇ The dragon and the bird From ancient times the bird and the serpent represented two of the mightiest cosmic entities. Their interaction harks back to well-known mythic themes recorded from ancient Mesopotamia,21 Greece22 and India.23 In these archaic accounts a bird is identified with the theft of the elixir of eternal life, whose custodian the serpent would once have been. Early Islamic histories, such as Balʿamī’s Â�Tarjumat-i tārīkh-i Ṭabarī, record several stories in which the rooster is considered apotropaic. In a passage of the chapter on Gayūmart (Av. gaya marətan “mortal life”), the mythical First Man of the Mazdean myth of creation, Balʿamī describes him witnessing the scene of a white rooster victoriously attacking a serpent in the middle of the day and then crowing gleefully. Gayūmart thence pronounces the rooster to be an auspicious portent. The same chapter notes that contrarily a rooster’s

Tr. after Redfield, 1975, repr. 1994, p.╯144. The motif similarly appears in Aristophanes’ Equites 218–22 and in Â�Plutarch’s Vita Themistocles 26. 23 ╇ In the Vedic myth recorded in the Rigveda (4.26–7) it is Indra mounted on an eagle or falcon who steals the elixir (Av. haoma, Ved. sóma) from the atmospheric gandharvas. The antagonism between the mythical bird or bird-like creature (often an eagle-like predatory bird), garuḍa, and the nāgas is a favourite theme in Indian literature which finds its origin in the ancient fable of the rival sisters, Kadrū, the mother of the nāgas, and Vinatā, the mother of garuḍa, recorded in the Mahābhārata (Ādi-parvan 25–34); Vogel, 1926, pp.╯51–3. Cf. Denis, E., La Lokapaññati et les idées cosmologiques du Bouddhisme ancien, Paris, 1977, vol.╯1, 170–1, as cited by Strong, 1992, pp.╯200–1, with further examples on pp.╯29, 188–9, 200–8.

dragons and animals of the natural and the mythical realms crowing at night was considered an inauspicious sign. Its crowing at such an unusual hour served as a warning that saved Siyāmak, Gayūmart’s grandson (according to the Bundahishn), from being attacked by a serpent.24 Cosmological concepts associating the bird with the serpent are recorded in the Meccan tradition, according to which Allāh sent:

77

Yet although birds and (serpent-)dragons are depicted as opposing symbols, the sky and water/ earth, above and below, they are often conjoined into a composite generated from the same stock of concepts and images.28 This visual conflation of the bird with the dragon can be seen in particular on textiles, which due to their portability and high status, were a prime luxury medium for the transmission of a visual vocabulary. The motif is exemplified, for instance, on a twelfthor thirteenth-century silk samite fragment from Samangan province in Afghanistan, of elongated rectangular form, featuring pairs of large and stately confronting birds, probably roosters, with prominent combs and florid tails terminating in

dragon heads, which flank elaborately rendered palmette trees (fig.╯68).29 Islamic textiles largely survive in relatively good condition because they were included in medieval church treasuries.30 The bird and dragon motif can be found on one such fragment which was preserved in the shrine of Saint Apollinaris at the church of Saint Servatius in Siegburg. Dragon heads project from spiralling volutes that emerge from the outer wing tips of a double-headed bird (probably an eagle) with “ears” set within shieldshaped medallions. The expressive crested dragon heads are rendered with large, circular eyes and particularly wide-open snouts tapering to pointed tips and revealing the tongue (figs.╯69a and b).31 The silk has been attributed by Friedrich Sarre to an Anatolian manufactory, where it might have served as a Saljuq coronation robe,32 and has been dated to the first quarter of the thirteenth century.33 Comparable imagery of the double-headed bird of prey is preserved in architectural ornamentation at the confines of the same cultural sphere. The same bird with “dragon-wings” can also be found on either side of the western gateway of the thirteenth-century Great Mosque of Divriği (Diwrīgī), situated on the borders of Armenia and Cappadocia. The dragon heads are rendered with elongated wide-open mouths emitting curving, tongue-like foliage, which in turn extends to the outspread wing tips of the double-headed birds (figs.╯70a and b).34 The mosque was built in 626/1228–9 (or later) by native craftsmen of Akhlāt (northwest of Lake Van) for the Türkmen Mengüjükid ruler Aḥmad Shāh and his wife, Tūrān Malik, who were under Saljuq suzeÂ�rainty.35 The epigraphic band above the main portal of the mosque mentions the name and title of ʿAlāʾ al-Dīn Kay Qubādh, presumably in recognition of his overlordship.36

╇ Recorded in Balʿamī’s Persian translation of Ṭabarī’s History; tr. and ed. Lazard, 1956, pp.╯206–7, 211. 25 ╇ Ibn Hishām’s edition of the Ibn Isḥāq’s Sīra, ed. Wüstenfeld, 1857, p.╯125; al-Diyārbakrī, Ḥusayn ibn Muḥammad, Taʾrīkh al-khamīs, vol.╯1, Cairo, 1283, p.╯112, cited after Wensinck, 1921, repr. 1978, p.╯47. Cf. Atallah, 1975, p.╯164. 26 ╇ Wensinck, 1921, repr. 1978, p.╯47. 27 ╇ Idem, p.╯46. 28 ╇ The theme of the transformation of a serpent into an eagle occurs for instance in the Alexander Romance; Pseudo-Callisthenes I, ch. 10, tr. and ed. Stoneman, 1991, p.╯42. 29 ╇ Related imagery occurs on a mid-thirteenth-century lampas weave which carries addorsed felines and doubleheaded eagles in the interstices grasping curled outer tail feathers that end in dragon heads; preserved in the Cleveland

Museum of Art, inv. no. 1990.2. Watts and Wardwell, 1997, cat. no. 43. 30 ╇ Cf. Shalem, 2004, pp.╯125–30. 31 ╇ Cf. Jacques and Wencker, 1967, pp.╯472–527 and 487–90. A smaller fragment (length 22.5 cm, width 21 cm) is preserved in Staatliche Museen Preussischer Â�Kulturbesitz, Kunstgewerbemuseum, Berlin, inv. no. 81.745. The Arts of Islam, 1976, p.╯79, cat. no. 14; von Wilckens, 1992, p.╯43, fig. 66, colour pl.╯on p.╯48; Gierlichs, 1993, p.╯30, pl.╯10, cat. no. 84. 32 ╇ Otto-Dorn, 1978–9, p.╯119. 33 ╇ von Wilckens, 1992, p.╯43, cat. no. 66. 34 ╇ Otto-Dorn, 1978–9, p.╯120, fig.╯18; Enderlein, 1990, p.╯88; Firat, 1996, no pagination. Gierlichs, 1996, pl.╯3, 1–4. 35 ╇ Sourdel-Thomine, “Diwrīgī,” EI2 II, 320a. 36 ╇ Van Berchem, 1910, pp.╯97–8; Otto-Dorn, 1978–9, p.╯121.

...from the cupola of heaven a bird in the form of an ʿuqāb, a bird with a black back, a white breast and yellow claws. While the serpent on the wall of the Kaʿba opened its mouth in order to repel the bird, the latter came, caught it by its head and took it to the mountain Ajjād.25

Wensinck interprets this tradition as a reference to the three cosmological powers, “the sun [symbolised] by the bird, the ocean by the serpent, and the earth by its navel, the sanctuary [the Kaʿba].”26 He further notes that the age-old symbolism of: ...the eagle with the serpent in its mouth or between its talons is a striking illustration of the idea of cosmic victory. … Eagle and serpent are the representatives of two of the mightiest cosmic entities, the sun and the ocean.27

24

chapter five

78

Here the upward-curling tips of the tail feathers of the confronted, regardant birds transform into characteristic dragon heads that peck at the tail tips (fig.╯74). The inverse mutation of the tail or wing ends of dragons transforming into raptorial bird heads is exhibited, for instance, on a copper alloy dragon knocker that was fastened to the wooden door of the Great Mosque (Ulu Cami) of Cizre (fig.╯83) or the dragon reliefs carved on the now destroyed Talisman Gate (Bāb al-Ṭilasm) in Baghdad (fig.╯139b), both discussed in the following.

The mutation of the bird into a dragon also occurs in the illuminations of Armenian manuscripts. In a marginal ornament in a ninth- or tenth-century missal from Ejmiatsin (“the Onlybegotten descended”) in Armavir province confronted mythical creatures have the bodies of birds, their long necks interlacing to terminate in dragon heads that turn backwards and appear to bite into the upright foliate tail tips (fig.╯71). Not the heads of the birds but more frequently their tails end in dragon heads as shown in the L’viv Gospel transcribed and illustrated in Cilician Armenia in 1198/1199.37 The eighth of the ten canon tables is framed by birds that are perched on vertically interlaced vines, their tails extending into a spiral from which grow dragon heads crowned by long rabbit-like ears,38 their open mouths revealing tongues with bifid tips that touch the tips of the birds’ beaks (fig.╯72).39 In the same manuscript on the opening page of the Gospel of Luke, a pair of long-legged birds flanking the central cross and surmounting the frame of a headpiece are shown in the same posture, their heads turned backwards, the beaks inserted into the open mouths of yet another type of dragon with a scaly body and long pointed ears. Importantly this depiction portrays the symbiosis of the entire dragon body with that of the bird, the dragon’s tail tapering towards the bird’s head and extending towards the beak which, held in the dragon’s jaws, closes the loop (fig.╯73). Similes associating the serpent and the dove are recorded in Agathangelos’ fifth-century literary composition entitled Teaching of St. Gregory, one of the earliest Armenian theological texts, which states that man will learn the wisdom of the serpent, strip off all ephemeral impurities as a serpent sloughs off its skin, and receive the pure simplicity of the dove.40 A comparable feature portraying a fusion of the bodies of the bird and the dragon can be observed on the stone relief above the southern outer door at the monastery of Mār Behnām/Deir al-Khiḍr.

The symbiosis of the dragon with a feline predator, two creatures carrying associations not only of great danger, but of royalty (discussed below in chapter 7), is alluded to in Iranian literature. The phenomenon is illustrated in a Shāh-nāma account which describes how Rustam, the chief epic hero of Iran, when fourteen years old, vanquishes in India the sea- and land-dwelling dragon known as Babr-i Bayān. Rustam has the slain beast flayed and then makes out of the skin or hide a coat for himself, the babr-i bayān.41 In nature when an animal dies its flesh decays while the skin and hair remain. This fact may have given rise to the association of animal hides with special powers. The dragon’s skin may be considered to represent the essence of the living beast. Being covered with its skin was believed to symbolise the acquisition of this fundamental nature and to induce a transformation of the wearer. The babr-i bayān was invulnerable, proof against fire and water and impregnable to any weapon. The hero thus wore it before going into battle, placing it above his chainmail tunic which was reinforced by iron-plated armour.42 The babr-i bayān was visualised as either the pelt of a tiger (babr) or that of a leopard (palangīna).43 The dragon Babr-i

╇ Prinzing and Schmidt, eds., 1997, pp.╯18–21. ╇ The extensive range of the repertoire of zoomorphic visual fusions resulting in fantastic beasts on artefacts from the Irano-Turkish region is exemplified, for instance, by birds with the heads of hares, depicted on the cavetto of a twelfth- or thirteenth-century silver-inlaid dish with polylobed cavity. See p. 75, n. 11. 39 ╇ Cf. Akinian, 1930, p.╯16, fig.╯7; Prinzing and Schmidt, eds., 1997, p.╯50, pl.╯VIII.

╇ Tr. and ed. Thomson, 1970, p.╯29, and chs. 602–10. ╇ The legend is known among the Mandeans of Iraq in which Rustam kills the Babr-i bayān in China when he is twelve years old. Petermann, 1860–1, vol.╯2, pp.╯107–9. See Khāleqī-Moṭlaq, “Aždahā II” and idem, “Babr-e bayān,” EIr. 42 ╇ Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯129, ll. 1470–6. Melikian-Chirvani, 1998, p.╯173. 43 ╇ Idem, vol.╯3, p.╯129, l. 1474; 227, l. 1035; this idenÂ� tification is confirmed by its reference in a Sogdian text

37 38

d.╇ The dragon and the feline

40 41

dragons and animals of the natural and the mythical realms

79

bayān thus seems to have acquired features of a feline predator endowed with dragon-like characteristics.44 By clothing himself in the babr-i bayān Rustam implicitly signalled his symbolic appropriation of the dragon’s qualities as well as his mastery over the hybrid creature. It is of note that such composite animal imagery associating the feline, in particular the lion, with the dragon is also apparent in ancient Near Eastern myth.45 The visual conflation of dragon and feline is expressed in Central Asian art from at least the eleventh century onwards. It appears on a matrix from the set of 77 late twelfth-century copper alloy matrices for belt/strap fittings, considered earlier in chapter 3, one of which bears an inscription in the name of the Ghurid sulṭān Ghiyāth al-Dīn Muḥammad ibn Sām. The matrix is symmetrically cast in relief with a pair of small, addorsed lions couchant. These are surmounted by ascending, exaggeratedly long, imbricated tails that interlace in a lozenge-shape before overlapping again at the tip and terminating in disproportionately large, confronted dragon heads. The creatures’ open snouts touch at the tips, thus forming another small lozenge, and are crowned by prominent curving horns with curling tips (fig.╯75). Another small matrix from the same set, with lozengeshaped outline topped by a trefoil finial, features a single seated lion in high relief whose upwardarching tail also ends in a large, horned, snarling dragon’s head with gaping mouth revealing the tongue (fig.╯76).46 The conflation of the king of animals, the personification of kingly power par excellence (lion hunting being the prerogative of Islamic rulers), with the dragon presented an apt symbolism for the strap fittings of a royal belt.

The iconography of a dragon-tailed lion is also shown on the portal of several thirteenth-cenÂ� tury Islamic and Christian monuments located in particular in southeast Anatolia and Â�northern Mesopotamia. It appears on the so-called royal door of the monastery of Mār Ḥūdéni/Mār Aḥūdēmmeh in Mosul,47 the monastery of Mār Behnām/Deir al-Khiḍr at the southeast of Mosul (fig.╯77),48 on the men’s entrance of the Chaldean church49 and on the main portal of the Jacobite church (fig.╯78),50 both located in Cizre. The seated lion is portrayed in profile, the head rendered frontally or in three-quarter view, his long tail winding around the hind legs, under the belly and extending over the back transforming into a dragon’s neck and head. The mythical creature is portrayed with a curved horn and gaping jaws, revealing a long tongue with bifid tip entwined at mid-section, snapping at the lion’s back. At Mār Ḥūdéni the lion’s tail again forms a pretzellike knot, and at the Jacobite church the dragons’ bifid tongues entwine at mid-section (the same feature can be observed on the addorsed entwined dragons at Mār Behnām, figs. 17a and b, 160). Of note is moreover the conspicuous placement and orientation of the dragon-tailed lions towards the opening of the doorways drawing attention to the threshold. Similar depictions appear also in the twodimensional medium of manuscript illumination such as in a Gospel book painted by Bartholemew (Bardagh) in the Armenian province of Siunikʿ, in which two confronted lions are seated on the tips of the Armenian initial “Ս” of the first word of€Mark’s Gospel€ (Սկիզբն, the beginning). The two€lions represent here the symbol of Saint Mark,51 their heads being rendered in frontal view

(Benveniste, E., Textes Sogdiens, Mission Pelliot III, Paris, 1940, pp.╯134–6), as cited in Khāleqī-Moṭlaq, “Babr-e bayān,” EIr. Cf. Melikian-Chirvani, 1998, p.╯173. 44 ╇ It is worth mentioning the analogy to the Greek legend of the pre-eminent hero Herakles, whose exploits included strangling a lion, which could not be slain in any other way because its pelt was invulnerable, and wearing its skin as a coat over his shoulders as though becoming part lion himself. 45 ╇ The story of “The Slaying of the Labbu” (Cuneiform texts from Babylonian tablets in the British Museum 13.33– 54; Heidel, repr. 1951, pp.╯141–3) recounts the victory of the god Tishpak over the dragon referred to variously as labbu (Akkadian for “lion”), bašmu and mušḫuššu. The latter two are listed as separate members of Tiamat, the personified primordial water-chaos, and her battalions in the Akkadian epic of creation, Enūma Elish (I.121; II.27; III.31,89), of Babylon. See Wilson, 2001, pp.╯30–2. 46 ╇ Kuwait, al-Sabāh Collection, Dar al-Athar al-Â� Islamiyyah, Kuwait National Museum, LNS 2558 J ‘a-x2’ ‘s2’ and ‘t2’; ‘t2’ being entirely covered with reddish corrosion

products; ‘s2’ height 2.69 cm, width 1.53 cm, ‘t2’ height 2.92 cm, width 1.87 cm. 47 ╇ Height 22,5 cm, width 27 cm. Sarre and Herzfeld, 1920, vol.╯2, p.╯294 (drawing), fig.╯281, and, eidem, 1911, vol.╯3, pl.╯106, fig.╯3. Fiey, 1959, pp.╯144–6, fig.╯11. 48 ╇ Preusser, 1911, pl.╯x; Hartner, 1938, fig.╯27; Fiey, 1959, fig.╯12; Gierlichs, 1996, pl.╯59.1. 49 ╇ The church no longer exists; recorded in situ by Â�Preusser, 1911, p.╯25, pl.╯35.2 (Preusser refers to it as Jacobite church); reproduced by Hartner, 1938, pp.╯143–4, fig.╯24; Willy Hartner also notes the fact that as “manly animal” par excellence, the lion reliefs were absent from the women’s entrance of the Chaldean church, fig.╯25. 50 ╇ Height 45 cm, width 56 cm. The church does not exist any more; recorded in situ by Preusser, 1911, p.╯25, pl.╯34.2 (reproduced by Hartner, 1938, pp.╯143–4, fig.╯23); the reliefs are preserved in the Archaeological Museum, Diyārbakr, inv. nos. 64, 65. Öney, 1969, p.╯209, and n. 50, fig.╯34; Gierlichs, 1996, pl.╯60.5, 6. 51 ╇ I would like to thank Rev. Dr Vrej Nersessian for this Â�information.

80

chapter five

at an angle, each with one raised foreleg meeting at the central axis and their ascending tails ending in dragon heads with open snouts revealing long tongues (fig.╯79). Artistic cross-fertilisation in plastic and pictorial art between Islamic Anatolia and Christian Armenian and Georgian regions is reflected in the depiction on a thirteenth-century relief on the hall portal of Kesikköprü Han (667/1268–9) in central Anatolia. It shows a lion in profile with frontally oriented head whose tail curves under its belly and then upwards to terminate in a dragon head just above the back. The head with small, pointed ears is oriented backwards to face a long-tailed bird which perches on the lion’s hindquarters and pecks at the dragon’s open mouth (fig.╯80).52 This type of imagery with a bird pecking with its beak at the open mouth of a dragon is a recurrent feature in Armenian miniatures (see figs. 72 and 73). It is probable that the dragon motif that was seen on the royal silks (parang or parniyān) or their representation on wall paintings in Central Asia, in particular Sogdiana, the later Māwarā al-Nahr of the early Islamic period, and Turkestan became a favoured emblem of royalty, next to the lion and the eagle, on medieval Islamic silks. Al-Ḥusayn ibn Muḥammad ibn ʿAlī al-Jaʿfarī al-Rughadī, known as Ibn al-Bībī al-Munajjima (“the son of the “lady,” the astrologer”), who was head of the chancellery of the Secretariat of State at the court of the Rūm Saljuqids,53 relates in his memoirs, al-Awāmir al-ʿAlāʾiyya fī ’l-ūmur al-ʿAlāʾiyya, that after his ascension to the throne,

the Rūm Saljuq sulṭān ʿAlāʾ al-Dīn Kay Qubādh I of Konya followed the ancient custom54 of beÂ�Â� stowing lavish gifts upon his emirs amongst which the receiving of robes of honour (khilʿa or tashrīfâ•›) constituted the greatest honour.55 A conflation of dragon and lion can be seen on a royal textile, a silk woven in gold, set against a pink ground, which bears the name of sulṭān ʿAlāʾ al-Dīn Kay Qubādh, probably referring to Kay Qubādh I. The textile is decorated with circular medallions enclosing a pair of addorsed regardant lions in rampant posture with frontally presented heads from whose open mouths spring volutes that form a central vegetal composition. The lions’ curling tail ends transform into dragon heads with wideopen jaws (figs. 81a and b).56 The conjunction of dragon and lion was also used on door handles, which seem to have had a special place in the tradition of southeast Anatolia. An example intended for the palace door at Diyārbakr is illustrated in a copy of the celebrated treatise on automata written by the master craftsman al-Jazarī (the nisba pointing to an affiliation with Cizre, also known as Jazīrat ibn ʿUmar), begun when in the service at the court of the Artuqid ruler Nūr al-Dīn Muḥammad ibn Qara Arslan (562/1167–581/1185) in the principality of Ḥiṣn Kayfā (Hasankeyf)57 (fig.╯134).58 In 577/1181 Nūr al-Dīn was awarded Diyārbakr as fief and transferred his court to the city. There al-Jazarī created a monumental door with cast brass plates inlaid with copper and silver for the ruler’s palace,59 now only surviving in draw-

52 ╇ To the left of the portal there is another lion relief, probably a spolium, which on account of its poor state of preservation (only the protome is extant) is difficult to assess; Gierlichs, 1996, pp.╯171–2, pl.╯17.3. When the author visited Kesikköprü Han in 2008, the building had just been renovated and the relief with the dragon-tailed lion with bird did not exist anymore. 53 ╇ Duda, “Ibn Bībī,” EI2 III, 737b. 54 ╇ The Patriarch Jacob is recorded as having chosen his favourite son Joseph from amongst his brothers to honour him with a ceremonial or royal robe, ketōnet passīm (Genesis 37,3). This custom was first recorded in Islam with the Prophet Muḥammad bestowing the burda he was wearing on the poet Kaʿb ibn Zuhayr; cf. Ibn al-Athīr, Kāmil, Cairo, 1301, vol.╯2, pp.╯133–4, cited after Stillmann, “Khilʿa,” EI2 V, 6a. 55 ╇ Al-Awāmir al-ʿAlāʾiyya, tr. Duda, 1959, p.╯95. Allsen (1997, pp.╯85, 87–9) remarks upon the “impressive continuity” of the ancient practice of the bestowing of garments. Gordon (2001, pp.╯5–6) hypothesises that the practice of investiture with “luxurious robes” began with the early nomads, “first perhaps as a semi-diplomatic relation with the sedentary culture but soon as a prerogative of a nomadic leader,” which set a prototype for later investiture ceremonies. 56 ╇ Cf. Diez and Arslanapa, 1956, pp.╯259–60; The Arts of Islam, 1976, p.╯79, cat. no. 13; Arrizoli-Clémentel, 1990, p.╯55.

57 ╇ Al-Jāmiʿ bayn al-ʿilm wa ’l-ʿamal an-nāfiʿ fī ṣināʿat al-hiyal, ed. Aḥmad Yūsuf al-Hassan, Institute for the History of Arabic Science, Aleppo, 1979; facsimile ed. of Ms. Ahmet III no. 3472, Istanbul, Topkapı Sarayı Museum, Library, Kültür Bakan-lığı, ed. Olağanüstü mekanik araçların bilgisi hakkında kitap, Ankara, 1990. Cf. Meinecke, 1996, p.╯62, n. 18. The treatise, which al-Jazarī compiled and illustrated, was completed at the beginning of the thirteenth Â�century for the Artuqid ruler of Diyārbakr. The original manuscript does not survive, but a copy was made shortly after its completion by another artist from Hisn Kayfā, a certain Muḥammad ibn Yūsuf ibn ʿUthmān al-Hiskafī, now preserved in Istanbul, Topkapı Sarayı Museum, Library. Â�Meinecke, 1996, p.╯63 and n. 26. 58 ╇ Cf. idem, 1996, p.╯139, pl.╯21a. This type of imagery is also used as marginal ornament in Armenian manuscript illumination, see the collection of sermons, copied on vellum, 607 fols., dated 1205, transcribed and illuminated in the monastery of Ghazarvankʿ by the scribe Vardan; Ms. 7729, fol. 237a. Mnatsakanyan, 1955, p.╯517, fig.╯1024. 59 ╇ Since Muḥammad ibn Qara Arslan had started only two years before his death in 581/1185 to furnish his residence, the doors may well have been made during this period, cf. Meinecke, 1989, p.╯57, and idem, 1996, p.╯62.

dragons and animals of the natural and the mythical realms ings.60 It consisted of two hinged leaves, each framed with a broad epigraphic band in Kufic containing eulogies of the ruler and fitted with cast copper alloy knockers,61 which are fashioned, according to al-Jazarī’s description:

81

Only a copy of al-Jazarī’s drawing of the door knocker and not the actual device survives. However the knockers were much imitated. Copper alloy examples of this type as well as brass plaques adorned the wooden doors of the Ulu Cami of Cizre, now in the Türk ve İslam Eserleri Müzesi (fig.╯82). According to the inscription on the upper panel of the door, it was endowed by the local Zangid atābeg Maḥmūd ibn Sānjar Shāh who resided in Cizre from 604/1208 to 638/1241.63 Hence the Cizre knockers were probably made about two decades after al-Jazarī’s creations.64 In contrast to the drawing of the knocker in the copy of al-Jazarī’s compendium of 602/1206 in which the heads of the dragons that flank the central lion-headed knob (serving as hinge and

attachment to the door) are facing each other, the dragon heads of the Cizre copper alloy knockers are portrayed addorsed, thus with heads turned away from the lion head. The dragon heads, rendered with small cusped ears and almond-shaped eyes outlined by curving lines, have characteristic wide-open jaws and a curled tip of the snout. They are shown to grasp or attack the small, curved wings which spring from an ornament that curls around their haunches. The dragons’ paired forelegs touch at the tips, their scaly ophidian bodies forming a loop at the centre, and tapering to the entwined tail ends from the tips of which spring small raptorial bird heads (fig.╯83).65 Several other dispersed examples of near-identical dragonknockers dating from the late twelfth to the early thirteenth century exist in slightly varying sizes.66 The imagery of the dragon guarding a treasure is a central theme in Niẓāmī’s romance Haft Paykar, completed in 593/1197.67 In the same work the poet, who spent most of his life in Ganja in present-day Azerbaijan, talks about “dragon-like locks on treasures rested”68 which may indicate that the dragon knockers designed by al-Jazarī and their counterparts at the Ulu Cami of Cizre and other mosques were part of a well-established iconographic genre which extended far beyond southeastern Anatolia and was well-known in the medieval world.69 As doorknockers the interlaced dragons were fastened to the doors at the entrance

60 ╇ Cf. al-Jazarī, Kitāb fī maʿrifat al-ḥiyāl al-handasiyya, tr. Hill, 1974, pp.╯192–3, 267. The earliest extant copy (Istanbul, Topkapı Sarayı Müzesi, Ms. Ahmet III, A.3472, fol. 165b) contains a colophon with the date “end of Shaʿbān 602” (about 10 April 1206) but as noted by David King in his review of The Book of Knowledge of Ingenious Mechanical Devices: Kitāb fī maʿrifat al-ḥiyāl al-handasiyya by Ibn al-Razzāẓ al-Jazarī (History of Science XIII, 1975, pp.╯284–9, esp.╯n. 4) this refers not to the date of Ms. Ahmet III, A.3472, but to the date of the manuscript from which the Istanbul manuscript was copied. 61 ╇ Height 450 cm, width 300 cm; Meinecke, 1996, pp.╯62 and 136, pl.╯18 a. Cf. al-Jazarī, Kitāb fī maʿrifat al-ḥiyāl alhandasiyya, tr. Hill, 1974, pp.╯191–5; Hauptmann von Gladiss, p.╯32, fig.╯10. For a modern reconstruction of the Artuqid palace door at Diyārbakr, see Meinecke, 1989, p.╯57, fig.╯6. 62 ╇ Kitāb fī maʿrifat al-ḥiyāl al-handasiyya, tr. Hill, 1974, p.╯194. 63 ╇ Cf. Hauptmann von Gladiss, ed. 2006, p.╯69. 64 ╇ Meinecke, 1989, p.╯57. 65 ╇ One of a pair of door knockers from the Ulu Cami of Cizre, now in Istanbul, Türk ve İslam Eserleri Müzesi, inv. no. 3749; see The Anatolian Civilisations, vol.╯3, 1983, pp.╯60–1, nos. D. 95–7; Ölçer, 1993, and idem, 2002, pp.╯98–9; Roxburgh, 2005, pp.╯130–1, 399–400, cat. no. 87. The other pair is preserved in Copenhagen, the David Collection, inv.

no. 38/1973; see The Arts of Islam, 1976, no. 194; Erginsoy, 1988, p.╯170, fig.╯141; von Folsach, 1990, p.╯196, cat. no. 323, idem, 1991, p.╯44, cat. no. 32, and idem, 1996, fig.╯110, cat. no. 362; Hauptmann von Gladiss, ed., 2006, p.╯68, fig.╯25. The lion-headed pin of the Copenhagen knocker is in Istanbul, Türk ve İslam Eserleri Müzesi, inv. no. 3750. The knockers were documented in situ by Preusser, 1911, pp.╯25–6, pl.╯36. Cf. Öney, 1969a, fig.╯17; Meinecke, 1989, p.╯56, fig.╯5, and idem, 1996, p.╯136, pl.╯18b. 66 ╇ Door knocker, preserved in Berlin, Museum für Islamische Kunst, inv. no. I.2242; see Kühnel, 1925, p.╯136, fig.╯100 and idem, 1950, p.╯13, fig.╯16; Meinecke, 1989, p.╯56, fig.╯4 and idem, 1996, p.╯139, pl.╯21b; Gierlichs, 1993, p.╯41, cat. no. 7; Grube and Johns, 2005, p.╯230, fig.╯77.6. Two examples are in the Nasser D. Khalili Collection of Islamic Art, London, inv. nos. MTW1407&1428; height 40 cm, width 40 cm; see Piotrovsky, ed., 2000, p.╯98, cat. no. 3; Bilici, 2006, figs. 9, 10. 67 ╇ Tr. Meisami, 1995. 68 ╇ Idem, p.╯221. 69 ╇ Another type of copper alloy knocker, now preserved in Paris, Musée du Louvre, features a circular openwork ring composed of five confronted pairs of dragon protomes with wide-open jaws held by a central lion-headed knob and in earlier publications has been attributed to Iran or Syria and the twelfth century (L’Islam dans les collections nationales, 1977, p.╯99, cat. no. 153) but has now been assigned a Jazīran

Then I made for each leaf [of the door] ring [i.e. a knocker] from cast brass in the shape of two connected serpents, the head of each facing the head of the other. Their mouths are open as if they wished to devour the head and neck of a lion. This lion’s head and neck are the extension of an iron staple which is nailed to the door. The fangs of the serpents are in two holes in the lion’s neck...62

chapter five

82

to a sacred building where they served to protect the building. In their function as “guardians” of entrances to sacred places or to “treasures” they were endowed with an apotropaic function.

...transversely between the centres of the pillars is an axle on which are two serpents, the claws of each one grasping the axle, its tail around the axle like a ring, its head tilted backwards, the mouth open as if to swallow the falcon, 71 the edge of the lower lip is touching the front of the castle [the tower], and the upper lip is spread, with the two fangs bared.72

It is interesting to note that the representation of the serpent body coiled around the central pole also has an analogy in ancient Hindu mythology, in which the body of the great cosmic dragon Vāsuki (“Possessor of Treasures”) serves as giant churning rope to rotate Mount Mandara in the well-known myth of the “Churning of the Ocean” (Samudramanthana) which produces the nectar (amṛta) of divine immortality.73 In the course of the story, the demons Rāhu and Ketu become involved, an episode which is discussed below in chapter 8. The motif of the serpent that appears to threaten the birds is based on a related design by an earlier craftsman, acknowledged by al-Jazarī,74 the late eighth-century Pseudo-Archimedes, 75 whose work included frequent representations of the dragon on clocks, knockers or as spouts of ewers. From this it appears that motifs ultimately derived from Hellenistic sources enjoyed widespread popularity in the medieval Islamic world. And although he has suggested no symbolic meaning for his creations, al-Jazarī may well have taken for granted that his audience was well aware of the overall symbolism, in which case the omission is inconclusive. The depictions may therefore have well been part of an apotropaic and, perhaps, talismanic tradition. On textiles of the early medieval period the imagery of the elephant, the giant among animals, was fairly frequent. It has been preserved on several Umayyad76 and Byzantine textiles77 made in imitation of late Sasanian examples.78 In the medieval Islamic period elephants had great value and regal status and as such were often richly caparisoned.79 Such noble pachyderms are shown together with disproportionately diminutive quadruped dragons on a saddle cloth,

provenance and a late twelfth- or early thirteenth-century date (L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯232, cat. no. 160). Paris, Musée du Louvre, Département des Antiquités orientales, Section Islamique, inv. no. MOA 97. Rouault and Masetti-Rouault, eds., 1993, cat. no. 471 (colour photograph). 70 ╇ Tr. Hill, 1974, p.╯48. 71 ╇ Idem, p.╯59 and 1984, p.╯238. 72 ╇ Idem, p.╯68. 73 ╇ Cf. Long, 1976, pp.╯177–81. 74 ╇ Kitāb fī maʿrifat al-ḥiyāl al-handasiyya, tr. Hill, 1974, pp.╯272, 279 and idem, 1998, ch. XII, “The Banū Mūsà and their “Book of Ingenious Devices,” p.╯73; Meinecke, 1996, p.╯63 and n. 27. 75 ╇ Possibly the work of a late eighth-century Muslim inventor who based his work on several earlier sources including probably Philon and Heron. Cf. Kitāb fī maʿrifat al-ḥiyāl al-handasiyya, tr. Hill, 1974, p.╯10. 76 ╇ Jerusalimskaja, 2000, p.╯114 and n. 13.

77 ╇ Grabar, 1951, p.╯35. Cf. the “Elephant Silk” from the tomb of Charlemagne, Treasury of Aachen Cathedral (Grabar, 1951, p.╯35, fig.╯3); or, an eleventh-century Armenian painting showing king Gagik-Abas of Kars (1029– 1064), sitting in state with his wife and daughter on a low throne covered with a textile decorated with pearl roundels containing elephants (illustration from the Gospel book of Gagik-Abas of Kars, perhaps painted in Kars, c. 1050. Jerusalem, Armenian cathedral of Saint James, Ms. 2556, fol. 135bis. Der Nersessian, 1945, p.╯119, pl.╯XXIII.2; Glory of Byzantium, 1997, p.╯353, photograph to the left); king Gagik himself is shown to wear a garment with a comparable pattern and a tiraz with Kufic letters at the upper arms, indicating an Islamic origin. This led André Grabar to suggest that the garment may have been a gift by a Muslim sovereign. It moreover shows that this pattern was evidently linked to both Armenian and Islamic contexts. 78 ╇ Grabar, 1951, pp.╯40–2. 79 ╇ Bosworth, 1963, p.╯117.

e.╇ The dragon and the elephant The ingenuity of al-Jazarī was further proven when later in his career he entered the service of Qutb al-Dīn Suqmān (581/1185–597/1201) and began inventing mechanical devices, such as the so-called “elephant water-clock.” This water-clock was amongst the devices originally compiled and illustrated by al-Jazarī, and is shown on a leaf from the Kitāb fī maʿrifat al-ḥiyāl al-handasiyya which was copied by Muḥammad ibn Yūsuf ibn ʿUthmān al-Ḥaṣkafī at Diyārbakr, and dated to the end of Shaʿbān 602/about 10 April 1206, now preserved in the Topkapı Sarayı Library in Istanbul (fig.╯84). “The passage of constant hours,”70 the concept of the ebb and flow of time and the seasons, is indicated by a scale adjusted by a small person seated atop an elephant passant within a tall tower that supports the device. Perched on top of the tall domed tower, a bird spins around every half an hour, whereupon the elephant driver strikes his elephant and another figure causes the falcon to release a pebble. In his description of the central mechanism al-Jazarī states:

dragons and animals of the natural and the mythical realms

83

known as the celebrated Khurasanian “Elephant Silk,” which was preserved as the shroud of Saint Â�Jodokus (Saint Josse) (fig.╯85a).80 The silk samite survives as a group of fragments which show two superimposed pairs of large, confronted eleÂ�phants in the central field. The great beasts are lavishly caparisoned with saddle blankets, their heavy feet with articulated toes being singled out for emphasis. These potent symbols of power are each paired with a disproportionately tiny quadruped dragon with a very long sinuous neck almost disappearing between its legs. The four-legged dragons rest on feet with separated toes and long, needle-pointed talons, their stiffly raised wings terminating in a small curled tip following post-Sasanian conventions, and their raised undulant tails terminating in a pointed tuft of hair (fig.╯85b).81 The depiction is framed above and below by a Kufic inscription in mirror image which includes glory and good wishes to the owner, Qāʾid Abu ’l-Manṣūr Bākh-tigīn, the Turkish āmir of Khurasan, whose execution in 349/960–1 by order of his Samanid sovereign ʿAbd al-Malik ibn Nūh presents a terminus ante quem for dating the textile.82 The silk is framed in the style of a carpet by a procession of camels with flying scarves and roosters with beaded neckbands in each of the corners. It thus provides a combination of Sasanian style elements with Central Asian aspects such as the dragons and the camels. Belenitskii and Bentowitsch note analogies in the motifs to Sogdian art, in particular with the eighth-century wall painting frieze of the Red Hall at the Sogdian palace at Varakhsha, northwest of Bukhara, featuring elephants being attacked by long-necked mythical creatures.83 The

authors suggest that the dynamic representation of dragons and elephants on the Varakhsha wall paintings may be a translation of the same theme onto a textile in a more static manner and without any apparent connection to the combat theme portrayed at the Varakhsha wall paintings.84 The name of the pachyderm certainly appears as title and opening verse of the early Meccan sūra 105 of the Qurʾān alluding to the failed expedition of the Yemenite king Abraha against Mecca. His troops had been accompanied by an elephant which on arriving at the frontier of Meccan territory, knelt down and refused to advance further towards the city. The episode is related to the tradition according to which the birth of the Prophet is said to have taken place at this time, in the “year of the Elephant.”85 These majestic quadrupeds had a special status and were covetously guarded by the sulṭāns. Rulers bestowed elephants only as a great favour or when a commander was appointed to a particularly responsible post.86 Among the Būyids, ʿAḍud alDawla had war elephants (fuyūl muqātila) which he used against his cousin Bakhtiar ibn Muʿizz al-Dawla in 366/977.87 It was also the Ghaznawid sulṭān Masʿūd’s favourite mount for hunting.88 On formal occasions involving solemn processions (mawākeb) the sulṭān rode on an elephant, as in September 422/1031, when Masʿūd proceeded to the plain of Shābahār outside Ghazna to preside over a session of the maẓālim court held for the redress of wrongs.89 The representation of lavÂ�ishly caparisoned elephants must therefore have served as a proclamation of power and authority for the Turkish āmir Qāʾid Abu ’l-Manṣūr Bākhtigīn. The unusual depiction of elephants together

80 ╇ The saint whose seventh-century tomb is in Runiacum, present-day Saint-Josse-sur-Mer, near Pas-de-Calais, Musée de Louvre, Paris, inv. no. 7502; Pope and Ackerman, eds., 1938–9, repr. 1964–81, vol.╯6, pl.╯981; text, vol.╯3, pp.╯1928– 39 and 2002; Ettinghausen, Grabar and Jenkins-Medina, 1987, repr. 2001, p.╯244, fig.╯260; von Wilckens, 1991, p.╯48, fig.╯43, and p.╯347, n. 47; technical analysis in Bulletin de liaison du CIETA 33, 1971, pp.╯22–57. A silk fragment of uncertain date featuring a related elephant pattern is found at Siegburg; cf. Meredith-Owens, “Fīl,” EI² V, 690b. Étienne de Blois, count of Bologne and commander of the first Crusade, is said to have procured the precious textile together with his two brothers Godefroy de Blois and Baudoin, and to have offered it to the Abbaye Saint-Josse of which the counts were benefactors. Cf. La France romane, 2005, p.╯177, cat. no. 124. 81 ╇ It is interesting to note that closely related depictions of these unusual, extremely long-necked quadruped dragons with their uplifted feathery wings and vertically raised tails are prominently portrayed flanking a tree-like composition in a headpiece of the Armenian Gospel book of Mughni, datable

to the second half of the eleventh century and probably executed somewhere in the Ani region (Yerevan, Matenadaran Ms. 7736, fol. 9r); see Gevorkian and Abgarian, eds., 1996, pl.╯54. 82 ╇ Combe, Sauvaget and Wiet, eds., 1933, vol.╯4, no. 1507, with bibliographical data. Cf. Ettinghausen, Grabar and Jenkins-Medina, 1987, repr. 2001, p.╯401, n. 66. 83 ╇ Belenitskii, 1980, p.╯228; Belenitskii and Bentowitsch, 2000, pp.╯44–5. 84 ╇ Eidem, 2000, p.╯45. 85 ╇ See al-Azraqī, Kitāb Akhbār Makkah, ed. Wüstenfeld, 1861, vol.╯4, German tr., pp.╯48–9. 86 ╇ Bosworth, 1963, p.╯117. 87 ╇ See Christensen, 1944, p.╯208. Anon., Taʾrīkh-i Sīstān, ed. Bahār, M.S., Tehran, 1314/1935, p.╯206; Miskawayh, Tajārib al-umam, tr. and ed. Margoliouth, 1921, vol.╯5, p.╯402. 88 ╇ Bosworth, 1963, p.╯118. 89 ╇ Bayhaqī, Taʾrīkh-i Masʿūdī, ed. Ghani and Fayyāz, Tehran, 1324/1945, pp.╯372–3, as cited in Bosworth, “Court and courtiers: In the Islamic Period to the Mongol Conquest,” EIr.

chapter five

84

with dragons appears thus to combine the royal symbolism accorded to the elephant with that of the dragon. The expression Zanda-pīl, “the furious elephant” is, moreover, a common metaphor used in Persian epic poetry for a pahlavān (hero),90 such as the well-known epithet of Rustam being Â�pīltan, “elephant-bodied.”91 In Wīs u Rāmīn Fakhr al-Dīn Gurgānī praises Wīs’ brothers, the trusted champions of Rāmīn, as “two elephants, two lions like no others,” thereby putting both creatures on a par with each other.92 A proverb from an Afghan manuscript which states that “only Zāl can manage the elephant” (Zāl most likely representing the nephew of Ẓaḥḥāk, the historicised hominoid dragon, in the late eleventh-century epic Kūsh-nāma), might furthermore point to the possibility that the elephant was, as Gianroberto Scarcia has shown, a significant component of the ancient legends of the eastern frontier province of Zābulistan.93 *** Aspects of the dragon’s association with other animals are encapsulated in a figural scheme porÂ�trayed on a gilded copper alloy lantern with a pyramidal cover in openwork which has been dated to the second half of the thirteenth century,

╇ Cf. Scarcia, 1967, p.╯42. ╇ Steingass, 1892, repr. 1981, p.╯269. See also Shāh-nāma, vol.╯3, tr. and ed. Mohl, 1838–1878, p.╯157, l. 202; Kowalski, 1939–49, p.╯94. 92 ╇ Tr. and ed. Davis, 2008, p.╯388. In a eulogy the twelfthcentury poet Rashīd-i Waṭwāṭ similarly associates the strength of both animals when he speaks of “lion-hearted and elephant-bodied warriors”; see p.╯39. 90 91

and was used in the Mevlevī Dergāhı in the Rūm Saljuq capital of Konya.94 The front face Â�features double doors with a lobed arch that allow for the insertion of a candle into the cuboid body of the lantern. This small lamp, which could serve as a miniature architectural model, shows two upÂ�right addorsed “Saljuq-style” dragons crowned with pointed ears with gaping mouths and looped bodies that flank a large, six-petalled star-rosette. The dragons face towards open-mouthed lions, each with a raised foreleg and tail that arches over the back ending in a dragon head with closed jaws. The composition is set just above the double doors, which are flanked by double-headed birds of prey with looped necks, outspread wings and legs whose tail feathers terminate in the form of a large, inverted palmette bud. A cursive inscription at the lower edge of the front face reveals the name of the maker, Ḥasan ibn ʿAlī al-Mawlawī. The sides, back and cover of the lamp are covered with floral and stellar ornaments. The composition of the rosette flanked by dragons and in turn by dragon-tailed lions surmounting doubleheaded eagles can be assumed to epitomise the power of the sulṭān.95 The fact that the scheme is portrayed on a lamp connects it, moreover, with the theme of light and hence also with the symbolism of the luminaries.

╇ Scarcia, 1967, pp.╯44–5. ╇ Erginsoy, 1978, pp.╯413–4, figs. 196 a and b (dragontailed lion); The Anatolian Civilisations, vol.╯3, 1983, p.╯75, cat. no. D. 138. 95 ╇ Konya Müze Müdürlüǧü, inv. no. 400. Erginsoy, 1978, pp. 412–9, figs. 196 a–e; Kalter and Schönberger, 2003, p.╯74; Turks, 2005, p. 394, cat. no. 70. 93 94

the dragon in scenes of combat

part three

the dragon and its royal and heroic associations

85

86

chapter six

the dragon in scenes of combat

87

chapter six

The dragon in scenes of combat

a.╇ The dragon combat in ancient myth Combat with the water-controlling dragon Myths of dragon-like creatures and the vanquishing of these dragons were well-known, if not universal, in the traditions both of the IndoEuropeans and of the Near Eastern civilisations with whom the Persian speaking peoples came into contact from at least the first half of the first millennium bc.1 In these contexts the mythical creatures are suited ideally to play the role of adversaries as they represent forces or elements that interfere with the correct order or functioning of the world, and they are defeated by deities, kings or heroes who shape and organise the cosmos. Through their victory the latter acquire authority and power over the newly ordered world. The iconography of the dragon combat or encounter, part of the Indo-Iranian literary theme of heroic mythological exploits, draws on the immemorially ancient epic theme of this quest, an ever-recurring motif even in cultures that are culturally and geographically far removed from one another. The cosmogonic quality of the dragon-slaying myths evidently lay in the fact that in order to construct or defend world order, the god or hero had to destroy the primeval or chaotic dragon. During the remote period of Indo-Iranian unity, an age that long played a key role in the later

1 ╇ Cf. Watkins, 1995, p.╯299. The most ancient known traditions about vanquishing dragons go back to the Sumerian, Akkadian and Egyptian mythologies of the first three millennia bc. The god Enlil defeats a monstrous dragon, the Labbu, in a Sumerian text. The god Marduk vanquishes Tiamat and her conscripts in the Akkadian epic of creation, Enūma Elish, of Babylon. In the mythology of the ancient Syrian city of Ugarit the god Baal overcomes the monsters Yamm and Mot. The dragon Apopis is dispatched by the god Seth in Egyptian mythology. In the Hittite texts of Bogazköy, the dragon Illuyanka fights the weather god. 2 ╇ Skjærvø, “Aždahā I,” EIr. In Greek mythology, Zeus slays the monster Typhaon/Typhon that has a hundred snake heads (Hesiod, Theogony 825–626). 3 ╇ Watkins, 1995, p.╯299.

cultures of the Near East and India, the IndoIranians may well have imagined dragons harbouring and restraining the heavenly waters, so causing drought, and not releasing them until overcome by a god or hero.2 The oldest texts of the Indo-Iranians, the Vedic texts (1500–1000 bc), are composed in Sanskrit and usually refer to the dragon by the most common word áhi-, while in the old Iranian Avestan texts the term used is azhi, which originally meant only “snake, serpent.”3 Intoxicated and strengthened by the ritual potion of Soma, the divine hero Indra fulfils a cosmogonic act by dismembering the primary denizen of the forces of chaos, the cosmic serpent-monster Vá¹łtra (“the enveloper”) whom the Rigveda calls “the first-born of dragons” (prathama-jấm áhīnām, 1.32.4).4 Indra, who carries the IndoIranian epithet “smashing resistance, obstacles” (ṷṛtra-ǰhan),5 is eulogised in the Sanskrit text of the Rigvedic hymn (1.32) with the words: I tell now the manly deeds of Indra, The foremost which he did armed with the cudgel. He slew the serpent, drilled through to the waters, He split the belly of the mountains.6

Importantly, the stone dragon is split into two halves, the upper half forming Dyáv- “the sky” and the lower kṣám- “the earth.”7 The defeated Vá¹łtra is referred to in the Rigvedic texts as áhi-, “serpent,”8 and dāsá, “the pent-up waters with the dāsá as husband, the Serpent as guardian”

4 ╇ Indra similarly defeats the monster Vala (valá-, meaning “enclosure”) who may have been conceptionally identical with Vá¹łtra at an earlier stage of the myth being derived from the same root, val-/var- “to cover, to enclose”; Vala thereupon frees the goddess of dawn, Ushas, whom he had imprisoned. Janda, 2010, pp.╯27, 65, 247, 266, 270. 5 ╇ Watkins, 1995, p.╯299. 6 ╇ Cited after idem, p.╯304. It is interesting to note the ambiguity surrounding the killing of the primordial dragon conveyed in the Purāṇic accounts in which, paradoxically, Vá¹łtra is said to be a brāhmaṇa and Indra is decried for committing brahmanicide, the most heinous of all sins. Cf. Long, 1976, p.╯172, n. 3, and p.╯192, n. 29. 7 ╇ Janda, 2010, pp.╯45–70, esp.╯27, 63, 79, 266, 270. 8 ╇ Rigveda 1.32.5, 1.32.8, 1.8.10, 1.61.8, 1.103.7.

chapter six

88

(Rigveda 1.32.11). The word dāsá is translated by Calvert Watkins as “hostile demon,” “enemy,” but also “non-āryan, barbarian,” as well as “slave.”9 Indra released the waters “surrounded by the serpent” (páriṣṭhitā áhinā) and struck down the dāsá (Rigveda 2.11.2).10 The serpent and the sobriquet in Indo-Iranian may be linked as aǰhi- dāsá-, from which originate the Vedic áhi- … dāsá- as well as the Iranian name of the dragon Azhi Dahāka.11 Significantly, one of the principal functions of the divinity was not necessarily to kill the huge dragon, but rather to destroy an obstacle, vәrәthra-, which tried to withhold the flow of the generative waters, thus to fulfil a cosmogonic task.12 The water-controlling aspect of the dragon has analogies in the figure of Vá¹łtra as well as Azhi Dahāka who becomes the hominoid Azhdahāk.13 As has been suggested, the mythological character of Azhdahāk may well be older than the Zoroastrian texts which first record his name, since figures of anthropomorphised dragons already appear in Bronze Age Central Asia.14 In the Avestan sources the mortal hero-king Thraētaona (Pahl. Frēdōn, N. Pers. Farīdūn, Arm. Hruden), whose abode was presumably the Central Asian steppes,15 is associated with the feat of conquering the dragon Azhi Dahāka with his mighty club and thus restoring order and

justice.16 In the Avesta the dragon is described as three-headed and six-eyed (Yasht 14.38.40; 9.8).17 Another great Avestan dragon-fighter is Kərəsāspa who slays Azhi Sruuara (or Azhi Zairita), “the horned Azhi, the horse-swallowing, man-swallowing, venomous, yellow-green …” (Yasna 9.11; cf. Yasht 19.40).18 In Zoroastrian Pahlawī literature the hero also defeats the sea monster Gandarəβa (Skr. gandharvá-) in Lake Vārukasha after having offered a sacrifice to the goddess Ardvī Sūrā Anāhitā.19 As has been recognised long ago, the hymns in both Indic and Iranian traditions relate to a traditional mythology which must be shared Indo-Iranian patrimony.20 The heroic feats associated with the two ancient Iranian festivals of Nawrūz and Mihragān, respectively, at the spring and autumn equinoxes of the solar year are of note. The vernal New Year festival, Nawrūz, the official beginning of the year in Sasanian times, was instituted according to legend by the primordial king Jamshīd (Jim), the legendary Kayānid emperor of Iran.21 Al-Bīrūnī records the Iranian tale of a demon called Iblīs22 who had caused a terrible drought and famine by his evil spells and threatened to destroy all life on earth. Jamshīd marched to the demon’s abode and conquered him, releasing the spell and causing new fertility to burst forth.23 On his victorious return

╇ Watkins, 1995, p.╯311. ╇ Idem, pp.╯311–2. 11 ╇ The material on Azhi Dahāka has been comprehensively discussed by Skjærvø, “Aždahā I,” EIr. 12 ╇ Boyce, 1975, repr. 1996, p.╯64; Gnoli, “Bahrām,” EIr. 13 ╇ Watkins, 1995, p.╯312. 14 ╇ Cf. Francfort, 1994, figs. 3–5; Kuehn, 2009, pp. 43–67. 15 ╇ Boyce, 1975, repr. 1996, p.╯101. The association of the dragon-fighter with that of a solitary figure fighting in a foreign country is interesting. In the Avesta the dragon-fighters, Thraētaona and Kərəsāspa, are associated with “frontier heroes” of the Central Asian steppes, in particular the grazing lands of southeast Afghanistan, Thraētaona being born in Varena and Kərəsāspa coming from the Pishin plain to the south Lōra river (lower Urvadhā). Sarkhosh Curtis and Stewart, eds., 2005, pp.╯43–4. Cf. Boyce, 1975, repr. 1996, p.╯101 and n. 58; Monchi-Zadeh, 1975, p.╯114. The great dragon-fighter of the national Iranian epic, Rustam, is similarly presented in Firdawsī’s Shāh-nāma as a non-Iranian, coming from outside Iran (Melikian-Chirvani, 1998, p.╯183), and was known as a Central Asian Saka/Scythian, Skythes, that is, Sagzī (idem, 1998, p.╯193; P’yankov, 2006, p.╯505. The term Sagzī was also used of a native of the province of Sīstān, originally called Sakastān/Sagistān, arabicised to Sijistān. Cf. the discussion of the term in Shahbazi, 1993, pp.╯157–8; also Bailey, 1958, pp.╯131–54, esp.╯132). 16 ╇ Monchi-Zadeh, 1975, pp.╯103–5; Boyce, 1975, repr. 1996, p.╯63. 17 ╇ Watkins, 1995, p.╯313. 18 ╇ Idem, p.╯319; Nyberg, 1933, pp.╯336–52, repr. 1975, pp.╯379–95. 19 ╇ Cf. Monchi-Zadeh, 1975, p.╯114.

╇ Watkins, 1995, p.╯314 and n. 3. ╇ See al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯13–4. Christensen, 1934, pp.╯146–8; Zaehner, 1961, p.╯138. 22 ╇ The name Iblīs is a Qurʾānic designation for the devil, although he is referred to as a jinn (sūras 18, 50; and 55, 15) and occurs less frequently than al-shayṭān which is used to designate the devil in the context of his maleficent intentions towards man. In the Shāh-nāma Iblīs is termed a demon (dīw), capable of transforming himself into another being including taking on the form of a dragon; dīws having a strong pre-Islamic (Zoroastrian) background (going back to the Old Iranian word daiva and the Avestan daēuua). Interestingly, among the few ḥadīth in which the name appears, there is one in which the Prophet speaks of Iblīs as having a throne “on the Waters” thus underlining the connection of Iblīs with the element of water (Ṣaḥiḥ Muslim IV, p.╯1472; cited after Algar, “Eblīs,” EIr). Azhi Dahāka/Ẓaḥḥāk is seen as leader of the demons (dīws) that corrupted Jamshīd who thereupon taught men, who were then vegetarians, to eat animal meat (whereas the Avesta forbids blood sacrifices), he thence became too proud, lost his purity, giving himself up to profane pleasures, and thus was forsaken by his glory (khvarәnah). Cf. Huart [Massé], “Djamshīd,” EI2 II, 438b. 23 ╇ On the problematic hypothesis that the ancient IndoIranian New Year festival contained, among other things, a ritual combat between a deity, or a ruler as his representative, and a dragon that has captured creative parts of nature such as rain, cattle or women and thus threatens procreation and life as such, see Widengren, 1965, pp.╯41–9. Cf. most recently Janda, 2010, pp.╯69–70, 102; also Nylander, 1974, pp.╯144–6.

9

10

20 21

the dragon in scenes of combat Jamshīd appeared before his people radiating like a second sun.24 Thereafter, al-Bīrūnī relates, Jamshīd’s rule was just and prosperous “until the time when Bīwarāsp [Ẓaḥḥāk/Dahāk]25 appeared who killed [him] and subdued his realm.”26 In turn, as al-Bīrūnī reports, the dragon Â�BīwaÂ�rāsp/Dahāk is vanquished by the dragonfighter Farīdūn (the Avestan Thraētaona) during the festival of Mihragān, which he then instituted to celebrate his victory over the usurping tyrant, whom he bound in fetters and imprisoned in Mount Damāwand,27 an extinct volcano in northern Iran as well as the highest peak in the country. Significantly, according to various traditions, the sun appeared for the first time on the day of Mihragān.28 Al-Bīrūnī further states that the Persian theologians conceived the prestigious ancient festival of Mihragān: ...as a sign of resurrection and the end of the world, because at Mihragān that which grows reaches its perfection.29

The festival was also “a time for rallying the forces of good to oppose the demons of coming winter and darkness.”30 It is also interesting to note that the celebration of Mihragān was accorded great importance at the courts of most rulers in the Turko-Iranian environment until the Mongol invasion.31 A typology of such festivals shows the constant need of society to struggle against marginal situations such as a savage exterior or 24 ╇ Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, p.╯202. Cf. Carter, 1974, pp.╯185–6 and n. 64. Al-Bīrūnī (idem, pp.╯199– 201) furthermore mentions presents being brought at the New Year festival, Nawrūz, which is evidenced by further sources; see Ehrlich, 1930, pp.╯95–101. 25 ╇ See al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯16–8. In the Bundahishn Dahāk is referred to as the one “whom they call Biwarāsp” (lit. “possessor of ten thousand horses”); cf. Yasht 5.28–31, in which Azhi Dahāka sacrifices “a hundred stallions, a thousand oxen, and ten thousand sheep” to to the fertility goddess Ardvī Sūrā Anāhitā, the Yazata (“venerable one”) of all waters and fertility, to obtain a wish. See al-Ṭabarī, Taʾrīkh al-rusul wa ’l-mulūk, tr. and ed. Brinner, 1991, p.╯18, n. 107. 26 ╇ Kitāb ’l-Āthār, tr. and ed. Sachau, 1876–8, p.╯202. 27 ╇ Idem, pp.╯207–10, 213–4. Also al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯35–6. See Widengren, 1966, pp.╯435, 439. Cf. Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯111–3, ll. 19–528, p.╯115, l. 10. Zaehner, 1961, p.╯139; Calmard, “Mihragān,” EI2 VII, 15a. 28 ╇ Calmard, “Mihragān,” EI2 VII, 15a. 29 ╇ Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, p.╯208. Cf. Boyce, 1983, p.╯802. 30 ╇ Boyce, 1983, p.╯802. 31 ╇ Calmard, “Mihragān,” EI2 VII, 15a. According to the Ghaznawid historian Bayhaqī’s narrative, the Â�Ghaznawid sulṭān Bahrām Shāh ibn Masʿūd regularly celebrated at court the two ancient Iranian seasonal festivals of Nawrūz and

89

awesome chaos in order to recreate social reality as Peter Berger and Thomas Luckmann have aptly pointed out:32 All societies are constructions in the face of chaos. The constant possibility of anomic terror is actualized whenever the legitimations that obscure the precariousness are threatened or collapse. And in such situations, or more regularly in ceremonially created periods of crisis – literally: separation between two eras, situations, periods – a ‘deep legiÂ�timacy’ is required, referring to a mythical reality outside ours, ‘the other reality’, lying beyond the borders of history and space, an eternal truth that existed before time but still exists behind it and behind our reality, and occasionally mingles with ours in ‘periods of exception’.

Zoroastrian myth also influenced ArmenianÂ� legÂ�ends. The early dragon-slaying myth of the Avestan victory god Vәrәthraghna (O. Iran. Vá¹łtraghan, N. Pers. Bahrām), “the best-armed of the heavenly gods, the strongest in strength, ... the most victorious in victory, ... the most glorious in glory,” may have been conserved in the Armenian national hero Vahagn (who carries the epithet vishapakʿagh, “who handles dragons”).33 According to the Armenian historian Moses of Chorene, Vahagn fought with and vanquished dragons, his deeds equalling those of Heracles.34 The golden aspect of the god is reflected in the song of the birth of Vahagn, the oldest testimony of Armenian literature,35 which also survived in Mihragān. Fallāḥ Rastgār, pp.╯431–3, as cited in Bosworth, “Court and courtiers,” EIr. Stripped of their original Zoroastrian religious significance, these traditional festival patterns had also survived under the ʿAbbasid caliphs, as attested by the verses of various contemporary Arabic poets; for instance, the Nawrūz poem by Ḥusain ibn Ẓaḥḥāk Khalīʿ and the Mihragān poem attributed to the caliph al-Maʾmūn (see Masʿūdī, Kitāb murūj al-dhahab, tr. and ed. Barbier de Meynard and de Courteille, 1917, vol.╯8, pp.╯277–8, 340–2). In the sulṭānate of his predecessor, Masʿūd ibn Maḥmūd of Ghazna, the exchange of customary gifts and tribute for this festival was accompanied by much drinking of wine (rasm) which was associated with the celebration of Mihragān since Achaemenid times (Fallāḥ Rastgār, pp.╯431–3; Browne, 1920, vol.╯1, pp.╯475–6; Â�Clinton, 1972, p.╯136; for the verses of Manūchihrī, see also Â�Hanaway, 1988, pp.╯69–80, as well as Bayhaqī’s Taʾrīkh-i Bayhaqī). 32 ╇ Berger and Luckmann, 1971, p.╯121. 33 ╇ Patmutʿiwn Hayocʿ, tr. Langlois, 1872, p.╯40 and n. 1. The dragon-slaying characteristic of Vahagn has also been interpreted as secondary or due to local elements (Benveniste and Renou, 1934, p.╯80). While Iran conserved the ancient deity, India may have fused the deity with the hero and thus Indra was associated with the characteristics and functions of Vá¹łtraghan; see Gnoli, “Bahrām,” EIr. 34 ╇ Patmutʿiwn Hayocʿ, tr. Langlois, 1872, p.╯41 and n. 1. 35 ╇ Cf. Russell, 2004, p.╯357.

90

chapter six

the records of Moses of Chorene.36 The Armenian Vahagn was identified with the Sun, as also reported by the classical Armenian commentator on Genesis, “some worshipped the Sun and called it Vahagn.”37 This Armenian hymn from the Zoroastrian tradition finds, as Watkins notes, a parallel in the Vedas, in particular the hymning of Apām Napāṭ, literally the “Son of the Waters” (Rigveda 2.35), as deity and as sacrificial fire, which is probably of Indo-Iranian date.38 After an inconclusive contest for the gleaming khvarәnah between the dragon Azhi Dahāka and Fire, the khvarәnah escapes to Lake Vārukasha and at the bottom of the deep lake Apām Napāṭ grabs it (Yast 19.45–54).39 In another early Armenian epic from Sasun, southwest of Lake Van,40 the hero Samsar (Sanasar), who was conceived after his mother drank from a magical spring,41 slays a dragon (vishap) that controls the spring of water which feeds the city and must receive a girl in sacrifice each year.42 Parts of Central Asia, in particular in the Kushāṇa empire, were of course Buddhist from the first centuries of the millennium. It is interesting to note that in the northwest of the empire – which seems to have been the only geographical region where he fulfilled such a role – the Buddha appears in the role of a dragon-fighter.43 He overcomes the great dragon-king (nāgarāja) Apalāla, the dragon who lived in the mountain spring that was the source of the Śubhavāstu (Swāt) river

and who terrorised the region in northwest Pakistan on the border with Afghanistan by causing periodic destruction to the harvest though violent storms and floods.44 There are parallels to the great Indo-Iranian dragon-fighter Indra vanquishing Vá¹łtra, for Apalāla like Vá¹łtra has aquatic features. The Buddha appears to be taking over the role of Indra when he strikes the mountain, within which the dragon resides, with Vajrapāṇi’s thunderbolt.45 The Buddha is also said to have tamed and converted another nāgarāja, named Gopāla, in “Chen-t’o-lo” (Gandhāra (?)).46 This feat is recorded in Indian texts that were translated into Chinese, such as the early fifth-century translation of the Buddhānusmṛtisamādhi Sūtra (“Sūtra of the Ocean-like Samādhis of Buddha Visualisation”),47 and is described as taking place at a cave located in a mountain at Nagarahāra near present-day Haḍḍa/Jalalābad in eastern Afghanistan.48 A pond near the cave is believed to have been the lair of a dragon-king (referred to as Gopāla in early Indian texts), his five ogress-consorts, who had been transformed into female dragons, and various small dragons that terrorised the region with adverse weather conditions. The Buddha was summoned to the area where he and his attendants tamed the dragon king. Once converted, the latter implored the Buddha to stay on for 1500 years whereupon the “World-Honoured One” left his “luminous mystical form” or “projection” (shadow) image in the cave.49 A related

36 ╇ Patmutʿiwn Hayocʿ, tr. Langlois, 1872, pp.╯40–1; Watkins, 1995, pp.╯253–4. Cf. Russell, 2004, pp.╯357–70. 37 ╇ Alishan, G., Hin hawatkʿ kam hetʿ anosakan krōnkʿ Hayokʿ (“The Ancient Faith or Pagan Religion of the Armenians”), Venice, 1910 ed., p.╯187, as cited in Russell, 1987, p.╯270. 38 ╇ Watkins, 1995, p.╯254. Also Janda, 2010, p.╯255. On Apām Napāṭ, see Oldenberg, 1894, repr. 1977, pp.╯118–20 and n. 1; Boyce, 1975, repr. 1996, pp.╯41, 44–50. Cf. Yasht 19; Boyce, 1984, pp.╯29–30. 39 ╇ Gershevitch, 1959, p.╯59. 40 ╇ Russell, 2004, p.╯879 and n. 8. 41 ╇ Azarpay, 1981, p.╯99. On the motif of the supernatural conception and birth of heros, see Schirmunski, 1961, pp.╯27–8. 42 ╇ Russell, 2004, p.╯1123. The Hebrew Bible similarly contains many traces of ancient mythology, wherein Yahveh, in primeval times, overcomes monsters that are extremely similar to the dragon-like beings defeated by the various Near Eastern gods. Added to names already present in the more ancient Ugaritic texts, such as Yamm, Mavet or Mot, were names such as Peten, Nahash, Rahab, Leviathan, Tannin, Behemot. 43 ╇ Strong, 1992, p.╯27. Dragon-slaying or -taming myths are not unknown in East Asian cultures. In the Chinese context the Buddha is also known in the role of “Huan Long Shi, the dragon-tamer↜” which may however be informed by the Central Asian prototypes. Cf. Wu Hung, 1986, p.╯270.

44 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, pp.╯121–3. A list of ancient Indian sources that were subsequently translated into Chinese recording the story of the Buddha’s subduing of Apalāla and Gopāla in the Kushāṇa kingdom is provided by Rhie, 1999–2002, pp.╯116–7. Cf. Strong, 1992, pp.╯26–7. The theme of the Buddha converting Apalāla was a very popular subject in Gandhāran art; see, for instance, Gandhara, 2009, cat.nos. 173 (from the Dharmarājikā stūpa complex in Taxila, Taxila Museum, inv. no. C.4ID.46) and 174 (Peshawar Museum, inv. no. PM-3133). The taming of the serpent king can perhaps also be seen in the light of the successful cultivation of the Swāt valley. 45 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, p.╯122 and n. 14. Cf. Carter, 1992, p.╯70. 46 ╇ Aśokarājāvadāna, I. See Rhie, 1999–2002, p.╯116. 47 ╇ Rhie, 1999–2002, p.╯118–27. 48 ╇ Caspani, 1945, pp.╯49–52; Soper, 1949, p.╯279; Rhie, 1999–2002, pp.╯113, 136–7, and ns. 227, 242; image section 2.b and figs. 1.7a–c. 49 ╇ Rhie, 1999–2002, pp.╯117–27, and n. 242 (with reference to Soper (1949, pp.╯314–30, and 13, nos. 1–2, 1950, pp.╯63–75) who relates the legend to Zoroastrian and even Manichean beliefs). On the discussion of the inappropriate common translation of the Chinese character “ying” as “mirror image of the Buddha’s body” (a fully-fledged Buddha image), his “projection” in other words rather than the com-

the dragon in scenes of combat version of this story is also given by the seventhcentury Chinese pilgrim Xuanzang who visited and described the dragon cave which at that time was well-known for the belief that the Buddha had once been there on a supernatural visit and had left his “luminous reflected image.”50 It is noteworthy that whereas Indra kills the dragon, the Buddha chooses to subdue, tame and convert Apalāla and Gopāla, by opening their eyes to the destruction they had caused and by “graphically showing [them] the world as it truly is – an intolerable place of pain.”51 Rather than physical dominance or even the destruction of the dragon, he advocates introspective paths and moral transformation. Another Buddhist story recorded by Xuanzang concerns a nāgarāja who lives in a lake on a mountain two hundred miles northwest of Kapisa. After many years of stirring up natural forces to destroy the monasteries and stūpas founded by the Kushāṇa king Kanishka, the creature is eventually overcome by the king. The story relates how the thunderous voice of the dragon “shook the earth, and the fierce winds tore up the trees, whilst stones and sand pelted down like rain,” whereupon the king summons all the accumulated merit of his past lives and “from both his shoulders there arose a great flame and smoke.”52 This form of imagery shows a solar ruler who tames the monstrous dragon-nāgarāja by making a pact with him: as soon as some unfavourable weather phenomenon appears, the ghantā (cymbal or drum) will sound to remind the dragon of his promise whereupon he will cause the danger to subside.53

monly used “shadow,” see Rhie, 1999–2002, p.╯113 and n. 227. Cf. Wang, 2005, pp.╯245–6. 50 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, pp.╯93–7. Cf. Strong, 1992, pp.╯28–30; Rhie, 1999–2002, pp.╯132–3. 51 ╇ Strong, 1992, p.╯27. 52 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, pp.╯62–6, esp.╯p.╯65. Cf. Carter, 1974, p.╯186, n. 67. For a discussion of the Kushāṇa royal solar cult, see also Rosenfield, 1967, pp.╯189–91. 53 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, p.╯66. 54 ╇ Dēnkard 9.12.19, tr. West, W.E., Sacred Books of the East, Oxford, 1880–7, repr. Delhi 1965, p.╯40. 55 ╇Yasht 5.34; 17.34. Watkins, 2005, p.╯464. In later traditions the female figures became Jamshīd’s (Yima) sisters or daughters; Zaehner, 1955, repr. 1972, p.╯413. The theme is also attested in Greek mythology, for instance, the hero Perseus saves Andromeda from an aquatic dragon.

91

The eschatological role of the hero Middle Persian/Pahlawī texts describe an eschatological myth which shows the dragon as the being responsible for the lapse into chaos and death that is to take place at the end of time. As a consequence of this metamorphosis the latter has to be fought by valorous characters symbolising the status quo. The dragon-fighter Frēdōn (the Avestan Thraētaona) is first mentioned as fulfilling an eschatological role: he defeats but does not kill the giant dragon, binding and imprisoning him “in the most grievous punishment of confinement” at Mount Damāwand.54 The conquering hero thereby also frees the royal women, Sauuaŋhauuāci and Arənauuāci, “the two most beautiful women in the world,” the two wives of Azhi Dahāka/Ẓaḥḥāk.55 Hence in later myths female figures appear to have replaced water and rain as symbols of fertility and life.56 The theme of the rescue of a princess or maiden by the dragon vanquisher, which became a regular topos in most Iranian dragon-slaying stories, was introduced when the dragon became identified with a historical person, such as the foreign tyrant Ẓaḥḥāk, who imprisoned the maiden.57 The dragon remains bound at Mount Damāwand until the end of the world approaches when the other great dragonfighter, Kərəsāspa (who figures as Kirsāsp in Mid. Pers. and Garshāsp in N. Pers.) will be awakened (resurrected) by the divine beings Srōsh and Nēryōsang.58 It is for him that the final victorious battle against the dragon is reserved, the final war between Good and Evil Spirits.59 His eschatological work takes place and Kərəsāspa defeats the dragon when he emerges near the end of time from his captivity, breaks free of his fetters, wreaks havoc and causes much devastation in the world.60

╇ Cf. Gershevitch, 1959, p.╯45; Remmer, 2006, pp.╯212–

56

25.

╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. ╇Yasht 3.60–2. Christensen, 1931, tr. 1993, p.╯95. 59 ╇ Cf. idem, pp.╯55, 95. 60 ╇ Bundahishn 29.9 and in greater detail in Dēnkard 9.21.8–10. Cf. Hintze, 1999, p.╯82. Like Garshāsp, Sām plays a part in eschatological events: on Mount Sagāwand he lies on his back in the cold covered by snow and guarded by ten thousand frawashis (Pahl. frawahr) of the righteous until the dragon breaks loose and the final battle commences. Bundahishn 197.14–98.6; Monchi-Zadeh, 1975, pp.╯108–9, 142–3. In the Avesta Sāma is the name of a clan, to which Thrita as well as his sons Kərəsāspa and Urvākhshaya belonged (Yasna 9.10); Monchi-Zadeh, 1975, p.╯109. From Sasanian times onwards, however, Sām and Garshāsp began to be taken as the names of separate persons. 57 58

92

chapter six

Zoroastrianism which is a prophetic religion with a revealed scripture offers salvation to the souls of its adherents and humanity can be redeemed. The dualistic worldview inherent in its belief system61 dictates that “Evil must be annihilated by Good, if God is to save his creation.”62 Hence in the Younger Avestan hymns the myth of the dragon-fighter inherited from Indo-European times acquired an additional semantic component, that of the coming world saviour or Saoshiiaṇt (Pahl. Sōshyans). This Saoshiiaṇt is characterised by the epithet “victorious” (Av. vәrәthrajan-)63 who, in the last days, “is conceived of as not only utterly defeating Evil but also as ushering in a new age.”64 “He is the one who brings about the Renovation of the world (Av. frashō.kәrәti) in which Ahura Mazdā’s good and perfect Creation is restored and freed from all Evil.”65 Significantly, his victory over the monstrous dragon became the “pre-condition for the resurrection of the dead and the beginning of a new era.”66 The Zoroastrian eschatological myth recounts the coming of the final cosmic saviour, who is believed to come from the region of the Helmand river (Haētumaṇt, “with dams” in the Avesta) in Sakastān/Sīstān in southeastern Iran and southwest Afghanistan.67 He is the son of Vīspa. tauruuairī, who became pregnant by the Prophet Zarathushtra’s own seed while bathing in Lake Kasaoiia, his seed being miraculously preserved in the depth of the lake guarded by “guardian spirits.”68 She will bear a son called Astuuaṱ.әrәta, the last and greatest of the ancient Avestan valiant warriors, who will brandish the victorious weapon borne by other heroes before him, especially Thraētaona when he slew the dragon Dahāka.69 With this weapon (which belongs to the terminology of the Indo-European myth of the hero killing a dragon, Av. azhi- “snake, serpent, dragon,” jan- “to kill”) he will drive out Falsehood from the world of Truth (Zamyād Yasht 19.93).70 The corresponding Vedic adjective vṛtrahán- (“slayer 61 ╇ Cf. the definition of the term “dualism” by Hintze (1999, p.╯75 and n. 19) referring to two separate cosmic powers. 62 ╇ Eadem, p.╯76. 63 ╇ Hintze (1995, p.╯94) suggests that this epithet was added because of “the connection of the concept of the Saoshiiaṇt with the myth of the hero slaying a dragon.” 64 ╇ Hintze, 1999, p.╯76. 65 ╇ Eadem, p.╯76; cf. eadem, 1995, p.╯96. 66 ╇ Eadem, 1999, p.╯86. 67 ╇ Eadem, 1995, p.╯96.

of Vá¹łtra,” lit. “breaking the defence”)71 is used especially by the Vedic god, Indra, who slays the dragon whose name became Vá¹łtra in the Indian tradition.72 The ancient myth of the dragon-fighter was thus reinterpreted in a religious way in the popular image of the Saoshiiaṇt overcoming Evil.73 b.╇ Iconography of the medieval dragon-fighter The visual representation of a fighter, whether mounted or on foot, doing battle with a dragon employs a traditional and enduring iconographical formula of some antiquity and wide diffusion throughout Central Asia and the Near Eastern world, part of a stock of popular imagery that survived into medieval times. The fighter takes aim at the dragon using a variety of weapons, including bow and arrow, sword, spear or long lance, while the dragon is shown either as a lively upright creature imbued with fighting spirit, or in the guise of a vanquished dragon lying on its back beneath the horse’s feet with gaping upturned jaws. The representation of the single equestrian dragonfighter may be divided into two basic groups: the rider distinguished by royal or divine attributes, as graphically depicted in the investiture relief of Ardashīr I at Naqsh-i Rustam, discussed below, or the figure of the hunter or warrior, prevalent in the medieval Islamic period, without such insignia. In Sasanian royal imagery Ahura Mazdā/ Ohrmazd is shown on horseback crushing the head of the Zoroastrian evil principle of the universe, Angra Mainyu, known in later times as Ahriman who is likened to the serpent in the Great Bundahishn (“Book of Primal Creation”), Pahlawī translations based on lost Avestan scriptures of the third century ad and before and their commentaries written after the Arab conquest.74 It describes him as having sprung:

68 ╇ Boyce, 1975, repr. 1996, p.╯282; Hintze, 1999, p.╯77. On Vīspa.tauruuairī, see Remmer, 2006, pp.╯57–8, 144–8, 200–5, 253. 69 ╇ Boyce, 1975, repr. 1996, p.╯283; Hintze, 1995, p.╯93 and 1999, pp.╯77–8. 70 ╇ Hintze, 1995, p.╯93 and eadem, 1999, p.╯77. 71 ╇ Eadem, 1995, p.╯94. 72 ╇ Eadem, 1999, p.╯77. 73 ╇ Eadem, p.╯78. 74 ╇ See Watkins, 1995, p. 58. On the Iranian or Great Â�Bundahishn, see also Klíma, 1968, pp.╯41–3.

the dragon in scenes of combat …like a snake, out of the sky down to the earth, … thereby the sky was as shattered and frightened by him, as a sheep by a wolf.75

The association of the serpent with Ahriman is perhaps best portrayed in the monumental third-century investiture relief sculptures of the Sasanian king Ardashīr I (r. 224–241) at Naqsh-i Rustam. It shows a bilateral equestrian scene. The conquering king Ardashīr appears on the left, facing an anthropomorphic Ahura Mazdā with turreted crown on the right, shown in the act of bestowing on Ardashīr the gift of khvarәnah, in the form of a ring to be hung over and secured around the royal crown. Ardashīr’s steed is shown trampling on the head of the last Parthian leader Ardavān (Artabanus) V, while Ahura Mazdā’s horse treads underfoot a creature held in the coils of serpents. The plastically sculpted reliefs draw on the age-old universally understood motif of the ruler placing his foot on a prostrate enemy as a symbolic gesture of physical as well as ideological supremacy, the horse serving in this instance as a visual extension of the ruler, his rider.76 This visual realisation of victory is of particular significance: the vanquished enemy, his head wreathed with serpents, one of which uprears its head at the front, presumably represents the anthropomorphic Ahriman, and thus the first dated example of a symbolic synthesis of the serpent and Ahriman (figs. 86a and b).77 The defeat of the Parthian leader is thus equated with triumphing over the powers of evil. Inscriptions in three languages, Middle Persian, Parthian and Greek, on the horses’ flanks identify the god and the conqueror.78 The representation of the paired mounted horsemen, imperial and divine, surmounting prostrate enemies, human and satanic, thus underscores the triumph by metaphorically alluding to a complex of eschatological beliefs. 75 ╇ Bundahishn 6.10–11 (Sacred Books of the East, tr. West, E.W., vol.╯5, Oxford, 1897). Cf. Zaehner, 1961, p.╯262; Boyce, 1984, p.╯50. 76 ╇ The motif appears in the Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯4, p.╯81, l. 237. See also the related ancient Oriental motif of “eating/licking dust” as simile for death and the humiliation of the vanquished enemy, which is exemplified in Genesis 3.14. Cf. p.╯23; also Martinek, 1996, p.╯65 with further examples. 77 ╇ Ghirshman, 1962, p.╯132, fig.╯168. Cf. Boyce, 1979, repr. 2001, p.╯107. 78 ╇ Boyce, 1979, repr. 2001, p.╯107. 79 ╇ For additional examples of Sasanian-period glyptics with this motif, cf. Gyselen, 2007, pls. 14.2 and 14.3; Gignoux and Gyselen, 1982, p.╯58, nos. 14.1–14.3; Gignoux, 1978,

93

A mounted dragon-fighter distinguished by a flaming halo appears on seal stones from the Iranian world79 such as a sixth-century brownred jasper, preserved in the State Hermitage, St. Petersburg (fig.╯87). The rayed halo that surrounds the rider’s head and shoulders probably symbolises the khurrak-i Kayan (the khvarәnah of the Avestan texts and Firdawsī’s farr-i kayani), the Royal Splendour of the Kayanids (the protagonists of a millenary struggle against Tūrān), the radiance that descends upon the heroic warrior and, above all, the ruler and renders him sacred.80 The rider thus probably represents one of the ancient Indo-Iranian epic heroes that fulfil a mythical quest. He is seen holding a spear and battling with what appears to be a seven-headed dragon whose body coils along the edge of the seal from below the horse’s hooves, rising upwards with one dragon head surmounting the other. He is flanked by the figure of a small scorpion and a star. The depiction visualises here the popular imagery of the hero as a “beneficial force” attacking a monstrous dragon representing a “malevolent force.” In the historical story in the Pahlawī text Kārnāmak-i Ardakhshīr-i Pāpakān (“Book of the Deeds of Ardashīr, Son of Pāpak”), written around 600, the founder of the Sasanian dynasty Ardashīr Pāpakān (224–241) himself is related to the valiant ritual of killing a dragon called Haftān-bōkht (the Haftwād of Firdawsī’s Shāhnāma) who lived in the village of Alār in the rustāq of Kojārān.81 Prominent depictions of the equestrian dragon-fighter are encountered on wall paintings discovered by Russian archaeologists in the city of Panjikent in Sogdiana (Zarafashān, north of the Hiṣār range), now at the State Hermitage, St. Petersburg. The depictions show pictorial epics from a variety of literatures, among them the Greek fables of Aesop and the Indian epic p.╯54, no. 6.6; Bivar, 1969, nos. BL 3, BL 4; Ghirshman, 1962, p.╯243, fig.╯300 (collection of H. Seyrig). 80 ╇ The iconography of farr(ah), khvarәnah has been associated with figures connected with light and fire, in other words, by having flames emanating from the human body or partially surrounding it. Cf. Christensen, 1944, p.╯146. For the “flaming shoulders” of divine beings or kings on Kushāṇa coins, see Rosenfield, 1967, pp.╯17, 23–4, 29, 157, 197–201. Cf. Gnoli, “Farr(ah), Xvarәnah,” EIr; Carter, 1974, pp.╯176–7 and ns. 18–20. On Firdawsī’s concern with the distinctive hereditary mark of Iranian splendour, farr-i kayani, see Rypka, 1968, pp.╯155, 159. See most recently, Soudavar, 2003. 81 ╇ Nöldeke (1879, ch. 6) associated the story of Haftwād with the ancient myth of Apollo and Hydra; Minorsky, “Lār,” EI1 III, pp.╯15–7.

94

chapter six

Mahābhārata, as well as episodes from the heroic cycle, the dragon-fighter being identified by Aleksandr Belenitskii as the hero-champion Rustam.82 The hero’s ancestors were Saka people who are part of the Scytho-Siberian cultural grouping and belong to the Indo-Iranian group that came to Sakastān/Sīstān and Zābulistān in the late second century bc,83 lands far from Sogdia. Saka heroic tales were nevertheless very popular with the Sogdians,84 although only a fragmentary Sogdian text survives85 and only in the tenth century was the tale taken up by Firdawsī in his magnum opus. The heroic cycle of Rustam’s Herculean seven feats (haft khwān) attains almost spiritual importance.86 Before reaching his ultimate goal the hero has to undergo these trials,87 which represent a kind of rite of passage.88 During his third feat, which is reminiscent of Herakles defeating the Hydra of Lerna,89 Rustam slays a magical dragon that is guarding a watering place which the ram has shown90 and comes out of the forest at night and approaches the sleeping hero.91 Twice he is woken by his formidable mount Rakhsh,92 but each time the dragon vanishes. On the third occasion the monster fails to conceal itself in time and with the help of the faithful Rakhsh the hero succeeds in killing the dragon. Interestingly, the Shāh-nāma portrays the dragon with human traits such as the power of reflection93 and speech: during the battle he declares himself master of the whole desert.94

Correspondingly, Rustam is shown confronting the dragon as if it were a human adversary.95 Igor P’yankov notes the archaic style of Firdawsī’s records of the Rustam cycle and has demonstrated its close parallels with ancient Greek records, particularly Herodotus’ accounts of one of the genealogical myths of the Scythians (Histories IV 8–9).96 According to the father of Greek historiography the hero, known by the Greeks as Herakles, comes to an arid region at the Pontus Shore (Black Sea) later inhabited by the Scythians where he loses his horses and in the search for them meets in a cave in the forest a mythical creature described as a woman with the lower body of a serpent.97 With this anguipede woman he engenders three sons, the youngest and worthiest of whom, named Scythes, becomes the first king of the Scythians. In Firdawsī’s account Rustam’s first feat is his victory over a lion whose pelt he wears just like the Grecian Herakles after the latter’s vanquishing of the Nemean lion.98 The second exploit is the discovery of a spring in the desert country.99 The third is the victory over the dragon. While there is no love theme in the third trial, the fourth episode recorded in the Shāh-nāma mentions a sorceress in the form of a beautiful girl who tries to seduce Rustam near a small river in the shade of some trees.100 Parts of the story are shown on a continuous frieze of the Sogdian wall paintings in Panjik-

82 ╇ Belenitskii, 1980, pp.╯103–5, 199. Cf. Azarpay, 1981, p.╯195. Guitty Azarpay (1981, pp.╯96–7) also points out that the “dramatis personae” were subject to change and not necessarily connected to any specific hero. 83 ╇ Boyce, 1975, repr. 1996, p.╯101 and n. 104; de Bruijn, “Rustam,” EI2 VIII, 636b; P’yankov, 2006, p.╯505. 84 ╇ P’yankov, 2006, pp.╯505–6. 85 ╇ The Sogdian fragment from Dunhuang which records part of the Rustam legend is fully quoted in Klíma, 1968, p.╯53; Azarpay, 1981, pp.╯6–7. Cf. Marshak, 2002, p.╯51. 86 ╇ Cf. Russell, 2004, p.╯543 and n. 30. For a discussion of the close parallels of the haft khwān (dragon slaying being one of the exploits) of Rustam and Isfandiyār, see Yarshater, 1983a, pp.╯469–70 and n. 5. 87 ╇ The number seven was specifically important to the “Avestan people,” and plays a significant role in the rites and customs of the Zoroastrians, for whom seven is the number of the creations and of the Amahraspands (Aməsha Spəntas), the positive creatures or “Bounteous Immortals,” who guard them. The number seven gained even greater prominence in the Islamic period, when it acquired additional symbolism. Cf. Hartmann-Schmitz, 1989, pp.╯12–20; Schimmel, 1994, p.╯27. Moreover, seven often conveys ideas of perfection and periodicity (for a list of examples, see Shahbazi, “Haft (seven),” EIr). It is a favourite number in eastern Semitic civilisations with magico-religious features; among the Israelites it was used in ritual incantations (2 Kings 13 and Joshua 6); and in the Old Testament seven is the number of completeness. Cf. Jeffers, 1996, p.╯87, n. 286.

╇ Cf. Omidsalar, 2001, pp.╯262, 265–6. ╇ For a discussion of the points of resemblance between Hercules and Rustam, see Melikian-Chirvani, 1998, p.╯178. See also p. 79, n. 44. 90 ╇ Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯515, ll. 336–41. 91 ╇ A view of the entire scene is reproduced in Azarpay, 1981, p.╯96, fig.╯42. Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯517–21. 92 ╇ On the motif of the horse as helper of the hero, see Schirmunski, 1961, pp.╯25–6. 93 ╇ Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯517, ll. 565–70. 94 ╇ Idem, vol.╯1, pp.╯519–21, ll. 396–400. This exploit is also recorded by the eleventh-century Armenian scholar Grigor Magistros who moreover notes that the battle took place near Mount Damāwand; see Tchukasizian, 1964, pp.╯321–2. Cf. P’yankov, 2006, p.╯507. 95 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯519–21, ll. 393–5, 400–2. 96 ╇ P’yankov, 2006, pp.╯505–11, esp.╯pp.╯506–7. 97 ╇ Cf. Sulimirski, 1985, p.╯168. 98 ╇ It is notable that the Nemean lion was one of the offspring of the drakōn Typhon and Echidna, who had the face and torso of a woman and the body of a serpent (Hesiod, Theogony 306–8). See West, 1962, p.╯161. 99 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯513–7. 100 ╇ Idem, vol.╯1, pp.╯521–3. 88 89

the dragon in scenes of combat

95

ent. The monster is depicted as a terrifying shedragon who has coiled her elongated serpentine tail around all four legs of the horse (probably to be understood as Rakhsh),101 her female upper body rises up, naked to the waist. With her long arms she is dragging the mounted Rustam-like hero’s head towards her. The hero has succeeded in wounding the dragon twice with his axe (fig.╯88).102 In the next scene the dragon, its wounds gushing blood, is in its death-throes. In contrast to the preceding images, it now lies prostrate on the ground (fig.╯89).103 Boris Marshak explains the depiction of the serpentine she-dragon with human arms and lion’s mane as a conflation of three trials mentioned in the Shāh-nāma, namely the fight with the dragon, the lion and the sorceress.104 While the genealogical aspect of the myth was apparently forgotten, some analogies with the original theme of the mythical anguipede progenitrix appear to have been retained.105 An equestrian dragon-fighter is also portrayed on a tympanum from the medieval city of Bunjikat (20 kilometres south of the modern town of Shahristan in northern Tajikistan) in the Sogdian principality of Ustrushana, probably dating from the seventh to the ninth century. A row of pearl roundels frame the rim of the monumental wooden arch-shaped tympanum (partly destroyed by fire) that adorned the top of the portal leading to the throne hall of the Qalʿa-i Qahqaha, now

preserved in the National Museum in Dushanbe in Tajikistan.106 In one of the roundels a mounted horseman is seen taking aim at a twice-knotted dragon.107 While there appear to be no surviving representations of the dragon-slayer around the turn of the millennium in the Islamic realm of Western Asia, the equestrian dragon-fighter can nonetheless be seen as a leitmotif which links the preIslamic Sasanian and the Sogdian times with the Islamic period. After an apparent lacuna in the tenth and perhaps the early part of the eleventh century (a period during which the motif occurs in the Christian art of the Caucasus, particularly in Armenia and Georgia, as examined below), it is depicted with great regularity on Islamic works of art from Central Asia to Anatolia and the Jazīra. The Abū Muslim-nāma,108 which records the life of the charismatic Abū Muslim Khurāsānī (d.╯c. 137/754–5) who led a popular movement for the ʿAbbasid cause and became a legendary figure after his assassination, recounts the heroic exploits of the fourth caliph ʿAlī ibn Abī Ṭālib (d.╯40/660), Muḥammad’s cousin who married the Prophet’s daughter Fāṭima. ʿAlī is portrayed as accomplishing the feat of vanquishing the dragon at a very early age and is eulogised as infant dragon-slayer with the words:

101 ╇ Comparable imagery is represented on a Sasanian seal in the British Museum, London, which shows a double-headed serpent coiled around each of the four legs of a bovine and rearing up above its head and inscribed with the name of the owner in Pahlawī. Mordtmann, 1864, pl.╯I, no. 4, republished in Ettinghausen, 1955, p.╯282, pl.╯XXXIX, no. 8. Cf. Bivar, 1969, p.╯8, pl.╯15.1; Marshak, 2002, p.╯43. 102 ╇ Azarpay, 1981, pl.╯6; Marshak, 2002, p.╯40, figs. 17–20, colour pl.╯3; Grube and Johns, 2005, p.╯233, cat. no. 78.5. 103 ╇ Azarpay, 1981, pl.╯7; Marshak, 2002, p.╯43, fig.╯20. 104 ╇ Idem, 2002, p.╯51. An anachronical but perhaps not entirely irrelevant parallel exists in Kushāṇa-period chthonic creatures, sometimes referred to as vyāla, which were represented as half-females, half-serpents, with a female upper body whose lower limbs transform into a long spiralling serpentine tail terminating in a fan-shaped caudal fin. Cf. Czuma, 1985, p.╯53, cat. no. 3. 105 ╇ Igor P’yankov (2006, pp.╯508–10) notes that traces of the ancient genealogical tales are still preserved in today’s folklore of southern Tajikistan where the Scythian people once lived and where he heard oral traditions from local people about a dangerous serpentine woman who lives in the river. 106 ╇ Unfortunately, the panel remains unpublished; it was not permitted to photograph it in the museum nor was it possible to obtain a photograph from the museum. For a description of the site, see Negmatov, 1996, repr. 1999, pp.╯259–74, and fig.╯41.

107 ╇ The first fire is associated with the time of the ʿAbbasid conquest of the town in 206/822, the second with the annexation to the Samanid state by Ismāʿīl ibn Aḥmad I (279/892– 295/907) providing a terminus ante quem for the dating of the wooden panel. 108 ╇ The Abū Muslim-nāma was written by Abū Ṭāhir Ṭarsūsī (Ṭūsī), who was part of the retinue of the Turkic Â�Ghaznawid sulṭān Maḥmūd (r. 389/999–421/1030). The legends of Abū Muslim are surveyed in Mélikoff, 1962. See also eadem, 1960, vol.╯1, p.╯43. 109 ╇ The translation is based on the manuscript in Paris, Bibliothèque Nationale, Ms. Pers. 843, dated 1145–6/1732–4; Mahdjoub, 1988, p.╯63. The motif of the infant dragon-slayer is repeatedly found in classical literature, for instance, in the depiction of the infant hero Herakles struggling with two serpents described by the classical Greek poet Pindar (c. 518–438 bc) in the Nemean Ode 1, 42–7. Another early classical example is given by the infant Apollo who when only a few days old shot arrows from the arms of his mother Leto at a multi-headed snake, the story being depicted on a fifth-century bc lekythos (predating Euripides’ Iphigeneia in Tauris, 1239–1251); see Fontenrose, 1959, repr. 1980, pp.╯16–7 and fig.╯1. For a discussion of the epic motif of the supernatural power and acquisition of “wisdom” and certain magical abilities of infant heros from dragons, see Schirmunski, 1961, pp.╯58–9.

Bare-handed in the cradle with his mighty arms he tore apart the dragon’s jaws.109

chapter six

96

It is interesting to compare this imagery with that of the infant Rustam whose arms are associated with courageous dragons in the Shāh-nāma account.110 The theme of the dragon-slayer appears in a mid-twelfth-century book on cosmography written in Persian and dedicated to the last Great Saljuq sulṭān of Iran and Iraq, Ṭoghrıl III ibn Arslan (r. 571/1176–590/1194). The story in Muḥammad ibn Maḥmūd ibn Aḥmad-i Ṭūsī’s “Book of Marvels,” entitled ʿAjāʾib al-makhlūqāt (“Wonders of Creation”),111 illustrates the power of talismanic images hidden in the pre-conquest Byzantine capital Constantinople (al-Qusṭantīniyya).112 Three bronze statues representing Muḥammad and two of his closest followers, ʿAlī and Bilāl (the first muezzin), were zealously guarded by the local people; they knew from past experience that damage to the statues would set off a devastating earthquake.113 Significantly, of the three statues only the one which shows ʿAlī on horseback striking a dragon with a spear is illustrated in the manuscript.114 While on the one hand this story is buttressed by apocalyptic traditions, such as the prophetic ḥadīth foretelling the city’s capture by an Islamic ruler who bore the name of a prophet,115 it is also noteworthy that it was the Byzantine capital which was associated with the figure of the equestrian dragon-slayer. However, the story also shows that the dragon-fighter represented a well-established iconographical theme in medieval Islam. It was indeed so firmly entrenched as to be deemed the most appropriate imagery for the representation of the Companion of the Prophet who for the Sunnī Muslims represents the fourth caliph and for the Shīʿites the divinely appointed successor (khalīfa) and heir (waṣī) of Muḥammad. By the twelfth century, dragon-fighter iconography was a prevalent part of a set narrative genre often inspired by textual sources such as the early eleventh-century Iranian national epic, the Shāh-nāma, and was a motif of choice depicted on many portable objects produced in medieval Western Central Asia. The theme

was often chosen as part of a visual narrative on metalwork, for instance on the two copper alloy buckets discussed above, the 559/1163 Bobrinski bucket probably from Herat, and the late twelfthto early thirteenth-century richly gilded bucket made by Muḥammad ibn Nāṣir ibn Muḥammad al-Harawī, also perhaps from Herat. The third figural relief that circumscribes the body of the Bobrinski bucket also comprises a procession of riders. Behind one of the mounted warriors, a “Saljuq-style” twice looped dragon with enormous gaping mouth rears up threateningly (fig.╯90), its tongue with bifid tip oriented towards the back of the horseman, the scaly body echoing the body of the dragon protome on the handle (fig.╯56). A second rider charges from behind to assist the beleaguered warrior, wielding what appear to be a shield and a club. The body of the Fould bucket is divided into twelve barely perceptible vertical facets, alternately enclosing a cartouche forming an angular figure of eight framed by a benedictory epigraphic band in Kufic and enclosing riders mounted on camels, horses and mules or donkeys. One of the horsemen, a curved sabre raised above his head, turns backwards to defend himself against an upright dragon of a type closely related to the one featured on the Bobrinski bucket, who threatens him with open jaws from behind (fig.╯91).116 The prevalence of the motif on Western Central Asian metalwork is further suggested by a related depiction in which the dragon is shown with a curved horn on an early twelfthcentury Herati-type copper alloy ewer inlaid with silver from a private collection (fig.╯92). A variant of the motif is found on the base of a well-known copper alloy penbox (qalamdān) inlaid in silver, gold and niello from Iran, which bears the name Maḥmūd ibn Sunqur and the date 680/1281–2 inscribed on the hasp.╯Here not the single equestrian fighter but a second genus, that of the paired horsemen, is reproduced. Two scenes of confronted fighting horsemen, separated by three large roundels filled with a geometric pattern, are shown: on the scene to the left one horseman attacks an upright double-headed dragon

╇ Tr. and ed. Mohl, 1838–1878, vol.╯5, p.╯353, ll. 1707–9. ╇ Ed. Sotūde, M., Tehran, 1345/1966, as cited in Pancaroğlu (2003, p.╯31 and n. 4) who dates the book to the years between 562/1167 and 573/1194. Cf. Radtke, 1987, pp.╯278–88. 112 ╇ First section of chapter six which includes descriptions on talismanic portraits, statues and tombs of prophets and kings (ed. Sotūde, M., Tehran, 1345/1966, pp.╯333–4);

Pancaroğlu, 2003, pp.╯33, 37 and appendix. 113 ╇ Eadem, pp.╯34, 37. 114 ╇ First section of chapter six (Aḥmad-i Ṭūsī, ʿAjāʾib al-makhlūqāt, ed. Sotūde, M., Tehran, 1345/1966, pp.╯333–4); eadem, pp.╯34, 37 and appendix. 115 ╇ See, for instance, Eisener, 1987, pp.╯129–37. Cf. Pancaroğlu, 2004, p.╯155. 116 ╇ Cf. Mayer, 1959, pl.╯X.

110 111

the dragon in scenes of combat

97

with a sword while the other turns backwards to shoot a lion with bow and arrow (fig.╯93).117 The vertically oriented body of the dragon is characterised by bilateral symmetry, the addorsed heads with open jaws revealing projecting tongues and the scaly body bifurcating to form two loops before uniting and thinning to a short pointed tip.╯On the right, two horsemen charge at each other with long lances. The depictions are set against a background filled with a dense interlace of foliate scrolls bearing long-eared animal heads that have been identified as “dragon progeny.”118 Another example with the motif survives in the form of a twelfth- or thirteenth-century richly gilt copper alloy polylobed openwork roundel, probably a fitting from a belt or horse-harness for attachment, now in the al-Sabāh Collection, Kuwait National Museum. The roundel depicts an elaborately dressed archer on the back of a prancing horse which has a knotted tail and carefully rendered trappings, studded with circular phaleÂ� rae and suspended crescents. A feline, probably a cheetah, crouches behind the warrior, who draws his bow and takes aim at a dragon as it writhes below the feet of his mount (fig.╯94). The serpentine body is enlivened by a dense dotted pattern and arranged in two loops. Of the same period and from the same wider geographical region stem related depictions on polychrome painted ceramic tiles or ves-

sels (so-called mīnāʾī ware) featuring a richly clad dragon-fighting rider, his mount decorated with magnificent trappings. On a star tile, now in the Freer Gallery of Art, Washington, DC, the sinuous speckled dragon is shown in a downward diagonal (fig.╯95),119 whereas on a bowl from the Vollmoeller Collection in Zurich it is rendered in the more conventional supine posture, jaws agape, tongue projecting and the serpentine body arranged in two loops (fig.╯96).120 The ophidian bodies are both demarcated with a dense pattern. On the tile the horseman seems to take aim with a bow, while on the bowl the cavalier probably brandishes a sword. Another example of the depiction of a horseman fighting an upright dragon is part of the main narrative frieze circumscribing the shoulders of a contemporary moulded flask in the Aga Khan Collection.121 The belief in the invulnerability of a dragon’s hide, which is impervious to water, fire, or any weapon, explains why the hero usually aims at the head, eyes or mouth of the dragon or has to attack him from within.122 This is particularly evident in one of the celebrated feats of Rustam in the legend of the battle with the Babr-i bayān.123 As a young man Rustam was fighting the beast in a distant land and managed to kill the dragon by making it swallow oxhides filled with quicklime and stones124 which he carried to the place where once a week the dragon came out of the

117 ╇ Cf. Furūsiyya, 1996, vol.╯1, p.╯173, fig.╯IVa detail to left, and IV view of entire lid, vol.╯2, p.╯232, cat. no. 194. 118 ╇ Hartner, 1959, pp.╯237–9, and idem, 1973–4, pp.╯112, 118. 119 ╇ In its unrestored condition the tile has been published in Atil, 1973, cat. no. 49, and Grube and Johns, 2005, p.╯233, cat. no. 78.7. 120 ╇ Another example of a mīnāʾī bowl with the same iconography featuring the dragon head rising above the horse’s head is preserved in the Calouste Gulbenkian Foundation, Lisbon. Curatola, 1989, fig.╯63. 121 ╇ Splendori a corte, 2007, p.╯149, cat. no. 117 (the dragonfighting theme is not featured on the side of the flask that has been reproduced in the catalogue). 122 ╇ In the Avesta (Yasht 9.10–1) a similar story of the invulnerability of the dragon’s hide is recorded, according to which Kərəsāspa/Garshāsp cooked his midday meal on the vast green flank of the sleeping Azhi Sruuara (Azhi Zairita):

As cited by Klíma, 1968, p.╯12, after the German translation of Wolff, F., Avesta: Die Heiligen Bücher der Parsen, Straßburg, first ed. 1910, repr. 1924; see also Boyce, 1975, repr. 1996, pp.╯102–3. The long time lapse before the heat began to permeate the dragon’s hide to finally wake him up (Skjærvø, “Aždahā I,” EIr) suggests, according to Khāleqī-Moṭlaq (“Aždahā II,” EIr), “that the belief in the invulnerability of the dragon-hide was a very old component of the story.” Having finally defeated the dragon, the hero, like Rustam, also made a coat out of its hide. It is noteworthy that similarly many early epic heroes were distinguished with a magical invulnerability apart from one vulnerable spot, for instance Isfandiyār in the Shāh-nāma who is invulnerable apart from his eyes, and so a double-pointed arrow has to be discharged into them in order to wound or kill him. Cf. Schirmunski, 1961, pp.╯36–7. 123 ╇ Two versions of the legend are found in a Shāh-nāma manuscript in the British Museum, London (Ms. Or. 2926, fols. 112b-115a and 118b-122b); it is also current in surviving Iranian oral folklore (Enjavī, A., Mardom o Shāh-nāma, Tehran, 1355/1976, pp.╯217–8, cited after Khāleqī-Moṭlaq, “Aždahā II,” EIr); a variant of the story is known among the Mandeans of Iraq (Petermann, 1860–1, vol.╯2, pp.╯107–9). 124 ╇ In the Shāh-nāma this method of killing the dragon is similarly employed by Iskandar (tr. and ed. Mohl, 1838– 1878, vol.╯4, pp.╯203–5, ll. 1230–1249). Analogies may be found in other mythological traditions such as the story of Bel and the Dragon in the book of Daniel LXX in which there was a giant serpent (drakōn) that was venerated by the Baby-

He was a young man famous for his strength, had curly hair and swung his club; he smote the horny dragon, the horse-swallowing and man-swallowing, full of poison, yellow of colour, over whom yellow poison flowed as high as a spear. On his back Kərəsāspa cooked his meal in an iron cauldron at the time of noon. And the monster felt warm and began to sweat. Then he rushed from under the iron cauldron and upset the boiling water. Affrighted rushed headlong the valiant Kərəsāspa.

98

chapter six

sea. When the dragon swallowed these, his stomach burst. Significantly, in stories where the dragon was presented as a historical person, the invulnerability of the hide was metaphorically transformed into the impregnability of the enemy’s castle.125 The depiction of a fighter on foot in direct combat with a dragon whom he attacks with his sword – a motif which appears to be absent from Christian iconography – is found on the outer walls of a celebrated plate whose interior is decorated with a large battle scene featuring the siege of a citadel and inscribed with the names of the warriors. These names incorporate Turkish elements, suggesting that they may have been Saljuq fighters.126 Datable to the early thirteenth century, the plate is housed in the Freer Gallery of Art, Washington, DC (fig.╯97).127 The outer walls show a richly clad warrior carrying bow and arrow, shown in three-quarter view and facing a writhing ophidian dragon in rampant posture. The dragon combat is one of five heroic feats portrayed on the plate, the vignettes being separated by trees. Other feats include the shooting of a wolf with bow and arrow, clubbing a feline, probably a panther, with a mace, and shooting a mythical creature with bow and arrow. A further element shows two confronted figures, one of whom is holding a feline, probably a cheetah, on a leash. An upper register contains an epigraphic band with good wishes in Kufic script. Following the Saljuq victory at the battle of Manzikert in 463/1071 and subsequent largescale Turkish penetration into Anatolia, the Saljuq sulṭāns of Rūm, being closely affiliated with Iranian cultural and artistic traditions, also perpetuated these semantic horizons. The image

of a mounted figure fighting a dragon was applied widely to objects and architectural decoration in Anatolia and the neighbouring Jazīra. The motif occurs on an important frieze fragment from the now destroyed pavilion (once part of the palace) of Qılıch Arslan II (r. 551/1156– 588/1192), one of the greatest Rūm Saljuq sulṭāns, noted both for his military achievements and his patronage of the arts. It shows two haloed horsemen charging each other and attacking respectively a dragon and a lion (fig.╯98). The rider on the left thrusts his long spear into the gaping jaws of the dragon. The beast is shown with curling goatee beard projecting below the chin, and a pronounced bristling crest running down the spine to the tapering tail. The second horseman turns around to grasp the lion’s mane with one hand while dealing him a blow with the sword held in the other. The Rūm Saljuq depiction thus provides a parallel to the version emblematised on the Iranian penbox made by Maḥmūd ibn Sunqur (fig. 93) almost a century after the making of the frieze. This composition, moreover, again features the fight of confronted horsemen against the most dangerous and deadliest of adversaries, the dragon and the lion. The fragment must have been part of a large frieze and is a valuable example of the type of decoration that presumably once adorned not only Saljuq-period pavilions but related secular buildings throughout the entire region as far as Central Asia.128 One of the earliest depictions of the motif of the single equestrian dragon-fighter west of Iran is found on the coinage of Turko-Islamic Anatolia. An equestrian warrior spearing a prostrate scaly dragon appears on the reverse of a copper coin of a type minted by the last Türkmen Dānishmendid

lonians; the Jewish Prophet Daniel killed it without sword or staff by brewing a concoction of pitch, fat, and hair and then feeding cakes made of it to the dragon (Gunkel, 1895, pp.╯320–3, cogently argues that this story is an adaptation of a passage of the Babylonian creation epic Enūma Elish, an Akkadian text; cf. idem, 1895, pp.╯412–3, tablet IV, ll. 93–104; Pritchard, ed., 1955, repr. 1968, p.╯67). The execution of enemies seems sometimes to have been inspired by the manner in which the dragon was killed in these traditions (Merkelbach, “Drache,” RAC IV, 1959, pp.╯234–5). Mithridates VI (120–163 bc), king of Pontus, for instance, gave orders to execute Manius Aquilius by pouring liquid gold down his throat (Appianos, Mithridates, XII 21; Pliny, Naturalis Historia, XXXIII 48; cf. idem, pp.╯234–5). This method of killing the dragon is described by al-Qazwīnī in his Kitāb ʿajāʾib al-makhlūqāt (ed. Wüstenfeld, 1849, repr. 1967, p.╯112). For the illustration of this story in the so-called Sarre Qazwīnī, now in the Freer Gallery of Art, Washington, DC, see Badiee, 1978, pl.╯32. The theme is also visualised in a fourteenth-century

miniature showing Ardashīr pouring molten lead down the dragon’s throat, illustrated in the St. Petersburg Shāh-nāma (Dorn 329, f. 243r), copied in Shiraz, 30 Jumādā I 733/16 February 1333. These references not only demonstrate that this method of killing was known but perhaps also testify to its ongoing popularity in the medieval period. See also p. 58, n. 96. 125 ╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. 126 ╇ Atil, 1973, cat. no. 50. 127 ╇ In 1983 the conservation department of the Freer Gallery of Art discovered that the original plate had some overpainting and subsequently restored it back to its original state. As a consequence part of the upper body and head of the warrior and a section of the snake’s protome are no longer visible after the restoration. A photograph of the plate that pre-dates the conservation was used in order to make the imagery more easily recognisable.€ 128 ╇ Cf. The Anatolian Civilisations, vol.╯3, 1983, pp.╯34–5, cat. no. D.38; Turks, 2005, p.╯392, cat. no. 58.

the dragon in scenes of combat

99

ruler of Malatya/Malaṭiyya (Melitene), Nāṣir al-Dīn Muḥammad ibn Ismāʿīl (r. 557/1162– 565/1170 and 570/1175–573/1178), whose name appears on the obverse, minted in the 1170s.129 Mounted on a galloping horse the rider’s right foot is placed on the body of a dragon and with his right hand he grasps the end of a lance which is thrust in the small uncoiled dragon’s open jaws; the left hand holds the reins. Importantly the armoured rider, clad in a short skirt and a long-sleeved coat, is here portrayed nearly in profile. The coinage was produced after the Dānishmendids’ capture of Malatya, situated not far from the upper Euphrates on an important stretch of the Arab-Byzantine frontier, in 494/1101 following a three-year siege. For much of its existence the dynasty of the Dānishmendids (463/1071–573/1178), one of the earliest Türkmen principalities established in Anatolia that reigned in northern Cappadocia, maintained a frontier ethos in which the dragon-fighting ghāzī was a pre-eminent symbol, as will be shown in the Epilogue. Qılıch Arslan II’s conquest of Malatya in 573/1178 brought about the end of the Dānishmendid dynasty. Not long after the Saljuq conquest of Malatya, the city of the frontier hero par excellence,130 a new copper coin with a horseman slaying a dragon was minted by the ruler of Malatya, Muʿizz al-Dīn Qayṣar Shāh (r. 582/1186– 597/1201 with an interruption in 587/1191), a son of Qılıch Arslan II (r. 551/1156–588/1192).131 As Nicholas Lowick tentatively suggests, the Turkish ghāzī rulers were perhaps “consciously or not, seeking to establish contact with an imaginary heroic world of the past, possibly under the stimulus of historical or quasi-historical works of literature.”132 The dragon-slayer iconography on Dānishmendid and Saljuq coins is thought to have been inspired by the Byzantine copper coins

struck by the crusader prince Roger (d.╯513/1119), who usurped the throne of Antioch in the guise of regent for Bohemond II between 1112 and 1119,133 and can ultimately be traced back to early Byzantine coinage.134 On Roger’s coin the equestrian rider is similarly portrayed nearly in profile and leaning forward on a galloping horse to stab the mouth of a serpent with his lance. The inscription identifies the figure as Saint George. In spite of the differences in the iconographic representation of this mounted dragon-fighting warrior, in particular the depiction of the figure in profile, there is an overall consistency with the image of contemporaneous dragon-fighters throughout medieval Western Asia. The cosmopolitan milieu and the generous patronage of Badr al-Dīn Luʾluʾ (618/1222–657/ 1259) in Mosul permitted the city to become a haven for master craftsmen from Central Asia who were fleeing the Mongol invasion.135 Among them were specialist inlay workers from the greater Khurasan region (where the technique was developed).136 Their skill is prominently displayed on a copper alloy candlestick base, probably made in the Jazīra in c. 1230, and inscribed with the names of the masters, Ḥājji Ismāʿīl and Muḥammad ibn Futtūḥ al-Mawṣilī.137 The dragon-fighting theme is shown in a large polylobed cartouche depicting a rampant knotted creature that rises above the horse’s rump, his large open maw with projecting tongue oriented towards the rider who turns towards it with raised sword in hand (fig.╯99).138 The same iconography of a horseman fighting a dragon reappears in a mid-thirteenth-century glazed fritware sculpture, now preserved in the National Museum of Damascus. Possibly a fountain element, it was discovered in Raqqa, a major city in the western part of the Jazīra.139 The warrior wears his hair in long braids under a small cap

╇ Whelan, 1980, pp.╯143–8, pl.╯16, 5b. ╇ Wittek, 1936, p.╯295. 131 ╇ Istanbul, Archaeological Museum, inv. no. 1068. Süslü, 1987, p.╯640, pl.╯118, ill. 5 (line drawing). Cf. Â�PanÂ�caÂ�roğlu, 2004, p.╯157, fig.╯7. Another copper coin with a horseman drawing a bow against a dragon, of uncertain date, is inscribed with the name of Muḥammad ibn Salduq (c. 570/1174–597/1200); Lane Poole, 1877, repr. 1967, p.╯114, cat. no. 310. 132 ╇ Lowick, 1985, p.╯170; cf. Pancaroğlu, 2004, p.╯156. 133 ╇ Schlumberger, 1878, repr. 1954, pp.╯46–9, pl.╯2, no. 12; Whelan, 1980, pp.╯147–8. 134 ╇ Related imagery appears already on the fourthcentury gold medal of the Roman Emperor Constantine II (337–361), struck after the victory over Magnentius in 353. The emperor is shown on horseback with right hand raised over a coiled dragon with the legend debellator hostium.

Paris, Bibliothèque Nationale, Cabinet des Médailles, inv. no. 462, Vitr. XVIII; cf. Merkelbach, “Drache,” RAC IV, 1959, p.╯255, fig.╯2; Lewis, 1973, fig.╯31 (pls. unnumbered); Grube and Johns, 2005, p.╯232, cat. no. 78.3. 135 ╇ Ettinghausen, 1962, p.╯92; Hillenbrand, R., 2006, p.╯20. 136 ╇ Pinder-Wilson, 1997, p.╯344. 137 ╇ The Arts of Islam, 1976, p.╯182, cat. no. 200. 138 ╇ The same motif is found on other thirteenth-century inlaid copper alloy candlesticks, cf. for example, one from Anatolia (?) in the Victoria and Albert Museum, London, inv. no. M. 711–910, and another from Siirt, Anatolia, in the Nuhad Es-Said Collection of Islamic metalwork (Allan, 1982a, repr. 1999, pp.╯59–61, cat. no. 7). 139 ╇ Glazed fritware in the form of a horseman fighting a dragon, datable to the mid-thirteenth century. Discovered by Eustace de Lorey in Raqqa, Syria. Height 46.5 cm. Damascus, al-Mathaf al-Waṭanī, inv. no. A.5819. A Concise

129 130

100

chapter six

and with his sword and round decorated shield defends himself against the dragon as it rears up to strike. Images of paired dragons are prominently depicted in the Mosul area, the heart of the Jazīra, on the gateway to the only remaining caravanserai located between Mosul and Sinjār, known as Khariyyāt al-Khān or Khān al-Harārāt, which is now partly destroyed.140 The reliefs on each half of the archivolts of the monumental archway show two horned dragons with backward-facing heads, fighting off warriors on foot distinguished by haloes and shown in three-quarter view. The dragons are closely related to those on the Bāb al-Ṭilasm in Baghdad. Curved horns project from the crown of the head. The strong forelegs end in feet with individual pointed talons and the slender arched wings have finely delineated plumes. The heads are also rendered in three-quarter view with wide-open jaws revealing sharp teeth; high vertically hatched ruff-like projections demarcate the necks. The scaly serpentine tails form a pretzel-shaped knot and a single loop before gradually tapering to terminate in another small dragon head projecting from the inward-curling tail tip.╯The fabulous creatures are being attacked by the lances of the long-haired bearded figures who grip the tips of the dragons’ noses with one hand while holding the lance in the other. Importantly, as will be further discussed below, the dragon-fighting reliefs surmount an imposing arch charged with nine equidistantly placed cusped medallions enclosing eight-petalled starrosettes (figs. 100a and b). The epigraphic frieze on the portal of the khān identifies the Zangīd successor Badr al-Dīn Luʾluʾ Abu ’l-Faḍāʾil al-Malik al-Raḥīm (618/1222– 657/1259) as patron, through which the khān can be dated between 631/1233–4, when Badr al-Dīn Guide to the National Museum of Damascus, 1969, fig.╯2 (35); Furūsiyya, 1996, p.╯222, ill. 31, pp.╯236–7, cat. no. 198. ii. Two further faience fountain sculptures in this series, representing a sphinx with tail and wings ending in dragon heads and a rooster with tail ending in a bird head; dated here to the late twelfth century, are preserved in Copenhagen, the David Collection, inv. nos. Isl. 56 and Isl. 57, respectively. See von Folsach, 1990, p.╯104, figs. 129 and 128. 140 ╇ Van Berchem, 1906, 2v, v1, pp.╯197–210, 203–4; Â�Preusser, 1911, pl.╯17 bottom, captions on this plate mistakenly reversed; Sarre and Herzfeld, 1920, vol.╯1, pp.╯11, 13–5; Reitlinger, 1938, pp.╯149–50; Kühnel, 1950, p.╯8; Ettinghausen and Grabar, 1987, p.╯302, fig.╯325; Gierlichs, 1998, pp.╯35, 199, fig.╯1; drawing of the right spandrel of the caravanserai with parts of the inscription by Ernst Herzfeld, 19 December 1907, reproduced by Hauptmann von Gladiss, ed., 2006, fig.╯19.

Luʾluʾ became official ruler (atābeg) of Mosul, and the Mongol invasion in 656–7/1257–8, when his rule was terminated although he succeeded in keeping Mosul as vassal of Hūlāgū.141 The inscription also gives the atābeg’s honorary titles and relates him in a eulogy to the Saka hero Rustam, “the Rustam-i Zāl of our time,” thus showing him in a long line of kings and underlining his legitimacy as ruler.142 It was fitting for Badr al-Dīn Luʾluʾ who governed Mosul for the longest time, and who was a freedman probably of Armenian servile origin, to carry Arab names, Persian titles and to be depicted as Turkic leader.143 The accompanying inscription also grants sanctuary to all comers144 which underlines the protective function of the motif. The depiction of the hand-to-hand combat of the dragon-fighter on foot on the gateway at al-Khān has a counterpart in the representation on the Mosul gate which secures the entrance to the small mountain town of ʿAmādiya, or al-ʿImādiyya, northeast of Mosul.145 Its monumental display echoes that on al-Khān, in particular the portrayal of the fighters who attack the dragon with a sword (although not with a lance as at al-Khān). One way in which the representations differ is that at ʿAmādiya the horned dragons are shown addorsed: the protomes project in rampant posture at the apex of the arched entrance from the top of a large knot, reminiscent of the “Syrian knot,” that ties their ophidian bodies. The gaping mouths reveal the bifid tongues that entwine at mid-section (the same feature can be observed on figs. 78 and 160), and again the fighters grip with one hand the tips of the dragons’ noses while with the other hand aiming their swords at the dragons’ necks. Significantly, the quadripartite knot encloses a composite hemispherical rosette formed of fine pointed petals arranged radially 141

p.╯15.

╇ Van Berchem (in Sarre and Herzfeld, 1911) vol.╯1,

142 ╇ Idem, pp.╯14–5 and fig.╯8; Hauptmann von Gladiss, ed., 2006, p.╯123 and fig.╯19. 143 ╇ See for instance Badr al-Dīn Luʾluʾ, identified by the inscription of his name on the tiraz, is depicted as mounted falconer on the frontispiece of the Kitāb al-Aghānī (“Book of Songs”), dated 616/1219, Muḥammad ibn ʿAlī Ṭālib ibn al-Badrī, now in Copenhagen, Royal Library, Ms. Cod. Arab. 169. Hauptmann von Gladiss, ed., 2006, p.╯8, fig.╯1; Hillenbrand, R., 2006, p.╯19. Cf. Cahen, “Luʾluʾ,” EI2 V, 820b. 144 ╇ Ibrāhīm, 1976, p.╯13. 145 ╇ The gate was destroyed during bombing in the 1980s and then reconstructed. Gierlichs, 1995, pls. 2–3 (watercolour of 1955), and idem, 1998, pp.╯35, 199, fig.╯2; Hauptmann von Gladiss, ed., 2006, p.╯122, fig.╯16.

the dragon in scenes of combat

101

in superimposed layers. The rayed outline of the rosette adds to the overall starburst effect, underlining the identification of the rosette as a solar symbol. The base extends to form an interlaced festoon that frames the arched aperture which includes a small eight-petalled rosette at the apex (fig.╯101).146 The solar iconography of the large central rosette adds weight to the identification of the small rosette, just like the closely related rosettes emblazoned on the gateway to al-Khān, as stellar symbol. As also shown in the above-discussed reliefs on the Bāb al-Ḥayyāt in Aleppo (fig.╯3a), on Karatay Han (figs.╯4a and b) and on the small “Kiosk Mosque” situated in Sultan Han (fig.╯12), the latter two examples being located near Kayseri, this once again associates dragons with stellar constellations. As noted, in contrast to the depiction on al-Khān, on which the mirror image of the dragon fight is rendered in a confronted manner, the dragons are shown as addorsed. Their bodies issue from the knotted configuration that encloses the solar symbol and then extends to frame the arch. The creatures may thus be seen to hold or bind the sun by means of the knotted enclosure. Yet at the same time, they appear to emerge from the solar symbol. The relief is thus an important example of the dual symbolism inherent in the dragon’s connection with the sun which has often been associated with the occurrence of the eclipse as will be seen below. Only fragments remain of the architectural inscription147 which again names Badr al-Dīn Luʾluʾ as benefactor and thus dates the relief, like the façade sculptures at al-Khān, to the period of his rule. At that time the treasury was kept at ʿAmādiya,148 which might have provided all the more reason for Badr al-Dīn Luʾluʾ to emblematise himself as a dragon-fighter set within the complex of stellar allusions on the town gate. At the same time it is interesting to consider the identification of the standing haloed fighters with the syncretistic figure of the Islamic Prophet Khiḍr,149 guide of wayfarers and patron saint of travellers and, moreover, identified with the Christian dragon-slayer Saint George.150 This is especially pertinent in view of the widespread

popular cult of Khiḍr at the regional monastery of Mār Behnām/Deir al-Khiḍr, located southeast of Mosul, an important place of pilgrimage for Jews, Christians and Muslims. The iconography of the dragon-slayer, regarded as Saint George as well as al-Khiḍr, played a pertinent role at the monastery. In the age of syncretism this however does not exclude the possibility that Badr al-Dīn Luʾluʾ wished to benefit from these different layers of identification of the dragon-fighter motif and to build his historical charisma upon a rhetorical association of his personality with religious figures – his conscious association with al-Khiḍr – as well as “mythical” figures associated with Iranian legendary history and national epic – he is called “the Rustam-i Zāl of our time.” Being an astute diplomat he thereby clearly relied on the potency and cultural resonance that these figures possessed at the time, arguably to enhance his position as powerful ruler and defender of his realm as well as to appropriate symbolically the religio-cultural space of his realm. It also demonstrates his ambition to create a cultural paradigm by embracing a multilayered symbolism which includes syncretic religious aspects as well as ancient Iranian traditions. This was visually anchored in the figure of the dragon-fighter as well as the symbol of the interlaced dragons at ʿAmādiya. It may be hypothesised that akin to the motif of the interlaced dragons, added around a century earlier to the otherwise late Byzantine style coinage of the Artuqid ruler Fakhr al-Dīn Qara Arslan (figs. 32a and b), these symbols were chosen in the spirit of a conscious revival of imagery carrying an association with the glorious Iranian past. The royal messages conveyed by the sculptures on the gateways of al-Khān and ʿAmādiya demonstrate visual expressions of both power and ideologies that are remarkably fluid, traversing geographical, religious and cultural boundaries. Yet another manifestation of the mounted dragon-fighter is found in a mid- to late thirteenth-century Anatolian manuscript, known as Daqāʾiq al-Ḥaqāʾiq, dedicated to the Saljuq sulṭān Ghiyāth al-Dīn Kay Khusraw III, which is variously dated Ramaḍān 670/April 1272 and

╇ Cf. Gierlichs, 1995, pp.╯195–7. ╇ On the right side, “ʿIzz li Mawlānā al-Sulṭān al-Mālik al-Malik ar-Raḥīm al-ʿĀlim al-ʿĀsil al-Muʾayyad al-Muẓaffar al-Manṣūr al-Mujāhid al-Murābiṭ al-Muṭaghir al-Ghāzī Badr al-Dunyā wa ’l-Dīn,” and on the left side, “Atābak al-Aʿẓam Abu ’l-Faḍāʾil Luʾluʾ …,” after Gierlichs, 1995, p.╯202.

╇ Idem, p.╯203. ╇ Sarre and Herzfeld, 1911, vol.╯1, pp.╯13, 37–8; Coomaraswamy, 1934, p.╯181; Tabbaa, 1997, p.╯76. 150 ╇ Friedlaender, “Khiḍr,“ ERE, vol.╯14, 1915, p.╯695; Fiey, 1965, vol.╯2, pp.╯575–6; Franke, 2000, pp.╯155, n. 512, and pp.╯159–60; Baumer, 2005, p.╯110. See also the discussion in the Epilogue, part 2.

146 147

148 149

102

chapter six

mid-Shawwāl 671/early May 1273,151 although the illustrations may be of a later date.152 It shows the warrior as a mounted angel with long beard, wearing a three-pointed Iranian crown and shawl that streams in the wind. The figure is identified as Shamhūrash (an angel known as judge and ruler of the jinn), who is about to pierce a dragon with a sword.153 The weapon of choice, a sword, recalls the stucco relief from the Saljuq palace in Konya. The dragon’s head was partly cut off when the margin was trimmed, leaving only the open snout with curled tip; the long tail forms a large pretzel-like knot and a simple loop (fig.╯102). The manuscript comprises a compilation of five different Persian texts on various topics related to astrology and magic, influenced to a certain extent by Byzantine prototypes. The painting is part of a treatise on geomancy and talismans which is further illustrated by the depiction of magic writing on the page, composed of rows of numbers and letters. It may be postulated that the representations of the single fighter, mounted or on foot, and of the paired horsemen of the medieval Islamic period did not solely fulfil a decorative purpose. Illustrating the belief of the magical power of images, they very likely served as prophylactic and apotropaic representations with a talismanic function, precisely because as André Grabar has observed, “in this domain the possibility of a ‘consubstantiality’ of the representation and the thing represented is implicitly acknowledged.”154 It is also worth noting that in the pictorial representations the dragon and the hero are generally shown in ongoing combat; the monster is sometimes depicted in a rampant posture engaging the hero in furious battle as if to test his valour, hence it is clearly not yet defeated. By contrast in Islamic literature of the medieval period, as also seen in the examples cited above, the dragon is generally overcome by the hero. A notable excep-

tion to this rule is the story of Isfandiyār’s son, the heroic king Bahman who features in the early twelfth-century eponymous epic Bahman-nāma which was probably written by Īrānshāh ibn Abi ’l-Khayr between 485/1092–93 and 501/1107–8. In this account the king abdicates in favour of Humāy, the daughter of the king of Egypt, and then during a hunting expedition “in Dayr-i Gachīn between Isfahan and Ray”155 he is killed by a dragon.156 This may however, just like the metaphor of being “caught in the dragon’s maw,” be a euphemism simply intended to indicate the fact that he perished. As such it would be indicative of man’s relation with the dragon as being interdependent and transformative.

151 ╇ Blochet, 1926, pl.╯XIX; Hartner, 1938, p.╯143, fig.╯22; Barrucand, 1990–91, pp.╯113–4, pl.╯3c. 152 ╇ Rogers, “Saldjūḳids,” EI2 VIII, 936a. 153 ╇ Winkler, 1930, p.╯102. 154 ╇ Grabar, 1957, p.╯140. 155 ╇ This detail is recorded by Mehrdad Shokoohy from the anonymous Mujmal al-tawārīkh wa ’l-qiṣaṣ, as noted by Bivar, 2000, p.╯22. On the basis of the geographic location of Bahman’s battle with the dragon, Bivar relates it to the campaign (much of which took place in Media and Susiana) between Eumenes of Cardia, whose name coincides etymologically with Bahman, and Antigonus the One-Eyed, both successors of Alexander the Great, as recorded by the Greek historian Diodorus Siculus (19.44) who wrote in the first century bc. Antigonus completely vanquishes Eumenes in 316

bc and since the latter had been more popular with the Ira-nians, Antigonus was allegorically equated with the dragon (azhdahā). See Shokoohy, 1983, pp.╯448 and 451; Bivar, 2000, p.╯22. 156 ╇ See Hanaway, “Bahman-nāma,” EIr. The dragon is named Abr-i Sīyāh (“Black cloud”); Khāleqī-Moṭlaq, “Aždahā II,” EIr. 157 ╇ Walter, 2003, p.╯140. 158 ╇ Although the horseman was one of the most distinctive figures in the pictorial repertory of the Coptic arts of Byzantine Egypt, it is notable that he does not battle with a dragon; an example of a rider piercing a dragon on a Coptic tapestry band of the early Islamic period forming a rare exception. Cf. Lewis, 1973, p.╯54, fig.╯28 (pls. unnumbered). 159 ╇ Cf. Walter, 2003, p.╯37.

c.╇ The Eastern Christian holy rider as dragon-fighter The idea of connecting the cult and iconography of the Eastern Christian warrior saints with the dragon can be traced to at least the early seventh century.157 The dragon motif in the Christian church developed in the eastern confines of Byzantium,158 where the so-called holy rider vanquishing a dragon was a well-established literary topos and was represented in early wall paintings.159 Depictions are found on portable items, ranging from magical amulets to luxury objects, as well as on sacred architecture, in particular churches and funerary settings. The motif fell on particularly fertile ground in the Transcaucasus region which was part of the pan-Iranian religio-cultural realm and was steeped in its artistic conventions. The antecedents of the victorious equestrian figure fighting a dragon have been sought further west where they have been linked to votive reliefs with Thracian horsemen, confronted or single, and a serpent sinuously rising towards the rider’s

the dragon in scenes of combat

103

feet, or coiled around his staff or around a tree.160 Similarly the god Mithras is depicted on horseback accompanied by a serpent.161 However, there is no evidence that would establish a direct connection,162 since, as Christopher Walter also notes, in none of the cases does the serpent seem to be a noxious beast nor does the rider seem to battle with the serpent.163 On the contrary, in Mithraism, which became a widespread religion in the Mediterranean basin, Europe and the Near East, the serpent appears to have been “a symbol of beneficial, life-giving force.”164 It has further been suggested that the dragon-slaying iconography grew out of the tradition of associating the saints with ancient Greek mythologies,165 in particular the legend of Perseus and Andromeda.166 However, this theory is based on the assumption that the story of a Christian saint rescuing a princess or maiden from a dragon was ancient, whereas it dates back no earlier than the eleventh century, as will be shown below.167 Conversely, the iconography of a triumphant rider trampling on or slaying a fallen enemy occurs frequently in antiquity and has been widely used in different contexts.168 The concept of killing a serpentine adversary was introduced on Roman

imperial imagery for the first time to represent the victory of Constantine I (r. 306–337) over his enemy, one hundred years after the above-discussed investiture relief of Ardashīr I (r. 224–241). According to Eusebius’ Vita Constantini (III, 3), a painting in the vestibule of Constantine’s palace at Constantinople showed the emperor and his sons with a dragon writhing under their feet, identified by Grabar169 as representing Constantine’s vanquished enemy, Licinius, his former co-emperor, who was defeated and killed in 324, portrayed as being pierced and cast down into the deep.170 The lost Constantinian composition was disseminated throughout the late empire by a widespread coin type represented by the mid-fourth-century gold medallion struck by Constantine II (317–361) after the victory over the usurper Magnentius in 353. It shows the emperor with raised right hand, mounted on a horse that rears up over a dragon framed by the legend debellator hostium.171 The horse is represented with hind legs parallel, hooves touching the ground, while the forelegs are raised high over the coiled reptile. However, as Grabar has suggested, even though this newly introduced iconography of the triumphant emperor striking down or trampling a dragon

160 ╇ Thierry, 1972, p.╯259, fig.╯22; Mazarov, I., “Opit za rekonstrukcija na hipomita v devna Trakija,” Izkustvo 35 III, 1985, pp.╯20–30, as cited in Walter, 1989a, p.╯664 and fig.╯2; 2003, fig.╯11; Furūsiyya, 1996, vol.╯2, pp.╯221–2. 161 ╇ Rostovtzeff, ed., 1939, pp.╯112–6, pls. XIV, XV; Cumont, 1937, pp.╯63–71. Cf. idem, 1939, p.╯74. On the relationship between the Iranian Mithra and the Roman Mithra, see Zaehner, 1961, pp.╯99–104. 162 ╇ Hinnells, 1974, pp.╯244–5. Khāleqī-Moṭlaq (“Aždahā II,” EIr) tentatively suggests that the feast of Mihragān (mentioned in the Shāh-nāma) held after the victory of Farīdūn (Thraētaona) over Ẓaḥḥāk may possibly be connected to the story of the dragon-slaying by the god Mihr (Mithra), although no direct association between the ritual and worship of the festival and the dragon-slayer has been found. A simulated dragon-slaying by the Emperor Commodus during the mysteries of Mithras is recorded in a passage from Lampridius (Commodus 9); see Loisy, 1930, repr. 1983, p.╯182. However, Jean Calmard (“Mihragān,” EI2 VII, 15a) calls this attempt “another attractive but faulty interpretation,” based on the ancient noun mithrakāna, associating the suffix kāna (no longer akāna) with a variant of ghna (Ved. han, Old Pers. jan) meaning to strike or kill; mithrakāna thus refers to the killing (or sacrifice) for Mithra, Â�analogous with the expression designating the Indo-Iranian god Â�Verethragna. Nevertheless, it is of note that with Christianisation, the festival of mithrakān was consecrated to Saint George. Cf. Boyce, 1981, p.╯67. 163 ╇ Walter, 1989b, p.╯664. 164 ╇ Hinnells, 1974, pp.╯244–5, 247, and idem, 1975, p.╯295; Skjærvø, “Aždahā I,” EIr. 165 ╇ The representation of a horseman surmounting a coiled snake features on Greek bronze coins struck at Isinda

in Pisidia during the first century bc. However the rider is generally depicted holding his lance as if to thrust it forward, rather than straight down, that is, he is rarely shown to aim at the serpent’s jaws, see Hill, 1897, pp.╯223–4, pl.╯XXXVI, 3, 4. Cf. Der Nersessian, 1965, p.╯24; Whelan, 1980, pp.╯146–7. 166 ╇ Cf. Sakisian, 1937, p.╯228; Fontenrose, 1959, repr. 1980, pp.╯515–20; Sharon, “Ludd,” EI2, V, 798b. 167 ╇ In spite of the analogies of the stories of Perseus rescuing Andromeda from a sea monster at Joppa located close to Lydda, and that of the eleventh-century story of Saint George rescuing the princess from the dragon, the enormous gap of time does not allow for the establishment of a connection. Cf. Walter, 2003, pp.╯121–2 and n. 82, p.╯140 and n. 195. 168 ╇ It is interesting to note, though, that in Coptic Egypt there appears to be only one rare example of a horseman vanquishing an enemy, in which case a prostrate human figure lies under the horse’s hooves (fragmentary tapestry panel, late sixth or early seventh century, Washington, DC, The Textile Museum, inv. no. 71.6). See Lewis, 1973, fig.╯7 (pls. unnumbered). 169 ╇ Grabar, 1936, pp.╯43–4 and 130. 170 ╇ The earliest representation of the symbol of the serpent(-dragon) being slain by a labarum as metaphorical victory over evil, a special imperial standard modelled on the cross after Constantine’s conversion to Christianity, appears on bronze coins minted in Constantinople in 326–7. Â�Eusebius, tr. and ed. Cameron and Hall, 1999, p.╯209, fig.╯2. Cf. Demougeot, 1986, pp.╯94–118, esp.╯94–6. 171 ╇ Paris, Bibliothèque Nationale, Cabinet des Médailles, inv. no. 462, Vitr. XVIII. Cohen, 1892, vol.╯7, p.╯443; Merkelbach, “Drache,” RAC IV, 1959, p.╯255, fig.╯2; Lewis, 1973, fig.╯31 (unnumbered pls.); Grube and Johns, 2005, p.╯232, cat. no. 78.3.

104

chapter six

under his horse’s hooves remained essentially a symbol of victory, it also introduced a moral and spiritual meaning.172 It is this very motif which provides a link with the hagiographical iconography of the holy rider vanquishing an enemy which appears on Jewish173 or Christian magical “amulets” that circulated in the Byzantine world. These frequently show on the obverse an anonymous rider saint spearing a prostrate figure, generally represented as female with long hair, who raises her hands in a vain attempt at supplication or defence. More rarely the figure appears to be portrayed with a bare female torso and what may perhaps be a serpentine lower body.174 The choice of a female figure as the vanquished enemy reflects perhaps the ancient belief in dangerous female demons which may go back to the Lilith of Jewish legend.175 Of note is that the term Druj,176 the demoness of deceit and treachery, is applied in later Avestan texts to a whole class of female demons.177 It has also been

shown convincingly that the prostrate figure on these amulets represents a female demon who harms children and pregnant women as attested by late antique magical sources and that the amulets served as prophylactic charms against a variety of illnesses.178 However, the systematic study of this iconography is complicated by difficulties in dating the amulets and by the frequent absence of legends which would make it possible to identify the equestrian figure. The rare accompanying legends invoke Solomon the horseman, the formidable enemy of demons, and/or Saint Sisinnios179 (who is thought to have been a Parthian),180 permitting identification with either or both of these figures. Through the use of his “magical” seal and other ritual techniques attributed to him, the Jewish king Solomon was considered to be the archetypal controller of spirits and demons,181 the quintessential warrior-magician, a figure that reflected a conflation of parallel Jewish, Christian, Greek

╇ Grabar, 1936, pp.╯44, 47. ╇ While monotheism and the prohibition of images restrained Jews from developing divine or demonic figural images, the Second Commandment was not strictly observed by all Jewish communities. Special highlights of Jewish figural art are, for instance, the wall paintings of the thirdcentury synagogue at Dura Europos (see Rostovtzeff, 1939, p.╯102) and the mosaics of the sixth-century synagogue of Beth Alpha (Sukenik, 1932). 174 ╇ For a detailed analysis of this type, see Bonner, 1950, pp.╯99, 208–21, pls. XIV–XVII, nos. 294–301, 306, 309, 311, 314, 315, 318, 319, 323–7. Cf. Perdrizet, 1903, and idem, 1922; Peterson, 1926, p.╯103, 107 (lower illustration); Engemann, 1975, p.╯25 and fig.╯1, p.╯37 and fig.╯6, n. 111 with references on the so-called seals with Solomon on horseback; Thierry, 1999, pp.╯238–9 and ns. 20, drawings 2 a-c; Alexander, 1999, pp.╯1076–1077, fig.╯32.2. Cf. also Bank, A.V., Gemma s Â�izobrazeniem Solomona, Vizantijskij Vremennik 8, Moscow, 1956, pp.╯331–8; as cited in Walter, 1989a, pp.╯665–6, n. 57; idem, 2003, fig.╯14. The reverse of the amulets depict various scenes, the most common being the representation of the “all-suffering eye.” The ancient symbol is usually depicted as an eye whose pupil is pierced by various sharp weapons and attacked by several fierce or noxious animals; as portrayed, for instance, at the Dura synagogue (Engemann, 1975, pp.╯22–48 and p.╯27, fig.╯5; Doro, 1941, pp.╯220– 1). Cf. Walter, 1989b; Thierry, 1999, p.╯240. For a recent study dedicated to Sasanian-period glyptics with this motif which however sees the subjects represented on the seals as having originated in the Mediterranean and the Near Eastern Â�civilisations, see Magistro, 2000, pp.╯167–94. For a discussion of Byzantine amulets with a rider saint spearing a prostrate she-demon, see Spier, 1993, pp.╯25–62, esp.╯pp.╯60–2, and pl.╯2, nos. a, 15, and b, 21, pl.╯3, no. a, 33 and pl.╯6, no. d. 175 ╇ Cf. Hutter, “Lilith,” DDD, pp.╯520–1; Bonner, 1950, p.╯210. 176 ╇ See, for instance, Bundahishn 2.3.3.23–4; Boyce, 1984, p.╯50. 177 ╇ Cf. Langton, 1949, p.╯70; Bonner, 1950, p.╯210.

178 ╇ Bonner, 1950, pp.╯210–8; Spier, 1993, pp.╯25–62, esp.╯pp.╯33–44. For the transmission from early Near Eastern to Jewish and classical Greek mythology of the belief in a female demon, bringer of harm to children and pregnant women, see Barb, 1966, pp.╯1–23. The aid of Saint Sisinnios is also invoked in Armenian talismanic scrolls against the female demon Āl, known throughout the Persian-speaking world, who is believed to be the personification of puerperal fever which strikes women in childbed. Cf. Russell, 2004, pp.╯447–8. On the Āl, see also Goldziher, 1896, vol.╯1, p.╯16; Winkler, 1931, pp.╯104–7; and, in particular, the monograph of Eilers, 1979. 179 ╇ The most elaborate visual representation of the mounted Saint Sisinnios appears on the seventh- or eighthcentury fresco found on the west wall of chapel XVII in the Apa Apollon monastery at Bawīṭ in Upper Egypt where he spears a prostrate female figure identified as the demoness “Alabasdria.” Excavated in 1901–2 by Clédat, 1904–6, xii.2, pp.╯79–81; Perdrizet, 1922, pp.╯13–5; Lewis, 1973, fig.╯30 (pls. unnumbered). 180 ╇ Perdrizet, 1922, p.╯13; Maguire, 1995, p.╯57. Â�Peterson (1926, p.╯118) considers the Sisinnios legend to have an Iranian origin. Rika Gyselen (1995, pp.╯89–90), moreover, remarks upon the resemblance of the name (and apparent function) of Saint Sisinnios and that of the Iranian “mage” Sāsān found on many Sasanian magical seals while excluding the possibility of an inverse influence (on the figure of Sāsān, pp.╯55–6). 181 ╇ Solomon evolved as legendary magician, endowed with exceptional wisdom and believed to be skilled in the art of exorcising spirits and demons. A detailed description of an exorcism is related by the first-century historian Josephus (Antiquitates Iudaicae 8.45–9) as well as by the late antique Testamentum Salomonis, an uncanonical religious text which appears to be a Christian adaptation of a contemporary demonological Jewish text entirely devoted to Solomon’s power over demons (ed. from manuscripts at Mount Athos, Bologna, Holkham Hall, Jerusalem, London, Milan, Paris and Vienna by McCown, 1922, esp.╯chs. 3–18). Cf. Bonner, 1950, p.╯209.

172 173

the dragon in scenes of combat

105

and Iranian traditions when moreover interest in extraneous doctrines and cults was very active.182 These visual images of triumph over the enemy or over the powers of evil were naturally adopted for the presentation of warrior saints,183 the actual systematisation of their cult probably having taken place in Constantinople.184 The image of the dragon being trampled by a potent rider’s horse, traceable to at least the early seventh century, thus mirrors the symbolic meaning of spiritual triumph over the persecutions of tyrannical pagan rulers, representing the crushing of evil and Satan through conversion and the destruction of pagan temples.185 The close parallels in iconography between the Iranian and the Judaeo-Christian traditions expressing the fundamental juxtaposition between victor and vanquished, however, may in large part be due to the influence of Iranian dualist notions on the religions of the Near East from the Achaemenid period to the early centuries of the present era.186 The strong dualistic character of Jewish eschatology, including the belief in the coming of a saviour (messiah),187 a final struggle between Good and Evil, and the resurrection of the dead, seems also to suggest the strong possibility of a borrowing from Zoroastrian doctrine,188 where according to Yasna 30.3–5 the good and bad spirits were “twins” (separate, independent

individuals as stated by orthodox Zoroastrianism),189 as is evidenced by the Qumran texts, early Christianity and all Gnostic religions (Hermeticism, Gnosticism, Manichaeism).190 Christianity, in turn, inherited most of its eschatological perspective from the Iranian-inspired apocrypha of the Old Testament and the Jewish writings of the period just before its emergence.191 The eschatological concept in the struggle against a dragon in Judaeo-Christian tradition has close analogies in ancient Iranian texts, cited above. In the Old Testament the resurrection is connected with the myth of Leviathan (Hebrew liwyātān; Ugaritic ltn) who is slain by Yahweh and thus plays a role in the eschatological struggle.192 In the Syriac Apocalypse of Baruch (29.4; perhaps early second century ad) Leviathan together with another mythical monster, Behemoth, rises out of the sea at the coming of the Messiah.193 In the Apocalypse of John (20.1–3) the “old dragon” is identified with Satan, evil incarnate. An angel descends from heaven with a key to the abyss, the underworld, and a chain to fetter the dragon there for one thousand years.194 Once this term has expired, the dragon is freed to take part in the great final struggle in which it will be decisively defeated, resulting in the resurrection of the dead and the coming of a new age in which death and evil are no more.195 Almut Hintze concludes:

182 ╇ Bonner, 1950, p.╯18; Alexander, 1999, p.╯1077; Walter, 2003, p.╯35; Pancaroğlu, 2004, pp.╯152–3. 183 ╇ Der Nersessian, 1965, p.╯24. 184 ╇ Walter, 1989a, p.╯663. 185 ╇ Especially during and after the exile of the Jews in Babylon, Iranian influence may have been instrumental in bringing about the change in the conception of Satan from a servant of God (for instance, in Zechariah, 3.1–3.3) to his adversary (Duchèsne-Guillemin, “Ahriman,” EIr). At least in part a Jewish apocalyptic movement, the early Christian faith inherited a worldview in which Satan played a vital role as the ultimate adversary or opponent of God and his agents. In the visions of John, Satan is portrayed as composite beast, which emerges from the sea (Daniel 7), drawn from the myth of a cosmic struggle between a god and a sea monster or dragon informed by a Mesopotamian mythic pattern (cf. also the general discussion in Gunkel, “Die Drachentraditionen,” 1895, pp.╯29–90, and “die Traditionen vom Urmeer,” 1895, pp.╯81–111) to which Satan “gave his power and his throne and great authority” (Revelation 13.2). This imagery was used in polemical discourses and played an important role in the “demonisation” of either external enemies or internal adversaries of which the Roman imperial power was a designated agent. 186 ╇ Boyce and Grenet, 1991, pp.╯361–490; Gnoli, “Dualism,” EIr; Hintze, 1999, pp.╯72–9, esp.╯pp.╯75–6. 187 ╇ In connection with the concept of messianism in Judaism it is of interest to note the identification of the saviour and the serpent, which is supported by the numerical equivalence of the Hebrew letters of the words naḥash

(“serpent”) and mashiaḥ (“messiah”), both being equivalent to 358, therefore by the rules of gematria these words are considered to have a close affinity. The idea that the serpent-messiah would destroy the evil serpent is expressed in the thirteenth-century writings of Isaac Cohen of Soria in Spain, published by Scholem, 1926, p.╯273. 188 ╇ The main locus of contacts between Jews and Mazdāyasnians was most likely to be Persia and Babylonia where Jews lived among the predominantly Persian population; Hintze, 1999, p.╯78. Cf. Shaked, “Eschatology,” EIr. 189 ╇ Boyce, 1975, repr. 1996, p.╯193. 190 ╇ Shaked, “Eschatology,” EIr; Gnoli, “Dualism,” EIr. Cf. Zaehner, “Zoroastrianism,” CELF, p.╯159. 191 ╇ Shaked, “Eschatology,” EIr. 192 ╇ Isaiah 27.1. Heidel, 1942, repr. 1951, p.╯103; Hintze, 1999, pp.╯89–90. 193 ╇ Hintze, 1999, pp.╯80–1 and n. 38. Cf. 1 Enoch 60.7. Related beliefs are found in Gnostic traditions. It is interesting to consider a later fourteenth-century Samaritan chronicle which reverts to earlier Jewish, Christian and Samaritan traditions and relates of the first-century Dositheans:€“These people believed that the Serpent will govern the lives of creatures until the day of resurrection.”€See Isser, 1976, p.╯80. 194 ╇ Cf. Hintze, 1999, p.╯81, n. 39 with further examples. 195 ╇ Eadem, p.╯81, and n. 41. Cf. parallels in the book of Revelation (20.1–6) where there appears the story of the leader of the powers of darkness, also referred to as “the dragon, the primeval serpent, which is the devil and Satan,” being overpowered and cast into the abyss by an angel who “sealed it over him, to make sure that he would not deceive

106

chapter six

The metaphysical struggle between fighter and dragon of course represents in a basic sense an “antagonism of the light, celestial (sun), and the dark, terrestrial, principles.”197 By extension traditional combat depictions convey the triumph of good over evil and light over darkness. As Kiti Mačabeli observes, the older eastern model of the iconography of the light-bearing horseman lent itself naturally to Transcaucasian Christian imagery, resulting in an identification between the mounted saint as the embodiment of light and the positive principle, seen in the act of crushing the dragon, the quintessential symbol of evil and darkness.198 One illustration of this concept in a Sasanian context is the third-century depiction of the god Ahura Mazdā/Ohrmazd on horseback crushing the head of the Zoroastrian spirit of evil, Angra Mainyu/Ahriman (figs.╯86a and b). The earliest surviving representations of the dragon-slaying model in Eastern Christian art are found, as mentioned above, on the eastern confines of the Byzantine empire in Transcaucasia. This region, and in particular Armenia, was profoundly influenced by Parthian Iranianism, hence it is safe to assume that the artistic traditions had to a certain extent developed in symbiosis with the Iranised world.199 The model underwent a com-

parable process of development in neighbouring Cappadocia which was located at the crossroads of the Byzantine, Arab and Transcaucasian worlds. Among the earliest known instances of bas reliefs with the motif is a seventh-century Georgian stone stele from Ekikilise, village of Agʿegui in Kartli region (fig.╯103)200 preserved in the Museum of Fine Arts in Tibilisi. The iconography of a single unnamed dragon-fighting equestrian figure spearing a large dragon, whose spotted body is rendered as twice coiled, succumbing at his horse’s feet, can be clearly made out.201 Just as in the later examples from the Islamic world considered above, the rider is often portrayed with his head encircled by a halo and his upper body turned so as to appear nearly frontally; the horse usually has its hooves firmly planted on the ground and the serpent is knotted or coiled beneath it. Yet significantly, the battle with the dragon is now almost exclusively reduced to the moment of triumph visualised by the weapon of choice being invariably plunged into the animal’s throat, thus killing it. An important, possibly seventh-century, wall painting with confronted warrior saints fighting two dragons is found in Cappadocia, now in a very poor state of preservation, in the rock-cut church known as Mistikan kilise in the village of Güzelöz (Mavrucan), region of Çavuşin.202 It is one of the earliest surviving examples of the conceptual doubling, in other words the paired

the nations again” for a thousand years. Cf. Boyce and Grenet, 1991, pp.╯421, 446. 196 ╇ Hintze, 1999, pp.╯86–7. 197 ╇ Hartner, 1938, p.╯143, n. 45. 198 ╇ Cf. Mačabeli, K., Pozdneantičnaja torevtika Gruzii: Pomaterialam torevtiki pervych vekov nashej ery, Tbilisi, 1976, p.╯85, translated into German by Scholz, 1982, p.╯245. 199 ╇ Sasanian-style iconography is found in reliefs with scenes of the hunt on the seventh-century church of Ptghni (Ptghnavankʿ) in Ararat province. On the right side, below the archivolt on a window of the south façade, is a galloping rider armed with bow and arrow taking aim at the protome of an excessively long-necked monster with large gaping mouth. The accompanying inscription identifies the rider as Manuel Amatuni (d.╯389), who has been portrayed in the same manner as was customary for the Sasanian kings who were overlords of this part of Armenia (Der Nersessian, 1945, p.╯89). The weathered condition of the carving makes it difficult to determine the long-necked mythical creature with long agape snout and forelegs. It was identified by Der Nersessian (1945, p.╯88) as a griffin; the description misÂ�Â� takenly confuses the attribution of the lion and the standing man with the griffin and the rider; however on account of its very long sinuous neck and the excessively long wide open mouth with what appears to be a projecting tongue, it might represent a dragon. Eadem, 1945, pl.╯X.1 (photograph represented mirror-inverted); Thierry, 1987, p.╯365, fig.╯199.

200 ╇ The stele bears the inscription of which one can only decipher the word “...assomtavrouli...” (“...serpent...”). I would like to acknowledge my debt to Nicole Thierry who provided me with this information (the inscription was read by Mrs. Kétino Abachidzé); she identifies the rider as an early representation of Theodore. Another contemporary, yet unpublished fragmentary relief with the same motif from Berdadzor is also preserved in the Museum of Fine Arts in Tibilisi. Thierry, 1999, p.╯236. An early example appears on one side of a fragmentary capital, dated to the sixth century, found at Dwīn, the ancient Armenian capital, now preserved in the Historical Museum, Yerevan, inv. no. 2604, featuring a serpent protome beneath the hooves of a horse. The rider’s hand is shown holding the horse’s reins but he is not shown holding a weapon, hence it is unclear whether he is intended to represent a mounted warrior. Khalʿpakhʿchian, 1980, p.╯92, pl.╯9. 201 ╇ Thierry, 1999, pp.╯236, 240, fig.╯4. The stone relief is also reproduced in Süslü, 1987, p.╯644, pl.╯119, ill. 27 (wrongly identified as coming from Ani and as preserved in the Historical Museum, Yerevan). 202 ╇ It is noteworthy that while the wall paintings are very faint, the overall composition could still be clearly discerned when the present writer visited the church in October 2008 (which is probably due to the fact that this particular church is hardly ever visited). Thierry (1972, pp.╯258–63; eadem, 1984–5, pp.╯293–302, esp.╯pp.╯300–2, fig.╯88, pl.╯156a;

It is very likely that the Jewish/Christian tradition took over the image [of the struggle against a dragon] from the Zoroastrian one in order to formulate its own eschatological myth.196

the dragon in scenes of combat portrayal, of the equestrian warrior saints, a significant characteristic, as will be seen in what follows, introduced to augment and reinforce the intended effect, making the paired dragon-fighters a doubly potent emblem. Here a unique feature is added to the symmetrical composition of the two horsemen in that they direct their spears at two horned dragons which are entwined around the “trunk” of a central tree-like composition (fig.╯104).203 Portrayed with frontally rendered curved horns, globulous eyes and long open muzzles whose slightly bulging tops are demarcated by small wrinkles and which exhibit the lupine characteristic of the steppe dragons, the dragons’ heads come to face the horses’ heads. The equines do not prance or rear up on their hind legs, but pose heraldically with one foreleg raised. Nicole Thierry sought to explain the composite aspect of the iconography by the fact that Cappadocia served as a large military encampment from where Byzantine troops, among whom were large mercenary contingents, in part of different nomadic origins,204 went to wage war against the Persians and later against the Arabs. The influence of the mercenary troops was also felt in the progressive adaptation of military techniques and equipment of a “Turanian cultural community,” which sometimes occurred directly or through Parthian or Sasanian mediation.205 The depiction is in the tradition of widely emÂ�ployed symmetrical Sasanian imagery exemplified in the third-century investiture relief of Ardashīr I at Naqsh-i Rustam with the equestrian scene of Ardashīr facing Ahura Mazdā whose horses’ hooves are shown to trample, respectively, the head of a dead enemy and the serpent-wreathed head of Ahriman (fig.╯86b).206 This conceptual pairing aspect is also found on different media, notably on Sasanian and Sogdian textiles.207 Here, too, the pairing of the iconography was presumably not purely intended to create the effect of eadem, 2002, no. 13) dates this painting as early as the seventh century, a date that Walter (2003, p.╯125 and n. 99) cautions as perhaps being too early. It is of note however that in a late sixth-century church, dedicated to Saint George, at Zindanönü, near Çavuşin, traces remain of a fighting warrior saint, identified by an inscription as Saint George (only the upper section featuring the upper body of the saint holding a long lance and the frontally represented head of his horse remain, so it is unknown what he is spearing below). 203 ╇ Cf. Thierry, 1999, p.╯242, drawing 4, and eadem, 2002, p.╯124; Walter, 2003, p.╯56, fig.╯27. 204 ╇ Cf. Darkó, 1948, pp.╯85–97, esp.╯pp.╯90–1. 205 ╇ Cf. idem, 1935, pp.╯443–69, esp.╯pp.╯463–9. See Thierry, 1972, pp.╯263 and n. 67 with further references. 206 ╇ Ghirshman, 1962, p.╯132, fig.╯168.

107

pictorial symmetry but rather aimed at reinforcing its potent benefit. Hence the visual symmetry of the dragon-fighters was also intended to bring into play the belief in a double protection through the agency of its essential meaning, the triumph of good over evil.208 Yet while the influence of the Iranian symbolic repertoire is certainly felt in the symmetrical composition of horsemen and the tree-with-serpents, it is represented here in a Christian context. The motif therefore may allude, just like the much later above-discussed miniature from the Yachakhapatum, or a book of homilies, of 1216 illustrated at Skevra monastery in Cilician Armenia (fig.╯44), to the subject of the serpent and the fruit-bearing tree in Paradise. Hence, the representation may be seen as a compound motif linking the dragon-fighting horsemen and the tree-with-serpents. Comparable symmetrical imagery of two confronted equestrian warriors also flanking a large stylised tree with curling foliage is found on a now lost stone relief plaque (extant only as a drawing) from a church in Ani. Mounted on caparisoned horses with large saddle blankets, the horsemen again mirror each other’s actions. Yet instead of directing their weapons at serpents entwined around a tree as in the early, possibly seventhcentury, Cappadocian wall paintings at Mistikan kilise, the warriors plunge their long lances down the upturned throats of the dragons whose prostrate bodies are arranged in three loops under the horses’ hooves (fig.╯105).209 The depiction in the Ani relief thus follows more common visual conventions current both in Christian and Muslim contexts. In another Cappadocian wall painting, found above the vestibule door of the late ninth- or tenthcentury funerary chapel known as Yılanlı kilise (“serpent church”) in the Ihlara valley, paired warrior saints thrust their lances into the gaping 207 ╇ See for instance Otavsky, 1998, figs. 65, 68, 83–8, 97–9, 103, 104, 106, 108–111; also, the zandaniji silk with the scene of Abraham’s sacrifice of Isaac from Moshchevaya Balka. Belenitskii, 1980, p.╯228, line drawing at the upper right; Jerusalimskaja, 2000, p.╯98, fig.╯11. 208 ╇ Pancaroğlu, 2004, pp.╯153–4. See also Henry Maguire’s discussion (1994) on the significance of regular repetition of geometric designs in early Christian floor mosaics and textiles as protective devices. 209 ╇ Strzygowski, 1918, pp.╯287–90, fig.╯329, drawing after Brosset, 1860, p.╯33, text, pl.╯XXXVII. The church is dated 622 and Josef Strzygowski suggests a contemporary dating for the relief; however Marie-Félicité Brosset, who recorded the relief in his Ruines d’Ani published in 1860 (after a drawing by M. Kästner), believed that it could not have been carved earlier than the Islamic conquest in 1072.

108

chapter six

mouths of a powerful bicephalous dragon, elaborately speckled, who rears up diagonally across the space between the two horses; the larger of the two dragon heads is horned (fig.╯106).210 Just above the two dragon heads is a cross with an inscription running on either side that establishes a semantic relation between the idea of Christ’s triumph over evil on the cross and the victory of the saints over the dragon.211 The pictorial and textual elements in the composition of this painting above an entrance have been interpreted as apotropaic in intention. A related composition is found on a relief frieze on the western entrance to the Georgian church of the Virgin in Martʿvili, founded by king George II of Aphkhasia (912–957). Here two confronted unidentified holy riders with fluttering cloaks are shown to spear the raised gaping heads of a massive twice looped bicephalous scaled dragon (fig.╯107).212 As in the wall painting at Yılanlı kilise, one of the dragon heads is larger, has a more pronouncedly curved upper snout and is crowned by what appear to be long pointed ears. A pair of winged figures, probably angels, hover on either side of the riders, extending to them crowns from the arc of heaven that were won by the martyrs for their courage in the struggle with demons and invisible enemies.213 Among the military saints Theodore and George were predominantly associated with the miracle of dragon-slaying and often appear together.214 In the hagiographical tradition, Saint Theodore clearly preceded Saint George in the

role of dragon-slayer.215 Prior to the eleventh century Saint George is almost invariably depicted in combat with a man. Then during the eleventh century the visual expression alters and the saint is increasingly represented slaying a dragon.216 The earliest dated and visually identifiable representation of Saint George, killing a man and not yet a dragon, seems to be a relief depiction on the façade of the Armenian church dedicated to the Holy Cross which stands on the small island of Aghtʿamar situated southeast of Lake Van (now in eastern Turkey), erected by king Gagik Artsruni, ruler of the southern Armenian kingdom of Vaspurakan between the years 915 to 921.217 In addition to Saint George, two further mounted warrior saints, Theodore and Sergios, are portrayed. The rendition of three equestrian saints, an example of the serialisation of the images, again illustrates the intent to amplify their role visually and thereby intensify their beneficial effect as apotropaic devices. The Armenian martyr Saint Sergios, traditionally considered the defender par excellence against all kinds of evil and dangers, was here added to the commonly paired military saints Theodore and George. He is depicted killing a feline, probably a panther, while Saint George tramples a supine fettered human figure that in eleventh-century legends is sometimes identified as his persecutor, the emperor Diocletian (fig.╯108).218 Saint Theodore is shown thrusting a lance into the upraised open mouth of a dragon, portrayed without wings or legs, its

210 ╇ This visual differentiation between the two dragonheads might also indicate two different identities. Cf. Thierry, 1999, p.╯234. See also eadem, 2002, p.╯155. A visit to the church in October 2008 showed that the imagery was effaced to the point of being barely recognisable. 211 ╇ An inscription above the cross states the words that are uttered by the pierced serpent heads: “Cross, who made you shine? The Christ, he who is struck in me!” Thierry (1999, p.╯234) interprets these words as the serpent, the personification of evil and death, recognising the defeat inflicted upon it by Christ through his death. 212 ╇ Cf. Aladaşvili, 1977, pp.╯48–56, 150–1, pls. 149, 150; Thierry, 1999, p.╯240 and n. 40, fig.╯5. 213 ╇ Walter, 1995, p.╯301; Thierry, 1999, p.╯243 and n. 64. 214 ╇ Walter, 1995, and idem, 1999. See also, idem, 2003, pp.╯44–66 and 109–44. For a discussion of the cult of saints, in particular the cult of Saint George, in Georgia, see Schrade, 2001, pp.╯169–98. 215 ╇ There are two saints called Theodore in the Orthodox tradition: Theodore Tyron (“the recruit”) and Theodore Stratelates (“the general”). On the confusion between the two Theodores, see Oikonomidès, 1981, pp.╯327–35. The legend of the Theodore Tyron mentions that he was “born in an eastern land” and died in Amaseia (Amasya), in presentday Turkey (from where his remains were taken to Euchaita,

present-day Avkat). His exploit of vanquishing a dragon with a spear only appeared in the second state of his Passio Prima (dated 890); Bibliographica hagiographica graeca, 1762d in Paris graec. 1470. Walter, 1995, p.╯309 and n. 87, and idem, 2003, p.╯50 and n. 38. Cf. Hengstenberg, 1912, pp.╯78–106, 241–80; “Theodore Teron,” ODB, vol.╯3, pp.╯2048–9. Some antecedents of Theodore’s dragon-slaying feat may be found in the seventh-century Passion of Marina of Antioch. Cf. Merkelbach, “Drache,” RAC IV, 1959, pp.╯246–7; Boulhoul, 1994, pp.╯255–304, esp.╯p.╯263; Thierry, 1999, p.╯242, n. 52. 216 ╇ Thierry, 1999, p.╯241. The earliest dated example of Saint George piercing a dragon is depicted in Cappadocia at the church of Saint Barbara at Soğanli (1006 or 1021). See De Jerphanion, 1925–1942, vol.╯2, pl.╯189. 2; Thierry, 1999, p.╯241. It is noteworthy, however, that in Cappadocia George was never represented as killing a man; cf. Walter, 2003, p.╯128. The fullest repertory of Georgian representations of Saint George piercing a human figure with his lance is given by Tschubinashvili, G.N., Georgian Repoussé Work, Tbilissi, 1957, cited after Walter, 1989a, p.╯665. 217 ╇ Cf. Der Nersessian, 1965, p.╯19, figs. 49, 50. 218 ╇ Scholz, 1982, p.╯242 and n. 1; Walter, 1989a, p.╯665, and idem, 2003, p.╯56 and n. 64, and p.╯129 and ns. 125 and 126.

the dragon in scenes of combat

109

body patterned with spots and tied into a heartshaped knot (fig.╯109).219 The warrior saints hold long cross-ended lances and are portrayed in full military attire according to iconographic types derived mainly from elements borrowed from antiquity, such as the chlamys tied on the right shoulder, the cuirass, short tunic and buskins.220 An early eleventh-century depiction of paired confronted horsemen, identified here as the two saints George and Theodore, spearing respectively an anthropomorphic figure and a twicelooped dragon and surmounted by a stern image of Christ, appears twice on the church of Nikʿorcʿminda built during the last years of Bagrat III’s reign (1010–1014) in the mountainous region of western Georgia, in the province of Racha (fig.╯110).221 The scene is once shown on the large tympanum over the western portal and again on the eastern façade. The belief in the saints’ power to vanquish evil probably inspired the representation of the motif on the interior and the exterior of churches and funerary structures. Its recurrent depiction in the Transcaucasian region, notably on Georgian churches, attests to the great importance accorded to the motif. Dragon-slaying riders were progressively identified as warrior saints and can conclusively be interpreted as exercising an apotropaic or protective function.222 The legend of Eastern Christendom’s best known and most venerated dragon-slayer, who also enjoyed considerable popularity in the Latin West, partly as a result of the Crusades,223 is first attested in the eleventh century when miracles were introduced into the Saint George cycle which includes the feat of rescuing a princess from a dragon.224 One such typical story is recorded in an eleventh-century Georgian manuscript preserved in the Patriarchal Library, Jerusalem.225

It describes how Saint George stopped at a lake where he saw a weeping girl about to be sacrificed to the dragon and there performed the miracle of vanquishing the dragon; thereafter the saint accompanies the princess who leads the dragon on a leash towards the city of Lasnia while her father, king Selenios, and the citizens look on. The miraculously subdued beast is tamed and symbolically tied to a leash and led about, a crucial detail that establishes the ceremonial and ritual of the subjection of the dragon. The earliest dated manuscript in which this story is illustrated is also found in Georgia at Pavnisi (c. 1170–1180).226 However there are other Georgian examples that can be dated earlier on stylistic grounds: Adish (late eleventh century),227 Bočʿorma (c. 1100),228 and possibly Ikʾvi (twelfth century).229 Thereafter the dragon miracle was incorporated as a regular feature in the iconographical repertoire of Saint George. Yet in Greek the miracle of the dragon is first attested in a late twelfth-century manuscript as an appendix to an account of the Passion of Saint George, and in a roughly contemporary illustration from Anargyroi in Kastoria (c. 1180).230 It is therefore very possible that the story of Saint George and the dragon miracle originated in the Transcaucasian region, probably in Georgia, from where it spread.231 The depiction of the saintly dragon-slayer on a funerary stele from Maraş in southeast Anatolia, now in the Archaeological Museum of Adana,232 attests to the continued importance of the motif. A beardless rider distinguished by a circular nimbus and with fluttering scarf is shown to hold a lance in both hands and thrust it into the large gaping jaws of a dragon, depicted with backward thrown head and short forelegs, that succumbs at his feet. The dragon’s body is rendered according to the

╇ The cross-ends of his lance are broken off. Cf. Der Nersessian, 1965, p.╯19. 220 ╇ Walter, 2003, p.╯22. 221 ╇ Baltrušaitis, 1929, pl.╯LXVII, fig.╯108; Neubauer, 1976, fig.╯57; Aladaşvili, 1977, pp.╯147, 158–9, pl.╯156; Thierry, 1999, pp.╯244, fig.╯6. The bas-relief on the eastern façade is reproduced in Scholz, 1982, fig.╯2. 222 ╇ Cf. Walter, 1989a, p.╯665. 223 ╇ The town Ludd (Lydda) in Palestine to the southeast of Jaffa is traditionally believed to be the birthplace of Saint George. Cf. Sharon, “Ludd,” EI2 V, 798b. 224 ╇ Walter, 1995, pp.╯320–2, and idem, 2003, pp.╯128–9; Schrade, 2001, p.╯177, n. 56. 225 ╇ Georgian manuscript, datable to the eleventh cenÂ� tury, Greek Patriarchal Library, Jerusalem, cod. 2. The GeorÂ�gian text has been translated into English by Walter (1995, pp.╯321–2 and idem, 2003, p.╯141) on the basis of

the translation into Russian (Privalova, 1977, p.╯73 and n. 57). 226 ╇ A thrice-looped dragon; Privalova, 1977, pp.╯16–7, 19, fig.╯5 (line drawing). Cf. Thierry, 1999, p.╯241, n. 50. 227 ╇ Undulant dragon; Privalova, 1977, p.╯77, fig.╯18 (line drawing), pl.╯XVIII, figs. 1, 2. 228 ╇ On account of the weathered surface, the drawing is too indistinct to identify whether the body was looped or knotted; eadem, p.╯83, fig.╯20 (line drawing). 229 ╇ Horned undulant dragon with long, flaring snout and goatee beard; eadem, p.╯80, fig.╯19 (line drawing), pl.╯XIX, figs. 1, 2. Reproduced in Walter, 1989b, p.╯357, fig.╯3. 230 ╇ Aufhauser, 1911, pp.╯31–3; Walter, 1989b, pp.╯350–1. 231 ╇ Walter, 1989b, p.╯351, and idem, 1995, p.╯322. 232 ╇ Height 56 cm, width 30 cm, thickness 4 cm. Adana, Archaeological Museum, inv. no. 76B. Süslü, 1987, p.╯638, pl.╯118, ill. 1.

219

110

chapter six

standard iconography with a pretzel-like knot at mid-section followed by a further loop.╯A pair of flying figures hovers over the canopy of the rider in a sign of glorification. A three-line inscription in Syriac as well as in Arabic associates the stele with the Christian minority and gives the date, 701/1301–2, as well as the name of the master, Ibn Bahri.233 Given that it is utilised as a symbol of victory in a funerary context, the presence of the dragon-slaying motif on this stele appears to echo the popular eschatological notions of the wall painting above the entrance of Yılanlı kilise in the Ihlara valley.234 The iconographic semantics of the equestrian dragon-fighter – from the greater Khurasan region to Asia Minor – in its heroic as well as saintly incarnation, thus owe much to ancient prototypes that germinated in the syncretistic melting pot of the great Near Eastern religions. These were probably inspired to a large extent by ancient Iranian dualist notions, and specifically eschatological thought systems, which resulted in close parallels between Iranian and Jewish concepts, inherited, in turn, by Christianity.235 As can be seen in both material culture and written sources, there is overwhelming evidence that the figure of the Iranian holy rider battling the dragon (more rarely also visualised as fighting on foot) largely served to articulate the many strands of this complex image. The latter thus informed a visual vocabulary that proved to be meaningful as an internationally recognisable symbol for adherents of different religions in the medieval Central Asian sphere. As a consequence it allowed for effortless re-contextualisation and seems to have served as connecting symbol between different cultural and religious spheres on a popular level, in some ways coming to symbolise a cultural syncretism. These strong syncretistic aspects are evident on a twelfth- or thirteenth-century nielloed silver

scabbard, which sets the dragon-slaying iconography alongside the dragon’s inherently prophylactic aspect. It moreover represents an example of a pastiche expressing in visual form the coexistence of themes from different religio-cultural repertoires. The scabbard is inscribed around the rim with customary good wishes in Arabic rendered in cursive script. Whereas the epigraphic frieze is worked in relief, an unusual, shallowly incised depiction is shown below. It differs not only in subject matter but also in style and execution from motifs shown on metal objects fabricated in the medieval Islamic world:236 a warrior saint in crusader attire is shown to aim at a prostrate dragon, its slender serpentine body forming a pretzel-like knot at mid-section, and the “Hand of God” projecting from the right to point at the warrior (fig.╯111). The dragon-slayer motif surmounts a register containing an ancient iconographic theme of almost worldwide currency, again executed in relief, showing a bird of prey attacking a quadruped, probably a deer. Below this a similarly time-honoured theme, a procession of animals of the hunt, is set against foliate scrolls, typical for the decorative repertoire of a local workshop which conceivably employed craftsmen belonging to one of the monotheistic religions of the Near East, where this sheath is thought to have been made for a Christian, probably crusader, client.237 In contradistinction however to the dragon-slaying motif on the scabbard decoration, the quillon block of the accompanying dagger, both preserved in the Furusiyya Art Collection, Vaduz, extends into downwardcurving prongs that terminate in dragon heads, a feature frequently employed on quillon blocks and imbued with an inherent apotropaic function.238 However, the rendering of the dragon heads is already highly stylised, perhaps suggesting that the heads had lost some of their original protective quality to become a more decorative feature.

╇ Erdmann, 1957, p.╯372, n. 32. ╇ Cf. Pancaroğlu, 2004, p.╯161. 235 ╇ Cf. Shaked, “Eschatology,” EIr; Hintze, 1999, pp.╯86–7. 236 ╇ I would like to thank Professor Robert Hillenbrand for drawing my attention to this point. 237 ╇ Cf. Chevaux et cavaliers arabes, 2002, pp.╯118–9, cat.

no. 57; L’art des chevaliers, 2007, pp.╯155–7, cat. no. 148. For a discussion of Islamic metalwork objects with Christian themes made for Muslim and Christian patrons, cf. Baer, 1989, pp.╯41–9. 238 ╇ See chapter 3. Also L’art des chevaliers, 2007, p.╯154, cat. no. 147.

233 234

the dragon in relation to royal or heroic figures

111

chapter seven

The dragon in relation to royal or heroic figures

a.╇ Royal and heroic associations of the dragon in literary accounts In Iranian legends the dragon combat was one of the wonders and heroic feats required as proof of the king’s or hero’s legitimacy, so becoming by extension an important device of royal or heroic ideology.1 A royal victory over the dragon was intended to manifest virtuous conduct and to endow the royal persona with heroic qualities. The visual enactment of this victory communicated mastery over the mighty mythical creature as well as implying metaphorically that through this deed of prowess the vanquisher was able to take on the formidable qualities of the dragon, that is to say, assume part of the dragon’s nature, as will be shown below. Just as in the visual arts the dragon’s powerful likeness was evoked, so it served also in literature as a simile in the formulation of praise and panegyric for rulers and their entourage. Drawing on the dragon’s qualities of instilling fear and dread as well as awe and reverence, the creature was evoked metaphorically as a lively expression of the heroic qualities of the mamdūḥ (“the praised one”). Repetitive reference to the dragon was thus a rhetorical device used by poets and historians, among them Firdawsī, the celebrated poet from a village near Ṭūs, near present-day Mashhad in Khurasan, who began writing the Shāh-nāma during the last decade of Samanid sovereignty and completed it during the rule of the Ghaznawids. The latter were ethnically Turkish but were deeply imbued with the Persian and Islamic courtly traditions. In the opening verses, Firdawsī eulogises the royal patron, sulṭān Maḥmūd of Ghazna (r. 389/999–421/1030), with the words: At his banquets (bazm), he is a heaven of fidelity In combat (razm), he is a dragon with sharp claws.2

╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. ╇ Tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯25, ll. 221–2. 3 ╇ Cf. Clinton, 1972, p.╯130. 4 ╇ Dīwān-i Hakīm-i Farrukhī-i Sīstānī, p.╯363, l. 7354, as 1 2

By thus creating a visual image like a frieze or bas-relief, the court poet uses his skill to publicly display and celebrate the virtues and accomplishments of the patron.3 Firdawsī’s flattering rhetoric is echoed by that of the foremost paneÂ� gyric poets at the Ghaznawid court, Abu ’l-Ḥasan Farrukhī Sīstānī (d.╯429/1037–8), who called his royal patron an “illuminating sun” at banquets and a “roaring dragon” in combat.4 The imagery of the dragon seems to have represented a heroic ideal and was frequently used in panegyrics in the Qābūs-nāma (“Book of Qābūs”), written in 475/1082–3 by the Ziyārid prince ʿUnṣur al-Maʿālī Kay Kāwūs ibn Iskandar ibn Qābūs, for the edification of his son. In this well-known “Mirror for Princes” the 63-year-old prince remarks upon the fashion for court poets to liken the “mighty” to a “dragon” or a “lion;” urging his son to weigh careÂ�fully such eulogies.5 Dragon symbolism was not only reserved for the sulṭān; Manūchihrī Dāmghānī, poet at the court of Maḥmūd’s successor, sulṭān Masʿūd of Ghazna (r. 421/1030–432/1040), also performed services for the sulṭān’s chief vizier, Aḥmad ibn ʿAbd al-Ṣamad, by extravagantly praising his virtues in an ode (qaṣīda) in which he too calls upon the symbolic meaning of the dragon: With such petty enemies why should the Khwājah do battle? The dragon is shamed who fights with a chameleon.6

That this metaphorical or allegorical use of the dragon image was in widespread currency is further attested by its use in a qaṣīda by Ẓahīr al-Dīn al-Fāryābī (550/1156–598/1201–2; as indicated by his nisba he was perhaps born at Fāryāb near Balkh). In his dīwān the poet glorifies the courage of the last Great Saljuq ruler in the west, sulṭān Ṭoghrıl III ibn Arslan with the words:

cited in Melikian-Chirvani, 1997a, pp.╯143–4. 5 ╇ Tr. Levy, 1951, p.╯133. Cf. Klíma, 1968, p.╯95. 6 ╇ Dīwān, p.╯25, cited after Clinton, 1972, p.╯45.

112

chapter seven When his enemies saw his dragon countenance They raised their hands above their heads like â•… scorpions.7

In a dramatic double simile in the Iskandar-nāma, Niẓāmī brings together the heroic qualities of the dragon with those of the lion to describe: ...an Alan knight, similar to a fierce lion, [who] came forth like a black dragon, with a mace in his hand, hard to bear for less than seventy men and able to split Mt. Alborz.8

The mythical creatures are thus invoked in order to endow the subject of eulogies, by association, with praiseworthy superhuman qualities. In the Haft Paykar Niẓāmī again jointly invokes the lion and the dragon in the allegorical description of king Bahrām seizing the crown from between two lions: The golden crown in two black lions’ jaws as in two dragons’ jaws a (lustrous) moonA moon escaping with a basin’s noise the cloud, but the basin eke a sword, The two vindictive lions lashed their tails upon the ground like dragons twain (in wrath); To say, Who’ll seize from us this golden crown? Who’ll dare a lion or a dragon rob? They knew not of that man of iron heart, who captured lions, dragons hunted too.9

Firdawsī evokes related heroic imagery in the Shāh-nāma in the story of the demon Akwān who at royal command was slain by Rustam who himself is repeated likened to a dragon.10 The hero assures the king that “whether demon, lion or male dragon, it will not escape my sharp sword,” and is himself described as arriving at the battle scene “like a male lion riding a dragon.”11 The court poet Masʿūd-i Rāzī was exiled to India by the Ghaznawid sulṭān Masʿūd ibn Maḥmūd for daring to admonish the sulṭān during the festival of Mihragān in Dhu ’l-Ḥijjah 430/September 1039 with the lines: Your enemies were ants, but now they have become vipers; hasten to destroy these ants turned vipers. 7 ╇ Ed. Bīnish, T., Tehran, 1958, p.╯43, as cited in Daneshvari, 1993, p.╯16, n. 7. 8 ╇ Muḥammad Ilyās ibn Yūsuf Niẓāmī, ed. and French tr. of the Russian episode by Spitznagel, L., Expédition d’Alexandre le Grand contre les russes, extrait de l’Alexandréide ou Iskèndèr-Namé de Nizâmy, St. Petersburg, 1828, cited after Alemany, 2000, pp.╯264–6 and n. 57. 9 ╇ Tr. Wilson, p.╯73. 10 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯187, l. 566. 11 ╇ Idem, vol.╯3, p.╯275, ll. 63–5.

Give them no more time, and do not tarry, for if given time a viper will become a dragon.12

Similar allegorical language is also used in the History of the Saljuqs by Ibn Bībī (d.╯after 684/ December 1285) who referred to rebellious youths as “young serpents that grow into dragons”13 and likened the victorious Saljuqs to “threatening dragons.”14 Rustam, the legendary prince of Sistān, is conspicuous for his role in the battles with his archenemy Afrāsīyāb, the legendary king of Tūrān in the Iranian epic. In the “Book of Qay Qubād” of Firdawsī’s Shāh-nāma, his father Zāl however warns him about going to war with the Tūrānian who is identified as a Turk and as a powerful “male dragon”: Zāl told him, “Listen my son Today, for once, think hard: In war, this Turk is a male dragon In skirmish and in combat he is a cloud of Â�calamity. 15

In literary tradition the dragon is hence used to emphasise the heroes’ and rulers’ superhuman qualities by allegorically transforming them into dragons whose likeness was often singled out as apt symbol of royalty and superiority.16 This is also noticeable in the visual language, for instance on a large marble slab relief-carved with addorsed regardant “Saljuq-style” dragons in rampant posture with forelegs and unfolded wings. Significant are their long interlaced tails. Their open mouths reveal enormous fangs with long bifid tongues thrust out, their forelegs with separated talon-toed paws rest on their long upward-curving sinuous tails. The plaque carries an epigraphic band inscribed with the formula al-sulṭān al-muʿaẓẓam (“the exalted sultan”), a title held by the Saljuqs (fig.╯112). It comes from the cenotaph of the Mamlūk ruler al-Muʾayyad Sayf al-Dīn Shaykh (d. 1421/824) in Cairo, which Max van Berchem has identified as (much earlier) spoils from the sulṭān’s Mesopotamian campaigns17 and for which Rogers suggests an Anatolian Saljuq provenance.18 The inscription ╇ Meisami, 1990, p.╯41. ╇ Al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, p.╯309. 14 ╇ Idem, p.╯212. 15 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯467, ll. 34–9. 16 ╇ Cf. Melikian-Chiravani, 1984, p.╯323. 17 ╇ Van Berchem, 1910, p.╯81, fig.╯34. 18 ╇ Rogers, J.M., Islamic Art in Egypt 969–1517, exhibition catalogue, Cairo, 1969, pp.╯203–4, no. 194, as cited in Ibrāhīm, 1976, p.╯11. Cf. Rogers, 1969, p.╯154, n. 14. 12 13

the dragon in relation to royal or heroic figures explicitly identifies the iconography of the paired dragons as a royal ensign of a now anonymous sulṭān. It is interesting to note the analogy of this type of imagery to the stone reliefs at the türbe of Emir Saltuq at Erzurum (fig.╯10) and at the church of Saint Gregory in Ani (fig.╯11). Closely related interlaced dragons are depicted on portable items such as the twelfth-century gold finger ring from the greater Khurasan region (fig.╯30) where they served to strengthen the protective function of the objects, and were chosen as emblematic motifs west of Iran for the coinage of twelfth- or Â�thirteenth-century rulers in the Jazīra (figs. 32a and b). This spolium is evidence that not only did the motif of the interlaced dragons prove to be meaningful beyond the Central Asian realm but that the iconography could be re-contextualised in a fifteenth-century Mamlūk setting where it was distinguished as an emblem that was deemed suitable for inclusion in a royal funerary setting. b.╇ Paired dragons with a central anthropomorphic figure Dragon symbolism and cosmic rulership

113

able beyond the caliphal court. According to the tenth-century philosopher Miskawayh, the ambitious Ziyārid ruler Mardāwīj ibn Ziyār (d.╯323/935) was imitating Sasanian custom when he sat on a golden throne and wore a crown.21 The historian Abu ’l-Faḍl Bayhaqī (d.╯470/1077), recorded in his Taʾrīkh-i Masʿūdī (“History of Masʿūd”) that the Ghaznawid sulṭān Masʿūd ibn Maḥmūd held court sitting on a dais and in 429/1038 had his throne, originally made of wood, replaced by a golden, jewel-studded throne, three years in the making, that he placed on a dais in his new palace.22 In the manner of the Sasanian kings he had a seventy-man crown suspended from golden chains.23 The grandeur of the house of the Ghaznawids was, in the words of Bayhaqī, “resplendent as the sun.”24 The twelfth-century Khurasani poet Awḥad al-Dīn Muḥammad Anwarī (c. 520/1126–583/1187–8 or 585/1189– 90), who was born near Abīward, had served sulṭān Sanjar ibn Malik Shāh and survived the collapse of the Great Saljuqs, expresses similar metaphorical notions for the enthroned sovereign: May the world be your servant and destiny your page The sky your throne and the sun your crown.25

Artistic representations developed during the ʿAbbasid period projected an idealised vision of rulership which included the creation of physical manifestations of their imperial pursuits, in particular with respect to court ritual. Iranian influence on court ceremonial (marasīm) became noticable reflecting “the profound iranisation of customs and society.”19 Considerable importance was attached to the throne (sarīr, kursī). The caliph sat on a throne placed on a dais (ṣuffa) when he presided over court ceremonies. Iranian features in these ceremonies sometimes included the caliph’s elevation on a raised platform or throne (Pers. takht), rather than just a dais.20 These influences were notice-

His words aptly express the celestial and solar associations of royal ideology, areas in which the dragon theme played an important role. The conventional medieval Islamic depiction of rulership shows a frontally rendered human figure, the head in three-quarter view, seated cross-legged on a raised platform between two poles, which serve as the arms of the “chair,” and which are topped by dragon heads, their wide-open jaws revealing tongues commonly oriented towards the head of the human figure. The figure is often shown with arms extended at either side, grasping the staffs just below the dragon heads. At times he holds a staff in one hand and what looks like a cup in the other, imagery suggestive of cosmic kingship.26

╇ Sourdel, 1960, pp.╯121–48; cf. Spuler, 1976, p.╯346. ╇ Sourdel, 1960, p.╯131. 21 ╇ Tajārib al-umam, tr. and ed. Margoliouth, 1921, vol.╯5, p.╯489. Firdawsī similarly speaks, for example, of the golden throne of Ẓaḥḥāk (Shāh-nāma, tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯109, ll. 449, 590), the golden throne given by Farīdūn to his successor Manūchihr (idem, vol.╯1, p.╯167, l. 627) or the golden throne of Kay Kāwūs (idem, vol.╯2, p.╯49, l. 526). 22 ╇ Ed. Ghanī and Fayyūḍ, Tehran, 1324/1945, p.╯438, as cited in Lambton, “Marasīm,” EI² VI, 518a. 23 ╇ Op. cit. Cf. Meisami, 1999, p.╯77. For the “hanging

crown” in the throne hall of the Sasanian kings at Ctesiphon which perhaps was used by the Umayyad owner of the desert residence Khirbat al-Mafjar (in this case in the form of a qalansuwa) and later also entered Byzantine court ceremonial; cf. Ettinghausen, 1972, pp.╯28–30. See also the crown suspended above the throne of Manuchehr; Shāhnāma, tr. and ed. Mohl, 1838–1878, vol.╯1, pp.╯211, ll. 1142– 143. 24 ╇ Meisami, 1999, p.╯296. 25 ╇ Dīwān, vol.╯1, p.╯200. Cf. Melikian-Chirvani, 1984, p.╯327. 26 ╇ Cf. Baer, 1983, pp.╯258–66.

19 20

114

chapter seven

The visual expression is particularly prevalent on mid-eleventh to thirteenth-century silver-inlaid metalwork and seems to have first appeared on objects produced in the greater Khurasan region. Dragon-headed staffs flanking this traditional medieval Islamic portrayal of rulership are portrayed in one of eight roundels on the lid of the covered copper alloy bowl, inlaid with silver, known as the Vaso Vescovali, made in the Khurasan region around 1200.27 The “ruler on a dragon-throne” is in turn symmetrically flanked by attendants while the dais is supported by a pair of addorsed feline protomes (fig.╯113).28 Since the other roundels enclose representations of the seven conventional planets, the depiction in the eighth roundel has been associated with the pseudo-planet jawzahar, illustrating here the eighth “planet.”29 The earliest dated instance of the “dragonthrone” is found on the 559/1163 Bobrinski bucket, where it is repeated four times in the main register separated by an epigraphic frieze of anthropomorphic and zoomorphic naskhī, the lower parts of the hastae often terminating in dragon and other animal heads, expressing good wishes for the owner (fig.╯114).30 These roundels are encircled by 28 radiating “rays,” alternately inlaid in silver and copper. The frieze is the uppermost of a series of three figurative registers that circumscribe the richly inlaid copper alloy bucket. It is interesting to investigate these dense pictographic friezes since they provide the wider context for the “ruler on a dragon-throne” iconography. The second register encloses figures engaged in various popular pastimes ranging from backgammon and banqueting to making music and fighting with staves. The third features a hunting party with several bow-bending, sword- or lance-wielding figures on horseback and on foot involved in fighting and hunting, depicted amidst animals in flight or in pursuit and mythical creatures such as a sphinx and a dragon, which will

be further discussed below. The second and third registers thus contain pictorial cycles visualising the outward declarations of very ancient Iranian royal pursuits: feasting or celebration (bazm) and hunting or fighting (razm). Such motifs, representing royal pleasures and pastimes, were evidently transmitted directly from Sasanian art.31 Within the Islamic cultural sphere hunting, feasting and fighting, considered the best preparation for war, are again important themes.32 These activities were associated with heroic figures, as is the case with Rustam in the Shāh-nāma:

27 ╇ Otto-Dorn, 1978–9, fig.╯32; Baer, 1983, p.╯255, fig.╯207. A colour reproduction of the lid is given in Ward, 1993, p.╯79, fig.╯57. 28 ╇ Hartner, 1973–4, pp.╯117–23, esp.╯pp.╯121. 29 ╇ Idem, pp.╯117–23, esp.╯pp.╯121 and fig.╯17.6 (drawing of the planets on the cover of the Vaso Vescovali), followed by Baer, 1981, p.╯14, and Carboni, 1997, pp.╯22–3. 30 ╇ Cf. Pope and Ackerman, eds., 1938–9, repr. 1964–81, vol.╯6, pl.╯1308; Ettinghausen, 1943, pp.╯193–208; Rice, 1955, pls. XIX–XX; Mayer, 1959, p.╯61; Hartner, 1973–4, p.╯122, fig.╯18; The Arts of Islam, 1976, pp.╯170–2, cat. no. 180; Loukonine and Ivanov, eds., 2003, pp.╯114–5, cat. no. 116.

31 ╇ Cf. the discussion of Dorothy Shepherd (1978, p.╯119) in which she traces these twin themes to “a very early date when … they symbolized heroisation and apotheosis… [and by analogy] were seen as symbols of paradise that Islam promised to all true believers.” 32 ╇ This association was already pointed out by Xenophon in his Kynegetikos (XII.1); see Davidson, “Haft kvān,” EIr. 33 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯3, p.╯358, ll. 782–3. 34 ╇ Reading after Ettinghausen, 1943, p.╯196. Cf. MelikianChirvani, 1982, pp.╯71, 82–3, n. 61; Ward, 1993, p.╯74, fig.╯54. 35 ╇ Melikian-Chirvani, 1982, p.╯83, n. 62. 36 ╇ Cf. idem, p.╯71 and p.╯83, n. 62.

Thus did Rustam ordain a banquet For he was one that ordains banquets (bazm) And that makes war (razm). 33

The rim of the bucket is circumscribed by a silverinlaid inscription in Persian, which reads: Ordered by ʿAbd al-Raḥmān ibn ʿAbdullāh al-Rashīdī, made by Muḥammad ibn ʿAbd al-Wāḥid, worked by ḥājib Masʿūd ibn Aḥmad the decorator of Herāt, for the exalted khwāja Rukn al-Dīn, the pride of the merchants, the most trustworthy of the faithful, grace of the pilgrimage and of [its] two shrines, Rāshīd al-Dīn ʿAzīzī, ibn Abu ’l-Ḥusayn al-Zanjānī, may his glory be lasting.34

On account of the kind of eulogies offered to the dedicatee and the date recorded on the object, the month of Muḥarram, during which the pilgrimage is performed, Assadullah Souren MelikianChirvani suggests that “the gift consecrated a collective ḥajj led by the dignitary. The gift of a bucket used to contain water, perhaps for ritual ablutions, would have been appropriate.”35 Evidently, the choice of the “ruler on a dragonthrone” motif on the bucket was part of an iconographic repertoire deemed suitable for an official gift perhaps to be offered by a religious community to a dignitary, whose nisba indicates that he may have been from Zanjān in northwestern Persia.36 However, since the other blessings inscribed on the bucket are directed at an anony-

the dragon in relation to royal or heroic figures

115

mous owner, it may be surmised that such luxury objects were designed for the market, with special inscriptions only being added to the rim or handle after purchase.37 Hence, one may presume that the “ruler on a dragon-throne” roundels which were so prominently displayed on the bucket were an iconographic theme which was meaningful and would have been understood by most potential clients. The motif also occurs on eleventh- or twelfthcentury objects of adornment, for example on a small, circular copper alloy pendant inlaid with silver reportedly from Herat in presentday Afghanistan. The pendant shows a human figure, the features of his moon-shaped face and the folds of his voluminous pantaloons finely incised, seated cross-legged on a “throne” flanked on either side by upright rods terminating in snarling dragon heads, and in turn by tall vessels. Here the “throne” is decorated at the centre with a pendant palmette, flanked on either side by horizontally-oriented half-palmettes. The reverse is decorated with a long-legged bird of prey attacking a hare that turns its head towards the bird. A zigzag border, giving the impression of solar rays, frames the scene.38 Analogous iconography of the seated figure flanked symmetrically by dragon-headed staffs is employed on several other pieces of metalwork, such as a late twelfth- or early thirteenth-century copper alloy inkwell inlaid with silver also from northern Afghanistan, formerly in the Minassian Collection, New York (fig.╯115). The figure, in a flat headdress from which seem to descend long fluttering ends, appears on the base of the inkwell in the same iconographical setting, his raised arms grasping dragon-headed staffs. The latter, in turn, are symmetrically flanked here by a pair of birds, instead of by palmettes as shown in the previous example or by attendants as on the Vaso Vescovali. The roundel is again set into a radiant frame, presumably conveying solar significance.39 This equating of the ruler with the Sun goes back to the pre-Islamic concept of the

divine glory conferred upon the king. The solar aspect of ancient Iranian kingship was associated with the Zoroastrian belief in the Iranian royal and divine attribute of farr, often rendered as farr-i kayānī, “the glory of the Kays (the ancient dynasty of Iran),” the distinctive hereditary mark of Iranian legitimacy.40 This may also be reflected in the Turkic convention of altun khān (qaghan), the “golden ruler.”41 The tenth-century Khwārazmian scholar al-Bīrūnī records that on the festival of Mihragān, the Sasanian king wore a radiant crown to resemble the Zoroastrian god Mihr, lord of the sun.42 The same scene is rendered on another inlaid copper-alloy inkwell of the same period, which features a cross-legged figure crowned by a pointed headdress and flanked by dragon-headed staffs whose gaping mouths have particularly long tongues oriented towards the figure’s head. Importantly, in this example the dragons’ undulant bodies descend diagonally from the staffs and thus directly associate the depiction with the entire body of the dragon, rather than – as shown above – with the staff mutating at the tip into a dragon head.43 If the imagery should be seen as bearing some astrological meaning, the latter could also be interpreted as an abbreviated reference to the astrological “head” (raʾs) and the “tail” (dhanab) of the dragon (fig.╯116).44 A point of resemblance with the motifs discussed above (figs. 113–116) occurs on seventhcentury stamp seals which depict a crudely rendered standing ithyphallic figure with arms and legs outstretched; the arms holding two upright staffs around which serpents are wound and which are topped by stellar motifs or birds.45 Phyllis Ackerman has associated this imagery both with the astronomical sign Serpentarius and with Gayūmart (Av. gaya marətan “mortal life”), the mythical first man of the Mazdean creation myth. The latter was considered a semi-divine giant who stood between gods and men.46 A different reading is offered by Marian Wenzel who attempts to link the motif of a seated or standing figure hold-

╇ Ward, 1993, p.╯74, fig.╯54. ╇ Height 2.61 cm incl. loop, thickness 0.17 cm. Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. no. LNS 2535 J. 39 ╇ Baer, 1981, p.╯14. 40 ╇ Melikian-Chirvani, 1994, p.╯149 and n. 39. See also p.╯93, n. 80. 41 ╇ Bailey, 1967, pp.╯95–6, 99. 42 ╇ Boyce, 1983, p.╯803. 43 ╇ A sizable representation of a crowned seated person holding on to dragon staffs which are rendered with a finely

imbricated ophidian pattern, the scales enlivened by central dots, is incised on a thirteenth- or fourteenth-century steel saddle plate from Afghanistan or Eastern Iran, housed the David Collection Copenhagen, inv. no. 30/1999; Chevaux et cavaliers arabes, 2002, p.╯113, cat. no. 48. 44 ╇ Cf. Pugachenkova and Rempel’, fig.╯196 (and fig.╯197, line drawing); Hartner, 1973–4, pp.╯110–1, 114, 121; Baer, 1983, pp.╯256–7. 45 ╇ Ackerman, 1936, p.╯127; Bivar, 1969, pls. 5–6; Wenzel, 2005, ills. 10–1. 46 ╇ Ackerman, 1936, p.╯127.

37 38

116

chapter seven

ing up serpent staffs with the process of firemaking.47 Yet the figure grasping staffs entwined with serpents and topped by birds or stellar symbols could indeed represent a mythical giant holding what might perhaps represent abbreviated visual formulae of a vegetal motif characterised as the “Tree of Life” with serpent motif.48 This interpretation gains weight through a depiction on a floor mosaic discovered in the nave of a church at Huarte, located fifteen kilometres north of Apamea in present-day Syria, and dated by inscription to the years 472 or 487.49 On the central axis of the pavement close to the sanctuary is the large frontally rendered figure of Adam, identified by an inscription above his head. He is portrayed as fully clothed and enthroned on a backless raised platform, leaning against a cushion, and holding an open book in his left hand, while extending the fingers of his right hand in a gesture of speech. Importantly, he is framed by two slender cypress trees around each of which is coiled a serpent, its head oriented towards Adam’s head, the open jaws revealing bifid tongues. Adam is surrounded by a variety of tame beasts and birds set amidst flowering vegetation indicating that the setting is the primordial Paradise before the Fall (fig.╯117).50 Since the image does not accord with the book of Genesis – Adam being represented not naked but clothed, enthroned and reading a book – Henry Maguire has proposed reading this paradisical scene not in a literal but a spiritual sense as allegory.51 Precedent for such an interpretation is found in the notion of the Robe of Glory, which is commonly found in the writings of the early Syriac fathers and ultimately derives from early Jewish sources.52 The imagery is repeatedly used in the hymns of the fourthcentury Syriac writer Ephrem, deacon in Edessa and Nisibis, who equates the white robes of those baptised with the Robes of Glory (which he also

reads as Robes of Light) lost by Adam and Eve at the Fall but regained through baptism which is understood eschatologically as re-entry into Paradise.53 Hence the scene may be understood as depicting the robed Adam not only as ruler over the world54 but as an image of one who through Christ has discovered the state of divine wisdom and glory.55 In sum, it may be noted that Adam, the first of men of the biblical creation myth, is represented just like Gayūmart, the mythical first man of the Mazdean creation myth, between two slender elongated plants around which coil serpents whose heads are oriented towards the central figure. On the late fifth-century Huarte mosaic the association of the cedars in the paradisical setting with the Tree of Life is apparent, whereas on the seventh-century stamp seals the staff-like motifs may presumably represent abbreviated visual formulae; the addition of the birds or stellar symbols at the apex of these verticals give further substance to this interpretation.56 It may thus be postulated that both depictions portray a central figure flanked on either side by a serpent-entwined Tree of Life motif. Ackerman has moreover noted an astronomical association for the imagery on the stamp seals.57 The conventional medieval Islamic representation of rulership features enthroned figures flanked by dragon-headed or dragon-entwined staffs; the iconography occurring predominantly on mid-eleventh to thirteenth-century silver-inlaid metalwork produced in the greater Khurasan region (figs. 113–116). Since these are often framed by a disc with pointed rays, Eva Baer has interpreted them as artistic formulae of princely or royal images transposed in a sense to a cosmic setting, and as emblematised luminaries, or panegyrical visual expressions of the “heavenly ruler,” which sometimes have an astrological

╇ Wenzel, 2005, pp.╯140–58. ╇ For the imagery of the “Tree of Life,” see also the discussion in chapter 9, part a. It may be notable to repeat here James L. Sauvé’s view (1970, p.╯181) on this time-honored symbol: “Indo-European mythology is a pan-cultural Eurasian phenomenon, and those studies that overlook the ‘tree of life,’ or dismiss it as marginal or incidental in importance, must fail to evaluate properly, or even approximately, many aspects of some Indo-European religions and mytholoÂ�gies.” 49 ╇ Canivet, M.T. and P., 1975, pp.╯49–70; Maguire, 1987, p.╯367. 50 ╇ Canivet, M.T. and P., 1975, p.╯61 and fig.╯3; Maguire, 1987, p.╯367. For similar motifs of a tree and serpent with an enthroned ruler or king, see Fodor, 1976, pp.╯159–60. 51 ╇ Maguire, 1987, p.╯368.

52 ╇ Ephraem, Hymni de Paradiso and Section 2 of Commnetarium in Genesium, tr. and ed. Brook, 1990, pp.╯31, 67, 226–7. 53 ╇ Idem, pp.╯31, 66–74, 94, 112, 226–7. 54 ╇ “God said…have dominion…over every living thing that moveth upon the earth…” (Genesis 1:28). In the Christian Syriac Meʿārath gazzē 2.11–25 (Cave of Treasures, tr. and ed. Budge, 1927, p.╯34) God gave Adam power over all beings. 55 ╇ Maguire, 1987, p.╯368, n. 23, and p.╯372, ns. 34, 38. In post-exilic Judaism, Midrash and Judeaeo-Christian tradition the figure of the enthroned Adam is seen as the source of supernatural wisdom, see Fodor, 1976, pp.╯166–7. 56 ╇ Of note may be the parallel given by the Greek poet Pindar of the drakōn Typhon who supports the column of the Etna on which perches the eagle of Zeus (Pythian Odes 1). 57 ╇ Ackerman, 1936, p.╯127.

47 48

the dragon in relation to royal or heroic figures character.58 An anecdote attesting to the astral qualities of the ruler in the face of the dragon is noted by Ibn Bībī, the Rūm Saljuq chronicler, in which he metaphorically likens the meeting of the armies of the Khwārazm-shāh and the Rūm Saljuqs to the day on which: … in accordance to servile custom the King of the Planets kissed the earth at the door of the King of the World and the encounter of the troops of the War-Dragon unfurled the crimson and yellow banner at the arena of the horizons.59

Luminary imagery is similarly used to describe the birth of Khurshīd shāh in one of the earliest Persian popular romances, the Kitāb-i Samak ʿAyyār, who throughout the story is portrayed in terms of solar metaphors and is even named “king Sun.”60 In sum, it may be stated that the visual tradition of the medieval Islamic representation of rulership involves the imagery of a central figure flanked by serpent-dragons that top or twine around verticals which might carry a vegetal association that could distantly associate them with the Tree of Life motif as ultimate prototype. Frequently the vertical is shown only with a projecting dragon head. This may thus present yet another example of the visual conflation that was so pervasive in the medieval Central Asian world and beyond. It would be an example of a visual development that merges the figure of the dragon with that of the staff in a unified whole. The Tree of Life motif exhibited a remarkable persistence and popularity in Near Eastern and Indo-European61 iconography which, as has been discussed above, could still be gauged in visual and textual sources of the eleventh to thirteenth century, the period when this imagery was employed with great regularity.62 This supports the hypothesis that the prototype of the vertical may be sought in a vegetal composition. It is moreover notable that the central figure often grasps with one or two hands the verticals just below the point from which the dragon heads project. To this may be added the luminary asso58 ╇ Baer, 1981, p.╯14, and eadem, 1983, pp.╯259–68, 274. See also the discussion on the astral transformation of the Sasanian throne in L’Orange, 1953, pp.╯37–47. 59 ╇ Al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, p.╯170. 60 ╇ Gaillard, 1987, pp.╯124–5. 61 ╇ Cf. Janda, 2010, pp.╯320–2. 62 ╇ See also the close association of the tree with the Iranian king in the Shāh-nāma account, for instance, the simile of the drying leaves of the tree of the Kayānids employed to describe the death of Farīdūn (tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯209, ll. 1128–129) or of the noble tree that stretches

117

ciation which transposes the motif to a cosmic setting – sometimes with an astrological character. The pivotal role the dragon plays in these images will be further elucidated in the following, concluding in a discussion of the time-honoured symbolism of the dragon-tamer. Issuing and devouring The iconography of a human figure, sometimes reduced to the image of a mask-like face, flanked on either side by dragons, was one of “great symbolic potency” in the Central Asian world and beyond and existed in manifold variations.63 The monsters not only frame the central motif but on account of their gaping jaws at first sight appear to threaten or perhaps even attempt to devour what they flank. The possibility of the latter interpretation will be discussed in the light of surviving examples of this enigmatic imagery. The motif is prominently recorded in the visual arts of the Caucasus region, for instance on the Armenian cathedral of the Holy Apostles in Â�present-day Kars erected by king Abas I between 930 and 943 in the adjacent secondary Bagratid kingdom of Vanand, near the border with eastern Anatolia. On the exterior of the drum above the blind arcades twelve full-length figures are carved in low relief that in folk tradition are associated with the twelve apostles. The figures have masklike faces and are clad in long robes. In what is perhaps a unique occurrence, some are shown to raise their bent arms upwards, others place their hands on their chests, in what may have represented ritually significant gestures and may have conveyed a canonical koiné.64 It is conceivable that the sequence of the movements played a role and even alluded to a visual narrative.65 One of the figures is flanked by a tree-like vegetal composition.66 Next to it is a figure with a looped serpent rising up at either side of his upper body, the heads oriented towards his ears. Significantly, the figure does not extend his arms towards the serpents but far into the shadow used to depict the death of Iskandar (idem, vol.╯4, p.╯255, l. 1849). 63 ╇ Cf. Russell, 2004, p.╯1169. 64 ╇ The folding of the arms was a deferential gesture known in the Iranised world since Zoroastrianism and was used by Christians in prayer towards God; see Russell, 2004, p.╯310. Cf. Colledge, 1986, p.╯16. 65 ╇ For a line drawing of all twelve figures, see Thierry, 1978, p.╯47, fig.╯26. 66 ╇ Idem, pl.╯V.3.

118

chapter seven

has the hands placed, one above the other, on his chest (fig.╯118).67 The repetition of the motif above the northwestern port-hole window (“oculus”) of the church underlines the importance accorded to this iconography. Here, instead of a full-length figure a frontally presented human bust flanked by giant looped serpents is portrayed (fig.╯119).68 Several interpretations have been proposed for the enigmatic reliefs. The portrayals are particularly puzzling since as decoration on a church façade they can be assumed to represent overall propitious and apotropaic motifs. Yet the readings of the figure flanked by serpents, in particular, have only multiplied the mystery. One explanation identifies the figure as Saint Gregory the Illuminator during his imprisonment in the snake-infested Khor Virap (“deep dungeon”).69 Another tentative suggestion is that the representation preserves an echo of the ancient myth of the Armenian dragon-fighter Hruden (the Avestan Thraētaona) who chained Azhi Dahāka.70 On the cathedral of the Holy Apostles in Kars, the figure is shown flanked by two serpents. An iconographical schema involving two serpents that grow out of the shoulders of a human figure was perhaps still known in medieval Armenia,

where even after Christianisation vestiges of the Zoroastrian faith – in which this was a wellknown form of imagery – continued to survive. In later Zoroastrian Pahlawī/Middle Persian sources (which were written down in the ninth century but incorporated material from third century ad writings and before), the formidable Avestan dragon Azhi Dahāka is transformed into an early Iranian historicised foreign king, Azhdahāk, or, as he is named in New Persian or Arabic narratives, Ẓaḥḥāk (al-Ḍaḥḥāk/Dahāk). According to an account in the Shāh-nāma, the dragon-king Ẓaḥḥāk was not originally evil but when Ahriman (Av. Angra Mainyu), the demonic half of ancient Zoroastrian dualist myth, tempted him, kissing his shoulders, a pair of serpents sprang from the place the Spirit of Evil had touched. These serpents, a reminiscence of Ẓaḥḥāk’s original reptilian nature, had to be fed the brains of young men every day, binding him to the perpetual sacrifice of humans, who eventually rebelled and overthrew him.71 An early example of the motif of a human figure with snakes growing out of each shoulder may be sought in greater Mesopotamia, where it is expressed in the figure of the chthonian god Nergal72 whose iconography may per-

67 ╇ Idem, pl.╯V.1 and 1987, p.╯154, pl.╯62; Russell, 2004, pp.╯1167–168, and p.╯1180, pl.╯2. See also the depiction of a stylised figure with hands similarly placed, one above the other, on the body and flanked by a pair of crosses on a relief from Samtavro (probably twelfth century), recorded by Baltrušaitis, 1929, pl.╯LXXXV, fig.╯144. 68 ╇ Jean Michel Thierry (1978, p.╯54, fig.╯30 (line drawing), pl.╯VIII, 2) identifies also the bust as that of Saint Gregory. See also Gierlichs, 1996, p.╯96. 69 ╇ Russell, 2004, pp.╯1168–9 and ns. 3 and 4, pp.╯1178, 1288. Russell (2004, p.╯631) tentatively perceives the figure to be Judas Iscariot fused with Azhdahāk and suggests that these figures are apotropaic symbols of evil. Another interpretation is given by Thierry (1978, p.╯49 and idem, 1987, pp.╯154, 544) who sees the figure as part of an ascension theme common in Byzantine art representing the Virgin; the paired serpents are explained as an artistic mishap which occurred when the artist copied the Byzantine model and mistakenly interpreted the wavy edges of the Virgin’s maphorion as serpents. 70 ╇ Russell, 2004, p.╯560 and n. 21. 71 ╇ Ṭabarī describes these as excrescences that resembled the heads of serpents which to Ẓaḥḥāk seemed like dragon heads each time he was taking off his clothes. The monstrous outgrowths were extremely painful but the application of brains appears to have assuaged the pain. See Balʿamī’s Persian translation of Ṭabarī’s History; Tarjumat-i tārīkh-i Ṭabarī, tr. Zotenberg, vol.╯1, pp.╯115–7; the story is recorded in greater detail in al-Thaʿālibī, Taʾrīkh Ghurar al-siyar, tr. and ed. Zotenberg, 1900, pp.╯19–27. In a manuscript of the Shāh-nāma (fol. 8r) featuring the enthronement scene of Ẓaḥḥāk, the tyrant is depicted with a pair of “Saljuq-style” serpent protomes springing from his shoulders in bilaterally symmetrical fashion whose gaping mouths with long curv-

ing snouts are turned towards his head. Copied in Shiraz, dated 752/1352. Present location unknown. Ferrier, ed., 1990, p.╯202, fig.╯7. An earlier known occurrence of the same scene in a miniature from the Great Mongol Shāh-nāma (“Demotte”; Washington, DC, Freer Gallery of Art, Ms. 23, 5), probably copied in Tabriz and often dated to around 735/1335, shows Ẓaḥḥāk with undulant serpents in asymmetrical arrangement featured with elongated, wrinkled and fleshy snouts, slightly agape, which exhibit more East and Central Asian characteristics. Grabar and Blair, 1980, p.╯59. 72 ╇ It is possible that the motif with the serpents was inspired by sculptural images of the Semitic Underworld God, Heracles-Nergal-Ahriman, as depicted on the bas-relief from a small house-temple in Parthian Hatra in northern Mesopotamia (which was an integral part of Iran in Parthian and Sasanian times). The composite figure shows the god of the realm of death and the underworld, who can be at once life- and death-giving (see Dhorme, 1949, pp.╯40–3, 51), clad in Parthian garb. The god’s attribute is the serpent, a pair of which springs from his shoulders and rise from either side of his waist, while another serpent rests at his feet. Ghirshman, 1962, p.╯87, fig.╯98; Bivar, 1975a, vol.╯2, pl.╯4a. See also pp.╯36–7. Cf. Drijvers, 1978, p.╯172. Comparable characteristics are likewise shared by the Palmyrene healing warrior god Shadrafa, distinguished by serpents and scorpions springing from his shoulders. On a beam from the peristylium of Bēl’s temple at Palmyra (dated c. 32 ad) Shadrafa is portrayed as one of the gods on foot, horseback and chariot who fight a snake-tailed female monster; significantly the combat scene is surmounted by a pair of winged creatures with human torso and serpentine coils as legs that probably represent beneficial beings. Seyrig, 1934, pp.╯165–8, pl.╯20. See also Drijvers, 1978, pp.╯176–7; Bonner, 1950, pp.╯124–5.

the dragon in relation to royal or heroic figures

119

haps have been appropriated for the depiction of Yima, the Indo-Iranian “first man,”73 who in the Iranian tradition becomes the ruler of the underworld after his death,74 as well as in much earlier Mesopotamian75 and Bronze Age Central Asian deities.76 It also appears in archaic Greek literature as shown by Hesiod’s anthropomorphic drakōn Typhaon/Typhon who is described as having a hundred snake heads growing from his shoulders (Theogony 825–626). Material evidence of this visual expression is found mainly in Western Central Asia and the Caucasus.77 A large terracotta figure, probably made in seventhor eighth-century Sogdia, thus before Islam had become firmly entrenched as the principal faith in the region, shows an enthroned crowned man, large-headed and with grinning mouth, from the base of whose neck a pair of serpents grow, curving upwards and around his ears before descending down to his cheeks. The man’s left hand is clasped to his chest while his raised right hand clenches a now lost object, perhaps a staff.78 As James Russell notes, it is unlikely that the figure represented an epic monster. Rather is it a supernatural figure that probably fulfilled an apotropaic function.79 The symbolism of an anthropomorphic figure with serpents springing from the shoulders is thus characterised by an element of ambiguity that allows for a multilayered interpretation.80 Russell remarks upon this ambiguity when he discusses the role of Yima about whom “Zoroastrian tradition preserves two separate narratives ...; in

one he dies and goes down to live in a happy underground abode, while in the other version, he commits sin, wanders unhappy and dies [Shāhnāma].”81 He therefore proposes that the motif of “Azhi Dahāka in the epic is contaminated by an image of Yima [the Jamshīd of the much later Shāh-nāma] appropriated from Nergal ([related by inference to] Ẓaḥḥāk [who] succeeds Jamshīd in the Shāh-nāma).”82 It is interesting that a closely related motif was chosen in the anonymous fifth-century Armenian Buzandaran Patmutʿiwnkʿ which portrays the unfortunate Armenian king Pap (369–374) as having serpents that sprang from his breasts (though not from his shoulders) and wove themselves around his shoulders (IV.44).83 According to the historian this was because the king was possessed by demons. The choice of symbolism may also be associated with the fact that the king, who was later assassinated by the Romans, had antagonised the Christian clergy and, in addition, had been accused of homosexuality.84 The description of king Pap with serpents may perhaps reflect the reformulation of the beneficial aspect of the dragon when it acquired an overall symbolic meaning as a satanic force in Christian imagery.85 The inversion of the beneficial association is graphically articulated in the later socalled “revenge miniatures” of the eleventh- or early twelfth-century Byzantine Metaphrastian Menologion volumes, in which persecuting pagan

73 ╇Yasht 13.130; Christensen, 1931, tr. 1993, p.╯15; Zaehner, 1961, p.╯134. 74 ╇ Russell, 1987, p.╯44. 75 ╇ Nergal’s iconographic characteristics link him with the ancient Sumerian Mesopotamian chthonic vegetation and healing god, Ningizzida, whose attribute is the horned dragon; when represented in anthropomorphic form two serpent heads grow from the god’s shoulders. Cf. Edzard, “Ningizzida,”WdM I, pp.╯112–3; Drijvers, 1978, pp.╯151–86, esp.╯pp.╯171–80. 76 ╇ See Aruz, 1998, figs. 3c and 3e; Francfort, 2002, p.╯132, fig.╯28; Kuehn, 2009, pp. 43–67. 77 ╇ It is interesting to observe further that the symbolic concept of serpents growing from Dahāk’s shoulders have a point of resemblance with the motif of the serpent crest that generally issues from the point of juncture between the neck and the shoulders of the anthropomorphic representation of the Indian nāgas that were frequently found in the Buddhist material culture of Central Asia; cf. Vogel, 1926, p.╯40. 78 ╇ Height 61.5 cm. The State Hermitage Museum, inv. no. GA 3053. D’yakonova, 1940, pp.╯195–200, fig.╯1. See also Russell, 1987, p.╯441. Grand Exhibition of Silkroad Buddhist Art, 1996, p. 37, cat. no. 22. The bust of a man with snakes growing from his shoulders is also featured on the wall paintings of Sogdian Panjikent; Belenitskii, 1980, p.╯203. 79 ╇ Russell, 1987, p.╯44. 80 ╇ In this context it is interesting to consider a note by al-

Qazwīnī in Āthār al-bilād wa-akhbār al-ʿibād (“Monuments of the Countries and History of their Inhabitants”), 437, 9–10, which states that serpents grew from the shoulders of the ṣūfī Sheikh al-Kammūnī (from whom al-Qazwīnī traces direct descent through five generations) in order to overturn the power of an unjust ruler. See von Hees, 2002, pp.╯36–7, 45. 81 ╇ Russell, 1987, p.╯44. See Zaehner, 1961, pp.╯134–7, esp.╯p.╯140; Boyce, 1975, repr. 1996, pp.╯92–5 and n. 69. 82 ╇ Russell, 1987, p.╯44. 83 ╇ Buzandaran Patmutʿiwnkʿ, tr. Garsoïan, 1989, p.╯202; Russell, 2004, p.╯62, n. 36, and p.╯130. 84 ╇ According to the Buzandaran, once when the king was young his mother entered his room [whilst he was engaged in sodomy] and saw that white snakes had twisted themselves around him (V.22); tr. Garsoïan, 1989, p.╯165. Cf. Â�Russell, 2004, pp.╯341–2. 85 ╇ In the Christian apocalypse, Satan as the Devil is called the “great dragon” and “ancient serpent” (Revelation 12.9 and 20.2). The use of the serpent as a symbol of Evil is exemplified, for instance, by a homily on the Armenian martyr Saint Sergios by Severus of Antioch, delivered at Chalcis on 1 October 514: “We must be watchful against Satan, the snake who with sleepless eye fixed on our heels lies waiting to push us into the pit of sin by our love for pleasure, such as stuffing one’s belly,” as cited in Fowden, 1999, p.╯23.

120

chapter seven

tyrant rulers who had martyred Christian saints were depicted with a serpent as a symbol of evil.86 What appears to be an abbreviated formula or contracted version of the motif of the fulllength human figure or bust flanked by two serpents as shown on the tenth-century cathedral of the Holy Apostles in Kars (figs. 118 and 119), is found three times on the façade of the central monument of the fortified monastic complex of Tatʿev, located in Siunikʿ province in southeastern Armenia. The church, dedicated to the Saints Paul and Peter (Surb Poghos Petros), was constructed by the order of prince Ashotʿ of Siunikʿ under the supervision of bishop Ter Hohannes between 895 and 906.87 Erected nearly half a century earlier than the imagery on the cathedral of the Holy Apostles in Kars, three frontally rendered masklike heads are framed on either side by horizontally oriented dragons whose bodies form a single loop.╯ On the eastern façade of the Paul and Peter church the human heads with prominent ears88 are flanked by long slender pairs of dragons, while on the northern façade a crowned moustachioed head is framed by two shorter and more voluminous dragons. All of the creatures are portrayed with mouths ajar, revealing a row of sharp teeth as well as projecting tongues, which touch the ears of the heads.89 The Armenian historian Stepʿanos Ōrbēlian has interpreted the heads as belonging to the donors of the church; the mustachioed head on the north façade, Grigor Supan, the ruler of Gegharkʿunikʿ (fig.╯120), as well as the two heads on the east façade, prince Ashotʿ (fig.╯121) and his wife Shushan (fig.╯122).90 The head identified as that of princess Shushan is framed by shoul-

der-length tresses. Her head is surmounted by a large composite star-rosette that has an elaborate centre, from which radiates one row of cusped petals, which may have conveyed a special significance, possibly with astral or royal connotations.91 Khalʿpakhʿchian has suggested a protective function of the motif of the human head between dragons, since the latter are regarded by Armenians as protectors of homes (shahapets), serpent genii of places.92 The widespread use of the symbol of openmouthed dragons flanking a human head is also attested by its use in twelfth-century Armenian manuscripts. It is shown as an element attached to ornate initials, such as in the loop of the initial in the Cilician Gospels of Luke and of Matthew (Baltimore, Walters Art Gallery Ms. 538), written in 1193 for the Bishop Ter Karapet, at the monastery of Paughoskan near the fortress of Katen in the region of Mlich (fig.╯123).93 Here the masklike human head is bracketed by dragon heads, which, like those flanking the heads at Tatʿev, are similarly portrayed with open jaws ending in prominently curled-up tips. The theme of symmetrically doubled dragons flanking a human head appears on other threedimensional constructions recorded in the Islamic world, such as at the thirteenth-century caravanserai Susuz Han, dated c. 644/1246, situated near Bucak, examined above. At the khān the symbolism is shown twice, mirrored above two muqarnas niches flanking the main portal, effectively once again doubling the visual impact of the potent motif (fig.╯124, detail of one muqarnas niche). Just like the above-discussed Armenian dragon heads, those at Susuz Han are portrayed in the so-called “Saljuq-style.”

86 ╇ When a first word of a Passion was the name of an emperor who had martyred a saint, the evil ruler might be depicted in the initial letter as being strangled by a serpent (Patterson Ševčenko, 1990, p.╯193), emperors or crowned figures may be encircled by a serpent (eadem, p.╯90), or, as in the case of the Roman emperor Trajan, portrayed as holding a serpent in his hand (eadem, p.╯155). 87 ╇ Cf. Garsoïan, 1994, p.╯11. 88 ╇ It is interesting to note that comparable mask-like heads, similarly cut off immediately above the neck, appear on several Armenian sacred buildings dating from the tenth to the thirteenth century. They appear for instance among the vine branches at the Armenian church of the Holy Cross at Aghtʿamar, a few are included among the animal friezes, and two continuous rows of mask-like human female and male heads appear under the eaves on the north and south walls of the west exedra of the church (Der Nersessian, 1965, p.╯26, figs. 16, 40). Sirarpie Der Â�Nersessian has identified these schematic heads as follow-

ing the tradition of Iranian motifs seen to decorate the Parthian palace of Hatra in the Mesopotamian desert (cf. Ghirshman, 1962, p.╯36, fig.╯49), which were also discovered at Hamadan (ancient Ecbatana) and at Qum south of Tehran (cf. idem, p.╯38, fig.╯52), which originally had an apotropaic character but became an ornamental motif in the classical period (Der Nersessian, 1965, p.╯26). 89 ╇ Khalʿpakhʿchian, 1980, pp.╯173–4, fig.╯6 (line drawing of three reliefs), fig.╯8 (forth relief). Russell, 2004, p.╯560 and n. 21, p.╯631, pp.╯1168, 1181, pl.╯3. 90 ╇ As cited in Khalʿpakhʿchian, 1980, p.╯164. 91 ╇ For a discussion of the significance of the rosette motif in the early Persian, Sasanian and Umayyad period, see Ettinghausen, 1972, pp.╯36–41. 92 ╇ Khalʿpakhʿchian, 1980, p.╯164. On the Armenian shahapets, see p. 56, n. 66 93 ╇ Der Nersessian, 1973, pp.╯6–9, 85–6, pl.╯C and figs. 13–29; for the ornate initial of the Gospel of Matthew, see fig.╯34.

the dragon in relation to royal or heroic figures

121

Paired dragons flanking a central motif are also featured on other architectural elements such as a thirteenth-century carved wooden door, discussed below, probably from the Tigris region, now preserved in the Museum für Islamische Kunst in Berlin. The gaping “Saljuq-style” dragons’ mouths similarly flank a small central medallion that contains a heavily abraded motif, perhaps once representing a human bust (fig.╯153). Instead of a human head as central element held by open-mouthed dragons, the mythical creatures similarly flank an inscription, a vegetal composition or an animal head, for instance that of a bull or a lion. The close link between the blessings expressed in an inscription and images of fructifying vegetation is evidenced in parallel depictions on the handles of the previously mentioned Bobrinski and Fould inlaid copper alloy buckets. The handle loops of the former are made in the form of dragon protomes from whose gaping mouths spring the central foursided arched section. Benedictory inscriptions appear in naskhī on two sides and in Kufic on the top band of the handle with the date muḥarram 559/December 1163. On the outside the loops are flanked by leaping lions (fig.╯56). The handle loops on the Fould bucket are made of arched dragons topped at the outside by projecting lionheaded knobs. From the dragons’ open mouths emerges the six-sided arched handle, decorated on all facets with scrolling foliage in place of the benedictory inscriptions shown on the Bobrinsky handle (fig.╯57). In a similar manner an inverted pendant palmette is flanked by open-mouthed dragon heads inserted in the loop of the initial letter on the first page of a twelfth-century Armenian Gospel of Matthew (fig.╯125). Further related imagery is shown in a marginal ornament portrayed in the important Mush Homiliary, discussed earlier, transcribed and illuminated at the monastery of Avagvankʿ near Erznga(n) in northeastern Anatolia in 1200 to 1202.94 In this illumination the excessively long looped ophidian bodies of two dragons form a pretzel-like knot at mid-section (fig.╯126).95 Their wide-open mouths revealing sharp teeth are oriented upwards, and in place

of the tongues, the creatures emit vegetal scrolls bearing split palmettes and buds. The dragons’ mouths thus seem to emit the foliage. This forms an interlaced composition surmounting the pretzel-knot which in turn intertwines with the ophidian bodies. The same manuscript yields further initial letters and marginal ornaments with variations of the same theme. An example of the former are two dragons in mirror image springing from vegetation and touching the foliage with the tips of their outstretched tongues (fig.╯127) and of the latter, a closely-related imagery showing the beasts touching each other’s tongue tips (fig.╯128); correspondingly, a scrolling vegetal stalk bearing palmettes and split-palmettes is shown with a single dragon head growing out of the foliage and biting into the stalk at mid-section (fig.╯129). Yet another analogous motif is represented by the previously discussed relief sculptures on the stone minaret buttresses of the mid-thirteenthcentury Çifte Minare madrasa at Erzurum. However here a pair of dragon protomes issues from the stem of a palm tree in such a way that they come to flank the central vegetation with its small birds and the double-headed eagle at its apex (fig.╯43). As has been shown above, in lieu of an inscription or vegetation dragons with gaping mouths flank Christian crosses on architectural representations such as on the vishap-type khatchkʿar (fig.╯46) or on the lintel of a door at the Monastery of Deir Mār Behnām in Mosul (figs. 17 a and b, 50) as well as on ornaments in Cilician Armenian Gospel books (figs. 45 and 47). Hence it appears less likely that the dragons framing the Christian emblems are intended to be read as threatening or even less as ingesting the crosses. It may rather be presumed that, conversely, as in the case of vegetation or benedictory inscriptions, the mythical creatures are depicted with crosses issuing from their mouths, perhaps suggesting the active association of the dragons with this symbol of spiritual deliverance. This would suggest a role for them not only as guardians but one may cautiously hypothesise perhaps even as rescuers or deliverers, drawing on associations with the idea of Christ the Saviour and Deliverer.

94 ╇ Cf. Izmailova, T.A., “Edesskaia rukopis 1171 goda (M. 313),” Kultura i iskusstvo nardo vostoka 8, Trudi gosudarstvennogo Orgena Lenina Ermitazha, 19, Leningrad, 1978, pp.╯98–9, fig.╯14, as cited in Der Nersessian and Agemian, 1993, p.╯31. The interlace has close analogies to a vishaptype khatchkʿar from Yovhannavankʿ, Ararat, dated 1171, on which the pair of stylised dragons form a central interlace

and their open snouts with curled tips emit what appears to be curving foliage. See Thierry, 1987, p.╯408, fig.╯362. 95 ╇ For a closely related marginal ornament, see the Gospel book, copied on paper, 275 fols., dated 1331, transcribed and illuminated in the village of Agheth (Bzhnunikʿ) by the scribe Hovhannes Kʿahanay (the priest). Mnatsakanyan, 1955, p.╯516, fig.╯1021.

122

chapter seven

By analogy, should the imagery of dragons flanking and apparently emitting crosses carry a positive symbolism, then this would have implications for other motifs in the same position. The closely related theme of symmetrically doubled dragons flanking a central animal head survives on architectural compositions. It appears on two of the round towers of the northern city wall of the medieval Armenian city of Ani, probably added under Shaddādid rule in the early twelfth century. Here the creatures’ heads with gaping jaws, revealing rows of teeth and tongues with bifid tips, frame a bovine head that in one instance holds a ring in its mouth (fig.╯130). The dragons’ necks carry ornamental bands as collars. The prominent depiction of the motif on the towers of the city walls certainly underlines its apotropaic intent. It may be surmised that the bovine head flanked by dragon mouths is therefore not threatened by the devouring aspect of the dragons, but, contrarily, is presumably guarded and protected. The Mush Homiliary yields further variations of the lion’s mask with paired dragon heads growing from or issuing vegetation. The mythical creatures are shown joined to the mask with the help of a knotted vegetal interlace bearing split-palmettes and buds (fig.╯131). This is also modified to include paired human heads. A marginal ornament shows the central lion mask to emit two leaves from its mouth which rise along either side reaching the mask’s crown and issue two human heads in profile (fig.╯132).96 It is notable that the human profiles, drawn in clear outline with large almondshaped eyes, bulbous nose and receding chin, are closely comparable to the head of a sphinx marÂ� chant on a star tile from the now destroyed small palace at Kubadabad (built three decades after the inception of the paintings in the Homiliary), which is rendered with a similar tight-fitting headdress with pointed tip (fig.╯65). From the apex of the leonine mask spring further vegetal stalks bearing palmettes and foliage which terminate at the top in two confronted open-mouthed dragon heads whose projecting tongues touch the Â�central stalk. In the same manuscript the theme of paired dragons cum lion’s mask can be observed. Significantly, this imagery takes the characteristic form of a pomegranate, a fruit which for its abundant seeds symbolises fertility and immortality both in

Armenia and throughout the Iranian world, thus underlining, once again, the auspicious character of the motif (fig.╯133).97 A closely comparable visual formulation is found in the drawing of the brass door handles of al-Jazarī, intended for the palace door at Diyārbakr in southern Anatolia, which survives only in a copy of the original illustration and a description in the treatise on automata written by the master craftsman. In the illustration the knockers are shown in the form of two confronted dragons with gaping mouths and outstretched tongues that frame the lion-headed knob (fig.╯134). The popularity and wide distribution of the theme of the lion’s mask with paired dragon heads is further underlined by its representation on architectural monuments such as the fine relief embellishing the upper corners just below the roof of the thirteenth-century church of Surb Karapet (John the Baptist) (1221–1227) on the east side of the monastic complex of Noravankʿ in Vayots Dzor in southern Armenia (fig.╯135). The complexity of the theme of confronted dragon heads is reflected in a zoomorphic composition at the monastic complex of Geghard, also known as Ayrivankʿ, in the mountainous region of Kotaykʿ, Armenia, that perhaps represents the coat of arms of the Proshian family.98 The large high relief covers most of the northern wall above the archways leading to the rock carved family sepulcher of the Proshians. The funerary chamber to the northeast of the gavit (narthex) was constructed in 1283 and houses the remains of Khaghbakian Prosh, a vassal of the Zakʿarid dynasty who had sold the monastery to the Prosh family. A stairway west of the gavit leads up to a burial chamber of his son Papak and his wife Ruzukan which according to an inscription on one of the central columns was hewn out in 1288. The plastically rendered relief sculptures show a monumental bovine head at the apex holding a large ring in its mouth which is fastened to chains that are attached to the collars of symmetrically doubled confronted lions marchant. The long curving tails of the imposing felines evolve at the tips into large upward-looking dragon heads. Rendered with gaping snouts terminating in tightly curled tips and revealing large fangs and projecting tongues as well as ornamental bands form-

96 ╇ Cf. Durnovo, Sargsian, and Mnatsakanyan, 1978, pl.╯31 (lower right corner). 97 ╇ See also pp. 65, 71, n. 208. On the pomegranate motif

in Greek mythology, see Elderkin, 1924, pp.╯1–3, 18, 25–7, 44–5, 118. 98 ╇ Khalʿpakhʿchian, 1980, p.╯325 and figs. 18 and 20.

the dragon in relation to royal or heroic figures ing collars around the necks, the dragon heads diagonally flank the bovine head from below in a triangular fashion. The composition thus offers a parallel to the two relief sculptures with paired dragons framing a bovine head (which in one instance also holds a ring in its mouth) that circumscribe two of the towers of the northern city wall at Ani (fig.╯130). At Geghard the sculptures are augmented further just below the lions’ frontally oriented heads by the sculpture of a commanding eagle with outspread wings whose claws grasp an ungulate – most probably a symbol of the princes’ power (fig.╯136). An elaboration of the motif of dragon heads flanking a central mask-like human face is featured on the türbe of Hüdavend Hatun in Niğde. The monument, with a sixteen-sided pyramidal roof on an octagonal body, is of cut stone and was built in 712/1312, during the lifetime of the princess who was buried there twenty years later, and during the rule of the Ilkhanid governor Sunqur Ağa. The motif is amalgamated with the relief of a frontally presented double-headed bird of prey, almost certainly symbolising an eagle, which fills one of the carved blind niches that decorate the façade of the octagonal building. The tips of the outspread wings of the eagle extend into upward-curving dragon heads with long, upwardcurved snouts and projecting tongues. A masklike frontally rendered human head springs from between the elongated interlaced birds’ necks that form a loop at the base, while the addorsed eagle heads peck at their own outspread wings (fig.╯137). As was shown earlier in chapter 5, animals can sometimes be seen merging visually with one another, thus fusing not only bodily but presumably also in terms of their innate characteristics. The motif here shows a fusion of the eagle with the dragon, both of which are rendered symmetrically with doubled heads and necks, to which the human aspect is added in the form of the central frontally represented face, resulting in a therianthropomorphic hybrid. The eagle heads are thus shown hacking with their sharp beaks at their own wings since these are at the same time an extension of the “bodies” of the dragons that spring from the wing tips. Moreover, this composite imagery offers an interesting parallel to that of the mask-like face flanked by dragon heads with wide-open mouths that appears on the 99

╇ Cf. Badiee, 1984, p.╯97.

123

above-mentioned thirteenth-century caravanserai Susuz Han near Bucak (figs. 7 and 124). However, at the türbe of Hüdavend Hatun the dragons appear visually merged with a double-headed eagle, the bifid tongues thrust upwards from their gaping jaws towards the eagle’s heads and then towards the human head at the apex. It may be presumed that just as in the case of the conceptual doubling of representations mentioned earlier, a device prominent throughout the medieval period and seen here in the double-headed eagle and dragon, the composite hybrid portrayed here on the türbe is similarly an example of an intention to reinforce and augment the visual impact and intended effect of the potent symbol. The motif shown on the türbe may thus be seen as a visually amplified version of that on the caravanserai Susuz Han. It is further noteworthy that the pictorial amalgamation of the dragon and the bird, as shown in the present example, is also illustrated for instance on the presumably royal Saljuq silk fragment, now preserved in the shrine of Saint Apollinaris in Siegburg (fig.╯69). The representation of a closely related motif on the türbe of Hüdavend Hatun, the daughter of the Rūm Saljuq sulṭān Rukn al-Dīn Qılıch Arslan IV, underlines its royal connotations. The examples above moreover show that even during Ilkhanid rule the iconography of the dragon, stylistically and presumably also iconologically, continued to follow “Saljuq-style” conventions. A similar conceptualisation governs a later illustration in a copy of Zakariyyāʾ ibn Muḥammad al-Qazwīnī’s cosmography ʿAjāʾib al-makhlūqāt wa gharāʾib al-mawjūdāt (“Wonders of Creation and Oddities of Existence”), written around 668/1270. It is represented in the Sarre Qazwīnī, now in the Freer Gallery of Art, Washington, DC, the suggested dates for which have varied from 1350 to the early fifteenth century.99 It is considered “a last living example of the long tradition of manuscript painting associated with the local iconographical tradition of Diyarbakir.”100 In the light of this attribution it may be interesting to take a closer look at what is a comparatively late representation of the motif. It shows a gigantic polycephalic scaly dragon (al-tannīn) with a large pretzel-like knot incorporating a small, mask-like human head. Two of the six projecting dragon heads, their confronted gaping jaws revealing pro╇ Eadem, 1978, p.╯406, pl.╯42, and eadem, 1984, p.╯103.

100

124

chapter seven

jecting tongues, encircle and surmount the human head. The small head thus appears almost to grow out of the dragon’s body, while being flanked at the same time by four other dragon heads that project from just below (fig.╯138). It is thus literally engulfed by the dragon heads and seems almost to fuse with the body of the giant polycephalic dragon. The text to the illustration states that the dragon was so large that it snapped off the top third of the minaret of Anṭākiya (Antioch).101 With regard to the paired reliefs at Susuz Han, Öney has identified the human heads as sun rosettes threatened by the “underground forces and the dark moon symbol” of the dragon,102 hence associating it with astrological functions, according to which the dragon is the cause of solar and lunar eclipses. It is important to note however, as Abbas Daneshvari has demonstrated, that the concept of the dragon solely in its role as eclipse monster threatening the light of the luminaries and, by extension, the rulers as their worldly embodiments, presents only one aspect of the multivalent symbolism of the dragon.103 As has been shown above there exists at the same time another possibility: that of perceiving the symbolism of the gaping dragons’ jaws flanking a central motif as beneficial and apotropaic. At Susuz Han this reading is supported by paired winged figures whose presence seems to bestow a honorific dimension upon the iconography of the gaping dragons’ jaws flanking the mask-like human faces (figs. 7 and 124). It thus appears reasonable to assume that the iconographic theme of dragons’ jaws flanking other central motifs, such as inscriptions, vegetation or animal heads may be similarly associated with a beneficial, apotropaic function.

The representation, examined above, of the ruler transposed to a cosmic plane flanked on either side by dragons, survived mainly on portable objects (figs. 113–116), in particular on silverinlaid metalwork from the greater Khurasan region, from where this visual expression spread westward. It is significant in this respect that the

central figure is portrayed with one or both of his arms extended to hold the staffs just below the dragon heads, in other words as if grasping the necks of the dragons. The fact that sometimes the figure does not hold the creatures’ necks with both hands may suggest that the actual holding of the necks is not a prerequisite. For the imagery to convey its iconological content to the contemporary beholder – namely the ruler as victor over mythical creatures which symbolise any and all hostile forces – it apparently sufficed to represent the key elements of a central figure flanked by two dragons. In its most detailed and perhaps most complete execution, this important iconographic expression is emblematised on an architectural structure. In a powerful parallel to the widely spread iconography of the ruler flanked by dragons the extraordinary representation of a frontally Â�portrayed figure seated cross-legged is shown grasping the tongues of a pair of mighty confronted dragons. The reliefs are carved on a now destroyed section of the Talisman Gate, the Bāb al-Ṭilasm (formerly Bāb al-Ḥalaba), one of the four gates at the east of Baghdad (figs.╯139a and b).104 The gate was part of the city wall of Baghdad, the capital of the ʿAbbasid caliphate, and according to the inscription was completed in 618/1221–2 under the caliph Abu ’l-ʿAbbās Aḥmad al-Nāṣir li-Dīn Allāh (577/1181–620/1223). In order to restore moral and political unity and to bring about a revival of the old grandeur of the ʿAbbasid empire,105 the caliph sought to develop and incorporate in the principal Sunnī states fraternities dedicated to the classical Islamic futuwwa (which may be translated as “youthful manliness”), whose aim was to promote a chivalric code of behaviour.106 Portrayed in high relief on the upper part of the spandrels of the arched gateway were two majestic confronted horned dragons whose wide-open snouts with turned-up tips, marked by rows of sharp teeth and fangs, reveal excessively long tongues with bifid ends that are grasped with both outstretched hands by the central frontally-portrayed figure. The latter has a “moon-shaped” face flanked by long tresses, which is distinguished by a halo and a three-pronged crown, and sits cross-

101 ╇ Kitāb ʿajāʾib al-makhlūqāt, ed. Wüstenfeld, 1849, repr. 1967, p.╯133. See Badiee, 1978, p.╯124. 102 ╇ Öney, 1969–70, p.╯200. Cf. Otto-Dorn, 1978–9, p.╯131. 103 ╇ Daneshavri, 1993, pp.╯15–25, esp.╯pp.╯20–1. 104 ╇ Preusser, 1911, p.╯16 top; Sarre and Herzfeld, 1920, vol.╯2, pp.╯153–6, vol.╯3, pls. 10–1; Hartner, 1938, fig.╯26;

Sarre, 1936, fig.╯26 (detail); Kühnel, 1950, p.╯11, fig.╯12; Gierlichs, 1996, pl.╯66.1; Meinecke, 1989, p.╯58, fig.╯7; Hauptmann von Gladiss, ed., 2006, p.╯114, fig.╯12. 105 ╇ For a translation of the inscription, see Sarre and Herzfeld, 1911, vol.╯1, pp.╯35–6. 106 ╇ Cahen and Taeschner, “Futuwwa,” EI2 II, 961a.

The dragon-tamer

the dragon in relation to royal or heroic figures

125

legged just above the apex of the arch. He is clad in an ornately patterned loose long coat, belted just below the slightly protruding belly over trousers, with short boots projecting from below the hem. The dragons have muscular forelegs depicted with toes and talonesque claws, the inner forelegs being raised. Their heads are crowned by a pair of curved horns that project from the top of the head and are flanked by small, cusped ears, folded to the back; the faces are rendered with small, almond-shaped eyes and trefoil-shaped motifs on the face and upper part of the neck, which is demarcated below the jaw line by small, contiguous curls, while larger curls accentuate the back of the neck. The enormous dragons’ scaly serpentine tails twine along the arch forming two impressive knots, the first combining a pretzel-shape with an additional loop, the second a straightforward pretzel-like knot, then extend into three more loops and taper to the tip (lacunae attest to the possible existence formerly of small heads whose identity, bird or dragon heads, can no longer be verified). The dragons’ feathery elegantly upswept wings project from the haunches and end in curls, the long uppermost tip curling inward and terÂ� minating in small, crested birds’ heads with wattles projecting from the chin and long feathers Â�sweeping down the back of the necks. With their small, curved beaks the birds peck at the dragons’ wings. The monumental depiction of the cross-legged central haloed and crowned figure (apparently small in size but in fact reaching about 60 cm in height)107 subduing the giant dragons by holding their tongues seems to reflect the ancient concept of the dragon-tamer. The fact that the royal figure, who is comparatively small in proportion to the dragons, manages to subjugate the great beasts only adds to the impression of dominance and prestige the presentation intends to convey. The gesture of tightly grasping the dragons’ tongues is probably the crucial aspect of the image, rep-

resenting not only the control of speech but also the ritual of public subjugation of the dragons and thereby the harnessing of their forces. This timehonoured motif of victory doubtless symbolises the caliph’s heroic feat of overcoming adverse forces, embodied by the dragons whose likeness was moreover sometimes used to emblematise personified historical foes.108 The pairing of dragons observed specifically on monumental depictions was presumably intended as much to reinforce and replicate the symbolic meaning as to create an effect of pictorial symmetry. This tendency to double single units is a well-known phenomenon among Near Eastern cultures.109 It may thus be presumed that the iconographic Â�elements of this composition have been selected in order to convey a certain meaning or to evoke a certain response in the beholder. It is in the consciousness of the latter that the full cultural meaning of the artwork unfolds. As Ernst Gombrich has pointed out: “The form of a representation cannot be divorced from its purpose and the requirements of the society in which the given visual language gains currency.”110 In order to provide an insight into the symbolic mindset and linguistic expression of the period during which the imagery of the dragon was commonly used as metaphor and allegory, Ernst Herzfeld refers to the report of the contemporary scholar Yāqūt, who compares the conquest of the fort of Ṭabaraq near Rayy by the Great Saljuq sulṭān Ṭoghrıl III in 588/1192 (previously occupied by the Khwārazm-shāh Takash) to a serpent with two heads, one in Iraq, the other in Khurasan, who opens its mouth because it wants to swallow both.111 This double-headed giant serpent or paired dragon whose bifid tongues are immobilised and are effectively bound by the restraining hands of the central seated figure is imbued with more effective talismanic power than anything else. As suggested many years ago by Herzfeld112 and more recently by Meinecke,113 the

╇ Meinecke, 1989, p.╯58. ╇ Friedrich Sarre interprets one of the dragons as symÂ� bolising the Khwārazm-shāh ʿAlāʾ al-Dīn Muḥammad ibn Takash, who had been overcome by the Mongols in the year before the erection of the monument, and the other as a personification of the Mongols themselves (Sarre and Herzfeld, 1911, vol.╯1, p.╯40). Max van Berchem however conceives the second dragon as representing the Ismāʿīlī Grand Master of Alamūt, Jalāl al-Dīn Ḥasan III ibn Muḥammad II who was subordinate to the caliphate since 608/1211–2. Al-Juwaynī, Taʾrīkh-i jahān-gushāy, tr. Boyle, 1912–37, vol.╯2, pp.╯364, 391, 699–701; van Berchem, “Das Baghdad Talismantor,” ed. Sarre and Herzfeld, 1911, vol.╯1, and idem, 1897, pp.╯474–7. This second interpretation may

be more likely since al-Nāṣir had succeeded in securing the return of the Syrian and Iranian Ismāʿīlīs to the fold of Sunnī orthodoxy in 608/1211–2. See however Marshall Hodgson (1955, pp.╯215–25, esp.╯pp.╯222–3, n. 31) who rejects van Berchem’s reading of Ḥasan symbolising the second dragon. Cf. Hartmann, 1975, pp.╯164–6; ul-Huda, 2003, pp.╯13–40, esp.╯p.╯35. 109 ╇ See Kuntzmann, 1983, esp.╯51–116; Anthony, 2007, pp.╯134–5; Kristiansen and Larsson, 2005, pp.╯264, 297. 110 ╇ Gombrich, 1960, pp.╯87–90. 111 ╇ Muʿjam al-buldān, III, p.╯507, cited after Sarre and Herzfeld, 1920, vol.╯2, p.╯153. 112 ╇ Sarre and Herzfeld, 1911, vol.╯1, pp.╯38–40. 113 ╇ The significance of the gate as potent victory

107 108

126

chapter seven

monumental sculptures almost certainly represented a victory commemoration on a triumphal gateway, which also served as entrance to the caliphal palace. Significantly, by employing the time-honoured symbolism of the dragon-tamer this victory commemoration aimed at invoking supernatural talismanic defences as a psychological safeguard against dangerous forces or catastrophic events. The associated royal charisma may moreover be rooted in the celestial,114 and possibly astrological, realms. Hartner sees in the central figure “the new-born child – symbol of the new moon.”115 However on account of the confronted seated lions in profile, which are carved in relief onto the imposts below the arch, the central figure could also be interpreted as a personification of the Sun whose house is in Leo.116 However, irrespective of the visual conflation of the figure with either the Sun or the Moon, its striking imagery probably represented the triumph of the “heavenly ruler” over his enemies by employing a highly potent symbolism of great antiquity.117 On yet another level, as has been pointed out by Herzfeld, the popularly used appellation “Talisman Gate” (Bāb al-Ṭilasm) is certainly no mere coincidence but, as noted above, may be associated with the overall propitious and apotropaic content of the iconographic visualisation.118 This is further corÂ�roÂ�Â�borated by the inscription, which refers to the caliph as “the imām, to whom the whole of humanity has to submit,” as well as “the caliph who is initiated by the master of the world and who is a proof for Allāh of the entirety of beings,” representing a further talismanic aspect alluding to defeated or yet to be defeated rivals of the caliph.119 It has been shown that the emblematic representation of the cosmic ruler flanked by and often grasping dragon-headed staffs, so ubiquitously employed on works of art, in particular on metalwork, from the mid-eleventh to the early thirteenth century, was associated with the ancient concept of the “Master of the Animals,”

represented here as the “Master of the Dragons.” The motif gives visual form to the idea of the cosmic ruler wrestling with and subduing threatening forces, represented by the dragons. What is more, the iconographic expression portrays the ruler as dragon-tamer whose supernatural powers enabled him to subdue, contain, tame and transform the dragons rather than annihilate them. Since this iconographical theme was used as Â�decoration on a wide range of artefacts accessible to a larger part of society (the Bobrinski bucket, for instance, was perhaps given to a religious dignitary), it seems probable that its semantic meaning was not seen as exclusively confined to the art of the court and the upper classes but rather that it was a multivalent symbol, transcending any specific level of society.120 This was made possible by the fact that the ethos of cosmic rulership was understood to comprise the idea of victory over adverse forces as a synthesis of the fundamental royal virtues and hence itself served as a visual metaphor imbued with a more generalised auspicious and apotropaic significance.121 It is moreover tempting to propose that a bust or a head flanked by two dragons represented an abbreviated reference, which has close analogies to the semantic meaning of the dragon-tamer, and whose allegorical content was clear to a medieval audience. The emblematic significance of the allegorical imagery that omitted the visual manifestation of the bodies would have conveyed the same meaning as the examples of the dragon-tamer motif on Khurasani metalwork or the Bāb al-Ṭilasm in Baghdad. Yet another possibility may be considered, namely that a twofold approach was possible in reading this imagery, and that a literal as well as an allegorical conception was attached to the “abbreviated” form. What is more, as already noted above, the representation alludes to a mythical vision of the ruler and refers to the age-old portrayal of the archetypal royal hero overcoming dangerous creatures in a mythical-heroic manner. It may be suggested with a degree of certainty that to the

monument was still relevant in not too distant memory which is reflected in the symbolic act of the Ottoman sulṭān Murād IV ibn Aḥmad I (1032/1623–1049/1640) who, after his occupation of Baghdad in 1638, walled up the gateway in order to prophylactically curtail the potency of the gate as victory monument, in other words, to preclude any future conquests. Meinecke, 1989, p.╯58. Cf. also Sarre and Herzfeld, 1911, vol.╯1, pp.╯34–6. 114 ╇ Cf. Hauptmann von Gladiss, ed., 2006, p.╯115. 115 ╇ Hartner, 1938, p.╯144. 116 ╇ Pancaroğlu, 2001, pp.╯171–2, n. 29. The lion’s asso-

ciation with the sun, which can be traced back to remote antiquity, as far as the fourth millennium bc (Hartner, 1938, pp.╯115, 119), apparently originated in a purely mythological or metaphysical conception (idem, p.╯138). The personification of the planets is discussed in chapter 8, part b. 117 ╇ Cf. Hauptmann von Gladiss, ed., 2006, p.╯115. 118 ╇ Sarre and Herzfeld, 1911, vol.╯1, pp.╯38–9. 119 ╇ Eidem, pp.╯39–40. 120 ╇ Cf. Pancaroğlu, 2000, p.╯246. 121 ╇ Eadem, pp.╯246–7.

the dragon in relation to royal or heroic figures

127

It is inherent in the dragon’s very nature that, owing to its multivalent qualities, it escapes definition. One of its many preeminent features is the conspicuous emphasis on the tongue, used to denote the supernatural quality of the creature. In most of the examples discussed earlier, the deep gaping throats of the fabulous creatures reveal tongues, which are invariably oriented towards the central motif. In the Armenian examples at the late ninth-century monastic complex of Saint Peter and Saint Paul in Siunikʿ (figs. 120–122) and the mid-tenth-century cathedral of the Holy Apostles in Kars (figs. 118 and 119), the outstretched ophidian tongues are clearly directed at the ears of the human heads. The serpent-headed staffs that flank the seated frontally portrayed figures depicted on twelfth- or thirteenth-century Islamic metalwork are also shown with the tongues projecting from the mouths towards the human heads (figs. 113–116). This may be related to the fact that a snake’s tongue is known to dart out before it strikes at a foe. At the Bāb al-Ṭilasm in

Baghdad of 618/1221–2 the central figure is hence portrayed as holding tightly on to the dragons’ tongues (figs. 139a and b). As has been shown, this gesture is associated with the symbolism of subjugation by bridling the creatures and thereby rendering them defenceless. Reference to taming and mastering the (serpent-)dragon by means of holding its tongue is also made in al-Kisāʾī’s story of Moses (Mūsa) and the pharaoh (firʿawn). To save the life of the latter from the attack of the Prophet’s rod turned giant serpent, Moses “cried to the serpent, which came to him as a tame dog comes to his master. Moses put his hand in its mouth and caught its tongue, whereupon it was again a staff.122 The vital organ of the tongue thus seems to epitomise the seat or the extension of power of the mythical creature. This underscores the inference that the use of control over this potent emblem permits the powerful creature to be overmastered. Establishing control over the dragon implies of course not only the domination of its dark and evil nature, but also of its inherent beneficial qualities. An attempt will be made in the next and subsequent chapters to describe the nature of these qualities. Tongues are generally associated with the power of speech, a quality of the dragon that will be addressed in chapter 13. According to the Qurʾān, the magician-king Sulaymān ibn Dāwūd, the biblical Solomon, was acquainted with the speech of birds and animals (sūra 27, 16–9), a tradition based on I Kings 4.33.123 In Judaism as well as in Islamic scriptural tradition the understanding of the language of animals was initially associated with the story of Paradise.124 The early Haggada speaks of Adam understanding the language of birds and beasts125 and of Eve126 speaking with the serpent127 who was considered to be the wisest, the cleverest and the most astute of all animals.128 In the late Haggada non-Jewish folklore was censored before it was assimilated; the

╇ Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯228. ╇ Walker [Fenton], “Sulaymān b. Dāwūd,” EI² IX, p.╯882b. 124 ╇ Al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, pp.╯36–46. Cf. Schwemer, 1994, p.╯137. 125 ╇ Adam lost this ability with the Fall of men but in the Messianic period this condition will be re-established. Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯5, pp.╯94, n. 54, 119–20, n. 113. 126 ╇ Idem, p.╯91, n. 48. 127 ╇ In Paradise all animals could speak, but as punishment for allowing Iblīs to seduce Eve (Ḥawwāʾ) the serpent lost not only its legs but also the ability to speak (a skill similarly denied to all the other animals). Al-Kisāʾī, Qiṣaṣ

al-anbiyāʾ, tr. Thackston, 1978, p.╯46. Cf. Book of Jubilees 3.28; Philo of Alexandria, De Opificio Mundi 55.156. Older rabbinic sources do not know of the original language spoken by man and the animals, but a passage in the Lekah on Genesis, 3.1, maintains that before the Fall only the serpent spoke the original language of man, Hebrew, whereas the rest of the animals spoke their own language which was only understood by Adam. See Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯1, p.╯181, n. 91, and vol.╯5, pp.╯91, n. 48, 94, n. 58, and 100–1, n. 83. 128 ╇ The wisdom of the serpents is also noted in the New Testament (Genesis 3.1; Matthew 10.16). Cf. Schwemer, 1994, p.╯137.

medieval mind the animal combat or the mastery of beasts was associated with the timeless image of a heroic and royal ideal associated with the ruler. Representations of the ruler between two dragons, as emblematised in the Bāb al-Ṭilasm relief, show him constraining the great powers of the dragons within the control of his realm, thereby demonstrating his mythical powers. It is conceivable that when commissioning the visual metaphor and sculptural vision of the mythical dragon-tamer for the monumental victory commemoration, the caliph al-Nāṣir’s aim was to evoke the supernatural forces associated with the time-honoured Indo-Iranian concept as well as to appropriate the associated royal charisma of ancient kingship. c.╇ Symbolism of the dragon’s tongue

122 123

128

chapter seven

biblical king Solomon took the place of the great serpent(s) and could understand and speak the language of animals as well as teach their language to others.129 However, in spite of this assimilation the connection of the serpent with the language of animals remained widespread in folktales throughout Asia and the Near East.130 One of the oldest recorded versions is a third-century Buddhist tale, in which a king rescues the daughter of a dragon-king who thereupon grants him a wish. The king says that he has already many precious objects but that he wishes to understand the language of the animals. The dragon-king grants him this request on condition that he will keep his ability a secret.131 In Armenian folktales serpents can speak and the touching of a serpent’s tongue serves as a source of knowledge.132 Vestiges of such myths may be discerned in Armenian popular traditions, which similarly record that a grateful serpent king gave the knowledge to cure all diseases to a boy who had saved his son. He transmitted this

knowledge by touching the boy’s tongue with his own; the boy then became the physician known also in the Turkish-speaking world as Luqmān al-ḥakīm.133 In another Armenian story, the grateful serpent-prince licks a dead person back to life.134 According to a further tradition the grateful serpent passes on the ability to understand the language of all animals through the power of its breath.135 Similarly in antiquity the ability to confer supernatural knowledge of the language of animals, that is to say, the art of divination, on human beings was most commonly attributed to the serpent.136 These scattered references to human mastery of the language of the animals functioned as metaphors for the extraordinary lucidity of prophets, saints and sages. Other widespread views held that the eating of serpent’s flesh transmitted supernatural wisdom to the consumer,137 and that knowledge of the language of animals was given to Arabs.138 This mantic art they are supposed to have acquired by

╇ In interpretation of 1 Kings 5.13. See Ginzberg, 1909– 38, repr. 1946 and 1955, vol.╯4, pp.╯138–40, 287–8, n. 34. In later Jewish folk tales, Leviathan replaces the great serpent (Noy, 1971, p.╯177), for instance, in the Jewish variant of the “Story of the Dutiful Son” (idem, p.╯196): “He (Leviathan) said to him: “Open your mouth!” He opened his mouth and Leviathan spat into it three times. At once, the spirit of wisdom and cleverness rested upon him, and he knew and understood the language of animals and birds, and he spoke seventy languages.” 130 ╇ For a comprehensive bibliography, see Noy, 1971, pp.╯171–208. 131 ╇ Chavannes, 1910, vol.╯1, pp.╯382–3. A comparable Jātaka story speaks of a king saving the life of a dragon-king who thereupon rewards him with a charm “giving knowledge of all sounds”; Cowell, ed., 1897, repr. 2000, vol.╯3, pp.╯174–7; see also Vogel, 1926, p.╯22. 132 ╇ Hoogasian-Villa, 1966, p.╯67. 133 ╇ Eadem, pp.╯426–9 and 531–2. 134 ╇ Surmelian, 1968, p.╯232. Cf. the parallel to the Armenian mythical dogs, called Arlez (Arm. aralez or yaralez), one of which is black, the other white, who live with invisible powers, and who are said to have licked the bodies of wounded war heroes back to life. See Eznik, Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, ch. 122; cf. Karamanlian, P.A., Die Aralezen bei Eznik (German Resumé, pp.╯705–7), as cited in Schlerath, 1954/58, p.╯39. 135 ╇ Hoogasian-Villa, 1966, pp.╯401 and 528. The motif occurs in the Shāh-nāma recounting how Buzurjmihr, the later minister of Khusraw Anūshirwān, was breathed on by a black snake which was interpreted by his companion who witnessed the scene as indicating that Buzurjmihr would attain a position of great power (tr. and ed. Mohl, 1838–1878, vol.╯6, pp.╯247–9, ll. 1037–1045). See also p. 61. At the same time the Iranian national epic repeatedly invokes the metaphor of the dragon’s breath to allude to an impending and potentially fatal calamity (tr. and ed. Mohl, 1838–1878, vol.╯1, p.╯153, ll. 450–3; vol.╯3, p.╯183, l. 510; vol.╯5, p.╯519, l. 1897). In like manner the potentially fatal power of the “slavering dragon”

and of his breath is mentioned in the romance Wīs u Rāmīn (tr. and ed. Davis, 2008, p.╯364 and 83, respectively). See also the related metaphorical imagery used by Matthew of Edessa, p. 9, n. 61. 136 ╇ References to such phenomena are preserved in Greek tradition. The Hellenistic/Roman author Apollodorus relates that Melampus acquired the gift of prophecy after he had performed funeral rites for some serpents which had been killed by his servants and thence reared the young serpents (Bibliotheca I. 9. 11–2, p.╯87; Scholiast on Apollonius Rhodius, Argonautica I.118. Cf. Eustathius on Homer, Odyssey XI 292, p.╯1685; Pliny, Naturalis Historia, X 137. See also Frazer, 1888, p.╯166). When these were fully grown, they crept onto each of his shoulders as he slept and cleansed his ears with their tongues. When he woke up he understood the voices of the birds and other animals, and thus acquired the gift of foretelling the future. Cassandra and her brother Helenus are said to have acquired their prophetic gift in the same manner. When young they were placed one night in the temple of Apollo and in the morning were found with serpents wreathed around their bodies, licking their ears (cf. Scholiast on Homer, Iliad, VII 44; Tzetzes, Scholiast in Lycophron, Introd. I, pp.╯266–8, ed. Müller, C.G.; Scholiast on Euripides, Hecuba, V 86). Likewise Porphyrios is recorded as saying that perhaps all men might understand the language of all the animals if a serpent had washed their ears (De Abstinentia, III 4). 137 ╇ “The reason why the serpent has the capacity to understand the language of the birds and the animals is according to Pliny (after a saying of Democritus) that they are generated from the mixed blood of diverse birds, hence anyone who eats a serpent will understand the bird Â�language” (Â�Naturalis Historia, X 70; see also XXIX 22), as cited in Frazer, 1888, p.╯180. 138 ╇ Cicero, De divinatione 1.92 and 94. Cf. Appian’s account of his successful escape from the Jewish revolt in Egypt (c. 116 ad) near Pelusium thanks to an Arab guide who correctly interpreted the three screeches of a crow. Appian, Roman History 24,19.

129

the dragon in relation to royal or heroic figures

129

eating the heart or the liver of large serpents.139 The lasting popularity of this type of account is evidenced in its appearance in the frame tale of the Alf layla wa-layla in the story of the Queen of the Serpents. Here the queen consents to be slain and sacrifices her flesh, which when boiled and eaten has healing properties.140 Moreover, the drinking of the elixir is said to give access to the fountains of knowledge, in other words knowledge of all sciences. Another story, the

Keys of Destiny, tells of a place that harbours the secret of transforming base metal into gold and ultimately of gaining immortality. Access to it can only be gained by killing a serpent in a black valley and preparing an ointment from the reptile’s heart mixed with other ingredients.141 In folk tales the imbibing, ingesting or application by any means of elixirs prepared from the heart or the flesh of a serpent initiates into hidden secrets.142

139 ╇ See Flavius Philostratus, who at the beginning of the third century ad, wrote a biography of the wandering philosopher Apollonius of Tyana in Cappadocia (Vita Apollonii 1.20, 2 vols., tr. Conybeare, F.C., London, 1980, p.╯57; also 3.9, p.╯249), who became master of the beasts by acquiring an understanding of the language and ideas of animals from the Arab tribes “by feeding either on the heart or the liver of the dragon.” The mantic significance attributed in antiquity to (serpent-)dragons in revelations and oracles can still be gleaned from the writings of the fifth-century Greek poet Nonnos (Dionysiaca XLI 340–1) when he reports that ophion has recorded the sayings of the gods about the world (thesphata kosmou). The second-century Greek traveller and geographer Pausanias (Graeciae Descriptio IV 10.5–6) even called a seer outright Ophioneus. It may not be irrelevant to note that in the Scandinavian version of the epic of Siegfried, the hero roasted the dragon’s heart, whereby some dragon blood dropped onto his tongue, which led to his acquisition of an understanding of the bird language; see Schirmunski, 1961, p.╯55. 140 ╇ 535th Night [1830 Calcutta ed. count]. The Book of the Thousand Nights and a Night, tr. Burton, 1885, vol.╯5, pp.╯407–9. Cf. Marzolph and van Leeuwen, 2004, pp.╯349–

50. In a Syrian tale a dervish is said to have drunk serpent water, following which the serpents cannot bite him and he is able to talk with serpents and birds in their respective languages; Syrische Sagen und Märchen, tr. Prym and Socin, 1881, pp.╯150–1. 141 ╇ Marzolph and van Leeuwen, 2004, p.╯251. 142 ╇ This has certain parallels to passages in the Old TesÂ� tament. Whereas Isaiah (6.6–8) and Jeremiah (1–9) are touched by God or an angel, Ezekiel (3.1–3) is ordered to eat a scroll with writing on it that tastes as sweet as honey. This intake is reminiscent of ancient initiation rituals that Â�similarly involve the touching of the mouth, that is to say, the tongue, with a liquid substance which is then swallowed. Van Lint, 2005, p.╯374. In the early Turkish Anatolian epic, the Baṭṭāl-nāma (“Book of Baṭṭāl”), the angel Gabriel/ Jibrāʾīl instructs the Prophet Muḥammad to place a drop of saliva in the mouth of the early Muslim frontier warrior ʿAbd al-Wahhāb. The warrior swallows it but it rests in his throat “to be held in trust.” He then passes the drop on to Jaʿfar, later named Sayyid Baṭṭāl Ghāzī, who on swallowing it “acquired perfect knowledge of seventy two different languages and of twelve sciences.” Dedes, 1996, pp.╯100, 117–8.

130

chapter seven

the dragon and astrology

part four

the dragon in astrology, alchemy, medicine and magic

131

132

chapter eight

the dragon and astrology

133

chapter eight

The dragon and astrology

a.╇ Astrology in medieval Central Asia The ancient practice of astrology,1 the interpretation of the movement of celestial bodies as reflecting divine powers and enabling prognostication of the future, had a deep and pervasive influence on early and medieval Islamic thought and culture.2 The history of astrology, which had been introduced into the Iranised world of Central Asia through Graeco-Babylonian influence, goes back to ancient times. Moreover, with the spread of Buddhism into Central Asia, Iran and China, Indian nakṣatra (lunar asterism) astrology was introduced.3 Interest in the science of astronomy, closely and in practice inextricably linked with astrology, is further corroborated in Central Asia by findings during excavations at the religious and funerary complex of Qoy-Qrylgan-Qalʿa (fourth/ third century bc to the third/fourth century ad) in the region of Toʿrtkoʿl in Khwārazm, which might indicate that the site could have been used for astronomical observations. The findings include clay fragments and rings that could be reconstructed to form an astrolabe with a circular alidade.4 The great tenth-century Khwārazmian scholar Abu ’l-Rayḥān Muḥammad ibn Aḥmad al-Bīrūnī, known as “the Master” (al-ustādh), who devoted more than half of his extensive writings to astronomy and astrology, writes in relation to 1 ╇ Astrology seems to have been widespread and practised at a very early time by all nations in the ancient East. Cf. Â�Jeffers, 1996, p.╯147. 2 ╇ The ʿAbbasid caliphs, in particular al-Manṣūr, accorded particular prominence to the study and practical application of astrology. Gutas, 1998, p.╯16 and n. 7, and p.╯33. 3 ╇ Pingree, 1963, pp. 230–1, 240–1. The Śārdūlakarṇāvadāna, which contains an exposition of this system, was widely diffused (idem, pp.╯240–1) and summarised in Chinese by the Parthian prince An Shih-kao in the second century ad (Śārdūlakarṇāvadāna of the Divyāvadāna, tr. Mukhopadhyaya, S., Santiniketan, 1954, pp.╯213–7; and on An Shih-kao, see Zürcher, 1959, vol.╯1, pp.╯32–4; cited after Pingree, 1963, pp.╯240–1) and fully translated twice in the third century ad (Śārdūlakarṇāvadāna, tr. Mukhopadhyaya, 1954, pp.╯xii–xiii, cited after Pingree, 1963, pp.╯240–1). 4 ╇ Schirmer, 1926–7, pp.╯43–6, 63–79. According to Boyce and Grenet (1991, pp.╯184, n. 133, 193, n. 173) the primary

ancient Khwārazmian astronomy and astrology that the Khwārazmians “knew them [the constellations] better than the [pre-Islamic] Arabs.”5 Later Parthian (250 bc–224 ad) and Â�SasaÂ�nian (224–652 ad) kings are recorded as having maintained a “chief of the star-gazers” Â�(Â�axtarÂ�mārānsālār) at court where a regnal horoscope would be drawn up for each king. However, only during the reign of the Sasanian king Shāpūr I (r. 241–271) was the study of Iranian astronomy and astrology known to have been encouraged. According to the Dēnkard (Book IV), the ninthcentury compendium of the Zoroastrian religion, the king is said to have gathered the astrological writings, “which were dispersed throughout India, the Byzantine empire and other lands.”6 Another Sasanian ruler who, according to al-Bīrūnī, enÂ�courÂ�aged Greek or Graeco-Syrian and Indian scholars in Iran was Khusraw I Anūshirwān (r. 531–579).7 According to a tradition reported by Firdawsī, the colossal throne (taq-i taqdis) of his grandson, the last Sasanian king Khusraw II Parwīz (r. 591–628), was embellished with images of the seven regions as well as the seven planets and the twelve signs of the zodiac.8 As the centre of the astrological throne the ruler represented the one who held the power to influence the stars.9 Political crises were regarded as inevitable at acute aspects of the constellations.10

purpose of the upper storey of the tower at the complex was not, as supposed by the excavators of the site, to function as an observatory; the use of the primitive astrolabes found at the site was only secondary. 5 ╇ Kitāb al-Āthār, tr. and ed. Sachau, 1876–8, p.╯226. 6 ╇ Zaehner, 1955, repr. 1972, p.╯8; Pingree, 1963, p.╯241; Gutas, 1998, p.╯36, see also p.╯41. 7 ╇ Kennedy, 1956, p.╯50. 8 ╇ Tr. and ed. Mohl, 1838–1878, vol.╯6, p.╯253. Comparable imagery is reported secondhand from Theophanes through Kedrenos that in Ganzaca (Ganjak) when Heraclius captured Khusraw’s palace in 624, he saw Khusraw’s image in the domed roof of the palace, as though enthroned in heaven and surrounded by the sun, moon and stars. Texts cited in full in Herzfeld, 1920a, pp.╯1–3; L’Orange, 1953, pp.╯18–27, esp.╯pp.╯19–21. Cf. also Carter, 1974, p.╯177 and n. 25. 9 ╇ Herzfeld, 1920, pp.╯1–24 and 103–47. 10 ╇ On court astrologers, cf. Christensen, 1944, p.╯396; on

134

chapter eight

Sanskrit astrological texts were translated into Pahlawī in the Central Asian world, according to David Pingree, in particular in Sind and Afghanistan.11 During the last centuries of Sasanian rule the influence of the sciences of astronomy and astrology, which were often a synthesis of Hellenistic and Indian theories, was particularly prevalent. The works of many early Islamic astrologers, many of whom were Iranians or Central Asians, in turn incorporated numerous Indian astronomical and astrological theories. This is reflected in the works of the ninth-century astrologer Abū Maʿshar Jaʿfar ibn Muḥammad ibn ʿUmar (d.╯272/886) from Balkh who combines in his Zīj al-hazārāt (“The Zij of the Thousands”) Hindu, Sasanian and Hellenistic astronomical and astrological traditions, claiming to have used an ancient Persian text from antediluvian times written during the reign of Ṭahmūrath (Av. Takhma Urupi), the second king of the Pīshdādian dynasty of legÂ� endary epic Iranian history.12 Ibn al-Nadīm quotes passages from Abū Maʿshar’s Kitāb ikhtilāf al-zījāt (“The Book on the Variations among zījs”), which contain calculations determining the movement of the planets:

The translation movement of the early ʿAbbaÂ� sid period saw an unprecedented level of activity in the sciences of astronomy and astrology. This had a profound effect on social attitudes.14 Indeed astronomy was viewed by scholars as the “mistress of all sciences.”15 The strong astrological tradition in the Sasanian period helps to explain why the field of Arabic astrological literature in the early Islamic period was dominated by Iranian

scholars,16 who in turn maintained contact with their Indian counterparts.17 In fact both sciences, astronomy (ʿilm al-hayʾa, the “science of the figure (of the heavens)” or ʿilm al-falak, “science of the spheres”) and astrology (ʿilm al-nujūm, the “science of the stars”), were for a long time so close that the word munajjim was used to designate both astrologer and astronomer.18 This is based on the fact that, according to Jābir ibn Ḥayyān (known to the Latins as Geber),19 one of the main representatives of earlier Arabic alchemy, “the astrologer must be a mathematician; he must have mastery of astronomy, this is a part of ʿilm al-nujūm (the “science of the stars” or astrology). For ʿilm al-hayʾa (astronomy) is the description of the situation of the state of the sky and what it contains (ṣūrat waḍʿ al-falak wa-mā fīhi), whereas astrology is the gift of the stars (ʿaṭāʾ al-kawākib).20 Astrology, which involves calculating the position of the planets and the mathematical production of horoscopes, is often referred to as judicial (or catarchic) astrology (ʿilm aḥkām al-hayʾa, the “science of the judgment of the stars”).21 Astrological predictions consisted not only of determining the fate of an individual (mawālīd, “genethlialogy,” or horoscopic astrology) and of the auspicious and inauspicious timing of events and actions (ikhtiyārāt, “choices”), but also of the application of continuous horoscopes for determining the course of events for a country or dynasty or to answer specific questions (masāʾil, “interrogations”).22 In spite of the fact that astrology stood in fundamental opposition to the tenets of the Islamic religion,23 it gradually established a role in the public life of Islamic rulers.24 Well-known astrologers included, for instance, Māshāʾallāh ibn Atharī, an Iranian Jew from Basra, and Abū Sahl al-Faḍl ibn Nawbakht, an Iranian, who converted from Zoroastrianism to Islam at the court of the second ʿAbbasid caliph Abū Jaʿfar al-Manṣūr (r. 136/754–158/775). Both

horoscopes, see Kennedy and Pingree, 1971, p.╯vi; see also Russell, 2004, p.╯85 and n. 11. 11 ╇ Pingree, 1963, pp.╯242–3. 12 ╇ The legendary history of king Ṭahmūrath is recorded, for instance, by al-Thaʿālibī in his Taʾrīkh Ghurar al-siyar (tr. and ed. Zotenberg, 1900, pp.╯7–10) in which he also describes the kings’s subjugation of Iblīs demonstrated by his using Iblīs as mount to perambulate the world. Cf. Pingree, 1963, pp.╯243–4 and idem, 1968, pp.╯3–4 and n. 3. The culture hero Takhma Urupi riding Angra Mainyu as his horse from one end of the earth to the other is mentioned twice in the Avesta (Yasht 15.11–2, 19.28–9). 13 ╇ Kitāb al-Fihrist, Cairo, n.d., pp.╯348–50, cited after

Â�Pingree, 1968, pp.╯3–4. 14 ╇ Cf. Gutas, 1998, pp.╯108–10. 15 ╇ Ullmann, 1972, p.╯277, n. 5. 16 ╇ Idem, pp.╯296–7 and n. 6. 17 ╇ Gutas, 1998, p.╯15. 18 ╇ Fahd, “Munadjdjim,” EI² VII, 557b. 19 ╇ On Jābir ibn Ḥayyān, see Sezgin, 1971, pp.╯132–269. 20 ╇ Idem. 21 ╇ Savage-Smith, 2004, p.╯xxxvii. 22 ╇ Fahd, “Nudjūm, Aḥkām al-,” EI² VIII, 105b; SavageSmith, 2004, p.╯xxxvii; Saliba, 1992, pp.╯56–63. 23 ╇ Saliba, 1992, pp.╯45–67; Michot, 2004, pp.╯277–340. 24 ╇ Gutas, 1998, pp.╯33–4.

The people of the time of Ṭahmūrath and the more ancient Persians called these the “cycles of the thousands”; and the wise men of India and their kings, the ancient kings of Persia, and the ancient Chaldeans who lived in Babylon determined the mean longitudes of the seven planets by means of them, preferring them over others because of their accuracy and brevity.13

the dragon and astrology astrologers were leading astrological advisors to this caliph and instrumental in drawing up a horoscope for him determining the day (4 Rabīʿ al-Thānī 145/30 July 762) on which the construction of the city of Baghdad should begin. The Iranophile ʿAbdallāh al-Maʾmūn (r. 198/813–218/833) was the son of Hārūn al-Rashīd by an Iranian slave girl and after his father’s death in 193/809 ruled over the eastern half of the caliphate residing in Khurasan. After overthrowing his brother al-Amīn, who ruled over the western half, in a civil war, al-Maʾmūn became caliph in Marw in 198/813 and ruled the caliphate from that city for another five years, during which time he gathered around him astronomers from Ferghana, Chach, Khwārazm and greater Khurasan.25 In his celebrated historical work al-Masʿūdī reports that under the influence of his Iranian vizier al-Faḍl ibn Sahl Zadhānfarūkh (d.╯202/818), al-Maʾmūn:

135

When he moved to Baghdad in 204/819, he was followed by a host of astronomers and astrologers.27 The pervasive role that astrology played may be seen in the example of the vizier Ibn Muqla (d.╯328/940) who, upon the advice of the astrologers, arranged his meeting with the caliph al-Rāḍī (322/934–329/940) when the Moon was in the zodiac sign Leo governed by the Sun, considered the most auspicious time to meet for secretive affairs. The science of the celestial bodies was thus of such significance for Islamic rulers that most of their decisions were governed by astrological considerations.28 This predisposition was equally prevalent among rulers of Turkic stock and is also attested for the Sogdians (who were closely associated with the Turkic empires and played the role of active agents of cultural interaction) in the

period immediately preceding Islam. The major assemblies of the Sogdians (anvāzak), whose responsibilities included proclaiming the new king, were composed not only of local rulers and ecclesiastical dignitaries but also of astrologers and soothsayers.29 During their campaigns, the Turkic Ghaznawid sulṭāns, Maḥmūd and Masʿūd were accompanied by their astrologers.30 At the battle of Dandānaqān in 431/1040, the victorious Saljuqs had an astrologer (munajjim) with them, who was rewarded, presumably on account of his correct prognostications, when Ṭoghrıl proclaimed himself ruler of Khurasan.31 Soothsaying qualities are also attested for members of the ruling house, for according to Ibn al-Athīr, even Ṭoghrıl’s cousin, Qutulmish ibn Arslan Isrāʾīl, had astrological skills.32 Astrologers from greater Khurasan are known to have moved to the courts of the Rūm Saljuqs of Anatolia. Among them was the mother of the thirteenth-century chronicler Ibn Bībī, known as al-Bībī al-Munajjima (“the lady, the astrologer”), who was from Nīshāpūr and working as astrologer at the court of the Khwārazm-shāh Jalāl al-Dīn. There she met an ambassador of the Rūm Saljuq sulṭān ʿAlāʾ al-Dīn Kay Qubādh I of Konya. The meeting made such an impression on the ambassador that he informed the Saljuq sulṭān and in due course the family was invited to the court in Konya where she successfully predicted a military victory.33 The services of the astrologer would frequently be requested before the undertaking of any important action. The astrologer Shams al-Dīn, for instance, is said to have determined the time of the military offensive under the Ayyubid al-Kāmil Nāṣir al-Dīn (r. 615/1218–635/1238).34 Like their eastern cousins, the Great Saljuqs, some Rūm Saljuq rulers, such as Malik ʿAlāʾ al-Dīn Dāwud Shāh, also practised the science of astrology themselves.35 Hence throughout medieval Islamic history the science remained popular, astrological practices being patronised not only by members of princely circles but by most segments of Islamic society.36

25 ╇ Al-Najaf, al-Maṭbaʿa al-Haydariyya, 1368/1949, p.╯133, cited after Saliba, 1992, pp.╯58 and n. 78. 26 ╇ Kitāb murūj al-dhahab (“Meadows of Gold”), tr. and ed. Barbier de Meynard and Pavet de Courteille, 1861–77, vol.╯8, pp.╯300–1. 27 ╇ Akhmedov, 2000, p.╯195. 28 ╇ Cf. idem, p.╯199. 29 ╇ Vessantara Jātaka 64, 1a–4a, cited after Widengren, 1969, p.╯117 and n. 61.

╇ Bosworth, 1963, p.╯118. ╇ Idem, p.╯217 and n. 40. 32 ╇ Idem, p.╯217. 33 ╇ Ibn Bībī, al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, pp.╯2–4, 187–8. 34 ╇ Ibn Shaddād, al-Aʿlāq al-khaṭīra, vol.╯3, 2, pp.╯552–3, cited after Saliba, 1992, p.╯58, n. 81. 35 ╇ Ibn Bībī, al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, p.╯150. 36 ╇ Saliba, 1992, pp.╯45–67.

used to spend time investigating astrological rulings and prognostications, following what the stars prescribed, and modeling his conduct on that of the past Sāsānian emperors like Ardashīr ibn Bābak [Ardashīr Pāpakān I, r. 224–41] and others.26

30 31

136

chapter eight

b.╇ The dragon in medieval Islamic Â�astrology The idea that eclipses of the Sun and the Moon were caused by the interference of an eclipse monster was widely held throughout the Eurasian continent and can be traced back to remote antiquity.37 The fearful monster, which quenched the light of the supreme luminaries by seizing them in its jaws, was generally conceived as a giant serpent or dragon, an iconography thought to be of oriental origin.38 Its function was thus seen to be that of threatening and “devouring,” as well as “delivering” and protecting the great luminaries at certain irregular intervals.39 A number of theories arose to explain the dragon’s role in the phenomena of solar and lunar eclipses and lunar waxing and waning. KhāleqīMoṭlaq offers the following summary: ...a dragon comes up from hell every month on the eastern side of the sky and swallows a piece of the moon’s disc every night until the night comes when no part of the moon can be seen. Then the moon-god kills the dragon from inside its belly and triumphantly re-emerges. In later times, however, the sun took over the moon’s role in the celestial combats, and it was the sun which slew the dragon and rescued the moon from the dragon’s belly twelve times every year.40

Vestiges of these and related beliefs survive in Persian poetry and will be cited below. 37 ╇ Hartner, “Al-Djawzahar,” EI² II, 501b. Cf. Massé, 1938, vol.╯1, p.╯172. 38 ╇ In ancient Babylon “the 28th of the month was a day of lamentations when prayers of penitence were offered, because the moon had disappeared from view and was to remain hidden for a few days in the power of the dragon.” Green, 1992, p.╯29; Hartner, 1938, p.╯132, n. 24. The “Chaldeans” considered the dragon to have been created even before the constellations and the planets, and guarding over the universe with its head towards the sunrise and its tail to the sunset. Mackenzie, 1964, p.╯525, and idem, “Gozihr,” EIr. 39 ╇ Hartner, 1938, p.╯131. Cf. Daneshvari, 1993, p.╯20. See also the Babylonian Talmudic tract Avodah Zarah (“Mishna on Idolatry,” VIII a) in which the dragon is portrayed as devouring the sun. See also Epstein, 1997, p.╯76. 40 ╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. 41 ╇ Khareghat, 1914, p.╯129. 42 ╇ See Panaino, 2005, pp.╯73–89, esp.╯74–5, who discusses the Zoroastrian practice of deducing omens through ophiomancy (that is to say, divination by serpents) which was linked with astral elements. In this connection it is interesting to consider the reference of the fifth-century Armenian theologian, Eznik of Koghb (Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, p.╯641, ch. 291) to the preChristian belief which saw the heavenly bodies as deities when they worshipped venomous creatures, whereby he implicitly appears to associate astrolatry with ophiolatry. 43 ╇ For an analysis of the origin of the concept of Rāhu, see Mallmann, 1962, p.╯81; Markel, 1995, pp.╯55–64; Pingree,

The idea that these phenomena were caused by a body whose head and tail intercept the Sun’s and the Moon’s light was probably related to the emergence of definite ideas as to the nature of the orbits of the Sun and the Moon and their opposite points of intersection between the Moon’s orbit and the ecliptic.41 The classical theory of the dragon myth seems to have been modified in accordance with developments in astrological doctrine at least from late Arsacid and Sasanian times onwards.42 Sasanian astrologers received from India the notion of Rāhu, a celestial serpent whose head (siras) and tail (ketu) cause solar and lunar eclipses.43 In Pahlawī Rāhu was referred to as Gōchihr, his head sar, and his tail dumb; in Arabic, the latter were respectively called raʾs and dhanab.44 The great treatise on horoscopic astrology of the first-century Hellenistic astrologer Dorotheus of Sidon, which was first translated into Persian in the third century and into Arabic in the eighth century, contains a chapter (V, 43) entitled “on clarifying the phases of the moon and the head of the dragon and its tail ...” It states that “the head is called the “ascending” and its tail the “descending” and the signs which those learned in the stars call “obscured” are from Leo to Capricorn …”45 Astrology also offered support for ZoroasÂ� trian apocalyptic ideas, according to which the planetary bodies were regarded as evil; the 2006, p.╯240. In the Rigveda (5.40.5–9) Rāhu is known as a demonic being, Svàr-bhānu-, which is said to have pierced the Sun with darkness. In post-Vedic mythology Svàr-bhānuis replaced by Rāhu-, his name sometimes being conferred upon the latter; Svàr-bhānu- perhaps meaning “who has the effulgence of the sun” or “who is affected by the effulgence of the sun.” Advanced knowledge of periodical eclipses of the sun and the moon led to the belief in two demonic beings, the red Rāhu- and the black Ketu-. See Scherer, 1953, pp.╯100–1. Representations of Rāhu in a narrative context begin to appear in Indian art slightly earlier than his iconic portrayal as a member of the planetary deities. One of the earliest known portrayals of Rāhu being in a relief of the “Churning of the Ocean” carved over the façade of the doorway of cave temple number nineteen at Udayagiri in the Vidisha district of Madhya Pradesh, which probably dates from c. 430 to 450. The planet is shown as a large horrific head with bulging eyes with a fierce, demonic expression turned to the right, his hands probably cupped together with palms facing upward in the gesture of scooping the elixir of immortality (on the legend, see the discussion below). Â�Williams, 1982, p.╯87 and pl.╯117. 44 ╇ Pingree, 2006, p.╯240. 45 ╇ Dorothei Sidonii Carmen Astrologicum, tr. and ed. Pingree, D., Leipzig, 1976, p.╯322, cited by Beck, 2004, p.╯172. Jews writing in Hebrew utilised the terms ro’sh or rather zanav hat-tᵉli or hat-tannin for raʾs and dhanab, whereas it was known in the Byzantine tradition as hē kephalē or hē oura tou drakontos. See Schlüter, 1982, p.╯138.

the dragon and astrology

137

“good” luminaries, the Sun and the Moon, were removed from the category of the seven planets whose intrusion brought injustice into the world.46 Consequently the Sun and the Moon were substituted by two “demonic” opponents, the head and tail of the dragon (Pahl. gōchihr which stems from the Avestan gao chithra, “holding the seed of cattle,” formerly the stock epithet of the Moon47).48 According to the Bundahishn, Gōchihr is portrayed as “similar to a snake with the head in Gemini (dō-pahikar) and the tail in Centaurus (nēmasp), so that at all times there are six constellations between its head and tail.”49 In contradistinction to the original meaning of gao chithra, the light and fecundity attribute of the Moon, the dragon’s head (gōchihr sar) and tail (gōchihr dumb) came to represent the demon of eclipses that intercepts the light of the luminaries, the personified dark principle and direct antagonist of the luminaries.50 This led to the concept of a polarity of good and evil throughout the cosmos, the eclipse demon being referred to as Dark Sun and Dark Moon, “dark” meaning “obscured,” and “eclipsed.”51 Thus, according to the Bundahishn, the serpent-like (mār homānāg) Gōchihr and Mūsh Parīg, with tail (dumbōmand) and wings (parrwar), are said to be the evil opponents of the stellar constellations and are therefore bound to the Sun’s path to restrain their capacity to cause harm.52 The expulsion of evil from the sky is manifested by the plunging to earth of Gōchihr,53 who sets the earth on fire and whose permanent body will only be destroyed by resurrection.54 The eclipse dragon also played a part in Manichaeism as Anabibazon and Katabibazon,

the head and tail of the dragon.55 According to Manichaean cosmogony the two dragons were hung up and fettered in the lowest heaven and two angels, male and female, were placed there to cause them to revolve continuously.56 In the history of ancient Indian astronomy throughout the pre-Siddhāntic period, only Rāhu (the grahaṇa, “seizer”) was held “responsible” for causing eclipses by devouring the Sun and Moon. Ketu (the tail of the dragon), understood as a planet57 that generates comets with its fiery tail, is first mentioned in the Atharvaveda (19.9.8-10).58 Both Rāhu and Ketu appear in the great epic Mahābhārata (1.5.15–7), in which the demon Rāhu allied himself with the celestial gods in the struggle against the world serpent, Ananta. After the victorious event, he assumed a disguise and thus succeeded in drinking from the most beneficial of substances, the amṛta (lit. “non-dying”; Av. haoma, Vedic Skt. sóma)59 drink containing the miraculous herb of immortality. But the Sun and the Moon having detected his deception denounced him to the gods, whereupon Vishnu swiftly threw his discus (sudarśanacakra) and severed Rāhu’s head. However, the drink had already produced its effect so that his head and tail both survive, immortalised, as invisible planets and intransigent enemies of the luminaries. As a consequence, the Sun and the Moon are periodically – in symbolic terms – “swallowed” or “disappear in” the vengeful monster that thus causes solar and lunar eclipses.60 In the later, “scientific” phase, when Indian notions were transmitted to Western Asia and the wider Iranian world, the two parts of the

46 ╇ Khareghat, 1914, p.╯129; Brunner, “Astronomy and Astrology in the Sasanian Period, s.v. Astrology and Astronomy in Iran,” EIr, pp.╯862–8. 47 ╇ Zaehner, 1955, repr. 1972, p.╯164, n. E; MacKenzie, 1964, p.╯515, n. 26. 48 ╇ Bundahishn ch. 5, A. 5, pp.╯52.12–53.1, cited after Â�Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr, pp.╯862–8. Cf. Hartner, “Al-Djawzahar,” EI² II, 501b. 49 ╇ Bundahishn ch. 5 A. 5. Skjærvø, “Aždahā I,” EIr. Cf. Zaehner, 1955, repr. 1972, p.╯164, n. E. Also MacKenzie, 1964, pp.╯515, 525. 50 ╇ Hartner, 1938, p.╯153. Cf. Duchèsne-Guillemin, 1990, pp.╯17–9. 51 ╇ Bundahishn (ch. 5.4, p.╯49.13–5) and the late ninthcentury catechism Shkand-gumānīg wizār (“Doubt Dispelling Exposition”) 4.46, cited after Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr. 52 ╇ “[The sun’s opponent, the “tailed Mūsh Parīg”] is tied to the sun’s chariot but occasionally becomes loose and does great harm”; Bundahishn ch. 5.4.5 A. 6–7, pp.╯50.6–7,

53.1–5, and Shkand-gumānīg wizār 4.46, cited after Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr. Cf. Zaehner, 1955, repr. 1972, p.╯164, n. E. MacKenzie, 1964, pp.╯513, 516; Hartner, 1938, p.╯151. 53 ╇ Bundahishn 34.17, p.╯225.1–3, cited after Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr. 54 ╇ Bundahishn 30.31, cited after Khareghat, 1914, p.╯128. 55 ╇ For instance in the Coptic Kephalaia (ch. 69), cited after Beck, 2004, pp.╯177–8. 56 ╇ Boyce, 1975, p.╯60 text y 1 with note; Skjærvø, “Aždahā I,” EIr. 57 ╇ For the evolution of the meanings of “ketu,” see Â�Hartner, 1938, pp.╯152–3. 58 ╇ Scherer, 1953, pp.╯101–3, for further names of Rāhu and Ketu, see esp.╯pp.╯102–5. Cf. Markel, 1995, pp.╯56, 65; Santoro, 2006, p.╯547. 59 ╇ For a brief resumé on the discussion of the etymology of the term amṛta, see Long, 1976, pp.╯181–2, n. 22. See also Janda, 2010, pp.╯29, 55. 60 ╇ Hartner, 1938, p.╯131, and idem, “Al-Djawzahar,” EI² II, 501b.

chapter eight

138

eclipse monster were identified with the lunar nodes, which play a crucial role in the eclipses.61 At the beginning of celestial motion the head or forepart, Rāhu, that is to say the ascending node of the Moon’s orbit upon the ecliptic, was in Gemini and Ketu, the tail or hindpart of the bisected serpent-monster, in other words the descending node, was in Sagittarius (al-qaws, lit. “bow”), often represented as an armed centaur.62 The 180° extent of the dragon reflects the fact that the nodes occupy diametrically opposite points of the ecliptic. Hence the dragon’s body is conceived as arched across the sky.63 The demon of the eclipses Rāhu is well-known not only in the Brahmanic tradition, but also in Buddhism.64 In the Buddhist Jātaka stories of the previous births of Gautama Buddha, which were familiar throughout the Central Asian region, repeated reference is made to the Moon gripped between Rāhu’s jaws, or being liberated from the latter.65 Thus, in the Gandhāra Jātaka, the king of Gandhāra chose to become an ascetic after observing a lunar eclipse, explaining that: Taking the moon’s orb seized by Rāhu as my theme I forsook my great kingdom and took the religious life.

Candimā-sutta, “Discourse on the Moon”) the Buddha reprimands Rāhu and directs him to release the Moon at once which Rāhu does, knowing that otherwise his head will be split into seven pieces.69 The Buddha thus delivers the Moon (the god dwelling in the Moon), who had appealed to him for refuge, from Rāhu’s clutches.70 The contextual and conceptual metamorphoses of the motif thus attest to a mechanism of continuity of these essential thought systems which governed the Central Asian world and beyond. Yet even when the scientific causes were clear, the mythological interpretation of the phenomenon survives. This fact and the ensuing syncretism is expounded by Hartner: We might suppose that clear insight into the physical causes of eclipses could have thrown mythological tradition into the background. But this has not been the case. What we observe is that mythological and astronomical elements contract an intimate fusion. The nodes of the moon’s orbit are simply identified with the eclipse monster itself: with the Hindus, Rāhu becomes the ascending, Ketu the descending node; with Persians and Arabs, the head and tail of the Â�Djawzahr play the same role.71

In the Buddhacarita Siddhārtha’s son is called Rāhula “with the face of Rāhu’s adversary.”67 It is of note that in some Pali texts the demon Rāhu is said not to devour the Sun and the Moon, “but merely to caress them with his hand.”68 In the well-known story of the Candrasūtra (Pali

In Islamic astronomy the Persian gōchihr, called al-jawzahar or al-tinnīn (also azhdahā, “the giant dragon”), was sometimes represented as a bipartite or double-headed dragon. It is the circumpolar constellation Draco, “represented as a very long serpent with many convolutions; it is coiled around the north pole of the ecliptic,”72 which is sometimes metaphorically applied to the Milky Way.73 The Viṣṇudharmottara Purāṅa (III.67) describes how the first sacred waist band or girdle (ayyaṅga which is closely related to the Iranian aiwiyaonghen that is worn by every devout Zoroastrian) was presented to the sun god

61 ╇ Cf. Hartner, 1938, p.╯131, and idem, “Al-Djawzahar,” EI² II, 501b. 62 ╇ Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr, II, p.╯867. The dragon head projecting from the centaur’s tail represents the descending node’s exaltation in Sagittarius; however, although the latter is the dragon’s “tail” (dhanab) and not its “head” (raʾs), and hence the representation of the “head” is an iconographic inconsistency, it has come to symbolise the astrological association. Hartner, 1973–4, p.╯110. 63 ╇ Bundahishn 5.4, p.╯49.13–5, cited after Brunner, “Astronomy and Astrology in the Sasanian Period,” EIr, p.╯867. 64 ╇ Cf. Santoro, 2006, p.╯547. 65 ╇ Cf. eadem, p.╯547. 66 ╇ Cowell, 1897, repr. 2000, vol.╯3, pp.╯222–3.

67 ╇ Johnston, 1936, repr. 2004, p.╯29. Cf. Santoro, 2006, p.╯547. The concept is evoked in yet another line: “Deliver Rāhula from grief for his parent as the full moon from eclipse by Rāhu.” See Johnston, 1936, repr. 2004, p.╯129. 68 ╇ Malalasekera, G.P., Dictionary of Pāli Proper Names, 2 vols., 1938, repr. 1974, vol.╯2, pp.╯735–7, cited after Strong, 1992, p.╯156. 69 ╇ Idem, p.╯156. The Candrasūtra was also translated into ancient Uighur, see Zieme, 2000, pp.╯65–80. 70 ╇ Waldschmidt, 1970. 71 ╇ Hartner, 1938, p.╯131. 72 ╇ Al-Bīrūnī, Kitāb al-Tafhīm, tr. and ed. Wright, 1934, p.╯71. 73 ╇ MacKenzie, 1964, pp.╯521–2, n. 53, 525. The Mystical and Visionary Treatises of Suhrawardi, tr. Thackston, 1982, p.╯113, n. 42.

because: There is the moon’s pure orb become dark by trouble from outside; now this kingdom is a trouble to me: I will take the religious life so that the kingdom does not make me dark as Rāhu does the moon’s orb.66

the dragon and astrology

139

by the king of serpents, Vāsuki, and represented the starry band of the Milky Way.74 In a verse of the late eleventh-century Iranian poet Labībī the seven heads of the dragon represent the heavenly spheres and the universe.75 Together with the acculturation of astronomical knowledge, astrological iconography emerged in the form of visual conceptualisations that were regularly featured in medieval imagery. These were emblematically transferred onto architectural sculpture as well as portable objects, in particular metalwork and ceramics. Astrological considerations also had a profound bearing on the artistic conventions of the iconography of the dragon. Its representation in medieval Islamic astrology has been addressed in a number of studies, foremost among which remains Hartner’s, in which he demonstrates the influence of the conceptualisation of the two “lunar nodes” (al-ʿuqdatāni) on Islamic artisans.76 As seen in sources that pre-date the Islamic period, the crucial aspect of al-jawzahar is that it consists of two nodes of the Moon’s orbit or “points at which (the) two [great] circles of the sphere intersect,”77 in other words the two points where the course of the Moon crosses the plane of the ecliptic from south to north: the “head of the dragon” (raʾs al-tinnīn) is formed by the ascending node of the Moon’s orbit, and, correspondingly, the “tail of the dragon” (dhanab al-tinnīn) by the descending node.78 This associates it with both solar and lunar eclipses; the latter were attributed to the occurrence of a conjunction, or opposition, of the Sun and Moon (New Moon or Full Moon, respectively) in or near the lunar nodes. In medieval Islam especially, the astrological influence of the eclipse of the Sun is considered one of the foremost signs of the impending destruction of the world.79 In the Qurʾān the latter is repeatedly described among the signs of perturbations of heavenly bodies (sūra 75, 8–9; and 81, 1). Early Islamic tradition frequently describes the occurrence of an eclipse of the Sun during

the period of the Prophet Muḥammad’s residence in Medina.80 On this occasion the Prophet is reported to have said “that he had never been so greatly filled with fear,” and the commentaries add to this, “that he thought that the Hour had come, and to illustrate it he is reported to have said that he saw Paradise and Hell so close to him that he could have gathered a bunch of grapes from the land of the blessed, had he stretched out his hand.”81 According to Abū Saʿīd al-Sijzī’s comprehensive tenth-century astrological compilation on the “Conjunctions and Revolutions of the Years of the World” which was based on earlier sources, a solar eclipse indeed indicated the death of Muḥammad as well as the accession of the first caliph Abū Bakr.82 The significance accorded to the eclipse is reflected in the bi-partite “dragon” that was seen as temporarily “devouring” the Sun and the Moon at certain irregular intervals, and then disgorging or “delivering,” them – since the two planets always appear to emerge unscathed from their temporary eclipse by the “dragon.” This nonPtolemaic concept played a prominent role in astrological associations whereby the two nodes were treated as though they were real celestial bodies, in other words extra, albeit invisible, “planets,” or fictitious nodes.83 They were conceived as an eighth and a ninth planet, the only difference between them and the original seven planets being that contrary to the others their movement was westwards or “retrograde,” rather than eastwards.84 Ghaznawid and Ghurid military campaigns in India brought not only extensive booty, but resulted at the same time in an influx of scholars, craftsmen and a variety of artisans, all of whom came with their own indigenous iconographies, contributing perhaps to the diffusion of the iconography of al-jawzahar. The great scholar Abū Rayḥān Muḥammad ibn Aḥmad al-Bīrūnī had accompanied sulṭān Maḥmūd, possibly as official astrologer, on several of his military expedi-

╇ Carter, 1981, p.╯80 and n. 27. ╇ M. Dabīrsīāqī, Ganj-i bāz yāfta, Tehran, 2535/1355 Sh., 1976, cited after Khāleqī-Moṭlaq, “Aždahā II,” EIr. 76 ╇ Hartner, 1938, and idem, “Al-Djawzahar,” EI² II, 501b. Cf. Öney, 1969a, pp.╯193–216; Otto-Dorn, 1978–9, pp.╯125– 36; Azarpay, 1978, pp.╯363–74. 77 ╇ See the definition of al-jawzahar in Abū ʿAbd Allāh MuḥamÂ�mand al-Khwārizmī’s Mafātīḥ al-ʿUlūm, cited after Hartner, 1938, p.╯120. 78 ╇ Cf. Kharegat, 1914, pp.╯126–8; MacKenzie, 1964, p.╯515.

╇ Wensinck, 1923, p.╯193. ╇ Al-Bukhārī, Mawāqīt al-Ṣalāt, b. 51, cited by Wensinck, 1923, p.╯193 and n. 6. 81 ╇ Al-Bukhārī, Kitāb al-ʿItq, b. 3, cited by Wensinck, 1923, p.╯193 and n. 7. 82 ╇ Al-Sijzī, Kitāb al-qirānāt wa taḥāwīl sinī al-ʿālam, as cited in Pingree, 1968, pp.╯70–127, 118–9 (the table of the horoscope of the solar eclipse foreboding the death of the Prophet). 83 ╇ Beck, 2004, p.╯161 and n. 29. 84 ╇ Idem, p.╯161.

74 75

79 80

140

chapter eight

tions to northwest India.85 In his Kitāb al-Tafhīm li-awāʾil ṣināʿat al-tanjīm (“Book of Instruction in the Elements of the Art of Astrology”), written upon his return to Ghazna in 422/1031, he refers to the two fictitious nodes, the eighth and a ninth planet, as knot (ʿuqda) and point of Â�crossing (majāz).86 In spite of his statement that “they are not real planets,” the same author does however record the position of the raʾs al-tinnīn and the dhanab al-tinnīn in the various astrological tables included in his texts.87 Much earlier, in the work On The Great Conjunctions, or the Aḥkām Taḥāwīl Sinī al-Mawālid, the astrologer Abū Maʿshar (d.╯272/886) had already referred to the points of exaltation for the nodes of the Moon which for the dragon’s head is in Gemini 3°, and for the tail in Sagittarius 3°.88 The “node of the Moon’s orbit” however is an integral part of the iconography of the eclipse monster, portrayed as loop or twisted knot, sometimes visualised as a pretzel- or heart-shaped knot. This is reflected in the symbolism of the personification of comets, Ketu, visualised on the navagraha reliefs that represent the nine Indian planetary deities, which are similarly illustrated with a human torso and a serpentine tail terminating in a knot.89 The earliest surviving representation of Rāhu and Ketu in India is carved on a navagraha lintel from Uttar Pradesh, dating from c. 600 or slightly later, in which Ketu is represented as a half-ophidian figure sitting on his coiled serpentine tail beside the cephalic Rāhu (fig.╯140).90 Individual depictions of jawzahar – Draco as eighth planet next to the seven traditional planets, comprising the Sun, the Moon, Saturn, Jupiter, Mars, Venus and Mercury,91 often portray a crosslegged figure holding a dragon in each hand. The

figure is shown to hold either a pair of upright dragons, their bodies forming a loop,92 or vertical staffs (figs. 113–116); both the coiling bodies and the allegorical staffs end in confronted dragon heads with gaping snouts. It is significant that representations of the planet jawzahar thereby make use of the emblematic portrayal of the cosmic ruler mentioned earlier, framed by dragon-headed staffs, ubiquitously employed on visual art from the mid-eleventh to early thirteenth century and associated with the ancient concept of the “Master of the Dragons.” The choice of this cosmic symbolism underlines the prominence accorded to jawzahar which gives an indication of the magnitude of the potential effects the planet could have on the course of human events. The conception of the central figure as dragon-tamer thereby perhaps reflects the apparent necessity to harness the forces of this planet. The planetary character of the two nodes of al-jawzahar indicates that they are traversing the plane of the ecliptic. With respect to the signs, the planets have a “domicilium” as well as a place of exaltation (sharaf) and dejection (hubūṭ). These astronomical terms respectively define the points most distant from and closest to the earth, especially in reference to the elliptical orbit of the Moon. In astrology they relate to the point of maximum and minimum influence of one of the seven traditional planets and of the two nodes of al-jawzahar when they find themselves in association with one of the constellations visualised as one of the twelve signs of the zodiac. In the Irano-Turkish territories, the eclipse pseudo-planet (al-jawzahar) is often shown at the point of exaltation of its head or tail in Gemini, as for instance on a silver- and copper-inlaid brass ewer from Herat, formerly in the Nuhad

╇ Boilot, “al-Bīrūnī (Bērūnī), Abu ’l Rayḥān Muḥammad b. Aḥmad,” EI² I, 1136a. 86 ╇ Tr. and ed. Wright, 1934, pp.╯91–2. 87 ╇ Idem, pp.╯255, 258. 88 ╇ Hartner (1938, p.╯133 and n. 30) refers to De magnis coniunctionibus, the Latin version translated by Johannes Hispalensis, printed at Augsburg in 1489 (repr. Venice, 1515), which contains a chapter dealing with the planetary influence of the nodes as a figure of the “dragon” with its head and tail twisted around two nodes (reproduced in idem, fig.╯10). Cf. al-Bīrūnī’s references in his Kitāb al-Tafhīm, tr. and ed. Wright, 1934, p.╯358. 89 ╇ Ketu’s serpent tail is alluded to in the Agnipurāṇa; see De Mallmann, 1962, p.╯86. 90 ╇ The earliest western Indian representation is found on a fragmentary lintel from Alwar district in Rajasthan, which

probably dates from c. 600 to 650 (Markel, 1995, fig.╯29; Government Museum, Alwar). For later depictions, see also Hartner, 1938, pp.╯134, 138, figs. 6–8. For a discussion of the navagraha reliefs, see Pingree, 1964–5, pp.╯249–67; Markel, 1995, pp.╯19–68 and 129–76. 91 ╇ Hartner, 1938, pp.╯114–38. In later medieval Indian Â�literature both nodes, Rāhu (raʾs al-tinnīn), and Ketu (dhanab al-tinnīn), were attributed the same importance as the other seven planets, hence there were a total of nine planets; idem, p.╯133, cf. also p.╯151. 92 ╇ For instance, on a late twelfth- or early thirteenthcentury copper alloy inkwell inlaid with silver from Western Central Asia (Pugachenkova and Rempel’, fig.╯196, and fig.╯197, line drawing), or on a thirteenth-century silverinlaid copper alloy candlestick from Mesopotamia (Baer, 1983, p.╯256, fig.╯208).

85

the dragon and astrology

141

Es-Said Collection, now in the National Museum of Qatar in Doha (fig.╯141).93 In Islamic tradition, the planetary eclipse in Sagittarius is generally rendered as a centaur taking aim with a bow at its long dragon-headed tail and shooting an arrow into the dragon’s mouth. On the Qatar ewer the sign is accordingly portrayed as the protome of a winged quadruped dragon with tongue protruding from the gaping mouth rising from the looped tail (fig.╯142). A sculptural example of the planetary eclipse in Sagittarius is depicted among eight astrological reliefs carved onto the pillars of the Tigris bridge, near the city of Jazīrat ibn ʿUmar (present-day Cizre), Anatolia. Among the reliefs is the upright knotted protome of a dragon with gaping mouth and curled-up snout tip, oriented towards the figure of a centaur shooting with a bow and arrow into its mouth. The bridge was commissioned by the wazīr of Mosul, Jamāl al-Dīn Muḥammad al-Iṣfahānī. The wazīr’s imprisonment in 558/1163 provides a terminus ante quem for the construction of the bridge and its astrological relief sculptures.94 As pointed out by Hartner, the reliefs are one of the earliestknown sculptural examples “in which the Islamic artist obviously grants the same rights to one or both of these pseudo-planets as to the seven real ones, while in India this had been the rule centuries before.”95 Similarly, depictions of al-jawzahar threatening the Sun and the Moon, or their respective zodiacal animals, the lion and the crab, became prevalent in the decorative programmes of objects, as evidenced in the depictions on the same ewer (figs. 143 and 144). The importance of the eclipse pseudo-planet is such that most of the roundels on this ewer show the signs of the zodiac and planets

inhabited with monster heads with long floppy ears growing from scrolling tendrils (as shown for example in the depiction of “Moon in Cancer”) that Hartner identifies with “the dragon progeny threatening the luminaries or, vicariously, their domicilia and exaltations.”96 However, he has qualified the astrological interpretation suggesting that “in all probability, various elements – astronomical, astrological and mythological – were here fused in one.”97 This shows that the astrological veracity of such details was less important than their exemplary significance. It is moreover noteworthy that the monster heads with long floppy ears issue from vegetation, which for its association with fertility generally has positive connotations. It may therefore be hypothesised that the depiction of “dragon progeny” together with the signs of the zodiac and the planets was intended to have a beneficial influence on the paths of the luminaries. A related large silver-inlaid brass ewer, housed in the Georgian State Museum, Tbilisi, Georgia, which is decorated on the shoulder with the signs of the zodiac and the planets, bears inscriptions that not only give the name of the maker, one Maḥmūd ibn Muḥammad al-Harawī and the date, the month of Shaʿban 577/19 December 1181–17 January 1182, but also state that the “seven heavenly bodies, however proud they may be, are protection for the one who works so.”98 As James Allan states, “from this poem it is evident that the inlayer saw those images as protection for himself against evil.”99 It further underlines the overall magical and prophylactic quality that was ascribed to the iconography of the signs of the zodiac and the planets. The Â�frequency of their

93 ╇ Jawzahar at the points of exaltation of its head or tail in Gemini is also depicted at the top of the lid of the penbox from Iran (signed by Maḥmūd ibn Sunqur on the hasp and dated 680/1281–2; on this penbox see also pp. 96, 98 and fig. 93), which bears roundels in three groups of four containing symbols of the zodiac with their ruling planets. Cf. Hartner, 1938, p.╯138 (misprint of dates), figs. 14 and 15 (roundel representing “Gemini”); idem, 1973–4, pp.╯115 and fig.╯9 (left); Pope and Ackerman, eds., 1938–39, repr. 1964– 81, vol.╯13, p.╯1336.B; Barrett, 1949, pl.╯33 top; Legacy, 2002, cat. no. 158, fig.╯46. The same motif is shown on an early thirteenth-century copper alloy inkwell, inlaid with silver, attributed to Iran or Syria, now in the Metropolitan Museum of Art, New York; cf. Baer, 1983, p.╯257, fig.╯209 (detail of p.╯79, fig.╯59). 94 ╇ Meinecke, 1996, p.╯60. On the reliefs of Jazīrat ibn ʿUmar, see Preusser, 1911, pl.╯40; Hartner, 1938, p.╯134 and fig.╯2 (photograph at bottom left), and idem, 1973–4, pp.╯108, 110; Gierlichs, 1996, pl.╯47.4. A Sagittarius rendered as a centaur shooting an arrow towards his own tail which terminates in a dragon head is depicted in a Georgian

astrological treatise, probably illustrated under Islamic influence in 1188; Amiranašvili, 1966, pl.╯56. See also p. 19, n. 36. A comparable figure is represented as centaurarcher shooting an arrow backwards at the dragon head emerging from its tail on the coinage of the Artuqid ruler of Mardin, Nāṣir al-Dīn Artuq Arslan ibn Il Ghāzī (599/1203–637/1239); Roxburgh, ed. 2004, p.╯398, cat. no. 86; What the Coins Tell Us, 2009, p.╯102; Hauptmann von Gladiss, ed., 2006, pp.╯107–8, figs. 15, 16. The same emblem also figures on the coinage of the ʿAbbasid caliph al-Nāṣir (577/1181–620/1223); eadem, 2006, p.╯107, cat. no. 15. 95 ╇ Hartner, 1938, p.╯132. 96 ╇ Idem, 1959, pp.╯237–9, and idem, 1973–4, p.╯112, 118. 97 ╇ Idem, 1973–4, pp.╯112–3. 98 ╇ Loukonine and Ivanov, 2003, pp.╯116–7, cat. no. 117 (with the inscription in Arabic and English). Cf. Allan, 1982a, repr. 1999, p.╯49; Atil, Chase, and Jett, 1985, p.╯17, fig.╯6. The translation given is based on the rendering by Allan. See also p. 36, n. 6. 99 ╇ Allan, 1982a, repr. 1999, p.╯49.

142

chapter eight

depiction, in particular on portable items, moreover emphasises the prodigious cultural significance of the signs of the zodiac and the planets in the medieval Islamic world. The prominent depiction of al-jawzahar on objects such as these Herati ewers “evidently originates,” as Hartner has underlined, “not in a doctrinal astrological conception, but in a purely metaphysical, one,” being associated with “the antagonism between the celestial luminaries and the terrestrial lightdevouring dragon.”100 As mentioned, a solar or lunar eclipse (al-kusūf) can occur only when the Moon is at one of the points of crossing (majāz), or nodes. In his Kitāb al-Tafhīm al-Bīrūnī, moreover, notes that the latter are perceived as having separate natures, the head being hot, auspicious, and indicating increase of wealth etc. and the tail being cold, bringing misfortune, and indicating diminution of wealth, etc.101 In addition he records the information that “some people say that the dragon’s head is male and diurnal and the tail female and nocturnal.”102 From about the twelfth-century symbolic persoÂ�nifications of Sol and Luna, often shown together with the dragon motif, were widely applied to portable objects, especially on metalwork, from greater Khurasan to the Anatolian region. By virtue of its very characteristic as an eclipse dragon al-jawzahar was directly linked to the Sun and the Moon. The two luminaries are among the representations of the eight planets (the pseudo-planet jawzahar is here represented as eighth “planet”) on the lid of a covered copper alloy bowl, known as Vaso Vescovali, made in the Khurasan region in about 1200.103 A threefaced Sun, akin to the one featured on the Qatar ewer (fig.╯143), surmounts a winged figure who, seated on a pointed support and holding up the luminary’s “dais,” is symmetrically flanked by two confronted attendants behind whom long-eared jawzahar-like heads grow out of stems which curl behind their waists (fig.╯145). The Moon consists of a human figure holding up with its four arms (an image probably informed by Indian proto-

types) a large crescent which frames the entire upper body, while squatting on a “dais” supported by quadruped protomes, probably horses. The addorsed attendants are related to those of the Sun but are clad in more angular attire with the jawzahar-like heads growing from their waists (fig.╯146). Significantly, as Hartner has observed, “the scene has no menacing character.”104 A remarkable depiction of a personified Moon in Cancer is shown on the so-called “Wade Cup,” dated 596/1200–622/1225, now in the Cleveland Museum of Art, featuring a winged figure holding up a crescent moon and with splayed legs surmounting the crab; the legs of the moon figure are held in the claws of the crab and both are flanked on either side by long-eared dragon progeny that springs from the base (fig.╯147).105 The personifications of the Sun and the Moon are also featured above a pair of addorsed knotted dragons, serving here as support for the luminaries, as part of a decorative programme on a large copper alloy basin inlaid with silver of the thirteenth-century atābeg Badr al-Dīn Luʾluʾ (618/1222–657/1259) of Mosul (fig.╯148).106 The depictions reveal an interest in the translating of entities beyond the domain of humankind, such as the two luminaries, into human guise.107 In this context it is interesting to recall that this also corresponds to the Turkish tradition of conceiving the two great luminaries as living beings.108 The selective visualisation of the Sun and the Moon and the menace posed to them in the form of solar and lunar eclipses, ascribed to al-jawzahar, is related to the daily relevance afforded to the two luminaries in human affairs and existence.109 While the dragon is mainly associated with the eclipses and, hence, the “devouring of light,” its positive aspect as giver of light and, consequently, as protector of light is often more difficult to gauge although references are found in Iranian poetry.110 The polymath Asadī Ṭūsī accordingly writes in his epic Garshāsp-nāma:

╇ Hartner, 1938, p.╯138. ╇ Tr. and ed. Wright, 1934, p.╯233. 102 ╇ Al-Bīrūnī introduced these concepts into Muslim literature, though not without misgivings as to their veracity, qualifying this information as “quite illogical” (idem, p.╯234). 103 ╇ Cf. Hartner, 1973–4, p.╯119; Ward, 1993, p.╯79. 104 ╇ Hartner, 1973–4, p.╯119. 105 ╇ Rice, 1955, pl.╯VII b; Hartner, 1959, p.╯235, fig.╯4.

106 ╇ Cf. Saxl, 1912, p.╯164 and fig.╯10; Sarre and van Berchem, 1907, pp.╯22, 27, figs. 1 and 13. 107 ╇ Pancaroğlu, 2000, p.╯197. 108 ╇ Roux, 1979, p.╯179. 109 ╇ Pancaroğlu, 2000, p.╯204. 110 ╇ Cf. Daneshvari, 1993, p.╯20. 111 ╇ Asadī Ṭūsī, Garshāsp-nāma, pp.╯475–6, cited after Daneshvari, 1993, p.╯21.

100 101

the dragon that gives the sun also takes it back by its poison.111

the dragon and astrology The simile “the sun is delivered from the dragon” in the romantic epic, Wīs u Rāmīn,112 almost certainly of Arsacid Parthian origin, expresses a related stance. It was translated and versified by Fakhr al-Dīn Gurgānī around 1050 for the first Saljuq sulṭān Ṭoghrıl I, his minister Abū Naṣr ibn Manṣūr, and his governor Abu ’l-Fatḥ ibn Muḥammad of Iṣfahān. The eleventh-century Iranian poet Masʿūd-i Saʿd-i Salmān (c. 440/1046–7–c. 515/1121–2) whose family came from Hamadān, enjoyed status and fame at the Ghaznawid courts of Lahore and Ghazna in his youth and again in his later years. But he also suffered the misery of some eighteen years of incarceration, resulting in the prisonpoetry (ḥabsiyya) for which he is renowned and in which he metaphorically employs both fire and dragon imagery: My heart has become like a fire temple, fearing it I don’t breath even for a moment, until from the heat of my dragon-like heart my mouth fills with fire.

143

break the resistance of the priestess and when she is brought in front of Iskandar, Balīnūs declares “this black dragon is the moon [a moon-faced beauty]” a pun that implicitly also refers to the dragon’s association with light and, by inference, perhaps his implicit role as the deliverer and protector of light.116 The esoteric conceptualisation of the dragon is illuminated in the allegory of a hero’s spiritual journey in A Tale of Occidental Exile written by the mystic Shihāb al-Dīn Yaḥyā Suhravardī (d.╯587/1191): If you desire to be delivered along with your brother [i.e., speculative reason, the guide (ʿāṣim)], do not put off traveling. Cling to your rope, which is the dragon’s tail (jawzahr) of the holy sphere that dominates the regions of the lunar eclipse [the realms of the eclipse denoting the world of ascetic practice].117

However he emerges from the dragon’s clutches “like a cool cypress in a garden”113 thereby employing the conventional metaphor which implies that he comes forth unscathed from an eclipse or other impending calamity. Comparable imagery governs Niẓāmī Ganjawī’s description in the first part of his Iskandar-nāma, the Sharaf-nāma, of Iskandar’s destruction of the fire temples of the Iranian Zoroastrians during his conquest of Iran (an action for which the historical Alexander is not responsible but that perhaps reflects Niẓāmī’s vague memory of an Iranian religious resistance to Hellenism).114 Iskandar arrives at a fire temple dominated by a powerful priestess, Āẓar Humā, who transforms herself into a fire-breathing black dragon to guard the holy fire of the temple, hence implying that the dark dragon protects the fire and therefore the light and, by association, the luminaries.115 Through his talismanic powers Balīnūs (Apollonius) helps to

The hero passes beyond the material world and reaches a light, the active intellect, which is the governor of this world. He places the light in the mouth of the dragon, the world of the elements, that “dwelt in the tower of the water-wheel [i.e., the sky which turns like a wheel], beneath which was the Sea of Clysma [i.e., the water below the sky] and above which are the stars the origin of whose rays was known only to the Creator and those “who are well-grounded in knowledge.”” 118 The metaphysical aspect of the bi-partite dragon is further evoked in a passage of the fables and anecdotes of the early thirteenth-century Marzubān-nāma with the allegorical allusion, “at dawn, when the black snake of night cast the sun’s disc out of the mouth of the east,”119 hence once again implying a double-headed dragon delivering the luminary and the creation of light. The luminary aspect of the dragon is also reflected, as Abbas Daneshvari has pointed out, by its flanking the mount of finger rings (fig. €30).120 In Farīd al-Dīn Aṭṭār’s Ilāhī-nāma (“Book of God”), the magic signet ring of Solomon, an

╇ Tr. cited after Daneshvari, 1993, p.╯21. ╇ Dīwān-i ashʿār-i Masʿūd-i Saʿd, qaṣīda 205, tr. Sharma, 2000, pp.╯94–5. On Masʿūd-i Saʿd-i Salmān, see also Rypka, 1968, p.╯196. Cf. the early Indian conception as expressed in the Vedic myth in which, after his defeat by Indra, the dragon Vala (valá-, meaning “enclosure”) frees the goddess of dawn, Ushas, whom he had imprisoned. See p. 87, n. 4. Janda, 2010, pp.╯45–70, esp.╯27, 63, 79, 266, 270. 114 ╇ Cf. Stoneman, 1991, p.╯2. 115 ╇ Niẓāmī, Sharaf-nāma, ed. Dastgardī, V., Tehran, 1936, p.╯244, cited by Daneshvari, 1993, p.╯21; Niẓāmī, Dīwān,

pp.╯974–5, cited in Pseudo-Apollonius of Tyana, tr. and ed. Weisser, 1990, p.╯27. 116 ╇ Ed. Dastgardī, V., Tehran, 1936, p.╯244, cited by Daneshvari, 1993, p.╯21. 117 ╇ The Mystical and Visionary Treatises of Suhrawardi, tr. Thackston, 1982, p.╯102 and ns. r and s. 118 ╇ Idem, p.╯105 and ns. uu, vv, ww. 119 ╇ Tr. Levy, 1959, p.╯51. Cf. Saʿīd al-Dīn Warāwīnī, Marzubān-nāma, ed. Rūshan, M., 2 vols., Tehran, 1978, pp.╯96–7, cited by Daneshvari, 1993, pp.╯20–1. 120 ╇ Daneshvari, 1993, p.╯21.

112 113

chapter eight

144

allusion to the sun, is guarded by a dragon.121 By analogy, dragons flanking the cosmic ruler (figs. 113–116), further increase the luminary symbolism of the ruler.122 Even more explicitly, Abu ’l-Ḥasan Farrukhī, the celebrated court poet of Sīstān (Seistan), links the luminous sun with the roaring dragon when he praises the character of his patron:

Reminiscences of ancient cosmogonical notions may be gauged from Armenian lore recorded by Khorenatsi relating to the Median king Astyages, the Armenian arch-enemy referred to as Azhi Dahāka, the archetype of evil misrule, whose first wife, Anoysh, was called the “mother of the

dragons.”124 Her name, Anoysh, however literally signifies “immortal, luminous, perfumed.”125 Moreover, her association with the monstrous dragon, to whom she gives numerous offspring, recalls certain cosmogonies in which one of the two primordial entities is “infinite light, serene and joyous” and the other “a frightening and dark obscurity, coiled up in twisting spirals akin to those of a serpent.”126 It is also interesting to note that the original meaning of the Sanskrit word ketu is “light,” “clarity” (synonymous with the etymologically related adjective citra of the Pahl. gōchihr), which is in apparent contradistinction to the light-devouring function of Ketu as eclipse demon.127 In Gnostic writings, as will be discussed shortly, the serpent is also associated with both darkness and light.128

╇ Ilāhī-nāma, pp.╯264–5, referred to by idem, p.╯21. ╇ Idem. 123 ╇ Dīwān-i Hakīm-i Farrukhī-i Sīstānī, p.╯363, l. 7354, cited by Melikian-Chirvani, 1997a, pp.╯143–4. 124 ╇ Mahé, 1995, p.╯183. 125 ╇ Acaryan, H., Hayeren armatakan bararan (“Diction-

naire étymologique arménien”), vol.╯1, p.╯206 b (in Armenian), cited by Mahé, 1995, p.╯183. 126 ╇ Poimandres, tr. and ed.€ Nock and Festugière, 1946, p.╯7 and 12, n. 9. 127 ╇ Hartner, 1938, pp.╯152–3. 128 ╇ Cf. Sheppard, 1962, p.╯89.

In your banquets, you are an illuminating sun In combat, you are a roaring dragon.123

121 122

vestiges of ancient dragon iconographies

145

chapter nine

Vestiges of ancient dragon iconographies

a.╇ The encircling dragon The iconography of the circular dragon biting its own tail, traditionally known by its Greek name as ouroboros, was also known in the Islamic tradition. This type of imagery is vividly described in surviving textual sources. The early medieval writer Muḥammad ibn ʿAbdullāh al-Kisāʾī who probably wrote not long before 1200 refers to the authority of Kaʿb al-Aḥbār, a Yemenite convert to Islam (probably in 17/638), when portraying the creation of the Canopy and the Throne of God: Then God created a great serpent to surround the Canopy. Its head is of white pearl and its body is of gold. Its eyes are two sapphires, and no one can comprehend the magnitude of the serpent except God. It has forty thousand wings made of different kinds of jewels, and on each feather there stands an angel holding a jeweled lance, praising God and blessing His name. When this serpent extols God, its exaltation overwhelms that of all angels...1

A related description of the girdling dragon is given by Abū ʿAbd Allāh al-Qurṭubī, the thirteenth-century expert in ḥadīth, or sacred tradition, in a commentary on sūra 40 of the Qurʾān: When God created the Throne, it said, ‘God has not created anything greater than myself,’ and exulted with joy out of pride. God therefore caused it to be surrounded by a serpent having 70,000 wings, each wing having 70,000 feathers in it, each feather having in it 70,000 faces each face having in it 70,000 mouths, and each mouth having in it 70,000 tongues, with its mouths ejaculating every day the praises of God …, the number of drops

╇ Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯7. ╇ Al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Â�Jayakar, 1906, vol.╯1, p.╯638; see also ʿArāʾis al-majālis fī qiṣaṣ al-anbiyāʾ, tr. and ed. Brinner, 2002, p.╯25. 3 ╇ Bet ha-Midrasch, 1853–73, vol.╯5, p.╯35. Cf. Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯4, pp.╯157–9; Wensinck, 1916, repr. 1978, p.╯63. 4 ╇ Bet ha-Midrasch, 1853–73, vol.╯2, pp.╯83–5. 5 ╇ Al-Thaʿlabī, Qiṣaṣ al-anbiyāʾ. Musammā bi ’l-ʿarāʾis al-majālis, Cairo, 1290, p.╯13, as cited in Wensinck, 1916, repr. 1978, p.╯62 and n. 3; see also ʿArāʾis al-majālis fī qiṣaṣ 1 2

of rain, the number of leaves of trees, the number of stones and earth, the number of days of this world, and the number of angels, – all these a number of times. The serpent then twisted itself round the Throne which was taken up by only half the serpent while it remained twisted around it. The Throne thereupon became humble.2

In the Jewish tradition a great silver serpent likewise encircles the machinery of the throne of king Solomon and by operating the wheelwork, activates the mechanism.3 It is of note that Solomon’s mechanical throne, which can be likened to a miniature universe, can only be put into motion by the serpent.4 In his Qiṣaṣ al-anbiyāʾ (“Tales on the Prophets”), Abū Isḥāq Aḥmad ibn Muḥammad ibn Ibrāhīm al-Thaʿlabī al-Nīsābūrī al-Shāfʿī (d.╯427/1035), describes the Kaʿba in Mecca, the central sanctuary of the Islamic world, as a divine throne that is circumscribed by a dragon: Then Allāh surrounded it by a serpent. … this serpent wound itself around the throne and the latter reaches to half the height of the serpent which is winding itself around it.5

In the biography of the Prophet Muḥammad, al-Ḥalabī similarly relates how the serpent that dwells in the pit of the Kaʿba to guard the treasures there, would: ...leave its dwelling place and appear glittering, viz. it exposed itself in the sun upon the wall of the Kaʿba while its colour assumed a glittering appearance; and often it wound itself on the wall so that its tail approached its head [emphasis added].6

al-anbiyāʾ, tr. and ed. Brinner, 2002, p.╯151. Wensinck (1916, repr. 1978, p.╯62 and n. 3) notes that there are also Greek images in which the serpent is wound around and ascends above the omphalos, which often has a sepulchral character (Elderkin, 1924, pp.╯109–16); for a discussion of the omphalos in literature, see Roscher, 1913, pl.╯IX, no.╯6; and idem, 1914, pl.╯I, no. 1, pl.╯II, nos. 3, 4, 14. See also p.╯59 and n. 102. 6 ╇ Al-Sīra al-Ḥalabiyya, Cairo, 1292, vol.╯1, p.╯189, 3–5, as cited in Wensinck, 1916, repr. 1978, p.╯64 and n. 1.

146

chapter nine

In the Islamic tradition (ḥadīth), as Wensinck points out, the description of the serpent is a metaphor for the ocean: ...as the Ocean, the Mekkan [sic] serpent is glittering in the sun and as the Ocean it is black and white.7

The motif of the “serpent whose tail approached its head” is well-known in Semitic cosmography. A key passage in the book of Job (26:12) states:

make covenant with him … This explains the story about the man who saw the huge snake that God had placed around Mount Qāf, which encircles the world. The head and the tail of this snake meet. The man greeted the snake, who returned his greeting and then asked him about Shaykh Abū Madyan, who lived at Bijāya in the Maghrib. The man said to it, “How do you come to know Abū Madyan?” The snake answered, “Is there anyone on earth who does not know him?11

The Pole is both the centre of the circle of the universe, and its circumference. He is the mirror of God, and the pivot of the world. … God is perpetually epiphanized to him. … He is located in Mecca, whatever place he happens to be in bodily. When a Pole is enthroned at the level of the quṭbiyya, all beings, animal or vegetable,

According to a saying of the Prophet Muḥammad, Mount Qāf is separated from the world “by a region which men cannot cross, a dark area which would stretch for four months walking.”12 It was thus a distant, marginal area at the boundaries of the “civilised” world.13 Such liminal regions were often inhabited by demons. Descriptions of dragons and other mythical creatures abound in such regions in the descriptions of medieval Islamic geographical and travel works.14 Their topical proliferation serves as a “cultural marker” (in James Montgomery’s words) indicating to the traveller that he is in a distant land.15 Together with other imaginary hybrid creatures, such as sphinxes and harpies, the presence of the dragon may have signified the outer reaches of the known earth. This vision of the fabulous distant lands at the remote ends of the world is also found in the Alexander Romance of Pseudo-Callisthenes. In this legend many wondrous feats are ascribed to Iskandar who made his way to the furthest west and furthest east, the end of the world, entering the “regions not illuminated by the Sun, the Moon and the stars and light as day” where he encounters creatures such as human-headed birds.16 However, Mount Qāf does not only encircle the earth: it also encloses the ocean which “forms a girdle around the earth.”17 The symbolism also occurs in the story of Solomon of the Alf layla wa-layla which recounts how Solomon on his flyÂ�

╇ Idem, p.╯64. ╇ Wakeman, 1973, pp.╯134–5. 9 ╇ Gunkel, 1895, p.╯47 and n. 1; Wakeman, 1973, p.╯135 and n. 1. 10 ╇ Eadem, 1973, p.╯64. Cf. Grünbaum, 1877, p.╯275. 11 ╇ The same story, in expanded form, of a man speaking to a serpent appears in the Risālat ruḥ al-quds. See Â�Chodkiewicz, 1993, p.╯55 and n. 32. 12 ╇ Cf. Streck [Miquel], “Ḳāf,” EI2 IV, 400a. 13 ╇ In ancient Greek lore the ends of the earth were inhabited by primeval and/or mythical creatures (for instance in Hesiod’s Theogony 270–6). Inaccessible by land, they could only be reached by the crossing of waters, often described as world-encircling. 14 ╇ Cf. Montgomery, 2006, p.╯72. 15 ╇ The fourth marvel of Ibn Faḍlān’s Risālat constitutes

the story of a tree being cut to size which then begins to move and crawl away in the form of a giant dragon. Risālat Ibn Faḍlān, ed. Dahhān, S., Damascus, 1959, pp.╯127–8 (fol. 4 206 wāw), as cited in Montgomery, 2006, p.╯72. Cf. the dangerous and monstrous creatures of Greek lore that dwell at the edges of the earth, the eschatiai, or “most distant lands,” and are very often guardians of treasure, for instance, the golden apples of the Hesperides in the far west and the golden fleece of Kolchis in the far east which are protected by giant serpents; see Romm, 1987, pp.╯45–54. 16 ╇ Pseudo-Callisthenes II, ch. 40, tr. and ed. Stoneman, 1991, p.╯121. Related conceptualisations of sphinxes and man-birds appear in the Kitāb-i Samak ʿAyyār; see Gaillard, 1987, p.╯120. 17 ╇ Streck [Miquel], “Ḳāf,” EI2 IV, 400a.

He has inscribed a circle on the face of the waters at the boundary between light and darkness.

The inscribed circle refers to the line of the horizon, which separates the inhabited world from the waters that surround it.8 These waters are symbolised by Leviathan, “the encircler,” who is primarily a sea monster.9 The name of the biblical monster Leviathan (Hebr. liwyātān) has been derived from lwh suggested by the Arabic lwy “turn,” “twist” and the Assyrian lamû “surround,” “encircle,”10 underscoring the probablility of an original serpent-like nature of Leviathan. The same motif is used by the great Andalusian Arab mystic Muḥyi ’l-Dīn Ibn al-ʿArabī (560/1165–638/1240) whose works draw on many sources, including Gnostic, Hermetic and Neoplatonic works. In his discussion of the Pole (quṭb; an elevated rank of sainthood in ṣūfī mysticism) that represents the living Messenger (rasūl) in the Kitab al-manzil al-quṭb (“Book of the Spiritual Dwelling of the Pole”), he describes an enormous serpent whose head and tail touch and that encircles Mount Qāf:

7 8

vestiges of ancient dragon iconographies ing carpet travels through the world and reaches the dragon that encompasses the world.18 According to a popular belief recorded by al-Qazwīnī, the earth is supported by the biblical monsters Leviathan and Behemoth.19 Later Jewish tradition similarly states that:

there would be no sowing and no growth and no motivation for the reproduction of all creatures. This serpent now stood originally outside the walls of the sacred precincts and was connected from the outside with the outer wall, since its tail was linked with the wall whereas its countenance was oriented inwards. It did not befit it to enter the inside, but its place and law was to affect the creation of growth and reproduction from the outside, and this is the secret of the tree and the knowledge of good and evil.23

...the Ocean surrounds the whole world as a vault surrounds a large pillar. And the world is placed in circular form on the fins of Leviathan.20

Similarly a large serpent is said to encircle the bier of a righteous person, a tradition which provides a microcosmic allegory of the whole world surrounded and supported by a giant serpent.21 It also shows that the Islamic conceptions are in some way connected with ancient biblical notions, which in turn have precedents in the Babylonian tradition of chaos.22 In his short tractate, the Sod ha-Nachasch u-Mischpato (“Mystery of the Serpent”), the thirteenth-century kabbalist, Joseph Gikatilla ben Abraham, a disciple of the Spanish mystic Abraham ben Samuel Abulafia (1240–c. 1292), sheds some light on the mystery of the mythical creature as a liminal symbol, situated upon the ambiguous dividing line between the divine and the demonic:

147

Know that from the outset of its creation the serpent represented something important and necessary for harmony so long as it stood in its place. It was the Great Servant who had been created to carry the yoke of both sovereignty and service. Its head surmounted the heights of the earth and its tail reached into the depths of hell. Yet in all worlds it had a befitting place and represented something extraordinarily significant for the harmony of all stages, each one in its place. And this is the secret of the serpent of heaven that is known from the Sefer Yezira, and that sets in motion the spheres and their cycle from east to west and from north to south. And without it no creature in the sublunar world had life, and

This world serpent, which likewise serves as liminal motif between order and chaos by encircling the cosmos, in other words the realm of order, was a symbol of great antiquity in the Mesopotamian world and beyond.24 The writings of mystics such as Ibn al-ʿArabī were also influenced by the esoteric science of alchemy (al-kīmiyāʾ), considered a form of reÂ�vealed knowledge that had both its spiritual goals and practical applications. A special alchemical symbol is that of the tail-eating serpent, known as ouroboros (the etymology is from oura, “tail,” and the root of bora, “food,” boros, “voracious”).25 Among the large pseudo-epigraphic literature of alchemical books composed in the medieval period, an Arabic alchemical treatise titled Muṣḥaf al-ḥakīm Usṭānis fī-l-ṣināʿat al-ilāhiyya (“Book of the Wise Ostanes on Divine Art”) is attributed to Ostanes (Uṣtānis), the renowned Median author of books on magic and gnosis of the Achaemenid period.26 It describes how in a dream a creature with serpent’s tail, eagle’s wings and elephant’s head devouring its own tail (like a serpent) guides Ostanes up to the seven gates of wisdom for which it gives him the keys.27 An important corpus of alchemical writings, compiled at the end of the ninth and beginning of the tenth century, is attributed to the celebrated alchemist Jābir ibn Ḥayyān (d.╯c. 196/812), allegedly from Ṭūs in Khurasan,28 who according to tradition was a personal friend of the sixth Shīʿite

╇ Marzolph and van Leeuwen, 2004, p.╯356. ╇ Cf. Streck [Miquel], “Ḳāf,” EI2 IV 400a. It is noteworthy that Behemot is known as Lawatyā (Leviathan, see Job 41:1), the patronymic part of the name (kunya) is Balhūt and Bahamūt (Behemot, see Job 40:15). Cf. al-Thaʿlabī, Qiṣaṣ al-anbiyāʾ, p.╯4, cited after Thackston (tr. al-Kisāʾī’s Qiṣaṣ al-anbiyāʾ), 1978, p.╯338, n. 9. 20 ╇ Bet ha-Midrasch, 1853–77, vol.╯2, p.╯63, 17–8. 21 ╇ Babylonian Talmud Bava Metsiah 84b-85a, as cited in Epstein, 1997, p.╯74. This is supported by the Hasidic verse zaddik yesod olam: “a righteous person is the foundation of the world” (Proverbs 10:25), cited after idem, p.╯74. 22 ╇ Streck [Miquel], “Ḳāf,” EI2 IV, 400a. 23 ╇ Scholem, 1957, repr. 1988, p.╯437.

╇ Schütt, 2002, p.╯106. ╇ Needham and Wang, 1965, p.╯374; Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯863. 26 ╇ Sezgin, 1971, pp.╯51–4; Ullmann, 1972, pp.╯184–5; Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯862; Needham and Wang, 1965, pp.╯333–5. In Zoroastrian pseudo-epigrapha which include those of Ostanes, the magus is said to have accompanied Khshayārshā (Xerxes) during the great Persian invasion of Greece. Cf. Boyce and Grenet, 1991, pp.╯494–6. 27 ╇ Reitzenstein, 1916, pp.╯3335; Ullmann, 1972, pp.╯184–5 and ns. 1 and 2; Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯862. 28 ╇ Ullmann, “Al-Kīmiyāʾ,” EI2 V, 110a.

18 19

24 25

148

chapter nine

imām Jaʿfar al-Ṣādiq (81/700–147/765) who initiated him into alchemy and secret teachings.29 In his Kitāb al-aḥjār (“Book of Stones”) Jābir ibn Ḥayyān includes several recipes for creating beasts such as serpents.30 The earliest texts of the corpus are thought to have originated in the scientific culture of the great cultural oases the Oxus (Āmū Daryā) and Jaxartes (Syr Daryā) rivers: Marw, Balkh, Khiva, Samarqand and Bukhara, the traditional roads of exchange between east and west, from where it was transmitted to the rest of the Islamic world.31 Another work was the Tabula Smaragdina (“Emerald Tablet”) attributed by Islamic sources to Hermes, which purported to reveal the key to the ultimate secret of alchemy, and which was included in the Kitāb Sirr al-Khaliqā wa Ṣanʿat al-Ṭabīʿa (“Book of the Secret of Creation and the Art (of Reproducing) Nature”), written in the mid-seventh century or later in about 205/820 under the caliph Maʾmun.32 Julius Ruska, who translated the text, was the first to suggest that its origin lay further east.33 He similarly proposes that its genesis should be sought in “the great cultureoases in the region of the Oxus and the Jaxartes rivers, of Merv and Balkh, or Khiva, Bokhara and Samarqand, those great cities which since ancient times have seen the exchange of material and intellectual goods between west and east, and where Greek traditions endured for such a surprisingly long time.”34 Ruska visualises these cities at the crossroads of Asia north and east of the Sasanian empire as filled with a mixed population of Iranians, Tūrānians, Syrians, Indians, and Chinese,35 places where Buddhism, Manichaeism, Nestorian Christianity and Chinese cults coexisted. Here the sciences of astrology, alchemy and macrobiotics were very much alive.36 The close analogies especially in the account of cosmic Creation with Chinese alchemy have been pointed out by Joseph Needham and Ling Wang.37 The text

expresses the alchemical concept of heat acting as male, cold as female, and their union producing humidity and dryness, which has close parallels to the Chinese theories of Yin and Yang.38 However, these correlations could only happen due to a “trans-Asian continuity” from ancient times that was greatly enhanced after the conquests of Alexander (r. 336–323 bc), son of Philip II, king of Macedonia, and later through the traveller and explorer Chang Chhien’s diplomatic and commercial expeditions (fl. c. 138–126 bc).39 In a text known as the “Congress of Philosophers” (surviving only in a twelfth-century Latin translation and hence lacking its original Arabic title), tentatively dated to c. 900 and attributed to a certain Abū Bakr Aḥmad ibn ʿAlī al-Kasdānī, better known as Ibn Waḥshiyya al-Nabaṭī and the Muṣḥaf al-jamāʿa of the Turba philosophorum, a symposium of philosophers and alchemists voice their divergent opinions.40 In the speech of “Socrates” chemical reaction is compared with generation, lead is male and orpiment female. According to “Diamedes” both male and female substances are needed, mercury relating to the former and sulphur to the latter. “Ostanes” makes copper female and mercury male, while “Theophilus” offers allegories of nights between husband and wife, etc.41 The great emphasis on sexuality in chemical substances and reactions that are detailed in the Turba philosophorum is one of the main characteristics of alchemy. This is based on the belief in a universal sympathy that ultimately unites all parts of the universe, metals and stones (which are alive), males and females.42 Further, the uniting of the opposite principles female/male, passive/ active, cold/hot, humidity/dryness finds expression in the coupling of the Sun and the Moon, a cosmological motif of central importance since it symbolises the generation of all things.43 It is believed that all metals are a result of their union

╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯866 and n. 40. Although the traditions about Jaʿfar al-Ṣādiq are persistent even in many books of the Jābirian Corpus itself, the claim that he himself is the instructor appears to be untenable. Cf. Needham and Wang, 1965, p.╯390, and ns. h-k. 30 ╇ Haq, 1994, pp.╯14–21. 31 ╇ Cf. Corbin, 1998a, pp.╯45–6. 32 ╇ Ruska, 1926. Cf. Needham and Wang, 1965, p.╯369. 33 ╇ Needham and Wang, 1965, p.╯370. 34 ╇ Ruska, 1926, p.╯167, as cited in Needham and Wang, 1965, p.╯370. 35 ╇ Eidem, p.╯370. 36 ╇ Eidem. 37 ╇ Eidem, p.╯373.

╇ Eidem. ╇ Chang Chhien’s expeditions are recorded by Ssuma Chhien in the Shih Chi (ch. 123) and the Chhien Han Shu (chs. 61, 96A, 96B). Eidem, pp.╯332, 387. 40 ╇ Sezgin, 1971, pp.╯60–3; Needham and Wang, 1965, pp.╯397–9. 41 ╇ Ruska, 1931, pp.╯200, 215–6, 229, 247, see also the summaries, sects. 54–5, 57, 59; Plessner, 1954, pp.╯331–8. Cf. Needham and Wang, 1965, p.╯401, n. A; Ullmann, “Al-Kīmiyāʾ,” EI2 V, 110a. 42 ╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯863 and n. 35. Cf. Needham and Wang, 1965, pp.╯469–70 and n. h with references. 43 ╇ Cf. eidem, 1965, p.╯373.

29

38 39

vestiges of ancient dragon iconographies

149

in different proportions and according to different modalities. Under the influences of the planets, these are formed in the heart of the earth by the union of the correlative hypothetical substances of sulphur, male, in which fire and air are present and which has hot and dry “natures,” and mercury, female, which contains water and earth, which is cold and wet.44 This relates also to the circular nature of the alchemical process and the agency of transformation that both devours and restores. The ouroboros symbol is depicted in Arabic alchemical texts, such as the writings of MuḥamÂ� mad ibn Umayl al-Ṣādiq al-Tamīmī (c. 287/900– 287/960, known in the west as “Senior Zadith”).45 His most renowned work was the Kitāb al-Māʾ al-Waraqī wa ’l-Arḍ al-Najmīya (“Book of the Silvery Water and Starry Earth”), known in Latin as the Tabula Chemica,46 in which the schematic depiction of a pair of winged creatures biting each other’s tails is shown (fig.╯149).47 The ouroboros motif is also illustrated in the work on alchemy, Kitāb al-Aqālīm al-sabʿa [dhāt al-ṣuwar wa ’l-tashābīh] (“The Seven Climes”) of the mid-thirteenth-century writer Abu ’l-Qāsim Aḥmad ibn Muḥammad al-ʿIrāqī.48 Similarly, in Abū Maslama Muḥammad al-Majrītī’s treatise Ghāyat al-ḥakīm the opposing principles of positive and negative bodily temperaments are associated with the two celestial nodes (knots), the head and the tail of the “hidden essence.”49 It is significant then that in medieval Islamic iconography the ouroboros dragon was doubled and is often pictured as two entwined dragons eating one another (or, in other words, threatening one another), in an act both self-destructive and at the same time parturient, the cycle that

is recreated in the self-devouring.50 In the symbolism of medieval Islamic alchemy, the paired interlaced dragons represented a fundamental polarity, on which the cosmic rhythm is based, the solve et coagula of the alchemical process: the sulphur and mercury of alchemy.51 This double aspect is echoed in the popular belief, mentioned earlier, that the earth is supported by the biblical monsters Leviathan and Behemoth.52 It is also found in rabbinic tradition where it is stated that:

44 ╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯866. Cf. Needham and Wang, 1965, p.╯373; Moulierac, 1987, p.╯88; Kraus [Plessner], “Djābir b. Ḥayyān,” EI² II, 357b. 45 ╇ Cf. Sezgin, 1971, pp.╯283–8; Needham and Wang, 1965, p.╯378; Ronca, 1998, pp.╯95–116, esp.╯pp.╯102–9. 46 ╇ Stapleton and Ḥusain, 1933, 12, no. 1, pp.╯117–213. 47 ╇ The highly stylised illustration is accompanied by an Arabic inscription in the Lucknow manuscript as:

╇ Joseph Needham and Ling Wang (1965, pp.╯378–9) consider this development to have been due to “Chinese influence” on the Hellenistic single tail-eating serpent motif. Cf. Schütt, 2002, pp.╯106–7. 51 ╇ Moulierac, 1987, p.╯88. 52 ╇ Cf. Streck [Miquel], “Ḳāf,” EI2 IV 400a. 53 ╇ Goldschmidt, 1892, p.╯83. 54 ╇ Grünbaum, 1877, p.╯275. 55 ╇ Leisegang, 1924, repr. Stuttgart, 1955, pp.╯111–3, 160, and idem, 1955, repr. Princeton, 1979, pp.╯218–20. On the pivotal function of the serpent for the Ophite sects, see Schlüter, 1982, pp.╯56–9. 56 ╇ Leisegang, 1955, repr. 1979, pp.╯218–20. 57 ╇ Saturnalia 1.9.12. Merkelbach, “Drache,” RAC IV, 1959, p.╯227. 58 ╇ Known as the Askew Codex, British Library, Ms. Add. 5114. See Sacred, ed. Reeve, 2007, cat. no. 72.

...two Birds [with an indication of the position of the respective heads and the tails]; the Male and the Female; Two in One. eidem, 1933, pl.╯I, A. A closely related, yet even more stylised, version of the motif is depicted in the Paris Ms. no. 2610, Bibliothèque Nationale, Paris, see eidem, pl.╯2 B. 48 ╇ Ullmann, 1972, p.╯237 and n. 1. 49 ╇ “Picatrix,” tr. and eds. Ritter and Plessner, 1962, p.╯45.16–9.

Behemoth and Leviathan are serpents (monsters) on the edge of the ocean who encircle the earth like a ring.53

Concerning nāḥāsh ʿāqallāṭōn in Isaiah 27:1 the Talmudic commentator Rashi (c. 1040–1105) similarly remarks that “this entwined Leviathan surrounds the whole earth” [emphasis added].54 The ouroboros symbol was especially important for the Ophite (from ophis, the Greek for “serpent”) Gnostics.55 In the ancient Orphic mystery the great dragon was the sphere of the Sun, the supreme cosmic force, the good spirit of light, stretched like a radiant ribbon around the rim of the heaven of fixed stars; closely related to the four winds, the four quarters of the cosmos and the four seasons.56 Writing in c. 400, Macrobius, moreover, describes the iconography of the dragon, which encompasses the world tail in its mouth, as like the forward and backward looking god Janus, a visible image of the universe which feeds on itself and returns to itself again.57 However in some Gnostic-Christian writings such as the Pistis Sophia (c. fourth century),58 in which belief in heavenly bodies and demonology are closely connected, the serpent became the

50

150

chapter nine

Antichrist and Satan, the Leviathan of the Old Testament.59 The Pistis Sophia, which claims to record the teachings of Jesus during the eleven years following his resurrection, states:

Significantly, in the apocalypse of the heavens the Sun appears as a great dragon accompanied by the seven planets and with the four horses61 that in Greek mythology drew the chariot of the sun god Helios (sol invictus), while the Moon is depicted as steered by a male and a female dragon and drawn by two white bulls.62 Nevertheless, the ancient conceptions of the serpent resurface in the Gospel of John in the New Testament (3:14). As noted earlier, the serpent lifted up by Moses is presented in the Epistle of Barnabas as a symbol of Jesus and of Christianity.63 The formula, “I am the alpha and the omega,” meaning “in my beginning is my end and in my end is my beginning,” also leads to the imagery of the ouroboros serpent, symbol of the unity and eternal renewal of life.64 The ouroboros motif reached its apogee during the Gnostic period, which lasted from the second century bc to the third century ad, when the Greek and West Asian cultures were subsumed by a form of pan-Hellenism.65 It is of significance, however, that the symbol had not been known in Classical Greek, Etruscan or early Roman culture.66 Thus the ouroboros symbol came to serve as an image for a complex of ideas that was known by a Greek name. An ouroboros serpent that encircles a lion, surmounted by a crescent and star, engraved on a

fifth-century Sasanian seal (fig.╯150),67 suggests that the Sasanian representations of the serpent biting its own tail may perhaps be based on conventional renderings of the dragon in contemporaneous star-maps inspired by astrological works of the Hellenistic east.68 In Hellenic astrology, the great celestial serpent (draco coelestis) dominated the heavens, encompassing all the spheres of the cosmos, coiling around the heavens and biting its tail.69 Yet it is important also to consider the reverse of the seal, which depicts a palm tree above a pair of outspread wings with small streamers surrounded either by a decorative border or a more stylised ouroboros.70 Since this represents a rare iconographic variant, Rika Gyselen tentatively suggests a magical interpretation of the seal.71 The motif of the paired encircling dragons also makes an appearance in the well-known legend of the Ascent of Iskandar. Certain editions of the romance of Pseudo-Callisthenes (c. fourth century) describe how, as Iskandar is carried to heaven by eagles, he sees a field encircled by a serpent which is explained as the world surrounded by the ocean.72 The depiction of this scene also reflects the force and ongoing cultural resonance which the Alexander Romance possessed at the time; in particular, the meaning associated with this imagery. It is portrayed on a small Byzantine enamelled medallion of the eleventh century from the Pala d’Oro on the high altar of St. Mark’s, now preserved in the Treasury of San Marco, Venice, representing the schematised composition of Iskandar’s view of the world as seen from on high (fig.╯151).73 The plaque is decorated with a large central tree inhabited by birds and flanked by a pair of addorsed peacocks with their heads turned backwards. The composition is encircled

╇ Leisegang, 1955, repr. 1979, p.╯218. ╇ Pistis Sophia, text ed. Schmidt and tr. Macdermot, 1978, bk. III, ch. 126, p.╯635; cf. Pistis Sophia, tr. Mead, 1896, p.╯320, sec. 319. See also Leisegang, 1955, repr. 1979, p.╯218; Reitzenstein, 1916, esp.╯p.╯41; Merkelbach, “Drache,” RAC IV, 1959, p.╯228; Sheppard, 1962, p.╯88; Beck, 2004, p.╯225. 61 ╇ Pistis Sophia, text ed. Schmidt and tr. Macdermot, 1978, bk. IV, ch. 136, pp.╯709–10; cf. Pistis Sophia, tr. Mead, 1896, p.╯296, sec. 359; Leisegang, 1955, repr. 1979, p.╯218 and n. 11; Sheppard, 1962, p.╯89; Beck, 2004, pp.╯225–6, n. 72. The motif of the flying chariot driven by dragons apppears also in the myth of Medea (Pseudo-Apollodorus, Bibliotheca I. 146; Diodorus Siculus, Library of History IV. 50. 6; Ovid, Metamorphoses VII 217, 350 and 391–3). 62 ╇ Pistis Sophia, text ed. Schmidt and tr. Macdermot, 1978, bk. IV, ch. 136, pp.╯709–10; cf. Pistis Sophia, tr. Mead, 1896, p.╯296, sec. 359; Beck, 2004, pp.╯225–6, n. 72. 63 ╇ Tr. Lake, 1914, vol.╯1, p.╯385. 64 ╇ Leisegang, 1955, repr. 1979, p.╯229.

╇ Needham and Wang, 1965, p.╯376. ╇ Eidem, p.╯376, n. g. 67 ╇ Cf. Gyselen, 1995, pp.╯52–3, fig.╯60 (photograph of seal recto, verso). 68 ╇ Bivar (1969, p.╯26) connects the ouroboros serpent with the Egyptian serpent deity Āpep that is regarded as the cosmic antagonist of the Sun and the Moon. AcÂ�Â� cording to Azarpay (1978, p.╯369 and fig.╯6) “representations of the ouroboros serpent on Sasanian seals of the fifth century suggest the currency of astrological notions about the ecliptic Dragon at an earlier period in Sasanian Persia.” 69 ╇ Leisegang, 1955, repr. 1979, p.╯217. 70 ╇ Bivar, 1969, p.╯26, pl.╯11, DL2; Gyselen, 1995, p.╯53, fig.╯60. 71 ╇ Eadem, 1995, pp.╯52, 76–7. 72 ╇ Pseudo-Callisthenes II, ch. 41, tr. and ed. Stoneman, 1991, p.╯123. Cf. Millet, 1923, p.╯94, ch. 102. See also p. 55. 73 ╇ Grabar, 1951, p.╯48, fig.╯10c (photograph).

The outer darkness is a great dragon whose tail is in its mouth, and it is outside the world, and it surrounds the whole world.60

59 60

65 66

vestiges of ancient dragon iconographies

151

by a pair of giant serpents with looped bodies and interlaced tails. The significance of this imagery is underlined by Grabar who identifies it as highly characteristic of the Iranian repertoire of motifs, stating that “this image of the Universe, a distinctly Iranian iconography, deserved to be introduced into the family of medieval representations of the Cosmos.”74 Similarly the tree and the birds represented on the plaque, as Grabar has observed, represent “Iranian symbols of the garden of the Earth; in the text of the legend they are documented with the words: “fruits of the Earth.”↜”75 The Latin version of the text states: “orbis terrarum sicut area, in qua conduntur fruges.”76 This form of imagery thus provides a bird’s eye view of the terrestrial universe current in the medieval Iranised Central Asian sphere. The motif was here adapted to a Christian context and, as Grabar notes, it is possible that the stylised cruciform motif on the apex of the tree represents a Christian cross which served to “convert” this oriental representation of the universe into a Christian one.77 Evidently the motif had wide currency in the medieval world, both Muslim and Byzantine. The origin of this iconography may plausibly be seen to lie in the Zoroastrian cosmological motif of the Saēna Tree, the perch of the great mythical Saēna bird (Av. saēna-mərəγa-, Mid. Pers. sēn murw, Pers. sīmorḡ).78 The tree on which the Saēna nests grows in the middle of Lake Vārukasha (Yasht 12.17) and is called Vīspō. bish (the Gaokərəna-Tree, Mid. Pers. Gōkirn), Tree of All Remedies, because it bears the seed of all healing herbs.79 According to the Bundahishn the revered Vāsī Panchā.sadvarā lives in the lake (Yasht 42.4). It appears to be a kind of sea dragon and is “so huge that if it were to rush swiftly along from sunrise to sunset it still would not have covered as much ground as the length of its own body;

it rules over all denizens of the waters.”80 Islamic cosmological notions speak of the “Lote Tree on the Boundary” (sidrat al-muntahā), also known as ṭubā (“blessed”).81 The Qurʾān describes how when the Prophet was carried on his night journey (miʿrāj) up to the seventh heaven, he went on as far as the sidrat al-muntahā beyond which no angel or any other being can pass. This tree is located to the right of the divine throne (ʿarsh) and near the garden of the eternal abode (sūra 53, 14–5).82 An emblematic illustration of the city of Babylon in the manuscripts of Beatus’ Commentary on the Apocalypse shows the theme of two giant serpents enclosing a city. Discussing this coincidence of imagery John Williams comments that “Christian culture shared iconographic traditions associated with the Muslim world…”83 Babylon, which must have been notorious for its dragons, was also known in Iranian tradition as the location of the giant dragon Dahāka.84 The illustration of the city of Babylon serves as frontispiece for the Daniel Commentary (1.1 and 14.23) in two tenth-century manuscripts, the Morgan Beatus, dated between c. 940 and 945 (fig.╯152), and the Girona Beatus, dated 6 July 975,85 as well as in one mid-eleventh-century manuscript, the Saint-Sever Beatus from the monastery of Saint-Sever-sur-l’Adour (on which the city is shown to be enclosed by two dragons whose heads and tails entwine at top and bottom).86 In the Morgan Beatus Babylon is shown as a towering city with a large horseshoe arch-shaped main city gate framed by an inverted pair of small dragons, who touch at the tail tips, in such a manner that their undulant bodies seem to form an ogee arch-shaped enclosure around the gate. The entire composition is encircled by a sequence of two giant dragons demarcated by a fine spotted pattern in anti-clockwise arrangement, the gaping

74 ╇ Idem, p.╯48. Similar cosmological concepts referring to the gigantic tree and the mythical bird sent by God to aid the hero are described in the Kitāb-i Samak ʿAyyār; see Gaillard, 1987, p.╯113. 75 ╇ Grabar, 1951, p.╯48. 76 ╇ Cited after idem, p.╯48. Cf. Millet, 1923, p.╯102. 77 ╇ Grabar, 1951, p.╯48, n. 17. 78 ╇ See Boyce, 1975, repr. 1996, pp.╯88–91, 137–8, 143. 79 ╇ Cf. Janda, 2010, pp.╯320–1. 80 ╇ Eadem, p.╯89. 81 ╇ On the Tree of Life in Islamic tradition, see Lechler, 1939, pp.╯369–419; Wensinck, 1921, repr. 1978, p.╯33; Widengren, 1955, pp.╯103–4, 208–9, 212–3. 82 ╇ The Qurʾān mentions also the Tree of Eternity (shajarat al-khuld) (sūra 20, 120–1) and another tree which appears in

Islamic eschatology as growing in Hell (sūras 37, 60–2; 44, 43; and 56, 52). 83 ╇ Williams, 1994, vol.╯1, p.╯139. 84 ╇ Skjærvø, “Aždahā I,” EIr. Cf. Yarshater, 1983a, p.╯426; Hintze, 1999, p.╯86, n. 62. 85 ╇ City of Babylon enclosed by two dragons, frontispiece, Daniel Commentary. The Girona Beatus. Kingdom of Léon, probably Tábara. Dated 6 July 975. Museu de la Catedral de Girona, Num. Inv. 7 (11). Williams, 1994, vol.╯2, ill. 376. 86 ╇ City of Babylon enclosed by two dragons, frontispiece, Daniel Commentary. The Saint-Sever Beatus. Monastery of Saint-Sever-sur-l’Adour. Mid-eleventh century. Paris, BiÂ�blioÂ�thèque Nationale, Ms. Lat. 8878, fol. 217. Williams, 1994, vol.╯2, ill. 466.

152

chapter nine

dragons’ snouts revealing a long tongue which nearly touches the tail of the preceding dragon. Williams refers to an ancient Slavic version of a Byzantine epic of the city of Babylon87 which may further elucidate the imagery of the dragons encircling the city. According to this story, thought to be of Indo-Iranian origin,88 Nebuchadnezzar rebuilt a new city on top of “seven rocks and seven stations,” the entire city being surrounded by the body of a giant dragon whose mouth and tail unite to form the gate.89 Nebuchadnezzar also had a large sword made for himself in the form of a dragon which had magical powers and could act autonomously.90 The image of the dragon was, in addition, chosen by the king as coat of arms of the city, and as such was applied to garments, weapons, military standards, horse harnesses and saddles, roof beams, doors and windows, vessels, plates, bowls and spoons.91 In case of war all the “heraldic” dragons would come to life and devour the army of the enemy.92 In the same account of Babylon it is said that following the king’s death, failure to observe his commandments led to an inversion of the protective powers of the dragon. The narrative later reveals that the singing of a hymn in praise of God made the initially hostile dragons recoil as if to form a knot.93 This version of the epic thus makes it clear that the dragon was seen not only as a creature that struck fear in people’s hearts, but as guardian of the city with inherent apotropaic powers.94 It suggests that the encircling dragons performed a protective function as well. It is interesting to consider, moreover, that the conspicuous knotting characteristic of many dragon representations, a topic examined in the following chapter, may be an indication that the fierce aspect of the dragon is contained once its body forms a knot. The encircling dragon iconography is also found on architectural elements such as a carved wooden door, once possibly part of a mausoleum.95 The door, dated to the first half of the thirteenth-century, is thought to come from the Tigris region, and is now preserved in the Museum für Islamische Kunst in Berlin (fig.╯153).96 It is carved with a pair of large dragons surrounding a cen87 ╇ Idem, vol.╯1, p.╯60, n. 64. The epic is known from six slightly varying text fragments of ancient Russian histories. Wesselofsky, 1876, pp.╯129–43, esp.╯133. 88 ╇ Idem, 1876, p.╯308. 89 ╇ Idem, p.╯133 and n. 11. It may further be noted that in another ancient Russian epic a dragon, his head touching the tip of his tail, is described as surrounding the city of Antioch. Idem, pp.╯325–6, n. 51.

tral medallion set within an arch-shaped frame. The latter contains an interlaced infinite star Â�pattern, in turn outlined by an interlaced pearled band which extends at its apex into a small medallion. It is touched on either side by the sinuous tongues projecting from the gaping mouths of the dragons whose scaly serpentine bodies with raised slender wings wind tightly around the medallion. Their bodies form a pretzel-like knot and then two loops, the tail tips tapering to a point to form a tight curl at the base. On account of the surface wear of the door only the frame of the small medallion is extant, one can therefore only speculate what it was that the dragons were protecting or threatening. A similar conceptualisation governs the depiction on a large bas-relief stone fragment carved with a pair of antithetically arranged dragons that frame a large mutlilayered composite rayed orb which carries clear solar associations, discovered near Alaeddin Tepe in Konya, now in the İnce Minare Müzesi in Konya (fig.╯154). It was probably part of a thirteenth-century Saljuqid monument, which no longer exists. Only the dragon head to the left is complete, portrayed with a long, curved wide-open snout revealing sharp teeth and fangs along with a prominent sinuous tongue the tip of which touches the edge of the star rosette. The head is punctuated with almond-shaped eyes framed by long, curved lashes and crowned by a small, rounded ear. At the back of the head the dragon’s neck is clasped by paired “collars,” the upper part braided, the lower marked with vertical hatching. The long scaly ophidian bodies form a loop and then a pretzel-like knot. The bottom section of the stone is broken off so the tail tips are lost. Likewise, only part of the pretzel-like knot of the dragon to the right has survived. A more complex form which may be related to the encircling motif graces the small “Kiosk Mosque” situated in the courtyard of the double-section caravanserai Sultan Han, near Kayseri (between 629/1232 and 633/1236) discussed in detail in chapter 2. Here the confronted serpent festoons end in small dragon heads, the necks bent inwards and the open jaws holding the tips ╇ Idem, pp.╯133, 135. ╇ Idem, p.╯133. 92 ╇ Idem, pp.╯133–4. 93 ╇ Idem, p.╯141, n. 28. 94 ╇ Idem, p.╯142. 95 ╇ Cf. Hauptmann von Gladiss, ed., p.╯95. 96 ╇ Cf. Meinecke, 1989, pp.╯54, 58, detail. 90 91

vestiges of ancient dragon iconographies

153

of the outer festoon-band. They thus appear to “swallow” or “deliver” their own tails (fig.╯13). An analogy may further be drawn to the interlaced dragons that enclose the ogival arches carved with figural compositions on the portals of the Imām Bahir Mausoleum in Mosul (fig.╯19) and the Mār Behnām Monastery southeast of Mosul (fig.╯18). The ouroboros aspect is further evident in the double frontispiece of the Kitāb al-diryāq (“Book of the Theriac,” often referred to as “Book of Antidotes”), dated 595/1199, in the collection of the Bibliothèque Nationale, Paris, on which the encircling dragons are juxtaposed with the personification of the Moon (fig.╯176), examined below in chapter 11. Although of course realised in an entirely different medium, the representations in stone and on paper probably share a relatively close geographic provenance and period of production. The astral personification on the Pseudo-Galen double frontispiece, moreover, might provide a link between what appear to be composite stellar symbols on the Berlin door (fig.╯153) and the Konya architectural stone fragment (fig.╯154). The motif also features conspicuously on several Kashan-style ceramic vessels that are overglaze painted with lustre; here the encircling imagery is rendered in the form of interlacing bands comprising multiple pairs of “Saljuqstyle” dragons with small pointed ears and gaping snouts with thin projecting tongues circumscribing a central composition.97 On one bowl, now preserved in Chicago, in the Harvey B. Plotnick Collection, the ophidian bodies of the mythical creatures encircle a well-known motif, often portrayed on early thirteenth-century Kashan-style ceramic bowls, of two seated human figures conversing. Portrayed with three-quarter facing large “moon-shaped” faces with long almond-shaped eyes, arched eyebrows and gently smiling lips, flanked by long tresses embellished with diadems and flowing over the shoulders, the figures conform to the standards of ideal beauty that prevailed in Islamic lands ruled by Western Asian dynasties from the tenth century onwards.98 The

composition is encircled by an interlaced band formed of six pairs of dragons terminating in paired confronting heads with gaping mouths and projecting tongues (composed of very fine flame-like lines). Their alternately spotted and finely scaled bodies form evenly spaced circular enclosures containing vegetal motifs. The serpentine interlace is in turn framed by epigraphic bands in Kufic and cursive script in Arabic and Persian (fig.╯155). The main theme of the inscriptions is love, the longing for the beloved and the anguish and suffering occasioned by love.99 These sentiments are reflected by the two human figures whose thoughtful interaction is underlined by a distinctive right-handed gesture with delicate finger movement by one of the figures, while the other has folded hands.100 The introspective experience of personal love was perceived as a crucial element in following the path to virtue and attaining moral perfectibility.101 This microcosmic ideal is aptly framed by the interlaced dragons. The tall cylindrical body of a Kashan-style ewer of the same period and in a closely related style, now preserved in Paris, Musée du Louvre, is similarly encircled by eight pairs of intertwining dragons whose heads with wide-open mouths and projecting tongues confront each other on the rounded shoulder. The bodies of the Â�fabulous beasts define five horizontal registers filled with cursive epigraphic bands in Persian Â�written in reserve on a lustre-painted ground.102 It may be inferred that, although generally perceived to be a mere decorative device on Â�Saljuq-period objects, the iconography of Â�encircling dragons may well have conveyed some cosmological and, possibly, mythological Â�significance. The encircling dragon motif is also found on textiles, which were a key luxury item for circulation. Their eminent portability and high status within the medieval Islamic and Eastern Christian cultures assured their important role in establishing an international visual vocabulary.103 This led to shared decorative themes in Islamic and Byzantine textiles, and consequently to frequent

╇ Grube, 1965, pl.╯28 (dated 607 H/1210, New York, Metropolitan Museum of Art). Enderlein et al., 2001, p.╯51 (Berlin, Museum für Islamische Kunst, inv. no. I. 1996.2). Pope, 1945, p.╯121, pl.╯84 (Kelekian Collection). 98 ╇ Cf. Holter, 1937, pp.╯11–2. 99 ╇ Pancaroğlu, 2007, p.╯133, cat. no. 86. 100 ╇ Cf. eadem, p.╯133, cat. no. 86.

╇ Cf. Meisami, 1987, pp.╯237–98. ╇ L’Islam dans les collections nationales, 1977, p.╯148, cat. no. 293. Watson, 1985, pp.╯104–6 and fig.╯76. L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯110, cat. no. 75. Paris, Musée du Louvre, Département des Antiquités orientales, Section Islamique, inv. no. MOA 444. 103 ╇ Cf. Hoffman, 2001, p.╯26.

97

101 102

154

chapter nine

difficulties in distinguishing between centres of production.104 The iconographic scheme of encircling serpents is clearly rendered on an eleventh- or early twelfth-century wine-red and gold samite, the socalled relic cover of Saint Amandus, one part of which is preserved in the Abegg-Stiftung in Â�Riggisberg, the other in the Cleveland Museum of Art.105 The textile fragments, which are thought to have a Western Asian provenance, possibly from the Saljuq realm, figure staggered beaded escutcheon-shaped medallions enclosing gigantic double-headed eagles with their wings outspread. The eagles surmount pairs of addorsed feline quadrupeds whose tails also terminate in dragon heads with gaping jaws. A pair of giant double-headed dragons encloses the entire composition (fig.╯156). An extended form of the encircling dragon motif is also found on the lining fabrics of the coronation mantle of Roger II (r. 1130–1154), the Norman ruler of Sicily and southern Italy. The outer face of the red silk cloak, which was woven in the royal workshop (dār al-ṭirāz) in Palermo and carries an inscription in Kufic dating it to the year 528/1133–4, is embroidered with gold thread and pearls with a central palm tree separating addorsed tigers attacking camels.106 This iconography not only represents the cosmological and emblematic motifs that symbolised Roger’s claim to power, but also creates a link beyond the western Norman domain to the Islamic sphere, which is underlined by the fact that the embroidered inscription along the lower hem of the garment is written in Arabic, a language used at Roger’s court and in which he was fluent.107 In

view of the king’s controversial status the imagery on the mantle would have been expected to convey the legitimacy of his position. The lining, however, would remain largely unseen, revealing only glimpses of the imagery which was therefore mainly reserved for the person of the king. The lining fabrics comprise five successive sections from three different textiles that are, according to the motifs, commonly referred to as the “Dragon,” the “Tree of Life,” and the “Bird” cloth (figs. 157a and b).108 All three types show a gridlike lattice composed of paired interlaced dragons with projecting tongues that define the overall surface and enclose groups of human figures, trees, birds and foliate motifs. A beaded rinceau runs along the body of one of the serpents while the body of the other is marked by evenly spaced stripes.109 Since this imagery was chosen for the lining of the coronation mantle, a highly official and ceremonial vestment, its symbolic content must have been imbued with a special significance. While the outer face was intended as a visual statement of the king’s claim to power and authority, the inner one may be seen as having an implicit talismanic intent. This may be compared to the use of talismanic shirts covered with Qurʾānic verses often in conjunction with astrological and magical practices and devices that were often worn under the outer garments (or armour) and of which a considerable number of Turko-Iranian examples are recorded.110 These shirts were destined to protect the owner, avert injury or illness, and thereby bestow invulnerability. The encircling dragon motif which decorates the lining fabric that covers the inside of the coronation mantle may therefore be presumed to

104 ╇ Soucek, 1997, pp.╯405–7; Glory of Byzantium, 1997, p.╯416 (catalogue entry no. 271 by Daniel Walker); Hoffman, 2001, pp.╯18, 26. 105 ╇ On the Cleveland fragment of the relic cover of Saint Amandus only the feathery tail and claws of the eagle survive. A further fragment is preserved in the collection of Rina and Norman Indictor, New York. 106 ╇ The mantle was subsequently used for the coronation of the German emperors of the Roman Holy Empire and entered the treasury of the Vienna Kunsthistorisches Museum in 1801, inv. no. XIII 14; height 146 cm, length 345 cm. Nobilis Officinae, 2005, pp.╯259–64, cat. no. 66; Tronzo, 2001, pp.╯241–53. 107 ╇ For a translation of the inscription, see Nobilis OffiÂ� cinae, 2005, pp.╯259–64, cat. no. 66 (in German); Hoffman, 2001, p.╯32 (in English). 108 ╇ Nobilis Officinae, 2005, pp.╯256–9, esp.╯p.╯257, cat. no. 65. A textile from the tomb of Roger I, Norman Count of Sicily from 1071 to 1101, in Palermo belongs to this group, of which today only a drawing of 1784 survives. Idem, cat. no. 65, fig.╯1 (after Daniele, F., I regali sepolcri del Duomo

di Palermo riconosciuti e illustrati, Naples, 1784). It may further be compared to an eleventh- or twelfth-century drawing from Egypt with continuous scroll patterns, the lower band of which is composed of paired dragons with raised confronted heads, gaping mouths and projecting tongues touching at the centre, now in Berlin, Museum für Islamische Kunst, inv. no. I.6608; von Wilckens, 1997, p.╯169, fig.╯96. Cf. Gierlichs, 1993, p.╯40, cat. no. 5; Grube and Johns, 2005, p.╯260, fig.╯94.5. 109 ╇ It may not be irrelevant to note that the motif of the serpent also appears in the pavement of the main apse, the Sancta Sanctorum, in the Cappella Palatina, inlaid on either side with a large twice looped serpent, which, according to William Tronzo (1997, pp.╯33–4, fig.╯26; see idem, p. 34 and n. 20, for comparable examples), may have apotropaic meaning. Its association with an “Oriental” repertoire of motifs is strengthened by the existence of a panel with addorsed lions symmetrically flanking a stylised tree at the nave entrance to the choir; idem, p.╯34, fig.╯27. 110 ╇ See Turks, 2005, 442, 458–9, cat. nos. 257, 322; and Tezcan, 2006.

vestiges of ancient dragon iconographies

155

have been intended not only to protect the king but also to endow him with certain abilities or powers. Research into the silk and gold thread textiles underlines on the one hand the characteristics of Byzantine court production, the panni imperiali, while on the other pointing to stylistic features pertaining to Central Asia and the Ancient Orient. The closely related iconographies indicate the use of pattern books that seem to have circulated among different workshops involved in making the “golden textiles.” In view of the complexity of cultural identification, Eva Hoffmann proposes the existence of a “shared textile vocabulary of the international courts between the tenth and the twelfth centuries from Islamic, Byzantine and Norman centres.”111 The possibility of an eastern, and perhaps Central Asian, provenance for the overall iconography is strengthened by the discovery of a silk fragment from Samangan province in Afghanistan, now preserved in the al-Sabāh Collection in Kuwait (figs. 158a and b). In the same manner as on the “Bird Cloth,” the Samangan fragment also shows a large, inverted pretzel-like interlace, formed by the ophidian bodies of four serpent-dragons. All the details of the iconographic expression, including palmette-like trees flanked by perched birds in the lobes of the interlace and the beaded or striped demarcation of the ophidian bodies as well as their projecting red tongues, are near-identical to those of the “Bird Cloth” (fig.╯157a and b). It would certainly be wrong to attempt to assign the motif to a single artistic tradition. However, it is worthy of note that the motifs enclosed by the confronted giant dragons that divide up the entire surface include the ancient Iranian cosmological motif of the tree with birds. This motif is closely related to the representation on the Byzantine enamelled medallion from the Pala d’Oro (fig.╯151) which, although it was made in Constantinople, was clearly inspired by Iranian and Central Asian visual expressions. This artistic tradition permeates also the more complex iconography on the “Bird Cloth” lining fabric found in the coronation mantle of Roger II. Even though it would at first sight seem likely that the lining fabrics were inserted at the time the coronation mantle of Roger II was made (dated 528/1133–4),

It is interesting to consider the exceptional motif of the drinking dragon which is depicted on the so-called “Dragon Cloth” section of the lining of the coronation mantle of Roger II. The openmouthed dragon heads are shown with projecting tongues, apparently lapping from the contents of the stemmed cups (fig.╯159).113 The same motif is found above the lintel of the southern outer door of the monastery of Mār Behnām/Deir al-Khiḍr, an important place of pilgrimage associated with miraculous healing by both Christians and Muslims. The two addorsed intertwined dragons are portrayed with wide-open mouths and long tongues with entwined bifid tips that also appear to drink from a stemmed cup (fig.╯160). The same visual formulation can be observed in an illustration in the Vani Gospels, a Georgian manuscript transcribed and illuminated in Constantinople at Romani Monastery by the Georgian monk Ioane for the celebrated Georgian queen Tʿamar (1184–1211/2) in c. 1200. A canon table set within an architectural frame of columns supporting arches is surmounted at the apex by paired serpents that lap with bifid tongues at the liquid

╇ Hoffman, 2001, p.╯34. ╇ Wardwell, 1988–9, p.╯110, with references, and fig.╯48. 113 ╇ Cf. Nobilis Officinae, 2005, pp.╯124, 126, figs. 1 and 2, pp.╯256–9, cat. no. 65 (“Dragon Cloth”). Pairs of drink-

ing griffins and lions antithetically placed on either side of a bowl-shaped vase are featured on the relief ornamentation of the Mshattā façade in present-day Jordan built during the early Islamic period. See Creswell, 1932, pls. 66–74.

111 112

Anne Wardwell has tentatively dated another silk and gold textile decorated with parallel curving vines which also forms part of the lining of the mantle to the period between the third decade and the last quarter of the thirteenth century,112 hence suggesting that a section of the lining was added at a later date. The al-Sabāh textile has been radiocarbon dated to between 1154 and 1282. Regrettably, the test results are not specific enough to be able to determine in which of the two centuries the textile was fashioned. The Samangan fragment with three-partite heart-shaped knots fashioned of serpent-dragons certainly underlines the wide circulation of this significant iconographic representation. The continued use of a motif associated with ancient Iranian ideas is not surprising given that, despite the triumph of Islam, the people of the eastern Islamic lands hung on to their preIslamic roots and associated iconographic notions more assiduously than those in most other areas of the Islamic world. b.╇ The drinking dragon

156

chapter nine

contained within a chalice (fig.╯161). The foot of the chalice is held in place by the interlaced reptilian bodies, patterned with scales, which spring from the base of a four-columned structure that holds up a fountain, probably representing the “Fountain of Life,” underlining the paradisiac symbolism associated with the illumination of canon tables.114 The composition is flanked by winged mythical creatures, a griffin to the left and a sphinx-like being playing a stringed instrument on the right. The symbolism of a dragon with its head placed close to the vessel as if eager to reach the contents while its body serves as handle attached to jugs and ewers had been known since ancient times.115 The motif became formalised in the post-Sasanian period with the body of the dragon becoming more stylised.116 This type of imagery in which the gaping snout of a dragon head holds the rim of a vessel above a handle or, alternatively, releases the liquid when it serves as spout (fig.╯35), is associated with the reading of the dragon as beneficent custodian, safeguarding the precious contents of the vessel. The dragon’s role as a beneficent creature, guardian of the liquid of vessels, evoked in the figure of a winged and horned dragon with elongated snout and open jaws, reaching towards the folded lip of a vessel while the frontlegs hold onto the side wall just below the rim, is featured on a late seventh- or early eighth-century pear-shaped gold jug of unknown provenance, preserved in the Moscow Historical Museum (fig.╯162).117 The

proximity of the dragon protome to the vessel’s rim, which is rendered as if eager to reach for the contents, was perhaps seen as token of the liquid’s protection. The motif of a serpent lapping from a vessel was a well-known theme in depictions of Greek mythology.118 It appears on a sixth-century Byzantine gilded silver dish, preserved in the State Hermitage Museum, St. Petersburg, an example of the reception of Dionysiac motifs in sixth-century Byzantine art.119 It shows a young woman with one knee bent at a right angle, opening with one hand the lid of a cylindrical container from which a serpent is emerging, while with the other holding a calyx-like vessel, from whose contents the serpent is lapping. The container has been identified as cista mystica, the iconographic emblem of the Eleusinian and Bacchic mysteries to which in late Hellenistic times a serpent issuing from the sacred cista was added.120 The female figure is probably associated with a mystery cult, perhaps representing a maenad, one of the female followers of the god Dionysos (fig.╯163).121 A similar motif is associated with Hygieia, goddess of medicine, whose cult spread with that of her mythical father Asklepios and who like her father is invariably associated with a serpent, sometimes shown as drinking from a bowl held in her hand.122 The motif of the lapping dragon, which seems to be rooted in Dionysiac lore, was also known in early Kushan-period Gandhāra (present-day Afghanistan and Pakistan). It is exemplified on a schist relief (fig.╯164) with a nude boy (eros or putto)

114 ╇ For a discussion of the overall iconography of Armenian canon tables (which however does not consider the role of the dragons), see Mathews and Sanjian, 1991, pp.╯169–73, esp.╯p.╯171. On the symbolism of the fons vitae, see Underwood, 1950, pp.╯43–138. 115 ╇ See chapter 3, “The dragon motif on vessels.” MelikianChirvani identifies the representation of a feline or mythical head griping the rim of a vessel as visual metaphor of “the king drinking wine as a substitute for sacrificial blood” which goes as far back as Achaemenid times. See idem, 1992, pp.╯101–34, idem, 1995, pp.╯47–97, esp.╯pp.╯54–7 for literary images of wine as substitute for blood libations, and idem, 1996, pp.╯85–139. 116 ╇ Cf. the parallel development of the feline, most often symbolising a panther, in post-Sasanian art, Ettinghausen, 1972, pp.╯3–10. 117 ╇ Marshak, 1971, T8, and 1986, pp.╯58–9, fig.╯68 (line drawing); Rempel’, 1987a, 63, fig.╯24b. 118 ╇ See the discussion in Elderkin, 1924, pp.╯11, 15–9, 37–8, 41–3, 137. The iconographic scheme of a serpent drinking from a vessel appears, for instance, on a marble relief from Tomis (Constanta), dated to the second century ad, preserved in Bucharest, Archaeological Institute, a serpent coiled around a tree moves to drink from a phialeheld out by a rider next to an enthroned goddess. See also

another marble relief, again perhaps from Tomis, dating to the second or third century ad, now in a private collection, London, where the reptile drinks from a phiale offered by a standing goddess flanked by an approaching rider. See Lane, ed., 1996, p.╯142, fig.╯K12, p.╯143, fig.╯K14. The motif of a goddess offering liquid from a vial to a serpent is also shown on a Roman relief in the National Museum, Palermo, inv. no. 1551 (ibid, p.╯157) as well as on a late Roman votive relief, now in the Musée du Louvre, Paris, inv. no. MA 3316 (ibid, p.╯148). A distant echo of the motif is found on fifth-century bc stone reliefs and terracotta plaques from Lakonia, attesting to its great antiquity. Cf. Salapata, 2006, pp.╯541–60. It also appears on an archaic Laconian relief from Gerakion featuring the descent of the dead to Hades. Elderkin, 1924, pl.╯II, fig.╯2, and p.╯11 with further references. 119 ╇ For a discussion of the reception of Dionysiac motifs in Sasanian art, cf. Ettinghausen, 1972, pp.╯3–10. 120 ╇ Burkert, 1995, p.╯84. See also Kelhoffer, 2000, pp. ╯364–5. 121 ╇ Cf. Iskusstvo Vizantii v sobraniiakh SSSR, 1977, vol.╯1, p.╯98, fig.╯133. 122 ╇ Cf. Thrämer, “Health and Gods of Healing (Greek),” ERE, vol.╯6, part 2, 1914, p.╯552.

vestiges of ancient dragon iconographies

157

attempting to quench the thirst of a dragon who seems to lap from the contents of a bowl that is offered while a second boy, also holding a cup, rides sideways on its back.123 In Christian iconography the Dionysiac symbol of the cup or the kantharos took on a new significance by becoming the Eucharistic chalice.124 The motif of the drinking serpent appears as the serpent-topped chalice of John the Evangelist.125 While it is likely that this form of imagery was initially imbued with ideas of healing and salvation, associated with the beneficial aspect of the serpent, the same motif later increasingly came to be reinterpreted and reformulated in a negative sense when the meaning of the myth associated with the serpent was inverted. As a result of this process the (serpent-)dragon assumed an overall meaning as symbol and instrument of a fiendish force in Christian imagery, thus frequently bearing the traces of the mental and cultural shift imposed by the new religious system.126 In the new Christian context, numerous hagiographic and other traditions created a restructured version of ancient mythical themes pertaining to the dragon. In consequence, there was a shift in mythological paradigm, the force of the earlier myth died, was changed or suppressed and only vestiges of it remain. This inversion of meaning is

attested in the eleventh- or early twelfth-century Byzantine encyclopaedia of the Metaphrastian Menologion volumes featuring so-called “revenge miniatures.”127 These show the crowned figures of Roman tyrants responsible for the deaths of martyred saints, associated with a serpent or dragon seen as an embodiment or emissary of Satan. One of these depictions portrays a crowned figure, probably representing the emperor Â�Maximian, holding a vessel from which a serpent is drinking.128 Hence it is a reasonable conjecture that the drinking dragons on the “Dragon Cloth” section of the lining of the coronation mantle of Roger II, those over the door at the monastery of Mār Behnām and those in the Vani Gospels all belong to the same symbolic group, representing the visual remnants that survived the mutations of ancient beliefs. The basic association between the vessel as container of liquid and the dragon was so strong that it survived in different forms even though the original iconographic association had perhaps long been forgotten.129 Vestiges of the importance of its former cultic associations can however still be gauged from the “revenge miniatures” in which the original intent of the iconography of the drinking serpent was inverted and survives in corrupted form.

123 ╇ Kurita, 1988, repr. 2003, vol.╯2, fig.╯737 (erroneously catalogued as lion). Cf. Ingholt, 1957, pp.╯156–7, no. 296; Boardman, 2003, pp.╯139–40, fig.╯10 (line drawing). 124 ╇ See Elderkin, 1924, pp.╯41–7. 125 ╇ The motif refers to the challenge given to Saint John by the High Priest of Diana at Ephesus to drink from a poisoned cup; in order to warn him, two criminals had been given the same poison to drink and they both died immediately. However the apostle made the sign of the cross, the poison departed from the cup in the form of a serpent and John did not suffer any ill effects. Recorded in Book V of the Acts of the Apostles of the Apocrypha Abdias, reportedly from the Bishop of Babylon; Fabricius, J.A., Codex apocryphus Novi Testamenti, ed. secunda, Hamburg, 1719, vol.╯1, p.╯577, see Barb, 1953, p.╯9; also Kelhoffer, 2000, pp.╯449–52. While it is not specified what kind of harmful substance was contained in the potion offered to Saint John, it is of note that this deed may not be as miraculous at it appears had it

contained snake venom. Already in classical antiquity it was quite well-known that a viper’s venom is harmful only if it enters directly into a person's blood stream (that is, through an open wound) but not if it is imbibed. See idem, pp.╯433– 452, with further examples of “miracles” involving the drinking of snake venom. 126 ╇ See p. 119, n. 85. 127 ╇ Cf. the example of an eleventh- or early twelfthcentury Byzantine encyclopaedia of the Metaphrastian Â�Menologion, n. 986. Cf. Walter, 2003, pp.╯52–3. 128 ╇ Sinai, Monastery of Saint Catherine, gr. 508, fol. 234v: Inde and Domna (6F7). See Patterson Ševčenko, 1990, p.╯156. 129 ╇ Cf. the parallel transmission of the Dionysiac animal, the panther or lynx, into Sasanian and post-Sasanian art and the concluding remarks of Ettinghausen, 1972, p.╯10. Note also the use of both serpent and feline (in this case interpreted as lion) on the handle of the Bobrinsky bucket (idem, fig.╯125).

158

chapter nine

the knotted dragon motif

159

chapter ten

The knotted dragon motif

a.╇ The ancient practice of knot tying From ancient times the tying of magical knots was a way of binding magical/supernatural powers.1 The ritual practice of tying knots in a cord and blowing on them was one of the most common forms of harmful or therapeutic magic in preIslamic times.2 The practice is connected with the concept of binding spells whose knots cannot be unloosed, thus drawing on the principles of sympathetic magic with the purpose of rendering the victim generally impotent.3 Serpent charming may be interpreted in a similar manner since the serpents are “bound” by the charmer.4 The knotted representations are used as symbols to conjure up the cosmic powers that are appropriate to effectively protect or, conversely, harm individuals as well as human communities.5 Such “binding spells” are frequently employed on Aramaic and Mandaic incantation bowls. In these the magician threatens to employ a spell such as he has used to bind the sea and the dragon Leviathan, and to 1 ╇ In Mesopotamian magic texts the verb “to bind” (Akkadian kasâru) was used to express the concept of “bewitchment” (Astour, 1968, p.╯18). Inversely, in lengthy litanies the help of the gods was invoked to undo or dissolve the knot of a magic spell or charm which was believed to have supernatural powers (Dhorme, 1949, p.╯261). In like manner the texts describe the role of the conjurer as the “knotter” (Akkadian kashu) or the “untangler” (Akkadian pâtaru) of nefarious or beneficial charms, respectively, both of which are symbolised by knots (Amiet, 1961, rev. and repr. 1980, p.╯116). The apotropaic function of the knot is a recurrent phenomenon in many countries and periods. Cf. Kitzinger, 1970, p.╯642 and n. 21. 2 ╇ Cf. Wellhausen, 1897, p.╯161. A scholium on a serÂ�mon of the fourth-century archbishop of Constantinople, Â�Gregory of Nazianze, defines incantations (epasmata) with the words:

The chants sung over young children by [...] old women, muttering to avert evil, and at the same time licking the babes’ foreheads with their tongues and spitting, blowing to each side. Migne, J.P., Patrologia Graeca 36, 907 B-C, cited after Bonner, 1950, p.╯4. 3 ╇ Jeffers, 1996, p.╯32. 4 ╇ Eadem, p.╯32. 5 ╇ This is corroborated by Porphyrios who in his letter to Anebon considered the tying and untying of sacred knots as

bind the victim with the same bonds as those constraining the sky, the earth and the seven planets.6 The Qurʾān warns of “the evil (arising) from those who blow (or spit) upon the knots” (sūra 113, 4), a practice analogous to that known as “tying the aglet,” designed to keep husbands and wives apart.7 It is notable that sūras 113 and 114 are called al-muʿawwidhatān (“the two who preserve”) and their verses are considered of special talismanic power if recited in tandem with the original signification of blowing upon the knots.8 Interestingly, sūra 113 refers to the evil women “who blow upon knots,” and for this reason it is believed to be “particularly efficacious against the ills of the flesh.”9 In his commentary on sūra 113, the thirteenthcentury Qurʾānic exegete al-Bayḍāwī refers to a related account according to which a Jew cast a magic spell on the Prophet Muḥammad by putting a cord with eleven knots into a well, whereupon he fell ill.10 After the revelation of sūras 113 and 114, the archangel Jibrāʾīl disclosed the cause of means by which the theurgists induced the gods to certain actions. Cf. Chwolsohn, 1856, vol.╯2, p.╯138, n. 144. 6 ╇ “Wenn ihr irgend etwas gegen … unternehmt, dann werde ich euch bezaubern mit der großen Bezauberung des Meeres und mit der Bezauberung des Ungeheuers Leviathan … biege ich den Bogen gegen euch und spanne die Sehne gegen euch ... bringe ich herab auf euch den Beschluß des Himmels und den Bann, den ich auf den Berg [Hermon] und auf das Ungeheuer Leviathan gelegt habe … [binde ich euch mit der Bindung], durch die Himmel und Erde gebunden wurden … binde ich euch mit der Bindung, durch welche die sieben Planeten gebunden wurden.” Niggemeyer, 1975, pp.╯70, 146. 7 ╇ Fahd, “Siḥr,” EI² IX, 567b. 8 ╇ It is however remarkable that the idiom “blowing on magical knots” was still metaphorically used during the tenth-century as attested by al-Thaʿālibī’s account of the literary life at the Samanid court where: ...men of letters ... [were] each offering to the other fragrant flowers of dialectic, and pursuing the perfumes of Culture, and letting fall in succession necklaces of pearls, and blowing on magical knots [emphasis added]. Yatīmat al-dahr, part IV, ch. 2, Cairo, 1377, tr. Browne, 1920, vol.╯1, pp.╯365–6, cited after Clinton, 1972, p.╯4. 9 ╇ Ruska and Carra de Vaux, “Tilsam,” EI² X, 500a. 10 ╇ Cf. Canaan, 1938, p.╯75 and n. 42.

160

chapter ten

the illness and the Prophet sent ʿAlī to read the two sūras. At each verse (āyaʿalāʾiyya) a knot was untied and the Prophet was cured.11 His third and favourite wife ʿĀʾisha is recorded as saying: As soon as he recited the Qurʾân over one of these knots into which a spell against him had been placed, that particular knot became untied.12

The Basran Muslim mystic, Abū ʿAbd Allāh al-Ḥārith al-Muḥāsibī (d.╯243/857), said that God helps the man who persists in his wrongdoing and is in need of something “which unties in his heart the knots of persistence in evil, in such a way that he may return to his Lord in repentance of his offence.”13 In Muslim religious science the act of loosing or untying a knot is considered the resolving of a difficulty (ḥulūl) which in Hellenistic philosophy denotes the inherent accident in an object as well as the substantial union of soul and body.14 This is also noticeable in the custom that the garments of pilgrims conducting the ḥajj must be knotless, which is in accordance with the widespread belief that those who officiate at either religious or magical ceremonies should have no knots on their person.15 Describing a magical act directed against an enemy, the historian and jurist Abū Zayd ʿAbd al-Raḥmān Walī al-Dīn al-Ḥaḍramī, known as Ibn Khaldūn (732/1332–784/1382), recounts how the magician created an image of the intended victim and then began to pronounce the spell: ...during the repeated pronunciation of the evil words he collected spittle in his mouth and spat upon (the picture). Then he tied a knot over the symbol in an object that he had prepared for this purpose since he considered tying knots and (making things) stick together to be auspicious (and effective in magical operations). In this manner the magician can inflict upon his victim what he had intended.16 11 ╇ Johnstone, 1976, pp.╯79–80. Cf. Muslim, Īmān, vol.╯2, p.╯275, cited after Fahd, “Sihr,” EI² IX, p.╯567b. 12 ╇ Muqaddima, tr. Rosenthal, 1958, vol.╯3, pp.╯160, 168. 13 ╇ In his Kitāb al-Riʿāya li-ḥuqūq Allāh (“Book of Observance of the Rights of God”), al-Muḥāsibī includes the study of the “repentant ones,” al-tawwābūn. Arnaldez, “al-Muḥāsibī, Abū ʿAbd Allāh al-Ḥārith,” EI² VII, 466b. 14 ╇ Massignon [Anawati], “Ḥulūl,” EI² III, 570b. 15 ╇ Frazer, 1913, repr. 1980. 16 ╇ Muqaddima, tr. Rosenthal, 1958, vol.╯3, p.╯161. Cf. Scheftelowitz, 1912–913, pp.╯15–6 and n. 1. 17 ╇ Cf. Chwolsohn, 1856, vol.╯2, pp.╯138–9, n. 144. In the Talmud the third-century rabbi Abaji says that when three, five or seven knots are tied to the left arm, they have power, respectively, to protect from ill-health, to heal and to protect from magic.

Contrarily, knots were also designed to protect from harm, and served to safeguard health and love relations.17 Release from binding spells, ḥall al-maʿqūd (“unbinding the tied”), is one of the stated uses of early magic-medicinal bowls. The word maʿqūd means one who is bound with knots, a commonly used term for a person on whom a spell has been cast. The knot is allegorically referred to in the Qurʾān, when Mūsā implores God to loosen a knot from his tongue (sūra 20, 27). The nightly recitations of the Qurʾān (tahajjud) are said to be justified on the basis that it loosens one of the knots which Satan ties in the hair of a sleeper (sūra 17, 81).18 It is notable that the same concept existed in the Turkic culture where the concept “magic” is also expressed in “bond, fetters” (bag).19 A magical association with knotting is also made in one of the earliest prose romances in Persian, the Kitāb-i Samak ʿAyyār, in which the magicians who are captured may be bound by a rope only if the binder knows the special types of knots and the appropriate manner of fastening them.20 More mundane are the allegorical associations made by the Rūm Saljuq chronicler Ibn Bībī in his al-Awāmir al-ʿAlāʾiyya fī ’l-ūmur al-ʿAlāʾiyya, who writes that in spite of the waning fortunes of the empire: ...it occurred to no one that the knot of this empire could dissolve and that the sun of this fortune could set.21

For the successful conclusion of the Rūm Saljuq sulṭān ʿIzz al-Dīn Kay Kāwūs’s marriage preparations he elaborates upon the commonly used metaphor “to tie the marriage knot” (ʿaqdi zanāshūʾī bastan)22 in the following manner: …when the knot of the agreement was tightened and the rope of the bond gained firmness.23

18 ╇ Abū Dāwūd, Tatawwuʿ, bāb 18. Wensinck, “Tahadjdjud,” EI² X, 87a. 19 ╇ The word bag is derived from the verb ba- “to bind,” hence a magic spell can be bound. I would like to thank Professor Dr Zieme for elucidating this point. 20 ╇ Samak-i ʿAyyār II, p.╯354; V, p.╯532, as cited in Omidsalar, “Magic in literature and folklore in the Islamic period,” EIr. Cf. Gaillard, 1987, pp.╯19–21. 21 ╇ “… es niemandem … in den Sinn [kam], daß der Knoten jenes Reiches sich auflösen und die Sonne jenes Glückes untergehen könne.” Ibn Bībī, al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, p.╯47. 22 ╇ Steingass, 1892, repr. 1981, pp.╯623, 857. 23 ╇ “… als der Knoten des Vertrages zusammengezogen war und das Seil der Verbindung Festigkeit erhalten hatte.” Ibn Bībī, al-Awāmir al-ʿalāʾiyya, tr. Duda, 1959, p.╯79.

the knotted dragon motif The allegory of the knotted dragon tail is used by the early thirteenth-century secretary in the administration of Jamāl al-Dīn Ay Ana Ulugh Bārbak,24 Abu ’l-Sharaf Nāṣiḥ Jarbādhqānī, who records that: Abu ’l-Ḥasan was commanded to Sistan and with cunning, boldness and skill to bring success to this travail, which had become knotted as the Dragon’s tail … to free the troops from the straits of exile, the locks and bolts of affliction25

161

ters (an important part of magical practice), was presumably intended to intensify their beneficial effect by invoking the concept of a continuous, endless protection.30 To this may be added the definition given by the fourteenth-century Ottoman theologian and biographer, ʿIṣām al-Dīn Ṭashköprüzāde, of the Arabic word ṭilasm, or talisman, as “indissoluble knot.”31 b.╇ The knotted dragon

hence alluding to these military difficulties as “knots” in the dragon’s tail. It should be remarked in this connection that the use of knots as apotropaic devices perpetuates earlier practice seen in the architecture of the early Islamic period.26 It is significant that these were also placed in or near thresholds. The desert residence of Khirbat al-Mafjar, possibly built by the eighth-century Umayyad prince and later caliph al-Walīd ibn Yazīd,27 for instance, had specially designed mosaics with magic knot designs placed on the thresholds to ensure the magical entrapment of any evil force trying to enter the throne hall (dīwān).28 Similar knot designs were employed in contemporary churches of Syria and Palestine suggesting that the apotropaic associations of ancient times were incorporated into Christian concepts as well.29 The knotting aspect can also be observed in Islamic epigraphy on portable objects, for instance on the Bobrinski bucket, dated 559/1163. Circumscribing the body at mid-section is an epigraphic frieze with a benedictory content which has constantly repeated knotted hastae (fig.╯34). In addition to their decorative function, the reduplication of potent signs such as complex knotted forms, in this case mainly of a quadripartite configuration, together with the reiteration of let-

This is illustrated on memorial steles, a celebrated example being the early eighth-century entwined dragons that crown the stele of the Turkic commander-in-chief Köl Tigin, inscribed with a bilingual text in Chinese and Turkic, in Chöshöö Cajdam in Mongolia (fig.╯165).34 Together with the memorial complex the stele was built in 731 with the help of Chinese palace artists sent by the Chinese emperor Xuanzong (r. 712–756). The Chinese artists worked under Turkic instructions overseen by Köl Tigin’s older brother Bilgä Qaghan (d.╯734), hence one may surmise that the iconographies figured in the memorial stele were intended, in accordance with Turkic cosmogonic and cosmological beliefs, to smooth the way of the deceased into the afterlife.35 The stele was

24 ╇ Jamāl al-Dīn Ay Aba Ulugh Bārbak played an important role in the politics that followed after the death of last Great Saljuq ruler in the west, Ṭoghrıl III ibn Arslan. 25 ╇ Meisami, 1999, p.╯260. 26 ╇ For an example of the use of knots of Solomon as an apotropaic device, see the eight Solomon knots that appear as isolated motifs in the mosaic pavement that covers the threshold block of the synagogue at Sardis, rebuilt in the fourth century. Cf. Dinkler, 1978, p.╯78, pl.╯XVII, 13. 27 ╇ Ettinghausen, 1972, p.╯42. 28 ╇ Cf. idem, pp.╯17–65, esp.╯pp.╯47, 63, pls. 17–27; Farès, 1959, p.╯33. 29 ╇ Ettinghausen, 1972, p.╯47; Farès, 1959, p.╯52. 30 ╇ Cf. p. 107, n. 209. 31 ╇ In his encyclopaedia Miftāḥ al-saʿāda wa-miṣbāh al-siyāda I, 277, 3 to 278, 3. Ullmann, 1972, p.╯362 and n. 3.

32 ╇ Azarpay, 1978, p.╯366, n. 20; Tabbaa, 1997, p.╯77; Â�Santoro, 2006, p.╯550. 33 ╇ Azarpay, 1978, p.╯366, n. 20. 34 ╇ The interlaced dragon motif has a very long history in the Central Asian region, as is exemplified by a pair of inlaid and pierced gold belt plaques with a pair of entwined lupine dragons in combat with birds, probably griffins, found at kurgan 3, datable to the second or first century bc, in Â�Tchaltyr, Miasnikovski district, a region bordering Kuban in Rostov province, near the Sea of Azov in south Russia, which are part of the stylistic repertoire of Scytho-Siberian art. Museum of Azov, inv. no. KP-24444/1 and 2. Schiltz, 2001, pp.╯178–9, cat. no. 198. 35 ╇ Scharlipp, 1992, p.╯51. Excavations of the Köl Tigin memorial complex revealed that its gates were oriented towards the east, the direction held sacred in Turkic belief

It has been suggested that the motif of entwined dragons with necks or bodies interlaced to form a single loop, occasionally a knot, consistently employed in medieval Islamic architecture on gates and portals, is subject to different interpretations.32 Guitty Azarpay more clearly specifies that: the theme of the entwined dragons finds Central Asian forerunners that doubtless contributed to the widespread use of the motif in Islamic art patronized by Turkish dynasties. 33

162

chapter ten

originally erected on the back of a pedestal in the form of a turtle, which was found nearby. The front of the stele oriented towards sunrise bears an inscription in runic Turkic script, whereas the back of the monument is inscribed in Chinese characters with the condolences sent by the Tang emperor. Another example is offered by a fragmentary monumental stone memorial stele carved with the arched body of a quadruped dragon (which formed the left side of an entwined pair of dragons) of the eighth Uighur qaghan (r. 808– 821) at the Uighur capital Qarabalghasun (Ordu Balïq), to the south of the palace remains, in the Ötükän region of Mongolia (fig.╯166). The dragon head is rendered with foreshortened snout ending in a tightly curled tip, holding a round object, perhaps a pearl, in its open jaws and with a thick curly beard at the lower jaw, extending into a prominently protruding forehead with bulging eyes and large, cusped ears swept to the back. The latter touch upon an X-shaped motif enlivened with small roundels in the interstitial areas that embellish the dragon’s neck, while the scaled muscular body is slightly awkwardly arranged with large projecting extremities terminating in unsheathed claws. The fragment almost certainly formed one side of an interlaced pair of dragons that once crowned a colossal memorial stele inscribed, as surviving fragments reveal, with a trilingual text in Chinese, Uighur and Sogdian. The doubling aspect of the interlaced dragons has already been touched upon in preceding chapters. The pairing of the dragons, aimed at buttressing and doubling the visual impact, was a noticeable feature that may have served to augment the intended effect of the potent symbol. This may moreover be associated with the inherently ambivalent aspect of the

The Kimek founder myth not only affirms the sacredness of the river Irtysh in Siberia to the Turkic tribe but also involves the motif of an ancestress that had been snatched by an aquatic dragon. Also noteworthy is the account given by the Armenian historian Mattʿēos Uá¹łhayetsi (d.╯1142) of the Qun migration in the mid-eleventh century, in which he mentions the “People of the Serpents” that attack the Pale Ones (khartēshkʿ = Cumans).40 To this may be added the story of a Qıpchaq chieftain, Tugorkan, recounted in Rus’

as source of the sun; a concept which is not adhered to in Chinese architecture. Cf. Hersek, 2002, pp.╯152, 156. The cosmic meaning of the dragon in the art of the pre-Islamic Türks of Central Asia was discussed by Emel Esin (1968–70 and 1973) and Gönül Öney (1969a). 36 ╇ Boratav, “Drache,” WdM VII, 1, p.╯207. See also Esin, 1970–1, pp.╯161–82, and a review by Rogers (1970–1, pp.╯267– 9), in which he disputes the cosmological significance of the dragon in Turkish art. In her reply (1973–4, pp. 151–2) Esin quotes, inter alia, from Yusūf Khāṣṣ Ḥājib’s Qutadghu-bilig (“Wisdom of Royal Glory”) completed in 462/1069–70, couplet 126: Yarattı, kör, evren, tuçi evrilür Anıng birle tezginç yime texginür (“See, He created evren [the dragon] which revolves continually, Together with it revolves the Ecliptic”). For a further dicussion, see Esin, 1981, p.╯834. 37 ╇ Cf. DeWeese, 1994, p.╯496, n. 8. See too the Hunnic legend of origin which is rendered, albeit imperfectly, in Iordanes’ sixth-century Getica (“Gothic History”) which comprises an account by the Byzantine historian Priscus who travelled among the Huns. According to “old traditions,” the Huns are descended from “witches” (haliurunnae), or

“magician women” (magas mulieres), who consorted with “unclean spirits” of the steppes and whose offspring dwelled first in “swamps.” See Getica, tr. and ed. Mierow, 1915, repr. 2006, pp.╯85–6; Maenchen-Helfen, 1944–5, pp.╯244–51; Pohl, “Hunnen,” RgA, vol.╯15, 2000, p.╯248; DeWeese, 1994, p.╯496. 38 ╇ Bartol’d, V.V., “Izvlechenie iz Zain al-Akhbar,” Â�Sochineniia, VIII, p.╯27 (text), 44 (tr.); Marquart, 1914, pp.╯89–91; Martinez, 1982, pp.╯109–75, esp.╯pp.╯120–1 on the Kimek. All references as cited in DeWeese, 1994, p.╯494 and n. 3. 39 ╇ DeWeese, 1994, p.╯494. See also Monchi-Zadeh, 1975, p.╯234. 40 ╇ Marquart, 1914, pp.╯54–5. Sharaf al-Zamān Ṭāhir Marwazī, tr. and ed. Minorsky, 1942, p.╯18 (Arabic), pp.╯29–30 (English). Minorsky (p.╯103) suggests that the “People of the Snake/Serpent” were the Qun, whereas Marquart (p.╯55) leans towards identifying them with either the Qitan or Qay. All references cited after Golden, 1997, p.╯89 and n. 13. On the eleventh-century migrations of nomadic peoples in the Eurasian steppe zone, cf. Vásáry, 2005, esp.╯pp.╯4–12.

dragon in ancient Turkish cosmology which saw the creature living underground in winter then re-appearing in the spring and soaring into the sky in the summer where it reigned at the zenith as a divine creature.36 The association of the Turks with the dragon is further corroborated by ethnogonic myths of different Turkic tribes which are characteristic of Central Asian founder myths and legends of origin (for instance, the mythic formulae of the origins of the ancient Scythian-Saka),37 such as the tale of the origin of the Kimeks, recorded in the eleventh-century account of the Ghaznawid historian Abū Saʿīd ʿAbd al-Ḥayy Gardīzī.38 The legend recounts how Shad, the ruler and “founder” of the Kimeks: …one day heard a voice coming from the water as he stood on the banks of the Irtysh; the voice said, Shad, give me your hand in the water. Shad, seeing only some hair floating on the surface of the water, tethered his horse and entered the water; he grasped the hair, which turned out to belong to his wife-to-be, Khatun, and upon enquiring, he learned that she had been seized from the riverbank by a water dragon.39

the knotted dragon motif

163

literary tradition (byliny), who carries the patronymic “Tugarin Zmeevich” (“the son of the serpent-dragon”).41 While it may not be possible to trace the chain of transmission of the interlaced dragons, they are considered to be variations in the context of a possible sensus communis. However, not only are the dragons portrayed as pairs with necks or bodies entwined to form a single central loop, but occasionally their own bodies incorporate a knot at mid-section. The shape of the knot varies, though it is often quadripartite in form. Yet before discussing the motif in the Western Central Asian region it is interesting to note that the iconography of dragons joined by means of a quadripartite knot also makes an appearance in a Buddhist context in the art of eastern Turkestan. Depictions appear in the wall paintings of cave monasteries in Shorchuk, dated between the sixth and the eighth century, and Bezeklik in Qocho (Chinese Gaochang, the former capital of the western Uighur kingdom, Dakianusšhahri, located c. 30 km east of present-day Turfan), which has been dated to the late tenth century.42 The Shorchuk wall painting shows donors facing confronted dragons knotted at mid-section in a similar manner (fig.╯167). An analogous composition of knotted dragons is portrayed on a now destroyed wall painting from Bezeklik which survives only in a drawing by the German archaeologist Albert Grünwedel made during the third German expedition to Chinese Turkestan in 1906 to 1907. The drawing shows a pair of superimposed vegetal stems each terminating in a blossom growing out of a water pool. At the point of juncture, a pair of addorsed Chinese-derived dragons project horizontally, their hind limbs entwined to form a central quadripartite knot, while the upper blossom forms the seat of Padmapāṇi, the Bodhisattva of Compassion (fig.╯168). The painting displays a fusion of Buddhist and Manichaean elements which evolved under Uighur patronage in the late tenth century. The fact that it was

executed under Uighur patronage adds further weight to the possibility that the interlaced dragon figure was introduced into Islamic art from Central Asia via the Turkic dynasties.43 It is significant in this regard that some of the earliest extant examples of the quadripartite knot in Islamic art seem to appear in the wider Khurasan region. A key document, the starting point for the discussion that follows, is a large, elongated copper alloy matrix with cusped arch terminals, distinguished by a pair of confronted dragons in mirror image whose scaly bodies form a central quadripartite interlacing knot (fig.╯169).44 Both arch-shaped ends are filled with the protomes of the paired confronted dragons. The latter have rounded heads with wide-open snouts showing finely demarcated flews and inward-curled tips and revealing projecting tongues that touch at the centre. Their elongated twisted horns with upward-curling tips curve towards the back above the slender cusped ears. The heads are separated from the curving bodies by a finely moulded “collar.” Their angled ribbed front legs with rounded feet nearly touch at the tips. The serpentine bodies of the four dragons form a large, central quadripartite knot that is enlivened at the centre by a small quatrefoil with elongated lozenge-shaped petals which may represent a stellar emblem. This matrix, now in the al-Sabāh Collection, Kuwait National Museum, is part of a set of 77 copper alloy matrices for belt/strap fittings, one of which bears an inscription in the name of Muḥammad ibn Sām above the figure of a lion.45 As discussed earlier, the inscription connects the set directly with the Ghurid sulṭān Ghiyāth al-Dīn Muḥammad ibn Sām or his younger brother, Muʿizz al-Dīn Muḥammad ibn Sām. This set of matrices could thus have appertained to either one, or both, of the brothers, although Ghiyāth al-Dīn clearly remained the elder statesman.46 The set was, moreover, reportedly found near the

41 ╇ Rybakov, B.A., Drevnjaja Rus’. Skazanija, byliny, letopisi, Moscow, 1963, pp.╯102–4. In a bylina verse Tugorkan is depicted as riding forth with two grey wolves preceding him; Rybakov, B.A., Kievskaja Rus’ i russikie knjazhestva XII–XIII vv., Moscow, 1982, p.╯156 (in which this verse is cited); both sources as cited in Golden, 1997, p.╯89. Cf. the discussion in Schirmunski, 1961, pp.╯23–4. 42 ╇ Cf. Russell-Smith, 2005, pp.╯230–2. 43 ╇ Azarpay, 1978, p.╯366, n. 20.

44 ╇ Three further matrices feature seated lions with upÂ�ward-arching tails terminating in large horned dragon-heads; Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. nos. LNS 2558 J a-x2 ‘x2,’ ‘s2’ and ‘t2.’ 45 ╇ Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. no. LNS 2558 J a-x2 ‘g’; length 4.42 cm, width 2.81 cm. 46 ╇ Cf. Flood, 2005, p.╯266.

164

chapter ten

minaret of Jām.47 The name and elaborate titles of sulṭān Ghiyāth al-Dīn Muḥammad ibn Sām are given on the three lower epigraphic bands of the minaret of Jām.48 Another important Ghurid document for assessing the evolving titulature of the sulṭān is a four-volume leather-bound lavishly illuminated Qurʾān dated 584/1188–9, now preserved in the Iran Bastan Museum in Tehran. The sulṭān’s extensive range of official and celebratory titles is recorded in the colophon,49 combining epithets used earlier by both the Saljuqs and the Ghaznawids.50 Even by the inflated standards of the time they are ostentatious, providing an insight into this sulṭān’s self-representation and imperial aspirations a decade before the Ghurid conquest of India.51 The choice of the dragon and the lion, both emblems of might and royalty, as symbols to represent Ghiyāth al-Dīn’s rulership, under which the Ghurid empire reached its apogee, is significant. Even though he was greatly concerned with orthodoxy52 a metrical version of the genealogy of the family, composed by Fakhr al-Dīn Mubarākshāh Marwarrūdhi, which was completed during the reign of Ghiyāth al-Dīn, traces the family back to the hominoid serpent of the ancient Iranian epic past, Azhi Dahāka. As is well-known, a pair of serpents are said to have sprouted from the latter’s shoulders. After Farīdūn (arabicised from Afrīdūn) overthrew the thousand-year reign of Ẓaḥḥāk/Dahāk, his descendants were presumed to have settled in Ghūr.53 The Ghurid descent from Dahāk seems to have enjoyed a favourable reputation. This descent was shared by Rustam, one of the greatest warriors of Iranian legends,54 by the Kushāṇas of the Yuezhi confederacy (c.

first–third centuries) who ruled over the region which comprises what is now Tajikistan, Afghanistan, Pakistan and northern India,55 as well as by the Armenians of a province near Lake Sevan.56 According to the genealogy of the Sām dynasty of Ghūr, the fourth caliph ʿAlī brought about the conversion of the family to Islam in the first century of the Islamic era, granting them the rulership of Ghūr.57 The knot that joins the two dragons is the socalled “knot of Solomon,” which was known in the Central Asian region from at least the first centuries ad through to the Islamic period. This knot succeeded the ancient so-called Herakles/Hercules knot, the most famous knot used in ancient times, considered to have apotropaic and magical qualities, which was often depicted in the form of both ophidian and quadruped dragons,58 and is a compelling example of continuity of meaning in spite of formal change. The symbolism of the Herakles knot still had currency in late seventh- or early eighth-century Sogdia. Moulded in the form of a pair of upright dragons that are linked by means of a large Herakles knot, the motif appears in a non-Buddhist context in the monumental sculptural decoration on the plastered mud-brick architecture of the portico of Temple II in the ancient city of Panjikent in Sogdiana (Zarafashān, north of the Hiṣār range) (fig.╯170). Like the Hercules knot, the knot of Solomon is a well-known apotropaic sign.59 It is commonly associated with the magical seal of Solomon (khātam Sulaymān)60 and attributed to the magician-king Solomon of post-biblical Jewish tradition61 who controls the winds, animals, and spirits. The biblical King Sulaymān is named

47 ╇ A pair of nielloed gold pendants, now in the al-Sabāh Collection, also inscribed with the name of the Ghurid sulṭān, “Muḥammad ibn Sām” was found in the same village. Kuwait, al-Sabāh Collection, Dar al-Athar al-Islamiyyah, Kuwait National Museum, inv. no. LNS 1890 J. 48 ╇ The name and titles of the sulṭān Ghiyāth al-Dunya wa ’l-Dīn Abu ’l-Fatḥ Muḥammad ibn Sām are given by Maricq and Wiet, 1959, p.╯27, no. 3; Pinder-Wilson, 2001, p.╯168 and n. 47. 49 ╇ Iran Bastan Museum, Tehran, Ms. 3507, fol. 198a. Cf. Flood, 2005, pp.╯267–9, fig.╯1. 50 ╇ Idem, p.╯268. 51 ╇ Cf. idem, pp.╯267–9. 52 ╇ Ghiyāth al-Dīn was closely affiliated with the Â�KarrāÂ�miyya sect that was dominant among the population of Ghūr and instrumental in persecuting the Ismāʿīlis, but later he shifted allegiance to the Shāfiʿī school of law, while his brother, Muʿizz al-Dīn, became a Ḥanafī; idem, pp.╯270, 281 and n. 66, p.╯287. 53 ╇ Quoted by Abū ʿAmr al-Jūzjānī, the historian of the

Ghurids; Bosworth, “Ghūrids,” EI² II, 1099a. Cf. Shahbazi, 1993, p.╯159. 54 ╇ See p. 8, n. 46. 55 ╇ See p. 8, n. 47. 56 ╇ See p. 8, n. 51. 57 ╇ See p. 8, n. 48. 58 ╇ See Belenizkii, 1980, fig.╯92. 59 ╇ The importance accorded to such plaited knots is underlined by their ubiquitous use in monumental epigraphic friezes on architecture, an early example being the tomb tower Rādkān West in the Alburz Mountains in Māzandarān finished in 411/1021. See Blair, 1992, pp.╯85–7, figs. 48–52. 60 ╇ Melikian-Chirvani, 1977, pp.╯367–406. The “Seal of Solomon” was portrayed as a five-pointed star or pentagram, interchangeably used with a six-pointed star or hexagram. Cf. Doutté, 1909, p.╯154; Dawkins, 1944, pp.╯145–50; Ittig, 1982, p.╯86; Moulierac, 1987, p.╯88; Shani, 1999, p.╯254 and n. 11. 61 ╇ Cf. Henninger, 2004, pp.╯20, 25. Evidence of Solomon’s influence on the magical traditions of the Graeco-Roman

the knotted dragon motif

165

seventeen times in the Qurʾān (especially sūra 38, 30–9).62 The khātam Sulaymān is thought to have been the symbol inscribed on the seal ring of Sulaymān, “which endowed his owner with power over both terrestrial and supernatural beings.”63 Melikian-Chirvani, moreover, has drawn attention to the possibility that the presence of the motif, which is “typical of the Eastern Iranian repertoire [on metalwork] seldom used in the West,”64 indicates an esoteric intention.65 The motif of the knotted dragons, rendered in virtually the same manner as on the Ghurid matrix, is pictured on a “sphero-conical” unglazed earthenware vessel found in the region of the Golden Horde (Juchi Ulus), datable to the late twelfth or early thirteenth century, now preserved in the State Historical Museum in Moscow (fig.╯171). The dragons are more stylised, however, and have lost their horns and forelegs, so that only their necks are shown to cross over. A great quantity of these “sphero-conical” vessels have been found throughout the Islamic world, and while the vessels probably served numerous different purposes,66 the prevalent view is that they contained liquids such as beverages,67 mercury or perfume,68 while some may have been used for chemical or alchemical operations.69 Several “sphero-conical” vessels with the same motif have been found, such as a bottle fragment from Dwīn in Ararat province, the former Armenian capital until the Arab conquest when it became the seat of the governor,70 an example from Iran, now in the collection of Michael Rogers, London,71 with stamped decoration and a large six-pointed star

(a motif also associated with the magical khātam Sulaymān) as well as another unprovenanced example from a private collection.72 A moulded band with eight almond-shaped medallions enclosing pairs of knotted dragons circumscribes a pyriform flask of unglazed earthenware, thought to come from Iran or Mesopotamia, which is datable to the same period, now in the Musée national de la céramique, Sèvres.73 Here the winged dragons are pictured addorsed. The heads, jaws agape, are twisted back so that their outstretched tongues meet at the centre while their forelegs are turned outwards and the bodies are linked by means of a quadripartite knot. Like most of the “sphero-conical” vessels, the body of the flask is made of dense almost stoneware-like high-temperature fired earthenware, which makes the vessel watertight. The specificity of the paste and the esoteric connotation of the motifs evoke alchemical procedures which were practised by great physicians such as Abū Bakr Muḥammad ibn Zakariyyāʾ al-Rāzī (the Rhazes of the Latin Middle Ages) from Rayy in Iran (c. 250/854–313/925 or 323/935), whose interest lay particularly in practical chemistry,74 and suggest the use of the flask in a laboratory context.75 In the symbolism of medieval Islamic alchemy, it seems the paired dragons connected with the use of a quadripartite knot represented a fundamental polarity on which the cosmic rhythm is based, the solve et coagula of the alchemical process: the sulphur and mercury of alchemy.76 This complex theory was developed in the writings of the Jābirian Corpus, Muḥammad ibn Umayl

period, revealing the extensive influence of Jewish magic, is found in the use of his name in several incantations in magical papyri and the specific reference to the use of his seal in the following appeal: “I adjure you by the seal which Solomon placed on the tongue of Jeremiah, and he spoke.” Preisendanz, 1973, p.╯102 (Pap.╯IV, 3039–041). Cf. Perdrizet, 1903, p.╯58; Aune, “Exorcism,” ISBE, vol.╯2, 1939, repr. 1994, p.╯243. 62 ╇ Sulaymān is acquainted with the speech of birds and animals (sūra 27, 16–9), he can control the wind as well as the demons and jinns who taught magic to humans (sūras 21, 81–2; 34, 12; 34, 12–3; and 38, 36–40). 63 ╇ Ittig, 1992, p.╯86; Shani, 1999, p.╯254 and n. 10. 64 ╇ Melikian-Chirvani, 1982, p.╯178. 65 ╇ Idem, 1977, pp.╯367–406. 66 ╇ On the possible usage of these vessels, see EttingÂ�Â�hausen, 1965, pp.╯218–29; Rogers, 1969, pp.╯147–58. A comprehensive list of functions based on the vessels’ forms is given by Savage-Smith, 1997, pp.╯324–33. The contents of “spheroconical” vessels excavated at Nīshāpūr were tested and were thought to “hold volatile liquids, water and oil without seepage.” See Hauser and Wilkinson, 1942, p.╯89; Wilkinson, 1973, pp.╯XXXII–XXXIII.

67 ╇ On the use of the vessels as beer gourds, containing fuqāʿ or fuqqāʿ, see Ghouchani and Adle, 1992, pp.╯72–92. 68 ╇ Ettinghausen, 1965, pp.╯218–29. 69 ╇ Rogers, 1969, pp.╯147–58; Moulierac, 1987, pp.╯86, 88. 70 ╇ The History Museum of Armenia, inv. no. 1682–8, height 12 cm, diameter 9 cm. Dzhanpoladyan, 1958, p.╯206, fig.╯5. Cf. Ettinghausen, 1965, p.╯228, fig.╯15 (drawing); Rogers, 1969, p.╯153, n. 12 (drawing p.╯154, fig.╯3); Moulierac, 1987, p.╯87 and n. 28. 71 ╇ Rogers, 1969, p.╯152, pls. 1 and 2; Moulierac, 1987, p.╯87; Savage-Smith, 1997, pp.╯324–5, type 6, drawing below (“private collection, London”). 72 ╇ Ettinghausen, 1965, p.╯224, pl.╯L A (photograph). 73 ╇ Sèvres, Musée national de la céramique, inv. no. MNC 19589; see À l’ombre d’Avicenne, 1996, p.╯143, cat. no. 56 (detail). 74 ╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯868. 75 ╇ Moulierac, 1987, p.╯86. Flasks of comparable shape are featured on fol. 15 and fol. 17 of the Pseudo-Galen treatise Kitāb al-diryāq (“Book of Antidotes”), dated 595/1198–9, now in the Bibliothèque Nationale, Ms. Arabe 2964. Cf. eadem, p.╯84. 76 ╇ Eadem, p.╯88.

166

chapter ten dragon being hot, auspicious and indicating increase of wealth and the tail being cold, bringing misfortune and signifying diminution of wealth.82 He also notes that:

al-Ṣādiq al-Tamīmī, and also the Turba PhilosoÂ� phorum. In the Kitāb al-Fihrist, a work composed in Baghdad at the end of the tenth century, Abu ’l-Faraj Muḥammad ibn Isḥāq ibn al-Nadīm (d.╯385/995 or 388/998) credits the Umayyad prince Khālid ibn Yazīd (d.╯c. 704/85) as the first Arab to have been interested in alchemy and to have ordered translations of some alchemical works.77 By the middle of the tenth century various philosophical currents circulated in Baghdad, which also included the magical and alchemical thought-world of the writings of the Jabirian cycle.78 In the writings of the Jabirian corpus of alchemy, the idea of a transmutational elixir (aliksir) used as medicine or life-giving force, which was probably transmitted from China,79 occurs for the first time. It could be prepared from animal, vegetal or mineral substances and could be used to prolong life or given as medicine to sick people. As already mentioned in the preceding chapter, Jābir ibn Ḥayyān (d.╯c. 196/812), purportedly an alchemist at the court of Hārūn al-Rashīd, believed all metals to be a result of their union in different proportions and according to different modalities. Under the influences of the planets, these are formed in the heart of the earth by the union of the hypothetical substances of sulphur, male, in which fire and air are present and which has hot and dry “natures,” and mercury, female, which contains water and earth, and is cold and wet.80 Astrological components may also inform the iconographic rendering. In Abū Maslama Muḥammad al-Majrītī’s magical treatise Ghāyat al-ḥakīm the opposing principles of positive and negative bodily temperaments are associated with the two celestial nodes (knots), the head and the tail of the “hidden essence [“verborgenem Wesen”].”81 As mentioned earlier, al-Bīrūnī reports that the nodes have separate natures, the head of the

It is further of note, as recorded by al-Damīrī in his bestiary, that in medieval Islam it was a popular belief that “serpents do not copulate in the usually known manner of copulation, but they twist themselves round each other (for that purpose).”84 A sexual character of the knotted dragon motif is further supported by the universal sympathy, discussed earlier, that ultimately unites all natural opposing forces, including those of the female and the male principle, which also finds expression in the coupling of the Sun and the Moon. The use of the twisted serpents’ motif as a simile for sexual intercourse appears in Gurgānī’s poem Wīs u Rāmīn which in its essence is a celebration of an erotic relationship.85 However, whereas the theme of the twisted serpents was frequently associated with sexual symbolism in pre-Islamic written sources,86 it is a relatively rare occurrence in medieval Islam. Exactly the same motif as shown on the Sèvres pyriform flask is recorded on a contemporary unglazed clay jug with filter from northern Syria, in the National Museum in Damascus. The vessel carries around the body a similar moulded band with contiguous almond-shaped medallions enclosing the same knotted dragons (fig.╯172). A similar depiction, but with the dragon heads in addorsed position, is found on a shallow ceramic bowl from Raqqa of the same period, now preserved in the Metropolitan Museum of Art in New York (fig.╯173).87

╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯864. ╇ Gutas, 1998, p.╯104. 79 ╇ Hill, 1990, p.╯335. Cf. Needham and Ling, 1954, pp.╯323–4; Needham et al., 1980, pp.╯330–425, esp.╯pp.╯339– 55. 80 ╇ Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯866. Cf. Moulierac, 1987, p.╯88; Kraus [Plessner], “Djābir b. Ḥayyān,” EI² II, 357b. 81 ╇ “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, p.╯45.16–9. 82 ╇ Kitāb al-Tafhīm, tr. and ed. Wright, 1934, p.╯233. 83 ╇ Idem, p.╯234. Of note is also the association ofÂ� Leviathan and Behemoth in the book of Job 40–1 where they appear as representatives of different orders, the one of Â�Â�liquids and the other of solids. In the book of Enoch 6:7–9

the Â�sex of the monsters is differentiated: the female Leviathan dwells in the depths of the sea and the male Behemoth occupies a dry wilderness on the east of Paradise. The Apocalypse of Abraham 10 speaks of male and female Leviathans. Cf. also the apocryphal book of Esther (1:4–11) in which Mordecai dreams of two combatting dragons. 84 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯634. Cf. Ruska, “Ḥayyā,” EI² III, 334b. 85 ╇ Tr. and ed. Davis, 2008, p.╯192. 86 ╇ See the forthcoming publication by the present writer which discusses the dragon iconography from 2500 bc to 650 ad; also Kuehn, 2009, pp. 43–67. 87 ╇ Lane, 1947, pl.╯81 B; Dimand, 1947, p.╯191, fig.╯124; Grube, 1963, p.╯59, fig.╯15; Grube and Johns, 2005, p.╯230,

77 78

...some people say that the dragon’s head is male and diurnal and the tail female and nocturnal.83

the knotted dragon motif An esoteric character of this dragon knot is suggested by its appearance on a hemispherical copper alloy magico-medicinal bowl, preserved in the Royal Ontario Museum in Toronto, which has been dated between the twelfth and fourteenth century.88 The interior of the bowl features a pair of knotted confronted dragons with paired slender sinuous tails, an undulating serpent, an eight-legged scorpion and a quadruped, probably a lion.89 The dragons are characterised by “Saljuqtype” fanged jaws with the conspicuous feature that their tongues are darting out to meet at the centre (fig. 174). The interesting serpent-lionscorpion combination is mentioned in Aḥmad ibn ʿAlī al-Būnī’s (d.╯c. 622/1225) widely circulated thirteenth-century writings on magic.90 It also occurs on Islamic talismans such as the “Lion Seal” described by Ibn Khaldūn.91 The bowl is circumscribed on the exterior in cursive script with a therapeutic inscription which reads: This holy bowl is useful for the sting of a scorpion and of a snake and the bite of a mad dog, for difficulty in childbirth, nosebleed, stomach ache and colic. The afflicted person is to drink from it three times. By the grace of god he will recover. For difficult labour drink saffron water. For stopping nosebleed and abdominal pain, snuff water from it. For colic, gulp down hot water. This has been proven by experience. The work of fig.╯77.9. For a line drawing of the knotted dragons, see Rogers, 1969, p.╯151, fig.╯2. 88 ╇ See a related bowl dated 565/1169–70 from Syria in the Nasser D. Khalili Collection of Islamic Art, London; SavageSmith, 1997, p.╯82, cat. no. 25. 89 ╇ Knotted dragons also occur on another type of early Islamic magic-medicinal bowl, the so-called “poison cups,” which are also of hemispherical form. In addition to the knotted dragons, they also consistently represent a serpent, a scorpion and a quadruped, in appearance something between a dog and a lion (the former would be more appropriate since the bowl serves to protect against the bite of a mad dog). These were used to provide an antidote for poison and to serve as protection against animal bites and stings. Rehatsek, 1873–4, pp.╯150–4; Canaan, 1936, p.╯105, 1938, p.╯146; Ittig, 1982, p.╯82; Savage-Smith, 1997, pp.╯73–4. 90 ╇ Canaan, 1936, pp.╯101–5; Ittig, 1982, pp.╯91–2; SavageSmith, 1997, pp.╯138–9; Porter, 2004, p.╯185. 91 ╇ Ibn Khaldūn states in the Muqaddima: Then there is the ‘Lion Seal’ which is also called the pebble seal. On a steel thimble, the sorcerer engraves the picture of a lion dragging its tail and biting on pebbles which is thus divided into two parts. A snake is represented in front of the lion. It is coiled at the feline’s mouth. Upon the lion’s back, a crawling scorpion is represented. In order to make the engraving, (the sorcerer) waits for a time when the sun enters the first or third decan of Leo, provided (further) that the two luminaries (the sun and the moon) are well and out of

167

Muhammad ibn Yūnus – may God have mercy on him. 92

In addition it is inscribed with Qurʾānic verses (comprising sūras 84, 1–4; 71, 18; and 94, 5–6), further invocations asking relief from labour pains and colic (requesting the help of four genii) as well as a series of nine letters that form the initial part of a different sūra (hence are endowed with a special “sympathetic quality,” or khawāṣṣ). It is significant that although the decorative programme and the inscription suggest that talismanic power was invested in this bowl, the inscription does not attribute this power directly to a planetary representation, but instead emphasises the curative properties of the bowl. The therapeutic efficacy for an ailing person to drink from a bowl that is inscribed with Qurʾānic verses and formulae of blessing is well-known. It may be presumed that just as the medicinal effect derives its power from the contact of water with a vessel inscribed with pious inscriptions and magical invocations, the symbolism of the three animals, the quadruped, the snake and the scorpion, as well as the knotted dragons on this magico-medicinal bowl93 serve a similar function through the agency of water. The charging of water with magico-medicinal healing through material contact with a magical object or talisman94 is also illustrated, for instance, in the magical manual of the Pseudo-Majrīṭī, Kitāb their misfortune … People assume that the person who holds onto it (the charm) has an indescribable power over rulers and is able to have close contact with them, to serve them, and to use them for his own ends… Tr. Rosenthal, 1958, vol.╯3, p.╯163. The same motif is menÂ� tioned in the Ghāyat al-ḥakīm (“The Philosopher’s Goal”), attributed to Abū Maslama Muḥammad al-Majrītī. The serpent-lion-scorpion motif was current in the Islamic east in the early Islamic period. A depiction of just the lion and scorpion, without the serpent, appears on the obverse of two circular pendants from the Nīshāpūr excavation of the Iranian expedition of the Metropolitan Museum of Art (1936 to 1947). Cf. Jenkins and Keene, 1982, cat. no. 7a and b; Allan, 1982b, pp.╯68–9, cat. nos. 60, 61. 92 ╇ Ittig, 1982, p.╯81. 93 ╇ The same “decorative programme” is given also on five other “magic bowls” that were examined by Annette Ittig (1982), but the knotted dragon figures on only one of these bowls. While Ittig suggests that the inscriptions may therefore not necessarily be directly related to the motif of the knotted dragons, the depiction certainly cannot be antithetical to the content of the inscriptions and may belong to a body of magico-talismanic symbols used for related purposes that do not necessarily have to concur. 94 ╇ The use of talismans in the region of Mosul that are written in ink and washed off with water which is then given as a potion to a person who has been bitten by a serpent, scorpion or mad dog, is described by Thompson, 1907, p.╯327.

168

chapter ten

ghāyat al-ḥakīm wa aḥaqq al-natījatain (the socalled Picatrix), which relates that the owner of a finger ring engraved on the bezel with a serpent surmounted by a scorpion will be protected from snake bite; or again that he will be healed from any such bite when the ringstone is placed into a liquid which is then imbibed.95 A comparable conceptualisation connecting the knotted dragons with the subject of medicine and healing governs the depiction on a marble relief plaque of a pair of confronted dragons linked by a quadripartite knot at mid-section with further knotted interlaces above and below. The plaque was found at the darüşşifa of the atābeg Lālā Jamāl al-Dīn Farrukh in Çankırı, outside Ankara, dated by the inscription to Muḥarram 633/1235,96 but is no longer extant (fig.╯175). The serpentine bodies are oriented to the left; their confronted heads, with gaping jaws, almondshaped eyes and cusped ears, reveal sharp teeth

and tongues (the lower head is less legible on account of the deteriorated condition but may be presumed to mirror the upper head).97 In his discussion of the Çankırı darüşşifa knotted double dragon motif Gabriel drew attention some time ago to the reminiscence of the ancient Asklepian symbol (consisting of a single serpent coiled around a staff).98 It is also of note that in some thirteenth-century star pictures the constellation Ophiūchus (Serpentarius), the “serpentholder,”99 is rendered knotted in such a way as to appear almost double.100 This is of special importance for the interpretation of the iconography of the knotted dragons, taking into account that in classical antiquity Ophiūchus was also referred to as Asklepios, the god of medicine par excellence, whose attribute is a pair of intertwined serpents, and as Hygieia, his mythical daughter.101 The constellation Ophiūchus governed physicians and pharmacologists as well as huntsmen and athletes.

╇ The German translation of the Arabic version of the Ghāyat al-ḥakīm, ed. Ritter, 1933, 2. Abh., 10. Absch., f115, states:

two closely related dragons, knotted in a similar manner, but with a large tripartite knot, are found in an Armenian Gospel book of 1304 by the miniaturist Vardan in Astapat, Nakhichevan, Ms. 3722, fols. 290b, 319a, 171a, 280a, 232a. Mnatsakanyan, 1955, p.╯517, fig.╯1023; Armenian Miniatures, eds. Gevorkian and Abgarian, 1996, pl.╯24 (upper left corner). 98 ╇ Gabriel, 1940, p.╯168, n. 1, and fig.╯137. Cf. Süslü, 1987, p.╯641. As noted earlier, the darüşşifa became a place of spiritual healing and snake charming by the Ottoman period. See p. 31, n. 75. 99 ╇ On Ophiūchus, see Scherer, 1953, p.╯184. 100 ╇ Saxl, 1932, vol.╯1, p.╯293, fig.╯344; “Ophis, Ophiuchus,” RE; “Sanitas et duo dracones perplexi” in Gundel, 1936, pp.╯174–6. See Rogers, 1969, p.╯156, n. 17. 101 ╇ Cf. idem, 1969, pp.╯156–7.

95

Man graviert auf einen Ringstein das Bild einer Schlange und darüber einen Skorpion ... wer diesen Ringstein trägt, den beißt keine Schlange und er wird geheilt von ihrem Biß, wenn er ihn in eine Flüssigkeit legt und sie trinkt. Cited after Barb, 1953, pp.╯17–8, n. 119. 96 ╇ Meinecke, 1976, vol.╯2, p.╯103. Cf. general references in Gabriel and Sauvaget, p.╯168, n. 1, fig.╯137; Kühnel, 1950, p.╯8. 97 ╇ The relief has been lost since 1940 and is only known from photographs and a drawing done before this date. Cf. Gierlichs, 1996, p.╯156. Marginal ornaments in the form of

the dragon and the magico-medical sphere

169

chapter eleven

The dragon and the magico-medical sphere

a.╇ The dragon as prophylaxis and cure In Iranian mythology Thrita/Trita was the first of the healers. He “drove back sickness to sickness, drove back death to death; he asked for an antidote (and) obtained it from Khshthra-Vairya to withstand sickness ... and to counter-act snakebite” (Vidēvdāt 20.2–3).1 Thrita, “the third man who pressed the Haoma” (Yasna 9.10), appears originally to have been closely associated (if not identical) with Thraētaona,2 for the invention of the miraculous gift of healing, in other words the granting of health, strength, fertility and fecundity, was also attributed to the latter.3 In the Farvardīn Yasht Thraētaona had the ability to cure certain illnesses and could: ...counteract pain, hot fever, humours, cold fever and incontinence, and […] the pain caused by the serpent.4

Knowledge of the secret causes of illness and the no less secret measures necessary to obtain a cure belonged to the duties of the healer. Accordingly, Thraētaona was also regarded as the inventor of magic.5 In living Zoroastrian observance it is ╇ Dubash, 1906, p.╯173. ╇ The Zend-Avesta, tr. Darmesteter, vol.╯2, 1880, p.╯549, n. 275; Boyce, 1975, repr. 1996, pp.╯98, 100. On the close etymological relation between Thrita/Trita and Thraētaona, see Watkins, 1995, pp.╯314–6; Remmel, 2006, pp.╯126–7. 3 ╇Yasht 13.131. See Boyce, 1975, repr. 1996, pp.╯98, 100. 4 ╇Yasht 13.131 (cited after Dubash, 1906, p.╯173). The Zend-Avesta, tr. Darmesteter, vol.╯4, 1880, p.╯219. Cf. Â�Sarkhosh Curtis, 1993, p.╯26. 5 ╇ When Thraētaona, on his march to Bawri, the capital of Aži [...], arrived at the Tigris (the Rangha); an angel then came and taught him magic to enable him to baffle the sortileges of Aži (Shāh-nāma). We have in this passage an instance of his talents as a wizard, and one which helps us to understand why Thraētaona is considered as the inventor of magic, and his name is invoked in spells and incantations. 1 2

The Zend-Avesta, tr. Darmesteter, vol.╯2, p.╯549, n. 275. Cf. Boyce, 1975, repr. 1996, pp.╯68–9 and n. 3; Bivar, 1967, pp.╯522–3. The association of Bawri (Bāvīr), the fortress of Dahāk, with Babylon seems to be a later tradition, cf. Â�Monchi-Zadeh, 1975, p.╯238 and n. 2. Two further fortresses of Dahāk were said to be located in Simbrān and in

Thraētaona who is invoked, as king Frēdōn/ Farīdūn, in prayers and on amulets to keep away or cure sickness.6 The tenth-century philologist Ḥamza al-Iṣfahānī also records that Farīdūn “constructed amulets, and introduced the antidote (made) from the body of vipers, and founded medicine, and pointed out those extracts of herbs which keep away pestilence from the bodies of animate beings...”7 Manichaean Middle Persian prayer and incantation texts reciting “names of power” frequently mention the “First Physician” prydwn (Frēdūn) in connection with other powerful names known in a magical context, such as Gabriel and “Sabaoth.”8 The parallels in the magical healing abilities of both Thrita and Thraētaona are mirrored by their heroic feats. Both are known to have overcome serpent-bodied, three-headed and six-eyed dragons, respectively known as Viśvarūpa and Azhi Dahāka, the difference being that the former is a celestial and the latter a terrestrial dragon.9 It is interesting to observe that the magical healers who are called upon to cure injuries caused by snake bites and to invent an antidote for snake venom are at the same time dragon fighters par Hindūkān (Bundahishn 209.11–2; idem, p.╯238). 6 ╇ Boyce, 1975, repr. 1996, p.╯98. Cf. Modi, 1894, pp.╯1–24. On the basis of Farīdūn’s attribute of a bull-headed mace with which he breaks the stronghold of the demon-king Ẓaḥḥāk, Bivar (1967, pp.╯522–4, pl.╯1, F) identifies the scene of a hero grasping a demon by the hair engraved on a Sasanian-period chalcedony seal-stone, probably an amulet, in the British Museum, inv. no. 1905-5-30, 1, with the legend of Farīdūn battling with Ẓaḥḥāk. The latter is portrayed in the process of devouring (or expectorating) a human being whose upper body, head and arms protrude from the demon’s mouth. Bivar surmises that the scene represents Farīdūn in his medical role, possibly combating a fatal illness, since the demon is shown as not merely wounding but devouring a human being. 7 ╇ Ḥamza al-Isfahani, ed. Gottwaldt, J.M.P., p.╯23 and ٣٣, as cited in Bivar, 1967, p.╯522, n. 25. Cf. Ḥamza alIsfahani, Taʾrīkh sinī mulūk al-arḍ wa ’l-anbiyāʾ, Beirut, 1961, p.╯34; Ṭabarī, I, p.╯226; Balʿamī, Tarjumat-i tārīkh-i Ṭabarī, ed. Bahār, p.╯148; Ibn al-Balkhī, I, p.╯36; and Tārikh-i guzīda, ed. Navāʾī, p.╯84; cited after Tafaẓẓolī, “Ferēdūn,” EIr. 8 ╇ Henning, 1947, pp.╯39–40. 9 ╇Yasna 9.7; Boyce, 1975, repr. 1996, pp.╯98–9.

170

chapter eleven

excellence.10 Another analogy between the two healer-heroes is provided in the Avesta which relates that the second mortal to press liquid haoma (Ved. sóma) was Āthwya, for which he received the boon of fathering Thraētaona, and the third haoma-presser was Thrita.11 The sacred haoma plant was regarded as chief of medicinal herbs.12 The drink, made of the haoma seed, intoxicated, gave heightened divine powers and mantic wisdom,13 and was used to counter the dragon “who tosses poison around.”14 Further manifestations of the dragon in the realm of prophylaxis and cure are found in epic narratives.15 In Asadī Ṭūsī’s heroic Garshāspnāma,16 Ẓaḥḥāk (the historicised mythological foreign usurper and avatar of the Avestan demon Azhi Dahāka) requests the legendary hero Garshāsp to vanquish a dragon that comes out of the sea and lives on Mount Shekāwand in Kābul.17 Before setting out to fight the dragon, Garshāsp took an antidote/theriac (diryāq), a kind of universal drug which often contained the flesh of serpents (see the discussion in the following section). He finally succeeded in killing the beast with a club carved in the form of a dragon head.18 By slaying the dragon with a weapon carved with his own likeness, the hero thus once again applied the homeopathic (or imitative) principle of similia similibus curantur. After the victory Garshāsp is said to have shed his skin and lost

consciousness. Hence as a result of his triumph over his mighty foe he assimilated some of his adversary’s positive ophidian characteristics, such as the shedding of the skin, an act that is symbolic of long or eternal life. Thus when the hero regained consciousness, he gave thanks to an angel. The dragon’s carcass was transported to the city where Garshāsp’s feat was celebrated and he was honoured as jahān-pahlavān (chief hero), a term closely associated with the Iranian heroic tradition. The event was commemorated in a flag embellished with the representation of a black dragon and a pole surmounted by a golden lion in turn topped by a moon.19 This was passed on to Garshāsp’s descendants and became his family’s coat of arms,20 another allegorical example which suggests that the formidable qualities of the dragon were appropriated by the vanquisher. Doses of diryāq against the dragon’s poison were also taken by Rustam’s son, Farāmarz, and his helper, Bīzhan, before they slew the hissing dragon that lived on the summit of a mountain in India.21 Similarly, the legendary hero Borzū, the son of Sohrāb, resorted to drinking diryāq and milk as an antidote to the dragon’s venom before entering his lair on Mount Zahāb to slay him.22 The ancient association of the serpent with the art of healing23 appears also in the popular tale of the serpent that introduced the healing herb,

╇ The Zend-Avesta, tr. Darmesteter, vol.╯4, 1880, p.╯219. ╇ Boyce, 1975, repr. 1996, pp.╯98–9. 12 ╇ Eadem, p.╯161. 13 ╇ Eadem, pp.╯158–9. 14 ╇Yasna 9.30; cf. Schwarz, 2006, p.╯216. 15 ╇ The serpent is an age-old emblem of chthonic gods of fertility, healing and divination in the ancient Near East. The association of the (serpent-)dragon with medicine and elixirs is also found in the ancient Jewish tradition. According to the Jewish Midrash the angel of healing, Raphael (Rᵉphâʾel, who acted as a physician as well as a binder of demons), was originally called Labbiʾel, who was ordered by God to provide healing for humanity (Midrashim Konen 26–7; Yerahmeʾel 14–5). In addition to his attribute as healer, Raphael also has a chthonic aspect being known as “Prince of Hades” (1 Enoch 20:2; 22:106). Significantly, the Babylonians reported of the constellation ᵐᵘᶥMuš, “serpent” (i.e., Hydra, the seaserpent), that it was the image of the ancient Mesopotamian dragon labbu (see Heidel, 1942, repr. 1951, pp.╯141–2). Astour, 1965, pp.╯236–7; Kuntzmann, 1983, p.╯97; cf. Wilson, 2001, pp.╯30–1. In the Ugaritic texts the Rephaim (Rᵉphâʾîm), the ghosts of the dead in the netherworld, are known for their healing abilities, Baal being known to drive out serpent demons (KTU 1.82:6), and another deity, Horon, to neutralise the effects of snake venom (KTU 1.100:61– 9). See also Becking, “El Rophe,” DDD, 1995, pp.╯292–3; Kuntzmann, 1983, p.╯216. 16 ╇ Ed. Yaghmāʾī, Tehran, 1354/1975; De Blois, “Garšāspnāma (or Karšāsp-nāma),” EIr.

╇ Khāleqī-Moṭlaq, “Aždahā II,” EIr. See also chapter 3. ╇ Asadī Ṭūsī, Garshāsp-nāma, p.╯269, l. 10, as cited in Khāleqī-Moṭlaq, “Aždahā II,” EIr. Cf. the related story of Mūsā killing a giant serpent with his serpent-staff; al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯223. See also the discussion on pp. 38–9. 19 ╇ Widengren, 1969, p.╯17, n. 35. Asadī Ṭūsī, Garshāspnāma, pp.╯49–63, cited after Khāleqī-Moṭlaq, “Aždahā II,” EIr. 20 ╇ Asadī Ṭūsī, Garshāsp-nāma, pp.╯49–63, as cited in idem. 21 ╇ Farāmarz-nāma, London, British Museum, Ms. Or. 2946, fols. 24, 25; Khāleqī-Moṭlaq, 1982, pp.╯22–45; idem “Aždahā II,” EIr. 22 ╇ Borzū-nāma, Paris, Bibliothèque Nationale, Ms. Supp. Pers. 1023, fols. 242, 243; idem. 23 ╇ Pliny (Naturalis Historia, XXIX 4) notes the widely held belief that the snake is full of “many remedies” (inesse ei remedia multa creduntur). This view was shared by the early Christian theologian Cyril of Jerusalem (c. 315–386 ad) in his Catechesis (9.14):

10 11

17 18

Can you know the efficacy of all herbs, or the benefit coming from every animal? Already even from poisonous vipers have come antidotes for the safety of humans. But you will say, “The snake is terrible.” Fear the Lord and he [that is to say, the snake] will not be able to harm you. Cited after Kelhoffer, 2000, p.╯440.

the dragon and the magico-medical sphere

171

From earliest times the serpent-dragon was assoÂ� ciated with poisons as well as antidotes. In different forms snake flesh and other ophidian elements were used as remedies for various kinds of ailment. Among the earliest extant illustrated Islamic manuscripts are two copies of an Arabic text on antidotes derived from snakes and used as a remedy for snake venom.26 As mentioned in its title, the Kitāb al-diryāq (“Book of the Â�Theriac”) concerns the preparation of the theriac (a derivative of the Greek thēriakos), a medicinal remedy

celebrated in antiquity that was considered efficacious against the poison of snakebites or the bites of other wild beasts (from ther, “wild animal”)27 dealing with the effects of snakebites. The earlier of the two copies, dated 595/1199, in the collection of the Bibliothèque Nationale, Paris, since 1853,28 has a double frontispiece.29 The doubling of the near-identical pages, once again, serves to increase and underpin the intended effect, underlining the double potency of the symbolism of the paired images. The pages depict two seated figures each holding up large crescents with two diminutive attendants in princely dress on either side, each enclosed by two confronted dragons and four framing figures that are equally splendidly clad (fig.╯176).30 Flanking the central section of both miniatures horizontally at top and bottom are epigraphic bands in Kufic which read: “Its owner and scribe is the meekest of God’s servants, be He praised, Muḥammad, son of the fortunate Abu ’l-Fatḥ, son of the rightly guided imām, Abu ’l-Ḥasan, son of the beneficent imām.”31 A second double page composition carrying an inscription on the right side states that the book was made for the library of a certain imām Abu ’l-Fatḥ Maḥmūd,32 possibly a nephew of the owner/scribe named on the frontispiece, and apparently a member of the Shīʿī religious class.33 Unfortunately the place of production is not stated. The paintings of the manuscript have been attributed to Iran, but most scholars concur that they originate in the north Mesopotamian (Jazīran) school of Mosul.34 The haloed and crowned female figure on the two miniatures, richly clad and bejewelled, her plaited hair falling over her shoulders, is seated with folded legs and crossed feet. One of the soles faces upwards showing the tips of the hennatinted toes, while the uplifted hands with henna-

╇ AlÂ�-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯639–40. 25 ╇ Hoogasian-Villa, 1966, pp.╯426–9 and 531–2. On Luqmān’s title of ḥakīm in Turklish folklore, cf. Eberhard and Boratav, 1953, p.╯346. 26 ╇ Ms. Arabe 2964 in the Bibliothèque Nationale, Paris, and Ms. A.F. 10 in the Österreichische Nationalbibliothek, Vienna. For a monograph on the Paris manuscript, see Farès, 1953; for a discussion on the Vienna manuscript, see Holter, 1937, pp.╯1–48, and Duda, 1992, pp.╯46–69. 27 ╇ A recipe for theriac is recorded in Moulierac, 1996, pp.╯102–3. Its production is visualised in Ms. Arabe 2964, p.╯5, Bibliothèque Nationale, Paris. 28 ╇ The date given in the colophon is Rabīʿ al-awwal of the year 595/31 December 1198–29 January 1199, copied by Muḥammad Abu ’l-Fatḥ ʿAbd al-Waḥīd. Paris, Bibliothèque Nationale, Ms. Arabe 2964. Farès, 1953, pls. III– IV.

╇ In the course of reassembling the manuscript with missing folios, the double frontispiece has been repaginated to pp.╯36 and 37, see Duda, 1992, pp.╯48–9. 30 ╇ Cf. Grube, 1967, pl.╯27 (colour illustration); Grube and Johns, 2005, p.╯230, fig.╯77.1. 31 ╇ Pancaroğlu, 2001, p.╯155 and n. 6, with the transcriptions of the Arabic text. In spite of the prestigious titles the name of the scribe is unknown to chroniclers and historians; Moulierac, 1987, p.╯84. 32 ╇ Farès, 1953, pp.╯8–9, pl.╯V; Pancaroğlu, 2001, p.╯155 and n. 7, with transcr. of the Arabic text and Â�English tr. 33 ╇ See Pancaroğlu, 2001, p.╯157 and n. 14. 34 ╇ An Iranian origin has been suggested by MelikianChirvani (1967) who however does not exclude the possibility of an Artuqid (Jazīran) provenance. Farès (1953) attributes the paintings to the “school of Baghdad.” Ettinghausen supports a Jazīran provenance (1962, pp.╯86, 92), which was strengthened by Nassar (1985) and Ward (1985).

called Persian basil, to the Sasanian king Khusraw I (r. 531–579), known as Khusraw Anūshirwān, recorded by al-Qazwīnī in his encyclopaedic treatise Kitāb ʿajāʾib al-makhlūqāt wa gharāʾib al-mawjūdāt. The king had saved a large serpent that crept under his throne and had sent one of his horsemen to protect it from the bite of a scorpion. The following year, the thankful serpent came again before the throne of the king and cast out of its mouth a black seed. This seed the king ordered to be sown and a basil plant grew out of it which the king used to cure his “cold in the head and pains in the brain.”24 More all-encompassing is the popular Armenian tradition which records that a serpent king endowed the archetypal physician Luqmān al-ḥakīm (after whom sūra 31 is named) with the knowledge to cure all diseases. But seeing that the knowledge would enable men to achieve immortality, God became jealous and ordered an angel to empty out Luqmān’s elixir and cast his books into the sea.25 b.╇ The dragon and the theriaca as illustrated in the Kitāb al-diryāq

24

29

172

chapter eleven

tinted fingertips hold a large crescent-shaped moon. The centrally facing figure is flanked on either side by two small attendants.35 The composition is encircled by two confronted dragons whose bodies are knotted at the four cardinal points; at the top the heads, with curved horns, wrinkled snouts and wide-open jaws revealing tongues with bifid tips, extend beyond the circumference, while the slender coiling tails form an additional loop before tapering to a point at the base. The corners of the composition are filled with four winged figures, of presumably honorific and celestial significance,36 that hold up the medallion. The central figures and the four framing figures are distinguished by a halo. The central frontally rendered figure, possibly a symbolic personification of the planet Moon (al-qamar), was a popular motif at this time37 and one that was “generally invested with astrological and semi-magical significance.”38 The now destroyed thirteenth-century Sinjār Gate of Mosul was decorated with a relief portraying a figure holding up a crescent moon.39 The emblem apÂ�pears to have been of some importance since it is shown on the coinage of Badr al-Dīn Luʾluʾ as well as on coins of other rulers of Mosul and Sinjār (between 585 and 657/1189 and 1258) and of Saladin (Mayyāfāriqīn, 587/1191).40 In his monograph on the manuscript, Bishr Farès perceives the lunar emblem as Nin-gal, the “Great Lady” and divine consort of the moon-god Sīn, whose attribute was the crescent, and proposes that the iconographic expression was trans-

mitted through the Ṣābian cult of the moon-god at Ḥarrān41 whose cult flourished well into the Islamic period. Farès associates the lunar figure with the content of the text, which he construed to be a prophylactic or talismanic device reinforced by healing powers associated with serpents against the evil powers of disease in ancient Mesopotamian and Graeco-Roman beliefs.42 The iconography of the Pseudo-Galen double miniature has been studied by Guitty Azarpay, who identified the depiction of the dragons juxtaposed with the anthropomorphic lunar emblem as eclipse dragons, the head and the tail of aljawzahar.43 This argument is strengthened by the fact that a partial solar eclipse did occur in the Near East on 29 Rabīʿ al-awwal 595/28 January 1199, a date which corresponds with that of the completion of the manuscript, indicating that the representations on the double miniature were intended as apotropaic devices against the potential astrological threats imminent at the very time of its production.44 Yet rather than associating the imagery of al-jawzahar, portrayed with the personified moon emblem, with the actual occurrence of a solar or lunar eclipse, Pancaroğlu suggests, as proposed earlier by Farès, that the depiction has in fact a wider meaning and, more precisely, may also be semantically linked with the contents of the manuscript.45 Although probably being a pseudo-Â�epigraphical original Arabic work by an anonymous author, the text of the Kitāb al-diryāq is falsely attributed

╇ Azarpay (1978, pp.╯364–5) relates the lunar personification to Graeco-Roman and Byzantine traditions in which the Moon is female, which, she suggests, is generally perpetuated in depictions of the symbol in Islamic art. On the female nature of the Moon in Roman astrological imagery, see also Markel, 1995, p.╯83, in contrast to it generally being thought of as male in Indian mythology, idem, pp.╯32–8 and n. 33 (for an exception see, idem, p.╯152); Pingree, 1964–5, p.╯250. The henna-tinted fingernails and tips of the fingers and toes of the figure clearly identify her as female. See also the depiction of female moons on a late twelfth-century Tell Minisstyle bowl, now preserved in the Paris, Musée du Louvre, Département des Antiquités orientales, Section Islamique, inv. no. OA 7872. L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯232, cat. no. 159. 36 ╇ Azarpay, 1978, p.╯366, n. 19; Pancaroğlu, 2001, p.╯163. 37 ╇ See, for instance, on twelfth- and thirteenth-century metalwork (Saxl, 1912, p.╯164, fig.╯10; Pope and Ackerman, eds., 1938–9, repr. 1964–81, vol.╯6, pls. 1327, 1331) and miniaÂ�ture painting (e.g. al-Qazwīnī, ʿAjāʾib al-makhlūqāt, Munich, Bayerische Staatsbibliothek, Cod. Arab. 464, reaÂ� lised during the author’s lifetime in Wāsiṭ in 678/1279–80; Saxl, 1912, fig.╯8). 38 ╇ Azarpay, 1978, p.╯364. 39 ╇ Sarre and Herzfeld, 1920, vol.╯2, pp.╯213–5; Azarpay, 1978, p.╯365, fig.╯3; Hauptmann von Gladiss, ed., 2006,

fig.╯22. The lunar figure was interpreted by scholars as the emblem of the city of Mosul as well as the coat of arms of Badr al-Dīn Luʾluʾ (whose sobriquet was “new moon” or “full moon of religion”). Rice (1957, pp.╯321–2) however has disputed this association on the basis that the motif was not restricted to Mosul; cf. Ettinghausen, “Hilāl. ii. – In Islamic art,” EI² III, 379a. 40 ╇ Such as the copper coinage of the Zangid rulers of Mosul, ʿIzz al-Dīn Masʿūd I ibn Mawdūd (576/1180– 589/1193) (cf. What the Coins Tell Us, 2009, p. 32, 15722 and 15719) or Nāṣir al-Dīn Maḥmūd ibn Masʿūd (616/1219– 631/1234) (dated 627/1229; American Numismatic Society; cf. Ettinghausen, “Hilāl,” EI² III, 379a, fig.╯4). See also Lane Poole, 1877, nos. 529–33, 567–9, 589–92. The lunar emblem is also shown on the coinage of the ʿAbbasid caliph al-Nāṣir (577/1181–620/1223); Hauptmann von Gladiss, ed., 2006, p.╯106, cat. no. 12. 41 ╇ Farès, 1953, pp.╯22–4, 26–7, 33. 42 ╇ Idem, pp.╯29–33. 43 ╇ Azarpay, 1978, pp.╯363–74, and eadem, 1991, pp.╯1–10. 44 ╇ Eadem, 1978, pp.╯363–74. On the apparent discrepancy of the choice of representing a lunar eclipse when in fact it was a solar eclipse that had occured, see Pancaroğlu, 2001, p.╯164. 45 ╇ Pancaroğlu, 2001, p.╯164.

35

the dragon and the magico-medical sphere

173

to the second-century physician Galen with the commentary of Yaḥyā al-Naḥwī (John the Grammarian/Johannes Grammatikos or Johannes Philoponus) of Alexandria (c. 490–565),46 and credited to the most renowned scholar of the translation movement, the Nestorian Christian Ḥunayn ibn Isḥāq al-ʿIbādī (c. 192/808–264/877).47 It is a literary “hybrid”48 comprising the biographical sketches of nine classical physicians – among whom are the physicians Galen (Jālīnūs), AndroÂ�machus the Younger (Andrūmakhus al-Qarīb al-ʿAhd), who together with his father was active at the court of the Roman emperor Nero, and the early fourth-century Philagrios (Aflaghuras) – their theriac recipes as well as their vitae, followed by a section on snakes and their classifications and further medicinal recipes.49 In the third painting of the Paris Kitāb al-Â�Â� diÂ�ryāq, Philagrios is shown in the process of preparing an antidote in a large vessel placed on a stand over a fire which is fanned by his assistant.50 The onlookers comprise men and women of different age groups. Seated at the lower right corner of the scene, an emaciated figure clad only in a loincloth, probably the patient, gestures to a dragon depicted just below the physician. The creature has large, wide-open jaws and its body is composed of two intertwining serpent coils. It may thus represent an ingredient in the preparation of the antidote, as indicated in the account of Andromachus the Younger that is given in the text, discussed below. Alternatively, the entwined coils may represent the caduceus-like staff, a symbol that was known in the medieval Islamic world. Writing in the thirteenth century, the physician Ibn Abī Uṣaybiʿa (599/1203–669/1270) records in his dictionary of over 380 biographies of physicians, ʿUyūn al-anbāʾ fī ṭabaqāt al-aṭibbāʾ

Other medieval Arabic sources record that Â�AskleÂ�pios’ staff was wreathed with entwined Â�serpents and, moreover, that Agathodaimon (AghāthūdhīÂ�mūn), who was considered a great authority in the occult sciences, was Asklepios’ teacher.54 In spite of their dangerous or even life-threatening qualities, serpents have long been considered to have medicinal benefits against a variety of afflictions and, as stated in the account of AndroÂ�machus the Younger included in the text of the Kitāb al-diryāq, were an essential ingredient in the theriac, into which he “added many drugs ... and made a single theriac … [he also]

╇ Meyerhofer, 1932, pp.╯1–21. ╇ Cf. Ullmann, 1970, p.╯49; Johnstone, “Summ,” EI2 IX, 872a; Pancaroğlu, 2001, p.╯156 and n. 9; Kerner, 2004, p.╯167. 48 ╇ Kerner, 2007, p.╯25. 49 ╇ In addition to the two frontispiece miniatures there are nine text miniatures and a set of illustrations of serpents and plants that have a direct bearing upon the content of the text. Farès, 1953, pls. VII–IX: portraits of physicians of antiquity, XI–XVI: themes and anecdotes treated in the text, XIX: a table of serpents, XVII–XVIII: specimens of plants. See also Moulierac, 1996, p.╯101; Kerner, 2004, pp.╯3–4 and n. 3. 50 ╇ Paris, Bibliothèque Nationale, Ms. Arabe 2964, p.╯15. Colour reproductions in À l’ombre d’Avicenne, 1996, p.╯156, cat. no. 87, bottom, and Pancaroğlu, 2007, p.╯26, fig.╯12. 51 ╇ Rosenthal, 1956, pp.╯67–71.

52 ╇ Idem, p.╯70 and n. 72. The denomination drakōn is thought to be derived from the Greek word derkomai, “to see” (see also Porphyrios, De Abstinentia, III. 3,8); the sacred serpent is said to be the guardian of the temple of Asklepios because it is the most watchful of animals; according to Cornutus, Lucius Annaeus, “snakes are symbols of Athene, because they look frightening and are vigilant and sleep little” (Theologiae Graecae Compendium, 20); similarly, it is associated with Asklepios because of its ability to self-rejuvenate, to discard age, as well as its vigilance; because patients require attentive care (Theologiae Graecae Compendium, 33). 53 ╇ The Greek gerās (“old age”) is also used to denote the cast-off skin of snakes. Rosenthal, 1956, pp.╯70–1 and n. 76. 54 ╇ Ullmann, 1972, p.╯176 and n. 6; Plessner, “Â�AghāÂ�thūÂ�dhīmūn,” EI² I, p.╯247a.

46 47

(“Sources of Information on the Generations of Physicians”), that according to Galen’s account: Asklepios is represented holding in his hand a carved staff with branches, made from the marsh mallow tree. … Upon it, there is represented a long-lived animal, wound [i.e., coiled] around it – a snake.51

He explains that the association of Asklepios with the serpent is due, firstly, to: ...the fact that the snake is a sharp-sighted animal which is much awake and never sleeps. Thus the student of the craft of medicine must not be detracted by sleep, and he must possess the keenest possible mind in order to be able to warn (his patients) in advance of present (conditions) and (those) likely to arise in the future.52

He offers a further explanation, namely that the serpent has a long lifespan, perhaps possessing eternal life and hence that: those who employ the craft of medicine are able to live long. … This animal – the snake – sloughs off its skin, called by the Greeks “old age.” Likewise people by employing the craft of medicine, are able to slough off old age, the result of disease and regain health.53

174

chapter eleven

mixed into the theriac the flesh of snakes ...”55 Three anecdotes attributed to Andromachus the Younger serve to substantiate the therapeutic properties of snake flesh. These comprise the use of snakes dissolved in wine, a mixture reported to have cured a man suffering from leprosy (known as elephantitis),56 the homeopathic effect of snakes dissolved in water upon a snakebite, and similarly a snakebite serving as an antidote against an opium overdose.57 The fourteenth-century scholar al-Damīrī records in his well-known bestiary that “it is unlawful to eat serpents on account of their injurious property, and an antidote prepared out of their flesh is also unlawful.” He based this on the tenth-century traditionist and Shāfʿī specialist Abū Bakr al-Bayhaqī, remarking that it is forbidden by the Shāfʿī school, “excepting in circumstances of great necessity.”58 In spite of these cautionary remarks snakes apparently continued to be a choice ingredient in the theriac or other concoctions, as evidenced also by the same scholar who cites that according to ʿĪsā ibn ʿAlī: The flesh [of snakes] preserves the senses, and soup (gravy) made of its flesh strengthens the sense of sight. The flesh of serpents from a part 55 ╇ Kerner, 2004, pp.╯232–3. The practice of consuming snake meat to cure a variety of illnesses has a very long history. The first-century medical writer Cornelius Celsus (d.╯c. 50 ad) reports the experience of farmers who cured abscesses in the lymphatic glands by eating snake flesh (De medicina 5.28.7b); cf. Hanson, 2006, p.╯497. The ingredient was formalised when the Roman emperor Nero (r. 54–68 ad) developed an interest in an antidote formulated by Â�Mithridates VI (120–163 bc), king of Pontus, for which a Â�Persian physician had probably supplied the recipe and which the king used daily as a preventive measure against poisoning. He asked his “Leibarzt,” Andromachus the Elder, to investigate and improve the compound, which subsequently was modified by the addition of snake flesh as a further ingredient, thus creating what is generally regarded as the first true theriac, the Theriaca Andromachi, being an early example of the homeopathic principle of similia similibus curantur (Bierman, 1994, p.╯5). Thereafter, the use of snake flesh as a cure is recorded by numerous physicians, such as Aretaios and Galen (Oberhelman, 1994, p.╯943 and n. 9). It is also reported in the writings of Eznik of Koghb (Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, pp.╯574–5, ch. 64). Moreover, it is known that the ingredient was used by Jewish physicians of the medieval period when they produced the universal drug:

...ṭariyāqā in which is mixed the flesh of the viper and the eggs of that animal that is called. …(?) mix it with dry ground leavened bread, roll it into oven loaves and dry them with the other ingredients in the ṭariyāqā. …then knead it with honey. Gil, 2004, pp.╯605–6. 56 ╇ Manuscript of the Kitāb al-sumūm (“Book of Poisons”)/

of their bodies in which there is much of it warms, dries, and clears the body, and removes diseases from it. [...] if it is burnt and kneaded with good olive oil, and a painful carious cavity in a tooth is stuffed with it, it will cure it; if it is pounded fine together with its head and applied over parts affected with alopecia, it will cause hair to grow again.59

AlÂ�-Damīrī notes further recipes from other authorities: If the fried slough of a serpent is taken and the bark of the root of capers, long birth-wort (aristolochy), and anacardium in equal parts are mixed with it, and then a person suffering from both external and internal hanging piles is fumigated (with the mixture), they will fall off. [...] if the slough of a serpent mixed with the reddish tinged bdellium both external and internal piles are fumigated, they will be cured. If an egg of a serpent is pounded with nitre and vinegar, and then applied over fresh patches of white leprosy, it will remove them. [...] [The serpent’s] heart if it is hung on the body, will cure quartan ague.60

There was a profitable trade in the importation of vipers and other snakes (afʿā) chiefly from Sijistān (Sīstān), a region known for its great Kitāb al-diryāq (a combination of the Kitāb al-diryāq with another toxicological treatise, the Kitāb al-sumūm), sold at Sotheby’s Arts of the Islamic World, London on 25 April 2002, lot 30; the unillustrated manuscript of the Kitāb al-diryāq is dated 8 Jumādā al-ūlā, 622/19 May, 1225, with a tentative provenance of Mesopotamia or western Iran, fol. 44b. Kerner, 2004, pp.╯17–8, 86, and eadem, 2007, p.╯32. The anecdote is also recorded by the eleventh-century Christian physician Ibn Buṭlān, see Browne, 1921, pp.╯72–3. A related account of a serpent “vomiting” into a bowl of maḍīra (a kind of broth made with sour milk) which is drunk by an epileptic boy who was subsequently cured is attributed to al-Rāzī; al-Faraj, vol.╯2, pp.╯103–4, and Ibn Abī Uṣaybiʿa, vol.╯1, p.╯312, cited after Browne, 1921, pp.╯75–6. Yet another anecdote is related by a man called Abū ʿAlī ʿUmar ibn Yaḥyā al-ʿAlawī, which recounts that a man was cured from leprosy by eating roasted snake meat after having cut off the heads and tails; al-Faraj, vol.╯2, p.╯100, cited after, idem, 1921, p.╯76. Cf. Massé, 1938, vol.╯1, p.╯207. 57 ╇ Kitāb al-sumūm/Kitāb al-diryāq, op. cit., fol. 45b, 46a. Kerner, 2004, pp.╯17–8, 88–9. Cf. Pancaroğlu, 2001, pp.╯160–1, figs. 6–8. Andromachus the Younger is shown together with a man who is bitten by a serpent in Ms. Arabe 2964, p.╯19, Bibliothèque Nationale, Paris. 58 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯649. In this connection it is worthy of note that in the Shāh-nāma account of Iskandar’s adventures the consumption of tinnīn, which in spring fall down from Heaven, is associated with Yājūj and Mājūj (Gog and Magog), the apocalyptic peoples known from biblical (Ezekiel 38.39, Apocalypse 20.7–10) and Qurʾānic eschatology (sūras 18, 93–8; and 21, 96); tr. and ed. Mohl, 1838–1878, vol.╯5, p.╯223, ll. 1467–1468. 59 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯654. 60 ╇ Idem, p.╯654.

the dragon and the magico-medical sphere

175

number of vipers,61 since theriac was prepared from their flesh.62 Moreover, the city of Arīhā (Jericho) is named “the home of the theriacal serpents, and the excellence of the theriac of Jerusalem owing to the use therein of the flesh of those serpents” in an addition to a thirteenthcentury manuscript copy of al-Muqaddasī’s geography, Aḥsan al-taqāsīm fī maʿrifat al-aqālīm (“The Best Divisions for Knowledge of the Regions”), preserved in the Süleymaniye Kütüphanesi, Istanbul.63 In the second extant copy of the Kitāb al-diryāq, in the Österreichische NaÂ�Â�Â� tionalbibliothek in Vienna, a miniature in the herpetological tract illustrates the procedure of serpent hunting for the flesh of snakes used in the remedy.64 The serpents were made to bite into leather-“dolls” with glass eyes for them to discharge their venom, making them easier to handle.65 The medicinal properties of serpent flesh were also known in Buddhist Central Asia where in the Swāt region Lord Śakra (Indra), an incarnation of the Buddha, sacrificed himself for the good of the people by transforming himself into a great serpent whose dead body filled the entire valley of Swāt. Since his flesh was a remedy for every kind of disease, the sick were thus delivered from their afflictions.66 This legend has certain affinities with an assortment of tales related to the Queen of the Serpents which is included in Alf layla wa-layla collection of fairy-tales and other Â�stories.67 In the tale the Serpent-Queen consents

to be slain and to sacrifice her flesh in order to heal the king. However just as in Andromachus the Younger’s theriac recipe recorded in the Kitāb al-diryāq text, the consumption of snake flesh is seen here specifically as a cure for leprosy with which the king is afflicted.68 In the tale the flesh of the reptile is to be boiled and then served to the leper. Moreover drinking the elixir, in other words the broth in which the flesh was cooked (after skimming off the “first scum” or foam which appears in the process of boiling), is said to give access to the fountains of knowledge, as well as various sciences.69 The consumption of snake flesh as an antidote to snake venom is recommended by ʿAlī ibn Waḥshiyya al-Nabaṭī in his ninth-century text on poisons. He notes an antidote attributed to a certain Ḥajujā which involves consuming the flesh of a viper (two ounces) cooked together with bread (samīd) and sipping the meat broth.70 The benefits of serpents, in particular the healing properties of snake flesh as general antidote to its venom, is also stated in the tenth-century texts of the Ikhwān al-Ṣafāʾ.71 The symbolism of the knots formed by the two dragons on the frontispiece may be related to the number symbolism of four which, as Meisami writes, “is the number by which the categories of nature are subdivided: the four elements, the four humours, and the four seasons.”72 Medieval IslaÂ� mic medicine, based largely on Greek medicine and natural history, in particular the Galenic

61 ╇ Cf. idem, p.╯56; Monchi-Zadeh, 1975, pp.╯117–8, see also n. 5. 62 ╇ Kopf, “Afʿā,” EI² I, 214b. 63 ╇ Copied in 658/1259. Istanbul, Süleymaniye Â�Kütüphanesi, Ayasofya 2971. Al-Muqaddasī, Aḥsan altaqāsīm fī maʿrifat al-aqālīm, tr. and ed. Collins, 1994, p.╯158. 64 ╇ Vienna, Österreichische Nationalbibliothek, Cod, A.F. 10, fol. 22b. Duda, 1992, vol. 2, fig. 39; À l’ombre d’Avicenne, 1996, p.╯100. Interestingly, as Jaclynne Kerner (2007, p.╯35, n. 92) has pointed out, the illustration sheds light on the common misperception that snake venom (which forms the basis for modern antidotes to snakebite/antivenins) was an ingredient in the theriac. However, snake venom was known to be beneficial and to act as an antidote in pre-Islamic times (see p. 157, n. 125). It was also administered in ancient India. In the chapter on the use of poison as drug recorded in the famous sixth- to seventh-century treatise Aṣṭāṅga Saṃgraha by Vāgbhaṭṭa the Elder, probably composed in Ujjain, in western Madhya Pradesh, it states that the person who has consumed vegetable/mineral poison should be treated with snake venom (48.3.5). The same is noted in ch. 36 of the Aṣṭāṅga Hṛdaya by Vāgbhaṭṭa the Younger. I am indebted to Dr Jouhar Kanjhirala Adam for this Â�information. The use of snake venom as antidote may furthermore be gauged from Indian lore, such as the episode of the boyhood of Bhīmasena and the tale of king Nala who, being possessed by an evil spirit, was healed by

the bite of the Nāga Karkoṭaka; see Vogel, 1926, pp.╯71–4, and 80–1, respectively. 65 ╇ Kitāb al-sumūm/Kitāb al-diryāq, op. cit., fol. 49a. Kerner, 2004, pp.╯17–8, 98. 66 ╇ Si-yu-ki, vol.╯1, tr. Beal, 1884, repr. 2000, pp.╯125–6. Cf. Carter, 1992, p.╯70, n. 24. 67 ╇ 483rd to 536th Night [1830 Calcutta ed. count]. The Book of the Thousand Nights and a Night, tr. Burton, 1885, vol.╯5, pp.╯324–411. See also Marzolph and van Leeuwen, 2004, pp.╯348–50. 68 ╇ 534th to 535th Night; The Book of the Thousand Nights and a Night, tr. Burton, 1885, vol.╯5, pp.╯404–10. Cf. Brandenburg, 1973, p.╯49. 69 ╇ 535th Night. The Book of the Thousand Nights and a Night, tr. Burton, 1885, vol.╯5, pp.╯407–10. Cf. Marzolph and van Leeuwen, 2004, pp.╯349–50. 70 ╇ Levey, 1966, p.╯69. Another form of consumption is recorded in the Aṣṭāṅga Saṃgraha. It states that a hooded serpent, which has been enraged and found emitting fumes from its mouth, should be made to bite many times on a piece of meat fastened to the tip of a stick. After carefully deciding the strength of the poison, the physician should administer the powder of this meat to the patient who has not been cured by any other anti-poison medicines (48.19.20). 71 ╇ Tr. and ed. Dieterici, 1858, pp.╯89–91. 72 ╇ Cf. Meisami, 1993, p.╯166.

176

chapter eleven

humoral physiology and pathology, was informed by a structure of sympathies. The four elementary qualities or natures (ṭabāʾiʿ) (hot, cold, moist and dry) were regarded as the crucial constituents of all things, whether metals, minerals, plants, or animals; these qualities traditionally correspond to other quaternary groupings such as the four Aristotelian elements of earth, fire, air, and water, and to the Hippocratic-Galenic cardinal humours (al-akhlāt) which consisted of phlegm, yellow bile (choler), black bile (melancholy) and blood.73 The eleventh-century historian of the Â�GhaznaÂ�wids, Bayhaqī, stresses the importance for man to “understand that he is composed of four elements, which must be maintained in equilibrium.”74 The actions of the poisons in the body were explained as an imbalance (iʿtidāl) of the four humours, or the entire constitution.75 Moreover, as Meisami states, “the four elements and humours (i.e., the bases of all creation), when maintained in equilibrium, are symbolised by the image of a square within a circle.”76 Since this text is largely associated with medicinal preparations, Oya Pancaroğlu argues that the four winged figures which surround the central figure may have been intended to symbolise the four elements and their four humoral counterparts which, by extension, may also be reflected in the four knots formed by the two dragons that may be read as a reinforcement of this number symbolism.77 However, since the pillars of Galenic humoral physiology and pathology are linked to worldly existence and its material manifestations, they may also be more closely associated with the motif of the dragons with fourfold knot. This motif would lend itself as a more likely association since the representation of the dragon often served to embody the world and its material expressions.78 The poet al-Fāryābī accordingly defines the four senses that bind humans to the world as four dragons:

...the dragon represents the “sky,” that is, the material world; its four feet the four elements, its seven heads the seven planets.80

It is, moreover, recorded that above the throne of the Ghaznawid sulṭān Masʿūd, who although being ethnically Turkish was deeply imbued with Persian and Islamic courtly traditions: ...a gold-plated chain [was] hung from the ceiling of the chamber containing the dais, and came down over the dais where the crown and the throne were. The crown was attached to this chain, and there were four bronze figures fashioned in the shape of human beings and mounted on columns which were secured to the throne itself, so that their hands were outstretched and thus held the crown safely.81

Hence, the number of four figures holding up the crown above the throne may be indicative of standard patterns for enthronement scenes that were current in the Central Asian world. It may thus be possible to view the four winged figures framing the personification of the Moon that is enclosed within interlaced dragons of the Kitāb al-diryāq as representing the heavenly sphere, while by contrast the serpent-dragon, with its symbolism evoking potent therapeutic talismanic devices directed against the “demons” of illness, is related to the medical sphere. c. Studies on the properties of serpent(-dragons) and the effects of their venom

In a similar vein, in Niẓāmī’s Haft Paykar:

Based largely on late antique forms of Greek medical knowledge with their magical practices and beliefs, medieval Islamic medical science, pharmacology and toxicology likewise maintained a close link between science and supernatural medical paradigms,82 as reflected in the Kitāb al-diryāq.83 Hence supernatural and magical explanations of diseases were often sought. However, while magic spells as remedies were usually prohibited they could exceptionally be resorted to in special cases such as snake or scorpion stings.84 The use of both natural (ʿilm sīmiyā) and

73 ╇ Cf. Browne, 1921, pp.╯120–1; Needham and Ling, 1954, p.╯459; Dols, 1984, pp.╯10–24, esp.╯10–1. 74 ╇ Meisami, 1999, p.╯82. 75 ╇ Johnstone, “Summ,” EI2 IX, 872a. 76 ╇ Meisami, 1993, p.╯166. 77 ╇ Pancaroğlu, 2001, p.╯164 and n. 30. 78 ╇ Daneshvari, 1993, p.╯19. 79 ╇ Al-Faryābī, Dīwān, ed., Bīnish, T., Tehran, 1958, p.╯43, as cited in Daneshvari, 1993, p.╯16, n. 7.

80 ╇ Tr. Wilson, 1924, vol.╯2, 1645, n. 1616, and Dastgirdī, V., Haft Paykar, Tehran, 1313, p.╯244, n. 1; as cited in Meisami, 1987, p.╯228, n. 56. 81 ╇ Bosworth, 1963, pp.╯135–6. 82 ╇ Moulierac, 1987, p.╯88.i. 83 ╇ Cf. Bürgel, 1988, pp.╯33–7. For the lands of the Â�Fertile Crescent, see also Farès, 1953, p.╯26, and idem, 1959, p.╯162. 84 ╇ Cf. Ullmann, 1978, pp.╯2–5.

This human who is the highest of creations Is constantly in the throes of the four dragons [that make up his senses]. 79

the dragon and the magico-medical sphere

177

supernatural (siḥr) magic which was thereby involved may largely be seen as a reflection of the beliefs and practices current in contemporary society.85 The relationship between pharmacology, toxicology and other occult sciences is evidenced by the attribution of one of the earliest and most complete known works in Arabic on the scientific study of poisons, their detection and actions, and the treatment of the conditions they cause, the Kitāb al-sumūm wa dafʿ maḍārrihā (“Book on Poisons and the Prevention of Their Harm”), to the renowned alchemist Jābir ibn Ḥayyān which was probably written about 900 or earlier.86 Another important text on poisons in Arabic ascribed to ʿAlī ibn Waḥshiyya al-Nabaṭī who lived in the second half of the ninth century similarly exhibits a mixture of science and magic.87 Texts on pharmacology and toxicology, in particular, contained ideas infused with late antique concepts of magic based on the Hermetic notion of a unified cosmos of independent forces.88 The subject of toxicology intersected above all with the Hermetic tradition of late antiquity, Â�forÂ�muÂ�Â�Â�lated in writings attributed to Hermes Â�Trismegistos (including a number of astrological treatises) which had great influence on magical, divinatory and alchemical discourse in the medieval era.89 This body of discourse emphasises the close relationship or cosmic “sympathy” between the divine and the physical world, spirit and matter, between the hidden and the seen, and was aimed at gaining intellectual and spiritual mastery of the cosmos by tapping into supernatural forces.90 An Arabic text ascribed to Hermes Â�Trismegistos on the venom of serpents and other poisonous

creatures, translated and compiled from the Greek most probably in the ninth century and in circulation throughout the medieval period, reveals the association between toxicology and the Hermetic notions of sympathies.91 The belief in a link between the terrestrial and the celestial informed in particular the description of the serpents, their venom and the theriac antidotes, which were systematised in the text according to their correspondence to the twelve constellations of the zodiac and the planets. Among the serpents corresponding to the description of the planets is a dark green or black and saffron-coloured serpent which represents the head and tail of the planet of the dragon (al-jawzahr).92 A list of fantastic serpents follows, among them a deathly red serpent with black wings living in the air93 and a large marine serpent with branching horns like those of a stag and a mane like that of a seahorse. We learn further that this marine serpent has a black neck, a white head, a red belly and a multicoloured back; that it can be caught with the help of music; that a stool made with its vertebra can both cure the sitter of podagra and serve to protect the house in which it stands from vermin, and that this serpent’s head yields stones that are used for talismans.94 Another deathly dark-green serpent from the Egyptian desert is humanheaded with a curly beard.95 A serpent from the mountains of Inner Armenia has a fishtail, a bird’s head and is adorned with many colours, while another from the same region has two breasts and, if not deathly, is able to inflict harm.96 The most vicious of all serpents is the pale yellow “Basiliskos” (“the queen”), whose head is crowned by tufts of hair. Whoever so much as sees it dies

╇ Levey, 1966, pp.╯10–1. ╇ Tr. Siggel, 1958, p.╯3; Levey, 1966, p.╯16. 87 ╇ Poison is clearly defined as “overpowering in its naÂ�Â� ture,” arising “from the mixing of the soul in its makeup with its uniting substances according to the influence of the stars,” destroying “that which is called the life-force,” affecting the bodily organs, prohibiting breathing, and ending ultimately in death. Levey, 1966, pp. 11, 15, 25–6; see also Sezgin, 1971, pp.╯318–29. 88 ╇ O’Connor, 1994, pp.╯21, 52–5, esp.╯53. 89 ╇ Cf. Peters, 2004, pp.╯189–90. 90 ╇ A large body of this literature probably originated in the city of Ḥarrān (ancient Carrhae) in northwestern Â�Mesopotamia, a major centre of ancient scholarship, in parÂ� ticular in astrology and astronomy (Ullmann, 1972, pp.╯289– 93; Massignon, 1950, repr. 1981, pp.╯384–400). During the ʿAbbasid period Ḥarrānian scholars were present in large numbers in Baghdad. Many citizens of Ḥarrān, claiming to be Ṣābians, resisted conversion to Islam and the city remained largely pagan until the early eleventh century. The Ṣābians, whose liturgical language was Syriac (Chwolsohn, 1856, vol.╯1,

p.╯159), provided some of the greatest translators and Â�scientists of Islam. In his classic study on the Ṣābians, Daniil Chwolsohn enumerates some thirty Ṣābian Â�scholars comprisÂ�ing astronomers, philosophers, doctors and mathematicians (1856, vol.╯1, ch. 12); al-Battānī al-Ḥarrānī al-Ṣābiʾ, one of the most renowned Arab astronomers, was born before 244/858, probably at Ḥarrān, into a family that formerly professed the Ṣābian religion. The theology of the Ḥarrānians, who came to be known as the Ṣābians after a visit by the caliph al-Maʾmun, is Babylonian in origin, and is a complex blend of polytheism, Gnosticism, Â�Mithraism, Hellenistic Â�Neopythagoreanism and perhaps even Indian cultural components in a synthesis catalysed by Hermeticism. Cf. Â�Marquet, “Ikhwān al-Ṣafāʾ,” EI2 III, 1071a. 91 ╇ Ullmann, 1994; for the dating of the manuscript, see idem, p.╯159. 92 ╇ Idem, p.╯18.24, p.╯54.187–90. 93 ╇ Idem, p.╯28.79. 94 ╇ Idem, pp.╯28–30.80–3. 95 ╇ Idem, p.╯30.84. 96 ╇ Idem, p.╯30.85–6.

85 86

178

chapter eleven

on the same day,97 and the sound of its hiss is equally deadly to the hearer. Any animal moreover that tastes the flesh of a person killed in such a way will immediately die in its turn.98 The Â�physicians Abū ʿAlī Aḥmad ibn ʿAbd al-Raḥmān ibn Mandawayh (d.╯410/1019) from Iṣfahān and Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh ibn Sīnā who came from a village near Bukhara (whose name was latinised as Avicenna; d.╯428/1037) both Â�mention that the “Basiliskos” serpent exists in the land of the Turks.99 In his text on poisons Ibn Waḥshiyya also mentions serpents which exist in the valley of the Khazluj in the land of the Turks noting that “if a man sees these snakes, he dies; also if the snake looks at him, he dies immediately.”100 Al-Damīrī similarly records in his Ḥayāt al-Ḥayawān that the gaze of some serpents exerts a baleful influence and can indeed be mortal.101 The same phenomenon is also described in the traditional tale of Alexander. There is said to have been a deep pit (biʾr) at a cemetery near Hamadān. Everybody who looked into it died at once. When Iskandar could not find a solution, he wrote to Aristotle who recommended him to place an iron mirror at the mouth of the pit and only to look into the pit when the mirror was blank. Iskandar followed this advice and for seven days the mirror was obscure, only appearing blank on the eighth day. He then ordered that someone should descend into the cavity, but his people refused. Eventually an animal was sent down and when it was clear that no harm had come to it, a slave 97 ╇ The fatal look of the basilisk is also noted by Eznik of Koghb (Elc alandocʿ, tr. and ed. Mariès and Mercier, 1959, p.╯594, ch. 140). 98 ╇ Ullmann, 1994, pp.╯30.87–8, pp.╯101–15.87–8. When disÂ�cussing the basilisk’s noxious influence, Pliny records that when struck with a spear the force of its poison was such that it could run up the spear and would kill not only the rider but also the horse (Naturalis Historia VIII 33). A reflection of this ancient belief may be found in Ibn Waḥshiyya’s text on poisons which states that:

There is a kind of snake called mādhyānā at the boundary of the countries of the Bājarmā [in the Arabian peninsula], between it and Media toward the side of Armenia. ... Its temper is that it bites the stone thrown at it. If this happens, then the man who threw the stone dies on the spot. This is when the thrower does not shoot at it; if he does, then he dies more quickly than the one who does not shoot. Tr. Levey, 1966, p.╯27. 99 ╇ Ullmann, 1994, pp.╯105–6, p.╯141.169. On Ibn Â�Mandawayh, see Sezgin, 1970, p. 328. 100 ╇ Tr. Levey, 1966, p.╯27.

descended in turn. At the bottom he saw a large coiled serpent with one central eye lying dead. Iskandar had it dragged from the pit. When he learned of all this, Aristotle explained to Iskandar that the serpent’s eye was the source of its venom, which was why all those who looked at it died immediately. When the giant serpent saw its own reflection in the mirror it was bound to perish as well.102 These references show that in particular the eyes of the creature are considered to be the seat of a specific fear-inspiring power which is believed to be potentially fatal. Further evidence of the persistence in Islamic times of a tradition relating to the legendary power of the serpent’s sharp-sightedness and penetrating gaze can be found in the saying that vipers and related snakes (af ʿā) live to an age of a thousand years, and that when they become blind, regain their sight by rubbing their eyes against the fennel plant (rāziyānaj).103 The tradition is probably distantly associated with the Greek drakōn, synonymous with ophis, serpens and other words for serpents104 which were used for living (real) as well as mythical serpents, said to derive its name from the root derk, meaning “to see.”105 Macrobius clearly derives the name draco from this root “for they say that this serpent has the keenest vision and counterfeits the nature of the star that is ever watchful; for this reason, furthermore, the protection of shrines, holy places, oracles, and treasures is entrusted to serpents.”106 In this context it is interesting to note the belief in the special property of the dhubābī type of 101

EI1.

╇ Tr. Jayakar, 1906, vol.╯1, p.╯633. Cf. Ruska, “Al-ḥaiya,”

102 ╇ Al-Zuhrī, Kitāb al-Jaʿrāfiyya, pp.╯598–600, as cited in Ullmann, 1994, pp.╯56–7. 103 ╇ Al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯57. According to the Egyptian encyclopaedist and historian al-Nuwayrī (Nihāya, XI, p.╯82), this happens when the reptiles leave their holes in spring. The same is noted repeatedly in al-Qazwīnī (ed. Wüstenfeld, 1849, repr. 1967, pp.╯284 and 428; Wiedemann, 1970, vol.╯2, pp.╯336, 386). The sympathy between the serpent and the fennel plant is also mentioned by the early medieval alchemical author Jābir ibn Ḥayyān (Kraus, 1942, p.╯67). A decoction of the fennel flower stalk mixed with wine was used as remedy against snake bites (Dietrich, “Basbās,” EI² supplement. Cf. Kopf, “Afʿā,” EI² I, 214b; Ruska, “Ḥayyā,” EI² III, 334b). The assertion that snakes use fennel for skin and sight is already mentioned in antiquity (cf. Pliny, Naturalis Historia XX 23.95). 104 ╇ Merkelbach, “Drache,” RAC IV, 1959, p.╯226. 105 ╇ See p. 173, n. 52. 106 ╇ Saturnalia 1.XX.1–4. Cf. Schlüter, 1982, pp.╯14–5; Van Henten, “Dragon,” DDD, p.╯265.

the dragon and the magico-medical sphere emerald (zumurrud). It was said that when this stone was brought close to the serpent’s eyes they would bulge from their sockets and burst.107 According to Ibn Waḥshiyya’s ninth-century text on poisons, it is the acorn-headed viper who responds in such a manner to the sight of the green emerald.108 The phenomenon was tested by the thirteenth-century Arabian mineralogist Aḥmad al-Tifāshī (580/1184–651/1253) and still found to be true.109 That the belief was wideÂ�spread in the medieval Islamic world is shown by its figurative use in philosophical arguments, for instance by the ninth-century Muʿtazilite Â�theologian Ibn al-Rāwandī in his book against prophecy, which he entitled Kitāb al-Zumurrud (“Book of the Emerald”) because as he informs the reader: One of the properties of the emerald is that the eyes of snakes dissolve and melt away if they look at it, and similarly, if an adversary peruses this book, he melts away [i.e., his arguments are silenced].110

The notion of the potency of the image, exemplified in the biblical story of the invasion of poisonous snakes which so devastated the people of Israel that Moses intervened on their behalf, setting a miraculous bronze image of a serpent on a pole (Numbers 21:4–9),111 is related to the magical principal of effecting a cure for snake venom by looking at the image of the serpent. In other words, it is a case of correspondence magic predicated on the principle that similia similibus curantur where a representation of the danger or of the disease helps in some way to exorcise it.112 In medieval Islamic toxicological treatises the section on the preparation of poisons was as prominent as that devoted to their prophylaxis and treatments. In Ibn Waḥshiyya’s text the preparation of poisons and their antidotes often also involves a mixture of magical rites, incantations and astrology.113 The description of the prepara107 ╇ Al-Masʿūdī, Kitāb murūj al-dhahab, tr. and ed. Barbier de Meynard and de Courteille, 1917, vol.╯3, p.╯46. 108 ╇ Levey, 1966, p.╯27. 109 ╇ Kitāb Azhār al-afkār fī jawāhir al-aḥjār, tr. and ed. Abul Huda, 1998, p.╯84. Cf. Kunz, 1913, p.╯158. 110 ╇ Al-ʿAbbāsī, Maʿāhid al-Tanṣīṣ, vol.╯1, p.╯156, cited after McKinney, 2004, p.╯270. 111 ╇ See also the the well-known copper alloy serpent column, the triple-headed serpent tripod of Delphi, in the Hippodrome of Constantinople, as an example of apotropaic sculpture against poisonous creatures including serpents. Cf. p. 23, n. 13.

179

tion of a poison with a snake as main ingredient and its symptoms reads as follows: There is a snake found in the desert called the black snake. It is one of the strong snakes called “the black which strips” because all snakes shed their skins once a year except the black which sheds four times every season of the year. It is so called because of the frequency of its shedding. This snake is taken and put into a wide, glass or clay vessel. It is cut into four pieces so that all its blood flows in the vessel. Its head is cut as one piece, and four fingers [width] below its head for every piece. This is carried out until the four or five pieces are finished and the blood comes together in the vessel. Set the pieces aside in something else so that the blood only will be in its vessel. Pay attention to the amount of its weight for the same amount of yellow sulfur is taken. This must be pure and good; it is pulverized finely and then sprinkled on the blood as long as the blood is fluid and mixes well with it. The head and tail [of the snake] are set aside but the pieces of its body are cremated with wood of mezereon until all becomes ash. This is then added to the blood and sulfur and mixed well by pulverization in the vessel until it needs moisture for its excessive dryness. It is moistened either with the blood of an animal or with the urine of a little boy until it is wet. Then the head and tail are thrown on it, side by side, and pulverized with it until all are mixed. It is gathered into a black lead vessel and covered with a well-fitting lid [i.e., of black lead], then buried in ass manure for twenty-one days. At the time of the hunting of the black [snake], its cutting, pulverizing, and mixing and at the time of its bleeding, the Moon must be adjoined to Mars or associated with it. When it is taken out after twenty-one days, it is found that it has become deeply black and has a very bad odor with a strong sharpness. Whoever carries out this work must fortify the atmosphere with two pieces of cotton soaked in violet oil which is so pure that there is nothing dearer than it or oil of nenuphar together with oil of pumpkin. He also smells sandalwood upon which

╇ Jeffers, 1996, p.╯234. Cf. Wakeman, 1973, p.╯86. ╇ For instance, while treatment for a viper bite is administered an incantation of Dābāth, a charm master, is to be recited while a hand-held rod is made to pass over the entire body of the victim. Levey, 1966, p.╯14, see also idem, p.╯69, for an unintelligible charm attributed to the same charm master. Ibn Waḥshiyya’s exposition of poisons (see p. 177, n. 87) gives an important insight into the medieval perception that harmful properties can be combatted by antidotes. See Levey, 1966, p.╯24. 112 113

chapter eleven

180

rosewater has been poured. When the cover of the black lead vessel is opened, air must be allowed to pass through for an hour to decrease the odor. If you wish that it have the power of killing within six hours, leave it so; if you wish that it be fatal immediately, without delay and tarrying, pulverize the penetrating things which are eaten and drunk. This reaches the core of the body. While one of the two is dry and the other moist, gather both together and add them to the poison you have macerated. Mix them all well. Certainly it can be lethal on the spot without delay. I recite the praise of the old one [God], the generous, the kind, the beneficent for those who know him and worship him and for those who do not know him and do not worship him! â•… When you want to use this to kill, weigh out one carat and mix it with any food, drink, gruel, or any odoriferous substance. It works when it reaches the belly or touches a spot on the body, wounds it and then, after a while, kills him. It is penetrating and sharp. â•… Whoever drinks this poison which is extracted from the black snake and sulfur, then his symptoms are that he becomes very restless, his eyes become red, his tongue swells, and he is so thirsty that a drink of water does not quench it [i.e., the thirst]; his tongue lolls and his voice is remote, his eyes pop out, all the veins of his body are gorged with blood, he becomes very anxious, and often he cries until he dies.114

Ibn Waḥshiyya, moreover, records an interesting operation which can be lethal by sound. It involves a compound agent, one of the central ingredients being three large vipers, used to anoint two hides and the wood of a drum, which when struck “kills by the sound when it is heard.” Once the agent has been applied three times and has dried, the drum has to be struck with a stick made of the branch of an olive tree. Drum and stick have to be set apart and when the Sun sets on a Tuesday evening, the drum and stick have to be placed facing Mars, preferably when the Moon is adjoining Mars or when Mars is rising. This is followed ╇ Tr. Levey, 1966, pp.╯58–9, also n. 297 for further references on this species. For other preparations of poisons with snakes as central ingredient, see idem, pp.╯52–5. Ibn Waḥshiyya also records the preparation of a compound poison which requires snake eggs, see idem, p.╯62, and specifies an antidote for someone who drinks gall of the viper (see idem, p.╯81). Snake gall also serves as ingredient for the preparation of a lethal agent (see idem, p.╯107). 115 ╇ Tr. Levey, 1966, pp.╯37–8. 116 ╇ The stone is the dracontias or dragon stone described by Pliny (Naturalis Historia, XXXVII 10.57) and Solinus (Collectanea Rerum Memorabilium, 30.16). According to 114

by a long incantation which repeatedly focuses on the drum and stick along with fumigations. Thereafter if one wishes to kill someone, one should strike the drum when the Moon is associated with Saturn. While striking the drum one should sing in praise of Mars and “whoever hears this voice dies either on the same day or after three hours following the darkness of the night have passed.”115 d.╇ The serpent(-dragon) stone Belief in the healing properties of a magic serpent stone (Pers. shāh-muhra) is ancient and widespread and refers to a precious stone located either in the head of the dragon116 or in its mouth, that is, one that would be expectorated by the ophidian creature.117 The idea that serpents have precious stones in their heads is recorded in the Indian Panchatantra, a collection of stories and fables, originally written in Sanskrit in the third century bc.118 It is propounded in the Koiraniden of Hermes Trismegistos and related texts in which the occult medical effects of the organs of animals are described.119 Examples include a large mythical serpent in the western sea with cervid-type dendritical antlers and a mane like that of the seahorse, a creature said to have stones in its head that are used for talismans.120 Such a serpent stone is also mentioned in Ibn Waḥshiyya’s ninth-century text on poisons and their antidotes. He specifies that: There is a bead (kharaza) in the neck of the viper between its head and body. It is not found and is not clear except in large vipers which have aged for hundreds of years. This bead is found on its neck. When you take this bead at the time of the rising of the sign of Aries, whatever climate it may be, with the left hand and fasten it tightly with parchment of the fawn by means of a white silk thread, and after he fastens it, if he wishes, he sews it in a tanned hide and hangs it fastened Pliny (Naturalis Historia, XXXVII 158) it is to be obtained by severing the head of a sleeping dragon. Cf. Merkelbach, “Drache,” RAC IV, 1959, p.╯228. See also Hasluck, 1929, vol.╯2, p.╯653 and n. 1; Massé, 1938, vol.╯2, p.╯326. 117 ╇ Steingass, 1892, repr. 1981, p.╯728, cf. also muhra, p.╯1354. Cf. Hoffmann-Krayer, “Schlange,” HdA, vol.╯7, 1936, p.╯1122, and idem, pp.╯1199–201. 118 ╇ Frazer, 1888, p.╯179; also Laufer and Walravens, 1987, p.╯138. 119 ╇ Ullmann, 1972, p.╯404, and idem, 1994, p.╯100, 83. 120 ╇ Idem, 1994, pp.╯28–30, 80–3, 100, 83.

the dragon and the magico-medical sphere on his left upper arm, then it serves to keep away from him the evils of the vipers and that of all snakes … When the vipers and any snakes touch his body, then they become soft; their hatred goes out. … â•… As to what Shūshā ordered, he said that the bead of the snake is taken during the rising of the sign of Aries and is tied in the slough of the snake, then sewed in a tanned hide. The man fastens it on his middle where he binds the trouser band. If snakes sting him, then it is not harmful to him. He requires only a mild remedy to avoid death, becoming black or blue in colour, and lessening in power.121

A “snake-stone,” often linked with a bezoar (bāzahr, a corruption of the Pers. pād-zahr, lit. “Â�protecting (against) poison”),122 that is, a concretion or “stone,” found in a snake or another aniÂ� mal, can be used in an amulet against the Evil Eye and illness.123 That it is also supposed to work as an antidote against snake venom is indicated by the twelfth-century poet Niẓāmī Ganjawī when he states that: from the thorny rose there comes rose water; life from the snake-stone.124

And about two hundred years later al-Damīrī (d.╯808/1405) records that there is one type of serpent, for whose bite “the bezoarÂ�stone (alÂ�diryâq) is useful.”125 Talismanic virtue is also ascribed to the guharmuhra:

╇ Tr. Levey, 1966, pp.╯69–70. ╇ Real bezoar stones, that is to say, hard round concretions, are also said to be obtained from the body of wild goats (the concretions are believed to have formed, for instance, in the head, the heart or the intestines, especially the stomach) indigenous to Iran and the lands on the borders of China that live chiefly on poisonous serpents; the bezoar is said to form when the animal has eaten too much snake flesh. Cf. Ruska, “Bezoar,” EI¹ I, p.╯710; for a detailed description, see al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, pp.╯222–5 and n. 1; Ettinghausen, 1955, pp.╯280–1 and n. 29, pp.╯283–5, and n. 52. The twelfth-century Jewish theologian Ibn Maymūn/Maimonides notes with regard to the bezoar: 121 122

Of the bezoar there is no mention in Galen’s writings. The bezoar stone, called animal bezoar, is an acorn-like object of green to blue-green hue. It is formed layer upon layer, like some shell heaped one upon another. People say its origin is in the medial eye-corner of the Oriental ram; others believe it is formed in the gallbladder, which is indeed the case. Treatise on Poisons and Their Antidotes, tr. and ed. Muntner, 1966, pp.╯17–8. According to other accounts, bezoar stones are also believed to be found in different animals as well as

181

a pearl said to be found in a serpent’s head and to secure the owner’s continual good luck.126

Pearls and their association with serpents also appear in a legend in the seventh-century Sanskrit text Harshacharita (“The Deeds of Harṣa”). According to the story a pearl necklace, “which shone ... like a cluster of stars,” born of the tears of the Moon god, became an antidote to poison and came into the possession of Vāsuki, the king of serpents. The latter always carried it with himself to soothe the burning heat of poison and eventually presented it to Nāgārjuna during his stay in the Netherworld.127 The belief in a rain stone (Turk. yai, Mong. yada),128 often a bezoar placed in water, which was widespread among the Altaic people of Inner Asia in the medieval period seems to have spread from the early Turks to the Mongols. The rain stone could be used in weather magic and by its means the holder could magically cause rain or snow to fall or to cease.129 In his treatise on mineralogy, Azhār al-afkār fī jawāhir al-aḥjār (“Best Thoughts on the Best of Stones”), written around 637/1240, Aḥmad al-Tifāshī describes the use of such a stone in a rain-making ritual conducted by an old Turkish weather-magician in the camp of the Khwārazm-shāh Muḥammad Khān, which took place under the personal supervision of the sulṭān: ...Then [the old Turk] took a live snake of the same colour as the [rain] stone, and fixed it [by human beings; cf. Molnár with an appendix by Zieme, 1984, pp.╯128–9. On the bezoar, see also Pseudo-Aristoteles, Das Steinbuch, tr. and ed. Ruska, 1912, pp.╯147–9. A thirteenthor fourteenth-century gold bowl with a mounted bezoar stone attached by a gold chain, perhaps produced in the Caucasus, is now housed in the Kunsthistorisches Museum, Vienna, inv. no. 1140, published by Ettinghausen, 1955, p.╯282, pl.╯XXXIX, no. 6. 123 ╇ Massé, 1938, vol.╯1, p.╯210. 124 ╇ Meisami, 1995, pp.╯25, 280, 7:54. 125 ╇ Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯636. 126 ╇ Steingass, 1892, repr. 1981, p.╯1107. For the association of dragons with pearls, see also the Iranian legend of Azhdahāk (that is, Azhi Dahāka) giving a huge pearl to a concubine. Tchukasizian, 1964, p.╯325. See pp. 62–3. 127 ╇ BaÌ—nÌ£a, Harshacharita, tr. Cowell and Thomas, 1897, pp.╯250–2. 128 ╇ See the linguistic discussion in Molnár with an appendix by Zieme, 1984, pp.╯104–16. 129 ╇ Bosworth, “Yada Tash,” EI² XI, 226a. For further discussion, see Rashīd al-Dīn, tr. and ed. Quatremère, 1836, pp.╯428–35. An in-depth study of the phenomenon is found in Molnár with an appendix by Zieme, 1984. Cf. DeWeese, 1994, pp.╯175–6, n. 24.

182

chapter eleven its tail to the tip of a stand of three reeds that surmount the bowl], hanging head down to the reed above the bowl [filled with water], so that there was a distance of two ells between the head of the snake hanging down and the surface of the water. After that he got two pieces of stone from the keeper of the stones, and put them into the water. ... He murmured some words, raised his head towards the sky, and prayed for rain. ... Then suddenly, clouds appeared at the edges of the sky, and it began to rain heavily, the air cooled down, and the men and the animals were eased.130

In his Taʾrīkh-i jahān-gushāy (“The History of the World-Conqueror”) ʿAlā al-Dīn ʿAṭa-Malik al-Juwaynī (623/1226–681/1283) reports how Genghis Khān’s son Toluy resorts to using the magical stone to produce snow as a means of concealment when hard pressed by the Tungusic Jurchen army during his invasion of northern China (Chīn) in 1232. Among his army was a Qanqlī Turk, a yaichi/yadachi (master of the art of yai/yada) who was well versed in the science of this stone and whom he ordered to make use of the weather magic in warfare: Ulugh Noyan commanded him to begin practising his art and ordered the whole army to put on raincoats over their winter clothes and not to dismount from their horses for three days and nights. The Qanqli busied himself with his yai so that it began to rain behind the Mongols, and on the last day, the rain was changed to snow, to which was added a cold wind. From this excessive summer chill, which was such as they had not experienced in winter, the Khitayan army

130 ╇ Molnár with an appendix by Zieme, 1984, pp.╯63–4. See also idem, pp.╯120–4. 131 ╇ Tr. Boyle, 1912–37, vol.╯1, p.╯193. This feat is also recorded by Ghiyāth al-Dīn Khwāndamīr (d.╯942/1535) in his chronicle Ḥabīb al-Siyār (“Friend of Travels”); see Ross, 1895, p. 33. The historian Muḥammad Ḥaidar Mīrzā Dughlāt (905/1499–1500–958/1551), a cousin of the first Mughal ruler, Bābur, notes the continued use of the magical rain stone in his Taʾrīkh-i Rashīdī. In the account of the Chagatayid Tughluq Tīmūr’s (d.╯764/1363) invasion of Mā warāʾ al-nahr, the Khān’s two forces were on the right bank of the Jayḥun (Āmū Daryā) but since they were faced with an army that: “exceeded them in number, [they] had recourse to magic, and sought aid from the Jadah stone, which possessed supernatural properties.

The army of Jatah had not strength for the fight, So they sought help from the magic stone. With the stone of Jadah, who was a magician,

were disheartened and dismayed and the Mongol army emboldened and exhilarated.131

The account thus details the military use of weather magic as a meteorological weapon emÂ�Â� ployed to enhance nomadic martial tactics. On yet another level, the “philosopher’s stone,” the universal remedy endowed with the power of transmutation, is allegorically compared to the serpent stone in the Kitāb Mānī rasūl Allāh, a short treatise attributed to Mani (216–c. 277), son of Pattēg, the founder of the dualist religion of Manichaeism, proclaiming the great secret: Blessed be he who knoweth the Tree of Beatitude. The distinguishing mark of this noble tree is the following: it only grows where goodness and blessing is to be found. Its fruit has a sweet, bitter, sour, bilious taste, from it emanates a tomb-like stench. Only few eat from it for it is deadly poison. However if you eat knowingly from it, you will not die, since it contains both poison and theriac [or theriac as well as poison], as is the case with the serpent. Our stone resembles the serpent, one can find it in the treasuries of kings and in the houses of sages, but one can also find it on the rubbish heaps, because fools spurn it and sages and scholars err with regard to its value. When you wish to pick from the fruits of the tree, it will cry and shed tears. Do not let yourself be frightened by this, rather pick them boldly and do not be frightened by it and its tears, because its tears banish suffering and illnesses, which it can generate, into the earth.132

They filled the world with wind and rain, The clouds roared with thunder and the winds howled. A thunderbolt fell upon the earth. Although the sun was in Orion, a host of dark clouds sudÂ� denly filled the sky. The thunder resounded and the lightning flashed. The elements rushed out from the amÂ�bush of destiny into the open plain of the ether, and the thunderclaps re-echoed round the azure vault of heaven. The arrows of lightning were shot out, in all directions, from the bow of the thunder-clouds, and the rain shot down its whistling darts. It seemed as if the Fates had again become prey to the love of rebellion and confusion. Such a quantity of water descended from the eyes of the stars, that the Deluge seemed to occur a second time.” Tr. Ross, 1895, pp.╯32–3. The use of weather magic in warfare appears also in the Shāh-nāma account, tr. and ed. Mohl, 1838–1878, vol.╯3, pp.╯35–9. 132 ╇ Ullmann, 1970, p.╯186, and idem, 1972, p.╯186 and ns. 1 and 2.

the dragon and the province of magic and divination

183

chapter twelve

The dragon and the province of magic and divination

a.╇ The dragon and talismans

pre-Islamic romance, Wīs u Rāmīn, translated and versified in the mid-eleventh century, mentions snake charming by means of sorcery.5 The magical aspect of the dragon is allegorically alluded to by ʿAsjadī Marwazī (as indicated by his nisba he was presumably a native of Marw), a court poet of sulṭān Maḥmūd of Ghazna, who writes:

As has been seen in preceding chapters, the range of symbolic meanings associated with the figure of the dragon extends to include an important role in those branches of knowledge that encompass the more esoteric or magical sciences, in particular sympathetic and talismanic astrological magic and the preparation of verbal and material talismans. The dragon’s ability to eclipse the great luminaries is allegorically treated in the quest for “mastery” of magic and the concomitant mystical illumination of life. This is apparent in medieval Islamic literature such as Niẓāmī’s Iskandar-nāma. The story recounts how the well-known sage and magician of the first century ad, Apollonius/Balīnūs, who in the Islamic Middle Ages became known as the great master of talismans (ṣāḥib al-ṭilasmāt),1 overcomes a powerful priestess, Āẓar Humā, who had transformed herself into a black dragon to guard the holy fire of the temple.2 However the wise Balīnūs not only breaks the resistance of the dragon-priestess but, interestingly, also marries her and manages in so doing to acquire knowledge of many of her magical practices. Niẓāmī’s account makes clear that it was only his association with Āẓar Humā that enabled Balīnūs to become a famous magician.3 It is notable that solely through union with a priestess who has the power to transform herself into a dragon can the great talisman-maker Balīnūs acquire knowledge of the magical sciences. In one of the oldest popular Persian prose narratives, the Kitāb-i Samak ʿAyyār, the magnanimous hero Samak is said to be taught two charms, one for dispelling serpents and another for calling them forth.4 Similarly, the

An important magical science was talismanic astrology. On the premise that stars can predetermine the course of future events, and that those well-versed in the motions of the heavens can accordingly foretell the future, the position of Islamic religious doctrine was, as already mentioned, fundamentally antithetical to astrological science. However, while astrology contented itself with mere prediction, magic on the other hand concerned itself with harnessing the forces of the cosmos to achieve predetermined ends. Between these two categories, there was also the magic of theurgy (ṭilsam), the adjuration of divine powers, which concerned itself among other things with the making and the use of talismans, a process dictated by the observance of specific conditions in the making of objects designed to protect the bearer.7 Islam ultimately accepted the use of “magic” and theurgy under certain strict conditions. Hence binding supernatural beings to human purposes was allowed as long as the aim was not to bring harm (so-called “white” or “natural” magic, ʿilm al-sīmiyā), which to a certain extent validated the use of amulets and talismans. The astrological iconography of the twelve zodi-

╇ On Apollonius of Tyana, see Sezgin, 1971, pp.╯77–90. ╇ Niẓāmī, Dīwān, pp.╯974–5, cited in Pseudo-Apollonius of Tyana, tr. and ed. Weisser, 1990, p.╯27. 3 ╇ Dastgirdī, V., vol.╯1, Tehran 1334/1956, pp.╯242–4, as cited in Bürgel, 2000, p.╯135. 4 ╇ Samak-i ʿAyyār III, pp.╯56–7, as cited in Omidsalar,

“Magic in literature and folklore in the Islamic period,” EIr. Cf. Gaillard, 1987, pp.╯19–21. 5 ╇ Tr. and ed. Davis, 2008, p.╯88. 6 ╇ Dīwān, ed. Shabāb, T., Tehran, 1955, p.╯31, as cited in Daneshvari, 1993, pp.╯21–2. 7 ╇ Ruska and Carra de Vaux, “Tilsam,” EI² X, 500a.

1 2

Don’t expect the world to be good to you The snake is only [made] kind by the power of magic.6

184

chapter twelve

acal signs and the seven planets fulfils a central function in talismanic design. The visual conceptualisation of the dragon thereby also plays a role, rendered as part of particular conjunctions which form the basis for astrological prediction. The Qurʾānic exhortation not to worship the Sun and Moon, two signs created by God, occurs once (sūra 41, 37).8 The codification of Rabbinic laws, Mishna, also specifically refers to images of the Sun, the Moon and, in addition, of the dragon, all of which must have been commonly represented in or before at least the fifth century ad (certainly the latest possible date of the Â�composition of the Mishna), resulting in the Â�Talmudic tract ʿAvodah Zarah (“Mishna on Idolatry” 3, 3) which contains the following prescription: Whenever a vessel is found on which the picture of the sun, or of the moon, or of a dragon (draqōn) is shown, it must be thrown into the salt sea.9

Rabbi Judah explains with regard to the imagery of the dragon (dᵉraqon) that it is “anything that has fringes between the joints ... of the neck.”10 Commenting on this prohibition of the use of idolatrous images of the luminaries and the dragon in Moreh Nebukhim (“The Guide for the Perplexed”) the great twelfth-century Jewish theologian Abū ʿImrān Mūsā ibn ʿUbaid Allāh ibn Maymūn/Maimonides (1135–1204) says: When the picture of the sun or the moon is mentioned, this does not mean that the picture of the sun is represented by a round disk, or that of the moon by a bow, but it refers to those figures which are called telesmata, and which are ascribed to the stars by the men who made them. Thus, for instance, they used to represent Saturn like a black old man, Venus like a gold-adorned fair young girl, the sun as a crowned king sitting on a chariot, and likewise they ascribe many figures to all of the constellations and stars though there is no agreement on it among them … But the picture of the dragon [d ᵉ raqon] which is ╇ Fahd, “Shams,” EI² IX, 291a. ╇ Hartner, 1938, pp.╯149–50. It is of note that the serpentdragon image is to be cast into the salt sea, that is the Dead Sea, presumably meant as symbol of death. A baraita (extracanonical Mishna) similarly maintains that “all planets except the sun and moon are “permissible” as are all faces except the human face and all figures except the dragon,” as cited in Epstein, 1997, p.╯142, n. 38. For a detailed discussion of rabbinic texts on the theme of the dragon and idolatry, see Schlüter, 1982, pp.╯62–129. 10 ╇ Wilson, 2001, pp.╯124–5. 11 ╇ Cited after the English translation by Hartner, 1938, p.╯150. For a German translation, see Chwolsohn, 1856, vol.╯2, pp.╯484–5; Schlüter, 1982, p.╯130. 8 9

mentioned in this Mishna is a scaled and finned figure like that of a fish. This figure was highly renowned with them because they attribute it to a certain part of the celestial sphere. And one of them who used to make such pictures told me that this one picture represents the dragon in the sphere of the moon [tᵉli] – called in Arabic “al-Djawzahar” – and that it is made after a certain model and in a certain hour. As I have never seen such a picture I asked him in what book I might find it mentioned. Whereupon he answered me that his teacher himself had devised that picture and confided it to him as a secret, together with many other things.11

It is interesting that Maimonides thereby equates the dᵉraqon with the dragon in the sphere of the Moon (tᵉli), the meaning of the latter however remaining unclear. The commentary thus shows, as Hartner put it: ...that still in the twelfth century the astrological doctrines of the Djawzahar had by no means become a generally known matter but used to be treated as a secret by the initiated, in such a way that even a highly erudite scholar like Maimonides could make only a rather vague statement about it.12

It reveals, moreover, the extent to which the fabrication of telesmata, or talismans, was shrouded in secrecy. It is also associated with the fact that, like astrology, the practice of magic (siḥr) was frowned upon by the religious establishment.13 As objects which were made to protect their owner and to ward off evil, the action of talismans is based on the concept of sympathetic magic,14 which operates through the connectedness and interdependence of all phenomena through their qualities and attributes.15 Magic was defined by the Khurasanian theologian and religious reÂ�former, Abū Ḥāmid al-Ghazālī (450/1058– 05/1111), as “based on a combined knowledge of the properties of certain terrestrial elements and of propitious astral risings.”16 12 ╇ Hartner, 1938, p.╯150. See however Schlüter’s (1982, pp.╯141–1) suggestion that Maimonides, who in spite of being well-versed in astrology was, as is generally known, ill-disposed towards the science, may thus intentionally have equated the Moon tᵉli with the dᵉraqon (which the Mishna so expressly proscribed) partly in order to take distance from his contemporaries who placed such great importance on the Moon tᵉli. 13 ╇ Fahd, “Siḥr,” EI² IX, 567b. 14 ╇ Idem. Cf. Porter, 2006, p.╯794. 15 ╇ O’Connor, 1994, p.╯67. 16 ╇ Iḥyāʾ, vol.╯1, pp.╯49–50, as cited in Fahd, “Ruḳya,” EI² VIII, 600a.

the dragon and the province of magic and divination The influential magical manual Ghāyat al-ḥakīm (“The Philosopher’s Goal”), attributed to Abū Maslama Muḥammad al-Majrītī, who wrote between 443/1052 and 448/1056, provides a synthesis of magic with astrology and stresses the importance of using talismans, whether material or verbal (spells).17 The compilation was inspired by Ṣābian sources,18 which played an important role in the transmission of esotericism from late antiquity into early Islam, and by “‘Indianised’ hermeticist astrology.”19 Taking inspiration from the tenth-century compilations of the Ikhwān al-Ṣafāʾ, al-Majrītī wrote: ...magic essentially comprises two parts, one theoretical and one practical. The first consists in knowledge of the positions of the immobile heavenly bodies (which is where, in fact, the forms are located), the modalities of their radiation on the planets and, finally, aspects of conjunctions of the celestial spheres at the precise moment that the successful outcome of a project is desired. Under this heading the ancients placed everything having to do with discernment of the beneficial and of the baleful and with theurgy. As for practical magic, it consists in the knowledge of three domains of the created being and of the qualities of the planets which would be disseminated there.20

Ibn Khaldūn, who knew the Ghāyat al-ḥakīm, goes further in his prolegomena on the history of the world, al-Muqaddima, stating that: … the souls of magicians possess the ability to exert influences in the universe and to tap into the spirituality of the planets, in order to use it 17 ╇ It was translated into Spanish in 1256 by order of Alfonso the Wise, king of Castile and Leon, and later became known in Latin under the title of Picatrix; Anawati, “Arabic Alchemy,” EHAS, 1996, vol.╯3, p.╯872. Cf. Ritter, 1933; “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962; also Ritter, 1921– 2. 18 ╇ The so-called Ṣābian inhabitants of Ḥarrān in northwestern Mesopotamia were particularly well-known for the practice of astronomy and astrology. Cf. p. 177, n. 90. 19 ╇ Cf. Fahd, “Ṣābiʾa,” EI² VIII, 675a. 20 ╇ “Picatrix ,↜” tr. and eds. Ritter and Plessner, 1962, p.╯9; tr. after Fahd, “Siḥr,” EI² IX, p.╯567b. 21 ╇ Idem (tr. after Muqaddima, tr. Rosenthal, 1958, vol.╯3, pp.╯156–8). 22 ╇ Citing three sources, specified as a lapidary by ʿUṭārid, a book without a title by Balīnūs (Apollonius) and a work by a certain Qrīṭūn (Kriton) on pneumatic talismans translated by Bu(i)qrāṭīs (the name Bu(i)qrāṭīs is identified by Ritter and Plessner as “Picatrix,↜” while Ullmann, 1972, p.╯420, and Strohmaier, 1989, p.╯267, relate it to that of Hippocrates). Varying descriptions of the planets are given in the text which is a result of different manifestations and a syncretist evolu-

185

in the exercise of their influence, by means of a psychic or satanic force.21

Among the images of the planetary gods described by al-Majrītī in the Ghāyat al-ḥakīm, one combines the Sun, the Moon and Jupiter with the imagery of the dragon.22 In one source cited by al-Majrītī the Sun is portrayed in the form of a standing male figure holding a shield in his left hand and having under his feet the image of a dragon;23 in another he appears as a crowned king, enthroned with a dragon under his feet and a raven in front of him.24 The Moon resembles a woman with a beautiful countenance, girded by a dragon. A pair of serpents crown her horned head, another pair encircle her wrists while above and below her head are seven-headed dragons.25 Again, in the first cited source the Moon is portrayed as a woman who sits on a pair of dragons, each biting the tail of the other.26 Jupiter is a male figure with lion head and raptor’s feet, with a polycephalic (or seven-headed) dragon under his feet.27 The significance of a talisman lies in its conjunction with the celestial bodies;28 hence talismanic astrology plays a central role in its prepaÂ�ration.29 The magician is warned of the eclipses of the Moon and the Sun and instructed to delay making the talisman until the Moon is free of the knot (node).30 He is further warned of worse calamity when the Moon “foregoes … the head or the tail (taqaddama); because then he will be ‘burnt’ through the conjunction (bimulāqāt) of his body with that of the Sun,”31 the reason being that when “passing from northern tion of the iconography of the celestial bodies; “Picatrix,” tr. and eds. Ritter and Plessner, 1962, pp.╯xliv, lxv. 23 ╇ Citing a lapidary by ʿUṭārid as source; eidem, p.╯115.10– 3. 24 ╇ Citing a work by a certain Qrīṭūn on pneumatic talismans translated by Bu(i)qrāṭīs; eidem, p.╯115.14–7 and n. 1. 25 ╇ Citing a lapidary by ʿUṭārid as source; eidem, pp.╯116.30– 117.1–5. 26 ╇ Citing a book without a title by Apollonius as source; eidem, p.╯117.6–8 and n. e. 27 ╇ Citing a work by a certain Qrīṭūn on pneumatic talismans translated by Bu(i)qrāṭīs; eidem, p.╯118.6–9. 28 ╇ Eidem, p.╯35.21. 29 ╇ Eidem, p.╯35.21. 30 ╇ Eidem, p.╯23.30. 31 ╇ “…dem Kopf oder Schwanz … vorhergeht (taqaddama); denn dann ist er ‘verbrannt’ durch das Zusammentreffen (bi-mulāqāt) seines Körpers mit dem Körper der Sonne.” “Picatrix,” tr. and eds. Ritter and Plessner, 1962, p.╯24.5, cf. pp.╯69.27–70.1; with emendation by Hartner, 1965, p.╯445.

186

chapter twelve refer to the knots of the jawzahr (tᵉli) as knots of love and mystical union:

to southern latitudes, the Moon develops maleficent qualities.”32 Instructions are given for the manufacture of an amuletic seal ring stone which has the extraordinary power of bringing about the disappearance of the wearer’s enemies, and of instilling fear and terror into enemies. The making of it requires great precision and should be done at the hour of the zodiacal sign of Jupiter and the first decan of the domicilium of his exaltation, which is Cancer.33 At this time green corundum should be engraved with the male figure of Jupiter rendered with a lion head and raptor’s feet, holding a lance (mizrāq) with which he stabs a dragon that lies under his feet.34 Another interesting association with the head of the dragon (raʾs al-tinnīn) is made in the Ghāyat al-ḥakīm describing in gruesome detail the severance of a man’s head, carried out on a living subject, for divinatory purposes.35 The soothsaying head is then posted in front of the raʾs al-tinnīn, denoting probably a navagraha relief.36 Basing himself on the idea that the Intellect and the Soul have intermediary roles in the Creation of the natural world, which again echoes the well-known Rasāʾil (Epistles) of the Ikhwān al-Ṣafāʾ, al-Majrītī allegorically states in the Ghāyat al-ḥakīm that if inclined towards the intellect, the soul will be illuminated by it, but if it moves away from the intellect the knot of the dragon’s tail will eclipse the Sun and the Moon.37 Associated notions are also known in Jewish mysticism, discussed by the thirteenth-century mystic Abraham Abulafia in a passage from the Otzar Eden HaGanuz, where he goes so far as to

The supernatural and magic power of the rod has been renowned since the dawn of classical antiquity. The shape of the caduceus (kerykeion) appears at least as early as the middle of the sixth century bc in the form of the figure eight or in that of twisted serpents.39 First and foremost in Greek mythology it was the attribute of the messenger of the gods, Hermes, whose chthonic character is related to his magic might.40 The symbol’s appearance in the Central Asian region is

╇ Hartner, 1965, p.╯446. ╇ The domicilium represents the position of a planet standing in a certain zodiacal sign, while the exaltation (sharaf) signifies the point of the maximum power of a planet when standing in a certain zodiacal sign. Cf. Nasr, 1964a, repr. 1993, pp.╯160–1, and table IV. 34 ╇ “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, p.╯129.7–13. 35 ╇ The implied human sacrifice is also recorded by Ibn Khaldūn, Muqaddima, tr. Rosenthal, 1958, vol.╯1, p.╯221. A comparable magical practice is ascribed to the Ṣābians. Cf. Ibn al-Nadīm, Kitāb al-Fihrist, ed. Chwolsohn, 1856, vol.╯2, pp.╯15, 19–21, and n. 121; and ed. Flügel, G., Leipzig, 1871– 2, p.╯321; Chwolsohn, 1856, vol.╯1, p.╯142, and idem, vol.╯2, pp.╯19–21, 130–2, 142–4. See also Dozy and de Goeje, 1884, pp.╯365–6; Green, 1992, pp.╯178–80. The soothsaying head of the Ḥarrānians is mentioned in several other sources, for instance, by the fourteenth-century author al-Dimishqī, Nukhbat al-dahr fī ʿajāʾib al-barr wa ’l-baḥr, book 1, ch. 10, ed. Chwolsohn, 1856, vol.╯2, pp.╯388–9, in the chronicle of the Syrian Jacobite patriarch Dionysius Telmaharensis

(d.╯848) and the Nestorian church chronicle by the fourteenth-century historian ʿAmr ibn Mattai, as cited in idem, p.╯VII and ns. 25, 26. 36 ╇ “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, pp.╯146.6–147.22; with emendation by Hartner, 1965, p.╯448. Green, 1992, p.╯179. 37 ╇ “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, pp.╯45.19–24–46.1; cf. introduction, p.╯lxi. 38 ╇ 55a Bodleian Ms. Or 606. This process of loosening and tying is conceived as the path towards enlightenment, see Idel, 1988, p.╯136. 39 ╇ Levi, 1941, p.╯227. On the origin and the history of the caduceus, see also Wilson, 2001, pp.╯183–94. 40 ╇ To this may be added the Greek myth which tells of the blind soothsayer Teiresias who once came across two serpents in the act of copulation and killed the female one with his staff. He was immediately transformed into a woman and remained so for seven years. After seven years he met again two copulating serpents and this time killed the male one whereafter he became a man again. See Astour, 1965, p.╯163.

32 33

And the cosmic axis (tᵉli) is none other than the knot of the spheres, and there is no doubt that this is the subject of their existence, like the likeness of the connection of the limbs within man, and the connection of the limbs in man which are suspended in the bones at the beginning are also called the axis in man as well. And its secret is that a magician bring this knot of desire and renew it in order to preserve the existence of this compound for a certain amount of time. And when the knot is undone, the matter of the testimony of the knot will be revealed, and one who cleaves to these knots (qešarim) cleaves to falsehoods (šeqarim), for as they are going in the future to be undone, the knots of his cleaving will also be undone, and nothing will remain with him any more, and therefore, before he loosens these, he must tie and cleave to the ropes of love those who have not loosened the knots of his love and the cleaving of his desire; and that is God, may He be exalted; and no other in any sense.38

b.╇ The dragon staff

the dragon and the province of magic and divination

187

testified by its representation as early as the late fourth century bc on gold coinage struck by Sophytes, who ruled over the Oxus region in northern Afghanistan.41 In the Semitic world the shepherd’s rod, stick, or staff (ʿaṣā) could function as a symbol of sovereign rule. If used as such it was believed to be a repository of the power of royalty, and when bestowed upon someone it both signified and helped to effect the transfer of this royal force.42 Following the example of the Prophet Muḥammad, the early caliphs carried a spear or staff on ceremonial occasions (ʿanaza).43 The word serpent or snake (ḥayya) appears in the Qurʾān only in allusion to the staff or rod that Mūsa carried as sceptre of his authority and miraculously transformed into a serpent.44 Of the nine miracles of which the Qurʾān speaks, Mūsā’s turning of his staff into a serpent is the first (sūras 20, 17–24; 27, 10; and 28, 31); he accomplishes this by throwing down the staff which takes the form of a crawling serpent. The transformation of Mūsa’s symbol of sovereign rule, into a dragon, is a magical act that proves the Prophet’s authenticity. In Mūsā’s duel with the magicians of the pharaoh (firʿawn) the rods and ropes of the magicians were devoured by Mūsā’s serpent-rod (sūras 7, 107 and 117; 26, 32 and 45). He thereby performed a miracle by which God provided evidence of his authenticity as a Prophet. In post-Qurʾānic tradition as well as in magical writings this miracle and the miraculous character of the serpentrod play an important part. The rod was directly associated with the shape of a serpent45 and could

also take the form of “a writhing serpent the size of a camel.”46 Mūsā is portrayed with such a rod ending in a dragon’s head in one of the surviving folios of Rashīd al-Dīn’s Jāmiʿ al-tawārīkh (“Compendium of Chronicles”) transcribed and illuminated in 714/1314–5 at the Ilkhanid capital, Tabriz, which is further discussed below.47 The complete transformation of the rod into a fiveclawed quadruped dragon of Chinese derivation is portrayed in the scene of Mūsā frightening the firʿawn illustrated in a copy of Balʿamī’s Tarjumati tārīkh-i Ṭabarī, probably made in the Jazīra and dated to c. 1300, now preserved in the Freer Gallery of Art, Washington, DC.48 Trees, and twigs as part of trees, were widely used for religio-magical purposes which relate the rod or wand to an artificial twig.49 This is reflected in the description in Arabic sources of Mūsā’s rod as a “two-pronged fork with a crook under the meeting point of the twigs and when it was turned into a serpent, the two twigs formed the mouth of the serpent with its forked tongue, while the crook took the shape of the crest.”50 The Qurʾānic story of Mūsā’s rod turning into a serpent is, moreover, an example of the living power of the rod.51 Al-Tirmidhī records the popular eschatological belief according to which Mūsā’s rod is one of the things that will reappear in the Last Days. When the Beast (al-dābbat al-arḍ) which is spoken of in the Qurʾān (sūra 27, 82)52 becomes manifest as one of the significant signs of the approaching Hour, it will bring with it Mūsā’s

╇ The dating is suggested relative to Seleucid coinage which dates from circa 305 bc. Reverse caduceus, obverse Sophytes in profile with helmet. See Bopearachchi, 1996, p.╯26. 42 ╇ Glaznov, 2001, p.╯92.€ 43 ╇ Miles, “ʿAnaza,” EI² I, 482b. 44 ╇ Ruska,“Ḥayyā,” EI² III, 334b. 45 ╇ Cf. al-Thaʿlabī, Qiṣaṣ al-anbiyāʾ, Cairo, n.d., p.╯189, cited after Fodor, 1978, p.╯13; see also idem, p.╯12 and n. 73. 46 ╇ Al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯227. Cf. Ibn Iyās, Badāʾiʿ al-zuhūr, Cairo, n.d., pp.╯125–6, as cited in Fodor, 1978, p.╯15 and n. 89. In the Old Testament, this rod belonged to Aaron, before it was turned into a serpentrod which swallows the rods of the Egyptian magicians, then it bursts into bloom and bears almonds. See Fodor, 1978, p.╯2. 47 ╇ Martin, 1912, repr. 1968, vol.╯1, p.╯24, fig.╯12; TalbotRice, 1957, p.╯61, cat. no. 11. The dragon finial of the staff held by Mūsā closely resembles a twelfth- or thirteenth-century copper alloy finial from Iran or Jazīra, now preserved in Paris, Musée du Louvre, which may have topped a ceremonial staff (L’Islam dans les collections nationales, 1977, p.╯102, cat. no. 161; L’Etrange et le Merveilleux en terres d’Islam, 2001, p.╯110, cat. no. 74. Paris, Musée du Louvre, département des

Antiquités orientales, section Islamique, inv. no. OA 6697). Another closely related finial is preserved in Copenhagen, the David Collection, inv. no. 47/1966 (von Folsach, 1990, p.╯197, cat. no. 325, and idem, 1991, p.╯44, cat. no. 33, ill. on p.╯17). For related imagery in Armenian manuscripts, see the figure of a priest holding an upright undulant serpent with gaping mouth as staff portrayed in an Armenian lectionary from Erznga(n), dated 1362 (Yerevan, Matenadaran MS 4519); Armenian Miniature, eds. Gevorkian and Â�Abgarian, 1996, pl.╯54. Such a dragon-rod would also have been carried by a Buddhist monk, see Whitfield, R., Whitfield, S., and Agnew, 2000, p.╯25. 48 ╇ Kadoi, 2008, p.╯147, fig.╯4.23. 49 ╇ Cf. Schimmel, 1994, pp.╯29–30. 50 ╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, I, Cairo, n.d., p.╯401; al-Thaʿlabī, Qiṣaṣ al-anbiyāʾ, Cairo, n.d., p.╯90, cited after Fodor, 1978, p.╯4. 51 ╇ Schimmel, 1994, p.╯30. 52 ╇ See also the the commentary on this verse of the twelfthcentury Khwārazmian scholar Abu ’l-Qāsim Maḥmūd ibn ʿUmar al-Zamakhsharī, known as Jār Allāh, verse in which he said that the beast will come forth from Ajyād without stating what or where this is; see Lane, 2005, p.╯xiii and n. 1. Cf. Abel, “Dābba,” EI2 II, 71a.

41

188

chapter twelve

rod as well as the seal of Solomon (khātam Sulaymān).53 The importance of the serpent-rod lies in its capacity to serve as an instrument for the performing of “magical” actions54 which may be linked to the widespread association of snakes with magic.55 In post-Qurʾānic histories of the Prophet Muḥammad, Mūsā’s miracles play an important part.56 According to al-Kisāʾī, the rod is said to come from Paradise and a long succession of pre-Islamic prophets had previously used it.57 When discussing the contest between Mūsā and the Pharaoh’s magicians, the same author states that the staff metamorphoses, as noted above, into “a writhing serpent the size of a camel.”58 According to al-Ṭabarī, an angel brought the rod.59 Al-Thaʿlabī speaks of it as a miraculous rod which performed wonders such as providing water in a drought, producing milk, honey and fragrant odours, and becoming a fruit-bearing tree when placed by Mūsā in the ground. The Qiṣaṣ al-anbiyāʾ further list the following endowments: that a double dragon would appear on the two twigs, in other words, that the rod transforms itself into a double dragon, to combat enemies; that the rod was able to divide mountains and rocks and to lead its owner over rivers and seas; that it warns of danger and protects Mūsā and his herd from beasts of prey and assassins;

finally that it was able to transport Mūsā at his will.60 In the eleventh-century magical treatise Ghāyat al-ḥakīm Abū Maslama Muḥammad al-Majrītī cites Plato’s Great Book of the Laws as a source for the performance of the magical transformation of rods or ropes into serpents that can devour anyone who is thrown in front of them.61 The magic power of Mūsā’s serpent-rod was attributed primarily to the words inscribed upon it,62 which consisted of names that were linked to the mystical divine name that was endowed with special power63 and in Arabic magic literature is generally referred to as ism Allāh al-aʿẓam (“the greatest name of God”).64 The same rod was associated with the symbol of a circle considered to have the power to ward off evil.65 In accordance with ancient Jewish and general Middle Eastern tradition, circles are a symbol of light and hence express the luminous character of the rod.66 In his writings, which draw on pre-Islamic (comprising Hellenistic and Jewish) magical practices, the “master of the art of magic,” Aḥmad ibn ʿAlī al-Būnī enumerates the mystical names on Mūsā’s rod with which “darkness disappeared” and “several strange things are accomplished.”67 Al-Thaʿlabī also records that the rod performed the miracle of shining in the darkness,68 a feature generally considered characteristic of objects of revelation.69

53 ╇ Al-Tirmidhī, Bāb al-Tafsīr on sūra 27; Musnad Aḥmad, vol.╯2, 295, as cited in Jeffery, “ʿAṣā,” EI² I, 680b. 54 ╇ Cf. Fodor, 1978, p.╯2 and n. 6. 55 ╇ Wakeman, 1973, p.╯77, n. 2. 56 ╇ See, for instance, the miracle recorded in al-Kisāʾī, Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯228. To a large extent these “miracles” are derived from rabbinic sources such as Yalqūṭ Shimʿonī, Midrash Wayyosha, Peṣiqta de-Rab Kahana, and Midrash Rabba, as cited in Jeffery, “ʿAṣā,” EI² I, 680b. 57 ╇ Ādam, Hābil, Shīth, Idrīs, Nūḥ, Hūd, Ṣāliḥ, Ibrahīm, Ismāʿīl, Isḥāq and Yaʿqūb; Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯222. 58 ╇ Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯227. 59 ╇ Al-Ṭabarī, Mukhtaṣar taʾrīkh al-rusul wa ’l-mulūk wa ’l-khulafāʾ, vol.╯1, 460–1, as cited in Heller, “Mūsā,” EI2 VII, 640a. 60 ╇ Al-Thaʿlabī, ʿArāʾis al-majālis fī qiṣaṣ al-anbiyāʾ, tr. and ed. Brinner, 2002, pp.╯294–5. See also Heller, “Mūsā,” EI2 VII, 640a; Fodor, 1978, pp.╯5–6. 61 ╇ “Picatrix,↜” tr. and eds. Ritter and Plessner, 1962, p.╯154.13–5. 62 ╇ Fodor, 1978, p.╯9.

╇ Idem, p.╯10. ╇ Idem, p.╯11. According to legend, Solomon’s ring was also inscribed with the divine names which gave it the miraculous power that Moses’ rod had possessed; Salzberger, G., Die Salomonsage in der semitischen Literatur I, Diss., Berlin, 1907, pp.╯117–9, as cited in Fodor, 1978, p.╯13, n. 77. For an analysis of the Islamic divine names, see Anawati, 1967, pp.╯7–58. 65 ╇ Canaan, 1937, p.╯109; Fodor, 1978, pp.╯13–5 and n. 87; Schimmel, 1994, p.╯91. The association of the serpent and the circle is evidenced on an Aramaic magic bowl from the collection of V. Klagsbald, Jerusalem, which is decorated at the centre with a figure standing in a circle that is drawn around the feet. Two serpents flank the figure at either side while a serpent encircles the entire composition. Naveh and Shaked, 1985, repr. 1998, p.╯198–214, drawing on p.╯200, pls. 30, 31. 66 ╇ Fodor, 1978, p.╯15 and n. 86. 67 ╇ Manbaʿ uṣūl al-ḥikma, pp.╯154–6, as cited in Fodor, 1978, p.╯6. 68 ╇ Al-Thaʿlabī, Qiṣaṣ al-anbiyāʾ, Cairo, 1325, pp.╯111–6, cited after Heller, “Mūsā,” EI2 VII, 640a. Cf. Fodor, 1978, p.╯5. 69 ╇ Idem, 1978, p.╯14 and n. 82. 63 64

the dragon and sound

part five

the dragon as mystical metaphor

189

190

chapter thirteen

the dragon and sound

191

chapter thirteen

The dragon and sound

In his biography of the Prophet Muḥammad, al-Ḥalabī describes how one day the serpent was on top of the wall of the Kaʿba when a large bird snatched it and cast it on the hill of al-Ḥajūn (where the cemetery of Mecca is located): […] where it was swallowed by the earth. Some people say that this is the creature, which will speak to mankind on the day of Resurrection. It is also said that the monster will come forth from the ravine of the mountain Adjyād.1

He thus implicitly seems to associate the serpent with the Beast of the Earth (al-dābbat al-arḍ) mentioned in the Qurʾān (sūra 27, 82).2 Here the serpent will raise its voice on the day of Resurrection. Citing the seventh-century Jewish convert Kaʿb al-Aḥbār, the early medieval writer al-Kisāʾī describes the voice of the great serpent: When the serpent extols God, its exaltation overwhelms that of all the angels. When it opens its mouth, the heavens and the earth are lit by the lightning that flashes. Were not this serpent tempered by extolling God, it would strike down all created things with the might of its voice.3

In this tradition its voice is likened to the terrifying sound of thunder, thereby metaphorically associating the dragon with climatological phenomena. The voice of a beneficial serpent saves the life of a mystic in a story recorded in Abū Ibrāhīm Mustamlī al-Bukhārī’s (d.╯434/1042–3) Persian commentary on the celebrated manual on Ṣūfism, Taʿarruf li-madhhab ahl al-taṣawwuf, by Abū Bakr al-KalāÂ�bādhī (d.╯380/990):

1 ╇ Al-Sīra al-Ḥalabiyya, Cairo, 1292, vol.╯1, p.╯192, 2–4, cited after Wensinck, 1916, repr. 1978, p.╯64. 2 ╇ See also pp. 187–8. 3 ╇ Qiṣaṣ al-anbiyāʾ, tr. Thackston, 1978, p.╯7. 4 ╇ Sharḥ-i al-taʿarruf, vol.╯4, ed. Raushan, M., Tehran, 1366/1987, pp.╯1792–93, as cited in Gohrab, 2000, p.╯86. 5 ╇ Tārīkh-i Ṭabaristān, compiled c. 613/1216 by MuḥamÂ�mad

I heard Abu ’l-Ḥasan al-Fārsī saying: “I became so extremely thirsty in a desert that I could not move any more. I had heard that the eyes of thirsty people burst before they die. I was waiting for the bursting of my eyes when I suddenly heard a voice. I turned my face and saw a white serpent, as bright as pure silver, approaching me. I ran fearfully because fear had engendered power in me. Then due to weakness I walked slowly while the serpent was still after me. In this way I walked, till I reached water. The voice grew silent and I could not see the serpent.4

The powerful voice of the dragon is also referred to in an account of the dragon fight of Sām/ Garshāsp, the legendary ruler of Sīstān and grandfather of the hero Rustam. When the dragon, who was fifty thousand cubits (gaz) in length, saw Sām, he jumped at him. Sām struck him with his mace so that the dragon fell to pieces and “uttered so fearful a cry that all of Sām’s companions fell to the ground in terror.”5 Al-Damīrī also reports that there is one species of serpent, found in abundance in the country of the Turks, whose hiss is deadly even at the distance of a bow-shot and yet another type of serpent, “the voice of which if a man hears, he dies.”6 On the other hand, he cites the interpretation of a dream according to which: He, who dreams as if a serpent has spoken to him, will obtain happiness.7

A speaking dragon appears in the epic Shāhnāma, during Rustam’s third trial. This creature lives underground on the road to Māzandarān8 and the hero inadvertently strays into its terrain. The dragon is endowed with the magical power of invisibility, and is portrayed as using his faculty

ibn al-Ḥasan ibn Isfandiyār (tr. Browne, 1905, pp.╯41– 2). 6 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯633. Cf. Ruska, “Al-ḥaiya,” EI1. 7 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯656. 8 ╇ For a discussion on the location of Māzandarān, see Monchi-Zadeh, 1975, pp.╯48–79.

192

chapter thirteen

of thinking, expressing surprise that the hero dares to sleep in the face of the impending danger of his advance.9 The mythical creature then approaches the sleeping hero, who is twice woken by his loyal steed Rakhsh, but the dragon vanishes. On the third approach it is unable to cloak itself and is seen by the hero. The exchange of words that ensues between the hero and the dragon serves to anthropomorphise the creature to some extent. Nevertheless, with the help of his horse, the hero manages to slay the dragon. The speaking anthropomorphised serpent is also known in the most famous Arabian collection of fairy-tales and other stories of the Alf layla wa-layla in the Journeys of Bulūqiyā.10 Â�Mythical serpents endowed with the power of speech appear similarly in Indian lore, such as in the great collection of fables known as the Â�Panchatantra.11 Armenian popular tradition also preserves several tales that contain the motif of the speaking serpent such as the story of the Magic Ring.12 To this may be added the speaking serpents of antiquity in Ptolemy I’s account of the Siwah expedition of Alexander the Great. Arrian (II.3.2) records the interesting story that Alexander was guided to and from the oracle by two giant dragons uttering speech. This aspect appears already in Hesiod’s account of the hundred serpent heads that issue from the shoulders of the dragon Typhon which spoke in many voices (phonai), at times those of different animals and, at times, sounds only the gods could understand (Theogony 825–52). The belief in the serpent-dragon’s receptivity to sounds that create a pattern such as tone,

rhythm and melody, that is to say, sounds transformed into music, seems to have roots that lie in deep antiquity and which appear to have survived into the Islamic period.13 According to a ninth-century Arabic text on the poisons of serpents, attributed to Hermes Trismegistos, music attracts mythological serpents such as the large serpent in the western sea, referred to earlier, with cervid-type dendritical antlers and a mane like the mane of a seahorse, which can be hunted with the help of music made with cymbals and shawms.14 A comparable phenomenon is recorded by Ibn Waḥshiyya in his late ninth-century text on poisons in which he states that people from Kasadān invented a particular type of bell which brings out serpents and vipers from their holes when they hear the sound.15 That the sound of music can lure serpents out of their caves is similarly reported by the late ninth-century author Abū Hilāl al-ʿAskarī.16 The serpent as well as its larger relative the dragon may thus have been considered susceptible to the transformative influence of the medium of sound and music.17 This aspect of the dragon is given expression by its depiction on musical instruments, which appear in a number of early Ilkhanid paintings portraying a court scene with the enthroned ruler and his consort, housed in the Staatsbibliothek zu Berlin and in the Topkapı Sarayı Library, Istanbul.18 The court ceremony shows the enthroned couple surrounded by princely relatives and courtiers, both male and female, accompanied by musicians playing a stringed instrument, which again resembles a harp, the arched string arm of which is fashioned in the form of a dragon’s neck terminating in its projecting head (fig.╯177). Like

╇ Cf. Omidsalar, 2001, pp.╯259–93, esp.╯p.╯269 and n. 14. ╇ Littmann, “Alf layla wa-layla,” EI2 I, 358b. 11 ╇ Cf. Vogel, 1926, pp.╯20–1, 28, 173–4. 12 ╇ Hoogasian-Villa, 1966, pp.╯193–7, for other tales with this motif, see pp.╯224–9, 401 and 426–9. 13 ╇ The Old Indic name gandharva represents a beneficent mythical being who is said to be surrounded by the heavenly waters, “which flow down at his look,” and who is sometimes, mainly in later literature, portrayed as heavenly musician (Skjærvø, “Aždahā I,” EIr; Panaino, “Gaṇdarəßa-,” EIr). In the Rigveda these celestial beings are compared to the luminosity of the Sun and the stars of the Moon’s orbit; Rigveda 8.1.2 (with the Sun), Vājasaneyi-Saṃhitā 9.7 (the stars of the Moon’s orbit; cf. Oldenberg, 1894, repr. 1977, p.╯245). Etymologically the name equals the sea Â�monster Gandarəβa (Gandarw, Gandarb, spelled gndlp) which is defeated by the hero Kərəsāspa/Garshāsp.╯According to Prods Oktor Skjærvø, “it is through “Iranian polarisation” of the inherited Aryan mythological concepts […] that gandarəβa has been turned into a sea monster” (“Aždahā I,” EIr. Cf. Oldenberg, 1894, repr. 1977, pp.╯245–50;Â�

Grassmann, 1873, repr. 1976, cols. 376–7). This associates it with the Manichean demoniacal beings called mazans which dwell in the ocean as sea dragons or dragon-like sea monsters. It is further notable that the gandarəβa survives in Sogdian as γntrw (Skjærvø, “Aždahā I,” EIr). 14 ╇ Ullmann, 1994, p.╯28.81. 15 ╇ Levey, 1966, p.╯36. 16 ╇ Ullmann, 1994, p.╯100.81. 17 ╇ See also the pact made with the tamed nāgarāja who lives to the northwest of Kapisa. As soon as some unfavourable weather phenomenon occurs, the sound of the ghantā (cymbal or drum) will remind the dragon of his pact with king Kanishka, whereupon he will cause the danger to subside. See p.╯91. 18 ╇ The dragon-headed string instrument is depicted on other fols. featuring an Ilkhanid court scene, one of which is also preserved in the Staatsbibliothek zu Berlin (Dschingis Khan und seine Erben, 2005, p.╯258, cat. no. 286) and another in an album compiled in the late 1400s in Istanbul, Topkapı Sarayı Museum, Library, Ms. H. 2153, fol. 166a (the painting is attributed to the same period).

9

10

the dragon and sound

193

the overall pictorial scheme of the painting, the dragon head is rendered in an east Asian manner, the closed jaws characterised by fleshy folds on the bridge of the snout which is terminating in a prominent curl and with flowing mane and beard springing from the chin. It is interesting to consider the possibility of music as a civilising force which in turn would have had a subsuming and, hence, taming effect on the nature of the dragon, a change suggested by its appearance at the tip of a musical instrument. The closest example to the Ilkhanid dragon harp can perhaps be found in seventh- or eighthcentury Panjikent in Sogdiana. In spite of the tremendous chronological hiatus this may be of some relevance. Among the epic cycles portrayed on the Panjikent wall paintings there are at least two that depict original Sogdian legends.19 One of these includes the well-known figure of a musician that serves as a caryatid supporting an arch. The tip of her large harp-like stringed instrument terminates in a dragon head with open mouth and elongated, curved upper proboscis-like snout

whose head is turned towards the harpist (figs. 178a and b). Another indirect association of the dragon with music is shown on one of the earliest wall paintings at Panjikent, which has been dated to the late fifth century, on the eastern wall of the northern chapel of Temple II. Here the dragon serves as zoomorphic throne for a goddess identified as deity of the river Zerafshan20 to whose right is a rectangular musical instrument with attached bells. The affiliation of the goddess with the instrument has led Martha Carter to point out an analogy with the ancient Hindu Saraswatī (who is also known in Buddhism), goddess of music and learning, who is also a river goddess.21 Vestiges pertaining to elements of both the musical and the aquatic qualities of the dragon can still be found in the more recent history of Central Asia. They recur in the songs of Kazakh shamans who also use a stringed musical instrument (kobyz) of which they say:

19 ╇ Marshak, 2002, p.╯145. For a drawing of the entire mural, see Belenitskii and Marshak, 1981, p.╯24, fig.╯3. 20 ╇ Belenitskii and Marshak, 1971, pp.╯42–5, and eidem, 1981, p.╯70; Azarpay, 1981, p.╯140 and n. 61. 21 ╇ Carter, 1992, p.╯75. Revealed wisdom, which in Brahmanism is personified in Vāch, the female Logos, and later appears as Saraswatī/Prajñā, is essentially an aquatic element and had extensive influence on the formation of the character of the nāgas. This association explains why the

nāgas are said to exhibit great musical skill and possess magic musical instruments (as well as excel in other arts). Cf. Bosch, 1960, pp.╯137, 174–5. See also the connection of tritons (serpent-legged mermen associated with Tritōn, the son of Poseidon) with music as visualised in Gandhāran art, for instance in a schist frieze from Andan Dheri near Chakdara with tritons playing musical instruments. Chakdara, Dir Museum, inv. no. 505 (AND 533); Gandhara, 2009, cat. no. 75. 22 ╇ Basilov, 1991, p.╯278.

I took in my hands a kobyz made of a pine-tree And wind like a water serpent.22

194

chapter thirteen

the dragon as symbol of transformation

195

chapter fourteen

The dragon as symbol of transformation

The dragon can be said to symbolise the liminal or transitional phase in rites of passage. As such it evokes a dual response, perceived both as beneficent and an object of fear. This results also in its twin functions as deliverer and destroyer, roles that link it to the elemental cycle of birth and death. Interestingly, the Targum Onkelos, an Aramaic translation of the Pentateuch, renders the west Semitic root nḥš (“serpent”) as Aramaic ḥiwyaʾ (“serpent”) whose Arabic cognate is ḥayya (“serpent”).1 There may be, moreover, an etymological association of nḥš and the Akkadian nêsu (naāšu) meaning “to live, to stay alive, to recover.” The temptation to see this cognate as an indication of a semantic relation between the terms “serpent” and “life” has repeatedly been succumbed to.2 The hypothesis appears to be substantiated by the fact that the root ḥyw apparent in the word ḥay[w]āt, as described by Ibn Manẓūr (d.╯711/1312) in his Lisān al-ʿArab (completed in 689/1290), entails the concept of life as opposed to death.3 He further states that “the plural of all that is alive is ḥay[w]āt (“life”) ... and that ḥayawān (“the animal kingdom”) implies a notion of life (ḥayāt).”4 The word ḥay[w]āt is once mentioned in the Qurʾān (sūra 29, 64) where it means “the true life” and is used of the afterlife. Ibn Manẓūr adds that “...the derivation of ḥayya (“serpent”) from ḥayāt is well-known...”5 The same author

more clearly specifies that “...when the life of a man or a woman is prolonged, he or she is called ḥayya; this is due to the long life of the ḥayya and that is why the man or the woman is described as ḥayya for his or her prolonged life...”6 That childbirth was associated with the dragon in its beneficent aspect seems to have been a view widely held in antiquity.7 Remnants of such notions seem perhaps to resonate in texts such as the Kitāb Tadbīr al-ḥabālā wa ’l-aṭfāl (“Book of Children and Pregnant Women”), written by Aḥmad ibn Muḥammad al-Baladī (d.╯c. 380/990) who served as physician to the Egyptian vizier Abu ’l-Faraj Yaʿqūb ibn Yūsuf ibn Killīs in the tenth century. There he gives the following instruction to ease birth:

╇ Wilson, 2001, p.╯97. ╇ Idem, p.╯50. See also Astour, 1965, p.╯194. 3 ╇ Beirut, n.d., vol.╯14, p.╯211. I am grateful to Layla al-Musāwī for elucidating this point. 4 ╇ Idem, p.╯214. 5 ╇ Idem, p.╯240. 6 ╇ Idem, p.╯221. Cf. Wallace, 1985, pp.╯143–72, esp.╯pp.╯144, 148, 151 and 160; also p.╯108. For further etymological associations, see Lecerf, “Ḥayy,” EI² III, 330a. 7 ╇ In ancient Greece women would go to the sanctuary of Asklepios where a serpent would appear in their dreams and they would then get pregnant (Plutarch, Life of Alexander 2.2–3.4). Cf. Ferwerda, 1973, p.╯107. Plutarch reports that:

Klodones and Mimallones, and imitated in many ways the practices of the Edonian women and the Thracian women about Mount Haemus, from whom, it would seem, the word thrêskeuein came to be applied to the celebration of extravagant and superstitious cereÂ�monies.

1 2

... women … were addicted to Orphic rites and the orgies of Dionysus from very ancient times being called

...a snake skin wound around the hip of a woman accelerates birth.8

According to another tradition, a snake’s scale placed on a pregnant woman, presumably when she is in labour, will ensure an easy delivery.9 In the epic romance Wīs u Rāmīn these notions are reflected in Fakhr al-Dīn Gurgānī’s allegorical imagery: ...the sun is delivered from the dragon.10

Serpents or dragons also play a symbolic role in the miraculous birth and apotheosis of numer-

Cited after Asirvatham, 2001, pp.╯96–7. 8 ╇ Ullmann, 1978, p.╯109. The notion that if attached to the loins of a woman in childbirth the sloughed off skin of a snake facilitates delivery is also recorded by Pliny (Naturalis Historia XXX 44) who adds that care must be taken to remove it immediately after the birth. 9 ╇ Donaldson, 1938, repr. New York, 1973, p.╯169. 10 ╇ Tr. cited after Daneshvari, 1993, p.╯21. The parable of a dragon helping a doe deliver its young is recorded in a Talmudic passage of the Bava Batra (“The Last Gate”):

196

chapter fourteen

ous heroes and kings of antiquity.11 Many traditions name them among the ancestors of heroes and as mythical ancestors of tribal confederacies and kingly dynasties.12 The classical author Herodotus reports that the ancient Scythians who lived north of the Black Sea regarded themselves as descendants of the greatest of the Greek heroes, Herakles/Hercules, and a woman with a serpent’s lower body (Histories IV 8–9).13 With this anguipede woman he engenders three sons, the youngest of whom, named Scythes (“the Scythian”), was the worthiest and became the first king of the Scythians.14 The story of the miraculous birth of the superhuman hero who issues from the encounter of a princess with a serpent appears also in one of the oldest recorded epic tales (bylina) compiled in the eighteenth century in west Siberia.15 Among the tribal confederacies, dynasties and heroes claiming their descent from Dahāk, the hominoid serpent of the ancient Iranian epic past are, as mentioned earlier, the Kushāṇas of the Yuezhi confederacy, the Armenians living in the region near Lake Sevan, the Islamic Sām dynasty of Ghūr as well as the hero Rustam and his descendants. These claims are surpassed in the Turkish epic Ṣaltūq-nāma (“Book of Ṣaltuq”), in which the first ruler of the world, Eslem, son of Adam, becomes the father of Ẓaḥḥāk

the Turk, ancestor of all Turkish sovereigns.16 The association of serpents with birth symbolism occurs also, albeit in a different manner, in the story of Prince Sayf al-Mulūk in the Alf layla wa-layla. The tale recounts how an aged childless ruler is advised by Solomon the Wise that, in order to bring about the birth of a prince, he must cook the flesh of two serpents that appear by a certain tree at noon and serve the dish to his wife.17 More often though, the dragon is known in its other function as the awful dragon of death. On the way to his execution the mystic and theologian Manṣūr ibn Ḥusayn al-Ḥallāj (“the wool-carder,” 244/857–309/922) from Ṭūr in Fars, is said to have faced his impeding martyrdom by reciting the following verses:

The doe has too narrow a womb [to permit it to give birth; therefore] when it crouches to give birth, I prepare for her a dragon that bites her belly so it grows slack and she gives birth.

╇ Cf. P’yankov, 2006, pp.╯505–11, esp.╯pp.╯506–7. ╇ This myth was modified by Valerius Flaccus to the extent that the Scythians were said to be descendants of Colaxes (the youngest son of Targitaos who reigned in Scythia was named Colaxais; Herodotus, Histories IV 5) and the anguipede earth-born maiden, and in the account of Diodorus Siculus (II 43) according to which the same woman was impregnated by Zeus. 15 ╇ Jakobson and Szeftel, 1949, pp.╯13–87, esp.╯21, 64; Schirmunski, 1961, pp.╯58–9. It is interesting that comparable notions occur in early Christian gnostic writings, such as a text entitled Baruch written by the second-century Christian gnostic Justin, which survives in summarised form in the early third-century antiheretical work, Refutatio omnium haeresium (“Refutation of All the Sects”) of Hippolytus of Rome. According to his recapitulation of the text, in the beginning creation results of the marriage of a male divine principle, Elohim, the God of creation and Lord of heaven, and a female principle, named Eden or Israel, the mother earth, who is described as looking like a woman as far as the groin and a serpent below (Refutatio omnium haeresium 5.24.2-3). Williams, 1996, pp.╯18–9, 37–9. 16 ╇ Mélikoff, 1960, vol.╯1, p.╯43 and n. 1; Dedes, 1996, p.╯29, n. 80. For further examples in South Slavic epics, see Schirmunski, 1961, pp.╯28–30. 17 ╇ Marzolph and van Leeuwen, 2004, pp.╯362–3. The story goes back to the age-old belief that pregnancy could be caused by ingesting magic food. See Astour, 1965, pp.╯171–2. 18 ╇ Browne, 1920, vol.╯1, p.╯435; see also p.╯363.

Bava Batra 16b, cited after Morgenstern and Linsider, 2006, p.╯85. Cf. Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯2, p.╯168. 11 ╇ Scipio Africanus (Livy IIVI 19.7), Alexander the Great (Plutarch, Life of Alexander II 4), the Messian hero Aristomenes (Pausanias, Graeciae Descriptio IV 14.7–8) and the future emperor Augustus after his adoption by Julius Caesar (Suetonius, Augustus 94.4), are said to have been born from the union of their mothers with a giant serpent or dragon. Cf. Ferwerda, 1973, p.╯107; also Küster, 1913, p.╯112. 12 ╇ The Greek god Zagreus was born of the union of Persephone and Zeus who had taken the form of a dragon (Nonnos, Dionysiaca V 562.564; VI 155–7). Pentheus, king of Thebes, was the son of Echion, “the serpent-man” (the name Echion being the male form of Echidna, the serpent-monster; Euripides, Bacchae 537–44). The Greek hero Kadmos kills the drakōn that barred the way to the site of the future city and then sowed its teeth in the earth, hence giving rise to the Spartoi (“sown men”) who became the first Thebans. It is moreover interesting to note that towards the end of his life both Kadmos and his wife Harmonia were changed into serpents and lived among Encheleians (Ovid, Metamorphoses IV 576–600). On the Kadmos myth, see Fontenrose, 1959, repr. 1980, pp.╯306–20; Astour, 1965, pp.╯156–61. Similarly, the Indian kings of Chhota Nāgpur claim origin from a nāga called Puṇḍarīka. Vogel, 1929, p.╯35.

My friend doth unrelated stand to aught of ruth or clemency: From His own cup He bade me sup, for such is hospitality! But when the Wine had circled round, for sword and [executioner’s] carpet called He Who with the Dragon drinketh Wine in [the heat of] Summer, such his fate shall be.18

The use of the dragon as metaphor for the inevitable fate of death occurs in a passage of the

13 14

the dragon as symbol of transformation

197

Mathnawī, in which the mystic poet Jalāl al-Dīn Rūmī (604/1207–672/1273) allegorically states:

saying of the Prophet on the torments in Hell inflicted by dragons:

When Destiny comes, the wide spaces are narrowed. A hundred ways and asylums may lie to left and right; yet they are all barred by Destiny, the invincible dragon.19

Concerning the chastisement of the truth-concealer in his grave: ninety-nine tinnīns will be given mastery over him. Do you know what a tinnīn is? It is a serpent. There will be ninetynine serpents, each of which has nine heads: They will gnaw at him, eat at him and blow into his body until the day he is raised up.23

A related conceptualisation of the “invincible dragon” appears not only in Arabic and Iranian literature, but in a saying of the military commander ʿAbd Allāh ibn ʿAmr ibn al-ʿĀṣ, a contemporary of the Prophet Muḥammad, in which he likens himself to the “inexorable serpent”: When others looked askance, I blinked not; Then I partially closed my eyes, but not in winking [at the sight of danger]. You saw me return [to the charge] and continue to dash forward. I support [equally well] good and evil, and am inexorable, Like the serpent at the foot of the trees.20

Although the Qurʾān makes no allusion to Hell being populated by huge serpents, such a belief is preserved in later traditions. It is attested in a ḥadīth by the tenth-century Ḥanafī jurist Abu ’l-Qāsim Isḥāq al-Samarqandī (d.╯342/953–4), also known as al-Ḥakīm (“the Wise One”), according to which: ... the Prophet, blessings and peace be upon him, said that in Hell are snakes each as large as a camel and the pain of their bite will last for forty years.21

Yet another tradition speaks of the punishment for insolence against God: ...as a black snake [that] winds itself around the impudent man’s neck and kills him after forty days.24

This belief in the dragon as “inexorable death” is still echoed in the popular culture of Iran, according to which there is a serpent in every grave that will torment the dead in proportion to the number of sins committed in life.25 Similarly, in popular belief in present-day Afghanistan it is said that following burial the mythical angel ʿIzrāʾīl/ʿAzrāʾīl, who has authority and power over death, appears and grips the tongue of the deceased for questioning.26 This idea may reflect the Qurʾānic tradition according to which the angel of death ʿIzrāʾīl creeps into the dying man’s throat to draw out his spirit (sūra 79).27 In the same way, the fourteenth-century encyclopedian al-Damīrī relates the interpretation of a dream according to which: He who dreams of a serpent coming out of his mouth while he is ill, will die, for that indicates his life which will have come out of his mouth.28

In his grave a sinner may be tortured by a serpent of fire which bites him until the day of judgement.22 Another ḥadīth records a more explicit

The early fourteenth-century Syrian traditionist Ibn Kathīr (d.╯774/1373) reports a ḥadīth on

╇ Tr. and ed. Arberry, 1961, p.╯271. ╇ Cited after al-Damīrī, Ḥayāt al-ḥayawān al-kubrā, tr. Jayakar, 1906, vol.╯1, p.╯634. 21 ╇ Tr. into Pers., ed. Ḥabībī, A., Tehran, 1969, p.╯83, cited after Daneshvari, 1993, p.╯18. 22 ╇ Wensinck and Tritton, “ʿAdhāb al-ḳabr,” EI² I, 168b. 23 ╇ Chittick, 1992, p.╯90, 9–15. 24 ╇ Al-Qalyūbī, Aḥmad ibn Salāma, Nawādir, no. 29, as cited in Ritter, 2003, p.╯183. 25 ╇ Bess Donaldson (1938, repr. 1973, p.╯168) remarks that this belief is mainly current “among the uneducated women.” 26 ╇ Private communication. A tradition connecting the serpent with the moment of the death of a human being was also well-known in antiquity. When a serpent passed through a hole in a wall, the ancients would say that the soul was definitely separated from the body and had started its descent into the underworld. Cf. Porphyrios, Vita Plotini (II, 27). According to Pliny (Naturalis Historia XVI 85.234), a draco lived in a cave near the grave of Scipio Africanus the Elder to watch over his soul. On the role of serpents as guardians of graves, see also Küster, 1913, pp.╯67–71. In the Testament of

Abraham (recension A ch. XVII) which is generally considered to be a Jewish work, datable to the first century ad (cf. Delcor, 1973, pp.╯63–5, 72, 76–8), when death comes to fetch Abraham’s soul, he shows him seven flaming drakōn heads as well as the faces of various poisonous serpents. In ch. XIX, death explains the function of the faces as the different manners of death, while the seven drakōn heads stand for death raging seven aeons long. Similar associations of the Greek words drakōn and ophis with death and the underworld appear in the Greek Apocalypse of Baruch. See Schlüter, 1982, pp.╯46–8. For traditions in which Rahav and Leviathan are identified with the angel of death, see Ginzberg, 1909–38, repr. 1946 and 1955, vol.╯1, p.╯40 and n. 187. 27 ╇ Cf. Wensinck, “ʿIzrāʾīl,” EI² IV, 216b. 28 ╇ Tr. Jayakar, 1906, vol.╯1, p.╯655. See also the troublesome dream of Kanaʿān ibn Kūsh, the father of Namrūd (the Nimrod of the Bible), in which his son is born and a snake enters his nose, an ominous sign which is interpreted to mean that his son will kill him. Heller, “Namrūd, also Namrūdh, Nīmrūd,” EI² VII, 952b.

19 20

198

chapter fourteen

authority of Abū Hurayrah according to which the Prophet Muḥammad said:

‘and to dust you shall return.’35 And to [the Â�serpent He said, ‘Dust shall become your food.’36,37

Anyone who does not remit the obligatory alms of the wealth which God has bestowed upon him, his wealth will appear to him on the Day of Resurrection as an ancient male serpent, bald from age and the accumulation of poison in its head, and with two projections dangling from its forehead. It shall coil itself around his neck and, grabbing him on both sides of the face, shall cry out, ‘I am your wealth, I am your treasure!’29

... Indeed, why did Justice, which interrogated Adam and Eve, not interrogate the serpent? And if the nature [of the serpent] was bound, why was it also condemned with those [Adam and Eve] who had free will? And if it had free will, why did [Justice] not interrogate it? [Justice] did not interrogate [the serpent] for the reason that It knew who It was with whom It was speaking. And that one who hears knows what he hears. And that the serpent was the first to receive punishment was so that [Adam and Eve] might repent, show compunction and make supplication to [God]. But when they were unwilling to turn [to Him] in contrition, [God] set upon them the decree of death, and they became dust, fodder for the serpent. ‘From dust you are,’ [God] said,

The close association of the mythical creature with the departure from life is underlined by its representation in the thirteenth-century barrelvaulted funerary chapel of Surb Grigor. Part of the monastic complex of Noravankʿ in Vayots Dzor in southern Armenia, it was added in 1275 by prince Tarsayich Ōrbēlian, governor of Siunikʿ province, at the north side of the church of Surb Karapet as mausoleum of the Ōrbēlian princes. The point of juncture of the wall arch and vaulted ceiling, above the semi-circular altar, is entirely horizontally circumscribed by two pairs of confronted dragons in low relief and highlighted with white and red pigment. The elongated serpentine bodies, which are enlivened with a spotted pattern, are arranged in evenly spaced loops (fig.╯179). In Iranian poetry death is sometimes metaphorically likened to the maleficent influence of the dragon, alluding to the belief that the moon, in eclipse, is swallowed by the eclipse dragon (aljawzahar). In Niẓāmī’s Haft Paykar the officer who brings the false report of the execution of the “bright-faced” luminous harp girl Fitnah to king Bahrām tells him that he has “given the Moon to the Dragon,” that is to say, eclipsed her life.38 The extent of the geographical and chronological dissemination of the dragon’s allegorical role in death is attested by the life stages of a parable called the “Man in the Well” in which it figures prominently. It is of great interest since the parable’s sequence of transmission can be documented. It was translated into Middle Persian (Pahlawī) from an early Sanskrit text, perhaps from the eleventh book of the great Indian epic Mahābhārata (chapters 5–6),39 together with other texts from the Mahābhārata and the Panchatantra, by Burzōe of Nīshāpūr, the personal

╇ Tafsīr al-Qurʾān al-ʿAẓīm, vol. II, pp.╯165–7, cited after Ayoub, 1992, p.╯390. 30 ╇ Cf. Epstein, 1997, p.╯74. 31 ╇ Babylonian Talmud, Bava Batra 74b, as cited in idem, p.╯74. 32 ╇ Idem. 33 ╇ The serpent appears as giver of life and possessor of the magic herb of life in Babylonian literature. Searching for the secret of immortality Gilgamesh succeeds in harvesting the magic plant from the bottom of the sea (Epic of Gilgamesh XI 287–9), only to have it stolen from him by the serpent. 34 ╇ For the story of Glaucos, a son of king Minos of Crete, who was restored to life by a herb placed by a serpent onto a dead serpent which thereupon was brought back to life,

see Ovid, Fasti, VI. 749–54, and Apollodorus, Bibliotheca III. 3.1–4. A similar story of a young dragon being raised from the dead through the agency of a plant and the same plant being used to resurrect the Lydian hero Tylon, who had been killed by a dragon, is recorded by Pliny (Naturalis Historia XXV 5). 35 ╇ Genesis 3,19. 36 ╇ Genesis 3,14. 37 ╇ Tr. and comm., Mathews, 1998, pp.╯34–5. 38 ╇ Tr. Meisami, 1995, 25:71. Cf. eadem, 1987, p.╯218. 39 ╇ De Blois (1990, pp.╯34–5) shows that since the parable was recorded in several Indian texts, if it was not taken from the Mahābhārata, it would have come from another contemporary Indian source available to Burzōe.

In the Jewish tradition it is also the serpent that has power over life and death.30 In the Talmudic tractate Bava Batra a serpent is said to be the keeper of a miraculous stone that has power of life or death.31 In the same text the angel of death himself is called Leviathan.32 In ancient Near Eastern33 and Greek34 lore the serpent is in the possession of a plant which can restore life. The Armenian commentary on the book of Genesis, attributed to the Syriac writer Ephrem, gives further insight into contemporary Christian understanding of the relation of the serpent with death:

29

the dragon as symbol of transformation physician of the sixth-century Sasanian king Khusraw I (Khusraw Anūshirwān, r. 531–579). About two centuries later Burzōe’s Pahlawī version40 of the Indo-Iranian political fable was translated into Arabic prose by ʿAbd Allāh ibn al-Muqaffaʿ (c. 102/720 – executed c. 139/756–7) as the celebrated Kalīla wa Dimna. In the tenth century the fables were translated into Persian under the Samanid Naṣr ibn Aḥmad (302/914– 331/943), but only sections of Rūdakī’s version survive in quotations.41 The text thus evolved to include stories that originated from different cultures. Less than two centuries later they were again translated into a Persian text (which survived) by Abu ’l-Maʿālī Naṣr Allāh ibn Muḥammad for his patron, the Ghaznawid Bahrām Shāh (r. c. 511/1117–552/ 1157), attesting to their continued popularity.42 A chapter in the Kalīla wa Dimna, entitled “The Perils of Life,” relates the following story (fig.╯180): A man fled from an enraged camel and out of necessity suspended himself in a well; his hands fell on two branches which were growing above the well and his feet obtained a foothold. Soon he could see better that his two feet were on the heads of two snakes who had emerged from their holes. His glance fell to the bottom where he saw a terrible dragon with its mouth open in expectation. At the top of the well he noticed a white and a black rat incessantly gnawing away at the roots of the branches. Meanwhile he considered 40 ╇ Burzōe’s Pahlawī translation is lost but survives in a Syriac version of about 570 by the Periodeut Būd. The transmission of the Kalīla wa Dimna from India is also described in the Shāh-nāma (tr. and ed. Mohl, 1838–1878, vol.╯6, pp.╯445–57); cf. Brockelmann, “Kalila wa-Dīmna,” EI2 IV, 503a. 41 ╇ Cf. Brockelmann, “Kalila wa-Dīmna,” EI2 IV, 503a. 42 ╇ Idem. 43 ╇ Mīnuvī’s 1343/1964 compilation of Naṣr Allāh’s Persian version of the Kalīla wa Dimna, p.╯56 l. 17 to 57 l. 13, as cited in O’Kane, 2003. 44 ╇ Idem. The desc