Studies in the Islamic Arts of the Book 1904597491, 9781904597490

Citation preview

Studies in the Islamic Arts of the Book

Studies in the Islamic Arts of the Book

Robert Hillenbrand

The Pindar Press London 2012

Published by The Pindar Press 40 Narcissus Road London NW6 1TH · UK

British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library

ISBN 978 1 904597 49 0

Printed by De Montfort Publishing Ltd. (trading as De Montfort Print) 18 Slater Street, Leicester LE3 5AY This book is printed on acid-free paper

Contents Prefaceiii I

The Uses of Space in Timurid Painting

1

II

The Iconography of the Shāh-nāma-yi Shāhī51

III

The Iskandar Cycle in the Great Mongol Šāhnāma79

IV Images of Muhammad in al-Biruni’s Chronology of Ancient Nations111 V

The Arts of the Book in Ilkhanid Iran

149

VI The Paintings of Rashīd al-Dīn’s “Universal History” at Edinburgh

207

VII Mamlūk and Īlkhānid Bestiaries: Convention and Experiment

211

VIII

290

The Qur’an Illuminated

IX The Relationship between Book Painting and Luxury Ceramics in 13th-Century Iran

307

X The Message of Misfortune. Words and Images in Sa‘di’s Gulistan327 XI

Literature and the Visual Arts

359

XII

New Perspectives in Shahnama Iconography

365

ii XIII Erudition Exalted: The Double Frontispiece to the Epistles of the Sincere Brethren

377

XIV The Shahnama and the Persian Illustrated Book

420

XV

452

Islamic Bookbinding

XVI The Schefer Harīrī: A Study in Islamic Frontispiece Design

465

 Bibliography Update507 Index 

509

Acknowledgements555

Preface

P

erhaps I may use this preface to sound a personal note and thereby explain something of the genesis of this book. I first encountered the arts of the Islamic book in the great exhibition of Persian miniature painting mounted by Basil Robinson at the Victoria and Albert Museum in 1967, but my collateral interest in Islamic architecture prevented me from undertaking any further exploration in the immediate aftermath of that exciting experience. Then, in 1975, I received out of the blue an invitation to curate an ambitious exhibition of Persian painting to be held in the course of the Edinburgh Festival two years later. The learning curve was dauntingly steep, and the two years spent selecting some 250 exhibits for the show, entitled (like its catalogue) Imperial Images in Persian Painting, were effectively a crash course in a new and utterly absorbing field of enquiry. The rewards were immediate, and it was an unforgettable experience to see so many paintings, most of them in their original manuscripts and others on detached leaves, in this short space of time. Thereafter Persian and Arab book painting became a core element in my undergraduate and postgraduate teaching at all levels, and this choice was justified by the enthusiastic response of generations of students to the multiple challenges of this material. The presence in Edinburgh University Library of two celebrated masterpieces of early Persian painting, namely the 1307 copy of al-Biruni’s The Chronicle of Ancient Nations and the 1314 copy of The World History of Rashid al-Din, enabled these students to experience some of the greatest of Persian paintings at first hand, and this made a profound impression on many of them. So perhaps it is not surprising that both these manuscripts figure in the articles or book chapters reprinted here. More generally, growing familiarity with the Biruni and Rashid al-Din manuscripts predisposed me to study the beginnings of Islamic book painting rather than its later periods. In the event, most of the courses on Arab and

iv

Persian painting that I taught in Edinburgh and elsewhere ended with the advent of the Safavids. The teaching included a series of lectures intended as a framework to orientate the students in an unfamiliar field, while the associated tutorials provided an opportunity to investigate individual manuscripts, themes or schools in greater depth. These tutorials, typically involving half a dozen students round a table, with relevant images in front of them and with every student given a short (and different) reading assignment in advance so that each could bring some particular angle to the general debate, lasted something over an hour and occurred twice a week. The students did not present papers; instead, the intention was that everyone present would always make a contribution, of whatever length, so that each student had a personal stake in the success of the class. Even when the main outlines of the topic under discussion were well enough known, there were still fresh challenges at every turn, and a free-wheeling discussion peppered with questions proved an effective way of tackling them. This formula helped to dispel students’ shyness and to build their confidence, and it was perennially exciting and unpredictable to see the directions that the discussions took. Ideas — from the excellent to the indifferent — would come thick and fast, and it was almost always a learning experience for me as the teacher, especially when the dearth of scholarship on a given subject forced students to think for themselves. The excitement of the chase was catching, and even if the quarry proved elusive, it was no matter. Essays and dissertations took the process of discovery further, and many of these students soon published the results of their undergraduate research. I count myself lucky to have had my own mind and knowledge stimulated and expanded by these experiences, which in their different ways have informed all the articles republished here. Several interlocking themes can be discerned in this collection of papers. They include the development of the frontispiece; the Shahnama and its iconography; the close study of individual manuscripts; and how Timurid painting achieves its most characteristic triumphs. These themes recur in several articles published after 2007 and thus after the closing date for items in this book. Six of the sixteen studies reprinted here deal with Persian material; the remainder are about Arab painting or have a have a strong Arab element. The articles also reflect different occasions. Some are closely focused on specific manuscripts; others review aspects of the field; still others handle specific problems of iconography or the formation of a distinctive style.

preface

v

I am very grateful indeed to Liam Gallagher and Tom Symonds of Pindar Press for the trouble they have taken to maintain the highest possible standards in the production of this book, from start to finish. Their commitment to the project and their unflagging attention to detail compel admiration. Robert Hillenbrand Edinburgh, March 2011

JOBNAME: ???? PAGE: 1 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 39B52A72 /production/pindar/booksxml/hillen/chapter−01

I The Uses of Space in Timurid Painting

N

O reasonably attentive observer could fail to notice, after looking at a representative series of fifteenth-century Persian book paintings, that they obey certain conventions governing the depiction of space.1 Many of these conventions, as it happens, differ significantly from those followed in Western painting, whether medieval or post-medieval, and therefore require those that look at them, at least if they are Westerners, to make certain visual and mental readjustments. Such readjustments are not in fact unduly difficult to make, which suggests that the thought world which generated the spatial conventions of Timurid painting is not as arcane as might at first appear. For those familiar with pre-Timurid Persian painting these conventions are even less foreign, since a surprising number of them may be detected, if only in embryonic form, in earlier Islamic book painting. Such comments are not intended to deny the originality of the early imperial Timurid style; but, if they are well founded, they do suggest that in spatial values — as perhaps in other elements of these paintings — it is principally such qualities as polish, refinement, and assurance2 that should

1 To the best of my knowledge there has been only one attempt to study these conventions in some detail: Dzul H. Zain, Formal Values in Timurid Miniature Painting (Kuala Lumpur, 1989). For similar work on Arab painting, see David James, “Space-forms in the Work of the Baghda¯d Maqa¯ma¯t Illustrators,” Bulletin of the School of Oriental and African Studies 37, 2 (1974): 305–20. 2 For an extended assessment of these qualities, see Ernst J. Grube, The Classical Style in Islamic Painting (Lugano, 1968), pp. 26–28; Eleanor G. Sims, “The Timurid Imperial Style: Its Origins and Diffusion,” aarp 6 (1974): 57–58, 64; Ernst J. Grube and Eleanor G. Sims, “The School of Herat from 1400 to 1450,” in Basil Gray, ed., The Arts of the Book in Central Asia (London and Paris, 1979), pp. 154, 156, and 158. See also Grube’s “Two Kalı#lah wa

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4E130F86 /production/pindar/booksxml/hillen/chapter−01

2 compel admiration, and not the discovery of something brand new. Of course, the cumulative effect of such improved execution in all aspects of the painting does indeed create a palpable gulf between Timurid and earlier Persian painting. Nevertheless, that gulf is by no means simply attributable to a difference in kind — a quantum leap, so to speak; much of it depends upon differences of degree. This point will emerge with greater clarity if some of the significant earlier examples of Persian, or for that matter Arab, painting are examined. Nevertheless, shortage of space excludes a fully detailed study of the evidence that bears on the uses of space in these and later paintings. It has seemed best, therefore, to highlight the salient developments in no more than four specific areas where spatial experiment is at its most intense, though this will not exclude passing references to some of the other relevant topics. The themes chosen here are architecture, the preference for solid blocks of color or form, the margin, and the use of empty space. The first of these categories to be examined in this paper, namely architecture, will furnish useful examples of the differences in degree noted earlier. In Arab painting, as in the Hariri3 and Dioscorides4 manuscripts, the buildings are deliberately simplified and shown in long section as a sequence of compartments, each box containing a person or a group. The Varqa va Gulshah manuscript proves that the same mode was known in early Persian book painting.5 In time, both Arab and Persian painters enriched this formula, as is shown by the frontispiece to the Rasa il of the Ikhwan al-Safa in the Süleymaniye Library, Istanbul,6 or by the paintings produced in the Rashidiyya scriptoria at Tabriz.7 They did so by breaking down the stark two-dimensional divisions of this nursery architecture — people lean over Dimnah Codices Made for Baysunghur Mı#rza¯: The Concept of the ‘Classical Style’ Reconsidered,” Problemi dell’età timuride. Atti del III Convegno internazionale sull’Arte e sulla Civilità Islamica (Venice, 1979), p. 115. 3 For relevant color illustrations, see Richard Ettinghausen, Arab Painting (Geneva, 1962), p. 121; P.J. Müller, Arabische Miniaturen (Geneva, 1979), pl. 13; and Desmond Stewart, Early Islam (Weert, 1975), pp. 92–93. 4 Bernard Lewis, ed., Islam and the Arab World (London and New York, 1976), p.197. 5 Ibid., p. 104. 6 Ettinghausen, Arab Painting, pp. 98–99. 7 See the copy of the Jami c al-Tawarikh in the Edinburgh University Library, ms. Arab 20, fols. 17a, 42a, 52a, and 54a (David Talbot Rice and Basil Gray, The Illustrations to the “World History” of Rashid al-Din (Edinburgh, 1974), pls. 21, 29, 34–35; and Basil Gray, Persian Painting (Geneva, 1961), p. 25.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 50CF3124 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

3

balconies,8 draperies curl around columns instead of being contained between them,9 groups of courtiers overlap the boundary between one spatial compartment and another.10 The scene of Samson destroying the temple of Dagon in the Edinburgh Jami c al-Tawarikh explores new ground even more boldly, in that it integrates architectural form and subject matter in a meaningful way.11 By the time of the Demotte Shahnama significant further advances had been made. The architecture, being now much more extensively articulated and embellished,12 no longer had the makeshift rickety air which characterized the stage-set buildings depicted earlier; and at times it overflowed with people. These people could be imagined as using it, perhaps even living in it. Nevertheless, the old sexpartite sectional division often still held good,13 even though it no longer defined or restricted spatial movement within the picture. In the also undated but probably later illustrations to the Kalila wa Dimna in Istanbul,14 and in some Shahnama illustrations of the later fourteenth century,15 this formula was finally discarded in favor of a more variegated set of interiors, often richly decorated with carpets, tiles, carved stucco and wall paintings — a sumptuous backdrop against which the action unfolded. All this gave late Jalayirid and early Timurid painters a solid base of achievement on which to build. They seized the opportunity with both hands, and greatly expanded the range of architectural types to be depicted

8 Ettinghausen, Arab Painting, pp. 98–99. This idea can in fact already be found in the Schefer Hariri of 634/1237 in the Bibliothèque Nationale, Paris, ms. arabe 5847, in the scene of Abu Zaid in the tavern, where a servant hands down a jug of wine from the upper floor. For a color plate, see Stewart, Early Islam, p. 95. 9 Ettinghausen, Arab Painting, pp. 98–99; here, too, parallels from an earlier Hariri ms. may be cited – ibid., pp. 106–7 (the Leningrad Maqamat), where figures hug the columns of the audience chamber. 10 See the Edinburgh Jami c al-Tawarikh, fols. 143a and 147a (Talbot Rice, Illustrations to the World History, pls. 69–70); and Basil Gray, The World History of Rashid al-Din: A Study of the ‘‘Royal Asiatic’’ Society Manuscript (London, 1978), pl. 31. 11 Fol. 11a (Talbot Rice, Illustrations to the World History, pl. 15). 12 Oleg Grabar and Sheila S. Blair, Epic Images and Contemporary History (Chicago, 1980), pls. 28 and 30. 13 Toby Falk, ed., Treasures of Islam (London, 1985), p. 53 (Garshasp enthroned). 14 Gray, Persian Painting, pp. 38–39; Jill S. Cowan, Kalila wa Dimna: An Animal Allegory of the Mongol Court (New York and Oxford, 1989), pl. 2. 15 Gray, Arts of the Book, pls. 20–22.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 5186BBF5 /production/pindar/booksxml/hillen/chapter−01

4 as well as their functions in the painting. Now, if a tripartite scheme were to be used, it might very well be decentralized and strikingly asymmetrical, with the ruler enthroned at the extreme right16 or left; or an interior might be divided vertically into two unequal sections;17 or it might consist of a single great chamber;18 or the central part might expand to such an extent that the side wings would all but disappear.19 If we constantly remind ourselves of what in theory we know already, namely that these pictures were not individually isolated works of art operating in a vacuum, as their reproduction in today’s art books subliminally implies, but were part of a sequence of illustrations in a book whose leaves would be turned over at intervals, so that each picture is enriched by those that precede it, another dimension of this variety will appear in its proper light. It can then be seen as an attempt to inject a shot of realism into this most hackneyed aspect of contemporary pictorial language. It drives home the point that one building is indeed different from another. Most strikingly of all, Timurid painters learned to suggest an architectural framework rather than to display it. In the Juki Shahnama scene of Tahmina entering Rustam’s chamber,20 it is the differences in plane within the architecture which help to

16

Karin Ådahl, A Khamsa of Nizami of 1439: Origin of the Miniatures — A Presentation and Analysis, Acta Universitatis Upsaliensis Figura Nova Series 20 (Uppsala, 1981), illustrations of fols. 42a, 75a, 82b, 202a, 267b, 302b, and 324b. Often there is a narrative quality in such placing, e.g., the conversational scenes in Baysunghur’s copy of the Chahar Maqala (Eleanor G. Sims, “Prince Baysunghur’s Chahar Maqaleh,” Sanat Tarihi Yıllıgˇı 6 [1976], fig. 3) or the game of chess in the British Library Anthology of 1468 (ms. Add. 16561), fol. 36b (Basil W. Robinson, Persian Miniatures [New York, n.d.], pl. 8). For a lesser degree of displacement of the royal figure in an otherwise centralized composition, cf. fol. 30b of the Juki Shahnama (London, Royal Asiatic Society, ms. 239: Ivan Stchoukine, Les peintures des manuscrits timurides [Paris, 1954], pl. 65). 17 As in the scene of the thief discovered in the bedchamber, from the Kalila wa Dimna of 834/1430 in the Topkapı Saray (see Filiz Çagˇman and Zeren Tanındı, trans., expanded, and ed. J. Michael Rogers, The Topkapı Saray Museum: The Albums and Illustrated Manuscripts [London 1986], pl. 57). 18 As in the scene of Luhrasp enthroned from Baysunghur’s Shahnama (Arthur Upham Pope and Phyllis Ackerman, A Survey of Persian Art from Prehistoric Times to the Present [London and New York, 1939], pl. 870). 19 Cf. Bahram Gur with the Indian princess in the black pavilion, from the Khamsa of Nizami dated 849/1445–46 in the Topkapı Saray, fol. 168a (Çagˇman, Tanındı and Rogers, Topkapı Saray Museum, pl. 60). 20 James V. S. Wilkinson and Laurence Binyon, The Shah-Namah of Firdausi: The Book of the Persian Kings (London, 1931), pl. opposite p. 24.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 5031F93C /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

5

structure the picture and above all to integrate it with the text (fig. 1). The vertical divisions of the architecture reinforce those of the text columns, and the blocks of color operate in harmony with that aim. Nevertheless, the correspondence is not exact, and this is designedly not so. Indeed, within the illustration proper the spatial divisions of the architecture, not the text columns, are echoed by an accumulation of seemingly insignificant detail — Rustam’s bow and pillows, the candles and the servant. Other details do not, so to speak, slot into this rectilinear grid, notably the carpet, the golden tableware, and the pose of Rustam himself. Thus the solid skeleton of the composition can easily accommodate minor divergences. Indeed, it is these so-called “minor” divergences which prevent the composition from becoming too obviously mechanical. Contrivance is there, of course, but it is decently masked. The key to the artist’s success lies in maintaining an unobtrusive balance between an intellectually imposed order and the casual disarray of daily life.21 Even though the architecture depicted in Timurid paintings has attracted steadily increasing interest in recent times,22 much still remains to be discussed about its spatial role. For example, many buildings embody an array of devices to draw the eye into the picture (fig. 2). Flights of steps are invitingly located near the bottom of the painting, their three-dimensionality suggested not by shading but via blocks of color.23 People lean inquisitively over balconies24 or peer in more timid fashion from half-shuttered windows25 or through grilles.26 Curtains are drawn aside to disclose figures

21

But cf. the destructive analysis of this painting by Eric Schroeder, Persian Miniatures in the Fogg Museum of Art (Cambridge, Mass., 1942), pp. 52–56. 22 Asma Serajuddin, “Architectural Representations in Persian Miniature Painting during the Timurid and Safavid Periods.” Ph.D. diss., University of London, 1968; Michele A. de Angelis and Thomas W. Lentz, Architecture in Islamic Painting (Cambridge, Mass.,1982). 23 Gray, Persian Painting, p. 123. 24 E.g., in the celebration of the day after the wedding from the Kulliyat of Khwaju Kirmani in the British Library, ms. Add. 18113, fol. 45b (Thomas W. Lentz and Glenn D. Lowry, Timur and the Princely Vision: Persian Art and Culture in the Fifteenth Century [Washington, D.C., and Los Angeles, Calif., 1989], p. 55) or in the fire ordeal of Siyavush from the Juki Shahnama (Stuart Simmonds and Simon Digby, The Royal Asiatic Society: Its History and Treasures [Leiden and London, 1979], frontispiece). Perhaps the best example is that from the Gulistan of Sa cdi in the Chester Beatty Library, Dublin, P. 119, fol. 9a (Gray, Persian Painting, p. 87). 25 Compare Khusraw before Shirin’s palace in a Nizami datable to 1405–10 in the Freer Gallery of Art, Washington, D.C. no. 31.36 (Gray, ed., Arts of the Book, pl. 33); or Laila

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 58F9E536 /production/pindar/booksxml/hillen/chapter−01

6 stepping in or out of doors.27 Drawbridges welcome the viewer into a castle and into the painting at the same time.28 Empty half-open doorways operate as a kind of repoussoir.29 The sides are often acutely angled, forming wings which unobtrusively nudge the attention towards the central space between them.30 Sometimes these devices are used in concert at different levels, compelling the eye to move upwards and further into the painting.31 Closely packed receding planes, often with people jammed between them — for

and Majnun in Paradise from a Khamsa dated 854/1450 in the Metropolitan Museum of Art, New York, no. 13.228.3, fol. 181b (Basil W. Robinson, Persian Drawings from the 14th to the 19th Century [Boston and Toronto, 1965], pl. 75). 26 E.g., the scene of Humay in the fairy palace from the Vienna ms. of the mathnavis of Khwaju Kirmani (Robinson, Persian Drawings, pl. 13) or the leaf showing Farhad being brought before Shirin, from a lost Khamsa of Nizami of ca. 1405–10 (Gray, Persian Painting, p. 54). 27 E.g., the scene of Majnun at the Ka cba from Baysunghur’s Anthology dated 823/1420 in the Museum für Islamische Kunst, Berlin I.4628, fol. 305a (Gray, ed., Arts of the Book, pl. 34); and the celebration of the birth of Majnun from the Khamsa of Amir Khusraw, dated 890/1485 in the Chester Beatty Library, P. 163, fol. 104b (ibid., pl. 58). 28 E.g., Faramarz mourning over the coffins of Rustam and Zavara, from Baysunghur’s Shahnama (Survey, pl. 871), the slaying of Arjasp in the same ms. (Ernst J. Grube, The World of Islam [Feltham, 1967], pl. 57) or the picture of the Mongols besieging a city, in the Tarikh-i Jahan Gushay of Juvaini dated 841/1438 in the Bibliothèque Nationale, supp. pers. 206, fol. 149a (Gray, “The School of Shiraz,” Arts of the Book, p. 143). 29 Cf. the scene of Khusraw before Hurmuzd from the British Library ms. of the Khamsa of Nizami dated 900/1494–95, Or. 6810, fol. 37b (Lentz and Lowry, Timur and the Princely Vision. p. 283), or fol. 135b of the same ms. (Gray, Persian Painting, p. 123). For an open door used in the same way, see the depiction of a funeral procession in an undated late Timurid painting in the Walters Art Gallery, Baltimore, no. 10.678 (Marie G. Lukens, “The Language of the Birds: The Fifteenth-Century Miniatures,” Bulletin of the Metropolitan Museum of Art [1967], p. 331, fig. 23). 30 Cf. the Nizami of 849/1445–46 in the Topkapı Saray, H. 781, fol. 168a (Çagˇman, Tanındı, and Rogers, Topkapı Saray Museum, pl. 60). Sometimes, especially when the royal figure is at one side of the picture, only one such angled wing is employed, thereby drawing attention to him (e.g., fol. 62a in the same ms.; see Lentz and Lowry, Timur and the Princely Vision, p. 108, or the manuscript of the mathnavis of Khwaju Kirmani in the Österreichische Nationalbibliothek, Vienna, N.F. 382, fol. 10b, for which see Robinson, Persian Drawings, pl. 13). 31 E.g., in the scene of Tahmina visiting Rustam from a lost early Timurid manuscript in the Fogg Museum of Art (Lentz and Lowry, Timur and the Princely Vision, p. 130). It is noticeable that the angled wings in the upper part of the picture suggest yet further recessions of space.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 44 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 541DDC7E /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

7

example, curtain walls (frequently concentric) in depictions of castles,32 of sieges,33 or of pilgrimage to Mecca34 — bring the third dimension to life. Sometimes nature is brought into the equation, as in the mountainous rocks which girdle Arjasp’s impregnable fastness in the Juki Shahnama (fig. 3).35 The prevalent contemporary fashion for composing a picture on several levels is often echoed in the architectural forms themselves, and these multiple stories may be stepped inwards from each other, again implying depth.36 A more dramatic version of the same idea is found in the sharp zigzag movement of successive flights of stairs in the celebrated scene of Yusuf pursued by Zulaikha in the Cairo Bustan of 893/1488 — a painting which implies the passage of time as well as a sequence of spaces (fig. 4).37 The scene of Isfandiyar slaying Arjasp in the Brazen Hold from the Gulistan Shahnama made for Baysunghur in 833/142938 also has this double dimension of time and space, for the forbidding maze-like architecture of the fortress, with its discordant geometry of violently juxtaposed angles or planes, vividly evokes the hero’s difficulties in penetrating the castle’s successive lines of defense before he can get his hands on Arjasp (fig. 5). In both these cases the explosive impact of the encounter between the two major protagonists owes much of its intensity to the earlier temporal and spatial buildup. The artist has responded to the accumulated suspense and eventual dramatic climax of the literary text with an extraordinarily apt visual 32

Such as the British Library Khamsa, Or. 6810, fol. 273a (Lentz and Lowry, Timur and the Princely Vision, p. 250). 33 For example, the siege of Gang Bihisht from the Juki Shahnama (Wilkinson and Binyon, Shah-Nameh, pl. 12). 34 As in the Topkapı Sarayı Khamsa of Nizami, H. 781, fol. 111b (Gray, ed., Arts of the Book, pl. 54; for a detail, see Lentz and Lowry, Timur and the Princely Vision, p. 219), or the Anthology dated 813/1410–11 in the British Library, Add. 27261, fols. 362b-63a (Lentz and Lowry, Timur and the Princely Vision, p. 118). 35 Lentz and Lowry, Timur and the Princely Vision, p. 134. 36 As in the British Library Khwaju Kirmani ms., fol. 26b (Gray, Persian Painting, p. 46). 37 Lentz and Lowry, Timur and the Princely Vision, p. 294 (Cairo, General Egyptian Book Organisation, Adab Farsi 908, fol. 52b). For a penetrating discussion of this painting, see Lisa Golombek, “Toward a Classification of Islamic Painting,” in Islamic Art at the Metropolitan Museum of Art, ed. Richard Ettinghausen (New York, 1972), pp. 28–29, to be supplemented by Lentz and Lowry, Timur and the Princely Vision, p. 292. 38 Grube, World of Islam, pl. 57. For a brief analysis of this picture, see Basil Gray, The Shahnameh of Ferdowsi. The Baysonghori Manuscript: An Album of Miniatures and Illuminations (Tehran, 1971), p. 94. The images are water-color copies of the originals.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4B697EEB /production/pindar/booksxml/hillen/chapter−01

8 equivalent whereby the principals of each tale confront each other at the very top of the picture — pictorially speaking, at the very last moment. Thus time is suggested by space. In order to ensure that the eye of the viewer does not get to this point too quickly, both pictures are punctuated by visual barriers: steps, walls, doorways, and abrupt changes of direction. All this slows down the action sufficiently to allow the dénouement its full impact. In both cases the victim of aggression is in the most literal sense backed into a corner. For Yusuf, escape lies just beyond the picture: God will open the closed door that seems to deny him deliverance.39 For Arjasp, not so. The architecture becomes an instrument of destiny, progressively closing off each successive avenue of escape until death finally checkmates him. And at the moment that he is toppled from his throne (a detail not to be found in the text), the walls of his stronghold are themselves rocked — for that seems the obvious conclusion to draw from their drunken angles, a potent emblem of disorder. Essentially literary terms like metaphor, conceit, and suspense are the concepts that come most naturally to mind when one examines such compositions. As for the spatial qualities which they embody, the execution of the zigzagging forms in both paintings might technically be described as two-dimensional, but of their three-dimensional effect there can be little doubt. Thus a form of intuitive perspective, enriched by temporal and narrative content, is being employed. Another kind of perspective commonly encountered in the architecture depicted in Timurid painting is akin to axonometric projection and might be termed parallel perspective. As with true axonometric projection, the effect is initially startling, but it takes only a little while to get used to it.40 Closely linked to this dramatic way of presenting a building is the deliberate exaggeration which is so standard a feature of the architecture in Timurid 39

Lisa Golombek rightly draws attention to the descending staircase to the left, suggesting that it refers to God’s speedy succour for the faithful. And this is not all. Given the frantic gesture of Yusuf, with his arm outflung up to the left, it is hard to ignore the unusual placing of the text panel, which dips down sharply from the horizontal to meet the line of his arm. Is it too fanciful to suggest that this is Bihzad’s way of harnessing the pictorial layout to suggest that the divine help alluded to in the text is on the way, that Yusuf has only to call on the name of God to be answered? 40 Examples include the depiction of Khavarnaq in the Khamsa dated 835/1431 in the Hermitage Museum, Leningrad, VR-1000, fol. 251a (Lentz and Lowry, Timur and the Princely Vision, p. 120), and the court scene in the Bustan of Sa cdi dated 893/1488 in the Egyptian National Library, Cairo, fols. 1b-2a, especially fol. 2a — the left-hand page (Lentz and Lowry, Timur and the Princely Vision, pp. 260–61).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 533FB893 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

9

miniatures. An almost theatrical attenuation was already earlier a marked feature of actual Ilkhanid architecture, but this is not enough to explain the paramount role of exaggeration in the architecture shown in Timurid paintings. An element of visual hyperbole is, of course, entirely appropriate for buildings described hyperbolically in poetry, but it also had spatial implications. These sheer cliff-like façades are often kept designedly bare of intrusive ornament so that the sense of mass which they generate is unimpaired, though glowing bands of tilework enrich their doorways and cornices.41 They shoulder their way into the painting and impose a sense of scale and space on the scenes enacted in front of them (fig. 6). In other words, like a vault in an actual building, they make a small space seem bigger than it is.42 This is especially so when, despite their negligible breadth, they stretch to the entire height of the picture.43 Very often they occupy the left-hand margin so that, for an eye trained to move automatically from right to left, they close off the picture space. These buildings typify the reliance of the Timurid artist on solid blocks of form or color — the second category to be explored in this paper. This emphasis deserves separate study. Like “space,” “block” is a somewhat loaded word, and in particular it has its implications for the treatment of the third dimension in Timurid painting. It is impossible to glance through a typical selection of these pictures without becoming aware of the reliance placed on such blocks by one artist after another. Often these blocks are 41

A typical example is fol. 26b of the Anthology of Baysunghur dated 831/1427 in the former Berenson collection at I Tatti, Settignano (Gray, Persian Painting, p. 86). The walls of such buildings also bear low-key detailing such as brick-end plugs, which ensures that they do not strike a discordant note of plainness amidst the visual splendors which surround them. 42 Cf. the remarkably original image on p. 125 of the Anthology dated 813/1410–11, in the Gulbenkian Foundation, Lisbon (L.A. 161), where Bahram Gur contemplates all seven of the princesses at once (Gray, Persian Painting, p. 75). 43 Compare the fire ordeal of Siyavush from the Juki Shahnama (Simmonds and Digby, Royal Asiatic Society, frontispiece, or the same scene from the Epics ms. dated 800/1397–98 in the Chester Beatty Library, P. 114, fol. 14b (Robert Hillenbrand, Imperial Images in Persian Painting [Edinburgh, 1977], pl. opp. p. 53), or the gardening scene on fol. 240a of the Shahnama of ca. 1420 made for Ibrahim Sultan and now in the Bodleian Library at Oxford, Ouseley Add. 176 (Laurence Binyon, James V. S. Wilkinson and Basil Gray, Persian Miniature Painting, [Oxford, 1933], frontispiece). Sometimes these buildings are so narrow that they become mannered, as in the case of yet another version of the fire ordeal of Siyavush, in the British Library Anthology of 813/1410–11, fol. 295b (Basil W. Robinson, Persian Miniatures [n.p., 1957], pl. 3).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 484C085B /production/pindar/booksxml/hillen/chapter−01

10 buildings, and especially ones whose wide expanses of relatively plain brickwork lend them a formidable presence in the picture (fig. 7). Other blocks of this kind may be created by rectilinear or multifoil pools of water, dadoes, doors, floors, balconies, windows, shutters, or platforms. It is a prerequisite of their compositional role that they should be as monochrome as is consistent with their nature. Thus they too establish a presence in the picture, something that is much more than mere decorative infill or background. They have an obvious spatial significance. When figures are set against such blocks of color they are pushed forward into a different space. The boldness of the chromatic contrast is enough to achieve this, as can be verified by comparing a typical painting in which such devices are employed with earlier paintings which use a single-color ground or merely the color of the paper itself as a backdrop. In such cases a distinct spatial ambiguity may be noted; the figures float.44 Similarly, when figures are embroiled amidst a fully detailed backcloth, as in the Demotte Shahnama, the very proliferation of detail may prove distracting and work against a full realization of the third dimension.45 A rather different example of how the idea of a block works is revealed by the contemporary practice of deliberately copying an earlier composition. Various degrees of copying may be noted, from a 1 : 1 correspondence to the reproduction of only a single element from an earlier painting. The background to this practice has been convincingly elucidated by Adel Adamova;46 here it will be enough to focus on an intermediate stage. In two versions of a scene from the Kalila wa Dimna of the thief discovered in the bedchamber, the later artist (fig. 8a)47 copies his predecessor’s painting (fig.

44 This is particularly noticeable in Arab painting; for a typical example, see Esin Atıl, Kalila wa Dimna: Fables from a Fourteenth-Century Arabic Manuscript (Washington, D.C., 1981), p. 44. 45 As in the case of Bahram Gur killing the dragon (Survey, pl. 839). 46 In her paper found elsewhere in this volume. Cf. Norah Titley, “Persian Miniature Painting: The Repetition of Compositions during the Fifteenth Century,” in Akten des VII. Internationalen Kongresses für Iranische Kunst und Archäologie, München 7.–10. September 1976, ed. W. Kleiss (Berlin, 1979), pp. 471–91. 47 Zain, Formal Values, p. 7. This manuscript (Topkapı Sarayı Library, R. 1022), is dated 833/1429 and is therefore earlier than the other Kalila wa Dimna made for Baysunghur (Topkapı Sarayı Library, H. 362), which is dated 834/1430–31. But the paintings of H. 362 are, as Ernst Grube has cogently argued, earlier than the text they adorn; he tentatively assigns a Jalayirid date to them (“Two Kalïlah wa Dimnah Codices,” pp. 116–19,121).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4A8EE718 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

11

8b)48 very closely, except that he removes the slab of text behind the bed and replaces it with an entire and, as it were, prefabricated unit comprising the whole wall, complete with windows and medallions. He thereby significantly enriches the visual texture of the painting. Whatever the motive for this change — and among other reasons one might suggest that the more tightly spaced hand employed in the text of the later manuscript required more room to be set aside for architecture and less for text, if the basic layout of the earlier page were to be preserved unchanged — it does suggest that this artist, at least, blocked out his compositions. Thus a component could be removed and replaced by something quite different without loss to the integrity of the painting. It is only a small step from this to regarding such a picture as being composed of potentially interchangeable parts. The parallel of movable scenery in a theater springs to mind. The placing of elements that cannot so easily be treated as blocks, such as individual figures and certain types of landscape elements49 or furnishings, then becomes critical for the definition of space. Nowhere, perhaps, is this system of blocks used with greater virtuosity than in the bathhouse scene in the British Library Khamsa of Nizami dated 900 (1494–95),50 where against all expectation the upper part of the picture, usually treated in a relatively neutral manner, has stolen the limelight (fig. 9). Spatial values, brilliantly evoked by the long pole of the bath attendant, interlock with the powerful chromatic accents of the bath towels, at one level emblems of humble daily life but, compositionally speaking, colored blocks set at angles to each other in real space. With all this visual interest concentrated in the central well of space, it is no wonder that the caliph and his barber at the far left, though ostensibly the subjects of the painting, are totally upstaged. Lisa Golombek has provided a possible explanation of this apparent lèse majesté.51

48

Çagˇman, Tanındı, and Rogers, Topkapı Saray, pl. 57 (Topkapı Sarayı Library, H. 362,

fol. 24a). 49

This can be seen by comparing, as Miss Titley has done, two versions of the lover outside the beloved’s castle (ibid., figs. 1 and 2). In the later version (the Khamsa in the Topkapı Saray, H. 781, fol. 73b) the trees in the far corner of the courtyard in Humayun’s palace (British Library ms. of Khwaju Kirmani, fol. 18a) have been replaced by a slab of text with nine bayts. The effect is precisely to destroy the sense of space which permeates the earlier painting. 50 Or. 6810, fol. 27b (Gray, Persian Painting, p. 117). 51 Elsewhere in this volume.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 48999071 /production/pindar/booksxml/hillen/chapter−01

12 It is now time to investigate the third category mentioned above, namely the spatial role of the margin. This provides a further (and endlessly instructive) example of how Timurid painters improved on their predecessors rather than fashioning an entirely new way of seeing. The invasion of the margin was by no means a novel idea. In some thirteenth-century Arab paintings, for example, the artist plots his composition within an invisible frame, and when that is broken, the effect is scarcely less striking than it would have been had such a frame existed — as in the case of the banners in the famous scene of the pilgrim caravan in the Schefer manuscript of the Maqamat of al-Hariri.52 Many Ilkhanid paintings reflect an awareness that the area outside the picture frame has a certain potential for the extension of the action within that frame. Yet they also betray uncertainty as to what to do with this discovery. Hence the rather timid employment of the broken frame at this period. Admittedly the artist may score minor coups of wit or surprise. Thus in the Morgan Bestiary’s image of entwined elephants, despite the sense of tremendous powers unleashed, only the extreme tip of one tusk breaks the margin, and that by millimeters.53 Conversely, in the Rashidiyya battle scenes, furiously galloping warriors are crammed into a narrow oblong frame well down the page with a substantial slab of text above, but their lance tips and pennants pop out at the top of the page.54 The purpose of this device is not clear; the idea feels somewhat tentative. And indeed, for the most part Ilkhanid painters were content to toy with this unfamiliar concept. The dangers of using it with more boldness than discretion or understanding are vividly illustrated by a slightly later Mamluk manuscript, the Sulwan al-Mut ca, where in two images the boar transfixed between picture and margin is not so much entering the paintings as gatecrashing it (fig. 10).55 Piggish behavior indeed.

52 53

Paris, Bibliothèque Nationale, ms. arabe 5847, fol. 94b. Ettinghausen, Arab Painting, p. 134, though the branches of the tree protrude far

further. 54

Talbot-Rice, Illustrations to the World History, p. 181 top. Esin Atıl, Art of the Arab World (Washington, D.C., 1975), p. 113. The parent ms. has an almost identical scene on fol. 39b (current foliation), also involving a boar. Other examples of frame-breaking in the same ms. are much more modest: the detached leaf published by Atıl (ibid., p. 114), and fols. 31a, 41a, 44a, 82b, 86b, 90a, 97a, 100b, 104a, and 104b; cf. the facsimile of the text and the accompanying book by A. S. Melikian-Chirvani, Sulwa¯n al-Muta¯ c f¯ı cUdwa¯n al-Atba¯ c: A Rediscovered Masterpiece of Arab Literature and Painting 55

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 55A9436F /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

13

Oddly enough, the Demotte Shahnama and the Istanbul Kalila wa Dimna make rather less use of the broken margin than might have been expected after the promising beginnings which this feature had known in slightly earlier Persian painting, though in both these manuscripts the artists experimented intensively with the related concept of the framed picture as a mere segment of a wider image, a window on a much bigger world — an idea which of course has spatial implications.56 It is therefore to late Jalayirid and early Timurid painting that one must turn for a serious exploration of the device of the broken frame or the area outside the principal image. The ideas are ingenious and their sheer range is exhilarating. Repeatedly the entire relationship between text and image on the page is turned topsy-turvy. In the Kulliyat of Khwaju Kirmani of 798/1396, tiny flags of text are engulfed by the picture itself, as if a few portentous words were a magic casement opening into another and brighter world.57 Sometimes in that manuscript the frame is, so to speak, not required because the picture proper comes to an end well short of it, leaving an almost unbroken cordon sanitaire of empty and ambiguous space between image and frame.58 The ambiguity is intensified because here and there the branches of trees brush the frame. The Divan of Sultan Ahmad, datable to ca. 1400–5, achieves a precarious balance between text and image. On one page (fol. 17a) an imposing fourteen-line panel of text occupies the central space (fig. 11). But life in the (Kuwait, 1985), pp. 82, 86, and 118. Nor is the clumsiness vis-à-vis the frame encountered occasionally in the Sulwa¯n al-Muta¯ c exceptional in Mamluk painting; cf. Atıl, Kalila wa Dimna, pp. 13, 20, 22, 24–25, 27–28, and 30, though here too it alternates with only tiny transgressions of the frame (ibid., pp. 16, 19, 31, 47–48, 51, and 53). Clearly both notions co-existed quite happily. 56 For the Demotte Shahnama, cf. Grabar and Blair, Epic Images and Contemporary History, pl. 28, and Pope and Ackerman, Survey of Persian Art, pl. 837; for the Kalila wa Dimna ms., cf. Cowan, Kalila wa Dimna, pls. 1–4, 13–14, and 26. 57 Gray, Persian Painting, p. 46: Lentz and Lowry, Timur and the Princely Vision, p. 55. 58 Gray, Persian Painting, p. 47. Teresa Fitzherbert, in an unpublished paper delivered at a symposium entitled “The Iconography of Islamic Art,” held in Edinburgh in October 1990, drew attention to (among many other aspects of this fascinating picture) the panel of text which it incorporates, whose second line contains a multilayered and therefore untranslatable play on words with the central idea of “the turning world.” The marked leaning to the left in this composition accentuates the impression of whirling speed, as if the world were spinning on its axis; even the birds are caught in the vortex. The moment of truth for the lovers is the still center of this turning world. It would be hard to find a subtler integration of text and image than this.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 44362D56 /production/pindar/booksxml/hillen/chapter−01

14 margin goes on regardless, as if indeed it were the text itself that had been marginalized. An ox whose body had been bisected and overlapped by the text panel looks back inquiringly into that panel. A couple with a baby plod off the page, while above them a flock of birds flies out into the same unknown land. The right-hand side of the page is left absolutely blank, again an ambiguous touch, for out of this emptiness there suddenly materializes a busy genre scene. A further outer margin line unconnected with the text panel is superimposed on this tableau but in no way constrains it.59 In another marginal painting from the same volume, which is one half of the only double-page spread in the manuscript (fols. 22b–23a), the ideas are bolder still: people or animals are moving off the page to right and left and two mounted men, both looking backwards, exchange glances across the central text panel. A similar interchange of glances between two figures, one strategically placed at the inner margin of each page so that they eye each other across the spine of the book, serves to unite the two pictures. On the right of fol. 23a, the hindquarters of several sheep can be seen diving into the space “behind” that panel, while to the left several horses’ heads peer out from its margins.60 In this very innovative manuscript, then — in which, moreover, the illustrations so to speak ostentatiously take second place to the text — the artist has invoked several conflicting notions of reality, and it is hard to suppress the suspicion that he is in fact playing with them. The text frame of fol. 17a is not broken at all; indeed, its multiple lines emphasize its separation from the rest of the page. The other margin, by way of contrast, serves no apparent purpose; it is not so much broken as totally ignored. Once again, therefore, the direction of the Arabic script from right to left can be seen to infiltrate the conventions of painting. Sometimes the frame is not physically broken at all, but the effect is as if it had been, as in the Gulbenkian Anthology in Lisbon, dated 813/1410–11, where King Nimrud is shown on the right-hand page with his catapult while Abraham occupies the left-hand page, seated in a flowery meadow within an aureole of flame.61 The double image (fig. 12) can only be understood as a near-punning conceit, with the gutter of the book symbolizing the space

59

Lentz and Lowry, Timur and the Princely Vision, p. 56. Cf. Esin Atıl, The Brush of the Masters: Drawings from Iran and India (Washington, D.C., 1978), pp. 18–19. 60 Gray, Persian Painting, p. 49 (fol. 23a). One horse too — like the sheep opposite — is even diving back into the area behind the panel. 61 Ibid., p. 79 (p. 646 of L.A. 161).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4CB9B84A /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

15

across which Abraham has been hurled into the fire. Our space, the reader’s space, is therefore pressed into service for the painting. It is a concept familiar in Byzantine wall decoration, especially in the post-iconoclastic period.62 In the Ilkhanid version of the attempted execution of Abraham, as found in the Jami c al-Tawarikh produced a century earlier, the story is also told seriatim from right to left, but it is compressed into a single image63 and thus excludes the spatial subtlety of the Timurid rendering of this tale. The frontispiece to Baysunghur’s Kalila wa Dimna in the Topkapı Saray Library, dated 833/1429, also employs the space outside the picture without needing to break the frame, but in this case the message is different (fig. 13).64 There is no need to dilate here on the numerous devices employed in this double frontispiece to establish a visual continuity between the two pages. Suffice it to say that the artist does his utmost towards this end, using color, directional indicators, body language, landscape features, and the way the composition is blocked out. The salient fact, however, is that the gutter of the manuscript, a notionally neutral element in the physical form of the book, is made to stress the gulf between ruler and ruled. For the time being the courtiers so patiently waiting in line65 are excluded from the royal presence and its all-too-obvious privileges of food, wine, music and Lebensraum. The frame-breaking device lends itself to many different uses in Timurid painting, and virtually all of them are bound up with space. On occasion the energy and dynamism generated by the painting spill through, indeed crash through, the frame. In a manuscript of the Khamsa of Nizami in the Topkapi Saray Library dated 849/1445–46, Khusraw is just riding out of the frame when he turns round and espies Shirin in the pool, back in the picture (fig. 14).66 Here the extension of the landscape into the left margin, into a spatially ambiguous world, is entirely appropriate to the idea of the prince making his way to a distant destination. Thus the margin is used to foster the notion of

62 See the extended discussion of this topic in Otto Demus, Byzantine Mosaic Decoration (London, 1948; reprt. London 1976), pp. 22–25, 27, 30–35. 63 See the Edinburgh Jami c al-Tawarikh, fol. 3b (Talbot Rice, Illustrations to the World History, pl. 7). 64 Lentz and Lowry, Timur and the Princely Vision, pp. 110–11 (Topkapı Saray, R. 1022, fols. 1b-2a). 65 The nature of the composition encourages the belief that yet more of them are waiting out of sight. 66 Çagˇman, Tanındı, and Rogers, Topkapı Saray, pl. 59 (H. 781, fol. 40a).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 44AAAF1D /production/pindar/booksxml/hillen/chapter−01

16 two separate but complementary worlds, and the prince’s backward glance is the link between them. But in many cases the expansion into the margin is unrelated to the action, as in the image of Bahram Gur killing the dragon in the British Library Khamsa of 848/1442 (fig. 15).67 It is as if the world of the picture bulges out into the margin simply because the picture is powerless to contain it. Riotous vegetation spills out of the frame and by virtue of this untrammeled growth becomes an image of freedom, an effect particularly beloved of Turkmen painters.68 Often a sense of fun can be detected in the way that the artist plays with contrasting spaces or conjures up a world outside the painting. Surely a broad pun is intended when, as in the Kalila wa Dimna of 834/1430–31 in the Topkapı Saray Library, the painting depicts a man who has fallen into a well that itself sinks the best part of three lines down into the text69 — an idea already encountered a century earlier in a Shahnama produced in 741/1340–41 for an Inju vizier.70 In the British Library Khamsa of 1494 at least one example of breaking the frame can be assigned a meaning with somewhat greater confidence (fig. 16). Among the seventeen mourners who bewail the death of Laila’s husband, only one breaks the frame, and that is the keening figure on the rooftop.71 He it is who announces the death to a wider world — a world in which readers of the manuscript are perhaps intended to be included. In this case the unmistakable visual association with a muezzin is susceptible of still further layers of meaning.

67

Lentz and Lowry, Timur and the Princely Vision, p. 282; this picture (Add. 25900, fol. 161a) was painted ca. 1492 and may fruitfully be compared with the depiction of the identical scene on fol. 157a of Or. 6810 in the British Library (ibid.) since the later version, though largely identical, allows margin and image to merge, not collide, by making the ground color of the scene the same as that of the margin. Image and margin thus blend imperceptibly together — yet another example of how Timurid painters used color as a means of playing with space. 68 E.g., in the Khamsa of Nizami (H. 762) in the Topkapı Saray dated 880/1475–76 and 886/1481, fol. 177b (Çagˇman, Tanındı, and Rogers, Topkapı Saray, pl. 72), or in the Dastan-i Jamal u Jalal of Asafi in the Uppsala University Library, O Nova 2, fol. 43a (Hillenbrand, Imperial Images, cover plate). 69 Topkapı Saray, H. 362, fol. 27b (Çagˇman, Tanındı, and Rogers, Topkapı Saray, pl. 58). 70 Grube, World of Islam, pl. 38. 71 Or. 6810, fol. 135b (Gray, Persian Painting, p. 122).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4BCFE1CA /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

17

More commonly the roofline and the upper margin are broken by a dome, usually silhouetted against the ground color of the paper.72 This is a standard formula for the revels of Bahram Gur with one of his seven princesses — the only concession to the outside world, for attention is entirely focused on the private paradise within. In the burial scene from the Mantiq al-Tayr of 888/1483 (also dated 892/1487-88 on f.28r),73 a tree breaks the margin, escaping from the somber mortality of the main scene and celebrating — perhaps in deliberate contrast — the birds and their nest, a promise of new life even in bleak mid-winter (fig. 17). Since, however, the margin is thickly dusted with flecks of gold, the artist has chosen to envelop the tree with its own private balloon of space, a space that is of a different color from that of the sky in the main picture.74 Here again a play upon reality may be suspected. A portion of the tree — the most obviously dead part of it — is in the picture proper and is hung with pennants which, to judge by the larger pennant held by the man at the door of the funerary compound, may refer to the burial. So much, at first sight, for the connection with death. How appropriate that the upper part of the tree, with its reference to life, should exist in an unmistakably different space. Yet so rich is the visual texture of the painting that even this division is precarious, for an emissary of death from the lower part of the picture threatens the birds in their fragile refuge. A black snake wriggles its way up the trunk towards the nest. The moral is tolerably clear: in the midst of life we are in death, an iron law not only for man but for all creatures.75 In other

72 For example, the British Library Nizami, Or. 6810, fol. 37b (Lentz and Lowry, Timur and the Princely Vision, p. 283); the Cairo Bustan, fol. 52b (ibid., p. 294), and the Dublin Gulistan, fol. 9a (Gray, Persian Painting, p. 87). 73 Richard Ettinghausen et al., Islamic Art (New York, n.d.), pl. on unnumbered p. 26. 74 It is hard not to connect this device with the fashion for abri painting (Richard Ettinghausen, “Abri Painting,” in Studies in Memory of Gaston Wiet, ed. Miriam Rosen-Ayalon [Jerusalem, 1977], pp. 354 — 56, reprt. in his Islamic Art and Archaeology: Collected Papers, ed. Miriam Rosen-Ayalon [Berlin, 1984], pp. 773–89). Oddly enough, Ettinghausen did not essay a connection between the examples of abri painting which he cited and the related phenomena in Persian painting. 75 The Sufi undertones in this painting have been admirably analyzed by Rachel Milstein in her article, “Sufi Elements in Late Fifteenth-Century Herat Painting,” in Studies in Memory of Gaston Wiet, pp. 362–64. I am grateful to Bernard O’Kane for pointing out this snake to me.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 40EE0593 /production/pindar/booksxml/hillen/chapter−01

18 miniatures, too, it is quite plausible that a tree — a living thing — swaying out into the margin is intended to suggest escape from a confined world.76 Most puzzling of all are the cases like fol. 20a of the Tarikh of al-Tabari in the Chester Beatty Library, Dublin, dated 874/1469 (P. 144), where a deliberate ambiguity is set up in the scene which shows King Jamshid teaching crafts to his subjects (fig. 18).77 An extension beyond the frame — in this case a tree-top — occupies a separate space, in effect a tiny box of its own with a beige ground color quite distinct from the ultramarine sky of the main scene. It is not a continuous extension of the picture space like the tree in the Mantiq al-Tayr scene, because a slab of text has already closed off that space, nor is it simply an extra feature attached to the frame. On the other hand, it is linked by its background color to the trees of the upper left margin, which in their turn are abruptly sundered from the main scene by a thick gold vertical line. In effect, therefore, the trees at the top and the right of the painting form a separate space which obeys its own laws, neither being a straightforward extension of the picture space nor belonging simply to the margin. Several distinct spaces, several separate perceptions of reality, thus co-exist on the one page.78 Numerous other examples, each subtly different from the next, could be cited to show how Timurid painters subverted the convention of the picture frame, and indeed recast its whole shape quite radically by allowing certain elements to drift in seemingly inevitable fashion beyond its confines. And so to the fourth and final category for discussion. Spatial values in Timurid painting are most effectively brought into play by allowing ample room for emptiness. In a literal, if paradoxical, sense, space is suggested by space. Important figures generally have plenty of room in which to move. Of course there are some exceptions. It is quite true that some figures are closely

76

Thus in the scene from the Topkapı Saray Kalila wa Dimna (supra, n. 65) of the man who has fallen into the pit, the plight of the victim is underlined by the two trees which, unlike him, break free from their confines. The same goes for the camel in this picture. 77 Robinson, Persian Drawing, pl. 20. 78 For other examples of multiple spaces in one picture, see the painting of Laila and Majnun in Paradise (above, n. 25) or the scene of Bahram Gur killing the dragon in a Khamsa of Nizami dated 1444–45 in the John Rylands Library, Manchester (Ryl Pers 36, fol. 157a; Robinson, Persian Drawings, pl. 73).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 5202EE25 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

19

embowered in blossoming plants79 — indeed, in some sense they are plants80 — and seem to be on parade or participating in some tableau vivant. Yet even these are each set in a corridor of personal space, virtually insulated from unwelcome overlap. This is not to deny that overlap occurs, though it is noticeable that its effect is often modulated by locating the overlapping figure at a different level from the one that is being overlapped.81 By dint of the most careful calculation Timurid painters were often able to load their picture space with large numbers of figures — a dozen, twenty, or thirty at a time — all suitably scaled in proportion to the other details in the picture.82 Yet miraculously such pictures manage to avoid looking overcrowded. The reason is precisely because the figures are allowed room for maneuver. Each such person clearly occupies his own space. Thus there is no danger of a group of figures being lumped into an amorphous and spatially ill-defined huddle. A further advantage of this system is apparent when figures are disposed serially along a diagonal (figs. 3 and 17) or vertical (fig. 12) plane. Each successive person in such a composition, whether isolated, overlapped or overlapping, can be understood as announcing a new plane.83 When this elegant solution is compared with the unthinking repetition of the coulisse device in Ilkhanid painting, with the intersecting lines placed millimeters apart and thus making a nonsense of their original intended function as definers of plane,84 it will be seen what giant strides had been made in the redefinition of pictorial space. It is thought-provoking, to say the least, that a

79

Such as the courtly personages on a leaf from a lost ms. of (or incorporating) the mathnavis of Khwaju Kirmani in the Musée des Arts Decoratifs, Paris, Inv. 3727 (Lentz and Lowry, Timur and the Princely Vision, p. 117). 80 Cf. Schroeder, Persian Miniatures, p. 11, and idem, “The Arts of the Book,” in Masterpieces of Persian Art, ed. Arthur U. Pope (New York, 1945), pp. 147–48. 81 E.g., the frontispiece to the Cairo Bustan (Adab Farsi 908, fols. lb-2a; Lentz and Lowry, Timur and the Princely Vision, p. 260). 82 As in the scene of the construction of Khavarnaq from the British Library Khamsa, Or. 6810, fol. 154b, (Gray, Persian Painting, p. 116). In this context it is noteworthy that Rustam, though consistently described in the Shahnama as “elephant-bodied,” is not depicted as substantially larger than other warriors. 83 For a typical example, see the prisoners before Khusraw in the Gulbenkian Anthology (L.A, 161, p. 47; Gray, Persian Painting, p. 77). 84 E.g., the painting of the simurgh in the Pierpont Morgan Bestiary, (Grube, World of Islam, pl. 36), or Muhammad, Abu Bakr, and the old woman with the herd of goats in the Edinburgh Jami c al-Tawarikh, fol. 57a (Talbot Rice, Illustrations to the ‘‘World History’’, pl. 37).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4E53CE96 /production/pindar/booksxml/hillen/chapter−01

20 similar process was taking place more or less contemporaneously in Western European painting. The new vision of space developed by the Van Eycks and Masaccio merely uses a different visual language to express much the same interests. It is as well to remember, too, the very different approach to spatial problems that is to be found in Yuan painting, to which Persian painters had increasing access in the early fifteenth century through the embassies exchanged between China and the Iranian world.85 Whether the new interest in spatial problems evinced by early Timurid painters was influenced by specific cross-cultural contacts or whether, in a looser sense, such ideas were merely in the wind at that time is a topic for future investigation.86 Most common of all, perhaps, are scenes in which the figures are caught in mid-activity, silhouetted in some distinctive pose with clear space all around them. The viewer is invited to move on to the next frame in his imagination and to complete the actions thus temporarily arrested. Perhaps this is why so many scenes of execution in paintings of the period depict the moment just before violent action.87 In scenes of murder or actual death, on the other hand, the moment that the blow is struck — the moment of maximum drama — is the one selected.88 The slaying of Arjasp in the Brazen Hold as rendered in the Juki Shahnama is one such example (fig. 3).89 The two protagonists are isolated in the center of the castle courtyard. Its startling white ground, in strong contrast to the tawny brick walls encircling it, allow it to function as an almost empty stage, with the spotlight trained on Isfandiyar and Arjasp. That the dramatic potential of this conflict is so fully realized

85

Cf., for example, A Persian Embassy to China. Being an Extract from the Zubdatu’lTavarikh of Hafizi Abru, trans. K. M. Maitra (Lahore, 1934; reprt. New York, 1970), and the next note. 86 For the subject of these cross-cultural relations at large, see Toh Sugimura, The Encounter of Persia with China: Research into Cultural Contacts Based on Fifteenth-Century Persian Pictorial Materials, Senri Ethnological Studies no. 18 (Osaka, 1986); for briefer guides, see Çagˇman, Tanındı, and Rogers, Topkapı Saray Museum, pp. 114–20, 153–56, and Lentz and Lowry, Timur and the Princely Vision, pp. 182–95. 87 Typical examples are the Div Akwan about to throw Rustam into the sea (from the Juki Shahnama, fol. 185b; Gray, Persian Painting, p. 90) or the Christian monk falling from the rooftop in the presence of cAli in the British Library Anthology, Add. 27261, fol. 305b (Glyn M. Meredith-Owens, Persian Illustrated Manuscripts [London, 1965], pl. 3). 88 E.g., Gustaham beheading Farshidward (Juki Shahnama, fol. 206b, Gray, Persian Painting, p. 91). 89 Fol. 278a (Lentz and Lowry, Timur and the Princely Vision, p. 134): “Childe Roland to the Dark Tower came.”

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4F5898F3 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

21

owes much to the imaginative use of empty space, itself heightened by color. Similar techniques are often employed in scenes of Rustam killing Suhrab.90 In the case of more crowded scenes, this concept of an action shortly to be completed naturally implies a three-dimensional world. A somewhat different way of approaching this notion of emptiness is to consider it in terms of interval. Although in architecture that concept is used principally to determine the balance of solids and voids, in Timurid painting it has a much wider application. It can cover the placing of complementary or contrasting colors; or the size of the color blocks themselves; or the shape, size, and location of text or inscription panels; or the manipulation of people, trees, furnishings, and so on as abstract elements of composition. Moreover, as in an actual building, this awareness of interval, this sensitivity to its nuances, does not necessarily imply regularity. In certain Ilkhanid paintings, such as those of the Rashidiyya manuscripts, the disposition of figures often follows a beat as regular as a metronome.91 The rhythms of imperial Timurid painting are much subtler than this. Space is denied to some areas only to be lavished on others. The contracting and expanding rhythms of design may owe something to the discipline of calligraphy, in which some painters92 were well versed, or even (one may suspect) to music. Mere onlookers to the action often overlap, though the group they form will itself tend to be isolated, so that it can function visually as a block, while those who really count towards the story or scene have plenty of room to move.93 Such gradations in the amount of space allotted to people, and for that matter to furniture or to certain natural forms, can have wider ramifications than mere design. They can reflect narrative concerns or 90 E.g. in the Juki Shahnama: cf. Basil W. Robinson, ‘‘Unpublished Paintings from a XVth century‘‘ Book of Kings’’, Apollo Miscellany (June, 1951), 18, fig.II (repr. in Basil W. Robinson, Studies in Persian Art. Vol. II [London, 1993], 159). 91 Cf. the Edinburgh Jami c al-Tawarikh, fols. 16b, 34a, and 138a (Talbot Rice, Illustrations to the ‘‘World History’’, pls. 20, 27, and 65). 92 Such as al-Wasiti, who was responsible for the Schefer Hariri, or al-Husain b. cAbd al-Rahman b. cUmar b. Muhammad al-Sufi, who produced the Kitab Suwar al-Kawakib al-Thabita in the Bodleian Library, Oxford, Marsh 144 (Melikian-Chirvani, A Rediscovered Masterpiece, p. 171). It remains questionable, however, given the degree of specialization in Timurid ateliers, whether this practice was followed to any significant extent in the mss. dealt with in this article, though cases do exist, e.g., Aqa Mirak (Clément Huart, Les calligraphes et les miniaturistes de l’Orient musulman [Paris, 1908, reprt., Osnabrück, 1972], p. 331). 93 This principle is well illustrated in the scene of the wrestling match on fol. 14a of the Dublin Gulistan (Gray, Arts of the Book, pl. 46).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 481666FF /production/pindar/booksxml/hillen/chapter−01

22 social hierarchies. So can degrees of centrality or the interplay of various levels along the vertical axis. In all of these aspects the notion of interval, with a corresponding spatial awareness, is of critical importance. Of course not all of these ideas are new, but their conscious manipulation on this scale is unprecedented. The various components of the composition naturally do not function entirely on this rarified and somewhat abstract plane. They also have their proper and obvious function as people, as trees, and so on. But there can be no doubt that they operate in two distinct modes. Moreover, these two modes affect each other, though they are not in conflict. The spacing of the columns on the façade of a classical Greek temple can be sensed instantaneously as “just right” and one may safely leave it to the specialists to justify that intuition scientifically by carrying out a battery of measurements. The placing of elements in many an early Timurid painting seems to be as precisely calibrated as the intercolumniations of the Parthenon. The aesthetic satisfaction to be derived from such apparently inevitable placing does not impair, but rather enhances, the narrative or ceremonial purpose of a specific painting, to say nothing of that painting’s wider ramifications as a luxury object in a royal book. Such reflections again take us far beyond the spatial values which operate in such paintings, further evidence, if any were needed, that these pictures were most thoroughly thought out, durchkomponiert. Thus each constituent element — in this case, even empty space or interval — had its part to play. Timurid paintings do not wear their hearts on their sleeves; their secrets must be prised out of them one by one. One last use of empty space is worth noting briefly, not least because it has a direct association with a Persian literary topos, the motif of the winter night.94 In studying early Timurid painting one is repeatedly conscious of looking in, almost eavesdropping, on an extraordinarily privileged, almost transfigured, world which has successfully shut out real life with its bustle, noise, disorder, privation, and squalor.95 This is surely no accident. Exclusion is part of the intent. It underlines the gulf between the haves and the

94

I am grateful to Professor Christoph Bürgel for this observation. In the scene from the Baysunghur Shahnama where Gulnar espies Ardashir one may note a staggering of spaces, perhaps arranged with narrative intent, in that the prince is enclosed within a fence (and thus privileged), but is still separated from the maiden high above him by a closed and guarded door (Gray, Album, pl. 31). 95

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 22

JOBNAME: ???? PAGE: 23 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 510B572D /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

23

have-nots. Hence the very visible presence of disappointed suitors,96 suppliants,97 beggars,98 and other unfortunates. The plight of the underprivileged is expressed in multiple ways. Gestures are a favorite method — a pleading upturned gaze,99 arms wrapped around a shivering torso,100 hands frantically extended to ward off attack101 or beseeching pity.102 Colors can serve to tell the same story: a glacial pale green, often set off by a cool violet, surrounds the various victims in one painting after another in the Chester Beatty manuscript of Sa cdi’s Gulistan, and is in poignant contrast to the hot vermilion used for the palatial buildings from which they are excluded. It is hard not to believe that, for this artist at least, the cool colors like pale green and mauve connoted distress, while the bright orange-reds suggested well-being. And of course spatial means — above all, empty space — are also employed to reinforce these subtle distinctions. Two further images from the Dublin manuscript of the Gulistan, both of them seascapes, are especially relevant in this context. In one painting, a drowning man clutches desperately on to his rescuer, both of them isolated in a vast expanse of water, while the passengers on a boat that is just too far away reach out unavailing hands (fig. 19). In the other, a poor wretch, marooned on top of a high pale-green tower that rises sheer out of the

96 To take one example among many, cf. Khusraw at the palace of Shirin in the Nizami ms. of ca. 1405 — 10 in the Freer Gallery of Art, Washington, D.C., no. 31.36 (Gray, ed., Arts of the Book, pl. 33). Even if this disappointment is only temporary, the fact of exclusion remains. 97 E.g., Khusraw before his father Hurmuzd in a Khamsa of Nizami dated 868/ 1463–64 in the Chester Beatty Library, Dublin (P. 137, fol. 31a; see Hillenbrand, Imperial Images, pl. opposite p. 35) or the same scene in the British Library Khamsa, Or. 6810, fol. 37b (Lentz and Lowry, Timur and the Princely Vision, p. 283). 98 Such as the dervish (formerly a vizier) in the Dublin Gulistan, fol. 9a (Gray, Persian Painting, p. 87). 99 As in the love scene from Baysunghur’s Anthology at Settignano, fol. 27b (Gray, Persian Painting, p. 86) and the numerous versions of Shaykh Sancan and the Christian maiden. 100 See above, n. 98. 101 As in the case of the distressed poet in the snow attacked by dogs (Dublin Gulistan, fol. 32b: Basil W. Robinson, Persian Miniature Painting from Collections in the British Isles [London, 1967], pl. 10). 102 E.g., Bizhan in the pit in the Juki Shahnama (Wilkinson and Binyon, Shah-nama of Firdausi, pl. 11). The same gesture can denote expostulation, as in the scene of Dara and the herdsman on fol. 20a of the Cairo Bustan (Lentz and Lowry, Timur and the Princely Vision, p. 293).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 23

JOBNAME: ???? PAGE: 24 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 4343EBF4 /production/pindar/booksxml/hillen/chapter−01

24 ocean, gestures in despair towards a boatload of people that draws slowly away from him (fig. 7).103 Some stretch out their arms to him, and as in the companion picture it is the unbridgeable gulf between the victim and his fellow-men that is underlined by these futile gesticulations. Space and color are therefore used in tandem to suggest a threatening mass. In their original state, when the pitiless sea would have been rendered as a sheet of glittering silver, the effect of these two pictures would have been much more subtle than it seems now, when oxydization has destroyed the balance of colors. A single example must suffice to show how all these methods worked smoothly in concert. In the Dublin Gulistan a deceptively simple painting illustrates a brief text describing how the poet, wandering in the noonday heat through the alleys of a town, chanced to encounter a maiden who appeared as if by magic from a doorway and offered him a much-needed drink of iced water (fig. 6). Sa cdi then intimates gracefully and in verse that the sight of her has afflicted him with a profounder thirst which no mere drink could assuage. The lady’s reply is not recorded, though no doubt she too quoted verse at him in abundance.104 This brief and ambiguous encounter is the point of departure for the artist, who builds on it his own commentary, full of wit and sly humor. In a telling reversal of the setting described in the text, with its reference to the poet sheltering against the heat in the shadow of a wall, he sets Sa cdi against a pale-green ground, while the girl comes out of an orange-red building which seems almost to glow with heat. These are the two dominant colors of the painting and can perhaps be seen as appropriate equivalents for the two major themes of passion and rejection. But why are they, so to speak, reversed? Closer examination suggests that the painter is playing a kind of cat-and-mouse game in his choice of colors. The warm-hued building has a marble dado of cool violet and the wall above and around the maiden is spangled with pale-green stars. The doorway in which she stands, silhouetted against a pure white ground, is framed in tilework of blue and green. She herself is clad, promisingly enough, in orange; but that is her outer garb only, and beneath it can be glimpsed a pale green shift. The iced beverage which she offers the poet is

103

Cf. Arthur J. Arberry, Edgar Blochet, Mujtaba Minovi, James V. S. Wilkinson, and Basil W. Robinson, The Chester Beatty Library: A Catalogue of the Persian Manuscripts and Miniatures, vol. 1 (Dublin, 1959). 104 Sadi: Gulistan or Flower-garden: Translated, with an Essay by James Ross (London, n.d.), pp. 221–22.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 24

JOBNAME: ???? PAGE: 25 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 51F8EF85 /production/pindar/booksxml/hillen/chapter−01

THE USES OF SPACE IN TIMURID PAINTING

25

contained in a bowl of the same jade-green color. His outer robe, in contrast, is blue — the color of renunciation — and is echoed, as just noted, in the doorframe. The message of this subtle interplay of colors seems to be that hopes are raised only to be dashed. The body language of the two figures bears this out. The setting conjures up an ample space, yet this contains only two diminutive figures. In one sense it dwarfs them; but in another, it ensures that attention is focused on them. Their hands almost touch — but not quite. She stays just behind the mauve threshold, which thus seems to bar his entry. She proffers the drink, but he has not yet received it. Thus gesture reinforces the message of color. But it is the contrast between their surroundings that is most explicit: unbroken emptiness for him, but for her a blossoming garden against a golden sky. Thus in his own domain the artist goes decisively beyond his literary source, and does so by means of a visual language every whit as allusive as the text itself. To summarize, then, the paradise garden, the hortus conclusus with all its connotations of bliss, privilege, luxury, the beloved, the palace and its good cheer — all of these are out of reach. They are behind fences, balconies, shutters, and protected by high walls105 and by closed106 or guarded107 doors. Much the same is true of the audience hall where the monarch sits enthroned.108 Thus spatial values are made to carry an emotional charge, whether of longing or loneliness, frustration or fear.

105 E.g., the scene from Baysunghur’s Shahnama where Zal and Rudaba embrace. Here the high wall in the foreground creates an impassable barrier between the lovers and the outside world (Gray, Album, pl. 18). 106 As in the scene from the Shahnama of Ibrahim Sultan (Oxford, Bodleian Library, Ouseley Add. 176, fol. 240a), which depicts a toiling gardener watched from above by a quintet of elegantly unemployed ladies (Binyon, Wilkinson and Gray, Persian Miniature Painting, frontispiece). 107 E.g., the love scene in the Anthology of 810/1407 (Topkapı Saray, H. 796, fol. 34b), where the hajib, a well-dressed medieval bouncer often shown wielding his stick, is indeed a key element in creating this notion of exclusiveness (Gray, ed., Arts of the Book, pl.xxxv). For further examples, see the Shahnama frontispiece of ca. 1444 in the Cleveland Museum of Art (Gray, Persian Painting, p. 102), and the British Library Khamsa, Or. 6810, fol. 214a (Gray, ed., Arts of the Book, pl. 62). 108 Thus in the scene of Luhrasp enthroned from the Shahnama of Baysunghur (Survey, pl. 870), the less favored courtiers sit out in the cold, totally cut off from the action by the outer wall of the chamber, while in the frontispiece to the Cairo Bustan (Adab Farsi 908, fols. 1b-2a) the only person to break the frame is the kneeling figure falling backwards out of the picture as he tries to ward off the blows of the hajib.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 25

JOBNAME: ???? PAGE: 26 SESS: 43 OUTPUT: Fri Feb 10 09:57:05 2012 SUM: 48849BA2 /production/pindar/booksxml/hillen/chapter−01

26 This paper has discussed only four of the many spatial devices wielded by Timurid painters: architecture; the use of blocks; the margin; and emptiness. They suffice, however, to show the total overhaul of spatial values in Persian painting which Timurid artists completed, though such values are often enough of only secondary importance in any given painting.109 Gone is the obsessive and rapidly meaningless piling up of intersecting linear planes so dear to many Ilkhanid painters. Gone too are the other overly literal spatial devices of that school, such as the coulisse system or the gnarled tree root which straddles two or more planes. So firm is the artist’s grasp of interval that a painting can be full of people and yet not seem overcrowded. Narrative or ceremonial scenes are given a credible three-dimensional environment by a few carefully chosen props (buildings, trees, furnishings), all of them placed with equal care, and also by the painter’s confident manipulation of planar contrasts: upper versus lower ground plane, interior versus exterior, foreground versus background. Even the most unlikely elements of a design contribute to creating a sense of space; the way that the paving of a floor is laid, or the direction of a water conduit that marks the central axis, or a tree of artfully random and uneven growth which nevertheless helps to parcel out the space. It is symptomatic of this total control of the conventions being employed that even logically conflicting planes cannot mar the spacious serenity of mature Timurid miniatures, as in the scene of Laila and Majnun at school from the British Library Khamsa of 1494, where the enclosure around the tree is shown in adequate isometric perspective at an angle of about 45° while the octagonal pool right beside it is presented at an angle of 90° (fig. 20).110 No doubt the artist, if challenged in these pettifogging literal terms, would have retorted that in both cases he had selected the viewpoint that would reveal the nature of the object with maximum clarity. This same tree,

109 In certain images from the Kalila wa Dimna in the Topkapı Sarayı Library (R. 1022), dated 833/1429, for example, space is suggested only technically or marginally, by means of overlap; the artist’s conception of the main theme is divorced from any sustained exploration of spatial values (e.g., the lion killing the bull [fol. 46b., Gray, Persian Painting, p. 84], or the owls and the crow [fol. 66a, ibid., title page]). 110 Or. 6810, fol. 106b: Gray Persian Painting, p. 123. Compare, too, the scene of the peeping Tom on fol. 47b of the Khamsa of Nizami datable to 1426 in the Metropolitan Museum of Art, New York (Ettinghausen et al., Islamic Art, pl. on unnumbered p. 25) and its close copy in the British Library Khamsa, Or. 6810, fol. 190a (Lentz and Lowry, Timur and the Princely Vision, p. 275).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01 Sequential 26

JOBNAME: ???? PAGE: 1 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0F0D4CC2 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

27

1. Tahmina enters Rustam’s chamber. Shahnama of Muhammad Juki, ca. 1440. London, Royal Asiatic Society, Morely 239, fol. 56b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 05E48998 /production/pindar/booksxml/hillen/chapter−01−artwork

28

2. Humay taken captive. Mathnavı¯s of Khwaju Kirmani, 831 (1427–28). Vienna, Nationalbibliothek, N.F. 382, fol. 53a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 087A3EB1 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

29

3. Isfandiyar kills Arjasp in the Brazen Hold. Shahnama of Muhammad Juki, ca. 1440. London, Royal Asiatic Society, Morley 239, fol. 278a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 06F92E75 /production/pindar/booksxml/hillen/chapter−01−artwork

30

4. Yusuf flees from Zulaykha. Sa cdi, Bustan, 893 (1488). Cairo, General Egyptian Book Organisation, Adab Farsi 908, fol. 52b Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 07D84E57 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

31

5. Isfandiyar kills Arjasp in the Brazen Hold. Shahnama of Baysunghur, 833 (1429). Tehran, Gulistan Palace Museum.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 062C555D /production/pindar/booksxml/hillen/chapter−01−artwork

32

6. The poet and the maiden. Sa cdi, Gulistan, 830 (1426). Dublin, Chester Beatty Library, P. 119, fol. 38b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 08360FB9 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

33

7. A boat and a tower. Sa cdi, Gulistan, 830 (1426). Dublin, Chester Beatty Library, P. 119, fol. 29b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 7

8a. The thief discovered in the bedchamber. Nasr Allah Abu’l-Ma cali, Kalila wa Dimna, 833 (1429). Istanbul, Topkapı Sarayı Library, R. 1022, fol. 24a. Schema of design. (After Zain).

8b. The thief discovered in the bedchamber. Nasr allah Abu’l-Ma cali, Kalila wa Dimna, 834 (1430–31); the paintings are earlier. Istanbul, Topkapı Sarayı Library, H. 362, fol. 24a. Schema of design. (After Zain).

JOBNAME: ???? PAGE: 8 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 1215DD5B /production/pindar/booksxml/hillen/chapter−01−artwork

34

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 08C716D3 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

35

9. Bathhouse scene. Nizami, Khamsa, 900 (1494–95), London, British Library. Or. 6810, fol. 27b. Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 07CDD215 /production/pindar/booksxml/hillen/chapter−01−artwork

36

10. The horse and the boar. Ibn Zafar, Sulwan al-Mut ca, 14th century. Washington, D.C., Freer Gallery of Art, 54.1.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 076481D4 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

37

11. The Valley of the Quest. Divan of Sultan Ahmad, ca. 1400–5, Washington, D.C., Freer Gallery of Art 32.30, fol. 17a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 06A5D403 /production/pindar/booksxml/hillen/chapter−01−artwork

38

12a. King Nimrud casts Abraham into the fire. Anthology of 813 (1410). Lisbon, Gulbenkian Museum, L.A. 161, fol. 327a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 07A17291 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

39

12b. King Nimrud casts Abraham into the fire. Anthology of 813 (1410). Lisbon, Gulbenkian Museum, L.A. 161, fol. 326b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0B654C39 /production/pindar/booksxml/hillen/chapter−01−artwork

40

13a. Frontispiece. Nasr Allah Abu’l-Ma cali, Kalila wa Dimna, 833 (1429). Istanbul, Topkapı Sarayı Library, R. 1022, fol. 2a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0829CD8E /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

41

13b. Frontispiece. Nasr Allah Abu’l-Ma cali, Kalila wa Dimna, 833 (1429). Istanbul, Topkapı Sarayı Library, R. 1022, fol. 1b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0675635A /production/pindar/booksxml/hillen/chapter−01−artwork

42

14. Khusraw espies Shirin. Nizami, Khamsa, 849 (1445–46). Istanbul, Topkapı Sarayı Library, H. 781, fol. 40a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0E7D6AA2 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

43

15. Bahram Gur Kills the dragon. Nizami, Khamsa, 846 (1442); this painting probably dates to ca. 1492. London, British Library, Add. 25900, fol. 161a. Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 066604AA /production/pindar/booksxml/hillen/chapter−01−artwork

44

16. Mourning scene. Nizami, Khamsa, 900 (1494–95). London, British Library, Or. 6810, fol. 135b. Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 087122E7 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

45

17. Scene in a cemetery. cAttar, Mantiq al-Tayr, 888 (1483). New York, Metropolitan Museum of Art, 63.210.35, fol. 35a. Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 05C8F587 /production/pindar/booksxml/hillen/chapter−01−artwork

46

18. Jamshid teaching the crafts. Al-Tabari, Tarikh (trans. into Persian). Dublin, Chester Beatty Library, P. 144, fol. 20a.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 07BFE5E5 /production/pindar/booksxml/hillen/chapter−01−artwork

THE USES OF SPACE IN TIMURID PAINTING

47

19. The rescue of a drowning man. Sa cdi, Gulistan, 830 (1426). Dublin, Chester Beatty Library, P. 119, fol. 15b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 30 OUTPUT: Tue Aug 10 10:23:06 2010 SUM: 0660849C /production/pindar/booksxml/hillen/chapter−01−artwork

48

20. Laila and Majnun at school. Nizami, Khamsa, 900 (1494–95) London, British Library, Or. 6810, fol. 106b. Schema of design. (After Zain).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-artwork Sequential 22

JOBNAME: ???? PAGE: 1 SESS: 43 OUTPUT: Tue Aug 10 10:24:08 2010 SUM: 4EFD07A5 /production/pindar/booksxml/hillen/chapter−01−ends

THE USES OF SPACE IN TIMURID PAINTING

49

incidentally, illustrates how a deliberate contrast may be set up between natural forms, which are treated with remarkable freedom and liveliness, and man-made forms characterized by an equally remarkable rigidity. The result is to evoke two adjoining but quite distinct environments — distinct even in their spatial concepts. The spatial implications of color still await a detailed examination.111 Complementary colors, repeated accents of identical colors in different places, or pairs of colors in rhythmic alternation, could reinforce spatial distinctions even if they could not quite create them, as shown by the widely accepted convention that color could be used anti-naturalistically (for example, orange mountains or trees with blue leaves) in obedience to laws of internal chromatic harmony which are still very imperfectly understood by modern scholarship. Sometimes, as in the tree with autumn leaves in the school scene just mentioned, a clutter of colors suggests three-dimensional space very successfully, but that particular illusion does not extend to the rest of the painting, though perhaps this illogicality would not have struck the painter himself. Gesture too plays its part, though its expressions are deliberately subdued. Heads inclined towards each other, eye contact established from one side of the page to the other, or between figures at windows and those at ground level, a slight twisting of the torso, an outstretched hand or arm, all of these can be seen to function as links in a compositional chain that has spatial implications.112 Majnun wringing his hands forlornly on the skyline as the rival clans battle it out in the foreground shows how the eloquence of mime can be harnessed to serve spatial ends.113 111

Cf. some acute remarks on this topic by Schroeder, Persian Miniatures, p. 53. Typical examples include the painting on silk depicting the encounter between Humay and Humayun (Survey, pl 878), the king and the beggar in the Mantiq al-Tayr ms. in the Metropolitan Museum of Art, New York, dated 888/1483 and 892/1487–88 (Bulletin of the Metropolitan Museum of Art, Autumn 1978, p. 22) and Humay at the court of the emperor of China from the British Library ms. of the Kulliyat of Khwaju Kirmani, fol. 12a (Gray, Arts of the Book, p. 65). More generally, one might note the tendency in such late Timurid manuscripts as the British Library Nizami, Or. 6810, for people to coalesce into pairs and groups. 113 This is from a Khamsa of Nizami in the British Library (Add. 25900, fol. 121b) and is datable to ca. 1492–93 (Ralph H. Pinder-Wilson, Persian Painting of the Fifteenth Century [London, 1958], p. 16 and pl. 7). It is instructive to note how this picture reworks and intensifies the ideas already present in nuce in the earlier depiction of the same subject in the Topkapı Saray Khamsa dated 866/1461, fol. 115a (Gray, ed., Arts of the Book, pl. 66). In the 112

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 43 OUTPUT: Tue Aug 10 10:24:08 2010 SUM: 30BF05CD /production/pindar/booksxml/hillen/chapter−01−ends

50 Given that everything about early Timurid painting is designed to cultivate a heightened sensibility to nuance, the deliberately coded spatial settings analyzed in this paper belong with the arrested poses, the intense palette, and the other hallmarks of the style. Space — like everything else in these paintings — has been transfigured and operates as a quasi-Platonic ideal. This paper has tackled only a few of its mysteries; more await the microscope. But the code can be broken.

Author’s note: Given the importance of color in the paintings under discussion here, the references to published reproductions which are given in the following footnotes will be to color plates wherever possible. Whenever possible, too, I have used the simplified schemata of Timurid paintings published by D’zul Haimi Zain, to whom I must express my warm gratitude for allowing me to do so. Of course such schemata can offer only a travesty of what the paintings look like, but for the topics discussed in this paper, they help to concentrate the mind and perhaps reveal some of the artist’s own thought processes. Finally, the illustrations have been carefully selected not only to complement passages in my text that refer specifically to them, but also with the more general intention of providing supplementary visual material for all four of the topics examined here.

later version the number of warriors has been pruned from fourteen to nine and the horizon is much lower, allowing Majnun to be silhouetted to much greater effect against the gold sky. In the earlier picture the large empty expanses of gray ground serve no such purpose. The tree beside him in the earlier version has also been removed. All these changes, obvious enough in themselves, have resulted in a much more poignant version of this tragic encounter.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-01-ends Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 48EBE0D5 /production/pindar/booksxml/hillen/chapter−02

II The Iconography of the Sha¯h-na¯ma-yi Sha¯hı# HE Sha¯h-na¯ma of Shah Tahmasb is full of puzzles. Much has been written about it,1 but most of this writing has been devoted to determining the identity of the artists responsible for each of its 258 paintings, a task of Herculean proportions given that only two of them are signed2 and that only one other bears a written attribution.3 That particular lode has been mined for all it is worth and has now been exhausted. But others remain to be explored.4 Just as its celebrated predecessor, the Great

T

1 Notably M.B. Dickson and S.C.Welch, The Houghton Shahnameh, 2 vols (Cambridge, Mass. and London 1981). See also S.C. Welch, A King’s Book of Kings. The Shah-Nameh of Shah Tahmasp (London 1972) [hereafter, KBK]; idem, Royal Persian Manuscripts (London 1976) [hereafter RPM], pp. 34–53; idem, Wonders of the Age. Masterpieces of Early Safavid Painting, 1501–1576 (Cambridge, Mass. 1979), pp. 38–117. 2 One is fol. 60vo (Mir-i Musavvir). See Dickson and Welch, II, pl. 47. The other is on fol. 105ro, which bears the signature of Mas cud Ahmad (J. Sellyer, Scribal notes on Mughal manuscript illustrations, Artibus Asiae 48/iii-iv [1987], p. 251, n.24. I am grateful to Dr. Sheila Blair for this reference). 3 Fol. 521 vo; signed by Dust Muhammad. For a colour plate, see Welch, KBK, p.173. For a detailed discussion of the confusion of identities between two artists named Dust Muhammad in the early Safavid period, see C. Adle, Autopsia, in absentia. Sur la date de l’Introduction et de la constitution de l’Album de Bahra¯m Mirza¯ par Dust-Mohammad en 951/1544, Studia Iranica 19/i (1990), 243–8 and idem, Les artistes nommés Dust-Mohammad au XVIe siècle, Studia Iranica 22/ii (1993), 219–96. Dr Adle’s view, like that of Dickson and Welch, is that the “signature” of Dust Muhammad here is an attribution (ibid., p. 269). 4 Chief among these, perhaps, is its impact on subsequent painting. The received opinion is that a spiritual crisis undergone by Shah Tahmasb around 952/1545 (the date may be reconstructed from an anecdote about Maulana Nizam al-Din Shah Mahmud Zarin-qalam Nishapuri in the Gulista¯n-i hunar of Qazi Ahmad – see the translation by V. Minorsky, Calligraphers and painters. A Treatise by Qa¯dı son of Mı#r-Munshı# (circa . # Ahmad, .

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 56A0E1F0 /production/pindar/booksxml/hillen/chapter−02

52 Mongol Sha¯h-na¯ma, generally and paradoxically called the Demotte Sha¯h-na¯ma after the man who cut it up for profit (admittedly in the Days of Ignorance before the First World War),5 so this manuscript is perversely known as the Houghton Sha¯h-na¯ma after the millionaire who presided over its dismemberment, and that in the more enlightened 1970s. Yet, as Chahriyar Adle has shown, the manuscript had an entirely appropriate name in its own time, Sha¯h-na¯ma-yi Sha¯hı# or Sha¯h-na¯ma-yi Sha¯h T.ahma¯sbı#.6 Unfortunately no facsimile was made of the book before it was broken up, so anyone working on the manuscript must do so through the massive two-volume publication of Dickson and Welch. Happily that does present all the illustrated pages at their actual size, though the almost entirely monochrome plates, for all their undoubted technical quality, can be no more than a pale reflection of the breathtaking, colourful splendour of the paintings themselves. The incontrovertible facts about the manuscript, unlike the all-toocontrovertible suppositions, can be quickly summarised. Its text of 759 folios, with a page size of some 30 by 45 centimetres (12 by 18 inches) lacks

A.H.1015/A.D.1606) [Washington 1959], pp. 135–6) made him turn his face against calligraphy and painting. There is no need to question the accuracy of this view, but it encourages a patron-dominated perspective on events, and that is not the only way of interpreting them. There is also the view of the painters themselves to take into account. An obvious contemporary parallel is the fate of Italian painting in the wake of the achievements of Leonardo, Raphael and Michelangelo. High Renaissance gave way to Mannerism, and one explanation for that process has been a failure of nerve and a consequent reluctance on the part of Italian artists working from the 1530s onwards to compete with the giants of the recent past. It is at least tenable to propose that something similar occurred in Iran in the 1540s. True, Shah Tahmasb’s change of heart towards painting provided an obvious catalyst, but those painters working on his Sha¯h-na¯ma can scarcely have failed to realise that they had reached a ne plus ultra in manuscript painting. One must reckon, in the aftermath of that colossal achievement, with overkill, with a certain fatigue, with a sense that the resources on which creative artists draw had been exhausted. Certainly the challenge posed by the Sha¯h-na¯ma-yi Sha¯hı# closed more doors than it opened for the next generation of painters. And its fame within the fraternity of painters would have been such as to make it an achievement impossible to ignore. 5 O. Grabar and S. Blair, Epic Images and Contemporary History. The Illustrations of the Great Mongol Shahnama (Chicago and London 1980). 6 C. Adle, review of Welch, Wonders of the Age, Abstracta Iranica 2 (1979), p.172. Cf. idem, “Autopsia”, pp. 236–7. The name Sha¯h-na¯ma-yi Sha¯h .Tahma¯sbı# is emblazoned in Persian on the cover of Welch, KBK. For a brief account of how the manuscript was broken up, see E. Munro, How to mangle a masterpiece. The sad story of the Houghton Shahnameh, Saturday Review (27 October 1979), 21–6.I am grateful to Dr. Eleanor Sims for this reference.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 572E6240 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

53

the customary colophon with the date or the name of the scribe,7 but an introductory rosette8 contains the name of Shah Tahmasb, and his name is repeated in an architectural inscription on folio 442 verso,9 while the only date in the book, 934/1527–8, is to be found on an architectural panel on folio 516 verso.10 The artist Mir Musavvir signed the painting on folio 60 verso and Mas cud Ahmad that on folio 105 recto, while folio 521 verso bears an attribution (not a signature) naming Dust Muhammad.11 And a clinching cross-reference in the celebrated potted history of Persian painting prefixed to the album of Bahram Mirza, dated 951/1544 and written by another Dust Muhammad, a calligrapher this time,12 says of Ustad Nizam al-Din Sultan Muhammad, “the zenith of the age”, that “Among the paintings by him in a Sha¯h-na¯ma of the Shah is an illustration of people clothed in leopard skins, such that the hearts of the boldest of painters were grieved and they hung their heads in shame before it.”13 This must surely be folio 20 verso of our manuscript, a masterpiece beyond compare in the book (fig.1).14 7

This may of course have been removed when the manuscript left the Ottoman royal collection, at some time between 1801 and 1903. It is hard to believe that a manuscript so finished in other respects lacked a colophon. 8 This too is an unusual feature, for it occurs not where one would expect it, i.e. at the very beginning of the book, but at fol. 16ro after a good deal of introductory material on the career of Firdausi. And yet that material in turn is preceded by a double-page geometrical frontispiece (fols 2vo – 3ro) which could easily have incorporated such a dedication. Finally, the absence of a single- or double-page figural frontispiece is noteworthy, especially in view of the popularity of such frontispieces during the previous century, and of the unprecedentedly lavish illustration of this particular Sha¯h-na¯ma. Perhaps a single explanation could account both for this omission and for the missing colophon – that these two elements, traditionally the last to be completed in an illustrated book, were never part of this book because it was never formally finished, for all that it was lavishly bound. 9 For a discussion of this, see Dickson and Welch, I, p. 542. 10 Ibid. 11 Ibid., I, pp. 87 and 118 respectively. 12 For a full elucidation of the careers of the two men who bore this name, see Adle, “Les artistes”. 13 L. Binyon, J.V.S. Wilkinson and B. Gray, Persian Miniature Painting (Oxford 1933), p. 186. For a different and somewhat fuller translation, see W.M. Thackston, A Century of Princes. Sources on Timurid History and Art (Cambridge, Mass. 1989), p. 348. 14 In the absence of a formal frontispiece, this, the first illustration to the Sha¯h-na¯ma text proper, could perhaps itself be interpreted as a kind of frontispiece. It accords the place of honour to Sultan Muhammad, then the doyen of painters in the royal atelier, and could even at the same time be intended as an act of homage to the Shah himself, the Gayumars of his rime. Whether this was Ismacil or Tahmasb remains an open question,

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 51EA4737 /production/pindar/booksxml/hillen/chapter−02

54 So much for fact. Everything else is speculation. Thus far speculation has, for the most part, confined itself to suggesting precisely who did what in the 257 unsigned paintings. And for good measure, the stylistic antecedents and affiliations of the manuscript have also been examined minutely and with authority. But – and it is a big “but” – scholars have so far not applied themselves in the requisite detail to such central questions as why the manuscript was ordered in the first place, especially at this particular time, why it was illustrated so lavishly, what governed the choice of illustrations or their frequency, what its principal themes are – essentially, what it means –what impact it had on subsequent painting,15 or indeed, why Shah Tahmasb should have presented it in 1568 to the Ottoman Sultan Selim II as an accession gift.16 The present article will try at least to tilt at some of these questions, but obviously shortage of space makes it impossible to develop the necessary arguments at the requisite length. This paper will therefore concentrate for the most part on two main topics: first, the place of this particular Sha¯h-na¯ma within the long-established tradition of Sha¯h-na¯ma illustration; and secondly, the message put across by the paintings themselves.

though it is tempting to see an allusion to both Ismacil and the young Tahmasb in the father/son tableau of Gayumars and Siyamak in this painting. 15 See note 4 above for a preliminary discussion of this topic, which deserves a far more extended treatment than this article can give it. Necessary constituents of any such discussion are the element of overkill already mentioned (a potent compound of the sheer size of the manuscript, the high quality of the images and the frequency of illustration); the exhaustion of resources which the project necessarily entailed; and, more intangibly, the demoralisation which afflicted later artists faced with this towering achievement, artists who were deprived by its very magnitude of the mental space they needed to develop their skills to the full. By this reckoning, then, Shah Tahmasb’s revulsion from painting (which might itself have had something to do with the sense that this Sha¯h-na¯ma was an unsurpassable achievement) was not the only factor which inhibited the development of book painting in Iran in the decades following the completion of the British Library Nizami (Or. 2265) in 950/ 1543. Nevertheless, it must be remembered that the direct impact of this Sha¯h-na¯ma on Persian painting ceased in 1568 when the manuscript left Iran for Turkey, although no doubt its fame remained legendary. Conversely, its subsequent impact on Ottoman painting would repay close attention. 16 This Sha¯h-na¯ma was not, it seems, the only princely gift from Shah Tahmasb to the Ottoman sultan. Accompanying it was a luxurious Qur’an, allegedly written by cAli himself, for which see the discussion by W. Robinson in Islamic Art, Indian Miniatures, Rugs and Carpets (Christie’s catalogue, 20 October 1992), no. 232, pp. 94–7.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4B433DF2 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

55

Several factors combine to make this Sha¯h-na¯ma in particular an unusual production. First, the fact that the entire enterprise was undertaken at all. This may seem rather paradoxical. Surely Shahs ordered Sha¯h-na¯mas – it went with the job? And yet the assumption that it was a matter of course for rulers to order splendid illustrated versions of this text is totally mistaken. Even the briefest survey is enough to show that by about 1520 the age of magnificent royal Sha¯h-na¯mas had long since passed. Fashions had changed. For the best part of a century, lyric poetry had supplanted Firdausi’s epic in the affections of princely patrons of painting.17 And that involved a significant shift in the themes of book illustration. The subject matter of poets such as Nizami, Sa cdi, Khwaju Kirmani, Amir Khusrau, Jamali, Nava’i, Hafiz and Jami focused on peace rather than war, on the pleasures of the court, fantastic adventures, religious topics and even on the humdrum details of everyday life. Nor were the repercussions of this change from epic to lyric poetry confined to subject matter. The much shorter texts of the lyric poets meant that these royal books had far fewer illustrations than did the typical Sha¯h-na¯ma. Where pictures were few, their significance, the load of meaning they had to bear, automatically increased. Perhaps as a result of this, but perhaps also quite independently, the style of Persian painting, as is well known, experienced a startling transformation around 1400. Everything became miniaturised. The grand scale, the furious energy of the best 14th-century painting gave way to decorum, finesse, precision: the so-called classical style of Persian painting.18 It became fashionable to load paintings with far more detail than before. And given their meticulous execution, they now took far longer to produce. To put it bluntly, they cost a lot more: in time, in craftsmanship, in materials – and hence, in money.

17 The fact that a recent article focusing on royal Sha¯h-na¯mas of the 15th century mentions only three such manuscripts is telling evidence in favour of this contention (E.G. Sims, The Illustrated Manuscripts of Firdausi’s Shahnama Commissioned by Princes of the House of Timur, Ars Orientalis 22 [1992], 43–68, esp. p. 55). 18 For concise statements of this process, cf. E.J. Grube, The Classical Style in Islamic Painting. The early school of Herat arid its impact on Islamic Painting of the later 15th, the 16th and 17th centuries. Some Examples in American Collections (Lugano 1968), pp. 25–8 and E.G. Sims, The Timurid Imperial Style: Its Origins and Diffusion, Art and Archaeology Research Papers 6 (1974), 56–67.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 53370A8B /production/pindar/booksxml/hillen/chapter−02

56 The effects of all this can be traced in 15th-century Sha¯h-na¯mas. Only two examples of the first rank19 – what might be termed metropolitan work – are known in the entire century: that made for Baysunqur in 1430, with 21 miniatures,20 and the one made about a decade later for his cousin Muhammad Juki, which has 31 illustrations.21 In both cases the paramount intention was clearly to emphasise quality above quantity. The inevitable result was a book which, visually speaking, was largely text. Provincial Sha¯h-na¯mas of the period, from that of Sultan Ibrahim (ca 1435)22 downwards,23 went the opposite way, stressing quantity rather than quality. As a result they tended much more to the opposite extreme, namely to look more like picture books with interpolated text. Among the better-quality provincial works, such as the Dunimarle Sha¯h-na¯ma of 850/1446,24 the gap between what was planned (in this case 153 miniatures) and what was completed (80 miniatures) tells its own story. Easily the most ambitious of the provincial Sha¯h-na¯ma projects brought to term in the 15th century is the so-called Big-head Sha¯h-na¯ma in two volumes produced for Sultan cAli Mirza of Gilan in the 1490s (the colophon of the second volume is dated 899/1494), originally with some 350 paintings in a striking and vigorous, if somewhat crude, style.25

19

Dr Sims, in the article mentioned in n. 17 above, argues in favour of treating the Sha¯ah-na¯ma of Sultan Ibrahim as royal, rather than “provincial”, the term applied to it by B.W. Robinson, e.g. in his Persian Miniature Painting from Collections in the British Isles (London 1967), pp. 91–2. By that reckoning there would be three, not two, such “royal” Sha¯h-na¯mas. 20 B. Gray, The Sha¯hna¯meh of Ferdowsi. The Ba¯ysonghori Manuscript. An Album of Miniatures and Illuminations (Tehran 1971). 21 J.V.S. Wilkinson, The Sha¯h-Na¯mah of Firdausı# (London 1931); B.W. Robinson, Unpublished Paintings from a Fifteenth Century Book of Kings, Apollo Miscellany (June 1951), 17–23; idem, The Sha¯hna¯ma of Muhammad Jukı#. RAS. Ms 239, in S. Simmonds and S. Digby (eds), The Royal Asiatic Society. Its History and Treasures (London 1979), 83–102. 22 For arguments justifying this date see E.G. Sims, The 101 Paintings of Sultan Ibrahim, in R. Hillenbrand (ed.), Studies in Persian painting. Festschrift for Basil Robinson, Pembroke Papers 3 (in press). 23 B.W. Robinson, A Descriptive Catalogue of the Persian Paintings in the Bodleian Library (Oxford 1958), pp. 16–22. 24 Idem, The Dunimarle Sha¯hna¯ma. A Timurid Manuscript from Mazandaran, in R. Ettinghausen (ed.), Aus der Welt der islamischen Kunst. Festschrift für Ernst Kühnel zum 75. Geburtstag am 26.10.1957 (Berlin 1959), 207–18. 25 Idem, The Turkman School to 1503, in B. Gray (ed.), The Arts of the Book in Central Asia (London and Paris 1979), p. 243 and colour pls. LXVIII-LXIX; and idem, A Survey of

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 522B0CB8 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

57

Numerous Sha¯h-na¯ma illustrations, either of isolated episodes or of substantial portions of the text, and of metropolitan workmanship, survive, and are to be found (or presumably once belonged) in anthologies or albums.26 Some, like the royal Turkman (or conceivably early Safavid) leaves in Leipzig and London,27 may best be interpreted as impressive torsos of projects left unfinished precisely because they were far too ambitious. These are indications of a sea-change in taste. But it is, above all, the total absence of a royal Sha¯h-na¯ma produced in the ateliers of Herat under Sultan Husain Bayqara that insistently demands an explanation.28 Here, where superlative visual interpretations of one Persian and Turkish lyric poet after another were produced in a long line of masterpieces, where the classical style of Persian painting found definitive expression in the work of Bihzad and his colleagues, where Persian culture was fostered with a taste and commitment simply not encountered elsewhere – here, of all places, Firdausi alone was ignored. And the most likely explanation is cost.29 The finest Herati Nizami, namely the British Library copy dated 1494, is, with its 22 paintings,30 the most lavishly illustrated manuscript of the entire school.31 In this case, the rate of illustration works

Persian Painting (1350–1896), in C. Adle (ed.), Art et Société dans le Monde Iranien (Institut Français d’Iranologie de Tehéran. Bibliothèque Iranienne no.26, Paris 1982), pp. 34–5. 26 These include, for example, the leaf in the Fogg Museum of Art depicting the visit of Tahmina to Rustam (the best account is still that of E. Schroeder, Persian Miniatures in the Fogg Museum of Art [Cambridge, Mass. 1942], pp. 51–74); the volume of Epics dated 800/1397 and divided between the British Library and the Chester Beatty Library, Dublin (I. Stchoukine, Les Peintures des Manuscrits Timurides [Paris 1954], p. 40); and the ex-Vever collection leaves datable ca. 1440 now in the Arthur M. Sackler Gallery, Washington (G. Lowry and S. Nemanzee, A Jeweler’s Eye. Islamic Arts of the Book from the Vever Collection [Seattle and London 1988], pp. 92–5 and colour pls. 15–16). For further examples, see Sims, “Illustrated Manuscripts”, p. 63, n.47. 27 B.W. Robinson, Origin and Date of Three Famous Shah-Namah Illustrations, Ars Orientalis 1 (1954), 105–12. 28 There is not even a sign of an abortive project of this type. 29 See L. Golombek, Discourses of an Imaginary Arts Council in Fifteenth-Century Iran, in L. Golombek and M. Subtelny (eds), Timurid Art and Culture. Iran and Central Asia in the Fifteenth Century (Leiden, New York and Cologne 1992), p. 13. 30 Counting the double-page frontispiece as two paintings. 31 F.R. Martin and Sir T. Arnold, The Nizami MS. illuminated by Bihzad, Mirak and Qasim Ali written 1495 for Sultan Ali Mirza Barla¯s ruler of Samarqand in the British Museum (Or. 6810) (Vienna 1926), is the only publication to date with illustrations of all the miniatures; its text is richly though unintentionally comic.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 583E4E91 /production/pindar/booksxml/hillen/chapter−02

58 out at an average of one painting per 13 folios. A Sha¯h-na¯ma which maintained that rate of illustration would have taxed the resources of the royal atelier as it was then constituted beyond breaking point,32 for it would have meant producing twice that number of illustrations.33 And given that the trend of 15th-century metropolitan painting was towards manuscripts whose paintings could not be hurried because they were immensely taxing from the technical point of view, the rate of illustration inevitably declined, whatever the manuscript was. In the particular case of the Sha¯h-na¯ma, moreover, the development throughout the 15th century, at a commereial level, of an ample and flourishing iconography, with many set-pieces which could scarcely be omitted by any self-respecting patron, presumably created a climate of expectation for richly illustrated Sha¯h-na¯mas. Certainly the ratio of illustrations to text was much higher than that of the Herat Nizami, Or. 6810.34 Thus it is unlikely that the somewhat parsimonious solution adopted in Baysunqur’s time, whereby the royal Sha¯h-na¯ma was allotted barely a score of paintings, would have recommended itself to Sultan Husain Bayqara 60 years later. A royal patron was therefore by now caught in a double bind: a sparsely illustrated royal Sha¯h-na¯ma would compare unfavourably with the plentifully

32 Technically, moreover, Or. 6810 is not a royal Nizami, since it was produced for the Amir Farsi Barlas and not for Sultan Husain Baiqara himself. This provides further evidence, if any were needed, that royal manuscripts simply held less illustration, page for page, than those made for a lower level of patronage – a paradoxical conclusion until one considers the remarkable discrepancies in quality between royal and commercial work. 33 The huge variety in the size of the text block, compounded by major differences in the size of the handwriting, makes straight comparisons between Nizami and Firdausi manuscripts an overly simplistic and hence misleading exercise. Moreover, it is obviously no easy matter to compare a text of lyric poetry, whose baits are typically divided into four columns, with a text of epic poetry whose baits are typically disposed in six columns per page. But raw length of text offers as neutral a standard of comparison as may be found. Bearing in mind that Firdausi himself, followed by other early authorities, estimated the length of the Sha¯h-na¯ma at 60,000 baits (A.S. Shahbazi, Ferdowsi: A Critical Biography [Costa Mesa 1991], pp. 107 and 123, n.38), a length maintained in many copies of the epic by interpolations and repetitions (cf. ibid., pp. 107–8), while Nizami’s Khamsa comprises about 30,000 baits (E.G. Browne, A Literary History of Persia. II. From Firdawsi to Sa cdi [repr. Cambridge 1969], p. 403), it seems justified to regard the Khamsa as roughly half the length of the Sha¯h-na¯ma. 34 More than a dozen of the non-princely 15th-century Sha¯h-na¯mas listed by Dr Sims in the most valuable Appendix B to her article (“Illustrated Manuscripts”, pp. 67–8) are more richly illustrated pro rata than Or. 6810.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 5D8D0821 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

59

illustrated commercial versions so readily available,35 while a royal copy with the requisite abundance of paintings would break the bank. Effectively, therefore, the Sha¯h-na¯ma, as an illustrated text with an appropriately larger complement of paintings, had priced itself out of the royal market. In its immediate art-historical context, therefore, the Sha¯h-na¯ma-yi Sha¯hı# was truly exceptional on two counts: that it was a royal Sha¯h-na¯ma (the first for almost a century), and that it was most lavishly illustrated. Both these factors involved major breaks with tradition and can therefore be seen as very deliberate choices. It was not a knee-jerk response for Shah Isma cil to embark on the project around 1522, if this was indeed what happened,36 or for the young Shah Tahmasb to continue it. Moreover, as the project unfolded, its sheer scale meant that it gobbled up the lion’s share of the resources available for painting. Naturally, this did not mean that all other work was entirely put aside, but it is noteworthy that the other roughly contemporary royal Safavid illustrated manuscripts, such as the Dı#va¯n of Hafiz in Cambridge, Massachusetts,37 the two-volume Dı#va¯n of Mir cAli Shir Nava’i in Paris38 and the Nizami in New York39 all have relatively few paintings – five, six and fifteen respectively.40 Measured by this standard, the

35

Dr Sims lists 48 illustrated non-princely Sha¯h-na¯mas in the period between 1427 and 1497 (ibid., pp. 67–8). In view of this very large number, the total absence of full-length illustrated Sha¯h-na¯ma manuscripts in the first three decades or so of the 15th century is worthy of note and challenges further investigation. 36 Welch, KBK, pp. 16 and 67; Dickson and Welch, p. 43. The style of some of the earlier folios does not exclude an earlier date. Nevertheless, there is no solid proof that Shah Ismacil (as distinct from Shah Tahmasb) did indeed initiate the project. For a succinct summary of his role as a patron of painting see I. Stchoukine, Les peintures des manuscrits safavis de 1502 à 1587 (Paris 1959), pp. 9–12 [hereafter, PMS]. For Stchoukine’s somewhat anodyne comments on the Sha¯h-nama-yi Sha¯hı#, see ibid., pp. 65–8. 37 Welch, RPM, pp. 62–9; Welch, Wonders, pp. 118–29; Stchoukine, PMS, pp. 60–2. 38 B. Gray, Persian Painting (Geneva 1961), pp. 129–31; Stchoukine, PMS, pp. 56–9; Welch, RPM, pp. 54–61. 39 In the Metropolitan Museum of Art; dated 931/1525. For the fullest scholarly publication of its illustrations to date, see P.P. Soucek, The Khamseh of 1524/25, in P.J. Chelkowski, Mirror of the Invisible World. Tales from the Khamseh of Nizami (New York 1975), 11–20; cf. Stchoukine, PMS, pp. 53–4. 40 For a detailed discussion comprising other illustrated manuscripts of the early Safavid period, see P.P. Soucek’s review of The Houghton Shahnameh in Ars Orientalis 14 (1984), 135–7 (for example, the Dı#va¯n of cAli Shir Nava’i, copied in Tabriz in 934/1528, in the Russian National Library, St. Petersburg, Turkic New Collection 36); and F. Çagˇman, The Miniatures of the Divan-i Hüseyni and the influence of their style, in G. Fehér (ed.), Fifth International Congress of

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 5B62AE2B /production/pindar/booksxml/hillen/chapter−02

60 Sha¯h-na¯ma-yi Sha¯hı#, with its 258 images, is a giant among pygmies. Why was the discrepancy so glaring? After all, it is totally out of proportion to the extra length of the Sha¯h-na¯ma text.41 Clearly the Sha¯h-na¯ma was being allotted a status far superior to that of all other texts – and this in turn is unexpected. Why should this long-familiar text suddenly be so special, so portentous? No such distinction operated in early Timurid painting. Indeed, one has to go back all the way to Ilkhanid times to discover a parallel for this kind of emphasis on the Sha¯h-na¯ma. For it was in the early 14th century that several lavishly illustrated versions of the epic, and incidentally the earliest such copies to survive, were produced within a few decades.42 The motive then seems to have been to reassert the values of Iranian culture in the wake of the Mongol holocaust and in the context of Mongol political dominance – though these manuscripts may have meant very different things to different people. And so the question arises – were broadly similar ideas at work in the very different political situation which obtained in the early Safavid period? At first sight, not at all – Iran was now under Iranian rule. Yet things were far less

Turkish Art (Budapest 1978), pp. 235–6 (for example, cAli Shir Nava’i, Ghara¯’ib al-sigha¯r, of which one copy is in the Istanbul University Library, T. 5669 – this is dated Rajab 930/May 1524 and was copied in Herat, while the other copy, also in Istanbul, is in the Topkapı Sarayı Museum, R. 803, and was copied in Tabriz in 939/1532–3). See especially Çagˇman, p. 236 n.9 for references to illustrated manuscripts of the Khamsa of Nava’i, the Khamsa of Jami and the Dı#va¯n of Amir Khusrau. One may perhaps add to this list a Khamsa of Amir Khusrau in the Russian National Library in St. Petersburg, PIS 267, with illustrations attributable to Tabriz in the first half of the 16th century (H. Suleimanov and F. Suleimanova,Miniatures Illuminations of Amir Hosrov Dehlevi’s Works [Tashkent 1983], pls. 53–60). Dr Çagˇman associates a series of illustrated manuscripts produced in Istanbul in the early 16th century with the developing style of Tabriz in this period (pp. 236–9). Certainly there was more than one style of painting current in Tabriz at the time that the Sha¯h-na¯ma-yi ShThı# was being produced. All this material, indeed, and the varied interplay between Herat, Tabriz and Istanbul which it suggests, indicates that what might be termed the stylistic hinterland of the Sha¯h-na¯ma-yi Sha¯hı# still requires much further study. 41 The first volume of the Nava’i manuscript in Paris, for instance, has only six illustrations in some 500 folios, while the second volume (Bibliothèque Nationale, suppl. turc 317) has no illustrations at all, though its illumination is very fine (E. Blochet, Les Peintures des manuscrits orientaux de la Bibliothèque Nationale [Paris 1914–20], p. 287; idem, Les enluminures des manuscrits orientaux – turcs, arabes, persans – de la Bibliothèque Nationale [Paris 1926], p. 96). For a detailed discussion of the paintings of Volume I, see Blochet, Enluminures, pp. 96–102 and idem, Peintures, pp. 282, 286–8. 42 M.S. Simpson, The Illustration of an Epic. The Earliest Shahnama Manuscripts (New York and London 1979).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 606D522B /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

61

cut and dried than such a statement implies. For in the 1520s and 1530s, precisely the decades when this Sha¯h-na¯ma was in course of production, the very existence of Iran as a political entity, ruled by Persians for Persians, was under continual threat.43 And that threat might be termed a creeping Turkification: from the west, from the east, and from within.44 Of course there is a danger of over-simplification here. The interplay between Turk and Tajik in the early Safavid state was complex and delicate,45 exemplified by a largely Turcoman army built on Turkic tribal principles,46 by Persian administrators, chancery officials47 and even provincial governors,48 and above all by Iranian Shahs who spoke a Turkic language at court, wrote poetry in Turkish dialects49 and used Turkish formulae in their farma¯ns,50 and yet conducted state correspondence in Persian with their Turkish foes. 43

A threat which modern specialists, enjoying the benefits of hindsight, are apt to underplay. 44 Cf. V. Minorsky, La Perse au XVe Siècle, Iranica. Twenty Articles (Tehran 1964), p. 326 and idem, Tadhkirat al-mulu¯k. A Manual of Safavid Administration (London 1943), Appendix I, “Iranians and Turks”, p. 188. 45 R.M. Savory, The emergence of the modern Persian state under the Safavids, Iranshinasi 3 (1971), pp. 25–6; idem, The significance of the political murder of Mı#rza¯ Salma¯n, Islamic Studies. Journal of the Central Institute of Islamic Research, Karachi 3 (1964), 181–91. Cf. also H.R. Roemer, Problèmes de l’histoire safavide avant la stabilisation de la dynastie sous Ša¯h cAbba¯s, in Actes du Ve Congrès International d’Arabisants et d’Islamisants (Brussels 1971), pp. 403–4. This process could even be seen as a class struggle; see R.M. Savory, Some Reflections on Totalitarian Tendencies in the Safavid State, Der Islam 53/ii (1976), p. 241. For an overview of this issue, see idem, Iran under the Safavids (Cambridge 1980), pp. 31–3. 46 Minorsky, Tadhkirat, p. 30 and Appendix II, “The Supporters of the Lords of Ardabil”, especially pp. 190–5. 47 2 I.e. “the men of the pen”; see R.M. Savory, “K . izilba¯sh”, Encyclopaedia of Islam , V, p. 244. Cf. idem, The qizilbash, education and the arts, Turcica 6 (1975), p. 169. A typical post was that of the khalı#fat al-khulafa¯: see idem, The office of “khalı#fat al-khulafa¯” under the Safawids, Journal of the American Oriental Society 85/iv (1965), p. 498. Cf. also idem, Some notes on the provincial administration of the early Safawid empire, Bulletin of the School of Oriental and African Studies 27/i (1964), p. 117. 48 Such as Qazi Muhammad Kashani; see Savory, “Some notes”, pp. 115–16. 49 V. Minorsky, The poetry of Sha¯h Isma¯ c#ıl I, Bulletin of the School of Oriental Studies 10/iv (1940–2), pp. 1010a-24a (for the poems themselves, see pp. 1030a-53a). As it happens, Isma cil did also write a little Persian verse (ibid., p. 1008a). By a neat reciprocity, his contemporary Sultan Selim wrote his own verses almost exclusively in Persian (E.G. Browne, A History of Persian Literature in Modern Times, 1500–1924 [Cambridge 1924], p. 12). 50 Minorsky, Tadhkirat, p. 199. For further examples of cultural Turkification, see Qazi Ahmad, Calligraphers and painters, pp. 94–6.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 48FC3E49 /production/pindar/booksxml/hillen/chapter−02

62 Recent research has shown, too, how complicated the demographic and linguistic picture was in Iran as a whole.51 Politically, however, matters are more clear-cut. The crushing defeat inflicted on Shah Ismacil by the Sunni Ottomans at Chaldiran in 1514 had permanently removed from Safavid control much of eastern Anatolia52 and had dramatically highlighted the exposed position of Tabriz as the Safavid capital, and with it much of Azarbaijan. On the northeastern marches of the Safavid state the equally Sunni Uzbeks under cUbaid-Allah Khan were vigorously engaged in what has been termed the duel for Khurasan,53 and at times it must have seemed likely that the Safavids would lose that province too. Furthermore, the 1530s saw territorial disputes over Qandahar and the eastern frontier with the Mughals54 – also Sunni, also Turkic. And the external attacks from east and west, which forced Tahmasb to fight a war on two fronts, were exacerbated and thus rendered much more perilous by the full-scale civil war which had erupted in the Iranian heartlands themselves in 932/1526 between the competing factions of Turkish tribal federations: Ustajlu, Shamlu, Rumlu, Tekelu and so on. This dragged on for at least eight years.55 It is this specific contemporary situation which may help to explain not only the decision to embark on the most ambitious Sha¯h-na¯ma in the annals of Persian painting, but also the particular choice of episodes for illustration. First, then, what light does this political situation shed on the decision to embark on the Sha¯h-na¯ma-yi Sha¯hı#? Surely it is not too bold to suggest that this was deliberately intended as a public patriotic gesture made by the head of state? If royal Sha¯h-na¯mas had long been two a penny, one could scarcely maintain such an argument; but that this was not the case and most specifically not the case with the two major dynasties which the Safavids

51

Cf. the paper by J.R. Perry in this volume and the remarks of Roemer, “Problèmes”,

p. 405. 52 Though of course the impact of these lands on the Safavid state, exercised by people native to these territories, long outlasted Chaldiran. Cf. R.M. Savory, The consolidation of Safawid power in Persia, Der Islam 41 (1965), p. 85. 53 Cf. M.B. Dickson, Sháh Tahmásb and the Úzbeks (the duel for Khurásán with cUbayd Khán: 930–946/1524–1540) (unpublished Ph.D. thesis, Princeton University 1958). For a brief summary, see H.R. Roemer, The Safavid Period, in P. Jackson and L. Lockhart (eds), The Cambridge History of Iran, VI. The Timurid and Safavid Periods (Cambridge 1986), pp. 235–40. 54 Roemer, “Safavid Period”, pp. 239 and 244. 55 Savory, “K . izilba¯sh”, p. 245; Roemer, “Safavid Period”, pp. 233–5.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4BE9DA27 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

63

supplanted, namely the Aq Qoyunlu and the Timurids, suggests that the news of this project might well have caused a stir. It should be borne in mind, too, that both these earlier dynasties had strong Turkic affiliations. This might in turn have given a certain piquancy to the report that virtually the entire resources of the royal kita¯bkha¯na – whose artists included the cream of the earlier Aq Qoyunlu and Timurid establishments, now under Safavid control by right of conquest56 – were being channelled into producing the definitive illustrated version of the most Persian of Persian texts. Could this manuscript have been intended as a celebration – somewhat belated, it is true, even at the time of commissioning – of the new dynasty, ordered by its founder when he had at last (in his own eyes) achieved political stability? The word ‘public’ was used advisedly in the previous paragraph. Of course the knowledge of this Sha¯h-na¯ma was broadly confined to court circles. But within those circles it must have made something of a splash simply by virtue of the prodigal expenditure of resources which it entailed, to say nothing of its unprecedented scale and magnificence. It cannot have failed to have put in the shade all other work being done in the kita¯bkha¯na.57 And of course the court – not the Iranian people at large – was the public at which it was aimed. That was audience enough. Yet more factors may also be at work here. Dickson and Welch have argued cogently that the project was begun in the closing years of Shah Ismacil’s reign.58 This does not mean, incidentally, that the choice of episodes for illustration was made then, but that issue is not the germane one here. It has often been remarked that Ismacil was a changed man after Chaldiran.59 He seems never to have taken the field seriously again. Instead, he applied himself to the hunt, the bottle and music, leaving to others much of the everyday business, both civil and military, of the state. Between 1514 and his

56 Welch, KBK, p. 44; for a lengthier assessment of the meeting of various styles of painting in the early Safavid court, see ibid., pp. 48, 50–5 and Dickson and Welch, I, pp. 15–26. 57 For the background to how the early Safavid kita¯bkha¯na worked, see Dickson and Welch, I, pp. 3–14. The close control kept over artists, whatever their medium of work, in late medieval courts in the Iranian world is illustrated by the celebrated Arzadasht; see Thackston, A Century of Princes, pp. 323–7. 58 Dickson and Welch, I, pp. 32–5. 59 Savory, Iran under the Safavids, pp. 45–7; G. Sarwar, History of Sha¯h Isma¯ c#ıl Safawı # . (Aligarh 1939), pp. 86 and 99.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4749860D /production/pindar/booksxml/hillen/chapter−02

64 death in 1524 he certainly had the leisure to turn his mind to painting.60 Did he have the inclination too? Dickson and Welch suggest, again quite plausibly, that the Sha¯h-na¯ma-yi Sha¯hı# was intended as a present from Ismacil to the young Tahmasb newly returned from Khurasan and known to be an enthusiastic amateur of the arts of the book.61 But there is room for other possibilities too. Some of the royal library seems to have been lost as loot after Chaldiran,62 and the new Sha¯h-na¯ma, along with the Nizami completed in 1524–5 (and presumably therefore begun a couple of years earlier), the Nava’i of 1526–7 and the presumably contemporary Hafiz, would have helped to make good these depletions More generally, the very ordering of a Book of Kings can also be seen as a symbolic assertion of royal authority – a dimension obviously missing in texts of lyric poetry. To that extent there was inescapably a propaganda element in the project, an element that bulked ever larger as the project grew in size. An Iranian ruler definitively defeated in actual battle by a Turkish foe may well have drawn solace from the contemplation of numerous Iranian victories over Turan, the land of the Turk, even if these took place only on the written page and in the theatre of the imagination. The stylistic evidence, for what it is worth, seems to point in much the same direction. The style of the early paintings (say most of the first 100 folios) consistently suggests the early 1520s, if not a slightly earlier period, and this consistency in turn indicates that for the most part the paintings were produced in a time-scale following the sequence of the text itself. In other words, assuming a starting date of ca. 1522, this concentration had the blessing of Shah Ismacil himself. It does not seem to have been the practice to allocate episodes from later on in the epic to the artists working in this early Safavid style. What this consistency reveals is that from the very outset it was firmly intended that this Sha¯h-na¯ma would be illustrated on a scale never before known in Sha¯h-na¯ma painting. And as the project developed, at least until just after halfway – that is, until the end of the Iran-Turan feud – this vision was broadly respected. The implications of this emphasis will be examined later in this paper.

60

For his interests in poetry, painting, calligraphy and music, see M.K. Yusuf-Jamali, The Life and Personality of Shah Isma cil I (1487–1524) (unpublished Ph.D. thesis, University of Edinburgh 1981), pp. 154–70 and 232–45. 61 Dickson and Welch, I, p. 34. 62 B. Gray, The Arts in the Safavid Period, in Cambridge History of Iran, VI, p. 879.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4BC152CA /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

65

The presumed dating of the early illustrated folios leads naturally to the second major topic of discussion and thus to the core of this paper. If the propaganda element mentioned above went beyond the most obvious level of cliché, the evidence must lie in the illustrations themselves. And while they have been intensively scrutinised for evidence that bears on style, narrative and attribution, their other aspects have been ignored. One thinks here of their cumulative impact, their interaction, the way they have been selected to highlight some themes and downplay others – for the omissions are as striking as what is actually there. First, it will be relevant to revert briefly to their sheer profusion. At one level, of course, this simply makes the book a more luxurious and pleasurable object. But beyond that, the presence of so many illustrations allowed the person or persons who actually planned the book – perhaps the term ‘project director’ will do – far more scope to stress the themes of their choice. The great physical length of Firdausi’s text and the range and variety of its subject matter offer ample opportunity to develop a sub-text. By a careful choice or manipulation of the scenes to be illustrated it is possible to emphasise certain sections or themes of the text at the expense of others, and to gloss over the latter, or even omit them entirely on the visual plane. By the same token, themes that are of secondary importance, or latent, can be brought into prominence. Thus the written text and the visual text can tell rather different stories. But all this needs plenty of illustrations. That is the prime condition for all such subtleties. They cannot be achieved with the few dozen images which over the centuries had become the standard kit of Sha¯h-na¯ma iconography. Most of those images had to be included anyway as a matter of course; it was the extra ones which gave the project director room for manoeuvre. So what is the message of the illustrations of the Sha¯h-na¯ma-yi Sha¯hı#? That question can best be treated under three headings: the rate of illustration, the subject matter of the illustrations and finally, very briefly, the images which seem to be unique to this manuscript. To begin, then, with the rate of illustration, this tells its own story. Once the introductory matter is out of the way and the narrative proper begins, namely at folio 20, it is quite remarkable to note that until folio 87, with only one exception,63 there is a painting on

63

The gap is between fol. 74vo (“Zal dictates a letter to Sam about Rudaba”) and fol. 76vo (“Rudaba confesses to Sindukht”). The text in between these two episodes describes the arrival of the messenger “upon a white steed of Zabulistan” (The Sha¯hna¯ma of Firdausı#, I, tr. A.G. and E. Warner [London 1905], p. 277), how Sam consulted the astrologers, the

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4D0C94E4 /production/pindar/booksxml/hillen/chapter−02

66 every single page. This means that the foundations on which Iranian history (as interpreted by Firdausi) rest, and in particular the Zahhak and Faridun cycles which show how the feud between Iran and Turan developed, are depicted in quite unprecedented detail. The rate of illustration slows down markedly once the Rustam story gets under way, though his role in the battles against Turan is well developed in the illustrative cycle. The rate picks up again with the story of Siyavush on folio 163 and the re-kindling of the feud with Turan (fig. 2), and continues at a steady rate of one painting every three or four sides (a side being one half of a complete folio) until the execution of Afrasiyab and the installation of an Iranian nominee on the Turanian throne at folio 385. And with that action this apparently everlasting war comes to an end. From the visual point of view, this war is beyond question the core of the book; the number and density of the illustrations proclaim the fact. As if to underline it, there is a palpable gap in illustration at this point, by far the largest so far – 32 sides without a picture. When the story resumes with the story of Gushtasp, we encounter in quick succession large gaps of 13, 17 and 15 sides with no pictures. Isfandiyar’s seven courses, admittedly, are illustrated rapidly one after another, but thereafter the pace slows down very dramatically. The tale of Darab, Dara and Iskandar – some 60 sides – is dismissed with a mere three illustrations64 and there are similar gaps in the Sasanian section. Indeed, for the last 320 folios (that is, almost half the book) – there are only 57 illustrations, which averages out at one for every 11 sides. The other 201 illustrations – four-fifths of the total – are crowded into the first half of the book, which averages out at one illustration for every two sides or so. And in that part of the text the major theme is the enmity between Iran and Turan.65 It might be added parenthetically at this point that the marked falling off in the rate of illustration once the Iran-Turan war has ended is susceptible of more than one explanation. And these various explanations are not mutually exclusive. The iconographic explanation is that, with the end of this longrunning feud, the principal purpose of the book had been achieved. A total return of the messenger to Zal, Zal’s conversation with the woman who acted as go-between for him and Rudaba, and finally Sindukht’s cross-questioning and mistreatment of this woman. There was thus no lack of possible subjects for a painting at this point. 64 Compare the Great Mongol Sha¯h-na¯ma, which devoted, it seems, 17 paintings to the story of Iskandar (Grabar and Blair, pp. 186–7). This can serve as a reminder of how readily the Sha¯h-na¯ma lent itself to special emphasis by pictorial means on a given section of the text. 65 Episodes irrelevant to this grand theme, such as the love affair of Bizhan and Manija, are glossed over and thus not permitted to break the tension.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4799CEFF /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

67

absence of illustrations after this point would naturally have caused a major visual imbalance in the whole book. But the project director might well have felt justified in severely reducing the rate of illustration after this point. Another explanation might be that, as the project progressed, the artists responsible for the later paintings might have found it increasingly difficult to match the achievements of their predecessors, and this sense of inadequacy could help both to explain the declining frequency of images per se and their gradual decline in quality. And finally, economic factors might have played their part. The royal exchequer may simply have become unable to sustain such a continuous drain on its resources, and the end of the Iran-Turan feud might have seemed the most rational place in Firdausi’s text to begin the process of retrenchment. The relatively greater prevalence of simple, even modular, compositions in the second half of the manuscript is a pointer in the same direction. Now to the subject matter of the illustrations. Battles account for 81 of the paintings, almost a third of the total (figs. 3 and 4). Enthronements, the time-honoured formula used to express royal authority, take up 48. And negotiations, embassies, councils of war and the like total another 30 – a quite disproportionate number in comparison with other Sha¯h-na¯mas. Twenty images depict murders, executions or violent deaths. The last significant category is that of beast- or monster-slayings, of which there are 19 examples illustrated. The remaining 60 paintings do not lend themselves to useful categorisation. It will be clear from this brief summary that the subject matter of the illustrations reinforces the message sent out by the rate at which the paintings occur. This is a Sha¯h-na¯ma devoted to a quite exceptional extent to war, including the preparations for it and the attempts to stave it off, and to the expression of royal authority, including royal power over life and death. It is deeply serious in tone. The elements of fantasy, romance and derring-do are peripheral to these central themes. Hence the otherwise inexplicable absence of such chestnuts of Sha¯h-na¯ma illustration as the birth of Rustam, Rustam and Tahmina, Rustam defeating Pilsam and the Khaqan of Chin, Bizhan in the pit, the dı#v Akvan, Kaykhusrau’s rash doings in Mazandaran, and most of the Iskandar cycle. And what of those images that are unique to this manuscript? The most striking aspect of these images is their sheer number – 79 in all, a little less than a third of the total. Although in any richly illustrated Sha¯h-na¯ma one might legitimately expect to encounter one or two brand-new images, it is far

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 410EED21 /production/pindar/booksxml/hillen/chapter−02

68 beyond the norm for the new images to attain 30% of the total. After all, the iconographic tradition was over two centuries old by this time and the programme of images for almost every Sha¯h-na¯ma was relatively predictable. They were, as already noted, produced almost commercially for wealthy (not royal) patrons by this time. The Sha¯h-na¯ma-yi Sha¯hı# breaks all these rules. And once again, the message of these new images is unmistakable. Overwhelmingly, they are devoted to fleshing out the details of the Iran-Turan war. Three-quarters of them fulfil this function; thus we experience the stately to-ing and fro-ing of negotiations between Faridun and his sons in the run-up to the fratricide itself and again before war breaks out. Such negotiations are frequently depicted and it must be admitted that they do not quicken the pulse. The deaths of Tur and Salm are celebrated by unique images, including one of Faridun receiving the head of Tur (fig. 5),66 a sinister counterpart to the familiar image of his grief over the head of Iraj. Perhaps the unwonted emphasis on the interlinked stories of Sam, Zal, Rudaba and Mihrab might conceivably relate to Safavid concerns in Afghanistan, though other explanations for this could be found.67 For the most part, however, as already noted, these extra images highlight visually unfamiliar aspects of the Iran-Turan conflict. The only other major emphasis is on Nushirvan the Just, who is given quite unprecedented coverage with seven hitherto unknown images68 emphasising his caution, zeal for reform, wisdom and diplomacy (fig. 6). Do we have here a role model for Shah Tahmasb?69 It will have been plain long ago where all this leads. The speculation is that this Sha¯h-na¯ma, conceived in the traumatic aftermath of Chaldiran and produced with immense labour in the dark days of the Qizilbash

66

Dickson and Welch, II, pl. 43. For example, the unwavering loyalty to the Shah of these great paladins might provide an instructive object lesson to the quarrelsome and rebellious Qizilbash amirs of the early Safavid court – or, at the very least, offer some consolation to the young Shah faced with the task of disciplining them. 68 For the Nushirvan cycle, see Dickson and Welch, II, pls. 237–49. 69 Nushirvan figures again in the opening illustration of the sparsely illustrated royal Khamsa of Nizami in the British Library, dated 946–50/1539–43 (Welch, RPM, p. 70, pl. 19). 67

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 76 OUTPUT: Tue Aug 10 11:33:05 2010 SUM: 4409B7C1 /production/pindar/booksxml/hillen/chapter−02

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

69

interregnum70 and, above all, of the five successive Uzbek attempts against Khurasan, was a commentary on its own times and embodied the hope of something better over the horizon. The war with Turan, in Firdausi’s as in Safavid times, is painfully long drawn out, and the illustrations faithfully reflect that. cUbaid-Allah Khan, in this interpretation, is the new Afrasiyab, and duly receives his deserts: the battle of Jam in 1528, where Tahmasb proved his valour and cUbaid-Allah was wounded, almost recreated a Sha¯h-na¯ma situation.71 The consonance between the Sha¯h-na¯ma text and contemporary Safavid realities can be discerned once more in the abiding sense of the integrity of Iran’s frontiers and the fierce determination to protect them at all costs. ‘Iran for the Iranians’ is the insistent message. And the recurrent emphasis – in text and image alike – on the sanctity of the ruler and of his person, even if he is under age (as Tahmasb was) or incompetent, is part of that same ‘national’ sentiment. Hence too the depictions of family treachery (which Tahmasb experienced at the hands of two of his three brothers, Alqas and Sam),72 and of rebellion. Indeed, the pictorial cycle of the manuscript ends – no doubt designedly – not with the familiar downbeat and anti-climactic image of Yazdigird, the last Sasanian ruler, squalidly murdered by the miller, but with the defeat of a usurper, Shah Farain Gudarz, an image found only in this Sha¯h-na¯ma.73 To the very end, therefore, the divine right of kings is vindicated. In one significant sense, however, the Sha¯h-na¯ma text and contemporary Safavid preoccupations were at cross purposes. This was, of course, in the matter of religion. The body of Firdausi’s text naturally gave no obvious scope for the expression of Shi ci or for that matter any Islamic belief, but a passage at the beginning of the epic clearly shows his confessional stance. It describes his

70 One should not lose sight of the possibility that the project was connected with Shah Tahmasb’s own sense of coming of age as a king – an event which he dated to the summer of 1527 (Michele Membré, tr. A.H. Morton, Mission to the Lord Sophy of Persia (1539–1542) [London 1993], p. xvi). 71 Roemer, “Safavid Period”, pp. 236 and 249. 72 Ibid., pp. 239 and 242. The revolt of Sam Mirza took place in 941/1534–5 and that of Alqas Mirza in 955/1548; the latter revolt thus does not fall within the time-frame of the Sha¯h-na¯ma-yi Sha¯hı#. 73 Dickson and Welch, II, pl. 261.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02 Sequential 19

JOBNAME: ???? PAGE: 1 SESS: 17 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 04E8E494 /production/pindar/booksxml/hillen/chapter−02−artwork

70

1. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 20vo: The court of Gayumars.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 0B04B602 /production/pindar/booksxml/hillen/chapter−02−artwork

THE ICONOGRAPHY OF THE SHA ¯ H-NA ¯ MA-YI SHA ¯ HI¯

71

2. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 179ro: Afrasiyab and Siyavush embrace.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 06C174D8 /production/pindar/booksxml/hillen/chapter−02−artwork

72

3. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 221ro: Kai Khusrau takes the castle of Bahman.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 088CC60F /production/pindar/booksxml/hillen/chapter−02−artwork

THE ICONOGRAPHY OF THE SHA ¯ H-NA ¯ MA-YI SHA ¯ HI¯

73

4. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 328ro: Bizhan kills Nastihan and stems the Turanian night attack.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 05C8EC87 /production/pindar/booksxml/hillen/chapter−02−artwork

74

5. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 56vo: Faridun receives the head of Tur.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 084B413E /production/pindar/booksxml/hillen/chapter−02−artwork

THE ICONOGRAPHY OF THE SHA ¯ H-NA ¯ MA-YI SHA ¯ HI¯

75

6. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 638ro: Nushirvan receives an embassy from the King of Hind.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 22 OUTPUT: Tue Aug 10 10:24:09 2010 SUM: 0A2CB523 /production/pindar/booksxml/hillen/chapter−02−artwork

76

7. Sha¯h-na¯ma-yi Sha¯hı¯, fol. 18vo: Firdausi’s parable of the Ship of Shi cism.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-artwork Sequential 7

JOBNAME: ???? PAGE: 1 SESS: 66 OUTPUT: Tue Aug 10 10:24:28 2010 SUM: 527DFFC8 /production/pindar/booksxml/hillen/chapter−02−ends

THE ICONOGRAPHY OF THE SHAH-NAMA-YI SHAHI

77

journey on a ship doomed to sink in the sea of eternity but where his fellow-passengers include the Prophet and cAli,74 Hasan and Husain, and the corresponding picture shows them all wearing the Safavid ta¯j (fig. 7).75 And Shi ci invocations fill some of the inscription panels on the architecture depicted in the paintings.76 This is really as far as the artists could go within the confines of a carefully un-Islamic text in alluding to the confessional loyalty of the Safavids. There is a curious and ironic epilogue to all this.77 In 1568 Shah Tahmasb actually gave his precious book away – and to the then Ottoman Sultan, Selim II, at that.78 It was one of several gifts to mark the accession of that monarch. Tahmasb’s youthful enthusiasm for painting had evaporated a quarter of a century earlier. Thus the one person with the authority to dispose of this great masterpiece no longer cherished it as he had done in the past, though no doubt his awareness of the value to be set on the book had not evaporated likewise. Yet in the wider scheme of things that did not matter. It would be mistaken to see Tahmasb as a seller of the family silver. It was more that his priorities had changed. He had laboured for most of his life to set Iran free from the Ottoman menace, and when he concluded the Treaty of Amasya with Suleiman the Magnificent in 1555 he had secured peace in his time.79 Now a new Ottoman sultan had come to power, and it was politic to cement good relations with him. And the political turmoil which had formed the urgent contemporary backdrop to this Sha¯h-na¯ma was now safely in the past. So political realities had changed, and with those changes his Sha¯h-na¯ma had

74

Firdausi, tr. Warner, I, pp. 106–7; see Shahbazi, pp. 57–9 for a revised translation and for a brief commentary on it. The problem of determining Firdausi’s confessional loyalty is complicated by the existence of forged interpolations designed to place him firmly in one camp or another (ibid., p. 49). 75 For a conveniently accessible colour plate, see Welch, KBK, p. 85. It is instructive that this painting, although it is placed near the very front of the book, is in a decidedly later style than its neighbours; this suggests that it was not part of the iconography planned at the outset of the book. Does it bear witness to the growing commitment to Shicism which Shah Tahmasb evinced in the 1530s? Welch ascribes it to the mid-1530s (KBK, p. 84). 76 Dickson and Welch, II, pp. 540–2. 77 I am grateful to Dr Chahriyar Adle for an illuminating conversation on this topic. 78 J.M. Rogers, F. Çagˇman and Z. Tanındı, The Topkapı Saray Museum. The Albums and Illustrated Manuscripts (London 1986), pp. 211–12 and colour pl. 157 from Loqman’s Shehname-i Selim Han dated 6 Dhu’l-Hijja 988/12 January 1581. For a detailed account of the circumstances of the gift, see Dickson and Welch, II, pp. 270–2. 79 Roemer, “Safavid Period”, p. 244.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 66 OUTPUT: Tue Aug 10 10:24:28 2010 SUM: 08D9ED36 /production/pindar/booksxml/hillen/chapter−02−ends

78 lost its topical edge. A couple of decades later, the Protestant Henri of Navarre, constrained to turn Catholic if he wanted to become King of France, murmured “Paris vaut bien une messe”. Tahmasb no doubt saw his book go with pain, but perhaps his feelings were not very different.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-02-ends Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 3BDEC80B /production/pindar/booksxml/hillen/chapter−03

III The Iskandar Cycle in the Great Mongol Ša¯hna¯ma

T

HE exploits and – even more – the idea of Alexander the Great were remembered and embellished just as much in Asia as in Europe. In the eastern tradition, too, his personality was made to bear a weight of significance, especially in religious and philosophical terms, that quickly left the historical Alexander far behind.1 The Alexander Romance of the so-called Pseudo-Callisthenes, probably based on a 7th-century Pahlavi original (Nöldeke 1890, 13–17), was translated into Syriac and Arabic, and its marvels lost nothing in the retelling.2 Persian national sentiment ensured that a bogus genealogy was fabricated for him so that the royal blood of the Achaemenids flowed in his veins and thus legitimized his conquest of the Persian empire. His subsequent campaigns in northern India and beyond were thus claimed for the credit of Iran rather than that of Macedon. The details of these later campaigns were much developed in the Islamic literary tradition and acquired their full quota of fantasy and wonder. Alexander encountered talking trees and talking birds, lands of perpetual fog and fire, climes inhabited by lion-bodied people, by fire-breathing black men or by one-breasted women. Thus the irresistible conqueror was gradually metamorphosed into the seeker after marvels. Nor was this mere self-indulgence. Like the somewhat lesser figure of James Elroy Flecker, it 1

For a brief summary of the Alexander legend in the Islamic world, see Anon. cols. 533b-534b, with bibliography, and the comments of Pfister 36–40. The Persian world is briefly covered in Darmesteter 1883, Darmesteter 1892, Abel 1966 and Abel 1978. 2 Budge translates, in addition to the Ethiopic version of the Pseudo-Callisthenes (1–353), Ethiopic versions of the life of Alexander by the Christian Arabs al-Makı#n (355–385) and Abu¯ Ša¯kir (387–401). Cf. Nöldeke 1890, 53, and, for translations of the relevant passages from al-Dı#nawarı# and al-T . abarı#, ibid. 35–47.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 48F7429F /production/pindar/booksxml/hillen/chapter−03

80 was “for lust of knowing what should not be known” that he trod “the golden road to Samarkand” and beyond. His endless journeyings thus became a metaphor of the quest for knowledge. It was for this reason that he sought out, or was accompanied by, wise men and philosophers – who are indeed later shown declaiming over his bier on the vanity of human wishes (pl. 1). But in the Middle Eastern tradition Alexander is still more than the supreme warrior and the seeker after knowledge. He also has a religious role. In popular belief he was equated with an enigmatic figure mentioned in the Qur an (18:83 onwards), namely Dhu l-Qarnain (“He of the Two Horns”). The term originates in the 6th-century Syriac legend in which Alexander says to God: “I know that thou hast caused horns to grow upon my head, so that I may crush the kingdoms of the world with them” (Pseudo-Callisthenes 156). According to the Qur an, God gave Dhu l-Qarnain power over the earth and he was thus enabled to travel to the furthest east and the furthest west. In response to the pleas of people living in the far north he built a rampart against their tormentors, the savage tribes of Gog and Magog (pl. 2). And he lectured the people of the west on God’s punishment which awaited the unrighteous, and His reward for the virtuous (Watt). Dhu l-Qarnain is thus regarded as a prophet, and the archetype of the just ruler whose power is the direct result of God’s will. Yet not all commentators were happy to identify him with Alexander,3 who for this reason occupies an ambiguous intermediate role between such indubitable prophets as Abraham or Moses and men who were merely distinguished. This uncertain status may explain several curious inconsistencies in the way that he is depicted. He lacks the veil, the green clothing, the cowl or the halo (whether circular, radiating or flame) which are so often the distinguishing marks of a prophet in Islamic painting; and yet his exploits have a supernatural rather than a merely heroic flavour. His propensity to found cities (the Muslim sources record a dozen, each named al-Iskandariya after him)4 and the tale of his ascension to Heaven – which inevitably presages that of Muhammad himself, though Alexander fails because of his .

3

Later al-T . abarı# says that Alexander and Solomon were the two believers who ruled the world; for a detailed discussion, see Abel 1951. 4 Anon, col. 534b; but cf. Guest for the information that, according to the Ta¯ˇg al- cAru¯s iii, 276, the number was sixteen.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 50EFA2C1 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

81

pride5- again stake out for him this middle ground between simple worldly ambition and true prophethood. It may be that this ambivalent status explains why only certain aspects of his life are developed in the pictorial cycles which were devoted to him in Islamic painting. It is one of these which forms the subject of this paper. Book painting, of course, implies a text. And as it happens, the story of Alexander captured the imaginations of some of the greatest poets in the Persian language, to say nothing of the writers of prose romances. The epic poet Firdawsı#, author of the Ša¯hna¯ma or “Book of Kings” completed around 1010, composed the earliest surviving lengthy account of the Alexander tale in Persian poetry.6 His version was the basis for the very different treatment of the theme by the lyric poet Niz.a¯mı#, which was completed in 1204, as the Iskandarna¯ma or “Book of Alexander,” and incorporated into his Hamsa or “Quintet” (Bürgel 586–587). He in turn inspired such 14-centuryypoets as Amı#r Husraw, writing in Delhi, or Ah.madı#, writing in Turkish Anatolia, to try y at this theme – in the latter case, with gratifying success, for the their hand Shaibanid ruler of Central Asia in the early 16th century, Uzbek Khan, slept with Ah.madı#’s Iskandarna¯ma under his pillow, and took it into battle with him – clear evidence of its power as a talisman. Two thousand years after his death, Alexander was still being remembered as the man who had brought civilisation to Central Asia and founded cities there. And in that same book the constant alternation between episodes of Muhammad’s life with those from Alexander’s makes plain the role of the latter as one of the many forerunners of the Prophet of Islam. Indeed, in the Hwa¯bna¯ma (“Book of y Dreams”) of the Ottoman writer Baylı#, Alexander appears alongside Jesus 7 and Muhammad as a dispenser of advice in dreams. As for works of history pure and simple, the Islamic literary habit of beginning such works at the very beginning, namely with the creation of the world and of man, ensured that as a matter of course Alexander was mentioned at roughly the correct juncture chronologically. Even in some of these prose accounts – such as that in the Ga y¯ mi c al-Tawa¯rı#h of Rašı#d al-Dı#n – legendary material creeps in, ˘ 5

Compare the impious king Kayka¯ u¯s in the Ša¯hna¯ma, who attempted the same exploit. The sinful nature of this attempt explains why, in some depictions of this scene, his face has been blanked out. 6 On the question of origins, see Spiegel; Nöldeke 1979, 29–30, 66 and 69–71; and Hanaway. 7 For this Turkish material, see the paper by C. Sawyer in this volume. For a brief summary of Ah.madı#’s work, see Gibb 270–284.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 49058CDB /production/pindar/booksxml/hillen/chapter−03

82 though it was in poetry rather than in prose that the fantasy element in his adventures was exploited (e.g. his visit to the land of Gloom; cp. pl. 11 with Rice, pl. 22). Even so, medieval Persian prose versions of the Alexander romance produced a further flood of legendary detail (Southgate). The purpose of this paper is to investigate how the story of Alexander, in his Persian incarnation of Iskandar, is recounted visually in one particular version of Firdawsı#’s Ša¯hna¯ma, namely the Great Mongol Ša¯hna¯ma (also known as the Demotte Ša¯hna¯ma from the name of a previous owner), and datable to the 1330s.8 It should be stressed at the outset that it is by no means self-evident that the Iskandar cycle would be illustrated in the same way, or at the same rate, or even at all, from one Ša¯hna¯ma manuscript to the next. Equally, the differences between one set of Ša¯hna¯ma illustrations and the next are not confined to matters of style, composition, illustrative technique and the like. All that one can say with confidence is that the text offered a more or less unchanging point of departure for successive generations of artists. What they did with it – or if they did anything at all – was their own affair. Modern theatre-goers know well enough that the plays of Sophocles or Shakespeare can be radically transformed by the use of modern dress or in a production driven by polemic. The text itself is the same as ever, but it can be made to bear a meaning that its author may never have intended. Just such a hidden agenda seems to lurk in the Iskandar cycle of the Great Mongol Ša¯hna¯ma. So much, then, by way of introduction. In what follows, this paper will attempt first to locate the Iskandar sequence in our Ša¯hna¯ma as precisely as possible in the context of other Ša¯hna¯mas; second, to account for the specific choice of scenes in that manuscript; and third, to suggest how this entire Iskandar cycle might have been interpreted at the Mongol court where it was produced.

8

For the significant periodical literature, see the only monograph on this manuscript, by Grabar and Blair, 191, notes 2–3. They suggest that the manuscript “was ordered, planned and begun by Rashid al-Din’s son Ghiyath al-Din between November 1335, when he organised the appointment of Arpa as sultan, and his death on May 3, 1336” (48). They do entertain the possibility of Abu¯ Sac#ıd as the patron, but for them “he is a less likely candidate than Ghiyath al-Din” (49). The present paper is an attempt to press the case of Abu¯ Sac#ıd as the patron of the manuscript, though that would not exclude the possibility that Ghiyath al-Din had a major part in planning it.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 529246A8 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

83

To assess the particular nature of the Iskandar cycle in our manuscript, it is obviously desirable to investigate how that cycle is treated in other Ša¯hna¯mas. Here a major problem presents itself. The Great Mongol Ša¯hna¯ma was produced before traditional cycles of images had fully developed.9 Moreover, it is by general consent the earliest de luxe royal Ša¯hna¯ma in the history of 14th century Persian painting – easily the largest in sheer scale, being physically very much larger than its nearest rival,10 and probably intended to contain (for it was never finished) many more paintings – some 190 in all (Blair 1989, 127). Firdawsı# recounts the tale of Iskandar in leisurely fashion, employing some 3000 lines which describe over 100 scenes suitable for illustration. Yet over the centuries Persian artists chose a grand total of only 37 out of all these possible scenes – roughly one third – and developed a marked tendency to confine themselves to a few chestnuts selected from this number. Ninety-two of the known illustrated Ša¯hna¯mas dated or datable between c. 1300 and c. 1650 contain images from the Iskandar cycle,11 and there are dozens more illustrated Ša¯hna¯mas which contain no images at all drawn from that cycle. Even among these 92, only 12 have four or more images which relate to the story of Iskandar. About half of the 92, in fact, have only a single image from that cycle. There is no point in complicating the argument here by employing too many statistics; the salient facts are that the entire Iskandar cycle was very frequently omitted altogether from the illustrative programme, and that even when it was not ignored it was treated only in a very cursory fashion.

9 For a discussion of the circumstances see Simpson 1979, 7, 221–248, 324, Simpson 1982, 43–47, 50–51, and Gray; for the contrary view, convincingly presented with hitherto unused evidence which bears both on wall painting and book painting, see MelikianChirvani 1988, 40–45. 10 For example, in Simpson’s proposed reconstruction the Freer Ša¯hna¯ma had at least 107 illustrations and the first “Small” Ša¯hna¯ma had at least 109 (1979, pp. 79 and 89), while the 1330 Ša¯hna¯ma in Istanbul had 98 illustrations (Simpson 1982, 49). The 1341 Inju Ša¯hna¯ma originally had as many as 140 miniatures (Simpson: in press). The Topkapı Šahnama of 731/1330 measures 37.5 x 29 cm (F. Ça˘gman and Z. Tanındı, Topkapı Palace Museum, Islamic Miniature Painting, tr. E. Atıl [Istanbul, 1979], 15). 11 Norgren and Davis contain the raw data which provide the basis for this statistic and those that follow. A similar inventory for illustrated Niz.a¯mı# manuscripts is now available (Dodkhudoeva). This shows that 127 subjects taken from the Iskandarna¯ma were illustrated over the centuries (235–287).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 532A54B1 /production/pindar/booksxml/hillen/chapter−03

84 Nor is this at all surprising, for the story of Iskandar – even if it is interpreted generously to extend to almost the entire reign of his half-brother and predecessor Da¯ra¯ – occupies only about one-twentieth of the Ša¯hna¯ma. And for reasons that cannot be rehearsed here, it never approached the popularity of the other major heroic cycles in this epic, such as those dealing with Rustam, Isfandiya¯r or Bahra¯m Gu¯r. So the point of departure for any detailed analysis of groups of Iskandar images is that such groups, whether large or small, are exceptional. Even the great Ša¯hna¯ma produced for Ša¯h Isma¯ c#ıl and Ša¯h Tahma ¯ sp in the early 16th century, whose . 258 paintings make it the most richly illustrated royal Ša¯hna¯ma to survive, has only three paintings to accompany the story of Iskandar. And among those Ša¯hna¯mas which allot a substantial role to the story of Iskandar, our manuscript is a whale among minnows. With 12 surviving illustrations it is clearly ahead of all the rest, and it originally had 17 on 15 folios, sometimes even two per page, which was unprecedented.12 Its nearest rivals are much later and have 10, 9 and 6 paintings apiece: British Library Or. 1403 (1437–38); St. Petersburg Public Library P.N.S. 266 (1598–99); Dublin, Chester Beatty Library, Per. 104 (c. 1300). The remaining 8 manuscripts in this group each have only four or five illustrations.13 So far as the Iskandar cycle is concerned, therefore, within the entire surviving corpus of Ša¯hna¯ma illustrations, extending as it does across three and a half centuries, the Great Mongol Ša¯hna¯ma is plainly a work apart: the only known version of the text which illustrates the Iskandar story more heavily than any other in the whole Ša¯hna¯ma. In the context of its own time it was, if possible, even more exceptional, for apart from the six images illustrating the Iskandar story in the perhaps contemporary copy in Dublin,14 12 For the evidence supporting this speculation, see Grabar and Blair 2–11, 80–81. But while the emphasis on Iskandar is noted in their book, it is not developed at length. 13 Those with five illustrations are British Library, Add. 27258 (1627–1628); New York, Metropolitan Museum of Art, 13.228.22 (1602–1603); and Windsor, Royal Library, Holmes 151 (A/6), dated 1648–1649. Those with four illustrations are Manchester, John Rylands Library, Ryl Pers 933 (15th-16th century); Dublin, Chester Beatty Library, Per. 214 (1548–1549); Istanbul, Topkapı Sarayı Museum, Hazine 1494 (16th century); Istanbul, Topkapı Sarayı Museum, Hazine 1504 (1509–10); and British Library, Add. 18188 (1846). 14 The scenes are: Iskandar receiving the daughter of the Indian king (f.44); F¯ur fighting Iskandar (f.45); Iskandar with Queen Qaida¯fa and her court (f.46); Iskandar with the talking birds (f.47); Iskandar and the Talking Tree (f.48); Iskandar sees the omen of his death, the baby at Babylon (f.49). It will be seen that half of these scenes do not occur in the Great Mongol Ša¯hna¯ma in its present form, which suggests that the Iskandar cycle had

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 53CEDAA4 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

85

there are only ten other Iskandar images known from the entire 14th century, and these are scattered among eight manuscripts – proof positive of how marginalised the Iskandar story was at this time. Nor can the special emphasis in the Great Mongol Ša¯hna¯ma on the Iskandar cycle be explained simply by the fact that it was a royal manuscript. Some other royal Ša¯hna¯mas, such as that produced for Baysung˙ur in 1429, have no paintings at all which feature Iskandar, while others have only one,15 two16 or, as in the case of the Ša¯hna¯ma of Ša¯h Tahma ¯ sp mentioned above, three. In other words, the . context of our Iskandar cycle in both contemporary and later Ša¯hna¯ma illustration serves to highlight the uniqueness of this group of paintings. It is now time to move from discussion of context to the second theme of this paper, the specific choice of scenes. Here again it will be revealing to investigate the choices made by the artists charged with illustrating other Ša¯hna¯mas. This paper has already drawn attention to the perennial popularity of a few scenes. Of these by far the most commonly encountered is that of Iskandar attending the dying Da¯ra¯, of which 36 examples are known. Next in order of frequency come images of Iskandar and the prophet Hid.r seeking y the Fountain of Life, but with 17 examples this is only half as popular. Only five other scenes are recorded in more than 10 examples; F¯ur dying in battle against Iskandar (15 versions), Iskandar before Queen Qaida¯fa, who holds his portrait (14 versions), Iskandar building the wall against Gog and Magog (14 versions), Iskandar at the Kacba (13 versions) and Iskandar killing a dragon (12 versions). Why are these scenes the most popular ones? This question poses itself with particular force in the case of the image of Iskandar attending the dying Da¯ra¯, because this is far and away the most popular. Perhaps the key factor is that it opens up an affecting human dimension of the inexorable theme of

by no means settled into a familiar pattern of illustrations by the early 14th century. It is also noticeable that in the Dublin manuscript the Iskandar cycle was illustrated as plentifully as could be, with a painting per folio. 15 E.g. London, Royal Asiatic Society, ms. 239; the image in question is, predictably enough, Iskandar comforting the dying Da¯ra¯ (Wilkinson, pl. XIX). It is conceivable that two other Iskandar scenes, depicting him in combat with Da¯ra¯ and F u¯ r respectively, were intended to be included (Sims 66), but this is very speculative. 16 E.g. Bodleian Library, Oxford, ms. Ousely Add. 176; here again the scenes are standard ones: Iskandar guided by Hid.r entering the Land of Gloom, and Iskandar in front of the Talking Tree (Robinson 21).y For a general discussion of the iconography of the Ša¯hna¯mas of Ba¯ysung˙ur, Muh.amamad G y u¯kı# and Sult.a¯n Ibra¯hı#m, see Sims.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 53E74556 /production/pindar/booksxml/hillen/chapter−03

86 royal legitimacy, that leitmotif of the entire Ša¯hna¯ma. Da¯ra¯ does not rail against fate or bewail the loss of his crown to an alien usurper. On the contrary, this scene is a carefully orchestrated tableau of personal and political reconciliation. Its message is more one of peaceful abdication than of warlike conquest – a willing transfer of power in which Da¯ra¯ bestows his blessing on Iskandar and legitimizes him. The scene is also a bridge between one reign and the next, and thus serves to illustrate the stories of both rulers. For an artist seeking to economise on images while still maintaining a reasonable coverage of the epic, the opportunity to kill two birds with one stone was an additional incentive. What of the other scenes? Two of them – F¯ur dying in battle against Iskandar, and Iskandar killing a dragon – are paradigmatic of our hero’s career, punctuated as it is in Firdawsı#’s account by recurrent military campaigns and monster-slayings. These two specific episodes do duty, so to speak, for the mass of similar events which are not illustrated: pars pro toto. More generally, they are leitmotifs of the whole Ša¯hna¯ma, and in that sense it is appropriate that such themes be taken up afresh as each new hero is introduced into the narrative. It is, after all, the business of heroes to perform heroic deeds. The episode of Iskandar before Queen Qaida¯fa can be considered alongside the battle with F¯ur and the dragon-slaying in that it falls into a familiar visual category, namely that of the royal audience scene. But it was familiar on quite another count too, for it owes much to the very popular scene in Niz.a¯mı#’s Husraw wa-Šı#rı#n, where Šı#rı#n holds Husraw’s portrait. The y ¯ fa holding the picture of Iskandar y appears in Ša¯hna¯ma image of Queen Qaida illustrations apparently only after it had become part of Niz.a¯mı# iconography; no 14th-century examples of this scene exist.17

17

See Norgren and Davis (unpaginated) which lists Or. 1403 (dated 1437–1438), f.323, as the earliest occurrence of this scene in Ša¯hna¯ma illustration. And since this manuscript has the fullest Iskandar cycle of any Ša¯hna¯ma apart from the Great Mongol Ša¯hna¯ma itself, it is scarcely surprising that this scene, a relatively popular one in Iskandar iconography, should feature in it. The earliest occurrence of the theme in Niz.a¯mı# iconography occurs in the earliest illustrated Hamsa of Niz.a¯mı# yet recorded, the copy in the British y f.43), and another five examples predating 1438 Library dated 1386–1388 (Or. 13297, survive (Dodkhudoeva 125). Admittedly, the compilation of Norgren and Davis could be greatly expanded on the basis of information now available (cf. Sims 64 and her impressively detailed Appendix B, which demonstrates how radically current knowledge can reshape the profile of Ša¯hna¯ma illustration), but the discrepancy is nonetheless striking.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4924245F /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

87

Thus three of the most popular of Iskandar scenes can be explained readily enough as stereotyped images which with minimal tailoring are made to fit into the Iskandar cycle. The three remaining scenes are more specific in nature. One is the celebrated episode of Iskandar building a wall in the Caucasus against Gog and Magog (pl. 2), the story which above all others caught the imagination of east and west alike and encapsulated to perfection the restless, adventurous spirit of this man for whose ambition the world was too small. But it is striking that among the 7 most popular scenes only this one ministers directly to that unabashed sense of wonder which is the connecting thread in the multiple stories of the Iskandar cycle. The others all concentrate on Iskandar the warrior, the statesman or the seeker after God. The last two scenes in this favoured group are those of Iskandar before the Ka cba and Iskandar accompanying the prophet Hid.r to seek the y Fountain of Life. As it happens, Firdawsı# confers the Christian faith on Iskandar (Nöldeke 51), presumably because he comes from Ru¯m, namely Byzantium. And he had no qualms in compounding the anachronism, either, for he seems to have intended also to present him as a prototypical Muslim, and the popularity of these two scenes indicates that book painters were not slow to pick up this nuance and to exploit it. Perhaps, as in the case of Iskandar’s portrait, the Hamsa of Niz.a¯mı# was laid under contribution to provide suitable images,y for the scene of Iskandar at the Ka cba is indistinguishable from the much more frequently encountered representation of Magˇnu¯n in the same setting. The known depictions of both these scenes date predominantly from the 15th century, namely the very period which saw the elaboration of the Niz.a¯mı# cycle.18 As for the episode of Iskandar and Hid.r seeking the Fountain of Life, y even though it lies beneath the here too the Muslim element is unmistakable, surface. First, Hid.r is associated by many commentators with the servant of y in the Qur an (18:59–81) and has (like Enoch in the Bible) Moses mentioned a somewhat nebulous supra-human status as a man exempt from mortality (see Wensinck 863–865 and Friedländer). And secondly, the relation between the young neophyte thirsting after wisdom and the venerable sage who

18

Nineteen of the 40 Niz.a¯mı# versions of this scene listed by Dodkhudoeva (75–77) date from the 15th century, while 8 of the 13 listings by Norgren and Davis of the scene of Iskandar at the Ka cba date from that same century. For a related example showing the merging of Ša¯hna¯ma iconography with that of Amı#r Husraw in the context of Iskandar, see Brend 1989, 291. For a translation of the Niz.a¯mı# text,y see Bürgel 187.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4477D512 /production/pindar/booksxml/hillen/chapter−03

88 guides him is one that irresistibly recalls the Sufi system where the muršid or postulant puts himself under the discipline and authority of his master, the pı#r. By these means, then, Iskandar is lent a Muslim colouring which greatly enriches the recital of his adventures, and which recurs also in Niz.a¯mı#. Nor is the colloquy with Hid.r an isolated case. Iskandar is repeatedly presented as a y – for example, from the Brahmins (pl. 6; Warner and seeker after wisdom Warner 6, 143–147) or from the women of Harum (ibid, 153–157) – and himself asserts that knowledge, not a large empire, is the aim of his journeys (ibid, 153). His adventures are thus transfigured from the physical to the spiritual plane and can be interpreted, in some sense, as a metaphor for the Sufi path. Finally, how might this Iskandar cycle have been understood at the Mongol court in the 1330s? The discussion so far has shown that the message of a text can be affected decisively by the particular choice and number of illustrations that accompany it. In this particular manuscript, the reader would have realised very quickly, from the dramatic increase in the rate of illustration at this point in the text, that Iskandar was more important than anyone else in the Ša¯hna¯ma - truly a momentous innovation, and one that was to have no successors at all.19 Furthermore, the distribution and content of the paintings might have suggested to such a reader that the principal message of the story is the vanity of earthly desires, whether for possessions or for wisdom. Iskandar in his roles as warrior and monster-slayer is overshadowed by Iskandar the mere man, whose wanderings across the face of the globe are doomed to disappointment. And these themes, it is worth repeating, are driven home by the illustrations independently of the text. Indeed, the text itself has a quite different emphasis, and that is the thoroughly, almost monotonously, familiar theme of Iskandar as the great conqueror, the boringly predictable victor of one battle after another. This aspect of the text cannot have given the painter much inspiration, though of course Iskandar would not be the numinous figure that the text describes if

19

This fact above all others indicates that right here, in the Iskandar cycle, the Ša¯hna¯ma text was being manipulated to a degree never encountered before or since. The present paper attempts an explanation of this without claiming that it is the only one possible; but the fact that the text is being made to serve some other purpose is beyond question and imperatively demands an explanation of some kind.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4F73F4E8 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

89

he were not sometimes shown winning battles or slaying monsters. But there is a sense in which text and images are at cross purposes, and one may be grateful for it. Now it is time to investigate whether contemporary political messages, or even cautionary tales, enrich these paintings. In doing so, it is well to remember the essential privacy of manuscript painting as an art form in contemporary court culture. This manuscript would almost certainly have been made for, and enjoyed by, an extremely select coterie, an inner circle consisting of the sultan and his closest advisers and intimates. Only a very few of them could look at the book at any one time. In view of this private context it would perhaps not be out of place to scan the paintings for (necessarily covert) allusions to recent events at court or in political life and to specific contemporary personalities and their doings.20 Awareness of this extra dimension would assuredly add extra spice to the enjoyment of the Ša¯hna¯ma text itself. Several distinct levels of meaning and allusion may be proposed. The first and most obvious relates to Iskandar in his role as a foreigner who rules Iran. It is not surprising that such a theme should have struck a responsive chord in the Mongols, and indeed the process of assimilating him to the Iranian past by means of a fake genealogy had already begun before Firdawsı#’s time (see Hanaway, Nöldeke and Darmesteter). Thus the alien destroyer of the Zoroastrian tradition21 became metamorphosed into the local hero.22 The

20 Grabar and Blair have done this (13–27) but have not systematically explored the Iskandar cycle in search of supporting evidence. For further contemporary parallels see Soudavar 42. One may note too the emphasis on Aristotle in the painting of the Bier of Iskandar in the Great Mongol Ša¯hna¯ma (pl. 1), which carries with it the idea that the ruler can profit from the guidance of a philosopher – an idea developed by Suhrawardı# (Soudavar ˙ iya¯t suggesting a role for himself here? 415). Was G 21 Darmesteter 1883, 231–245; Nöldeke 1890, 34 n. 1. For the relevant passages in Niz.a¯mı#, see Bürgel 157–162, 188, 245. Cf. Warner and Warner vol. 1, 59, 61–62 for the Zoroastrian memory of Alexander as one of the three arch-enemies of their faith, alongside D . ahha¯k and Afra¯siya¯b, and for his destruction of the Zoroastrian scriptures. The echoes of the outrage which this action provoked could still be heard among the Parsees of 17th-century Bombay, as Chardin noted (Pfister 6, n.3). Zoroastrian legend also credited him with destroying the seven treasures of G y amšı#d (Christensen 1934, 77) and the bridge which G y amšı#d built over the Tigris (ibid. 92, 107–108, 121). And it is not just in the Zoroastrian tradition that this negative side of Alexander was remembered. The Mugˇmal at-Tawa¯rı#h, for example, mentions that he destroyed the fire temple of Minu Dizh outside Isfahan ˘(Christensen 1918, 202).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4F1FB46E /production/pindar/booksxml/hillen/chapter−03

90 longer the Mongols remained in Iran, the more they came to identify themselves with its culture. A persuasive case has been built up by Melikian-Chirvani to suggest that the first major stage in this process, so far as the visual arts were concerned, was marked by Aba¯qa¯ Ha¯n’s refurbishing y later 1260s. But of the hallowed Sasanian site of Shı#z/Taht-i Sulayma¯n in the ˘ this was overshadowed by the conversion to Islam of the later Ilkhanid ruler ˙ a¯za¯n Ha¯n in 1295, which triggered a vast building campaign throughout G y (Wilber 16–19 and 100–102) – almost as if the Mongols were the country seeking to make amends for the destruction they had wrought earlier. On the ideological plane, the World History written by the vizier Rašı#d al-Dı#n and ˙ a¯za¯n, a noted amateur of history, in the early years of the commissioned by G 14th century (Boyle 1971: 19–22, 26), offered reparation of a more subtle and psychological nature by integrating the Mongol past, insofar as it was known (Boyle 1970, 2–5; Boyle 1962, 133–135), into the much more imposing edifice of the history of mankind as seen through Muslim eyes. A vital bridging element here was the retrospective interpretation of the Mongols as being of Turkish stock, the seed of Afra¯siya¯b and thereby linked to the remote Iranian past (Melikian-Chirvani 1984, 289–290 and MelikianChirvani 1988, 32–33). The preparation of a de luxe Ša¯hna¯ma was a further logical step down the same road.23 It could be interpreted as a gesture of solidarity implying Mongol acceptance not only of Islamic culture – and of a Mongol role in that culture – but also of the specifically Persian past.24 This was a pre-Islamic past of heroic deeds and far-flung conquests. From the Mongol point of view, its misty legends conveniently blurred the intractably sharp edges of history. There was room here too for the expression of their contemporary concerns. Now it was not a matter of rewriting history from

22

Firdawsı# incorporates both these contradictory emphases into the Ša¯hna¯ma without resolving them. Compare Warner and Warner 6, 224 and 240, where Iskandar is regarded as infamous and the bearer of an evil name, with ibid., 55, where he undertakes to marry Ru¯šanak so that an heir may be born to restore the Zoroastrian faith. 23 ˙ a¯za¯n associates him with a Thus a panegyric by Huma¯m-i Tabrı#zı# celebrating G series of kings from the Ša¯hna¯ma, including Iskandar (Soudavar 36–37). 24 Melikian-Chirvani 1984, 261–263, n. 25 shows that the process had begun as early as the reign of Aba¯qa¯ – and, to make assurance doubly sure, at the quintessential pre-Islamic site of Taht-i Sulayma¯n. The Ša¯hna¯ma text found on a series of tiles from that ˘ site is even modified occasionally to address him directly (279, 292, 304; Melikian-Chirvani 1991, 110–111).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 5A24BBD7 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

91

scratch (as in the World History, where Mongol affairs assume a disproportionately significant role), but of reinterpreting a familiar text in a new way. Bald assertion gave way to nuance. As it happened, among the various narrative cycles in the Ša¯hna¯ma the story of Iskandar was quite unusually à propos to these new Mongol attitudes.25 Moreover, the connection between the Mongol ruling house and ¯ ljaytu¯, for example, is Iskandar easily predates the Great Mongol Ša¯hna¯ma. U termed a “Second Alexander” in an inscription at the Bast. a¯m shrine (Combe et al., 80–81) and “Alexander-like in judgement” by the chronicler Qa¯ša¯nı#.26 References to Alexander have been noted in the iconography of contemporary Ilkhanid metalwork.27 But the pivotal example is furnished by the palatial complex at Taht-i Sulayma¯n, begun by Aba¯qa¯ possibly at the very ˘ start of his reign (Melikian-Chirvani 1991, 79) although the work seems to have continued into the brief reign of Taku¯da¯r, who on his conversion to Islam took the name Ah.mad. To his patronage has been attributed the spectacular wall revetment of glazed tiles with Ša¯hna¯ma inscriptions28 in which the Talking Tree, so prominent in the story of Iskandar (pl. 4), is a leitmotif.29 The Iskandar cycle is also quoted here (Melikian-Chirvani 1991, 17). A second series of star tiles bore Ša¯hna¯ma inscriptions of a pronounced religious character (ibid., 89, 94 and 97). And at the centre of Taht-i Sulayma¯n, of course, was a natural phenomenon – a hill-top lake fed˘by perpetual springs: perhaps perceived at the time as the very Fountain of Life

25

Melikian-Chirvani 1984, 284 shows that the Iskandar cycle figured at Taht-i ˘ Sulayma¯n. But it was already a cliché to compare rulers with Iskandar. Ha¯qa¯nı# does so in the y case of the Atabeg Ildegiz (Melikian-Chirvani 1984, 293–294), Kama¯l applies the parallel to ˙ azna (Melikianthe Khwa¯razmsha¯h Tekesh (294–295) and Farruhı# both to Mah.mu¯d of G ˘ (ibid., 18). Indeed, the comparison is Chirvani 1988, 17) and to the latter’s son Muh.ammad intended to refer to contemporary events: the victorious campaigns of Mah.mu¯d in India are associated with those of Alexander (ibid., 22). 26 Melikian-Chirvani 1984, 321. Compare also the royal title Ha¯qa¯n-i Sikandar ša¯ n – “the ha¯qa¯n who has the splendour of Alexander” (Melikian-Chirvani y1991, 146). ˘27 Melikian-Chirvani 1984, 297. In his earlier and fuller publication of the same object (1982, 202–207) he does not make this point. 28 Melikian-Chirvani 1991, 139 suggests that the termini are the accession of Ah.mad ˘ uwaynı# (5 March 1283). (6 May 1282) and the death of cAt.a¯-Malik G 29 Melikian-Chirvani 1984, 304–307 and 1991, 100–102. Tiles with Ša¯hna¯ma verses are to be found slightly earlier in the Ru¯m Saljuq palace of Quba¯da¯ba¯d (Melikian-Chirvani 1984, 289, n. 54).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4D5F1655 /production/pindar/booksxml/hillen/chapter−03

92 so eagerly sought by Iskandar (Melikian-Chirvani 1984, 308)? A contemporary poet, Dhu l-Faqa¯r Širwa¯nı#, took up these resonances in a panegyric in ˘ uwaynı# to Hid.r and, by which he likens the s. a¯h.ib dı¯wa¯n cAt.a¯-Malik G y association, the newly Islamicised Taku¯da¯r to Iskandar (Melikian-Chirvani 1991, 131). At the very least, the ambitious programme of works undertaken at Taht-i Sulayma¯n from the reign of Aba¯qa¯ onwards – a programme in which,˘ as the penetrating analysis of Melikian-Chirvani has shown, some of the leading intellects and personalities of late 13th-century Iran had a hand30 – shows that by this time there was nothing new in exploiting both the visual arts and literature so as to associate contemporary Mongol rulers with the heroes of Iranian antiquity. And for good measure the work at Taht-i ˘ Sulayma¯n is saturated with Sufi resonances (Melikian-Chirvani 1991, 99–109); indeed, they were applied directly to the Mongol ruler himself, not least through the medium of the Ša¯hna¯ma itself. Here, then, Iran’s heroic past and Islamic mysticism are both harnessed to the task of acculturating the Mongol royal family to their new realm. Two generations later, the Great Mongol Ša¯hna¯ma took up that same task, but with new emphases. The Ilkhanid ruler associated with the Great Mongol Ša¯hna¯ma, the 30-year old Abu¯ Sac#ıd (one need not go so far as to suggest that his youth was seen as a further link with the historical Alexander) was, as it happens, a man of decisively different stamp from most of his predecessors (Jackson). He was a prince of cultural attainments to whom the Ša¯hna¯ma would have meant something. Unlike his illiterate forebears, he wrote Persian poetry (see Soudavar 40, 54) and was an accomplished calligrapher in the Arabic and Uighur scripts.31 The sources record that he was also a musician and that he took a lively and informed interest in theological discussions, and his precocious military exploits induced him to assume the title of baha¯dur (“hero”) at the tender age of fourteen.32 Such a monarch might well be disposed – in suitably vainglorious mood – to note certain “similarities”

˘ uwaynı#, Nas.#ır al-Dı#n T Namely cAt.a¯’-Malik G . u¯sı# and Qut.b al-Dı#n Šı#ra¯zı# (MelikianChirvani 1991, 76). 31 Boyle 1968, 413. The use of Uighur on the coins minted in his name in the 1330s has been described as “a throwback to early Ilkhanid coinage (…) designed to connect Abu¯ Sac#ıd to his more successful and powerful predecessors” (Blair 1983, 305; cf. Blair 1982, 213). 32 Boyle 1968, 409. It took three years before the new title appeared on Abu¯ Sac#ıd’s coinage (Blair 1982, 214), whereas its first appearance on a monumental inscription, at 30

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 511CEE4F /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

93

between his line and the charismatic figure of Iskandar.33 And Iranian bureaucrats anxious to exculpate themselves from the charge of collaborating with the enemy could personally profit from such a connection (Soudavar 42). The text of Firdawsı# is at pains to stress the legitimacy of Iskandar’s claim to be ruler of Iran; indeed, one might argue that he protests too much, which is suspicious. The Great Mongol Ša¯hna¯ma illustrations bear this out quite explicitly. In this particular manuscript only Iskandar, of all the monarchs mentioned, is depicted both at the beginning and at the end of his reign, seated on his throne (pl. 5) and encoffined on his bier (pl. 1). Writers such as al-Mascu¯dı# and H.amza al-Is. faha¯nı# record the Sasanian custom of producing such formal ruler portraits (Schaeder 232); indeed, al-Mascu¯dı# personally saw a purple-dyed codex at Is. t.ahr with the portraits of 27 Sasanian kings and queens as each looked on the˘ day of their death.34 It is remarkable, to say the least, that the proper forms seemingly denied by the artists to all the other kings in the Great Mongol Ša¯hna¯ma are so punctiliously extended to the one ruler whose legitimacy is not entirely self-evident. In other words, the Iskandar cycle in this manuscript gives visual expression to the idea of how a primordial, innate and conveniently undefined legitimacy – not a narrowly legal one – takes precedence over the rights conferred by birth and family.35 And in the context both of the Mongol irruption into Iran and of the succession squabbles that plagued the closing years of Ilkhanid power, such a mysterious legitimacy would have a potent attraction (Morgan 173; Jackson 376; Boyle 1968, 408–416; Spuler 121–137).

Nasmus, dates to 8 October 1320, a mere fifteen months after the battle which gained him the title (Combe et al. 152–153). The same inscription calls him pa¯dı#ša¯h-i isla¯m, “Emperor of Islam.” 33 It is a pleasing coincidence that like Alexander (and, more to the point, like his ¯ ljaitu) he was a city-founder. The city in question, named after him Abu¯ Sac#ıdı#ya/Bu¯ father U Sac#ıdı#ya, has not yet been located but occurs as a mint name (Blair 1982, 224). Nor was Iskandar the only Ša¯hna¯ma hero with whom he was associated. The necrology of Abu¯ Sac#ıd ˘ amšı#ds, Farı#du¯ns, by cAbd al-Razza¯q al-Samarqandı# compares him with a hundred G Kayhusraws and Afra¯siya¯bs (Melikian-Chirvani 1984, 300). ˘34 Mas cu¯dı#, 150–151. Every king was shown either standing, as a warrior, or seated, dealing with state business. 35 Compare the account in the Karnamak-i Artakhshatr-i Papakan (written c. 600 A.D.) of how the kingly glory (farr) departs from the ruling monarch Ardashir to his challenger Ardavan (Browne I, 143, 145).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 484B3262 /production/pindar/booksxml/hillen/chapter−03

94 A second aspect of the Iskandar story which is of equally obvious relevance in the context of Ilkhanid Iran is the spectacular success of his military campaigns. He is, quite simply, invincible. And while his personal valour is expressed, like that of Rustam or Isfandiya¯r, in single-handed exploits, this is overshadowed by his role as a subduer of nations, a world-conqueror. The parallel with the Mongols is striking. In early 14th-century Iran the memory of the irresistible Mongol hordes who had so quickly wiped out the eastern Islamic world was still green. Indeed, the ˘ uwaynı#, whose patriotism and hatred of the Mongols can be Persian vizier G detected in the ironic sub-text of a panegyric he wrote for his masters,36 chose as his title for the history of those campaigns Ta¯rı#h-i ˇgaha¯n guša¯ – “The History of the World Conqueror.” Moreover, this˘ was scarcely an exaggeration. At their fullest extent the lands under Mongol rule constituted the largest continuous land mass ever controlled by a single power in the history of the world – an empire stretching from the Sea of Japan to the Baltic, from Korea to East Germany. In recounting the career of Iskandar, the Ša¯hna¯ma stresses how he penetrated to the furthest confines of the known world – to Andalusia in the distant west, to India and China, and, most evocatively of all, to that Ultima Thule where civilisation peters out before the menace of chaos, symbolised by Gog and Magog. To an Ilkhanid ruler ensconced at Tabriz and contemplating the wellnigh resistless spread of the Pax Mongolica across Asia in the previous century, it may well have seemed that legend had come to life. Thus Iskandar, as the foreign conqueror par excellence in Iranian history, was the perfect role model for the Mongols.37 It is a clear indication of how the historical personality of Alexander of Macedon had become submerged in a seductive ideal of world dominion. To convey this sense of limitless domain through the medium of pictures was no simple task. The obvious solution – to depict Iskandar defeating one enemy after another – would have been too painfully literal. In the case of this Ša¯hna¯ma, the artists manage vividly to evoke Iskandar’s endless campaigns and journeys by concentrating not on the battles themselves – though two images capture the heat and dust of Hindustan as Iskandar 36 Compare Melikian-Chirvani 1984, 317 for a hidden anti-Turkish message in Suhrawardı#. 37 This exemplary role of Alexander among the Mongols can already be noted before Ilkhanid times. Rašı#d al-Dı#n (212) tells the story of how Mah.mu¯d Yalawacˇ used an anecdote from the life of Alexander to drop a broad hint to Möngke Qa’an on how to deal with refractory notables.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 448453B2 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

95

defeats King Porus – but on the adventures that befall him en route. Constant changes of scene ensure that a distinct sense of place is created for each adventure, and frequent exotic touches – the dark faces of the Hindus, the ascetic Brahmins (pl. 6), the unkempt goggle-eyed savages from the land of Gog and Magog (whose hairstyles and facial colouring [pl. 2], to say nothing of their connotations of chaos and aggression, or their geographical location, all bring the Mongols themselves to mind)38 – serve as reminders of the sheer geographical scope of Iskandar’s realm. In fact, apart from the first and last (deliberately formulaic) paintings of the cycle as it survives, every single Iskandar image in this Ša¯hna¯ma refers to his travels outside the Iranian world. But it is also possible to draw a more ambiguous moral of very contemporary relevance from the material. Admittedly, the early conquests of the Mongols had been spectacular. Chinggis Ha¯n, like Alexander, had run y definitively ended by the out of worlds to conquer. But these conquests had 1260s. Thereafter, Mongol power unravelled in a sorry tale of decline and internal strife. The later Ilkhanids did not have the option of giving up a career of world conquest, since they had never been strong enough even to begin one. So to remove the theme to a safely literary plane would have an obvious attraction. In this specific and somewhat inglorious context, the valetudinarian quality of the Ša¯hna¯ma text as it nears the end of Iskandar’s life would acquire unexpectedly felicitous resonances in the context of contemporary politics, in which the Ilkhanids were reduced to petty squabbles with the Mamluks on matters of protocol and prestige. The example of Iskandar renouncing war and conquest would furnish an excellent excuse for the relative inactivity of the later Ilkhanids in military matters – a subtle and flattering manipulation of the Ša¯hna¯ma text. What no-one could have foreseen – and this is the richest irony of all – is that the premature death of Abu¯ Sac#ıd himself was, like that of Iskandar, just round the corner. Unfortunately, the Mongol sultan had no talking tree to give him decent warning of his end.

38

I am grateful to Professor Kugler for a suggestion pointing me in this direction. Cf. also Carey 130–131. Hence the assertion of the Dominican missionary Ricoldo di Monte Croce that the very name Mongol derives from “Magogli,” and that this explains why the Tatars cannot abide any reference to Alexander (Saunders 223).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 4635B35A /production/pindar/booksxml/hillen/chapter−03

96 A third point of contact between the Iskandar story and the Ilkhanid dynasty is the theme of religion. As noted earlier, Firdawsı# presents Iskandar as a Christian, and of course several Ilkhans professed that faith or took part in Christian services (Spuler 214–219; Bausani 541; Boyle 1968, 379). But Iskandar also visits the Kacba – which at the time the manuscript was being planned might well have brought to mind the Ilkhanid attempt to seize the Holy Cities of Arabia in 1317 (Blair 1987, 71–73) – and the mere mention of the Kacba is enough to attest Muslim loyalties. Indeed, this direct reference to Islam is at odds with the determinedly pre-Islamic setting of the entire epic, as is the fact that Iskandar, alone among Ša¯hna¯ma monarchs, is permitted to converse with Isra¯fil, the Angel of the Last Trump (Warner and Warner 6, 161–162). Firdawsı# himself is even more inconsistent in this matter, since the later sections of the Ša¯hna¯ma – which no doubt reflect lost Sasanian sources – betray a settled hostility to Alexander as a killer of legitimate Iranian kings.39 At all events, the implied change of faith to Islam is extraordinarily apposite in view of the actual practice of Ilkhanid rulers, who sampled, as if in a restaurant, most of the major religious persuasions of their time.40 The early Ilkhans were by turn pagan, Christian and Buddhist; and although after ˙ a¯za¯n’s conversion they remained Muslim, the sympathies of his successor G ¯ lgˇa¯ytu¯, for example, veered first between different schools of Sunni U orthodoxy and later turned to Shicism.41 Sufi teachings also exerted their appeal, and indeed the Illuminationist school of Suhrawardı#, which flourished in the 13th and 14th centuries, developed an elaborate Sufi interpretation of the Ša¯hna¯ma.42

39

Warner and Warner vol. I, 59, 61, 63; Warner and Warner vol. 6, 224, 240. Cf. Darmesteter 1883, 241, 243–244. Niz.a¯mı# also mentions his destruction of fire-temples (ibid., 245). 40 For Mongol beliefs in general, see Morgan 40–44; Spuler 198–249; Bausani 538–549. 41 Spuler 243–244; Boyle 1968, 401–402; Bausani 543–544; Blair 1987, 70. For the specific numismatic evidence, see Blair 1982, 212 and Blair 1983, 297. For the evidence of monumental inscriptions, see van Berchem 1909, 680–681. 42 Melikian-Chirvani 1984, 312–321. It is even suggested that the Mongol sultan was presented at Taht-i Sulayma¯n as the Master of the Initiated (321–323), though it remains an ˘ much he understood of these subtleties. Melikian-Chirvani 1988, 46–49, open question how suggests that such Sufi undertones were intentionally included in the text by Firdawsı#.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 45F68533 /production/pindar/booksxml/hillen/chapter−03

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

97

It is precisely this openness to varied forms of religious experience that motivates Iskandar’s unremitting, indeed desperate, search for the meaning of life.43 In the context of the conversion of the Mongol élite to Islam a mere generation earlier – an event too recent to have allowed that élite to strike deep roots into the Muslim faith (Bausani 542–544; Boyle 1968, 378–380 et passim) – one may note as further significant factors the broad-brush nature of Iskandar’s concept of the divine and the lack of specifically Islamic religious colouring. To interpret the experiences he undergoes in the process of spiritual enlightenment merely as colourful adventures is to downgrade the high seriousness of his quest. Iskandar’s opponents, notably Queen Qaida¯fa of Andalusia and later the Fag˙fu¯r of China, teach him humility and lead him to accept that his power comes from God. Only by degrees does he establish that control over himself which is the necessary preliminary to his becoming both a just king and a true Sufi (Melikian-Chirvani 1984, 309). There is a dimension of personal growth, too, an affecting acceptance of solitude and vulnerability. His penchant for disguising himself as his own ambassador on his visits to rival monarchs is met with repeated failure; his military triumphs are constantly counterpointed by oracular utterances from those whom his power cannot suborn – exotic sages, philosophers, angels or even talking trees, all harping on the inevitability of his death and his need to give all glory to God alone. Thus by degrees Iskandar is transformed into an apostle of monotheism. Here too there was a message for the Mongol ruler, if he had ears to hear it. Other Sufi elements could be cited. The riddle that confronts Iskandar’s troops in the Land of Gloom (pl. 6) – should they pick up the stones in the darkness, since they are warned that unhappiness will be their destiny whether they do so or not? – has the typically paradoxical quality of a Sufi parable. The magical element in this tale, incidentally, can be matched by numerous other episodes in the Iskandar story,44 and a good many of these are illustrated in the Great Mongol Ša¯hna¯ma: the talking tree (pl. 7), the fabulous beasts which Iskandar slays, the fire-breathing cavalry which he

43 Illustrated in the Great Mongol Ša¯hna¯ma by the scenes of Iskandar with the Brahmins (pl. 3); with the talking birds; in the Land of Gloom (pl. 6); and before the Talking Tree (pl. 2; see Grabar and Blair pls. 32, 35, 36 and 38). 44 Warner and Warner vol. 6, 147–148, 150, 152–153, 157, 161–162, 166 and 177. ˘ am˘s#ıd (Christensen Zoroastrian tradition also associates him with the magical cup of G 1934, 131–132 and Zick-Nissen).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 60 OUTPUT: Mon Feb 13 11:54:04 2012 SUM: 415FFA27 /production/pindar/booksxml/hillen/chapter−03

98 unleashes against King Porus,45 and the speaking birds. These magical scenes account for half the extant images in this cycle, and are thus clearly a vital element in the way that it was conceived. Here again it is possible to suggest a connection with the cast of mind displayed by successive Ilkhanid rulers. They were noted for their superstition and for their dependence on astrology (Melikian-Chirvani 1988, 34–35 and Melikian-Chirvani 1991, 34–35, 76–78). They were apt to demand impossible feats of their astrologers and have them executed when they failed to give satisfaction (Spuler 193). The reliance of their immediate forebears on the pseudo-magical practices of shamanism is recorded in detail by Western friars like Rubruck and Carpini, and it is likely enough that vestiges of this system of beliefs and rites survived into the 14th century (Warner and Warner 6, 161–162). And one might add that for a dynasty whose conquerors had seen all that there was to see in the sublunary world, the supernatural was all that was left to wonder at. But the most surprising link of all between Alexander the Great and the Ilkhanids has nothing whatever to do with the Ša¯hna¯ma, even though that link may well have provided the unspoken ideological foundation for the extraordinary visual emphasis on the Iskandar cycle in this particular manuscript. This link hinges on the absurd fiction that the Chingizid line was descended from Olympias, the historical mother of Alexander of Macedon. Incredible as it may sound, an array of sources – al- cUmarı¯ (Herzfeld 318), Abu’ 1-Fad.l (Skelton 1988, 180), the lengthy inscription on Tı˜mu¯r’s tombstone,46 Mirhwa¯nd (Herzfeld 318) and cAlı# Yazdı# (ibid., 319) as well as a ˘ the Yuan-shao-Pi-shea (Herzfeld 319–320) – take a tale Chinese chronicle, borrowed from Plutarch about the miraculous birth of Alexander and apply it (changing the name of Olympias to Alamfiya or Alanquwa)47 to the ancestor of Chinggis Ha¯n, one of the Nurani or “light-beams”; and y Chinggis himself was termed for that reason the Son of the Sun (Herzfeld 319). In the version of Abu’ l-Fad.l, the story runs as follows:

45

I am grateful to Teresa Fitzherbert for the demonstration that this image derives from Tibetan sources. 46 Herzfeld 318. Note too that Ha¯tifi’s Tı#mu¯rna¯ma associated Tı¯mu¯r with Iskandar (Lentz and Lowry 1988, 284, 357–358). 47 For the process of transference see Nöldeke 11 and Herzfeld 324.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03 Sequential 20

JOBNAME: ???? PAGE: 1 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 0A87AF79 /production/pindar/booksxml/hillen/chapter−03−artwork

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

99

1. The Great Mongol Sˇa¯hna¯ma: The Bier of Iskandar (after Gray)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 073F46D1 /production/pindar/booksxml/hillen/chapter−03−artwork

100

2. The Great Mongol Sˇa¯hna¯ma: Iskandar builds the wall against Gog and Magog (after Lowry and Beach)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 07D5EE7C /production/pindar/booksxml/hillen/chapter−03−artwork

101

3. The Great Mongol Sˇa¯hna¯ma: Iskandar visits the Brahmins (after Lowry and Beach)

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 090B7038 /production/pindar/booksxml/hillen/chapter−03−artwork

102

4. The Great Mongol Sˇa¯hna¯ma: Iskandar and the Talking Tree (after Brian)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 06BC1FF6 /production/pindar/booksxml/hillen/chapter−03−artwork

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

103

5. The Great Mongol Sˇa¯hna¯ma: Iskandar enthroned (after Brian)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 25 OUTPUT: Tue Aug 10 10:24:29 2010 SUM: 05F5EDCD /production/pindar/booksxml/hillen/chapter−03−artwork

6. The Great Mongol Sˇa¯hna¯ma: Iskandar enters the Land of Gloom (after Robinson)

104

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-artwork Sequential 6

JOBNAME: ???? PAGE: 1 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 5206A259 /production/pindar/booksxml/hillen/chapter−03−ends

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

105

One night a glorious light cast a ray into the tent and entered the mouth and throat of that fount of spiritual knowledge and glory [i.e. Alanquwa]. The cupola of chastity became pregnant by that light as did her Majesty H.azrat Mariyam [= the Virgin Mary].48 No doubt the implied connection with Christianity fell upon receptive ears, given the accommodating religious policy of the Mongols. From Chinggis in turn descended the Mongol Ilkhans of Iran, including Abu¯ Sa c#ıd. And a whole clutch of later Islamic rulers joined the party. They include the Mughal emperor Akbar (Skelton 1988, 180) and – as already noted – Tı#mu¯r, who for good measure also claimed descent from the Prophet’s son-in-law cAlı# (Herzfeld 322–323). The Ottoman sultan Ba¯yazı#d I claimed to be descended from Alexander and in 1396 took carpets with Alexander scenes as a ransom for John, son of the Duke of Burgundy.49 Mehmed II, the conqueror of Constantinople, had the Alexander Romance read to him regularly so as to learn how to conquer the world,50 and was routinely compared to Alexander.51 One of his successors, Selim I, wrote to the Sultan

48

Abu’ l-Fad.l, Akbarna¯ma I: 180, cited by Skelton 1988, 180 and 187, n. 33. This is much the same story as told by the inscription on Tı#mu¯r’s tombstone. 49 Froissart 237 reads “he sayde he wolde reygne lyke Alysaunder of Masydone, who was twelve yere kynge of all the worlde, of whose lynage he sayde he was dysceneded. All that herde him agreed to his sayenge,” and on p. 244 we read that “Lamorabaquy [Ba¯yazı#d I] toke gret pleasure in clothes of arras, made of olde auncyent hystories”; he therefore sent as a partial ransom for John, Earl of Nevers, the son of the Duke of Burgundy, “clothes of arras, of the goodlyest that coude be gote, wrought with the storie of Alexandre, of his lyfe and conquestes, ryght pleasount to beholde” (245). 50 Babinger 1951, 139, quoting Jacopo Tedaldi. Phrantzes notes that he also read about the deeds of Augustus, Constantine the Great and Theodosius (ibid., 140). Niccolo Sagundino records that he had the deeds of Alexander and Caesar translated into his own language, in quibus legendis vel audiendis mirum in modum delectatur (ibid., 138). Ferducci adds Cyrus to the list (ibid., 140–141). Jacopo Languschi quotes Mehmed as saying that Alexander had marched into Asia with a small force, whereas now times had changed and the East was coming to the West (ibid., 140). Mehmed had advisors conversant with Latin and Greek who made him familiar with the contents of classical texts, and there was even a copy of Arrian’s Life of Alexander in his library (ibid., 141–2). 51 Thus Kritovoulos, admittedly in his official history of the high points of his patron’s reign, tells Mehmed that his accomplishments are “in no way inferior to those of Alexander the Macedonian” (3) and notes how Mehmed was resolved to overrun the whole world “in emulation of the Alexanders and Pompeys and Caesars and kings and generals of their sort” (14). Geoffrey of Rhodes says in a letter that after 1453 Mehmed sought to excell

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 4B7B36A3 /production/pindar/booksxml/hillen/chapter−03−ends

106 of Egypt that he planned to become ruler of east and west like Alexander the Great (Inalcik 315). Sult.a¯n H.usayn Bayqara, the last Tı#mu¯rid sultan of Herat, had his portrait painted in the guise of Alexander seated amidst the Seven Sages (Brend, frontispiece), and many a Muslim ruler, no matter whether his status was exalted or modest, styled himself “The Second Alexander,”52 “Dhu l-Qarnain” or “The Alexander of the Time.”53 No doubt Olympias, however bewildered, would have been proud of them all. Works Cited Abel, Armand. “Iskandar Na¯ma.” Encyclopaedia of Islam. 2nd ed. Leiden, 1978. IV, 127b-128b. __________. “Dhu’l-Qarnayn, Prophète de l’Universalité.” Annuaire de l’Institut de Philologie et d’Histoire Orientales et Slaves 11 (1951). 6–18. __________. “La Figure d’Alexandre en Iran.” Accademia Nazionale dei Lincei. Convegno sul tema La Persia e il mondo greco-romano. Rome, 1966. 120–134. Anon. “Al-Iskandar.” Encyclopaedia of Islam. 1st ed. Leiden, 1921. Cols. 533b-534b. Babinger, Franz. “Mehmed II., der Eroberer, und Italien.” Byzantion 21 (1951). 127–170. __________. Mehmed the Conqueror and his Time. Trans. Ralph Mannheim. Ed. William C. Hickman. Princeton, 1978. Bausani, Alessandro. “Religion under the Mongols.” The Cambridge History of Iran 5. The Saljuq and Mongol Periods. Ed. John A. Boyle. Cambridge, 1968. 538–549. Alexander (Werner 295). The Italian Zorzo Dolfin says of Mehmed, in his Cronaca, that he is “as eager for fame as Alexander of Macedon” (Babinger 1978, 112). Michael Dukas, who attended Mehmed the Conqueror on occasion, refers to him as a new Alexander (Werner 295, citing Phrantzes, Chronicon). 52 E.g. the Inju sultan Šaraf al-Dı#n Mah.mu¯d (Ša¯h) of Fars in 1332–1333 (MelikianChirvani 1982, 148 and Ivanov et al. 21–22 and illustration on p. 52 of Arabic text); and the Mamluk sultans Muh.ammad b. Qala¯ u¯n (Van Berchem 1900, 368), al-Malik al-Ašraf Barsbay (ibid, 367) and, above all, Baibars (Cohen et al. 103, 104–106, 142, nos. 4554, 4556, 4557 and 4612). In the latter case this element is found in his titulature only between 1266 and 1270, and in each case a rhyming formula of the type so common in the 13th century is used: Iskandar al-zama¯n .sa¯.hib al-qira¯n, “Alexander of the time, Lord of the favourable conjunction.” 53 Such as a minor princeling of Sultanate India, Nus.rat Sha¯h in far-off Bengal (Skelton 1978). The Danishmendid amı#r al-wa¯thiq (1152–1161), whose coins even bore Greek inscriptions alongside Arabic ones (Vryonis 474), styled himself “Dhu l-Qarnain.”

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 40A60581 /production/pindar/booksxml/hillen/chapter−03−ends

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

107

Blair, Sheila S. “The Coins of the Later Ilkha¯nids: Mint Organization, Regionalization, and Urbanism.” American Numismatic Society Museum Notes 27 (1982). 211–230. __________. “The Coins of the Later Ilkhanids: A Typological Analysis.” Journal of the Economic and Social History of the Orient 26 (1983). 295–317. __________. “The Epigraphic Program of the Tomb of Uljaytu at Sultaniyya: Meaning in Mongol Architecture.” Islamic Art 2 (1987). 43–96. __________. “On the track of the ‘Demotte’ Sha¯hna¯ma manuscript.” Les Manuscrits du Moyen-Orient. Essais de codicologie et de paléographie. Actes du Colloque d’Istanbul, (Istanbul, 26–29 Mai 1986). Ed. François Déroche. Paris, 1989. 125–131. Boyle, John A. “Juvaynı# and Rashı#d al-Dı#n as Sources on the History of the Mongols.” Historians of the Middle East. Eds. Bernard Lewis and Peter M. Holt. London, 1962. 133–137. __________. “Dynastic and Political History of the I¯l-Kha¯ns.” The Cambridge History of Iran 5. The Saljuq and Mongol Periods. Ed. John A. Boyle. Cambridge, 1968. 303–421. __________. “The Significance of the Ja¯mi c al-Tawa¯rı#kh as a source on Mongol history.” Iran-Shenasi 2 (1970). 1–8. __________. “Rashı#d al-Dı#n: the First World Historian.” Iran 9 (1971). 19–26. __________. “The Alexander Legend in Central Asia”. Folklore 85 (1974). 217–228. Brend, Barbara C. “Akbar’s Khamsah of Amı#r Khusrau Dihlavı#. A Reconstruction of the Cycle of Illustration.” Artibus Asiae 49 (1989). 281–315. Browne, Edward G. A Literary History of Persia. From the Earliest Times Until Firdawsí. Cambridge, 1928. Vol. 1 of 4 vols. 1928–1934. Budge, E. A. Wallis. The Life and Exploits of Alexander the Great. Being a series of Ethiopic texts edited from manuscripts in the British Museum and the Bibliothèque Nationale, Paris, with an English translation and notes. 2. The English Translation. London, 1896. Bürgel, J. Christoph. Nizami. Das Alexanderbuch. Iskandarname. Persian Heritage Series 37. Zürich, 1991. Carey, George. The Medieval Alexander. Cambridge, 1956. Christensen, Arthur. Les types du premier homme et du premier roi dans l’histoire légendaire des Iraniens. I. Stockholm, 1918; II. Leiden, 1934.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 40D7BF60 /production/pindar/booksxml/hillen/chapter−03−ends

108 Cohen, M. et al. Répertoire chronologique d’épigraphie arabe. Cairo, 1943. Vol. 12 of 16 vols. 1932–1964. Combe, Etienne et al. Répertoire chronologique d’épigraphie arabe. Cairo, 1954. Vol. 14 of 16 vols. 1932–1964. Darmesteter, James. “La légende d’Alexandre chez les Parses.” Essais Orientaux par James Darmesteter. Paris, 1883. 227–250. __________. “Alexandre le Grand dans le Zend Avesta.” Revue des Etudes Grecques 5 (1892). 189–196. Dodkhudoeva, Larissa N. Poemy Nizami v srednevekovoi miniatiurnoi zhivopisi. Moscow, 1985. Friedländer, Israel. Die Chadhirlegende und der Alexanderroman. Leipzig, 1913. Froissart, Sir John. The Chronicle of Froissart. Trans. Sir John Bourchier, Lord Berners. Vol. 6. London, 1903. Gibb, Edward J.W. A History of Ottoman Poetry. London, 1900. Vol. 1 of 6 vols. 1900–1909. Grabar, Oleg and Sheila S. Blair. Epic Images and Contemporary History. The Illustrations of the Great Mongol Shahnama. Chicago and London, 1980. Gray, Basil. “Sha¯hna¯ma illustration from Firdausı# to the Mongol invasions.” The Art of the Salju¯qs in Iran and Anatolia. Proceedings of a Symposium held in Edinburgh in 1982. Ed. Robert Hillenbrand. Costa Mesa, 1994. 96–105. Guest, Rhuvon. “Al-Iskandariya.” Encyclopaedia of Islam. 1st ed. Leiden, 1921. Hanaway, William H. “Alexander and the Question of Iranian Identity.” Iranica Varia: Papers in Honor of Professor Ehsan Yarshater. Textes et Mémoires, 16. Acta Iranica 30. Leiden, 1990. Herzfeld, Ernst E. “Alongoa.” Der Islam 6 (1916). 317–327. Inalcik, Halil. “The Rise of the Ottoman Empire.” The Cambridge History of Islam. I. The Central Islamic Lands. Eds. Peter M. Holt, Ann K.S. Lambton and Bernard Lewis. Cambridge, 1970. 295–323. Ivanov, Anatoly A. et al. Masterpieces Of Islamic Art In The Hermitage Museum. Kuwait, 1990. Jackson, Peter. “Abu¯ Sa c#ıd.” Encyclopaedia Iranica. 1.4. Ed. Ehsan Yarshater. London, 1983. Cols. 374b-377a. Juvainı¯, cAt.a¯-Malik. The History of the World-Conqueror. Trans. John A. Boyle. 2 vols. Manchester, 1958. Kritovoulos. History of Mehmed the Conqueror. Trans. Charles T. Riggs. Princeton, 1954.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 44E56AE1 /production/pindar/booksxml/hillen/chapter−03−ends

THE ISKANDAR CYCLE IN THE GREAT MONGOL {SA z HNA z MA

109

Lentz, Thomas W. and Glenn D. Lowry. Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century. Los Angeles and Washington, 1989. al-Mas cu¯di. Le Livre de l’Avertissement et de la Révision. Trans. Carra de Vaux. Paris, 1896. Melikian-Chirvani, Assadullah S. Islamic Metalwork from the Iranian World, 8–18th Centuries. London, 1982. __________. “Le Sha¯h-Na¯me, la gnose soufie et le pouvoir mongol.” Journal Asiatique 272 (1984). 249–337. __________. “Le Livre des Rois, miroir du destin.” Stadia Iranica 17 (1988). 7–46. __________. “Le Livre des Rois, miroir du destin. II – Takht-e Soleyma¯n et la symbolique du Sha¯h-Na¯me.” Studia Iranica 20 (1991). 33–148. Morgan, David O. The Mongols. Oxford, 1986. Nöldeke, Theodor. “Beiträge zur Geschichte des Alexanderromans.” Denkschriften der Akademie der Wissenschaften, Wien. Phil.-hist. Klasse 38, Abhandlung 5 (1890). 1–56. __________. The Iranian National Epic or the Sha¯hna¯mah. Tr. Leonid T. Bogdanov. Repr. Philadelphia, 1979. Norgren, Jill and Edward Davis. Preliminary Index of Shah-nameh Illustrations. Ann Arbor, 1969. Pfister, Friedrich. Alexander der Grosse in den Offenbarungen der Griechen, Juden, Mohammedaner und Christen. Berlin, 1956. Pseudo-Callisthenes. The History of Alexander, being the Syriac Version of the Pseudo-Callisthenes. Ed. and tr. E.A. Wallis Budge. Cambridge, 1889. Rashı#d al-Dı#n. The Successors of Genghis Khan. Trans. John A. Boyle. New York and London, 1971. Rice, D. Talbot. The Illustrations to the ‘World History’ of Rashı#d al-Dı#n. Edinburgh, 1976. Robinson, Basil W. A Descriptive Catalogue of the Persian Paintings in the Bodleian Library. Oxford, 1958. Ross, David J.A. Alexander Historiatus. A Guide to Medieval Illustrated Alexander Literature. Warburg Institute Surveys I. Eds. Ernst H. Gombrich and Joseph B. Trapp. London, 1963. Saunders, John J. The History of the Mongol Conquests. London, 1971. Schaeder, Hans H. “Ücber das ‘Bilderbuch der Sasaniden-Könige.’” Jahrbuch der Preussischen Kunstsammlungen 57 (1936). 231–232. Simpson, Marianna S. The Illustration of an Epic. The Earliest Shahnama Manuscripts. New York and London, 1979.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 63 OUTPUT: Mon Feb 13 11:54:08 2012 SUM: 44663190 /production/pindar/booksxml/hillen/chapter−03−ends

110 __________. “The Pattern of Early Sha¯hna¯ma Illustration.” Studia Artium Orientialis et Occidentalis 1 (1982). 43–53. __________. “A Reconstruction and Preliminary Account of the 1341 ¯Inju¯ Sha¯hna¯ma with Some Further Thoughts on Early Sha¯hna¯ma Illustration.” Studies in Persian Painting. Festschrift for Basil Robinson. Ed. Robert Hillenbrand, (in press) Sims, Eleanor G. “The Illustrated Manuscripts of Firdausı#’s Sha¯hna¯ma Commissioned by Princes of the House of Timu¯r.” Ars Orientalis 22 (1993). 43–68. Skelton, Robert. “A Royal Sultanate manuscript dated 938 A.H. 1531–2 A.D.”, in Indian Painting. Mughal and Rajput and a Sultanate manuscript. London, 1978. 133–152, 155–157. __________. “Imperial Symbolism in Mughal Painting.” Content and Context of Visual Arts in the Islamic World. Papers from a Colloquium in Memory of Richard Ettinghausen. Institute of Fine Arts, New York University, 2–4 April 1980. Planned and Organized by Carol Manson Bier. Ed. Priscilla P. Soucek. University Park and London, 1988. 177–192. Soudavar, Abolala. Art of the Persian Courts: Selections from the Art and History Trust Collection. New York, 1992. Southgate, Minoo S. A Persian Medieval Alexander Romance. New York, 1978. Spiegel, Friedrich von. Die Alexandersage bei den Orientalen. Leipzig, 1851. Spuler, Bertold. Die Mongolen in Iran. Berlin, 1955. al-T.abarı#, Abu¯ Jarı#r. The History of al-Tabari (Ta’rı#kh al-rusul wa’l-mulu¯k) II. Prophets and Patriarchs. Trans. William M. Brinner. Bibliotheca Persica. SUNY Series on Near Eastern Studies. Albany, N.Y., 1987. van Berchem, Max. “Une inscription du sultan mongol Uldjaitu.” Mélanges Hartwig Derenbourg. Recueil de travaux d’érudition dédiés à la mémoire d’Hartwig Derenbourg par ses amis et ses élèves. Paris, 1909. 367–378. Repr. in Max van Berchem. Opera Minora. Geneva, 1978. 673–684. __________. Matériaux pour un Corpus Inscriptionum Arabicarum. Première Partie. Egypte. Fascicule troisième. Le Caire. Mémoires publiés par les Membres de la Mission Archéologique Française au Caire 19. Paris, 1900. Vryonis, Speros, Jr. The Decline of Medieval Hellenism in Asia Minor and the Process of Islamization from the Eleventh through the Fifteenth Century. Berkeley and Los Angeles, 1971.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-03-ends Sequential 6

JOBNAME: ???? PAGE: 1 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 3E52BA4C /production/pindar/booksxml/hillen/chapter−04

IV Images of Muhammad in al-Biruni’s Chronology of Ancient Nations L-BIRUNI, perhaps the single most outstanding Muslim polymath,1 wrote The Chronology of Ancient Nations in c.1000 A.D.2 To judge by surviving evidence, the text was never a favoured vehicle for illustration; among the manuscript versions known,3 only two are illustrated, of which the second and later one (Paris, Bibliothèque Nationale, ms. arabe 1489, datable to the seventeenth century and probably of Ottoman provenance) copies the earlier version in Edinburgh.4 This Edinburgh manuscript (Edinburgh University Library, Ms. Arab 161), dated 707/1307–8 and copied by a certain Ibn al-Kutubi,5 (there is no indication as to where the work was produced)6 has in fact a greater claim to attention than merely the circumstance that it is the earliest illustrated copy of this text (pls. 5–6, 12–14,

A

1 See The Encycopaedia of Islam (1st ed.), s.v. (C. Brockelmann) and The Encyclopaedia of Islam (2nd ed.), s.v. (D.J. Boilot). 2 C.E. Sachau, The Chronology of Ancient Nations. An English version of the Arabic text of the Athar-ul-Bakiya of Albiruni, or “Vestiges of the Past”, collected and reduced to writing by the author in A.H. 390–1, A.D. 1000 (London, 1879), viii. 3 Which number at least five, namely the three used by Sachau (Chronology, xiv) plus the illustrated versions in Edinburgh and Paris (for which see n. 84 below). 4 Sir T.W. Arnold and A. Grohmann, The Islamic Book. A Contribution to its Art and History from the VII-XVIII Century (n.p., 1929), pl. 40. 5 M. Ashraful Hukk, H. Ethé and E. Robertson, A Descriptive Catalogue of the Arabic and Persian Manuscripts in Edinburgh University Library (Hertford, 1925), 136–7. 6 The traditional ascription is to Tabriz; more recently, Dr Stefano Carboni has suggested Mosul (see his paper “The London Qazwini: an early 14th-century copy of the ‘Aja’ib al-Makhluqat”, Islamic Art III [1989], 17).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 540D23AF /production/pindar/booksxml/hillen/chapter−04

112 16), for – with the images in the Edinburgh copy of the World History of Rashid al-Din, dated to 714/1314 (pls. 1–4, 7–11, 15, 17–18),7 which have an extra public dimension – it contains the earliest set of images of the Prophet Muhammad in Islamic art.8 An obvious question to arise at this point is why this long-dead text should have been resuscitated at this particular time, and why it should have had conferred on it the extra éclat of illustrations (pls. 5–6,12–14) and illumination (pl. 16). Here the content of the text is plainly crucial. It deals with the various systems of computing time known to al-Biruni’s world, and backs this up with a complement of astronomical calculations. There were two grounds on which such a text could arouse interest in Ilkhanid Iran. One was the focus on astronomy. The Ilkhan Hulegu had built a state-of-the-art observatory at Maragha,9 and subsequently, under the Ilkhan Abaqa, there was completed in 667/1265 or 670/1271–2 a set of astronomical tables known as the Zij-i Ilkhani.10 Other Ilkhans shared this fascination with the lore of the stars.11 The second focus of the text, equally well calculated to appeal to Mongol taste, was the well-nigh ecumenical scope of al-Biruni’s material. He ranged very widely in time and space, commenting not only on the calendars but also on the history, festivals and customs of many exotically remote nations.12 In his tolerance and open-minded curiosity – which was 7

D. Talbot Rice, The Illustrations to the ‘World History’ of Rashid al-Din (Edinburgh, 1976), 2. 8 Though not the earliest image of him per se; I have discussed this issue in an unpublished paper delivered at a conference on Rashid al-Din held at the Warburg Institute, University of London, and entitled ‘‘Images of the Prophet Muhammad in the World History of Rashid al-Din’’. 9 For the physical remains of this observatory, see P. Vardjavand, “La découverte archéologique du complexe scientifique de l’observatoire de Maraqe”, in Akten des VII. Internationalen Kongresses für Iranische Kunst und Archäologie. München 7.–10. September 1976, ed. W. Kleiss (Berlin, 1979), 527–36. 10 Ibid., 527. 11 Ibid., 528; E.S. Kennedy, “The exact sciences in Iran under the Saljuqs and Mongols”, in The Cambridge History of Iran. 5. The Saljuq and Mongol Periods, ed. J.A. Boyle (Cambridge, 1968), 668–70, 672–3; B. Spuler, Die Mongolen in Iran. Politik, Verwaltung und Kultur der Ilchanzeit 1220–1350 (Berlin, 1955), 192–3. 12 P.P. Soucek, “An Illustrated Manuscript of al-Biruni’s Chronology of Ancient Nations”, in The Scholar and the Saint. Studies in Commemoration of Abu’l-Rayhan al-Biruni and Jalal al-Din Rumi, ed. P.J. Chelkowski (New York, 1975), 103–68. This excellent study contains much information on the miniatures discussed in this paper. I owe much to it, as to the comments of generations of students in Edinburgh with whom I have discussed these images.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4DFF2F2C /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

113

evidenced by his decision to live in India for many years, learn Sanskrit and write a book about the non-Islamic part of the country13 – al-Biruni was centuries ahead of his time. Not for him the instinctive conviction that his own culture, and all that it entailed, was simply the best. And this lack of dogmatism, this willingness to learn about other cultures and other systems of belief, seems to have struck a chord with the unknown Ilkhanid patron of this manuscript. There can be no question that al-Biruni’s text was uncannily relevant in the Ilkhanid period. It presents a remarkable array of systems of belief, and this inevitably brings to mind the farrago of religious practices in Ilkhanid times.14 Nevertheless, since the text itself remained a production of the period c.1000, the task of making it specifically relevant to the fourteenth century devolved upon the person who selected the subjects for illustration, and on the illustrator himself. They could, of course, have been one and the same person. Thus the text and its illustrations are not necessarily at one; indeed, the content of the illustrations can change the emphasis of the text. A prime example of this process is given by the last two images. Admittedly, al-Biruni seems to have had Shi cite sympathies,15 yet that is not enough to account for the spirit of intense religious particularism which animates these two paintings. While it would be too strong to maintain that they put the text and the illustrations virtually at cross purposes, they certainly show how paintings can place in the forefront of attention ideas which are only latent in the text itself, and how this heightening of emphasis can be used to affect the reader. Thus the fact that the pictorial cycle begins and ends with images of Muhammad places the whole book as it were under his protection, and sanctifies its contents. The point of departure of al-Biruni’s book was revolutionary within its own society because the author in some respects implicitly placed Islam on a level with other religions rather than in a category all by itself, while still

13 See the translation by C.E. Sachau: Alberuni’s India (London, 1910). Cf. also A.T. Embree, “Foreign Interpreters of India: The Case of al-Biruni”, in Chelkowski, op. cit., 1–16; F.E. Peters, “Science, History and Religion: Some Reflections on the India of Abu’l-Rayhan al-Biruni”, ibid., 17–27; and B.B. Lawrence, “The Use of Hindu Texts in al-Biruni’s India with Special Reference to Patanjali’s Yoga-Sutras”, ibid., 30 and 44. 14 Spuler, op. cit., 198–249. 15 Sachau, Chronology, xiii says “he inclined towards the Shi‘a”, and the book ends with the author calling down the mercy and blessing of God on the Prophet “and upon his holy family” (ibid., 365).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4C0487B9 /production/pindar/booksxml/hillen/chapter−04

114 writing as a Muslim.16 The Edinburgh manuscript drives home this point by means of its illustrations. This is not to say, of course, that the mindset of al-Biruni was that of the Mongols. He displays a genuinely broad mind; his tolerance is based on wide knowledge and careful observation. That of the Mongols was based, it seems, on a profound lack of commitment to any single body of religious dogma. Nevertheless, despite these fundamental differences of approach, a freak of history saw to it that al-Biruni found his real audience for this book some three centuries after his death and among people who had done much to destroy the culture he had helped to build. The pervasive emphasis on false prophets, and on false religion generally, attains such prominence in the paintings that it must be recognised as one of the driving forces behind the entire cycle of illustrations. The images of Muhammad fit well into this programme, for they include scenes where he is set against representatives of other systems of belief and is shown to be superior to them. To that extent there is a proselytising flavour to the images of Muhammad; but there can be no doubt that Jesus and the prophets of the Old Testament are also objects of veneration in these paintings.17 This particular emphasis, which of course recurs in the World History of Rashid al-Din, corresponds very closely to the spirit of the generation before c.1300, which saw Christians and Jews rise to positions of power and influence unparalleled in the earlier history of Iran.18 The religious tolerance of the Mongol elite, and their readiness to espouse a variety of faiths, helps to explain this ecumenical spirit, while the conversion to Islam of the Ilkhan Ghazan and his amirs19 gives an apt context for the unprecedented emphasis on Muhammad. But it was precisely in a milieu of such religious ferment that accounts of some of the discredited sects and systems of belief in past ages or in distant lands would arouse a lively interest. Hence the number of illustrations devoted to false prophets; and orthodoxy could be appeased by sandwiching these images between pictures asserting the supremacy of Muhammad and thus of Islam. Thus, seen as a whole, the 25 illustrations of 16

Embree, op. cit., 2, 4–8; cf. Peters, op. cit., 21 and 25 and Lawrence, op. cit., 30. T.W. Arnold, The Old and New Testaments in Muslim Religious Art (London, 1932), passim but especially 15, 17, 22 and 35–6. 18 Spuler, op. cit., 207–10, 212–8 and 246–9. 19 J.A. Boyle, “Dynastic and Political History of the Il-Khans”, in CHI 5, 378–80; A. Bausani, “Religion under the Mongols”, ibid., 541–3; and, above all, C. Melville, “Padshah-i Islam: The Conversion of Sultan Mahmud Ghazan Khan”, in Pembroke Papers I. Persian and Islamic Studies In honour of P. W. Avery, ed. C. Melville (Cambridge, 1990), 159–77. 17

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4CC17D41 /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

115

the Edinburgh al-Biruni manuscript propel to the forefront of the reader’s attention themes that are only latent in the text itself. The overwhelmingly scientific, and specifically astronomical, emphasis of al-Biruni’s book has been ousted by themes more in sympathy with the religious trends of the Ilkhanid realm around 1300. In that sense the text has been turned upside down and furnishes a striking example of how a cycle of pictures can go much further than merely illustrating the major themes of the accompanying text, or even providing a sub-plot; they can give it an entirely new and unexpected twist, rendering it relevant for contemporary concerns. Clearly, then, this text is an unlikely one to serve as a basis for a set of images of Muhammad. The sine qua non for any iconographic cycle, namely continuity, is simply not there. The images showing Muhammad, though totalling only 5 out of the 25 in the book, span the entire text, from beginning to end. Thus the whole book, and its illustrations, is presented within an overall Islamic context. But even so, it is more appropriate to speak of a set of images rather than of a cycle. Such a cycle would have had to be fashioned in the teeth of the text’s undeclared opposition. Nor does this group of images respect any chronological continuity; indeed, the very first scene, which depicts the Prophet forbidding intercalation (pl. 5), renders an event which occurred at the very end of his life, during his farewell pilgrimage.20 It would be more accurate to explain the choice of images as the result of the desire of the painter (or whoever else dictated the selection of episodes for illustration) to pounce on any event which involved the Prophet. Since the life of the Prophet was very far from being the concern of al-Biruni in writing his book, there were very few such events. On occasion even a relatively insignificant episode is chosen to exalt the Prophet, as in the case of the watchman and the two riders (pl. 6). That this somewhat undramatic event rated a painting points to the desire to use the text as a vehicle for images of the Prophet, and it seems that few opportunities were missed.21 There is no evidence that any attempt was made to impose unity on the scenes from his life that were selected for illustration. That would have 20

A. Guillaume, The Life of Muhammad. A translation of Ibn Ishaq’s Sirat Rasul Allah (repr. Karachi, 1980), 620. 21 Al-Biruni interprets Deuteronomy xxxiii:2 as referring to Muhammad rather than to Jehovah (Sachau, Chronology, 23); gives the lineage of Muhammad (ibid., 46); describes the efforts of “those Muslims who try to derive mystical wisdom from the comparison of the name of Muhammad … with the human figure. According to them the Mim is like his head, the Ha like his body, the second Mim like his belly, and the Dal like his two feet” (ibid., 293);

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 48EE3D25 /production/pindar/booksxml/hillen/chapter−04

116 demanded at least some rearrangement of the text. As it is, these scenes plainly have no inherent unifying theme apart from the person of Muhammad. In short, no-one seriously attempting to create a connected cycle of images of Muhammad would have chosen either this particular combination of scenes or their present order. Hence these images can be called a cycle by default. The images of Muhammad in the World History of Rashid al-Din (pls. 4, 7, 9–11, 15)22 are far more carefully selected to tell a continuous story and they focus on some of the major events of Muhammad’s life; but they have the advantage, unlike al-Biruni’s text, of being based on a coherent historical narrative which focusses closely on his life and respects the chronology of events. In no sense, then, can this be regarded as a cycle of deliberately religious images, like the Siyer-i Nebi for instance.23 After all, this is emphatically a secular manuscript. Hence any orthodox opposition to these images – assuming that such opposition voiced itself in the Ilkhanid period – could have been deflected by the argument that there was no question of these images having any religious function at all, least of all worship. In both these books, then, the religious image enters as it were by the back door: under cover of history in the case of the text of Rashid al-Din, and of calendrical issues in the case of al-Biruni’s text. Once these images had taken shape, the way was open for a genuinely religious cycle, and the fragmentary Istanbul Mi crajnama of c.135024 shows that it was not long in coming.

and refers to a pair of white stones in the Mosque at Jerusalem with inscriptions mentioning Muhammad (ibid., 294). None of these four references to Muhammad lend themselves to depiction. 22 Pending the appearance of the article mentioned in note 8 above, see Talbot Rice, op. cit., 95–113; B. Gray, The World History of Rashid al-Din. A Study of the Royal Asiatic Society Manuscript (London, 1978), 24–5; and P.P. Soucek, “The Life of the Prophet: Illustrated Versions”, in Content and Context of Visual Arts in the Islamic World. Papers from a Colloquium in Memory of Richard Ettinghausen. Institute of Fine Arts, New York University, 2–4 April 1980. Planned and organized by Carol Manson Bier, ed. P.P. Soucek (Philadelphia and London, 1988), 199–206, prefaced by a brief discussion of images of the Prophet and ‘Ali in the Edinburgh Biruni manuscript on pp. 198–9. Most recently, see S.S. Blair, A Compendium of Chronicles. Rashid al-Din’s Illustrated History of the World (London and Oxford, 1995), 69–72. 23 Z. Tanındı, Siyer-i Nebi, Islam Tasvir Sanatında Hz Muhammed’in Hayatı (Istanbul, 1984). 24 R. Ettinghausen, “Persian ascension miniatures of the fourteenth century”, Accademia Nazionale dei Lincei, XII Convegno “Volta”, promosso dalla classe di scienze morali, storiche

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4AC2EC46 /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

117

Since the guiding thread between the images of Muhammad is not absolute chronology but the sequence dictated by the text, this is the most convenient order to follow in discussing the paintings. The first picture (pl. 5), on f.6b (dimensions: 67 by 134 mm.) depicts the Prophet forbidding intercalation.25 This may sound an abstruse subject, and it certainly does not lend itself naturally to illustration; but for an Islamic author, the Prophet’s own views on time as they affected Muslims were a natural place to begin a disquisition on that subject. Hence, perhaps, the choice of this scene for the first painting in the book. Its immediate context is the controversy about observing a sacred month. Some argued that it could be postponed in a given year and then reinstated in a later year. In the Sira’s version of Muhammad’s last address to his people – in which he ranged widely over the practices to be followed by pious Muslims – Muhammad is reported as saying “Time has completed its cycle and is as it was on the day that God created the heavens and the earth. The number of months with God is twelve; four of them are sacred …”26 This is a more circumstantial account than that given by al-Biruni. In the same section, according to the Sira, he quoted Sura 9:37 in which postponement of a sacred month and making up the total later is roundly condemned. The importance of Muhammad’s stand on this matter may be gauged from the fact that two of the Five Pillars of Islam – fasting and pilgrimage – are involved in this ruling. Intercalation, then, was a practice with very serious repercussions27 and justified a pronouncement by the Prophet ex cathedra. Hence the location of the scene in a mosque setting and the presence of a minbar, whereas according to tradition the event occurred in the open air and Muhammad was seated on his camel.28 Thanks to the plain white ground, the scene acquires a certain universality, as is indeed fully appropriate for an image dealing with the enunciation of doctrine. Here the minbar adds – and neatly blends –

e filologiche. Tema: Oriente e Occidente nel Medioevo (Rome, 1957), 360–83, repr. in Richard Ettinghausen: Islamic Art and Archaeology. Collected Papers, ed. M. Rosen-Ayalon (Berlin, 1984), 244–67. 25 Sachau, Chronology, 14. 26 Ibn Ishaq, 651; The History of al-Tabari (Ta’rikh al-rusul wa’l-muluk). Volume IX. The Last Years of the Prophet, tr. I.K. Poonawala (Albany, 1990), 112–3. 27 One may compare the reaction of a British mob to an eighteenth-century attempt at calendar reform: they rioted, shouting “Give us back our eleven days!” 28 T. Andrae, Mohammed. The man and his Faith, tr. T. Menzel (New York, Hagerstown, San Francisco and London, 1960), 170.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 50E430B4 /production/pindar/booksxml/hillen/chapter−04

118 secular and religious authority, the latter aspect reinforced by the mosque environment. That environment is also suggested by the hanging lamp, which could also be interpreted as a symbol of religious enlightenment.29 It certainly sanctifies the scene. As for the minbar, its form is unmistakably anachronistic, for it has much more in common with the developed ‘Abbasid type as at Qairawan30 than with the version known in the seventh century, of which Umayyad coins seem to preserve an echo.31 Muhammad is shown grasping one of the posts,32 painted red, at the top of the minbar, perhaps in order to support himself since at this time he was close to his final illness. The source of this image is presumably one of the series of enthroned qadis, khatibs and rulers which are so frequently encountered in Mesopotamian painting of the thirteenth century, especially in Maqamat scenes.33 Such images were probably part of this artist’s own training and background. The artist contents himself with isolating Muhammad from the rest of the congregation (again as in Maqamat scenes), and although he depicts the Prophet as slightly bigger than the men listening to him, he does not distinguish him further by giving him a special head-dress, colour or other clothing. All the figures have haloes, thereby depriving this feature of any meaning, and four wear the same turban as Muhammad. Indeed, this painting is full of information about turban fashions and it suggests that certain hierarchical principles were followed in this item of dress.34 Thus the turbans of the two youngest figures have no loops under the chin, which suggests that the loop is an adult fashion, and the central cross-band of the turban is undecorated, whereas the turbans of the older men, including the tailpiece, have panels decorated with a lozenge grid.35 Two other figures have 29 A. Shalem, “Fountains of Light: The Meaning of Medieval Islamic Rock Crystal Lamps,” Muqamas 11 (1994), 5–8, 10–11. 30 K.A.C. Creswell, Early Muslim Architecture. II. Early ‘Abbasids, Umayyads of Cordova, Aghlabids, Tulunids, and Samanids. A.D. 751–905 (Oxford, 1940), pl.89. 31 B. Spuler and J. Sourdel-Thomine, Die Kunst des Islam (Berlin, 1973), pl.71b. For the early history of the minbar, see C.H. Becker, “Die Kanzel im Kultus des alten Islam”, Islamstudien (Leipzig, 1924), 450–71. 32 Such posts can be seen on the minbar at Qairawan. 33 See R. Ettinghausen, Arab Painting (Geneva, 1962), 106–7. The iconography continued in use after this period; one may cite the Epics of 1397 in the British Library – Or. 2780, f.61a – where Chingis Khan, holding a sword, replaces Muhammad (see I. Stchoukine, Les Peintures des Manuscrits Timurides [Paris, 1954], pl.XII). 34 Cf. L.A. Mayer, Mamluk Costume (Geneva, 1952) 13, 15–17, 30–1, 49–54 and 65–7. 35 For information on the turban, see W. Björkman, EI (1st ed.), s.v.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4AB6058F /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

119

a different type of turban with a central red conical cap round which the folds of the material are wound. Yet in the scene depicting the Prophet’s family with the envoys of Musailima, Hasan and Husain do have the loop under their chins even though they are both clearly children. The tight, expressive grouping of the figures, the simple and dramatic spatial relationships – fully legible even though they are technically all awry – owe much to the formulae worked out during the previous century in the illustrated Maqamat manuscripts.36 The next scene to be illustrated (f.10b; dimensions: 87 by 133 mm.),37 though not strictly about calendrical systems, takes up what might be called the other leitmotif of al-Biruni’s text: false religion (pl. 6). Ostensibly it is directed against Babylon and its idols, for it illustrates a passage in Isaiah 21:6–9 in which a watchman hears the news that “Babylon has fallen, has fallen! All the images of its gods lie shattered on the ground!” In the Biblical version, this message is given by “a man in a chariot with a team of horses” (Isaiah 21:9). The artist here, however, has preferred to depict an earlier verse, in which the watchman is admonished “when he sees chariots with teams of horses, riders on donkeys or riders on camels, let him be alert, fully alert” (Isaiah 21:7). Even so, he has removed any visual reference to chariots, or to multiple riders, and has compressed the image into one rider on a camel and one on a donkey. Thus the emphasis in the Biblical account, which is on the destruction of Babylon’s idols,38 has been subtly altered to stress the pre-eminence of the man on the camel, and, more generally, the contrast between him and the rider on the donkey. The complexities of numerical calculation and the symbolism of numbers which characterise al-Biruni’s text39 would of course have appealed to the author’s mathematical, astronomical, astrological and generally scientific interests, but they emphatically do not lend themselves to visual depiction. This is not enough,

36 E.g. Ettinghausen, AP, 113, 118—9. Cf. D. James, “Space-forms in the work of the Baghdad Maqamat illustrators, 1225–58”, Bulletin of the School of Oriental and African Studies XXXVII/2 (1974), 305–20. 37 Sachau, Chronology, 22. 38 And the artist could perfectly well have tackled such a scene, for there is in fact a scene in the Biruni manuscript in which Abraham destroys the idols of the Sabians, f.88b; compare too the destruction of the Temple of Jerusalem by Bukhtnassar/Nebuchadnezzar, f.134b. For illustrations, see Soucek, “Illustrated manuscript”, figs. 5 and 20. 39 E.g. Sachau, Chronology, 21–2.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4C9DA01D /production/pindar/booksxml/hillen/chapter−04

120 however, to explain the form which the picture has taken, and in particular its references to Muhammad and “the Messiah”, which are made explicit in a later marginal note. The starting point for any interpretation must be the text itself. It mentions that Babylon has fallen, and that is a phrase that would have resounded in the ears of a contemporary reader of the Edinburgh manuscript, by whom the trauma of the fall of Baghdad – itself a great capital in the tradition of Babel, and located near the site of its great predecessor40 – in 1258 would not easily have been forgotten. The aftermath of that cataclysmic event, an event like the sack of Rome or the capture of Constantinople, was precisely a recrudescence of Christian influence and political power in Iran.41 Only after the mass conversions of 1295 did this situation end; and that was a still fresher memory. Seen in this light, then, the picture is a commentary on recent events, a sigh of relief that Islam is back in the saddle again. And there is little doubt that the picture has personalised this issue by depicting Muhammad and Jesus together. It is a message that could scarcely be more contemporary.42 It is instructive to see how the artist puts this message across. Visually speaking, the camel – the Muslim beast par excellence43 – shoulders out the donkey, the Christian beast, and its head held high contrasts with the submissive drooping of the donkey’s head. Muhammad, who is slightly the larger figure, occupies the foreground, relegating Jesus to the next plane and overlapping him; he has a slightly more elaborate turban than Jesus, complete with decorated tailpiece, cross-band and loop under the chin. For his part, Jesus indicates Muhammad in a way that broadly recalls John the Baptist indicating Jesus in Christian images. The cloud above underlines this gesture. Muhammad alone wears green, which was to be the colour of sanctity in 40

Cf. EI (1st ed.), s.v. “Babil” (G. Awad). Spuler, Mongolen, 205–24; D.O. Morgan, The Mongols (Oxford, 1986), 159–60. 42 Perhaps the images of Nebuchadnezzar and Eli, with their message of tragedy for the Jews, drive home the idea that for the Jews too the year 1295 marked a new and less favourable dispensation. 43 See in general R.W. Bulliet, The Camel and the Wheel (Cambridge, Mass., and London, 1975), especially 105–10. See too the use of the camel to symbolise Muslims in defeat on the coronation robe of Roger II of Sicily (W. Hartner and R. Ettinghausen, “The conquering lion, the life cycle of a symbol”, Oriens 17 [1964], 164–5 and fig.2). Since the animal normally shown under attack by the lion is the bull, its replacement by a camel in this case can be seen as a deliberately topical detail. Why the victorious feline is shown as a tiger rather than a lion remains obscure. 41

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4968A3F8 /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

121

Islam and may already have taken on that meaning. Muhammad’s halo is above that of Jesus. Nor can there be any doubt that Jesus is making way for Muhammad. What of the watchman? Certainly he fits rather uneasily into the rest of this picture. He wears the cowl of a Christian monk, and that might suggest that his tower is a Christian hermitage (saumi ca). One is irresistibly reminded of a comparable scene in Rashid al-Din’s World History where the Christian monk Bahira, again looking out of a tower, recognises Muhammad as a prophet (pl. II).44 It is hard to doubt that these two contemporary scenes are related. Jesus, on the other hand, wears a turban, presumably because he is reckoned to be a Muslim whereas the watchman is not. In the scene from the World History which has just been quoted, it is a Christian who (like Jesus himself here) acknowledges the supremacy of Islam. It is not hard to believe that a message to contemporaries is contained in this scene. It would be mistaken, however, to interpret the juxtaposition of Muhammad and Jesus here in a spirit hostile to Christianity.45 The close parallelism of pose encourages the idea that the two religions are intimately linked, though with Islam having a distinct edge. It was not yet time for a persecution of Christians, what with the intense diplomatic and commercial activity at this period between the Ilkhanids and Western Europe,46 not to mention the probable survival of Christians and Christian sympathisers at the Ilkhanid court. The next image to depict Muhammad (f.92a; dimensions 83 by 133 mm.) is located well over half-way through the book, at such a distance from its predecessor that it is hard to defend the notion that they are intended as part of a connected cycle. It shows Muhammad, surrounded by his family, receiving the envoys of the false prophet Musailima (pl. 12).47 The two figures at his own right hand have presumably been added for symmetry to counterbalance ‘Ali and his two sons. The resultant compositional format is clearly derived from a standard audience scene of the kind familiar from the 44

Talbot Rice, op. cit., pl.30 (f.43b). Nor can one easily detect any hostile flavour in the scenes of the Annunciation and the Baptism in this same manuscript, for all that these scenes are thoroughly Islamic (D. Talbot Rice, “Two Islamic manuscripts in the Library of Edinburgh University”, Scottish Art Review VII/I [1959], 4). Note, however, the implications of choosing for depiction the moment after the Baptism. 46 Spuler, Mongolen, 224–35; Morgan, op. cit., 183–7. 47 Sachau, Chronology, 192. 45

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4BECCD91 /production/pindar/booksxml/hillen/chapter−04

122 frontispieces of thirteenth-century book painting and related images.48 But this visual cliché is used to excellent effect here to suggest instead – mainly by the body language of ‘Ali and his children – the solidarity and intimacy of the Prophet’s family, and how they revere him, for all of them gaze at him. Al-Husain is placed closest to Muhammad, presumably a reference to his favoured status. All in all, the exclusiveness of the ahl al-bait is introduced into a scene in which it has no place, for the standard account of this episode, an account also followed by al-Biruni, mentions merely the Prophet himself and the two envoys from the rival prophet Musailima. By thus transforming an official ceremonial scene into an intimate family portrait, the artist has injected into the image an unmistakable hagiographical fervour. He has, so to speak, raised the emotional temperature and brought a Shi‘ite emphasis (of which the original text is innocent) to an otherwise neutral ceremonial image. There is no sign of the letter which the envoys brought, in which Musailima proposed that he and Muhammad should divide half of Arabia between them, leaving the other half to the Quraysh. Ibn Ishaq in his Sira or Life of the Prophet states that Muhammad was so incensed with this message – to which the envoys, on being asked their opinion, subscribed – that he said to them, “By God, were it not that heralds are not to be killed I would behead the pair of you!”49 The artist’s rendition of this scene makes it clear that there was no discussion with the envoys. Muhammad’s rigid pose bespeaks a stern and fixed resolve to yield no ground. He ignores the pleading hands of the ambassadors, silhouetted for extra effect in mid-air, and stares straight ahead. His own left hand makes an ambiguous gesture, denoting perhaps his dismissal of what they are saying, unless indeed he is dictating an answer to them.50 A narrow but critical gap has opened up between the envoys and the people clustered around Muhammad. As already noted, the gazes of ‘Ali and

48

Such as the scene depicting the Queen of Waqwaq enthroned with her attendants (al-Qazwini, ‘Aja’ib al-Makhluqat, Munich, Staatsbibliothek, cod, arab. 464, 678/1279, f.60b [Arnold and Grohmann, op. cit., pl.35c]) or the frontispieces of the Rasa’il Ikhwan al-Safa in the Süleymaniye Mosque Library, Istanbul (Esad Efendi 3638, 686/1287, ff.3b and 4a [Ettinghausen, AP, 98–9]) and the Kitab al-Hasha’ish or Hiyula cilaj al-tibb of Dioscurides in the Ayasofya Library, Istanbul (No. 3704, 7th/13th century, f.1b [E.J. Grube, “Materialien zum Dioskurides Arabicus”, in R. Ettinghausen, ed., Aus der Welt der islamischen Kunst. Festschrift für Ernst Kühnel zum 75. Geburtstag am 26.10.1957 (Berlin, 1959), fig.6]). 49 Ibn Ishaq, op. cit., 649; al-Tabari, History, Volume IX, tr. Poonawala, 106–7. Al-Biruni has a slightly terser version of these events, without the isnad. 50 Soucek, “Illustrated Manuscript”, 120; but there is no sign of a scribe.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 454F72FC /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

123

his two sons are fixed on the Prophet and away from the envoys, on whom they pointedly turn their backs. Musailima’s men are isolated and alienated at the edge of the picture – space is used aggressively to diminish and indeed to marginalise them. The issue is already settled. Indeed, ‘All’s sword, Dhu’l-Faqar, juts into the ankle of the foremost ambassador, perhaps an allusion to the coming campaign in which Musailima’s followers were defeated by the Muslims and he himself killed.51 Illustrating (and perhaps initiating) the iconography that was to become standard, though mistaken, the sword has two points, though in fact it was merely double-edged.52 Muhammad is the only personage among the eight figures here who is depicted frontally, and this uncompromising pose helps to drive home his authority. The Chinese cloud (tai) motif – for motif it is, since it is unrelated to any attempt to create a landscape setting for this image – is used to double as an honorific canopy over Muhammad’s head. The Prophet is again clothed in green, but clearly this was not yet a colour especially associated with sanctity, for one of the envoys also wears a robe of this colour. Similarly, as in the two previous scenes featuring Muhammad, all the figures – not just the Prophet and his family – are haloed. Given the Prophet’s pronouncements against the use of silk unless one sat on it, it is intriguing to note that the patterns of the throne covering on which he is seated identify this as a piece of Chinese silk.53 The infusion of an extra degree of emotion and seriousness, noted already in the scene with the envoys, is taken still further in the next two images depicting Muhammad. These strike a note of exaltation and veneration of the ahl al-bait which sets them apart as expressions of intense Shi‘ite sentiment. In some respects they are the key to the whole iconographic cycle, and they offer incontrovertible evidence as to how the text of al-Biruni was made to serve new purposes in the specific circumstances of the contemporary political situation in the Ilkhanid realm.

51

Ibn Ishaq, op. cit., 377. H. Halm, Shiism (Edinburgh, 1991), 27, n. 20. 53 One may recall the hadith which recounts how ‘A’isha had a patterned figural silk which the Prophet permitted her to keep only on condition that it was used for cushions (T.W. Arnold, Painting in Islam. A Study of the Place of Pictorial Art in Muslim Culture, repr. Toronto and London, 1965, 7). 52

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 48225810 /production/pindar/booksxml/hillen/chapter−04

124 These are grandiose claims and of course they require substantiation. The evidence is in fact manifold. First, these two pictures are the last illustrations and they come near the end of the book,54 thus constituting a kind of khatima, in other words saying in visual form that Muhammad and ‘Ali are the seal of the prophets, the last in line. They have been saved for the end of the volume, thereby creating a lasting impression that the reader takes away from the book. Second, these images appear on successive folios (ff.161a and 162a), a practice found several times in the book55 and normally used for two scenes of very similar import.56 This suggests that the two Shi‘ite images are intended to support each other and thus to intensify the impact of the message they transmit. Third, they are physically the largest images in the entire book (dimensions: 105 by 136 mm. and 128 by 176 mm.). This alone sets them apart from all the rest of the paintings, and is a silent warning that something extraordinary is afoot here. To the reader of the book it comes as a dramatic surprise. Fourth, their colour scheme dramatically intensifies the palette employed for most of the other images. In particular, they use a striking indigo blue, a deep purplish black and a blazing orange – shades that are employed from f. 129b onwards, but never before on this dramatic scale.57 Finally, both images have a much more fully realised landscape setting – i.e., involving both the earth and the sky – than all but two of the other 22 images.58 And their pronounced emphasis on the skyscape, with its lowering, threatening clouds, big with meaning, is unique in the book. It has a more thoroughly Chinese flavour than any other landscape in the cycle, even though the putative Chinese model has been thoroughly re-interpreted. This marshalling of portentous forces in the sky is a signal that something of moment is happening, and that the heavens themselves are involved in it. Chinese landscape elements are thus manipulated to drive home a polemical

54 There remains the very last chapter, XXI (16 folios long), which deals with the lunar stations and is unillustrated, though it contains numerous subjects that would have been suitable for illustration (e.g. Sachau, Chronology, 338 and 343–50). The colophon occurs on f.178a. 55 Namely ff. 91a, 92a, 92b and 93b (all about false prophets); 94a and 95a (both execution scenes); 100a and 101a (Sasanian kings enthroned); 103a and 103b (Feast of Sada); 133b and 134b (disasters for the Jews); and 140b and 141b (Christian scenes). 56 But not 101a and 101b or 103b and 104b. 57 Clearly a second artist was responsible for the last eight images. 58 Namely, ff. 48b and 140b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4DE84224 /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

125

religous message, a novel and unexpected reuse of borrowed material.59 All this points to the hand of another artist who was employed especially, though not, perhaps, exclusively, for these more significant scenes – and was certainly given (or took) more leeway in these last two paintings in the book. But of course all these factors, important and diagnostic as each of them is, are secondary to the actual content of these two scenes; and it is the content which clinches the matter. Folio 160a depicts the imminent prayer contest (mubahala, sometimes rendered as “mutual ritual cursing”) between Muhammad and the Christians of Najran (pl. 13). On this occasion, Muhammad formally recognised his own grandchildren in lieu of his sons, Fatima in lieu of his wives and ‘Ali as his closest friend.60 While the other three Muhammad scenes discussed so far, and all the 13 scenes in the Muhammad cycle in the Edinburgh and London manuscripts of Rashid al-Din’s World History, follow in their essentials the classical account of the relevant events in the Sira of Ibn Ishaq, this scene does not figure at all in that source. Its presence here is an indication of al-Biruni’s Shi‘ite proclivities. He has of course drawn on a Shi‘ite source for this event; Sunnis regard the Shi‘ite versions of this and the next scene as outright falsifications of historical reality.61 On the other hand, Ibn Ishaq does say in the Sira that three of the Najranis entered into a theological dispute with Muhammad, and that is exactly what the painting in the Biruni manuscript depicts – except that it adds

59 This is a foretaste of how landscape was employed only a generation later in the Great Mongol Shahnama. 60 Under 4 Shawwal, al-Biruni summarises the whole story very concisely: “Muhammad and the Christians of Najran argued with each other. Muhammad installed Hasan and Husain in the right of sons of his, and Fatima in the right of his wives, and ‘Ali b. Abi Talib he made his intimate friend, complying with the order of God in the verse of the cursing” (Chronology, 332). 61 Thus Ibn Ishaq (op. cit., 270–77) places the deputation from the Christians of Najran before the first raid, while the Shi‘ite Shaikh al-Mufid dates it to the last year of Muhammad’s life (Kitab al-Irshad. The Book of Guidance into the Lives of the Twelve Imams, tr. I.K.A. Howard [Horsham and London, 1981], 116). Andrae, op. cit., 170, quotes Ibn Sa‘d, Tabaqat, i, 2, 84, who says that the Najranis sent an emissary in the last year of Muhammad’s life to make a treaty. But this is not the formal ritual exchange in which the two groups are about to engage in the Biruni painting. Ibn Ishaq’s account of the deputation simply records that the Najranis consulted their ‘aqib or chief adviser on whether to resort to mubahala with Muhammad. He advised them not to do this but to make their peace with him, which they did (Ibn Ishaq, op. cit., 277).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4E8D8CFE /production/pindar/booksxml/hillen/chapter−04

126 Muhammad’s family to the dramatis personae, and thereby fundamentally changes the configuration and meaning of the image.62 This is the clue which reveals that the painting is based on a Shi‘ite version of this event, such as that of Shaikh al-Mufid. Here the Najranis engage Muhammad in a dispute about the nature of the Messiah; he thereupon recited Sura 3:61 to them, which contains the statement that God created Jesus from earth and if there is any dispute about this, each party should call their sons and women and then “call on God to witness against each other, and let … the curse of God fall on those who lie”. Muhammad then challenged the Najranis to a contest of prayer to God (mubahala) and states that God has informed him “that dread torment will come down on him who has spoken falsely after the contest of prayer”. In the account of Shaikh al-Mufid, the key factor which causes the Najranis to make up their minds is the presence of Muhammad’s family. The Najrani bishop, Abu Haritha,63 reasons that if Muhammad enters the contest alone, they should proceed with the mubahala. But if he is prepared to expose his own family to the outcome of such a contest, he must have right on his side and conflict with him should be avoided. Sure enough, “the next morning, the Prophet, may God bless him and his family, came and took ‘Ali b. Abi Talib by the hand, while al-Hasan and al-Husayn, peace be on them, were walking in front of him and Fatima, peace be on her, walked behind him”64. When the members of Muhammad’s family – each of them described in turn as being the dearest of creatures to him – are identified for the bishop, he advised the three principal Najranis (the ‘aqib or deputy, the sayyid or chief, and ‘Abd al-Masih) to make peace with Muhammad. They duly did so, and Shaikh al-Mufid then gives the details of the treaty,65 which broadly conform with those related in the standard Sunni sources.66 At the end of the narrative, Shaikh al-Mufid provides a brief commentary on the

62

Ibn Ishaq in fact seems to contradict himself; first he states that there were 60 men (Shaikh al-Mufid says there were 30) in the Najrani delegation, very elegantly dressed, (Shaikh al-Mufid, op. cit., 116, relates that “they were wearing robes of silk and crosses”; of this there is no sign in the painting) and these included their 14 principal men, of which three spoke to Muhammad (Ibn Ishaq, op. cit., 271). But a paragraph later (p. 272), he mentions only “two divines [who] spoke to the apostle.” 63 Not the ‘aqib who advises the Najrani delegation in Ibn Ishaq’s text. 64 Shaikh al-Mufid, op. cit., 117. 65 Shaikh al-Mufid, op. cit., 118. 66 For his text, see W.M. Watt, Muhammad at Medina (Oxford, 1956), 359–60; for a discussion of it, see ibid., 127–8.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 4B5804B2 /production/pindar/booksxml/hillen/chapter−04

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

127

entire episode,67 stressing that it shows ‘Ali to have been equal in stature to the Prophet, that it is a miracle which indicates Muhammad’s prophethood, and that it conferred on Muhammad’s family “a merit which no one else of the community shares with them, nor even approaches them in it, nor has anything like it in significance. It is associated with the outstanding special qualities of the Commander of the Faithful”.68 How, then, is this story depicted? The painting, the largest so far in the whole manuscript, deploys the full resources of landscape in favour of Muhammad and his family. Thunderous clouds – harbingers, perhaps, of the “dread torments” with which the Najranis are threatened – hover over Muhammad’s head and indicate that the heavens are on Muhammad’s side.69 The fire and smoke in these clouds recall the experiences of Moses on Mount Sinai, where his role as a prophet to whom God spoke was underlined by these same phenomena.70 Two tongues of fire point balefully at the Christians, like emissaries of divine wrath. Quite aside from the portentous landscape, however, the scene projects a powerful sense of family solidarity that is quintessentially Shi‘ite in feeling. This sense of ‘asabiyya is underlined by the alarmingly wide gap that has opened up between Muhammad and his opponents, as in the scene depicting Musailima’s envoys. Fatima’s arched eyebrows meet at the bridge of her nose, a diagnostic sign of beauty.71 Hers are the only eyebrows in the picture to form this gracefully joined double arch or bow (Persian kaman). Her hand is laid protectively on al-Husain’s shoulder, as is ‘Ali’s, and once again, as in the scene with Musailima’s envoys, al-Husain is closest to Muhammad. His hands are so positioned as to be a reverse echo of Muhammad’s own hands. The disparity in size is enough to differentiate al-Husain from his elder brother, al-Hasan, who functions as the link figure between the two groups; the two landscape planes point from him to the Najranis, while his gaze turns away from them to encounter that of his younger brother.72 The tension in the image is palpable as the opposing

67 68

Shaikh al-Mufid, op. cit., 118–9. To Shi‘ites, ‘Ali is the only person who can legitimately bear this title; see Halm, op.

cit., 9. 69 Soucek notes that certain Shi‘ite sources mention that there was thunder and lightning on this occasion (“Illustrated Manuscript”, 154; cf. 168, n. 188). 70 They are faithfully depicted in the corresponding illustration of the World History of Rashid al-Din. See Talbot Rice, op. cit., pl.10, f. 8a. 71 J. Nurbakhsh, What the Sufis say (New York, 1980), 57–8. 72 Does this perhaps point to a certain ambiguity in his status?

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 50 OUTPUT: Tue Aug 10 10:24:34 2010 SUM: 48318E2C /production/pindar/booksxml/hillen/chapter−04

128 sides eye each other up; something awesome is about to occur. Presumably it is the very moment that the Prophet challenges the Najranis to the prayer contest. Colour is used to dramatic effect in distinguishing Muhammad’s family, who wear an array of intense deep colours, from the plain white worn by the Najranis. Muhammad, who on folios 6b, 10b and 92a is attired more or less like the other figures in these scenes, now wears his prophet’s tarha, its radiating patterns reminiscent of a rayed nimbus,73 though he already has a halo. This, as in the next picture, is of a specific type: it encircles the head but not the turban, which seems to swallow it up so that the halo is reduced to a pair of golden ear-muffs. In the earlier paintings of the manuscript, where Mesopotamian influence is strong, the halo encircles the turban as well as the head. The largely obliterated halo is a signature of the second painter. There remains the pressing question of why this particular episode was selected for illustration, and treated in this emotionally charged style. The crucial factor here seems to have been that Muhammad’s opponents were Christians, and that – by the transparent device of a theological debate – they were challenged head-on as to the Prophet’s claims about his mission. As with the three previous Muhammad scenes, then, the issue of a rival to Islam is raised, and once again, in the person of Muhammad, Islam is presented as triumphing over its enemies. Clearly Christianity is the principal enemy to be overcome, for two of the five pictures involving Muhammad deal with it. As before, the theme of Islam victorious takes on an extra resonance in the religious context of the Ilkhanid realm c.1300, and the very recent triumph of Islam over rival faiths. But an extra dimension is added here in the form of a Shi‘ite emphasis that was first detectable in the scene with Musailima’s envoys but is now much more pronounced – for whereas in that scene ‘Ali and his sons are not integral to the story, but serve merely to give the Prophet moral support, here their presence, now reinforced by Fatima, is crucial to the immediate victory over the Christians. This victory, as Shaikh al-Mufid’s tafsir explains, sets the seal on the unique status of the ahl al-bait in the Muslim community. The very last painting in the book (pl. 14; f. 162a) is also the largest (128 by 176 mm.) – with only three lines of text beneath it, it is almost a full-page illustration – and of course its power is intensified by its proximity to the

73

One may compare the curious combination of a halo and drapery in one of the frontispieces to a multi-volume copy of the Kitab al-Aghani made for Badr al-Din Lu’lu’ in c.614/1218–19 (Ettinghausen, AP, 65).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04 Sequential 18

JOBNAME: ???? PAGE: 1 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 1E18E156 /production/pindar/booksxml/hillen/chapter−04−artwork

129

2. Rashid al-Din, Jami‘ al-Tawarikh, f.48v: early converts to Islam undergo torture.

1. Rashid al-Din, Jami‘ al-Tawarikh, f.149v: bismillah.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 064EAECF /production/pindar/booksxml/hillen/chapter−04−artwork

3. Rashid al-Din, Jami‘ al-Tawarikh, f.54r: the Quraish consult about the proscription of their kinsmen who support Muhammad.

130

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06B5AE0E /production/pindar/booksxml/hillen/chapter−04−artwork

131

4. Rashid al-Din, Jami‘ al-Tawarikh, f.42r: the birth of Muhammad.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 051629E8 /production/pindar/booksxml/hillen/chapter−04−artwork

5. Al-Biruni, Al-Athar al-Baqiya, f.6v: Muhammad forbids intercalation.

132

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06D14EB6 /production/pindar/booksxml/hillen/chapter−04−artwork

133

6. Al-Biruni, Al-Athar al-Baqiya, f.10v: Isaiah’s prophecy about Muhammad.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06493188 /production/pindar/booksxml/hillen/chapter−04−artwork

7. Rashid al-Din, Jami‘ al-Tawarikh, f.45r: Muhammad arbitrates over lifting the Black Stone into position in the Ka‘ba.

134

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 0845BF70 /production/pindar/booksxml/hillen/chapter−04−artwork

135

8. Rashid al-Din, Jami‘ al-Tawarikh, f.52r: the Negus of Abyssinia refuses to yield up the Muslims who sought asylum with him.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 04EF54E6 /production/pindar/booksxml/hillen/chapter−04−artwork

9. Rashid al-Din, Jami‘ al-Tawarikh, f.57r: Muhammad, Abu Bakr, the herdswoman and the goats.

136

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 07DAF17F /production/pindar/booksxml/hillen/chapter−04−artwork

137

10. Rashid al-Din, Jami‘ al-Tawarikh, f.55r: the Night Journey of Muhammad on Buraq.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 05ADD73E /production/pindar/booksxml/hillen/chapter−04−artwork

11. Rashid al-Din, Jami‘ al-Tawarikh, f.43v: the young Muhammad recognised by the monk Bahira.

138

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06799D88 /production/pindar/booksxml/hillen/chapter−04−artwork

139

12. Al-Biruni, Al-Athar al-Baqiya, f.92r: Muhammad with the envoys of Musailama.

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 050D3C01 /production/pindar/booksxml/hillen/chapter−04−artwork

13. Al-Biruni, Al-Athar al-Baqiya, f.161r: the Day of Cursing.

140

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06F80EB5 /production/pindar/booksxml/hillen/chapter−04−artwork

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

141

14. Al-Biruni, Al-Athar al-Baqiya, f.62r: the investiture of ‘Ali at Ghadir Khumm.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 06D67E0C /production/pindar/booksxml/hillen/chapter−04−artwork

15. Rashid al-Din, Jami‘ al-Tawarikh, f.45v: Muhammad receives his first revelation through the angel Jibra’il.

142

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 07585017 /production/pindar/booksxml/hillen/chapter−04−artwork

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

143

16. Al-Biruni, Al-Athar al-Baqiya, f.1r: the title page.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 0A68D04E /production/pindar/booksxml/hillen/chapter−04−artwork

17. Rashid al-Din, Jami c al-Tawarikh, f.149r: shamsa

19. Rashid al-Din, Jami c al-Tawarikh, f.137r: decorative cartouche.

144

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 32 OUTPUT: Mon Feb 13 11:56:55 2012 SUM: 08396DFA /production/pindar/booksxml/hillen/chapter−04−artwork

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

145

18. Rashid al-Din, Jami c al-Tawarikh, f.41r: cAbd al-Muttalib and al-Harith about to discover the well of Zamzam.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-artwork Sequential 17

JOBNAME: ???? PAGE: 1 SESS: 56 OUTPUT: Mon Feb 13 11:58:01 2012 SUM: 4C79C05C /production/pindar/booksxml/hillen/chapter−04−ends

146 previous image. It is as if the artist had reserved the most important and eloquent image to act as a fitting climax for the whole book. Al-Biruni’s text does not lend itself to anything like this degree of manipulation,74 yet another reminder that illustrations and text do not necessarily follow the same agenda. The theme depicted here is of critical importance in Shi‘ite belief, for it is the culminating moment in a long sequence of statements made by the Prophet to indicate that ‘Ali was his designated successor. Shaikh al-Mufid gives four separate stories in this vein,75 adding that many more could be cited.76 But he reserves a lengthy separate section for the episode illustrated here, when in the course of his farewell pilgrimage Muhammad, in obedience to a revelation he had received, stopped at a place called Ghadir Khumm77 to make a formal announcement acknowledging ‘Ali as his heir: “The time has come for me to depart from you. I leave behind me among you two things which, if you cleave to them, you will never go astray – that is the Book of God and my offspring from my family” (ahl al-bait) … “taking both arms of the Commander of the faithful, peace be on him, and raising them so that the white of his armpits could be seen, he said ‘Whoever I am the master (maula) of, this man, ‘Ali, is his master. O God, befriend whoever befriends him, be hostile to whoever opposes him, support whoever supports him and desert whoever deserts him.’’’78 The setting is briefly evoked by Shaikh al-Mufid: Ghadir Khumm “lacked water and pasturage”,79 “it was a scorching day of intense heat” and Muhammad ordered ‘Ali “to go and stand under a great tree”80 – as indeed is shown in the painting, though the effect is vitiated by a second smaller tree to the left – and “the travellers to be gathered in that place and to be put in (rows) one after another”.81 Muhammad made ‘Ali stand at his right hand,82 74 Sachau, Chronology, 333. As usual, this is a compressed narrative. Even so, it adds a detail not reflected in the painting itself: “He (Muhammad) gave orders to collect the saddles and all the riding-instruments into one heap; this he ascended, supported by the arm of ‘Ali b. Abi-Talib …” The painting shows Muhammad standing on a flat piece of ground. 75 Kitab al-lrshad, 27–9. 76 Ibid., 29. 77 This place was chosen because it was the last spot before “many of the people would separate from his party (heading) for their towns, homes and valleys” (Kitab al-Irshad, 123). 78 Kitab al-Irshad, 124. 79 Ibid., 123. 80 Ibid., 124. 81 Ibid.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 56 OUTPUT: Mon Feb 13 11:58:01 2012 SUM: 4A7EA510 /production/pindar/booksxml/hillen/chapter−04−ends

IMAGES OF MUHAMMAD IN AL-BIRUNI’S CHRONOLOGY

147

but here ‘Ali is shown at Muhammad’s left hand. Moreover, Muhammad lays his left hand on ‘Ali’s shoulder instead of raising both of ‘Ali’s arms. Similarly, the gathering of many Muslims in rows has been reduced to three bystanders. ‘Umar is mentioned as one of those present at Ghadir Khumm – indeed, he congratulates ‘Ali83 – and given that the faces of the three attendant figures have been comprehensively obliterated while those of Muhammad and ‘Ali have remained untouched, it is possible that they were interpreted as, and may even have been intended by the painter to represent, the other three Rashidun caliphs.84 No matter what the identity of the bystanders might be, there is no mistaking the solemnity of the moment, with Muhammad and ‘Ali regarding each other intently and the secondary figures pressing close. The way the landscape is used in this investiture image recalls the technique employed in the Temptation scene on f.48b.85 In both cases a double landscape is developed, though the technique has evolved further in the investiture scene. Here the artist uses the lower plane to give his scene an ambiente and the upper plane to crown it. But there is no middle plane; it is left entirely white so that it does not get in the way of the action. As with the previous picture and the scene with Musailima’s envoys, one may note an effective and dramatic use of a caesura of empty space, in this case to underline the scene’s significance. The clothing chosen for Muhammad differs from that employed in the previous image. Over a pure white robe, seemingly spangled with rosette designs in its lower part, he wears a dark brown combination of cowl and

82

Ibid. Ibid., 125. 84 But the subsequent fate of these images, once the painter had completed them, is a treacherous guide to their significance. How, for example, is one to explain the defacement of the Prophet and his family in the scene with Musailima’s envoys, when their images escape unscathed on ff. 161a and 162a? The original aspect of the damaged painting can perhaps be reconstructed with the aid of the later (perhaps Ottoman, 11th/17th century) copy in Paris whose illustrations seem to be based on those of the Edinburgh manuscript (Bibliothèque Nationale, fonds arabe 1489, f. 87 [Arnold and Grohmann, op. cit., pl.40]). Professor Marianne Barrucand is preparing a detailed study of this cycle of paintings. Cf. too M. V. Fontana, Iconographica dell’ Ahl al- Bayt. Immagini di arte persiana dal XII al XX secolo, Istituto Universario Orientale, Annali 54/1 (1994), Supplemento 78, figs. 6 and 8; for the Biruni manuscript, see ibid., 15–17. 85 Since this double-tiered landscape depicts Paradise, is it possible that a similar meaning is indirectly suggested in f. 162a (Soucek, “Illustrated Manuscript”, fig. 4)? 83

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 56 OUTPUT: Mon Feb 13 11:58:01 2012 SUM: 3EA999AA /production/pindar/booksxml/hillen/chapter−04−ends

148 mantle, perhaps his burda. As in f. 161a, the effect is to single him out. Perhaps the choice of different clothing in the two cases is due to the fact that sacred iconography was still at that time in the making, and standard images had not yet evolved. Again as in f. 16la, the side figures have their hands muffled in their deep, full sleeves in true Chinese fashion. Both Muhammad and ‘Ali are dark-complexioned, as if the artist had carefully observed the colouring of certain Arabians. These last two images were to remain exceptional in cycles of Muhammad’s life and their presence here is an obvious example of pictures with a message. They were painted at a time when the current Ilkhan, Oljeitu, who definitively embraced Shi‘ism three years later,86 in 1309, was possibly already turning over in his mind the possibility of conversion. With their fervid religiosity they are prophetic of the ghulat. Did the painter perhaps exceed his brief ? Admittedly the scenes depicting both figure in al-Biruni’s text, but the author himself did not write as a partisan Shi‘ite and did not highlight them in the way that the painter did. It is not even certain that the patron, whoever that was, intended this particular emphasis; for, as already suggested, it is entirely possible that the book was ordered because the unprecedentedly wide horizons of the Mongol empire made at least some people unusually ready to interest themselves in other systems of belief. But whereas the images illustrating Rashid al-Din’s Jami c al-Tawarikh are appropriately integrated with the historical text which they accompany, three of the five Muhammad images in the Biruni manuscript are already permeated by strong sectarian feeling. Thus the polemical potential of such subject matter is there right at the beginning of religious painting in Islam: proof, if any were needed, that it was a sound instinct which had steered earlier painters away from such themes.

86

M. van Berchem, “Une inscription du sultan mongol Uldjaitu”, in Mélanges Hartwig Derenbourg. Recueil de travaux d’érudition dédiés à la mémoire d’Hartwig Derenbourg par ses amis et ses élèves (Paris, 1909), 374–5, repr. in his Opera Minora, II (Geneva, 1978), 680–1; see also S.S. Blair, “The coins of the later Ilkhanids: a typological analysis”, Journal of the Economic and Social History of the Orient XXVI/III (1982), 297.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-04-ends Sequential 3

JOBNAME: ???? PAGE: 1 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 3B6A211E /production/pindar/booksxml/hillen/chapter−05

V The Arts of the Book in Ilkhanid Iran ‘‘Then, the custom of portraiture flourished so in the lands of Cathay and the Franks until sharp-penned Mercury scrivened the rescript of rule in the name of Sultan Abusa cid Khudaybanda [i.e., until Sultan Abusa cid Khudaybanda came to the throne]. Master Ahmad Musa, who was his father’s pupil, lifted the veil from the face of depiction, and the [style of] depiction that is now current was invented by him. Among the scenes by him that lighted on the page of the world in the time of the aforementioned emperor, an Abusa cidnama, a Kalila u Dimna, a Mi crajnama calligraphed by Mawlana cAbdullah Sayrafi, and a Tarikh-i Chingizi in beautiful script by an unknown hand were in the library of the late emperor Sultan-Husayn Mirza. —Dust Muhammad, Preface to The Bahram Mirza Album1

T

O judge by what has survived, there can be no doubt that the luxurious decorated and illustrated books produced from about 1280 to about 1336 were, along with architecture, the principal achievement of the visual arts in Ilkhanid times. Ceramics and metalwork, which had developed so dramatically in technique, subject matter, and expressive range during the twelfth and early thirteenth centuries, failed to sustain quite the same momentum under the Ilkhans. Perhaps there was less call for them, which may reflect changes in the pattern of patronage, especially in mercantile circles, arising from the cataclysmic Mongol invasions. The massive loss of life in northern Iran and Khurasan would have greatly thinned the ranks of

1

Muraqqa c-yi Bahram Mirza. Dibacha-yi Dust Muhammad, translated by Wheeler M. Thackston. Thackston 2001, pp. 12–13. I am very grateful to Sheila Blair and Linda Komaroff for taking the trouble to read my text carefully and to improve it with their astute and trenchant, but thoroughly helpful, criticism.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 43914C2A /production/pindar/booksxml/hillen/chapter−05

150 patrons2 and also to some extent craftsmen,3 causing a loss of continuity in technical expertise that could not readily have been rectified. Yet Abu’l -Qasim Kashani’s treatise on ceramic techniques, written about 1300;4 such sumptuous new wares as lajvardina; and pictorially complex metalwork all indicate that these arts were still vibrant. The sudden rise to prominence of book painting may stem in part from its capacity to carry more subtle messages than any other medium (fig. 2). Political developments are of direct relevance in this regard. The rapid conversion to Islam of the Mongol elite after 1295 had the most direct impact on the arts5 (significantly, the year before, the death in China of Khubilai Khan, the supreme Mongol ruler, had given greater independence to the Mongols in Iran.) A building boom, concentrated on religious structures, was matched by a corresponding revival in the output of top-quality Qur ans. Of exceptionally large scale thanks to a special so-called Baghdadi paper, the finest then available,6 these works also featured calligraphy and illumination of the highest possible standard. The new emphasis on massive scale continued in the field of book painting and also characterized several buildings ordered by the most exalted patrons in the land. Among these were the mosque of cAlishah in Tabriz, the tomb of Öljeitü at Sultaniyya, and the mausoleum of Ghazan Khan himself,7 whose liberal endowment paid for multiple copies of the Qur an and also for regular readings from it, which gave the structure a religious function. Such high-profile emphasis on traditional Islamic piety found a somewhat less traditional counterpart in the creation of multiple images of the prophet Muhammad and narrative scenes from his life. Other gestures of ostentatious piety among the Mongol elite included building mosques and charitable institutions for the poor, honoring holy men, wearing Sufi garb,

2

The subsequent economic collapse would also have been disastrous for the visual

arts. 3

The Mongol practice of sparing and relocating artists (see Morris Rossabi’s chapter 1, James Watt’s chapter 3, and Linda Komaroff ’s chapter 7 in this catalogue) must have reduced the death toll of artists and craftsmen but would still have removed them from the Iranian sphere. 4 Allan 1973. 5 Melville 1990b. 6 Bloom 2001, pp. 53, 62, 64, 70, 112, and p. 54, fig. 22; Loveday 2001, pp. 50, 59, 81. 7 O’Kane 1996, pp. 506–10.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 46F50261 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

151

undertaking the chilla (a forty-day ascetic retreat), visiting popular shrines, and ordering public Qur an readings. Ghazan Khan did all of these things.8 The Rise of Book Painting A crucial side effect of the very public Mongol conversion, and of the Ilkhans’ desire to draw propaganda capital from it, was a sudden surge in the production of manuscripts, made possible by the development on a grand scale of all the expensive facilities required to sustain such production. Papermaking factories were obviously the key element, and surviving Ilkhanid manuscripts document how wide-ranging were the size and quality of their output. However, calligraphers and workers in their ancillary trades—those who prepared the paper, pens, and ink, bookbinders and leatherworkers, illuminators and painters—also benefited from the new direction taken by royal and vizierial patronage. In fact, the extremely rapid development of book painting in this period can be seen as an unexpected by-product of calligraphy, which contemporary culture regarded as its principal art form. But it also represents a distinct change of direction in artistic priorities, and thus a redeployment of the artistic energies then available. Why book painting? An answer is perhaps more likely to be found in the purpose of the newly commissioned volumes than in the medium itself. All these books had a closely focused aim: to assert and promote either religion or heritage. The needs of the first were met principally by Qur ans, but also by prestigiously produced volumes of hadith (sayings of the Prophet) and tafsir (commentaries on the Qur an). Heritage was served by history books such as cAta’ Malik Juvaini’s Tarikh-i jahan-gusha (History of the World Conqueror, fig. 37), Abu’l-Fadl Muhammad Bal cami’s translation of al-Tabari’s Ta’rikh al-rusul wa’l-muluk (Chronicle of Prophets and Kings), and of course Rashid al-Din’s Jami c al-tawarikh (Compendium of Chronicles, fig. 16), as well as by a flood of illustrated copies of the Shahnama (Book of Kings), the Persian national epic (fig. 4). In such works, the hitherto stubbornly alien rulers of Iran were expressing a new and public commitment to the religion and cultural heritage of the very lands that they

8

Bausani 1968, pp. 542–43; Boyle 1968b, pp. 389, 393, 395; Melville 1990b, pp. 163, 168. He may also have been the patron of one of the finest of Ilkhanid Qur ans; see James 1988, pp. 12, 78, 96, 235.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 43F62E29 /production/pindar/booksxml/hillen/chapter−05

152 themselves had devastated some two generations previously—and doing so with an urgency that suggested they were making up for lost time. The Extant Manuscripts Some three dozen illustrated manuscripts have survived in whole or in part from the Ilkhanid period. Clearly this amount of material allows for adequate study—enough to avoid skewed generalizations founded on too little evidence. It is possible to isolate significant trends while remaining alert to the remarkable variety of these manuscripts. Major centers of production can be identified, as can the most popular texts. But in order to make sense of what happened in this period, and to analyze how it relates to earlier and later production, a close, detailed focus on key manuscripts is required rather than brief and thus possibly superficial comments on all or most of them. This account therefore omits the various manuscripts of the Kalila va Dimna (a book of animal fables), none of which seem to be crucial to the argument;9 the fascinating scientific anthology Mu nis al-ahrar with its highly individual illustrations and precious evidence that a school of book painting flourished at Isfahan (fig. 39);10 and other books on star lore,11 cosmography (fig. 38),12 history,13 and ethics, several of them with a secure or at least likely Iraqi provenance. The range of texts chosen for illustration in Iran during the Ilkhanid period is much wider than that of earlier Arab painting. In addition to the subject matter just cited, there are bestiaries, fables, epics, encyclopedias, works of philosophy, calendrical systems, lyric or panegyric poetry by various authors (figs. 5, 6), and even a prose romance (figs. 7, 8). This largely new material posed problems of illustration that could not be solved simply by reshuffling the standard iconographic components of earlier painting. Instead, it called for a fundamental re-evaluation of the functions of illustrated books to keep abreast of the dramatic expansion in their subject

9 For these, see the comprehensive monograph by Bernard O’Kane now in press; meanwhile, for the great Istanbul Kalila va Dimna and its context, see Cowen 1989. 10 Swietochowski and Carboni 1994, esp. pp. 8–66. 11 Carey 2001, p. 56. 12 Carboni 1992; see also Carboni 1988–89, pp. 15–31, and color pls. IV–VII. 13 Richard 1997, p. 41; Fitzherbert 2001.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 41E3C727 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

153

matter. They became fashionable, collectible objects of high status, popular in many fields of learning and leisure, perhaps as a result of the educational function of several key manuscripts. The variety of styles employed to illustrate this remarkable array of subjects is correspondingly wide, as is the range of centers at which the books were made. These styles were further enriched by borrowings from western Europe and from the Far East (fig. 9), and the latter in particular decisively broadened the pictorial (as distinct from the purely iconographic) repertoire. A readiness to experiment was, it seems, fostered by patrons who sought to use illustrated books for new purposes. The Morgan Bestiary The sequence of key manuscript opens with a bestiary by Ibn Bukhtishu c, an c Abbasid court physician, that was originally written in Arabic and then translated into Persian as Manafi c-i hayavan (On the Usefulness of Animals). The Persian copy in the Pierpont Morgan Library, New York, was made in Maragha in 1297–98 or 1299–1300.14 Though smaller (14 by 11 inches) than some of the other ambitious Ilkhanid manuscripts, it is typical of the school in its innovative approach to page layout, with illustrations big and small distributed throughout the available space and in different formats from page to page. The painters lift the depiction of living creatures out of the doldrums of stereotype and investigate the potential of naturalism, expressiveness, and even sheer fantasy. In many images, this new interest in the depiction of specific rather than prototypical creatures goes hand in hand with an unprecedented awareness of the potential of landscape. This is an unmistakable response to Far Eastern works of art, although precisely what they were remains uncertain. Hence the Morgan Bestiary is an early but classic example of the Iranian encounter with Chinese art.15 The impact of Chinese handscrolls can be sensed in the many pictures in which the frame abruptly terminates the scene, thereby forcing viewers to complete it in their imaginations.16 Often enough, Chinese landscape features serve merely as a miniaturized

14

Schmitz 1997, p. 11. See de Lorey 1935, an article that has worn surprisingly well. 16 The extension of the picture space by deliberately breaking the frame and thereby bringing it into “our” space is obviously a related device. 15

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 3FD2261C /production/pindar/booksxml/hillen/chapter−05

154 decorative backdrop for the large-scale creatures occupying center stage, but they also occasionally suggest specific habitats, from misty mountain ranges to a bird’s-eye view of a tropical island (fig. 10). The spatial relationships so integral to the types of Chinese art that inspired the Iranian artists (probably at one remove or more) are reversed, so that, for example, gigantic blossoms and birds contradict the proportions of the tree in which they are set.17 Strong, assertive, clashing colors and powerful outlines replace the restricted palette and deliberately ambiguous modeling of the putative Chinese originals, and thus their distinctive sense of illimitable space is lost. Rather than simply copying Chinese elements, then, the Iranian artists harnessed them to their own way of seeing. China was less a model than a catalyst. The “Chronology of Ancient Nations” That scientific manuscripts remained as popular in Ilkhanid times as in thirteenth-century Iraq is shown by the surviving manuscripts datable between 1275 and 1315. Along with two bestiaries, two cosmographies, and a book on the stars, there is the Kitab al-athar al-baqiya (Chronology of Ancient Nations) by the eleventh-century polymath Abu’l-Rayhan al-Biruni, dated 707 (1307–8) and now in the Edinburgh University Library. Despite the rather forbidding nature of its contents—the calendrical systems of the known world, past and present, and their associated festivals—this lengthy, three-hundred-year-old text offers numerous if fleeting opportunities for illustration. The twenty-five illustrations of this copy present al-Biruni’s text as a survey of the world’s religions, true and false. The patron who ordered it made, while obviously sharing the author’s fascination with other systems of belief, did not share his tolerance—or at any rate the painters commissioned to illustrate the book did not. An underlying charge of fervent sectarian belief can be detected throughout the cycle of illustrations; in particular, three of the five scenes depicting Muhammad have a markedly Shi cite emphasis. Because of the wide-ranging scope of the text, entirely different subjects could have been chosen for illustration, and, by the same token, the choices actually made must have been deliberate. Thus it is appropriate to

17

Linda Komaroff has pointed out to me that the landscapes depicted in Chinese textiles have many of the same spatial devices as those in Iranian painting. The Iranian artists thus obviously had media other than paintings before them.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 446E965A /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

155

look for evidence that the illustrations were an integrated cycle projecting specific ideas or at any rate to inquire what motivation lay behind the choices made. Distinct emphases immediately assert themselves in the subject matter and placing of the pictures. The scenes involving the prophet Muhammad, for example, are strategically located at roughly the beginning of the book, on folios 6v (fig. 11), 10v; the middle, on folio 92r; and the end, on folios 161r, 162r—a symmetrical pattern of distribution that recalls the use of full-page illuminations at the beginning, middle, and end of Qur ans. This arrangement may have been intended as a means of sanctifying the entire book, or at least of asserting the dominance of Islam amid so many other conflicting systems of belief. In the number of illustrations, too, Islam takes pride of place among the religions represented. Other marked attitudes detectable in the paintings are a hatred of heresy and freethinking, in both Islam and other faiths, and an interest in pre-Islamic Iranian beliefs and festivals. The treatment of Buddhist, Judaic, and Christian images is particularly revealing. Special favor had been extended to these three faiths under the earlier Ilkhans, some of whom espoused Buddhism and Christianity; at that time all three faiths flourished as never before in Iran,18 and Jews and Christians alike rose to high office. But the Mongol conversion to Islam was followed by state discrimination against, and even persecution of, other religions.19 The razing of Buddhist temples, monasteries, and statues is referred to obliquely here in a painting of Abraham destroying the idols of the Sabians, which are anachronistically depicted with the shaven heads, ample bellies, cross-legged poses, mudras, and (in one case) gilded surfaces of Buddhist devotional images. Judaism and Christianity, however, are approached in a somewhat more complex manner. Two images of Judaic content depict key moments in which God punished his chosen people by allowing foreigners to destroy or capture the symbols of his presence among them. The scene of Bukhtnassar (Nebuchadnezzar) overseeing the destruction by fire of Solomon’s Temple in Jerusalem—here represented as the Dome of the Rock—records one of the great catastrophes of Jewish history, itself followed by the Babylonian exile. The other scene represents a disaster of comparable magnitude from 18 19

Spuler 1955, pp. 198–249, gives a clear overall survey. Bausani 1968, p. 542; Boyle 1968b, pp. 379–80; Melville 1990b, p. 170.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 446F73EC /production/pindar/booksxml/hillen/chapter−05

156 the earlier era of the judges. The High Priest Eli, old, blind, and fat, falls back in shock, breaking his neck, as he hears the news that the Philistines have captured the Ark of the Covenant and that his two wicked sons have been killed (fig. 12).20 When the images allocated to any given faith are so few, each carries a greater freight of meaning, and in this case the highlighting of two major disasters for the Jews could be interpreted as a sign of hostility on the part of the painter and perhaps also of his patron. The treatment of Christian scenes reflects similar preoccupations, although here emphasis is secured by omission and commission alike. The omissions, in a text devoted to calendars and their associated festivals, are indeed striking. The central festivals of the Christian faith—Christmas, Good Friday, and Easter, normally represented by scenes of Christ’s birth, Crucifixion, and Resurrection—are absent. The core of Christianity has, so to speak, been airbrushed out of the picture. Instead, the two images devoted to traditional Christian themes depict the Annunciation, an event alluded to in the Qur an, and the Baptism of Christ—Nestorians believed that this was the moment when Jesus became divine. The latter is of course an event of central theological importance, but the artist has cleverly contrived to drain away its significance (while still ensuring that the scene is recognizable) by opting to illustrate not the Baptism itself but the moments after it. In thoroughly humdrum fashion John the Baptist helps Jesus to put on his coat, while the Savior’s slippers float off downstream. Not surprisingly, the hand of God is absent, though related concepts were known to Islamic painters at the time;21 the Holy Spirit, now metamorphosed into a bird of Chinese appearance, is not placed above Christ’s head but in the middle distance to his right, thereby robbing this motif too of its traditional significance. In much the same spirit, Jesus is overlapped by Muhammad in the scene where they ride together, while on folio 161r (fig. 36) Muhammad triumphantly outfaces the Christians of Najran. The stylistic hallmarks of these paintings are consistent: a preference for a strong graphic line, bright colors, and formulaic compositions involving single figures or tightly massed groups disposed serially along the frontal

20

Eli’s elaborately decorated chair subverts the standard format of an enthronement scene, while also suggesting a Torah shrine. For a brief discussion of this image, see Soucek 1975, pp. 141, 143. 21 See the image of Bahira and Muhammad in the Compendium of Chronicles of Rashid al-Din; D.T. Rice and Gray 1976, pl. 30.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 3F767AC3 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

157

plane. The figures take up most of the space, with only intermittent attempts to develop a more sophisticated setting by means of Far Eastern landscape elements. Instead, the principal decorative accent is provided by obsessively convoluted drapery that seems to take on a life of its own and for which the closest parallels are in Byzantine rather than Islamic art. This feature suggests that the manuscript may have been produced in Mosul, with its preponderantly Christian population. The “Compendium of Chronicles” Nothing in earlier Islamic painting foreshadows the striking innovations found in the next major surviving manuscript, the Compendium of Chronicles of Rashid al-Din (1247–1318), the outstanding Ilkhanid vizier of his time. The earliest, though fragmentary, portion to survive is divided between the Edinburgh University Library and the Khalili Collection in London and is dated 714/1314. The reasons for its importance have been laid out with exemplary clarity and scholarship by Sheila Blair in a series of pioneering publications.22 The Compendium is indeed a phenomenon in itself, in which an ambitious concept was realized with remarkable discipline. Tabriz, the Ilkhanid capital, was perhaps the major metropolis of the contemporary world, a multicultural, multiconfessional, political, and commercial center that served as a bridge between Europe and East Asia. Rashid al-Din had the vision to capitalize on its unique advantages and to set in motion a colossal intellectual enterprise of which the Compendium was merely one facet. He founded a suburb, named Rab c-i Rashidi after himself, that contained within a walled precinct his tomb, a hospice, a Sufi center, a hospital, and the so-called rauda, which had living and teaching accommodation, plus a small mosque, arranged around a courtyard. The entire complex employed more than three hundred people,23 and all its expenses were shouldered by Rashid al-Din himself. The detailed provisions of his endowment deed reveal the meticulous planning that underpinned the entire project.

22

Conveniently listed in Blair 1995, p. 121. Blair 1984, pp. 70, 74, 79.The inflated description (taken from the so-called Letters of Rashid al-Din) of the facilities that is so often quoted and is summarized by Wilber 1955, pp. 20–21, has been shown to be a Timurid forgery. See Morton 1999. 23

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 448F9BB6 /production/pindar/booksxml/hillen/chapter−05

158 The Compendium was commissioned successively by two Ilkhans, Ghazan (r. 1295–1304) and Öljeitü (r. 1304–16), and was nominally written by Rashid al-Din himself, though it is far more likely, given his onerous political and administrative responsibilities, that his major input was financial and editorial rather than authorial. In its final form, dating to about 1310,24 it comprised a history of Ghazan Khan (volume 1), a universal history (volume 2) probably some four hundred folios in length, a survey of the genealogies of the Arabs, Jews, Mongols, Franks, and Chinese (volume 3), and a geographical compendium (volume 4).25 The entire Compendium probably had no fewer than twelve hundred large-format folios and perhaps 540 paintings.26 What was the overall purpose of this work? The original commission makes sense as an attempt to display the Mongol commitment to Iran’s cultural heritage, while also exalting the history of the Mongols themselves.27 Those motives had obvious educational and propaganda dimensions that would be served (as well as contemporary technology allowed) by copying the text as widely as practicable. In addition, the exceptionally large size of the manuscripts made them well suited for display and even for study by several people at once. Rashid al-Din’s plan was to produce a copy, alternately in Arabic and Persian, every six months, not just of the Compendium but of each of his other six works, which dealt respectively with theology, philosophy, zoology, culture, industries, and architecture (four of them in several volumes).28 It is worth considering in practical terms what is implied in executing such a colossal undertaking. After the initial translations had been done, an entire army of scribes, papermakers (there was a mill nearby),29 burnishers, binders, leatherworkers, specialists in materials (inks, brushes, pigments, pens), and painters would have had to work in the closest harmony and under very tight supervision to achieve success. Moreover, they actually had less than six full

24

Boyle 1971b, p. 21. For volume 2, see Blair 1995, p. 27. For volume 4, see Boyle 1971b, p. 21; Blair 1995, p. 115. 26 The severely trimmed original pages of the Edinburgh manuscript measure 17!× 11# inches. The estimated number of paintings excludes entirely formulaic images such as those of the Chinese emperors. 27 The Timurid versions of Rashid al-Din’s text suggest, incidentally, that this section of his original work was indeed illustrated. 28 Blair 1995, p. 115. 29 Wilber 1955, p. 21. 25

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 4440C29B /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

159

months for the job, since this mountain of texts needed to be checked before being dispatched to the Arab or Persian city for which they were destined. These highly unusual production circumstances strongly affected the nature of the illustrations. They made the development of a house style virtually mandatory. They encouraged speed at the expense of meticulous technique, and copying or adapting at the expense of originality.30 Formulaic solutions were sought to the problems of composition, although the images still had to be impressive. To maintain the taxing production quotas, individual artists or groups of artists probably developed special time-saving skills for depicting particular subjects, such as landscapes (fig. 3), groups of almost identical figures, battle scenes (fig. 13), enthronements (fig. 14), or clothing; a given painting might therefore be subcontracted to several painters. The presence of certain extremely repetitive details suggests that the artists were working from painted prototypes which they subjected to successive minor modifications in order to avoid monotony. Both the nature and the placement of the illustrations demand comment, and the two issues are intimately linked. The Edinburgh fragment, the most important one, will serve as our focus. A few singletons apart, its images can be easily divided into only four categories: scenes from the Old Testament, incidents from the life of the prophet Muhammad, battles or sieges, and enthronements. The images are by no means of equal interest visually: those in the first two categories are as varied as those in the last two are stereotypical. That distinction, which is quickly apparent, raises interesting questions about how the book was meant to be viewed, as distinct from being read. Where the illustrations occur is no less interesting. In the Edinburgh fragment, the section on pre-Islamic Arabian prophets and Iranian mythical kings (which comes at the beginning of the manuscript in its present form) and that on the Old Testament are very lavishly illustrated. The opening folios are especially so, with images on both sides of successive folios and sometimes even two on a given side. Thus there are twenty-six illustrations in the first twenty-four folios. Suddenly, however, after folio 24r, this rapid rate of illustration falls off; the next illustration occurs at folio 34r, and the one after at folio 41r. At this point, the text takes a new direction as the section on

30

This could explain some of the bunching discussed below, in which case it seems that speed of production counted for more than the visual consistency achieved by a steady rate of illustration.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 47623A4D /production/pindar/booksxml/hillen/chapter−05

160 the life of Muhammad opens. Here again illustrations are relatively plentiful. After folio 57r there are no illustrations at all for the fifty-one folios of text covering the period of the Rightly Guided caliphs, the Umayyads, and the golden age of the cAbbasids. Yet this is the most eventful and politically successful period of Islamic history up to the time of Rashid al-Din himself. The images begin again only when the text reaches the history of the eastern Iranian world from the late tenth century on—a period in which the Buyids, Samanids, Ghaznavids, and Seljuqs (to say nothing of lesser power blocs like the Simjurids and Qara-Khanids) were successively contending for mastery of these territories. Between folios 108 and 147 there are thirty-three pictures, a rate of illustration almost as intense as at the beginning of the manuscript.31 This bunching probably cannot be explained by regarding the Edinburgh fragment as part of an individualized copy intended for a city in eastern Iran. Nor does it seem plausible that its highly uneven rate of illustration reflects poor organization. It is most likely that the Edinburgh manuscript was “standard issue” and that its bunching of illustrations was part of an overall strategy. So what could this strategy have been? The emphasis on the Old Testament can be explained in several ways. Such scenes would have been entirely appropriate to begin the story of mankind, spotlighting with a kind of visual fanfare the heroic figures of the key prophets and patriarchs of the Judaeo-Christian and Islamic traditions. The episodes depicted were also familiar from the popular commentaries that embroidered the tales of these personages with picturesque detail (fig. 15). Moreover, since Rashid al-Din was himself of Jewish birth, these Old Testament episodes might well have enjoyed a special place in his affections. Finally, Byzantine,32 western European, and Jewish manuscripts (such as the nearly contemporary Golden Haggadah with its extensive Exodus images)33 all provided ample models for such scenes—an especially important factor

31 There are also imbalances in the coverage of the unillustrated sections. The portion on early Islamic history spans a period of well over three hundred years and fills fifty-one folios, while that narrowly devoted to the eastern Islamic (i.e., Persian-speaking) world is substantially shorter (thirty-five folios) and covers a much smaller geographical area and a shorter period of time (some two hundred years).This very uneven approach to both text and illustrations implies a deliberate manipulation of emphasis. In that case, the process of production must have been much more complex than simple copying. 32 Allen 1985. 33 Golden Haggadah 1970.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 42D1198A /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

161

given that the Rashidiyya artists were, it seems, always pressed for time. In the contemporary West, too, the fashionable genre of world histories, as evidenced by the Weltchronik of Rudolf von Ems of about 1260, also featured a lengthy opening section from the Old Testament.34 The numerous scenes of violence and enthronement in the Compendium make some sense, since the major turning points of history are so often key battles and key usurpations of power. But here, too, a breakdown of the material is revealing. From folios 108v to 135r there are twenty-six scenes of actual or implied violence—four sieges, four executions or maimings, and eighteen images of war (fig. 13)— some of them as extreme as the visual language of the time allowed. Such illustrations, conspicuously absent in the earlier part of the manuscript, represent a decisive change in tone. In particular, the effect of this repetition on the reader, whether casually leafing through the text or studying it carefully, should not be minimized. Successive large-scale and vivid images of warriors wreaking havoc—warriors clad in Mongol armor, carrying Mongol weapons and yak-tail standards, and mounted on wiry Mongol ponies—would have rammed home the reality of the Ilkhanids’ military might. They may also be seen as attempts to “Mongolize” the past, imposing a Mongol pattern on pre-Mongol history. It may well be, as Sheila Blair has argued,35 that the artists drew on illustrated histories of the Ghaznavids for this section of the text, just as the illustrations for the Old Testament episodes derived in part from already well-established models. As suddenly as they erupted onto these pages, the warriors disappear, to be replaced in the last six images of the Edinburgh fragment by a series of bland enthronement scenes, possibly deriving from lost illustrated Seljuq histories (fig. 14).36 The rulers are presented as quasi-divine beings, dispensing justice and representing the state. Solemn icons of power, they epitomize the absolute authority of the monarch. Yet here again the message often has a secondary purpose, for the presence of Persian advisers and scribes implies that the ruler does not exercise his power arbitrarily but calls on local people for counsel. An inclusive rather than exclusive pattern of government is thus advertised.

34 35 36

Kratzert 1974; Gunther 1993. Blair 1995, pp. 55, 93. I am grateful to Sheila Blair for this suggestion.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 42BA6BF4 /production/pindar/booksxml/hillen/chapter−05

162 What of the cluster of images dealing with the life of the Prophet? The flirtations of earlier Ilkhans with Buddhism and Christianity had accustomed them to religious images, and thus they had none of the traditional inhibitions of Muslims on this score. Instead, they might well have wished to celebrate, indeed broadcast, their recent conversion by honoring the Prophet and singling him out in this way. The very fact that the images in the Compendium are not isolated but, as in the Biruni manuscript, form a cycle signals a decisive change of pace and intention. However, for all their unmistakable religious content, even spiritual exaltation, these images of the Prophet are found in several different categories of manuscripts. Their location in books of historical, scientific, or epic character is of defining importance: religious painting entered Islamic art by the back door, and this may well have helped to secure its acceptance. In the Iranian world, at least, the moment for images of the Prophet had finally come. The “Mi crajnama” The next and culminating stage of this particular fashion was a predictable one: a manuscript of exclusively hagiographic character. The first illustrated religious manuscript in the history of Islamic art is the fragmentary Mi crajnama (Book of the Ascension) preserved in the Topkapı Palace Library, Istanbul, album H. 2154. Its date is very hard to fix. Its images may well be the work of the artist Ahmad Musa, whom the Safavid librarian Dust Muhammad identifies as the chief painter for Abu Sa cid and to whom he attributes a Mi crajnama (see the passage quoted at the head of this essay). While this identification might date the illustrations to the late Ilkhanid period, they do not correspond in style to any work from that time and could indeed be later.37 The images are for the most part remarkably large: they have been removed from their parent manuscript and are now textless, but some of them might have occupied a full page. They are also brightly colored and of simple composition, which made them well suited for a devotional purpose. Their sustained emphasis on perhaps the most miraculous and otherworldly episode of Muhammad’s life, his nocturnal journey to the seven heavens, underlines the special status of the Prophet. Parallels with

37

Since painting at this time was in a state of flux, and since the subject matter of the Mi rajnama was almost entirely unprecedented in Islamic painting, this discrepancy is not strange. c

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 44BD44E7 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

163

Buddhist apsaras, contemporary Italian (especially Sienese) paintings, and Byzantine icons furnish yet another example of the fruitful interplay of East and West in fourteenth-century painting. But the unmistakable naïveté of this cycle, so different from the pictorial sophistication displayed in the Great Mongol Shahnama, betrays the uncertainty of the artists in tackling such a new and highly charged genre of painting. The stakes were high. Illustrated Copies of the “Shahnama” By far the most popular illustrated text of the period was the Shahnama (Book of Kings), an epic poem of some sixty thousand couplets written by Abu ’l-Qasim Firdausi about 1010. The ten illustrated Shahnamas datable from approximately 1300 to 1350 are all individual creations in the sense that their iconography does not derive from a single model but was fashioned anew for each successive manuscript. However, the obvious need to make the paintings reflect the main stories and emphases of the epic in an appropriate manner ensured that there was a great deal of overlap between the various pictorial programs. The variety in the illustrations was a direct outgrowth of the nature of Firdausi’s text and of the medieval approach to it, which was by no means reverential. The verses were, it seems, orally transmitted, and a standard text did not exist; thus no two medieval Shahnama manuscripts are textually identical. Firdausi himself, after all, did not compose his epic from scratch but assembled a motley series of oral narratives, recasting them as the Shahnama. It is only to be expected that some of the alternative versions, occasionally further lengthened or shortened by the scribes themselves, survived in the oral repertoire and generated illustrations. The differences between the extant Ilkhanid cycles of this epic also partly result from its adaptable nature, which is furthered by the medium of painting, with its capacity to privilege certain aspects of the text above others. The poem can be interpreted as a succession of adventure stories, battles, fantastic episodes, or romances, but also as a guide to ethics, a chronicle, a celebration of royalty, and a manual for royal conduct. In fact, Firdausi’s text is inherently suited to being adjusted, even manipulated, to fit the personal tastes of a patron or to transmit a particular message. The illustrations thus function as a parallel text, highlighting a given theme. And, as the familiar adage has it, a picture is worth a thousand words.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 4BE5966D /production/pindar/booksxml/hillen/chapter−05

164 The surviving manuscripts, most of which have been broken up, have been attributed to various centers—Baghdad, Isfahan, Shiraz, and even India—and not surprisingly reflect several distinct styles. But in format and approach they have much in common. The most frequent pattern is that of a two-volume text containing about a hundred illustrations, each usually taking up between a fifth and a third of the framed text block. The standard illustration format, a horizontal oblong strip, encourages the disposition of figures along the frontal plane in a simple narrative sequence (figs. 17, 18). Yet square and stepped designs (figs. 19, 20) also occur, and sometimes a visual pun is attempted, as when the shape of the painting mimics a plunging shaft in a scene that depicts a well or pit. Since the human figure, in formulaic poses and gestures, dominates the pictorial space, little room is left for ambitious landscapes or highly detailed interiors: mere indications suffice. Bright, strong colors and simple, direct, though often powerful, compositions capture the surface verve of the Shahnama narratives (fig. 21) but are less well equipped to plumb their subtleties. By the same token they discourage experiment in the third dimension, the creation of subplots, and the development of a subsidiary focus of visual interest. Despite isolated references to earlier examples,38 the concentration of illustrated Shahnamas in the first half of the fourteenth century signals a dramatic new departure that cannot be explained by the previous history of the genre. Several theories have been proposed to explain the sudden fashion for such works, which did not all serve identical purposes. Marianna Shreve Simpson has argued convincingly in favor of a teaching, even a propagandist, aim for the Shahnamas that she attributes to Baghdad, a motive that can be invoked for several other Ilkhanid manuscripts.39 After the cultural havoc wrought by the Mongol invasion, in which cities and especially their

38

A. S. Melikian-Chirvani has unearthed a reference from the poet Suzani Samarqandi that he interprets as pointing to the existence of illustrated Shahnamas under the Qara-Khanids in the early twelfth century; he also proves conclusively that Shahnama scenes were painted on the walls of royal palaces in that century. Melikian-Chirvani 1988, pp. 43–45. 39 Simpson points out that the illustrations of the Shahnamas that she assigns to Baghdad stick closely to the text, which may imply an intended readership not very familiar with the poem, whereas the Shirazi painters seem to have felt free to stray from the text and thus may have assumed that their readers knew it well. For a penetrating discussion of the whole topic of text-image relationships, see Simpson 1982a. For other Ilkhanid manuscripts that seem to have served a teaching function, see Fitzherbert 2001, pp. 366–73.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 421C2396 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

165

educational establishments suffered grievously, there was much ground to be recovered in the world of learning. These Shahnamas could also have expressed a new Mongol commitment to the country and culture that they ruled, as suggested by the tiles with Shahnama inscriptions at Takht-i Sulaiman.40 Moreover, the lavish use of illustrations would make the book accessible to Mongol patrons who, even if not illiterate, would derive more pleasure from images than from words. Those Shahnamas made in areas outside direct Mongol political control, most notably in southern Iran, might have been intended to boost national sentiment. The vizier of the quasi-independent Inju dynasty in Shiraz, for example, ordered a celebrated Shahnama that is dated 1341 (fig. 22). Firdausi’s original project, after all, had as one of its aims the assertion of Iranian identity vis-à-vis the alien Arabs—a motive that helps to explain why the poem ends with the last Sasanian king, rather than continuing into the period of Arab domination, and why Firdausi so carefully purged its language of Arabic elements. Thus the illustrated Shahnama could serve the purposes both of a Mongol elite seeking to ingratiate itself with hostile Iranians and of Iranians seeking to reassert their hitherto oppressed but millennial culture, of which this epic is the distillation. The Great Mongol “Shahnama” Given the dominant role of the Shahnama in Ilkhanid painting, it is entirely appropriate that the supreme masterpiece of that school should be the only royal copy of Firdausi’s epic. Not only is it by far the largest of these Shahnama manuscripts, but its paintings in general are by far the most complex and sophisticated, and again the largest, of the entire school (fig. 25). Controversy has surrounded the Great Mongol Shahnama for almost a century. Now only a torso, for it was cut up and mangled for the sake of profit in the early twentieth century, its folios are so widely scattered that a comprehensive study of the manuscript is difficult (figs. 24, 27). Moreover, it was never finished in the first place.41 This may have been due to the hubris of those responsible for its design and execution, who perhaps simply took

40

See Tomoko Masuya’s chapter 4 in this volume as well as the works of recent scholarship cited in it. 41 See Grabar and Blair 1980, pp. 11–12; Blair 1989, p. 128.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 42BE519B /production/pindar/booksxml/hillen/chapter−05

166 on more than they could handle; maybe time or money ran out. Or perhaps the project was cut short prematurely by entirely external factors such as political events, like the death of the vizier Ghiyath al-Din or indeed the fall of the Ilkhanid dynasty itself, a disaster that would have had a direct impact on the royal ateliers. The majestic physical scale of the work, with trimmed pages measuring 16 by 11#inches and originally, when the margins were intact, very much more,42 enabled its illustrators to think big (fig. 26). The original plan for the illustrations was probably equally ambitious and indeed unprecedented, calling for some two hundred paintings, of which fifty-seven survive.43 Since Shahnama iconography was presumably still in its infancy, the painters had no models for scores of the projected images. Even existing models would have required thorough reworking to fit naturally into the unusually large spaces set aside for painting; mere proportional enlargements would not do. The continual pressure to be innovative propelled the painters into unfamiliar territory in search of fresh inspiration. It also required that they react more than mechanically or literally to the text. Such pressures perhaps account for the fact that their paintings speak with many voices. This much may legitimately be inferred from the physical evidence provided by the manuscript itself. But the disappearance of much of the actual text, which may have been discarded by the Belgian dealer Georges Demotte when he broke up the bound manuscript, means that information vital for more detailed analyses is not at hand. For instance, the precise balance between illustrated and unillustrated folios as well as the exact subject matter of the lost or projected paintings remain matters for speculation. No colophon survives to provide key information as to date, provenance, patronage, and perhaps even the identity of the artists. These are formidable losses, but scholarship over the last sixty years or so has crystallized in favor of certain propositions regarding these issues. Among them are a date in the 1330s, though strong arguments have been made for both earlier and later

42 Sarah Bertalan’s remarks on issues of paper and pigments in her technical study in this catalogue suggest that the Great Mongol Shahnama may have had more in common with the Compendium of Chronicles manuscripts than was previously thought; wide margins are only one aspect of this. 43 One of the fifty-eight to survive into the twentieth century (no. 16 in the standard numbering) was destroyed by fire in 1937.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 42DBC79D /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

167

dates;44 a provenance in Tabriz, the Ilkhanid capital; a royal patron, probably Abu Sa cid; and the participation of Ahmad Musa,45 the artist mentioned in the somewhat enigmatic and minimalist account of the history of Persian painting by the Safavid royal librarian Dust Muhammad (the Vasari of Iran), writing in the 1540s (see above). Most previous accounts of this manuscript have emphasized its key significance in fourteenth-century painting without seeking to probe its illustrative program for deeper meanings. In keeping with the Eurocentric approach of scholars from the 1930s onward, which was preoccupied with questions of date, provenance, patronage, and style,46 the tacit assumption was that the purpose of the paintings was to illustrate the accompanying text. Only two scholars, Oleg Grabar and Abolala Soudavar, have made a concerted effort to explain the paintings as something more than a succession of ambitious textual illustrations. Grabar proposed that the cycle of paintings revolved around four themes: death and mourning, legitimacy, human frailty, and divine revelation.47 Soudavar suggested a still more comprehensive interpretation, namely that the Great Mongol Shahnama is the manuscript described by Dust Muhammad as the Abusa cidnama (Book of Abu Sa cid) and that it is nothing less than a daring attempt to reconfigure the Shahnama as a chronicle of the royal Mongol house.48 By this reckoning, every episode chosen to be depicted in the illustrative program was selected because it also served to present some event in recent Mongol history and thus brought Firdausi’s text right up to contemporary Iran. It is a most audacious theory and has thus not

44

For an earlier date, see Soudavar 1996, p. 179. For a later one, see Schroeder 1939,

p. 131. 45 Schroeder admits this possibility but believes that the principal artist among the several painters who contributed to the manuscript was Shams al-Din. Schroeder 1939, pp. 131–32. 46 Both Schroeder and Ivan Stchoukine, for instance, assumed, with somewhat inconclusive results, the connoisseur’s task of distinguishing a series of different hands in these paintings; see Schroeder 1939 and Stchoukine 1958. 47 Grabar 1969. He developed this interpretation further in Grabar and Blair 1980, pp. 13–27. 48 Both of these propositions are controversial, but the second does not depend on the first. See Soudavar 1996.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 45BC1EBC /production/pindar/booksxml/hillen/chapter−05

168 won universal acceptance.49 Nevertheless, Soudavar has been able, thanks to his impressive familiarity with the historical sources, to propose a sequence of remarkably exact correlations between episodes in Firdausi’s epic and Mongol history in the thirteenth and fourteenth centuries. To interpret all of these as accidental would be to stretch the proverbially long arm of coincidence well beyond breaking point. In its essentials, Soudavar’s theory is compelling and has far-reaching implications for the study of Persian painting. In particular, attention must henceforth be paid to the political and propagandist (and perhaps at times religious) dimension in depictions of scenes from familiar texts. (Soudavar also marshals ample evidence to suggest that the finest calligrapher of the time, cAbd Allah Sayrafi, renowned for his Qur ans, was the principal scribe for the manuscript.) This elusive masterpiece also provides copious material for other types of investigation. It pioneers a new complexity in storytelling techniques, with plot and subplot artfully juxtaposed or interwoven, or the main event richly embroidered with complementary detail. The emotional range is wide and expertly orchestrated. The pictures capture the clangor, confusion, and carnage of battle (figs. 1, 25, 29), the gore-spattered scenes of monster slayings evoke a frisson of horror (figs. 23, 28), and the images of death and mourning speak of grief by turns measured and frantic (fig. 40). Yet there is room too for images love and passion, betrayal and fantasy. Many an image is shot through with conflicting emotions, with wonder, or with sly humor. In general, all the surface is painted and all the space is used, thereby actively contributing to the narrative. The three-dimensional world thus created is made all the more credible by the meticulous rendering of detail, from costumes to carpets, from patterned floors to tiled dadoes, from frescoes and balconies to window grilles and lacquer thrones (fig. 30). In this one manuscript the largely lost court arts of the Ilkhanids are on display in concentrated splendor, complementing and enhancing each other to achieve a whole that is much more than the sum of its parts. They bring the technicolor ambience of the Ilkhanid court to vivid life. The incomplete state of the manuscript means that not one of the illustrative cycles that punctuate Firdausi’s poem can be followed pictorially in all its fullness. Yet even these fragments suggest that, quite apart from the

49

Some scholars doubt that Dust Muhammad would have had access to this manuscript and that illustrations so disparate could be made to fit into an overall program; see, for example, Blair 2002b.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 4764F034 /production/pindar/booksxml/hillen/chapter−05

THE ARTS OF THE BOOK IN ILKHANID IRAN

169

pointed references to Mongol history suggested by Soudavar, such cycles worked very successfully within the primary context of Firdausi’s epic. This double life is typical of the multi-layered nature of the manuscript. Thus the paintings speak of patriotism, of succession disputes, of a fierce desire for justice, of the painful road from pride to humility, of the vanity of human wishes. They glorify kings rather than heroes, and sometimes, as in the case of the cycle of Iskandar (Alexander the Great; figs. 31, 32), they do so to an extent not equaled in earlier or later Shahnamas. Taken together, they create a memorable “Mirror for Princes,” a pictorial equivalent to this fashionable literary genre.50 Perhaps nothing captures the boldness of these painters better than their confident appropriation of ideas from other cultures. Tabriz was thronged with European missionaries, Chinese officials, and merchants and diplomats from all over the old world. This cosmopolitanism must account in part for the truly unique openness to Christian images found in the Great Mongol Shahnama. Gospel archetypes such as the Adoration of the Magi, Entry into Jerusalem, Flagellation, Carrying of the Cross, Crucifixion, Deposition, Lamentation, and Entombment are freely adopted and adapted, to say nothing of images of Dominicans, Franciscans, warrior saints, and the Virgin Mary. Along with the recycling of these images, so familiar in another context, come sudden unmistakable echoes of contemporary European fashions and conventions, whether for plaited hair, the depiction of drapery, or the stylized mime of grief. This willingness to echo contemporary Italian and French art is balanced by an equal readiness to copy and refashion elements of Chinese art (fig. 33). The paintings contain details taken from Buddhist images, among them mudras, the trailing leg, the recumbent pose, and the triratna (Three Jewels); dragons; the sacred fungus; architecture; lacquer thrones; screens; and all kinds of Far Eastern textiles, including a Chinese imperial robe for the dying Rustam. Above all, Chinese landscape elements are widely used to create an atmosphere or to comment on the action, rather than simply to provide a background (fig. 34). Their integration with large-scale figures, a flouting of Chinese convention, is nonetheless pictorially and emotionally compelling.

50

This genre, popular in Arabic and Persian literature alike, could take many forms, ranging from ethical guides in the form of animal fables to manuals detailing how kings should behave. All of them used narratives, whether fictional or nonfictional, to convey a moral regarding statecraft.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 45 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 457559F1 /production/pindar/booksxml/hillen/chapter−05

170 The Great Mongol Shahnama is a thoroughly appropriate place to end this survey. At once a climax and a coda, it built on past achievements but at every turn unveiled new vistas. The team of masters who worked on it forged, with intuitive mutual understanding, a new style. But that style was less important in itself than for what it attempted to convey, for it embraced depths of meaning and expression hitherto unknown in Islamic book painting. So ambitious were these artists that they effectively broke the bounds of the medium, taking book art into areas for which it was perhaps unsuited and from which their successors recoiled. Within the covers of this two-volume book, one can trace the sequence from paintings that are simple illustrations51 to ones that are commentaries, then metaphors, and finally independent works of art operating confidently on several levels of meaning. More and more content—descriptive, emotional, historical, symbolic—is gradually pumped into these paintings, and only an absolutely assured command of pictorial language enables the greatest of these painters to control the forces that they unleash (figs. 1, 35). All this visual splendor and intellectual complexity are, however, destined for the eyes of only a very few. Thus to modern viewers there may seem to be an inherent mismatch between the medium and the message. It could be argued that the extremely limited intended readership makes it absurd to suggest that the images of the Great Mongol Shahnama carried a complex freight of politico-historical, let alone symbolic, meaning. But that would be to underestimate the despotic, unaccountable power of the Ilkhan as ruler, and indeed to misunderstand the very nature of Muslim panegyric, whose prime purpose was to exalt that ruler.52 Precisely because the text illustrated in the Great Mongol Shahnama was not written to glorify the Ilkhan Abu Sa cid but rather was written for the ages, the images in this manuscript—however relevant to Mongol history—have a timeless validity. So long as Iran has rulers, so long will the Shahnama remain relevant to its people. Ironically enough, it is only in the past century that these images have come into their own, and then in a way that could scarcely have been foreseen by their original patron. Extracted one by one from their

51

It remains to be seen whether these different approaches were undertaken in any particular order; it is too early, for example, to assume that those at the beginning of the poem were also the first to be painted, as can be suggested grosso modo for those at the beginning of the Shahnama-i shahi of Shah Tahmasp. 52 Meisami 1987, pp. 40–76.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05 Sequential 22

JOBNAME: ???? PAGE: 1 SESS: 32 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0BAEEEEE /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

171

1. Iskandar’s Iron Cavalry, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Harvard University Art Museums, Cambridge, Mass., Gift of Edward W. Forbes (1955.167).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 09B8F459 /production/pindar/booksxml/hillen/chapter−05−artwork

2. Elephant and Rider, illustration from the Diez Albums, Iran (possibly Tabriz), early 14th century. Ink, colors, and gold on paper. Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung (Diez A fol. 71, S. 56).

172

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0F893C18 /production/pindar/booksxml/hillen/chapter−05−artwork

173

3. Mountains between Tibet and India, from the Jami cal-tawarikh (Compendium of Chronicles) by Rashid al-Din, Iran (Tabriz), A.H. 714/A.D. 1314–15. Fol. 261r; ink, colors, and gold on paper. The Nasser D. Khalili Collection of Islamic Art, London (MSS 727).

THE ARTS OF THE BOOK IN ILKHANID IRAN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0BDD29C6 /production/pindar/booksxml/hillen/chapter−05−artwork

174

4. Ardavan Captured by Ardashir, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Arthur M. Sackler Gallery, Smithsonian Institution, Washington, D.C.; Purchase, Smithsonian Unrestricted Trust Funds, Smithsonian Collections Acquisition Program, and Dr. Arthur M. Sackler (S1986.103).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 16F31C3B /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

175

5. Prince, Poet, and Courtiers, from the Anthology of Diwans, copied by Abd al-Mu min al- cAlawi al-Kashi, Iran (possibly Tabriz), A.H. 713–714/A.D. 1314–15. Fol. iv; ink, colors, and gold on paper. The British Library, London (Manuscript 132).

c

6. Sultan, Poet, and Courtiers, from the Anthology of Diwans, copied by Abd al-Mu min al- cAlawi al-Kashi, Iran (possibly Tabriz), A.H. 713–714/A.D. 1314–15. Fol. 88v; ink, colors, and gold on paper. The British Library, London (Manuscript 132).

c

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 5

7. Scene from the Samak-i cayyar (Samak the Knight-Errant) by Sadaqa ibn Abu l-Qasim Shirazi, Iran, ca. 1330s. Fol. 217v; ink and colors on paper, Bodleian Library, Oxford (MS Ouseley 379).

8. Page from the Samak-i cayyar (Samak the Knight-Errant) by Sadaqa ibn Abu al-Qasim Shirazi, Iran, ca. 1330s. Fol. 257r; ink and colors on paper. Bodleian Library, Oxford (MS Ouseley 380).

JOBNAME: ???? PAGE: 6 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 1766ACC4 /production/pindar/booksxml/hillen/chapter−05−artwork

176

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0D21DFC8 /production/pindar/booksxml/hillen/chapter−05−artwork

177

9. Landscape, illustration from the Diez Albums, Iran, 14th century. Ink and colors on paper. Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung (Diez A fol. 71, S. 10).

THE ARTS OF THE BOOK IN ILKHANID IRAN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0D7C3F79 /production/pindar/booksxml/hillen/chapter−05−artwork

178

10. The Simurgh (a mythical bird), from the Manafi c-i hayavan (On the Usefulness of Animals), copied by cAbd al-Hadi ibn Muhammad ibn Mahmud ibn Ibrahim al-Maraghi for Shams al-Din ibn Ziya al-Din al-Zushki, northwestern Iran (Maragha), A.H. 697 or 699/A.D. ca. 1297–1300. Fol. 55r; ink, colors, and gold on paper. The Pierpont Morgan Library, New York (MS M. 500).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 16D0A772 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

179

11. Muhammad Forbids Intercalation, from the Kitab al-athar al-baqiya can al-qurun al-khaliya (Chronology of Ancient Nations), copied by Ibn al-Kutbi, northwestern Iran or northern Iraq, A.H. 707/A.D. 1307–8. Fol. 6v; ink, colors, and gold on paper. Edinburgh University Library (MS. Arab 161).

12. The Death of Eli from the Kitab al-athar al-baqiya can al-qurun al-khaliya (Chronology of Ancient Nations), copied by Ibn al-Kutbi, northwestern Iran or northern Iraq, A.H. 707/A.D. 1307–8. Fol. 133v; ink, colors, and gold on paper. Edinburgh University Library (MS. Arab 161).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 9

13. Mahmud of Ghazna’s Conquest of India, from the Jami c al-tawarikh (Compendium of Chronicles) by Rashid al-Din, Iran (Tabriz), ca. A.H. 714/A.D. 1314–15. Fol. 179v; ink, colors, and gold on paper. Edinburgh University Library (MS Arab 20).

JOBNAME: ???? PAGE: 10 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 1982C1E0 /production/pindar/booksxml/hillen/chapter−05−artwork

180

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 13C1152F /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

181

14. Sultan Berkyaruk ibn Malikshah (r. 1093–1105), from the Jami c al-tawarikh (Compendium of Chronicles) by Rashid- al-Din, Iran (Tabriz), A.H. 714/A.D. 1314–15. Fol. 187v; ink, colors, and gold on paper. Edinburgh University Library (MS Arab 20).

16. Jurjays Miraculously Protected When Tortured by the King of Mosul for Refusing to Worship Idols, from the Jami c al-tawarikh (Compendium of Chronicles) by Rashid al-Din, Iran (Tabriz), A.H. 714/A.D. 1314–15. Fol 26r; ink, colors, and gold on paper. Edinburgh University Library (MS Arab 20).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 11

15. The Death of Moses from the Jami c al-tawarikh (Compendium of Chronicles) by Rashid al-Din, Iran (Tabriz), A.H. 714/A.D. 1314–15. Fol. 294v; ink, colors, and gold on paper. The Nasser D. Khalili Collection of Islamic Art, London (MSS 727).

JOBNAME: ???? PAGE: 12 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0AF1B30F /production/pindar/booksxml/hillen/chapter−05−artwork

182

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0D9D55F4 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

183

17. Buzurjmihr Masters the Game of Chess, page from the First Small Shahnama (Book of Kings), northwestern Iran or Baghdad, ca. 1300–1330. Ink, colors, gold, and silver on paper. The Metropolitan Museum of Art, New York, Purchase, Joseph Pulitzer Bequest, 1934 (34.24.1).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0F50BF00 /production/pindar/booksxml/hillen/chapter−05−artwork

184

18. Zal Visits Rudaba in Her Palace, page from the First Small Shahnama (Book of Kings), northwestern Iran or Baghdad, ca. 1300–1330. Ink, colors, gold, and silver on paper. The Trustees of the Chester Beatty Library, Dublin (Per 104.5).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 14

19. Combat between Rustam and Isfandiyar, page from a copy of the Shahnama (Book of Kings), Iran (Shiraz), A.H. 741/A.D. 1340–41. Ink and colors on paper. Cleveland Museum of Art, Purchase from the J. H. Wade Fund (44.479).

20. Rustam Escapes from Isfandiyar, page from a copy of the Shahnama (Book of Kings), Iran, early 14th century. Ink and colors on paper. Cincinnati Art Musuem (1947.498).

JOBNAME: ???? PAGE: 15 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 12B16A66 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

185

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 15

21. Farud before His Mountain Fortress, illustration from the Diez Albums, Iran (probably Isfahan), ca. 1335. Ink, colors, and gold on paper. Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung (Diez A fol. 71, S. 29, no 2).

JOBNAME: ???? PAGE: 16 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 08E745BB /production/pindar/booksxml/hillen/chapter−05−artwork

186

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0D0ECD97 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

187

22. Execution of Afrasiyab, page from a copy of the Shahnama (Book of Kings) dedicated to the vizier al-Hasan al-Qavam al-Daula wa l-Din, copied by Hasan ibn Muhammad ibn cAli ibn Husain al-Mawsili, Iran (Shiraz), A.H. Ramadan 741/A.D. February–March 1341. Ink, colors, and gold on paper. The Walters Art Museum, Baltimore (W.677b).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 17

23. Bahram Gur Fighting a Wolf Monster, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Harvard University Art Museums, Cambridge, Mass., Bequest of Abby Aldrich Rockefeller (1960.190).

JOBNAME: ???? PAGE: 18 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0A748421 /production/pindar/booksxml/hillen/chapter−05−artwork

188

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0AAE1143 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

189

24. Bahram Gur Hunting Onagers, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Worcester Art Museum, Worcester, Mass., Jerome Wheelock Fund (1935.24).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 19

25. Ardashir Battling Bahman, Son of Ardavan, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Detroit Institute of Arts, Founders Society Purchase, Edsel B. Ford Fund (35.54).

JOBNAME: ???? PAGE: 20 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 099BFE26 /production/pindar/booksxml/hillen/chapter−05−artwork

190

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 09C7FF11 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

191

26. Page of text, from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. The Trustees of the Chester Beatty Library, Dublin (Per 111.8).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 07EDC5E4 /production/pindar/booksxml/hillen/chapter−05−artwork

192

27. Iskandar Emerging from the Gloom, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Keir Collection, England (PP3).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 22

JOBNAME: ???? PAGE: 23 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0ACB0FAC /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

193

28. Bahram Gur Slaying a Dragon, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. The Cleveland Museum of Art, Grace Rainey Rogers Fund (1943.658).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 23

JOBNAME: ???? PAGE: 24 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 094740FF /production/pindar/booksxml/hillen/chapter−05−artwork

194

29. Rustam Shooting an Arrow into Isfandiyar’s Eye, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Harvard University Art Museums, Cambridge, Mass., Gift of Edward W. Forbes (1958.288).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 24

JOBNAME: ???? PAGE: 25 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0B9625FE /production/pindar/booksxml/hillen/chapter−05−artwork

195

30. Zal Approaching Shah Manuchihr, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. The Trustees of the Chester Beatty Library, Dublin (Per 111.4).

THE ARTS OF THE BOOK IN ILKHANID IRAN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 25

31. Taynush before Iskandar; The Visit to the Brahmins, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Arthur M. Sackler Gallery, Smithsonian Institution, Washington, D.C.; Purchase, Smithsonian Unrestricted Trust Funds, Smithsonian Collections Acquisition Program, and Dr. Arthur M. Sackler (S1986.105).

JOBNAME: ???? PAGE: 26 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0DE25CF4 /production/pindar/booksxml/hillen/chapter−05−artwork

196

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 26

JOBNAME: ???? PAGE: 27 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0E1CC4F2 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

197

32. Iskandar Building the Iron Rampart, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Arthur M. Sackler Gallery, Smithsonian Institution, Washington, D.C.; Purchase, Smithsonian Unrestricted Trust Funds, Smithsonian Collections Acquisition Program, and Dr. Arthur M. Sackler (S1986.104).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 27

JOBNAME: ???? PAGE: 28 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 09C09F55 /production/pindar/booksxml/hillen/chapter−05−artwork

198

33. Mihran Sitad Selecting a Chinese Princess, page from the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Museum of Fine Arts, Boston, Helen and Alice Colburn Fund and Seth K. Sweetser Fund (22.392).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 28

JOBNAME: ???? PAGE: 29 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0AC62365 /production/pindar/booksxml/hillen/chapter−05−artwork

199

34. Rustam Slaying Shaghad, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. The Trustees of the British Museum, London (1948.12–11.025).

THE ARTS OF THE BOOK IN ILKHANID IRAN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 29

JOBNAME: ???? PAGE: 30 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 08D3D315 /production/pindar/booksxml/hillen/chapter−05−artwork

35. Faramarz Pursuing the Kabulis, from a page of the Great Mongol Shahnama (Book of Kings), Iran (probably Tabriz), 1330s. Ink, colors, and gold on paper. Muse´e du Louvre, Paris (7095).

200

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 30

36. The Day of Cursing, from the Kitab al-athar al-baqiya can al-qurun al-khaliya (Chronology of Ancient Nations), copied by ibn al-Kutbi, northwestern Iran or northern Iraq, A.H. 707/A.D. 1307–8. Fol. 161r; ink, colors, and gold on paper. Edinburgh University Library (MS Arab 161).

JOBNAME: ???? PAGE: 31 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0D8BDDD7 /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

201

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 31

37. The Author with a Mongol Prince and A Horse and Groom, double frontispiece from the Tarikh-i jahan-gusha (History of the World Conqueror), copied by Rashid al-Khwafi, probably Iraq (Baghdad), finished on A.H. 4 Dhu’l-hijja 689/A.D. December 8, 1290. Fols. 1v, 2r; ink, colors, and gold on paper. Bibliothe`que Nationale de France, Paris (MSS or., Suppl. persan 205).

JOBNAME: ???? PAGE: 32 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 0CE1A778 /production/pindar/booksxml/hillen/chapter−05−artwork

202

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 32

38. The Beast Called Sannaja, from a Kitab c aja ib al-makhluqat wa ghara ib al-mawjudat (Book of the Wonders of Creation and the Peculiarities of Existing Things), Iraq (possibly Mosul), ca. 1295–1310. Fol. 129v; ink, colors, and gold on paper. British Library, London (Or. 14140).

39. Enthroned Couple, left side of a double frontispiece from the Mu nis al-ahrar fi daqa iq al-ash car (Free Men’s Companion to the Subtleties of Poems), copied by the author and compiler, Muhammad ibn Badr al-Din Jajarmi, Iran (Isfahan), A.H. 741/A.D. 1341, Fol. 2r; ink, colors, and gold on paper. Kuwait National Museum, The al-Sabah Collection, Dar al-Athar al-Islamiyyah, Kuwait City (LNS 9MS).

JOBNAME: ???? PAGE: 33 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 1FEE88CD /production/pindar/booksxml/hillen/chapter−05−artwork

THE ARTS OF THE BOOK IN ILKHANID IRAN

203

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 33

JOBNAME: ???? PAGE: 34 SESS: 34 OUTPUT: Tue Aug 10 10:25:22 2010 SUM: 08BDF696 /production/pindar/booksxml/hillen/chapter−05−artwork

40. The Bier of Alexander, from a page of the Great Mongol Shahnama (Book of Kings), probably Tabriz, 1330s. Ink and colours on paper. Freer Gallery of Art, Smithsonian Institution, Washington, D.C. (38.3-1)

204

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-artwork Sequential 34

JOBNAME: ???? PAGE: 1 SESS: 46 OUTPUT: Tue Aug 10 10:26:47 2010 SUM: 4396B8C5 /production/pindar/booksxml/hillen/chapter−05−ends

THE ARTS OF THE BOOK IN ILKHANID IRAN

205

parent volumes, reproduced in color in books and as posters, projected as slides, scanned as images on the Web, viewed by thousands of people at exhibitions, they can now at last be given their due as supreme masterpieces, not of an age but for all time. Conclusion Perhaps the most striking characteristic of the extraordinary half-century or so of achievement in Ilkhanid book painting is the sense of barely contained energy. Artists of this school responded vigorously and imaginatively to the new challenges posed by unfamiliar subject matter and hitherto alien ways of seeing. Content and style became the matrix for the transformation of Persian painting during this brief formative period. The two greatest undertakings of Ilkhanid book painting – the Compendium of Chronicles and the Great Mongol Shahnama—can both be seen as attempts (one public, the other rigorously private) to harness the expressive powers of manuscript illustration to political and propagandist ends. In both cases the attempt miscarried, defeated by the nature of the medium itself, which does not lend itself to being experienced by many people at once. Manuscript painting is an exquisitely selfish art form; to be reminded of that fact one has only to try reading a book together with someone else. Even when the huge dimensions of books made large-scale paintings an option, even when the volume was displayed on a lectern in a mosque or a madrasa and thus visible to a group of readers, it remained in the very nature of a book for the memory of one image to be obliterated by the sight of the next. Referring back to earlier images to refresh the memory or for purposes of comparison is possible for only one reader at a time. The ambitions of Ilkhanid patrons and painters, then, to some extent outran the means at their disposal. The enormous enterprise set in motion by Rashid al-Din, and the measures he took to ensure that it had the necessary financial backing, leave no doubt that the constraints were neither intellectual nor monetary. The sheer size of the books produced under these circumstances was crucial, giving them a palpable presence from the moment one saw them.53 It opened new perspectives in book painting and

53

I have had the pleasure of verifying this proposition by means of the immediate reactions of generations of Edinburgh students who were seeing the Rashid al-Din manuscript for the first time.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 46 OUTPUT: Tue Aug 10 10:26:47 2010 SUM: 1654FB3D /production/pindar/booksxml/hillen/chapter−05−ends

206 pushed that medium to its very limit. Indeed, the next logical step would have been to emancipate painting altogether from the confines of the book by turning to fresco and easel painting. But this was not to be. Sadly, later generations of painters gracefully declined that implied challenge and chose to work on a reduced scale, loading more and more visual content into less and less space. Such pictures invite prolonged, absorbed meditation. Rather than explore the potential for expression and narrative offered by very large and lavishly illustrated books, artists preferred to reduce the number of images and to refine their techniques. And given the beguiling mix of intellectual complexity and visual splendor that marked mature Timurid painting in the following century, who is to say that they were wrong?

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-05-ends Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 42 OUTPUT: Mon Feb 13 12:17:05 2012 SUM: 341ECB7B /production/pindar/booksxml/hillen/chapter−06

VI The paintings of Rashı#d al-Dı#n’s “Universal History” at Edinburgh

E

DINBURGH University Press is shortly to publish a substantial monograph on one of the great treasures of the University Library—a manuscript of the “Universal History” by Rashı#d al-Dı#n, the vizier of the Mongol rulers of Persia, copied and illustrated in 1306. This was the last book written by the late Watson Gordon Professor of Fine Art, David Talbot Rice, and was the culmination of his long-standing interest in the manuscript. The text has been seen through the press by Basil Gray, the noted expert on Persian painting who was formerly Keeper of Oriental Antiquities at the British Museum. The volume will reproduce all seventy miniatures not only in monochrome plates but also in colour, in microfiche form, an unusual feature in fine art publishing. The historical context of these illustrations explains why they have fascinated generations of art historians. They probably originated in Tabriz, which for a brief period of less than a century was one of the key cities of the East and in which Persians and Mongols mingled with Arabs, Turks, Chinese, Byzantines and Venetians. This unique conjunction of races and cultural traditions naturally produced a highly individual art. In the field of manuscript illumination this art, though of striking power and originality, could be termed a hothouse product, owing its very existence to the official scriptoria established at Tabriz by Rashı#d al-Dı#n. Each year copies of his encyclopaedic work, written in Arabic and Persian, were sent to one of the major cities of the Mongol realm. This practice implies a series of highly organised workshops, and indeed contemporary historians have left detailed accounts of the extensive scale of the vizier’s patronage. Such unusual circumstances explain why the style of the Edinburgh manuscript so closely

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-06 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 42 OUTPUT: Mon Feb 13 12:17:05 2012 SUM: 41408254 /production/pindar/booksxml/hillen/chapter−06

208 approximates to that of the copies of the same work in London and Istanbul, which are roughly contemporary with it. They also explain an otherwise puzzling feature of this style—its virtual disappearance within less than twenty years. Its demise coincides with the downfall and execution of Rashı#d al-Dı#n himself in 1318. With him collapsed the whole administrative apparatus which had underpinned his scriptoria. The Tabriz style could not survive the loss of the patronage which had created it, for it did not have its roots in the native Persian tradition. This latter conclusion applies also, though with diminished force, to the subject-matter of the manuscript. In keeping with its title—literally “Compendium of Histories” —the text ranges freely over the Old and New Testaments and the Apocrypha, the Qur’an, the life of Muhammad, the legendary Persian past, and recent Islamic history. The London manuscript also illustrates Buddhist and Indian legends. No subsequent Persian illuminated manuscript attempted to encompass such a variety of subjects. Even when the artists of the “Universal History” illustrated subjects which were later to be the mainstay of Persian painting their approach was marked by a vigour, inventiveness and sense of grandeur which makes later versions of these subjects seem rather tame. It is fascinating to see the incidents of the Sha¯hna¯ma—the “Book of Kings”, the Persian national epic—transfigured in this way, even though one sees them in the distorting mirror of an alien tradition. For once the heroes of Persian legend are indeed heroic, not feeble puppets. Enthronements and Battles In the later part of the manuscript, which deals with Islamic history, the artists were no longer obliged to ransack their imaginations for suitable illustrations to the text. Two types of scene dominate this section: enthronements and battles. For depictions of the enthroned ruler the artists could draw on a millennial Near Eastern tradition, familiar in Persia from pre-Achaemenid to Sasanian and early Islamic times. The seated ruler was always shown as larger than the serried rows of standing courtiers flanking him, and was distinguished by the appurtenances of rank. Apart from an almost Western appreciation of the positive qualities of empty space, which dictates the grouping of figures, the rendering of these tableaux in the Rashı#d al-Dı#n manuscript adds little to this tradition. It is quite another story with the battle scenes. Dominated as they are by Mongol faces, costume,

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-06 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 42 OUTPUT: Mon Feb 13 12:17:05 2012 SUM: 41B3C681 /production/pindar/booksxml/hillen/chapter−06

PAINTINGS OF RASHID AL-DIN’S “UNIVERSAL HISTORY” AT EDINBURGH

209

weapons and horses, they testify to the traumatic effect on the Persian mentality of repeated invasions by Mongol hordes. Galloping horsemen erupt with explosive force from the sides of the page and crash together in a hideous tangle of lances, swords and maimed limbs. In some details one is reminded of Uccello’s “Rout of San Romano”, also a careful rendering of the field of battle but one which, with its toy figures and dream-like atmosphere, lacks the frightening urgency of these scenes. Not since the Sasanian bas-reliefs had Persian artists devised so powerful a formula for battle, and in later years they never succeeded in matching it. In some ways the varied subject-matter offers the key to the style of the manuscript. Even if all the artists were Persians—and this has often been disputed—there would have been no suitable models in the native tradition for many scenes of this manuscript. It is thus scarcely surprising that the inevitable borrowing should have extended also to style and technique. Very few examples of Persian painting of the pre-Mongol period exist, but they suffice to highlight the startling novelty of the Tabriz style. The unusually large size of many of the pictures invites speculation that the presumably Chinese models included not only book illustrations but also paintings on silk or even frescoes. This generous scale—the pages measure about 17 inches by 12 inches, and some of the paintings take up half that area—allows the artists to achieve a monumentality that would be almost impossible in small-scale paintings, and they take advantage of it to infuse their compositions with drama and action. The large format also facilitates the introduction of new landscape elements taken from Chinese sources, such as cloud forms resembling scrolls, water rendered as a series of interlaced curlicues, mountains of pyramidal form receding one behind the other into the distance, and trees in which knotted branches and gnarled boles are deliberately emphasised. A landscape infused with these elements need no longer serve simply as the background to the action. It can mirror the fantasy, solemnity, or, occasionally, as the “Death of Sha¯gha¯d” shows, the suffering of the main scene. No wonder that the earliest pure landscape in Persian painting, without any figures, appears in one of the Rashı#d al-Dı#n manuscripts. A century later, all the savage power of these landscapes had been forgotten, to be replaced by the delicate fairytale world thenceforward synonymous with Persian painting.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-06 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 42 OUTPUT: Mon Feb 13 12:17:05 2012 SUM: 32617EAE /production/pindar/booksxml/hillen/chapter−06

210 Calligraphic Mastery of Line A less obviously Chinese trait in these miniatures is the calligraphic mastery of line. Figures are drawn with the utmost economy and finesse, but the draughtsmanship is often nervous rather than smooth and they can be instinct with tension. This is because the artist’s treatment of line is unpredictable. He may interrupt the sweeping fall of a robe by a knot of drapery with an outline like broken waves, or reinforce a gesture by a swathe of jagged drapery aligned in the same direction. This style depends on drawing rather than painting, for colour is confined to areas where it can have maximum impact. Many of the miniatures resemble tinted drawings. The artist makes two colours do the work of a dozen simply by placing them selectively and by varying their intensity. Red and blue are the commonest primary colours. The eyes of the figures are rendered in silver, now tarnished. They have no pupils—apparently a minor detail, but one which invests the figures, otherwise expressionless and devoid of emotion, with a sinister, spectral air. It is difficult to overrate the importance of these miniatures as documents of cultural history, iconography and style. They transcend the multiple borrowings of subject-matter and technique from Western and Far Eastern sources which their unique historical context encouraged. It was the high ambition of the artists and patrons of Tabriz greatly to extend the range of Persian miniature painting not only in subject-matter but also in emotional depth. With their first efforts they scored an unqualified success. Subsequent artists shrank from emulating this extraordinary achievement and kept to a safer and lesser range of subjects and tone. It is a sad irony that the very qualities of the Tabriz style should have made it effectively a dead end for later Persian painting.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-06 Sequential 4

JOBNAME: ???? PAGE: 1 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 450CE246 /production/pindar/booksxml/hillen/chapter−07

VII Mamlu¯k and I¯lkha¯nid Bestiaries: Convention and Experiment ERY few miniatures from Mamlu¯k manuscripts have been analyzed in detail, which makes it difficult to generalize about that school as a whole.1 Yet it should at least be possible to situate a given cycle of Mamlu¯k

V

1

The exceptions are as follows: 1)The Automata of al-Jazarı#, dated 755/1354, of which the mutilated original is in Istanbul: Süleymaniye Library, no. 3606. See the translation and commentary by D. R. Hill: The Book of Knowledge of Ingenious Mechanical Devices (Kita¯b fi ma crifat al-h.iyal al-handasiyya) by Ibn al-Razza¯z al-Jazarı#, Dordrecht, 1974, and the review of it by J. M. Rogers, Biblioteca Orientalis, v. 33, no. 5–6, 1976, pp. 358–63. 2) The Niha¯yat al-su cl wa l-umniyya fi ta callum a cma¯l al-furu¯siyya by Muhammad al-Aqsara¯ #ı, British Library, Ms. Add. 18866. See G. Tantum, “Muslim Warfare: A Study of a Medieval Muslim Treatise on the Art of War,” in R. Elgood, ed. Islamic Arms and Armour, London, 1979, pp. 187–201, and G. R. Smith, Medieval Muslim Horsemanship: A Fourteenth-Century Arabic Cavalry Manual, London, 1979; and, on the Chester Beatty ms., D. James, “Mamluke Painting at the Time of the ‘Lusignan Crusade’, 1365–70,” Humaniora Islamica, v. 2, 1974, pp. 73–87. 3) The Maqa¯ma¯t of al-H . arı#rı# in the Nationalbibliothek, Vienna (A.F.9). See K. Holter, “Die Galen-Handschrift und die Makamen des H . arîrî der Wiener Nationalbibliothek,” Jahrbuch der Kunsthistorischen Sammlungen in Wien, N.F. 11, 1937, pp. 1–48, Special Number 104. 4) The Maqa¯ma¯t of al-H . arı#rı# in the British Library, Or. 9718. See J. D. Haldane, “A H . arı#rı# Manuscript and Its Relevance to Mamlu¯k Painting,” unpublished Ph.D. thesis, University of Edinburgh, 1973, pp. 127–70, figs. 45–122. 5) The Oxford Kalila wa Dimna of 755/1354 (Bodleian Library, Pococke 400). See E. Atıl, Kalila wa Dimna: Fables from a Fourteenth-Century Arabic Manuscript, Washington, D.C., 1981. 6) The undated Munich Kalila wa Dimna. See H.-C. Graf v. Bothmer, Kalila und Dimna. Ibn al Muqaffa’s Fabelbuch in einer mittelalterlichen Bilderhandschrift, Wiesbaden, 1981. 7) The manuscript of the Sulwa¯n al-Mut.a¯ c in a private collection in Kuwait has been

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 551E7CED /production/pindar/booksxml/hillen/chapter−07

212 miniatures fairly precisely within three distinct contexts: first, other roughly contemporary Mamlu¯k painting; second, its own probable models; and finally, thematically comparable cycles of the period produced elsewhere in the Islamic world. Perhaps this method will throw into sharper focus a hitherto neglected cycle of Mamlu¯k animal paintings: the illustrations of the copy of the Kita¯b al-H.ayawa¯n of al-Ja¯h.iz., now in the Biblioteca Ambrosiana in Milan (Ms. 140 Inf. S. P. 67), first published by Löfgren and Lamm in 1946.2 The fragmentary nature of the manuscript has not been sufficiently considered in assessing the thirty-two paintings it contains. Its eighty-seven large quarto folios comprise less than one-tenth of the text, and a pro rata calculation would therefore suggest that it once had, or was intended to have, more than three hundred illustrations.3 It is worth emphasizing that no Mamlu¯k manuscript of this quality and number of illustrations is known. The Kita¯b al-H.ayawa¯n is undated but is customarily placed at about 1350.4 Its provenance is similarly unrecorded, as are the names of its scribe and patron. However, the number of illustrations, their unusually large size, and the generous use of gold all point to a wealthy patron. Löfgren even suggests that he might have been a sult.a¯n.5 But so far no evidence indicates that a single published by A. S. Melikian-Chirvani: Sulwa¯n al-Mut.a¯ c fi cUdwa¯n al-Atba¯ c. A Rediscovered Masterpiece of Arab Literature and Painting, Kuwait, 1985. 8) The Escorial Bestiary, a manuscript of the Mana¯fi c al-H . ayawa¯n (The Usefulness of Animals) by Ibn al-Duraihim al-Maus.ilı#, has been published in color facsimile in a limited edition by a Spanish bank. For a good survey of some of the wider problems posed by Mamlu¯k painting, see Rachel Astor (now Ward), “The Sources of Mamluk Painting,” unpublished M.Phil. thesis, Oxford, 1981. 2 “Ambrosian Fragments of an Illuminated Manuscript Containing the Zoology of ˘ a¯h.iz.,” with a contribution by C.J. Lamm, ‘‘The Miniatures: Their Origin and Style,” al-G Uppsala Universitets Årsskrift, v. 5,1946, pp. 7–39; all the miniatures are illustrated, and Löfgren translates the relevant portion of the text for each. 3 Löfgren, “Ambrosian,” p. 20. 4 On p. 35 Lamm claims—without giving the necessary evidence—that it is characteristic of the fourteenth-century Mamlu¯k style. On p. 38 the implication is that he dates the ms. between ca. 1300 and ca. 1350. R. Ettinghausen, Arab Painting, Geneva, 1962, p. 156, dates the paintings “to the second quarter of the fourteenth century,” and, without resolving the contradiction, to “the middle of the fourteenth century.” J. D. Haldane dates them “mid to late 14th century,” (Mamluk Painting, Warminster, 1978, p. 78), and A.R. Zaky to the fourteenth century (“Medieval Arab Arms,” in Elgood, Islamic Arms, p. 210). 5 Löfgren, “Ambrosian,” p. 21.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 5A4F6498 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

213

one of the thirty Mamlu¯k manuscripts currently known was produced for the ruler himself,6 although several Mamlu¯k sult.a¯ns ordered lavishly illuminated copies of the Qur a¯n.7 The implications of this contrast are indeed far-reaching. They cannot be pursued here, but it does seem clear that, in the case of secular book painting, patronage was exercised at a markedly lower social level than it was in Iran during the same period. Indeed, only one of the thirty extant Mamlu¯k manuscripts was indisputably produced for an amı#r—and his rank was not of the highest.8 In the absence of specific textual evidence to the contrary, one must conclude that the Ambrosian Ja¯h.iz. manuscript was probably not made for a member of the ruling elite. Following the approach outlined above, this paper will attempt—after a brief consideration of the text and its effect on the pictorial cycle—first to assess the place of these paintings within the Mamlu¯k school of the fourteenth century. It will then demonstrate their marked, indeed crippling, dependence on thirteenth-century Mesopotamian models. Finally, it will analyze the slightly earlier but much more innovative animal paintings in the Pierpont Morgan Bestiary of the 1290s in order to show how in the I¯lkha¯nid pictorial tradition this apparently moribund genre could be renewed and by about 1370 produce such subtle and complex masterpieces as the Istanbul Kalı#la wa Dimna. In order to stress the truly international character of such images, at least as far as their literary origins are concerned, material from Western medieval bestiaries will be cited where relevant. Although these Western bestiaries depend ultimately on the Physiologus,9 whereas the Kita¯b

6

Though E.J. Grube suggests that the Kita¯b al-Makhzu¯n Ja¯mi c al-Funu¯n, Paris, Bibliothéque Nationale, ms. arabe 2824, dated 875/1470, might have been produced for Sult.a¯n Qa¯ it Bay; see his “Pre-Mongol and Mamluk Painting,” in B. W. Robinson, ed., Islamic Painting and the Arts of the Book, London, 1976, p. 116. 7 All earlier studies have now been superseded by D. James, Qur a¯ns of the Mamlu¯ks, London, 1988. 8 This was Na¯s.ir al-Dı¯n Muh.ammad walad al-jana¯b al- ca¯lı# al-marh.u¯m H . usa¯m al-Dı#n T arant a ¯ # ı , who ordered the Oxford Maqa ¯ ma ¯ t, dated 738/1337, Bodleian Library, Marsh 458. . . See Haldane, Mamluk Painting, p. 83, with full bibliography. 9 See M. R. James, ed., The Bestiary: Being a Reproduction in Full of the Manuscript Ii.4.26 in the University Library, Cambridge, with Supplementary Plates from Other Manuscripts of English Origin, and a Preliminary Study of the Latin Bestiary as Current in England, Oxford, 1928, p. 2; and F. McCulloch, Mediaeval Latin and French Bestiaries, Chapel Hill, 1962, pp. 15–20 (the Greek Physiologus) and pp. 21–44 (the Latin Physiologus). For the text, see Physiologus, tr. from the Latin by M.J. Curley, Austin, 1979.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4520E189 /production/pindar/booksxml/hillen/chapter−07

214 al-H.ayawa¯n derives at multiple removes from Aristotle,10 the origin of both texts in ancient Greek literature is itself a significant bond between them, and the additional material that both texts subsequently acquired is sometimes very similar. The link between the standard Western bestiary text and that of Ibn Bukhtı#shu¯ c (as reproduced for example in the Pierpont Morgan manuscript) is closer still.11 Relation of the Illustrations to the Text How much light can the text of the Kita¯b al-H.ayawa¯n shed on the pictures that accompany it? The answer seems to be surprisingly little. Like much other work by al-Ja¯h.iz., it is an idiosyncratic mixture of anecdote, verses, gossip, reasoned discussion, moral saws, and humorous comment.12 This is not to say that the book’s tide is entirely misleading: there are several chapters, for instance, devoted entirely to the dog.13 But the discussion focuses only intermittently on animals. Instead, there is much material on human foibles, theological controversies, philosophy, history, and what may loosely be described as science. The passages on animals are enough to fill a respectably long book, but they have to be winkled out from a mass of entirely unrelated or only vestigially related material. Thus, the inordinate length of the Kita¯b al-H.ayawa¯n—seven books in all—is deceptive from the zoological point of view.

10

Löfgren, “Ambrosian,” p. 18. For examples, compare details from the standard Western bestiary as translated by T. H. White, The Bestiary: A Book of Beasts, New York, 1954, with others from the text of Ibn Bukhtı#shu¯ c, partially translated in D. Stewart, Early Islam, Weert, 1967, s.v. unicorn/ rhinoceros (White, p. 21; Stewart, p. 139), elephant (White, p. 26; Stewart, p. 134), and camel (White, p. 80; Stewart, p. 137). Cf. also nn. 101–2 below for a more extended correspondence. 12 It is most easily accessible in the substantial anthology from it translated by C. Pellat, The Life and Works of Ja¯.hiz., tr. from the French by D. M. Hawke, London, 1969, pp. 130–85, to which may be added his comments on the work in the Encyclopedia of Islam, 2nd ed., s.v. al-Dja¯h.iz., col. 386. 13 See F. Viré, EI2, s.v. Kalb, pp. 490–91; G. R. Smith and M. A. S. Abdel Haleem, The Book of the Superiority of Dogs over Many of Those Who Wear Clothes by Ibn al-Marzuba¯n, Warminster, 1978, pp. xxvii, xxix-xxxi, xxxiii-xxxiv; Pellat, Life, pp. 136–47; and Löfgren, “Ambrosian,” pp. 15,17, 25. 11

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 5C7A9C41 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

215

It could be argued that this prolix and chaotic arrangement of the subject matter presents problems only to the reader. It need not shackle the painter. Even in lavishly illustrated medieval Islamic texts, after all, the pictures do not succeed one another at precisely predictable intervals.14 As long as the text contains enough material appropriate for illustrations, runs the argument, the painter can be well satisfied; the arrangement of that material is a secondary concern. Such an argument is especially appropriate for manuscripts that contain no more than a few highly finished pictures.15 There the artist can afford to wait for the right subject. As it happens, the Kita¯b al-H.ayawa¯n is not such a manuscript, for it is liberally furnished with illustrations. Manuscripts in this category were subject to certain conventions. One such convention dictated that, as in any modern coffee-table book, the illustrations should be spaced fairly evenly throughout the book. Another laid a premium on images that were immediately legible. A third governed their size, ensuring that the picture space rarely exceeded half a page and thereby interlocking text and illustration. It would not be difficult to find exceptions to these rules, but their general validity can be safely asserted. Thus, it would not have been sensible for the painter of the Kita¯b al-H.ayawa¯n to aim at producing lavishly finished full-page pictures. Indeed, his manuscript subjected him to constraints that were utterly at variance with that aim. He was committed to 14

To cite only examples of Mamlu¯k date, the Sulwa¯n al-Mut.a¯ c manuscript in a private collection in Kuwait, which occasionally allows a dozen to fifteen pages to elapse without a picture though usually only three folios or fewer separate the pictures; the unclassified Furu¯siyya ms., dated 767/1366, in The Chester Beatty Library, Dublin, where the gap is usually two or three folios but can sometimes be as wide as ten; and the Oxford Maqa¯ma¯t, dated 738/1337, Bodleian Library, Marsh 458, where there are two substantial gaps of fourteen to sixteen folios between illustrations otherwise never more than six folios apart—all have illustrations at irregular intervals (see Haldane, Mamluk Painting, pp. 59,43,83, respectively). In the Kita¯b al-H . ayawa¯n, while the pictures normally follow one another on successive folios or with a gap of one folio, there are also gaps of eight, seven (twice), twelve, and twenty-three folios. These gaps are only partly due to the fact that the manuscript has been rebound in some disorder. Cf. Löfgren, “Ambrosian,” p. 16. In the case of the Maqa¯ma¯t genre, the rate of illustration can be plotted from Appendix 2 in O. Grabar, The Illustrations of the Maqamat, Chicago and London, 1984, pp. 161–65. 15 Such as the early thirteenth-century manuscripts of the Kita¯b al-Agha¯nı# or that of the Rasa¯ il Ikhwa¯n al-S . afa¯, dated 686/1287 (Ettinghausen, Arab Painting, plates on pp. 58, 65, 98–99, 101). Such choice manuscripts with deliberately few illustrations are much more common in the Byzantine than in the Islamic tradition. This may reflect the perennial Islamic desire for lavish rather than sparse decoration.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4D5DAAD6 /production/pindar/booksxml/hillen/chapter−07

216 turning out a steady flow of somewhat simplified illustrations, probably several hundred in all, and this accentuated his dependence on the text. Unfortunately, it was not only the layout of the contents but the contents themselves that caused him difficulties.16 Such a statement may sound paradoxical and therefore demands closer examination. The Kita¯b al-H.ayawa¯n contains numerous lengthy digressions on subjects that do not lend themselves to illustration, such as the nature of good and evil, translation and bilingualism, prolixity and conciseness, happiness, and the reasons for Zoroaster’s success.17 Such digressions may be interrupted by a few sentences mentioning an animal. Then the painter pounces.18 Curiously enough, this decision to illustrate asides has sometimes meant ignoring a more detailed account of an animal that on the face of it would be more suitable for illustration. Such a procedure invites the suspicion that the artist did not welcome the pictorial challenge offered by these detailed descriptions. Perhaps he preferred a summary and casual reference because it could most appropriately be illustrated in an equally summary manner. The foregoing remarks may help to explain why the relationship between text and picture in this manuscript is neither as intimate nor as fruitful as might have been expected. This conclusion is all the more disappointing because no other medieval illustrated versions of the Kita¯b al-H.ayawa¯n are known. But it is precisely the lack of a directly relevant model that would go far to explain the exemplary and unspecific nature of some of these illustrations. The contrast with contemporary Iranian book painting is instructive. There the period from about 1290 to 1340 saw the production of the earliest illustrated versions of numerous texts, notably the Usefulness of Animals by

16

The distribution of pictures in the Kita¯b al-H . ayawa¯n is as follows: ff. 1–10 have eleven pictures; ff. 11–20 have three; ff. 21–30 have seven; ff. 31–39 have one; ff. 40–50 have six; ff. 51–60 have one; and ff. 61–64 have three. Thus, of the three heavily illustrated areas the first, ff. 1–10, has easily the most pictures. In the case of the Pierpont Morgan Bestiary (Pierpont Morgan Library, New York, Ms. 500), the same trait can be seen taken to excess, for the first eleven folios of that manuscript contain the pictures that are generally considered the finest. In the small Sha¯hna¯ma manuscripts customarily dated in the early fourteenth century it seems to have been a standard operating procedure to place the miniatures three pages apart. See M. S. Simpson, The Illustration of an Epic: The Earliest Shahnama Manuscripts, New York, 1979, pp. 250–51. 17 For a representative list of headings, see Pellat, Life, pp. viii–ix. 18 For example, Löfgren, “Ambrosian,” p. 31.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 70F314B0 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

217

Ibn Bukhtı#shu¯ c,19 al-Bı#ru¯nı#’s Chronology,20 the World History of Rashı#d al-Dı#n,21 and the Sha¯hna¯ma,22 but including also such lesser manuscripts as the T.a¯nsu¯qna¯ma,23 the Ta rı#kh-i Jaha¯n Gusha¯ of Juvainı#,24 the Marzuba¯nna¯ma,25 the Dı#wa¯ns of Mu cizzı# and others,26 and the Mu nis al-Ah.ra¯r.27 These texts

19

Among the illustrated versions of this text, the undoubted masterpiece is the Persian translation produced in Mara¯gha in the 690s/1290s (Pierpont Morgan Library, New York, Ms. 500). The manuscript contains two dates. One clearly reads 690/1291, but in the other date, while there is no doubt about the tens, the letters that give the unit are illegibly blurred. This manuscript is mentioned in most of the handbooks on Persian painting but is still largely unpublished. The earliest account of it to appear in print was that of A. Yohannan, “A Manuscript of the Mana¯fi c al-H . aiawa¯n in the Library of Mr. J. P. Morgan,” Journal of the American Oriental Society, v. 36,1917, pp. 381–89. For the related manuscripts, see R. Ettinghausen, Studies in Muslim Iconography I: The Unicom, Washington, 1950, p. 163. Particularly important are the illustrated leaves from a dispersed manuscript of the Mana¯fi c-yi H . ayava¯n datable to ca. 1300. In a bibliographical tour de force, E. J. Grube has assembled the extremely scattered literature on these individual leaves; see his Persian Painting in the Fourteenth Century: A Research Report, Supplemento n. 17 agli Annali dell’ Istituto Orientale di Napoli, v. 38, no. 4,1978, p. 12. This work is perhaps the one indispensable guide to this complex period, and it contains the fullest discussion yet published of the Morgan Bestiary, pp. 5–11. To Grube’s references for the leaves of the dispersed manuscript may be added Sotheby & Co., Catalogue of Highly Important Oriental Manuscripts and Miniatures: The Property of the Kevorkian Foundation, 7 April 1975, p. 9, no. 17, and 12 April 1976, p. 8, no. 11, with further references; M. L. Swietochowski, “Persian Painting,” Metropolitan Museum of Art Bulletin, v. 36, no. 2, 1978, pp. 6–7; and E. J. Grube, in Islamic Painting, ed. Robinson, p. 133 and pl. 13 (lot no. 17 of the Sotheby sale of 7 April 1975). 20 See P. P. Soucek, “An Illustrated Manuscript of al-Bı#ru¯nı#’s Chronology of Ancient Nations” in P.J. Chelkowski, ed., The Scholar and the Saint, New York, 1975, pp. 103–68. 21 For general treatments of the two major manuscripts of this text, see D. Talbot Rice, The Illustrations to the “World History” of Rashı#d al-Dı#n, ed. B. Gray, Edinburgh, 1976, and B. Gray, The World History of Rashid al-Din: A Study of the Royal Asiatic Society Manuscript, London, 1978. For a discussion of the related manuscript in Istanbul see G. Inal, “Artistic Relationship between the Far and the Near East as Reflected in the Miniatures of the ˘ ami c at-Tawa¯r ¯ıh . . .,” Kunst des Orients, v. 10, no. 1–2, 1976, pp. 115–20,122–40, and Gray, G ˘ for related leaves in the same museum, in the album H.2153, see B. World History, p. 18; Karamagˇaralı, “Camiu ct-tevarihcin bilinmeyen bir nüshasina âit dört minyatür,” Sanat Tarihi Yıllıgˇı, v. 2, 1966–68, pp. 70–80 and German text on pp. 81–86. Finally, there are the similar leaves in the Berlin albums: M. S. Ips¸irogˇlu, Saray-Alben. Diez’sche Klebebände aus den Berliner Sammlungen, Wiesbaden, 1964, pp. 15–32. 22 The whole problem of the earliest illustrated Sha¯hna¯mas has now been tackled by Simpson, Illustration. 23 B. Gölpinarlı and A. Süheyl Ünver, Tansuknamei Ilhan der fünunu ulûmu hatai mukaddimesi, Istanbul, 1939.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 5E93C11B /production/pindar/booksxml/hillen/chapter−07

218 generated a flood of new images. Again and again the artist was stimulated by the text to produce some novel composition or some unexpected combination of familiar elements.28 New relationships between text and picture were explored. Moreover, the existence of several versions of the World History, some now broken up and scattered, and the numerous copies of the Sha¯hna¯ma, culminating in the Demotte manuscript, show that certain texts were already popular. Despite this popularity, however, which in other Islamic painting traditions—notably that of Mamlu¯k Egypt—was apt to occasion a certain staleness and repetition,29 the inventive powers of the Iranian artists scarcely slackened. Formulaic images certainly occur, especially in battle30 and enthronement31 scenes, but much is new. Viewed against this wider contemporary context, the painter of the Kita¯b al-H.ayawa¯n can clearly be seen to have squandered, not to say funked, a golden opportunity.

24 R. Ettinghausen, “On Some Mongol Miniatures,” Kunst des Orients, v. 3, 1959, pp. 44–52 and figs. 1–2. 25 Simpson, Illustration, pp. 273–80, 283–87, 290, 292–97, 299–301, 303–8 and pls. 109–12; M. S. Simpson, “The Role of Baghda¯d in the Formation of Persian Painting,” in Art et Société dans le Monde iranien, ed. C. Adle, Paris, 1982, esp. pp. 94–106. 26 B. W. Robinson, Persian Paintings in the India Office Library, London, 1976, pp. 3–10. 27 Mirza Muhammad Qazvini, “An Account of the Mu nis ul-Ah.ra¯r: A Rare Persian Manuscript Belonging to Mr. H. Kevorkian,” Bulletin of the School of Oriental Studies, v. 5, 1928–30, pp. 97–108; E.J. Grube, Muslim Miniature Painting, Venice, 1962, pp. 39–42 and pls. 30–31; B. W. Robinson, The Kevorkian Collection: Islamic and Indian Illustrated Manuscripts, Miniature Paintings and Drawings, unpublished report prepared for the Trustees of the Kevorkian Foundation at the Metropolitan Museum of Art, New York, 1953, p. 13. 28 The most innovative manuscripts in this respect were the Sha¯hna¯ma, al-Bı#ru¯nı#’s ¯ tha¯r al-Baqı#ya, and the Ja¯mi c al-Tawa¯rı#kh by Rashı#d al-Dı#n. al-A 29 Examples are the Furu¯siyya and Automata texts. 30 This is true not only within a single manuscript (e.g., the selection of battle scenes in the Edinburgh Rashı#d al-Dı#n manuscript, Rice, Illustrations, pls. 38–41, 43–44, 48–49, 51, 53–61) but also from one manuscript to another. Compare the battle scenes from the Rashı#d al-Dı#n manuscript with typical ones from the various illustrated Sha¯hna¯mas (Simpson, Illustration, pls. 55 and 67; 56, 68, and 72; 97 and 98) or with the depiction of a siege on f. 93b of the Bı#ru¯nı# manuscript (color plate in D. Barrett, Persian Painting of the Fourteenth Century, London, 1952, pl. 1). 31 The generalizations expressed in n. 30 apply with still greater force in this case. Even sparsely illustrated manuscripts, such as the Ta rı#kh-i Jaha¯n Gusha¯ or the Marzuba¯nna¯ma, manage to find room for images of seated figures of authority.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4B7A034F /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

219

One may note in passing another consequence of the painter’s somewhat confined imagination—the undue prominence given to human beings in these illustrations. It is, of course, hazardous to generalize on the basis of less than 10 percent of the projected total of illustrations. Nevertheless, it does seem slightly inappropriate to illustrate a book entitled Kita¯b al-H.ayawa¯n with so many pictures of human beings.32 Some 30 percent of the pictures33 are devoted wholly or in part to human beings, and it must be conceded that the incidents that these pictures depict are on occasion quite trivial (fig. 1).34 The text gives ample warrant for depicting only animals. It therefore seems possible that the artist chose to paint scenes with human beings in them because models for such scenes were ready to hand, notably in illustrated manuscripts of the Maqa¯ma¯t.35 His practice in this respect accords entirely with the desire, manifested so obviously elsewhere in the book (fig. 16),36 to save time and effort whenever possible. The Context of Fourteenth-Century Mamlu¯k Painting The position of the Kita¯b al-H.ayawa¯n within fourteenth-century Mamlu¯k painting may be quickly summarized. The apparent lack of royal, or even high official, patronage for these manuscripts has already been mentioned. Probably as a direct result of this, almost no Mamlu¯k manuscripts of the fourteenth century can match the quality of the paintings produced by the contemporary royal ateliers of Iran. Such ateliers ensured not only the best quality of work that could be had at the time but also its continuity. No such continuity can readily be detected in the best Mamlu¯k work of the period. To identify this “best work,” of course, demands value judgments that are intrinsically hazardous; but they must be made. Probably only one manuscript of the period would be generally accepted as being of the first

32 The illustrations depict a total of twenty-three people, thirty animals, twenty-three birds, and twelve fish. 33 Namely, those on ff. 18a, 19b, 20b, 25a, 26a, 29a, 36a, 40b, 41a, 63b. 34 E.g., f. 36a, showing an Umayyad governor reading a message; or f. 40b, showing a domestic interior with three women, illustrating a single sentence in the text: “A woman complained about her husband and told of his ignorance in intercourse with women.” For these scenes, see Löfgren, “Ambrosian,” p. 26 and pls. I and II, respectively. 35 F. R. Martin, The Miniature Paintings and Painters of Persia, India and Turkey, London, 1912, pls. 21 (below), 22 (below), and 24 (above). 36 Especially audience and court scenes.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4F75CF54 /production/pindar/booksxml/hillen/chapter−07

220 rank—the Vienna Maqa¯ma¯t. This very isolation raises serious questions about its provenance and patronage. Its lavish use of gold at a time when gold was in short supply37 is shared by only one other fourteenth-century Mamlu¯k manuscript, the Oxford Maqa¯ma¯t, which indeed most closely rivals its quality. A third manuscript, the Escorial Bestiary, is related to both of these works in that its miniatures have a plain, completely colored background (fig. 2).38 To this trio should probably be added the Paris Kalı#la wa Dimna, which is easily the most inventive and accomplished version of this picture cycle in Mamlu¯k painting.39 These four manuscripts seem to stand well above other contemporary Mamlu¯k manuscripts in quality. The Kita¯b al-H.ayawa¯n may be placed in the category immediately below these and alongside such work as the Oxford Kalı#la wa Dimna, the damaged British Library Maqa¯ma¯t of 1323,40 and the Sulwa¯n al-Mut.a c in Kuwait and Washington. Its most successful pictures, such as the depictions of the ostrich, the giraffe, and the mottled goat (figs. 3–5), rank with the best of Mamlu¯k animal painting; but this level is not maintained, for the artist is too prone to repeat himself in his choice of animals and poses alike.

37

E. Ashtor, A Social and Economic History of the Near East in the Middle Ages, London, 1976, p. 291 and, for an analysis of the changed situation in the fifteenth century, p. 324. This lavish use of gold is therefore another reason for dating the Kita¯b al-H . ayawa¯n ms. to the fourteenth century rather than later. 38 Far too little attention has been paid to this important manuscript of the Mana¯fi c al-H . ayawa¯n (The Usefulness of Animals) by Ibn al-Duraihim al-Maus.ilı#. The major study of its paintings so far is now somewhat outdated: E. de Lorey, “Le Bestiaire de l’Escurial,” Gazette des Beaux Arts, 6e série, v. 14, 1935, pp. 228–38. See also Haldane, Mamluk Painting, pp. 50–51 and n. 1 above. 39 No specialized study of this manuscript has yet been published, though its place vis-à-vis other contemporary versions of the text has been established; see M. S. Walzer, “The Mamlu¯k Illuminated Manuscripts of Kalı#la wa-Dimna,” in R. Ettinghausen, ed., Aus der Welt der islamischen Kunst. Festschrift für Ernst Kühnel zum 75. Geburtstag am 26.10.1957, Berlin, 1959, pp. 195–206. 40 These paintings are substantially unpublished; the stray references are gathered in Haldane, Mamluk Painting, p. 64. Of the manuscripts in this second category, the Bodleian Kalı#la wa Dimna is certainly the closest to the Kita¯b al-H . ayawa¯n: the pictures are of about the same size and other similarities include the use of a plain ground and a high standard of animal drawing. See the study by Esin Atıl referred to in n. 1 above.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4845D0B7 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

221

Use of Mesopotamian Models This repetition might be predictable in an artist who was illustrating a book that had never been illustrated before and whose forte was not inventiveness. If this were the case, there would be no need to embark on a marathon search for his models. Nevertheless, even if no pictorial tradition had developed for the Kita¯b al-H.ayawa¯n itself, there were models aplenty in other illustrated texts for all kinds of animal depictions. A painter of this quality and technical assurance would surely have had access to such books. He would also, one might add parenthetically, have encountered a quite different source of animal depictions, namely those on pottery. From Fa¯t.imid times onwards Egypt and, to a lesser extent, Syria had produced strikingly effective images of animals in several styles. Some of these animals played a secondary role in narrative scenes and were depicted with a competent degree of realism.41 More impressive, however, are the dishes in which the animals are the central if not the only theme and are subject to extremes of abstraction. The constant aim is to formulate a recognizable and lifelike image that can nevertheless be accommodated without visual strain within the plunging circular picture space of a dish. Two major modes may be distinguished: the heraldic and the silhouette. The heraldic mode adopted a formal, static, frontal pose well exemplified in a Fa¯t.imid luster dish depicting an eagle with outstretched wings (fig. 6).42 This convention clearly did not recommend itself to miniature painters. The silhouette manner was more promising from the manuscript painter’s point of view (fig. 7). Unlike the heraldic mode, it frequently employed stylized vegetal forms as a background accompaniment to, and almost a commentary on, the animal itself.43 The animal or bird is shown in profile, frequently in lithe movement

41

E.g., the celebrated luster dish depicting a cock-fight: E.J. Grube, Islamic Pottery of the Eighth to the Fifteenth Century in the Keir Collection, London, 1976, pp. 138–41 and color plate opposite p. 136. 42 Color plate in R. Ettinghausen et al., Islamic Art, Bulletin of the Metropolitan Museum of Art, N.S. 33, Spring 1975, tenth unnumbered page. The fullest development of this mode of ceramic decoration was reached in fifteenth-century Valencia; see anon., “Valencian Lusterware of the Fifteenth Century: An Exhibition at The Cloisters,” Bulletin of the Metropolitan Museum of Art, N.S. 29, no. 1, 1970, pp. 25, 29–32, and color photographs on front and back covers. See also G. Scavizzi, Maioliche dell’ Islam e del Medioevo occidentale, Milan, 1966, pls. 41–42, 44–45, 47. 43 A. Lane, Early Islamic Pottery, London, 1947, pls. 28B, 29B, 79B.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4E64F0C1 /production/pindar/booksxml/hillen/chapter−07

222 (fig. 8).44 These characteristics would naturally commend such images to book painters, yet there is no sign that such cross-fertilization took place. Perhaps the idea of transferring such images from one medium to another did not occur to book painters. Or they may have felt that the animals on pottery were too stylized to be employed in book painting, with its more realistic bias. They may have shied away from the problem of recomposing for use on a flat page images whose extreme disproportion was appropriate for the compositionally awkward three-dimensional form of pottery. Perhaps such pottery was produced for a different clientele; if so, this alone may have hindered free transmission of motifs between the two media. Moreover, such types of pottery were not current in Egypt or Syria in 1350—though it seems significant that thirteenth-century Syrian book painting on animal subjects betrays little if any sign that the painters were familiar with contemporary Raqqa pottery. All and all, it looks as though pottery and painting developed in separate compartments in thirteenthcentury Egypt and Syria. Exactly the opposite case prevailed in Iran.45 If, then, the Ambrosian painter was not by nature an original and inventive artist, often sidestepping the challenge of devising entirely new images, and did not learn from painted pottery, his most likely source of inspiration would have been other illustrated books. On what sources did he depend? And how did he use them? The Syrian and Iraqi schools of the thirteenth century are the only prototypes that merit serious consideration. The H.arı#rı# manuscripts that were the staple of the Baghdad and Syrian schools are rich in animal depictions;46 in fact, the paintings of camels by Yah.ya¯ al-Wa¯sit.#ı in the so-called Schefer H.arı#rı# of 1237 in Paris (B.N. Arabe 5847) rank among the finest animal paintings in the Islamic world.47 Animal depictions also occur sporadically in the many manuscripts of the text De Materia Medica of

44

E. Atıl, Art of the Arab World, Washington, 1975, pp. 76–77. Perhaps the narrative element found quite frequently in Salju¯q times in luster and mı#na¯ #ı pottery and in luster tile-work facilitated artistic interchange with book painting. For typical examples, see A. U. Pope and P. Ackerman, eds., A Survey of Persian Art from Prehistoric Times to the Present, London and New York, 1939, pls. 642, 660, 664, 672, 674–75, 686–89, 703A, 706–9. For related book painting, see A. S. Melikian Chirvani, Le Roman de Varqe et Gol˘sa¯h, Arts Asiatiques, v. 22 (numéro spécial), Paris, 1970, figs. 1–65. The subject warrants more detailed attention than it has so far received. 46 The best convenient survey is to be found in Ettinghausen, Arab Painting, pp. 104–24. 47 Ettinghausen, Arab Painting, pls. on pp. 116–17, 119. 45

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 6C0CA2A0 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

223

Dioscorides,48 the Book of Fixed Stars (Kita¯b Suwar al-Kawa¯kib al-Tha¯bita),49 and the Book of Antidotes (Kita¯b al-T.irya¯q)50—the Byzantine Theriaca of Nicander51 in Islamic dress. But there were also several illustrated texts largely devoted to animals: the Book of Farriery (Kita¯b al-Bait.ara),52 the Description of Animals (Na ct al-H.ayawa¯n),53 the Wonders of Creation ( cAja¯ ib al-Makhlu¯qa¯t),54 and of 48

See H. Buchthal, “Early Islamic Miniatures from Baghdad,” Journal of the Walters Art Gallery, v. 5, 1942, pp. 21–22, 27, 30. For the best group of illustrations in color, see Atıl, Art, pp. 53–60, especially 59, with bibliography. This was the principal but not the only illustrated work of this kind; in the twelfth century Mans.u¯r al-Su¯rı# wrote a work on botany and commissioned a painter to accompany him on his field trips and to paint the plants in color at the different stages of their growth. See A. Mohiuddin, “Muslim Contribution to Biology,” Proceedings of the International Symposium on Islam and Science, Islamabad, 1981, p. 92. 49 E. Wellesz, “An Early Al-S . u¯fı# Manuscript in the Bodleian Library in Oxford: A Study in Islamic Constellation Images,” Ars Orientalis, v. 3, 1959, figs. 18, 19, 26–34. 50 For illustrations from the manuscripts in Vienna and Paris, see B. Farès, Le Livre de la Thériaque, Cairo, 1953, second unnumbered color plate, figs. 9–10 and pls. XIII–XV. 51 For the famous illustrated version of this text in the Bibliothèque Nationale, Paris, see H. Omont, Miniatures des plus anciens manuscrits grecs de la Bibliothèque Nationale, Paris, 1929, pp. 34–40 and pls. LXV–LXXII. The manuscript is undated but is commonly attributed to the tenth century. 52 Ettinghausen, Arab Painting, pp. 97, 100. Cf. also n. 84 below. 53 The paintings of this manuscript, which are probably of thirteenth-century date, remain virtually unpublished. See K. Holter, “Die islamischen Miniaturhandscriften vor 1350,” Zentralblatt für Bibliothekswesen, v. 54, no. 1–2, 1937, p. 14; H. Buchthal, O. Kurz, and R. Ettinghausen, “Supplementary Notes to K. Holter’s Check List of Islamic Illuminated Manuscripts before A.D. 1350,” Ars Islamica, v. 7, 1940, p. 53; and Buchthal, “Early Islamic Miniatures,” pp. 34–37, 39. 54 Perhaps the finest thirteenth-century version is that in the Bayerische Staatsbibliothek in Munich, C. arab. 464 (Aumer 616), which is dated 678/1280 (color illustrations most easily accessible in Ettinghausen, Arab Painting, pp. 138–39). An even more relevant example in the case of the Milan Kita¯b al-H . ayawa¯n is the so-called “Sarre Qazwı#nı#,” now divided between the Freer Gallery and the New York Public Library; see Atıl, Art, pp. 115–31 and J. A. O. Badiee, “An Islamic Cosmography: The Illustrations of the Sarre Qazwini,” unpublished Ph.D. thesis, University of Michigan, 1978. Finally, see the copies in the British Library (Ms. Or. 14140) datable to the early fourteenth century; the slightly later copy in Gotha (ca. 1330–40; Forschungsbibliothek, Ms. A1506); and the fragmentary copy recently sold at Sotheby’s in London, convincingly dated to the early fourteenth century and attributed to a Mamlu¯k painter. (For these three manuscripts, see S. Carboni and A. Contadini, “An Illustrated Copy of al-Qazwı#nı#’s The Wonders of Creation,” in Sotheby’s Art at Auction, London, 1990, pp. 228–33; for their comments on the Ambrosiana manuscript, see pp. 232–33.) For an attribution of the fragmentary copy to the third quarter of the thirteenth century and to Syria, see Sotheby’s Oriental Manuscripts and Miniatures. 26 April 1990, London, 1990, Lot 182, pp. 112–20.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4D4051C3 /production/pindar/booksxml/hillen/chapter−07

224 course the fables of Bidpai (Kalı#la wa Dimna).55 These would be even more obvious sources for this painter to use (indeed, it would have been hard for any animal painter of the fourteenth century to have ignored them), although they did not represent—as did the illustrated Maqa¯ma¯t—the high-water mark of Mesopotamian painting. Despite the different emphases of these various texts, their common focus on animals encouraged the establishment of certain pictorial conventions for such pictures.56 These were inherited by Iranian and Mamlu¯k painters alike. The varying responses of the painters to these inherited conventions expose the basic character of each school. In Iran these conventions served merely as a base upon which to build. In Mamlu¯k paintings, however, they were apt to define the painter’s scope and ultimately to circumscribe his vision. Thus, a misplaced fidelity to Mesopotamian pictorial conventions could result in fossilized technique and iconography. This can be demonstrated at least in the case under discussion as well as in the fourteenth-century Mamlu¯k Kalı#la wa Dimna manuscripts in Munich, Paris, Oxford, and Cambridge.57 Happily, this tendency does not affect all Mamlu¯k animal painting; the Escorial Bestiary, for instance, is an honorable exception, showing that the use of inherited conventions did not preclude the creation of an assured and splendid image. But the trend seems all too clear.

55 See the studies by Atıl and von Bothmer mentioned in n. 1 above. For treatments of other relevant illustrated versions of this text see: M. S. Walzer, “Mamlu¯k Illuminated”; M. S. Walzer, “An Illustrated Leaf from a Lost Mamlu¯k Kalı#lah wa-Dimnah Manuscript,” Ars Orientalis, v. 2, 1957, pp. 503–5; H. Buchthal, “‘Hellenistic’ Miniatures in Early Islamic Manuscripts,” Ars Islamica, v. 7, 1940, pp. 128–33 and figs. 27, 33, 42, 44, 46; P. Waley and N. Titley, “An Illustrated Persian Text of Kalila and Dimna Dated 707/1307–8,” The British Library Journal, v. 1, no. 1, 1975, pp. 42–61; J. M. Rogers, F. C q agˇman, and Z. Tanındı, The Topkapı Saray Museum: The Albums and Illustrated Manuscripts, London, 1986, pp. 47–48, 50–51; J. S. Cowen, “Drama and Morality in Two Early Mongol Illustrated Sequences from Kalila wa Dimna,” Oriental Art, N.S. 30, no. 2, 1984, pp. 167–77; eadem, Kalila wa Dimna. An Animal Allegory of the Mongol Court: The Istanbul University Album, New York, 1989. See also n. 98 below. 56 Such conventions might, of course, be much older than the thirteenth century and might have originated elsewhere than in Mesopotamia. For the purposes of the present argument, however, the salient fact is that they did exist in mature form in thirteenthcentury Mesopotamian book painting and are thus the most obvious source that can be postulated for similar Mamlu¯k painting. 57 The case made by Sofie Walzer, “Mamlu¯k Illuminated,” passim, is irrefutable.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 477872BF /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

225

Some of these conventions may now be examined in greater detail. Perhaps the most striking feature is the absence of a background. The picture is painted directly onto the paper, without frame or background color (fig. 9). This practice—a standard one in illuminated Byzantine scientific works,58 which themselves perpetuated classical models—had numerous side effects. Most significant was its impact on the principal subject matter. It rendered the image neutral and timeless, distancing it and frequently investing it with a heraldic quality, which is, incidentally, echoed in certain Mamlu¯k blazons.59 In a sense this formality was, to borrow a term from literary criticism, an “objective correlative” to the type of text—usually a bestiary—in which the animal is characterized by a few Theophrastic phrases and deliberately not set in a real-life context. Certain practical advantages accrued from the absence of a closely defined background. Borrowing from other media such as ceramics and textiles was facilitated. Attention was made to focus exclusively on the image. Obviously, the entire picture was quicker and easier to execute than one of similar quality but with a detailed background. Occasionally—as in the Escorial Bestiary—the image was formalized still further by the use of a monochrome background, a device that implied a more carefully finished painting and thus laid a certain onus on the artist. A second widespread convention governed the size of the animal vis-à-vis the picture as a whole. Here a distinction must be drawn between narrative and non-narrative scenes. In narrative scenes—such as abound in Kalı#la wa Dimna manuscripts, where the function of the text is to tell a story rather than describe an animal—the presence of two or more protagonists inevitably requires a reduction in scale, often accentuated by the need to provide at least a rudimentary setting.60 In images with no narrative intention, however, such as many illustrations of the Kita¯b al-H.ayawa¯n, the great size of the animal in relation to the whole picture space can be most marked (fig. 10). The same process can be seen in illustrated Byzantine scientific manuscripts: compare the Theriaca of Nicander, which uses small

58

K. Weitzmann, Late Antique and Early Christian Book Illumination, London, 1977, fig. VI and pls. 18–19. 59 E.g., that of Sult.a¯n Baibars; see L. A. Mayer, Saracenic Heraldry, Oxford, 1933, pp. 9, 106–10, and pl. 1. 60 For typical examples see Haldane, Mamluk Painting, pls. 50, 59–63.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 44C93C16 /production/pindar/booksxml/hillen/chapter−07

226 figures in a realized landscape,61 with the large, dominating, and essentially didactic pictures of plants in a typical Dioscorides manuscript.62 In the case of the Kita¯b al-H.ayawa¯n it is as if the artist had nothing else to say. Certainly, the image is too big to allow any secondary theme to develop, even though the text often gives ample occasion to introduce such themes. Only in a single case is a second animal depicted—without the express authorization of the text63—perhaps because the models that the artist was using did not have one, though the desire to keep secondary detail to a minimum is an abiding characteristic of this artist. The image has a larger-than-life quality. This convention too, then, operates as a distancing agent. A third convention with the same effect is the preference for a profile depiction (fig. 11). At one stroke this makes the animal a topos and cuts it off from the viewer. It eliminates the possibility of depicting a varied interaction between animals, for example a female with her young. The artificiality of this convention is emphasized by the fact that the animals almost always face left (fig. 12). It must be admitted, however, that this convention excellently suits the linear style of the Ambrosian manuscript. Interestingly, while the profile mode is virtually de rigeur for animals, it is sedulously avoided in depicting human beings (fig. 13). Of the thirty-two pictures in the Ambrosian manuscript, ten show people. Twenty-three people are shown in these pictures, and, although their poses frequently demand a profile rendering (fig. 14), only one of them is shown thus (fig. 15). Byzantine influence, though obviously operating at several removes, must be reckoned with here. It was above all in Byzantine art that various shades of meaning had gradually come to be associated with full-face, three-quarterview, and profile renderings of people. Clearly, the negative connotations that the profile view had long since acquired in Byzantine religious iconography64—it is the mode normally used for Judas—were sufficiently strong to restrict its use in Islamic manuscripts such as this one. It would be

61 The manuscript is in the Bibliothèque Nationale, Paris (Ms. grec Suppl. 247) and may date from the tenth or eleventh century. For a color illustration, see A. Grabar, Byzantium: Byzantine Art in the Middle Ages, tr. B. Forster, London, 1966, p. 150, pl. 8. 62 Atıl, Art, p. 54; Ettinghausen, Arab Painting, pp. 72–73. 63 This is the bird perching in a “tree” on f. 26a, Löfgren, “Ambrosian,” pl. VI. 64 O. Demus, Byzantine Mosaic Decoration, London, 1948, p. 8.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 52146057 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

227

possible to infer from this evidence of familiarity with Christian iconography65 that the artist used models of the Mosul school, whose connection with Syro-Jacobite painting has long been known.66 Painters of the Baghdad school, on the other hand, which represents an alternative tradition, freely used the profile mode. But the grouping of the Ambrosian figures is, paradoxically enough, reminiscent of works of the Baghdad school.67 There seems no reason to doubt that works of both schools could well have been available to a Mamlu¯k painter, and he took from each tradition what suited him. The profile is also associated with certain stereotyped poses. Thus, in animals of canine type the extended nearside hind leg is always cocked in the form of a rectangle, and the right front paw is nearly always lifted off the ground (fig. 16). The artist is not above purloining this pose for other animals, either in its entirety (e.g., the hare on f. 5a—fig. 17) or in part (e.g., the lion on f. 3a—fig. 18). The excessive use of such formulas brings into question the much-vaunted realism of Islamic animal painting, at least in the Ambrosian manuscript. The artist is curiously cavalier in his use of the ground-line, so much so that it seems he has not fully grasped what that convention implies. Frequently plants grow directly on the ground-line while animals float uneasily above it (fig. 19); sometimes their feet cut right through it. It is hard to see why, having adopted this admittedly arbitrary but already longhallowed schema,68 the artist should fail so signally to use it in the accepted way. The ground-line itself is interpreted in the most literal way as a thin straight line, though in a couple of folios at the end of the manuscript this is 65 Features in this manuscript with an arguably Byzantine ancestry include: knotted curtains, haloes, serried ranks of figures, bolsters, drapery conventions, fantasy architecture, and so on. But it seems likely that many stages of transmission cannot be plotted with precision because of the lack of relevant dated manuscripts. 66 H. Buchthal, “The Painting of the Syrian Jacobites in Its Relation to Byzantine and Islamic Art,” Syria, v. 20, 1939, pp. 136–50; G. de Jerphanion, Les Miniatures du Manuscrit Syriaque No. 559 de la Biblliothèque Vaticane, Vatican City, 1940; J. Leroy, Les manuscrits syriaques à peintures conservés dans les bibliothèques d’Europe et d’Orient, Paris, 1964. 67 This is especially true of the tendency to cluster the figures in tightly-knit groups; for parallels in paintings of the Baghdad school, see D. James, “Space-forms in the Work of the Baghdad Maqa¯ma¯t Illustrators, 1225–58,” Bulletin of the School of Oriental and African Studies, v. 37, no. 2, 1974, fig. 1 (cf. p. 315) and Ettinghausen, Arab Painting, pp. 118–19. 68 It is found in the earliest Byzantine manuscripts, such as the purple codices; see Weitzmann, Late Antique, pp. 24–33.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 41EB910A /production/pindar/booksxml/hillen/chapter−07

228 replaced by an equally straight thick slab (fig. 20)—perhaps an inspiration that came too late to be much use. Admittedly, this ground-line convention is hostile to the depiction of multiple planes; indeed, illusionism is better served when it is absent. The limited and unimaginative way in which the ground-line functions in this manuscript exemplifies the persistent desire to pare the picture to the bone. Economy of time and effort seems to have been the artist’s watchword.69 Mention of the ground-line invites a more general discussion of the nature and function of the landscape elements in this manuscript. Unless the text calls for some particular feature of this kind—such as the pool of water in the scenes showing a boar or crocodile and various fishes (figs. 21 and 22)—landscape is rendered almost telegraphically by a few large, stylized plants, never more than four in all. Thus, the images seem to represent an uneasy marriage of a bestiary with a Dioscorides manuscript. This is most obvious when the plants grow taller than the elephants or horse and ass that they flank (fig. 23).70 A similar comment applies to the unintentionally absurd way in which a large vegetal organism, not yet recognizably a tree or a plant, seems at first glance to function as a luxuriant tail for the giraffe on f. 26a (fig. 4). The plants cannot be recognized in the way that the animals can, which emphasizes their symbolic and compositional role and suggests that in executing them the artist metaphorically kept his eyes shut. This is even truer of one form that seems to be this artist’s own invention—a curious type of vegetation that resembles a tall, flexible post crowned with a bud (fig. 24).71 Its effect is to provide a much more dominant visual accent than do the other plants; indeed, it functions in an almost architectural fashion. The fact that the artist was not constrained to render nature, as he was in Dioscorides manuscripts for instance, may help to explain the varied forms of these ideographic plants.

69

Exactly the same characteristics can be traced in the principal Mamlu¯k furu¯siyya manuscripts, especially those in London, Dublin, Paris, and Istanbul. For a useful general discussion of these manuscripts, with detailed bibliographical references, see Grube in Islamic Painting, ed. Robinson, pp. 115–17. 70 Folios 64a and 42a respectively, corresponding to Löfgren, “Ambrosian,” pls. XXIII and IV. 71 Found on folios 23a, 29a, 44b, and 63b; see Löfgren, “Ambrosian,” pls. XIb, VII, Va, XXII.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 47C59CB6 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

229

The pictures usually contain two plants, one on each side of the central image (fig. 25). Never are they allowed to cluster, as in nature, or to obscure the image. Even when several of them are laid out at intervals throughout the picture, it is the animal that overlaps them, not vice versa (fig. 26). Their placing is itself enough to establish that their principal role is to frame the scene.72 Indeed, they often sway and curl in obedience to the outline of the animal in question—though this feature could also be interpreted as an attempt to disguise their otherwise patently obvious framing function. These plants are most obtrusive in depictions of a single creature whose outlines form a square or a vertical rectangle—that is, in a strongly centralized composition.73 Conversely, they are used with most freedom, are most summary, and indeed are most likely to be omitted altogether, in images that are elongated horizontally (fig. 27). Occasionally they occur singly, especially if they are placed opposite massed blocks of color (fig. 28).74 It seems clear that when the artist laid out his pictures, he thought of these plants as blocks, and the number and density of their leaves reinforce this image. The almost exclusive emphasis on plants as a decorative accompaniment to the main theme, and the fact that these plants are subjected to a high degree of formal abstraction in which their compositional role is often paramount, suggests that the depiction of landscape was not the artist’s prime concern. These features find parallels in other media of Islamic art, where the question of landscape does not even arise. Thus, it may be a mistake to interpret the plants in the Ambrosian paintings primarily as landscape indicators. The similarities between the Ambrosian paintings and those on the pottery of Fa¯t.imid Egypt and Raqqa have already been noted in the context of animal painting. These similarities are relevant also to the use of vegetation in the Ambrosian pictures in that the latter also illustrate, but in a new or at least an abbreviated form, the ancient motif of an animal amidst foliage. In the Islamic minor arts of pre-Mamlu¯k times, for example in ivory and metalwork, this foliage obeys a circular or undulating rhythm and thus

72 This is most tellingly shown by their use in a picture showing three fishes (folio 28b, corresponding to Löfgren, “Ambrosian,” pl. XXb); apparently the artist did not trouble to reflect whether such plants were appropriate in an aquatic scene. 73 Folio 8a; see Löfgren, “Ambrosian,” pl. XIV. 74 Folios 40b and 18a respectively, corresponding to Löfgren, “Ambrosian,” pls. II and VIII.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 454DFD4B /production/pindar/booksxml/hillen/chapter−07

230 provides a convenient frame for the image.75 It is tempting to derive such forms from the classical and early Byzantine “inhabited” or “peopled” scroll,76 and this seems especially justifiable when the vigorous survival of Coptic traditions in early Islamic Egypt is taken into account. Such plants may therefore have a largely compositional function as filler devices or space dividers. Clearly, the vegetation in the Ambrosian pictures can be interpreted in several different ways. These are not, however, mutually exclusive. In particular, its use as a framing device seems beyond dispute. However, only the sides of the image are framed in this way; there is no corresponding closing-off device at the top, though sometimes the tips of the plants incline inwards to suggest a roof.77 It is perhaps worth asking why the artist was content to let the text serve this function of providing the upper “margin” and yet on the whole chose not to use the text below the picture in the same way. It can scarcely be maintained that the scenes that actually are so treated (and incidentally lack the symbolic vegetation) suffer as a result. Even in such cases, however, the base of the picture is apt to be so straight that the artist seems to have worked from a faint ruled line that he later erased—so that a ground-line implicitly divides text from picture even when absent.78 One may conclude that the ground-line was an inherited convention that the artist used almost absent-mindedly and indeed seems occasionally to have forgotten to use altogether. Elements that functioned as notations of rhythm, interval, and space division were more congenial to him. Where the opportunity presents itself, the framing plants are replaced by secondary— and textually redundant—human figures flanking the major one.79 Clearly,

75

Notably in oliphants; see E. Kühnel, Die islamischen Elfenbeinskulpturen, Berlin, 1971. For metalwork, see D. Barrett, Islamic Metalwork in the British Museum, London, 1949, p. 32. 76 J. M. C. Toynbee and J. B. Ward-Perkins, “Peopled Scrolls: A Hellenistic Motif in Imperial Art,” Papers of the British School at Rome, v. 18, 1950, pp. 2–43; C. M. Dauphin, “Inhabited Scrolls from the IVth to the VIIth Century A.D. in Asia Minor and the Eastern Provinces of the Byzantine Empire,” unpublished Ph.D. thesis, University of Edinburgh, 1974. 77 E.g., f. 29a; see Löfgren, “Ambrosian,” pl. VII. 78 E.g., f. 36a; see Löfgren, “Ambrosian,” pl. I. 79 This is true of the enthronement scenes in both folios 25a and 36a; see Löfgren, “Ambrosian,” pls. XIII and I.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 48D967C4 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

231

the tripartite composition came naturally to this artist. It was one of several formulas on which he could ring a few changes and thus avoid having to rethink a composition afresh. A word about the picture space available to the artist concludes this analysis of conventional elements in the Ambrosian manuscript. It seems likely that at the outset a decision was made to keep the illustrations of this text compositionally simple and accordingly to pare the image to its essentials. It would follow, then, that they could usually be accommodated without hardship in the central one-third of the page, with substantial bodies of text above and below. This is exactly what happened. Any much larger picture space might have revealed their essentially simple—not to say simplistic—quality in none too favorable a light. This generous allocation of space gave the image room to breathe, so to speak, in contrast to the small illustrations in a cramped space that are so often found in Kalı#la wa Dimna manuscripts. Occasionally, two separate images occupy a single page, though divided by a slab of text; in such cases they take up the whole length of the page between them. Occasionally, too, an image takes up most of the page. Although such pictures would lose very little by being reduced to half their present size, they would look absurd if they occupied twice as much space on the page. There is no need to labor the difference between these images and, say, the more ambitious Maqa¯ma¯t manuscripts of thirteenth-century Syria and Iraq. In such Maqa¯ma¯t illustrations the text is frequently reduced to a single line, which functions as the notional base line for the picture. With this minor exception, the picture becomes essentially a full-page image, with a concomitant complexity of layout. Moreover, in such manuscripts the picture size often varies significantly from page to page, thereby encouraging a whole range of relationships between text and illustration, and incidentally making the book visually more stimulating, for the dimension of surprise is introduced. By contrast, the ratio of text to picture on any given page of the Ambrosian manuscript is fixed, if the surviving illustrations provide a fair sample. This practice is quite understandable in the case of such an extensive text, all the more so since the evidence suggests that a single artist was responsible for the surviving pictures and may well have planned to execute the entire corpus of illustrations. If this were the case, a consistently simple format would have suited his purposes admirably. The discussion so far has emphasized the elements that the artist inherited from Mesopotamian painting and was content to reproduce, fundamentally unenriched, in his turn. One can go further, however, and note a positive

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4D63A29B /production/pindar/booksxml/hillen/chapter−07

232 reduction in scope and variety. Received traditions served to cripple the artist’s imagination. The figure types clearly derive from H.arı#rı# illustrations; but here they are jerky, their grouping formalized, their gestures wooden.80 Fewer people are depicted—no picture has more than three, a far cry from the bustling masses of the earlier H.arı#rı# manuscripts—and the background of architecture and furnishings that enlivens these Maqa¯ma¯t paintings is cut down to a few symbolic references.81 The variety of movement that is so marked in Mesopotamian animal painting has been drastically reduced. The animals do not run at full speed, look over their shoulders, or move abreast of one another—in fact only one of them in the whole manuscript is seated.82 None is shown affronted or addorsed. The artist misses the opportunity of depicting males and females together so as to explain their different characteristics,83 thereby showing that he does not have a specialized zoological interest. The exception to this, and the only relatively novel feature, is the frequent depiction of copulation and related sexual themes—an iconographic quirk only partially explained by the text.84 Thirteen of the thirty-two illustrations

80

E.g., folios 19b, 20b, and 40b; see Löfgren, “Ambrosian,” pls. IX, X, and II. E.g., folios 18a and 20b; see Löfgren, “Ambrosian,” pls. VIII and X. 82 This is the cat on f. 9a; see Löfgren, “Ambrosian,” pl. XVb, lower picture. The ostrich on f. 10a should perhaps also be mentioned here; see color plate in Ettinghausen, Arab Painting, p. 157. 83 An opportunity taken by the artists of two comparable Mamlu¯k manuscripts—the Kashf al-Asra¯r in Istanbul (Süleymaniye Library, Lala Ismail 565) and the Escorial Bestiary (Cod. arab. num. 898); see Haldane, Mamluk Painting, pp. 50–54. 84 These number five in all—folios 9a (upper picture), 42a, 44b (upper picture), 44b (lower picture), and 64a; see Löfgren, “Ambrosian,” pls. XVa, IV, Va, Vb, and XXIII. This theme occurs much more rarely in other comparable manuscripts. For example, among the thirty-nine pictures of the Kita¯b al-Bait.ara of 605/1209 produced in Baghdad (Cairo, National Library, Cod. med. VIII), it occurs only once (on f. 157) while among the fifty-six pictures of the same text in the Topkapı Saray Library in Istanbul (Cod. coll. Ahmet III, no. 2115) it does not occur even once. The omission is remarkable given the variety of equine poses that occur in the manuscript; by the same token, its frequent presence in the Ja¯h.iz. manuscript, where no animal is allowed very many illustrations, is all the more exceptional. For a discussion of the versions of the Kita¯b al-Bait.ara, see E. J. Grube, “The Hippiatrica Arabica Illustrata: Three 13th Century Manuscripts and Related Material,” in Pope and Ackerman, eds., Survey, v. 14, pp. 3138–55. 81

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 22

JOBNAME: ???? PAGE: 23 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4DC9A12B /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

233

refer directly or indirectly to sexual matters (figs. 23, 27–29, and 35),85 and in this it seems justifiable to detect a personal idée fixe of the painter; the subject certainly holds no such dominating position in the text of al-Ja¯h.iz.. The painter, however, eagerly grasped at the flimsiest pretext offered by casual anecdote or digression in the text. Perhaps the extreme case is f. 36a, which seems at first sight to be a standard dı#wa¯n scene, with an enthroned dignitary holding a scroll and flanked by attendants (fig. 30). It transpires, however, that the picture illustrates an anecdote concerning a slip of the pen. The Umayyad Caliph Hisha¯m, it seems, wrote to one of his governors ordering, “Let the weaklings who are with you be castrated,” which was simply a scribal error for “Let the weaklings who are with you be counted.”86 The artist actually regarded this footnote as worthy of illustration. Other examples of the progressive atrophy of Mesopotamian conventions may owe something to the shadow theatre.87 Virtually all the creatures are shown in profile, facing left, and in poses that are sometimes repeated quite mechanically. The same reduction makes itself felt in landscape: rocks, trees, mountains, and rivers are on the whole conspicuously absent. Water, the sole exception, is treated with geometrical patterning as if it were costume.88 The selection of animals depicted also points in this direction. In these thirty-two illustrations may be found three lions, three roosters, three donkeys, and no less than nine dogs or closely related animals. It is true that al-Ja¯h.iz. himself expatiates on the dog at quite inordinate length, but none of these nine depictions can refer to the main chapter on the dog because the Ambrosian text does not contain that section,89 while another chapter on the 85 Folios 5a, 9a (upper picture), 18a, 19b, 20b, 36a, 40a, 40b, 41a, 42a, 44b (upper), 44b (lower), and 64a; see Löfgren, “Ambrosian,” pls. XIXa, XVa, VIII, IX, X, I, XXIVa, II, III, IV, Va, Vb, and XXIII. 86 Löfgren, “Ambrosian,” p. 26 and pl. I. 87 See G. Jacob, Geschichte des Schattentheaters im Morgenland und Abendland, Hanover, 1925; P. Kahle, “Islamische Schattenspielfiguren aus ägypten,” Der Islam, v. 1, 1910, pp. 264–99, and v. 2, 1911, pp. 145–95; P. Kahle, “Das arabische Schattentheater im mittelalterlichen Ägypten,” Wissenschaftliche Annalen, v. 3, 1954, pp. 748–76. Further links with the shadow theatre may be the deliberately plain background, like a screen against which the simplified image is projected; the use of stagey pictorial devices like plants or trees flanking the central image; and, in the case of many animals, the selection of a pose that in the simplest and most economical way denotes movement. 88 Folios 6a, 29a, and 51a; see Löfgren, “Ambrosian,” pls. XXa, VII, and XXIVb. 89 Löfgren, “Ambrosian,” pp. 15, 17, 19, 25.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 23

JOBNAME: ???? PAGE: 24 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 40E3DE88 /production/pindar/booksxml/hillen/chapter−07

234 dog comes from the unillustrated section of the manuscript.90 Thus, once again the artist chose to illustrate digressions from the main theme at the expense of the main theme itself. The blame for this impoverishment must be laid at the door of the artist, for the text gave him endless opportunities for experiment and innovation. One can imagine what the artist of the Schefer H.arı#rı#, Yah.ya¯ al-Wa¯sit.#ı, would have made of it. To judge by the animal depictions alone, the Ambrosian text might just as well have been the briefest of bestiaries. All this no doubt sounds very negative and ignores the charm that bright colors,91 fluent draftsmanship, and, intermittently at least, an instinctive insight into the nature of the animal lend to these miniatures. They are fun, and they admirably fulfill their role as illustrations. But they can scarcely claim to be masterpieces, for, as in nearly all Mamlu¯k paintings, the artist has been concerned principally to produce an illustration. The ambition to produce a fine self-contained picture has been secondary—if it has been there at all. These two aims are of course not mutually exclusive. Indeed, it might be fairer to say that much of the finest Iranian painting of the fourteenth century is also primarily book illustration but that the Iranian artist took his duty as an illustrator very much more seriously than his Mamlu¯k counterpart. It was because he thought out his pictures so carefully and built up a wealth of corroborative and expressive detail in them that he created, so to speak incidentally, masterpieces. This series of illustrations begins with some of the paintings in the Pierpont Morgan Bestiary of the 1290s, which may now be examined.

90 This section would have begun just after the present f. 50 (f. 40 in the original text), but unfortunately the next nine folios are missing; see Löfgren, “Ambrosian,” pp. 19, 25. This is entitled ba¯b a¯khir fi l-kalb washana; see Löfgren, “Ambrosian,” pp. 14–15. A third section of the text bears the only superscription in the whole manuscript; this reads, “description of the symptoms of the sagacity of dogs,” yet in what follows there are, amidst two lacunae, pictures of an eagle, a snake, a eunuch, birds, and two elephants—but no dog. 91 The artist often uses the presence of two animals in a composition as a means of variegating his palette; this is especially true in mating scenes.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 24

JOBNAME: ???? PAGE: 25 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 54FE3168 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

235

The I¯lkha¯nid Tradition Because the illustrations in the Pierpont Morgan Bestiary are much better known than the Ambrosian ones,92 there is no need to dwell on them at such length. They were produced at Mara¯gha, in northwest Iran, for an unknown patron who may have been connected with the court. The illustrations accompany a Persian translation of a popular bestiary text of medieval Islam—the Mana¯fi c al-H.ayawa¯n (The Usefulness of Animals), written by the eleventh-century physician Ibn Bukhtı#shu¯ c. In content this bestiary belongs firmly within the tradition of those practical treatises, scientific in their own time, that had long been popular in Mesopotamia and were, via Byzantium, a legacy from the Graeco-Roman past.93 The painting of Adam and Eve, however, strikes a new note (fig. 31).94 It is placed, significantly enough, at the very beginning of the manuscript and is followed by one of Cain killing Abel.95 Other contemporary Iranian manuscripts share this interest in Old and New Testament themes96 and thus, it seems, reflect the persistent curiosity and tolerance that the I¯l-Kha¯ns showed toward other faiths. It may also be connected with the Indian summer of prosperity and power enjoyed briefly by the Jews and the Oriental Christian minorities under the Mongol rulers from Hu¯legu¯ to Gaykha¯tu¯.97

92

For color plates see C. Anet, “The Manafi-i-Heiwan-II,” Burlington Magazine, v. 23, 1913, p. 261 and pl. IV; Ettinghausen, Arab Painting, p. 134; B. Gray, Persian Painting, Geneva, 1961, pp. 20–21; D. Stewart, Early Islam, Weert, 1967, pp. 131–39; E. Kühnel, Persische Miniaturmalerei, Berlin, 1959, pl. 1; and E. J. Grube, The World of Islam, Feltham, 1967, pls. 35–36. For a representative series of black-and-white plates, see Martin, Miniature Painters, pls. 21–26. The material on this manuscript assembled for publication by Rudolf Riefstahl has never appeared in print. 93 For the transmission of such manuscripts to the Islamic world, see K. Weitzmann, “The Greek Sources of Islamic Scientific Illustrations,” in G. C. Miles, ed., Archaeologica Orientalia in Memoriam Ernst Herzfeld, Locust Valley, 1952, pp. 244–66. This article has been reprinted with updated footnotes in H. L. Kessler, ed., Kurt Weitzmann: Studies in Classical and Byzantine Manuscript Illumination, Chicago, 1971, pp. 20–44. 94 Color illustration in Grube, World, pl. 35. 95 Color illustration in Kühnel, Persische, pl. 1. There is no doubt as to the identification of this scene, and its location so close to the previous picture goes far to justify the identification of that scene as Adam and Eve, and thus related to a Genesis cycle, instead of merely “man and woman.” 96 T. W. Arnold, The Old and New Testaments in Muslim Religious Art, London, 1932, pp. 14–16, 21–22, 24, 28, 31–32, 35–36, and pls. III–V, XII–XIII, and XVI. 97 B. Spuler, Die Mongolen in Iran, Berlin, 1955, pp. 198–224, 245–49.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 25

JOBNAME: ???? PAGE: 26 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 51F02B81 /production/pindar/booksxml/hillen/chapter−07

236 This kind of connection with contemporary society would be hard to demonstrate in the case of the Kita¯b al-H.ayawa¯n. Stylistically, the painting of Adam and Eve is recognizably of the Baghdad school. The haloes, the vegetation, the plain background, even the notably corpulent Eve all betray Arab taste. But the careful shading of Adam’s robe is a Chinese feature,98 while the bland, jowly faces are survivals of the Salju¯q tradition in Persia as expressed in painted pottery. The stress on calligraphy, a trait shared with the slightly later Bı#ru¯nı# manuscript,99 is evident in the way that the huge black letters stand out like banner headlines against their white ground. Their importance is such that the calligrapher has plunged deep into the picture space rather than deprive a single letter of its redundant terminal flourish—thereby hampering the artist severely when he came to paint Adam’s halo. But the very fact that the painter did not recompose his picture to avoid having the calligraphy near the halo epitomizes the purely decorative role to which the halo had sunk in the Arab tradition perpetuated here. The artist’s eagerness to exploit calligraphy also helps to explain the unusual variety of scripts employed in the text. Sometimes it is used as a make-weight for, or even as a commentary on, the ostensible subject—as in the case of the battling moufflon, where the calligraphy partakes of the excitement of the action (fig. 32). This bestiary inhabits a different imaginative world from that of its Arab equivalents (fig. 33). The Ambrosian manuscript, for instance,

98 M. S. Dimand, Persian Miniatures, Milan, n.d., pp. 5, 14, and pl. III. The same drapery convention is used on the Cain and Abel miniature in the same manuscript (n. 95 above) and, with notable success, to indicate the loose folds of skin on the elephants; see Ettinghausen, Arab Painting, p. 134. Although this detail is probably a decorative cliché, it is conceivable that the artist was familiar with the texture of elephant skin or copied a model that embodied such knowledge. India is a tempting source, the more so as it is known that an Indian Buddhist, Kamalashri Bakhshi, was employed to help with the Indian section of Rashı#d al-Dı#n’s Ja¯mi c al-Tawa¯rı#kh; see K. Jahn, Rashı#d al-Dı#n s History of India, The Hague, 1965, p. xxxii. In the context of Indian influences in this bestiary, note also that the only other popular text in the Islamic world with numerous animal depictions, the Kalı#la wa Dimna, was based on the Indian Panchatantra, which was illustrated in various media in medieval India; see H. Buchthal, “Indian Fables in Islamic Art,” Journal of the Royal Asiastic Society, 1941, pp. 317, 323–24. For further Indian connections, cf. J. Raby, “Between Sogdia and the Mamluks: A Note on the Earliest Illustrations to Kalı#la Wa Dimna,” Oriental Art, v. 33, no. 4, 1987, pp. 381–98. 99 Barrett, Persian Painting, pl. 1.The manuscript uses elaborate Ku¯fic calligraphy, often in red ink, for chapter headings throughout most of the text.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 26

JOBNAME: ???? PAGE: 27 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4A61BD33 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

237

though written in a very elegant hand, makes no attempt to exploit calligraphy by means of elaborate chapter headings and captions. The clue to the success of the Iranian painter lies partly in his fondness for drama. He invests essentially undramatic subjects with a portentous power wholly at variance with the stiff, woodenly articulated animals of most Arab bestiaries. A vivid sense of tremendous forces unleashed informs the miniature of two elephants entwined (fig. 34).100 The circular lines modeling them create such a formidable vortex that the animals seem quite naturally confined by the frame. It may be a deliberate effect of wit that the foliage alone breaks out. In such a concentrated composition any stress on calligraphy would be superfluous, and the artist has wisely omitted even an explanatory heading for the picture. Interestingly, the text gives no particular warrant for this remarkably original image or for its details:101 An elephant lives three or four hundred years; the animals with the longer tusks have a longer life. The elephant is afraid of a young pig and a horned ram but he is annoyed most of all by the gnat and the mouse. When an elephant is tired, people rub his feet with oil and warm water and he gets well. One dram of his ivory is good for leprosy; a piece of an elephant’s skin tied on the body stops the ague; his fat relieves headaches when it is burned and the patient sits on the fumes.102 Both text and illustration on this subject in the Kita¯b al-H.ayawa¯n bring the reader down to earth with a bump (fig. 35). The text reads, “When [one among] their males are [is] in a strait, they help him in such wise that another 100

Ettinghausen, Arab Painting, p. 134. Such as the anklets worn by both animals, a feature that recurs in the image of the giraffe in the Kita¯b al-H . ayawa¯n, f. 26a; see Löfgren, “Ambrosian,” pl. VI. Could this have been regarded as the appropriate ornament for these especially exotic creatures? The motif occurs elsewhere in Mamlu¯k painting, but only rarely; in the Pierpont Morgan Bestiary, in an image where Indian influence has been postulated (see n. 98 above), the same anklets are worn by Eve. An elephant very similar to the ones depicted here—violet in color and complete with anklets and cap—occurs in a picture datable ca. 1300 in an album in the Topkapı Sarayı in Istanbul (H. 2152, f. 60b); see E. Esin, “A Pair of Miniatures from the Miscellany Collections of Topkapı,” Central Asiastic Journal, v. 21, 1977, pls. Ib, IIb. Esin associates the images of this picture with Buddhist iconography (p. 16). For a color plate, see M. S. Ips¸irogˇlu, Das Bild im Islam, Vienna and Munich, 1971, pl. 24. 102 Stewart, Early Islam, p. 134. 101

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 27

JOBNAME: ???? PAGE: 28 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 458BD1DD /production/pindar/booksxml/hillen/chapter−07

238 male covers him, and this done he will calm down completely.”103 It is curious that the artist should have chosen this particular section of the text for illustration since al-Ja¯h.iz. discusses this animal for several pages. One can only guess at the effect such an image might have had in Western Europe, where a twelfth-century English bestiary, for example (fig. 36),104 would begin the corresponding chapter as follows: “There is an animal called the elephant, which has no desire to copulate,”105 a remarkable statement which has stung one translator of the text, T.H. White, into producing a learned and at times scabrous footnote quoting the views of Solinus, Pliny, and Albertus Magnus on the matter and concluding, “The copulation of elephants was a matter for speculation in the dark ages, and still is, as it is rarely witnessed. … In fact, they copulate in the ordinary way and, according to Lieut.-Colonel C. H. Williams, more gracefully than most.”106 There one may perhaps leave the subject. Drama is again the keynote in the depiction of the Chinese phoenix, identified elsewhere on the page as the Persian sı#murgh and the Arabic canqa (fig. 37). Here the sinuous rippling line of the bird’s head and tail calls for a correspondingly cursive and seemingly relaxed script. The formal, angular Ku¯fic hand used earlier would be inappropriate here. No miniature in the bestiary is more infused with the Chinese spirit. This spirit expresses itself in the wayward curves of the frame, curling like a parchment scroll, in the calligraphic jottings that so readily suggest random clumps of grass, and in the shaded overlapping planes that give the design depth. Even the blue surrounding ocean, with its languid irregular scrolls, is of Chinese inspiration. Yet the artist’s relish of color, especially his daring juxtaposition of tones, presages developments later in the 14th century and could be regarded as a local trait. To some extent this decorative bias has led the artist astray, for it is hard to equate this gaudy fowl with the information given in the text: “The Sı#murgh, found in inaccessible islands and near the equator, is

103

Löfgren, “Ambrosian,” p. 32. The illustration shows a somewhat glum elephant bent double under the weight of a gigantic howdah crammed with military men, and being led most literally by the nose (Harley MS. 4751, f. 8). See A. Payne, Medieval Bestiaries, London, n.d., p. 6; cf. also the bestiary in the Westminster Chapter House Library, MS. 22, f. 20b, mentioned in B. Rowland, Animals with Human Faces, London, 1974, p. 72. 105 White, The Bestiary, p. 24. 106 White, The Bestiary, p. 25. 104

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 28

JOBNAME: ???? PAGE: 29 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 58026D3B /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

239

fearless beyond all other animals.107 He can carry off exceedingly large animals like the elephant and the rhinoceros and when he does, rocks quake and tremble.”108 But it is precisely his untrammeled imagination109 that has made it possible for the artist to depart so far from his text. As in the case of the elephants, the text did not restrict him; it served merely as a point of departure. Hence the wide range of poses in the manuscript: the animals are shown singly or in pairs, face left or right, are shown frontally or from behind, upside down, turning the head, sitting on their haunches or intertwined in scenes of violence or tenderness. Secondary planes are established by the liberal use of overlap. The background varies from picture to picture, displaying a wide variety of landscape elements—water, grass, trees, rocks, bushes, flowering plants, and coulisses (fig. 38). Birds perch among the plants.110 As is natural with such an experimental approach, the picture size varies dramatically,111 and at times the artist seems to go out of his way to choose a difficult subject (fig. 39).112 One example must suffice: ‘‘One of the wonderful traits of the mountain goat is that he leaps down from places that are at a height of about a hundred spears, and stands on his horns.”113 The artist, nothing daunted, does his best to render this marvel—which is, incidentally, sober fact—and if the effect is more curious than beautiful, its originality is not in dispute (fig. 40). Not so that of the text, for the appropriate entry in a typical 107

A far more formidable creature is conjured up by the artist of the cAja¯ ib al-Makhlu¯qa¯t in the Bayerische Staatsbibliothek in Munich, dated 678/1280. This shows a convincingly authentic roc carrying off Sindba¯d; see Ettinghausen, Arab Painting, p. 139. 108 Stewart, Early Islam, p. 131. 109 This very quality irritated a later reader of the manuscript, who scrawled on the margin of this page, “Thou foolish son of a burned father, if nobody has seen the Sı#murgh then how didst thou portray it?” (Yohannan, “A Manuscript,” p. 389, slightly adapted). 110 These birds constitute perhaps the most formulaic element of these paintings, for they are invariably disposed in groups of three and colored in red and yellow or related hues. In the early part of the manuscript, which is the section with illustrations in a style akin to Mesopotamian painting, every painting has at least one group of three such birds. 111 In the later part of the manuscript very small square pictures involving lesser creatures like monkeys, birds, crustaceans, and fish take over from the earlier, more ambitious paintings of large mammals. 112 The treatment of the giraffe, for example, makes an interesting contrast with the same subject in the Kita¯b al-H . ayawa¯n (f. 26a; see Löfgren, “Ambrosian,” pl. VI). In the Morgan Bestiary the artist exploits the size of the giraffe by making its head break through the margin; the artist of the Kita¯b al-H . ayawa¯n keeps it confined within the frame. 113 Stewart, Early Islam, p. 138.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 29

JOBNAME: ???? PAGE: 30 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 5F09F100 /production/pindar/booksxml/hillen/chapter−07

240 twelfth-century English bestiary reads, “There is an animal called IBEX the chamois, which has two horns. And such is the strength of these that, if it is hurled down from a high mountain peak to the very depths, its whole body will be preserved unhurt by these two.”114 In view of the great popularity of the Latin bestiary, and the numerous Western diplomats and merchants who thronged the I¯lkha¯nid capitals, it might not be entirely fanciful to suggest the possibility of some cultural interchange in this particular sphere. But despite many seductive visual115 and textual analogies between Islamic and Western bestiaries,116 two basic differences of approach require emphasis. First, the Latin bestiary drew moral and especially theological lessons from the animals it described;117 and secondly, a great many of its creatures were fabulous.118 In Islamic bestiaries the medicinal value of the animal is stressed,119 and very few fabulous creatures occur. The Kita¯b al-H.ayawa¯n does emphasize the

114

White, The Bestiary, pp. 29–30. This section of the text is not derived from the Physiologus, the Greek source on which the Latin bestiary is ultimately based; see M. R. James, The Bestiary, p. 2. Could this information, then, which entered the mainstream of the Latin bestiary tradition only in the twelfth century (White, The Bestiary, p. 238), derive from an oriental source? But James notes that relevant texts other than the Physiologus were available in twelfth-century Britain: Isidore’s Etymologiae, Solinus’ Liber Memorabilium, the Hexameron of Ambrose, and probably others (The Bestiary, p. 28). 115 E.g., the treatment of lions as a tender family group; compare the scene of the animals nuzzling each other in Gray, Persian Painting, p. 20, with that of lions and their young in a twelfth-century bestiary in the British Library, Royal MS. 12 C XIX, f.6 (see Payne, Medieval Bestiaries, pp. 4–5), which can be explained by the statement in the text that lion cubs are born dead and remain so for three days—‘‘until their father, coming on the third day, breathes in their faces and makes them alive” (White, The Bestiary, p. 8). 116 I am very grateful to my friend John Higgitt for much helpful advice on Western bestiaries in general. For a brief survey, see McCulloch, Medieval Latin, pp. 70–77. 117 E.g., White, The Bestiary, p. 41 and passim; Physiologus, tr. Curley, pp. xxii–xxvi; and B. Rowland, Birds with Human Souls: A Guide to Bird Symbolism, Knoxville, 1978, passim. 118 Rowland, Birds, pp. 22–24 (griffin), p. 33 (bonnacon), pp. 43–44 (monoceros), pp. 51–52 (manticora), and pp. 52–53 (parandrus). 119 “The flesh of a wolf, beaten in a mortar, and cooked in the juice of celery and honey, then mixed with saffron, galangale and white pepper, a little bit of each, and taken with the juice of mouse ear is good for fever. … The brain of a female camel, dried and liquified with vinegar is good for epilepsy resulting from melancholia. The hump of a camel, taken internally, is good for dysentery; and the milk is useful for dropsy and for trouble of the liver and spleen. The shin bone, pounded small and with an admixture of water, exterminates mice when placed in their holes; the melted marrow, taken with date wine four times, will help epilepsy and cure diphtheria; the saliva, in vinegar, is given to an insane man who is as violent as an infuriated camel” (Stewart, Early Islam, pp. 132, 137).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 30

JOBNAME: ???? PAGE: 31 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4FA2B353 /production/pindar/booksxml/hillen/chapter−07

MAMLUK AND ILKHsNID BESTIARIES

241

character traits of certain animals,120 but it does not take the crucial further step of drawing the obvious analogy with mankind for homiletic purposes. In fact the European connections are insignificant compared to the intrusive, even at times dominating, influence of Chinese art in this manuscript. In the tiny pictures of magpies (fig. 41),121 the artist has used landscape indicators as stereotyped as any comparable feature in Mesopotamian painting. In this case they comprise a close-up of a tree whose almost leafless branches grow at random intervals, their forms suggested by a few sure, impressionistic strokes. But not only are these of Chinese, not Iraqi, origin; their use in combination with birds is itself a quintessentially Chinese idea. Similarly, such features as the mare’s head cut off by the frame122 imply a knowledge of Chinese handscrolls and exploit the idea of treating the frame as a window on the world—a convention new to Islamic painting. It is symptomatic that when Mamlu¯k painting of this very period uses the related device of the deliberately broken frame (cf. fig. 42), the resultant clumsiness betrays a profound misunderstanding of the spatial implications of the broken frame. In the Sulwa¯n al-Mut.a¯ c manuscript, the animal that breaks into the frame is simply superimposed upon it with no serious attempt to interlock the areas inside and outside the frame (fig. 43).123 In other paintings in the Morgan Bestiary tree boles, flowering branches, rocks, marshy terrain, and a host of other features also have a Chinese origin. Rather than itemize these features, however, it might be more interesting to ask whether Persian painters really learned much from Chinese sources. The answer is certainly not a clear affirmative. The Chinese conventions for landscape elements were taken over virtually wholesale, but in a somewhat uncomprehending and uncritical fashion.124 They function as a kind of visual shorthand. Moreover, the changes these conventions undergo are 120

Pellat, Life, pp. 146, 175, 182–84. E.g., Martin, Miniature Painters, pl. 26, top left. 122 Gray, Persian Painting, p. 21. 123 Cf. the illustration in Haldane, Mamluk Painting, p. 59, pl. 17. The artist repeats this schema almost exactly in at least one other miniature from this manuscript, a leaf now in the Freer Gallery, Washington; see Atıl, Art, p. 113. 124 It is of course impossible to be certain about the range of Chinese paintings and related work available to I¯lkha¯nid painters. If, for example, they had access only to provincial rather than metropolitan Chinese models, or encountered Chinese painting through the distorting mirrors provided by Central Asian temple frescoes or woodblock prints, it is little wonder that their chinoiserie should seem somewhat coarse. Cf. W. Watson, “Chinese Style in the Paintings of the Istanbul Albums,” Islamic Art, v. 1, 1981, pp. 69–76. 121

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 31

JOBNAME: ???? PAGE: 32 SESS: 71 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 4D984EB6 /production/pindar/booksxml/hillen/chapter−07

242 most significant: features that should be randomly disposed are made symmetrical, muted tones are replaced by vivid color contrasts, blurred outlines are sharpened. These transformations are already evident in the Morgan Bestiary, and they stamp the borrowings as chinoiserie rather than expressions of a sinicizing style. By the end of the fourteenth century, in the illustrations to the mathnavı#s of Khwaju¯ Kirma¯nı#, dated 798/1396 (British Library, Add. 18118),125 Chinese landscape notations have finally been incorporated into the visual vocabulary of Persian artists and lost much of their Far Eastern flavor. In the process, however, Chinese works of art had opened the eyes of Persian painters to the potential of landscape as a means of expression126 and—at least to some extent—as a theme in its own right.127 If Persian landscapes became less Chinese as the fourteenth century wore on, they nevertheless developed a distinctive local idiom in this field that did ultimately owe a good deal to Chinese inspiration. Moreover, at intervals throughout that century and in key manuscripts, Persian artists were able to take up Chinese landscape themes and use them with surprising freshness, integrating them into their pictures with signal success. In the manuscript of Rashı#d al-Dı#n’s World History dated 714/1314, the depiction of the sacred grove in which the Buddha received enlightenment well illustrates this ability.128 From a slightly later version of the same text comes a picture of Mongol cavalry breasting a foaming torrent whose swirls, eddies, and spume-flecked waves are remarkably close to contemporary Chinese models.129 Similarly, in the Demotte Sha¯hna¯ma, the lowering hillside, thickly forested, gashed by ravines and strewn with boulders—from which a dragon spits defiance at Iskandar and his men130—is again a noteworthy attempt to 125 For typical illustrations, see Gray, Persian Painting, pp. 46–47, and B. Spuler and J. Sourdel-Thomine, Die Kunst des Islam, Berlin, 1973, pl. LV. 126 This point can be demonstrated repeatedly in the case of the Demotte Sha¯hna¯ma. 127 This occurs as early as the Rashı#d al-Dı#n manuscript of 714/1314 in the depictions of the mountains of India and the sacred grove of the Buddha; see Gray, World History, pls. 19, 26. 128 Gray, Persian Painting, p. 24. Compare the image “An arhat in the forest” attributed to Fan-lung and dated to the twelfth century (Freer Gallery) illustrated in J. Cahill, Chinese Painting, Geneva, 1960, p. 94. 129 Ips¸irogˇlu, Das Bild, pl. 37. For a Chinese parallel dated 1244, see I. E. Lodge, “Ch’en Jung’s Picture of Nine Dragons,” Bulletin of the Museum of Fine Arts, Boston, v. 15, 1917. 130 O. Grabar and S. Blair, Epic Images and Contemporary History, Chicago, 1980, pl. 34.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07 Sequential 32

JOBNAME: ???? PAGE: 1 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 28B83B9F /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

243

1. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , harem scene, f. 40b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 08585FCD /production/pindar/booksxml/hillen/chapter−07−artwork

244

2. Kita¯b Mana¯fi c al-H.ayawa¯n (the Escorial Bestiary) by Ibn al-Duraihim al-Maus.ilı¯, herons, f. 80a. After Ettinghausen, Arab Painting, p. 3.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06D8FB41 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

245

3. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , ostrich, f. 10a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 060C7FFE /production/pindar/booksxml/hillen/chapter−07−artwork

246

4. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., giraffe, f. 26a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07ED6BBC /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

247

5. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , he-goat, f. 4a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 070433D3 /production/pindar/booksxml/hillen/chapter−07−artwork

248

6. Fa¯t. imid luster bowl depicting an eagle, Metropolitan Museum of Art, acc. no. 63.178.1. After M. Jenkins, Islamic Pottery: A Brief History, New York, 1983, p. 12.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0825328C /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

249

7. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , eagle and snake, snake and chicken, f. 63a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 064FB6F6 /production/pindar/booksxml/hillen/chapter−07−artwork

250

8. Syrian bowl, ca. 1200, depicting a heron, Freer Gallery of Art, acc. no. 47.8. After Atıl, Art, p.76.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06D9CB83 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

251

9. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., rooster, f. 40a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0620C835 /production/pindar/booksxml/hillen/chapter−07−artwork

252

10. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., rooster, f. 8a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07F8C321 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

253

11. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., rooster, vulture, and hoopoe, f. 6b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 05F55C2C /production/pindar/booksxml/hillen/chapter−07−artwork

254

12. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., three fishes, f. 28b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 098D12A9 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

255

13. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , the female slave T.ughya¯n, al-Jamma¯z, and the eunuch Sı¯na¯n, f. 19b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06D697E4 /production/pindar/booksxml/hillen/chapter−07−artwork

256

14. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., Maisu¯n, Mu ca¯wiya, and the eunuch, f. 20b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 08117626 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

257

15. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , eunuch releases birds from cage, f. 41a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 072F1B2A /production/pindar/booksxml/hillen/chapter−07−artwork

258

c 16. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , the sim (a cross of wolf and hyena), f. 22b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07788613 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

259

17. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., hare, f. 5a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06797624 /production/pindar/booksxml/hillen/chapter−07−artwork

260

18. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , lion eating the carcass of a cow, f. 3a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07C057D7 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

261

19. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., three dogs, f. 7a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 05D1670A /production/pindar/booksxml/hillen/chapter−07−artwork

262

20. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , a fowler, f. 63b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07CE106B /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

263

21. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , boar and fish, f. 6a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0645A979 /production/pindar/booksxml/hillen/chapter−07−artwork

264

22. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , crocodile, bird, and fish, f. 51a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 22

JOBNAME: ???? PAGE: 23 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 082B2090 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

265

23. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., horse mating with ass, f. 42a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 23

JOBNAME: ???? PAGE: 24 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0650496B /production/pindar/booksxml/hillen/chapter−07−artwork

266

24. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., Umm Ja cfar and her fish pond, f. 29a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 24

JOBNAME: ???? PAGE: 25 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 084EB451 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

267

25. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., dogs eating the carcass of an ass, f. 1a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 25

JOBNAME: ???? PAGE: 26 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06E54A00 /production/pindar/booksxml/hillen/chapter−07−artwork

268

26. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., the daisam (a cross of wolf and dog), f. 23a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 26

JOBNAME: ???? PAGE: 27 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 085729EF /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

269

27. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., rooster mating with hen; cat and lizard, f. 9a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 27

JOBNAME: ???? PAGE: 28 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 05EF2CC4 /production/pindar/booksxml/hillen/chapter−07−artwork

270

28. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., scene of bestialism, f. 18a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 28

JOBNAME: ???? PAGE: 29 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 09453A13 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

271

29. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., mating scenes between he-goat and ewe, ram and she-goat, f. 44b, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 29

JOBNAME: ???? PAGE: 30 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06C0529B /production/pindar/booksxml/hillen/chapter−07−artwork

272

30. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz., dı¯wa¯n scene, f. 36a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 30

JOBNAME: ???? PAGE: 31 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 083DC380 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

273

31. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, Adam and Eve, f. 4b, Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 31

JOBNAME: ???? PAGE: 32 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 079936E0 /production/pindar/booksxml/hillen/chapter−07−artwork

274

32. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, battling moufflon, f. 37a. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 32

JOBNAME: ???? PAGE: 33 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 08DB970C /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

275

33. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, stag and deer under willow tree, f. 35b. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 33

JOBNAME: ???? PAGE: 34 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0727579D /production/pindar/booksxml/hillen/chapter−07−artwork

276

34. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, elephants, f. 13a. Courtesy of the Pierpont Morgan Library, New York

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 34

JOBNAME: ???? PAGE: 35 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07F67C9D /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

277

35. Kita¯b al-H.ayawa¯n by al-Ja¯h.iz . , copulating elephants, f. 64a, Biblioteca Ambrosiana, Milan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 35

JOBNAME: ???? PAGE: 36 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0451BEE0 /production/pindar/booksxml/hillen/chapter−07−artwork

278

36. Harley MS. 4751, elephant, f.8, British Library, London.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 36

JOBNAME: ???? PAGE: 37 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 096EF0B5 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

279

37. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, canqa, f. 55a. Courtesy of the Pierpont Morgan Library, New York

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 37

JOBNAME: ???? PAGE: 38 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 07FE6350 /production/pindar/booksxml/hillen/chapter−07−artwork

280

38. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, herdsman with animals, f. 44b. Courtesy of the Pierpont Morgan Library, New York

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 38

JOBNAME: ???? PAGE: 39 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 08A49C98 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

281

39. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, giraffe, f. 16a. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 39

JOBNAME: ???? PAGE: 40 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06D6E309 /production/pindar/booksxml/hillen/chapter−07−artwork

282

40. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, ibexes, f. 37b. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 40

JOBNAME: ???? PAGE: 41 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 08792D5A /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

283

41. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, magpies, f. 60b. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 41

JOBNAME: ???? PAGE: 42 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 06CF506E /production/pindar/booksxml/hillen/chapter−07−artwork

284

42. Mana¯fi c al-H.ayawa¯n by Ibn Bukhtı¯shu¯ c, stag and deer, f. 33b. Courtesy of the Pierpont Morgan Library, New York.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 42

JOBNAME: ???? PAGE: 43 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 087CD7B2 /production/pindar/booksxml/hillen/chapter−07−artwork

MAMLU ¯ K AND I¯LKHA ¯ NID BESTIARIES

285

43. Sulwa¯n al-Mut. a c, The Tale of the Horse and the Boar, Freer Gallery of Art, acc. no. 54.IV. After Atıl, Art, p. 113.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 43

JOBNAME: ???? PAGE: 44 SESS: 29 OUTPUT: Tue Aug 10 10:26:48 2010 SUM: 0801E0CD /production/pindar/booksxml/hillen/chapter−07−artwork

286

44. Kalila wa Dimna, the deposed king of the monkeys throws figs to the tortoise, Istanbul University Library, F. 1422, f. 19b. After Gray, Persian Painting, p. 35.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-artwork Sequential 44

JOBNAME: ???? PAGE: 1 SESS: 71 OUTPUT: Tue Aug 10 10:28:55 2010 SUM: 520F509A /production/pindar/booksxml/hillen/chapter−07−ends

MAMLUK AND ILKHANID BESTIARES

287

recreate the mood of a certain type of Chinese landscape setting.131 For all that, it is a backcloth—an excellent complement to the action, to be sure, but no longer a subject in its own right. Man and his deeds have usurped the limelight. In the next generation this process repeats itself, but in a completely different way. The surviving leaves of the great Kalı#la wa Dimna manuscript in Istanbul (University Library, F.1422), tentatively attributed by Grube132 to Ah.mad Mu¯sa¯, consistently create a narrow well of space at the side of the picture and fill it with a closely observed study of a tree or branches (fig. 44).133 Usually these branches bear blossoms that stand out with absolute clarity against the unpainted surface of the page. Executed as they are with surpassing sensitivity and precision, these botanical studies cannot but recall a well-established genre of Chinese painting—plum branches in blossom, clumps of bamboo, and so on. However, by a simple process of transposition, a theme that in Chinese painting was self-sufficient has been integrated into a wider setting, and as a secondary theme at that. These remarks call into question the idea of a linear development of Persian landscape painting in the fourteenth century, with an initially strong Chinese influence gradually becoming weaker and overlaid by native modes. Any overview of the period must also take account of the cyclical nature of Chinese influence, a phenomenon that can be noted again in the following century.134 In a broadly similar way, though admittedly for a much longer period, the artists of medieval Byzantium could draw at will upon a surprisingly untainted classical heritage.

131

In spirit, though not in color, it seems to have much in common with the grand landscapes of Chinese painters in the eleventh century (Cahill, Chinese Painting, pp. 30, 31, 33, 36; J. Cahill, Chinese Paintings: XIth–XIVth Centuries, London, n.d., pp. 31–32, pl. 20; W. Cohn, Chinese Painting, London, 1948, p. 38), in the tenth century (P. C. Swann, Chinese Painting, Paris, 1958, p. 56), and even—most relevantly—in the thirteenth and fourteenth centuries, for example “Bullock Carts in the Mountains” by Chu Jui, in which the abrupt reversals of planes echo the Demotte painting (Cohn, Chinese Painting, p. 143). 132 E.J. Grube, “The Kalilah wa Dimnah of the Istanbul University Library and the Problem of Early Jalairid Painting,” Akten des VII. Internationalen Kongresses für Iranische Kunst und Archäologie. München 7–10. September 1976, ed. W. Kleiss, Berlin, 1979, p. 504. 133 Gray, Persian Painting pp. 38–39. 134 Notably in a series of paintings on silk; see E. J. Grube, “Studien zur Malerei der Timuriden: I. Zur Frühstufe von Herat,” Kunst des Orients, v. 5, no. 1, 1968, pp. 1–23. For a color plate of one such painting (Huma¯y meets Huma¯yu¯n) see Pope and Ackerman, eds., Survey, pl. 878.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 71 OUTPUT: Tue Aug 10 10:28:55 2010 SUM: 50AAE9EA /production/pindar/booksxml/hillen/chapter−07−ends

288 With the growing importance of landscape in fourteenth-century Iran came a reappraisal of animal painting. Clearly, the Mesopotamian outline style, which was well adapted to a plain background, could not coexist with a developed landscape.135 It was obviously desirable to show animals in natural movement against such a background, and, for that matter, to depict their characteristics—their color, their typical stance, and the texture of their fur or skin—as accurately as possible.136 In these respects, as in the exploitation of a pronounced vein of caricature,137 Chinese painting could give useful leads.138 The precise medium and mode of transmission that brought such ideas to Iran is by no means clear. There can be no doubt, however, as to the impact of such borrowings. By the middle of the fourteenth century, or very soon after, and following the achievement of a developed landscape style, the peaceful or violent interaction of animals could at last be presented convincingly.139 The stage was set for the Istanbul Library Kalı#la wa Dimna leaves, usually dated ca. 1360–70. Here all the picture space is used. In fact large portions of the picture repeatedly burst out of the actual or implied frame.140 Moreover, the detail is often carefully chosen to augment and

135

This is very clearly seen when the plates of Mamlu¯k images of the lion attacking the bull are juxtaposed with the illustration of the same subject in the Istanbul Kalı#la wa Dimna; see J. David-Weill, “Sur Quelques Illustrations de Kalila et Dimna,” in O. Aslanapa, ed., Beiträge zur Kunstgeschichte Asiens. In Memoriam Ernst Diez, Istanbul, 1963, pp. 258–63 and pls. II, IV. 136 Gray, Persian Painting, p. 35; Ips¸irogˇlu, Das Bild, p. 50. 137 E.g., the enraged camel in the Istanbul Kalı#la wa Dimna; see Grube, “Kalilah wa Dimnah,” fig. 3. 138 Cahill, Chinese Painting, pp. 71–73, 77, 97, 100; cf. also the picture of an emaciated horse by Kung K’ai (late thirteenth century) in a private collection in Japan, or the animal studies of Chao Mêng-fu (1254–1322) in the Freer Gallery (Cohn, Chinese Painting pls. 134–36). 139 This was an achievement that still eluded the painter of the Kalı#la wa Dimna of 744/1343–44 in the National Library, Cairo; see E. Kühnel, “A Bidpai Manuscript of 1343/4 (744 H.) in Cairo,” Bulletin of the American Institute for Iranian Art and Archaeology, v. 5, no. 2, 1937, figs. 1–2. Even if, as Grube suggests in “Kalilah wa Dimnah,” p. 497, these illustrations were repainted towards the end of the fourteenth century, the basic layout of the original composition—notably the balance between animals and landscape—was presumably respected. 140 Four examples of this may be seen in B. Gray, “Fourteenth-Century Illustrations of the Kalilah and Dimnah,” Ars Islamica, v. 7, 1940, p. 140, fig. 7; Spuler and Sourdel-Thomine, Die Kunst, pl. XLVIII.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 71 OUTPUT: Tue Aug 10 10:28:55 2010 SUM: 3C0027AE /production/pindar/booksxml/hillen/chapter−07−ends

MAMLUK AND ILKHANID BESTIARES

289

intensify the main theme. Landscape has become a vehicle for drama, for the realistic presentation of narrative, and for secondary comment on the principal subject.141 Manuscripts like the Morgan Bestiary provided an indispensable bridge between works like these and the Mesopotamian school. Yet the differences between these two contemporary schools of manuscript painting can scarcely be overemphasized. In Iran, the range of books, subject matter, patrons, and centers for painting was very much wider than in Egypt. Reciprocal influences between the two countries in the fourteenth century have been noted in such fields as metalwork and architectural tilework,142 but seem not to have operated to any significant degree in manuscript painting.143 Perhaps the explanation for this lies in the fact that Mamlu¯k book painting illustrated texts that were not part of the repertoire of Iranian artists. Quite aside from these various differences, however, an essentially inward-looking orientation marks much of Mamlu¯k painting. Egypt, which for political reasons was sundered from contact with the Mongol realm and thus with China, experienced no such transfusion as was vouchsafed to Iranian painting in the late thirteenth century. Its painters seemed unable to look beyond the horizons of their grandfathers. Hence the archaic nature of the Ambrosian manuscript. There were no clues here for further developments. For all its charm and color it is, when judged by the highest standards of the time, an epilogue.

141

Gray, Persian Painting p. 35; Ips¸irogˇlu, Das Bild, pls. 48–49. A. S. Melikian-Chirvani, Le bronze iranien, Paris, 1973, pp. 80–81 and pls. on pp. 70, 72, 74; M. Meinecke, “Die mamlukischen Fayencemosaikdekorationen: eine Werkstätte aus Tabrı#z in Kairo (1330–1350),” Kunst des Orients, v. 11, no. 1–2, 1976–77, pp. 85–144. 143 The exception was once thought to have been a manuscript published by N. Atasoy: “Un manuscrit mamlu¯k illustré du Ša¯hna¯ma,” Revue des Etudes Islamiques, v. 31, 1969, pp. 151–58, pls. I–XIV. It now seems more likely, however, that this manuscript (Topkapı Saray, Hazine 1519) is an early Ottoman work. 142

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-07-ends Sequential 3

JOBNAME: ???? PAGE: 1 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 31DB8E8E /production/pindar/booksxml/hillen/chapter−08

VIII The Qur’an Illuminated

F

IRST things first: these three catalogues propel the study of medieval Qur’anic calligraphy and illumination to new levels of scholarship. And it was high time. For even though calligraphy is the premier Islamic art within its own culture, it is perhaps the field most neglected by those Western art historians dealing with Islam. That is a painful paradox, and the dominant role enjoyed by Westerners in the scholarship of Islamic art has exacerbated it. A perennial problem here is that calligraphy is the least accessible of all these arts to Westerners, and on two counts especially: they lack that instinctive familiarity with the Qur’an which is the birthright of so many Muslims, and they approach the Arabic script from the outside rather than from within. It was not the first alphabet which they learned to write. Most Westerners (and that includes many seasoned scholars) write a childish hand in the Arabic script, and there is very little they can do about it. Like many other ‘simple’ skills, it is best learned when one is too young to ask questions. But the scholar who approaches Islamic calligraphy with the eye and mind alone, and not the hand, misses a vital dimension. This is particularly true when one encounters calligraphy which bends the rules without breaking them, which sets up unexpected rhythms and counter-rhythms, which delights in playfulness, puns, allusions, quotations. In one’s own alphabet and script, the ability to appreciate such fine tuning gradually becomes instinctive among those who possess it. But it is not a skill that can be transferred to another alphabet. One is reminded of the epigrammatic definition of poetry as ‘what gets lost in translation’. Perhaps it is the inescapable realization of this built-in deficiency that has so inhibited the progress of Western scholarship in this field.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 41639A05 /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

291

These remarks are true above all of modern hands. As one penetrates further into the past, however, the yawning gap between the foreigner and the native begins to narrow. That gap was at its narrowest in the period when the rules governing Qur’anic calligraphy took shape. Hence the most valuable contribution of Western scholars to the study of Islamic calligraphy has been in the field of the scripts generally — if inaccurately — lumped together under the heading of ‘Kufic’, for these are the earliest hands of all. Moreover, Kufic in its varied forms was used to the exclusion of all other scripts for Qur’ans produced in the first four centuries or so of the Islamic era — i.e., until about the year 1000 — and even when its monopoly had been breached, it still dominated Qur’anic calligraphy for a further century. This is the period covered by The Abbasid Tradition. Nevertheless, even here those raised in the traditional culture of the Middle East have the edge. Professional calligraphers to this day cultivate an expertise in writing Kufic, and indeed generate new variations on it, as modern mosques from Rabat to Karachi and beyond show; and the success of modern forgers in palming off their creations (which can be masterly) even on specialists alert for fraud is a further pointer in this direction. The foreigner deciphers; the native reads. There is all the difference in the world between the two activities. All this may sound rather bleak. Is there then any role at all for Western scholars in this field? Are they condemned to plod, to break a butterfly upon a wheel? Happily not. For all the wealth of expertise, both academic and practical, available in the Islamic world, it has yet to achieve commensurate published form. Picture books cannot replace analysis. Skilled practitioners of calligraphy abound, but so far they have not articulated their insights satisfactorily — and that, of course, is a very different skill. As for the histories of calligraphy and calligraphers produced over the centuries in the Muslim world,1 they are packed with incidental detail of all kinds but their evidential value for identifying calligraphers or even styles of calligraphy is limited. By and large, it is simply not possible to correlate their accounts of

1

For example, Qadi Ahmad, Calligraphers and Painters. A Treatise by Qadi Ahmad Son of Mir-Munshi (c.A.H.1015/A.D.1606). Freer Gallery of Art Occasional Papers, III/2, trans. V. Minorsky (Washington, DC, 1959; B. Dodge (ed. and trans.), The Fihrist of al-Nadim. A tenth-century survey of Muslim culture, 2 vols. (New York, 1970).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 52EEBEAB /production/pindar/booksxml/hillen/chapter−08

292 manuscripts, calligraphers or styles with the material that survives.2 This confers a baffling sense of unreality on these texts and makes them exquisitely frustrating to read. In this situation, Western scholarship has a major contribution to make. For all the riches of Qur’anic calligraphy preserved in the libraries of Morocco, Tunisia, Egypt, Yemen, Turkey, Iran and India (for these are the only countries with holdings both extensive and significant), the material in question remains for the most part unpublished; indeed, more than half of it is even uncatalogued. By contrast, the material in the West is much more in the public domain. Yet even in the West, only one of the major collections has been published in extenso.3 Instead, scholarship has had to make do with handlists,4 picture books5 and exhibition catalogues,6 happily supplemented by studies devoted to individual Qur’ans,7 larger groups of manuscripts8 or even entire schools.9 But the volumes reviewed here represent the first attempt to catalogue a major collection in detail; and ‘detail’ in this case means basic information about the physical state of each item, a description of it and at least one colour illustration. For the more significant items, the discussion is lengthy and there are several colour plates. Moreover, care has been taken to provide a context for the catalogue proper; hence there are several essays in each volume which raise issues of general import. Every single one of these repays close study, for all of them (even the shortest) contain new material, while the more substantial ones — such as those on the 2

As shown by the controversy about the badi’ script: see E. Schroeder, ‘What was the Badi Script?’, Ars Islamica IV (1937), pp. 232–48; M. Minovi, ‘The so-called Badi‘ Script’, Bulletin of the American Institute for Iranian Art and Archaeology V (1937), pp. 143–6. 3 F. Déroche, Les manuscrits du Coran. Aux origines de la calligraphie coranique and Les manuscrits du Coran. Du Maghreb à l’Insulinde. Bibliothèque nationale. Catalogue des manuscrits arabes. Deuxième partie: Manuscrits musulmans, I/I (Paris, 1983) and I/2 (Paris, 1985). 4 For example, that drawn up for the Keir Collection: see B.W. Robinson (ed.), The Keir Collection. Islamic Painting and the Arts of the Book (London 1976), pp. 277–83. 5 A. Khatibi and M. Sijelmassi, The Splendour of Islamic Calligraphy, trans. J. Hughes (New York, 1977). 6 M. Lings and Y.H. Safadi, The Qur’an (London, 1976). 7 J.M. Bloom, ‘The Blue Koran. An Early Fatimid Kufic Manuscript from the Maghrib’, in F. Déroche (ed.), Les Manuscrits du Moyen-Orient, Varia Turcica, VIII (Istanbul and Paris, 1989), pp. 95–9; R. Ettinghausen, ‘A signed and Dated Seljuq Qur’an’, Bulletin of the American Institute for Persian Art and Archaeology IV (1935), pp. 95–102. 8 E. Whelan, ‘Writing the Word of God: Some Early Qur’an Manuscripts and Their Milieux’, Ars Orientalis XX (1991), pp. 113–47. 9 D. James, Qur’ans of the Mamluks (London, 1988). c

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4348B891 /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

293

early ‘Abbasid scripts, the New Style, the ‘Baisunghur’ Qur’an and Yaqut al-Musta‘simi — are major pieces of research. And they leaven the loaf. The range of these topics is sufficient indication of the sheer variety of the material published here. There is, in fact, no reason to doubt that the claim made in these volumes — that the Khalili collection of Qur’ans is the most comprehensive in private hands — is accurate. And for that reason alone, these volumes will be very warmly welcomed. Visually speaking, as books, they put all other publications of medieval Qur’ans in the shade. The design, printing, colour plates, binding, jacket and box are all absolutely excellent. Is this circumstance apt to mislead? Is there a danger of being beguiled by the gorgeous colour plates into an exaggerated estimate of the importance of the Khalili holdings in the field of medieval Qur’ans? A close statistical comparison with the only public collection that represents one man’s accumulation in this field, namely the Chester Beatty Library in Dublin — long a touchstone of quality for students of medieval Qur’ans — proves that this is not so. The Khalili material consistently matches or outranks the Dublin collection in most categories except Mamluk Qur’ans, where it has eight complete Qur’ans and eleven fragments against the nineteen complete Qur’ans and fifty-one fragments in Dublin. Dr Khalili came late to the field, but clearly not too late. In the method of publication, there is simply no comparison. Chester Beatty’s collection of Qur’ans was celebrated in book form by a single octavo volume with a text of 83 pages in which 104 complete Qur’ans and 141 fragments were catalogued in entries of usually half a dozen lines; and this was accompanied by 71 plates (11 of them in colour), many of them dingy or on too small a scale.10 Similarly, the British Library collection, one of the two or three finest in the world, has had only a single major book devoted to it. This, for all its visual splendour, presents no more than a sample of the Library’s holdings in this area.11 Great collections, then, are travestied or at best undersold by their published catalogues — that is, if they have a catalogue at all. Conversely, a lesser collection can have a proportionately greater impact by virtue of comprehensive and sumptuous publication. In art history as in other fields, then, the medium is the message. The three

10

A.J. Arberry, The Koran Illuminated. A Handlist of the Korans in the Chester Beatty Library (Dublin, 1967) — a book to which I owe the title of this review. 11 M. Lings, The Qur’anic Art of Calligraphy and Illumination (London, 1976).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 44DF4DAE /production/pindar/booksxml/hillen/chapter−08

294 books reviewed here, more than any yet published, ensure that medieval Qur’ans can take their rightful place in the panorama of Islamic art. And the scope and depth of this collection mean that ample raw material is available for detailed specialist studies. In Volume I pride of place goes to the two quires from Palermo dated 372/982–3 (cat. 81), which come from the only Qur’an positively known to have been produced in Muslim Sicily; but special attention should also be drawn to a part (juz’) of a North African or Egyptian Qur’an datable before 900 (cat. 24), a set of five folios from a Qur’an written in gold and preserved in the Nuruosmaniye Library in Istanbul (cat. 41), three folios employing dramatically elongated letters (cat. 66), a juz’ of an early ‘New Style’ Qur’an, richly illuminated (cat. 78), a folio of a Buyid Qur’an from Isfahan dated 383/993 (cat. 83; a document of cardinal importance), 122 folios of a seven-part Qur’an of the eleventh century (cat. 84) and two ‘New Style’ fragments (cats. 93 and 95, the latter conceivably from India). Among the ‘cursive’ manuscripts that are obvious candidates for further study one might single out II, cats. 1, 6, 17, 18 (both Indian?), 29 and 36 and III, cat. 1 — written, incidentally, not by Karamshah-i Tabrizi but by Abu Muhammad ‘Abd al-Qayyum ibn Muhammad ibn Karamshah-i Tabrizi.12 François Déroche, whose text has been translated with great fluency and finesse, gives an admirably lucid description of the history of the early scripts, and his codicological expertise is employed to good effect in his account of how the books were written and actually put together (I, pp. 11–24). This section repays repeated reading; it distils a great deal of information drawn from very disparate sources. Subsequent general introductions cover the major varieties of script. Here, as mentioned earlier, the historian of early Qur’anic calligraphy is in a double bind. Arabic literary sources gloss over or ignore altogether the individual peculiarities of the numerous scripts they list, thereby preventing significant correlations with the very varied scripts that actually survive. Faced with this insoluble problem, Dr Déroche prudently adopts codes — ‘Hijazi I-II’, ‘Groups A to F’ — and supplements his verbal descriptions of each of the latter groups with tables of the scripts they employ. For the later period he adopts a similar method. He elegantly explains the rapid popularity of what he terms the ‘New Style’ (i.e., Eastern Persian or Qarmatian Kufic) by showing how it was 12 A. Soudavar, Art of the Persian Courts. Selections from the Art and History Trust Collection (New York, 1992), p. 50.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 45B6D2CA /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

295

rooted in everyday cursive scripts13 and was therefore inherently legible, not requiring the marginal or interlinear cribs found in some ‘Abbasid Qur’ans. He argues that it managed to accommodate enough “Abbasid Kufic’ features to sanctify itself for Qur’anic use. He enumerates too, with admirable clarity, the sea-changes that affected Qur’anic calligraphy in the course of the tenth century; the change from parchment to paper, from the horizontal to the vertical format, the adoption of the quire composition and the production both of very large and very small Qur’ans. But the distinctions between New Styles I and III (there is no mention of II) are couched in rather general terms, and do not sufficiently reflect the impact of ‘personality’ exerted by a full page of script. The basic tool used by Dr Déroche to classify his material is a table of letter forms, which channels our concentration towards the diagnostic variations in the shapes of the individual letters of the alphabet. The idea is not new. It was pioneered some eighty years ago by Samuel Flury, who refined it steadily for a quarter-century.14 He is mentioned in passing in a footnote and does not appear in the bibliography; but more to the point, the experience that he built up so laboriously is not fully exploited here. Dr Déroche has chosen to limit his tables to six letters of the Arabic alphabet, where Flury typically used eighteen; and he gives only one form per letter, where Flury gave up to fifteen. Dr Déroche’s reduced tables inevitably simplify complexities, toning down or even ironing out diagnostic flourishes and variations; above all, they suggest (quite fallaciously) that a given letter of the alphabet was always written in the same way. Thus the value of these tables as a guide to changes in orthography is greatly reduced; indeed, in a sense they have been abused as a kind of Procrustean bed. And fuller tables of the Flury kind might throw into question the definition of entire categories of material (e.g. D.VC). As things stand, too many of the key diagnostic features required to pigeonhole a given Qur’an are not placed before the reader, though one is confident that Dr Déroche has them at his fingertips. It is precisely such ‘buried’ information that would be made accessible by means of fuller tables. 13 S.M. Stern, ‘A Manuscript from the Library of the Ghaznawid Amir ‘Abd al-Rashid’, in R. Pinder-Wilson (ed.), Oriental Studies IV. Paintings from Islamic Lands (Oxford, 1969), p. 18, n. 44. 14 For bibliographical details of a representative selection of his articles, see K.A.C. Creswell, A Bibliography of the Architecture, Arts and Crafts of Islam to 1st Jan. 1960 (Cairo, 1961), cols. 61, 301 and 633.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 43479612 /production/pindar/booksxml/hillen/chapter−08

296 Two issues remain on which one would dearly like an opinion from Dr Déroche. They concern the central questions of date and provenance. Too little is said about either; and yet if anyone is in a position to comment on these matters it must be Dr Déroche. It is he who has worked on the Qur’an of Amajur, one of the very rare Qur’ans of this period to bear a date (an endowment notice in this case) and which has a Damascus connection. Dating clues are consistently left unexploited. Thus the floriated Kufic ‘unwan of I, cat. 68 is not exploited; Dr Déroche rightly says that it was inspired by epigraphy, but ventures neither a date nor a provenance, nor does he suggest parallels. Obvious comparators are provided by the Iranian mausolea of Rasgat and Safid Buland, both datable to the late eleventh century.15 But the sad fact is that detailed arguments laying out criteria for dating and provenance are never marshalled, and that one therefore closes the book with no clear idea of where most of these Qur’ans were produced, what local schools there were (for most major categories of material are not associated even tentatively with a given locality), or what broad chronological development they followed. As Qur’ans written in New Style scripts are known, for example, from Sicily to eastern Iran, in the timescale 905 to 1223, it is clearly a priority to set up a basic framework for the chronology and provenance of Qur’ans written in this script, founded on an intensive scrutiny of dated or localized specimens. In such an enterprise the role of illumination, which is somewhat underplayed in Volume I, is crucial. More photographs of the sparse illumination found on these early Qur’ans would have been very useful; a model of such presentation is Storm Rice’s treatment of the illumination of the Qur’an of Ibn al-Bawwab,16 and Graf von Bothmer has achieved substantial results from prolonged study of the illuminated fragments from San‘a’.17 Dr Déroche provides useful codes for verse markers between one, five and ten verses, but here too — as in the alphabetical tables — the codes are too generalized to cover the numerous variations, even though it is precisely in these variations that one might

15 S.S. Blair, The monumental inscriptions from early Islamic Iran and Transoxiana (Leiden, 1992), figs. 79 (Safid Buland) and 155–6 (Rasgat). 16 D.S. Rice, The Unique Ibn al-Bawwab Manuscript in the Chester Beatty Library (Dublin, 1955). 17 H.-C. Graf von Bothmer, ‘Frühislamische Koran-Illuminationen’, Kunst und Antiquitäten I (1986), pp. 22–33.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4AEB34AB /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

297

expect to find the key clues.18 And there is no overall treatment of the palmette (shujaira), the principal focus of illumination in this early period, or of the ‘unwan. The savage trimming which has destroyed so much of the illumination of these early Qur’ans, and yet which has carefully preserved the text itself, is clear proof that illumination was of secondary importance. It was very rare for remargining to be undertaken so as to rescue it. But that does not affect its value as a dating control. Once the evidence of script, illumination and codicology for the crucial dated or localized manuscripts is available in extenso, all other material can be measured against that yardstick. One may note in passing that eleventh-twelfth century dates, technically outside the purview of the book’s subtitle, are proposed for cats. 84–98. So different are the approaches of the two authors under review here that they have to be considered separately. Dr Déroche is a remarkable amalgam of the connoisseur and the archaeologist. For him, ‘le dieu est dans les détails.’ It is typical of his approach that he painstakingly cites the sura and verse for every colour plate in Volume I, a practice not consistently followed by David James. Dr Déroche disdains pat answers to complex problems. His slow, step-by-step approach instinctively inspires confidence. Generalizations are, as it were, wrung out of him. But as he knows this early ‘Kufic’ material better than anyone else alive, it is sad that he does not bring his formidable expertise to bear on the many questions of wider import raised by these Qur’ans. The date of the Hijazi script, for example, warrants closer discussion, as shown by Marcus Fraser’s recent attribution of an exceptionally large leaf, written in an early hand, to the period of the Rashidun caliphs.19 And any discussion of these early scripts must make more use than Dr Déroche does of the chronological benchmark established by the Dome of the Rock inscriptions, to say nothing of the supplementary evidence provided by coins, milestones and graffiti. The difference of material cannot disguise the numerous points of contact. Indeed, the absence of corroborative evidence drawn from other media is striking — for example, the discovery of a 3 : 2 proportional ratio of width : length in the text block of ‘Abbasid Qur’ans might usefully have been correlated with the employment 18 Thus Jonathan Bloom was able to establish a Maghribi origin for the Blue Qur’an on the basis of the abjad system employed in the system of verse counts (J.M. Bloom, ‘Al-Ma’mun’s Blue Koran?’, in L. Kalus (ed.), Mélanges D. Sourdel (= Revue des études islamiques, LIV [1986]), p. 63. 19 M. Fraser, ‘An important early Qur’an leaf in Hijazi script’, in Sotheby’s. Islamic and Indian Art. Oriental Manuscripts and Miniatures London, 22–23 October 1992, pp. 254–9.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4622382F /production/pindar/booksxml/hillen/chapter−08

298 of the same proportional ratio in ‘Abbasid architecture (for example, the mosques of Qairawan and Samarra). Similarly, Samanid epigraphic pottery, the parallel par excellence for New Style Qur’ans, is never used for specific comparisons. Nor are tombstones. And Dr Déroche’s extensive knowledge of other great Qur’an collections could have been used more often to contextualize the Khalili material by citing folios from the same manuscripts, or closely related items, in other collections (for example, leaves of the Blue Qur’an in San Francisco and Bahrain; indeed, no reference is made to the spectacular holdings of the Bait al-Qur’an there, which include leaves from the Nurse’s Qur’an of 1022, a leaf dyed red [cf. I, cat. 11], and a diminutive complete Qur’an from Cordova, probably the earliest known and very richly illuminated). Questions of the functions of the various types of Qur’an, including the rare single-volume copies (I, cats. 82 and 96), the systems of vocalization employed, the aids to recitation, the numbering systems and so on all remain to be tackled. But these lacunae should not simply be laid at the door of the authors; they are merely an index to the underdeveloped state of the field. Dr Déroche frequently notes the practice of elongating letters (mashq). An obvious task for future research is to determine its canons and the permissible variations. Was it normally confined to certain letters? I, cats. 65–66 show that it could be applied to every line, and even, on occasion, as in I, cat. 70, to almost every ligature. But here, as in the mannerist staccato emphasis of I, cat. 93, it is the exception that proves the rule. The effects of excessive mashq are counter-productive, as if every sentence on a page were underlined. Were certain ground rules for mashq generally respected? What of its use to set up diagonal or zigzag rhythms down a page (I, cat. 50)? Or to secure a consistent double-decker effect (I, cat. 44, f. 4b shows a good half-dozen examples)20 — which shows that new strategies for securing links between lines as well as along them were being tried out? Was its function aesthetic, as the multiple parallel bars of I, cat. 66 suggest, or was it related to content, like a forte marking or other types of musical notation which might bear on techniques of recitation? Could it have been intended, like the unnatural division of words found so often in early Qur’ans, to slow down

20

Perhaps the outstanding example of this device can be seen in a curious set of leaves in the Metropolitan Museum of Art, New York, attributed to eleventh-century Iran (40.164.2a and others); see The Metropolitan Museum of Art. The Islamic World (n.a.) (New York, 1987), colour plate on p. 31.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4BB2AE10 /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

299

the pace of reading and thereby increase a sense of awe for holy writ? Dr Déroche’s text and its accompanying plates furnish much raw material for such investigations. Much the same can be said for the conventions which dictated how the ending of one sura dovetailed with the beginning of the next. Sometimes the scribe’s desire to begin a line with the illuminated heading (‘unwan) of the sura constrained him, against all textual sense, to follow that heading with the tail end of the previous sura. Or he might indeed begin the new heading halfway along the line. In one case (I, cat. 84) the last word of the opening sura, the Fatiha, is carried over in stark black letters on to the next page, thereby displacing (in sense, in space, in colour, in script) the outburst of gold illumination that heralds Sura 2. To a Western eye, this irregularity wrecks the next page, and with it the entire double-page spread which introduces this Qur’an. But of course the work was not prepared for the delectation of a Western eye. Such piercing glimpses into the nature of an Islamic aesthetic could be multiplied. On occasion the ‘unwan omits the name of the sura altogether (I, cat. 35). Yet another device, which is in fact quite common, is to drop the last words of the previous sura into the dead centre of the ‘unwan for the next, thereby bisecting it. This too runs counter to the sense of the text, though as a design motif it is an unqualified success. The two volumes by David James (II and III) may conveniently be considered together, as they have a shared approach to the material, and complement each other chronologically — though in II, as in I, the material extends well beyond the cut-off dates of the contents page. ‘Kufic’ Qur’ans are excluded; all the material is in one or other of the so-called ‘cursive’ hands, principally naskh, thulth or muhaqqaq — though Dr James very properly draws attention to the distinctions made by the Arab calligraphers themselves between ‘dry’ scripts (muhaqqaq, naskh and raihan) and ‘dampened’ ones (riq‘a, tauqi‘ and thulth). In an absolutely model exposition Dr James constructs a context for these Qur’ans: the way the manuscripts were put together, the types of calligraphy, illumination and the respective roles of vegetal and geometric elements, the bindings, and (not least) the history of Western scholarship in this field. For good measure, the characteristic development of Qur’an production over a period of six centuries is confidently integrated into the historical background. As with the introductory material in Volume I, all this is required reading for those wishing to orientate themselves in this difficult field. Dr James’s decades of experience in studying Qur’ans, and his profound knowledge of comparative

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 47450C70 /production/pindar/booksxml/hillen/chapter−08

300 material in other, mainly public, collections, including especially unpublished material, reveals precious nuggets of unexpected information at every turn. It is good, too, after the mass of Egyptian material presented in his magisterial Qur’ans of the Mamluks, to have his assessment of the major collateral traditions to the east and west of Egypt. And even where current knowledge is at a rather primitive stage, and where the arguments about provenance, authorship and date tend to go round in circles for want of hard documentation, Dr James clarifies the situation. Thus, he provides lists of work attributed to Yaqut al-Musta‘simi or of folios purporting to belong to the gigantic ‘Baisunghur’ Qur’an, here identified on the basis of scientific paper analysis.21 And if he does not wield Occam’s razor himself, at least he indicates where the cuts are likely to be needed. The identification of Anatolian, Indian and Yemeni Qur’ans illustrate cases where hard data are in very short supply (though a persistent misplacing of diacritical marks is apparent in some Qur’ans identified here as Anatolian), and where Dr James’s tentative efforts to pin down a local manner are full of pointers to further research. One might add that it is too often forgotten that Yemen and India are near neighbours along the sea lanes, a connection which can be demonstrated in fields as diverse as architectural ornament, woodwork, tombstones — and Qur’ans.22 In several cases Dr James’s researches have revealed hitherto unsuspected or undervalued centres of Qur’anic illumination, and indeed book production generally, such as Damascus c. 1330 (II, cat. 43) or Shiraz between 1330 and 1375 (II, cats. 29–32). Here, illumination of electric energy, so brightly coloured as to seem enamelled, gives the lie to the glib assumption that a major court is required to sustain such work. As a result of his labours it is now possible to form a much clearer idea of the wider significance for the interested non-specialist of ‘cursive’ Qur’ans within the art of their times. They can be seen to have played a key role in the secular development of the frontispiece (many thirty-part Qur’ans had a double-page frontispiece for each volume, and many had finispieces too), of the ex libris and of the fashion for margin-breaking which gathered 21

See III, 18–23. Yet it is disturbing to note that the leaf sold at Sotheby’s in 1988 (lot no. 168), the authenticity of which is a pivot of Dr James’s argument for distinguishing genuine fragments from copies, is identified by Abolala Soudavar, on the basis of cogent reasoning, as a copy (op. cit., p. 81, n. 16). 22 In the case of architectural ornament, this connection has persisted right up to the present day. See G.R.D. King, The Historical Mosques of Saudi Arabia (Harlow, 1986), pp. 68–73.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 44B2DF2B /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

301

momentum throughout the fourteenth century.23 Properly used, they can provide a yardstick for the slow evolution of canons of ornament, including its localization and chronology. And the prestige of the Qur’an ensured that the ornament lavished on it was the best that the age could produce. The care which could be devoted to these Qur’ans beggars the imagination. For example, II, cat. 5, with three huge lines of muhaqqaq script marching majestically across a page of some 40 × 32 cm, comes from an Iranian Qur’an which must originally have comprised about 2,760 folios of polished handmade paper. At the other extreme is II, cat. 3, an Ayyubid leaf of c. 1200 with nineteen lines to the page and so crammed that the words and letters keep tumbling over each other whilst remaining entirely legible. It is a tour de force whose cumulative power, built up steadily line by line, terminates in a notable display of wit with the letter kaf written in extreme mashq so as to take up half the line — the scribe reminding us of what he could have done had he so wished. Volumes II and III are punctuated by mini-chapters where Dr James selects a specific issue or problem for extended discussion. It is an excellent idea to spotlight such self-contained topics in this way; it breaks up the body of the catalogue and reminds the reader of the wider implications of the material presented here. Thus the close analysis of Ruzbihan Muhammad of Shiraz, who in the early sixteenth century ran a multi-generational family business specializing in all aspects of book production, brings into focus some of the trademarks of Shirazi illumination. A good example is the habit of devoting a double-page spread to Sura 1 written in two medallions, followed by another such opening in which the first verses of Sura 2 are similarly presented. In earlier traditions, Sura 1 and the early verses of Sura 2 faced each other across a single opening. Equally valuable and pioneering is the extended section on Valencian Qur’ans, which removes the confusion that has surrounded the identity of the major scribe recorded there in late twelfth-century Spain. Here, too, a dynasty of book specialists, in this case scribes, seems to have flourished. It would have been useful to have had even a brief analysis of the diagnostic differences between the so-called ‘Andalusi’ and ‘Maghribi Kufic’ scripts, as the ambiguities surrounding them have bedevilled the establishment of a secure provenance for Western Islamic material. Nevertheless, new 23 See B. Brend, ‘Beyond the pale: meaning in the margin’, in R. Hillenbrand (ed.), Studies in Persian Painting: Festschrift for Basil Robinson (in press).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 35F57BC5 /production/pindar/booksxml/hillen/chapter−08

302 perspectives are opened here. Recent discoveries in the libraries of Rabat and Paris promise to shed further light on these problems, as does the crucially important Western Islamic material in the Bait al-Qur’an in Bahrain. More generally, an interpretive guide to the thicket of orthographic markings found in so many Qur’ans (for example, II, cat. 40) would have been useful. So would a table of specimens of the major scripts side by side, and some discussion of how Qur’anic verses were used as mottoes or otherwise out of context in quality Qur’ans. A special feature of Volumes I and III is the appendix on documentation, which transcribes and translates colophons, commissioning notices, endowment records and the like. These are crucial documents and the absence of accompanying photographs for the great majority of them is a great pity. As all the plates in these books are in colour, it would presumably have been the policy to use colour for this material too, although mostly it would not be needed. In the case of the colophon mentioning ‘Ali Sayrafi, a drawing is given because the gold in which the text is written does not yield sufficient contrast to allow successful photography. One may hope nonetheless that, wherever possible, subsequent volumes in this series will present photographs of such key information and thereby open the way for corrections and supplementary comment. To take a single example, the damaged certificate of commissioning for II, cat. 7 — a Zangid manuscript datable between 1198 and 1219, whose seminal importance depends precisely on that certificate — is given only in a line drawing, which is also transcribed and translated. A good photograph would have made all the difference; but, to judge by the description on II, p. 44, the damage which this page has suffered made this impossible. As it is, the drawing shows that the Arabic transcription should have a lacuna after the word mujahidin in line 3, and that the fifth word in line 4 should read Shahriyar rather than Shahriyad. Comparison with other northern Iraqi protocols of the early thirteenth century suggests that the gaps in lines 5 to 7 could be filled as follows:

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4792C878 /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

303

… sharaf al-muluk wa’l-salatin qahir al-khawarij qatil al-kufra wa’l-mulhidin nasir al-mujahidin wa’l-murabitin hami thughur al-muslimin falak al-ma‘ali iftikhar al-‘iraq …24 Fakhr al-‘iraq would be another possibility for the last phrase; it has the same sense and would take up a little less room. But the contemporary fashion for rhyming pairs of titles suggests iftikhar al-‘iraq as suitably antiphonal for the next phrase, shahriyar al-sham (‘Emperor of Syria’). While the fullest possible visual documentation of such records is desirable, it is nevertheless a great bonus to have this kind of information collected in one place. Similarly, the importance of illumination as a control for dating and provenance, independent of its intrinsic beauty, makes one regret the absence of sets of drawings — of palmettes, frontispiece layouts, hasps, roundels with verse counts, rosettes, ‘unwans — reference to which would obviate the need for lengthy and ultimately unsatisfactory verbal descriptions. They would also give the reader much concrete information about all the material that is not illustrated. The colour plates here, after all, represent only the very tip of the iceberg. Here again one may hope that drawings will play a much larger part in subsequent volumes. Reading through these volumes, one is left with an abiding sense of wonder at the capacity of Muslim scribes and illuminators to breathe new life into this, the most traditional and hallowed of art forms. The self-same artist, the one-armed, left-handed ‘Umar-i Aqta’, could pen a Qur’an small enough to fit under a signet ring, and when that found no favour with his royal patron, the monarch was prevailed upon to allow him a virtual monopoly of the state paper-making enterprise so as to produce the most colossal Qur’an on record — no less than 340 sheets each measuring some 7 × 5 ft, so that the open book was over fourteen feet long. The result weighed half a ton and had to be brought to the ruler on a cart.25 A single line of this Qur’an is in the Khalili collection (III, cat. 2). That particular swing of the pendulum is unique. But less dramatic contrasts abound, and all were aimed at introducing 24

‘‘Honour of the kings and sultans, Conqueror of the dissidents, Slayer of the unbelievers and the heretics, Helper of the fighters for the faith and of those from the frontier fortresses, Guardian of the frontiers of the Muslims, Star of the heights, Pride of Iraq.’’ 25 Following here the reconstruction by Abolala Soudavar (op. cit., pp. 59–62 and 81) rather than that of Dr James, who postulates that there were 1800 single sheets, which in turn implies (by Mr Soudavar’s reckoning) a weight of almost three tons!

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 462F9082 /production/pindar/booksxml/hillen/chapter−08

304 a variety of visual accents into the design of Qur’ans and thereby continually re-invigorating the tradition. The book might be in a ‘cursive’ script throughout but then end with a page written in Kufic (II, cat. 18) or bear a double-page colophon in gold (II, cat. 20); or the opening and closing verses might be enclosed within clouds, as if haloed (II, cat. 11; it would be interesting to know when this motif — known as abri — was first used in Qur’ans, and what it signified in that context).26 The first and last lines of a page are sometimes writ large, or in different colours or script from the rest, and sometimes the middle line introduces yet another colour or script (II, cat. 40). Sometimes individual verses (for example, Sura 2:256, the ‘Throne Verse’) are given in gold (II, cat. 25). In sixteenth-century Iran the text was divided into blocks, each containing a script different from the last, and often executed in different colours and sizes too. Basically similar, even identical, pages of illumination were played off against each other by switching round the colours (III, cat. 81). Apparently inexhaustible variations were played on this theme. Different colours of paper were used side by side for decorative effect (III, cat. 24). Pallisades of tree-like forms hedged round the illumination, which in turn protectively enclosed the text itself. Huge empty margins played on the notion of the text as a relic. In Safavid Qur’ans of the sixteenth century the illumination was, so to speak, drained from the body of the book and loaded onto the opening pages, Sura 1 being written within two medallions on a double-page spread, like an ex libris; the central opening (Sura 18); and the closing pages, now supplemented by prayers and material on divination (the falnama). The consequent imbalance within the book as a whole is palpable. The introductory texts in Volume III make it clear that quite distinct traditions of producing fine Qur’ans developed among the three Islamic superpowers — the Ottoman empire, Safavid Iran and Mughal India — from the sixteenth century onwards. Thus thirty-part Qur’ans continued to be popular in Turkey while Iranian patrons seem to have preferred single-volume Qur’ans. In Iran, just as the decorated pages were saturated with illumination to an unprecedented degree, so by the late sixteenth century the text proper was kept far simpler visually than hitherto. III, cat. 48 shows how difficult it had become to control the overpowering richness of the illumination, perhaps because it used too many colours: gold, yellow, dark 26 Abri is not to be confused with the cognate Turkish term ebru, which means ‘marbled paper’ (see M.U. Derman, Türk Sanatinda Ebru [Istanbul, 1977]).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 437036D6 /production/pindar/booksxml/hillen/chapter−08

THE QUR’AN ILLUMINATED

305

blue, light blue, white, vermilion and green. The new types of illumination that became fashionable in the fifteenth century are carefully noted and analysed in Volume III, but the principles that governed Ottoman, Safavid and Mughal illumination respectively in the following century are not explored. In sixteenth-century Iran, the desire to miniaturize every element found expression in subdividing the page into a collage of panels of variegated size, a device which made ornament easier to control but also created new problems of balance. In II, cat. 33, for example, folio 16 has ten major panels of which seven are further subdivided into independent fields of ornament.27 In short, as III, cat. 36 shows, the entire Qur’an became a kind of album (muraqqa‘) of mingled calligraphy and ornament, often with carefully calculated reciprocities which linked facing pages. Whether this album treatment originated with Qur’ans or in secular books remains an open question. At all events, the virtuosity of the artist was liable to take over from the text itself, whose rhythm and thus perhaps even sense might be distorted by the constant changes of colour, ornament, size, pen and script. But the flawless design of the best pages compels admiration; the eye is by turns rested and challenged by changes in scale and layout within which the other variables interact. Ottoman practice, by contrast, confined the illumination to the opening and closing pages of the text itself. Indeed, Ottoman Qur’ans in the hand of Ahmed Qarahisari, for example, are notably more sober and austere than their Iranian counterparts. For example, II, cat. 56 displays the same fastidious appreciation of empty space as characterizes contemporary Ottoman architecture. This review has tried to highlight some of the areas of medieval Qur’ans in which further research is needed. But it has to be stressed once again that those areas would not have emerged so distinctly if Dr Déroche and Dr James had not done their work so well. Theirs has been a pioneering labour triumphantly brought to fruition. And they have been extremely well served by the editor and the production team. Misprints and errors are minimal (see I, 72–3; I, 84–5; II, 82; II, 200; II, 208; III, 14; and III, 96). As for the books themselves, no pains have been spared to make them worthy of the manuscripts they celebrate. Much thought has been devoted to the sequence of illustrations, with a fair sample of plates reproducing the actual 27 For similar modes in secular manuscripts, see A. Welch, Calligraphy in the Arts of the Muslim World (Austin, 1979), pp. 98–9 and colour frontispiece.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 37 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 2125CC38 /production/pindar/booksxml/hillen/chapter−08

306 size of the object, a just balance maintained between text and illumination, many double-page spreads and the occasional doublures and bindings to underline the importance of the exterior casing of the text itself. Typeface and paper bear witness to the same unfailing dedication to quality. And of course none of this kind of detail looks after itself. But the end result is a set of books which the scribes themselves would have saluted. * The Nasser D. Khalili Collection of Islamic Art. General Editor Julian Raby, Oxford: Nour Foundation in association with Azimuth Editions and Oxford University Press, 1992, Vol. I: The Abbasid Tradition. Qur’ans of the 8th to the 10th Centuries A.D. by François Déroche, 191 pp., 138 col. plates, 12 b. & w. figs., £135.00; Vol. II: The Master Scribes. Qur’ans of the 11th to 14th Centuries by David James, 240 pp., 111 col. plates, 3 b. & w. figs., £135.00; Vol. III: After Timur. Qur’ans of the 15th and 16th Centuries by David James, 256 pp., 140 col. plates, £135.00 This review dispenses with all diacritical marks except ‘ain and hamza

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-08 Sequential 17

JOBNAME: ???? PAGE: 1 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 3076D713 /production/pindar/booksxml/hillen/chapter−09

IX The Relationship between Book Painting and Luxury Ceramics in 13th-Century Iran

A

LTHOUGH at first sight book painting seems to be something of a Cinderella among Salju¯q art forms, the evidence of the finest pottery of the period suggests quite the opposite. The close relationship between these very disparate media has long been recognised, though the details of that relationship have not been explored in depth. Yet a major theme of this symposium has been the frequent migration of themes from one medium to another in the Salju¯q period. Textiles, metalwork and architecture have all been mentioned in this connection.1 The central role of the arts of the book in this process seems to be generally acceptable, but the koine created by painters and designers can be most fully traced not directly in their own but indirectly in another medium – luxury ceramics. It is primarily this pottery that allows conclusions to be drawn not only as to the subject matter of the largely vanished school of Salju¯q book painting, but also as to its stylistic quirks, its clichés – such as filler motifs – and possibly also its favoured compositions. Since books are so much more vulnerable than pottery to the malice or negligence of men, the ceramic record constitutes a vast reservoir of otherwise lost information about the characteristics of contemporary book painting. This may seem a large claim to make, and it must be substantiated before the argument proceeds any further. In Islamic art generally, ceramic decoration and book painting do not show much overlap in style and even

1 See the papers by Drs. Allan, Ewert, Melikian-Chirvani, Raby and Watson published in this volume.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4F61EEDC /production/pindar/booksxml/hillen/chapter−09

308 their subject-matter is not very closely related. The principal exception that comes to mind, apart from the present case, is the close parallel between the script on Sa¯ma¯nid epigraphic pottery and that on contemporary eastern Iranian Qur’a¯ns.2 But this case remained an isolated one until the late Salju¯q period and subsequent centuries produced no further examples of such striking correspondence. Nor is the basic incompatibility of style and subject matter the only reason why close connections between pottery and book paintings are intrinsically unlikely. At the most basic level, the labour involved in painting a piece of pottery is in no way comparable to that involved in preparing an illustrated book. The element of mass production which bulks large in painted ceramics is generally absent in book painting. The huge difference in time and therefore price which naturally results reduces the likelihood that the two arts were supported by the same patrons. Luxury ceramics could scarcely compete in status with the illustrated book. Moreover, the disadvantages which the ceramic painter faced in trying to emulate book painting were legion. His conditions of work were less privileged than those enjoyed by a painter securely ensconced in a royal kita¯bkha¯na. The texture of the surface on which he painted was substantially less smooth than paper. A coating of glaze would not be enough entirely to eradicate this unevenness. Such a surface clearly discouraged precise draughtsmanship. Available evidence suggests that pottery was sold by the piece and not the set,3 which – the Freer beaker notwithstanding4 – discouraged the painter from planning an extensive iconographic cycle with each image deriving an added charge from those which came before. He was committed to producing a single self-contained image per piece. Besides, and most significantly of all, he was forced to operate within the daunting physical constraints imposed by the very shape of the ceramic pieces he was 2

I hope to explore the relationship between these two art forms in detail elsewhere; it embraces not only the type of script but also, it would appear, the use of the qalam in both media, the colour scheme used for the works themselves and their background, and the type and placing of additional ornament of palmette and other types. 3 An exception to this – if a pair may be regarded as a set – is the pair of early 13th-century bowls in the Freer Gallery depicting a king and queen respectively: E. Atıl, Ceramics from the world of Islam (Washington, D.C., 1973), 118–21 and nos. 52–3. 4 For a colour plate, see ibid., 100. For the most recent discussion of the beaker, in which it is argued persuasively that the text illustrated is not that of Firdausı#, see M. S. Simpson, “The Pattern of Early Sha¯hna¯ma Illustration”, Studia Artium Orientalis et Occidentalis I (1982), 46–7; cf. eadem, “The Narrative Structure of a Medieval Iranian Beaker”, Ars Orientalis XII (1981), 15 and 21.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4BB0D569 /production/pindar/booksxml/hillen/chapter−09

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

309

painting. Convex ewers, shallow concave dishes, deep bowls – these were forms intrinsically ill-attuned to the skills of the book painter, and they posed teasing problems of composition and emphasis. Thus while the picture space offered by a leaf from a codex, or a portion of such a leaf, might actually be smaller in surface area, it would be much more straightforward to paint. These are, when weighed together, substantial reasons for doubting the existence of very close links between Salju¯q luxury wares and book painting. But the notable rapport between the illustrations of, say, the Varqa va Gulsha¯h manuscript5 and Salju¯q luxury wares is surely sufficient to dispel these doubts (Figs. 1 and 2). The resemblances embrace the lavish use of cursive script, the figural types employed, the use of the strip cartoon format, the use of either a plain background6 or of one that is dense and busy with curling scrolls,7 and finally the courtly flavour which permeates pot and book alike. Significantly enough, these resemblances apply principally to lustre and mı#na¯’ı# wares; hence the much more abundant everyday pottery of the period is not very relevant to the present enquiry. Though this luxury ware was expensive, the sheer quantity to survive is substantially more (given the vastly greater quantity that must have disappeared) than court patronage alone could have produced. Perhaps, indeed, lustre and mı#na¯’ı# wares served as the poor man’s illustrated book. Centuries before, Sa¯ma¯nid epigraphic pottery (which survives in comparable abundance) had shown in its script, applied ornament and chromatic range that in Iran the arts of the book could usefully serve as a source for fine pottery (Figs. 3 and 4). It may even be that in both cases the connection was deliberately emphasised, so that the pottery might acquire some of the prestige of the codex. The presence of lengthy inscriptions on the pottery could only accentuate this connection. One may conclude that, in Sa¯ma¯nid epigraphic ware and Salju¯q lustre alike, the humbler medium was not simply expanding its range but making available more widely and more cheaply a type of art distinguished by an exclusiveness and costliness which confined it to court patronage. Salju¯q lustreware could, in short, be regarded as an exercise in haute vulgarisation.

5

A.S. Melikian-Chirvani, Le roman de Varqe et Golša¯h, Arts Asiatiques XXII, numéro spécial (1970). 6 Ibid., lower colour plate opposite pp. 98–9; M.S. Ips¸irogˇlu, Das Bild im Islam (Vienna, 1971), pls. 15–16. 7 Melikian-Chirvani, op. cit., upper colour plates opposite pp. 98–9.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4C76E36F /production/pindar/booksxml/hillen/chapter−09

310 The discussion so far has admittedly rested on some rather cavalier assumptions about the close dependence of Salju¯q luxury wares on the arts of the book, or perhaps rather on painters and designers who worked on paper. It is now time to demonstrate in detail the grounds for these assumptions, and the rest of the paper will be devoted to this. En passant, it may be noted that the features to be discussed, though they may be hackneyed in book illustration, are broadly speaking new to ceramic decoration. This should in itself be enough to alert one to the probability that influences or designs from outside the medium are at work. Among these novel features perhaps the most important of all is the particular use of the written word. With the significant exception of Sa¯ma¯nid epigraphic pottery, the inscriptions found on earlier pottery are formal in script and content alike. They are all in Arabic. Various types of Ku¯fic are used and it is normal for the inscriptions to be allotted an important role in the decoration, laid out as it were in capital letters8 and mostly occupying a broad outer band in the case of a plate, bowl, or dish. Their content is undeniably formal and stereotyped, although an earlier categorisation of these messages as “formules banales” has been shown to be too sweeping.9 Good wishes phrased in paired catchwords – “lasting glory”, “perfect felicity”, “eternal power” – reverberate in these inscriptions, forming an almost incantatory sequence. Such inscriptions, which are even more standard in metalwork and may indeed derive from that source,10 continue throughout the Salju¯q period and well beyond. While such inscriptions are common enough on Salju¯q lustre and mı#na¯’ı# ware, a completely different type of inscription also occurs commonly on these ceramic types, and only rarely elsewhere. These latter inscriptions are in a rapidly executed cursive hand, in other words they use the normal contemporary medium of communication (Fig. 5). Thus at a stroke this pottery establishes an association with normal writing as distinct from formal epigraphy.

8

It is remarkable to note how the Arabic alphabet seems never to have been disadvantaged by the absence of a majuscule. 9 Cf. J. David-Weill, “Une inscription banale attribuée au XVe siècle”, Bulletin des Musées de France XI/3 (1946), 15–16 and S. Flury, “Une formule épigraphique de la céramique archaîque de l’Islam”, Syria V (1924), 53–66. 10 The largest selection of these yet published in one volume is to be found in A. S. Melikian-Chirvani, Islamic Metalwork from the Iranian World. 8–18th centuries (London, 1982). For general remarks on this topic see ibid., 16–18.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4EA1D117 /production/pindar/booksxml/hillen/chapter−09

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

311

The content of these cursive inscriptions is if possible even more significant than their style, and it too represents an innovation. For the first time Persian is used, and the text is lyric poetry.11 This novel and fateful combination of an everyday hand (mistakes and all) with Persian poetry, and its use in association with figural compositions, clinches the case in favour of influences from book painting. A brief survey of the pottery bearing such inscriptions is enough to show that it is principally lustre and mı#na¯’ı# pieces that are involved.12 Clearly the association between poetic inscriptions and the figural style of these wares was deliberate: the presence of such inscriptions virtually implied that style, though of course much lustre and mı#na¯’ı# pottery has no inscriptions. The parallel with the manuscript page extends further, however, than the use of Persian poetry written in cursive script. In some lustre ceramics of the period the role of writing is greatly extended until it becomes a major, perhaps even the principal, theme of the piece in question (Fig. 6). This is particularly true of the so-called Jurja¯n wares.13 Occasionally such pottery may even be dated, a detail which implies a certain pride on the part of the craftsman. He at any rate regarded such work as something special. Since the handwriting is too carelessly executed to qualify as a piece of decoration in its own right – in marked contrast to the script on a typical Sa¯ma¯nid epigraphic dish – one must conclude that the content of the inscriptions rather than their style was the focus of attention. This would also help to explain why there is so much writing on such a piece. Here again, then, pottery is usurping the function of a manuscript page. In some cases of this kind the resemblance to a leaf from a book extends even to the presence of multiple marginal lines and of decorative geometric roundels of a kind familiar from contemporary and earlier Qur’a¯ns (Fig. 7).14 Sometimes there is even a

11

For a representative selection see the texts read by Oliver Watson and published in E. J. Grube, Islamic Pottery of the Eighth to the Fifteenth Century in the Keir Collection (London, 1976), 221, 223 and 230. 12 For example, in the Keir collection, which is remarkably representative of medieval Iranian pottery, virtually all the pottery with poetic inscriptions is of lustre or mı#na¯’ı# type, though the occasional piece of underglaze-painted ware is the exception that proves the rule. 13 G. Fehérvári, The Gurga¯n Finds (London, 1976), nos. 40, 45, 50, 67 and 91. 14 M. Bahrami, Gurgan Faiences (Cairo, 1949), pls. 34, 48–9 and 53–5. For parallels in roughly contemporary Qur’a¯ns, see M. Lings, The Qur’anic art of calligraphy and illumination (London, 1976), pls. 24–7.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4A3FCBC8 /production/pindar/booksxml/hillen/chapter−09

312 suggestion of an ‘unwa¯n,15 or of the rows of precisely deliniated finials, like the tassels of a carpet, which are so characteristic of contemporary manuscript illumination (Fig. 8).16 The way that strips of text are arranged over the surface of a dish to form a striking design in themselves (Fig. 9)17 is curiously prophetic of a style of page layout that became popular in the early Tı#mu¯rid period and is exemplified in the poetic anthologies in the British Library and the Berlin museum (Fig. 10).18 In such manuscripts, which are themselves forerunners of the muraqqa‘ fashion, the pages are criss-crossed by adjacent asymmetrical blocks of calligraphy. The texts themselves, the styles of writing and even the direction of the script all vary from one block to the next. Yet this layout is not as random as it might seem at first glance, for any given text reads continuously from one page to the next in its appropriate block. This correspondence between a given text and the shape of its frame obviates any confusion and allows the calligrapher to multiply the number of separate texts on a single page. The use of strips of writing to divide up the compositional space in Salju¯q luxury ware is in some ways akin to this practice, although the circular format encourages rather different expressions of this idea, such as the depiction of a wheel whose hub and spokes are formed by writing (Fig. 11).19 A second element suggesting the influence of book illustration on this pottery is the preference for certain well-defined compositional schemata. Chief among these is the arrangement of successive lines of cursive text in horizontal tiers, in vertical columns, or both.20 The background is white, in contrast to the scrolling background often used for Ku¯fic inscriptions on the same ware, and this makes it natural to suggest the parallel of a book page. The same strip format is often employed for figural compositions (Fig. 12),

15

Bahrami, op. cit., pls. 51 and 54–5. It is noticeable that these echoes of the ‘unwa¯n use the Ku¯fic script, as do actual ‘unwa¯ns, in deliberate contrast to the cursive script used elsewhere on the same pieces of pottery. 16 Ibid., pls. 28b, 47b and 49–51. 17 Ibid., pls. 59, 65–7, 69–70, 83 and especially 87. 18 B. Gray (ed.), The Arts of the Book in Central Asia, 14th-16th Centuries (London, 1979), pl. X and fig. 76; V. Enderlein, Die Miniaturen der Berliner Ba¯isonqur-Handschrift (Leipzig, 1970). 19 Bahrami, op.cit., pls. 25, 28b, 59 and 69–70. 20 Ibid., pls. 80, 82 and 88 and Fehérvári, op.cit., no. 45 (horizontal); Bahrami, op.cit., pl. 27 (vertical); and ibid., pls. 66, 79 and 83 and Fehérvári, op. cit., nos. 44, 84–5 (both).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 50A914F7 /production/pindar/booksxml/hillen/chapter−09

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

313

especially a row of figures passant. Here again book illustration is an obvious analogy – one might cite the Varqa va Gulsha¯h manuscript21 and many a scene from the small Sha¯hna¯mas.22 However, it is not the only parallel, for this device was common in metalwork.23 Especially diagnostic is the use of this strip format within the central roundel of a bowl or dish, since this creates much compositional dead space in the form of an upper and lower exergue.24 This arrangement is not a natural one for a circular format and its use therefore implies external influence of some kind. Sometimes as many as three such strips, one above the other, are found;25 it would be hard to cite a more illuminating example of influence from the arts of the book, an influence so dominant, moreover, that it compels the artist to violate the norms of ceramic decoration. The themes of Salju¯q lustre and mı#na¯’ı# ware are a topic which has yet to be explored in adequate detail. It is therefore premature to draw definite conclusions about the putative influence of book illustration on the subject-matter depicted in these ceramic paintings. Besides, much of the repertoire of lustre and mı#na¯’ı# ware is the stock-in-trade of the minor arts in medieval Islam: scenes of enthronement, hunting, banqueting, musicmaking and similar courtly pastimes; and various depictions of animals and birds. There is nothing particularly revealing about such themes in the present context. What is rather unusual, however, is the self-imposed limitation of subject-matter in mı#na¯’ı# ware. Here one may note an almost total absence as the major motif of the animal themes which are such a standby in other Salju¯q pottery; nor is abstract design, whether geometrical or floral, often allowed the dominant role it so frequently takes in other contemporary ceramics. When such designs are used, their kinship with contemporary modes of book illustration is striking (Fig. 13). It is as if the painters who decorated mı#na¯’ı# ware were at pains to emphasise the similarity between their repertoire and that of book painters, and by the same token to deny their

21

Melikian-Chirvani, Varqe et Golša¯h, passim. M.S. Simpson, The Illustration of an Epic. The Earliest Shahnama Manuscripts (New York and London, 1979), pls. 2–3, 8, 11–12, 18, 20 and so on. 23 Melikian-Chirvani, Islamic Metalwork, pls. 56, 75–6 and 99 among many others. For the occurrence of epigraphic bands in horizontal tiers and vertical columns on the same object see ibid., pl. 45. 24 Bahrami, op.cit., pls. 38, 49–50, 53, 56, 73. 25 Ibid., pl. 31; A.U. Pope and P. Ackerman (eds.), A Survey of Persian Art from Prehistoric Times to the Present (London and New York, 1939), pl. 660A. 22

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 35 OUTPUT: Tue Aug 10 10:28:56 2010 SUM: 4BCE8B80 /production/pindar/booksxml/hillen/chapter−09

314 own association with other media. It is therefore not suprising that mı#na¯’ı# rather than lustre pottery betrays the most unmistakable signs of dependence on book illustration, as the celebrated Freer beaker, with its superposed tiers of Sha¯hna¯ma scenes, shows. Mention of that beaker brings to mind one aspect of this ceramic iconography for which the most though not the only satisfactory explanation is the acceptance of prototypes in book illustration – namely the presence of narrative scenes. There is little need to labour the point that a single object of limited surface area is an unsuitable vehicle for narrative. It is significant that such scenes do not on the whole occur in metalwork or textiles, and in medieval pottery they are virtually confined to Salju¯q lustre and mı#na¯’ı# ware. Often the degree of strain is such that the scene barely fits within the confines of the dish (Fig. 14). In that strain may lie the clue to the source of the image. An image conceived specifically for that picture space would presumably fit into it more comfortably. As it is, the Freer mı#na¯’ı# dish depicting a battle and the Freer lustre dish dated 1210 (Fig. 6) both bear the marks of a composition transferred from a large area to a small one.26 The elisions required to accommodate a complex narrative within a limited picture space strain the artist’s storytelling technique to the utmost,27 and indeed sometimes render the scene incomprehensible. The same applies to such complex and formal (though not narrative) compositions as enthronement or processional scenes.28 Aside from the use of cursively written Persian poetry, compositional clichés and iconographic themes, a fourth link between ceramic and manuscript painting deserves further discussion here, namely style. Perhaps the most significant feature in this respect is the technique of draughtsmanship employed. At once hurried and confident, it is the natural pendant to the calligraphy on this luxury ware and therefore invites the speculation that scribe and painter were one and the same person, as indeed was sometimes the case in the production of illustrated manuscripts, for example the Schefer H.arı#rı#. Moreover, this same boldness and rapidity characterises much of 26

E.J. Grube, The World of Islam (Feltham, 1967), colour pls. 17–18. Ibid., colour pl. 16; idem, Islamic Pottery, colour pls. opposite pp. 200 and 201. 28 Idem, Islamic Pottery, pl. opposite p. 217; D. Jones and G. Michell (eds.), The Arts of Islam (London, 1976), no. 350; A. Lane, Early Islamic Pottery (London, 1947) pls. 62B and 63A; Survey, pls. 641B, 642 and 689; and F. Bagherzadeh et al., The World’s Great Collections. Oriental Ceramics. Vol.4. Iran Bastan Museum, Teheran (Tokyo, New York and San Francisco, 1981), pl. 35. 27

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09 Sequential 8

JOBNAME: ???? PAGE: 1 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 1B1EDF3B /production/pindar/booksxml/hillen/chapter−09−artwork

315

1. Gulsha¯h reveals himself to Varqa who is led away as a prisoner. From Varqa va Gulsha¯h (after E.J. Grube, The World of Islam).

2. Mı¯na¯’ı¯ bowl with cavaliers (after A.U. Pope and P. Ackerman, eds., A Survey of Persian Art).

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 0952485C /production/pindar/booksxml/hillen/chapter−09−artwork

3. Sa¯ma¯nid epigraphic dish.

4. Page from Qur’a¯n written in 485/1092, in Iraq or Iran (after M. Lings, The Qur’anic Art of Calligraphy and Illumination).

316

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 094FF666 /production/pindar/booksxml/hillen/chapter−09−artwork

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

317

5. Lustre albarello with Persian poetic inscriptions (after G. Fehe´rva´ri, Exhibition of Islamic Art. Iranian Lustreware of the Thirteenth Century).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 0447FC19 /production/pindar/booksxml/hillen/chapter−09−artwork

318

6. Ka¯sha¯n-style ewer (after O. Watson, Persian Lustre Ware).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 0BFF8408 /production/pindar/booksxml/hillen/chapter−09−artwork

319

7. Mina¯’ı¯ bowl with a Sha¯hna¯ma scene (after E. Atıl, Ceramics from the World of Islam).

8. Lustre plate dated 607/1210 (after M. Bahrami, Gurgan Faiences).

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 043A4891 /production/pindar/booksxml/hillen/chapter−09−artwork

320

9. Lustre jug (after M. Bahrami, Gurgan Faiences).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 31 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 09F1163B /production/pindar/booksxml/hillen/chapter−09−artwork

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

321

10. Shaikh .San‘a¯n and the Christian girl, f.96 of an Anthology made for Ba¯isunghur in 823/1420, now in the Islamisches Museum in Berlin (after V. Enderlein, Ba¯isunqur-Handscrift).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 081AE4C9 /production/pindar/booksxml/hillen/chapter−09−artwork

11. Lustre bowl (after M. Bahrami, Gurgan Faiences).

12. Lustre dish (after O. Watson, Persian Lustre Ware).

322

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 0AE29F7F /production/pindar/booksxml/hillen/chapter−09−artwork

323

13. Mı¯na¯’ı¯ dish (after A.U. Pope and P. Ackerman, eds., A Survey of Persian Art).

14. Lustre bowl (after O. Watson, Persian Lustre Ware).

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 30 OUTPUT: Fri Feb 10 09:57:24 2012 SUM: 086C4A4D /production/pindar/booksxml/hillen/chapter−09−artwork

15. Lustre bowl (after E. Atıl, Ceramics from the World of Islam).

16. Lustre plate dated 607/1210 (after E. Atıl, Ceramics from the World of Islam).

324

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-artwork Sequential 10

JOBNAME: ???? PAGE: 1 SESS: 36 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 56DE7050 /production/pindar/booksxml/hillen/chapter−09−ends

RELATIONSHIP BETWEEN BOOK PAINTING AND LUXURY CERAMICS

325

12th- and 13th-century miniature painting, including the Varqah va Gulsha¯h manuscript. The emphasis on perfection of technique and purity of colour which was to characterise Persian book painting from the later 14th century onwards was not yet fully developed, though the best examples of the style are entirely worthy of fine book illustration. Thus the drawing style of Salju¯q lustre and mı#na¯’i ware would not have looked out of place in contemporary book painting. By way of contrast, it could scarcely have been more out of sympathy with the drawing technique of other contemporary or earlier Islamic pottery in Iran, which is altogether coarser and employs deliberate distortion and caricature as a regular device.29 Attention should also be drawn to the miniature scale of the painting on luxury Salju¯q wares, especially in mı#na¯’ı# ware. Even when the design is dominated by a single large figure, as is common in lustre ware (Fig. 15), the details of costume and background are treated with a precision which is alien to other contemporary Salju¯q pottery and finds its natural analogy in book painting. An obsession with surface patterning applied impartially to foreground and background, principal and secondary subject matter characterises both lustre painting and some of the illustrations of Varqa va Gulsha¯h.30 Facial types are perhaps the most obvious and frequently noted similarity between the two types of painting, and so there is no need to expatiate further on this subject here.31 It is perhaps worth noting that lustre and book painting share the habit of making a compositional feature of rows of bland white circular faces set in haloes (Fig. 16).32 Mı#na¯’ı# painting does not emphasise this feature, evidence perhaps that it draws from a collateral but distinct source of book painting. It is more likely, however, that at least one clue to the use of these serried faces may lie in the nature of the background employed. Where it is a deep, dark 29 This coarseness is often laid at the door of technique, in that cutting rather than painting is employed to execute the design. Nevertheless, even painted wares, whether they are ‘Abba¯sid lustre, Nı#sha¯pu¯r figural ceramics or the pottery of Ma¯zandara¯n, exhibit the self-same coarseness and distortion. 30 Compare the plate showing a warrior stabbing a leopard (Grube, Islamic Pottery, pl. opposite p. 216) with the scene in Varqa va Gulsha¯h where the father of the heroine buries a sheep (Melikian-Chirvani, Varqe et Golša¯h, pl. 46). 31 A.S. Melikian-Chirvani, “Le legs littéraire du bouddhisme iranien”, Le Monde Iranien et l’Islam II (1974), 1–71. 32 Compare the Freer lustre plate dated 1210 (Grube, World of Islam, pl. 18) with the frontispiece of the Kita¯b al-Agha¯nı# in Istanbul or the scene of the royal pavilion in the Kita¯b al-Dirya¯q in Paris illustrated in R. Ettinghausen, Arab Painting (Geneva, 1962), pls. on pp. 65 and 85 respectively.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 36 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 2B0F247B /production/pindar/booksxml/hillen/chapter−09−ends

326 colour such faces obviously have the function of providing accents within the composition; against the white background so standard in mı#na¯’ı# wares they could play no such role. On the other hand, that same white background provides the ideal chromatic foil for the numerous bright colours used in mı#na¯’ı# ware, and in this respect the influence of book painting is manifest. Other stylistic tags shared by luxury ceramics and book painting which cannot be investigated in detail here include the use of space-filling plants; continuous arabesques, palmettes, split-leaved plants and other vegetal scrolls all forming a seamless backcloth to the action; the presence of animals and birds, often coloured or patterned in bizarre fashion, within such a background; the preference for the frontal plane and the silhouette mode; and finally the use of a coloured background in shades of red, yellow and brown. Discussion of these features in the requisite detail will provide the material for another and more extensive paper to be published elsewhere. The evidence marshalled here is by no means exhaustive, but it will perhaps suffice for the present. Shortage of space makes it impossible to embark here on the obvious next stage of this enquiry, namely discussion of the clues which the luxury wares offer as to the probable nature of the vanished book-painting of late Salju¯q times. But at least one may hope that the “cascades of speculation” necessarily generated by such an enquiry will now be monitored by a slightly clearer understanding of the debt which late Salju¯q ceramic painting owed to the traditions of book illustration.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-09-ends Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 2E2EE9B4 /production/pindar/booksxml/hillen/chapter−10

X The Message of Misfortune* Words and images in Sa cdi’s Gulistan HE need to investigate closely the relationship between words and images in Persian miniature painting has long been recognised, but the principle is one that has been more honoured in the breach than in the observance. Literary historians, regrettably, have so far shown virtually no inclination to do such work; and most art historians have lacked the high degree of linguistic competence required.1 The present paper, then, ventures into unfamiliar territory.2 The manuscript to be investigated is an early Timurid copy of Sa‘di’s Gulistan in the Chester Beatty Library, Dublin.3 It is what reviewers call a slim volume – a mere 56 folios – and its covers measure ten inches by six. The written surface is significantly smaller, being contained within a space of seven inches by four, with some 23 lines per page. The text is written in

T

* I should like to thank David Roxburgh, who took working photographs of the manuscript on my behalf. 1 For an exception, see the article on a 16th-century illustrated manuscript of the Bukhara school by A.S. Melikian-Chirvani, in R. Hillenbrand (ed.), Studies in Persian painting, Festschrift for B.W. Robinson (in press). 2 A fair amount of what is presented here is admittedly speculative; but it should be a step in the right direction. 3 For a basic description of the manuscript, see A.J. Arberry, E. Blochet, M. Minovi, J.V.S. Wilkinson and B.W. Robinson, The Chester Beatty Library. A Catalogue of the Persian Manuscripts and Miniatures, I, Dublin 1959, p.39.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 5F5227C6 /production/pindar/booksxml/hillen/chapter−10

328 nasta‘liq, then a newly fashionable script,4 and was copied in 830AH (1427AD) by the leading calligrapher of the time, Maulana Ja‘far Tabrizi, who took the sobriquet of Baisunghuri after the Timurid prince whose atelier (kitabkhana) he ran.5 This circumstance, together with the dedication to Baisunghur6 (1397–1433) and the unusual emphasis on illumination, suggests that the painter, too, was one of the outstanding exponents of his craft. Among its forty artists, the royal atelier counted gilders, gold-sprinklers, musicians, poets, calligraphers, painters, mystics, bone-carvers and singers; and some of them were polymaths who excelled in several of these fields.7

4 Evidence suggesting that this script was developed in the 1370s was presented by Elaine Wright in a paper delivered at the symposium The Art of the Mongols, Edinburgh, July-August 1995. See also W.M. Thackston, A Century of Princes. Sources on Timurid History and Art, Cambridge, Mass. 1989, p.341. 5 See Arberry et al., op.cit., p.39. For Baisunghur’s kitabkhana, see T.W. Lentz, Painting at Herat under Baysunghur ibn Shahrukh, unpublished Ph.D. dissertation, Harvard University, 1985, and T.W. Lentz and G.D. Lowry, Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century, Los Angeles and Washington 1989, pp.159–236, esp. 165–169. For further information on Ja‘far Tabrizi, see Thackston, op.cit., p.340. 6 The ex libris reads “For the treasury of the books of the eminent sultan Baisunghur Bahadur Khan, may Allah prolong his reign for ever”, Arberry et al., op.cit., p.39. 7 The principal evidence for this is contained in the so-called Arzadasht, literally ‘petition’ but effectively a progress report on the activities of a royal Timurid atelier (Istanbul, Topkapı Sarayı Library, H.2153, f.98a), datable ca. 833AH/1430AD; for a translation and commentary, see Thackston, op.cit., pp.323–7. Of particular interest is the reference right at the beginning of the Arzadasht, and thus in the position of honour, to one Amir Khalil, who “has finished the waves in the two sea scenes of the Gulistan and will begin to apply colour”. He was the supervisor of painting in Baisunghur’s atelier, while the chronicler Daulatshah refers to him as a “second Mani” (Mani being the legendary exemplar of the arts of painting) and as one of four “craftsmen at the Shahrukhid capital who had no equal in their time in the inhabited quarter [of the world]” (Thackston, op.cit., p.323). Since the Dublin Gulistan has two sea scenes, and is dated within three years of the suggested date of the Arzadasht, this is almost certainly the manuscript in question. A few lines later there are references first to one Khwaja Ghiyath al-Din, who is not only engaged on illustrating a manuscript entitled Rasa’il but is also “at present … busy repairing a scene that was spoiled in the Gulistan” (ibid.), and then to Maulana Shihab, who “at present is busy with another scene in the repair of the Gulistan” (ibid., p.324). Thus there seem to have been at least three artists working on these paintings. Still further on, there is a statement that “Khwaja Ata has finished the sections of the Gulistan” (ibid., p.325), and since his other work is described as illumination, the likelihood is that it is his work on the illumination rather than on the calligraphy of the Gulistan that is meant here. Amir Khalil was not only the chief painter but also the boon companion of Baisunghur, as is revealed by a lengthy anecdote recounted by Dust Muhammad in his preface to the Bahram Mirza album (Istanbul, Topkapı Sarayı

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 51AC1A78 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

329

Thus there is nothing untoward in allowing for the possibility that the illustrator of the Dublin Gulistan had wider horizons than painting; that he might, indeed, have had professional views of his own on the text that he was called upon to embellish. Several members of the atelier survived Baisunghur by thirty, fifty, and in one case, some seventy years,8 which suggests that some members of this team of stars were taken on as quite young men, perhaps with young ideas to match. The manuscript contains eight paintings, spaced very unevenly throughout the book at folios 3, 9, 14, 15, 16, 29, 32 and 36. Thus there are significant gaps of 13 and 20 folios respectively without illustrations.9 But the sheer plethora and variety of illumination on all the pages which lack paintings ensures that these still make a sumptuous impression. The illustrator of this manuscript had a number of problems to overcome, and all of them affected his approach to his text. First of all, and most important, the Gulistan had apparently not been illustrated before.10 In this respect the painter was at an obvious disadvantage vis-à-vis his colleagues charged with illustrating the texts of Firdausi or Nizami. There was no established cycle to follow. He himself – at least in theory – had to make up suitable images. And of course the difficulties thus created were not

Library, H.2154; tr. Thackston, op.cit., p.346). For further discussion of Amir Khalil, see B.W. Robinson, Fifteenth-Century Persian Painting. Problems and Issues, New York and London 1991, p.7–8 and idem, ‘Prince Baysunghur and the Fables of Bidpai’, Oriental Art XVI, 1970, especially. pp.150–154. 8 Compare Thackston, op.cit., p.342, for a discussion of Sultan ‘Ali, a calligrapher whose working life – as documented from signatures – appears to have lasted 69 years. 9 It is strange that such a long passage – more than a third of the entire text – should have remained unillustrated. When coupled with the evidence of the Arzadasht that work on the manuscript continued after the date of its colophon, this fact might suggest that the volume was never finished. The reference in the Arzadasht to damage sustained by at least one of the paintings might have something to do with the book’s apparently unfinished state. 10 V.R. Prentice, The Illustration of Sa‘di’s Poetry in Fifteenth-Century Herat, unpublished Ph.D. dissertation, Harvard University, 1977, 1. I am most grateful to Dr Prentice for generously lending me her own personal copy of her dissertation and allowing me to make a copy of it. Sa‘di lived from ca.1213–1292 (see Prentice, op. cit., p.2, for the range of dates which have been suggested, which extend from 1184 to 1219), and the Gulistan had been recognised as a classic for well over a century by the time that Baisunghur ordered his copy. The Dublin Gulistan is preceded by a copy with Shiraz-style illumination in the India Office Library dated 819AH/1416AD (B.W. Robinson, A Descriptive Catalogue of the Persian Paintings in the Bodleian Library, Oxford 1958, p.16).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 4FD892D7 /production/pindar/booksxml/hillen/chapter−10

330 confined to the invention of new scenes. Much of the strength and effectiveness of traditional sequences of illustrations lay in the way that individual images balanced, complemented or enriched each other. Gifted painters knew how to develop leitmotifs so that a given picture would contain resonances, references to other paintings in the cycle, that would be picked up by an attentive observer.11 To achieve that from scratch was a very tall order, since much of the success of this game depends on familiarity. Accordingly, it is only to be expected that some of the paintings in this Gulistan should betray signs that this is a pioneering effort. The repetition of similar scenes, for example the two seascapes (figs. 3, 11),12 or the preference for figures in isolation outside a cliff-like facade,13 is one such sign, and a very marked one given that there are only eight paintings. Another is the tailoring of familiar iconography to slightly different ends. Thus the scene of the lustful prince lunging at the Chinese maiden (fig. 8) owes much to images of amorous encounters between Rustam and Tahmineh, or Khusrau and Shirin.14

11

The self-conscious and self-referential nature of court painting at this time is perhaps best revealed by the custom of including free copies of earlier masterpieces in each new cycle. See A. Adamova, ‘Repetition of compositions in manuscripts: the Khamsa of Nizami in Leningrad’, in Timurid Art and Culture. Iran and Central Asia in the Fifteenth Century, L. Golombek and M. Subtelny (eds.), Leiden, New York and Cologne 1992, pp.67–75; Lentz and Lowry, op.cit., pp.376–379, and N.M. Titley, ‘Persian miniature painting: the repetition of compositions in the fifteenth century’, in Akten des VII. Internationalen Kongresses für Iranische Kunst and Archa¨ologie, W. Kleiss (ed.), Berlin 1979, pp.471–91. 12 The seascape was a theme which had quite recently attracted the interest of progressive painters, perhaps because it fitted the new fashion for juxtapositions of colour. Examples include the Gulbenkian Anthology of 1410 in Lisbon, LA. 161 (B. Gray, Persian Painting, Geneva 1961, p.76), the Anthology of 1410–11 in the British Library, Add.27261 (Titley, op.cit., fig.16) and the Khamsa of Nizami in St Petersburg, dated 1431, VP-1000 (ibid., fig.17). 13 This device, which is employed three times in the Dublin Gulistan, takes up a theme which was first exploited by the painter of ‘Humay before Humayun’s castle’ in the Kulliyat of Khwaju Kirmani in the British Library, Add.18113, dated 1396, and was thereafter appropriated by one fashionable painter after another (for typical examples, see Lentz and Lowry, op.cit., p.376). 14 See respectively Lentz and Lowry, op.cit., p.130 and N. Atasoy, ‘Four Istanbul Albums and some Fragments from Fourteenth-Century Shah-Namehs’, Ars Orientalis VIII, 1970, fig.25; ibid., fig.21 and B. Gray, The Shahnameh of Ferdowsi. The Baysonghori Manuscript. An Album of Miniatures and Illuminations, Teheran 1971, pl.XVIII; and I. Stchoukine, Les peintures

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 4F3CD26E /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

331

A second and quite different problem was posed by the prodigal wealth of illumination which this particular manuscript contained. So ubiquitous is it, and so varied, that one wonders what precisely the status of the paintings themselves could have been. Was there a degree of competition in the royal kitabkhana about the respective roles of illumination and illustration in a fine manuscript? This Gulistan is dated in the early years of Baisunghur’s activity as a patron of the arts of the book.15 Perhaps, then, it documents a transitional stage in the formation of his taste. Eventually, of course, he came to favour books in which paintings, though still relatively few in number, came to dominate, asserting their presence by their sheer density of content.16 The Gulistan has not arrived at this point, but illustrates a different way of creating a luxury book – and one which has unmistakable associations with fine Qur’ans.17 As it happens, this way proved to be a dead end. But en route it manifestly produced a masterpiece of Timurid ornament.18 At intervals throughout the fifteenth century, moreover, the odd especially select manuscript was produced in which paintings were subordinated to, or even replaced by, pages tinted mauve or powder blue, sprinkled with gold or silver, with the central panel of text surrounded by scenes in a kind of grisaille.19 In Baisunghur’s Gulistan the mix of painting, calligraphy and illumination in markedly different styles may have been intended as a deliberate response to the variety of Sa‘di’s text, with its lightning

des manuscrits de la “Khamsah” de Nizami au Topkapı Sarayı Mu¨zesi d’Istanbul, Paris 1977, pl.IIa. Images of Bahram Gur in pavilions of various hues are another possible source: Lentz and Lowry, op.cit., p.377. 15 For a list of manuscripts attributed or dedicated to Baisunghur, see Thackston, op.cit., pp.325–326. 16 His Shahnama of 833AH/1430AD in Teheran (Gulistan Museum, 4752) is the classic case. See Lentz, Painting at Herat, pp.100–109 and 385–387. 17 It is no accident that this period represents the high-water mark of Timurid Qur’ans. For a brief overview, see D. James, After Timur. Qur’ans of the 15th and 16th centuries. The Nasser D. Khalili Collection of Islamic Art, General Editor J. Raby, vol.III, London and Oxford 1992, pp.14–45, Lentz and Lowry, op.cit., pp.72–79 and M. Lings, The Qur’anic Art of Calligraphy and Illumination, London 1976, pp.171–187. 18 Moreover, the manuscript explores roles for illumination which are not encountered in Timurid Qur’ans. Indeed, the illumination deserves a study to itself. The experimental nature of the Baisunghuri atelier is revealed by the Chahar Maqala and by the illustrations to the Kalila wa Dimna (notes 67–68). 19 E.g. Lentz and Lowry, op.cit., p.121; for cognate experiments, see ibid., pp.128–129, 190, 202, 204–205, 270.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 4D3A9210 /production/pindar/booksxml/hillen/chapter−10

332 alternations of mood and mode as the author switches from verse to rhymed prose, from Arabic to Persian or from narrative to lyric.20 A third problem confronting the painter of the 1427 Gulistan was the nature of the text which he was charged to illustrate. It is at the opposite pole to the leisurely pace at which the action so often unfolds in Firdausi’s Shahnama or even in the Khamsa of Nizami – the two most familiar illustrated texts of the Timurid period. The Gulistan moves swiftly from one tale to the next. For the most part the story is told without factual embellishment. It begins in medias res and continues in the same brusque fashion. Sa‘di fleshes out the skeletal narrative not by anecdotal detail but by a helter-skelter succession of verses in Arabic and Persian. One aphorism succeeds another. The literary structure is often antithetical or repetitive; thus a moral is drawn by three or four examples. These enrich the texture of the passage and illumine the central theme from a series of different angles.21 It seems reasonable to look for evidence that the painter attempted to translate some of these devices into visual terms. This would be wholly in accordance with the way that the entirety of the text is presented in the Dublin manuscript, which shows a consistent and lively awareness of how the component elements of the text should be presented – and distinguished from each other.22 Yet at the same time it would be idle to expect any consistently exact parallelism between the literary and linguistic devices used by the poet and those available to the painter. They speak different languages, and that is in the nature of things. What can be expected is that the painter will try to complement and enrich the text in his own fashion. Sometimes he may embroider upon it so assiduously that the original text is obscured.

20

In this respect, the manuscript is strangely prophetic of the muraqqa‘ which was to be so distinctive a feature of 16th-17th-century taste, and in which new relationships between text, calligraphic panels, illumination and illustration were constantly being explored. See the study of this form by A. Adamova in Parthian and Sasanian Themes in Iranian Art, V. Curtis, R. Hillenbrand and J.M. Rogers (eds.), in press. 21 For a brief conspectus of the literary nature of the Gulistan, see A. Schimmel, ‘The Genius of Shiraz: Sa‘di and Hafez’, in Persian Literature, E. Yarshater (ed.), Albany 1988, pp.216–217. See also the discussion of overlap at the end of the present article. 22 It is especially ingenious in ringing the changes from one folio to the next in the page layout. Thus the degree of indentation employed for quotations is in a state of constant flux, with the announcement that a new tale is coming (hikayat) being located at the left edge of the margin, perfectly centred or inclining to the fore-edge.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 455A34AA /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

333

So far as the story itself is concerned, then, the painter’s marching orders are clear. It is his job to spotlight the moral, which is exactly what the text itself is designed to do. And the nature of Sa‘di’s text both helps and hinders him. It helps him in that the poet’s extremely pithy narrative style – if indeed the word narrative is appropriate for so bald a means of expression – tends to focus unwaveringly on the point of the story. When Sa‘di digresses, it is not to pad out the story by telling us, for instance, when and where a king ruled, or his name, or his physical appearance, or what kind of a person he was. Instead he prefers to insert verses which sententiously point the moral and adorn the tale. So the painter can be in no doubt as to the proper object of his attention. But Sa‘di’s text is that much more difficult to illustrate precisely because of the absence of supplementary detail. Does the scene take place inside or outside a building? Are the principal actors alone, or is there an audience? What else is happening at the same time?23 These are matters which the painter has to decide, and he has to tread carefully. If there is too little going on visually, the painting can be in danger of looking trite or disagreeably empty.24 But at the opposite extreme lies confusion, so much detail that the eye cannot order nor the mind digest it.25 And in steering a safe course between these poles, any sensitive artist would presumably have sought a means of visual expression that dovetailed not only with the actual words of his text but also with its mood. The Dublin Gulistan clearly possesses that general unity of style which is only to be expected in a royal manuscript, and especially in one illustrated by a single painter. But this does not stop the artist developing several quite different approaches to the text chosen for illustration. Sometimes he tells the whole story straightforwardly; at other times he so manages matters that we are left in suspense as to what will happen; and then on occasion he vouchsafes clues as to how the story will end. Each of these three responses to the text involves him in a degree of interpretation, but his intellectual

23

Another area of ambiguity is what the weather or season is, though sometimes, when this is a crucial factor, it is made clear – for example, the distressed poet, and Sa‘di encountering the maiden. 24 As in much Muzaffarid painting: e.g. B. Gray (ed.), The Arts of the Book in Central Asia, Paris and London 1979, pls.72–73 and XXXIV. 25 A textbook case is the battle scene from the copy of Nizami’s Khamsa made for Shah Tahmasp and dated 1539–43 (Gray, 1961, p.134).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 4AFF8670 /production/pindar/booksxml/hillen/chapter−10

334 input varies from one to the next. This will become clear as we look at examples of each of these three approaches in turn. Let us begin with an instance of straightforward exposition. The garden scene on folio 3 (fig. 2) closely follows the relevant accompanying text and depicts a nocturnal conversation between Sa‘di and a friend. His own age is not mentioned in the text,26 and it may be that the choice of an old man was dictated by an already established model – the youth consulting a sage or Majnun visited in the wilderness – which the artist happily adapted to his purpose.27 This scene is the odd one out among these eight paintings. Obviously it is intended to depict not action, as do the other seven pictures, but contemplation. The poet’s meditations are about to be made still more comfortable by means of the extra pillow which a servant is bringing him. In some sense this is a display piece intended to show that the painter can manage this kind of composition, which had been popular for at least a generation.28 The text on the picture page mentions “the dew-fallen pearl on the damask rose” (the reference to the rose, gul, is perhaps intended as a pun, as will shortly appear) “heart-gladdening groves” and “clusters of fruit like the pleiades [which] hung aloft from its [the garden’s] boughs”.29 This natural abundance finds its ample counterpart in the painting. But on the very next page Sa‘di falls to musing that the flowers of such a garden are doomed to fade; and so he determines to “write such a Kitab-i Gulistan as neither the rude storm of autumn shall be able to lay the hand of usurpation upon its leaves, nor the revolution of the season convert the serenity of its summer into the gloom of winter … this rose-bower must bloom to all eternity!”30 Thus the

26

Sa‘di himself refers to his reason for writing the Gulistan as follows: “Every moment a breath is expiring of my life; when I curiously inspect it, I find that only a little is left. O man! fifty years of thy life are gone …” (J. Ross, Sadi: Gulistan or Flower-garden, London, n.d., p.67). 27 E.g. M.S. Ips¸irog˘lu, Saray-Alben. Diez’sche Klebebände aus den Berliner Sammlungen, Wiesbaden 1964, pl.XIV or R. Ettinghausen, Persian Miniatures in the Bernard Berenson Collection, Milan 1961, pl.VI. It could also be a free adaptation of a royal ‘picnic’ of the kind which was popular at this time for frontispieces (e.g. that for Baisunghur’s Kalila wa Dimna of 833AH/1429AD in the Topkapı Sarayı Museum, Istanbul, R.1022: see Lentz and Lowry, op.cit., pp.110–111). 28 E.g. Gray, 1979, pl.XXXIV. For a close contemporary parallel, from the Chahar Maqala of 835AH/1431AD made for Baisunghur (Istanbul, Türk ve Islam Eserleri Müzesi, no.1954, f.22a) see ibid., pl.L. 29 See Ross, op.cit., p.70. 30 Ibid., p.71.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 4E2C2AD7 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

335

context explains alike the placing and the mood of this picture – although it might be argued that the painting would have fitted more relevantly into the following passage of text. It is a visual preface, an extended pun; and furthermore, it helps to explain the absence of a painted frontispiece. Here on folio 3 is the frontispiece;31 but it is more than just an author portrait, for its close integration with the content of the text makes it express the raison d’être of the manuscript. Thus a hackneyed image has been turned to new account. Given the long history of the author portrait in manuscript painting, this is no mean feat.32 Let us turn now to the second category, namely scenes in which the artist leaves us guessing. In the pictures depicting the beggar before the King (fig. 5) and the distressed poet attacked by dogs (fig. 6), the text incorporated into that page ends at a critical point in the story, and the illustration accurately mirrors this state of suspense. This is, incidentally, evidence suggesting that the artist did not necessarily see his task as illustrating the entirety of the story in question. Instead, he seems to have made his own choice of the most appropriate part of the tale to depict. Will the poet escape the fangs of the village curs? Is the courtier opening – or closing – that portal so auspiciously inscribed with the invocation “O Opener of gates!”? Will the shivering dervish receive charity or the royal elbow? The action is finely balanced between triumph and disaster, and the picture itself gives nothing away. For the answer, the reader has to turn to the next page. Let us look at one of these two pictures in more detail to see how the story is told. Attention focuses on the half-naked dervish who crouches with hunched shoulders, clasping his knees against the cold, his ribcage distressingly visible (fig. 7). He is set against a largely barren landscape coloured a frigid blue and relieved by one large and a few tiny red and green flowers. These same two colours are taken up three more times in a diagonal slant marked in turn by the courtier lounging at the corner of the palace, the prince leaning over the balcony and finally a figure at the upper window. One is reminded here, as in the other examples of colour parallelism in these paintings, of the use of rhyme – and, as in literary rhyme, there is room for variety: not only aabb but 31 Cf. Prentice, op.cit., p.38. It was not rare for the frontispiece to be sited a folio or two after the beginning of the book (e.g. the double frontispiece of the Rasa’il Ikhwan al-Safa in the Library of the Su¨leymaniye Mosque, Istanbul, Esad Efendi 8638, which occupies ff.3b and 4a: R. Ettinghausen, Arab Painting, Geneva 1962, pp.98–99). 32 E.R. Hoffman, ‘The Author Portrait in Thirteenth-Century Arabic Manuscripts: A New Islamic Context for a Late-Antique Tradition’, Muqarnas X, 1993, pp.6–20.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 44BDD883 /production/pindar/booksxml/hillen/chapter−10

336 abab and other combinations.33 The russet-hued palace radiates warmth, and vaunts its own beauty in the golden inscription which runs along the cornice: O loftiness higher than the gold-adorned turquoise-blue palace, Shall I call you the rosebed of the soul or the garden of the Abode of Permanence?34 Elsewhere in the Gulistan references may be found to moralising inscriptions, as on the palace of Feridun, the tombstone of Bahram Gur and the crown of Kaikhusrau.35 This one seems somewhat flatulent by contrast, but it too has its purpose when seen from the worm’s-eye view of the dervish. It serves to widen the gap between the haves and the have-nots – a leitmotif of the illustrations in this manuscript. As it happens, the ready wit of the dervish wins him a thousand-dinar purse from the monarch, who (on the next page) tells him “‘O dervish! Hold up your skirt!’ He replied ‘Where can I find a skirt, who have not a garment?’’’ and was duly given a robe of honour as a bonus.36 So the story ought to have ended, but it did not. The dervish contrived to squander his thousand dinars in a few days, and rashly came back for more; but his wit had degenerated into chutzpah, and he was sent away empty-handed. Thus his situation at the end of the story is the same as it was at the beginning, and in that sense the message of the picture – which exudes pain rather than comfort – is entirely appropriate. The third category, in which the painter offers clues as to how the story will end, is the most interesting of all. In the only interior scene of the whole manuscript, a wasp-waisted prince is trying to overpower a Chinese maiden (fig. 8). Sa‘di’s language is clinical. “In a state of intoxication he wanted to copulate with her. The damsel resisted. The king flew into a passion” – here the text on the illustrated page ends, and continues on the next page as

33 The Anthology of Sultan Iskandar in the British Library (Add.27261), dated 813–4/ 1410–11, is an apt illustration of experiments with complex antiphonal colour rhythms. 34 A. Serajuddin, Architectural representations in Persian Miniature Painting during the Timurid and Safavid Periods, unpublished Ph.D. thesis, University of London, 1968, p.80. The “turquoise-blue palace” is the sky. 35 Ross, op.cit., pp.78, 113 and 166. 36 Ibid., p.93. The text on the page with the painting ends with the dervish pleading with the monarch: “0 thou who in good fortune hast not thy equal in the world, I admit that thou hast no cause of care for thyself, but hast thou none for us?” It is only on the next page that we learn that “the king was pleased with this speech”.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 423A4B00 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

337

follows: “… and forced her upon a negro, whose upper lip projected along the sides of his nose, and the lower one doubled down upon his chin. He had such a hideous appearance that the demon Sakhr would have been terrified at seeing him, and fountains of liquid pitch trickled from his armpits.” This fearsome character, Sacdi says, “the slave of his libidinous and carnal appetites … gave way to his lust, and deflowered the damsel.”37 These coming events cast their shadow over the painting. The ardour of the prince has pushed the Chinese maiden almost to the limit of the picture frame; she is cornered in the most literal fashion between two bolsters. At the opposite extreme of the painting is the negro lurking in the doorway, a diabolus ex machina awaiting his opportunity. This parallelism is, as we shall see, developed further. For the moment it is enough to notice that this visual connection between the girl and the negro, which embraces the entire width of the picture, sets up a tension between the ostensible subject matter and the hidden agenda, between the present and the future. The chamber is lavishly decked out with patterned textiles and tilework, and all this sumptuous decoration offers a poignant contrast to the squalid exercise in sexual politics which is unfolding before us. It acts as a gilded prison for her. The starkly confronted horizontal and vertical lines of the bedding seem to fence her in and may be intended as a visual metaphor of her plight. Outside, and out of reach – as in the scene of Sa‘di and the girl (fig. 9) – is a paradise garden. Here, very clearly, is a story of two pairs, and this is true not only of the positioning of the two groups. Both the inner figures are seeking to impose their will on the outer ones, the prince assaulting the girl and the courtier restraining the negro with an admonitory gesture. Perhaps the negro is expostulating at being forced to remain outside. The club he carries may be seen as a sinister portent of the impending rape. Finally, the distribution of colour accents is so managed as to create a link and also a diametrical opposition between the girl and the negro, and only between these two. As is the case elsewhere in this manuscript, the colours in question are pale green and vermilion, and here too it seems at least possible that they carry a

37

Ibid., pp.123–124. The de´nouement is chilling. Acting on the advice of his vizier, who points out that “now the orange is soiled by having fallen in the mire, how can it again grace the hand of a king?” the monarch presents him with the negro, to whom in turn the vizier advises that he should give the damsel, “for it is meet that a dog should eat the leavings of a dog”.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 42D4EF6A /production/pindar/booksxml/hillen/chapter−10

338 message, as may the curtains in the selfsame colour, rolled up and down respectively. The lady wears a red outer garment and a green inner one; with the negro the colours are reversed, green outside, red inside. If red can be taken, naturally enough, to connote heat, and green to imply coolness, the relevance of these colours to the action would need no further comment. One might only add that his clothes, masking as they do most of the nature that lurks beneath, also smack of a literary device – litotes: extreme understatement for the sake of effect.38 Next, a deceptively simple painting illustrates a brief text describing how the poet, wandering in the noonday heat through the alleys of a town, chanced to encounter a maiden who appeared as if by magic from a doorway and offered him a much-needed drink of iced water (fig. 9).39 Sa cdi then intimates gracefully and in verse that the sight of her has afflicted him with a profounder thirst which no mere drink could assuage. The lady’s reply is not recorded, though no doubt she too quoted verse at him in abundance. This brief and ambiguous encounter is the point of departure for the artist, who builds on it his own commentary, full of wit and sly humour. In a telling reversal of the setting described in the text, with its reference to the poet sheltering against the heat in the shadow of a wall, he sets Sacdi against a pale green ground while the girl comes out of an orange-red building which seems almost to glow with heat. These are the two dominant colours of the painting and can perhaps be seen as appropriate equivalents for the two major themes of passion and rejection. But why are they so to speak reversed?40 Closer examination suggests that the painter is playing a kind of cat-and-mouse game in his choice of colours. The warm-hued building has a marble dado of cool violet and the wall above and around the maiden is spangled with pale green stars. The doorway in which she stands, silhouetted against a pure white ground, is framed in tilework of blue and green. She herself is clad, promisingly enough, in orange; but that is her outer garb only, and beneath it can be glimpsed a pale green shift. The iced beverage which

38

One can only admire the way the painter has invested a stereotypical composition with such dramatic intensity. 39 Ross, op.cit., p.222. The analysis which follows has already been published in substantially the same form: R. Hillenbrand, ‘The Uses of Space in Timurid Painting’, in Golombek and Subtelny, op.cit., p.95. Cf. Prentice, op.cit., pp.34–38 for an eloquent assessment of this image. 40 It is worth noting that this reversal goes against the grain of the text itself.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 41B9A525 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

339

she offers the poet is contained in a bowl of the same jade-green colour. His outer robe, in contrast, is the colour of renunciation, blue, and is echoed, as just noted, in the door-frame. The message of this subtle interplay of colours seems to be that hopes are raised only to be dashed. The body language of the two figures bears this out. The setting conjures up an ample space, yet this contains only two diminutive figures. In one sense it dwarfs them; but in another, it ensures that attention is focused on them. Their hands almost touch, but not quite. She stays just behind the mauve threshold, which thus seems to bar his entry. She proffers the drink but he has not yet received it. Thus gesture reinforces the message of colour. But it is the contrast between their surroundings that is most explicit: unbroken emptiness for him, but for her a blossoming garden against a golden sky. That garden is sensuously lush in comparison with the starkness outside. And only a tiny, enticing fragment of it, resplendent in gold and bright green, is visible. Even the shutters of her house are firmly closed, so to speak, against him. Thus in his own domain the artist goes decisively beyond his literary source, and does so by means of a visual language every whit as allusive as the text itself. The third picture which involves a development of the text on the painted page concerns the story of a youth with more brawn than brain who, having no money, tries to beg a passage on a boat (fig. 11).41 The boatman refuses and, for good measure, reviles him, but changes his mind when the youth offers him his garment in payment of his fare. Once aboard, the youth beats up both the boatman and his colleague. Worsted, they hypocritically profess amity and peace, and the voyage proceeds. Then they arrive at a pillar which juts out of the waves; here someone must climb up with a hawser so that the boat can be turned round. Sa cdi’s text continues, at the foot of the text panel, “The youth, in the pride of that courage which was uppermost in his thoughts, disregarded the rancour of an injured rival” – and here we must turn to the next page to learn that he duly shinned up the pillar with the cable. Immediately, says the text, “the boatman dragged it from his hand and pushed off the vessel. The deserted wretch stood aghast.” It is therefore this next page that is being illustrated, but as we read the text on the picture page we can see well enough where it will all end.

41

Ross, op.cit., pp.186–198, especially pp.191–193.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 44412FA2 /production/pindar/booksxml/hillen/chapter−10

340 Here again the painter makes colour, gesture and empty space do his work for him. The landscape behind the youth materialises in a bleak palette of white, faded blue, mauve and pale green. The trees are bare of leaves and silhouetted against an unframed sky. The tower, whose hard-edged lines emphasise the sheer drop to the ocean below and seem to seal his fate, is also pale green. This chromatic austerity, with its unmistakable emotional implications, lends extra poignancy to the wide range of bright cheery colours in the boat to which he is denied access.42 The message of isolation is equally driven home by placing and gesture. The ten figures in the boat are huddled together, enjoying the privileges of human company and warmth. They are plainly indifferent to his fate and their vigorous hand gestures show them talking animatedly to each other. Only three figures turn towards the marooned youth with triumphant or mocking gestures. In the wordless colloquy between the boatman and the abandoned young man, empty space is made to play its full part.43 It amplifies the huge gulf fixed between the fortunate and the unfortunate, across which the mute semaphore of the youth’s pleading outstretched arms sends its unavailing message. It would be possible to analyse each of the remaining five pictures in similar terms, but the law of diminishing returns would certainly operate. At this point, therefore, it may be useful to move from the specific to the general in an attempt to discover what principles governed the way the artist responded to his subject matter. Among those that can readily be identified, three stand out: the painter’s penchant for drama; his choice of misfortune as a motif; and his concern to make text panels play their full role in the finished picture. First, then, the painter’s taste for drama. Again and again the artist reveals the mainspring of the action, the moment that triggers the dénouement. Occasionally it is the dénouement itself. It seems natural to use the vocabulary of the theatre – however anachronistic that analogy may be – since so many of the scenes are conceived primarily in dramatic terms. Spectators are ranged tightly packed like an audience; sometimes they even 42 The sky, being potentially limitless, emphasises his plight, while his white drawers make him appear still more vulnerable. 43 The gestures of the figures facing the youth follow a hierarchy whereby the first gesticulates more energetically than the second, who in turn is more active than the third. See also Hillenbrand, ‘Uses of Space’, pp.92–95, and D’zul Haimi bin Muhammad Zain, Formal Values in Timurid Miniature Painting, Kuala Lumpur 1989, p.76.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 46CF8C77 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

341

applaud or jeer.44 The way the composition is organised, too, is often reminiscent of the theatre, with angled facades or diagonal files of onlookers operating like wings to nudge the attention towards centre stage, and a gorgeously decorated palace exterior as the backdrop.45 Empty space is used in the most calculated manner to isolate the protagonist and focus attention upon him. Indeed, deliberate contrasts of colour – for example, the way the key figure is made to stand out vividly against a monochrome background – function in much the same way as spotlights in the theatre.46 Gesture is manipulated to the same end.47 Frequently the artist contrives to load a gesture with meaning by almost literally stretching it out, as in the case of the youth stranded on the pillar. He thus marshals the empty area between the figures linked by the gesture. Dead space comes to life. The second leitmotif is no less obvious. When the subject matter of these paintings is critically examined, the major theme to emerge is quite plainly misfortune. That is certainly not the lasting impression left by a reading of the Gulistan as a whole. It is therefore an emphasis introduced by the artist – unless, indeed, he were following orders from above. The generally ironic and good-humoured tone which permeates the Gulistan leaves little trace in these paintings. Instead, scenes of a decidedly black cast have been selected for illustration and impart a downbeat, even pessimistic atmosphere to the book. This creation of a consistent mood at variance with that of the text itself is quite remarkable and no parallels for it suggest themselves. People are drowned, marooned, betrayed, sexually assaulted, attacked by dogs, disappointed, spurned and frustrated. The exceptions to this bleakness are the somewhat anodyne scene at the beginning, which depicts Sa‘di

44

As in the wrestling match, the man drowning, the distressed poet (whose dress employs the traditional iconography used for slaves and prisoners) and the dervish. Cf. the scene of the royal lover in Baisunghur’s Anthology of 830AH/1427AD in I Tatti, Florence, f.44a (Ettinghausen, Persian Miniatures, pl.III). 45 As in the encounter between Sa‘di and the maiden. Cf. the scene of Farhad brought before Shirin in the Nizami of ca.1405–1410 in the Freer Gallery of Art, Washington DC, no.31.34 (Gray, 1961, p.54). 46 Again the scene of Sa‘di and the maiden is the obvious example. The lack of interest in depicting shadow in this tradition of painting correspondingly inhibited experiments with shafts of light. 47 For example, the boatmen in f.29v.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 52338CF1 /production/pindar/booksxml/hillen/chapter−10

342 conversing with a youth in a garden (fig. 2), and the tableau of the wrestlers, where virtue narrowly triumphs over vice (fig. 10).48 It remains uncertain whether it was the similarity of theme which generated the repetitive compositions or whether, on the contrary, the artist was most at ease with a certain compositional formula which lent itself most readily to scenes depicting victims of misfortune. At all events, our painter left a series of his trademarks on such scenes, and he uses them constantly to put across his message.49 It may be useful to run through them quickly. All of them express the painter’s imaginative response to the story, his desire to develop and intensify the narrative structure.50 Thus, for example, in every case except that of Sa‘di and the girl (fig. 9), where privacy is the essence of the encounter, he adds extra people to the scene. They are not mentioned in the text. Their purpose is not mere ballast – they are too carefully placed for that – but to act as a mute commentary on the main theme.51 And they are not on the victim’s side. By this unobtrusive means, then, the painter quietly weights the scales against the disadvantaged, following in this, it seems, his own personal interpretation of the Gulistan. Life is not a bed of roses. In the same spirit Sa‘di favours the device of isolating, almost silhouetting, the victim against an unfriendly expanse of plain monochrome background, such as the sea or a barren landscape.52 The victim’s vulnerability is exposed 48

In the light of this, the generally acceptable dictum of Lentz (Painting at Herat, p.100) that “there is little room for individual expression” in the house style of the Baisunghuri atelier may perhaps be modified in the case of the Gulistan. Even in the garden scene, though, despite the festive notes struck by the landscape, the carpet and the servant bringing refreshment, the atmosphere of regret is palpable. Prentice argues that Sa‘di himself is the the real subject of these paintings (op.cit., p.42). 49 The problems alluded to earlier, which were consequent on illustrating a text which had not yet acquired a standard repertoire of images, must also be kept in mind here, since this situation might have predisposed the artist to recycle his compositions whenever he could do so. 50 In other words, he is never content baldly to tell the story in the way that Sa‘di does. There is always something extra. 51 This use of spectators as chorus is particularly marked in the scenes of the wrestlers and the drowning man. 52 This is not a new device in court painting of the early Timurid period; it can already be seen in such Jala’irid work as a combat scene in the Kulliyat of Khwaju Kirmani (see M. Gorelik, ‘Oriental Armour of the Near and Middle East from the eighth to the fifteenth Centuries as shown in Works of Art’, in Islamic Arms and Armour, R. Elgood [ed.], London 1979, p.48); in the British Library Anthology (Add.27261), as in the fire ordeal of Siyavush (f.295v: see B.W. Robinson, Persian Miniatures, New York n.d., pl.III); and in the same scene

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 49D3AAD2 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

343

by his pathetic nakedness – the poet, the dervish, the marooned youth – and further emphasised by his feeble physique (figs. 7, 13). Even the wrestlers have bodies unworthy of their calling, and here again their nakedness and their dun-coloured breeches reflect their lowly status. The unspoken message of these pictures is that it pays to be rich; woe to the poor and the unlucky. “Laugh, and the world laughs with you; weep, and you weep alone”. Nothing subversive here,53 for all that the Gulistan would have given ample warrant for it; and that is perhaps as it should be, seeing that the artist depended on a prince for his livelihood. The rich and powerful, secure in their luxurious palaces, look disdainfully down from their windows on the suppliants below, like Roman emperors idly contemplating events in the arena. The gulf between them is an index of the social as well as the spatial distance which separates them. Bright or warm colours are reserved for the privileged; cold greys and blues encompass the unfortunate. Body language plays its part too. In the scene of the drowning brothers (3), one has only to contrast the phlegmatic posture of the spectators in the boat, safe and snug, with the frantically waving arms of the man sinking below the surface. The dervish vainly trying to warm his naked torso, the poet fumbling desperately to dislodge a stone he can throw at the dogs slinking towards him (figs. 6, 13), are both watched with indifference or schadenfreude by courtiers from the shelter of the nearby building. Neither of them is offered any help. Indeed in the case of the poet, even nature conspires against him, for winter has frozen the stones to the ground, prompting him to rail in the text panel “What bastards these men are, for they untie their dogs and tie up their stones!” This witticism, forced out of him by sheer panic, proves his salvation because it amuses the bandit chieftain watching at the window. But the picture itself does not betray the outcome, and the stark juxtapositions in these two paintings typify this cycle of illustrations.

from a Shahnama in the Chester Beatty Library, Dublin, dated 800AH/1397–8AD, f.14v (R. Hillenbrand, Imperial Images in Persian Painting, Edinburgh 1977, colour pl. opposite p.53). Perhaps, though, one might argue that the painter(s) of the Gulistan imbued the theme with a poignancy not found in other contemporary manuscripts. The presumably slightly later painters of the Juki Shahnama (London, Royal Asiatic Society, MS 239) are constantly using this device, but they do not invest it with the emotional power found in the Dublin Gulistan, preferring dramatic effects (Lentz and Lowry, op.cit., pp.127 and 133–134; J.V.S. Wilkinson, The Shah-Namah of Firdausi. The Book of the Persian Kings, London 1931, pls.III, V and XV). 53 As there would have been if Bihzad had treated these themes.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 48FA968B /production/pindar/booksxml/hillen/chapter−10

344 The third constant of our artist’s approach is the way that he incorporates blocks of text into the picture space. The pervasive presence of these text boxes creates something akin to the speech bubble in a modern cartoon or comic – though the analogy underlines how crude a device the speech bubble is in comparison with the language of mime.54 The use of text panels in these pictures marks a new stage in the interplay between text and image in Persian painting. For almost a century – since the time of the so-called Demotte Shahnama, probably executed in the 1330s – artists had realised that their options were wider than merely varying the breadth of the text panel which was set above or below the picture.55 But while the more adventurous artists had indeed experimented with flags of text engulfed by the picture,56 or with paintings banished to the margins while text panels claimed pride of place,57 the idea of varying the size, the placing and the role of the text panel from one painting to the next in a single manuscript was still unfamiliar. And this is just what is done in the Gulistan. Equally striking, and of course intimately related to this phenomenon, is the way that each picture is set on the page differently from the next. Thus in the dimensions and shape of the painted surface and text panel alike, continuous variety is the rule, with a consequent enlivening effect on the entire manuscript. Nobody should underestimate the sheer professionalism of the artists who together created a fine manuscript of this period. The labours of the bookbinders, leather-workers, paper-cutters and paper-polishers, though integral to the finished product, are not strictly relevant in this context. But the scribe, painter and illuminator had to work in the closest concert if the grand design were not to miscarry. Their interdependence derived from the way the space on the page was parcelled up. The relative emphasis placed on

54

The comparison with a speech bubble is most relevant when the text panel is reduced to a few verses, as in the Freer Nizami (Gray, 1961, p.54; id., Arts of the Book, pls.XXXIII and p.68). 55 See O. Grabar and S. Blair, Epic Images and Contemporary History. The Illustrations of the Great Mongol Shahnama, Chicago and London 1980, pls.9, 17, 34, 35, 38 and 46. 56 The most spectacular examples of this practice occur in the Kulliyat of Khwaju Kirmani, for example the unmasking of Humayun by Humay (Gray, 1961, p.47; see the analysis by T. Fitzherbert, ‘Khwaju Kirmani (689–753/1290–1352): an éminence grise of fourteenth century Persian painting’, Iran XXIX (1991), pp.142–143. 57 Such as the Divan of Sultan Ahmad Jala’ir in the Freer Gallery, Washington D.C., datable to the first decade of the 15th century (Gray, 1961, p.49).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 494ABF35 /production/pindar/booksxml/hillen/chapter−10

THE MESSAGE OF MISFORTUNE

345

text, illustration and illumination varied dramatically from page to page, but could afford to do so only because policy decisions about the layout of every page were taken early on. The textbook case of this practice is furnished, as Teresa Fitzherbert has demonstrated, by the Khwaju Kirmani manuscript of 1396 in the British Library, produced a generation before the Gulistan under discussion here.58 Here a poem many thousand baits (couplets) in length has been transcribed with such total control of pace that it finishes on the dot of the last line of the last page of the manuscript. Yet in the process it has managed to integrate into the rhythm of that text – whose number of baits per full page of text remains absolutely even throughout the book – a set of paintings notably uneven both in size and in the amount of text which they incorporate. All this argues the most careful planning in advance. Not only the presence of a text panel in a picture, but the very content of that same panel, had to be fixed before the painter seriously began work. The many examples of unfinished paintings whose text panels remain empty should not be misconstrued. The text that was eventually intended to fill such panels was known in advance, and we have seen that it frequently dictated the precise content of the painting. Even a full-page illustration could only be inserted at a later stage if it fulfilled two conditions; that its verso also contained a full-page painting, and that neither painting incorporated any of the text to be illustrated.59 These remarks should make it clear that the underlying concept of an illustrated manuscript, at least in the academy founded and funded by Baisunghur, was emphatically not that of a self-contained text to which an elastic number of illustrations could be added later and at will. Instead, text, illustration and even illumination were intended to form a seamless whole. Ornament, too, is not mere decorative infill because artist or calligrapher have left some visually irksome gaps on the page. It is placed with a finely tuned sense of interval; in musical terms, one might speak of absolute pitch. 58

See Fitzherbert, op.cit., p.143, for a telling example of how and why the text was tampered with by adding extra verses. 59 In point of fact there are no full-page paintings in the Gulistan, and indeed artists of this period seem to have gone to great lengths to make a picture ‘technically’ contain text even though it was effectively a full-page illustration. As late as 1556, the paintings of the Haft Awrang were largely of this kind (M. S. Simpson, Sultan Ibrahim Mirza’s Haft Awrang [New Haven and London, 1997]). The obvious exception to this principle is the frontispiece, which in Timurid times was often extended to a double-page spread.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 32 OUTPUT: Tue Aug 10 10:29:19 2010 SUM: 49F9F38A /production/pindar/booksxml/hillen/chapter−10

346 A destructive domino effect would inevitably result if a painted or illuminated surface took up too much room on a given page, for the finely wrought balance between text, painting and illumination on subsequent pages would then be displaced. There is only one right place for a given bait. In other manuscripts of earlier or later times, or of lesser quality, this is patently not so, because the calligraphy is not of uniform size. But once the decision has been taken to execute a manuscript in a uniform size of script, very precise planning is imperative.60 Even an untrained eye can detect the signs of a calligrapher cramming words too closely together because he was running out of space, or conversely expanding the space between words and letters because he had too much of it. What is the precise role of the painter in this situation? There is no certainty about whether he agreed with the patron or the calligrapher on the exact number of pictures for the manuscript. But he must certainly have maintained close liaison with the calligrapher as to the exact shape of the boxes that would hold the paintings; and, in his turn, the calligrapher would have had to know what part of the text would need to be accommodated on such pages. And of course, the less text that appeared on the page, the more constricted did the options of the painter become – not least because that text took on proportionately greater significance as it became less and less. Most challenging of all were the cases where the flag of text was no more than a bait or two. Then the painter was on his mettle to ensure that, while his illustration accurately captured the spirit of the story as told in the adjoining pages, it also carried a special charge of relevance linked to the panel of text in the painting itself. In the case of the Gulistan’s wrestling scene, for example, the painter had ample scope since the text panel contained virtually the entire story. As it happens, this Gulistan – unlike, say, the Khwaju Kirmani manuscript in the British Library – has no case where the textual presence in the picture has been whittled down to a pair of baits, and so there is no opportunity for a final flourish of virtuosity, as there is when a painting is so constructed that it turns on the significance of a few words.

60 In some cases it is plain that the exact length of the text, down to the number of lines, was calculated, as were the dimensions of each picture. The Dublin Gulistan would not have been able to exploit its most teasing devices of suspense without such careful calculation. The text block, too, maintains consistent dimensions, and if these are transgressed, it is only by pictures. See B. Brend, ‘Beyond the pale: meaning in the margin’, in Studies in Persian Painting. Festschrift for B.W. Robinson, in press.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10 Sequential 20

JOBNAME: ???? PAGE: 1 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 10DF2FB6 /production/pindar/booksxml/hillen/chapter−10−artwork

THE MESSAGE OF MISFORTUNE

347

1. Two courtiers (detail of 5).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 088E919C /production/pindar/booksxml/hillen/chapter−10−artwork

348

2. The poet Sa cdi converses by night with a young friend in a garden, from The Gulistan of Sa cdi, Herat, Persia, 830AH (1427AD), Chester Beatty Library, Dublin, inv.no. P.119, f.3r.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 065EFCCA /production/pindar/booksxml/hillen/chapter−10−artwork

THE MESSAGE OF MISFORTUNE

349

3. One brother is rescued, the other drowns, Sacdi, Gulistan, f.15v.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 0A157E98 /production/pindar/booksxml/hillen/chapter−10−artwork

5. The prince and the dervish, Sacdi, Gulistan, f.9r.

4. A drunken prince assaults a Chinese maiden (detail of 8).

350

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 099569B0 /production/pindar/booksxml/hillen/chapter−10−artwork

351

6. A distressed poet is attacked by dogs, Sacdi, Gulistan, f.32v.

7. A dervish begs for charity (detail of 5).

THE MESSAGE OF MISFORTUNE

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 05185515 /production/pindar/booksxml/hillen/chapter−10−artwork

352

8. A drunken prince assaults a Chinese maiden, Sacdi, Gulistan, f.16r.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 06FAED64 /production/pindar/booksxml/hillen/chapter−10−artwork

THE MESSAGE OF MISFORTUNE

353

9. A chance encounter between a maiden and the young Sacdi, Gulistan, f.36r.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 04E6B12E /production/pindar/booksxml/hillen/chapter−10−artwork

354

10. The wrestling match, Sacdi, Gulistan, f.14r.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 0702F7C4 /production/pindar/booksxml/hillen/chapter−10−artwork

THE MESSAGE OF MISFORTUNE

355

11. A youth is marooned by an angry boatman, Sacdi, Gulistan, f.29v.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 30 OUTPUT: Tue Aug 10 10:29:20 2010 SUM: 09B491F6 /production/pindar/booksxml/hillen/chapter−10−artwork

12. Illuminated text page, Sacdi, Gulistan.

13. The poet is threatened by dogs (detail of 6).

356

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-artwork Sequential 10

JOBNAME: ???? PAGE: 1 SESS: 31 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 525D3009 /production/pindar/booksxml/hillen/chapter−10−ends

THE MESSAGE OF MISFORTUNE

357

As it happens, there are three cases in the Dublin Gulistan where the text panel adjoining the painting, and on the same page, contains a portion of the previous story. Was this a matter of indifference to the makers and reader of the book? Or was it tolerated as just a minor blemish? Or was it, on the contrary, a piece of deliberate planning? The latter seems the most likely explanation, for in all three cases – ff.9r,61 14r,62 and 16r63 (5, 10, 8) – the presence of the immediately preceding episode (which is of course potentially intrusive and distracting) was intended to drive home the moral which the painting expressed. Overlap, then – or, for that matter, its absence64 – was part of the overall plan.65 When the Dublin Gulistan is compared with the great Persian manuscripts of earlier or later periods, it can be seen to lack their range of expression, as exemplified, say, in the Demotte Shahnama or in the British Library Nizami of

61

In the tale of the wrestlers, the page begins with the quotation which ends the previous tale, namely the verses inscribed on the crown of Kaikhusrau. These stress that no one can hold on to power for ever. And the story which they round off features a king whose dishonest dealings prove his downfall. That is the immediate context for the account of the wrestler who passed on all but one fragment of his professional knowledge to his younger colleague and pupil, and used that against the youth when the latter treacherously challenged him to public combat. 62 In the tale of the amorous prince, the Chinese maiden and the negro, most of the text written on the painted page comprises verses from the previous story; their burden is that fortune bestows wealth upon the ignorant, the foolish and the illiterate. These verses are quoted in the context of the ignorant negro whom Harun al-Rashid appointed to rule over Egypt, and who suggested, in a manner worthy of George III, that wool should be planted instead of cotton so that the Nile should not be able to sweep it away. The following tale is about an unworthy negro on whom Fortune smiles. 63 The story of the king and the beggar is accompanied by a painting whose text panel contains all of the previous tale and the closing verses of the one before that. All three stories are linked by the theme of a ruler who behaves badly. 64 In the story of the wrestlers, the connection with the previous tale, about the king whose dishonesty brings him to grief, is patent – for here too the dishonest apprentice fails of his purpose and is publicly shamed. 65 What do these examples of overlap suggest? At the very least it may be said that in none of them does the text of the earlier stories weaken or obscure the episode that is actually being illustrated. In the cases of the Chinese maiden and of the dervish, a good deal of text from the previous story is included on the painted page. And here the connection is crystal clear. The incorporation of ostensibly irrelevant text can thus be seen to play its part after all.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 31 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4A220A4D /production/pindar/booksxml/hillen/chapter−10−ends

358 1494.66 Instead, it adopts a deliberately restricted compass, reworking a single basic idea in several different guises. Hence the repetition of a single figure type – thin, bloodless, narrow-waisted – or of similar architectural facades. Two illustrations – the poet in the garden and the amorous prince – are evidently lifted or adapted from earlier manuscripts. In one sense, then, this manuscript represents a step backwards from the confident full-page pictures of the Khwaju Kirmani manuscript, for even its apparently full-page illustrations prove on closer inspection to be nothing of the kind and they contain significantly less action, colour, detail. Yet to dismiss it in these terms would be a serious error. Its true context is in the activity of the entire Baisunghuri atelier over a period of some seven years between 1426 and 1433, the year of his death. Baisunghur clearly ordered copies of many of the major literary masterpieces of Persian literature, and they emphatically do not conform to a single house style. Indeed, even successive versions of the same text, as in the case of the Kalila wa Dimna and probably the Shahnama, attest significant differences in style.67 This, then, was not merely a period of experiment in the normal sense, but rather a time when the appropriate style of presentation for a given text, including its page layout, calligraphy, painting and illumination, with the degree of emphasis to be allotted to each of these elements, was a matter for careful thought. It seems likely that different teams were each allotted a manuscript and encouraged to devise a tailor-made solution for the particular problems it posed. It was inevitable that some solutions should be perceived as dead ends – perhaps the Chahar Maqaleh was one of these – and others as only partial successes.68 But their felicities would nevertheless have something to offer towards the formation of the celebrated classical style of Persian painting, traditionally associated with Baisunghur.69 And in that category the Dublin Gulistan deserves a place of honour.

66

For the latter manuscript, see I. Stchoukine, Les peintures des manuscrits timuˆrides, Paris 1954, pp.78–81; Gray, 1961, pp.113–118. 67 See E.J. Grube, ‘Two Kalilah wa Dimnah codices made for Baysunghur Mirza: the concept of the “classical style” reconsidered’, Atti del III Convegno Internazionale sull’Arte e sulla Civilità Islamica. ‘Problemi dell’età timuride’, Venice 1980, pp.115–122 plus Appendix. 68 See E.G. Sims, ‘Prince Baysunghur’s Chahar Maqaleh’, Sanat Tarihi Yıllıgˇı VI, 1974–5, pp.375–409. 69 E.J. Grube, The Classical Style in Islamic Painting, n.p. 1968.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-10-ends Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 32CD828A /production/pindar/booksxml/hillen/chapter−11

XI Literature and the visual arts

A

RABIC secular literature plays only a minor role in Islamic art, although of course Koranic inscriptions in their thousands permeate all media – architecture, glassware, metalwork, ceramics and so on – and sanctify even the most mundane object. With the exception of book painting, which saw the principal expression of Arabic literature in visual terms because the pictures illustrate a specific text, the impact of Arabic literature on Islamic art was in general restricted to a few lines of poetry quoted on buildings, on metalware and on ivories, while Mamlu¯k textiles and Sa¯ma¯nid ceramics also bear rhyming proverbs. Only rarely is this poetry of literary value, as in the case of the multiple quotations from the works of the Grenadine poet Ibn Zamrak (and possibly Ibn Jayya¯b) at the Alhambra. In these inscriptions the building or the object often apostrophizes itself. A Córdoban ivory likens itself in ramal metre to a maiden’s breast, while other Spanish ivories bear love-poems or vaunt their value as safe deposits. The Marinid madrasa of Sale in Morocco bears verses praising its door; the niche for a Nasrid vase compliments itself (presumably the vase is intended) on its beauty and splendour; the dome of the Hall of the Two Sisters in the Alhambra describes itself in terms of the Pleiades and the constellation of Gemini; a Mamlu¯k brass bowl states, ‘My colour and workmanship are beautiful and flattering’, and extols the excellence of its construction; while the fountain in the Court of the Lions in the Alhambra bears a long poem replete with references to royal victory and to Solomon. In other contexts the verses are of Shı# c#ı content or are benedictory, as in the case of a Mamlu¯k brass lunch box:

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 41B618F2 /production/pindar/booksxml/hillen/chapter−11

360 On the owner be happiness and peace, And may his life last as long as the dove coos; May he receive of the outpouring of the spiritual world; The owner is Ah.mad ibn Muh.ammad ibn Idrı#s. Several pieces of Mamluk metalwork bear inscriptions with all or part of the following poetic text: ‘You have reached the highest rank as regards greatness, and good fortune has associated with you on every side; may you not cease to be in demand and to stretch forth your right hand in the world by obtaining your wishes.’ And a Zangid lid for a cylindrical box bears a corrupt text of two verses from the dı#wa¯n of the Ja¯hiliyya poet al-Na¯bigha al-Dhubya¯nı#, written for the king of H.¯ıra, al-Nu cma¯n ibn Mundhir: Don’t you see that God has granted you a degree of power which makes all the kings (grovel at your feet)? For you are a sun; the kings, stars. When the sun rises no star will be seen. Arabic poetry also became a favoured type of decoration in the private houses of the wealthy in the Levant during the eleventh/seventeenth and twelfth/eighteenth centuries, and is also found on later Islamic woodwork. Curiously enough, some of the most popular classics of medieval Arabic literature, such as the Sı#rat ‘Antar and The Thousand and One Nights (Alf layla wa-layla), were never chosen by artists for book illustration, although some of the Sı#rat ‘Antar images have found a ready market in modern folk art, for example in underglass paintings from Tunisia. Arabic poetry, too, conspicuously unlike Persian poetry, did not develop a tradition of cycles of illustration. Nor, for that matter, did the vast majority of adab literature. One must conclude that there was no expectation on the part of writers of poetry or belles-lettres that their work would be illustrated. A rare exception is the Ja¯mi c al-tawa¯rı#kh of Rashı#d al-Dı¯n, composed c.700/1300 and copied at the behest (and expense) of the author twice each year for transmission to the major cities of the Ilkhanid realm; contemporary fragments of this text survive in Edinburgh and London. Texts of a scientific tenor, on the other hand, offered numerous popular subjects for painters of the seventh/thirteenth and eighth/fourteenth centuries, the period when most Arab book painting of quality was produced. Particularly favoured texts ranged over such topics as astronomy

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 51DC9847 /production/pindar/booksxml/hillen/chapter−11

LITERATURE AND THE VISUAL ARTS

361

(al-S.u¯fı#’s Kitab suwar al-kawa¯kib al-tha¯bita), medicine and pharmacology (the Arabic translation of the De Materia Medica of Dioscorides), automata (works by al-Jazarı#), bestiaries (by al-Ja¯h.iz. and Ibn al-Durayhim), the cosmology of Zakariyya¯’ al-Qazwı#nı# (Kita¯b ‘Aja¯’ib al-makhlu¯qa¯t), handbooks of military exercises (al-Aqsara¯ c#ı’s Niha¯yat al-su’l) and farriery (Ibn Akhı# H.iza¯m’s Kita¯b al-Bayt. a¯ra and a book of the same name by Ah.mad ibn al-H.usayn ibn al-Ah.naf). Numerous other works known in only one or occasionally two illustrated copies fall into this same ‘scientific’ category – treatises on snakebite, medicine, calendrical systems, zoology and cosmology. In such texts, whether they were popular with painters or not, the illustrations had a specific purpose, namely to explain or amplify what the text said. They were not intended to serve as decoration. In this respect the painters of these manuscripts, some of whom may have been Christians, followed Byzantine precedent, in which the illustrations of secular manuscripts were subordinate to the text, and were indeed often reduced to scrawled explanatory diagrams. Adab literature offered far fewer texts that were deemed suitable for illustration, and it remains puzzling why some adab texts were illustrated and others not. Here, too, several unica are known. One is the H.adı#th Baya¯d. wa-Riya¯d., a romance of courtly love originating in the Muslim West, whose paintings evoke the Andalusian cultural milieu most vividly (possibly Ma¯likı# influence may explain why the tally of medieval illustrated manuscripts from the Maghrib and Muslim Spain is tiny in comparison with that from the central Islamic lands). Others are al-Mubashshir ibn Fa¯tik’s Mukhta¯r al-h.ikam, an anthology of wisdom literature with appropriate images depicting the sages of times past, and Ibn Z.afar’s Sulwa¯n al-mut. a¯ c, a collection of animal fables. But two texts in this category stand out for their popularity with illustrators, a factor that resulted in the evolution of sophisticated pictorial cycles. They are the animal fables known under the title Kalı#la wa-Dimna, and the Assemblies (itself a disputed translation) or Maqa¯ma¯t of al-H.arı#rı#. Over a score of illustrated manuscripts of these two texts are known, and they take us to the heart of the painter’s response to Arabic literature. That response, it has to be said, is not very profound; wherever possible, painters tend to seize on the obvious, and when the text is too complicated verbally they fix on peripheral details from it or draw directly on their own imaginations, sometimes to powerful effect. Thus the painting may be a commentary on the text rather than a direct illustration of it.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 499D1AA3 /production/pindar/booksxml/hillen/chapter−11

362 Kalı#la wa-Dimna is in origin an Indian, not an Arabic, text whose great popularity ensured its translation into most of the major languages of the medieval world. The text is also noteworthy for its function as a Mirror for Princes, which gave it a political and moral function, despite its notional purpose of ‘straight’ entertainment. Nevertheless, the painters usually focused on animal rather than human subjects, even though the text offered ample scope for the latter. Presumably the lively, if not narrowly naturalistic, depiction of a wide variety of creatures constituted a major attraction of these manuscripts. Bold colours, lavish use of the silhouette mode and of the frontal plane only, and simplified conventional notations for landscape elements – all of them features that facilitated the legibility of the image – contribute to the ready appeal of these Kalı#la wa-Dimna manuscripts, produced mostly in the seventh/thirteenth and eighth/fourteenth centuries in Egypt, Syria and Iraq. The simple narrative framework of the text, with its succession of brief independent tales, was ideally suited for illustration. The main protagonists are often identified by inscriptions, although it is not always clear whether these are contemporary. Once the iconography for a given scene had been established – a process that usually involved pruning the image of extraneous detail – it tended to become fixed and was copied with minimal alteration by subseqent painters. Thus it seems that generations of painters were not each challenged afresh by the text to produce their own personal interpretation of a scene, and this strong undertow of conservatism also explains why the same scenes were copied from one century to the next, even though the text offered hundreds of opportunities for new images. Happily, the most popular book for illustration in all of Arabic literature was itself quintessentially Arab. This was the Maqa¯ma¯t of al-H.arı#rı#, which within a century of the author’s death in 516/1122 was being transcribed in multiple illustrated copies, first in Syria and Iraq and then in Egypt. The plethora of illustrated versions should not, however, be attributed to the abilities of the painters, but to the runaway popularity of the text itself, a fashion that began in the lifetime of the author and lasted for centuries. Nevertheless, there is a major paradox here. The attraction of the text centred on its form rather than on its content, and specifically on its verbal pyrotechnics, for each tale or ‘assembly’ is the merest peg on which to hang grammatical disquisitions of stultifying length and complexity, puns, outlandish vocabulary and other features of narrowly linguistic and not generally literary interest. By and large, these are beyond the powers of the painters to illustrate. Accordingly, artists approached the text from another angle

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 3ACAC649 /production/pindar/booksxml/hillen/chapter−11

LITERATURE AND THE VISUAL ARTS

363

altogether, and used its minimal references to a geographical location or a social milieu as a foundation for the close (and frequently satirical) observation of contemporary society. The fact that the Maqa¯ma¯t stories revolve around a confidence trickster, Abu¯ Zayd, provides the necessary justification for this approach. As a result, the best of the illustrated Maqa¯ma¯t manuscripts, like those of Paris and St Petersburg, hold up a mirror to high and low life alike, to the world of the palace and the caravansarai, the judge’s court and the tavern. These manuscripts are thus an unrivalled source for the minutiae of daily life in the seventh/thirteenth and eighth/fourteenth centuries, from the daily grind of village life to the great public processions and parades, with the fabled Eastern Isles adding a touch of exotic fantasy. Despite the survival of a dozen medieval illustrated versions, no standard iconography was ever evolved for this text, unlike the case of Kalı#la wa-Dimna; the reasons for this remain obscure, but these internal differences point to the existence of numerous independent ateliers probably working for the market rather than for specific patrons. Possession of a copy of the Maqa¯ma¯t may well have been desirable as an indication of the owner’s high level of education and culture – a matter of snob appeal – but in the case of illustrated copies one may further suggest that their main attraction was visual rather than intellectual. Their patrons may have been incapable of appreciating the finer points of the text; but when it came to pictures, they knew what they liked.

Further reading Allan, J.W., ‘Later Mamluk metalwork, II. A series of lunch-boxes’, Oriental Art 17(2) (1971), 1–9. Atıl, E., Renaissance of Islam: Art of the Mamluks, Washington DC (1981). Ettinghausen, R., Arab Painting, Geneva (1962). Ferrandis, J., Marfiles Arabes de Occidente, Madrid (1935–40). Ghouchani, A., Inscriptions on Nishabur Pottery, Tehran (1986).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 42 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 0FDAA4E1 /production/pindar/booksxml/hillen/chapter−11

364 Grabar, O., The Alhambra, London (1978). _____, The Illustrations of the Maqamat, Chicago and London (1984). Grube, E.J. (ed)., A Mirror for Princes from India: Illustrated Versions of the Kalilah wa Dimnah, Anvar-i Suhayli, lyar-i Danish, and Humayun Nameh, Bombay (1991). Haldane, D., Mamluk Painting, Warminster (1978). Rice, D.S., ‘The brasses of Badr al-Din Lu’lu’’, BSOAS 13(3) (1950), 627–34. _____,‘Two unusual Mamlu¯k metal works’, BSOAS 20 (1957), 487–500.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-11 Sequential 6

JOBNAME: ???? PAGE: 1 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 387548A8 /production/pindar/booksxml/hillen/chapter−12

XII New Perspectives in Shahnama Iconography

T

HE papers published in this volume admirably illustrate the range of possible art-historical approaches to the Shahnama and its iconography in the current state of the field and in the wide range of media in which Shahnama themes appear. It will be both convenient and appropriate to begin with contributions on book painting. An excellent way to orientate oneself is the overview of the field provided by Marianna Shreve Simpson, the doyenne of Shahnama studies in the field of art history. Drawing on the experience of decades of highly focused research on Shahnama images, she is an assured and reliable guide not only to art-historical publications on the topic but also to the other disciplines involved, which for the art historian are ancillary but which are of course central to Shahnama studies in general. These include textual criticism and the problems of editing the text, literary criticism, and the complex historical dimensions of the text. It is worth noting that she has deliberately restricted her horizon to the last quarter-century or so, and the extensive documentation – itself a selection from a much larger body of published material – which she has assembled for this period alone amply justifies her decision. This is not to deny, of course, as she herself acknowledges, that much important work on Shahnama images was published in previous years, but there can be no doubt that scholarship in this field has taken a quantum leap forward in the period which she covers. It is especially useful to have her assessment of the range of approaches that have been developed over this time for the better understanding of the meaning and function of illustrated Shahnamas in medieval Iran. It is worth considering briefly at this stage the question of why illustrated Shahnamas suddenly became popular around 1300. It could perhaps have been a belated reaction to the destructive impact of an invasion even more

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 43D171C4 /production/pindar/booksxml/hillen/chapter−12

366 alien than that of the Arabs – that of the Mongols – and could have been intended to celebrate the revival of Persian culture from that trauma. Yet another possibility is that the Shahnama simply benefited from a more general and intense fashion, inaugurated by Ghazan Khan and continued by his successors, for books, among which illustrated books were especially popular. This can be seen as an aspect of a more general revival of the arts after the conversion of Ghazan in 1295, a revival which marked not only the definitive recovery of Persian culture, but also a new-found Mongol commitment to that culture. The element of take-over discernible in such an attitude has been interpreted by Soudavar as the principal motive behind the Great Mongol Shahnama. That commitment found expression in renewed productivity in architecture, metalwork and textiles among other art forms. The development of manuscript painting in particular, which was so marked that this is the only art, alongside that of the luxury Qur’an – so far as one can tell from surviving evidence – which could rival architecture, might also have had something to do with the exposure of artists to paintings from Europe and from the Far East, and with the role of the Shahnama as an instrument of teaching. It could also have provided entertainment for the largely illiterate Mongol elite. So much for preliminaries. It is now time to turn to the papers devoted to specific manuscripts. Jonathan Bloom and Sheila Blair take up a theme on which both of them, but especially the latter, have already published important and far-reaching studies that go well beyond the traditional identification of provenance, patron and date. The manuscript on which they focus is, moreover, one of the supreme masterpieces of Persian painting, the “Demotte” or “Great Mongol” Shahnama. It has been one of their particular achievements to shed new light on the afterlife of this complex masterpiece, whose meaning has proved so elusive and continues to provoke contention. Their papers here, which should be read together, approach from quite different angles the issue of its fate in the nineteenth century at the hands of Qajar restorers. Jonathan Bloom tackles the questions of how and why. He looks critically at the kind of paper employed for the remargining of the leaves and demonstrates not only its Russian provenance but also a date of no earlier than 1832. He then links this restoration to a general cultural revival which transformed the literary scene in Qajar times. To modern tastes, some of the Qajar alterations – especially the repainting in heavy, matt colours and thick outlines – were intrusive; but they demonstrate the continuing fascination which this manuscript exerted

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 48D8326D /production/pindar/booksxml/hillen/chapter−12

NEW PERSPECTIVES IN SHAHNAMA ICONOGRAPHY

367

on later generations. Sheila Blair homes in on a topic which only she and Abolala Soudavar have treated in any detail – the penmanship of the Great Mongol Shahnama. In an impressive display of forensic skill, she distinguishes no less than four separate hands, and proposes a plausible dating for each in turn, from the early fourteenth to the early twentieth century. She thus settles once and for all the vexed question of this manuscript’s calligraphy. In so doing, she lays bare the difficulties which faced later scribes who attempted to mimic the calligraphy of past masters. Elaine Wright’s paper, another triumph of patient yet imaginative detective work, uses somewhat similar investigative techniques to unravel the equally complex afterlife of another celebrated masterpiece, the Epics now divided between the British Library and the Chester Beatty Library in Dublin. The content of these volumes, essentially an anthology of selections of Persian poetry in the epic mode of the Shahnama and written in imitation of it, speaks for the sophisticated literary tastes of some Timurid patrons, and illustrates the genre of illustrated poetic anthologies which enjoyed a fashionable cachet at this particular time. The major conclusion of this paper, based on very close scrutiny of the two now sundered manuscripts, is that these texts were produced to one specification for a patron in 1397 and then, within less than two decades, significantly retooled to quite another specification for a second patron. Yet these changes, involving different paper, a different calligraphic hand, new illustrations and a revised layout to accommodate these major changes without destroying the natural sequence of the text or the aesthetic integrity of the manuscript as a whole, were made with the art that conceals art. They were not meant to be noticed, and the fact that none of the earlier scholars who had dealt with the Epics had spotted them is a tribute to the virtuoso skills of the artists who created these books. Tim Stanley’s paper also tracks the vicissitudes undergone by a manuscript after its completion, this time over a period of centuries and over an equally wide geographical space. His discussion centres partly on the reconstitution of the manuscript to its original state with its full complement of pictures, and partly on the evidence of inscriptions left by subsequent owners. These detail its fortunes in Mughal India and the cash value that it had there in the seventeenth century. But his paper also includes illuminating discussions of the break line and its function, of visual puns (such as the beak of the Simurgh being placed just by the text “thus spake the Simurgh”, so that Firdausi’s text becomes a kind of banderole or speech bubble), and of how this Shirazi manuscript modifies the canonical layout proposed by Grace Guest for that school of painting.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 463FCCCD /production/pindar/booksxml/hillen/chapter−12

368 Mention of the break line – the line or lines of text most relevant to the illustration, which follows directly after that passage – raises more general questions which deserve a brief airing here. In particular, the fact that a given episode may have different break lines from one manuscript to the next (a matter which still awaits detailed discussion in print) prompts questions about the consistency of the text itself. In fact, that text was in a state of constant flux. To this day there is lively debate about the authenticity of any number of verses ascribed to Firdausi, for subsequent poets freely but furtively interpolated their own pastiches as Firdausi’s own work. Moreover, the text dictated to a given scribe – for Islamic practice favoured dictation over direct copying, though such copying was not rare – would reflect the personal preferences of the reciter as well as the accuracy either of his memory or of the copy he was using, or both. Thus given episodes of the poem might be extended or contracted at will, or even – though more rarely –omitted altogether. It is for these reasons that no two medieval Shahnama manuscripts have the same text. What follows from all this? And what was the modus operandi of scribe and painter? More than one scenario presents itself. The practice of timing the illustration so that it followed directly after the break line(s), its trigger, suggests the possibility that, in view of the not entirely predictable sequence of the text from one reciter or copy to the next, a space might have had to be made for the painting, and not in advance, at a specific place on the still blank surface of the paper, but only at the very moment that the reciter dictated the break line. Yet any lengthy break in the dictation, with the consequent loss of rhythm and concentration, was clearly undesirable – particularly if other scribes were taking down the text at the same time for unillustrated versions, which would have been a sensible way of deriving maximum advantage from the mammoth and expensive task of dictating the entire Shahnama. The commercial element in the production of many Shahnamas should not be forgotten, for it obviously encouraged economies of scale wherever they could be managed. How then to minimise the loss of momentum? It would have been too much to expect the scribe to be able within seconds to devise an appropriate size and format for the painting, and put it in the correct place on the page while simultaneously keeping in mind the need to vary picture size and format throughout the whole book. A possible (though no doubt not the only possible) solution to this impasse is that before the dictation of the book began, the artist(s) worked out in general terms the subjects for

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4228B63C /production/pindar/booksxml/hillen/chapter−12

NEW PERSPECTIVES IN SHAHNAMA ICONOGRAPHY

369

illustration, including the precise dimensions of each painting. The use of wire moulds whose impress mapped out the shape of a text-only page, including interstitial columns as well as the desired number of lines, provided a serviceable grid which would have made it easy to plan the painting in modular fashion and to ensure that no picture transgressed the subdivisions of that grid. Thus no painting results in the standard space allocated for a hemistich being trimmed, while the top and bottom of the picture always correspond to the upper or lower line of the text panels. Thus once the co-ordinates of each painting had been determined – without the need for any actual painting to have been done – and passed on to the scribe in the order that the paintings would follow (whether in the form of numbered co-ordinates or a series of pages lightly marked to indicate the exact shape of each painting, and suitably identified by title) the scribe merely had to remain alert to the trigger of a suitable breakline. That breakline, as already noted, might vary for the identical scene from one manuscript to the next. Thereupon he would rapidly mark the boundaries of the painting and then continue writing. The same principle would be followed for the ‘unwans, whether these were titles of books, sub-headings or captions. This system would also ensure that, at a glance, the artists could determine the successive outlines of a sequence of paintings, and thus secure an agreeable variation in this respect. Moreover, since Firdausi’s descriptions of a given event tend to be lengthy and indeed to provide a number of possible breaklines, it would in most cases not have been difficult for the scribe to make a rapid decision about the optimum placing for a given painting, depending on how far down the page the copying had progressed. The system was therefore flexible enough to accommodate quite significant variations in the text, and in the large majority of cases it would work smoothly. Conversely, in the case of those relatively few paintings for which only one breakline would do and where, in the course of dictation, that breakline appeared too far down the page for the desired picture to fit in – a rare case of double jeopardy – the obvious solution would have been to lay that page aside for the moment and later to recopy it, tweaking the text slightly by omitting a verse or two. But at that moment of “bad fit” during the dictation itself, the scribe would turn to begin a new verso or recto side as the case might be. Such a method of planning the illustrations as a whole also helps to explain the differences between the various cycles.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 44BCC8DD /production/pindar/booksxml/hillen/chapter−12

370 A refinement of this scenario – and here I must acknowledge the contribution of Dr Farhad Mehran in an illuminating discussion during the Edinburgh conference – might lay rather more stress on the element of long-term and short-term planning in a project as ambitious and time-consuming as producing an illustrated Shahnama. The first priority would be to determine the desired level of production values – was it to be a relatively modest-looking volume or a sumptuous one? Obviously this would involve specifying the dimensions of the finished volume, the size of the text block, the number and scale of its constituent paintings and the amount of illumination to be included. This would be an essentially financial decision. Next, a programme of illustrations (and perhaps also of illumination) would be drawn up so as to ensure a balance of these elements that corresponded to the vision of the person responsible for the book, or of the patron. Such a programme would certainly allocate images to specific narrative sequences of the Shahnama, but might leave until later the detail of how to interpret the chosen episodes, and what their exact content should be. Once these policy decisions had been made, it would become a matter of fixing the daily rate of work. On any given day the (say) 300 baits to be written down might generate perhaps a single picture (assuming a Shahnama with about 100 illustrations, a common pattern in the 14th centrury). At that stage – in other words, immediately before the scribe embarked on copying the relevant portion of the text – the nature of the scene to be illustrated and the painter’s interpretation of it would together determine its format. It could also be that the format for the individual paintings (though probably not their exact content) was fixed at the time that the overall programme of illustrations was decided. But this is a secondary issue. The main point is that the scribe would know the required dimensions of the day’s picture from the outset of the day’s work. For his part, the painter would be working on pictures located earlier in the text; the scribe would always be ahead of him. Moreover, in most pre-modern illustrated Shahnamas the number of paintings was too small to affect the scribe’s rate of work – all that was needed was for the scribe to mark the previously agreed parameters of each illustration when he arrived at the relevant point in the text. Thus to make room for the illustrations as dictation or direct copying proceeded apace (the constant minor variations from one copy of the text to another points to dictation rather than transcription) would pose no special problems for the scribe and would not impact significantly on his momentum. The dawning awareness that a break-line verse was imminent

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 478467B5 /production/pindar/booksxml/hillen/chapter−12

NEW PERSPECTIVES IN SHAHNAMA ICONOGRAPHY

371

might lead the person dictating the text to warn the scribe to move from the 4- or 6-column format to a 2- or 4- column format to accommodate the painting. Minor “nicking” or “shaving” of the painting’s format (e.g. by raising or lowering one step or more in a stepped format) would tailor the painting to the space available and would reflect an instant response to the coming break-line. In the Small Shahnama intensively studied by Dr Mehran the break-line verse is found substantially after the painting only at the very beginning of the manuscript – that is, at the one place where the scribe and painter were left with no room for manoevre. That said, it remains to be demonstrated – as suggested earlier – exactly how the scribe, or the person dictating the text, reacted when the break-line threatened to come uncomforably close to the base of the page. It should be stressed again that no claim is being made here that the process just described was the standard one; but the practicalities of how an illustrated Shahnama was produced in the early fourteenth century do demand somewhat more attention than they have received so far. The message of the four papers just discussed is that it pays to look extremely closely, and then look again, at Persian manuscripts if their secrets are to be revealed. It has to be said that, by and large, this has so far been done only rarely in the history of Western scholarship on Persian book painting. The results achieved by such closely focused scrutiny in these papers help to blaze a trail for future research, and drive home the point that the traditional concerns of Western scholarship on Persian painting need to be supplemented by quite different approaches. It is no longer enough to answer the standard questions of who, what, where and when – questions that, rooted as they are in the evolution of Western art history, are inappropriately Eurocentric for an art forged in quite another culture. The fact that one or two articles have been written about a given manuscript, or that it has figured in the general surveys of Persian painting, does not by any means indicate that the last word has been said about it. The combined evidence of the Great Mongol Shahnama and the Epics in particular suggests that it would be prudent for scholars of Persian painting henceforth to be more alive than their predecessors were to the fact that books, like buildings, are apt to have an afterlife. To assume that they have, as it were, remained in the deep freeze since the moment of their creation is to minimise the importance of the reception that they had in later times. That reception is often very revealing about later taste, as in the case of the Great

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 47EE1BEE /production/pindar/booksxml/hillen/chapter−12

372 Mongol Shahnama, where the somewhat coarse repainting of certain leaves implies an indifference to visual consistency, colour balance and surface texture, to say nothing of a fundamental lack of respect for the integrity of an ancient masterpiece executed in a very different style from that which was considered fashionable in Qajar times. Yet the Qajar remargining of the leaves shows that, even at such a late stage in the history of Persian painting, it was still perceived as appropriate that the text-block should have ample blank margins on at least three sides. In much the same way, the attributions attached to unsigned paintings, a practice particularly marked in the case of late Timurid work, reveals a Safavid taste in which connoisseurship, as displayed for example by the recognition of the hands of named painters, had become fashionable. But as contemporary sources show, the critical vocabulary needed to justify those attributions – however acute the critical perceptions which prompted them – was not yet in place. Adel’ Adamova’s paper, like Sheila Blair’s, is devoted to a manuscript on which she has already written a co-authored book. Here, with the advantage of hindsight, she takes a measured look at her earlier conclusions and builds on them. She rightly characterises the Inju style as sketchy, but convincingly repudiates the common accusation that its draughtsmanship is coarse and inaccurate, showing that this judgement does not take into account the fact that the lavish application of thick black lines represents much later overpainting. Here, then, is a parallel for the process noted earlier a` propos the Great Mongol Shahnama, namely the transformation of a fourteenth-century idiom to align Ilkhanid paintings with later taste. On the iconographic side, she notes that heroic scenes are more than twice as frequent as those which depict “historical events” (interpreting that phrase generously so as to include mythical and legendary material). This, like her discussion of the pre-Mongol antecedents of this Shahnama, opens the way to a more nuanced view of the special contribution of this dated Injuid manuscript to the history of early illustrated Shahnamas. No less than four papers in this book – almost half of the total – deal with Shahnama themes divorced not only from the context of the book, but also from any of Firdausi’s text. At most, some of these images bear brief captions identifying the personages depicted. These images, then, for all that they are independent of Firdausi’s text, are intended to be recognised. The implication is clear: these people, and the narratives in which they figure, were familiar outside the book itself. They had entered a popular world of myth, legend and folklore which was perpetuated by oral transmission – in

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 47DFE942 /production/pindar/booksxml/hillen/chapter−12

NEW PERSPECTIVES IN SHAHNAMA ICONOGRAPHY

373

many cognate versions – perhaps even more than by literary means. Given the length of the Shahnama text, that oral transmission would favour abbreviated references to the narratives of the epic, and these in turn would tend to encourage the gradual development of generic, simplified, speaking images. Such images become visual spotlights of the key passages in the epic, rendered instantly recognisable by dint of frequent repetition. They carry a freight of both narrative and symbolic significance. Sylvia Auld decodes the elements of Shahnama origin on a bronze bowl of fourteenth-century date in the Victoria and Albert Museum. Its figural style excludes a connection with the Fars school of metalwork, in which associations with pre-Islamic Iran were vaunted. The bowl is an interesting example of how Shahnama images could readily be accommodated to the format so standard for inlaid metalwork of the thirteenth and fourteenth centuries, in which the emphasis lay on a sequence of roundels rather than on continuous narrative. Only minor changes in mount and weaponry were needed to transform a generic rider into an image of Faridun. Similarly, Zahhak, Bahram Gur and Azada, and perhaps Zal, can also be identified thanks to minor iconographical clues. Yet the neutral setting of these images — they are placed against a vegetal ground – excludes any marked narrative intent. Thus the Shahnama complexion of these figural roundels is not strongly marked, which in turn enables them to operate simultaneously as examples of a standard iconography of royal pursuits which was widespread in the metalwork of the thirteenth and fourteenth centuries. In a further twist of ambiguity, Sylvia Auld notes that many of these images can be read, thanks to their surrounding animal and waqwaq imagery, as references to the fixed stars and their constellations. The bowl thus shows how readily stories from the Shahnama could fit into the royal and astrological iconography which was the fashion in Iran, Anatolia, Iraq and Syria in these centuries. This collateral development of Shahnama themes independently of book painting and thus of Firdausi’s text enriches our understanding of how these traditional stories fashioned new roles and meanings for themselves over time. Those roles extended, moreover, beyond the Islamic world – if only through the agency of images reused in new and unheralded contexts. Avinoam Shalem’s paper documents how an Iranian or Central Asian silk with gold thread, depicting Bahram Gur and Azada, was reused as a chasuble in which Bishop Hartmann of Augsburg (d.1286) was buried. So fragmentary are the remains that it is not certain whether this cloth was

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 471C8DD2 /production/pindar/booksxml/hillen/chapter−12

374 originally intended to serve as a wall hanging or as a garment. But in either event this textile – and it is by no means unique; another episode from the story of Bahram Gur to figure on textiles is his seizing of a lion cub – is a reminder of a further medium for Shahnama images and yet another use of Shahnama iconography both within and beyond the Islamic world. As Avinoam Shalem notes, emancipation from the constraints of the written text allowed the artist to reinterpret the familiar scene in quite radical fashion, for example by adding redundant winged quadrupeds or an extra archer to the core scene, or by juxtaposing it with the figure of an enthroned ruler. Nor should the effect of sheer repetition – a sequence of identical or mirror-image medallions in the formula so commonly adopted for medieval Islamic textiles – be underestimated. All in all, then, this textile shows how pervasive certain Shahnama images had become even before the fashion for illustrated Shahnamas had gathered momentum. The continued vitality of Shahnama images in the nineteenth and twentieth centuries and in unexpected media is the subject of two further papers. Ulrike al-Khamis and Godfrey Evans analyse a collection of brass plaques executed in the repoussé technique on sheet brass manufactured in Liverpool between c.1853 and c.1864. They depict, among other subjects, a series of shahs, princes and paladins from the Shahnama – Faridun, Salm, Tur, Iraj, Minuchihr, Kay Ka’us, Kay Khusrau, Afrasiyab, Zahhak, Siyamak, Rustam, Hushang and the Black Div. Possibly produced in Isfahan or Shiraz, they were probably used as wall decoration and the connection with the figural tiles, paintings and frescoes of similar content so fashionable in nineteenthcentury Iran is unmistakable. The iconography of these plaques resembles that of the other media just cited in its preference for single, almost emblematic figures rather than for detailed narratives. Here too, then, as in the medieval inlaid metalwork and textile just discussed, it was enough to evoke the great names and selected episodes of the Shahnama; the fact that these were familiar to the viewer made it unnecessary to do more. The figures have become symbols, mere pegs on which to hang memories. Rather more ambitious is the material from the fort of Bandar-i Tahiri discussed by Jennifer Scarce. Datable to the period around the First World War, it comprises a sequence of eighteen carved stucco panels which operate like framed pictures in a gallery and decorate the quarters of the ruling shaikh of the al-Nasuri family which controlled the port. Six of these images might be described as formulaic, for they represent seated figures. The remaining scenes, however, are narrative and they focus above all – though not

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 49E54547 /production/pindar/booksxml/hillen/chapter−12

NEW PERSPECTIVES IN SHAHNAMA ICONOGRAPHY

375

exclusively – on a succession of popular scenes from the cycle of Rustam. These feature the deaths of Isfandiyar, Suhrab and Rustam himself; Suhrab attempting to identify Rustam, Rustam killing the White Elephant, his combat with the White Div, the Div Akhvan throwing Rustam into the sea, Rustam rescuing Bizhan from the pit, Rustam kicking away the rock rolled down on him by Bahman, and Rustam with Zal. These ten scenes constitute a most unusually full cycle of the life and exploits of the Shahnama’s most popular hero, and thus shed light on a little-studied aspect of Shahnama iconography, namely the rise and fall of interest in particular themes, aspects or personalities of the Shahnama over the course of centuries. It remains to be seen why this particular scion of the al-Nasuri family which ruled Bandar-i Tahiri in the twilight of Qajar power should have held the story of Rustam so dear – to the extent of privileging it above the royal and political themes of the epic. Visually these images derive from the illustrated lithographed Shahnama tradition of Qajar Iran, and from Qajar oil and lacquer paintings. But they are also a very late example of the ancient custom of using Shahnama scenes as wall decoration. The limitations of applying the broad style so fashionable in Qajar Iran to the medium of stucco carving sufficiently explain the minimalist conventions employed for details of landscape, architecture and furnishings, though (as in Qajar painting) clothing, jewellery, armour and weaponry are treated in more detail. An unusually original approach to the problems of reconstructing how Persian illustrated manuscripts were designed is explored in the paper by Amin Mahdavi. Using a significantly large sample of 84 Shahnama manuscripts, he employs mathematical techniques to calculate the principles followed by the Persian designers in such matters as the ratio of margins to text block, the location of the text block on the page, the possibility that proportional ratios were applied, and the exact operation of grid divisions. The result of these calculations is to suggest a method for determining the original dimensions of a page even when it has been cropped. It should eventually be possible to correlate variations in grid layout with specific periods and provenances. It is a sobering thought that this is the first time that advanced mathematical (especially geometrical) calculation has been brought into art-historical scholarship on Persian painting. Yet the exciting results achieved here – whose implications extend to the arts of the book in general, and in particular to the production of Qur’ans – give reasons to hope that problems earlier regarded as intractable may now become less so. One thinks here of the use of statistical techniques to establish date and

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 30 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 1C585B54 /production/pindar/booksxml/hillen/chapter−12

376 provenance or to define the characteristics of a given school on the basis of iconography or manuscript design rather than, as hitherto, on the basis of style alone. These papers, then, demonstrate that current understanding of the evolution of the illustrated Shahnama over the centuries is increasing at a well-nigh exponential rate. The time-honoured focus on patronage, date, provenance and style is as relevant as ever, for it provides the basic building blocks for more elaborate investigations. But to a steadily increasingly extent it is being supplemented by new approaches which can be sampled in this volume. These pose entirely new questions, tackle old problems from new angles, and give promise of greater precision in areas where earlier scholars had to be content with guesswork. The study of the illustrated Shahnama, then, is set for another quantum leap forward, and the database currently being assembled under the auspices of the Cambridge – Edinburgh Shahnama Project bids fair to provide the necessary launching pad.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-12 Sequential 12

JOBNAME: ???? PAGE: 1 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 31C28D83 /production/pindar/booksxml/hillen/chapter−13

XIII Erudition Exalted: The Double Frontispiece to the Epistles of the Sincere Brethren

T

HE two paintings to be discussed in this paper (figs. 2–3) mark the end of an era. Painted in Madinat al-Salam (Baghdad) in Shawwa¯l 686/November 1287, they introduce1 a copy of an early cAbbasid text from Bas. ra commonly known by its short title: Rasa¯’il Ikhwa¯n al-S.afa¯’. The manuscript is preserved in the Library of the Süleymaniye Mosque in Istanbul (Esad Efendi 3638).2 The text is a kind of Shi cite encyclopedia;3 in early medieval times, some parts of southern Iraq had strong Shi cite loyalties, or a tendency to oppose the ruling authority.4 The work was a composite enterprise, and the five authors are duly identified on the first leaf of the frontispiece, fol. 3 verso, above the image (fig. 3). The phrasing of this text is sufficiently curious to be worth noting: ‘It has been handed down from the Tatimma .Siwa¯n al-H.ikma to Z.a¯hir al-Dı#n Abu l-Qa¯sim al-Bayhaqı# that five of the wise men assembled and they compiled The Epistles of the Brethren of

1

After a preface which extends from fol. 1 recto to fol. 3 recto; Farès (1957), 78. The most convenient publication of these leaves is that by R. Ettinghausen (1962), 98–103 with three color plates. The fullest publication is that by Farès (see n. 1, above), 78–86. 3 Marquet (1971), 1073–75. 4 E.g., Djait describes Shi cism as ‘the quasi-unique ideology’ of Kufa at the end of the ninth century; Djait (1986), 359. 2

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 60609055 /production/pindar/booksxml/hillen/chapter−13

378 Purity. And they were Abu¯ Sulayma¯n Muh.ammad b. Miscar al-Bustı# known as al-Maqdisı#,5 and Abu’l -H.asan cAlı# b. Zahru¯n6 al-Zanja¯nı# and Abu¯ Ah.mad al-Nahraju¯rı# and al- cAwfı# and Zayd b. Rifa¯ ca. And the editor of the book was al-Maqdisı#.’ So the scholar who was mentioned first in the list of the five authors and is thus specially honored (and is further singled out by being given his full name, with kunya, patronymic, ism and laqab, plus the name by which he was commonly known—al-Maqdisı#) clearly stands out from among his colleagues. The corresponding space above the painting on fol. 4 recto bears the full (rhyming) title of the work: Rasa¯’il Ikhwa¯n al-S.afa¯ wa khulla¯n al-wafa¯ (The Epistles of the Brethren of Purity and the Friends of Loyalty) (fig. 2). From the outset, then, the reader is made aware by means of both writing and image that this is a work of multiple authorship. Whether these men were Isma¯ c#ılı#s has been a matter of some contention, as has the identity and number of the authors themselves.7 The two paintings, both measuring 20.5 × 18 cm, are meant to be read together. They constitute a double frontispiece, a format which was indeed known in the Byzantine world, but which Arab painters in the thirteenth century made very much their own. Indeed, they tossed off a succession of strikingly original variations on this neglected theme. Sometimes these double frontispieces worked by repetition, with two virtually identical images on facing pages,8 much in the style of so many Qur a¯nic frontispieces.9 In other cases the artists, as will shortly appear, struck out in a more radical 5

The name appears somewhat differently as ‘Abu¯ Sulayma¯n Muh.ammad b. Macshar al-K . udsı# al-Bustı#, who, according to Abu¯ Sulayma¯n al-Mant.ik.#ı, was the author of the Rasa¯’ı#l of the Ikhwa¯n al-S . afa¯ ’; Goitein (1986), 330a. Yet the text in the Istanbul manuscript clearly has .s ¯ın, not shı#n, in the father’s name, and reverses the order of the two nisbas, while clarifying the distinction between Muh.ammad’s place of origin and his commonly used name. 6 Marquet (1971), 1071, gives this name as Ha¯ru¯n, while Ettinghausen has Zahra¯n with a long alif, but the name is clearly written as Zahru¯n with a long wa¯w. Ettinghausen’s transcription ([1962], 100) has several minor errors. 7 Marquet (1971), 1071–76; see also Stern (1964), 405–28; idem (1946–47). 8 Some of these double frontispieces, such as those of the Paris Kita¯b al-dirya¯q of 595/1199 (Farès [1953], pls. 3–4), or of the undated but probably early thirteenth-century Mukhta¯r al-h.ikam wa mah.a¯sin al-kalim (The Choicest Maxims and Best Sayings) of al-Mubashshir (Ettinghausen [1962], 75; see my figs. 24–25) are so similar that the immediate visual impression that they give is that they are identical. In the Kita¯b al-dirya¯q the principal differences between the two halves of the double frontispiece are confined to details of dress and of the set of the head. The same goes for the Mukhta¯r al-h.ikam; in this manuscript, moreover, there is a double finispiece whose design is closely related to that of the

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4E7D7B7A /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

379

way.10 The two Rasa¯’il leaves under discussion conclude the creative evolution of the Arab frontispiece. As such, they are a marker for the astonishingly rapid rise, flowering, and decline of Arab painting itself. And it all happened in well under a century. This is not to say that Arab painting was finished after 1287. But in the matter of frontispiece design,11 it marched vigorously backwards. Conventional wisdom has long held that Arab painting as a whole went into permanent decline in the fourteenth century,12 though the occasional dissenting voice has been raised.13 And, by a telling coincidence, just as the Arab frontispiece ran out of steam, the Persian frontispiece, which was destined to have a much longer history, made its debut. The leaves that introduce the Paris copy of Juwaynı#’s History of the World Conqueror, dated 1290 (fig. 1), are infinitely less knowing and subtle than the Rasa¯’il leaves; they have only three figures and their landscape is more cerebral than real (fig. 4).14 But they carry the seed of a rich future. For that, they have China to thank. And China was essentially a closed book to Arab painters. Yet the Paris frontispiece is transitional in more ways than in its hesitant experiments with a visual vocabulary derived from the Far East. For it looks both backwards and forwards in its subject matter. Backwards in that the principal seated figure, the vizier Juwaynı#, is an author portrait, which thus taps into an already millennial tradition whose earliest frontispiece (fig. 26). I am grateful to Dr. Elizabeth Lambourn for lending me her slides of these images, and to Dr. Jaclynne J. Kerner for allowing me to reproduce her slides in this volume. 9 An early example is the double geometric frontispiece of the Qur’a¯n of Ibn al-Bawwa¯b, made in Baghdad and dated 391/1000–1, fols. 8v–9r, which themselves follow two successive pairs of decorative openings in which writing figures prominently (fols. 6v–7r and 7v–8r); after the end of the text there is a double geometric finispiece (fols. 284v–285r), followed by another double decorative opening with multiple lines of writing on each leaf. See, for illustrations of these double-page compositions, Rice, D. S. (1955), pls. 1–5. 10 Here again, Qur’a¯ns showed the way: see the double frontispiece of the luxury Qur’a¯n in S.an ca’ attributed by Graf von Bothmer to the Umayyad period (von Bothmer [1987], color pls. 1–2). 11 c c c See, for example, the frontispiece of Ibn Z . afar’s Sulwa¯n at-mut.a¯ fi udwa¯n al-atba¯ , datable to the first half of the fourteenth century, in the color facsimile edition (Ibn Z . afar [1985]), fols. 1v–2r. 12 Ettinghausen (1962), 179; Ward (1996), 16, 311. 13 See the arguments presented, on the basis of the Ibn Z . afar manuscript, by A. S. Melikian-Chirvani (1985), xi–xiii and 166–68. 14 Richard, F. (1997), 41.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4FA89FBD /production/pindar/booksxml/hillen/chapter−13

380 surviving traces are to be found in late classical times in Rome,15 and which was then taken up enthusiastically in both Western16 and Eastern17 Christianity for evangelist portraits. But this frontispiece also looks forwards in that Juwaynı#, as S.a¯h.ib-i Dı#wa¯n (the title inscribed, probably later, beside his image), was—after the Mongol Ilkhan himself—the most powerful man in the land. So this is also a ruler portrait, and in that sense the Paris frontispiece foreshadows, not least in its landscape setting, the long line of ruler images that was to dominate the tradition of frontispiece design in Persian painting.18 None of this is to deny that the standing figure is of greater importance than the vizier;19 the contrast in body language is enough to ensure that, to say nothing of the striking disparity in costume. This figure could be Hülegü or Abaqa Kha¯n;20 there is at all events a high likelihood that this is one of the Mongol rulers or a high Mongol dignitary.21 The Rasa¯’il leaves, as befits their role as the summative experiment in Arab frontispiece design, are exceptionally rich in content. Accordingly, they lend themselves to quite diverse approaches and interpretations. The present paper will confine itself to five distinct headings only, though plainly these leaves would sustain a much more detailed analysis than this. First, their core subject matter: how they work as author portraits. Secondly, their use of pictorial space, which breaks much new ground. This leads naturally to the third theme, the architectural setting, which will be treated at length since no 15

Weitzmann (1977), 11 and figs. II (the Vergilius Romanus) and VIII (a Carolingian copy of a fifth-century Terence manuscript). 16 Friend (1929); Nees (1987), 83 ff. 17 Friend (1927). Some of these formulaic images made their way into other media, for example ivories; see a probably Egyptian example of the sixth century in which the figure, dressed in tunic and pallium, and holding a book in the crook of his arm, strides illusionistically out of a picture space defined by a knotted curtain (Weitzmann, ed. [1979], 542–43, no. 487). 18 A tradition that perhaps culminated in the depiction of the court of Sultan H . usayn Ba¯yqara¯ in the 1488 copy of Sacdı#’s Bu¯sta¯n (Lentz and Lowry [1989], 260–61). For a full analysis of this frontispiece, see Prentice (1977), 45–54. 19 Though the fact that his face has been rubbed out (as has that of his groom), while Juwaynı#’s face remains as painted—which suggests the anti-Mongol feeling of some later user of the manuscript—reduces his visual impact. 20 These are the two suggestions made by Richard, F. (1997), 41. 21 Ettinghausen (1959), 48, transmits the opinion of J. A. Boyle that the figure is likely to be either Hülegü or the Amı#r Arghun. Ettinghausen notes that this formulaic representation has as its prototype the classical author inspired by the muse standing beside him (ibid., 49).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 45E443AF /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

381

other Arab frontispiece makes so much of this feature. The fourth theme again focuses on a pioneering aspect of the Rasa¯’il paintings, namely the interplay of figures to construct a narrative more complex than is to be found in any other Arab picture of comparable function in this period. Finally, the paper will attempt to characterize the distinctive style of these paintings and to assess their importance in the history of the Arab frontispiece in the thirteenth century. The Multiple-Author Portrait The obvious place to begin is with the principal problem that the painter had set himself—how to depict not one but five authors.22 The inherent difficulty of that problem is well illustrated by the relative failure of earlier attempts to solve it. These attempts are on the whole characterized by wooden, lifeless, and implausible arrangements of the relevant figures.23 From the way that the painter of the Rasa¯’il leaves met this challenge springs most of what is remarkable in this frontispiece. How was he to put over in convincing fashion the concept of multiple authorship, let alone the hierarchies within the team, the milieu within which the book was written, or the readership for which it was intended? Moreover, in the more ambitious Arab frontispieces of the thirteenth century, the emphasis was not confined to the author himself, as was the case in the classical, medieval Western and Byzantine traditions, but extended also to the acquisition and transmission of knowledge. That is a far more complex theme, and it took the author portrait a long way from its origins. This change of direction injected narrative, and with it far more human interest, into a genre that had hitherto tended to the static and formulaic. This broadening of the scope of the author portrait populated the picture space with an entire dramatis personae, from students to servants. There was room too for some quirkiness in their interaction.

22

Five authors are mentioned in the text on fol. 3v but six figures dominate the frontal plane. Hoffman (1993), 7 and 17, n. 7, explains the discrepancy as due to reliance on a late antique group portrait. But the desire to create a satisfying visual symmetry may also be at work here, and the addition of the figure identified by Ettinghausen ([1962], 102) as a scribe (fig. 28) was an elegant way of achieving this. No doubt other possibilities could be canvassed. 23 See Hoffman (1982), 283–84 and, more generally, 255–59 and figs. 86a–91.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 550DB4EC /production/pindar/booksxml/hillen/chapter−13

382 All this, then, brought a moribund genre back to life. No wonder that no two Arab frontispieces of the thirteenth century that depict authors are the same.24 No wonder that one experiment crowded on the heels of another. No wonder that the author himself seems to change his nature as readily as a chameleon. In one image, he gives a one-on-one practical demonstration to an absorbed student.25 In another he is the practical businessman, ready—at a price—to grant formal permission to a pair of deferential students to go out into the wide world and teach the material he has dictated to them.26 The commercial aspect is again to the fore as he sits in his ostentatiously well-stocked shop preparing a drug for an affluent customer (fig. 6).27 Or he

24

Whereas in Byzantine evangelist portraits, for example, the sheer quantity of repetition is astonishing. See Friend (1927), pls. 1–16, 103–6, 125–28, 132–35, 142–43, and 173–75. 25 Dioscorides, Hayu¯la¯ cila¯j al-t.ibb, an Arabic translation of De Materia Medica, 626/1229 (probably made in Northern Mesopotamia), now in the Topkapı Palace Museum Library (Ahmet III, 2127), fol. 2v. For a color plate, see Ettinghausen (1962), 71. 26 See ibid., 67 and 70; color plates on 68–69. For a summary of recent discussion of this double frontispiece, see L. Komaroff in Evans and Wixom, eds. (1997), 429, 432–33. The practical process whose culmination is depicted in these two paintings is described by J. Pedersen (1984), 24–32. For evidence that the artist was a Muslim by the name of cAbd al-Jabba¯r (?) b. cAlı#, see O. Grabar’s review of Farès (1957): Grabar, O. (1959). 27 This is from another partially dispersed and incomplete Dioscorides manuscript, dated Rajab 621/June–July 1224, now in the Süleymaniye Mosque Library, Istanbul (Ayasofya 3703); see Grube (1959), 173 and Abb. 1 on 171. Professor Grube interprets the central figure as a druggist serving two customers, and his is indeed a plausible suggestion. Nevertheless, such a workaday scene would mark a major departure from the longestablished tradition of depicting the author in the opening pages of certain manuscripts (notably of this very type) and it should be remembered that in this case the image (nearly the only one left in this ruthlessly plundered manuscript) occupies fol. 2 recto and thus precedes the beginning of the extant text on fol. 2 verso, which begins with Book 4, so that it is in pole position in the manuscript rather than taking its place in the sequence of later illustrations. Comprising today 155 fols. that encompass Books 4–7, it is not possible to say whether Books 1–3 were removed or were bound as a separate volume, or if it was planned as a truncated version of the text. (I am grateful to Dr. Linda Komaroff for this information.) Moreover, all the other Arabic Dioscorides manuscripts of the thirteenth century that survive and that have figural frontispieces depict Dioscorides himself in them. It may seem preferable, therefore, to interpret the elderly white-haired man rendered in profile to the right, his hand imperiously outstretched, as Dioscorides himself. The composition can be seen in absolute clarity in the colored drawing made by A. Süheyl Ünver in 1939 and published in Brandenburg (1982), 115.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4DE26566 /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

383

is enthroned in majesty, set on high above his fellow scholars28 or his pupils or attendants (fig. 7).29 Sometimes he is shown lecturing to an audience craning to hear his every word, while he himself keeps a weather eye on his lord and master, the amı#r enthroned on the opposite page.30 No question here about where the power lies—it is with the eagle, not the pigeon; with the frontal image, not the three-quarter or profile one. So while there is the odd image of the standing author holding his book in a Byzantine style which is only lightly Islamized, as in the Oxford Dioscorides,31 that is exceptional. So, too, is a frontispiece whose theme is exclusively royal, as in the Paris Kalı#la wa Dimna, probably of the 1220s (fig. 8).32 Indeed, the intellect rather than political power is exalted in the frontispieces of the thirteenth century. For the most part they proclaim that the pen is mightier than the sword. But the balance was to shift decisively—and for good—in the opposite direction in the fourteenth century and thereafter. This contrast between Arab veneration of learning and Persian veneration of power is, to say the least, instructive. But that is certainly the message of the frontispieces. Where exactly, then, do the Rasa¯’il leaves belong in this sequence? In simple terms, they are the most developed realization of the concept of multiple authorship in frontispiece design. This theme was itself rare enough. Perhaps its earliest expression is in the little-known double frontispiece to an undated manuscript entitled Risa¯lat al-S.u¯fı# fi’l-kawa¯kib, which ends with a Poem of the Constellations sometimes attributed to Ibn al-S.u¯fı#

28

As in the Dioscorides in Bologna (Biblioteca Universitaria, Cod. arab. 2954, dated 642/1244–45, fol. 141 recto), identified by a later user as Dioscorides flanked by Luqma¯n and Aristotle (Grube [1959], 182). For a color plate, see Brandenburg (1982), 209. 29 For a color plate of this image (undated; Istanbul, Süleymaniye Mosque Library, Ayasofya 3704, fol. 1 verso), see ibid., 95. 30 Farès (1953), 16–17, figs. 3–4. The layout here is as misleading as the caption (‘anges baghdadiens’), for it reverses the actual layout in the manuscript itself. 31 For a good color plate of this image (Bodleian Library, Cod. or. d. 138 [Baghdad, 637/1239–40], fol. 2 verso), see Evans and Wixom, eds. (1997), 402. 32 Bibliothèque Nationale, arabe 3465. For a brief description, see O’Kane (2003), 49; and Hamid (1966), vol. 1, 224–25 and vol. 2, fig. 55. Hamid notes an inscription on the writing board of a pupil in the stylistically related Maqa¯ma¯t of al-H . arı#rı# (B.N., ms. arabe 6094, fol. 167) which can be deciphered as stating that the book was made in Damascus, and on this basis concludes that arabe 3465 was also made in Damascus (ibid., vol. 1, 224 and 236 and vol. 2, fig. 64).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 4878D73D /production/pindar/booksxml/hillen/chapter−13

384 but is more likely to date from the twelfth century.33 Mehdi Mahboubian’s theory that it depicts the Buyid ruler cAd.ud al-Dawla on the right and al-S.u¯fı# himself on the left34 is unconvincing, partly because the style puts the paintings around 1200 (and who would wish to honor a potentate who had died some two centuries previously?), and partly because the right-hand figure holds an astrolabe, scarcely an attribute of authority. He is also smaller than the other man holding the book. And neither figure wears a crown. This painting, then, depicts two scholars—but they do not communicate directly apart from more or less facing each other, as one is in profile and the other in three-quarter view. Each is treated as an independent personage, with the figure to the left treated as the more important one. Vesel’s suggestion that the figure holding the astrolabe is Ptolemy is plausible, for al-S.u¯fı# s work was based on that of Ptolemy, which al-S.u¯fı# revised.35 There may indeed be an attempt here to distinguish between Islamic and Greek culture, with the taller, more imposing figure of al-S.u¯fı# wearing a turban and thus contrasting with his opposite number, who wears a fillet (traditionally associated with classical culture) around his brow. The segmental arch that crowns each picture recurs in other Arab frontispieces,36 and, like the form of the throne back, the blue-tiled wall and the huge blue scrolls that unfold languidly against a gold backdrop, creates a distinctly Islamic ambience for these images. The theme of multiple authorship is developed a good deal further in the double frontis- and finispieces to a copy of The Choicest Maxims and Best Sayings by al-Mubashshir, a manuscript attributed to Syria and datable to the first half of the thirteenth century (figs. 9–11).37 This repeats or rings minor changes on the theme of setting six or seven authors each within a separate compartment of an overall ornamental framework, but all so depicted that they connect with each other or with the outside world. The message is plain:

33 Now in the Riza-yi cAbbasi Museum, Tehran, ms. 570. I am grateful to Dr. Moya Carey, who is currently working on this manuscript, for directing me to a color illustration of this double frontispiece: Vesel (2001), 268–69. 34 Mahboubian (1970), no. 913 and unnumbered 16. 35 Vesel (2001), 262, 269. 36 E.g., in the Dioscorides manuscripts of 626/1228–29 and 637/1239–40 (see nn. 25–26 and 31, above). It is only fair to add, though, that it is also found in Syriac manuscripts of this period, as in the evangelist portrait in a Lectionary from Mar Mattai of 1220 (Vatican, Biblioteca Apostolica, ms. Siriaco 559). 37 Ettinghausen (1962), 74, 76–79; color pl. on 75.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 44553E47 /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

385

erudition involves discussion with your peers, and brings with it the responsibility to inform the rest of the world about your findings. All this is schematically achieved by the poses and gestures of the scholars (presumably themselves the authors of the maxims and sayings of the title). But the simple gold background, lightly embellished with vegetal scrolls, distances these images from reality, although it does serve to simplify the message. This could be summarized, moving from bottom to top, in the words ‘discuss,’ ‘transmit’, and ‘broadcast.’ Although the influence of Old Testament prophets or New Testament apostles has been proposed as an explanation of these figures,38 any such religious origin has been hijacked, so to speak, for secular purposes and for the dissemination of a secular text. It is this very theme—the acquisition and communication of knowledge—that is taken up and so smoothly and confidently developed in the two Rasa¯’il leaves. They exude a strong sense of lively debate, of a seminar. There is both humor and urgency here. The debate is largely confined to the left-hand page, fol. 4 recto, while the right-hand page, fol. 3 verso, seems rather to emphasize silence, meditation, and deep thought. Perhaps it is this subdivision between speech and silence that at least partly explains why the two halves of the frontispiece, while genetically similar, are not identical. In fol. 4 recto, a servant brings a book as further ammunition in the argument, while another book lies open on the knee of the reclining sage and a third book is held open in the hand of the man with corkscrew eyebrows who is eagerly challenging the central sage. A fan disperses the hot air that is no doubt being generated. Pictorial Space It is now time to turn to the second theme, the treatment of pictorial space. This frontispiece is full of spatial adventurousness. Faces are typically in three-quarter view, the angle that best meets the demands of narrative while also involving the viewer. As in Byzantine religious art, the figures are usually not in profile,39 even though that is the pose that their interchanges would lead us to expect. An all-but-frontal view is also used. People either interact directly through the direction of their glances, which cancel each other out,

38

Ibid., 78. For a concise account of the theological reasons behind this preference, see Demus (1948), 7–9 (page citations are to the 1976 reprint edition). 39

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 475F004B /production/pindar/booksxml/hillen/chapter−13

386 so to speak, or they turn to the facing page, again keeping the direction of their gaze within the frontispiece as a whole. Poses are even more varied. People turn their backs, step out towards us, lean back, stoop forward or recline, stand or sit, and overlap. In other words, the key central strip is in constant motion. The figures themselves, then, are the major instrument of spatial innovation, and their hand gestures, which are often reciprocal and are broadly at the same level, create a further ripple of movement. Textiles play their part too. Curtains swirl, loop, and undulate, with internal patterning employed to highlight the sense of movement. But even more noticeable than this is the treatment of clothing, including turbans. The body beneath the drapery is simplified into large ovals, but these are filled with furious activity expressed in churning, eddying folds. The Rasa¯’il leaves illustrate only one variation of an interest in intensely expressive ways of depicting drapery folds with which thirteenth-century Arab painters experimented. Related modes can be seen in the Tehran Ibn al-S.u¯fı#, the Mukhta¯r al-h.ikam, the Wa¯sit. Qazwı#nı#,40 and the London Na ct al-h.ayawa¯n.41 In the Rasa¯’il frontispiece, the outlines of sleeves and turban tailpieces resolve themselves into a flurry of cascading pleats and ruffles. It is all rather excitable, even flamboyant. This emphasis on crumpled cloth counterpoints, admittedly in a minor register, the broader movement of the figures themselves. The drapery, then, is part of the overall scheme of spatial mobility; it radiates energy, and a powerful sense of controlled movement expressed through a series of strictly contained vortices. This drapery technique, like that of the slightly later Marzuba¯nna¯ma,42 is very closely related to metalwork, perhaps inlaid metalwork43 as is suggested by the large oval patches denoting the lining of the sleeves or the books held by various figures, but also, perhaps, Byzantine cloisonné enamels.44 The chrysography—the ‘writing in gold’ that was such a hallmark of the

40

Ibid., 138. Contadini (1992). 42 Simpson (1982), 100, 102, and 104 (figs. 49–51 respectively). 43 On this general question, see Komaroff (1994), especially figs. 10b–c, 11, and 18a–d, the latter (a brass bowl in Lyons) with the same distinctive use of multiple parallel lines to assert a form. 44 For a representative selection of relevant images, see Wessel (1969). For superb color close-ups of such work, see Evans and Wixom, eds. (1997), front and back cover (ibid., 88, cat. no. 41). 41

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 48DF71DF /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

387

Byzantine manner—is more marked here than in any other known example of Mesopotamian or Syrian painting. The limited range of colors (black, white, orange, brown, blue, gold) also recalls metalwork, although in this respect too the immediate parallel is probably the Byzantine cloisonné enamel tradition. The distinction between the sages, squatting gigantically at the center of each composition (and not just at the center of each lower story) in the frontal plane, and their tiny servants, placed well back behind the lateral balustrades, is also a triumph of spatial subtlety. The many different overlaps underline this three-dimensional complexity. And part of the spatial interest of these figures lies in their sheer number. No other thirteenth-century frontispiece has so many active figures. They never coalesce into an anonymous mass, like the audiences in the frontispiece to the Paris Maqa¯ma¯t of 634/1237. Everyone has something to do. There are no walk-on parts. And that theatrical allusion is perhaps not entirely misplaced, for the influence of the contemporary shadow theater may be sensed here,45 especially in the well-lit but neutral central space occupied by the authors themselves, a space carefully singled out from its surroundings by its size and coloring. The Rasa¯’il leaves are the most stagy of all thirteenth-century frontispieces, cunningly choreographed and with plenty of sub-plot provided by the servants and the studious occupants of the upper story. The entire space is conceived as a shop front46 for how learning is acquired. This represents a quantum improvement on the mechanical sexpartite division of architectural space which was a cliché of earlier Arab painting.47 That said, size is still a function of status, which results in some figures inhabiting rather awkwardly the space allotted to them. Here the contrast between the large fan-waver and the pygmy proportions of the servant opposite, holding a book, is particularly marked, though the former intrudes into the space reserved for the scholars, which may help to explain his greater size. One may note too the elaborate—and perhaps rather labored—spatial device whereby his left arm pushes forward into the frontal plane as he grasps the column to steady himself, while most of his body is behind it.

45

The principal source here is P. Kahle; see especially Kahle (1910–11); idem (1940); idem (1954). See also Jacob (1925); Moreh (1987) and idem (1992), Part 3, 87–151, passim. 46 Cf. the pharmacy discussed in n. 26, above. 47 Cf. another pharmacy, in the Dioscorides manuscript of 621/1224 (Ettinghausen [1962], 87), or the parturition scene in the Paris Maqa¯ma¯t (ibid., 121).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 490127E4 /production/pindar/booksxml/hillen/chapter−13

388 The side curtains break through the notional frame established by the architecture itself, and this is typical of the spatial energy and expressiveness which permeates both paintings. The effect of these folds of cloth wreathed and wrapped around the columns is to emphasize the third dimension, rather like the column-hugging figures in the St. Petersburg48 and Paris49 Maqa¯ma¯t manuscripts. And yet any literal interpretation of space is excluded, because the curtains, in being wrapped around the outer columns, project well beyond the frame.50 The same spatial verve makes itself felt in the treatment of the balconies: three of the four people seated there lean their bare elbows casually over the balcony in a device strangely prophetic of the portraiture conventions of the early and High Renaissance with the protruding elbow claiming the frontal plane. Drapery billows out behind their elbows, thus drawing extra attention to the device. By this means, the figures in the balcony serve to link the ground floor with the story above it, and thereby assert their interest and involvement in what is going on below. Similar bridging motifs are known in thirteenth-century Maqa¯ma¯t manuscripts.51 In fol. 3 verso, the forward-projecting leg and haunch of each of the two men flanking the senior sage perform much the same spatial function, leading the eye back into the picture (fig. 2). Even the body language is dynamic and expressive, as in fol. 3 verso, where one figure is scribbling furiously with furrowed, knotted brow as though he were a stenographer under pressure. Note too the clenched fists in fol. 4 recto (fig. 1).52 Clearly, then, the spatial inventiveness of these paintings extends beyond the figures being depicted, for it includes the architecture and its accessories. Moreover, the artist has succeeded very effectively in so depicting the structure as to suggest three-dimensional space, and especially the recession

48

Ibid. 106 and 107. Ibid. 114. 50 A similar device is known in contemporary Armenian manuscript painting, as in the Queen Keran Gospels of 1272 (Evans and Wixom, eds. [1997], 355). 51 E.g. the scene of Abu¯ Zayd in the tavern in the Paris Maqa¯ma¯t; Stewart (1967), 95 (page citation is to 1975 reprint edition). 52 While there is a long tradition of such intensely dramatic depictions of the act of writing, going back indeed to Carolingian times—e.g. the celebrated image of St. Matthew in the Ebbo Gospels (Epernay, Bibliothèque municipale, datable before 823, ms. 1, fol. 18 verso; see Robb [1973], 111, fig. 60)—no continuous chain of transmission between such an image and the Rasa¯’il manuscript can be traced, and it is likely enough that the Arab version of this theme was an original invention. 49

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 44B8EBB8 /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

389

of the building at its sides. This leads naturally to the third theme of this paper, the architecture itself and the purpose behind its meticulous depiction. The Architecture There is a long prehistory for the role of architecture in author portraits, and some review of this rich past is necessary in order to contextualize the Rasa¯’il paintings. It was of course standard practice in both the Western medieval and the Eastern Christian tradition to place the author figure in some kind of architectural context. Typically, however, this setting has a strong flavor of fantasy. One cannot readily imagine the author actually living in that specific building. This is partly because of the gross discrepancy of scale—after all, the author figure dominates the frontal plane, and his spatial relationship to the building behind him is uncertainly defined—and partly because of the minute and fussy detail of the architecture. It has the air of a model, a doll’s house, rather than of an actual building. Yet the Christian tradition demanded a grand building as a backdrop for the more elaborate type of author (usually evangelist) portrait, and the power of that tradition can be sensed even in the most unlikely places. In the only known non-abstract Qur anic frontispiece, namely that of the Umayyad luxury Qur’a¯n in S.anca ,53 the human element has of course disappeared. But the architecture remains, and in spades. Even in the most minimalist author portraits, in the Western and Eastern Christian tradition alike, the backdrop, whether it be a simple arch or a sparsely furnished room, suggests an architectural space, though that space has little chance to assert itself given the overpowering scale of the author portrait itself. Thus, whether the setting is elaborate or simple, it is dwarfed and downgraded by the human element. It is plainly of secondary interest and is very seldom conceived as a natural setting for the author’s daily work. The cupboard of books seen in the background of the Codex Amiatinus, fol. 5 recto, as Ezra scribbles busily is one of the rare exceptions.54 Thus the conventions inherited by the Arab painters charged with devising author portraits made it mandatory for them to construct some kind of architectural framework for the figure which was the principal focus of interest.

53 54

See n. 10, above. Weitzmann (1977), pl. 48.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 415E5151 /production/pindar/booksxml/hillen/chapter−13

390 There was, however, a major difference of meaning between these Arab frontispieces and the Christian ones from which they initially drew inspiration. The Arab author portraits illustrated secular manuscripts, whereas the Christian evangelist portraits introduced a Gospel text, and even when the text was not a Gospel it was much more likely to be religious than secular—a lectionary, perhaps, a set of homilies, or a menologion. That religious purpose so central to the image seems to have affected the way that the architectural surround was perceived. It seems to have been regarded as inappropriate to depict an evangelist writing his Gospel in a workaday setting. His task was a sacred one; he himself was a saint; the text that he was composing was the word of God. Hence, of course, his halo, the gold background, and the overall sense of otherworldliness and the supernatural. It was thus only proper that the architecture, too, should partake of that heightened atmosphere: grander, purer, whiter than the buildings of this earth. Whether these Christian painters were deliberately trying to evoke the New Jerusalem is an open question. But when the architecture depicted in these Christian frontispieces rises above minimalism, it tends to acquire a quality at once charged and abstract: charged in the sense that it becomes iconographically meaningful, suggesting as it does a sacred setting, and abstract in the sense that it is an intellectual construct, only incidentally based on real architecture, even if a specific architectural type was the remote inspiration.55 It is not easy to determine how far Arab painters were attuned to the deeper resonances of the elaborate architecture in the east Christian frontispieces that they encountered—Byzantine, Syriac,56 Coptic,57 Armenian.58 It seems implausible that the resolutely (and thoroughly archaic) classical vocabulary of these imaginary buildings struck the same kind of chord with them as it did with Byzantine artists. For while Islamic architecture too had its Graeco-Roman heritage, a heritage which as it happens was actively being revived in twelfth-century Syria,59 this could not 55

For a detailed discussion of this issue and related ones, see Smith, E. B. (1956); and see ibid., figs. 68–69. 56 Evans and Wixom, eds. (1997), pl. 253. 57 Ibid., pl. on 370, 380–81, and pl. 251. 58 Der Nersessian (1963), figs. 17, 19, 21, 23, 46–47, 49, 51, 182, 196, 198, 200, and 202. 59 The principal study of the topic is Allen (1986); see also Rogers (1971) and, more generally, Grabar (1971) and Hillenbrand, R. (1986).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 47253E28 /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

391

compare with the unbroken classical tradition so jealously guarded in so many different fields in the Byzantine world, a tap-root which nourished the Byzantine imagination and sense of identity for a millennium and more.60 Whatever the reason, whether it was mere lack of empathy or an active pious distaste for Christian symbolism, Arab painters gradually discarded the more elaborate architectural confections of East Christian frontispieces. A thirteenth-century Arab Dioscorides (fig. 30) employs a much scaled-down architectural setting, though the echo of a consular diptych is palpable, notably in the predella.61 Several experiments in a minimalist idiom were made. By degrees, too, features derived from Arab domestic interiors and the built environment began to infiltrate these frontispieces, while a geometrical setting with no hint of architecture might be employed. There is no telling what other experiments of this kind were conducted in this exceptionally creative period, when Arab painters were busily rewriting a rule book for frontispiece design that had held sway for at least a thousand years. When one studies the Arab frontispieces dated or datable between 1199 and 1244 as a group, the immediate impression is of an absolute ferment of change. Barely two of them are alike. Most unfortunately, there is no dated Arab frontispiece between 1244 (the date of the Bologna Dioscorides) and 1287 (the date of The Epistles of the Sincere Brethren). Who can doubt that in the last decades of the cAbbasid caliphate, when (under the enlightened patronage of al-Na¯s. ir62 and al-Mustans. ir63 ) Baghdad was fast reclaiming its intellectual primacy in the Muslim world, far more illustrated manuscripts were produced than have survived? For such manuscripts seem to have been at the very peak of fashion at just that time. And while the sack of Baghdad in 1258 was a calamity comparable to that of the sack of Rome in 451, this metropolis (like Rome) obviously recovered quite quickly. So it would be strange if there were no predecessors for the many innovations of design in the 1287 frontispiece. What, then, is new about the architectural setting here? To begin with, this is a credible environment for scholarly work: an octagonal pavilion, it seems, a scriptorium which evokes the atmosphere of the bayt al-h.ikma, the intellectual powerhouse of the golden age of cAbbasid power. The sense of

60 61 62 63

See Mullett and Scott, eds. (1981). E.g. Delbrueck (1929), pls. 9–12, 16–21, 23–25, and 32. Hartmann (1975), 162–68 and 198–205. Hillenbrand, C. (1993).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 484044C9 /production/pindar/booksxml/hillen/chapter−13

392 corporate endeavor is very strong. But this building also functions as an instrument of design. As in so many Arab paintings of the thirteenth century, architecture serves as a device to subdivide space.64 It obligingly expands and contracts to accommodate not just the people that fill it but also the poses that they adopt, and—most importantly—their size. When this is badly done, and the idea is crudely applied, the architectural setting resolves itself into an equal six-part grid in two stories dumped into the frontal plane with no suggestion of depth. But in the Rasa¯’il frontispiece spatial divisions are managed with much more subtlety. The sense of a grid has disappeared altogether, with three unequal divisions on the lower floor and four on the upper story, which is barely half as high as the floor below and thus clearly of lesser importance. And on each of these two floors the spaces themselves are uneven, with the outer flanking compartments taking up markedly less room. The subdivisions here are therefore distinctly asymmetrical. Yet the painter has taken care to provide a visual justification for this, in that he has boldly attempted to give this building, which is all but identical in both parts of the frontispiece, a solid three-dimensional presence. He has done so by suggesting a polygonal (probably octagonal) structure, perhaps an open kiosk or pavilion, of a kind which has not survived in thirteenth-century Iraqi architecture but is known in the Iranian world four centuries later, for example the Namakda¯n (‘salt cellar’) at Herat.65 The painter achieves the likeness of a polygonal building by giving a marked slant to the lateral walls and emphatically foreshortening their volume. This beveling is accentuated by the triangular cornerpieces on the baseline. These are painted an eye-catching light blue and thus visually lift the side walls away from the main wall which takes up the frontal plane. That same diagonal line continues in the outer wings of the upper story, which are visually distinguished from the inner ones by being left blank and—a painter’s trick, this—filled with a gold background. Such close interest in the third dimension implies mathematical calculation, which the casually overlapping figures do not. An equally self-conscious spatial device is the way that the servant waving the fan ducks behind the column and grips it to steady himself as he leans forward to fan the authors. This burgeoning interest in the third dimension foreshadows developments in the next century, but in Persian, not in Arab painting, which regressed in this respect. Indeed, such sophisticated awareness of pictorial 64 65

For a general survey of this subject, see Barrucand (1986); eadem (1994). L. Golombek in Sourdel-Thomine and Spuler (1973), 355 and pl. 348.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 45BA7963 /production/pindar/booksxml/hillen/chapter−13

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

393

space was historically half a century or so ahead of its time. The same can be said for the emphasis on architecture as a complex and coherent setting for group scenes. Here the obvious successor to the Rasa¯’il frontispiece is not the World History of Rashı#d al-Dı#n, where the architectural framework is repeatedly a simplified tripartite one at ground floor level only,66 but the Great Mongol Sha¯hna¯ma. Here too the multiple subdivisions of a complex two-story building are crammed with action and human interest,67 and even a wall depicted at an angle appears once,68 though this feature, so confidently used in the Rasa¯’il frontispiece, does not come into its own until Jalayirid times.69 The precise treatment of the architecture in the Rasa¯’il frontispiece extends to both the overall structure of this scriptorium, with its distinctively recessed sides, and the details of its execution, which testify to the painter’s close familiarity with contemporary architectural techniques. This up-to-date knowledge is at the opposite pole to the purely imaginary content of the elaborate architecture in so many evangelist portraits. He is alert to the decorative effect of brickwork with painted rising and horizontal joints, though he favors black rather than the white used in actual buildings. Indeed, these walls may represent a plaster coating incised and painted to resemble actual brick coursing; this was a commonly employed technique in Ilkhanid architecture. The treatment of a coping by bricks in vertical bond is equally authentic. So is the fanciful segmental arch profile at bottom center, repeated in the outer niches of the upper floor, or the segmental saucer arch of the two central niches on that same floor. These arches caused the painter some difficulty; every one of them is lopsided to some degree. Indeed, not one of the arches in the Rasa¯’il leaves has a normal arched profile. But this is a rare lapse. The three different types of arch are all part of a fashion for decorative arch profiles that enjoyed a brief flurry of popularity in the century following the Mongol invasion of 1220: for example, in the mosque of cAlı# Sha¯h at

66 Rice, D. T. (1976), pls. 29, 35, and 69. Sometimes this is expanded, somewhat unconvincingly, to a quadripartite or sexpartite arrangement (ibid., pls. 70 and 34 respectively). 67 Grabar, O. and Blair (1980), pls. 1, 14, 15, 17, and 52. 68 Ibid., pl. 46. 69 Gray (1961), pls. 46 and 54; Sourdel-Thomine and Spuler (1973), color pl. LV.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 105 OUTPUT: Tue Aug 10 10:29:47 2010 SUM: 55DEF00E /production/pindar/booksxml/hillen/chapter−13

394 Tabrı¯z,70 at Vara¯mı¯n,71 Takht-i Sulayma¯n,72 and repeatedly at Ka¯sha¯n73 and Bast. a¯m.74 At the meeting places of the upper bays there is a curious pendant ornament. This building—whether it is a palace, a pavilion, an octagon, or a kiosk (perhaps riparian, to judge by the blue triangles in the lower corners, presumably a reference to water)75 —is a pretty elaborate and luxurious piece of architecture. The source for this inspiration could be local buildings, as indeed the brickwork suggests, with some quite specific details like the well-marked horizontal and rising joints, the use of specially diminutive bricks for the lintel above the lowest arch, the vertical lay of the coping above it, the use of lighter-colored brick for the base of the balcony, and the fact that this brickwork in general is further highlighted by the use of glaze or dark paint for the joints.76 Nor is this all. The painter has noticed that in buildings of high quality a brick incorporating two planes set at an obtuse angle is used in alternating courses when two walls meet at an angle, though his imagination runs ahead of him when he creates unevenly kinked bricks. He has also reproduced accurately the appearance of a wall in typical Ilkhanid common bond, down to the use of headers instead of stretchers in alternating courses at the edge

70

Wilber (1955), 69 and 71. See Pope and Ackerman, eds. (1938–39), pl. 400. It is of course likely enough that this, like similar luster mih.ra¯bs at Mashhad and Qum (Watson [1985], pls. 104a–b and 109; and 103), was actually made in Ka¯sha¯n (see n. 74, below). The same is likely to be true of another luster mih.ra¯b fragment from Qum now in Berlin (Kühnel [1931], fig. 13) and a luster mih.ra¯b in the Ja¯mi c Zı#r Da¯la¯n at Najaf, which like the Berlin fragment has a decorative arch formed out of calligraphy (Aga-Ogˇlu [1935], fig. 1), and equally of the tiled (not luster) mih.ra¯bs in the Masjid-i cAlı# and the Masjid-i Kalah, Quhru¯d (Watson [1975], pls. VIIIa and b respectively). More generally on such forms, see Wilber (1955), 69 and 71. 72 Melikian-Chirvani (1984), figs. 3–14, 16–17. 73 Watson (1985), color pl. N and pls. 111, 113, and 125–26. 74 See Pope and Ackerman, eds. (1938–39), pl. 395 (in the very early fourteenthcentury mosque adjoining the tomb tower known as the Ka¯sha¯na¯). For similar arch forms in the interior of that tomb tower, see Hillenbrand, R. (1982), 252–53, figs. 95–96, and—in the porch of that tower—255, fig. 98. 75 Among the riverine buildings in thirteenth-century Baghdad, where the Rasa¯’il manuscript was produced, the Madrasa al-Mustans.iriyya, the most famous madrasa in the Islamic world at that time, and dedicated—like the building in this frontispiece—to learning, took pride of place (Schmid [1980], Abb. 1–2). 76 For the quality of Iraqi brickwork of late cAbbasid times, as seen in the Mustans.iriyya madrasa and the so-called cAbbasid palace (probably the Bashı#riyya madrasa), see ibid., Abb. 1–15, 25–31, and 35–48. 71

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13 Sequential 18

JOBNAME: ???? PAGE: 1 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 16B35511 /production/pindar/booksxml/hillen/chapter−13−artwork

395

1. Double frontispiece, Ta rikh-i jaha¯n gusha¯y of cAla¯ al-Din cAt. a¯ Malik Juwaynı¯, probably Iraq (Baghdad), dated 4 Dhu¯ al-h.ijja 689/8 December 1290. Bibliothe`que Nationale, Paris (suppl. persan 205, fols. 1v–2r).

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 0B56E4C2 /production/pindar/booksxml/hillen/chapter−13−artwork

396

2. One leaf of double frontispiece, Rasa¯ il Ikhwa¯n al-S.afa¯ , Baghdad, Shawwa¯l 686/November 1287. Su¨leymaniye Mosque Library, Istanbul (Esad Efendi 3638, fol. 4r). After Ettinghausen (1962), 98.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 0D446260 /production/pindar/booksxml/hillen/chapter−13−artwork

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

397

3. One leaf of double frontispiece, Rasa¯ il Ikhwa¯n al-S.afa¯ , Baghdad, Shawwa¯l 686/November 1287. Su¨leymaniye Mosque Library, Istanbul (Esad Efendi 3638, fol. 3v). After Ettinghausen (1962), 99.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 0C4E65C2 /production/pindar/booksxml/hillen/chapter−13−artwork

398

4. Detail from one leaf of double frontispiece, Ta rı¯kh-i jaha¯n-gusha¯y of cAla¯ al-Din c At. a¯ Malik Juwaynı¯, probably Iraq (Baghdad), dated 4 Dhu¯ al-h.ijja 689/8 December 1290. Bibliothe`que Nationale, Paris (suppl. persan 205, fol. lv).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 4

5. Detail from one leaf of double frontispiece, Rasa¯il Ikhwa¯n al-S.afa¯ , Baghdad, Shawwa¯l 686/November 1287. Su¯leymaniye Mosque Library, Istanbul (Esad Efendi 3638, fol. 3v). After Ettinghausen (1962), 101.

6. Frontispiece, Hayu¯la¯ cila¯f al-t. ibb of Dioscorides, dated 621/1224. Su¨leymaniye Mosque Library, Istanbul (Ayasofya 3703, fol. 2r).

JOBNAME: ???? PAGE: 5 SESS: 25 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 12B85A2D /production/pindar/booksxml/hillen/chapter−13−artwork

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

399

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 5

7. Frontispiece, Hayu¯la¯ cila¯f al-t. ibb of Dioscorides, 13th century. Su¨leymaniye Mosque Library, Istanbul (Ayasofya 3704, fol. 1v).

8. Frontispiece, Kalı¯la wa Dimna of Ibn al-Muqaffa c, 1220s. Bibliothe`que Nationale, Paris (arabe 3465, fol. 34r).

JOBNAME: ???? PAGE: 6 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 0D1D5068 /production/pindar/booksxml/hillen/chapter−13−artwork

400

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 6

9. One leaf of double frontispiece, Mukhta¯r al-h.ikam wa mah.a¯sin al-kalim of Abu¯ al-Wafa¯ al-Mubashshir ibn Fa¯tik, probably early 13th century. Topkapi Palace Museum Library, Istanbul (Ahmet III, 3206, fol. lv).

10. One leaf of double frontispiece, Mukhta¯r al-h.ikam wa mah.a¯sin al-kalim of Abu¯ al-Wafa¯ al-Mubashshir ibn Fa¯tik, probably early 13th century. Topkapi Palace Museum Library, Istanbul (Ahmet III, 3206, fol. 2r).

JOBNAME: ???? PAGE: 7 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 15525090 /production/pindar/booksxml/hillen/chapter−13−artwork

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

401

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 7

11. One leaf of double finispiece, Mukhta¯r al-h.ikam wa mah.a¯sin al-kalim of Abu¯ al-Wafa¯ al-Mubashshir ibn Fa¯tik, probably early 13th century. Topkapı Palace Museum Library, Istanbul (Ahmet III, 3206, fol. 173v).

12. Frontispiece, Kita¯b quwa¯’id al-ah.ka¯m fı¯ ma‘rifat al-ara¯m. of al-H.illı¯, probalby early 14th century. Private collection.

JOBNAME: ???? PAGE: 8 SESS: 29 OUTPUT: Tue Aug 10 10:29:48 2010 SUM: 0F502D18 /production/pindar/booksxml/hillen/chapter−13−artwork

402

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-artwork Sequential 8

JOBNAME: ???? PAGE: 1 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 400A23CD /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

403

of a wall, or of a row of headers as a lintel above an arch. The arabesque ornament in blue and white tilework, or a combination of tilework and terracotta, is also a feature familiar in Ilkhanid architecture, for example at Sult. a¯niyya77 and Bast. a¯m.78 Given the demonstrable accuracy of these details, it seems reasonable to suppose that other features of architecture and its decoration for which no parallels survive may also be founded upon observation rather than fantasy. The balustrades, for example, ebony in color and possibly also in material, are inset with arabesque scrolls, perhaps in ivory although there is a bluish tinge to this inlay work. Arrow- or hourglass-shaped brass (or gilded brass) fittings of a kind familiar from Ilkhanid woodwork, for example the cenotaph of Shaykh S.afı# al-Dı#n at Ardabil, are set at the short edges of these panels. The columns of the lower story, apparently made of wood, are also very complex.79 They appear to have t.ira¯z bands. They have the same silver vase-shaped designs for bases as for capitals and bear patterns that look like Chinese seal script or, more likely, a kufic version of that style. The central part of these columns is painted silver, and may indeed have been silver-plated. Their upper and lower extensions bear successive bands of rectilinear geometric ornament in gold and black, which also occur at the junction of the column and the upper and lower walls. No wooden columns from courtly buildings survive from Ilkhanid times, so this contrast between plain silvered surfaces and golden decoration furnishes precious evidence for contemporary taste. The few columns illustrated in Ilkhanid painting of the

77

Seherr-Thoss and Seher-Thoss (1968), pls. 40–43 and 45. Noted in an unpublished paper delivered by J. M. Rogers at the Seventh International Congress of Iranian Art and Archaeology held in Munich in 1976. 79 Columns with similarly extravagant decoration can be seen in early fourteenthcentury Iranian book painting, e.g. the World History of Rashı#d al-Dı#n (Rice, D. T. [1976], pls. 15, 21, 34, and 69; there are also seven porphyry columns depicted in the Edinburgh manuscript), or the Great Mongol Sha¯hna¯ma (e.g. Pope and Ackerman, eds. [1938–39], pls. 836–37). 78

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 4A153D26 /production/pindar/booksxml/hillen/chapter−13−ends

404 early fourteenth century, however, indicate clearly enough—more than does thirteenth-century Arab painting80 —that it was the taste of the time to decorate them richly.81 But another possible source of inspiration for this architecture takes us right out of the Muslim sphere. Can one perhaps detect here a faint memory (transmitted through East Christian book painting, and therefore perhaps imperfectly understood) of the classical theater? Stage backdrops which distantly recall the scenae frons do, after all, often feature in Byzantine religious iconography.82 A lavish use of curtains, as here, was part of that tradition; so was the use of flanking figures, of perspectival recession and slanted, beveled projection, and of course elaborate architecture. Indeed, it was not rare for Byzantine portraits of evangelists or saints to make great play with a complex architectural setting,83 while in other Byzantine religious paintings such architecture is inhabited by a group of people.84 Yet the overall impression created by this fictive architecture is unmistakably Islamic. Clearly the decoration in the spandrels is welldeveloped tilework, and that at a time (and in a place) well before such work survives. Similarly, the columns rank as perhaps the most luxuriously decorated examples in thirteenth-century Arab painting. This is in harmony with the ebony and ivory of the gilded brass-plated balustrades, the curtains, the tilework, the fancy arch profiles, and the refined brickwork technique. Altogether an unmistakable message is being sent: the life of scholarship pays. Finally, what of the curtains? Their elaborate patterning, with pearling at the rims and lavish use of the three-dot motif, their rich coloring, dominated by gold and blue, and their crinkled folds all show that they served as a major constituent element in the architecture, their sumptuousness set off by the 80

Despite the occasional appearance of lightly decorated columns in Maqa¯ma¯t manuscripts of this period (Ettinghausen [1962], 79 [gold bands] and 93 [gold roundels]; Farès [1957], pl. Vb [bands at top and bottom of the shaft]). 81 See n. 79, above. 82 E.g. in the early fourteenth-century Kariye Cami in Istanbul (Grabar, A. [1953], 133), or in the Bible of Leo Sakallarios, Constantinople, c. 940 (Evans and Wixom, eds. [1997], 89). 83 For evangelists, see Friend (1927), figs. 95–98 (Mount Athos, Stauronikita, ms. 43, early tenth century); for a saint, see Evans and Wixom, eds. (1997), 4 (portrait of St. Gregor)’ from The Liturgical Homilies of St Gregory of Nazianzos, c. 1150). 84 Ibid., 108 (Homilies on the Life of the Virgin by the monk James, c. 1125–50); 111 (a liturgical roll of c. 1125–50); and 282 (Izbornik Sviatoslava, 1073).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 417C98F4 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

405

plainness of the brickwork. They served equally well to open up a space or to close it off, or to form an honorific canopy over the key figures. They curl luxuriantly around each of the outermost columns and with a fine rhetorical flourish they take the whole scene into the outer margins. This marginbreaking technique creates a certain spatial ambiguity and was later to develop into a means of bringing the action of the painting into the viewer’s own space and thus involving that viewer much more closely in what was being depicted. Here it precociously foreshadows the ingenious experiments with this device made by Persian painters in the fourteenth and fifteenth centuries.85 Their agitated folds and ruffles betray their expressive intent. The central gold and black curtain descends in curvilinear tiers to settle over the central figure in each painting, thereby singling him out from his colleagues. Perhaps this use of the curtain preserves memories of the chatr, the parasol used by those of high rank not only to furnish shade but also to single them out and to suggest that they enjoy a divinely bestowed legitimacy.86 At all events, these curtains—like the architecture of which they are an integral if ephemeral part —are used, as is their color (gold at the center and blue for the outlying sections) to assert a hierarchy. And while there is no serious attempt to depict the interior of this pavilion, so that the figures effectively float against a ground of pale yellow or gold, the architecture is not merely a backcloth for those figures but is properly integrated with them. Narrative So much, then, for the architecture. It is now time to consider the figures that inhabit it, and the story that they tell. Very few thirteenth-century frontispieces are as heavily populated as this one, with fourteen people in all, each of whom has his part to play That in itself is a major achievement, and it indicates how far Arab painting had traveled from the days when a single figure, that of the author, dominated the composition, as was the approved fashion in East Christian art. True, the Paris Maqa¯ma¯t double frontispiece has more people, but only two of them are singled out and the rest (angels excepted) constitute an amorphous audience with their backs to us. Here, by contrast, everyone is in the active mode; and in each leaf every person is

85 86

Brend (2000), 39–55; Hillenbrand, R. (1992). Sims (1973), 263; L’Orange (1953), 134–38.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 42FE2D5D /production/pindar/booksxml/hillen/chapter−13−ends

406 treated differently from the rest. Thus the sense of individuality is strong. It may be that the presence of precisely seven people in each leaf is an oblique reference to the fact that the Ikhwa¯n al-S.afa¯ were Isma¯ c#ılı#s; the tradition of using symbolic language was well established among Seveners generally. Alternatively, it may reflect classical precedent,87 as suggested by Pliny’s account of the Hebdomades of Marcus Varro (late first century B.C.), whose text, comprising biographies of seven hundred famous people, was illustrated by one hundred pictures of seven notables apiece. The obvious place to begin is with the principal figures, who effectively identify themselves by virtue of their larger size and by the fact that they occupy the central opening on each leaf. They are, incidentally, sumptuously dressed, their robes flashily lined with ochre-colored material. These are prosperous scholars, not starving schoolmen pauperized by their thirst for knowledge. Even their servants are well-dressed. This is an aristocracy of the intellect whose members are men of presence and dignity. It is noticeable that not one of them adopts the standard author pose so familiar from the Western and the East Christian traditions, namely standing or seated in a chair. Their poses are quintessentially Islamic: all are on the ground. One reclines, his back against a column; the rest are kneeling or seated with one knee raised. There is an implicit respect here for the man of learning, and the very process of acquiring, setting down, checking, and transmitting learning is faithfully reproduced. There is no way of telling for sure whether this double frontispiece was intended to be read from the left or from the right, and indeed both practices were followed, rather randomly it seems, in Islamic painting. In the left leaf, fol. 4 recto, several stages of scholarly activity are represented simultaneously. One sage, with a bushy white beard, reclines comfortably on a cylindrical yellow bolster, his back supported against a column. Presumably he has been meditating, though his snug, tightly curled pose might suggest that he has nodded off. If so, he is wide awake now and has obviously dreamt up some contribution to the debate, for he is prodding his neighbor in the side to attract his attention, and his other hand is outstretched in a gesture of speech. Nevertheless, his recumbent posture makes it reasonable to regard him as ‘The Thinker.’ An open book is balanced somewhat precariously on his wrist, as it were in limbo. 87 Weitzmann (1959), 116–20. On the connection between literary and pictorial cycles see idem (1970), 41.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 47A4A0D7 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

407

The two other figures that share center stage with him—for there is a strong sense of intellect on display here—are engaged in some sort of textual debate, to judge by the open book held by the younger man. His elder, who sports a long but trim pointed beard, puts across a cool objectivity. His rich, long brown-and-gold robe distinguishes him from his two flanking colleagues clad in blue. He does not turn to his interlocutor but instead looks diagonally out of the painting with a tranquil, faintly quizzical expression. His controlled, economical hand gesture suggests a take-it-or-leave-it authority. What a contrast to the black-bearded man, who in his eagerness seems to thrust the book towards his senior. This younger man is distressingly anxious to make his point: hence the pleading angle of his gaze, not to mention his popping eyes, tightly pursed protruding mouth, mobile eyebrows, and furrowed forehead. Scholarly controversy clearly awakens strong emotions in him. He is a direct descendant of the chaffering merchants and disputatious litigants of al-H.arı#rı# s Maqa¯ma¯t half a century earlier. Discussion, debate, controversy, teaching—there is something of all of these in the tense confrontation pictured here. A somewhat different sequence of scholarly activity meets the eye in the right frontispiece, the first picture in the book. Here attention focuses on the old man in the center. He is the key figure in this entire double frontispiece, marked out as he is by his rich brown robe worn above a yellow undershirt and decorated with a pattern of triple dots and striated with long vertical lines in gold. The two men flanking him wear blue and thus set him off visually. He alone of all fourteen figures does not wear a turban; instead, his robe covers his head to form a cowl (t.arh.a),88 as is common practice for prophets in Islamic painting.89 This, like the curtain terminating directly above his head, or the way that the flanking figures incline their heads towards him, confers unchallengeable authority upon him, and is a reminder—here, at the very beginning of the book—of its religious content. His words alone are being taken down by a man we must call ‘The Scribe,’ who balances his codex on his raised knee and writes with a calamus (fig. 28). His frown of anxious concentration, staring eyes, puckered eyebrows, and knotted forehead proclaim the importance of what the old man is dictating.

88

Cf. Simpson (1982), 99. Cf. Moses and Joshua in the Edinburgh Rashı#d al-Dı#n manuscript (Rice, D. T. [1976], pls. 10–11 and 14 respectively) and Muh.ammad in the Edinburgh al-Bı#ru¯nı# manuscript (Hillenbrand, R. [2000], color pl. 13). 89

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 43A579B7 /production/pindar/booksxml/hillen/chapter−13−ends

408 This is ‘The Lecturer.’ The speaking gesture of his right hand underlines this. He turns away from the scribe as if to maintain his concentration better. His face, as is the case with the other two old men, expresses the serenity conferred by wisdom. His colleague, distinguished by his spade beard, luxuriant moustache and long black hair escaping from his turban, clutches his raised knee while deeply absorbed in a half-open book; he is obviously ‘The Reader.’ This man, like the youth on the balcony above, has almond eyes of East Asian origin; the youth has in addition the rounded cheeks and tiny mouth which had long been shorthand for the moon-faced or Buddha-like beauty associated in the eastern Islamic world with East Asia.90 What is the point of depicting extra figures in the balcony? All four of them are engaged in intellectual pursuits: they read, whether from scrolls or books, or they listen intently to what is being recited below. In both leaves of the frontispiece one of the figures in the balcony casually rests his arm on the balustrade so that his sleeve billows down towards the main scene, thereby establishing a link with that scene that is both physical and intellectual. They could be interpreted as representing the general audience, young and of mature age alike, for whom these authors wrote. They bring a sense of the wider world to the rarefied intellectual activities taking place below them. Physically they are to some extent excluded from those activities, but their body language proclaims that they too partake of them. This idea was further developed in early fourtenth-century Persian painting. In the Great Mongol Sha¯hna¯ma, for instance, the windows in the upper stories of the palatial buildings are crammed with women who evince the keenest interest in what is going on below; one can almost hear their excited muttered commentary.91 Yet the diminutive scale of the men in the balconies, and even more of the servants on the ground floor vis-à-vis the authors themselves, reveals a crucial element of the Islamic world of books: the reverence for intellectual achievement. Many regarded the pen as mightier than the sword. In more senses than one, then, the authors of the Rasa¯’il are depicted as giants. Moreover, they are being waited on hand and foot. The servants function as a foil to the main actors, and their humble deportment and diminutive scale underline the importance of their masters. Like the men busily studying in the upper story, they are an innovation in frontispiece design. At one level 90 91

Melikian-Chirvani (1974), 34–37. Grabar, O. and Blair (1980), pls. 6 and 17.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 450113B8 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

409

they are space-fillers, and the fact that their space is so much narrower and more confined than that of the centerpiece in part accounts for this. But those same narrow spaces also have to accommodate the tempestuous undulations of the blue and gold curtains which, being looped, take over the upper portion of those spaces. That said, in the left-hand frontispiece the alert, watchful servant holding a brown book is tiny in comparison with his fan-wielding colleague at the far right, who rises to almost the full height of the opening in which he stands, and yet is crouched over the reclining scholar to whom he tends. And even this latter servant would in turn be dwarfed by the two seated scholars if they stood up. The discrepancies in size are thus not entirely consistent. In the right-hand frontispiece the two servants are again of pygmy stature. One, with a slightly furtive look on his face and his back turned to the viewer, seems about to sneak off for an unscheduled break.92 He has had enough of high-octane intellectual discourse. The other—paint loss has largely obliterated his face—approaches his masters with his hand raised, ready to help. Thus all four servants are differently engaged, another sign of that unobtrusive attention to detail which is the hallmark of this double frontispiece. Nevertheless, they are dressed in the same splendiferous robes and turbans as everyone else; nothing in their apparel betrays their modest status. Clearly the desire for a uniform visual tone overrode the demands of naturalism. Style Now that the major themes of this paper have been discussed, it remains to consider briefly the hallmarks of this painter’s style. He has a distinct preference for the play of line rather than for the bold blocking out of colors, though he does show an acute sensitivity to colors that complement each other, such as blue and gold. His figures naturally arrange themselves in sweeping, flamboyant curves and ovals—turbans, faces, thighs, shoulders, knees, ballooning sleeves. This painter dearly liked to abstract and simplify the human form, and he was fully alive to the expressive potential of these 92

For a comparably subversive vignette from Western medieval painting, depicting a student snoozing during a lecture, see L. de Voltolina, Liber ethicorum of Henricus de Allemania, showing the author in Arab dress haranguing his students, Bolognese school, later fourteenth century (Sievernich and Budde [1989], pl. 150). I am grateful to Dr. Stefano Carboni for this reference.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 460D2ED1 /production/pindar/booksxml/hillen/chapter−13−ends

410 curvilinear volumes. Here too the heritage of the Paris Maqa¯ma¯t is instantly recognizable.93 He rejoices in the rhetoric of a fluttering turban-end, the squiggles of its tailpiece, or the cascading folds and shooting sprays of a robe and its secondary internal patterning. These drapery folds employ the distinctive chrysography so familiar in Byzantine mosaics and enamels. Earlier Arab painting had used it confidently, though only on rare occasions, as in the frontispieces of the Kita¯b al-agha¯nı#.94 There the impact of Byzantine modes is plain to see, extending even to the choice of a deep ultramarine as a foil to the gold. Here, on the other hand, two complementary colors are employed—sky blue and gray-brown. The lines of the chrysography are thicker than in Byzantine art, and sometimes they acquire a different and more assertive texture altogether because the gold line is bracketed between two dark lines, for example in the delineation of turban folds.95 But it is less the technique whereby drapery folds are executed than their complexity which defines the mannerisms of this painter and furnishes a clue to his antecedents. For this obsessive fascination with the intricacies of the folds into which clothing, and in this case headgear as well, falls, fits neatly into a fashion that was already over a century old by this time and was destined to linger on into the fourteenth century. It was, moreover, a cross-cultural fashion, as likely to be encountered in Norman Sicily96 as in the Mamluk97 or Ilkhanid98 realms. The artists who experimented with it did so apparently for sheer love of ornament rather than using these convoluted folds as a vehicle for rendering emotion. They whip up a storm, no question about that, but it is a storm in a teacup. It should be no cause for surprise that artists working in a city that, until a generation before, had been the prime metropolis of the entire Islamic community should have been au courant with contemporary fashions in the world beyond.

93

E.g. the cupping scene, fol. 156 recto. Rice, D. S. (1953); for color images, see Ettinghausen (1962), 65 and von Folsach (2001), pl. 23. 95 There was a parallel contemporary fashion for this two-tone effect in luxury Qur’a¯ns (James [1988], 86, fig. 53). 96 Demus (1947), Pls. 64, 67–70, 73, 75, 79, 83–87, 101–5, and 109a. 97 Haldane (1978), 17, 20, and 32 and associated plates. 98 ¯ tha¯r al-ba¯qiya (Chronology of Ancient Nations); see Hillenbrand, As in al-Bı#ru¯nı# ’s al-A R. (2000), color pls. 5–6 and 12–14. 94

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 437D78A0 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

411

What is the role of color here? By the standards of Arab painting earlier in the century, the palette here is indeed reduced. The predominant tonalities are yellow in various shades, grayish brown, and blue, with touches of black and white to accentuate and even highlight key areas or to provide contrast. In chromatic terms, the two halves of the frontispiece mirror rather than complement each other, and this helps to pull together the entire two-page composition. So too does the device of using blue to create a horizontal accent along the top, middle, and bottom of each picture and thus right across the whole double-page spread. Overall, the effect of this particular choice of colors, especially the profusion of a dull gold, is to lend a touch of somber understated magnificence to the entire frontispiece. The awareness of other cultures and their artistic traditions is not limited to the drapery technique. The hairstyle of the beardless fan-waving youth, with his carefully barbered lock of hair strategically placed in front of his ear and tapering down to his chin, and the elegant thin arch of his eyebrow, are both in full accord with medieval Persian canons of beauty.99 Yet the two intensely animated figures to the left of each central personage are entirely within the local Iraqi tradition, complete with corrugated brow and pop-eyed stare. And the dusky faces of the servants in the right-hand frontispiece suggest that they are Indians.100 Other details tell a similar story. Three different types of halo can be seen in these two paintings. The commonest one, which recurs six times, is a large golden roundel with a double black rim. Next in popularity, with five examples, is a much smaller gold roundel embellished with discreet decorative curlicues on both rims along the central horizontal axis. On one occasion these curls sprout a long leafy tendril which snakes its way to the top of the halo. Such haloes, with their distinctive clustering of ornament in a few carefully chosen places, are already to be found in the Kita¯b al-agha¯nı# frontispieces,101 and probably derive from Buddhist sources.102 The third type of halo, which occurs only once, significantly around the head of the cowled figure who is probably meant to represent al-Maqdisı#, has a single thick black rim for the large gold roundel.

99 100 101 102

Nurbakhsh (1980). For parallel types, see Ettinghausen (1962), 108 and 121. Ibid., 65. For a general treatment of this theme, see Oh (2003).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 49A0F110 /production/pindar/booksxml/hillen/chapter−13−ends

412 Conclusion It is now time to summarize what claims to importance these leaves of the Rasa¯’il can make. They are legion. The discussion so far has shown that, even after the disastrous sack of Baghdad in 1258 and the fall of the Caliphate, Arab painting was still full of sap103 and that its artists had the mental capacity to reconfigure the millennial tradition of the frontispiece in new and unexpected ways. In the Rasa¯’il leaves the painter does so by allotting the architecture, which is now wholly contemporary, a far more active role than before, promoting it from a mere backdrop to a believable setting for intellectual activity and an ordering device for very literally putting people in their place and thereby asserting hierarchical distinctions. He also succeeds in developing, more creatively than ever before, the concept of multiple authorship, of intellectual teamwork,104 with due attention to the processes involved: thinking, reading, writing, discussing, dictating, and teaching. And all this is achieved within a credible built and human environment, rather than in an abstract vacuum. The effect is to humanize learning while maintaining intact its prestige, indeed its charisma. All this involves some quite radical experiments within the very conservative and artificially restricted domain of author portrait iconography, building too upon earlier (though distinctively different) Arab experiments with the double frontispiece. In style, too, these leaves are nothing less than a radical departure from earlier Arab painting. The principal innovations are in line and color. Figures and textiles in particular are for the most part (faces excepted) doubly outlined in black and gold. The infill of the draperies is a sumptuous swirl of gold sprays, striations, and ovals which infuse the entire painting with dynamism. As in the slightly later Marzuba¯nna¯ma, the combination of black, gold, and silvery white suggests that the painter was influenced by the

103

Cf. Simpson (1982), 93–94. This sense of teamwork is precisely what is missing in the group portraits of the early Byzantine period, as in the Vienna Dioscorides of 512 (Weitzmann [1977], pl. 16), in which each physician meditates in isolation, with no interaction with his peers. And while such interaction is indeed depicted in the little-known frontispiece (fig. 12) to the Kita¯b qawa¯’id al-ah.ka¯m fı# ma crifat al-h.ala¯l wa’l-h.ara¯m tas.nı#f (The book of the bases of the ordinances on the subject of the knowledge of what is licit and what is forbidden, compiled by [al-H . illı#]), probably of the early fourteenth century (Farès [1957], 87–93 and pls. VI–VII; cf. the review by Grabar, O. [1959], 225), it does not progress far beyond stark juxtaposition. 104

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 4681EBB4 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

413

tonality and the visual impact of inlaid metalwork.105 Another connection with early Persian painting is the absorbed attention which the artist devotes to the fussy intricacies of drapery, a theme which he takes to new heights of expressiveness, tapping into a clearly defined cross-cultural style that spans well over a century. The chromatic range is uncharacteristically muted for Arab painting, but the most striking colors in this limited palette are also the most expensive: lapis blue and gold, both of them carefully distributed over most of the painted surface. The emphasis on chrysography points to a continuing respect for the Byzantine tradition. The style of these leaves, then, confirms that Arab painting continued to develop after 1258, and did so within the lines laid down in the previous half-century or so, in everything from figural types to the clothing they wear or the buildings that they inhabit. It remained resolutely deaf to the siren call of Far Eastern art,106 whereas, at this very time, as the Paris History of the World Conqueror of Juwaynı# shows, Iran had fallen once and for all under its spell. In form the Rasa¯’il leaves continue the double frontispiece design with which the Arabs had experimented so creatively for almost a century. At first glance it seems to return to the simple repetition of the Paris Kita¯b al-dirya¯q of 1199. But the central horizontal band of the design, where most of its meaning is concentrated, develops numerous variations within a consistent uniformity of tone, and these variations challenge the viewer to look more closely. Next, for the first time Arabic script plays a major role in the layout. The name of the book and its author (here, authors) has been moved inwards from the title page. Moreover, the order of those two elements has been reversed, and not by accident either. The names of the authors are embedded in three lines of closely written small scribal naskh scarcely vocalized at all, at the top of fol. 3 verso, the first painted page. There is no attempt to highlight their names in any way. Yet the opposite page, fol. 4 recto, has the full title of the book calligraphed in large riqa¯ c script as a single banner headline, with every vocalization in place. There could be no clearer way of stating that the book was a team effort, and that teamwork matters more than individual achievement. The way that the script is used therefore teaches a moral lesson.

105

Cf. for example Ward (1993), 82, pl. 60. For a minor exception (the Oxford Maqa¯ma¯t of 738/1337), see Ettinghausen (1962), 152. 106

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 4152A4A6 /production/pindar/booksxml/hillen/chapter−13−ends

414 There is more. This double frontispiece, the swan song of the golden age of Arab painting, celebrates intellectual achievement rather than political power. It is the last double frontispiece to do so: an elegy, then, for the cultural achievement of a people destined to outlast their political eclipse. Its successors, which are mostly Iranian,107 sing the praises of the ruler and his pastimes. That is a sea change in the subject matter of Islamic painting, and it should not be suffered to go unrecorded. Bibliography Aga-Oˇglu, M., “Fragments of a Thirteenth-Century Mih.ra¯b at Nedjef ”, Ars Islamica II [1935], 128–31 Allen, T. A Classical Revival in Islamic Architecture (Wiesbaden, 1986) Barrucand, M., “Les representations d’architectures dans la miniature islamique en Orient du début du XIIIe au début du XIVe siècle”, Cahiers Archéologiques 34 (1986), 119–41 Barrucand, M., “Architecture et espaces architectures dans les illustrations des Maqa¯ma¯t d’al-H.arı¯rı¯ du xiiie siècle”, in R.Hillenbrand (ed.), The Art of the Saljuqs in Iran and Anatolia (Costa Mesa, CA, 1994), 79–88 Bothmer, H.-C. Graf von, “Architekturbilder im Koran. Eine Prachthandschrift der Umayyadenzeit aus dem Yemen”, Pantheon XLV (1987), 1–20 Brandenburg, D., Islamic Miniature Painting in Medical Manuscripts (Basle, 1982) Brend, B., “Beyond the Pale: Meaning in the Margin”, in R.Hillenbrand (ed.), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London, 2000), 39–55 Contadini, A., The Kita¯b na‘t al-h.ayawa¯n (Book on the characteristics of animals, British Library Or. 2784) and the “Ibn Bakhtishu¯‘” illustrated bestiaries (unpublished Ph.D. thesis, University of London, 1992) Delbrueck, R. Die Consulardiptychen und verwandte Denkmäler. Studien zur spätantiken Kunstgeschichte 2 (Berlin and Leipzig, 1929) Der Nersessian, S. Armenian Manuscripts in the Freer Gallery of Art (Washington, D.C., 1963) Djait, H., EI (2 ed.), V (Leiden, 1986), s.v. “al-Ku¯fa” Ettinghausen, R. Arab Painting (Geneva, 1962)

107 Neither Turkey nor India came close to matching the Iranian achievement in this field, a matter which would bear close examination.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 44F1C5B5 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

415

Ettinghausen, R., “On Some Mongol Miniatures”, Kunst des Orients III (1959), 44–65 Evans, H.C. and W.D.Wixom, The Glory of Byzantium. Art and Culture of the Middle Byzantine Era A.D.843–1261 (New York, 1997) Farès, B. “Philosophie et jurisprudence illustrées par les arabes. La querelle des images en Islam”, Mélanges Louis Massignon, (Damascus, 1957), 77–118 Farès, B., Le Livre de la Thériaque. Manuscrit arabe à peintures de la fin du XIIe siècle conservé à la Bibliothèque Nationale de Paris (Cairo, 1953) Folsach, K. von, Art from the World of Islam in the David Collection (Copenhagen, 2001) Friend, A.M., Jr., “The Portraits of the Evangelists in Greek and Latin Manuscripts”, Art Studies 5 (1927), 115–47 Friend, A.M., Jr., “The Portraits of the Evangelists in Greek and Latin Manuscripts, II”, Art Studies 7 (1929), 3–29 Goitein, S.D., EI (2 ed.), V (Leiden, 1986), s.v. “al-K.uds” Golombek, L., “Safaviden”, in B.Spuler and J.Sourdel-Thomine, Die Kunst des Islam. Propyläen Kunstgeschichte 4 (Berlin, 1973), 348–58 Grabar, A., Byzantine Painting, tr. S.Gilbert (Geneva, 1953) Grabar, O., review of Farès, “Philosophie”, in Ars Orientalis III (1959), 225–6 Grabar, O., “Survivances classiques dans l’art de l’Islam”, Annales Archéologiques Syriennes 21 (Proceedings of the 9th International Congress of Classical Archaeology) (1971), 371–80 Grabar, O. and S.Blair, Epic Images and Contemporary History. The Illustrations of the Great Mongol Shahnama (Chicago, 1980) Gray, B., Persian Painting (Geneva, 1961) Grube, E.J., “Materialien zum Dioskurides Arabicus”, in R.Ettinghausen (ed.), Aus der Welt der Islamischen Kunst. Festschrift für Ernst Kühnel zum 75. Geburtstag am 26.10.1957 (Berlin, 1959), 163–94 Haldane, D., Mamluk Painting (Warminster, 1978) Hamid, I.S., Mesopotamian School and the Place of Painting in Islam (unpublished Ph.D. thesis, University of Edinburgh, 1966) Hillenbrand, C., EI (2 ed.), VII (Leiden, 1993), s.v. “al-Mustans. ir” Hillenbrand, R., “The Flanged Tomb Tower at Bast. a¯m”, in C. Adle (ed.), Art et Société dans le Monde Iranien (Paris, 1982), 237–60 Hillenbrand, R., “The Classical Heritage in Islamic Art: The Case of Medieval Architecture”, The Scottish Journal of Religious Studies 7 (1986), 123–40

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 4096BEF4 /production/pindar/booksxml/hillen/chapter−13−ends

416 Hillenbrand, R., “The Uses of Space in Timurid Painting”, in L.Golombek and M.Subtelny (eds.), Timurid Art and Culture. Iran and Central Asia in the Fifteenth Century (Leiden, 1992), 76–102 Hillenbrand, R., “Images of Muhammad in al-Biruni’s Chronology of Ancient Nations”, in R.Hillenbrand (ed.), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London and New York, 2000), 129–46 Hoffman, E.R., “The Author Portrait in Thirteenth-Century Arabic Manuscripts: A New Islamic Context for a Late-Antique Tradition”, Muqarnas 10. Essays in Honor of Oleg Grabar contributed by his students (Leiden, 1993), 6–20 Hoffman, E.R.F., The Emergence of Illustration in Arabic Manuscripts: Classical Legacy and Islamic Transformation (unpublished Harvard University Ph.D. thesis, 1982) Ibn Z.afar, Sulwa¯n al-Mut. a¯‘ fı¯ ‘Udwa¯n al-At. ba¯‘, colour facsimile edition (Kuwait, 1985) Jacob, G., Geschichte des Schattentheaters im Morgen- und Abendland (2nd enlarged ed., Hannover, 1925) Kahle, P., “The Arabic Shadow Play in Medieval Egypt (Old Texts and Old Figures)”, Journal of the Pakistan Historical Society II (1954), 85–97 Kahle, P., “Islamische Schattenspielfiguren aus Ägypten”, Der Islam I (1910), 264–99 and II (1911), 143–95 Kahle, P., “The Arabic Shadow Play in Egypt”, Journal of the Royal Asiatic Society (1940), 21–34 Komaroff, L., “Paintings in Silver and Gold: The Decoration of Persian Metalwork and Its Relationship to Manuscript Illustration”, Studies in the Decorative Arts II/1 (1994), 2–34 Kühnel, E., “Dated Persian Lustred Pottery”, Eastern Art III (1931), 221–36 Lentz, T.W. and G.D.Lowry, Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century (Los Angeles, 1989) L’Orange, H.P., Studies on the Iconography of Cosmic kingship in the Ancient World (Instituttet for Sammenlignende Kulturforskning, series A: Forelesninger, 23 (Oslo, 1953) Mahboubian, M., Treasures of Persian Art after Islam (New York, 1970) Marquet, Y., EI (2 ed.), III (Leiden, 1979), s.v. Ikhwa¯n al-Safa¯’ Melikian-Chirvani, A.S., “L’évocation littéraire du bouddhisme dans l’Iran musulman”, Le Monde Iranien et l’Islam II (1974), 1–72

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 42B4D234 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

417

Melikian-Chirvani , A.S., “Le Sha¯h-na¯me, la gnose soufie et le pouvoir Mongol”, Journal Asiatique CCLXXII/3–4 (1984), 249–337 Melikian-Chirvani, A.S., Sulwa¯n al-Mut. a¯‘ fı¯ ‘Udwa¯n al-Atba‘. A rediscovered masterpiece of Arab literature and painting (Kuwait, 1985) Moreh, S., “The Shadow Play (Khayal al-Z.ill) in the Light of Arabic Literature”, Journal of Arabic Literature XVIII (1987), 46–61 Moreh, S., Live Theatre and Dramatic Literature in the Medieval Arabic World (Edinburgh, 1992) Mullett, M. and R. Scott (eds.), Byzantium and the Classical Tradition: University of Birmingham Thirteenth Spring Symposium of Byzantine Studies, 1979 (Birmingham, 1981) Nees, L., The Gundohinus Gospels (Cambridge, Mass., 1987) Nurbakhsh, J., What the Sufis Say (New York, 1980) O’Kane, B., Early Persian Painting. Kalila and Dimna Manuscripts of the Late Fourteenth Century (London, 2003) Oh, L.J., “The East Asian Characteristics of Ilkhanid Royal Manuscripts”, Persica 19 (2003), 69–105 Pedersen, J., The Arabic Book, tr. G.French, ed. R.Hillenbrand (Princeton, 1984) Pope, A.U., and P.Ackerman (eds.), A Survey of Persian Art from Prehistoric Times to the Present (London and New York, 1938–9) Prentice, V.R., The Illustration of Sa‘di’s poetry in fifteenth-century Herat (unpublished Harvard University Ph.D. thesis, 1977) Rice, D.S., The Unique Ibn al-Bawwa¯b Manuscript in the Chester Beatty Library (Dublin, 1955) Rice, D.T., ed. B.Gray, The Illustrations of the ‘World History’ of Rashid al-Din (Edinburgh, 1976) Richard, F., Splendeurs persanes. Manuscrits du XIIe au XVIIe Siècle (Paris, 1997) Robb, D.M., The Art of the Illuminated Manuscript (Cranbury, N.J. and London, 1973) Rogers, J.M., “A Renaissance of Classical Antiquity in North Syria (11-12 Centuries)”, Annales Archéologiques Syriennes 21 (Proceedings of the 9th International Congress of Classical Archaeology) (1971), 347–61 Schmid, H., Die Madrasa des Kalifen al-Mustansir in Baghdad. Eine baugeschichtliche Untersuchung der ersten universalen Rechtshochschule des Islam. Mit einer Abhandlung über den sogenannten Palast in der Zitadelle in Baghdad (Mainz, 1980) Seherr-Thoss, Sonia P. and H.C.Seherr-Thoss, Design and Color in Islamic Architecture. Afghanistan. Iran, Turkey (Washington, D.C., 1968)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 3D62AF8F /production/pindar/booksxml/hillen/chapter−13−ends

418 Sievernich, G., and H.Budde, Europa und der Orient, 800–1900 (Berlin and Gütersloh, 1989) Simpson, M.S., “The role of Baghdad in the formation of Persian painting”, in C.Adle (ed.), Art et Société dans le Monde Iranien (Institut Français d’Iranologie de Téhéran. Bibliothèque Iranienne No.26. Editions Recherche sur les civilisations. Synthèse No.9) (Paris, 1982), 91–115 Sims, E.G., The Garrett Manuscript of the Zafar-Name. A Study in Fifteenth-Century Timurid Patronage (unpublished Ph.D. dissertation, New York University, 1973) Smith, E.B., Architectural Symbolism of Imperial Rome and the Middle Ages (Princeton, 1956) Spuler, B., and J.Sourdel-Thomine, Die Kunst des Islam. Propyläen Kunstgeschichte 4 (Berlin, 1973) Stern, S.M., “New information about the authors of the “Epistles of the Sincere Brethren””, Islamic Studies III/4 (1964), 405–28 Stern, S.M.,, “The authorship of the epistles of the Ikhwan-as-Safa’”, Islamic Culture 20 (1946), 367–72 Stern, S.M., “Additional Notes to the Article ‘The authorship of the epistles of the Ikhwan-as-Safa’” Islamic Culture 21 (1947), 403–4 Stewart, D., Early Islam (Weert, repr. 1975) Vesel, Z., “Science and Scientific Instruments”, in C.Parham (ed.), The Splendour of Iran. Volume III. Islamic Period (London, 2001), 260–99 Ward, R.M., Islamic Metalwork (London, 1993) Ward, R.M., “Painted book illustration, c. 1250 – c. 1500: Egypt and Syria”, in The Dictionary of Art, ed. J.Turner (London, 1996), 16, 310b-313a Watson, O., “The Masjid-i ‘Alı¯, Quhru¯d; an architectural and epigraphic survey”, Iran XIII (1975), 59–74 Watson, O., Persian Lustre Ware (London, 1985) Weitzmann, K., [ed.], Age of Spirituality. Late Antique and Early Christian Art, Third to Seventh Century (New York, 1979) Weitzmann, K., Ancient Book Illumination (Cambridge, Mass., 1959) Weitzmann, K., Studies in Manuscript Illustration. Illustrations in Roll and Codex. A Study of the Origin and Method of Text Illustration (Princeton, 2nd ed., repr. 1970) Weitzmann, Late Antique and Early Christian Book Illumination (London, 1977) Wessel, K., (tr. I.R.Gibbons), Byzantine Enamels from the 5th to the 13th Century (Shannon, 1969)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 104 OUTPUT: Tue Aug 10 10:29:55 2010 SUM: 058B92F7 /production/pindar/booksxml/hillen/chapter−13−ends

DOUBLE FRONTISPIECE TO THE EPISTLES OF THE SINCERE BRETHREN

419

Wilber, D.N., The Architecture of Islamic Iran. The Il-Khanid Period (Princeton, 1955)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-13-ends Sequential 17

JOBNAME: ???? PAGE: 1 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 36BA9AD2 /production/pindar/booksxml/hillen/chapter−14

XIV The Shahnama and the Persian Illustrated Book1

T

HIS chapter will begin with a substantial introduction that deals with the general issue of the illustrated epic and the effect on it of a tradition of oral performance. While illustrated epics in other cultures are not unknown – the Ambrosiana Iliad and Mughal Ramayana manuscripts are cases in point – the huge corpus of illustrated Shahnama (‘Book of Kings’) manuscripts is a thoroughly atypical phenomenon in the context of world literature; the instability of the text of the Shahnama has further impacted on that tradition. After a brief discussion of these issues, attention will focus on three themes that bear on the role of the illustrated book in medieval Iran. The first is the Persian renaissance of the tenth century. This was a backlash against the forcible imposition of Arab language and culture, a backlash in some senses both spearheaded and symbolized by the Shahnama, a heroic epic of some fifty to sixty thousand couplets (the uncertainty is due to the difficulty of establishing an authentic text) by the eastern Iranian poet Firdausi, completed c. 1010 CE. The second theme deals with the reasons for the perennial popularity of the Shahnama both within Iran and throughout a much wider cultural sphere. The final theme concerns the way that artists responded to the challenge of depicting the stories of the national epic. Here the discussion will focus on the treatment of a figure who, though

1

Given the global scope of the present volume, I have opted to do without the diacritical marks required for full transliteration of the kind aimed at a specialist audience. Special thanks go to Ian Willison, who has taken a lively interest in this paper from the outset and has made many useful suggestions for its improvement.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 514D594D /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

421

world-famous, is not generally familiar in his Persian guise, namely Alexander the Great, who has maintained his charisma intact to this day over improbable gulfs of space and time. Some introductory remarks will be in order. It is perhaps natural for an art historian to approach the world of books in Muslim culture through the medium of illustrations. But books with pictures were the exception, not the rule, in medieval Islamic society, where literacy was widespread and was further encouraged by numerous public libraries. Books were big business; let us recall that the legendary library of the tenth-century caliph al-Hakam II at Cordoba allegedly contained some four hundred thousand volumes2 at a time when the great monastic libraries in the West had holdings better measured in the hundreds than the thousands. This truly startling contrast had much to do with the Islamic preference for producing books via dictation to large groups rather than via one-to-one copying.3 Even illustrated books were apt to be created in this way.4 Moreover, the continuing commercial dynamism of the book trade throughout most of the Islamic middle ages and in many of the major Muslim centres ensured that even illustrated books (among which, in Iran, the Shahnama was predominant) were at times produced in substantial numbers for a market of the educated and mercantile elite.5 Such illustrated books were overwhelmingly secular in character. 2

See al-Maqqari, The History of the Mohammedan Dynasties in Spain, tr. P. de Gayangos (London: Royal Asiatic Society of Great Britain and Ireland, 1840 and 1843; repr. London and New York: Routledge Curzon, 2002), I, 139, 418, n.I and Appendix C, p. xl. Even greater numbers are bandied about in the context of the great libraries of the eastern Islamic world, such as that of the Lord of Amida/Diyarbakir in eastern Turkey in 1183, which allegedly ran to 1,040,000 volumes (Y. Eche, Les Bibliothèques arabes publiques et semi-publiques en Mésopotamie, en Syrie et en Égypte au moyen âge [Damascus, 1967], p. 251). 3 For a detailed account of the process, see J. Pedersen, The Arabic Book, tr. G. French, ed. by R. Hillenbrand (Princeton: Princeton University Press, 1984), pp. 28–36; M. G. S. Hodgson, The Venture of Islam: Conscience and History in a World Civilisation (Chicago and London: University of Chicago Press, 1974), II, 444–5. 4 For a tentative attempt to explain how this might have worked in practice, see R. Hillenbrand, ‘New Perspectives in Shahnama Iconography’, in Shahnama: The Visual Language of the Persian Book of Kings, ed. by R. Hillenbrand (Aldershot: Ashgate, 2004), pp. 2–4; cf. M. S. Simpson, ‘The Pattern of Early Shahnama Illustration’, in Studium Artium Orientalis et Occidentalis. I: Problems in the Relation Between Texts and Illustration (1982), p. 48. 5 A tentative attempt to identify likely groups of patrons for the lesser Ilkhanid Shahnamas is given by M. S. Simpson, The Illustration of an Epic: The Earliest Shahnama Manuscripts (New York and London: Garland Publishing, 1979), pp. 318–21. See also M. L.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4C69E740 /production/pindar/booksxml/hillen/chapter−14

422 As with the epics of certain other cultures, fairly standardized copies of the Shahnama were produced in quantity, and this occurred from the fourteenth century onwards. But there was one key difference between the epic in Iran and the epic elsewhere. In addition to the adoption of certain formulae — for page layout, illuminated captions or chapter headings, and types of hand — such as operated in other, non-Islamic, traditions in which multiple copies of certain epics were made, mass production in the case of the Shahnama often had the added dimension of scores of pictures. Mass production is of course a relative concept here. Such a book was still a luxury object. And since most surviving illustrated Shahnamas were made for educated people to buy, the issue of whether the illustrations were intended merely as pictures — what one might term optional extras — or as something more ambitious, a parallel visual exegesis, in which the painter explores some of the implications of the text, and thus is apt to go off at a tangent from it, poses itself with some insistence.6 But there were certain constraints to be obeyed here. While no two Shahnamas have exactly the same iconographic programme, certain scenes became standard fare for the illustrator,7 and custom constrained their composition and format by ever narrower parameters. In this process one may again recognize the influence of the mass market, with its tendency towards uniformity, and a consequent saving of time and therefore money. There was, in short, a conveyor belt in operation, and artists did not scruple

Swietochowski and S. Carboni, with essays by A. H. Morton and T. Masuya, Illustrated Poetry and Epic Images: Persian Painting of the 1330s and 1340s (New York: Metropolitan Museum of Art, 1994), pp. 50–1, where Morton tentatively suggests that the patron of the Mu’nis al-Ahrar was an Isfahani notable. For further discussion about the identity of the patrons of the slightly earlier Arabic Maqamat manuscripts, see O. Grabar, ‘The Illustrated Maqamat of the Thirteenth Century: the Bourgeoisie and the Arts’, in The Islamic City, ed. by A. H. Hourani and S. M. Stern (Oxford: Cassirer, 1965), pp. 214–18; and O. Grabar, The Illustrations of the Maqamat (Chicago and London: University of Chicago Press, 1984), pp. 149 and 182, n.80. 6 See O. Grabar, ‘Pictures or commentaries: The Illustrations of the Maqamat of al-Hariri’, in Studies in Art and Literature of the Near East in Honor of Richard Ettinghausen, ed. by P. J. Chelkowski (Salt Lake City: Middle East Center, University of Utah, 1974), pp. 85–104. 7 This becomes very plain as one studies the only attempt made so far in written form to list the subject matter of Shahnama illustrations: J. Norgren and E. Davis, Preliminary Index of Shah-Nameh Illustrations (Michigan, 1969; stenograph), a listing of some 4000 paintings. The Cambridge-Edinburgh database lists substantially more than twice as many.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 52DB7596 /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

423

to repeat themselves, as becomes increasingly evident by the fifteenth century. This repetition makes itself felt partly in the way that illustrations are limited to a small group of especially popular episodes from the poem, and partly in the use of formulaic compositions which by minor tweaking can be made to fit a variety of episodes.8 This process is quite distinct from the deliberate copying of earlier paintings as an act of homage or as a way of demonstrating skill. The market did not encourage painters to spend three months on an image when three days (or less) would do. This key role of commerce, rather than princely patronage, was the factor which ensured that the illustrated Shahnama flourished throughout the Persian-speaking world, from Turkey to India, Kashmir and central Asia. The implications of that fact have been slow to percolate into modern scholarship. Even manuscript copies of Firdausi’s text, whether illustrated or not, by no means give us the whole picture, for no account of the impact of the Shahnama could afford to ignore the impact of the epic as an oral text. (The Qur’an, it will be remembered, also has both a literary and an oral tradition.) As was the case with the Homeric epics,9 there existed an elite reading public,10 but also a mass audience for the performed text of the epic,11 so that it became a remarkably familiar text at all levels of society. The rhapsode (naqqali) is still alive and well in Iran;12 and before the coming of radio and television, apart from the ta‘ziya or Passion play, which was religious and performed only in the sacred month of Muharram,13 the principal form of

8 For this practice, see A. Adamova, ‘Repetition of Compositions in Manuscripts: the Khamsa of Nizami in Leningrad’, in Timurid Art and Culture: Iran and Central Asia in the Fifteenth Century, ed. by L. Golombek and M. Subtelny (Leiden, New York and Köln: Brill, 1992), pp. 67–75. 9 See the chapter by Christopher Carey in this volume. 10 In this context it is important to remember that most surviving Shahnamas are not illustrated. Thus it could not be argued that the images are integral to the text or that the owner of a Shahnama without illustrations was missing something essential. 11 See K. Yamamoto, The Oral Background of Persian Epics: Storytelling and Poetry (Leiden and Boston: Brill, 2003), pp. 20–5; M. E. Page, Naqqali and Ferdowsi: Creativity in the Iranian National Tradition (unpublished Ph.D. thesis, University of Pennsylvania, 1977, pp. 19–20, 35 and 39–47); and Page, ‘Professional Storytelling in Iran: Transmission and Practice’, Iranian Studies, 12 (1979), 195–215. 12 P. Briant, Darius dans l’ombre d’Alexandre (Paris: Fayard, 2003), pp. 501–2 and fig. 49. 13 See Ta‘ziyeh: Ritual and Drama in Iran, ed. by P. J. Chelkowski (New York: New York University Press, 1979). There are in fact certain thematic links between the ta‘ziya and the Shahnama, such as the challenge declaimed between two champions (P. Mamnoun, ‘Ta‘ziyeh

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4BE6566E /production/pindar/booksxml/hillen/chapter−14

424 public entertainment throughout the country was the coffee-house recitation of the Shahnama. This persistent tradition of public performance, when considered alongside the relatively late date of the earliest surviving written version of the text (the copy in Florence dated 1217)14 begs the question of how reliable the text is. Here an example from quite another culture may be helpful. One theory about the Homeric epics is that the texts survived for some two centuries (possibly longer) and with minimal changes, in oral form, recited by transmitters who were not significantly creative artists. But before the text crystallized, c. 150 BCE, in more or less the form we now know, it acquired – as the Homeric papyri datable up to the early third century BCE show – many so-called ‘wild lines’ which were themselves perhaps the result of continued performances, a kind of secondary orality. There are clearly arresting parallels here with the preservation of Firdausi’s text and with the role of orality in that process. In the context of an illustrated epic they are particularly relevant, for they suggest the likelihood that a passage of spurious text, or a parallel text from another source, could very well generate illustrations.15 It is time now to turn to the question of images. Let us dispose at the outset of the preconception that the Qur’an forbids the making of images. It does not. True, clerics have squabbled inconclusively over the issue for centuries,16 and there was indeed a widespread distaste for figural images in a religious context. But no similar aversion makes itself felt in the secular milieu at large. In fact, the illustrated book in the Islamic world has a history that can be traced back for a millennium on the evidence of surviving codices alone, and textual references take the story back another two hundred years.17 This lengthy time-frame is liable to conceal the fact that the floruit of from the Viewpoint of Western Theatre’, in Chelkowski, p. 166) and the elegy (Z. Eqbal, ‘Elegy in the Qajar Period’, in Chelkowski, p. 194). 14 On this manuscript, see A. M. Piemontese, ‘Nuova luce su Firdawsi: uno ‘Šah-nama’ datato 614 H./1217 a Firenze’, Annali dell’Istituto Orientale di Napoli, 40 (1980), 1–38 and 189–242. 15 See n. 9 above; and the remarks on oral transmission and its impact on the illustrative programme in M. S. Simpson, ‘The Pattern of Early Shahnama Illustration’ (note 4 above), 46–8; and M. S. Simpson, ‘The Narrative Structure of a Medieval Iranian Beaker’, Ars Orientalis, XII (1981), 15. 16 The classic discussion is that by K. A. C. Creswell, ‘The Lawfulness of Painting in Early Islam’, Ars Islamica, XI-XII (1946), 159–66. 17 Sir Thomas W. Arnold, Painting in Islam: A Study of the Place of Pictorial Art in Muslim Culture (Oxford: Clarendon Press, 1928), pp. 52-66.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 50498E72 /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

425

this art form was relatively short, extending as it did from c. 1200 to c. 1600. In that period, however, it established itself as one of the major genres of Islamic art, and – still more important – as the major conduit for painting within that culture. Frescoes were always of secondary importance. These circumstances give book painting a much greater role in the Islamic lands than it had in the West, which is the obvious benchmark because Europe like the Islamic world favoured the codex above other forms of book. Europe, by contrast, sank its energies into altarpieces and easel painting – forms effectively unknown to Muslims – and into frescoes, with illustrated books trailing (rightly or wrongly) far behind in the hierarchy of western art history. In Iran, while royalty and the aristocracy provided over the centuries the patronage required for luxury illustrated books, there was also (as in Europe; one has only to think of fourteenth-century Paris) a vibrant market at a much lower level, and it is this commercial market that accounts for the majority of the illustrated Persian books that have survived. The themes of those books even made their way to other media, notably ceramics.18 The briefest glance at the history of book painting in the Islamic world makes it plain that it is Iran – not the Arab world, not Turkey, not India – that has the longest continuous tradition, as well as the richest, most varied and most inventive one, in this medium. The prime vehicle for that inventiveness was a single text which was illustrated far more than any other text in the entire canon of Islamic painting: Firdausi’s Shahnama. The history of Persian painting, in other words, can be traced by means of the illustrations of this one text.19 Nearly four hundred illustrated copies of this text, dating from c. 1300 to c. 1900, and produced in the lands between Egypt and India, 18 For the general issue in the medium of ceramics, see R. Hillenbrand, ‘The Relationship between Book Painting and Luxury Ceramics in thirteenth-century Iran’, in The Art of the Saljuqs in Iran and Anatolia, ed. by Robert Hillenbrand (Costa Mesa, Cal.: Mazda Publishers, 1994), pp. 134–45; for detailed discussion of specific examples, see B. Schmitz, ‘A Fragmentary Mina’i Bowl with Scenes from the Shahnama’, ibid., pp. 156–64 and Simpson, ‘Narrative Structure’; for a related example in metalwork, see S. J. Auld, ‘Characters out of Context: The Case of a Bowl in the Victoria and Albert Museum’, in Hillenbrand (ed.), Shahnama, pp. 99–116. The subject of Shahnama scenes in both ceramics and metalwork is covered in M. S. Simpson, ‘Narrative Allusion and Metaphor in the Decoration of Medieval Islamic Objects’, in Pictorial Narrative in Antiquity and the Middle Ages, ed. by M. S. Simpson and H. L. Kessler, special issue of Studies in the History of Art, 16 (1985), 131–49. 19 See B. W. Robinson, ‘Persian Painting and the National Epic’, Proceedings of the British Academy, LXVIII (1982), 275–97.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 47806A3A /production/pindar/booksxml/hillen/chapter−14

426 between Turkey and central Asia, survive. No other text can marshall anything like this total.20 It is likely, moreover, that these represent a mere fraction of what was originally created. But these survivals are more than enough to create a critical mass that permits scholars to develop a variety of approaches to how this text was illustrated.21 A five-year project (1999–2004), sponsored by the Arts and Humanities Research Board (AHRB) and carried out at the universities of Cambridge and Edinburgh, has resulted in the creation of an illustrated database in which these manuscripts are catalogued and where information about the relation between text and image, the precise location of a given painting in a given manuscript, the changing interpretations of each episode over the years, and the nature of iconographic cycles, can easily be accessed. The Princeton database, containing 277 illustrations from 5 manuscripts, with some supplementary material, was a first attempt in this direction (www.princeton.edu/–shahnama/). Now for the first theme, the cultural and political context within which the Shahnama was created, a context of burgeoning national sentiment. This was itself a long-delayed reaction to the trauma of the Arab conquest of Iran between 637 and 651. Of course it would be an exaggeration to maintain that Firdausi’s purpose in writing the Shahnama was to counter an ignominious Arab present with a glorious Persian past. But in fact it has seldom been the fate of a single book so to offset actual political defeat by a virtual victory in the mind, so there is more than a grain of truth in such a premise. Let us begin by considering the cultural setting. Firdausi gave classic shape to his countrymen’s wistful memories of a world without Arabs and Arabic, and even without Islam. In so doing he reversed the humiliation endured nearly four centuries previously by a proud and ancient civilization at the hands of lizard-eating22 desert nomads. Just before the fateful battle of Qadisiyya in

20 Perhaps the closest rival is the Khamsa of Nizami; but the only comprehensive attempt so far to list the surviving miniatures of this text conflates single miniatures with complete manuscripts, and thus the number of parent manuscripts is hard to determine. See L. N. Dodkhudoeva, Poemi Nizami v Srednevekovoi Miniatiyurnoi Zhivopisi (Moscow: Izdatel’stvo ‘Nauka’, 1985); her index (pp. 288–304) lists 245 items. 21 For a judicious assessment and a comprehensive survey of work on these themes since 1975, see M. S. Simpson, ‘Shahnama as Text and Shahnama as Image: A Brief Overview of Recent Studies, 1975–2000’, in Hillenbrand (ed.), Shahnama (note 4 above), pp. 9–23. 22 This is a familiar gibe; see I. Goldziher, Muslim Studies, ed. S. M. Stern, tr. C. R. Barber and S. M. Stern (London: Allen and Unwin, 1967), I, 152.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 52069CDB /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

427

637, which opened Iran to the Arab invaders, the Sasanian Persian generalissimo Rustam met his Arab counterpart, al-Mughira. ‘We have the kingship,’ he boasted; but was silenced by the retort, ‘But we have the religion.’23 If it didn’t happen that way, it should have, for that dichotomy (which resonates in Firdausi’s own account of the diplomatic preliminaries to the battle) crystallizes the perennial conflict between these two cultures. That conflict took literary as well as military shape in the so-called Shucubiyya, the ninth-century ‘battle of the books’ in which Islamized Persians – writing, let us note, in Arabic – disputed intellectual supremacy in literature and philosophy with the Arabs;24 and that conflict can still be sensed in the Shahnama, admittedly in a more implicit rather than explicit manner. The illustrations evoke the vanished glamour of Persian kingship just as they suggest nothing of Islam: in that sense they belong to the past rather than the present. For the Arabs brought their religion, and with it their language, to all the countries that they conquered. The Qur’an contained the very words of God and could therefore not be translated; and so its language, and often the calligraphy whereby it was transcribed, took on a numinous character. Thus Arabic readily became the lingua franca of the world of Islam, not only for the religious sciences such as Qur’anic commentary, the traditions of the Prophet and law, but also for philosophy, science and even literature. In the process it inevitably exerted a repressive effect on the high languages that it supplanted.25 Middle Persian, the language in which

23 There are numerous versions recounted in al-Tabari of the verbal exchanges between the Arabs and the Persians at this time; see The History of al-Tabari, vol. 12 (Albany: State University of New York Press, 1989). For a condensed narrative see J. B. Glubb, The Great Arab Conquests (London: Hodder and Stoughton, 1963), pp. 193–202. For the tendency of the Arab accounts of the early conquests in Iraq to inflate and conflate events, see F. M. Donner, The Early Islamic Conquests (Princeton: Princeton University Press, 1981), pp. 195–9. 24 H. A. R. Gibb, ‘The social significance of the Shu‘ubiyya’, in his Studies in the Civilization of Islam (Boston: Beacon Press, 1962), pp. 62–3; R. Mottahedeh, ‘The Shu‘ubiyah controversy and the social history of early Islamic Iran’, International Journal of Middle East Studies, VII (1976), 161–82; and Encyclopaedia of Islam (2nd edn), IX (Leiden: Brill, 19977), s.v. ‘al-Shu‘ubiyya’, cols. 513–16 (S. Enderwitz). 25 That did not always imply a lack of respect for other cultures. Goldziher quotes Diya’ al-Din Ibn al-Athir (1163–1239; brother of the famous historian), whom he terms ‘the best aesthetic critic of Arabic literature’, as follows: ‘I found that in [one point] the Arabs are outdone by the Persians. Persian poets write poetical books which from beginning to end contain well-ordered descriptions of stories and events and which move in the

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4DF66234 /production/pindar/booksxml/hillen/chapter−14

428 manuals of statecraft and administration had been composed, was no exception. Thus the Persian sense of disfranchisement was acute. Their empire, which under the Achaemenids, Parthians and Sasanians had dominated much of central and western Asia for most of the previous millennium and more, had gone with the wind. So much for the cultural context. What of the political situation? The Arabs had settled en masse in large tracts of Iranian territory, displacing the local landed aristocracy. But their loyalties were naturally directed westwards, to the Arab caliph, first in Damascus and then in Baghdad. Those caliphs, however, found it increasingly difficult to control the centrifugal tendencies of an empire stretching from the Pyrenees to the Pamirs, from central France to the borders of China, and it was perhaps inevitable that such an empire should begin to fray at the edges. Spain broke away in the west in 756, and in 827 it was the turn of the Iranian east, when the caliph gave one of his most trusted governors hereditary control there. While the outward forms and symbols of loyalty to the caliph – such as the mention of his name on the coinage and at the Friday service – were still strictly observed, de facto independence was only a matter of time. Indeed, occasional rebellions against the central power erupted in this area in the ninth century. The most serious was that of the Saffarids, whose leader, one Yacqub the Coppersmith, is presented by one of his panegyric poets, Ibrahim ibn Mamshadh, as possessing the imperial Iranian banner, known as the banner of Kaveh.26 This is a calculated reference to one of the popular heroes of the Shahnama, the blacksmith Kaveh, who hoisted his leather apron as a standard to rouse the common people against the tyrant Zahhak.27 That same poem, written in the 870s, vaunts Yacqub, a man of lowly birth, as a descendant of the highest levels of the eloquence of the national language. Thus, for example, al-Firdawsi wrote his book Shahnama in 60,000 lines; it contains the whole history of the Persians and is the Koran of the nation, since their most important rhetoricians are in agreement that there is nothing in their literature to excel this work in elegance. There is nothing comparable in the Arabic language despite its wealth and versatility, and despite the fact that the Persian language is but a drop in the sea in comparison with it.’ Goldziher adds ‘In other words: the Persians excel the Arabs by having an epical literature which the latter lack’ (Muslim Studies, I, 160). 26 S. M. Stern, ‘Yacqub the Coppersmith and Persian National Sentiment’, in Iran and Islam: In Memory of the late Vladimir Minorsky, ed. by C. E. Bosworth (Edinburgh: Edinburgh University Press, 1971), pp. 541–2, 544. 27 A. Christensen, ‘The smith Kaveh and the ancient Persian imperial banner’, Journal of the K. R. Cama Oriental Institute, 5 (1925), 22–39.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 47906BF1 /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

429

mythical Persian king Jamshid; it boasts that the inheritance of the noble kings of Persia has fallen to his lot, and that he is reviving their lost glory. Ironically enough, for all its anti-Arab sentiment, this poem is written in Arabic, even though the poet’s patronymic (Mamshadh – ‘the mother is joyful’) proclaims his Persian descent.28 In the ninth and tenth centuries, under the patronage of Persian dynasties, New Persian developed as a literary language, cutting its teeth on translations from Arabic of Islamic universal histories29 and of the fables of Bidpai, a phenomenally popular book of animal tales whose text was rendered during the Middle Ages into over thirty languages from Icelandic to Indonesian.30 The tenth century was the age of the first great Persian poets, most of whom made their careers as panegyrists.31 Contemporary ceramics vividly illustrate a society poised between two cultures. It is worth noting that the very material of glazed pottery — cheap earthenware when all is said and done — suggests patrons who were affluent townspeople rather than aristocrats. On the one hand we encounter Arabic proverbs executed in stylish Kufic calligraphy with a peerless sense of interval and an intellectual appreciation of the visual potential of empty space. Such work, known from literally thousands of sherds, implies widespread literacy, an important factor in the contemporary flowering of literature. Other dishes, by contrast, capture the stately majesty of a ceremonial Sasanian royal hunt. Admittedly these images are reduced, provincialized, and in some ways only a shadow of the glamour that had gone; but their evocative power was undimmed. It shines through their awkward vernacular idiom.32 The celebrated Persian renaissance of the tenth century, which culminated in Firdausi’s Shahnama, took many different forms, visual as well as textual. Kings wore Sasanian crowns, mounted Sasanian thrones, and took Sasanian

28

Stern, ‘Coppersmith’, p. 553. c A. C. S. Peacock, Abu Ali Bal cami’s Translation of al-Tabari’s History (unpublished Ph.D. thesis, University of Cambridge, 2003). 30 See A Mirror of Princes from India: Illustrated Versions of the Kalilah wa Dimnah, Anvar-i Suhayli, Iyar-i Danish, and Humayun Nameh, ed. by E. J. Grube (Bombay: Marg Publications, 1991), endpapers. 31 E. G. Browne, A Literary History of Persia (Cambridge: Cambridge University Press, repr. 1959–1964), I, 445–80. 32 For typical examples of each genre, see C. K. Wilkinson, Nishapur: Pottery of the Early Islamic Period (New York: Metropolitan Museum of Art, 1973), p. 116 (plate 20) and p. 45 (plate 62a) respectively. 29

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4D567E21 /production/pindar/booksxml/hillen/chapter−14

430 titles such as shahanshah (‘king of kings’).33 Such details suggest that they subscribed enthusiastically to the ancient Iranian notion of the royal radiance (farr-i kiyani) which was popularly held to surround the legitimate king.34 They visited the magnificent ruins of Persepolis, the Achaemenid capital built largely in the sixth century BCE, and had its inscriptions deciphered for them (so they said) by Zoroastrian priests, before leaving their names on its walls.35 They forged genealogies for themselves which linked them to the great kings of the pre-Islamic past, and many noble Persians kept the memories of those monarchs green by adopting their personal names.36 They maintained such ancient customs as the celebration of the traditional spring and autumn festivals and the annual review of the army, and many Zoroastrian customs were still observed.37 Well into the eleventh century buildings bore inscriptions in Middle Persian alongside those in Arabic, and used the Sasanian solar calendar alongside its Muslim lunar equivalent.38 The Caspian provinces, where old traditions died hard, minted coinage of Sasanian type long after the rest of the Muslim world had abandoned it in favour of purely epigraphic currency, and continued to pay their taxes in silver plate as they had done under the Sasanians.39 The Iranian world, in short, clung fiercely to its pre-Islamic past and, in the cultural as distinct from the religious sphere, refused to bow the knee to the pervasive Arabization which by the year 1000 had engulfed most of the Muslim lands. It was

33 For a general discussion of these trends, see W. Madelung, ‘The Assumption of the title Shahanshah by the Buyids and “The Reign of the Daylam (Dawlat al-Daylam)”’, Journal of Near Eastern Studies, 28 (1969), 84–108 and 169–83. 34 See A. Soudavar, The Aura of Kings: Legitimacy and Divine Sanction in Iranian Kingship (Costa Mesa, Cal.: Mazda Publishers, 2003). 35 S. S. Blair, The Monumental Inscriptions from early Islamic Iran and Transoxiana (Leiden: Brill, 1992), pp. 32–7, 118–22 and plates 31–2 and 68. 36 In the Caspian provinces, the local chieftains habitually bore Shahnama names; see H. L. Rabino di Borgomale, Mazandaran and Astarabad (London: Luzac, 1928), pp. 133–49. The same is true of almost all the Saljuq sultans of Anatolia. See C. Cahen, La Turquie Pré-Ottomane (Istanbul and Paris: Institut français d’études anatoliennes d’Istanbul, 1988), p. 217. 37 G. Wiet, Soieries Persanes (Cairo: Imprimerie de l’Institut français d’archéologie orientale, 1948), pp. 113–36. 38 Blair, Monumental inscriptions, cat. nos 19, 31–2 and plates 33–4 and 48–54. 39 For the local background to these customs, see R. Vasmer, ‘Die Eroberung Tabaristans durch die Araber zur Zeit des Chalifen al-Mansur’, Islamica, III/1 (1927), 86–150.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 51ED99DC /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

431

Firdausi who rescued40 and codified the scattered national traditions, partly from oral informants but mainly from written sources which were rendered largely in prose,41 and gave them unified form in verse.42 Firdausi’s decision to end that history with the defeat of the last Sasanian monarch at the hands of the Arabs definitively separated in the national consciousness the period before Islam from what followed the coming of the new religion.43 The barrier between the two periods was not nearly as impassable or impermeable as Firdausi makes out, but he, if anyone, crystallized the sense of such a barrier, and it seems likely that he intended all along that his epic would create strong national sentiment. In this regard there is perhaps a parallel to be made here with the role of epic literature in the political process whereby regional kingdoms emerged, as Sheldon Pollock’s chapter in this volume demonstrates, from the break-up of the Gupta empire. This, then, is the context within which the Shahnama was created, and that leads to the second theme of this paper, though it overlaps with the first: why was the Shahnama so popular? This is of course a chapter in itself, so let us stick to essentials. First and foremost, language. Here too Firdausi’s was a work of rescue.44 His mighty line, with its regular beat and its end-stopped rhymes, lends itself perfectly to declamation and, more specifically, to popular oral performace. His set pieces, such as battles and laments, are especially well suited to that mode. To this day a special tone, at once nasal and resonant, and instantly recognizable, is adopted by those who quote the Shahnama, and particularly by the professional coffee-house reciters.45 The verses are made much easier to remember not only because they rhyme but

40

A. J. Arberry, Classical Persian Literature (London: Allen and Unwin, 1958), p. 46, quoting Mirza Muhammad ‘Ali Furughi. See also A. S. Shahbazi, Ferdowsi: A Critical Biography (Costa Mesa, Cal.: Mazda Publishers, 1991), pp. 63–75. 41 Arberry, Classical Persian Literature, pp. 43–4. For the argument that these sources were principally of eastern Iranian origin, see V. V. Bartold, tr. H. H. Schaeder, ‘Zur Geschichte des persischen Epos’, Zeitschrift der Deutschen Morgenländischen Gesellschaft, 98 (1944), 121–57. 42 R. Levy (tr.; rev. by A. Banani), The Epic of the Kings: Shah-Nama the National Epic of Persia by Ferdowsi (London and Boston: Routledge and Kegan Paul, 1973), p. xvi. 43 For a discussion of how legitimately Firdausi can be seen as a historian, see G. E. von Grunebaum, ‘Firdausi’s concept of history’, 60. d˘ogum yili münasebetiyle Fuad Köprülü armˇgani Mélanges Fuad Köprülü (Dil ve Tarih-Co˘grafya Fakültesi tarafindan ne˛sredilmi˛stir) (Istanbul: 0. Yalçin, 1953), pp. 177–93. 44 Arberry, Classical Persian Literature, p. 47, quoting Furughi. 45 Browne, Literary History of Persia (note 31 above), II, 143.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 59F8AA00 /production/pindar/booksxml/hillen/chapter−14

432 also by the nature of the metre itself: U – – | U – – |U – – | U – , which follows the pattern of the English line: ‘The Pharaohs of Egypt, the Caesars of Rome’.46 Or, to give a specimen of the original:47 ágar tónd bádi bár ayad ze kónj be khák áfkanad nárasidé torónj setámakaré khanimesh ár dadgár honarmánd danímesh ár bí honár ágar marg dádast bídad chíst ze dád in hame báng o faryád chist

What if a fierce and vagrant wind springs up, And casts a green unripened fruit to earth. Shall we call this a tyrant’s act, or just? Shall we consider it as right or wrong? If death is just, how can this not be so? Why then lament and wail at what is just?

Just as, in his role of historian, Firdausi broke with Islamic precedent in weeding out the Semitic elements in the early history of mankind, removing the Hebrew and Arabian prophets,48 so he took immense pains to expurgate from his vocabulary as many as possible of the Arabic elements which by this time had well-nigh infested the Persian language. Grossly exaggerated claims have been made as to the extent of this purification; in fact, the percentage of words of Arabic origin has been calculated at between 4 per cent and 5 per cent.49 But this is far lower than was the norm with other roughly contemporary poets or prose writers in Persian. To a considerable extent, therefore, his vocabulary was contrived and recondite. Hence the Shahnama has become a thesaurus for later generations. More than that, it instilled a pride in Persian as a language of literature and culture. Persian was the court and administrative language of Anatolia and especially northern India for centuries, and thus it is no cause for surprise that court poets of those areas, among many others, should promote, indeed exalt, local rulers through the medium of either actual Shahnamas, perhaps lightly doctored, or entire pastiches of them. The latter was common practice. In Turkey, Ahmad Qani‘i and Khwaja Dahhani both wrote their own versions of the Shahnama, tailoring them to glorify the deeds of the Rum Saljuq sultans in the thirteenth century,50 while Yarcani did the same in the fourteenth century for the 46 A. G.Warner and E. Warner (tr.), The Shahnama of Firdausi, vol. I (London: Kegan Paul, Trench, Trübner, 1905), p. 47. 47 The Tragedy of Sohrab and Rostam from the Persian National Epic, the Shahname of Abol-Qasem Ferdowsi, tr. J. W. Clinton (Seattle and London: University of Washington Press, 1987), p. xxii. For a similar specimen, see Browne, Literary History of Persia, I, 115. 48 Shahbazi, Ferdowsi (note 40 above), pp. 122, 135. 49 Browne, Literary History of Persia, II, 146. 50 M. F. Köprülü (tr. and ed. by G. Leiser), The Seljuks of Anatolia: Their History and Culture According to Local Muslim Sources (Salt Lake City: University of Utah Press 1992), pp. 15–19.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4E87EE2E /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

433

Qaramanids.51 Two hundred years later the exploits of Sultan Sulaiman the Magnificent, the scourge of Eastern Europe, were extolled in a ‘Book of Sulaiman’ (Süleymanname) loosely modelled on the Shahnama.52 Indeed, the post of shahnameji (a term which could perhaps best be rendered as ‘epic poetaster’) was a fixture of the Ottoman court.53 As late as the nineteenth century, in the obscure Himalayan state of Chitral, a local chieftain, although he was in political terms assuredly no more than a pimple on the rump of the Raj, employed a court poet to sing his praises in the manner of Firdausi.54 But of course the popularity of the Shahnama for Iranians rests on far more than its language. The Shahnama has remained the catalyst for the definition of what it means to be an Iranian. Its stories are the prized birthright of the entire nation. It has fostered pride of race,55 patriotism and national revivals. It has embodied to perfection the Persian sense of the divine right of kings. It integrated Iranian myth, legend and history into classic form, much as the Mahabharata did in India.56 Its particular brand of meta- or para-history has hijacked the actual events of the historical past, so that Persepolis becomes Takht-i Jamshid, ‘the throne of Jamshid’, while the identifiable Achaemenid tombs nearby, with the reliefs of equally identifiable Sasanian kings below them, become in the popular imagination something that is much vaguer but also much more evocative: Naqsh-i Rustam, ‘the picture of Rustam’. By such means the Shahnama has remained a bastion against what many Persians have regarded as creeping Arabization.57 It is a

51

Ibid., pp. 20–1, 71–2. E. Atıl Süleymanname: The Illustrated History of Süleyman the Magnificent (Washington, DC: National Gallery of Art; New York: H. N. Abrams, 1986). 53 C. Woodhead, ‘An Experiment in Official Historiography: The Post of ¸Sehnameci in the Ottoman Empire, c. 1555–1605’, Wiener Zeitschrift für die Kunde des Morgenlandes, LXXV (1983), 157–82. 54 This was discussed by Bert Fragner at the first Cambridge Shahnama conference in 2001. 55 Browne, Literary History of Persia, II, 143. 56 These various levels of discourse have prompted Nallino to compare the Shahnama with the Mahabharata: C.A. Nallino, Firdusi: Discorso per il millenario della sua nascita: Pronunziato nella R. Accademia d’Italia il 20 Decembre 1934 — XIII (Rome: R. Accademia d’Italia, 1935), p. 22. For a fuller discussion of what might be termed the Indian connection, see S. Wikander, ‘Sur le fonds commun indo-iranien des épopées de la Perse et de l’Inde’, La Nouvelle Clio, 1–2 (1949–50), 310–29. 57 Shahbazi, Ferdowsi (note 40 above), pp. 134–5. 52

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4A76826E /production/pindar/booksxml/hillen/chapter−14

434 succession of adventure stories, true; but it is also a guide to ethics58 – notably to personal conduct – an instrument of teaching and a mirror for princes.59 This unique status should not be taken to imply a text set in stone. In fact, the text was in a state of constant instability. To this day there is lively debate about the authenticity of any number of verses ascribed to Firdausi, for subsequent poets (and reciters who thought they were) freely but furtively interpolated their own pastiches as Firdausi’s original work. These could run to thousands of lines – and pose (though on a vastly greater scale) the same problem as the Homeric ‘wild lines’ mentioned above.60 And the practice of dictation rather than of one-to-one copying meant that given episodes of the poem might be extended, recast or contracted at will, or even omitted altogether. Hence no two medieval Shahnama manuscripts have the same text. In Iran there was no lack of later poets who tried to rival him under their own names,61 but there can be no doubt that Firdausi saw off all his rivals in recounting the story of Iran in verse. In the field of the visual arts, the popularity of certain episodes of the Shahnama, and of key events within those episodes, have been fixed in the popular imagination by the increasingly streamlined formulae which painters have developed for them. Those formulae triumph with ease over any number of textual inconsistencies. Here again one can detect the impact of the market. It is of course the non-generic and innovative images which are the most interesting. Here the new Cambridge-Edinburgh Shahnama database comes into its own, for it enables the researcher to fix the balance between standard and non–standard images. Indeed, it can be used to reveal not only how the illustrative programme of any one manuscript worked but also the fine tuning of exactly what part of the text was being illustrated. Artists naturally observed differing degrees of fidelity to the text. They might choose, or be directed, to depict, say, a combat scene in a formulaic way or instead to

58

Nallino, Firdusi, pp. 24–5. J. S. Meisami, ‘The Šah-Name as Mirror for Princes: A Study in Reception’, in Pand–o Sokhan: Mélanges offerts `a Ch.-H. de Fouchécour ed. by C. Bala¨y et al. (Teheran: Institut Français de Recherche en Iran, 1995), pp. 265–73. 60 An example is the Barzunama, on which Dr Gabrielle van den Berg is preparing a detailed study. 61 W. L. Hanaway, ‘The Iranian epics’, in Heroic Epic and Saga, ed. by Felix J. Oinas (Bloomington and London: Indiana University Press, 1978), pp. 76–98; M. Molé, ‘L’épopee iranienne après Firdausi’, La Nouvelle Clio V (1953) (Mélanges A. Dupont-Sommer), 377–93. 59

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 579047A3 /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

435

capture a precise moment in that combat – a parry, a fall, a clinch or a maiming are all capable of being rendered as separate visual elements. It is here that the concept of the breakline – the line which triggers a given image – is vital. Recent research suggests that painters often looked at the text with much more care than was previously thought.62 Finally, what of the treatment of Alexander the Great in the Shahnama – a topic shortly to become a staple of Persian romantic poetry, for example by Nizami? With his entry on the stage, the epic leaves the world of myth and, for the first time, enters a factual chronological framework. It is quickly clear, however, that the Iranian tradition as retailed by Firdausi comes closer to the confections of Baron Münchhausen than to sober history.63 This is of course equally true of many other literary traditions about Alexander,64 including Iranian ones.65 In this case, Firdausi was for once forced to depart from his normal practice of relying on Iranian sources. The indigenous Zoroastrian tradition saw the conqueror as a limb of Satan,66 while the Muslims awarded him the status of a prophet.67 Neither of these approaches 62 The key research in this field so far has been carried out by F. Mehran. See his article ‘Frequency Distribution of Illustrated Scenes in Persian Manuscripts’, Student 2 (1998), 351–79. 63 See the richly illustrated exhibition catalogue Alessandro Magno: Storia e mito (Fondazione Memo) (Rome: Leonardo Arte, 1995). 64 For a recent survey of the subject, see Alexandre le Grand dans les littératures occidentales et proche-orientales, ed. by L. Harf-Lancner, C. Kappler and F. Suard (Nanterre: Centre des sciences de la littérature de l’université Paris X, 1999). 65 See J.-C. Bürgel, ‘Conquérant philosophe et prophète: L’image d’Alexandre le grand dans l’épopée de Nezami’, in C. Bala¨y et al. (eds), Pand-o Sokhan, pp. 65–78; M. Gaillard, ‘Alexandre dans la littérature “semi-populaire” de l’Iran médiéval (le Darab-nameh d’Abu Taher Tarsusi)’, in Harf-Lancner et al., Alexandre le Grand, pp. 367–9; A. Abel, ‘La figure d’Alexandre en Iran’, in La Persia e il mondo Greco-romano, Accademia Nazionale dei Lincei, Anno CCCLXIII, Quaderno 76 (Rome, 1966), 119–36. 66 J. Darmesteter, ‘La légende d’Alexandre chez les Parses’, Mélanges publiés par la Section historique et philologique de l’École des Hautes Études (Paris: Vieweg, 1878), pp. 83–99; C. A. Ciancaglini, ‘Alessandro e l’incendio di Persepolis nelle tradizioni greca e iranica’, in La diffusione dell’eredit`a classica nell’et`a tardoantica e medievale: Forme e metodi di trasmissione, ed. by A.Valvo (Alessandria: Edizioni dell’Orso, 1997), pp. 59–81; and G. Gnoli, ‘La demonizzazione di Alessandro nell’Iran sassanide (III-VII secolo d.C.) e nella tradizione zoroastriana’, in Alessandro Magno, p. 175. 67 Y. Yamanaka, ‘“From Evil Destroyer to Islamic Hero”: The Transformation of Alexander the Great’s Image in Iran’, Annals of the Japanese Association for Middle-East Studies, 9 (1993) 55–87; and ‘Ambiguité de l’image d’Alexandre chez Firdausi: les traces des traditions sassanides dans le Livre des Rois’, in Harf-Lancner et al., pp. 341–53.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 458BA798 /production/pindar/booksxml/hillen/chapter−14

436 fitted Firdausi’s plan. Compelled to turn elsewhere, he elected to base his account on the Islamized version of the pseudo–Callisthenes.68 Hence its strong family resemblance to the numerous medieval versions of the Alexander Romance in east and west.69 These are by no means the end of the story: right into modern times creative artists have felt empowered to take diabolical liberties with the facts of this particular hero’s life. In 1731, for instance, Johann Adolf Hasse wrote an opera called Cleofide which has our hero vainly pursuing the intended wife of the Indian king Porus.70 And his exploits are a staple of the Italian peplos or sword-and-sandal films that clog up the schedules of daytime television. In Firdausi’s work Alexander has many faces. To begin with, his name itself is Persianized into Iskandar, perhaps as an expression of national vanity after national defeat.71 Next, he has a marked religious dimension; Firdausi says he is a Christian, but makes him behave like a Muslim ruler with Sufi leanings, for example in his readiness to humble himself before his spiritual superiors, naked and homeless though they be. The poet emphasizes his role as a philosopher, tirelessly seeking wisdom; and then he is the very exemplar of the undefeated warrior king. But he is also a peace-maker, a subtle diplomat and a paragon of justice. These aspects of his personality transfigure his adventures from the physical to the political, metaphorical and spiritual planes. Primarily, though, he is presented as an orientalist before his time, tirelessly exploring the strange and the exotic world of the east. The Islamic world, like medieval Europe, equated the east with marvels.72 Real and imaginary lands are promiscuously jumbled together; Firdausi’s grip on geography is tenuous, and besides, who cares? The tale’s the thing, and to use a hand lens on its topography is like taking a spade to a soufflé. So in the space of a few lines Iskandar leaves his prolonged Indian adventure, with its

68

Shahbazi, Ferdowsi (note 40 above), p. 134. For the medieval period, see G. Cary, The medieval Alexander (New York and London: Garland, 1987) and D. J. A. Ross, Alexander historiatus: A Guide to medieval illustrated Alexander Literaure (London: Warburg Institute, University of London, 1963). 70 This Dresden-based composer lived from 1699 to 1783. Cleofide has recently been revived on disc. 71 Browne, Literary History of Persia (note 31 above), I, 118. 72 J. M. Rogers, Myth and Ceremony in Islamic Painting (London: n.p., n.d.), pp. 3–4. 69

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4EC8194F /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

437

ascetic soothsaying sages and its emaciated Brahmins (figure 1),73 to make for Andalusia and its glamorous queen, stopping en route at the Ka‘ba in Mecca, where – a Muslim avant la lettre – he reveres the Black Stone. China and the Punjab, Yemen and Babylon, are also on the itinerary, but it is the marvels that really dominate this account. Indeed, Iskandar explicitly confesses that he is after novelties. Some, such as his search for the Water of Life, guarded by the prophet Khizr, have an obvious metaphysical dimension. Others, like his encounter with the kargadan, the winged rhinoceros (figure 2), or with a dragon that is terrorizing the local countryside, exercise his brawn rather than his brain. Yet others reflect a historical reality, like his building of a wall against the savages of Gog and Magog (figure 3), who symbolize the perennial nomadic menace from the Eurasian steppe;74 indeed, the remains of a fortification still called Alexander’s Wall (sadd-i Iskandar) can be traced to this day to the east of the Caspian Sea. He leads his men through the Land of Gloom (figure 4), despite hearing a spectral voice that all of them will be sad when they leave it. And indeed they are – for they stumble across some stones in the dark, and neglect to pocket enough of them, only to discover later that they were jewels. His sightseeing takes in a topaz palace enshrining a corpse, and also a visit to the Talking Tree at the furthest limit of the world (figure 5), a theme which also figures, though in slightly different form, in the medieval Western illustrations of the Alexander story.75 Here its fruit consists of heads whose normal practice is to utter the words waq waq, waq waq, waq waq.76 But the sight of Iskandar inspires them to heights of eloquence far beyond this stunted vocabulary; in solemn tones they warn him that his time is up and that he has wasted his life in meaningless conquests. 73 And possibly other strange encounters; see J. J. Modi, ‘The story of Alexander the Great and the poison-damsel of India: A trace of it in Firdousi’s Shah-Nameh’, Journal of the Bombay Branch of the Royal Asiatic Society, N.S. 3 (1928), 212–30. 74 C. E. Wilson, ‘The wall of Alexander against Gog and Magog; and the expedition sent out to find it by the Khalif Wathiq in 842 A.D.’, Hirth Anniversary Volume (Asia Major … Introductory volume), ed. by Bruno Schindler (London: Probsthain, 1922), pp. 575–612; Encyclopaedia of Islam (2nd ed.), XI (Leiden: Brill 2002), cols. 231a-234a, s.v. ‘Yadjudj wa-Madjudj’ (E. van Donzel and C.Ott). 75 In the Western tradition this becomes the Tree of the Sun and the Moon; see Penelope Lively, The Mythical Quest: In Search of Adventure, Romance and Enlightenment (London: The British Library, 1990), plate 22. 76 See Encyclopaedia of Islam (2nd ed.) XI (Leiden 2002). cols. 103b–109a, s.v. cWak.wak.’ (G. Ferrand, G. S. P. Freeman-Grenville, G. R. Tibbetts, S. M. Toorawa and F. Viré).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 453BC903 /production/pindar/booksxml/hillen/chapter−14

438 How was the Alexander story interpreted by the painters who illustrated it? Here again the Michigan database mentioned earlier comes into its own, for it reveals the frequency with which certain episodes of that story were depicted over the centuries. Far and away the most popular image, repeated 74 times, was the affecting, undeniably big emotional scene in which Iskandar cradles in his lap the head of his dying enemy, the Persian emperor Dara – the historical Darius. Next in popularity (27 scenes) is the physical feat whereby Iskandar splits his foe, the Indian Rajah Porus, in two (figure 6).77 Thereafter come, in short succession, with almost the same frequency, the building of the wall against Gog and Magog (figure 7), and the prophet Khizr leading Iskandar on his quest to find the Water of Life; and finally, with 15 occurrences, Iskandar at the Ka‘ba in Mecca. The most interesting aspect of these statistics is that painters by tacit consent avoided the most challenging scenes of the text, namely those full of fantasy and wonder featuring lands of perpetual fire, climes inhabited by lion-bodied people, men whose bodies were sandwiched by their ears, fire-breathing blacks and one-breasted women. Moreover, most manuscripts limited themselves to the illustration of one or two scenes, and here again the themes of choice were safe rather than challenging: battles, receptions, enthronements and of course the scene of the dying Dara with its obvious subtext of legitimizing the accession of Iskandar, foreigner though he was, to the throne of Iran.78 Painters, it seems, were conservative and went for themes which could fit a familiar, and often commercially-inspired, template, as already noted above. The three thousand-odd lines of Firdausi’s tale of Iskandar contain no fewer than a hundred incidents that would lend themselves well to illustration. Yet during the great centuries of Persian miniature painting – from c. 1300 to c. 1650 – two-thirds of these were never illustrated at all. The general impression, then, is of a series of missed opportunities. The salient facts are that the entire Iskandar cycle was very frequently omitted altogether from the illustrative programme, and even when it was not ignored it was treated only in very cursory fashion. 77

H. von Mzik, ‘Die Schilderung der Schlacht zwischen Alexander dem Grossen and dem Inderkönig Porus in Firdousis “Königsbuch” ’, ZDMG, 104 (1954), 357–61. When it is complete, the Cambridge/Edinburgh database will obviously contain many more examples of scenes illustrating the story of Iskandar. 78 For this episode, see P. Briant, Darius dans l’ombre d’Alexandre, pp. 497–502 and figs 49–53. See also pp. 502–3, 506 and 508, and figs 54–5, for a discussion of how this same iconography was made to serve the very different story of Rustam killing his son Sohrab.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 41 OUTPUT: Fri Feb 10 09:57:13 2012 SUM: 4AD5B7FC /production/pindar/booksxml/hillen/chapter−14

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

439

But there is one great exception, and I will end with it.79 The fullest version of the Iskandar story in Persian painting is the so-called Great Mongol Shahnama, an unfinished but spectacularly ambitious royal manuscript produced at the behest of the Mongol ruler of Iran, Abu Sa‘id, perhaps in the 1330s. As such it is an impressive testimony of how, in the space of a little over a century, Iran had bested its conquerors, as (long before) captured Greece had taken Rome captive. Of the Mongol invasion under Genghis Khan from 1220 onwards, the chronicler says laconically: amadand, kandand, sukhtand, kushtand, bardand, raftand (‘they came, they uprooted, they burned, they slaughtered, they plundered, they departed’).80 Three generations later, a Mongol prince fluent in Arabic and Persian as well as his ancestral tongue ordered a physically huge version of the Persian national epic with some two hundred illustrations:81 a truly remarkable volte face. Even more astonishingly, most of those paintings were carefully correlated to events in recent Mongol history.82 This bald statement requires some glossing. Firdausi’s text, despite the vagaries attendant on oral transmission, with its opportunities for omission and pastiche, was basically not negotiable. But it did provide an articulated skeleton which could be fleshed out in various ways. Of course Firdausi did not write the Shahnama as a seer with Mongol history in mind, and any richly illustrated Shahnama demands that many of the major episodes of the epic be depicted, even if their relationship to events of (say) Mongol history are at best peripheral and might well be non-existent. But the painters could – and did – fill some of the gaps left by the text, and thereby help to make a Mongol connection plausible.

79 See R. Hillenbrand, ‘The Iskandar Cycle in the Great Mongol Shahname’ , in The Problematics of Power: Eastern and Western Representations of Alexander the Great, ed. by M. Bridges and J. –C. Bürgel (Bern and New York: Peter Lang, 1996), pp. 203–29. 80 ‘Ata Malik Juvaini, Tarikh-i Jahan Gusha, in the translation of Browne, Literary History of Persia, II, 12. 81 The latest estimate is 205 paintings: see A. Mahdavi, An Event-Driven Distribution Model forAutomatic Insertion of Illustrations in Narrative Discourse: A Study Based on the Shahnama Narrative (unpublished Ph.D. thesis, University of Edinburgh, 2004), Chapter 8 (‘Reconstructing The Great Mongol Shahnama’). 82 This strikingly innovative theory has been put forward by A. Soudavar in ‘The Saga of Abu-Sa‘id Bahador Khan. The Abu-Sa‘idname’, in The Court of the Il-khans 1290–1340, ed. by J. Raby and T. Fitzherbert (Oxford Studies in Islamic Art XII) (Oxford: Oxford University Press, 1996), pp. 95–218.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14 Sequential 20

JOBNAME: ???? PAGE: 1 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 0C742FBC /production/pindar/booksxml/hillen/chapter−14−artwork

440

1. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar and Taynush: detail showing Brahmins.

2. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar fights the kardagan.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 121AE841 /production/pindar/booksxml/hillen/chapter−14−artwork

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

441

3. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar builds a wall against Gog and Magog.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 06E88447 /production/pindar/booksxml/hillen/chapter−14−artwork

442

4. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Alexander consults the Talking Tree.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 086E67E4 /production/pindar/booksxml/hillen/chapter−14−artwork

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

443

5. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar enters the Land of Gloom.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 0CF6290F /production/pindar/booksxml/hillen/chapter−14−artwork

444

6. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar in single combat with the Fur of Hind.

7. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar builds a wall against Gog and Magog: details of savages.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 07DAD7BA /production/pindar/booksxml/hillen/chapter−14−artwork

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

445

8. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Iskandar enthroned.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 06F775B4 /production/pindar/booksxml/hillen/chapter−14−artwork

446

9. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: the Iron Warriors of Iskandar rout the army of the Fur of Hind.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 34 OUTPUT: Tue Aug 10 14:34:53 2010 SUM: 085D9938 /production/pindar/booksxml/hillen/chapter−14−artwork

447

10. Great Mongol Shahnama, probably Tabriz, c. 1320–c. 1335: Bier of Iskandar.

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-artwork Sequential 8

JOBNAME: ???? PAGE: 1 SESS: 43 OUTPUT: Tue Aug 10 10:30:02 2010 SUM: 47CC2BC8 /production/pindar/booksxml/hillen/chapter−14−ends

448 There was a more timeless strand in the Shahnama text which pictures were well suited to highlight. Without any need to manipulate the text itself, pictures could serve to single out recurrent themes in the grand sweep of history – for it is unlikely that contemporary readers would have made the fine distinctions that a modern reader makes between myth, legend and historical fact. Thus a Shahnama produced, like this one, for a princely patron could be made into a paean of royal power simply by a careful choice and placing of key images which proclaimed that theme. This would not contradict the text itself; it would simply single out that theme among many others. To suggest that the Great Mongol Shahnama celebrates royal power – a visual panegyric composed for an audience of one (or at any rate not many: at most the ruler’s inner circle; the enjoyment of book painting is apt to be marred if it is shared) – may seem banal, and might also be taken to imply that the images are being wielded as a rather blunt instrument. But royal power has many aspects, as do these images. Much of what might be termed the nuts and bolts of a monarch’s life pass before us here, and their human context is realized in a fully nuanced way. This is a pictorial version of a literary genre much in favour in medieval Iran, namely the Mirror for Princes. The works of that genre ram home a given point by means of an appropriate and memorable anecdote. The Great Mongol Shahnama does the same, but all the more effectively because key stories are given a pointed visual embodiment. Thus the human cost exacted when a king doubts the loyalty of those closest to him, his vizier and his wife, finds moving and dramatic expression in successive images – surely intended to serve as pendants to each other. We see first the shattered cup of poison which reveals the queen’s duplicity, and then in contrast the self-castrated vizier presenting the king with a little bag, poignant physical proof of his sacrificial loyalty.83 In the course of the book virtually every aspect of royal life passes under review, often seen from different angles or at different stages, as in the case of succession disputes – broken faith, usurpation, rebellion and so on. Sometimes the unfolding of a monarch’s reign is made to serve, by means of carefully chosen episodes, didactic purposes. Thus we can trace Bahram Gur’s gradual redemption from a heedless, self-seeking playboy to a virtuous

83

For the story and the related images, see O. Grabar and S. Blair, Epic Images and Contemporary History: The Illustrations of the Great Mongol Shahnama (Chicago and London: Chicago University Press, 1980), pp. 142–5.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 43 OUTPUT: Tue Aug 10 10:30:02 2010 SUM: 43878C92 /production/pindar/booksxml/hillen/chapter−14−ends

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

449

king repeatedly ready to risk his life for the good of his people, and humbly ready to accept rebuke from the lowest of his subjects, acquiring a halo en route.84 Such selective correlation of text and image highlights the homiletic element in the poem. Still other stories reveal the link between the Shahnama and wisdom literature; and the search for wisdom takes Iskandar, to whom probably the fullest cycle of images was devoted, to the very ends of the earth. Indeed, it was in this particular manuscript that the story of Iskandar, with its twelve surviving scenes out of a probable seventeen, a record in Persian painting, was singled out for special pictorial treatment.85 The reader would have realized very quickly, from the dramatic increase in the rate of illustration at this point in the text, that Iskandar was more important than anyone else in the Shahnama – truly a momentous innovation. In this particular manuscript only Iskandar, of all the monarchs mentioned in the text, is depicted according to Sasanian protocol, both at the beginning and at the end of his reign, seated on his throne (figure 8) and encoffined on his bier (figure 10).86 Furthermore, the distribution and content of the paintings might have suggested to an attentive reader that the principal message of the story is the vanity of earthly desires, whether for possessions or for wisdom. Iskandar in his roles as warrior and monster-slayer is overshadowed by Iskandar the mere man, whose wanderings across the face of the globe are doomed to disappointment. But the fantasy element which is so integral to Firdausi’s version of the Alexander story is not forgotten either, and indeed it figures more largely in the Great Mongol Shahnama than in any other illustrated version of the epic: the talking tree, the fabulous beasts which Iskandar slays, the fire-breathing cavalry which he unleashes against King Porus (figure 9), and the speaking birds. Indeed, these magical scenes account for half the extant images in this cycle, and are thus clearly a vital element in the way that it was conceived. Successive Ilkhanid rulers were noted for their superstition and for their

84

Ibid., 150–63. See the detailed discussion of this issue in F. Mehran, ‘Missing Paintings in Dismantled Persian Manuscripts’, Student (Special Issue on Statistics in Art), 4/I (2001), pp. 61–78. 86 Grabar and Blair, Epic Images, pp. 112–13 and 134–5. It is of course possible that some of the missing pictures would also have shown monarchs being both crowned and buried. 85

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 43 OUTPUT: Tue Aug 10 10:30:02 2010 SUM: 45131B29 /production/pindar/booksxml/hillen/chapter−14−ends

450 dependence on astrology. For a dynasty whose conquerors had seen all there was to see in the sublunary world, perhaps the supernatural was all that was left to wonder at. The unprecedented and subsequently unmatched pictorial emphasis on Iskandar suggests that – unlikely as it may seem – the former inhabitants of outer Mongolia saw themselves in the mirror of Alexander the Great. The reasons are not far to seek. Iskandar too was a foreigner. Iskandar too had conquered most of the known world. Iskandar too, as Firdausi depicts him, was enquiring and tolerant in matters of religion. Iskandar too had a taste for the supernatural. Best of all, Mongol rulers – some of whom took the title ‘The Second Alexander’ – fabricated a genealogy which linked them to Alexander himself via his mother Olympias.87 The paintings rose to this challenge. Within this single manuscript one can trace the sequence from paintings that are simple illustrations88 to those that are commentaries, then metaphors and finally independent works of art operating confidently at several levels of meaning. More and more content – descriptive, emotional, historical, symbolic – is pumped into these paintings, and only an absolutely assured command of pictorial language enables the greatest of these painters to control the forces that they unleash. Yet all this visual splendour and intellectual complexity is destined, as already noted, for the eyes of only very few people. Book painting is an exquisitely selfish art form. Thus to modern eyes there is an inherent mismatch between the medium and the message. But this extremely limited readership does not exclude a complex freight of politico-historical, let alone symbolic, meaning. We should not underestimate the despotic, unaccountable power of the ruler, or misunderstand the nature of Muslim panegyric. Yet precisely because the text illustrated in the Great Mongol Shahnama was not written to glorify some Mongol prince who had hijacked Firdausi’s words, Abu Sa‘id, but was written for the ages, a summative heroic epic of the generations of Persian kings, the images of this manuscript – however relevant to the history of the Mongol ruling house over the previous century – also illustrate that much more public document written by Firdausi long

87

Hillenbrand, ‘Iskandar cycle’ (note 79 above). It remains to be seen whether a chronological sequence in real time can be demonstrated in these paintings; it is too early, for example, to assume that those at the beginning of the poem were also the first to be painted, as can be suggested grosso modo for those at the beginning of the Shahnama-yi Shahi of Shah Tahmasp. 88

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-ends Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 43 OUTPUT: Tue Aug 10 10:30:02 2010 SUM: 07F9B84C /production/pindar/booksxml/hillen/chapter−14−ends

THE SHAHNAMA AND THE PERSIAN ILLUSTRATED BOOK

451

before. Their resonances extend alike far into the past and far into the future, just as Firdausi’s text does. So long as Iran has rulers, so long will the Shahnama remain relevant to its people.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-14-ends Sequential 4

JOBNAME: ???? PAGE: 1 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 34642B06 /production/pindar/booksxml/hillen/chapter−15

XV Islamic Bookbinding

I

N all of Islamic art there is perhaps no clearer case of the dependence of Islamic craftsmen on Western methods than is afforded by bookbinding. By the time the earliest Islamic manuscripts were being written, the format of the roll (rotulus, volumen) had long been superseded in the West by the codex, the ancestor of the modern book. It was this Western invention, itself derived successively from Roman tablets and diptychs, that was taken over by the Muslims, possibly via Ethiopian codices. This is not to say, however, that the rotulus was totally unknown to them. In the Christian and Muslim worlds alike, the roll form was employed spasmodically until the tenth century; but it is principally in the Islamic Near East that its influence on the codex form can be detected. The preferred format employed for early Qur’ân manuscripts is an extended oblong, stressing the horizontal rather than the vertical. The connection with the rotulus form is self-evident. However, such manuscripts are in fact codices since they are bound within covers. Despite the prestige conferred on this format by its use for copies of the sacred book, its popularity was comparatively short-lived. Qur’âns using the vertically conceived codex format were produced simultaneously with those in which the horizontal emphasis was dominant, and by the 11th century the latter had effectively died out. In the three centuries of its floruit this distinctively Islamic transitional form, part rotulus, part codex, generated what appear to be correspondingly original developments in binding technique and decoration. No description of these, however, can hope to be definitive in view of the almost total disappearance of so many kinds of early Islamic (and, for that matter, early Christian, Byzantine and Coptic) manuscripts and bindings. The surviving sample derives principally from only two areas, Egypt and Tunisia,

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 437ED32F /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

453

themselves very closely linked in style and technique as well as location, and thus cannot sustain any claim to be representative. If follows that features which seem to be original, or exclusively Islamic, may be nothing of the kind. Before these early bindings of oblong format are examined with these caveats in mind, certain general remarks may be in order. All of these bindings appear to be Egyptian or Tunisian, and the vast majority of pre-Islamic codex bindings are from Egypt too. Indeed, the history of the Islamic codex binding until the early 13th century must, insofar as it depends on surviving examples, perforce be based almost exclusively on Egyptian or Egyptian-influenced material. Ample literary evidence attests that bookbinding was widely practised elsewhere in the Islamic world in this early period, notably in Spain and in Syria, Iraq and Iran. However, while the names of these bookbinders, and the places where they worked, have been preserved, the records give no clue as to regional variations in bookbinding methods or ornament. Even in the period after 1200 it is hazardous to identify certain types of bindings (e.g. those from Syria and Iraq) on the basis of style alone. For many centuries it has been common practice to repair, renew or even entirely to replace bindings and it is therefore rash to assume that the binding of a dated manuscript will be contemporary with the text. Comparatively few Islamic bindings have been subjected to the detailed scrutiny, and occasionally even laboratory analysis, which should ideally underpin a dating attribution. Against a background of such lacunae, which are at once numerous and varied, generalisations are obviously risky. The directly or indirectly Egyptian origin of so many early Islamic bindings means that the immediately pre-Islamic bindings of Coptic Egypt are the obvious starting-point for a general survey of them. A great many of these Coptic bindings have survived and they testify to the high technical quality which was reached and long maintained in monastery workshops. The deliberate isolation of many of these monasteries fostered a notably conservative approach to this as to other crafts. Thus the ingrained expertise of Coptic bookbinders could be passed on virtually intact to their Muslim colleagues. It is not surprising, therefore, that many standard features of later Islamic bindings find their prototypes in this Coptic material, whose suggested dates range from the 4th to the 9th century. Examples include the preference for large geometrical designs filling most of principal field; the contrasting plainness of the spine; the use of multiple borders, also of geometrical type; and the extension of the back cover to form a flap (though in Coptic bindings all four sides were so extended, whereas in Islamic

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 43918329 /production/pindar/booksxml/hillen/chapter−15

454 bindings it is the right-hand side of the back cover which is so treated). The use of different designs for the front and the back of the binding, with the back cover being the simpler one, and of a blue linen backing pasted on the cover below the leather, are features based on Coptic practice. Standard Coptic techniques are used to sew the leaves together – chainstitching is used, for example, to join the quires. The 10th-century geographer al-Muqaddasî, himself a bookbinder, describes a variant of this technique whereby the leaves were sewn and the quires then glued together before the case itself was glued on with asphodel or wheat-starch. But the important fact is that such alternatives as stacking the pages together and tying wooden boards around them in the Indian way, or gumming strips of parchment or paper together to form long rolls, were ultimately rejected by the Muslims. Coptic influence was decisive in this process. Coptic bindings exhibit a wide range of technical devices which were borrowed and developed by Islamic craftsmen: sewn leather appliqué; painted, repoussé, stamped, incised, blind-tooled, punched and even inlaid fret- and filigree work; stencil patterns; and painted vellum end-papers. Most Coptic covers were, it seems, approximately square, and these are of little relevance to later Islamic work. But a rarer type is also known; this uses the vertically disposed rectangular format which was standard in later Islamic bindings, even to the detail of multiple framing bands. While Coptic bindings of wood are known, the majority use the less durable material of millboard or pasteboard made from papyrus pulp, an expedient easily explained by the lack of local wood. Prepared leather was then glued on to these boards. Thongs were passed through the projecting flaps to close the book. Such thongs are also found on 9th-century Islamic bindings from Qairawân. Other fastenings which became popular in Islamic bookbindings were metal clasps or loops which were passed over and secured by a peg. In some public libraries books were further protected by being chained to the shelves on which they were laid flat. Before dealing further with Egyptian bindings it will be convenient to comment on the only other substantial body of early Islamic bookbindings. A chance find in the 1940s revealed a treasure trove of discarded medieval Islamic bindings which had miraculously survived in a lumber room in the Qairawân Jâmi c. They total 179 of which 58 are attributed to the 9th century, 19 to the 10th, 71 to the 11th and 31 to the 12th and 13th centuries. Their close relationship to Egyptian Islamic and ultimately Coptic bookbinding is patent. If these bindings are indeed local, Tunisian bookbinding was clearly an offshoot of that of Egypt. All 77 of the bindings attributed to the 9th and

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 420227B8 /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

455

10th centuries are wider than they are high, but a few examples of putative 11th-century date reverse this relationship. Wooden covers with fastening clasps were common; Aleppo pine, fig, white and black poplar, laurel and tamarisk have been identified. The leaves were fastened to the boards in Coptic fashion by threads passed through eyelets or by glued strips of cloth, the latter a technique not often found in Coptic bindings. However, in general the technique of these Qairawân bindings, and often enough their decoration too, is closely dependent on Coptic work of the 8th-9th centuries. They exhibit a variety of designs in blind tooling on leather laid over wooden boards. Their decorative repertoire includes large languid palmettes, Kufic inscriptions and, above all, interlace designs for both the rectangular field and the broad borders. The latter often recall classical floor mosaics or coffered ceilings. Raised decoration is managed by gluing cords onto the underlying border and then tooling the moistened leather spread over them, a technique of Coptic origin. Some 240 separate dies or stamps have been identified; they include a craftsman’s signature and pious phrases. The latter theme is taken up on a grander scale in Kufic inscriptions on the covers themselves. Such short phrases as ‘What God wills’ or ‘in the name of God’ are so laid out that the phrase is divided between the front and then back cover. Certain features of the Qairawân material cannot, however, be so readily explained by reference to the Coptic tradition. The predominantly horizontal format, for example, which suggests that these bindings contained Qur’âns, is alien to Coptic work. So too is a novel treatment of the flaps. On three sides of the lower cover the leather is folded back on itself to form a double thickness in a continuous strip bent at right angles at the corners and glued to the lower cover there. These raised strips of stiffened leather effectively formed a box for the book and the upper cover acted as a lid. Some 125 Qairawân bindings were of this type. Both the oblong format and the box design, however, fell out of fashion comparatively quickly and therefore did not affect the subsequent development of Islamic bookbinding. Bindings of 12th-13th century date found in Qairawân bear simple geometric designs in medallion form and generally confine interlace motifs to a subsidiary role as spacefillers and border decoration. They are remarkably close to contemporary and slightly later Persian bookcovers, another proof of the conservative and international character which long marked Islamic bookbinding.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 41969778 /production/pindar/booksxml/hillen/chapter−15

456 With the exception of Tunisian bindings, virtually all medieval Islamic books, like their Coptic predecessors, use some kind of pasteboard for the core of the front and back covers, but among the oldest Egyptian bindings are several in wood. One of these (Cairo, Dar al-Kutub, Masâhif 188) has an interlace design whose bands are punctuated by tiny slits; this detail, like the motif itself, is directly dependent on Coptic prototypes. Others have somewhat coarse vegetal motifs with a generic likeness to Samarran decoration in the bevelled style. The contrast between these unpretentious covers and the text itself, carefully inscribed on vellum and enhanced by illumination in gold, red, green and other colours, is most marked. Perhaps for religious motives Islamic craftsmen consistently eschewed the bejewelled bindings in precious metals which were popular in Byzantium and the medieval West. It might indeed be conjectured that, as in Western Europe, bookbinding long remained a humble craft in the Islamic world perhaps because it lacked the glamour which attached to the other arts of the manuscript. It was, however, the Muslim artists who first broke free of these constraints and, in so doing, blazed a trail for their European counterparts. A special characteristic of pre-13th century Egyptian bindings is the attention paid to the doublures. They exhibit very varied decoration. They may be painted, or lined with strong saffron-coloured parchment with a clear varnish and geometrical ornament or even animals painted in yellow, red and black. Eventually paper supplanted parchment for the endpapers as for the leaves of the book itself, and it became the practice to line the inside cover with paper pressed into relief patterns or with Kufic inscriptions in filigree work, presaging the filigree doublures of the 15th century. Sometimes the endpapers were printed from a wooden block; yet others are painted or stamped. The most ambitious resemble antique ceilings or floors in their grid design with emblemata comprising animals or rosettes. A large group of these endpapers dating from the 10th-11th centuries and found at al-Ushmûnain is preserved in the Austrian National Library in Vienna. The bindings of Qur’âns in safina form, where the oblong is so emphasised that the ratio of height to breadth approaches 1:3, constitute a striking advance on the simplicity of earlier Egyptian Islamic covers, and announce many a leitmotif of later bindings. In an example in Vienna (Archduke Rainer collection, Inv. Chart. Ar. 28002, probably Egypt, 10th11th century), multiple frames enclosing a continuous palmette border surround a central field divided at intervals by roundels of seven inscribed circles apiece. A large teardrop medallion issues from either side of the

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 47B4172D /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

457

central roundel and emphasises the strong horizontal axis. Already the close relationship with carpet design is unmistakable. In this example the cover is made of multiple sheets of tough brown paper. Side by side with the growing emphasis on the central field in such bindings there developed – as in contemporary east Iranian pottery – an appreciation of the aesthetic qualities of blank space. The texture of highly worked but undecorated leather could also yield its full effect in such bindings. Intricate borders would set off the quiet splendour of the plain monochrome field. Medieval texts speak of books bound in yellow or blue satin or silk, which shows that there was a lively appreciation of colour in this medium. Nevertheless, although the technology for dyeing leather was readily available (indeed, shoes, saddles and other leather goods were often brightly dyed), Islamic bookbinders – or their patrons – preferred to respect the natural colours of the leather. Perhaps this was a factor in the preference for blind rather than gold tooling even when the latter technique had been mastered. Irrespective of their provenance, most early Islamic bindings are characterised by simple decoration in which plain knot patterns predominate. This suggests that the idea of a lavishly decorated book cover was rather slow to take root. Indeed, bookbinding could fairly be described as one of the most conservative of Islamic crafts, a factor which bedevils dating and provenance alike. Until the 15th century it seems to have been a markedly international art. Striking confirmation of this is provided from an unexpected source – the Manichaean bindings from Turfan in Central Asia, generally dated to the 8th and 9th centuries. When compared to contemporary work in Egypt, they can be recognised as very advanced for their time, for they employ such techniques as stamping, polychromy and filigree work. It is highly likely that knowledge of these techniques percolated to Iran. Further evidence of the international character of earlier medieval bookbindings is provided by a book cover from Khara Khoto in southern Mongolia, now in the Asiatic Museum, Leningrad, which is datable to the 13th century and whose bold thick knotted centrepiece finds its closest parallels in contemporary and earlier bindings from Tunisia. With such unlikely correspondences in mind it is not surprising that the earliest Iranian bindings to survive do not seem to express a distinctively local style. A Qur’ân cover in the Chester Beatty Library dated 292/905 has a central knotted rosette easily paralleled in Tunisian or Egyptian bindings. Another in the same collection, dated 361/972, is unusual in that it has no flaps, but its lozenge grid with a circle at

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 42ADFEB0 /production/pindar/booksxml/hillen/chapter−15

458 each vertical junction of the lozenges is in line with contemporary Tunisian work. Each circle bears the text of Sura 112 in miniature Kufic; such stamped inscriptions are also found on Tunisian, Egyptian and Yemeni covers. Unfortunately the dearth of 11th-13th century Iranian bindings makes it impossible to be sure whether the close connection with Egyptian work continued, but the notable consonance between Egyptian and Iranian bindings throughout much of the 14th century, when both schools shared a preference for geometric designs, does suggest that the links were never broken. It is only in the later 14th century that Iranian bindings shed the taste for angular geometric patterns and begin to develop along lines alien to the Egyptian tradition, although both schools developed filigree work to the full. But in the later medieval period the disparate traditions which had crystallised in Iran and Egypt remained the two dominant forces in Islamic bookbinding, and it was only in the bindings of Ottoman Turkey that a certain rapprochement between them was eventually achieved. A brief comparison between the bookcovers of Egypt and Iran in the 14th century is enough to show that Egypt had established a far more sophisticated industry by this time. Here as in Iran bookcovers took their cue from manuscript illumination. Unfortunately very few Mamlûk bindings are dated and their development is therefore hard to trace. Virtually all of them are used on Qur’âns. Perhaps the most telltale Mamlûk device is the use of an overall geometric framework of strapwork of stellar or lattice type which fills the entire field, as in Coptic bindings. This is a taste virtually unrepresented in Iran after the 14th century, as is the emphasis on blind-tooled knots and interlace patterns. Inscriptions pack the borders. The ground is often heavily punched and gold-tooled dots or other elements further embellish the design. Rosettes and guilloche patterns alleviate the severity of the geometrical layout. Later Mamlûk covers are closer to Iranian examples in their use of an openwork central medallion with floral or arabesque infill, tapering finials above and below and quadrants with similar decoration, the whole set against a coloured background. Egyptian binders were much bolder in their use of filigree than their Iranian counterparts, using it freely for the outer covers as well as for the doublures. The bindings of the medieval Yemen, like its metalwork, were offshoots of the Mamlûk tradition but have some distinctive local features. Chief among these is a marked respect for the natural qualities of the leather, a factor explicable by the highly developed Yemeni tanning and leatherwork industry. This was already well established in early ‘Abbasid times and

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 45A391C8 /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

459

resulted in leather being brought from great distances (from as far afield as Kirman in south-eastern Iran) to be worked there for export all over the Islamic world. The major leatherworking centres were Sa‘da, Najrân, Zabîd, al-Tâ’if and San‘a’, the latter being particularly renowned for its bookbindings. Published examples – certainly a mere fraction of those that survive – date mainly from the 15th-16th centuries. Yemeni covers use exclusively blind tooling and this is kept very flat. Their main feature is a central rayed roundel; the quadrants have long diagonal finials and the double borders are packed with repetitive stamped designs including overlapping lozenges, grid patterns and naskhî inscriptions with such messages as blessings to the owner or expressions of faith in God. The quarto format permits a large field with plenty of space for the grain of the leather to become a decorative feature in its own right. Medieval Maghribî bindings are substantially rarer than those of Egypt or Iran. Typologically they are closest to Egyptian models, for the core of the cover is made up of several sheets of paper glued together, and strapwork or stellar designs with floral infill predominate. However, the quality of the strapwork is poorer and the endless patterns of Egyptian covers are rejected in favour of a closed independent roundel or a centralised interlace motif. The flap too has this kind of circular design, and round punches predominate. Thus Maghribî bindings do form an independent school; but the historical importance of the school lies rather in the way its influence extended northwards to Moorish Spain and thence to Naples and eventually Venice, thereby helping to trigger the great expansion of the Western bookbinding industry in Renaissance times. Venice was in fact the key middleman in this process, for it was through the agency of a group of Islamic bookbinders settled there that arabesques, morocco leather and gold tooling were introduced into the European craft of bookbinding. The Maghrib also has the distinction of having produced the earliest surviving example of gold tooling (i.e. impressing designs on the leather by applying heated tools to gold leaf). This technique is found on the binding of a Qur’ân copied in Marrakesh c. 1256 for an Almohad ruler. Despite this apparently precocious technical advance, Maghribî covers usually confine gilding to a few lines or dots as grace notes of the design. The later medieval Iranian school of bookbinding opens modestly enough with the cover of the Manâfi‘-yi Hayawan in the Pierpont Morgan Library in New York, dated 1290. This constitutes a rather humble prelude to a splendid manuscript, with its slender central mandorla, grid-like corner

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 425C476B /production/pindar/booksxml/hillen/chapter−15

460 triangles and foliate roundel on the flap. Probably no more than ten punches are used, and in earlier Persian bindings, such as an austere, perhaps 13th century, example from Nâ’în, the number is as low as four. This rather stark style, which finds its natural analogue in Saljûq book illumination, recurs on an Iranian Qur’ân dated 1304 and bearing the signature-stamp of the binder, but already extra detail has crept in, and a glance at the series of dated or datable 14th century Iranian bindings does show a steady accumulation of decorative detail, notably in the expansion of central medallion and corner-pieces alike, which clearly paves the way for the great Timurid bindings. Notable steps in this development are the celebrated Qur’ân produced for Oljeitû at Hamadan in 1313 (now in the Egyptian National Library, Cairo), another Qur’ân produced in Marâgha in 1338–9 (now divided between the Chester Beatty Library, Dublin and the Museum of Fine Arts, Boston) and the cover of the Mu’nis al-Ahrâr of 1341 (formerly in the Kevorkian Foundation, New York). Further significant advances are marked by two bindings formerly in the Evqaf Museum, Istanbul – one from Tabriz dated 1334, which has the earliest polylobed medallion yet recorded, and one from Shîrwân dated 1379, with baroque, mobile plant forms whose Chinese flavour presages Timurid work. The time lag vis-à-vis book painting in this respect is almost a century, a reminder of the innate conservatism of this art. The series ends with a Qur’ân of 1387 in Philadelphia. Seen as a group, these bindings illustrate an increasing tendency towards integrating the design by making each corner-piece a quarter of the central medallion, and they also document the gradual invasion of a geometrical framework by floral infill. From this it was obviously a natural step to jettison the geometrical designs altogether, and this is what occurs in the following century. Blind tooling is the standard decorative technique for much of the 14th century in Iran, but gilding and multiple pigments were increasingly employed for extra emphasis and finesse. It is generally accepted that the high-water mark of Islamic bookbinding was reached in 15th-century Iran under Timurid patronage. The range of techniques and ornament expanded dramatically in this period. The house style especially favoured bindings of dark brown kid; their cynosure was typically a central mandorla whose interior designs were repeated in miniature in the corner quadrants. Large expanses of smooth plain leather were set off by multiple borders packed with interlace, inscriptions, floral patterns or combinations of these elements. Thus the

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 437187D1 /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

461

predominantly geometric quality of earlier book-covers now gave way increasingly to floral designs and arabesques, a process most clearly traceable in Shiraz bindings of c. 1400–1450. Such stately and somewhat severe medallion designs were supplemented by a wide range of covers which have quite different emphases but maintain the same proportional ratio between field and border. Sometimes the entire field is filled with tightly scrolling stems which at intervals produce blossoms and leaves. Or the familiar sunken medallion and corner-pieces may be lobed in a style resembling the Chinese cloud-collar, occasionally with a second medallion inside the first. The medallion itself may take the form of a star, or a rayed circle like the shamsiyya used for the ex libris in contemporary manuscripts. While the flap often repeats a segment of the cover decoration, it is not uncommon for the front cover, back cover, flap and doublure all to have their own individual designs, and of course this adds greatly to the sumptuousness of the binding. Much of the ornament is executed in gold tooling; indeed, so popular was gold that it was used in leaf, powder and liquid form. This virtuosity in the application of gold is a new feature in Islamic bookbinding. As in contemporary Egypt, the better protected inside covers frequently bore fragile arabesque designs executed in leather, and eventually paper, filigree (munabbat kârî). This technique, though admittedly a late motif of Timurid bindings, had already been used in Coptic Egypt and Persia’s 16thcentury Vasari, Dust Muhammad, is therefore certainly wrong in attributing its invention to one Ustâd Qiwâm al-Dîn whom Baisunghur b. Shâh Rukh brought from Tabriz to Harât. Its use for figural designs is, however, an important innovation. One of the earliest and finest examples of the technique is found on a copy of the poems of ‘Attâr produced in Harât in 1438 (Topkapı Sarayı Library, Istanbul, AIII 305g). The pre-eminence of Harât in this technique is, incidentally, typical – for although Timurid bindings from Yazd, Shîrâz, Abarqûh and Isfahân have survived, it was Harât which now became the major centre for bookbinding in the eastern Islamic world. Indeed, by the time that the Harât scriptorium was reconstituted at Tabrîz by Shâh Ismâ‘îl in the early 16th century, the bookbinders were sufficiently important and independent to be organised in a guild, as their counterparts in Cairo had been for centuries. Without doubt, the sharpest break with previous conventions was the introduction of animals disporting themselves within a landscape, even if the latter was at times suggested only by profuse foliage. The choice of

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 4343681C /production/pindar/booksxml/hillen/chapter−15

462 animals – fabulous creatures like the dragon, feng-huang and ch’i-lin – betrays the close dependence on Chinese designs, which presumably reached Iran through the medium of textiles or paper. Even in bindings of traditional medallion format with ansate cartouches in the borders, these animals – with other Chinese motifs like cloud-collars, ju-i heads and lotuses – are likely to invade almost every available space. Human beings and still more animals such as bears, deer, hares, foxes and monkeys are frequently found in wrestling and especially hunting scenes. These echo contemporary miniatures and slightly later carpets, and even the age-old theme of the lion bringing down the bull reappears (Türk ve Islam Esereli Müzesi, Istanbul, no. 1541). These figural scenes bring the iconography of book painting (especially marginal decoration) onto the cover itself, and often announce the content of the volume. This integration of text and cover is altogether novel. The introduction of figural motifs into Timurid bookbindings had a profound effect upon technique. The abstract designs favoured by earlier generations had been produced by individual die-impressions which in the more sumptuous bindings, such as the cover in the Chester Beatty Library, Dublin, dated 1434, could take years to complete. It has been calculated that this binding required c.550,000 blind stamps as well as c.43,000 gold stamps. Free-hand cut filigree work employing multiple layers of leather was also very laborious; the finer details were modelled with a needle. The new figural motifs, however, generated a new and labour-saving technique, that of the metal block, which could be as large as the cover itself. This was a development of large-unit plaque moulding in which stone was used as the matrix. Moulded abstract designs were characteristic of early Timurid bindings (e.g. TKSL no. 976, produced in Yazd in 1407, or British Library Qr. 2833, produced in Shîrâz in 1405) and were the forerunners of the more elaborate pictorial designs developed later in the century. Metal of course had much greater potential than stone. Indeed, by Safavid times these large matrices of copper or steel were applied ready-heated and –gilded to the leather, so that several processes were telescoped into one. The substitution of paper for leather in Safavid filigree decoration was another labour-saving device. One early result of this metal-block technique was a new emphasis on the third dimension, so that the design acquired the quality of an embossed relief, with the raised portions often picked out in gold. Apart from the latter detail, the resemblance to the multi-layered carving of certain Timurid stone

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 3E7F0F17 /production/pindar/booksxml/hillen/chapter−15

ISLAMIC BOOKBINDING

463

sarcophagi is striking. Smaller brasses were also used for these designs and they varied greatly in size and shape; they included long cartouches, right angles, roundels, lobed medallions and corner-pieces, thus allowing the designer maximum flexibility and thereby encouraging new types of composition which recycle familiar elements. Further contrast was afforded by painting on leather, and by filigree work with coloured silk as a backing (e.g. TVIEM ms. 1537, dated 1483). Some of the later Timurid bindings in fact exhibit a polychrome splendour which surviving earlier Islamic bindings cannot match (although the literary evidence mentions the white and yellow striped leathers of Cordoba, the red kid leather of Zabîd in the Yemen and ‘Abbasid bindings in brightly coloured silk); their shades include Venetian red, emerald and ultramarine, often with black used as a foil (TKSL no. R.1405). Moreover, it was standard practice for the doublures to have colour schemes as well as designs completely different from those of the outer covers. Nor did this exhaust the inventiveness of Timurid craftsmen, for there is even an example – unique in Iran though later Turkish examples are known – of a typical medallion cover executed in embroidery of blue silk and gold thread on black leather, datable c.1500 (Istanbul University Library, no. 2782/32). The complexity and virtuosity of this example, which cannot but recall Chinese craftsmanship, suggests that it was made in Harât, for this centre seems to have been unrivalled for the technical excellence of its bindings. Perhaps the most significant innovation of the period so far as later generations were concerned was the introduction of lacquered painting (itself a technique known in Iran as early as the 12th century) into the decorative repertoire of bookbinding. At first the technique was reserved for bindings of the utmost distinction, such as that of the copy of Rûmî’s Mathnawi completed for Sultan Husain Baiqara in Harât in 1483 (TVIEM, no. 1905). The process involved placing a layer of chalk on the pasteboard front cover (the back cover was of leather) and then coating it with a thin veneer of lacquer. This served as the ground for the painting and was subsequently overlaid with several films of clear lacquer. The topmost coat of lacquer was then further embellished with gold and silver paint and powdered mother-of-pearl. The latter substance was also frequently used as inlay material in conjunction with lacquer (TKSL no. H.676, dated 1496).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 41 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 3A8B91DE /production/pindar/booksxml/hillen/chapter−15

464 The smooth surface afforded by lacquer opened a new field to the miniature painter, though the full development of these possibilities was not achieved until Safavid times. Hunting scenes and animal combats in a sombre-hued forest setting were the commonest themes. Both monochrome and polychrome lacquer were now popular, and naturally enough some celebrated miniature painters, such as Sultân Muhammad and Muhammdî, turned their hands to lacquered painting. Hence, no doubt, the popularity of busy figural compositions in such book covers. This close connection between two related arts is established, as it happens, on the basis of literary evidence, but it is also a salutary reminder that bookbinding was not a craft on its own. Designs for bookcovers have survived in the Istanbul albums and it is clear that a designer (naqqâsh) might see his work realised in any one of several media – painting, textiles, carpets, woodwork, glazed tiles or book-covers. In the previous century the case of Zain al-‘Abidin b. Muhammad shows that a single craftsman could be at once scribe, illuminator and bookbinder, and that this continued in later times is shown by an Ottoman Qur’ân of 1848 in the Chester Beatty Library, Dublin. Conversely, the bookbinder might be essentially a leatherworker and therefore leave the actual sewing of the book to the warrâq, who concerned himself with numerous other aspects of book production and sale. Since the vogue for lacquer extended to other objects, including large boxes and even doors, it is not surprising that Safavid artists developed an unrivalled expertise in the use of this medium. Later Safavid polychrome lacquer book covers favoured overall floral patterns against gold or russet grounds, and exerted significant influence on Turkish and Indian bindings. In Turkish lacquered book covers of the 18th and 19th centuries, which are found mostly on Qur’âns, the colour scheme was slightly different, for a black or lemon ground was common, and there as in India the local variants of floral motifs (e.g. carnations and serrated leaves) supplanted the cognate Iranian motifs. Lacquerwork experienced a vigorous revival under Fath ‘Ali Shah (d. 1834) and book covers in somewhat garish tones such as carmine and malachite enjoyed their full share of this popularity. The range of subjects was now larger than ever and besides the ever-popular naturalistic flower pictures included familiar set-pieces from Nizami and Firdausi as well as scenes of Qajar court life.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-15 Sequential 13

JOBNAME: ???? PAGE: 1 SESS: 38 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 3833F5AC /production/pindar/booksxml/hillen/chapter−16

XVI The Schefer Harı¯rı¯: A Study in Islamic Frontispiece Design T has long been generally recognised that the copy of the Maqa¯ma¯t of al-H.arı#rı# in the Bibliothèquc Nationale at Paris, dated 634 (1237)—BN ms. arabc 5847—is by far the most distinguished of the many illustrated versions of that text (fig. 1).1 It holds that position by virtue of its peerless assemblage of 99 pictures and its fine calligraphy, all the work of a single artist, Yah.ya¯ b. Mah.mu¯d b. al-Wa¯sit.#ı. Given the embarras de richesse provided by the illustrations to the text, it is perhaps not surprising that the frontispiece of the manuscript should have been somewhat neglected; but it richly repays detailed scrutiny in its own right. It draws discriminatingly on a wide range of visual sources and wields their motifs with an enviable certainty of touch. The aim of the present paper is to interpret this unusual pair of pictures and at the same time to analyze some of the conventions that dictated the form of images of authority in the medieval Islamic world. To do this it will be necessary to study first the contemporary conventions which governed how the opening pages of an illustrated manuscript should be treated; next, the general layout of this particular double page, both as a whole and, more specifically, as regards the panels enclosing the principal figures; then the compositional principles followed in each of the two main pictures; and finally, the often minute, but nevertheless diagnostic, differences between the two images of authority. It is these differences that will provide evidence to justify the interpretation of their meaning which is proposed below. What follows is essentially a formal analysis, for that promises the best chance of solving the problem of interpretation posed by the two main

I

1

The manuscript is now available in facsimile form; Maqa¯ma¯t al-H . arı#rı# 2004.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16 Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 38 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 4ABBF25C /production/pindar/booksxml/hillen/chapter−16

466 figures; besides, the fact that scarcely any detailed analyses of any 13th-century Islamic frontispieces have been published to date2 suggests that it is high time that such an attempt was made. Contemporary conventions of frontispiece design in the context of the Schefer H.arı#rı# In a manuscript so variously original it is small wonder that the opening pages should infuse time-honoured conventions with new life. The volume opens in an acceptably traditional manner; fol. 1 recto is largely blank and its sole decoration is a large tabula ansata3 whose central panel bears the words al-maqa¯ma¯t al-H.arı#riyya in cursive script (fig. 2). The geometrical interlace and scrolling arabesques of the whole composition, as well as the endpieces of the tabula ansata and the colour scheme of blue, gold and white, are entirely in the spirit of su¯ra headings in contemporary Qur’a¯n illumination4 and show the artist’s mastery of that idiom, though these echoes of the sacred book occur in an emphatically non-sacred context. The principal visual impression which this page makes is one of emptiness. Yet contemporary and earlier Qur’a¯ns show that ample precedent existed for an opening page5 (or pages)6 totally devoted to illumination. Perhaps the artist was understandably chary of adopting this latter procedure here, as it might have brought his manuscript inappropriately close to a style which carried unmistakable religious associations. This relative emptiness does, however, have one positive advantage: it acts as a curtain-raiser for the following two pages. Its plainness provides an excellent foil for their prodigal wealth of design and ornament. Turning the page for the first time from the Spartan austerity of folio 1 recto to the visual plenitude of folios 1 verso and 2 recto, where all the space is painted, most readers might well experience a start of surprise. Most

2

The most comprehensive attempts in this direction were those of Richard Ettinghausen in his classic Arab Painting, for which see Ettinghausen 1962, pp. 62–71, 74–8, 92, 98–103; for a discussion of the Schefer H . arı#rı# frontispiece, see pp. 111 and 114 -5. See also the analyses of individual frontispieces in Hoffman 1982, pp. 254–305; she deals with the Schefer H . arı#rı# on pp. 193 and 259–60. Cf. also Hoffman 1993, p. 8. The fullest published account of the Schefer H . arı#rı# frontispiece is that of Grabar 1984, pp. 21–3. 3 Herzfeld 1916, 189–99. 4 See James 1992, pp. 43 and 57; Parham 2001, pp. 62–3. 5 James 1992, pp. 46–7, 53 and 61. 6 James 1992, p. 25; cf. too the Qur’a¯n of Ibn al-Bawwa¯b in the Chester Beatty Library (Rice 1955, pl. III).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16 Sequential 2

JOBNAME: ???? PAGE: 1 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 12A0BD67 /production/pindar/booksxml/hillen/chapter−16−artwork

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

467

A. Frontispiece. Dioscorides, Hiyu¯la¯ cila¯j al-t. ibb, Baghdad or North Jazı¯ra, dated 621 (1224). Istanbul, Süleymaniye Library, No. 3704, fol. lv.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 085D6E97 /production/pindar/booksxml/hillen/chapter−16−artwork

468

1a. Double frontispiece. Al-H.arı¯rı¯, Maqa¯ma¯t, Baghdad, dated 634 (1237). Paris, Bibliothe`que nationale de France, arabe 5847 (Schefer H.arı¯rı¯), fol. 2r (Copyright of the Bibiothe`que nationale de France).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 0ADCE772 /production/pindar/booksxml/hillen/chapter−16−artwork

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

469

1b. Double frontispiece. Al-H.arı¯rı¯, Maqa¯ma¯t, Baghdad, dated 634 (1237). Paris, Bibliothe`que nationale de France, arabe 5847 (Schefer H.arı¯rı¯), fol. 1v (Copyright of the Bibiothe`que nationale de France).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 0871005B /production/pindar/booksxml/hillen/chapter−16−artwork

470

2. Title page. Al-H.arı¯rı¯, Maqa¯ma¯t, Baghdad, dated 634 (1237). Paris, Bibliothe`que nationale de France, arabe 5847 (Schefer H.arı¯rı¯), fol. 1r (Copyright of the Bibiothe`que nationale de France).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 092CE0B9 /production/pindar/booksxml/hillen/chapter−16−artwork

471

3. Double frontispiece. Ibn al-S.u¯fı¯, Risa¯lat al-S.u¯fı¯ fi’l-kawa¯kib, Baghdad or North Jazı¯ra, ca. 1220–25. Tehran, Reza Abbasi Museum, M. 570, pp. 2–3 (Courtesy of the Reza Abbasi Museum, Tehran).

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 07CE7672 /production/pindar/booksxml/hillen/chapter−16−artwork

472

4a. Double frontispiece. Dioscorides, Hiyu¯la¯ cila¯j al-t. ibb, North Jazı¯ra?, dated 626 (1229), Istanbul, TKS Ahmet III, 2127, fol. 2r. (192 × 140mm) (Courtesy of the Topkapı Sarayı Museum, Istanbul).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 09BDBE74 /production/pindar/booksxml/hillen/chapter−16−artwork

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

473

4b. Double frontispiece. Dioscorides, Hiyu¯la¯ cila¯j al-t. ibb, North Jazı¯ra?, dated 626 (1229), Istanbul, TKS Ahmet III, 2127, fol. 1v (192 × 140mm) (Courtesy of the Topkapı Sarayı Museum, Istanbul)

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 079C0C7D /production/pindar/booksxml/hillen/chapter−16−artwork

474

5. Frontispiece. Ibn Bakhu¯shu¯ c, Kı¯ta¯b na‘t al-h.ayawa¯n, North Jazira?, ca. 1220–25. London, British Library, Or. 2784, fol. 4r (Copyright of the British Library).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 097500E6 /production/pindar/booksxml/hillen/chapter−16−artwork

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

475

6. Frontispiece. Pseudo-Galen, Kı¯ta¯b al-dirya¯q, North Jazı¯ra?, ca. 1250. Vienna, Nationalbibliothek, A.F. 10, fol. 1r. (320 × 255 mm) (Courtesy of the Nationalbibliothek, Vienna).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 0755EC39 /production/pindar/booksxml/hillen/chapter−16−artwork

476

7. Frontispiece. Dioscorides, Hiyu¯la¯ cila¯j al-t. ibb, Oxford, Bodleian Library, Cod.or. d. 138, fol. 2v (Courtesy of the Bodleian Library).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 0A85E06F /production/pindar/booksxml/hillen/chapter−16−artwork

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

477

8. Finispiece. Al-Mubashshir, Mukhta¯r al-h.ikam wa mah.a¯sin al-kalim, Syria?, early 13th century Istanbul TKS, Ahmet III, 3206, fol. 173v. (250 × 142 mm) (Courtesy of the Topkapı Sarayı Museum, Istanbul).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 31 OUTPUT: Tue Aug 10 10:30:03 2010 SUM: 056D4DA9 /production/pindar/booksxml/hillen/chapter−16−artwork

B. Dina¯ withi ‘standing caliph’, Umayyad, 74 (693–4) (drawing by Dzul haimi b. Muhammad Zain)

478

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-artwork Sequential 12

JOBNAME: ???? PAGE: 1 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 520C0602 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

479

significant of all, the decision to use the opening leaf of the book as a title page, and not as a frontispiece, presented the artist with the possibility of creating a frontispiece which spread over two pages instead of being confined to one.7 The challenge was, of course, self-imposed. Yet it is important to be aware that, so far as the extant evidence suggests, no ready-made iconography for a double frontispiece existed in the Islamic world.8 In later manuscripts, double frontispieces feature spacious alfresco settings in which the delights of princely life unfold in leisurely fashion. Attention shifts from the prince’s image itself9 to an imaginative evocation of his entire lifestyle. Ample space is a prerequisite for this somewhat freewheeling approach to royal iconography. In the body of the Schefer H.arı#rı#, al-Wa¯sit.#ı does occasionally show his awareness of treating a double-page spread as a unit, with the action spilling over from one page to the other in as natural a way as he can devise. But these ideas,10 which sometimes reveal an engaging wit, are not developed very far. In his double frontispiece they are present only as a dim premonition. Moreover, his was not, as it turned out, the way of the future. Perhaps the extreme formality and sense of hierarchy which stamps his images of authority did not recommend themselves to later painters because his approach excluded so much of the

7 Naturally, it was always open to the designer of a book to place the frontispiece further into the volume, and thereby to secure a double-page spread, and this was indeed done. One must also reckon with the possibility that the present state of the opening pages of a given codex does not reflect its original appearance; often enough, it is precisely the opening pages that have vanished and that have been replaced with later leaves—see, for example, a couple of manuscripts of al-S . u¯fı#’s Kita¯b .suwar al-kawa¯kib: TKS 4293 of 525 (1130–1) (Holter 1937, p. 3) and a copy dated 519 (1125) (Sotheby’s 1988, p. 35). 8 Hence the very different and frankly experimental solutions to this challenge that artists devised at this period, as shown by the three manuscripts with double frontispieces that predate the Schefer H . arı#rı#: the Istanbul Dioscorides (Ettinghausen 1962, pp. 68–9), the Paris Kita¯b al-Dirya¯q (n. 16 below) and the Tehran al-S . u¯fı# (Parham 2001, pp. 268–9; see fig. 3). 9 See Bernus-Taylor et al. 1983. 10 For example in the preaching scene at Rayy in maqa¯ma 21 (fols. 58v-59r), where the khat.#ıb’s voice is projected as it were across the gutter of the manuscript to the more distant section of his audience (for a colour illustration, see Hillenbrand 1999, pl. 2 and unnumbered plate at end of the section of colour plates; cf. Grabar 1984, pp. 64–5).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 1

JOBNAME: ???? PAGE: 2 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 440CB707 /production/pindar/booksxml/hillen/chapter−16−ends

480 normal contemporary repertoire of royal iconography. Within its selfimposed limits it was indeed an impressive achievement—but also in some sense a dead end. Al-Wa¯sit.#ı, then, seems to have taken an unusual though not unprecedented step in giving his book a double frontispiece. It is worth remembering that he did not have to use both pages for frontispiece themes. He could, for example, have painted a traditional frontispiece on folio 1 recto; or at folio 1 verso, and then begun the text proper on the opposite page. Perhaps it was precisely because this visual juxtaposition of frontispiece and text page was regarded as undesirable—since it obviously detracted from the pre-eminence of the frontispiece image—that such images were normally placed on the opening folio, though occasionally they were located as late as folio 4 recto. The reason why al-Wa¯sit.#ı chose to break with convention in this matter will probably never be known. His inclusion of an image of power certainly suggests at first sight that the manuscript was produced at the behest of a specific patron, as indeed common sense would indicate. Yet the colophon, full as it is, makes no mention of a patron. The fact that al-Wa¯sit.#ı copied the manuscript as well as illustrating it, and seems thus to have been an independent operator rather than a member of a production team, speaks for his originality and should prepare one for distinctly personal solutions to problems of layout and iconography. If it is strange enough for the manuscript to have a double frontispiece, it is still stranger—and no Islamic parallels come to mind at this period—for that frontispiece to comprise two images of authority and not one. If one figure were of lesser authority than the other, it would have been open to the artist to place it on a succeeding page. As will shortly appear, the figure on the viewer’s right is probably a ruler and the figure on the left is of lesser importance. Yet placed as it is opposite an image of power, it competes with that image and indeed cannot help doing so. The difference in their faces indicates clearly enough that these are two different people, not one and the same person seen in different guises or activities. The artist has therefore created a serious problem for himself, being forced to reconcile the irreconcilable. Other Arab frontispieces are either not double or, if they are, take good care to ensure that the design of the subordinate (usually left-hand) page does not lure the eye away from the pre-eminent figure of authority on the opposite page. Al-Wa¯sit.#ı’s ‘double’ frontispiece seems essentially to comprise two adjoining single frontispieces. Yet, as we shall see, there is rather more to it than that.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 2

JOBNAME: ???? PAGE: 3 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 5020E890 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

481

A brief survey of Islamic illustrated manuscripts produced before 1300 is enough to reveal that figural frontispieces were by no means the norm. Among the eight 13th-century illustrated Maqa¯ma¯t manuscripts, for example, only one, apart from the 1237 example, perhaps had a frontispiece.11 Most of the 13th-century illustrated texts existing in one or two copies have no frontispieces at all. Among the more popular books, or those which have been preserved in several copies, two texts are of special interest in this connection: the Kı#ta¯b al-Agha¯nı# and the De Materia Medica of Dioscorides in its Arabic version. The former is known in an incomplete luxury edition prepared for Badr al-Dı#n Lu’lu’ of which six volumes, each with a one-leaf royal frontispiece and no other illustrations, have survived.12 The Dioscorides text exists in eleven illustrated copies predating 1300, of which five have frontispieces.13 None of these are royal, but one is a double frontispiece, with the seated author on the right and two students on the left presenting him with a book (fig. 4). The total of illustrated Islamic manuscripts and detached leaves datable before 1300 was computed over sixty years ago at 49,14 and on the basis of that published evidence it seems that only 11 had frontispieces. Five of these are to be found in Dioscorides manuscripts and two at most in Maqa¯ma¯t ones. Some of these frontispieces are too damaged to be interpreted in detail. Others may have been double but, if so, only one leaf has survived. In this list, only six are indisputably double frontispieces.15 Clearly these statistics are not entirely satisfactory, since other illustrated manuscripts have come to

11

This is in Paris 6094, of which only traces of a frame—which could equally well have held a dedicatory inscription or a title—remain; see Grabar 1984, pp. 8–9. 12 The best general discussion of these images is still that of Rice 1953. See also Johnson 1975, pp. 39–128. 13 See Grube 1959, pp. 163–80. 14 The sources for this figure are Holter 1937, supplemented by Buchthal, Kurz and Ettinghausen 1940. 15 These are al-Mubashshir’s Mukhta¯r al-H . ikam wa Mah.a¯sin al-Kulim (Istanbul, TKS Ahmet III, 3206), the Dioscorides or 626/1229 (Istanbul, TKS Ahmet III, 2127), the Schefer Maqa¯ma¯t, the Rasa¯’il Ikhwa¯n al-S . afa¯’ (Istanbul, Library of the Süleymaniye Mosque, Esad Efendi 3638), the Ta¯rı#kh-i Jaha¯n Gusha¯ (Paris, Bibliothèque Nationale, MSS or., Suppl. Persan 205) and a textless royal frontispiece (Istanbul, TKS Hazine 2152, fols. 60v–6lv; in the substantial bibliography on this frontispiece, three items stand out for their careful and original scholarship: Schulz 1914, I, pp. 78–81; Esin 1963, pp. 143–61; and Fitzherbert 2001, 1, pp. 35–52). One might add to these six a further two manuscripts that came to light after 1940: the Paris Kı#ta¯b al-Dirya¯q (see next note) and the Tehran al-S . u¯fı#.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 3

JOBNAME: ???? PAGE: 4 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 49AAD258 /production/pindar/booksxml/hillen/chapter−16−ends

482 light since 1940,16 but they do serve to give an order of magnitude, and to underline the relative rarity of the particular frontispiece under discussion. The contemporary context suggests that frontispieces imply royal patrons, or at any rate figures of secular authority (fig. 5), for the books in question; their mere presence in a manuscript conferred on it a certain éclat. Finally, the context of the 1237 H.arı#rı# suggests that no norm had yet developed for frontispieces, a point made with particular forcefulness by the totally different frontispieces of the Vienna Kita¯b al-Dirya¯q (fig. 6) and its counterpart in Paris. So much, then, for contemporary practice in the matter of figural frontispieces. The layout of the Schefer H.arı#rı# frontispiece as a whole The next topic for analysis is the aspect presented by the double page as a whole. As the introduction to the whole book and the reader’s first exposure to its figural style, its importance is critical. The painter lavishes on these two pages all his powers as a figural painter and as a designer of ornament. The richness of effect thus achieved is breathtaking. A welter of ornament bids fair to engulf the pictures themselves. Yet already here, at the very opening of the book, the painter serves notice of his readiness to make creative use of ideas taken from quite disparate media and artistic traditions. The upper and lower margins of the two pages are ornamented at their outer edge by a series of tessellated forms which recall the outermost surrounds of Qur’a¯n leaves of this period and earlier,17 and also 13th-century metalwork,18 though they also bring to mind the fringes of rugs. Whichever of these ideas is at work, there can be no doubt that the associations of this motif were with costly artefacts. It is even possible that these two pages represent an attempt to combine two quite separate traditions of frontispiece design. One, obviously enough, is the princely image. The other is the full-page ornamental frontispiece, a 16

For example, the Paris Kı#ta¯b al-Dirya¯q (Farès 1953); the Tehran al-S . u¯fı# (n. 8 above); and a Kalı#la wa Dimna of c. 1260–85 in Istanbul (TKS, H.363; see O’Kane 2003, pp. 49 and 228). Note that the Paris manuscript (like the Mukhta¯r al-H . ikam), essentially repeats two virtually identical scenes on facing pages, rather in the manner of Qur’a¯nic frontispieces. The Mukhta¯r al-H . ikam, like several Qur’a¯ns (James 1992, p. 27 and Rice 1955, pl. IV) also has a double finispiece whose two halves are virtually identical. 17 James 1992, pp. 53–5. 18 Baer 1983, figs. 75, 91 and 199; Ward 1993, pp. 75 and 81.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 4

JOBNAME: ???? PAGE: 5 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 4E0C536E /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

483

concept already ancient in the 13th century, and here perhaps even further enriched by a visual association with a fringed carpet with many borders. Within the outer tessellated fringe each of the two pictures is enclosed by a thick band of intersecting vegetal scrolls. Inside these is a menagerie of animals disporting themselves—rabbits, hares, felines, dogs and birds of prey. The ultimate source of this motif is of course the inhabited scroll of late antiquity. It is a theme which enjoyed wide popularity in various Islamic media such as textiles and ceramics, and can also be found in neighbouring Christian contexts, as in Crusader art, where the outstanding examples are the ivory covers of the Psalter of Queen Melisende, datable c. 1135,19 and the frontispiece of a copy of the Histoire Universelle produced in Acre c. 1285.20 Here in the Schefer H.arı#rı# it is firmly relegated to secondary status, confined as it is to a border surrounding the figural compositions that are the main focus of interest. The continuous inhabited scroll is one of the most popular border designs in medieval Islamic metalwork.21 A favoured motif here is a key image surrounded by a field of dense ornament which, being greater in area, engulfs it, but, being lesser in significance, gives way to it visually.22 The obvious visual purpose of the border is to frame the two central panels and not to compete with them. Hence its density, its repetitive quality and its predominantly gold tonality (another metalwork effect), which tends to make an immediate visual impact as a mass of undifferentiated colour rather than to emphasise the individual features of its design. Since each of the two pages is laid out in the same way, the effect is to create a thick quadruple border between the two central images. Within the shimmering backcloth thus created, the two figural panels are encased—again, metalwork provides the natural analogy—like jewels in a rich setting. It is worth insisting on this point, since other contemporary frontispieces make far less play of ornament, preferring to devote to the figural composition itself the extra space thus made available. It should be noted, too, that the impact of the surrounding ornament is magnified by the multiple blue and gold borders which set off the inhabited scroll design on either side. These contrive to distance still further the separate, enclosed 19 Buchthal 1957, especially pp. 1–14; Folda 1995, pp. 137–59 (for the ivory covers, see especially pp. 152–3, pls. G, 10a–b, and pp. 157–8); and Hoffman, Emergence, 217–20. 20 Buchthal op, cit., pp. 80–2 and pl. 83; Folda 1976, pp. 77–80; Hoffman 1982, pp. 213–6; Folda 2005. 21 Baer 1983, figs. 22, 55, 59, 61, 74, 126–8, 151. 155–6; Ward 1993, pp. 70, 87 and 89. 22 Marchal et al. 1971, pp. 103–4 and pls. 150–1.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 5

JOBNAME: ???? PAGE: 6 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 3EBBE581 /production/pindar/booksxml/hillen/chapter−16−ends

484 world of the central panels and thus to enhance the authority of the images depicted in them. By that account, then, the decision to devote most of the double frontispiece to ornament should be interpreted less as evidence of a preference for ornament above figural images and more as a characteristically original attempt to make ornament serve new purposes. Clearly there is a spatial element in all this; and indeed, a sensitivity to nuances of space can be detected throughout the pictures of this manuscript. Certainly the notorious Islamic horror vacui is at work in the border panels, and might be thought to exclude the possibility of any spatial subtleties; but this density itself highlights the comparatively freer space within the panels. Thus a deliberate contrast is being set up between the borders and the panel itself. Both are needed if the effect—to concentrate the viewer’s gaze on the main figures—is to work. These details apart, the cumulative impact of the borders—frame within frame within frame-—is as if the figural images were being studied from the wrong end of a telescope. The eye is drawn ineluctably to the centre of each page. This use of successively smaller frames like Chinese boxes is essentially a spatial device, for in the end they create a three-dimensional impression for the central images. It may be no more than an illusion, a trick of perception rather than of perspective; but its success can scarcely be denied. Further devices within the panels themselves, to be discussed in a moment, foster this illusion. Given so subtle a deployment of illusionistic techniques, it is not surprising that the diminutive size of the central panels is not immediately noticeable. This is all the more remarkable since the figural panel takes up less than 23% of each page in its current state. With so little space at his disposal, the artist had to make every detail tell; and, that decision made, he could scarcely allow himself the luxury of, say, an expansive account of court life. In sum, then, a whole battery of devices is employed to focus attention on the central panels, and all work in concert. Such deliberation presages an even greater concentration of meaning in the images themselves, and excludes the presence of casual or random detail. Even in a richly illustrated manuscript such as this, after all, it was still logical to expect the attention of the patron to home in on the frontispiece which showed him enthroned. The nature of the right-hand page gives us a clear enough idea of what kind of patron the artist had in mind.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 6

JOBNAME: ???? PAGE: 7 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 4858AC45 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

485

The layout of the Schefer H.arı#rı# frontispiece: the figural panels From the page as a whole one may turn to the figural images that dominate it. The compositional layout here illustrates the selfsame ability of the artist, already noted elsewhere in the picture, to ring the changes on very familiar themes. These changes are subtle, and individually they are no doubt minor; but collectively and cumulatively they invigorate an enfeebled iconographic theme. At the same time, the restricted compass within which these changes are effected is instructive. Whether it betrays the artist’s personal reluctance to embark on a thorough-going reworking of the frontispiece convention or rather reflects an exhausted tradition which could not accommodate any more radical change than this remains an open question. Beyond any doubt, though, it is the work of a man who understood intimately the tradition within which he was working and yet had not allowed familiarity to dull the edge of his perceptions. Hence his ability to manipulate familiar ideas in unfamiliar ways, and to make every detail tell. This was a hard act to follow. In a sense it is the intellectual density of this frontispiece that most clearly attests its maturity and that makes it natural that the more innovative artists of the next two generations should develop the frontispiece along appreciably different lines. The H.arı#rı# frontispiece, then, may claim to be a coda to an old tradition rather than a point of departure for something new. Broadly speaking, the layout of the figural panels uses a straightforward three-tier system. These tiers may conveniently be considered in turn. Each of the lower tiers contains standing people, apparently all men, crammed close together. An obvious mirror symmetry or duplication operates here. This symmetry is evident not only in the almost exact consonance of numbers from one panel to the other, but also in the poses of the men and the order in which they occur. In each panel the two central figures among the ‘lower orders’—that cliché has a compositional justification here—have their backs to the viewer. The remaining figures in this tier are all depicted in three-quarter view. The turned backs represent one of the very earliest uses of the repoussoir device in Islamic painting. Its purpose here is to reinforce the three-dimensional effect already secured by the multiple frames. Indeed, that effect is so accentuated as to draw the viewer still further into the picture. The three figures as it were climbing up into the middle tier are conceived in the same spirit. Nevertheless, it must be conceded that the full spatial potential of this group of figures is not realised. In other artistic traditions such a grouping is often turned into a semi-circle by means of a whole gamut of

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 7

JOBNAME: ???? PAGE: 8 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 486BE903 /production/pindar/booksxml/hillen/chapter−16−ends

486 poses.23 These extend, by successive nuances of three-quarter view and profile, all the way from total frontality to a back view. As befits its greater importance, the middle level is physically larger than the tier below, but is not utterly divorced from it. The trio of figures encroaching into the middle tier obviously forms a link between the lower and middle levels. Yet their significance extends beyond this. Indeed, it is central to the entire interpretation of the central panels. The left-hand image makes sense only if the slightly turned head of the man with the turban has some figures nearby with which to make contact. Once this is accepted, the emptiness of the space to his left is self-explanatory: the presence of figures there would be counter-productive. Hence the blue triangle of empty space in that very place. The principle of mirror symmetry dictates that the same triangle should occur to the right of the enthroned ruler. This has a further consequence—that the visual function of these two blue triangles, which draw the eye because of their concentrated mass of colour, is to integrate the two panels and to affirm that they occupy the same continuous space. Certain basic facts about the layout of this central tier stand out. They may seem painfully obvious, but that should be no reflection on their importance; rather the contrary. For the sake of convenience these elements will be considered in the context of the image of the amı#r; the image opposite shares the same characteristics. The differences between the two images are another matter altogether and will be analysed in detail in the next section of this paper. What, then, are these basic compositional features?24 To begin with, the amı#r is seated. Everyone else stands. It is a simple detail, but one that establishes his dominance in unmistakable fashion. Body language of this kind has lost virtually none of its significance in our own century and society; indeed, it transcends most barriers of time and culture. The prominence of the ruler is further stressed by the fact that he sits on a bolster with an elaborate kind of throne behind it.25 More significantly, he is allotted ample space whereas the people below him are crammed uncomfortably close together. This juxtaposition of ample and confined space has a long history in the royal iconography of the Middle East, finding expressions as varied as

23

Cf. too the St Petersburg Maqa¯ma¯t: Akimushkin el al. 1994, pp. 117–27. It is instructive to note how many of them recur in the Agha¯nı# images. 25 There is nothing particularly royal about this feature; it recurs frequently in everyday scenes in this very manuscript. Here the bolster (and turban?) bears a .tira¯z inscription. 24

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 8

JOBNAME: ???? PAGE: 9 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 4813CABE /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

487

Sasanian bas-reliefs26 and al-Mans. u¯r’s Baghdad. Whatever the local idiom, the basic message remains unchanged: it is a privilege, indeed a sign, of the ruler to enjoy the Lebensraum denied to his subjects. The features noted so far by no means exhaust the examples of an iconographic shorthand of power in this image. The main figure, for example, is central and this makes him the natural focus of the composition. The greatly exaggerated size of the ruler, which is of course intended to enhance his importance, is so familiar a characteristic of royal iconography in Islamic, Byzantine, classical and ancient Near Eastern art as to call for no further comment. Frontality is an equally hackneyed signifier in such contexts. The figure opposite, though not entirely frontal, is sufficiently close to that pose to partake of some of its advantages. Finally, in both cases the central figure is set physically above everyone else, and thus symbolically beyond the ken of the common man. In sum, then, the figure of authority is exalted by a battery of visual devices using a terse, direct idiom which even today has lost little of its power. While the differences between the two main figures are of crucial relevance to a correct interpretation of their meaning, it must be remembered that they are minor. They should not deflect attention from the raison d’être of these figural panels, which is to exalt the two images of authority by all the means at the artist’s disposal—except, curiously enough, inscriptions. Despite the crowded nature of the two main pictures, room could still have been found for identifying .t ira¯z bands or an inscribed panel between the winged figures or below the audience. Yet it would be mistaken to read very much into the absence of inscriptions, for a brief glance at other contemporary frontispieces is enough to reveal that inscriptions were simply an optional extra in such compositions (cf. fig. A).27 There remains the upper tier of the composition. A degree of colour symbolism is brought to bear in this part of the picture: above the head of each figure is a patch of blue.28 Given the visual connection with the empty triangle on one side of each figure, the obvious conclusion is that this blue area is intended to signify the sky. This segment of sky is enclosed within a decorative arch profile. Here again an ancient tradition is at work: the idea of

26

Herrmann 1977, pp. 87–96. Brandenburg 1982, pp. 25 (Istanbul Dioscorides) and 209 (Bologna Dioscorides); and Ettinghausen 1962, p. 65 (Kı#ta¯b al-Agha¯nı#). 28 Cf. Rice 1953, p. 134. 27

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 9

JOBNAME: ???? PAGE: 10 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 3E9AF8DD /production/pindar/booksxml/hillen/chapter−16−ends

488 setting the figure of authority, usually divine or royal, within an arcuated niche.29 The theme is protean in its expressions—the cult statue in a pagan temple, the bishop in his apse, the ruler in a throne recess in his palace. This motif had established itself firmly in the Islamic world in the course of the first century AH,30 and had already entered the repertoire of Islamic frontispiece design (fig. 7). The red area above highlights this arch and may be intended to represent, admittedly only in symbolic fashion, a curtain or canopy, as is suggested by the presence of a winged figure in each spandrel. In some roughly contemporary frontispieces such winged figures are depicted actually holding a canopy or drawing a curtain aside, and it is therefore possible that this was the immediate association conjured up by such figures at this time. The location of the winged figures —be they genii, victories, angels or yet other beings —above the notional sky is clear enough indication that they inhabit another, heavenly world. It may be of interest to note that in both panels the two upper corners of the throne on which the main figure sits project into this uppermost space. Perhaps this detail is to be understood purely at the compositional level, as a linking device between tiers. Nevertheless, it could be intended in a symbolic sense too and would thus express the relationship between the figure of authority and the heavenly realm. Compositional principles: general considerations So much, then, for the treatment of each of the three levels of the figural panels. What of their organisation as a whole? Two major compositional principles may be observed at work here. One is a pronounced emphasis on axiality. The centrality of the main figure is underlined by features both above and below him. This creates a kind of spinal column that controls the layout of the entire picture. Above the ruler the chord of the picture is marked by the arch and by the empty space flanked (and therefore stressed) by the two winged figures; below him, the same caesura drives down between the two figures with their backs turned. Even the creatures in the margin are so

29

Miles 1952, pp. 159, 160 [fig. 6], 161–4. Miles 1952, pp. 162–4; Fehérvári 1972, p. 241. In architecture obvious examples are to be found at Qusayr cAmra and at Mshatta. 30

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 10

JOBNAME: ???? PAGE: 11 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 42E1E076 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

489

disposed as to accentuate this imaginary line bisecting the entire composition. It is of course the central figure who gathers up into himself, and becomes the focus of, this axial emphasis. The second compositional principle is the division into tiers (cf. fig. 8). It seems very likely that the intention behind this device is to mark off separate levels within an overall hierarchy. In that case, the higher the image is, the more importance attaches to it. Like so many other visual devices in these panels, this idea has a long history. That history may most conveniently be taken up at the point when the Byzantine consular diptych was reaching its apogee.31 In these objects, which were of course also symbols of power, the significance of the threefold lateral division was driven home by the content of each panel: servants or tribute-bearers in the predella, the figure of authority in the centre and winged figures in the upper panel.32 It will be noted that the Schefer H.arı#rı# closely follows this layout, even maintaining the traditional distinction between small oblong panels above and below and a larger rectangle at the centre. In time this image was Christianised and then employed in a distinctively new way: as a book cover. The Murano and Echmiadzin Gospels provide typical examples of this innovation.33 It seems a persuasively simple transition from external to internal decoration, from such a carved book cover to a painted frontispiece of identical theme and layout just inside the book cover. Ample other evidence exists to prove firstly that the craftsmen concerned with the arts of the Islamic book were well abreast of the practices followed by their Byzantine counterparts,34 and secondly that Islamic artists delighted in transposing Byzantine ideas into new media or locations. Paradoxically, it was the Islamic world that seems to have kept alive the smoothly constructed, economic and expressive iconography of the consular diptych at a time when it had fallen out of fashion in its parent civilisation.

31

Delbrueck 1929, pls. 9–25, 32, 43 and 48; Capps 1927. Delbrueck 1929, pls. 17–18, 20–1. 33 For the Murano diptych, see Weitzmann 1979, p. 403; for the Echmiadzin diptych, see Beckwith 1970, pl. 116 – but it is worth noting that this is only half the complete diptych, whose two parts complement each other in format and subject matter (Grabar 1968, III, pl. 146c). 34 See the general discussion of this topic in Ettinghausen 1962, pp. 67–80. 32

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 11

JOBNAME: ???? PAGE: 12 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 41BAC845 /production/pindar/booksxml/hillen/chapter−16−ends

490 The best evidence for this surprising survival is in the Vienna Kita¯b al-Dirya¯q. Here, as in the consular diptych, the whole of the surface is imbued with iconographic meaning, whereas in the Schefer H.arı#rı# the key motifs occupy less than a quarter of the page; and the picture surface is subdivided by actual lines into five unequal compartments, exactly on the analogy of a late antique model like the Athanasius diptych.35 Diagnostic distinctions between the principal figures and their environments It is now time to note the key differences between the two panels. An awareness of the motivation behind these distinctions will enable each picture to yield its full quota of meaning. The tally of such differences is surprisingly long, which is itself instructive in that it reveals how much calculation was required to build up an effective image of authority. Almost every detail counts towards the cumulative effect. The obvious place to begin is of course the figures themselves. The ruler is larger than the man in the turban. By the accepted conventions of the day, that made him more important. His pose takes up a good deal more space than that of the turbaned man, and the size of his throne expands to accommodate this. By a kind of domino effect, the people standing beside the ruler’s throne are much more tightly packed than those in the comparable space opposite. His face is fully frontal, whereas his counterpart is shown in three-quarter view. Such frontality was the norm for royal and religious images in Near Eastern art long before the coming of Islam.36 It had therefore come to connote an image of authority. The three-quarter view, by contrast, carried associations of narrative and speech. It made contact with others in the picture space, whereas the frontal view directly addressed the viewer.37 Other details of the body language reinforce this distinction. Both men are seated cross-legged à la turque, but there the resemblance ends. The man wearing a turban sits comfortably relaxed, if not slumped, his right hand raised in an age-old gesture denoting speech38 and his left arm resting

35 36

Weitzmann 1979, p. 34. Hopkins 1936; cf. the discussion, with further bibliography, in Perkins 1973, pp.

123–4. 37 38

For the theological background, see Demus 1976, pp. 7–9. King 1960, pp. 14–16.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 12

JOBNAME: ???? PAGE: 13 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 492A657E /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

491

casually on his thigh. The pose of the ruler tells a very different story. His carriage is erect, alert, commanding. He leans forward slightly, his right hand crooked at his chest in a gesture which suggests that he is holding a goblet, as was indeed standard practice in comparable Islamic images of authority.39 His left arm is clamped aggressively onto his left thigh, thereby accentuating the already marked dominance of his pose. The ruler is resplendent in scarlet and gold; indeed, armbands, lapels, frogging and piping are all in gold. Touches of gold for wrists and neck, and gold armbands, suffice for the figure opposite. He wears a powder-blue robe and thus makes a much more modest impression. His turban is of a type standard for private citizens, and indeed encountered continually in the later illustrations of this very book. The ruler, by contrast, sports a black fur hat of imposing proportions, very much in the style of a modern busby. Though not identical to the fur hats worn by the amı#rs depicted in the Agha¯nı# frontispieces and the Vienna Kita¯b al-Dirya¯q, it is recognizably of the same family. Whether this kind of hat (spectacularly unsuited for the climate of Baghdad) was a distinctively Turkish emblem of authority, or whether on the contrary it derived from cAbbasid prototypes, is unclear.40 The golden badge at its centre bears a design related to the third style of Samarran stucco. Spatial values are again pressed into service in depicting the immediate environment of the principal figures. While the man in the turban floats comfortably above the heads of his audience, with the bolster on which he sits acting as a kind of buffer between the two tiers, and thus giving the audience more room, the ruler squats directly on the heads and haloes of those beneath. Visually speaking, they carry him—the ultimate expression of their subordination. The idea itself is of very ancient lineage in the Near East, but commonly it is the litter or throne on which the ruler sits that is carried by his subjects, rather than the ruler shorn of any other attribute. Both figures have their heads outlined by a halo. Appropriately enough, the ruler has a much bigger halo, and its importance is further emphasised by the successive concentric lines which define it. Both men have a secondary ‘halo’ of cloth, a feature which is also found in the slightly earlier Agha¯nı# frontispieces. There it is of the same material as the robe worn by the amı#r, a detail which suggests that its origins may lie in Byzantine and ultimately classical iconography. Such a shawl is a standard attribute of the female 39 40

Marchal et al. 1971, cover. Ibn Jubayr 1952, p. 237; Ibn Jubayr specifies this as ‘Turkish dress’.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 13

JOBNAME: ???? PAGE: 14 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 4CDDC755 /production/pindar/booksxml/hillen/chapter−16−ends

492 personification of Night (nux) in numerous Byzantine manuscripts.41 Even in the Byzantine tradition, however, this meaningful symbol eventually lost its significance and became an empty rhetorical cliché, as can be seen in the shawl hovering redundantly over the maidservant attending the infant Virgin Mary in the apocryphal cycle of the Virgin’s life decorating the exonarthex of the Kariye Camii at Istanbul.42 How was this motif understood in the Islamic world? Personifications of Graeco-Roman type had no locus there, since the system of ideas that had produced them found no sympathetic echo in Islamic thought. That did not, however, prevent the Islamic artist from devising a new way of using the old idea. In his hands the motif lost its traditional association with a woman and became an element of royal iconography evoking the idea—ancient in the Near East as in the Graeco-Roman and Byzantine worlds—of the canopy or baldachin over a ruler’s head. Islamic court ceremonial adopted the idea so readily that the canopy (chatr) became a literary topos for the ruler himself. In the frontispiece of the Vienna Maqa¯ma¯t of 1334, for example, two flying angels hold a canopy of embroidered cloth over the head of the monarch.43 Let us turn to the architectural setting within which the two main figures are placed: an arch whose profile is shouldered, segmental and re-entrant, with a pronounced ogee character. Such flamboyantly non-structural arch profiles enjoyed a certain vogue in 13th- and 14th- century architecture;44 as might be expected, however, they reached their peak of fantasy in the flat manuscript page45 rather than in the three dimensions of architecture itself. They are an Islamic revision of a standard late antique theme. They reflect a hierarchy in that the profile of the arch framing the ruler is given four outlines instead of the two which suffice for the picture opposite, and encloses a red tricorne shape (not found in the other image) which works as a 41

As in the 10th-century Paris Psalter (BN Ms grec 139), for example (Grabar 1953, p.

169). 42

Underwood 1966, I, pp. 68–9 and II, pl. 105. Ettinghausen 1962, p. 148. 44 For examples in real architecture, see Wilber 1955, pp. 69 and 71 (Tabrı¯z); Pope and Ackerman 1938–9, pl. 400 (Vara¯mı¯n); Watson 1975, pl. VIIIa (Masjid-i ‘Alı¯, Quhru¯d) and, pl. VIIIb (Masjid-i Kalah, Quhru¯d); Melikian-Chirvani 1984, figs. 3–14, 16–17 (Takht-i Sulayma¯n); Watson 1985, colour pl. N and pls. 111, 113 and 125–6 (Ka¯sha¯n); Pope and Ackerman 1938–9, pl. 395 (the Masjid-i Ja¯mic, Bast.a¯m); and Hillenbrand 1982, pp. 252–3, figs. 95–6 (Bast.a¯m tomb tower) and p. 255, fig. 98 (the porch of that tower). 45 E.g. in the double frontispiece to the Istanbul copy of the Rasa¯’il Ikhwa¯n al-Safa¯’ (Ettinghausen 1962, pp. 98–9). 43

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 14

JOBNAME: ???? PAGE: 15 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 495FE429 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

493

symbolic reference to a crown, and thus as a kind of visual pun. The ruler is already marked out by a large halo, an aureole of cloth and a huge busby. No room remains to give him an actual crown; but the composition gives him a symbolic one. Its form has the three peaks characteristic of Saljuq crowns. The arch form, therefore, has been exploited in a double sense: as a reference to the arcuated niche traditionally associated with figures of authority, and as an evocation of an actual crown. No such visual ambiguity operates in the panel opposite. Even the inner border, with its inhabited scroll, has its part to play, subtly highlighting one page at the expense of the other. Each panel has a slightly different inner border. On the right-hand page the base of the inner border contains six animals neatly divided along the chord of the composition into two groups of three. The same three animals, in the same order and in identical poses, are set in mirror symmetry to each other on either side of the imaginary central line, a line neatly enclosed by two leaves forming an oval and reinforced by a pair of addorsed rabbits. Since the left-hand page has only five animals, a comparable symmetry there is impossible. The looser spacing places them off-axis and sets up a discontinuity vis-à-vis the central axis of the main panel. It should come as no surprise that the corresponding central position within the upper inner border should be equally loaded with meaning. Directly above the ruler is set a heraldic eagle, legs planted well apart, body ramrod-straight and wings outspread—a fully frontal image but for its head turned to the left46 and thus making contact with the opposite page. It is the image of royal power incarnate. In this upper band of the inner border there are five creatures instead of the six in the lower band, but perfect symmetry is again ensured, this time by making the significant creature frontal and central. As with the rabbits below, this centrality is stressed by the closely serried pairs of leaves flanking the centrepiece. This is the only bird in the entire border of this page, and it is placed with appropriate symbolism in the 46

One should bear in mind here the inherent difficulty of devising a convincing frontal pose for such a creature; it is noticeable that the various Islamic silks with the theme of the frontal eagle all depict the head (or heads) in profile (see von Falke 1913, I, figs. 144, 157–8, 163, 176, 178, 184–5, 200 and 202). The same is true of Byzantine silks with the eagle theme (von Falke, II, figs. 249, 251 and 253–4). The odd exception, where the frontal mode is attempted (not in a silk but in stone sculpture), is less than a complete success (von Falke, I, fig. 177), and the silk closest to this sculpture in design makes the key alteration of turning the bird’s head (von Falke, I, fig. 178).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 15

JOBNAME: ???? PAGE: 16 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 46B12F70 /production/pindar/booksxml/hillen/chapter−16−ends

494 upper band, just as the rabbits, who make their homes underground, are at the symbolically lowest part of the border. Whether some further subtlety is intended here, perhaps exploiting the spiritual significance of certain birds of prey to people of Turkic stock,47 is a matter for future research to resolve. The lordly status of this eagle is thrown into even sharper relief by its counterpart opposite. Here again the design falls short of the symmetry that the observer instinctively expects. True, there is a bird at roughly the centre point of the upper band. Yet it is distinctly—and unnecessarily—off-centre vis-à-vis the main panel immediately below it, and this awkward discrepancy extends to the placing of the leaves in the border. This lopsided layout jars with the consummate precision and balance displayed in the inner border of the opposite page. Superficially, this part of the design has enough in common with its opposite number to establish a basic visual equivalence. Again, there is a roughly central bird, it too is the only bird in the whole border, and it too is placed in the upper band.48 But one must be content with this perfunctory nod in the direction of parallelism. In spirit the left-hand border is uncertain and imitative, falling decisively short of its counterpart in its essentials. This bird could well be a dove rather than a bird of prey, and is depicted in three-quarter view rather than frontally (the ‘speaking’ rather than the ‘official’ pose), with its head turned in profile to the right and the body inclined rather than erect. These differences suggest that the birds are a deliberate mirror of the principal figures below them, and that the deference expressed in the lesser bird’s pose is a metaphor of the dominant role of the amı#r in this double frontispiece. Analysis of minutiae should not be allowed to obscure a fact of central importance: that from the reader’s viewpoint the man wearing a turban is placed on the left-hand page while the ruler occupies the right-hand page.49 There is no need to dilate here on the deep-seated dichotomy between left

47

Roux 1984, pp. 43–4. See also Gierlichs 1993, pp. 27–9 and 60–1. Grabar (1984, pp. 21–2) suggests that these contrasts indicate that fol. 2r was executed by an artist of lesser calibre; hut the need to distinguish the two folios may provide sufficient rationale for the differences. This combination of bird and rabbits in the border of a frontispiece recurs in a Fatimid figural image of the 11th to 12th century which prefaced a collection of the poems of Kuthayyir and ‘Azza; see Hoffman 1982, p. 198 and fig. 59. 49 The actual order is reversed in some reproductions. 48

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 16

JOBNAME: ???? PAGE: 17 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 49C37E06 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

495

and right in Islamic culture,50 a dichotomy inherited from the classical and Byzantine world, from which it passed into Western culture. The symbolic resonances of left and right are already exploited with assurance in Umayyad art51 and had presumably lost none of their evocative power by the 13th century. This is by no means to suggest that the left implied evil;52 for in any double frontispiece there must needs be a left-hand page, and indeed any centralised composition must have a left as well as a right side. One must also reckon with the possibility that the choice of the right-hand page for the image of the ruler could well have been reached at a subliminal level. At all events, the right had particular associations of honour, power and privilege which the left lacked. In a composition dictated by notions of mirror symmetry, any obvious and sustained attempt to glorify the right at the expense of the left would be undesirable, for the imbalance thus created would destroy the surface equality between the two images. Nevertheless, the placing of the ruler panel on the right rather than the left is of a piece with the mass of unobtrusive but cumulatively significant details which draw the eye to the right rather than the left.53

50

See the brief discussion in the Encyclopaedia of Islam (2nd ed.), s.v. yamı#n (IX, cols.

280–1). 51

For a more extended discussion of the topic, see Ettinghausen 1972, pp. 45–6. As a corrective it is worth noting that the basic meaning of the root y.s.r., whose elative gives the meaning ‘left’, is ‘ease’, with a common association of ‘prosperity’. 53 There is an apparent exception to this rule—the huge double frontispiece datable to c. 1300 and remounted in an album in the Topkapı Sarayı Müzesi, Hazine 2152, fols. 60v61r (see n. 15 above). Here the right-left dichotomy operative in the Schefer H . arı#rı# frontispiece is reversed. While the image of the ruler himself does not exclude the possibility that it was intended for a right-hand page, the direction of movement in the companion leaf implies that the ruler image was indeed on the left. Yet there is a key to this apparent discrepancy: the right-hand page contains no single image which could be construed as rivaling the enthroned ruler opposite. It is precisely the presence of a competing image in the Schefer H . arı#rı# frontispiece which makes the choice of the right-hand page for the amı#r image a significant one. Where no competition exists, the notion of dualism loses its power, and it is no longer of such critical importance that the figure of authority should occupy the right-hand page. For a colour reproduction of this key frontispiece, see Ips¸irogˇlu 1971, pls. 23–4. 52

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 17

JOBNAME: ???? PAGE: 18 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 459E4F23 /production/pindar/booksxml/hillen/chapter−16−ends

496 The meaning of the two images For all the minor discrepancies between the two panels which have been discussed so far, the first impact of the 1237 frontispiece is of two equal panels set within borders decorated in identical fashion. Yet there also develops a continual tension between the demands of symmetry and the need to exalt the official figure at the expense of his counterpart opposite. Why should the artist set himself this almost insoluble difficulty? The answer to this question should also clarify the last of the five topics outlined at the beginning of this paper: the meaning of the two images. It is important at the very outset to emphasise that the iconography of this double frontispiece is innovative. No similar combination of figures is known in Islamic, or even for that matter Byzantine, book painting, though in the latter tradition there is at least one example (the Vienna Dioscorides of 512) of two important images of seated figures following each other in close sequence. Thus it might be argued that the novelty lies in the combination—that is, in the close juxtaposition of the two images—rather than in their component parts. At the simplest level, the three-tier composition comprising, in descending order, angels or victories, a lone figure of authority, and a group of lesser figures, was an iconographic topos in Byzantine as in early Islamic art. The ruler page of the 1237 frontispiece is entirely explicable as a variation on this familiar theme, with its main interest lying in the idea of incorporating an audience scene into this iconography of power. Nevertheless, the whole point of such images is that they should enjoy unchallenged predominance in the page opening allotted to them. That is an integral part of their message. If there is a picture on the opposite page, as is sometimes the case in Byzantine frontispieces, some of which indeed have a whole sequence of such portraits;54 the picture opposite will conduce, usually in a very direct way, to the impact of the main image.55 Hence Byzantine double frontispieces contain a plethora of worldly and spiritual references to the emperor’s power. It is not a question of competition but of complementarity between one page and another. Yet there can be little doubt 54 As in the 11th-century Homilies of St. John Chrysostom in Paris (BN Cod. Coislin 79), which has four portraits of the Emperor (Omont 1929) or the 14th-century Typikon in the library of Lincoln College, Oxford, Gr. Ms. 35, which has no fewer than ten portraits of the imperial family (Rice 1968, p. 357). 55 The solution adopted in the 1199 Paris Kita¯b al-Dirya¯q is at once simple and direct: repetition.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 18

JOBNAME: ???? PAGE: 19 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 49F2A0F4 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

497

that the idea of competition is built into the Schefer H.arı#rı# frontispiece. The two images vie for attention and their similarity makes it impossible to interpret one as the handmaiden, so to speak, of the other. This very similarity proclaims that each has its own separate existence, not that they constitute obverse and reverse of a single coin. It might be worth noting, parenthetically, that it is precisely in the field of Muslim numismatics that an analogy to the double royal frontispiece may be found. Such few coins as do use figural images invariably have a bust on the obverse and some less concentrated royal image on the reverse (fig. B). This practice may be understood as the numismatic equivalent of the symbolism inherent in the contrast between right and left. The preceding discussion has established beyond question that models for the ruler page abounded, and indeed that it was common practice, both in Byzantium and in the Islamic world, to reserve the frontispiece for an image of the ruler. The surviving volumes of the Kita¯b al-Agha¯nı# show how varied the iconography of kingship could be in such contexts. Thus the repertoire of royal iconography contained enough ideas at this time to have accommodated without strain a demand for a double royal frontispiece. Yet that is not what al-Wa¯sit.#ı provides here. The similarity between the two pages of the 1237 frontispiece excludes the possibility—as indeed does the difference in facial type—that the artist is representing the same person twice. Who, then, can the person on the left-hand page be? It will be as well to dispose of the lesser possibilities first. The suggestion that the figure hitherto identified in this paper as the ruler is in fact the wife of the man with a turban, and that the latter is the ruler himself, has been put forward56 but is not worth serious consideration. The right-hand figure is, after all, the possessor of two drooping moustaches, while the left-hand figure is heavily bearded. The numerous devices employed to exalt the right-hand figure would also accord ill with an identification of this figure as the ruler’s wife. A suggestion that these are Baghdadi angels,57 whatever that phrase may mean, is also not worth entertaining. But it is not impossible that the figures refer generically to the text of the Maqa¯ma¯t.58 It would be strange to single out a particular maqa¯ma above the others, and stranger still not to

56

Hamid 1966, I, p. 251. Farès 1953, pp. 16–17 and 28. 58 Grabar 1984, p. 23, following Ettinghausen 1962, pp. 111 and 114, suggests that they may refer to the qa¯.d¯ı and the amı#r, stock protagonists of many maqa¯ma¯t; he further 57

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 19

JOBNAME: ???? PAGE: 20 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 475FF229 /production/pindar/booksxml/hillen/chapter−16−ends

498 identify it. If a scene from a specific maqa¯ma is to be ruled out, perhaps the figures represent the two chief characters of the Maqa¯ma¯t, namely Abu¯ Zayd and al-H.a¯rith. This idea is extremely unlikely, not only because the right-hand panel is brim-full of images with royal connotations but also because Abu¯ Zayd is consistently described as an old man with a white beard, and therefore corresponds to neither of the two main figures in the frontispiece. Even less persuasive is the possibility that the left-hand page has some symbolic significance; not only are such symbolic images rare in Islamic frontispieces, but they bear attributes which invite interpretation. No such attributes can be detected on the left-hand page. The clues to the meaning of this page are to be found both in the composition itself and in the models from which it derives. Only one tradition rivalled that of the ruler image as a subject suitable for frontispiece decoration, and this tradition was familiar in classical, Byzantine, early Western medieval and Islamic art alike: the author portrait. Admittedly, the conventions which had been developed for this subject in the first millennium CE were somewhat different from those that operate in the H.arı#rı# frontispiece. Very often the author is entirely alone on the page and stands out against a uniform gold background. It is customary for him to hold his book or to be actually engaged in writing it. Standing images outnumber seated ones. Occasionally the author will be shown in the act of dictating to a scribe. Other secondary figures found from time to time include the author’s inspiration, who takes human form as a female figure bending over to speak into his ear, and other personifications. Several of these iconographic types had made their way into Islamic art by the early 13th century, among them the solitary seated author and the author accompanied by his disciples. How, then, is the left-hand page of the H.arı#rı# frontispiece to be fitted into this group of images? If it is indeed an author portrait, why is there no sign of a book? And what is the point of all the people below him? Why, finally, does the artist risk lèse-majesté by placing him so nearly on a par with the ruler opposite? The answers to these questions will shed new light on the originality and resourcefulness of the artist responsible for this manuscript, al-Wa¯sit.#ı, and will highlight his qualities as a painter well ahead of his time. suggests that the figure on fol.lv could be al-H . arı#rı#’s forerunner in the maqa¯ma¯t genre, al-Hamadha¯nı#. But the princely accoutrements and attributes of this figure argue against such an interpretation.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 20

JOBNAME: ???? PAGE: 21 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 47C6BEF0 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

499

It was nothing short of a brainwave to make double capital out of an idea which was itself new: to show the main figure exalted above a mass of lesser figures, but nevertheless occupying the same space. Earlier imperial iconography had favoured the placing of secondary figures on either side of the principal one. On the rarer occasions when they were placed below, a system of self-contained compartments prevented them from intruding into the world of the main figure. In the 1237 frontispiece the two groups, though so similar as to be virtually identical, have distinctively different functions. As argued above, it seems plausible that the ruler panel depicts a scene of royal audience – perhaps the dı#wa¯n al-‘a¯mm. Sufficient literary descriptions of such ceremonies exist to establish that serried ranks of onlookers or participants were a normal feature of these occasions.59 Within the limited space available, al-Wa¯sit.#ı has done his best to evoke such a scene. What made excellent sense in one panel was, however, a possible embarrassment in the other. It certainly required an adaptation of existing iconography. The smoothness with which this potentially awkward transition has been accomplished excites admiration. In the right-hand panel the lower figures make most obvious sense as participants in a royal audience; in the left-hand panel they become an audience of a different and more usual kind, for they seem simply to be listening to what the main figure is saying. This is where the three-quarter view adopted for that figure comes into its own. It is proper in a formal royal audience for the ruler to be both physically and metaphysically above his people, and the frontal view employed for the ruler in the right-hand panel is an apt expression of that remoteness. Equally appropriate to its context is the three-quarter view used for the main figure in the left-hand panel, for it expresses in an effortlessly natural way the interplay between the narrator and his audience. This emphasis on narrative explains the unprecedented gathering of listeners, a feature that has no place in the traditional iconography of the author portrait, and the speaking gesture of the raised right hand made by the man with the turban which, incidentally, absolves him from the need to hold a book. Moreover, the practice of public recitation of famous works of literature is well attested in the Islamic world, and the evidence suggests that at times such performances took place in a court setting.60 For those whose competence in Arabic as a literary language was not enough to appreciate the finer points of this notoriously difficult 59 60

Kennedy 2004, pp. 139–41. It is the context of a good deal of the Kı#ta¯b al-Agha¯nı#.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 21

JOBNAME: ???? PAGE: 22 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 4DA47373 /production/pindar/booksxml/hillen/chapter−16−ends

500 text—and it is readily conceivable that Turkish amı#rs might belong in this category—the readiest access to the stories of the Maqa¯ma¯t might well be to see an illustrated version rather than to read it. The pose of the man with a turban, incidentally, can be read to imply deference and to suggest that the book was intended for the ruler. Nor should one lose sight of the possibility that some appropriate pun is intended on the very title of the work itself, for the word maqa¯ma¯t means assemblies, and that is exactly what we see depicted here. Indeed, the same kind of interaction between storyteller and audience is replicated in several of the paintings later on in the volume.61 One detail in particular speaks for the hypothesis that the main left-hand figure is al-H.arı#rı# himself: he was no longer alive. Thus he does not present the same kind of challenge to the amı#r that a living figure would. Moreover, the author portrait was the obvious choice for a frontispiece theme if a ruler figure was not being depicted. Some author portraits, both Byzantine and Islamic, also feature the distinctive detail of a hand raised in a gesture of speech,62 though admittedly there are not crowds of people in attendance. The winged beings in the spandrels above the figure make good sense if he is the author, for their connotations of glory and exaltation would be entirely appropriate for this most popular writer, instead of a distracting irrelevance. In this detail, as in the case of the audience itself, one can observe the artist’s use of an identical theme in two quite different ways. Thus the potentially conflicting needs of visual symmetry and of iconographic function are met and resolved without the slightest jar. Here, then, perhaps uniquely among Islamic manuscripts, is a case where the author is placed on a well-nigh equal footing with the representative of secular authority. It is a momentous statement. What possessed Yah.ya¯ al-Wa¯sit.#ı to make it? Attention has already been drawn to the fact that he was both scribe and illustrator, as he himself attests. This would have predisposed him to consult his own preferences for the layout of the

61 Using Grabar 1984, see—for images of a speaker raised above his audience – maqa¯ma¯t 17 (fol. 46v, microfiche 3/F8), 21 (fol. 58v, microfiche 4/El), 28 (fol. 84v, 5/F1), 33 (fol. 103r, 6/G9), 37 (fol. 114v, 7/E4), 40 (fol. 125r, 8/A7) and 47 (fol. 156r, 9/C9); cf. also 42, fols. 131v and 133v (8/C10–11) for images of a seated speaker addressing a seated audience. The connection between this frontispiece and the text of the manuscript is pointed out by both Grabar 1984, pp. 22–3 and O’Kane 2003, p. 49. 62 For Byzantine examples, see Weitzmann 1977, pls. 16 and 35 and Hoffman 1993, fig. 10; for Islamic ones, see three manuscripts in Istanbul: the 1229 Dioscorides, the Mukhta¯r al-H . ikam and the Rasa¯’il Ikhwa¯n al-S . afa’ (Ettinghausen 1962, pp. 68–9, 75 and 98–9).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 22

JOBNAME: ???? PAGE: 23 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 42FEB1ED /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

501

opening pages, as for the rest of the manuscript. He was thus in an unusually privileged position and could profit from this independence to devise a new kind of formal visual introduction to the text. Once this freedom from constraint is recognised, the full significance of al-Wa¯sit.#ı’s solo performance clicks into place. To transcribe the entire text, to insert explanatory glosses, to provide it with a remarkably full and varied complement of illustrations, and to carry out both of these tasks without collaborators, bespeaks an unusual commitment to the Maqa¯ma¯t as a text and as a quarry for visual material. Al-Wa¯sit.#ı would have had to commit many months of his life to this project. It should therefore not stretch credulity to suggest that he might well have cherished a particular affection for this work,63 and that al-H.arı#rı# might indeed have been one of his favourite authors. What more natural token of his regard than to represent that author in the frontispiece? Presumably the realities of daily life, and the constant need to find a patron to finance the expensive undertaking represented by such a book, dictated the presence of the ruler image in the traditional place of honour. Yet al-Wa¯sit.#ı set up the author portrait as near as he dared to that of the still undetermined patron, and by the various shifts detailed above made it clear to the discerning viewer that the ruler image took precedence. The colophon makes it clear that no high-ranking patron commissioned the manuscript. Al-Wa¯sit.#ı therefore had no specific patron to flatter, though of course he had to bear a prospective patron in mind. Presumably he produced the manuscript speculatively, for the market; nevertheless, he took out sufficient insurance against any future charge of lèse-majesté. The ruler is conceived as a generic image that implies respect for the office rather than for a specific individual. There was never any question that the dead author should be allowed to triumph over the living embodiment of secular authority; but their relative equality in visual terms underlines the interdependence of the men of the pen and the men of the sword. Bibliography Baer, Eva. Metalwork in Medieval Islamic Art, Albany: State University of New York Press, 1983.

63 It is relevant to point out here how much care al-Wa¯sit.¯ı has taken, in his role of scribe, to gloss certain passages of the text (in red ink).

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 23

JOBNAME: ???? PAGE: 24 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 3E488D70 /production/pindar/booksxml/hillen/chapter−16−ends

502 Beckwith, John. Early Christian and Byzantine Art, Harmondsworth: Penguin, 1970. Bernus-Taylor, Marthe & Bittar, T. & Charritat, M. & Dubois, S. & Gayraud, G. and Joël, M. Le Prince en Terre d’Islam, Paris: Louvre, 1983. Brandenburg, Dietrich. Islamic Miniature Painting in Medical Manuscripts, Basle: Roche, 1982. Buchthal, Hugo. Miniature Painting in the Latin Kingdom of Jerusalem, Oxford: Clarendon Press, 1957. Buchthal, Hugo, Kurz, Otto and Ettinghausen, Richard. “Supplementary Notes to K. Holter’s Check List of Islamic Iluminated Manuscripts”. In Ars Islamica, VII, pp. 147–64, Ann Arbor: University of Michigan Press, 1940. Capps, Walter H. “The Style of the Consular Diptychs”. In Art Bulletin, 10, pp. 61–101, New York: College Art Association of America, 1927. Delbrueck, Richard. Die Consulardiptychen und verwandte Denkmäler, Berlin: Walter de Gruyter, 1929. Demus, Otto. Byzantine Mosaic Decoration. Aspects of Monumental Art in Byzantium, London: Routledge & Kegan Paul, repr. 1976. Esin, Emel. “Two Miniatures from the Collections of Topkapı”. In Ars Orientalis, V, pp. 143–61, Washington: Smithsonian Institution, 1963. Ettinghausen, Richard. Arab Painting, Geneva: Skira, 1962. ———. From Byzantium to Sasanian Iran and the Islamic World, Leiden: Brill, 1972. Falke, Otto von. Kunstgeschichte der Seidenweberei, Berlin: Wasmuth, 1913. Farès, Bichr. Le Livre de la Thériaque. Manuscrit arabe à peintures de la fin du XIIe Siècle conservé à la Bibliothèque Nationale de Paris, Cairo: l’Institut français d’archéologie orientale, 1953. Fehérvári, Géza. “Tombstone or Mihrab? A Speculation”. In Islamic Art in the Metropolitan Museum of Art, ed. Richard Ettinghausen, pp. 241–254, New York: Metropolitan Museum of Art, 1972. Fitzherbert, Teresa. “‘Balcami’s Tabari’, An illustrated manuscript of Balcami’s Tarjama-yi Ta¯rı¯kh-i T.abarı¯ in the Freer Gallery of Art, Washington (F59.16, 47.19 and 30.21)”, unpublished Ph.D. thesis, University of Edinburgh, 2001. Folda, Jaroslav. Crusader Manuscript Illumination at Saint-Jean d’Acre: 1275–1291, Princeton: Princeton University Press, 1976. ———. The Art of the Crusaders in the Holy Land, 1098–1187, New York: Cambridge University Press, 1995.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 24

JOBNAME: ???? PAGE: 25 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 42D3E291 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

503

———. Crusader Art in the Holy Land, From the Third Crusade to the Fall of Acre, New York: Cambridge University Press, 2005. Gierlichs, Joachim. Drache, Phönix, Doppeladler. Fabelwesen in der islamischen Kunst, Berlin: Gebr. Mann Verlag, 1993. Grabar, A. N. Byzantine Painting, trans. S. Gilbert, New York: Skira, 1953. Grabar, André. L’Art de la Fin de l’Antiquité et du Moyen Age, Paris: Collège de France, 1968. Grabar, Oleg. The Illustrations of the Maqamat, Chicago and London: University of Chicago Press, 1984. Grube, Ernst J. “Materialien zum Dioskurides Arabicus”. In Aus der Welt der islamischen Kunst: Festschrift für Ernst Kühnel zum 75. Geburtstag am 26.10.1957, ed. Richard Ettinghausen, pp. 163–194, Berlin: Gebr. Mann Verlag, 1959. ‘Isa¯ Salma¯n H.amid, “Mesopotamian School and the Place of Painting in Islam”, unpublished Ph.D. thesis, University of Edinburgh, 1966. Herrmann, Georgina. The Iranian Revival, Oxford: Elsevier-Phaidon, 1977. Herzfeld, Ernst. “Die Tabula ansata in der islamischen Epigraphik und Ornamentik”. In Der Islam, VI, pp. 189–200, Berlin: De Gruyter, 1916. Hillenbrand, Robert. “The Flanged Tomb Tower at Bast. a¯m”. In Art et Société dans le Monde Iranien ed. Chahriyar Adle, pp. 237–261, Paris: Editions Recherche sur les civilizations, 1982. Hillenbrand, Carole. The Crusades. Islamic Perspectives, Edinburgh: Edinburgh University Press, 1999. Hoffman, Eva Rose F. “The Emergence of Illustration in Arabic Manuscripts: Classical Legacy and Islamic Transformation”, unpublished Ph.D. thesis, Harvard University, 1982. Hoffman, Eva Rose F. “The Author Portrait in Thirteenth-Century Arabic Manuscripts: A New Islamic Context for a Late-Antique Tradition”. In Muqarnas, 10, pp. 6–21, Leiden: Brill, 1993. Holter, Kurt. “Die islamischen Miniaturhandschriften vor 1350”. In Zentralblatt für Bibliothekswesen, 54, 1–2, pp. 1–34, 1937. Hopkins, Clark. “A Note on Frontality in Near Eastern Art”. In Ars Islamica, III, 2. pp. 187–96, Ann Arbor: University of Michigan Press, 1936. The Travels of Ibn Jubayr, trans. and ed. Broadhurst, Ronald J. C., London: Jonathan Cape, 1952. Ips¸irogˇlu, Mazhar S¸evket. Das Bild im Islam. Ein Verbot und seine Folgen, Vienna: Schroll, 1971.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 25

JOBNAME: ???? PAGE: 26 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 44C53CFC /production/pindar/booksxml/hillen/chapter−16−ends

504 James, David Lewis. The Master Scribes. Qur’ans of the 10th to 14th centuries AD, Oxford: Nour Foundation in association with Azimuth Editions and Oxford University Press, 1992. Johnson, Elizabeth J. “A Study of the Miniatures of the Illustrated Kitab al-Aghani: Their Relation to Contemporary Seljuk Painting and to Central Asian Traditions”, unpublished Ph.D. dissertation, University of California, Los Angeles, 1975. Kennedy, Hugh. The Court of the Caliphs: The Rise and Fall of Islam’s Greatest Dynasty, London: Weidenfeld & Nicolson, 2004. King, Noel Q. There’s such Divinity doth hedge a King. Studies in ruler cult and the religion of sacral monarchy in some late fourth century Byzantine monuments, Edinburgh: Thomas Nelson and Sons Ltd for the University College of Ghana, 1960. Maqa¯ma¯t Al-H.arı#rı#. Illustrated by Y. Al-Wa¯sit.#ı (facsimile edition), London: Touch Art, 2004. Marchal, Henri et al. Arts de l’Islam des origines à 1700 dans les collections publiques françaises, Paris: Réunion des musées nationaux, 1971. Melikian-Chirvani, Assadullah Souren. “Le Sha¯h-na¯me, la gnose soufie et le pouvoir mongol”. In Journal Asiatique, CCLXXII, 3–4, pp. 249–339, Paris: Société Asiatique, 1984. Miles, George Carpenter. “Mih.ra¯b and ‘Anazah: A Study in early Islamic Iconography”. In Archaeologica Orientalia in Memoriam Ernst Herzfeld, ed. George C. Miles, pp. 156–172, Locust Valley, NY: J. J. Augustin, 1952. O’Kane, Bernard. Early Persian Painting. Kalila wa Dimna Manuscripts of the Late Fourteenth Century, London: I. B. Tauris, 2003. Omont, Henri Auguste. Miniatures des plus anciens manuscrits grecs de la Bibliothèque Nationale du VIe au XIVe siècle, Paris: H. Champion, 1929. Parham, Cyrus, ed. The Splendour of Iran. Volume III. The Islamic Period, London: Booth-Clibborn Editions, 2001. Perkins, Ann. The Art of Dura-Europos, Oxford: Clarendon Press, 1973. Pétrosyan, Yuri A. De Bagdad à Ispahan. Manuscrits islamiques à la Filiale de Saint-Pétersbourg de l’Institut d’Etudes orientales, Académie des Sciences de Russie, Catalogue of an international exhibition held 1994–95, Paris: Foundation ARCH, Paris-Musées, Electa, 1994. Pope, Arthur Upham and Ackerman, Phyllis, eds. A Survey of Persian Art from Prehistoric Times to the Present, London and New York: Oxford University Press, 1938–9.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 26

JOBNAME: ???? PAGE: 27 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 30CF9AF3 /production/pindar/booksxml/hillen/chapter−16−ends

THE SCHEFER HARIRI: A STUDY IN ISLAMIC FRONTISPIECE DESIGN

505

Rice, David Storm. “The Aghani Miniatures and Religious Painting in Islam”. In The Burlington Magazine, 95, pp. 128–34, London: Burlington Magazine Publications, 1953. ———. The Unique Ibn al-Bawwa¯b Manuscript in the Chester Beatty Library, Dublin: Emery Walker, 1955. Rice, David Talbot. Byzantine Art, revised ed., Harmondsworth: Penguin, 1968. Roux, Jean-Paul. La religion des Turcs et des Mongols, Paris: Payot, 1984. Schulz, Philipp Walter. Die persisch-islamische Miniaturmalerei. Ein Beitrag zur Kunstgeschichte Irans, Leipzig: K. W. Hiersemann, 1914. Sotheby’s Oriental Manuscripts and Miniatures, London: Sotheby’s, 29 April 1988. Underwood, Paul Atkins. The Kariye Djami, New York: Bollingen Foundation, 1966. Ward, Rachel. Islamic Metalwork, London: British Museum, 1993. Watson, Oliver. “The Masjid-i ‘Alı¯, Quhru¯d: an architectural and epigraphic survey”. In Iran, XIII, pp. 59–74, 1975. ———. Persian Lustre Ware, London: Faber, 1985. Weitzmann, Kurt. Late Antique and Early Christian Book Illumination, London: Chatto and Windus, 1977. ———. ed. The Age of Spirituality. Late Antique and Early Christian Art, Third to Seventh Century, catalogue of the exhibition at the Metropolitan Museum of Art, November 19, 1977, through February 12, 1978, New York: the Metropolitan Museum of Art in association with Princeton University Press, 1979. Wilber, Donald Newton. The Architecture of Islamic Iran. The Il-Kha¯nid Period, Princeton: Princeton University Press, 1955.

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 27

JOBNAME: ???? PAGE: 30 SESS: 42 OUTPUT: Tue Aug 10 10:30:38 2010 SUM: 009B1A48 /production/pindar/booksxml/hillen/chapter−16−ends

Kerrypress Ltd – 01582 451331 – Typeset in XML Division: chapter-16-ends Sequential 30

Bibliography Update Several publications described as being “in press” have since appeared. They are: p. 13, n. 58 The proceedings of the symposium Iconography in Islamic Art have appeared under the title Image and Meaning in Islamic Art, ed. R. Hillenbrand, London 2005 p. 56, n. 22 E. G. Sims, “The Hundred and One Paintings of Ibrahim Sultan”, in R. Hillenbrand (ed.), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London and New York, 2000), 101–27 p. 83, n. 10 M. S. Simpson, “A Reconstruction and Preliminary Account of the 1341 Shahnama, with Some Further Thoughts on Early Shahnama Illustration”, in R. Hillenbrand (ed.), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London and New York, 2000), 215–47 p. 147, n. 84 M. Barrucand, “Kopie – Nachempfindung oder Umgestaltung am Beispiel arabischer mittelalterlicher Bilderhandschriften und ihrer osmanischen Kopien”, in B. Finster, C. Fragner and H. Hafenrichter (eds.), Bamberger Symposium. Rezeption in der Islamischen Kunst. Vom 26. 6.–28. 6. 1992 (Beiruter Texte und Studien, Band 61) (Beirut 1999), 19–41 and Taf. II/3–IX/2 p. 152, n. 9 B. O’Kane, Early Persian Painting. Kalila and Dimna Manuscripts of the Late Fourteenth Century (London and New York, 2003)

508

p. 301, n. 23 and p. 346, n. 60 B. Brend, “Beyond the Pale: Meaning in the Margin”, in R. Hillenbrand (ed.), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London and New York, 2000), 39–55 p. 327, n. 1 A. S. Melikian Chirvani, “The Anthology of a Sufi Prince from Bokhara”, in R. Hillenbrand (ed. ), Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson (London and New York, 2000), 151–85 p. 332, n. 20 A. Adamova, “The Repetition of Compositions and the Problem of the Identification of Artists in Persian Painting”, in V. S. Curtis, R. Hillenbrand and J. M. Rogers (eds.), The Art and Archaeology of Ancient Persia. New Light on the Parthian and Sasanian Empires (London and New York, 1998), 175–81

Index Note: The presence or absence of transliteration in this index reflects the usage in each particular article. Abāqā Khān 90, 90 n22, 91, 92, 112, 380 Abarqûh 461 Abbasid(s) 118, 153, 160, 391, 458, 491 architecture 298, 394 n76 bookbindings 463 ceramics 325 n29 Kufic 295 script 293 texts 379 Qurʾans 295, 297 ʿAbd al-Hadi ibn Muhammad ibn Mahmud ibn Ibrahim alMaraghi 178 Fig. 10 ʿAbd al-Jabbār b.ʿAli 382 n26 ʿAbd al-Qayyum ibn Muhammad ibn Karamshah-i Tabrizi 294 ʿAbd Allah Sayrafi 168 ʿAbd al-Masih 126 ʿAbd al-Muʾmin al-ʿAlawi al-Kashi 175 Figs. 5, 6 ʿAbd al-Muttalib 145 Fig. 18 ʿAbd al-Razzāq al-Samarqandī 93 n33 Abel 236 n98 Abel, Armand, 79 n1, 80, n3, 106, 235

abjad 297 n18 Abraham 14, 15, 38 Fig. 12a, 39 Fig.12, 80, 155 abri 17 n74, 304, 304 n26 Abu Bakr 19 n84, 136 Fig. 9 Abu Haritha 126 Abuʾ l-Faḍl Muhammad Balʿami 98, 105 n48, 151 Abuʾ l-Qasim Kashani 150 Abū Saʿīd 82 n8, 92, 92 n31, n32, 93 n33, 95, 105, 162, 167, 170, 439, 450 Abū Saʿīdīya/Bu Saʿīdīya 93 n33 Abusaʿīd Khudaybanda 149 Abūsaʿidnāma 149, 167 Abū Šākir 79 n2 Abū Zayd 3 n8, 363, 388 n51 Achaemenids 79, 208, 428, 430 tombs 433 Ackerman, Phyllis 4 n18, 13 n56, 222 n45, 232 n84, 287 n134, 315, 323, 394 n71, n74, 403 n79, 492 n44, 504 adab 360, 361 Ǻdahl, Karin 4 n16 Adam 235, 235 n95, 236, 273 Fig. 31 Adamova, Adel 10, 330 n11, 332

510 n20, 372, 423 n8, 507 Adle, Chahriyar 51 n3, 52, 52 n6, 53 n12, 57 n25, 77 n77, 218 n25 Adoration of the Magi 169 ʿAḍud al-Dawla 384 aesthetic(s) 22, 298, 299, 367, 457 Afghanistan 68 Afrāsiyāb 66, 69, 71 Fig. 2, 89 n21, 90, 93 n33, 187 Fig. 22, 374 Aghānī 486 n24 Aga-Oğlu, M. 394 n71, 414 ahl al-bait 123, 146 Ahmad ibn al-Husayn ibn al-Ahnaf 361 Ahmad ibn Muhammad ibn Idrīs 360 Ahmad Jalaʾir, Sultan 91 Divan13, 37 Fig. 11, 344 n57 Ahmad Musa 149, 162, 167 Ahmad Qaniʿi 432 Aḥmadī 81, 81 n7, 91 n28 Ahmed Qarahisari 305 ʿAʾisha 123 n53 Akbar 105 Akbarnāma 105 n48 Akimushkin, O. F. 483 n23 Akvan 67 Alamfiya 98 Alanquwa 98 albarello 317 Fig. 5 Albertus Magnus 238 albums, 57, 305,464 (H.2152) 237 n101, 481 n15, 494 n53 see also: Bahram Mirza; Diez Albums Alexander the Great 79, 80, 80 n3, 81, 82, 89 n21, 91, 91 n25, n26, 92, 93 n33, 94, 94 n37, 95, 95 n38, 96, 98, 105, 105 n50, 106, 169, 204, Fig. 40, 421, 435, 436, 450

see also: Iskandar Alexander Romance79, 79 n1, 81, 82, 105, 436, 437,438, 449 Ethiopic version 79 n1 Alexander’s Wall 437 Alf layla wa-layla 360 Almohad dynasty 459 al-Hasan Qavam al-Daula wa ʾl-Din 187 Fig. 22 ʿAli 20 n87, 54 n16, 77, 121–129, 125 n60, 147–148 ʿAli b. Abi Talib see: ʿAli ʿAlishah mosque 150 ʿAli Mirza of Gilan 56 ʿAli Sayrafi 302 ʿAli Shah 393 ʿAli Shir Navaʾi Divan59 n40 Gharāʾib al-sighār 60 n40 Khamsa 60 n40 ʿAli, Sultan 329 n8 ʿAli Yazdī 98 Allan, J. W. 150 n4, 307 n1, 363 Allen, T. 160 n32, 390 n59, 414 Alqas Mirza 69, 69 n72 Amajur 296 Amasya, Treaty of 78 Ambrose, Hexameron 240 n114 Ambrosian Iliad 420 Amida/Diyarbakir 421 n2 amīr 68, 114, 383 Anatolia 81, 373, 430 n36, 432 Anatolian Qurʾans 300 Andalusi script 301 Andalusia 94, 97, 437 Andalusian culture 361 Andrae, T. 117 n28, 125 n61 Anet, C. 235 n92 Angelis, M. A.de 5 n22 angels 97, 405, 488, 492, 496, 497 Annunciation 121 n45, 156 ʿanqa 238, 279 Fig. 37

INDEX

anthologies 9 n42, 14–15, 19 n83, 25 n107, 38 Fig.12a, 39 Fig. 12b 57, 152, 214 n12, 361, 367 Anthology of Baysunghur 6 n27, 9 n41, 23 n99, 312, 321 Fig. 10, 341 n44 Anthology of Sultan Iskandar 7 n34, 9 n43, 20 n87, 330 n12, 336 n33, 342 n52 Anthology of Divans 175 Figs. 5, 6 Aqa Mirak 21 n92 ʿaqib 125 n61, 126 Aqsarāʿī, al Nihāyat al-suʾl 211 n1.2, 361 Apocrypha 208 Aq Qoyunlu 63 Arab bestiaries 236, 237 calligraphers 299 culture 159, 236, 383, 420, 425,429, 430, 432 painters 378, 379, 386, 389, 390, 391, 405, 410, 411, 412, 413, 414 painting 2, 10 n44, 12, 152, 360, 379, 381, 382, 384, 387, 390, 391, 392, 404, 405, 410, 480 Arabia 96, 122 Arabian prophets 159, 432 Arabians 148 Arabic language 79, 153, 158, 165, 207, 238, 302, 310, 332, 420, 426, 427, 427–428 n25, 429, 430, 432, 439, 481, 499 Arabic literature 290, 359–364 passim Arabic script 14, 92, 290, 295, 299, 310, 413 Arabs 158, 165, 207, 365, 426, 427 conquest of Iran 426, 427, 428, 431 Arberry, Arthur J. 24 n103, 293 n10,

511 327 n3, 328 n6, 431 n40, n41, n44 archer 374 architecture 2, 3, 4, 5, 7, 8, 9, 21, 26, 27, 149,158, 169, 227 n65, 232, 298, 305, 307, 366, 375, 380, 388, 389–394, 394 n76, 404, 405, 412, 492 Abbasid 298, 394 n76 Ardabil 403 Armenian 390 Coptic 390 Ilkhanid 9, 158, 169, 393, 394, 396, 403 in author portraits 389–394, 405 Iraqi 392, 394 n76 architectural framework 4, 389, 393 Ardabil 403 Ardashir 22 n95, 93 n35, 174 Fig. 4, 190 Fig 25 Ardavan 93 n35, 174 Fig. 4, 190 Fig. 25 Arghun, Amir 380 n21 Aristotle 89 n20, 214, 383 n28 Arjasp 6 n28, 7, 8, 20, 29 Fig. 3, 31 Fig. 5 Ark of the Covenant 156 Armenian architecture 390 manuscript painting 388 n50 Arnold, Sir Thomas W. 57 n31, 111 n4, 114 n17, 122 n48, 123 n53, 147 n84, 235 n96, 242 n17 Arpa 82 n8 Arrian 105 n50 Arzadasht 63 n57, 328 n7, 329 n9 ʿasabiyya 127 Ashtor, E. 220 n37 Aslanapa, O. 288 n135 assemblies 361, 362 astrology 98, 450 astronomy 112, 360

512 Atabeg Ildegiz 91 n25 Atasoy, N. 289 n143, 330 n14 ateliers see: workshops ʿAṭā-Malik Ǧuwaynī 91 n28, 92 n30 sāhib dīwān 92 Tarikh-i jahan-gusha 93, 202 Fig. 37, 151 Athens, Parthenon 22 Athir, Diyaʾal-Din Ibn al- 427 n25 Atıl, Esin 10 n44, 12 n55, 14 n59,42, 50, 83 n10, 211 n1.5, 220 n40, 222 n44, 223 n48, 224 n55, 241 n123, 308 n3, 319, 324, 363, 433 n52 ʿAttār, Mantiq al-Tayr 45 Fig. 17, 461 audience scenes 25, 86, 121,333, 340, 383, 387, 405, 487, 491, 496, 499, 500 Augustus 105 n50 Auld, Sylvia 373, 425 n18 author portraits see: portraits Automata of al-Jazarī 211 n1.1, 218 n29, 361 Awad, G. 120 n40 ʿAwfī, al- 378 Ayyubid Qurʾan 301 axonometric projection 8 Azada 373 ʿAzza 494 n48 Babel 120 Babinger, Franz 105 n50, 106, 106 n51 Babylon 119, 120, 437 Babylonian exile 155 Badiʿ script 292 n2 Badiee, J. A.O. 223 n54 Baer, Eva 482 n18, 483 n21, 501 Baghdad 120, 164, 164 n39, 183 Fig.17, 184 Fig. 18, 202 Fig. 37, 222, 232 n84, 377, 379 n9, 391,

394 n75, 395 Fig. 1, 396 Fig. 2, 397 Fig. 3, 398 Fig. 4, 399 Fig. 5, 412, 428, 487, 491, 497 Baghdad school 222, 227, 227 n67, 236 Baghdadi paper 150 Bagherzadeh, F. 314 n28 bahadur 92 Bahira 121, 156 n21, 138 Fig. 11 Bahman 72 Fig. 3, 190 Fig. 25, 375 Bahrain, Bait al-Qurʾan 298, 302 Bahram Gur 4 n19, 9 n42,10 n45, 16, 17, 18 n78, 43 Fig. 15, 84, 188 Fig. 23, 189 Fig. 24, 193 Fig. 28, 331 n14, 336, 373, 374, 448 Bahram Mirza Album 53, 149, 149 n1, 328– 329 n7 Bahrami, M. 311 n14, 312 n15, n16, n17, n19, n20, 313 n24, n25, 319, 320, 322 Baibars, Sultan 106 n52, 225 n59 baits 11 n49, 58 n33, 345, 346, 370 Bakhshi, Kamalashri 236 n98 Balay, C. 434 n59, 435 n65 balconies 5, 10, 25,168, 335, 338, 394, 408 Baltimore. Walters Art Museum Shahnama (W677b)187 Fig. 22 Timurid painting (no.10.678) Banani, A. 431 n42 Bandar-i Tahiri 374, 375 banqueting scene 313 Baptism 121 n45 Barrett, D. 218 n30, 230 n75, 236 n99 Barrucand, M. 147 n84, 392 n64, 414, 506 Barsbay, al-Malik al-Ašraf 106 n52 Bartold, V. V. 431 n1 Barzunama 434 n60

INDEX

Bashīriyya Madrasa 394 n76 Baṣra 377 Baṣtām 91, 394, 403 Masjid-Jāmiʿ 492 n44 tomb tower 492 n44 battle scenes 12, 49, 64, 66, 67, 69, 81, 85, 86, 88, 89, 94, 159, 161, 163, 168, 171 Fig. 1,190 Fig. 25, 194 Fig. 29, 180 Fig. 13, 208–209, 218, 218 n30, 314, 426, 427, 431, 438 Bausani, Alessandro 96 n40, n41, 97, 106, 114 n19, 151 n8, 155 n19 Bawwab, Ibn al- 296, 379 n9, 366 n6 Bāyazīd 105 Bayhaqī, Ẓāhir al-Dīn Abuʾl-Qāsim al- 377 Baylī 81 Bāyqāra, Sultan Ḥusayn 380 n18 Baysunghur b. Shah Rukh 2 n2, 56, 58, 61, 328 n74, 331, 331 n15 Anthology 6 n27, 9 n41, 23 n99, 321 Fig. 10, 312, 341 n44 Chahar Maqala 4 n16, 331 n18, 334 n28, 358 Gulistan 321 Fig. 10, 328, 329, 331, 345, 358 Kalila wa Dimna 15, 16, 16 n69, 18 n76, 34 n27, Qurʾan 293 Shahnama 4 n18, 6 n27, n28, 7, 22 n95, 25 n105, 85 n16, 31 Fig.5, 331 n16 Baysunghuri atelier 328, 328 n5, n7, 331 n18, 342 n48, 358 bayt al-ḥikma 391 Beach, Milo C. 100 Fig. 2, 101 Fig. 3 Becker, C. H. 118 n31 Beckwith, John 502 beggar 335, 350 Fig. 5, 357 n63 Bengal 106 n53 Berchem, Max van 96 n41, 106 n52,

513 110, 148 n86 Berg, Gabrielle van den 434 n60 Berkyaruk ibn Malikshah, Sultan 181 Fig. 14 Berlin. Museum für Islamische Kunst Anthology of Baysunghur(I.4628) 6 n27, 321 Fig. 10 Berlin. Staatsbibliothek Preussischer Kulturbesitz, Orientabteilung. Diez Albums 217 n21 (Diez A fol.71, S.56) 172 Fig 2 (Diez A fol.71, S.10) 177 Fig 9 (Diez A fol.71, S.29) 186 Fig 21 Bernus-Taylor, Marthe & Bittar T. 479 n9, 502 Bertalan, Sarah 166 n42 bestiaries 152,211, 236, 237 361 see: Kitab al-Hayawān of al-Jahiz Escorial Bestiary Ilkhanid Mamluk Manafiʿ-i hayavan of Ibn Bukhtīshūʿ Morgan Bestiary see: Manafiʿ-i hayavan of Ibn Bukhtīshūʿ Western medieval Bible, Leo Sakallarios 404 n82 Biblical account 119 see also: Gospels, Old Testament, NewTestament Bidpai 224, 429 Bihzad 8 n39, 57, 343 n53 Binyon, Laurence 4 n20, 7 n33, 9 n43, 23 n102, 25 n106, 53 n13 birds 14, 17, 26 n109, 79, 98, 154, 178 Fig. 10, 156, 219, 221, 226 n63, 234 90, 238, 239, 239 n109, n110, n111, 241, 244 Fig. 2, 245 Fig. 3, 248 Fig. 6, 249 Fig. 7, 250 Fig. 8, 251 Fig. 9, 252 Fig. 10, 253 Fig. 11, 257

514 Fig. 15, 262 Fig. 20, 264 Fig. 22, 269 Fig. 27, 279 Fig. 37, 283 Fig. 41, 313, 326, 449, 483, 493 n46, 494, 494 n48 Biruni, Abuʾl-Rayhan al Kitab al-Athar al-Baqiya ʿan al-qurun al-khaliya 111–148, 115 n21, 116 n22, 122 n49, 132–133 Figs. 5, 6, 139–143 Figs. 12–19, 154, 156, 162, 179 Figs. 11, 12, 201 Fig. 36, 217, 218 n28, n30, 236, 407 n89, 410 n98 bismillah 129 Fig. 1 Bizhan 23 n102, 67 n65, 68, 73 Fig.4, 375 Björkman, W. 118 n35 Black Div 374 Black Stone 437 Blair, Sheila S. 3 n12, 13 n56, 51 n2, 52 n5, 66 n64, 81 n8, 83, 84 n12, 89 n20, 92 n31, n32 93 n33, 96, 96 n41, 97 n43, 107, 108, 116 n22, 148 n86, 149 n1, 157, 157 n22, 23, 158 n25, n28, 161, 161 n35, n36, 165 n41, 167 n47, 168 n49, 242 n130, 296 n15, 344 n55, 366, 367, 372, 393 n67, n68, 408 n91, 430 n35, n38, 448 n83, 449 n94, n86 Blochet, Edgar 24, 60 n41, 103, 327 n3 blocks and colour 2, 5, 9, 10, 11, 21, 229, 409 and space 10, 11, 15, 21, 26 and text 164, 297, 299, 304, 312, 344, 370, 372, 375 Bloom, Jonathan 150 n6, 292 n7, 297 n18, 366 boar 12, 12 n55, 36 Fig.10, 228, 263

Fig. 21, 285 Fig. 43 boatman 339, 340, 341 n47, 355 Fig. 11 Bogdanov, Leonid T. 109 Boilot, D. J. 111 n1 Bologna, Biblioteca Universitaria. Hayūla ʿilaf al-ṭibb (Cod. arab. 2954) 383 n28, 391, 487 n27 Bombay 89 n21 Book of the Ascension see: Mi´raj Nama Book of Antidotes see: Kitab al-Tiryāq Book of the Bases of the ordinances on the subject of the knowledge of what is licit and what is forbidden see: Kitab qawāʾid al-aḥkam fī maʿrifat al-ḥalāl waʾl-ḥarām taṣnīf Book of Farriery see: Kitab al-Baitara Book of Fixed Stars see: Kitab Suwar al-Kawākib al-Thābita Book of Guidance into the Lives of the Twelve Imams see: Kitab al-Irshad Book of Ingenious Mechanical Devices see: Kitab fi maʿrifat al-ḥiyal alhandasiyya Book of the Wonders of Creation and the Peculiarities of Existing Things see: Kitab ʿajaʾib al-makhluqat wa gharaʾib al-mawjudat book-binders 151, 158, 344, 453, 454, 457, 459, 461, 464 bookbinding, Abbasid 463 Byzantine 452, 456 Chinese influence 461, 462 Christian 452 Coptic 452, 453, 454, 455, 456, 458, 461 Egyptian 452, 453, 454, 456, 457, 458, 459, 461 Indian 454, 464

INDEX



Iranian 453, 457, 458, 459, 460, 462, 463, 464 Islamic 452–464 Maghribî 459 Mamluk 458 Manichaean 457 Mongolian 457 Ottoman Turkey 458, 464 Persian 455, 460 Qairawân 454, 455 Timurid 460, 461, 462, 463 Tunisian 452, 453, 454, 456, 457, 458 Western European 452, 456, 459 Yemeni 458, 459 see also: Istanbul. Topkapi Saray Library; Istanbul. Türk ve Islam Eserli Müzesi Borgomale, L. Rabino di 430 n36 Boston, Mass. Museum of Fine Arts. Great Mongol Shahnama (22.392)198 Fig. 33 Qurʾan 460 Bosworth, C. E. 428 n26 Bothmer, H.-C.Graf von 211 n1.6, 224 n55, 296, 296 n17, 379 n10, 414 bow 5 bowls 25, 248, 249, 309, 310, 313, 315, 319, 322, 323, 324, 339, 359, 373, 386 n43 Boyle, John A. 90, 92 n31, n32, 93, 96 n41, 97, 107, 108, 109, 112 n11, 114 n19, 151 n8, 155 n19, 158 n24, n25, 380 n21 Brahmins 88, 95, 101 Fig. 3, 196 Fig 31, 437, 440 Fig. 1 Brandenburg, Dietrich. 382 n27, 383 n28, 414, 487 n27, 502 Brazen Hold 7, 20, 29 Fig. 3, 31 Fig. 5

515 Brend, Barbara 87 n18, 107, 301 n23, 346 n60, 405 m85, 414, 507 Brian, Doris 102 Figs. 4, 5 Briant, P 423 n12, 438 n78 Bridges, M. 439 n79 Brinner, William M. 110 Brockelmann, C. 111 n1 Browne, E. G. 58 n33, 62 n49, 107, 429 n31, 431 n45, 432 n47, n49, 433 n5, 436 n71, 439 n80 Buchthal, Hugo 223 n48, n53, 224 n55, 227 n66, 236 n98, 481 n14, 483 n19, n20, 502 Budde, H. 409 n92, 418 Buddha 242, 242 n127, 408 Buddhism / -ists apsaras 163 artist 236 n98 Ilkhans 96, 162 imagery 155, 169, 237 n101, 242 n127, 411 legends 208 Budge, E. A. Wallis 79 n2, 107 Bukhara School 327 n1 Bukhtnasar 155 Bukhtīshūʿ, Ibn Manafiʿ-i hayavan 153, 214, 214 n11, 217, 234, 235–242, 287–289, 273–276 Figs. 31–34, 273–276 Figs. 31–34, 279–284 Figs. 37–42 bull(s) 26 n109, 120 n43, 288 n135, 462 Bulliet, R. W. 120 n43 Buraq 137 Fig. 10 Bürgel, J. Christoph 22 n94, 81, 87 n18, 89 n21, 107, 435 n65, 439 n79 burial scene 17 see also:funerary practices; Iskandar, bier of Bustan of Saʿdi 7, 7 n37, 8 n40, 17

516 n72, 23 n102, 25 n108, 30 Fig. 4, 380 n18 Bustī, Abū Sulaymān Muḥammad b. Misʿar al- 378 Bustī, Abū Sulaymān Muḥammad b. Maʿshar al-Ḳudsī al- 378 n5 Buyids 160, 384 Buzurjimihr 183 Fig. 17 Byzantine architecture 390 art 15, 157, 226, 410, 489, 496, 498, 500 artists 390 bookbinding 452, 456 emperor 496 n54 enamels 386, 387, 410 icons 163 iconography 226, 404, 487, 491 influence 226, 227 n65, 230, 361, 387, 410, 489 manuscripts 160, 223, 225, 227 n68, 452, 492 mosaics 410 religious art 385, 404 tradition 215 n15, 381, 383, 387, 413, 492 world 378, 391, 492, 495 Byzantines 207 Byzantium 87, 235, 287, 456, 497 Caesar 105 n50 Çağman, Filiz 4 n17, n19, 6 n30, 11 n48, 15 n66, 16 n68, n69, 20 n86, 59–60 n40, 77 n78, 83 n10, 224 n55 Cahen, C. 430 n36 Cahill, J. 242 n128, 287 n131, 288 n138 Cain 235, 236 n98 Cairo 456, 461 General Egyptian Book Organ-

isation. Bustan of Saʿdi (ms. Adab Farsi 908) 7, 7 n37, 19 n81, 23, n102, 25 n108, 30 Fig. 4 National Library of Egypt 460 Bustan of Saʿdi (fols. 1b–2a) 8 n40, 17 n72 Kalila wa Dimna (744H) 288 n139 Kitab al-Baitara (cod.med. VIII) 232 n84 calendrical systems 112, 116, 117, 119, 132 Fig. 5, 179 Fig. 11, 152, 154, 156, 361, 430 calligrapher(s) 53, 92, 151, 158, 168, 236, 291, 292, 299, 312, 328, 345, 346 calligraphy 21, 52 n4, 64 n60, 150, 151, 210, 236, 236 n99, 237, 238, 290–306 passim, 332 n20, 367, 413, 427, 429, 465 Cambridge, University, Arts and Humanities Research Board 426 Cambridge-Edinburgh Shahnama project 376, 422 n7, 434, 438 n77 Cambridge. Kalila wa Dimna 224 Cambridge, Mass. Divan of Hafiz 59, 64 Cambridge, Mass., Fogg Museum of Art 6 n31, 57 n26, Cambridge, Mass., Harvard University Art Museums. Great Mongol Shahnama (1955.167) 171 Fig. 1 Great Mongol Shahnama (1960.190) 188 Fig. 23 Great Mongol Shahnama (1958.288) 194 Fig.29 camels 18 n76, 119, 120 n43, 214 n11, 222, 240 n119, 288 n137

INDEX

Capps, Walter H. 502 captions 237, 369, 372, 422 Carboni, Stefano 111 n6, 151 n10, n12, 223 n54, 409 n92, 422 n5 Carey, Christopher 423 n9 Carey, George 95 n38, 107, 152 n11, 436 n69 Carey, Moya 384 n33 carpets 5, 105, 168, 312, 457, 462, 464, 482, 483 Carpini 98 Carrying of the Cross 169 Caspian provinces 430 sea 437 catalogues 293, 301, 426 Cathay 149 Central Asia 81, 423, 426, 457 Central Asian frescoes 241 n124 textiles 373, 374 ceramics 149, 150, 165, 221, 222, 225, 236, 359, 425, 425 n18, 429, 457, 483 ʿAbbāsid 325 n29 and book painting 307–326 passim epigraphic 298, 308,309, 310, 311, 316 Fig. 3 Iranian 307, 311 n12, 317 Fig. 5, 325, 457 Jurjān 311 lustre ware 220 n41, n42, 222 n45, 309, 310, 311 n12, 314, 317 Fig. 5, 319 Fig. 8, 320 Fig. 9, 322 Figs. 11, 12, 323 Fig. 14, 324 Figs. 15, 16, 325 n29, n32, 394 n71 Māzandarān 325 n29 mīnāʾī ceramics 222 n45, 309, 310, 311, 311 n12, 313, 314, 315 Fig. 2, 319 Fig. 7, 323 Fig.

517 13, 326 Nīshāpūr pottery 325 n29 Raqqa pottery 222, 229 Samanid 298, 308, 309, 310, 311, 316 Fig. 3, 359 Seljuk 222 n45, 307, 308, 309, 310, 312, 313, 314, 325, 326 Valencian 221 n42 Chahar Maqaleh 4 n16, 331 n18, 334 n28, 358 Chaldiran 62, 62 n52, 63, 64, 68 Chao Meng-fu 288 n138 Chardin 89 n21 Charritat, M. 502 chatr 404 Chelkowski, P. J. 59 n39, 112 n12, 113 n13, 217 n20, 422 n6, 423–4 n13 chilla 151 China 20, 94, 97, 150, 154, 289, 379, 428, 437 Chinese art 153, 209, 241, 242 birds 156, 238, 241 calligraphy 210 costume 148, 169, 236 emperor 49 n112, 158 n26 handscrolls 153, 241 influence 169, 238, 241, 241 n124, 242 n129, 287, 463 landscapes 124, 153, 154 n17,169, 209, 241, 242, 287 n131, 287 maiden 330, 336, 337, 352 Fig. 8, 357 n62, n65 manuscript 98 motifs in: Al-Athar al-Baqiya 123, 124, 148, 156 bookbindings 460, 461, 462 Gulistan 330, 336, 337, 352 Fig. 8

518 Great Mongol Shahnama 169 Jamiʿ al-Tawarikh 209 Morgan Bestiary 154, 236, 238, 241, 242, 287 officials 169 peoples 158, 207 princess 198 Fig.33 seal script 403 silk 123 textiles 154 n17 Chitral 433 Choicest Maxims and Best Sayings see: Mukhtār al-ḥikam wa maḥāsin al-kalim Christensen, Arthur 89 n21, 97 n44, 107, 428 n27 Christian bookbinding 452 frontispieces 390, 391 girl 23 n99, 321 Fig. 10 iconography 227 imagery 120, 155, 156, 169, 227, 391, 489 minorities 235, 361 monk 20 n87, 121 painters 390 Christianity Eastern 380, 389, 390, 391, 404, 405, 406 in Iskandar cycle 87, 96, 105, 436 in Al-Athar al-Baqiya 120,121,124 n55, 125–128, 125 n60, 155, 156, 157 in: Great Mongol Shahnama 169 in : Jamiʿ al-Tawarikh 160, 162 Western 380, 389, 406 Chronicle of Prophets and Kings see: Tarikh al-rusul waʾl-muluk Chronology of Ancient Nations see: Kitab al-Athar al-Baqiya

Chu Jui 287 n131 Ciancaglini, C. A. 435 n66 Cincinnati Art Museum. Shahnama (1947.498) 185 Fig. 20 classical influence 22, 55, 57, 225, 230, 235, 287, 380, 381, 384, 390, 391, 404, 406, 455, 487, 491, 492, 495, 498 Cleofide 436, 436 n70 Cleveland, Museum of Art Shahnama 25 n107 Shahnama (44.479) 185 Fig. 19 Great Mongol Shahnama. (1943.658) 193 Fig. 28 Clinton, J. W. 432 n47 cloud(s) 120, 123, 124, 127, 209, 304, 461, 462 cloud collar 461, 462 Codex Amiatinus 389 Cohen, M. 106 n52, 108 Cohn, William 287 n131, 288 n138 coins106 n53 Ilkhanid 92 n31, 148 n86 Sasanian 430 Umayyad 118 colour 210, 229,325, 326, 358, 359, 362, 366, 372, 409, 466, 483, 486, 487 in bookbinding 456, 457, 458, 463, 464 in illuminated Qurʾans 299, 300, 304, 305 in images of the prophet 120, 123, 124, 128 in the Gulistan 335–341, 343, 358 in Timurid book painting 2, 5–11, 16 n67, 21, 24–25, 49, 50 columns 3, 22, 387, 388, 399, 403, 404, 404 n80, 405, 406 Combe, Etienne 108 Compendium of Chronicles see: Jamiʿ

519

INDEX

al-Tawarikh Compendium of Histories see: Jamiʿ al-Tawarikh composition 5, 8, 10, 11, 12, 13, 15, 19, 21, 22, 49, 67, 82, 121, 156, 159, 162, 164,209, 218, 222, 228, 229, 230, 231, 237, 295, 307, 309, 311, 312, 313, 314, 325, 326, 330, 334, 341, 342, 387, 405, 411, 422, 423, 463, 464, 465, 466, 483, 485–489, 493, 496, 498 Contadini, A. 223 n54, 386 n41, 414 connoisseurship 372 Constantine the Great 105 n50 Constantinople 105, 120 Bible, Leo Sakallaros 404 n82 convention(s) 1, 14, 18, 26, 49, 169, 211, 215, 221, 224–228, 230–236, 241, 299, 362, 375, 388, 389, 461, 465, 466, 480, 485, 490, 498 Coptic architecture 390 bookbinding 452, 453, 454, 455, 456, 458, 461 tradition 230 Cordova 298, 359, 421, 463 cosmography / ology152, 154, 361, 384 costume 53, 80, 118–119, 118 n35, 123, 126 n62, 147, 148, 159, 168, 169, 208, 233, 236, 325, 338, 373, 374, 375, 380, 380 n17, 386, 406, 407, 408, 409, 409 n92, 410, 413, 486 n25, 491 n40 coulisse device 19, 26, 239 Cowan, Jill S. 3 n14, 13 n56, 152 n9, 224 n55 Creswell, K. A. C. 118 n30, 295 n14,

424 n16 Crucifixion 156, 169 cup 448 Curley, M. J. 213 n9, 240 n117 Curtis, V. 332 n20, 507 Cyrus 105 n50 dado(es) 10, 24, 168, 338 Dagon, temple of 3 Ḍahhāk 89 n21 see also: Zahhak Damascus 296, 300, 383 n32, 428 Danishmendid amīr al-wāthiq 106 n53 Dar al-Kutub 456 Dārā 23 n102,66, 84, 85, 85 n15, 86, 438 Darius 438 Darab 66 Darmesteter, James 79 n1, 89, 89 n21, 96 n39, 108, 435 n66 darvish 23 n98, 335, 336, 341 n44, 434, 350 Fig. 5, 351 Fig. 7, 357 n65 Dastan-i Jamal u Jalal of Asafi (O Nova 2) 16 n68 Daulatshah 328 n7 Dauphin, C. M. 230 n76 David-Weill, J. 288 n135, 310 n9 Davis, Edward 83 n11, 86 n17, 87 n18, 109, 422 n7 Death of Shaghad 209 Delbrueck, Richard 391 n61, 414, 489 n31, n32, 502 Delhi 81 De Materia Medica of Dioscorides 2, 223, 361, 382 n25, 384 n36, 412 n104, 481 Demotte, George. Shanama 3, 4, 10, 13, 52, 82, 166, 218, 242, 344, 357, 366 see also: Shahnama, Great Mongol Demus, Otto 15 n62, 26 n64, 385

520 n39, 386 n40, 410 n96, 490 n37, 502 Deposition 169 Derman, M. U. 304 n26 Der Nersessian, S. 390 n58, 414 Déroche, Franҫois 292 n3, n7, 294, 295–299, 305, 306 Description of Animals see: Naʿt al-Hayawān Detroit Institute of Arts Great Mongol Shahnama (35.54) 190 Fig. 25 Deuteronomy 115 n21 Dhubyāni, al-Nābigha al- 360 Dhuʾl-Faqār123 Dhuʾl-Faqār Širwānī 92 Dhuʼl-Qarnain 80, 106, 106 n53 Dickson, M. B. 51 n1, n2, n3, 52, 53 n9, n10, n11, 59 n36, 62 n53, 63, 63 n56, n57, n58, 64 n61, 68 n66, 69 n68, 77 n76, n78 Diez Albums (Diez A fol.71, S.56) 151, 172 Fig 2 (Diez A fol.71, S.10) 153,177 Fig 9 (Diez A fol.71, S.29) 186 Fig 21 Digby, Simon 5 n24, 9 n43, 56 n21 Dimand, M. S.236 n98 dinar 336 Dīnawarī, al- 79 n2 Dioscorides De Materia Medica 2, 223, 361, 382 n25, 384 n36, 412 n104, 481 Hayūla ʿilaf al-ṭibb 122 n48, 226, 228, 382 n25, n27, 383, 383 n28, 384 n36, 391, 399 Fig. 6, 400 Fig. 7, 479 n8, 481 n15, 487 n27, 500 n62 dishes 321, 309–314, 316 Fig. 3, 322 Fig. 12, 323 Fig. 13, 429

diptychs 489, 489 n33 Div Akhvan 20 n87, 375 Divan 13, 37 Fig. 11, 59, 64, 217, 272 Fig. 30, 233, 360 Anthology of Divans (ms 132) 152, 175 Figs. 5, 6 Divan of Ahmad Jalaʾir 344 n57 Divan of ʿAli Shir Navaʾi 59 n40 Divan of Amir Khusrau 60 n40 Divan of Hafiz 59, 64 Divan of MirʿAli Shir Navaʾi 59 Divan of al-Nābigha al-Dhubyāni, 360 Djait, H. 377 n4, 414 Djāḥiẓ,al- 214 n12 Dodge, B. 291 n1 Dodkhudoeva, Larissa 83 n11, 86 n17, 87 n18, 108, 426 n20 dogs 23, n101, 233, 234 n90, 261 Fig. 19, 267 Fig. 25, 335, 343, 351 Fig. 6, 356 Fig. 13, 483 Dolfin, Zorzo 106 n51 Dominicans 169 Donner, F. M. 427 n23 Donzel, E. van 437 n74 doors 6, 6 n29, 8, 9 10, 17, 24, 25, 116, 162, 337, 338, 339, 359, 464 dragons 10 n45, 16, 18 n78, 43 Fig. 15, 85, 86, 169, 193 Fig. 28, 242, 437, 462 drapery 3, 128 m73, 153, 19, 210, 227 n65, 236 n98, 386, 388, 410–413 Dublin, Chester Beatty Library. Epics (P.114) 9 n43, 57 n26, 367 First Small Shahnama (Per.104.5) 84, 184 Fig. 18 Furūsiyya 215 n14 Great Mongol Shahnama (Per.111.8) 85, n14, 191 Fig. 26 Great Mongol Shahnama

INDEX

(Per.111.4) 195 Fig. 30 Gulistan (P.119) 5 n24, 17 n72, 21 n93, 23 n98, 23 n101, 23–24, 32 Fig. 6, 33 Fig. 7, 47 Fig. 19, 327–358, 328 n7, 329 n10, 347–356 Figs. 1–13, 346 n60 Khamsa of Amir Khusrau (P.163) 6 n27 Khamsa of Nizami (P.137) 23 n97 Qurʾans 293, 457, 460, 462, 464 Qurʾan of Ibn al-Bawwāb 466 n6 Shahnama (Per.214) 84 n13 Shahnama 343 n52 Tarikh (P.144)18, 46 Fig. 18 Dubois, S. 502 Dukas Michael 106 n51 Dunimarle, Shahnama 56 Durayhim, ibn al- 361 Dust Muhammad, Dibacha-yi 51 n3, 53, 149 149 n1, 162, 167, 167 n49, 328 n7, 461 Eastern Isles 363 Eastern Persian script 294 Ebbo Gospels 388 n52 ebru 304 n26 Eche, Y. 421 n2 Echmiadzin. Diptych 489 n33 Gospels 489 Edinburgh, University, Arts and Humanities Research Board 426 Edinburgh, University of Edinburgh Library Kitab al-Athar al-Baqiya ʿan alqurun al-khaliya (ms. Arab 161) 111–148, 111 n3, 116 n22, 132–133 Figs. 5, 6, 139–143 Figs. 12–19, 154, 156,

521 179 Figs. 11, 12, 201 Fig. 36, 407 n89 Jamiʿ al-Tawarikh (ms. Arab20) 2 n7, 3 n10, 3, 15 n63, 19 n84, 21 n91, 157–162, 158 n26, 180 Fig. 13, 181 Figs. 14,16, 207– 210, 218 n30, 360, 403 n79 Egypt 218, 221, 222, 229, 230, 289, 292, 300, 357 n62, 362, 407 n89, 425 Egyptian bookbinding 452, 453, 454, 456, 457, 458, 459, 461 ivories 380 n17 Qurʾan 294, 300 elephants 12, 172 Fig. 2, 214 n11, 228, 234 n90, 236 n98, 237– 239, 237 n101, 238 n104, 276 Fig. 34, 277 Fig. 35, 278 Fig. 36, 375 Elgood, R. 211 n1.2, 212 n4, 342 n52 Eli 120 n42, 156, 156 n20, 179 Fig. 12 Embree, A. T. 113 n13, 114 n16 Ems, Rudolf von, Weltchronik 161 enamels, Byzantine 386, 387, 410 Enderlein, V. 312 n18, 321 Enderwitz, S. 427 n24 Enoch 87 enthronements 3 n13, 4, 4 n18, 25 n108, 93, 103 Fig. 5, 121 n48, 156 n20, 159, 161, 181 Fig. 14, 203 Fig. 39, 208–209, 218, 233, 313, 314, 374, 383, 429, 438, 495 n53 see also: thrones Entombment 169 Entry into Jerusalem 169 Epernay. Bibliothèque Municipale, Ebbo Gospels 388 n52 epic(s)55, 60, 64, 69, 81, 84, 86, 96, 118 n33, 151, 152, 162, 163,

522 165, 168, 169, 208, 367, 371, 373, 375, 420, 422, 423, 424, 428 n25, 431, 433, 435, 439, 449, 450 epigraphy 296, 430 see also: ceramics Epistles of the Brethren of Purity and the Friends of Loyalty 377, 378 see: Rasāʾil Epistles of the Sincere Brethren 377– 419 passim Eqbal, Z. 424 n13 Escorial Bestiary see : Manafiʿ-i hayavan Esin, Emel 237 n101, 481 n15, 502 Ethé, H. 111 n5 ethics 152, 163, 434 Ethiopian codices 452 Ethiopic versions Alexander Romance 79 n1 Ettinghausen, Richard 2 n3, n 6, 3 n8, n9, 7 n38. 12 n53, 17 n73, n74, 26 n110, 56 n24, 116 n24, 118 n33, 119 n36, 120 n43, 122 n48, 128 n73, 212 n4, 215 n15, 217 n19, 218 n24,220 n39, 221 n42, 2 n6, n47, 223 n52, n53, 226 n62, 227 n67, 232 n82, 235 n92, 236 n98, 237 n100, 239 n107, 244 Fig. 2, 292 n7, 325 n32, 334 n27, 335 n31, 341 n44, 363, 377 n2, 378 n6, n8, 379 n12, 380 n21, 381 n22, 382 n25, n26, 384 n37, 385 n38, 387 n47, 388 n48, n49, 396 Fig. 2, 397 Fig. 3, 404 n80, 410 n94, 411 n100, n101, 413 n106, 414, 415,466 n2, 481 n14, 487 n27, 489 n34, 492 n43, n45, 495 n51, 497 n58, 500 n62, 502 European influence / tradition 1, 20, 98, 121, 153, 169, 208, 238, 240, 456, 459, 495

Evangelists see: portraits Evans, Godfrey 374 Evans, Helen.C. 382 n26, 383 n31, 386 n44, 415 Eve 235, 235 n95, 236, 237 n101, 273 Fig. 31, 388 n50, 390 n56, 404 n82, n83, n84 Ewert, Christian 307 n1 ewers 309 Exodus 160 Ezra 389 fables 152, 169 n50, 224, 361, 429 Fagfūr of China 97 Falk, Toby 3 n13 Falke, Otto von 493 n46502 falnama 304 Fan-lung 242 n128 Farain Gudar 69 Faramarz 6 n27, n28, 200 Fig. 35 Far Eastern influence 153, 157, 158, 169, 210, 242, 366, 379, 413 Farès, Bichr 223 n50, 377, n1, n2, 378 n8, 382 n26, 383 n30, 404 n80, 412 n104, 415, 482 n16, 497 n57, 502 Farhad 6 n26, 341 n45 Faridun 66, 68, 74 Fig.2, 93 n33, 373, 374 farr 93 n35 farriery 361 farr-i kayani 430 Farruhī 91 n25 Farshidward 20 n88 Farsi, Adab 30 Fig. 4 Farsi Barlas, Amir 58 n32 Fars school 373 Farud 186 Fig. 21 Fath ʿAli Shah 464 Fatiha 299 Fātik, Abū ʾl-Wafāʾ al-Mubashshir ibn 361, 378 n8, 384, 401 Figs.

INDEX

9, 10, 402 Fig. 11, 481 n15, 482 n16 Fatima 125, 125 n60 126, 127, 128 Fātimid era 221, 229, 230, 494 n48 Fehér, G. 59 n40 Fehérvári, Géza 311 n13, 312 n20, 317, 502 Ferducci, Lillo 105 n50 Ferrand, G. 437 n76 Ferrandis, J. 363 finispiece 300, 379 n9, 384, 402 Fig. 11, 477 Fig. 8 Finster, B.506 Firdausi, Abu ʾl-Qasim 53 n8, 55, 57, 58 n33, 65, 66, 67, 69, 76 Fig. 7, 77, 77 n74, 81, 82, 83, 86, 87, 90 n22, 93, 96, 96 n42, 163, 165, 168, 169, 308 n4, 329, 332, 367, 368, 369, 372, 373, 424, 425, 426, 427, 428 n25 429, 431, 431 n43432, 433, 434, 435, 436, 438, 439, 449, 450, 451, 464 Fitzherbert, Teresa 13 n58, 98 n45, 152 n13, 164 n39, 344 n56, 345 n58, 345, 439 n82, 481 n15, 502 Flagellation 169 Flecker, James Elroy 79 Florence Shahnama 424 Flury, Samuel 295, 310 n9 Folda, Jaroslav 483 n20, 502 Folsach, K. von 410 n94, 415 Fontana, M. V. 147 n84 Forster, B. 226 n61 Fragner, Bert 433 n54 Fragner, C. 506 France 428 Franciscans 169 Franks 149, 158 Fraser, Marcus 297, 297 n19 Freeman-Grenville, G. S. P. 437 n76

523 Free Men’s Companion to the Subtleties of Poems see: Muʾnis al-ahrar fi daqaʾiq al-ashʿar French art 169 French, G. 421 n3 frescoes 168, 206, 209, 241 n124, 374, 425 Friedländer, Israel 87, 108 Friend, A. M. 80 n16, n17, 382 n24, 404 n83, 415 Froissart, Sir John 105 n49, 108 frontal plane 164, 326, 362, 381 n22, 387–389, 392 frontispiece, 25 n107, 128 n73, 334 n27, 335 n31 Arab 378, 379, 380, 381, 382, 384, 390, 391 Bustan of Saʿdi 380 n18 Christian 390, 391 double frontispiece 379 n9, 382 n26 double frontispiece to Rasāʾ il 122 n48, 377–419 passim. 378 n8, double frontispiece Risālat al-Ṣūfī fiʾl-kawākib 384 n33 Hayūla ʿilaf al-ṭibb of Dioscorides 382 n27 Persian 379 Maqamat (Schefer ms) 465–505 Qurʾanic 378, 389 Shāh-nāma-yi Shāhī 53 n8, n14, 57 n30 Sulwan al-Muṭāʿ 379 n11 funerary practices 6 n29, 17, 80, 85, 99 Fig. 1, 204 Fig. 40, 373, 447 Fig. 10, 449 Fūr 84 n14, 85, 85 n15, 86, 444 Fig. 6, 446 Fig. 9 furnishings/furniture 11, 21, 26, 156 n20, 232, 375, 389 Furughi, Mirza Muhammad ʿAli 431

524 n40, n44 Furūsiyya 215 n14, 218 n29, 228 n69 Gaillard, M. 435 n65 Gang Bihisht 7 n33 gardens 9 n43, 25, 25 n106, 334, 336, 337, 339, 340, 342 n48, 348 Fig. 2, 358 Garshasp 3 n13 Gayangos, P. de 421 n2 Gaykhātū 235 Gayraud, G. 502 Gayumars 53–54 n14, 70 Fig. 1 Gāzān Ḫān 90, 90 n23, 96 Gemini 359 Genesis 235 n95 Genghis Khan 95, 98, 105, 118 n33, 439 Geoffrey of Rhodes 105 n51 Ghadir Khumm 129, 147 Gharāʾib al-sighār 60 n40 Ghazan Khan 114, 150, 158, 366 Ghaznavids 160, 161 Ghiyath al-Din 82 n8, 166 Ghouchani, H. 363 Gibb, Edward J. W. 81 n7, 108 Gibb, H. A. R. 427 n24 Gierlichs, Joachim 494 n47, 503 gilding 459, 460, 462 giraffe 220, 228, 237 n101, 239 n112, 246 Fig. 4, 281 Fig. 39 Ġiyāt 89 n20 glassware 359 Glubb, J. B. 427 n23 Gnoli, G. 435 n66 goats 19 n84 goblet 491 Gog and Magog 80, 85, 87, 94, 95 100 Fig. 2, 437, 438, 441 Fig. 3, 444 Fig. 7 Goitein, S. D. 378 n5, 415

gold background 384, 385, 390, 392, 498 gold tooling 457, 458, 459, 461 Golden Haggadah 160, 161 n33 Goldziher, I. 426 n22, 427–428 n25 Golombek, Lisa 7 n37, 8 n39, 11, 57 n29, 330 n11, 339 n39, 392 n65, 415, 423 n8 Gölpinarlı, B. 217 n23 Gombrich, Ernest H. 109 Gorelik, M. 342 n52 Gospels 169, 390, 489 Ebbo Gospels 388 n52 Echmiadzin 489 Murano 489 Queen Keran Gospels 388 n50 Gotha, Forschungsbibliothek Kitab ʿajaʾib al-makhluqat wa gharaʾib al-mawjudat (ms. A1506) 223 n54 Grabar, A. 404 n82, 415 Grabar, Oleg 3 n12, 13 n56, 52 n5, 66 n64, 82 n8, 84 n12, 97 n43, 108, 165, n41, 167, 167 n47, 215 n14, 226 n61, 242 n130, 344 n55, 364, 382 n26, 390 n59, 393 n67, n68, 408 n91, 412 n104, 422 n5, n6, 415, 448 n83, 449 n84, n86, 466 n2, 479 n10, 481 n11, 489 n33, 492 n41, 494 n48, 497 n58, 503, 500 n61 Graeco-Roman era 235, 390, 492 Granada, The Alhambra 359 Court of the Lions 359 Hall of the Two Sisters 359 Gray, Basil 1 n2, 2 n7, 3 n10, n14, n15, 5 n23, n24, n25, 6 n26, n27, n28, 7 n34, n36, n38, 9 n41, n42, n43, 11 n50, 13 n57, n58, 14 n60, 16 n71, 17 n72, 19 n82, n83, 20 n87, n88, 21

525

INDEX

n93, 22 n95, 23 n96, n98, n99, 25 n105, n106, n107, n108, 26 n109, n110, 49 n112, n113, 53 n13, 56 n20, n25, 64 n62, 83 n9, 89 n20, 99 Fig. 1, 108, 116 n22, 156 n21, 207, 217 n21, 235 n92, 240 n115, 241 n122, 242 n125, n127, n128, 286, 287 n133, 288 n136, n140, 289 n141, 312 n18, 330 n14, 333 n24, n25, 334 n28, 341 n45, 344 n54, n56, n57, 358 n66, 393 n69, 415 Great Mongol Shahnama see: Shahnama, Great Mongol see also: Demotte, Georges Shahnama Greece 439 Greek literature 214 Grohmann, A. 111 n4, 122 n48, 147 n84 Grube, Ernest J. 1 n2, 6 n28, 7 n38, 10 n47, 16 n70, 19 n84, 54 n18, 122 n48, 213 n6, 217 n19, 218 n27, 220 n41, 228 n69, 232 n84,235 n92, n94, 287, 287 n132, n134, 288 n137, n139, 311 n11, 314 n26, n27, n28, 315, 325 n32, 358 n67, n69, 364, 382 n27, 383 n28, 415, 429 n30, 481 n14, 503 Grunebaum, G. E. von 431n43 Guest, Grace 367 Guest, Rhuvon 80 n4, 108 Guillaume, A. 115 n20 Gul 334 Gulistan-i hunar of Qazi Ahmad 51 n4 Gulistan of Saʿdi 5 n24, 7, 17 n72, 21 n93, 23 n98, n101, 23–24, 32 Fig. 6, 33 Fig. 7, 47 Fig. 19, 327–358 passsim 328 n7, 329

n10, 332 n21, 334 n26, 342 n48, 343 n52, 345 n59, 346 n60, 347–356 Figs.1–13 colour in 337–340, 343 space, use of 341, 344 Gulnar 22 n95 Gulshāh 315 Fig.1 Gunther, Ursula161 n34 Gupta empire 431 Gushtasp 66 Gustaham 20 n88 hadith 123 n53, 151 Ḥadīth Bayāḍ wa Ritāḍ 361 Hafenrichter, H. 506 Hafiz, Divan 59, 64 hajib 25 n107, n108 Hakam, al- III 421 Haldane, J. D. 211 n1.4, 212 n4, 213 n8, 215 n14, 220 n38, n40, 225 n60, 232 n83, 241 n123, 364, 410 n97, 415 Haleem, M. A. S. Abdel 214 n13 Halm, H. 123 n52, 127 n68 haloes 80, 118, 121, 123, 128, 128 n73, 227 n65, 236, 304, 325, 390, 411, 449, 491, 493 Hamadan 460 Hamadhānī, al- 498 n58 Hamid, I. S. 383 n32, 415, 497 n56 Ḥamza al-Iṣfahānī 93 Hanaway,William A. 81 n6, 89, 108, 434 n61 Ḫāqān-ī Sikandar šāʾn 91 n26 Harf-Lancner, L. 435 n64, n65, n67 Harith, al- 145 Fig. 18 Hariri, al-Maqamat 1 n1, 2, 3 n. 8, n9, 12, 12 n52, 21 n92, 118, 119, 211 n1.3, 1.4, 213 n8, 215 n14, 219, 220, 222, 224, 231, 232, 233, 314, 362, 363, 382 n32, 387 n47, 388 n51. 405,

526 407, 413 n106, 422 n5, n6, , 465–507 passim Harley MS. 4751 278 Fig. 36 Hartmann, Bishop of Augsburg 373 Hartmann, Angelika 391 n62 Hartner, W. 120 n43 Harum 88 Hārūn 378 n6 Hasan, al- 77, 119, 122, 125 n60, 126, 127 Hasan ibn Muhammad ibn ʿAli ibn Husain al-Mawsili 187 Fig. 22 Hasse, Johann Adolf 436 Hātifi 98 n46 Hawke, D. M. 214 n12 Hayūla ʿilaf al-ṭibb of Dioscorides 122 n48, 226, 228, 382 n25, 383 n28, 391, 399 Fig. 6, 400 Fig. 7, 481 n15, 479 n8, 487 n27, 500 n62 Hebdomades 406 Hebrew prophets 432 Henri of Navarre 78 Henricus de Allemania, Liber ethicorum 409 n92 heraldic style 221 Herat, 57, 58, 60 n40, 106, 348 Fig. 2, 392, 461, 463 Herrmann, Georgina 503 Herzfeld, Ernest E. 98, 98 n46, n47, 105, 108, 466 n3, 503 Hickman, William C. 106 Ḫiḍr 85, 85 n16, 87, 88, 92 Higgitt, John 240 n116 Hijazi script 294, 297 hikayat 332 n22 Ḥīlī, al- 412 n104 Hill, D. R. 211 n1.1 Hillenbrand, C. 391 n63, 415, 503 Hillenbrand, R. 9 n43, 16 n68, 23 n97, 56 n22, 108, 110, 112 n8, 301 n23, 327 n1, 332 n20, 338

n39, 340 n43, 343 n52, 390 n59, 394 n74, 405 n85, 407 n89, 410 n98, 414, 415, 416, 417, 421 n3, n4, 425 n18, 426 n21, 439 n79, 450 n87, 479 n10, 492 n44, 503, 506, 507 Himalayas 433 Hind, King of 75 Fig. 6, 444 Fig. 6, 446 Fig. 9 Hindus 95 Hindustan 94 Hīra 360 Hishām, Caliph 233 History of the World Conqueror see: Tarikh-i jahan-gusha Hodgson, M. G. S. 421 n3 Hoffman, Eva Rose F. 335 n32, 381 n22, n23, 416, 466 n2, 483 n19, n20, 494 n48, 500 n62, 503 Holter, Kurt 211 n1.3, 223 n53, 479 n7, 481 n14, 503 Homeric epics 423, 424, 434 Homilies of St. John Chrysostom 496 n54 Hopkins, Clark 490 n36, 503 horses 14, 14 n60, 36 Fig. 10, 119, 202 Fig. 37, 209, 228, 241, 265 Fig. 23, 285 Fig. 43 Houghton Shahnama see: Shāhnāma-yi-Shāhī Hourani, A. H. 422 n5 Howard, I. K. A. 125 n61 Huart, Clément 21 n92 Hughes, J. 292 n5 Hukk, M. Ashraful 111 n5 Hulagu 112, 235, 380, 380 n21 Humām-i Tabrizi 90 n23 Humay 6 n26, 49 n112, 28 Fig. 2, 287 n134, 330 n13, 344 n56, Humayun 11 n49, 49 n112, 287 n134, 330 n13, 344 n56 hunting 63, 189 Fig 24, 313, 429,

INDEX

462, 464 Hurmuzd 6 n29, 23 n97 Husain, al- 119, 125 n60, 126, 127 Husain Bayqara, Sultan 57, 58, 58 n32, 77, 106, 463 Husain Mirza 149 Hushang 374 Ḫusraw 86 Ḫwārnāma 81 Ibrahim, Sultan 9 n43, 25 n106, 56, 56 n19 Icelandic 429 Iconography 86, 91, 115, 123, 148, 152, 153, 163, 166, 210, 224, 226, 227, 232, 308, 314, 330, 362, 363, 390, 404, 412, 422, 462, 479, 480, 485–487, 489– 492, 496–500 in Shahnama 365–376 passim in Shāh-nāma-yi Shāhī 51–78, 70–76 Figs. 1–7, 77–78 Iliad 420 Ilkhanid dynasty 90, 92, 93, 94, 94 n37, 95, 96, 98, 105, 112– 116,121, 123, 128, 148, 164 n39, 360, 403, 410, 449 architecture 9, 158, 169, 393, 394, 396, 403 astronomy 112 bestiaries 235–242, 287–289 book painting 12, 15, 19, 21, 26, 149–206, 151 n8, 213, 235–242, 241 n124, 287–289, 379, 380, 403 ceramics 149, 165 coinage 92 n31, 148 n86 metalwork 91, 149 religion 96, 235 Shāh-nāma 60, 61, 421 n5 woodwork 403 illumination 112, 150, 155, 207,

527 208, 213, 225, 290–306 passim, 312, 313, 316 Fig. 4, 328, 329, 329 n10, 331, 332 n20, 345, 346, 356 Fig. 12, 358, 370, 422, 456, 458, 460, 466 Illuminationist School 96 illuminators 151, 344, 414 Inal, G. 217 n21 Inalcik, Halil 108 India 79, 91 n25, 94, 106 n53, 113, 164, 173 Fig. 3, 180 Fig.13, 242 n127 292, 294, 300, 304, 367, 414 n107, 423, 425, 432, 433, 464 Indian bookbinding 454, 464 influence, 237 n101, 433 n56 legends 208, 362, 436 princess 4 n19 Qurʾans 300 Indians 236 n98, 411, 436, 438 Indonesian 429 Inju dynasty 16, 83 n10, 106 n52, 165, 372 inscriptions 21, 53, 77, 91, 92–93 n32, 96 n41, 98, 106 n53, 165, 297, 309, 310, 311, 312, 317 Fig. 5, 336, 359, 360, 362, 367, 430, 455, 456, 458–460, 487 see also: epigraphy Ipşiroğlu, Mazhar Şevket 217 n21, 237 n101, 242 n129, 309 n6, 495 n53, 503 Iraj 68, 374 Iran 60, 61, 62, 63, 79, 89, 92, 114, 120, 213 291, 304, 308, 316, 365, 373, 413, 421, 422, 423, 425, 434, 435, 438 artists 218, 224, 237, 289 bookbinding 453, 457, 458, 459, 460, 462, 463, 464 bookpainting 54 n15, 149–206,

528 154 n17, 219, 222, 224, 235, 288, 289, 292, 309, 365, 375, 403 n79, 414, 420–451 ceramics 301, 311 n12, 317 Fig. 5, 325, 457 conquest by Arabs 426, 427, 428, 430 culture 20, 60, 89, 90,392, 413 eastern 296 Ikhanids 94, 105, 112, 149–205 invasion by Mongols,439 medieval 420, 448 metalwork 374 northwestern 183–184 Figs. 17–18, 201 Fig. 36, 235 poet 420 Qurʾans 296, 301, 304, 305, 308, 460 rulers 61, 64, 89, 92, 93, 94 textiles 373 Iran-Turan feud 61–70, 77–78 Iraq 152, 154, 202 Fig. 37, 203 Fig. 38, 222, 231, 241, 316 Fig. 4, 362, 373, 395 Fig. 1, 398 Fig. 2, 453 architecture 392 northern 149, 179 Figs. 11, 12, 201 Fig. 36, 302 southern 377 tradition 411 Isaiah 119 ʿIsa Salman Hamid 383 n32, 415, 503 Isadore, Etymologiae 240 n114 Isfahan, Iran 89 n21, 152, 164, 186, Fig 21, 203 Fig. 39, 294, 374, 422 n5, 461 Isfandiyar 7, 20, 29 Fig. 3, 31 Fig. 5, 66, 84, 94, 185 Figs.19, 20, 194 Fig. 29, 375 Ishaq, Ibn, Sira 117 n26, 122, 122 n48, 123 n51, 125, 125 n61,

126 n62, n63 Iskandar 66, 66 n64, 67, 84 n12, 91 n25, 98 n46, 242, 450 and the baby at Babylon 84 n14 bier of 80, 89 n20, 93, 99 Fig. 1, 204 Fig. 40, 447 Fig. 10, 449 builds wall against Gog and Magog 80, 85, 87, 100 Fig. 2, 437,438, 441 Fig. 3, 444 Fig. 7 builds Iron Rampart 169, 197 Fig. 32 and Dara 66,85 n15, 438 enthroned 103 Fig. 5, 445 Fig. 8, 449 fights the kardagan 437, 440 Fig. 2 and the Indian princess 84 n14 Iron cavalry 171 Fig. 1, 446 Fig. 9 and the Fur of Hind 84 n14, 438, 444 Fig. 6, 446 Fig. 9, 449 at the Kaʿba 87 n18 and Land of Gloom 82, 85 n16, 97, 97 n43, 104 Fig. 6, 192, Fig. 27, 437, 443 Fig. 5 and Queen Qaidāfa 84 n14 and talking birds 84 n14, 97 n43 and Talking Tree 79, 84, 84 n14, 85 n16, 91, 95, 97,97 n43, 102 Fig. 4, 437, 442 Fig. 4 portraits 87 visits Brahmins 88, 95, 97 n43, 101 Fig. 3, 169, 196 Fig. 31, 436, 437 440 Fig. 1 and Water of Life 437, 438 Iskandar cycle context 83–85 choice of scenes 85–88, 449 and Mongols 88–98, 94 n37, 95 n38, 105–106 in Great Mongol Shahnama 52,

529

INDEX

79–110, 438, 438 n77, 439, Iskandar, Sultan 336 n33 see: Anthology of Iskandariya, al- 80 Iskandarnāma of Aḥmadī 81 of Amīr Ḫusraw 81 of Firdawsī 81, 82, 83, 86, 87, 93, 96 ism 378 Ismāʿīlīs 378, 406 Ismaʿil, Shah 53–54 n14, 59, 59 n36, 61 n49, 63, 64, 84, 461 Isrāfil 96 Iṣṭaḫr 93 Istanbul, Evqaf Museum 460 Istanbul Jamiʿ al-Tawarikh 208 Istanbul, Kariye Cami 404 n82 Istanbul, Nuruosmaniye Library Qurʾan 294 Istanbul, Süleymaniye Mosque Library Hayūla ʿilaf al-ṭibb (Ayasofya 3703) 399 Fig. 6, 382 n27, 400 Fig. 7 Hayūla ʿilaf al-ṭibb (Ayasofya 3704) 122 n48, 383 n29 Kashf al-Asrār (Lala Ismail 565)232 n83 Kitab fi maʿrifat al-ḥiyal al-handasiyya (no. 3606) 211 n1.1 Rasāʾil (Esad Efendi 3638) 2, 122 n48, 335 n31, 377, 396 Fig.2, 397 Fig. 3, 399 Fig. 5, 481 n15, 492 n44, 500 n62 Istanbul, Topkapi Saray Library Album (H.2152) 237 n101, 481 n15, 495 n53 Album of Bahram Mirza (H.2154) 328 Anthology (H796) 25 n107 Arzadasht (H2153) 328 n7



bookbindings TKSL (no.976) 462 TKSL (no. 1405) 463 TKSL (no. H.676) 463 Gharāʾib al-sighār (R. 803) 60 n40 Hayūla ʿilaf al-ṭibb of Dioscorides (Ahmet III, 2127) 382 n25, 479 n8, 481 n15 Jamiʿ al-Tawarikh (H.2153) 217 n21 Kalila wa Dimna 4 n17, 152 n9, 288 n135, n137 Kalila wa Dimna of Baysunghur, (H362) 10 n47, 11 n48, 15 n64, 16 n69, 18 n76 Kalila wa Dimna (H363) 482 n16 Kalila wa Dimna of Nasr AbuʾlMaʿali (R.1022) 10 n47, 15n64, 26 n109, 34 Figs. 8a, 8b, 40 Fig. 13a, 41 Fig. 13b, 213, 334 n27 Khamsa of Nizami (H.781) 4 n19, 6 n30, 7 n34, 11 n49, 15, 49 n113, 42 Fig 14 Khamsa of Nizami (H.762) 16 n68 Kitab al-Baitara (Cod. Coll. Ahmet III no. 2115) 232 n84 Kitab Suwar al-Kawākib al-Thābita (TKS 4293) 479 n7 Miʿraj Nama (H.2154) 116, 162–163, 329 n7 Mukhtār al-ḥikam wa maḥāsin al-kalim of Abū ʾl-Wafāʾ alMubashshir ibn Fātik(Ahmet III 3206) 361, 401 Figs. 9, 10, 402 Fig. 11, 481 n15, 500 n62 poems of ʿAttār (AIII 305g), 461 Shahnama 83 n10 Istanbul, Topkapi Saray, Museum Bestiary (Hazine 1519) 289 n143

530 Shahnama (Hazine 1494) 84 n13 Shahnama (Hazine 1504) 84 n13 Istanbul, Türk ve Islam Eserli Müzesi 462 bookbinding (TVIEM ms. 1537) 463 Chahar Maqala (no. 1954) 334 n28 Mathnavīs (TVIEM, no.1905) 463 Istanbul University Library Gharāʾib al-sighār (T5669) 60 n40 Kalila wa Dimna (F.1422) 286 Fig. 44, 287, 288 Italian painting 52 n4, 163, 169 Ivanov, Anatoly A.106 n52, 108 Ivories Echmiadzin. Diptych 489 n33 Egyptian 380 n17 Murano Diptych 489 n33 Spanish 359 ivory 229, 237, 403, 404, 483 Jackson, Peter 92, 93, 108 Jacob, G. 233 n87, 387 n45, 416 jāhilliya poetry, 360 Jāḥiẓ, al- Kitab al-Hayawān 211–234, 232 n84, 236, 238, 243–247 Figs.1–5, 249 Fig. 7, 251–272 Figs. 9–30, 277 Fig. 35, 361 Jahn, K. 236 n98 Jalayirid era 10 n47, 342 n52, 392 painters 3, 13 Jam 69 Jamali 55 James, David Lewis 1 n.1, 119 n36, 151 n8, 211 n1.2, 213 n7, 227 n67, 292 n9, 297, 299, 300, 300 n21, 301, 303 n25, 305, 306, 410 n95, 466 n3, n4, n5, n6,

481 n16, n17, 504 James, M. R. 213 n9, 240 n114 Jami 55 Khamsa 60 n40 Jamiʿ al-Tawarikh 2 n7, 3 n10,3, 15, 15 n63, 19 n84, 21 n91, 81, 90, 112, 114, 121, 125, 127 n70, 129–131 Figs. 1–4, 134–138, Figs. 7–11, 142 Fig. 15, 144–145 Figs.17–19, 148, 151, 156 n21, 157–162, 166 n42, 173 Fig. 3, 180–182 Figs.13–16, 207–210, 217, 217 n21, 218, 218 n28, n30, 236 n98, 242, 242 n127, n128, 360, 392, 407 n89 Timurid version 157 n23, 158 n27 Jamshid, King 18, 46 Fig. 18, 89 n21, 93 n33, 97 n44, 429, 433 Jayyāb, Ibn 359 Jazarī, Ibn al-Razzāz al- 211 n1.1, 361 Jehovah 115 n21 Jenkins, M. 248 Jerphanion, G. de 227 n66 Jerusalem Dome of the Rock 116 n21155, 297 Entry into Jerusalem 169 Temple of Solomon 119 n38155 Jesus Christ 81, 114, 120, 121, 126, 156 jewellery 375 Jews 114, 120 n42, 124 n55, 155, 156,158, 160, 235 Joël, M. 502 John the Baptist 120, 156 Johnson, Elizabeth J. 481 n12, 504 Jones, D. 314 n28 Joshua 407 n89 Jubayr, Ibn 491 n40 Judaeo-Christian imagery 160

INDEX

Judaic imagery 155, 160 Judaism 155 Judas 226 Juki, Muhammad, Shahnama 4 n16, 5 n24, 4–5, 7, 7 n33, 9 n43, 20, 20 n87, n88, 21 n90, 23 n102, 27 Fig. 1, 29 Fig. 3, 56, 85 n16, 343 n52 Jurjān ceramics311 Jurjays 181 Fig. 16 Juvainī/ Juwaynī, ʿAlā al-Din ʿAṭā Malik Tarikh-i jahan-gusha 6 n28, 108, 217, 379, 380, 380 n19, 395 Fig. 1, 398 Fig. 4, 413, 439 n80 juz 294 Kaʿba 6 n27, 85, 87, 87 n18, 96, 437, 438 Kabulis 200 Fig. 35 kaf 301 Kahle, P. 233 n87, 387 n45, 416 Kai Khusrau 67, 72 Fig. 3, 93 n33, 336, 357 n61, 374 Kalila wa Dimna 3, 10, 10 n47, 11 n48, 13, 13 n56, 15, 16, 16 n69, 18 n76, 26 n109,34 n27, 34 Figs. 8a, 8b, 40 Fig.13a, 41 Fig. 13b, 149, 152, 152 n9, 211 n1.5, n1.6, 213, 220, 220 n40, 224, 225, 231, 236 n98, 287, 288, 288 n135, n139, 331 n18, 334 n27, 358, 361, 362, 363, 383, 383 n32, 400 Fig. 8, 482 n16 Kalila wa Dimna codices 1 n.2, 10 n47 Kalus, L. 297 n18 Kamāl 91 n25 kaman 127 Kappler, C. 435 n64 Karachi 291

531 Karamağaralı, B. 217 n21 Karnamak-i Artakshatr-i Papakan 93 n35 Kāshān 318 Fig 6, 394, 394 n71, 492 n44 Kāshānā 394 n74 Kashani, Qazi Muhammad 61 n48 Kashf al-Asrār 232 n83 Kashmir 423 Kay Kaʾus 81 n5, 374 khalīfat al-khulafā 61 n47 Kennedy, E. S. 112 n11 Kennedy, Hugh 499 n59, 504 Kerner, Jaclynne J. 379 n8 Keran, Queen Gospels 388 n50 Kervokian, H. 218 n27 Kessler, H. L. 235 n93, 425 n18 Khalil, Amir 328–329 n7 Khamis, Ulrike al- 374 Khamsa 8 n40 Khamsa of Jami 60 n40 Khamsa of Khusrau 6 n27, 60 n40 Khamsa of Navaʾi 60 n40 Khamsa of Nizami 4 n19, 6 n26, n29, n30, 7 n32, n34, 11, 11 n49, n50, 15, 16 n68, 71, 17, 17 n72, 18 n78, 19 n82, 23 n97, 25 n107, 26, 26 n110, 35 Fig. 9, 42 Fig. 14, 43 Fig. 15, 44 Fig. 16, 48 Fig. 20, 49 n112, n113, 54 n15, 57, 58, 58 n32, n33, 59 n39, 64, 69 n69, 86 n17, 330 n12, 332, 333 n25, 426 n20 Khatibi, A. 292 n5 khatibs 118, 479 n10 khatima 124 Khaqan of Chin 67 Khara Koto, Mongolia 457 Khavarnaq 8 n40 19 n82 Khizr 437, 438 Khubilai Khan 150

532 Khurasan 64, 69, 149 Khusrau 5 n25, 6 n27, n29, 15, 19 n83 23 n96, n97, 55, 81, 87 n18, 42 Fig. 14, 60 n40, 330 Khwaja Ata 328 n7 Khwaja Ghiyath al-Din 328 n7 Khwaju Kirmani Kulliyat 5 n24, 7 n36, 11 n49, 13, 49 n112, 330 n13, 55, 342 n52, 344 n56, 345, 346, 358 Mathnavīs 6 n26, n30,19 n79, 28 Fig. 2, 242 Khwārazmshāh Tekesh 91 n25 King, G. R. D. 300 n22 King, Noel 490 n38, 504 Kirman, Iran 459 kitābkhāna 63, 63 n57, 308, 328, 331 Kitāb ʿAjāʾib al-makhlūqāt wa gharaʾib al-mawjudat 122 n48, 154,223 n54, 203 Fig. 38, 223, 239 n107, 361 Kitāb al-aghānī 128 n73, 215 n15, 325 n32, 410, 411, 499 n60 Kitab al-Athar al-Baqiya ʿan al-qurun al-khaliya 111–148, 111, n3, 116 n22, 132–133 Figs. 5, 6, 139–143 Figs. 12–16, 154, 156, 179 Figs. 11, 12, 201 Fig. 36, 218 n28, 407 n89, 410 n98 Kitab al-Baitara 223 232 n84 Kitab al-Baytāra of Ibn Akhī Hizām 361 Kitab al-Baytāra of Ahmad ibn alHusayn ibn al-Ahnaf 361 Kitab al-Hayawān of al-Jāḥiẓ 211– 234, 215 n14, 216 n16, 220 n37, n40, 223 n54, 236, 237, 237 n101, 239 n112, 240, 243–247 Figs.1–5, 249 Fig. 7, 251–272 Figs. 9–30, 277 Fig. 35 relationship of text and illustra-

tions 214–219 Kitab al-Makhzūn Jāmiʿ al-Funūn 213 n6 Kitab al-Irshad 125 n61, 146 n75– n81, 147 n82, n83 Kitab al-Tiryāq 223, 223 n50, n51, 225, 325 n32, 378 n8, 413, 481 n15, 482 n16, 496 n55 see also: Theriaca Kitab fi maʿrifat al-ḥiyal al-handasiyya 211 n1.1 Kitab qawāʾid al-aḥkam fī maʿrifat al-ḥalāl waʾl-ḥarām taṣnīf 412 n104 Kitab suwar al-kawākib al-thābita 21 n92, 223, 361, 479 n7, 481 n15, 482 n16 Kleiss, W. 10 n46, 287 n132, 330 n11 koine 307 Komaroff, Linda 149 n1, 150 n3, 154 n17, 382 n26, n27, 386 n43, 416 Köprülü, M. F. 432 n50, 433 n51 Kratzert, Kacie 161 n34 Kritovoulos, 105, n51, 108 Kufa 377 n4 Kūfic script 236 n99, 238, 291, 294, 295, 296, 297, 299, 301, 304, 310, 312, 312 n15, 403, 429, 455, 456, 458 Kugler, Bernard 95 n38 Kühnel, E. 230 n75, 235 n92, n95, 288 n139, 394 n71, 416 Kulliyat of Khwaju Kirmani 5 n24, 7 n36, 11 n49, 13, 49 n112, 55, 330 n13, 342 n52, 344 n56, 345, 346, 358 Kung Kʾai 288 n138 kunya 378 Kurtz, Otto 223 n53, 481 n14 Kuthayyir 494 n48

INDEX

Kutubi, Ibn al-, 111, 179 Figs. 11, 12, 201 Fig. 36 Kuwait, Sulwan al-Mutʿa (private collection) 211 n1.7, 215 n14, 220 Kuwait City, Kuwait National Museum Muʾnis al-ahrar fi daqaʾiq al-ashʿar (LNS 9MS) 203 Fig 39 Laila 5 n25, 16, 18 n78, 26, 48 Fig. 20 lajvardina 150 laqab 378 lacquer 168, 169, 375, 463, 464 Lambourn, Elizabeth 379 n8 Lambton, Ann K. S. 108 Lamentation 169 Lamm, C. J. 212, 212 n2, n4 landscape 11, 15, 113, 124, 125 n59, 127, 147 n85, 153, 154 n17, 157, 159, 164, 169, 177 Fig. 9, 209, 226, 228, 229, 233, 239, 241, 242, 287–289, 287 n131, 288 n139, 335, 340, 342, 342 n48, 362, 375, 379, 380, 461 Lane, A. 221 n43, 314 n28 Languschi, Jacopo 105 n50 Lawrence, B. B. 113 n13, 114 n16 leatherwork 151, 158, 344, 454, 455, 457–464 passim. Leiser, G. 432 n50 Leningrad see: St Petersburg Lentz, Thomas W. 5 n22, n24, 6 n29, n30, n31, 7 n32, n34, n35, n37, 8 n40, 13 n57, 14 n59, 15 n64, 16 n67, 17 n72, 19 n79, n81, 20 n86, n89, 23 n97, n102, 26 n110, 98 n46, 109, 328 n5, 330 n11, n13, n14, 331 n16, n17, n19, 334 n27, 342 n48, 343 n52, 380 n18, 416

533 Leonardo da Vinci 52 n4 Levant 360 Lewis, Bernard 2 n. 4, 5, 108 Life of the Prophet see: Sira Lings, M. 292 n6, 293 n11, 311 n14, 316 lions 26 n109, 120 n43, 227, 233, 240 n115, 260 Fig. 18, 288 n135, 359, 374, 462 Lisbon, Gulbenkian Museum. Anthology (L.A.161) 9 n42, 14–15, 19 n83, 38 Fig.12a, 39 Fig. 12b, 330 n12 literature 92, 427, 429, 439, 449, 499 Arabic 359–364, 427 n25 Greek 214 Persian 358, 427 n25 Lively, Penelope 437 n75 Liverpool 374 Lodge, I. E. 242 n129 Löfgren, Oscar 212, 212 n2, n3, n5, 214 n10, n13, 215 n14, 216 n18, 219 n34, 226 n63, 228 n70, 71, 229 n72, n73, n74, 230 n77, n78, n79, 232 n80, n81, n82, n84, 233 n85, n86, n88, n89, 234 n90, 237 n101, 239 n112, London, British Library Add.18188 84 n13 Add. 27258 84 n13 Anthology (ms. Add.16561) 4 n16 Anthology of Sultan Iskandar (ms. Add.27261) 7 n34, 9 n43, 20 n87, 330 n12, 336 n33, 342 n52 Anthology (of 813/1410–11) 9 n43 Anthology of Divans (ms 132) 175 Figs. 5, 6 Bestiary (Royal Ms. 12C XIX)

534 240 n115 Epics (Or. 2780) 57 n26, 118 n33, 367 Harley MS. 4751 278 Fig. 36 Khamsa of Nizami (Add.25900) 16, 16 n67, 43 Fig. 15, 49 n113 Khamsa of Nizami (Or. 6810) 6 n29, 7 n32, 11, 11 n50, n71, 17, 17 n72, 19 n82, 23 n97, 25 n107, 26, 26 n110, 35 Fig. 9, 44 Fig. 16, 48 Fig. 20, 49 n112, 58 n32, n34, 332, 357 Khamsa of Nizami (Or. 2265) 54 n15, 69 n69 Khamsa of Nizami (Or. 13297) 86 n17 Khwaja Dahhani 432 Khwaju Kirmani ms. (fol.26b) 7 n36, 11 n49, 49 n112, Kitāb ʿajaʾib al-makhluqat wa gharaʾib al-mawjudat (Or. 14140) 203 Fig. 38, 223 n54 Kulliyat of Khwaju Kirmani (Add18113) 5 n24, 49 n112, 330 n13, 345, 346, 358 Maqāmāt (Or. 9718) 211 n1.4 220 Nihāyat al-suʾl waʾl-umniyya fī taʿallum aʿmāl al-furūsiyya (Add. 18866) 211 n1.2, 361 Or 2833 462 Qurʾans 293 Shahnama (Or. 1403) 84, 86 n17 London, British Museum, Great Mongol Shahnama (1948.12– 11.025) 199 Fig. 34 London, India Office Library, Gulistan 329 n10 London. Keir Collection 292 n4, 311 n12 Great Mongol Shahnama (PP3)

192 Fig. 27 London. Nasser D. Khalili Collection of Islamic Art Jamiʿ al-Tawarikh (MSS 727) 157, 173 Fig. 3, 182 Fig. 15, 208, 360 Naʿt al-ḥayawān 223, 386 Qurʾans 294–306 Vol. I The Abbasid Tradition 291, 294–299, 306 cat. 11 298 cat. 24 294 cat. 41 294 cat. 50 298 cat. 65 298 cat. 66 294, 298 cat. 68 296 cat. 70 298 cat. 78 New Style Qurʾan 294 cat.81 Qurʾan from Palermo 294 cat 82 298 cat. 83 Buyid Qurʾan 294 cat. 84 294 cat. 93 New Style Qurʾans from India, fragments 294, 298 cat. 95 New Style Qurʾans from India, fragments 294 cat. 96 298 Vol. II The Master Scribes 299– 304, 306 cats.1, 6, 17, 18 Qurʾans from India 294 cat. 3 Ayyubid Qurʾan 301 cat. 5 301 cat. 7 302 cat. 11 304 cat. 18 304 cat. 20 304 cat. 25 304 cat. 29, 36 294 cat. 29–32 300 cat.40 302

INDEX



cat. 43 300 Vol. III After Timur 304–306 cat. 1 Qurʾan of ʿAbd alQayyum ibn Muhammad ibn Karamshah-i Tabrizi 294 cat. 2 303 cat. 24 304 cat. 48 304 cat. 81 304 London, Royal Asiatic Society, Shahnama of Muhammad Juki (Morley 239) 4, 4 n16, 5, 7, 9 n43, 27 Fig. 1, 29 Fig. 3, 85 n15, 343 n52 London, Victoria and Albert Museum bronze bowl 373 London, Warburg Institute, University of London Hillenbrand, R. Conference paper.112 n8 London, Westminster Abbey Chapter House Library bestiary (ms. 22) 238 n104 Loqman 77 n78 L’Orange, H. P. 405 n86, 416 Lorey, Eustache de 153 n15, 220 n38 lotus 462 Loveday, Helen150 n6 Lowry, Glenn D. 5 n24, 6 n29, n30, n31, , 7 n32, n34, n35, n37, 8 n40 13 n57, 14 n59, 15 n64, 16 n67, 17 n72, 19 n79, n81, 20 n86, n89, 23 n97, n102, 26 n110, 57 n26, 98 n46, 100 Fig. 2, 101 Fig. 3, 109, 328 n5, 330 n11, n13, n14, 331 n19, 334 n27, 343 n52, 380 n18, 416 Luhrasp 4 n18, 25 n108 Luʾluʾ, Badr al-Din 128 n73 Luqmān 383 n28 lustre ware 221 n41, n42, 222 n45,

535 309, 310,314, 311 n12, 317 Fig. 5, 319 Fig. 8, 320 Fig. 9, 322 Figs. 11, 12, 323 Fig. 14, 324 Figs. 15, 16, 325 n29, n32, 394 n71 Lyons. brass bowl 386 n43 McCulloch, F. 213 n9, 240 n116 Macedon 79, 94, 98 Madelung, W. 430 n33 Madinat al-Salam 377 madrasa 205, 359, 394 n75, n76 Madrasa al-Mustanṣiriyya 394 n75, n76 Maghrib 297 n18, 361 Maghribî bookbinding 459 Kufic script 301 Mağnūn 87 Magog see: Gog and Magog Magogli 95 n38 Mahabharata 433 Mahboubian, Mehdi 384, 384 n34, 416 Mahdavi, Amin 375, 439 n81 Mahmud of Ghazna 91 n25, 180 Fig. 13 Mahmud Yalawač 94 n37 Maisūn 256 Fig. 14 Maitra, K. M. 20 n85 Majnun 6 n25, n26, n27, 18 n78, 48 Fig. 20, 49, 50 n113, 334 Makīn, al- 79 n2 Mālikī 361 Mamluks 95, 106 n52, 218, 410 bestiaries 211–234 bookbinding 458 painting 13 n55, 212 n1.8, 212 n4, 215 n14, 219–221, 224– 289, 224 n56, 237 n101, 288 n135 Qurʾans 293, 300

536 manuscript 12, 228 n69 metalwork 359, 360 textiles 359 Mamnoun, P. 423 n13 Mamshadh, Ibrahim ibn 428, 429 Manafiʿ-i hayavan of Ibn Bukhtīshūʿ 12, 19 n84, 153–154, 178 Fig. 10, , 213–216, 217 n19, 234, 235–242, 237 n101, 287–289, 273–276 Figs. 31–34, 279–284 Figs. 37–42, 459 Manafiʿ-i hayavan (Escorial Bestiary) 212 n1.8, 220 n38, 224, 225, 232 n83,244 Fig. Manchester, John Rylands Library Khamsa of Nizami (Ryl. Pers. 36) 18 n78 Shahnama (Ryl. Pers.933) 84 n13 Mani 328 n7 Manichaean bookbinding 457 Manija 66 n65 Mannheim, Ralph 106 Manṭiḳī, Abū Sulaymān al- 378 n5 Mantiq al-Tayr 17–18, 49 n112, 45 Fig. 17 Manuchihr, Shah 195 Fig. 20 Maqamat of al-Hariri 1 n1, 3 n8, n9, 12, 12 n52, 118, 119, 211 n1.3, n1.4, 213 n8, 215 n14, 219, 220, 222, 224, 231, 232, 233 314, 362, 363, 382 n32, 387 n47, 388 n51, 387 n47, 388 n51, 403 n80, 405, 407, 413 n106, 422 n5, n6, 465–505 passim Maqdisī, al- 378, 411 Maqqari, al- 420 n2 Maragha, Iran 112, 153, 178 Fig. 10, 217 n19, 235, 460 marble 24, 338 Marchal, Henri 483 n22, 491 n39,

504 mare 241 margins 2, 9, 12–18, 26, 166, 166 n42, 230, 295, 297, 300, 304, 311, 344, 366, 372, 375, 405, 462, 482, 488 Marinid dynasty 359 Mar Mattai 384 n36 Marquet, Y. 377 n3, 378 n6, 416 Marrakesh 459 Martin F. R. 57 n31, 219 n35, 235 n92, 241 n121 Marzubānnāma 217, 218 n31, 386, 412 Masaccio 20 Mashhad 394 n71 masâhif 456 mashq 298, 301 Masʿud Ahmad 51 n2, 53 Masʿūdi, al- 93, 93 n34, 109 Masuya, Tomoko 165 n40, 422 n5 Mathnavīs 6 n26, n30, 19 n79, 28 Fig. 2, 242, 463 maula 146 Maulana Jaʿfar Tabrizi 328, 328 n5 Maulana Shihab 328 n7 Mausilī, Ibn al-Duraihim alManafiʿ-i hayavan 212 n1.8, 220 n38, 224, 225, 244 Fig. 2 Mawlana ʿAbdullah Sayrafi 149 Mayer, L. A. 118 n34, 225 n59 Mazandaran 56 n24, 67, pottery 325 n29 Mecca 7, 437, 438 medicine 361 Mehmed II 105, 105 n50, n51 Mehran, Farhad 370, 371, 435 n62, 449 n85 Meinecke, M. 289 n142 Meisami, Julie Scott 170 n52, 434 n59 Melikian-Chirvani, Assadullah S.12

INDEX

n55, 21 n92,83 n9, 90, 90 n24, 91, n25, n 26, n27, 28, n29, 292, 94 n36, 96 n42, 97, 98, 106 n52, 109, 164 n38, 212 n1.7, 22 n45, 289 n142, 307 n1, 309 n5, n6, n7, 310 n10 313 n21, n23, 325 n30, n31, 327 n1, 379 n13, 394 n72, 408 n90, 416, 417, 492 n44, 503, 507 Melville, C. 114 n19, 150 n5, 151 n8, 155 n19 Menzel, T. 117 n28 Membré, Michele 69 n70 menologion 390 merchants 240, 407 Meredith-Owens, Glyn M. 20 n87 Mesopotamia influence 118, 128, 235, 239 n110, 241, 288, 289, 387 models for Kitab al- Hayawān of al-Jāḥiẓ 221–234, 224 n56 northern 382 n25 metalwork 91, 149, 150, 229, 289, 307, 310, 313, 314, 359, 360, 366, 373, 374, 386. 387, 413, 425 n18, 482, 482 in bookbindings 458 Michell, G. 314 n28 Michelangelo 52 n4 Michigan database (of Shahnama illustrations) 422, 438 Mihrab 68 miḥrāb 394 n71 Mihran Sitad198 Fig. 33 Milan. Biblioteca Ambrosiana Kitab al-Hayawān (MS.140Inf. S.P.67) 211–234, 23 n54, 243–247 Figs.1–5, 249 Fig. 7, 251–272 Figs. 9–30, 277 Fig. 35 Miles, George Carpenter 235 n93, 488 n29, n30, 504 military exercises 361

537 Milstein, Rachel 17 n75 mīnāʾī ceramics 309, 310, 311, 311 n12, 313, 314, 315 Fig. 2, 319 Fig. 7, 323 Fig. 13, 325, 326 minbar 117, 118, 118 n31, n32 Minorsky, V. 52 n4, 61 n44, n46, n49, 62 n50, 291 n1 Minovi, Mujtaba 24 n103, 292 n2, 327 n3 Minu Dizh, temple of 89 n21 Minuchihr 374 Mi´raj Nama 116, 149, 162 n37, 162–163 MirʿAli Shir Navaʾi, Divan 59 Mirḫwānd 98 Mir Musavvir 53 Mirror for Princes 448 Modi, J. J. 437 n73 Möngke Qaʾan 94 n37 Mohiuddin, A. 223 n48 Mongol(s) 6 n28, 95 n38, 365, 380, 380 n21 bookbinding 457 cavalry 242 conversion to Islam 150, 151, 155 invasion of Iran 149, 393, 439 political dominance 158, 161, 165, 167, 170 religious attitude 96 n40, n42, 114, 148, 235 influence 60, 207–209, 289 in Kitab al-Athar al-Baqiya 112– 114 in Iskandar cycle Mongol influence 60, 79–110, 94 n37 see also: Shahnama, Great Mongol monkey(s) 239 n111, 286 Fig 44, 462 Monza Cathedral, Murano Diptych 489 n33

538 Moreh, S. 387 n45, 417 Morgan Bestiary see: Manafiʿ-i hayavan of Ibn Bukhtīshūʿ Morgan, David O. 93, 96 n40, 109, 120 n41 Morocco 291 Morton, A. H. 69 n70, 157 n23, 422 n5 mosaics, Byzantine 410 Moses 80, 87, 127, 182 Fig. 15, 407 n89 mosques 117, 118, 150, 157, 205, 291, 298, 393 see also: Istanbul. Süleymaniye Mosque Library Mosul, Iraq 111 n6, 203 Fig. 38, 157, 227 Mosul, King of 181 Fig. 16 Mottahedeh, R. 427 n24 moufflon 236, 274 Fig. 32 mountain goat 239 mountains 49, 154, 159, 173 Fig. 3, 186 Fig. 21, 209, 233, 240 Mount Athos, Stauronikita (ms.43) 404 n83 Mount Sinai 127 mourning 44 Fig. 16, 167, 168 Muʿāwiya 256 Fig. 14 mubahala 125 n61, 126 mudras 155, 169 Mufid, al-Shaikh 125 n61, 126, 126 n62, n64, n65, 127 n67128, 129 Mughals 51 n2, 62, 105,304, 367, 420 illumination 305 Muğmal al- Tawārīh 89 n21 Muhammad ibn Badr al-Din Jajarmi, Muʾnis al-ahrar fi daqaʾiq al-ashʿar 203 Fig 39 Muhammad, the Prophet 19 n84, 80, 81, 115 n21, 118 n33, 125 n61, 126 n62, 162, 378 n5



images of in: Kitab al-Athar al-Baqiya 111–148, 116 n22, 132–133 Figs. 5, 6, 139–143 Figs. 12–19, 146 n74, 154, 155, 156, 407 n89 ʿAli at Ghadir Khumm 128, 141 Fig. 14, 146–148 Day of cursing 125–128, 140 Fig 13,201 Fig. 36 forbids intercalation 115, 117, 132 Fig. 5, 179 Fig. 11 Watchman and two riders 115, 133 Fig. 6 images of in: Jamiʿ al-Tawarikh 159, 160, 162 and Abu Bakr 136 Fig. 9 birth of 131 Fig. 4 and lifting the Black Stone 134 Fig. 7 night journey on Buraq 137 Fig. 10 recognised by Bahira 156 n21, 138 Fig. 11 Muhammad, Sultan 53 n14, 464 Muhammadi 464 muhaqqaq 299, 301 Muʿizzi 217 mujahidin 302 Mukhtār al-ḥikam wa maḥāsin al-kalim of Abū ʾl-Wafāʾ alMubashshir ibn Fātik 361, 378 n8, 384, 385, 401 Figs. 9, 10, 402 Fig. 11, 481 n15, 482 n16, 500 n62 Müller, P J. 2 n. 3 Mullett, M. 391 n60, 417 munabbat kârî 461 Munich. Bayerische Staatsbibliothek Kalila wa Dimna (Cod. Arab. 616) 211 n1.6, 224 Kitab ʿAjaʾib al-Makhluqat (Cod. Arab.464) 122 n48, 223

INDEX

n54, 239 n107 Muʾnis al-ahrar fi daqaʾiq al-ashʿar 152, 203 Fig 39, 217, 422 n5, 460 Mundhir, al-Nuʿmān ibn 360 Munro, E. 52 n6 Muqaddasi, al- 544 Murano diptych 489 n33 Gospels 489 muraqqaʿ 305, 312, 332 n20 Muraqqaʿ-yi Bahram Mirza 149 n1 Mūsā, Ahmad 287 Musailima 119, 121, 122, 123, 127, 139 Fig. 12, 147, 147 n84 Musavvir, Mir-I 51 n2 Mustanṣir, al- 391 Muzaffarid painting 333 n24 Mshatta 488 n30 mystics and mysticism 92, 328 Mzik, H. von 438 n77 Nahrajūrī, Abū Aḥmad al- 378 Nâʾîn 460 Najaf, Jāmiʿ Zīr Dālān 394 n71 Najran 125, 459 Najranis 125–128, 125 n60, n61, 126 n62, n63, 140 Fig. 13 156 Nallino, C. A. 433 n56, 434 n58 Namakdān 392 Naples 459 naqqâsh 464 naqqali 423 Naqsh-i Rustam 433 Nāṣir, al- 391 Naṣīr al-Dīn Muhammad walad al-janāb al-ʿālī al-marhūm Ḥusām al-Dīn Taranṭāʾī 213 n8 Naṣīr al-Dīn Ṭūsī 92 n30 naskh 299, 413, 459 Nasmus 93 n32 Nasrid dynasty 359

539 nastaʿliq script 328 Nastihan 73 Fig. 4 Nasr Allah Abuʾl-Maʿali, Kalila wa Dimna 10, 26 n109, 34 Fig. 8a, 8b, 40 Fig 13a, 41 Fig 13b Nasuri family 374, 375 Naʿt al-ḥayawān 223, 386 Nebuchadnezzar 119 n38, 120 n42, 155 Nees, L. 417 Negus of Abyssinia 135 Fig. 8 Nemanzee, S. 57 n26 Nestorians 156 New Style Qurʾans 294, 295, 296, 298 New Testament 208, 235, 385 New York, Metropolitan Museum of Art. Fātimid lustre bowl 248 Fig. 6 Kervorkian Foundation 218 n27, 460 Mantiq al-Tayr (63.210.35)17– 18, 45 Fig. 17, 49 n112 Khamsa of Nizami (no. 13.228.3) 6 n25, 26 n110, 59, 59 n39 Shahnama (13.228.22) 84 n13 First small Shahnama (34.24.1)183 Fig. 17 New York, Pierpont Morgan Library Manafiʿ-i hayavan (ms. M.500) 19 n84, 153, 178 Fig. 10, 216 n16, 217 n19, 234, 235–242, 287–289, 273–276 Figs. 31–34, 279–284 Figs. 37–42, 459 New York, Public Library Sarre Qazwīnī 223 n54 Nicander, Theriaca 223, 223 n50, n51 225 Night 492 night scenes 22, 73 Fig. 4, 137 Fig. 10, 348 Fig. 2

540 Nihāyat al-suʾl waʾl-umniyya fī taʿallum aʿmāl al-furūsiyya of Muhammad al-Aqsarāʿī 211 n1.2, 361 Nimrud, King 14, 38 Fig. 12a, 39 Fig. 12b Nishapur ceramics 325 n29 Nishapuri, Maulana Nizam al-Din Shah Mahmud Zarin-qalam 51 n4 Niẓāmī 83, n11, 96 n39, 329, 435, 464 Khamsa 4 n19, 6 n26, n29, n30, 7 n32, n34, 11, 11n49, n50, 15, 16, 16 n68, n71, 17, 17 n72, 18 n78, 19 n82, 23 n97, 25 n107, 26 26, n110, 49 n112, n113, 54 n15, 58 n32, n33, 35 Fig. 9, 42 Fig. 14, 43 Fig. 15, 44 Fig. 16 48 Fig. 20, 57, 58, 59 n39,64, 69 n69, 80, 81, 87 n18, 332, 357, 426 n20 Ḫusraw wa-Šīrīn 5 n25, 86, 23 n96 Iskandarnāma 81 Nöldeke, Theodor 79, 79 n2, 81 n6, 87, 89, 89 n21, 98 n47, 109 nomads 426 Norgren, Jill 83 n11, 86 n17, 87 n18, 109, 422 n7 North African Qurʾan 294 Nurani 98 Nurbakhsh, J.127 n71, 411 n99, 417 Nushirvan the Just 68, 68 n68, n69, 75 Fig. 6 Nuṣrat Shāh 106 n53 Oh, L. J. 411 n102, 417 oil painting 375 Oinas, Felix J. 434 n61 O’Kane, Bernard 17 n75, 150 n7, 152 n9, 383 n32, 417, 482 n16,

504, 506 Old Testament, imagery 159, 160, 16, 208, 235, 385 Olympias 98, 106, 450 Omont, H. 223 n51, 494 n54, 504 Onagers 189 Fig. 24 On the Usefulness of Animals see: Manafiʿ-i hayavan Ott, C. 437 n74 Ottoman architecture 305 bookbinding 458, 464 court 433 illumination 305 Qurʾans 305, 464 Ottomans 53 n7, 54 , 54, n15, n16, 62, 69, 77, 81, 105, 111, 147 n84, 289 n143, 304 overlaps/overlapping 3, 14, 19, 21, 156, 163, 229, 238, 239, 307, 357, 386, 387 Oxford, Bodleian Library. Kalila wa Dimna (ms Pococke 400) 211 n1.5, 220, 220 n40, 224 Kitab Suwar al-Kawākib al-Thābita (Marsh 144) 21 n92, 223, 361 Dioscorides ms (Cod. or. d. 138) 383, 383 n31 Maqāmāt (ms Marsh 458) 213 n8, 215 n14, 220, 413 n106 Samak-i ʿayyar (ms. Ouseley 379) 176 Fig. 7 Samak-i ʿayyar (ms. Ouseley 380) 176 Fig. 8 Shahnama (ms Ouseley Add.176) 9 n43, 25 n106, 85 n16 Oxford, Lincoln College, Typikon (Gr. Ms. 35) 496 n54

INDEX

pādīšah-i islām 93 n32 Page, M. E. 423 n11 Pahlavi 79 Palermo 294 Pamirs 428 Panchatantra 236 n98 paper/ papermaking10, 17, 150, 151, 158, 295, 300, 301, 303, 304, 306, 308, 310, 344, 366, 367, 368, 456, 457, 459, 461, 462 parchment 238, 295, 454, 456 Parham, Cyrus 479 n8, 504 Paris 302, 425 Paris, Bibliothèque Nationale de France Divan of MirʿAli Shir Navaʾi 59 Homilies of St. John Chrysostom 496 n54 Kalila wa Dimna of Ibn al-Muqaffaʿ(arabe 3465) 220, 224, 383, 383 n32 400 Fig. 8 Kitab al-Athar al-Baqiya (ms arabe 1489)111–148, 111 n3, 132–133 Figs. 5, 6, 139–143 Figs. 12–19, 147 n84 Kitab al-Makhzūn Jāmiʿ al-Funūn (arabe 2824) 213 n6 Kitab al-Diryāq 479 n8, 481 n15, 482 n16, 496 n55 Maqamat of al-Hariri, (Schefer ms Arabe 5847) 3 n8, 12, 12 n52, 21 n92,118, 119, 219, 220, 222, 234, 314, 363, 387, 388, 387 n47, 388 n51, 405, 465–507 passim Maqamat of al-Hariri (ms Arabe 6094) 383 n32, 481 n11 Psalter (ms grec 139) 492 n41 Tarikh-i jahan-gusha (ms or., Suppl. persan 205)202 Fig. 37, 378 n8, 379, 395 Fig. 1, 398 Fig. 4, 481 n15

541 Theriaca (ms arabe 2964) 223, 223 n50, n51, 225, 413 Theriaca (ms suppl. grec. 247) 226 n61 Paris, Musée des Arts Decoratifs Mathnavis of Khwaju Kirmanī (Inv. 3727)19 n79 Paris, Musée du Louvre Great Mongol Shahnama (7095) 200 Fig. 35 Parsees 89 n21 Parthians 428 patronage 91, 149, 151, 166, 167, 207, 208, 213, 219, 220, 309, 376, 391, 423, 425, 429, 460 Payne, A. 238 n104, 240 n115 Peacock, A. C. S. 429 n29 Pedersen, J. 382 n26, 417, 421 n3 Pellat, C. 214 n12, 216 n17, 241 n120 Perkins, Ann 490 n36, 504 Persepolis 430, 433 Persian bookbinding 455, 460 culture 57, 208, 366, 383, 411, 420, 433 language 81, 332, 427, 428 n25, 429, 430, 432 manuscripts 347, 358 painting 17 n74, 207–210, 325, 327–358 passim, 366, 371, 372, 380, 392, 404, 408, 413, 420– 451 and Western scholarship 371, 376 pre-Timurid 1, 2, 13, 20, 22, 26, 53, 55, 57, 62, 63, 83 poetry 81, 92, 311, 314, 317 Fig. 5, 360, 367, 427 n25, 435 Perry, J. R. 62 n51 perspective 8, 26, 484

542 Peters, F. E. 113 n13 Pétrosyan, Yuri A. 504 Pfister, Friedrich 79, n1, 89 n21, 109 pharmacology 361 Philadelphia 460 Philistines 156 philosophy 152, 158 phoenix 238 Phrantzes, 105 n50, 106 n51 Physiologus 213, 213 n9, 240 n114 Piemontese, A. M. 424 n14 pilgrims/pilgrimage 7, 12, 115, 117, 146 Pilsam 67 Pinder-Wilson, Ralph 49 n113, 295 n13 Pleiades 359 Pliny 238, 406 Plutarch 98 Poem of the Constellations see: Risālat al-Ṣūfī fiʾl-kawākib poetry 9, 55, 61, 64 n60, 64, 81, 82, 92, 152, 290, 311, 312, 314, 317 Fig. 5, 359–363, 367, 432, 435 poets 23 n101, 24, 32 Fig. 6, 55, 57, 81, 92, 175 Fig. 5, 328, 332–335, 338, 339, 341, 341 n44, 343, 348 Fig. 2, 351 Fig 6, 356 Fig. 13, 358–360, 368, 420, 432–434, 428, 429, 434, 436, 494 n48 Pollock, Sheldon 431 Poonawala, I. K. 117 n26, 122 n49 Pope, Arthur Upham 4 n18, 13 n56, 19 n80, 222 n45, 232 n84, 287 n134, 313 n25, 315, 323, 394 n71, n74 403 n79, 417, 492 n44, 504 portraits Arab 390 authors 335, 379–385, 381n22,

389, 390,412, 498–501 Evangelists 380, 382 n24, 384 n36, 390, 404, 404 n83 Renaissance conventions 388 rulers 85, 86, 93, 106, 380, 496 saints 404 see also: Iskandar, Muhammad Porus, King 95, 98, 436, 438, 449 pottery see: ceramics prayer 125, 126, 128, 304 pre-Islamic Iran 90, 96, 155, 159, 373, 430 Prentice, V. R. 329 n10, 335 n31, 338 n39, 342 n48, 380 n18, 417 prophets 80, 81, 85, 87, 114, 121, 159, 160, 385, 407, 432, 435, 437, 438 Pseudo-Callisthenes 79, 79 n2, 80, 109, 436 Ethiopic version 79 n1 Ptolemy 384 Pyrenees 428

Qadi Ahmad 291 n1 qadi 118, 497 n58 Qadisiyya 426 Qaidāfa, Queen of Andalusia 84 n14, 86, 97 Qairawan 118, 118 n32, 298 bookbinding 454, 455 Jâmiʿ 454 Qāʾit Bay, Sultan 213 n6 Qajar dynasty 366, 372, 375, 464 qalam 308 n2 Qalāʾūn, Muhammad b. 106 n52 Qara-Khanids 160, 164 n38 Qaramanids 433 Qarmatian Kufic script 294 Qazi Ahmad 51 n4, 62 n50 Qazvini, Mirza Muhammad 218 n27 Qazwīnī, Zakariyyāʾ al-, Kitab ʿAjāʾib al-makhlūqāt 122 n48,

543

INDEX

361 Qizilbash 68, 68 n67 Qum 394 n71 Quraysh 122 Qurʾan 54 n16, 80, 117, 150, 151, 155, 168, 208, 213, 311, 311 n14, 316 Fig. 4, 331, 366, 375, 378, 423, 424, 427, 482 n16 Anatolian 300 Ayyubid 301 Blue Qurʾan 297 n18, 298 of Bawwab, Ibn al- 296, 379 n9 of Baysunghur 293, 300 bindings 452, 455, 456, 457, 458, 459, 460, 464 Egyptian 294, 456 Qurʾan of Ibn al-Bawwāb 466 n6 Ilkhanid 151 n8 illumination 290–306 Indian 300 Inscriptions 359 Iranian 296, 301, 304, 305, 308, 460 Mamluk 293, 300 North African 294 Nurse’s Qurʾan 298 Ottoman 305, 364 Timurid 331 n17 Umayyad 379 n10, 389 Valencian 301 Western scholarship 290–306 Yemeni 300 see also: London. Nasser D. Khalili Collection of Islamic Art. Qurʾans 294–306 Qubādābād 91 n29 Quṭb al-Dīn Šīrāzī 92 n30 Quhrūd. Masjid-i ʿAli 394 n71, 492 n44 Masjid-i Kalah 394 n71, 492 n44

Qusayr ʿAmra 488 n30 Rabat 291, 302 Rabʿ-i Rashidi 157 Raby, Julian 236 n98, 306, 307 n1, 439 n82 raihan 299 ramal 359 Ramayana 420 Raphael 52 n4 Raqqa pottery 222, 229 Rasāʾil Ikhwan al-Ṣafa wa khullān al-wafāʾ2, 122 n48, 215 n15, 328 n7, 335 n31, 378 n5, 394 n75, 396 Fig. 2, 397 Fig. 3, 399 Fig. 5, 377–419 passim, 388 n52, 481 n15, 492 n45, 500 n62 Rasgat 296, 296 n15 Rashid, Harun al- 357 n62 Rashidiyya ms 12, 21, 161 Rashidun caliphs 297 Rašīd al-Dīn 82 n8, 94, n37, 109, Jamiʿ al-Tawarikh 2 n7, n8, 3 n10,3, 15, 15 n63, 19 n84, 21 n91, 81, 90, 112, 114, 121, 125, 127 n70, 129–131 Figs. 1–4, 134–138, Figs. 7–11, 142 Fig. 15, 144–145 Figs.17–19, 148, 151, 156 n21, 157–162, 205 n53, 173 Fig. 3, 180–182 Figs.13–16, 205, 207–210, 217, 218, 218 n28, n30, 236 n98, 242, 242 n127, 360, 393, 403 n79, 407 n89 Timurid version 157 n23, 158 n27 Rashid al-Khwafi 202 Fig. 37 rauda 157 reliefs 209, 433, 456, 462, 487 religions and religious themes 69, 91, 96, 97, 105, 113, 115, 116, 118, 119, 121, 128,148, 150, 151,

544 154,162, 168, 226, 235, 385, 390, 404, 407, 423, 424, 427, 430, 431, 436, 450, 466, 490 Renaissance, portraiture conventions 388 repoussoir 6, 485 Resurrection 156 Rice, David Storm 296, 296 n16, 364, 379 n9, 410 n94, 417, 466 n6, 481 n12, 482 n16, 487 n28, 505 Rice, David Talbot 2 n7, 3 n11, 12 19 n84, 21 n91, 54, 82, 109, 112 n7, 116 n22, 121 n44, n45, 127 n70, 156 n21, 207, 217 n21, 393 n66, 407 n89, 417, 496 n54, 505 Richard, F. 152 n13, 379 n14, 380 n20, 417 Riefstahl, Rudolf 235 n92 riqʿa 299, 413 Rifāʿa, Zayd b. 378 Riggs, Charles T. 108 Risālat al-Ṣūfī fiʾl-kawākib 383 Robb, D. M. 388 n52, 417 Robertson, E. 111 n5 Robinson, Basil W. 4 n16, 6 n25, n26, n30, 9 n43, 18 n77, n78, 21 n90, 23 n101, 24 n103, 54 n16, 56 n19, n21, n23, 57 n27, 85 n16, 104 Fig. 6, 109, 213 n6, 217 n20, 218 n26, n27, 228, n69, 292 n4, 327 n3, 329 n7, 329 n10, 342 n52, 425 n19 rocks 2, 233, 239, 21, 375 Roemer, H. R. 53 n54, n55, 61 n45, 62 n51, 69 n71, 77 n79 Roger II of Sicily 120 n43 Rogers, J. Michael 4 n17, n19, 6 n30, 11 n48, 15 n66, 16 n68, n69, 20 n86, 77 n78, 211 n1.1, 224 n55,332 n20, 390 n59, 403

n78, 417, 436 n72, 507 Rome 380, 391, 439 Rome, Vatican. Biblioteca Apostolica Lectionary (ms. Siriaco 559) 384 n36 Roman tablets 452 Rosen-Ayalon, Miriam 17 n74, 117 n24 Ross, David J. A. 109, 436 n69 Ross, James 24 n104, 334 n26, n29, n30, 336 n35, n36, 337 n37, 338 n39, 339 n41 Rossabi, Morris 150 n3 Roux, Jean-Paul 494 n47, 505 Rowland, B. 238 n104, 240 n117, n118 Roxburgh, David 327 Rubruck 98 Rudaba 25 n105, 66 n63, 68, 184 Fig. 18 Rūm 87 Rum Saljuq 91 n29 432 Rûmî 463 Rumlu 62 Rūšanak 90 n22 Rustam 4, 5, 6 n28, n31, 19 n82, 20 n87, 21, 27 Fig. 1, 57 n26, 66, 67, 84, 94, 169, 185 Figs. 19, 20, 194 Fig. 29, 199 Fig. 34, 330, 374, 375, 427, 433, 438 n78 Ruzbihan Muhammad 301 Sabians 119 n38, 155 Sachau, C. E. 111 n2, n3, 113 n13, n15, 115 n21, 117 n25, 119 n37, n39, 121 n47, 124 n54, 146 n74 Saʿd, Ibn 125 n61 Sada, Feast of 154 n5 Saʿda 459 Sadaqa ibn Abu ʾl-Qasim Shirazi,

INDEX

Samak-i ʿayyar (ms Ouseley 379) 152, 176 Fig. 7 (ms Ouseley 380) 152, 176 Fig. 8 Saʿdi Bustan 7, 7 n37, 8 n40, 17 n72, 30 Fig. 4, 380 n18 Gulistan 5 n24, 23–24, 32 Fig. 6, 33 Fig. 7, 47 Fig. 19, 327–358 passim, 329 n10, 333 n23, 334 n26, 341 n45, n46, 342 n48, n50 sadd-i Iskandar 437 saddles 120, 457 Safaʾ, Ikhwan al-, Rasāʾil 2, 396 Fig. 2, 397 Fig. 3, 399 Fig. 5, 377, 406, 481 n15, 492 n45 Safadi, Y. H. 292 n6 Safavids 51 n3, 57, 59, 59 n40, 60, 61 n47, 62, 62 n52, 63, 63 n56, n57, 64, 68, 68 n67, 69, 77, 167, 304, 372 bookbinding 462, 464 illumination 305 Qurʾans 304 Saffarids 428 Ṣafī al-Din 403 Safid Buland 296, 296 n15 safina 456 Sagundino, Niccolo 105 n50 Ṣāḥib-i Dīwān 380 St. Gregor 404 n83 St Matthew 388 n52 St. Petersburg Academy of Sciences. Institute of Oriental Studies Maqamat (S23) (Leningrad Maqamat) 3 n. 9, 363, 388 486 n23 Asiatic Museum 457 St. Petersburg, Hermitage Museum Khamsa of Nizami(VR-1000) n40, 330 n11, n12, 423 n8

545 St Petersburg, Public Library, Shahnama (P.N.S. 266) 84 St Petersburg, Russian National Library Khamsa of Khusrau (PIS 267) 6 n27, 60 n40 Divan of ʿAli Shir Navaʾi 59 n40 (Turkic New Collection 36) Sakallarios, Leo Bible 404 n82 Sakhr 337 Sale 359 Salm 68, 374 Salmān, Mīrzā 61 n45 Sam Mirza 66 n63, 68, 69, 69 n72 Samak the Knight Errant see: Samakiʿayyar Samak-i ʿayyar (ms Ouseley 379) 152, 176 Fig. 7 (ms Ouseley 380) 152, 176 Fig. 8 Samanids 160 ceramics 298, 308, 309, 310, 311, 316 Fig. 3, 359 Samarkand 80 Samarqandi, Suzanne 164 n38 Samarra 298, 456, 491 Samson 3 Sanʿa 296, 379 n10, 389, 459 Ṣanʿān, Shaikh 23 n99, 321 Fig. 10 San Francisco 298 Sannaja 203 Fig.38 Sarwar, G. 63 n59 Sarre Qazwīnī 223 n54 Sasanian(s) 66, 69, 93, 96, 165, 124 n55, 208, 427, 428, 429, 430, 433, 449 coinage 430 Saunders, John J.95 n38, 109 Savory, R. M. 61 n45, n47, n48, 62 n52, n55, 63 n59 Sawyer, C. 81 n7 sayyid 126

546 Scarce, Jennifer 374 Scavizzi, G. 21 n42 Schaeder, Hans H. 109 Schefer ms see: Maqamat of al-Hariri Schimmel, A. 332 n21 Schindler, Bruno 437 n74 Schmid H. 394 n75, 417 Schmitz, Barbara 153 n14, 425 n18 Schroeder, Eric 5 n21, 19 n80, 49 n111, 57 n26, 167 n44, n45, n46, 292 n2 Schulz, Philipp Walter 481 n15, 505 Scott, R. 391 n60, 417 script 290–306 passim see also: Andalusi, Arabic; Badiʿ, Eastern Persian, Hijazi, Kufic ; Maghribi, nastaʿliq; New Style; riqʿa, Uighur seascapes 330, 330 n12, 339, 349, Fig. 3, 355 fig. 11 Seherr-Thoss, H. C. 403 n77, 417 Seherr-Thoss, Sonia P. 403 n77, 417 Selim II, Sultan 54, 77, 62 n49, 105 Seljuks 160, 161, 236, 430 n36, 432 book illumination 460 ceramics 222 n45, 307, 308, 309, 310, 312, 313, 314, 325, 326 Sellyer, J. 51 n2 Serajuddin, Asma 5 n22, 336 n34 Settignano, Villa I Tatti, Biblioteca Berenson, Anthology of Baysunghur 9 n41, 23 n99, 341 n44 Shahbazi, A. S. 58 n33, 77 n74, 432 n48, 433, n57, 436 n68 Shaghad 199 Fig. 34 shahanshah 430 Shahnama 19 n82, 60, 61, 163–165, 84 n14, 164 n38, n39, 185 Figs 19, 20, 187 Fig. 22, 208, 216 n16, 217, 218 n28, n30, 313, 314, 319 Fig. 7, 332, 343 n52,

358, 421 n5, 424 Baysunghur 4 n18, 6 n27, n28, 7, 22 n95, 25 n105, 31 Fig. 5, 85 n16, 331 n16 Demotte 3, 4, 10, 13 13 n56, 52, 82, 166, 218, 242, 242 n126, 287 n131, 344, 347, 366 Dunimarle 56, 56 n24 First small 83 n10, 164, 183 Fig. 17, 184 Fig. 18, 371 Great Mongol Shahnama 52, 66 n64, 79–110, 81 n5, 84 n14, 97 n43, 125 n59151, 163, 165– 170, 166 n42, 205, 171 Fig. 1, 174 Fig. 4, 188 Fig. 23, 189 Fig. 24, 190 Fig. 25, 191 Fig. 26, 192 Fig. 27, 193 Fig. 28, 194 Fig. 29, 195 Fig. 30, 196 Fig. 31, 197 Fig. 32, 198 Fig. 33, 199 Fig. 34, 200 Fig. 35, 204 Fig. 40, 218, 366, 367, 371, 372, 392, 403 n79, 408, 439, 440–447 Figs. 1–10, 448–451 Ibrahim, Sultan 9 n43, 25 n106, 56, 56 n19, 85 n16 iconography of 365–376 passim and illustration 420–451, 421 n5, 422 n7, 423 n10, n13, 425 n18 Inju 83 n10 Juki 4 n16, 5 n24, 4–5, 7, 7 n33, 9 n43, 20, 20 n87, n88, 21 n90, 23 n102, 27 Fig. 1, 56, 85 n16, 343 n52 Tahmasb, Shah 51, 52 n4, 53, 53–54 n14, 54, 54 n15, n16, 59, 59 n36, 64, 69, 69 n70, 77, 77 n75, 78, 84, 85, 170 n51 shahnameji 433 Shāh-nāma-yi Shāhī 170 n51, 450 n88 iconography 51–78, 52 n4, n6,

INDEX

70–76 Figs. 1–7, 77–78 Shahriyar 302 shahriyar al-sham 303 Shakespeare, William 82 Shalem, Avinoam 118 n29, 373, 374 shamanism 98 Shamlu 62 Shams al-Din ibn Ziyaʾ al-Din alZushki 167 n45, 178 Fig.10 shamsiyya 461 Shehname-i Selim Han 77 n78 Shiʿism 70, 76 Fig.7, 77, 77 n75, 96, 113, 122–127, 125 n61, 127 n68, n69, 129, 148, 154, 359, 377, 377 n4 Shiraz, Iran 164, 164 n39, 165, 185 Fig. 19, 187 Fig. 22, 300, 301, 329 n10, 367, 374, 461, 462 Shirin 5 n25, 6 n26, 15, 23 n96, 42 Fig. 14, 86, 330, 341 n45 Shîrwân 460 Shīz/Taḫt-i Sulaymān 90 Shuʿubiyya 427 shujaira 297 Sicily 120 n43, 296, 410 Palermo 294 Sienese painting 163 Sievernich, G. 409 n92, 418 signed work 51, 51 n3, 53, 455, 460 Sijelmassi, M. 292 n5 silhouette style 50 n113, 221, 326, 362 silk(s) 123, 209, 373, 457, 463, 493 n46 Simjurids 160 Simmonds, Stuart 5 n24, 9 n43, 56 n21 Simpson, Marianna Shreve 60 n42, 83 n9, n10, 109164, 164 n39, 216 n16, 217 n22, 218 n25, n30, 308 n4, 313 n22, 365, 412 n103, 418, 421 n3, n5, 424

547 n15,426 n21, 506 Sims, Eleanor G.1 n2, 4 n16, 52 n6, 55 n17, n18, 56 n19, n22, 57 n26, 58 n34, 59 n35, 85 n15, n16, 86 n17, 110, 345 n59, 358 n68, 386 n42, 405 n86, 407 n88, 412 n103, 418, 425 n18, 506 Simurgh 19 n84, 178 Fig. 10, 238, 239 n109, 367 Sīnān 255 Fig. 13 Sindbād 239 n107 Sindukht 66 n63 Sira of Ibn Ishaq 117, 122, 125 Sīrat ʿAntar 360 Siyamak 54 n14, 374 Siyavush 5 n24, 9 n43, 66, 71 Fig. 2,342 n52 Siyer-i Nebi 116 Skelton, Robert J. 98, 105, 105 n48, 106 n53, 110 Smith, E. B. 390 n55, 418 Smith, G.R. 211 n1.2, 214 n13 snake 17 n75, 234 n90 Solinus 238 Liber Memorabilium 240 n114 Solomon 80 n3, 359 Solomon, Temple of 119 n38, 155 Sophocles 82 Soucek, P. P. 59 n39, n40, 112 n12, 116 n22, 119 n38, 122 n50, 127 n69, 147 n85, 156 n20, 217 n19 Soudavar, Abolala 89 n20, 90 n23, 92, 110, 167, 167 n44, n48, 168, 169, 294 n12, 300 n21, 303 n25, 366, 367, 430 n34, 439 n82 Sourdel-Thomine, J. 18 n31, 242 n125, 288 n140, 392 n65, 393 n69, 418 Southgate, Minoo S. 110 space, use of in illustration 368, 375,

548 380, 384–389 use of in bookbinding 457 spatial devices 26, 387, 392, 484 Spain 301, 361, 428, 453, 459 Spiegel, Friedrich von 81 n6, 110 Spuler, Bertold 93, 96 n40, n41, 98, 110, 112 n11, 113 n14, 114 n18, 118 n31, 10 n41, 121 n46, 155 n18, 235 n97, 242 n125, 288 n140, 392 n65, 393 n69, 418 stage-set 3, 20, 404 Stanley, Tim 367 Stchoukine, Ivan 4 n16, 57 n26, 59 n36, n37, n38, n39, 118 n33, 167 n46, 330 n14, 358 n66 Stern, S. M. 295 n13, 378 n7, 418, 422 n5, 426 n22, 428 n26, 429 n28 Stewart, Desmond 2 n. 3, 3 n. 8, 214 n11, 235 n92, 237 n101, 239 n108, n113, 240 n119, 388 n51, 418 stucco work 3, 374, 375, 491 Subtelny, M. 57 n29, 330 n11, 338 n39, 423 n8 Suard, F. 435 n64 Sufi 17 n75, 88, 92, 96, 96 n42, 97, 150, 157, 436 Ṣūfī, ibn al-, Kitab Suwar al-Kawākib al-Thābita 21 n92, 223, 361, 383, 384, 386, 479 n7, n8, 481 n15, 482 n16 Sugimura, Toh 20 n86 Suhrab 21, 375, 438 n78 Suhrawardī 89 n20, 94 n36, 96 Suleiman the Magnificent 77, 433 Suleimanov, F. 60 n40 Suleimanov, H. 60 n40 Süleymanname 433 Sulṭāniyya 150, 403 Sulwan al-Muṭāʿ 12, 13 n55, 36 Fig.

10, 211 n1.7, 215 n14, 220, 241, 241 n123, 285 Fig. 43, 361, 379 n11, n13 Sunnis 62, 96, 126 sura 117, 126, 297, 299, 301, 304, 458 Sūrī, Manṣūr al- 223 n48 Swietochowski, Marie Lukens 152 n10, 217 n19, 422 n5 Syriac manuscripts 384 n36 Syria 221, 222, 231, 362, 373, 384, 390, 453 emperor of 303 Syrian architecture 390 painting 387 Syro-Jacobite painting 227 Tabaqat 125 n61 Tabari, al- 79 n2, 80 n3, 122 n49 Tabari, al- Tarikh al-rusul waʾl-muluk 18, 46 Fig. 18, 110, 151, 427 n23 Tabriz, Iran 169, 171–175, Figs. 1–6, 180–182, Figs.13–15, 188–200 Figs. 23–35, 204 Fig. 40 150, 157, 394, 440–447 Figs 1–10, 460, 461, 492 n44 Tabriz, Rashidiyya scriptoria 2, 59–60 n40, 94, 111 n6, 207, 208–210 tafsir 128, 151 Tahmasb, Shah Khamsa of Nizami 333 n25 Shahnama 51, 51–52 n4, 53, 53–54 n14, 54, 54 n15, n16, 59, 59 n36, 64, 69, 69 n70, 77, 77 n75, 78, 84, 85, Shāh-nāma-yi Shāhī 450, n88 Tahmina 4, 6 n31, 57 n26, 27 Fig. 1, 330 Taḫt-i Sulaymān 90, 90 n24, 91, 91

INDEX

n25, 92, 96 n42, 165, 394, 492 n44 Tâʾif, al- 459 Takht-i Jamshid 433 Tajik 61 Takūdār 91, 92 Tanindi, Zeren 4 n17, n19, 6 n30, 11 n48, 15 n66, 16 n68, n69, 20 n86, 77 n78, 83 n11, 116 n23, 224 n55 Tantum, G. 211 n1.2 Tarikh al-rusul waʾl-muluk 18, 46 Fig. 18, 151 Tarikh-i Chingizi 149 Tarikh-i jahan-gusha 6 n28, 94, 202 Fig. 37, 151, 217, 218 n31, 395 Fig. 1,398 Fig. 4, 413, 439 n80, 481 n15 ṭarḅa 407 tauqiʿ 299 Tānsūqnāma 217 Tatars 95 n38 Tatimma Ṣirwān al-Ḥikma 377 Taynush 196 Fig. 31, 440 Fig. 1 taʿziya 423, 423 n13 Tedaldi, Jacopo 105 n50 Tehran, Gulistan Palace Museum Shahnama of Baysunghur (4752) 7, 31 Fig. 5, 331 n16 Tehran, Kitab Suwar al-Kawākib al-Thābita 386, 479 n8, 481 n15, 482 n16 Tehran, Riza-yi ʿAbbasi Museum Risālat al-Ṣūfī fiʾl-kawākib (ms. 570) 383, 384 n33 Tekelu 62 temples 3, 22, 155, 488 tents 105 Terence 380 n15 text block 8 n39, 164, 297, 299, 304, 312, 344, 346 n60, 370, 372, 375

549 textiles 123 n53, 154 n17, 169, 209, 225, 307, 314, 337, 359, 366, 373,374, 386, 404, 405,412, 457, 462, 463, 464, 483, 493 n46 Thackston, W. M. 53 n13, 63 n57, 149 n1, 328 n4, n5, n7, 329 n8, 331 n15 theatre 11, 82, 233, 233 n87, 340, 341, 387, 404 Theodosius 105 n50 theology 158 Theophrastus 225 Theriaca 223, 223 n50, n51, 225, 226 n61 see also: Kitab al-Tiryāq Thousand and One Nights see: Alf layla wa-layla thrones 8, 66, 93, 123, 149, 168, 169, 304, 384, 429, 433, 438, 449, 486, 488, 490, 491 see also: enthronements thulth 299 Tibbetts, G. R. 437 n76 Tibet 98 n45, 173 Fig. 3 tiles and tilework 3, 9, 24, 91, 91 n29, 165, 168, 337, 338, 374, 384, 403, 404, 464 tigers 120 n43 Tigris 89 n21 Timūr 98, 98 n46, 105, 105 n48 Timurids 56 n24, 63,157 n23, 158 n27, 328, workshop 328 n7 bookbinding 460, 461, 462, 463 book painting 1–50, 60, 206, 312, 327, 331, 332 342 n52, 345 n59, 367, 372 and architecture 2–9 and blank space 18–25 and colour 2, 5–11, 21, 24–25, 49–50

550 and role of margins 12–18 Qurʾans 331 n17 Timūrnāma 98 n46 ṭirāz 403, 486 n25 triratna 169 Titley, Norah 10 n46, 11 n49, 224 n55, 330 n11 tombs and tombstones 98, 150, 157, 298, 300, 336, 433 Toorawa, S. M. 437 n76 Torah 156 n20 Toynbee, J. M. 230 n76 Trapp, Joseph B. 109 tripartite division 4, 231, 393 Tughyān, al-Jammaz 255 Fig. 13 Tunisia 291, 360, 452 Tunisian bookbinding 452, 453, 454, 456, 457, 458 Tur 68, 74 Fig. 5, 374 Turan 64, 66 Turfan 457 Turkey 54 n15, 292, 304, 414 n107, 423, 425, 426, 432, 458 Turkish influence 61–70, 77–78 see also: Iran-Turan feud Turks 207 Typikon 496 n54 ʿUbaid-Allah, Khan 69 Uccello, Rout of San Romano 209 Uighur script 92 Uljaitu 91, 93 n33 96, 148, 150, 158, ʿUmar 147 ʿUmar-i Aqtaʾ 303 ʿUmarī, al- 98 Umayyads 160, 219 n34, 233, 389 coins 118 Umm Jaʿfar 266 Fig. 24 Underwood, Paul Atkin 492 n42, 505 ʿunwan 296, 297, 299, 303, 312,

312 n15, 369 unica 361 unicorn 214 n11 Universal History of Rashid al-Din see: Jamiʿ al-Tawarikh Ünver, A. Süheyl 217 n23, 382 n27 Uppsala, Uppsala University Library, Dastan-i Jamal u Jalal of Asafi (O Nova 2) 16 n68 Usefulness of Animals see: Manafiʿ-i hayavan Ustad Nizam al-Din Sultan Muhammad 53 Ustâd Qiwâm al-Din 461 Ustajlu 62 Ushmûnain, al- 456 Uzbek Khan 81 Uzbeks 69 Valley of the Quest 37 Fig. 11 Valvo, A. 435 n66 Van Eyck brothers 20 Varāmīn 394, 492 n44 Vardjavand, P. 112 n9 Varqa 315 Fig. 1 Varqa va Gulash 2, 308, 309, 313, 315 Fig. 1, 325, 325 n30 Varro, Marcus Hebdomades 406 Vasari 167, 461 Vasmer, R. 430 n39 Venetians 207 Venice 459 Vergilius Romanus 380 n15 Vesel, Z. 384 n33, n35, 418 Vienna. Archduke Rainer Collection. (Inv. Chart. Ar 28002) 456 Vienna, Österreichische Nationalbibliothek De Materia Medica (Cod. Vindob. Med. gr. 1) 412 n104) endpapers 456 Maqāmāt (ms A.F.9) 211 n1.3,

INDEX

219 Mathnavīs (N.F.382) 6 n26, n30, 28 Fig. 2 Theriaca 223 n50 Viré, F. 437 n76 Virgin Mary 169 Voltolina, L. De 409 n92 Vryonis, Speros Jr 106 n53, 110 Waley, P. 224 n55 wall painting 3 Walzer, M. Sophie 220 n39, 224 n55, n57 Waqwaq 122 n48, 373 Ward, Rachel M. 212 n1.8, 413 n105, 418, 482 n18, 483 n22505 Ward-Perkins, J. B. 230 n76 Warner, A. G. 66 n63, 89 n21, 90 n22, 96 n39, 97 n44, 432 n46 Warner, E. 66 n63, 89 n21, 90 n22, 96 n39, 97 n44, 432 n46 Warner, I. 77 n74 warrâq 464 Washington, D.C., Arthur M. Sackler Gallery Henri Vever papers 57 n26 Washington, D.C., Freer Gallery of Art ceramics 308, 308 n3, 314, 325 n32 Divan (32.30) 13, 37 Fig. 11, 344 n57 Khamsa of Nizami (no. 31.34) 341 n45, 344 n54 Ḫusraw wa-Šīrīn (no.31.36) 5 n25, 23 n96 Jamiʿ al-Tawarikh 242 n128 Sarre Qazwīnī 223 n54 Shahnama 83 n10 Syrian bowl 250 Fig. 8 Sulwan al-Mutʿa (54.1) 12, 36

551 Fig. 10, 241 n123, 285 Fig. 43 Washington, D.C., Smithsonian Institution Great Mongol Shahnama (S1986.103) 174 Fig. 4 Great Mongol Shahnama (S1986.104)197 Fig.32 Great Mongol Shahnama (S1986.105) 196 Fig.31 Great Mongol Shahnama (38.3– 1) 204 Fig. 40 Wāsiṭ Qazwīnī 386 Wāsitī, Yahyā al- 21 n92, 222, 233, 501 n63 water 10, 23, 24, 26, 146,209, 228, 233, 237, 238, 338, 394, 437, 438 Water of Life 437, 438 Watson, Oliver 307 n1, 311 n11, 318, 322, 323, 394 n71, n73 418, 492 n44, 505 Watt, James 150 n3 Watt, W. M. 126 n66, 241 n124 Weitzmann, K. 225 n58, 227 n68, 235 n93, 380 n15, n17, 389 n54, 406 n87, 412 n103, 418, , 489 n33, 490 n35, 500 n62, 505 Welch, S.C. 51 n1, n2, n3, 52, 52 n46, 53 n9,n10 n11, 59 n36, n37, n38, 63, 63 n56, n57, n58, 64 n61, 68 n66, 69 n68, n69, n73, 77 n75, n76, n78, 305 n27 Wellesz, E. 223 n49 Wensinck, A. J. 87 Werner, Franz von 10 n51 Wessel, K. 386 n44, 418 Western bookbinding 456, 459 Western scholarship and Persian book painting 371, 376 and the Qurʾan 290–306 and the Shahnama 365–376

552 Western medieval bestiaries 213, 214,214 n11, 238, 240 Western medieval tradition 381, 389, 409 n92, 437 Whelan, F. 292 n8 White Div 375 White, T. H. 214 n11, 238, 238 n105, n106, 240 n114, n115, n117 Wiet, G. 430 n37 Wikander, S. 433 n56 Wilber, Donald Newton 157 n23, 158 n29, 394 n70, n71, 419, 492 n44, 505 Wilkinson, C. K. 429 n32 Wilkinson, James V. S. 4 n20, 7 n33, 9 n43, 23 n102, 24 n103, 25 n106, 53 n13, 56 n21, 85 n15, 327 n3, 434 n52, Williams, C. H. 238 Willison, Ian 420 n1 Wilson, C. E. 437 n74 Windsor, Royal Library Shahnama (Holmes 151) 84 n13 wine 15 Wixom, W. D. 382 n26, 383 n31, 386 n44, 388 n50, 390 n56, 403 n82, n83, n84, 415 women 79, 88, 126, 408, 438 The Wonders of Creation and Oddities of Existence see: Kitāb ʿAjāʾib al-makhlūqāt wa gharaʾib almawjudat Woodhead, C. 433 n53 woodwork 300, 360, 403 in bookbinding 456, 464 Worcester, Mass. Worcester Art Museum. Great Mongol Shahnama (1935.24) 189 Fig. 24 World History of Rašīd al-Dīn see: Jamiʿ al-Tawarikh

workshops 21 n92, 53 n14, 57, 58, 166, 207, 219, 328, 328 n5, n7, 329, 358, 363, 453 worship 116, 181 Fig. 16 wrestlers 21 n93, 341 n44, 342, 346, 354 Fig. 10, 357 n61, n64 Wright, Elaine 328 n4, 367 Yamamoto, K. 423 n11 Yamanaka, Y. 435 n67 Yaʿqub the Coppersmith 428 Yaqut al-Mustaʿsimi 293, 300 Yarcani 432 Yarshater, Ehsan 108 Yazd 461 Yazdigird 69 Yemen 292, 437, 463 Yemeni bookbindings 458, 459 Qurʾans 300 leatherwork 458 Yohannan, A. 217 n19, 239 n109 Yuan painting 20 Yuan-shao-Pi-shea 98 Yusuf 7, 8, 8 n39, 30 Fig. 4 Yusuf-Jamali, M. K. 64 n60 Zabîd 459, 463 Zabulistan 66 n63 Zafar, Ibn Sulwan al-Mutʿa 36 Fig. 36, 361, 379 n11, n13, 416 Zahhak 66, 373, 374, 428 see also: Ḍahhāk Zahrān 378 n6 Zain al-ʿAbidin b. Muhammad 464 Zain, Dʾzul Haimi bin Muhammad 1 n.1, 10 n47, 30 Fig. 4, 34 Figs. 8a, 8b, 35 Fig.9, 43 Fig. 15, 44 Fig. 16, 45 Fig. 17, 48 Fig. 20, 50, 340 n43 Zaky, A. R. 212 n4 Zal 25 n105, 66 n63, 184 Fig. 18,

INDEX

195 Fig. 30, 373 Zamrak, Ibn 359 Zamzam 145 Zangid dynasty 360 Zanjānī, Abuʾl HasanʿAlī b. Zahrūn 378 Zavara 6 n28 Zij-i Ilkhani 11

553 Zick-Nissen, Johanna 97 n44 zoology 158, 361 Zoroaster 216 Zoroastrianism 89 n21, 90 n22, 97 n44, 430, 435 Zulaikha 7, 30 Fig. 4

Acknowledgements The articles reprinted in this volume first appeared in the following works I

Timurid Art and Culture. Iran and Central Asia in the Fifteenth Century, eds. L. Golombek and M. Subtelny (Leiden, 1992), 76–102

II

Safavid Persia. The History and Politics of an Islamic Society, ed. C. Melville (London and New York, 1996), 53–78

III

The Problematics of Power. Eastern and Western Representations of Alexander the Great, eds. M. Bridges and J. Ch. Bürgel (Bern, 1996), 203–30

IV

Persian Painting from the Mongols to the Qajars. Studies in Honour of Basil W. Robinson, ed. R. Hillenbrand (London, 2001), 129–46

V

The Legacy of Genghis Khan. Courtly Art and Culture in Western Asia, 1256 – 1353, eds. L.Komaroff and S.Carboni (New York, 2002), 134–67

VI

University of Edinburgh Journal XXVI (1974), 331–33

VII

Ars Orientalis XX (1990), 149–87

VIII

Art History 17/1 (1994), 116–26

IX

The Art of the Saljuqs in Iran and Anatolia. Proceedings of a Symposium held in Edinburgh in 1982, ed. R. Hillenbrand (Costa Mesa, 1994), 134–45

556

X

Asian Art 2, ed. J. Tilden (London, 1996), 32–45, 186–88

XI

Encyclopaedia of Islamic Literature, eds. J. S. Meisami and P. Starkey (London, 1998), 475–7

XII

Shahnama. The Visual Language of the Persian Book of Kings, ed. R. Hillenbrand (Aldershot, 2004), 1–7

XIII

Beyond the Legacy of Genghis Khan, ed. L. Komaroff (Leiden and Boston, 2006), 183–212

XIV

Literary Cultures and the Material Book, eds. S. Eliot, A. Nash and I. Willison (London, 2007), 95–119

XV

Dictionary of the Middle Ages 8, eds. J. Strayer, A. Udovitch et al. (New York, 1987), 98–100

XVI

Arab painting: text and image in illustrated Arabic manuscripts, ed. A. Contadini, (Leiden 2007), 117–134