Le Coran de Gwalior - polysémie d'un manuscrit à peintures 2701804434, 9782701804439

Citation preview

ORIENT

MÉDITERRANÉE

19

UMR 8167, Orient et Méditerranée – Textes, Archéologie, Histoire CNRS, Université Paris-Sorbonne, Université Paris 1 Panthéon-Sorbonne, École pratique des hautes études, Collège de France

L

n September 24, 1398, the army of Timur Lang or Tamerlane took the city of Delhi by storm. This date marked the fall of the Delhi Sultanate; it had already been weakened by the emancipation of some of its territories, which had been transformed into small independent sultanates, and by endless conflicts with neighboring Hindu kingdoms. It would take nearly fifty years to recover from the ransacking of its capital and to free itself from Timurid control. It was precisely at the end of the 14th century, in this context of chaos, that a copy of the Qurʾan was completed in the fortress of Gwalior, a few hundred kilometers from Delhi. This manuscript, today kept in the collections of the Aga Khan Museum in Toronto, contains remarkable decorative elements and a hermeneutic system of commentaries. The glosses, which refer to various levels of reading, suggest the existence of complex memorization techniques and of multiple functions for the Qurʾanic manuscript. The flamboyant illuminations, which have no known equivalent in any other works, reveal a number of different influences. Between 2008 and 2014, a research program (UMR 8167- Medieval Islam) focused on this extraordinary work, the analysis of which required the skills of specialists in fields including Qurʾanic studies, history, art history, codicology, paleography and divinatory and magical practices. Their hypotheses and conclusions are gathered in this book.

ISBN 978-2-7018-0443-9

Polysémie d’un manuscrit à peintures

O

LE CORAN DE GWALIOR Polysémie d’un manuscrit à peintures sous la direction de

Éloïse Brac de la Perrière Monique Burési

LE CORAN DE GWALIOR –

e 24 septembre 1398, Delhi est prise d’assaut par les troupes de Tı¯mu¯r Lang, notre Tamerlan. Cette date marque la chute du sultanat de Delhi, déjà ébranlé par l’émancipation d’une partie de ses territoires, transformés dès lors en de petits sultanats indépendants et affaibli par d’incessants conflits avec les royaumes hindous voisins. Il lui faudra près de cinquante ans pour se relever du terrible sac de sa capitale et s’affranchir de la tutelle timouride. C’est précisément en cette même fin du xive siècle, dans ce contexte chaotique, qu’est achevée, dans la forteresse de Gwalior, à quelques centaines de kilomètres de Delhi, la copie d’un coran dont l’existence même soulève un grand nombre de questions. Ce manuscrit, aujourd’hui conservé dans les collections de l’Aga Khan Museum (Toronto), est doté d’extraordinaires décors et d’un savant système de commentaires herméneutiques. Ces gloses, qui font appel à différents niveaux de lecture, supposent des techniques de mémorisation complexes et des usages variés du manuscrit coranique. Les flamboyantes enluminures, dont on ne connaît aucun équivalent, renvoient à de multiples origines. Entre 2008 et 2014, un programme de recherche (UMR 8167- Islam Médiéval) a été consacré à cette œuvre hors du commun, dont l’analyse conjugue les compétences de spécialistes des études coraniques, de l’histoire, de l’histoire de l’art, de la codicologie, de la paléographie, des pratiques divinatoires et magiques. Leurs hypothèses et leurs conclusions sont réunies dans le présent ouvrage.

19

Éditions de Boccard

Le coran de GwaLior PoLysémie d’un manuscrit à Peintures

Illustration de couverture

Coran, fols. 40v-41r. Gwalior (Inde), 801/1399. Toronto, Musée Aga Khan [© avec la permission de l’Aga Khan Museum / By courtesy of the Aga Khan Museum]

Ouvrage publié avec le concours du Labex RESMED (ANR-10-LABX-72) dans le cadre du programme Investissements d’avenir ANR-11-IDEX-0004-02 Ouvrage publié avec le concours du Centre national de la recherche scientifique dans le cadre et avec le soutien du projet Imperial Government and Authority in Medieval Western Islam (IGAMWI) financé par le 7e PCRD European Research Council : FP7-ERC-StG 263361

UMR 8167, Orient et Méditerranée – Textes, Archéologie, Histoire CNRS, Université Paris-Sorbonne, Université Paris 1 Panthéon-Sorbonne, École pratique des hautes études, Collège de France

Directeur de la collection Véronique BOUDON-MILLOT, CNRS - UMR 8167, Orient et Méditerranée Responsable éditoriale Fabienne DUGAST, CNRS - UMR 8167, Orient et Méditerranée Comité scientifique Françoise BRIQUEL CHATONNET Sylvie DENOIX Vincent DÉROCHE Olivier MUNNICH Pierre TALLET Création de la maquette et mise en pages Fabien TESSIER © Éditions de Boccard - 2016 ISBN : 978-2-7018-0443-9 ISSN : 2101-3195

ORIENT

MÉDITERRANÉE

19

Le coran de GwaLior PoLysémie d’un manuscrit à Peintures sous la direction de Éloïse BRAC DE LA PERRIÈRE et Monique BURÉSI

Éditions de Boccard

11 rue de Médicis, 75006 Paris 2016

TABLEAU DE TRANSLITTÉRATION ‫آ‬

ā

‫ض‬

ḍ

‫ب‬

b

‫ط‬

ṭ

‫پ‬

p

‫ظ‬

ẓ

‫ت‬

t

‫ع‬

ʿ

‫ث‬

th

‫غ‬

gh

‫ج‬

j

‫ف‬

f

‫چ‬

ch

‫ق‬

q

‫ح‬

ḥ

‫ک‬

k

‫خ‬

kh

‫گ‬

g

‫د‬

d

‫ل‬

l

‫ذ‬

ḏ

‫م‬

m

‫ر‬

r

‫ن‬

n

‫ز‬

z

‫و‬

w

‫ژ‬

zh

‫ه‬

h / -a

‫س‬

s

‫ي‬

ī / y

‫ش‬

sh

‫ء‬

ʾ

‫ص‬

ṣ

‫ة‬

-at / -a

Voyelles (sauf quelques cas exceptionnels) : a, ā, i, ī, u, ū Diphtongues : aw, ay L’iḍāfa est noté par : -i / -yi. L’article arabe : al- (même devant les lettres solaires)

ABRÉVIATIONS

Bibliographiques Indian Archaelogy. A review: An Annual Publication on Archaeological Reports of Archaeological Survey of India. Encyclopédie de l’Islam / The Encyclopaedia of Islam, nouvelle édition, P. Bearman et al. (dir.), Leyde / Paris : E. J. Brill / Maisonneuve et Larose S. A., 1977-2007.

ASI EI2

D’institutions AKM BL BnF CBL TIEM WAM

Aga Khan Museum, Toronto British Library, Londres Bibliothèque de France, Paris Chester Beatty Library, Dublin Türk ve Islam Eserleri Müzesi, Istanbul Walters Art Musem, Baltimore

RESSOURCES EN LIGNE (CONSULTÉS LE 01/08/2015) Aga Khan Museum, Toronto AKM00281

http://www.e-corpus.org/fre/notices/105296-Tughluq-Qur-an-Gwalior.html.

Bibliothèque nationale de France, Paris Arabe 12 Arabe 3465 Arabe 5844 Arabe 6041 Arabe 6043 Arabe 6073 Arabe 6716 Arabe 7260

http://gallica.bnf.fr/ark:/12148/btv1b84192173 http://gallica.bnf.fr/ark:/12148/btv1b84229611 http://gallica.bnf.fr/ark:/12148/btv1b84192299 http://gallica.bnf.fr/ark:/12148/btv1b8433296d http://gallica.bnf.fr/ark:/12148/btv1b84061661 http://gallica.bnf.fr/ark:/12148/btv1b84061661 http://gallica.bnf.fr/ark:/12148/btv1b8433294k http://gallica.bnf.fr/ark:/12148/btv1b84272593

Library of Congress, Washington Manuscrit 1-84-154.21 

http://international.loc.gov/cgi-bin/ampage?collId=ascs&fileName=070/ ascs070.db&recNum=0

The Digital South Asia Library, Chicago http://dsal.uchicago.edu/reference/gazetteer/pager.html?objectid=DS405.1.I34_V20_110.gif.

Walters Art Musem, Baltimore W563

http://www.thedigitalwalters.org/Data/WaltersManuscripts/html/W563/

REMERCIEMENTS Le projet de recherche consacré au coran de Gwalior dont cet ouvrage est le fruit, a été soutenu par de nombreuses institutions et a bénéficié depuis ses débuts, en 2008, de l’aide de nombreux acteurs, collègues et étudiants. Nous nous devons d’abord d’exprimer toute notre reconnaissance à son Altesse l’Aga Khan à qui appartient cet extraordinaire manuscrit coranique. En la personne de Benoît Junod (Aga Khan Trust for Culture), nous avons trouvé un interlocuteur toujours disposé à faciliter nos recherches, alors que le manuscrit était encore en Europe. C’est grâce à son assistance, puis à celle que nous a prodiguée Heather Ecker (Aga Khan Museum), que le manuscrit a pu être accueilli et exposé à la Bibliothèque nationale de France lors du colloque qui lui a été consacré en juin 2012. Nous adressons tous nos remerciements à Annie Vernay-Nouri, conservateur en chef à la Bibliothèque nationale de France, pour son accueil. Nous avons eu la chance de pouvoir rassembler autour de ce manuscrit des étudiants et des chercheurs confirmés. Ces derniers se sont profondément impliqués dans le projet, tant dans les recherches scientifiques que dans le volet pédagogique : c’est le cas d’Yves Porter et de Francis Richard, avec nous dès les débuts, puis de Nourane Ben Azzouna, Patricia Roger-Puyo et Simon Rettig, qui nous ont rejoints un peu plus tard. Dans des domaines où nous étions plus démunis, pour leur connaissance des pratiques ésotériques et divinatoires, nous devons beaucoup à Christiane Gruber, Constant Hamès et Anne Regourd, et dans celui des textes coraniques et de l’exégèse, à Asma Hilali. D’autres collègues ont bien voulu encore éclairer notre approche en présentant leurs hypothèses et conclusions sur ce manuscrit hors du commun lors du colloque de 2012 : Nalini Balbir en ouvrant la réflexion sur les liens entre les manuscrits islamiques et les manuscrits jaïnas, Johanna Blayac en recontextualisant la genèse de l’ouvrage dans la forteresse de Gwalior à la fin du xive siècle, Finbarr Barry Flood en proposant un éclairage nouveau sur les liens avec les manuscrits ghurides et Fabrizio Speziale en revenant sur la transmission des savoirs indiens dans l’Inde des sultanats. Nous devons aussi remercier John Seyller et Manijeh Bayani, qui nous ont épaulés dans la lecture des sceaux figurant dans le manuscrit. Il nous faut aussi rappeler l’implication des étudiants de l’Université Paris-Sorbonne, Frantz Chaigne, Mathilde Cruvelier, Sabrina Alilouche, Ghazaleh Esmailpour Qouchani, auteurs de deux articles, celle d’Isabelle Imbert et d’Alice Leblanc dans la phase préparatoire du projet. Enfin, il faut évoquer l’extraordinaire travail de coordination qu’a effectué Sandra Aube lors du colloque. Cet ouvrage étant le fruit des très riches discussions que nous avons eues durant cette rencontre scientifique, nous exprimons encore toute notre reconnaissance aux institutions qui en ont permis la réalisation : l’UMR 7528 – Mondes iranien et indien, l’Institute of Ismaili Studies, le Labex RESMED et le musée des Beaux-arts d’Orléans. L’équipe Islam Médiéval – UMR 8167 Orient et Méditerranée nous a offert un constant appui scientifique et matériel depuis la genèse du projet jusqu’à la publication de l’ouvrage. Que Françoise Micheau qui a soutenu le projet dès l’origine et Sylvie Denoix qui nous a guidés et encouragés, puissent trouver ici toute l’expression de notre reconnaissance. Nous remercions enfin Fabienne Dugast qui a coordonné la publication de cet ouvrage et Fabien Tessier pour s’être si consciencieusement investi dans la mise en page. Cet ouvrage a été publié avec le concours du Labex Resmed, l’ERC StG 263361 « Imperial Government and Authority in Medieval Western Islam » et celui du Laboratoire d’archéologie médiévale et moderne en Méditerranée (LA3M). Les images du coran de Gwalior ont toutes été gracieusement fournies par l’Aga Khan Museum et celles du coran W563 par le Walters Art Museum (Baltimore). Pour finir, nous remercions Ruba Kanaan qui nous a grandement facilité l’accès aux reproductions du coran de Gwalior, Stephanie Allen pour son efficacité, ainsi qu’Amy Landau pour les photographies du manuscrit du Walters Museum.

des sultanatsÀà la du XIVe L’INDE DESL'Inde SULTANATS LAfinFIN DUsiècle XIVe SIÈCLE

Sultanat du Cachemire Peshawar Lahore Multan

Sultanat de Delhi s

du

In

Tsangpo

Delhi

Sultanat de Jaunpur

Gwalior

Royaume du Samma

Ju

mn

Patan

Sultanat du Gujarat

a

Jaunpur

Sultanat du Malwa

Gange

Gaur

Sultanat du Bengale

Sultanat de Khandesh

Gondwana

Dawlatabad

Orissa

Goda

Sultanat Bahmanide

Bidar

vari

Telingana

Gulbarga na

sh Kri

Empire Vijayanagar

Royaumes hindous Sultanats musulmans Zone disputée entre dynasties rajputes et sultanat de Delhi

0

500 km © H. Renel, Cnrs UMR 8167

INTRODUCTION Éloïse Brac de la Perrière (Université Paris-Sorbonne)

C’est une véritable surprise que réserve la première rencontre avec « le coran de Gwalior », un manuscrit qui dissimule au sein d’une modeste reliure de cuir rouge, stigmate d’une restauration tardive, 554 feuillets déployant des trésors d’inventivité. Les extraordinaires décors qui ornent cet ouvrage n’ont pas d’équivalent dans l’histoire du manuscrit islamique, l’alliance des textes qu’il recèle non plus. Déconcertant et fascinant, tels sont certainement les adjectifs qui siéent le mieux à ce codex aujourd’hui conservé dans les collections du Musée Aga Khan. La production de manuscrits à peintures dans l’Inde des sultanats demeure dans l’ensemble mystérieuse et ce jusqu’au xve siècle. Les ouvrages enluminés datant de cette période qui couvre plus de 300 ans (de 1206 jusqu’à l’instauration progressive de l’empire moghol au xvie siècle) semblent bien rares, surtout si l’on tient compte du fait que certains sultanats de l’Inde, notamment celui de Delhi jusqu’au début du xive siècle, ont attiré un grand nombre de savants et de lettrés venus de toutes les terres d’Islam. Textes et vestiges archéologiques témoignent d’une vie intellectuelle et artistique foisonnante dans ces cours indo-musulmanes et le patronage des sultans est attesté à de multiples reprises par les sources et l’épigraphie monumentale. Les quelques manuscrits dont nous disposons aujourd’hui ne constituent qu’un éventail peu représentatif d’une production livresque qui fût certainement bien plus étendue. Le coran de Gwalior se trouve être l’unique manuscrit à peintures pouvant être attribué avec certitude à l’Inde islamique avant le xvie siècle. Il est ainsi nommé car sur sa dernière page figure un colophon qui donne le nom du lieu où l’ouvrage a été copié, et très vraisemblablement enluminé, Kālyūr, Gwalior, place forte située à quelques 300 km au sud de Delhi, ainsi que la date exacte d’achèvement de la copie, le « 7 du mois de ḏū al-qaʿda 801 de l’hégire », soit le 11 juillet 1399 de notre ère1. Sept mois seulement séparent la fabrication de ce manuscrit des évènements politiques majeurs qui ont marqué un tournant décisif dans l’histoire de l’Inde : le 17 décembre 1398 en effet, Tīmūr Lang, notre Tamerlan, a mis Delhi à sac, précipitant la chute de ce qui a été  pendant deux siècles le plus puissant sultanat de l’Inde. Ainsi l’atmosphère dans laquelle ce codex a vu le jour est bien particulière, complexe et chaotique. La forteresse de Gwalior, alors sous la coupe des Tomara, une lignée rajpute qui l’a ravie aux Musulmans seulement

1.  « Tamām  shud  in  jāmiʿ-i  kalām-i  masjid  bi-khaṭṭi  bandi-yi  gunāh  kār-i  umidwār  bi-raḥmat-i  parwardigār Maḥmūd Shaʿbān sākin-i qalʿi-yi Kālyūr ghafara Allāh lahu wa li-wālidayh wa li-man qaraʾa wa ḥafaẓa rūz-i dūshanbih waqt-i namāz-i pishin haftum azqr māh-i ḏū al-qaʿda hashṣad wa yik az  hijrat. Khaṭṭī yuḍakkiranī li-aḥbāb iḍā ṣāra al-turāb ʿalā saqfān ḥājibān » : ici s’achève cet ensemble de  glorieuses paroles, calligraphiées par l’esclave pécheur qui implore la miséricorde divine, Maḥmūd  Shaʿbān, habitant la forteresse de Gwalior. Que Dieu lui accorde son pardon, à lui et sa famille,  ainsi qu’à celui qui lira [ce texte] et l’apprendra par cœur. Lundi, à l’heure de la prière de midi, du septième jour du mois de ḏū al-qaʿda de l’an huit cent un de l’hégire. Que ma calligraphie rappelle mon souvenir à mes proches lorsque, comme un toit, la terre me dérobera aux regards. En réalité, le 7 ḏū al-qaʿda de l’année 801/11 juillet 1399, correspond à un vendredi et non à un lundi. Il est donc envisageable que la copie ait été achevée le 17 ḏū al-qaʿda, un lundi, 21 juillet 1399 du calendrier grégorien. Les articles de cette étude, dans leur grande majorité, s’en tiennent à la date donnée par le colophon sans prendre en compte le jour de la semaine. Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 11-14

12 • ÉLOÏSE BRAC DE LA PERRIÈRE

quelques mois plus tôt, subit inévitablement les conséquences des évènements qui bouleversent le cours de l’histoire indienne. Elle accueille provisoirement dans l’enceinte de ses murailles les exilés qui fuient vers le Sud, où les sultanats du Deccan pourront constituer  un ultime refuge. Le  colophon  porte  aussi  le  nom  de  Maḥmūd  Shaʿbān qui a probablement supervisé l’exécution du manuscrit dont les calligraphies et les peintures présument l’intervention de plusieurs artistes. Sur la page du colophon, plus bas, figure l’un des deux cachets qui  apparaissent dans le codex mais qui n’offrent malheureusement que très peu d’indices sur le parcours du livre après sa création2. Ce sceau est circulaire, polylobé et couronné d’un motif trifolié. Il affiche une date dont seuls les deux premiers chiffres sont lisibles ; on  peut la situer entre 1200/1785 et 1299/1881. On y déchiffre encore le nom « Muḥammad  Amīn » sans être en mesure d’identifier de quel personnage il s’agit précisément. L’autre  cachet3 se trouve à l’ouverture du manuscrit et comme l’ont remarqué Manijeh Bayani et John Seyller, il peut être rapproché de celui de ʿAbd al-Ḥaqq b. Qāsim Shīrāzī, alias Amānat  Khān (m. 1644), célèbre calligraphe de l’Inde moghole, dont il partage la forme et la date :  1037/1627. Bien qu’aucun nom ne soit inscrit sur cette empreinte initiale, on peut y lire le texte suivant : « Zāhid Sūhrawardī sharaf yāftih / Makhdūm-i maḥbūb-i rabb-i raḥmān wa raḥīm » (Qui a accédé à l’honneur d’être hermite de Suhrawardī, [un] maître aimé de  [notre] Créateur, Clément et Miséricordieux). Le manuscrit a donc appartenu au xviie siècle à  un  disciple  de  l’ordre  suhrawardī,  l’une  des  principales  confréries  soufies  implantées  dans le sous-continent depuis l’instauration du sultanat de Delhi. Or, comme le proposent plusieurs auteurs dans notre ouvrage, il est fort possible que le coran de Gwalior ait été fabriqué au sein de telles confréries ou, au moins, grâce à leur patronage. Le codex se présente comme un épais volume de 22,2 × 29 cm, des dimensions relativement modestes pour une œuvre de cette qualité. Le papier a certainement coûté cher : de couleur ivoire il est d’une texture très fine, parfaitement polie, brillante et homogène.  Le texte principal est disposé sur treize lignes alternant en rouge ou bleu ; il est copié en  écriture bihārī. Cette bichromie se retrouve dans d’autres corans en écriture bihārī, mais elle est plutôt rare en général ; au-delà des considérations d’ordre esthétique, il est possible  qu’elle ait servi de béquille à la mémorisation du texte et sa récitation. En effet, la calligraphie bihārī est celle d’un vaste ensemble de corans indiens dont le coran de Gwalior partage plusieurs caractéristiques, prouvant qu’il n’est pas, comme certains l’ont avancé, un objet isolé, un accident de l’histoire. Bien que nous n’ayons que trop peu d’informations sur leur genèse, l’existence de ces corans semble s’être étendue du xive siècle au xixe siècle. Au cœur du manuscrit de Gwalior, la calligraphie bihārī, comme un fil d’Ariane, fait le lien entre les peintures hétéroclites qui scandent l’ouvrage et forment une collection d’images bigarrées, florilège de motifs appartenant indifféremment aux mondes islamique  et indien et agencées ici en d’insolites compositions. Parfois talentueuses, parfois moins expérimentées, les mains qui ont collaboré à l’exécution de l’ouvrage sont donc nombreuses et le soin qu’on lui a porté incontestable. C’est bien d’un manuscrit de luxe qu’il s’agit même si le caractère expérimental qu’il présente, l’aspect parfois naïf de certaines peintures, bouleversent  nos  critères  de  jugement  influencés  par  une  meilleure  connaissance  des  manuscrits coraniques proche et moyen-orientaux du xive siècle dont les normes ornementales sont plus aisément déchiffrables. Pourtant, le coran de Gwalior demeure dans le contexte particulier où il a vu le jour, celui du crépuscule des sultanats de l’Inde, une  œuvre d’une qualité esthétique exceptionnelle.

2.

Je tiens à citer ici l’aide fournie par Manijeh Bayani, John Seyller, puis Asyeh Ghafourian pour la lecture de ces sceaux ; je les en remercie encore. Voir figure 1 de la communication de J. Blayac. 3.  Voir figure 4 de la communication de J. Blayac.

INTRODUCTION • 13

Comme un écho textuel à la matière picturale, le système herméneutique du manuscrit, riche et complexe, est un autre puzzle à reconstituer. Les gloses multiples jointes au texte coranique forment un ensemble inédit, le livre de divination (fālnāma) qui conclut le manuscrit s’avérant même tout à fait exceptionnel pour l’époque. Cette combinaison de différents textes dans le même codex ouvre des perspectives sur les cercles de réception, ainsi que sur l’usage de tels manuscrits. Le coran de Gwalior démontre par sa seule existence combien les pistes à explorer pour parfaire notre connaissance du codex coranique, particulièrement, et du codex en écriture arabe, plus généralement, demeurent nombreuses. Une partie de l’histoire de l’Inde islamique pré-moghole – celle de sa production livresque et artistique, celle de son patronage, de ses artisans et de ses ateliers – est également dissimulée derrière ces énigmes. C’est en s’intéressant aux dynamiques, aux échanges entre l’Inde islamique et le monde qu’elle côtoie, à la circulation des objets et, bien sûr, à celle des idées que ces objets mêmes véhiculent, qu’il sera possible de reconstruire la trame d’un récit inscrit au fil des pages du coran de Gwalior. L’analyse d’un tel objet invite  à renouveler notre manière d’aborder l’histoire de l’art islamique en redonnant pleinement leur place à certains domaines peu explorés, comme l’Inde des sultanats. Les approches arbitraires qui reposent sur une géographie dessinant un monde rayonnant autour d’un noyau central, montrent là leurs limites : en déplaçant l’angle de vue, en repositionnant notre regard à partir de ces soi-disant périphéries, de nouveaux paysages s’ouvrent à nous. Le projet mené sur le manuscrit de Gwalior durant quatre ans, au sein de l’université Paris-Sorbonne et de l’UMR Orient et Méditerranée (UMR 8167 – Islam médiéval), a permis de rassembler autour de ce codex peu commun des spécialistes de plusieurs disciplines. De cette alliance transdisciplinaire, de cet effort commun, sont nées les réflexions et conclusions  développées dans cet ouvrage. Ce dernier est divisé en trois parties, trois faces d’un unique prisme réfléchissant trois manières d’approcher le manuscrit. La première partie est consacrée à l’analyse détaillée du coran de Gwalior, le codex même et le texte qu’il renferme. Frantz Chaigne et Mathilde Cruvelier présentent pour commencer une étude approfondie des décors, comme une première entrée dans l’histoire de ce codex. Ils s’interrogent sur les possibles origines des peintures, le choix des répertoires iconographiques, le caractère expérimental de ces décors, l’agencement insolite des motifs et la nouvelle grille de construction de l’ornement que ces enluminures proposent par rapport aux développements voisins de l’enluminure islamique. Cette approche iconographique et stylistique est éclairée dans un second chapitre par les analyses spectrométriques menées sur les peintures par Nourane Ben Azzouna et Patricia Roger-Puyo. Décomposant les différentes phases d’intervention sur le manuscrit, l’article émet quelques hypothèses sur la distribution des tâches qui a prévalu à la réalisation de ce codex, envisageant une organisation en atelier qui reposerait sur une division hiérarchique du travail. Concluant cette première partie de l’ouvrage, Sabrina Alilouche et Ghazaleh Esmailpour Qouchani explorent les gloses, et définissent la teneur des textes en examinant  leurs possibles usages. Le livre de divination qui clôt le manuscrit est finalement comparé  à  ceux  figurant  dans  d’autres  corans  indiens  avec  lesquels  le  coran  de  Gwalior  partage  plusieurs correspondances formelles et textuelles. La deuxième partie du volume veut offrir au lecteur différents éclairages pour mieux cerner les contextes dans lesquels un tel ouvrage a pu voir le jour. Johanna Blayac d’abord, en synthétisant et en analysant les sources historiques, procède à une mise au point sur la présence musulmane à Gwalior avant le xvie siècle, question directement liée à celle du patronage à l’origine d’une œuvre de cette envergure. Le chapitre suivant que l’on doit à Nalini Balbir, permet de revenir sur les liens formels ayant pu exister entre le codex coranique et d’autres formes de codex, également religieux mais non-islamiques. En effet, le coran de Gwalior a été exécuté dans un haut lieu de la culture jaïne, une place forte non-musulmane. Il présente des liens stylistiques avec certaines copies du livre jaïn des rituels, le Kalpasūtra, exécutées dans le Gujarat voisin. L’exhibition des manuscrits du Kalpasūtra lors de cérémonies publiques, « leur mise en spectacle », a-t-elle pu avoir un quelconque impact sur un codex comme le coran de Gwalior ?

14 • ÉLOÏSE BRAC DE LA PERRIÈRE

La présentation détaillée du système herméneutique établie précédemment par Sabrina Allilouche et Ghazaleh Esmailpour Qouchani permet enfin à Asma Hilali de questionner  dans un sixième chapitre les usages du manuscrit coranique, en comparant les manifestations matérielles les plus anciennes du Coran à ces développements indiens de la période prémoderne. Il s’agit d’examiner le Coran comme l’élément central d’un tableau complexe et vivant, toujours en mouvement, celui des cercles de transmission des textes religieux. Le lien entre Coran, ḥadīth et textes qu’Asma Hilali désigne comme « textes intermédiaires », est ici interrogé. Enfin, la dernière partie de ce livre présente les réseaux transrégionaux qui tissent la toile de fond de l’histoire de l’architecture et des arts mobiliers de l’Inde pré-moghole. Considérant que c’est à travers deux dynamiques croisées, « horizontale » ou transrégionale d’une part, et « verticale » ou diachronique d’autre part, que doit être envisagée l’étude d’une œuvre telle que le coran de Gwalior, Finbarr Barry Flood revient sur la portée de l’héritage ghuride du xiie siècle, tout en proposant de possibles chaînons manquants.  Dans le chapitre qui suit, Yves Porter se consacre aux liens formels et aux transferts stylistiques entre les manuscrits indo-islamiques et d’autres productions artistiques. La fleur  de lotus, issue du répertoire décoratif de la Chine, qui a fait une apparition massive dans les arts de l’Iran mongol aux xiiie et xive siècles, figure dans plusieurs décors du coran de  Gwalior. Partant de ce constat, il s’intéresse aux motifs sinisants dans les arts du livre et les décors architecturaux de l’Inde des sultanats. Deux voies de transmission se dessinent : l’une, directe, par les importations d’objets chinois. L’autre, indirecte, par les arts de l’Iran ou du monde mamelouk, en particulier dans le domaine des manuscrits enluminés. Simon Rettig, qui conclut l’ouvrage, met en avant les liens subtils mais profonds qui lient les productions indiennes aux ateliers de l’Iran voisin. Il s’appuie sur l’examen d’un superbe coran indien conservé au Walters Art Museum à Baltimore exécuté dans le goût timouride, mais qui présente tout en même temps des ressemblances indéniables avec le manuscrit de Gwalior. Ce dernier serait au cœur d’un processus partagé par d’autres manuscrits indiens du xve siècle, ultime représentant de lignées passées et précurseur d’une production à venir. Les études réunies dans cet ouvrage rendent compte de la richesse des problématiques soulevées par le coran de Gwalior et, plus généralement, par les manuscrits enluminés, coraniques et non-coraniques, de la période pré-moghole. Synchronisant les voix, croisant les regards, les trois parties sur lesquelles est construit le présent recueil s’interrogent et se répondent sans cesse et déchiffrent de concert l’une des plus passionnantes énigmes de l’histoire du manuscrit islamique.

I

LE CORAN DE GWALIOR THE GWALIOR QURʾAN

THE ORNAMENTATION OF THE GWALIOR QURʾAN, BETWEEN DIACHRONIC LEGACIES AND GEOGRAPHIC CONFLUENCES Frantz chaigne (Université Paris-Sorbonne, UMR 8167) Mathilde cruvelier (chercheuse indépendante)*

Abstract Just like the manuscript considered as a whole, the ornamentation of Qurʾan AKM00281 gives rise to many questions and the accumulation of these leads us to consider it as a unicum among the production of the Indian sultanates. There are two complementary approaches to understanding its decoration. The first is based on a study sui generis. With this purpose, successive focalizations highlight the main divisions of the text as well as the structures that define or articulate them (inaugural frontispiece, frontispieces indicating the different  quarters of the Qurʾan, frontispieces marking the ajzāʾ (sing. juzʾ), titles of suras, marginal vignettes and ‘yokes’, inter alia). This approach continues with the outline of a taxonomy of the elements which fill in these structures, these elements being either geometric, vegetal  or ‘vegetalized’. In the second section, this preliminary study – thanks to which we are now able to characterize the ornamental language inherent in this singular manuscript which was copied in a troubled context – is put into perspective. A comparison of these observations with the productions of some poles of the Islamic world, such as the Mamluk sultanate, the Ilkhanid empire and its successors, or the Turkish principalities, shows us at which point the Indian sultanates had become deeply integrated into the umma and shared aesthetic options developed in the aforementioned regions, even if it involved adapting them. It is also worth noting that the Jain Indian substratum is by no means a solitary island in the choice of filling patterns; which raises new questions about the workshops’ compositions.  All these elements suggest that this region was a major centre of convergence, interaction, mixing and assimilation, as is well attested by many historical sources.

Résumé L’ornementation du coran de Gwalior, entre héritages diachroniques et confluences géographiques À l’instar du manuscrit considéré dans sa globalité, l’ornementation du Coran AKM00281 soulève de multiples questions dont l’accumulation l’érige au rang d’unicum au sein de la production de l’Inde des Sultanats. La compréhension du décor peut être envisagée selon deux approches complémentaires. La première repose sur son étude sui generis.  À  cette  fin,  des  focalisations  successives  permettent  de  mettre en évidence les divisions principales de ce coran et les structures qui les *

Both authors express their gratitude to Dr. Brac de la Perrière for giving them the opportunity to take part in this long-term program. Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 15-56

18 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

sous-tendent ou qui soulignent leurs articulations (frontispice inaugural, frontispices de division en quatre parties, frontispices de juzʾs, bandeaux de sourates, vignettes et empiècements marginaux…) ; cette approche se poursuit par une  esquisse de taxonomie des éléments de remplissage accueillis par ces structures, éléments tant géométriques que végétaux. Dans un deuxième temps, cette étude préliminaire est mise en perspective. En effet, les comparaisons des résultats de l’étude précédente avec des productions de certains pôles du monde islamique, tels que le sultanat mamelouk, l’empire ilkhanide et ses successeurs, et les principautés turques montrent que l’Inde des sultanats partageait des solutions développées dans l’ensemble des sphères précitées, quitte à les adapter. Toutefois, il convient de remarquer que le substrat indien jaïn n’est nullement occulté dans la multiplicité des motifs de remplissage, ce qui oblige à s’interroger sur la composition des ateliers. La concomitance de ces faits tend à montrer que cette région était tout à la fois un lieu de rencontres, de convergences, de mélanges et in fine d’assimilation, ce que confirment au demeurant de nombreuses sources historiques.

Although the regions of Delhi and Gwalior suffered invasions and ravages by Timurid armies, the artists there were nonetheless able to produce a truly fascinating Qurʾan, whose ornamentation has up to now no previously dated analogue. We hope that this detailed study will offer a tool to illustrate the potentialities of an intrinsic close examination, which may make it possible to characterize a local production which took a comparative approach including ornaments originating from other cultural areas. The second aim of this paper is therefore to understand the existence of this amazing artifact and to characterize the artistic, commercial or even diplomatic aspects that such illuminations presuppose.

eLements of contextuaLization Islam first appeared in India as early as 711, when  the province of Sind was taken over by Arabs sent by the Umayyads (41-132/661-750). From then on, India was often subjected to invasions from Islamized dynasties such as the Ghaznavids (366-582/ 977-1186). For instance, in 1022, under Ghaznavid domination, Lahore became a Muslim capital where Persian culture flourished. At first glance, one might think that the outlying  position of India in the Islamic world during the late medieval period would have kept it isolated from the main centres that were the Mamluk Empire and the Ilkhanate and its followers. In fact, though, this significant  distance  never  inhibited  close  relations  between the Delhi Sultanate with the western regions of the Islamic world. Links were formed in all kinds of fields: trade, diplomacy and culture. Furthermore,  the geographic area of the Indian sultanates was also shared by non-Islamic cultures, such as the Jains, for whom the arts of the book constituted a primordial cultural aspect too. We must therefore examine this

Qurʾan, completed on 7 Dhū al-Qaʿda 801/11 July 1399 in the fortress (qalʿa) of Gwalior,1 as the result of exchanges between many centers and by no means as the product of a ‘lost country’, cut off from any reference.2 India was indeed a node for the trade in many costly goods, such as spices, ambergris, ivory, camphor, indigo, pepper, etc., and some merchants from Iraq and Egypt, the nawākhiḏat (i. e. ship captains), even specialized in the spice trade. More specifically,  relationships with Egypt can easily be illustrated in many ways: for instance, trade flourished during Antiquity and Alexandria served as an emporium for goods from China, India and the Occident. As for the Islamic period, archaeological evidence in the city of Fustat shows that textiles were imported from Gujarat between the 12th and the 15th centuries. Letters written to their families by merchants who sailed to India have been discovered among the documents of the Genizah in Cairo.3 In the 8th/14th century, the famous traveler Ibn Baṭṭūṭa mentions that in Alexandria he met a hermit whose three brothers were living in India, Sind and China, respectively, and who described in turn the permanent presence of an Indian settlement in Egypt.4 Commercial relationships between Egypt and China are recounted by the thirteenth-century historian Chau Ju-Kua in a chronicle, in which India plays the role of a middle

1.

The city of Gwalior was identified as “kālyūr” in the colophon of the manuscript, an orthography also found in  Ibn  Baṭṭūṭa’s  Riḥla, among other written forms: Ibn Baṭṭūṭa [1976], 43-45, 152, 162, 163. 2. Brac de la Perrière, Chaigne and Cruvelier (2010). 3.  Al-Qūṣī [1976], 95; Regourd (2012), 1-19. 4.  Ibn  Baṭṭūṭa  [1958-2000],  vol. I,  23-24,  228.  Evidence  is  also provided by the architecture of Fatimid and Ayyubid mosques in Cairo: Salem (1993), 32-35.

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 19

station.5 Diplomatic and scholarly exchanges ought to be mentioned as well: for instance, according to Ibn Ḥajar,6 Muḥammad b. Tughluq (r. 725-752/1325-  1351), the sultan of Delhi, sent a ship full of presents to the Mamluk sultan al-Nāṣir Muḥammad b. Qalāwūn  (r., with interruptions, 693-741/1294-1340). Such embassies were certainly quite common, for Ibn Baṭṭūṭa  wrote  that  the  Indian  sovereign  sent  the  Shaykh Rajab al-Burquʿī to the Abbasid caliph in Cairo  al-Hākim II (r. 741-753/1341-1353),7 and al-Maqrīzī8 wrote in 1304 that Indian emissaries had reached the Mamluk court. Finally, we must not neglect the role played by the circulation of Indian scholars. We  can  at  least  mention  al-Sarrāj  al-Hindī  ʿUmar b. Isḥāq’s (d. 1395) travels to Egypt, or those of al-Sāfī  al-Hindī Muḥammad b. ʿAbd al-Raḥmān, who also  visited Egypt in the 12th century.9 Long-term relationships between India and Egypt are clearly illustrated by these few examples, which reveal the depth of commercial, diplomatic and cultural links between these two geographic poles. Obviously, the relationships between the Delhi sultanate, founded at the turn of the 13th century, and the Middle East constitute an essential key for understanding the context of the production of this Qurʾan and its decorations. Although these links had existed since Antiquity, we will focus here on the period involving the Mongols and their followers.10 At the beginning of the 13th century, the first waves  of Mongol invasions led to the migration of the Turks who had established themselves in Khurasan, and these population transfers continued under pressure from the Chaghatai khanate. Once the Ilkhanid empire was established, it and the Delhi sultanate developed a policy of appeasement, illustrated by their frequent exchanges of embassies. An example of this is the case of the Ilkhanid Ghāzān (r. 694-703/1295-1304) sending  his  famous  vizier  Rashīd  al-Dīn  to  ʿAlāʾ  al-Dīn  Khaljī (r. 695-715/1296-1316).11 As this Rashīd al-Dīn  is known for being one of the most enlightened patrons of the arts of the book, it is reasonable to imagine that he carried some manuscripts with him.

5. 6. 7.  8.  9. 10.  11. 

Ibid., 26. Unfortunately, we were not able to consult the primary source. Ahmad (1999), 17. Ibn Baṭṭūṭa [1958-2000], vol. I, 227-228. Al-Maqrīzī (1958), 712-713. Both authors thank J. Pfeiffer  for her help. Ibid., 37. Here too, we were unable to find any primary  source. For  a  comprehensive  approach,  see:  Jackson  (1975);  id.  (2009); Aḥmad (1998); Hambly, Asher, 1994. This  fact  is  related  by  Rashīd  al-Dīn  himself  in  his  Kitāb-i mukātibāt-i rashīdī: Jackson (1975), 129.

After  the  rule  of  Ghāzān,  his  successors  Öljeitü  (r. 703-717/1304-1317)  and  Abū  Saʿīd  (r. 717-736/  1317-1335) continued building relationships with the Indian sultanate, and the last Ilkhanid ruler received in  1328  an  embassy  from  Muḥammad  b. Tughluq.12 During his reign, as he was ordered to leave the court, the son of Rashīd al-Dīn met the Sayyid Adad coming  back from Delhi with many goods for him and for Sultan Abū Saʿīd.13 Furthermore, the Tarīkh-i Firūz Shāhī  points  out  that,  ca. 1345,  Sultan  Muḥammad,  who was at that time in Delhi, used to honouring the numerous Mongol dignitaries, giving them many gifts.14 Of course, this alliance with the Ilkhanate was very convenient in helping Delhi to stamp out the Chaghatai danger.15 A  later  wave  of  invasions,  led  by  Tīmūr,  caused  an important migratory influx toward the Delhi sultanate, whose capital fell in 1399. The Qurʾan of Gwalior is contemporary with this troubled period and to this day remains the first illuminated Qurʾan copied in the Indian sultanates16 that we are aware of, but no analysis of its ornament can be carried out without considering the fact that this region was for a long time in contact with the Middle and Near East17 – both already leading zones of experimentation in the arts of the book – and was also an area where several religions cohabited, including the Jains, for whom books were a significant part of their culture.18 If Delhi’s geographic position may be considered as outlying, the sultanate can nevertheless be seen as a cultural node in the Islamic culture of this period.

ProbLems As this study has two main aims, i. e. giving a detailed description of the manuscript per se and highlighting the networks of exchanges to which it bears witness, our approach has to rely on two methodological tools. The first one was developed by  Marianne Barrucand, who showed the potentialities of a deconstructive approach in the understanding

12.  Aḥmad (1998), 65. 13.  Rashīd  al-Dīn  al-Ṭabīb  [1836],  introduction,  part I,  LV.  Nothing else has been written about this Sayyid Adad. 14. Jackson (1975), 149-150. 15.  For  instance,  Amīr  Chūpān  inflicted  a  defeat  on  Tamarshīrīn in 1326: Boyle (dir.) (1968), 410. 16. Brac de la Perrière (2008), 133. 17.  In this specific article, the authors consider the Near East  as the region that gathers medieval Bilād al-shām and  Egypt whereas the Middle East includes the Persian speaking area and ʿIrāq-i ʿarab. 18. See Johanna Blayac’s article in this volume. And also http://www.jainology.org/projects/cataloging-of-jainmanuscripts.

20 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

of the ornamentation of a manuscript.19 According to this process, we will undertake a detailed description of the illumination of this Qurʾan, following a protocol that starts with its general programme, then moves on to the structures and ends with the filling in of  the structures. But this approach would lack interest without a constant comparison with the solutions in use in other Islamic spheres or even in India. This may indeed help us to understand on the one hand the role India played among other empires or cultural areas and, on the other, how the Indian sultanates developed their own aesthetics. On this point, the considerations developed by Flood have been highly useful to our work. In his stimulating book Objects of Translation,20 he establishes the role of transcultural exchanges in the Indian context by underlining “relations rather than essences”21 and by exploring “the dialectic between alterity and identity”. Admittedly, the period Flood studies runs from the early 8th to the early 13th century, but most of the conditions are similar: the presence of an extra-Islamic substratum, participation in an international network, military threats from the north. A careful comparison between the ornamentation of this Qurʾan and artifacts from other cultural poles, Islamic or not, should therefore provide a clue to the role that sultanates played as centres of the arts, diplomacy and commerce, at the very least.22

frontisPieces Diversity in function and construction This Qurʾanic manuscript has no less than thirtyfour double frontispieces that can be classified as follows: an opening one (fols. 1v-2r), twenty-nine others placed at the beginning of the ajzāʾ – the frontispiece of juzʾ 13 is inexplicably missing – and four that indicate suras 2, 7, 19, and 38.23 It should also be noted that juzʾ 30 is devoid of any illuminated decoration, its title being simply written in the margin.

19.  This  approach  has  been  developed  and  exemplified  by  Marianne Barrucand (1992), 235-248; id. (1995), 164-172  and 194-195, both papers concerning the Maghribi illuminated manuscripts. 20. Flood (2009): see especially the introduction, 1-14. 21. Or, in the words of Clifford Flood, ‘roots’ and ‘routes’ (Ibid., 3). 22. In a preliminary paper written with Brac de la Perrière, we proposed that this decoration, issuing from many models but nevertheless with its own unity, be qualified  as “kaleidoscopic”: Brac de la Perrière, Chaigne, Cruvelier (2010). 23. Respectively sura 2, al-Baqara,  (fols. 3v-4r);  sura 7,  al-Aʿrāf, (fols. 143v-144r);  sura 19,  Maryam, (fols. 274v-275r) and sura 38, Ṣād (fols. 406v-407r).

The construction of the inaugural double frontispiece (fols. 1v-2r) (figure 1) follows quite traditional  models from the Middle or Near East, with its tripartite division and the presence of marginal vignettes. For instance, such a structure can be compared with frontispieces of earlier famous Persian Qurʾans such as the Qurʾan copied in Bust in 505/1111-1112 (BnF Arabe 6041).24 However, its fillings of stars and polygonal patterns, which are unique in this manuscript, are much closer to what we can usually observe in Mamluk production from the 14th century. The most famous example of such a composition, involving a geometric development of polygons centered on an eight-pointed star, is given by the double frontispiece opening one of the seven parts of a Qurʾan copied for the future sultan Baybars (r. 658-676/1260-1277) under the supervision of master Muḥammad b. Mubādir  Abū  Bakr,  known  as  Ṣandal  (BL Add. 22406-12,  Egypt or Syria, 1306-1308).25 However, among the structural differences between both Qurʾans, we can notice in Baybars’ one the absence of tripartite division in the frontispiece. This aesthetic option in fact occurs in Mamluk manuscripts from the second part of the 14th century onwards,26 even if few earlier examples are known.27 References in the Middle East seem to be harder to find but all present the special feature  originating from ʿIrāq-i ʿArab.28 In fact, at that time the shamsa was the most common solution for the opening pages of manuscripts in the Persianate and the Indian worlds.29 Even if all the other frontispieces belong to a unique family of aesthetics, they do not belong to the same group of functionality. As we have already remarked, four of them, situated at the beginning of suras 2, 7, 19 and 38, indicate in fact the opening of the arbāʿ (sing. rubʿ), i. e. the quarter divisions of a Qurʾan.

24. This manuscript is available on http://gallica.bnf.fr/ark:/ 12148/btv1b8433296d. 25. See James (1988), 50-51. 26. BnF Arabe 12 (753/1353): Vernay-Nouri, Guesdon (2001), 73;  Cairo,  Dār  al-kutub:  MS  8  (Cairo,  757/1356):  James  (1988), no. 24, MS 6 (Cairo, ca. 1366-1368): ibid., no 28, MS 54 (ca. 1369-1372): ibid., no. 29; Khalili collection, QUR595  (dated 1382-1383): James (1992), no. 47. 27. TIEM 450 (Cairo, 713/1313): James (1988), no. 6, CBL 1476 (Cairo, ca. 1332-1336): ibid., no. 18, TS 138.M5 (Damascus?, 741/1341): ibid., no. 21. These nevertheless differ from the Qurʾan of Gwalior in their chromatism, the details of the ornamental elements and, more significantly, in the  relative proportions of the divisions. 28. Frontispieces without tripartition: TS E 249 (Baghdad, 1302-1308); TIEM T.486 (dated 1318). Frontispieces with  tripartition: ex-Christie’s October 20th-22nd 1997, lot 237 (Iraq,  1307-1308);  Cairo,  Dār  al-kutub 72  (Hamadan,  dated 1313). 29. e.g., BnF Supplément Persan 1332 (ca. 1360) and Arabe 3368  (1366-1368);  Walters  Art  Museum  W563  (India, attr. 1420-1440).

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 21

Though this option is quite frequent in the Islamic West, it seems to be more difficult to find manuscripts  which follow this rule in the Near or the Middle East. However, the Qurʾan copied in 1188-1189 for the Ghurid sovereign Ghiyāth al-Dīn Muḥammad (r. 558-  599/1163-1203) has this particularity.30 Later, we encounter this solution in Mamluk productions as illustrated by a manuscript kept in the Vever Collection.31 Of course this division into four parts was also, as a structure, particularly convenient for another kind of manuscript produced in Egypt, i.e. the Evangeliaries, copied for the Coptic community which lived then in convivence32 with the Muslims. Such books can indeed have a similar construction as they open with an illuminated double frontispiece but have a b simple frontispieces at the beginning of the other Gospels.33 With regard to the Middle East region we have to admit that we have never yet seen a Qurʾan divided into arbāʿ that are clearly indicated by frontispieces from the 13th and the 14th centuries. After putting forth an initial classification of these  frontispieces based on their functionality, we may approach them according to the rules governing their construction, as shown in the graphic 1. a/ This design, which results from what we can call ‘simple’ or ‘canonical’ tripartition, is the most usual in this Qurʾan (seventeen cases for thirty-three frontispieces)34 (figure 2). This construction is common in the productions of the Near and Middle East of the 13th and the 14th centuries. b/ The extension of both vertical bands into the two ʿunwāns (i. e. title panels) thus creating four cornerpieces35 to be illuminated, as observed in only two frontispieces.36

30. Tehran, Iran Bastan MS 3496. This Qurʾan is also noteworthy for it is one of the first with a Persian interlinear  translation. For a recent bibliography, see Blair (2006), 24-25, 214; see also F. B. Flood’s publication in the present  volume. 31. S86.0025 (14th century). 32. This term belongs to the history of the countries surrounding the Mediterranean Sea where Christians, Jews and Muslims cohabited. It describes the peculiar situation of these three communities living altogether in a territory under Islamic domination. This word entered the dictionary of the Académie française in 2004 but was already used in other romance languages. 33.  BL Or. 1327  (Syria  or  Egypt,  1334);  BL Add. 11856  (Syria  or Egypt, 1337): Reeve (2007), 55. 34. Folios 1v-2r, 2v-3r, 59v-60r, 78v-79r, 97v-98r, 251v-252r, 274v-275r, 289v-290r, 305v-306r, 323v-324r, 341v-342r, 358v-359r, 395v-396r, 406v-407r, 415v-416r, 433v-434r and 471v-472r. 35. Square zones at the intersection of the horizontal and vertical bands. 36. Fols. 208v-209r and 270v-271r.

a

b

c

d

Graphic 1 - Double pages format

c/ This option, which concerns seven double frontispieces,37 is characterized by its tripartition and by vertical lateral bands placed on each side of the central zone. d/ Fols. 490v-491r offer a unique example of the fusion of the vertical and horizontal bands, thus creating a unified frame around the central zone  of each page. This formal solution encountered long-term success for it was common during the 12th century and was still in use in the 15th century, as illustrated by the Qurʾan W563 kept at the Walters Art Museum in Baltimore.38 The last frontispiece (fols. 530v-531r) (figure 3) offers a structure that has nothing comparable with all the precedents: tripartition is here abandoned in favor of a text written in clouds, with the illuminator filling in the residual spaces with a grid pattern. To  the best of our knowledge, no manuscript produced during this period in the Persian or Arabic context presents such a construction. Since no less than sixteen double frontispieces have a framing pattern, the former taxonomy can be scrutinized in a close examination of the borders around each illuminated field. These borders can be  as fine as a simple stroke (or line) or thick enough to  receive an elaborate ornamentation, either geometric or ‘vegetalized’.39 However, once more, we shall not

37. Fols 3v-4r, 21v-22r, 40v-41r, 233v-234r, 377v-378r, 452v453r and 510v-511r. 38. http://art.thewalters.org/detail/39217/quran/; see Simon  Rettig’s publication in this volume. 39. By this term we designate an ornament inspired by the botanical realm but presenting unnatural aspects.

c

22 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 23

Figure 1 – Gwalior Qurʾan, double-frontispiece, fols. 1v-2r. [© By courtesy of the Aga Khan Museum]

24 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 25

Figure 2 – Gwalior Qurʾan, fols. 78v-79r. [© By courtesy of the Aga Khan Museum]

26 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 27

Figure 3 – Gwalior Qurʾan, fol. 530v-531r. [© By courtesy of the Aga Khan Museum]

28 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

tackle their fillings in this section but we shall instead  follow a structural approach that focuses on their functionality, i. e. on the way they build a relationship between both pages of each double frontispiece. It so happens that borders in this manuscript are constructed in one of four different ways so as to play different roles as depicted in the graphic 2:

b’

The borders can be ‘linking’ as depicted in case “a”;40 these are the most numerous in this Qurʾan (about 60 %). Denis Muzerelle – in the field of Western  codicology – calls this type of border ‘demi-encadrement’.41 This solution, which was as much in favour in the Persianate world42 as in the Mamluk Sultanate,43 was also adopted in the Indian world, even if it was some time later. The Qurʾan of Gwalior can by

no means be considered as a unicum for this region since it can be illustrated by at least two other manuscripts.44 The second most frequent type of border is the ‘lateral’ one, as in case “b”45 (figure 2); they represent  about 25 % of the occurrences. Whereas the presence of just one single part of a ‘virtual border’, placed in the outer margin – i.e. the side of the fore-edge — is uncommon in the Arabic art of the book, it can be found in some medieval Turkish or Central Asian copies of the Qurʾan.46 In this respect it is worth pointing out that relationships between the Turkish and the North Indian worlds had existed since at least the 9th century.47 In the third case the borders can be called ‘isolating’;48 they represent 19 % of the occurrences. Each c’ page is thought out as an individual unit,d’ preventing the eye from joining them together. Indeed each illuminated surface is visually separated from its counterpart. Such a construction is seldom used in Mamluk Qurʾans49 but it can be considered a veritable leitmotiv of Persian illumination, as illustrated, among many other examples, by the Majmūʿa al-rashīdiyya copied for Rashīd al-Dīn (BnF, Arabe 2324; dated 1307-  1310) or by the frontispieces of all the ajzaʾ of Ilkhanid imperial Qurʾans copied for Ghāzān or Öljeitü.50 This type of border reached India as well as shown by the ornamentation of a manuscript copied in 1394 (BnF, Persan 36; figure 4). Finally, only one instance of an ‘upper-and-lower’ border occurs in the whole manuscript (fols. 274v275r). This option, quite rare in Mamluk manuscripts,51 seems to be absent from the Persian territories. However, it enjoyed a certain success in India, showing here once again that the arts of the book in the Indian sultanates were not simple derivates of Persian models.52

40.  Fols. 1v-2r;  59v-60r;  233v-234r;  251v-252r;  323v-324r;  358v-359r; 377v-378r; 415v-416r; 490v-491r. 41. This ‘demi-encadrement’ is “un ensemble de trois bordures occupant les marges supérieure, extérieure et inférieure de la page, susceptible de former, avec le demi-encadrement symétrique de la page qui fait face, un encadrement complet de la double page”: Muzerelle (1985). 42. For some examples, see: TIEM K 430 (1338-1339), TIEM K 452 (1340-1341), a dispersed Qurʾan copied in Maragha (1338): James (1988), no. 61, ex-Sotheby’s, October 15th 1998, lot 20 (attributed to Tabriz or Maragha, 702/1303), TS K.3 (Tabriz or Baghdad, 709/1309-1310), TS EH 248 (Tabriz, 715/1315), TS K 452 (Baghdad, 741/1341-1342), Al-Sabah collection LNS 44 MS (Iran or Iraq, 14th century). 43. This kind of border was in use throughout the 14th century, from Baybars’ Qurʾan (BL Add. 22406-13, ca. 1306), see James (1988), no. 1 to the manuscripts copied under the rules of Shaʿbān  (r. 764-778/1363-1376)  and  Barqūq  (r. 784-791/1382-1389). This is by far the preferred ornamental solution during this century.

44.  BnF Supplément Persan 1332 (ca. 1360); Keir Collection,  Pl.1 (ca. 1400-1450): Clause-Peter (2007), no. 19. 45.  Fols. 78v-79r; 289v-290r; 305v-306r; 452v-453r. 46. e.g., a dispersed Qurʾan kept in the following collections: John Rylands Library 760-773; LACMA 73.5490; CBL 1606;  CBL 1630 (Anatolia or Central Asia, ca. 1335). 47. Among many references about the relationships between Turkish and Indian culture during the medieval period, see for instance: Bosworth (2001); Siddiqui (1998), 110-124. 48.  Fols. 341v-342r; 433v-434r; 471v-472r. 49.  Fraser, Kwiatkowski (2006), no. 27 (ca. 1298-1310); Cairo,  Dār al-Kutub 10 (dated 1372): see James (1988), no 32. 50. See for instance: ibid., nos. 39, 40, 42 (no. 41, copied in Baghdad, is an exception to this rule) and Richard (1997), no. 12. 51. We can nevertheless mention ex-Christie’s, April 23rd 2002, lot 12 (dated 1401). 52. For examples among Indian manuscripts see W563 (Walters Art Museum) and Sam Fogg (2000), nos. 14, 36-37.

a’

b’

c’

d’

Graphic 2 - Framing formats

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 29

Figure 4 – BnF Persan 36, fol. 1v (India?, 1394). [© Bibliothèque nationale de France]

30 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Double frontispieces losing their symmetry Some double frontispieces deserve special attention for they show a noticeable lack of symmetry, and this lack can be classified into two main types  originating from different causes. First, there is what can be called ‘filling dissymmetry’: elements such as  flowers  or  arabesques  may  avoid  respecting the  rules of symmetry on both pages or even on a single page if their own construction is based on a symmetrical axis. Then we have dissymmetry that can affect the frames. This ‘macrostructural dissymmetry’ can be generated by three different factors. First, it can come from a difference in the models of frames as in the case of fols. 289v-290r and 305v-306r. To our knowledge, this option is very rare and we can in fact only indicate one other manuscript following this kind of dissymmetry, a Mamluk Qurʾan, dated 1429 and copied in Cairo.53 Second, it can be explained by a difference in the dimensions of the frames54 (figure 2). It may  have been brought about by the lack of attention of a copyist who did not place the lines symmetrically or, even worse, did not respect the same ruling on both pages. The illuminator had therefore no choice but to construct the frames imposed by these constraints. In effect, the illuminated frame must be adapted to a preponderant respect of the written area rather than to the simple conservation of visual symmetry. Finally, it can result from a difference in the positioning of both pages (fols. 274v-275r; 377v-378r; 415v-  416r; 471v-472r).

ʿunwān: structuraL and ornamentaL diversity ʿunwān for juzʾ As already mentioned, the tripartition of the illuminated frontispieces led to the existence of upper and lower bands devoted to ʿunwān. Our purpose here is not to focus on the different scripts such as muḥaqqaq or kufic but to examine the ways in which  these bands are structured. As usual, several formal solutions have been adopted: a/ A vegetal background receives a cartouche in which the text is placed, consequently marking three informal zones, with the cartouche surrounded by two vegetal zones55 (figure 2). The first occurrences 

53. CBL Is. 1496: James (1980), nos. 37, 53. 54.  Such as in fols. 3v-4r; 78v-79r; 97v-98r; 208v-209r; 289v-  290r; 323v-324r; 358v-359r; 452v-453r; 510v-511r. 55. Fols. 2v-3r, 21v-22r, 59v-60r, 78v-79r, 135v-136r, 153v-154r, 171v-172r, 274v-275r, 339v-340r, 356v-357r, 404v-405r, 412v-415r, 433v-434r, 471v-472r.

of this model in Mamluk illumination seem to have occurred around 1320 and from that point on never left the stage.56 Persian illuminators, and especially Ilkhanid ones, often used this construction57 but it seems to be less frequent in the Indo-Persian arts of the book.58 The ʿunwān undergoes a multi-part division, in clearly drawn geometric compartments.59 The construction is geometrically symmetrical, with two or three compartments placed on each side of the central cartouche, and the whole pattern finally surrounded by two corner-pieces. This organization can be seen in Turkish and Persian contexts during the 14th century60 and made a discrete appearance around 1320 in the Mamluk production, however without the occurrence of the corner-pieces. In fact, the Qurʾans illuminated at the very end of the 14th century by Ibrāhīm al-Amīdī, whose nisba suggests he originated in the city of Amid (nowadays Diyarbakır in Anatolia), are among the only  ones to benefit from such complexity.61 b/ As for Indo-Persian manuscripts, we may note a division  among  several  clearly  identified  zones  (fols. 208v-209r  and  270v-271r)  (figure 5).  This  occurs for instance in the Pl.1 bihārī Qurʾan of the Keir Collection (early 15th century).62 c/ In some double frontispieces, the title is directly written on the background (figure 6).63 This kind of construction was widely used in the Near East at the beginning of the 14th century and during the 15th century with an apparent collapse during the second half of the 14th century.64 As for the

56. CBL 1481 (720/1320): see James (1988), no. 9 and BnF Arabe 12 (753/1353): http://gallica.bnf.fr/ark:/12148/ btv1b84192173 are some examples among many others. 57. Among many cases, see the Qurʾan  of  Rashīd  al-Dīn,  TS EH 248, (715/1315): James (1988), no. 46 or the Qurʾan kept at the BL Or.1339 (undated). 58. The Walters Art Museum Qurʾan W563, albeit more recent, is however a notable illustration of such a case in this precise context. 59. Fols. 208v-209r, 270v-271r and 323v-324r. 60. The dispersed Qurʾan of Öljeitü (Baghdad, ca. 1307); Khalili  collection QUR671 (Mossul, 1300); TIEM 540 (Mossul, 1307);  BnF Arabe 2324  (Tabriz,  1307-1310);  BnF  Supplément  Persan 95 (Shiraz, 1317); BnF Arabe 6073 (Turkey?, 14th century, http://gallica.bnf.fr/ark:/12148/btv1b84061661);  Khalili collection QUR283 (Anatolia, 13th-14th centuries). 61. BL Or. 1009 (end of the 14th century): Lings, Safadi (1976), 56; Cairo, Dār al-Kutub MS 9 (Cairo, ca. 1367-1369): James  (1988), nos. 31 and MS 10 (dated 1372): ibid., no. 32 illuminated by al-Amīdī himself. His style, which can be  considered as the opposite of the ‘classical’ Mamluk style, is very likely impregnated with the ornamentation of Seljukid manuscripts. 62. Clause-Peter (2007), no. 19. 63. Fols. 3v-4r, 40v-41r, 395v-396r and 510v-511r. 64. As examples, see BL Add. 22406-13 (ca. 1306): James (1988), no. 1; BnF Arabe 1130 (dated 1337).

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 31

Persian production, this construction was adopted very early since it had been used at least since the 12th century65 and it continued to be largely favored in the Ilkhanid imperial copies.66 It is worth noting that this aesthetic option can coexist with the former one in the Ilkhanid production as attested by the case of Öljeitü’s Qurʾan, copied in Baghdad in 1307 for his mausoleum.67 A continued focus on the structural elements now brings us to examine the typology of the previously mentioned cartouches: a/  Some cartouches have a biconvex outline (figure 2).  Two-thirds of the cartouches (i. e. 19/34) have this outline, an option that became quite usual in the illuminated Mamluk Qurʾans as early as 133068 but was rather put aside in Persia where rectangular frames were preferred.69 b/ Some cartouches adopt a ‘kilīl arch’-like-outline:70 two frontispieces (fols. 208v-209r and 233v-234r) (figure 5) have cartouches whose extremities follow this typical Persian architectural pattern. Examples can therefore be found, but only rarely, in Persian illuminated manuscripts.71 c/  In one case, the cartouches adopt a profile we could  qualify as a reversed medallion ending with pendants, or ‘toronj sar toronj’ (fols. 97v-98r; figure 7).  The only analogous pattern we encountered is in an Indian Qurʾan attributed to the 15th century (Walters Art Museum, W563). d/ Another unusual cartouche outline can be seen in fols. 490v-491r. Even if this ‘morning star’-profile  cartouche (the morning star being a medieval weapon consisting of a spiked club resembling a

mace)72 seems to have undergone subsequent modification.  Some  analogs  participated,  even  rarely, in the ornamentations of Ilkhanid ʿunwān.73 e/ ‘Interlocked circles’ or ‘almond’ are used in fols. 270v-271r (figure 8). This original solution was  mainly appreciated in the Turkish area74 and its presence in the Gwalior Qurʾan could be explained by the secular exchanges between both cultures as briefly depicted in the introduction. This versatility in the outlines should not lead us to forget that five double frontispieces have no cartouche (fols. 3v-4r; 40v-41r; 116v-117r; 395v-396r;  510v-511r). In fact, this situation is rather frequent in the Mamluk production of the 14th century with a predilection during its first half.75 As for the Persian world, this lack of a cartouche can be considered as a long-term option for we know cases from the 12th century onwards. The Qurʾan of Bust, which was kept in use by the Seljuks and the Ilkhanids, had already followed this aesthetic.76

As for the ʿunwān of the suras, which of course are placed inside the text, they provide it with a structure but with a lower hierarchical importance. Their construction almost always follows a unique scheme, based on a tripartite division, the title of the sura being written in a cloud drifting over the mostly vegetal background of the central rectangle. However, for a very few cases, the illuminator only drew the central rectangle, so that the first words of these  suras are written on the same line as the title.

65. Khalili collection QUR87 (1175-1225). 66. This kind of construction was in use throughout most of the Ilkhanid period, from the manuscript TIEM 3548 (Baghdad, 1302-1308) to the Qurʾan TIEM K 452 (Baghdad, 1340). 67. James (1988), no. 40. 68.  See Khalili collection, QUR317 (1329): James (1992), no. 41;  Al-Sabah collection, LNS 47 MS (746/1346): id (1988), no. 22; Süleymaniye Kütüphane, Ayasofia 3606 (1354);  BL Or. 848 (attr. 1382-1399): ibid., no. 35. 69. See nevertheless for instance CBL 1494 (Baghdad, ca. 13001335);  Khalili  collection  QUR182  (Shiraz,  ca. 1336-1357;  figure 16) and TS K.3 (Tabriz or Baghdad, 1310). 70.  The ‘kilīl-arch’ can be defined as a trilobed arch with  a short broach, presenting lateral lobes in the form of a segmental arch, basket arch or pointed arch; the upper  lobe, originally a corbel, varies from a simple drop curve to a pointed one. The etymology of this terminology is not obvious. It may stem from the Persian word kalīl, which is related to something low and short: Laleh (1989), 467 and n. 120, 502. Laleh also gives some English equivalents

(‘segmental arch’, ‘shouldered arch’ or ‘broken headed arch’) with afferent references. A pertinent example can be seen in the Majmūʿa al-rashīdiyya, by Rashīd al-Dīn (BnF, Arabe 2324), copied  707-710/1307-1310 at fol. 284v. Newman (2001). TS E.H. 74 (Iraq, 1295) and TIEM K.452 (fols. 1v-2r) (Baghdad, 1340-1341) are note-worthy examples. See BnF Arabe 6043 (Turkey, 14th c., http://gallica.bnf.fr/ ark:/12148/btv1b84061661); LACMA 73.5490 (Anatolia or  Central  Asia,  ca. 1335);  ex-Sotheby’s,  October  16th 2002, lots 6-9 (Anatolia or Central Asia, attr. 1300-1350); Khalili  collection, QUR228 (Iran or Anatolia, ca. 1280-1320). TIEM 450  (Cairo,  713/1313):  James  (1988),  no. 6;  Khalili  collection QUR349 (attr. 1330-1340): James (1992), no. 36;  Khalili collection, QUR580 (1330-1350): ibid.,  no. 42;  ex-Christie’s, April 23rd 2002, lot 12 (1401). BnF Arabe 6041  (Bust,  555/1111-1112);  Iran  Bastan  Museum, Tehran, 4277 (1286), CBL 1494 (beginning of the 14th century), dispersed Qurʾan copied in Maragha in 1338: James (1988), no. 61.

ʿunwān for suras

71. 72. 73. 74.

75. 

76. 

32 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 5 – Gwalior Qurʾan, fol. 208v (detail). [© By courtesy of the Aga Khan Museum]

Figure 6 – Gwalior Qurʾan, fol. 40v (detail). [© By courtesy of the Aga Khan Museum]

Figure 7 – Gwalior Qurʾan, fol. 99r (detail). [© By courtesy of the Aga Khan Museum]

Figure 8 – Gwalior Qurʾan, fol. 270v (detail). [© By courtesy of the Aga Khan Museum]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 33

Figure 9 – Gwalior Qurʾan, fol. 190r. [© By courtesy of the Aga Khan Museum]

34 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

in the marGins... For now, we only take into account the ornamentation of the frontispieces and the written text but the margins as well receive a varied ornamentation which can be classified into three main categories:  vignettes, semi-vignettes and more massive elements we call ‘yokes’77 by analogy with this textile piece. But, before focusing on the typology of these elements, we can notice that each double-frontispiece is given at least one of them, the possibilities of distribution being quite diverse. In seventeen occurrences78 (e.g., figure 2),  there  is  only  one  marginal  vignette  on  each page, which is a very usual solution in the Arabic and Persian arts of the book. In contrast, we have to notice that there is no occurrence of two marginal vignettes per page, which is quite surprising insofar as it was widely used in the whole Islamic world. Three marginal vignettes per page are used in two occurrences (fols. 21v-22r and 406v-407r). This composition was fashionable in the Mamluk production around 1345.79 A so-called ‘yoke’ is added to two marginal  vignettes  in  four  occurrences  (fols. 2v-3r;  97v-98r;  189v-190r;  274v-275r)  (figure 9).  This  aesthetic option can be observed in some Ilkhanid manuscripts80 and is still present in the Indo-Persian world in the 15th century as demonstrated by the ornamentation of the Indian Qurʾan of the Walters Art Museum (W563, attr. 1420-1440). In the folios 452v453r, the external margins are ornamented with two semi-vignettes and one vignette. Formal analogs can be found in Turkish or Central Asiatic manuscripts of the 13th and 14th centuries81 but are absent from the Mamluk production. Then, two semi-vignettes and one yoke are in use in two double frontispieces (fols. 305v-306r, 395v-396r) and, finally, two semivignettes with one yoke and two vignettes only occur in fols. 289v-290r.

77. These ornaments share an extended segment with the central field of text, just as textile yokes are in close contact  with the garment they decorate. They differ from the marginal vignettes and semi-vignettes by this very kind of linkage to the central field and by their bigger size. 78.  See  the  fols. 1v-2r;  3v-4r;  40v-41r;  59v-60r;  78v-79r;  208v-209r;  251v-252r;  270v-271r;  323v-324r;  341v-342r;  358v-359r;  377v-378r;  415v-416r;  433v-434r;  471v-472r;  490v-491r; 510v-511r. 79.  TS 138.M5  (Damascus?  741/1341):  James  (1988),  no. 21;  Al-Sabah collection, LNS 47 MS (746/1346): ibid., no. 22. 80.  For instance Öljeitü’s Qurʾan (Leipzig, Albertina, XXXVII, K1 (Baghdad, 1307) and TS EH 243 (Baghdad, 1307); BnF  Supplément Persan 906 (?, ca. 1308); National Library of  Russia, Dom. 255 (?, ca. 1330). 81. Khalili collection QUR228 (Iran or Anatolia, ca. 12801320); BnF Arabe 6073 (Turkey, 14th century).

A short commented typology of these elements, yokes and marginal vignettes, can be briefly displayed. The yokes present in this Qurʾan have either a semicircular82 or a triangular83 outline (e.g. figure 10). The semicircular ones are found in Mamluk84 as well as in Indo-Persian contexts.85 The same parallels can be made for the triangular ones.86 It is also worth noting  that  the  yokes  and  the  central  field  can  be  either separated or unified by the borders, which leads  once more to a new formal option. In the Persian area, such patterns were at first used alone (figure 11)87 but were later associated with two marginal vignettes,88 both solutions being then sometimes able to coexist in the same manuscript. In the Near East, they participated in the illuminations of manuscripts from Ayyubid period (564-648/1169-1252)89 and went on during the Mamluk one.90 The vignettes can also first be classified according to their main outlines, which offer quite a large number of formal possibilities (figure 12): spinning  top-like – i.e. almond-like with two highly-sharpened ends –, pseudo-circular, with an undulating border only in fols. 40v-41r, hastate91 and pearshaped – which are the more numerous for they are present in eleven double frontispieces. The imagination of the illuminator allowed him to propose two different constructions for this last type of vignette.

82. The semicircular yokes are used on three double pages (fols. 153v-154r; 189v-190r and 305v-306r) and on one page  from a dissymmetrical double frontispiece (fol. 290r). 83. The triangular yokes are used on four double pages (fols. 2v-3r; 97v-98r; 171v-172r and 395v-396r) and on  the other page of the preceding dissymmetrical double frontispiece (fol. 289v). 84. See the Qurʾan of the Library of Congress (1-84-154.16 from the 14th century). 85. As in the Walters Art Museum Qurʾan (W563). 86.  They can be seen in the Mamluk production (Cairo, Dār  al-Kutub MS 81 (Cairo, 734/1334): James (1988), no. 17 and BnF Arabe 1130 (1337) as well as in the Indo-Persian context (Walters Art Museum, W563). 87.  TIEM  507  (Baghdad,  1286);  Khalili  collection,  QUR29  (Baghdad, 1282). 88.  BnF Supplément Persan 906 (?, ca. 1318); for an Inju Qurʾan, see Khalili collection, QUR182 (Shiraz, attr. 1336-1357;  figure 16). 89. See a Kalīla wa Dimna that may have been copied in Egypt, ca. 1230 (BnF Arabe 3465, fol. 35v) (http://gallica.bnf.fr/ ark:/12148/btv1b84229611). 90.  BL Or. 3025 (712/1312); TIEM 450 (Cairo, 713/1313); Cairo,  Dār al-Kutub MS 81 (Cairo, 734/1334): James (1988), no. 17;  BnF Arabe 1130 (1337) and TS 138.M5 (Damascus, 741/ 1341): ibid., no. 2. 91. Although quite rare, this type can nevertheless be found in the Indo-Persian production as illustrated by BnF Persan 36 (dated 1394; figure 4).

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 35

a: Semi-circular yoke (fol. 154r) Figure 10 – Marginal yokes of the Gwalior Qurʾan. [© By courtesy of the Aga Khan Museum]

b: Triangular yoke (fol. 395v)

36 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 11 – Ethnographical Museum, Ankara, MS 10118 (Maragha, 738/1338). [© Ankara Etnoğrafya Müzesi, photograph: F. Chaigne, 2011]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 37

a: Half-vignette (fol. 448v)

b: Pear-shaped vignette (fol. 97v)

d: Spinning top-like vignette (fol. 433v)

c: Pseudo-circular vignette (fol. 117r)

e: Hastate vignette (fol. 453r)

Figure 12 – Marginal vignettes in the Gwalior Qurʾan. [© By courtesy of the Aga Khan Museum]

38 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Some are indeed inscribed in a pear-shaped stroke92 but most of them result from the superposition of a circle and a curvilinear triangle. Some parallels to these pear-shaped vignettes in the Mamluk production can be find in prestigious Qurʾans illuminated by Ibrāhīm al-Amīdī whose nisba, as mentioned above, suggests he originated in Anatolia.93 It is worth noting that some of these vignettes are vertically oriented, calling to mind some Shirazi94 and Mamluk models.95 This model was still in great favor in the WAM Qurʾan. Half-vignettes are only present in five double frontispieces96 but, once more, this low number of occurrences does not inhibit variations, both in their outlines97 and in their fillings. Analogs for comparisons  are quite easy to find for such elements occurred in  many cultural areas, including the Indian,98 Central Asian,99 Persian and Turkish100 spheres, ancient and contemporary. The final point to examine for these structures is  their linking, or their articulation, with the border or with the main field and, once again, heterogeneity is  the  key  word.  Four  different  modes  can  finally  be  characterized according to the relationship between the border and the vignette. Some of them are ‘isolated’, which is to say that there is no physical link between the vignette and the border. This option, which is the more frequent for it occurs in ten double frontispieces,101 concerns the smallest vignettes. Next, some marginal ornaments can be linked to the border, a pattern in which either a fragment of arabesque

92.  Fols. 1v-2r; 270v-271r; 510v-511r; all showing once again  very diversified ornaments. The first series is limited to  the Mamluk production of the second half of the 14th century with golden arabesques on a blue background; the  second group presents the same floral ornament as the  pear-shaped vignettes, the background being however gilded rather than black, and the last family is obviously a later addition. 93. BL Or. 1009 (end of the 14th century): Lings, Safadi (1976), no. 84; Cairo, Dār al-Kutub, MS 9 (Cairo, attr. 1367-1369):  James (1988), no. 31 and MS 10 (dated 1372) illuminated by al-Amīdī himself: ibid., no. 32. 94. CBL 1494 (Iran, beginning of the 14th century). 95.  Cairo, Dār al-Kutub, MS 15 (Cairo, 776/1374): ibid., no. 34. 96.  Fols. 153v-154r; 289v-290r; 305v-306r; 395v-396r; 452v-453r. 97.  In fact we can find half-vignettes or half-yokes of any of  the types we mentioned previously. 98.  BnF Persan 36,  fols. 1v-2r  (dated  1394)  (figure 4);  ex-Christie’s, 29 April 2005, lot 505 (13th century) and ex-Christie’s, May 1st 2001, lot 19 (Gujarat, ca. 1450). 99. AKM00310 (ca. 1300-1350). 100. The so-called ‘Qarmatian’ Qurʾan, AKM00256 (Iran, ca. 1150);  BnF Arabe 6073  (Turkey,  14th century), BnF Supplément Persan 906 (?, ca. 1318) and Khalili collection QUR228 (Iran or Anatolia, ca. 1280-1320). 101. Fols. 40v-41r, 78v-79r, 208v-209r, 251v-252r, 323v-324r, 341v-342r, 377v-378r, 433v-434r, 471v-472r, 510v-511r.

(fols. 1v-2r and 452v-453r), a fleuron (fols. 490v-491r)  or a lanceolate pattern102 carries out the contact. Third, they can be just ‘detached’103: a slight space has been left between the marginal element and the border of the main field. In the case of ‘joined’ ornaments,  the vignette’s base is tangential to the border, both sharing a rectilinear segment. The rare occurrence of this construction (only four cases, in fols. 2v-3r, 59v-60r, 305v-306r, 395v-396r) might be explained by the simultaneous presence in these folios of quite large yokes. Some tenuous reasons sometimes seem to govern the choice of one of these modes. When the available space is small, for instance because it is mostly occupied by a yoke, or if the vignette is quite big for it is bordered by leaves, the favorite links are quite naturally the detached and joined ones. Furthermore, when there is no border but simple braids and lines, the lanceolate pattern is preferentially chosen. Finally, it seems difficult to underscore some general rules in  Mamluk or Persian productions. Only peculiar cases can be pointed out, such as the use of the ‘linked’ mode by the mediation of a fleuron under the rules  of Shaʿbān (r. 764-778/1363-1376) and Barqūq (r. 784-  791/1382-1389) but some examples are known as early as the 1350s.

a sPace for the text The jadwal After this analysis of the macrostructures of the field  of  the  illumination,  let  us  go  into  the  space  devolved to the script. In the simple pages of text it is framed by what is called a jadwal in arts of the Persian and Indo-Persian book (figure 13). Here, it is  of a rare complexity because it is composed of seven elements, namely, from the inside to the outside: a double red line, a simple blue line, a golden braid, a black line, a red line and a blue line. To our knowledge, no other Indo-Persian manuscript nor even another manuscript from the Near or the Middle East displays such a succession of frames from the same period. The jadwal of the other IndoPersian manuscripts are usually composed of three or four elements on average, chosen among a red line, a blue line and a golden braid; as with Mamluk  or Ilkhanid manuscripts, they mostly present only one or two lines. In addition, in this Qurʾan, the simple

102. This articulation, which can be considered as a direct avatar of some vignettes, is only present in pages surrounded by a line and a braid but without borders. 103. As in the cases of fols. 21v-22r, 97v-98r, 270v-271r, 274v275r, 289v-290r, 415v-416r.

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 39

Figure 13 – Jadwal in the Qurʾan of Gwalior. [© By courtesy of the Aga Khan Museum]

pages of text as well as the double pages of frontispieces are framed by a simple blue line, which is characteristic of Qurʾans written in bihārī script (figure 13).  However, here, it seems to be a later addition related to the marginal vignettes among which many are not original.

Backgrounds The illuminated programme of the double pages of frontispieces is of a matchless visual richness based on the combinatorial diversity of its ornaments and, first of all, its backgrounds, which consist of settings  and their possible adventitious elements. The majority of the settings – 24 out of 33 – are made of a red square pattern104  drawn  with  a  fine  tracer and in a very light depiction that is set at 45° to the frame (figure 2). Whereas in a number of other  manuscripts,  the  whole  interlinear  field  is  covered  in a single movement, here, every interlinear space is considered without regard to the others. Moreover, they are drawn freehand, not with a ruler, and

104. This covering is found as a background decoration for heads of suras in the simple pages of text.

with a quill of rather irregular ink flow. All this produces a rather unconvincing aesthetic result. After the backgrounds marked out in squares come the plain ones – 7 out of 33 – including the white ones – in four cases out of seven the paper base is left plain –, then the celadon backgrounds – 2 out of 7 (fols. 208v-209r; 305v-306r) – and finally only one black  (fols. 116v-117r). The red square-pattern background is one of the main features of Mamluk105 and Iranian106 illuminations. It can be frequently found in the Indo-Persian context as attested by Indian Qurʾans from the BnF (Arabe 7260)107 and the British Library (Add. 554851;  figure 14)  both  attributed  to  the  16th century, that of the Walters Art Museum of Baltimore (W563)108 or else the Sam Fogg Qurʾan (fol. 4v).109 105. It was present since the beginning of the 14th century, as illustrated by the Qurʾan CBL 1457 (attr. 1306-1311). 106. See a dispersed Qurʾan produced in Baghdad, ca. 13021308 (Iran Bastan Museum, Tehran, 3548; TS, EH 249). 107. The complete manuscript is available on http://gallica. bnf.fr/ark:/12148/btv1b84272593. 108. About this manuscript, see Simon Rettig’s article in this volume. 109. Sam Fogg (2000), lot 14. This manuscript is now kept in the Museum of Islamic Arts, Doha, MS.259.2003.

40 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 14 – BL Add. 5551, fol. 136r (India, 16th century). [© British Library]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 41

This is in fact a very old ornamental element which had already adorned the backgrounds of the folios of the Seljuk Qurʾan produced in the city of Bust in 505/1111-1112 (BnF Arabe 6041).110 The black background is also interesting in that it was seldom used. As far as we know, only one example of it can be found in the Indian Qurʾan of the Walters Art Museum (fols. 8v-9r). Among the backgrounds that decorate the double pages of the Qurʾan of Gwalior, we can also notice a single case of what we are in the habit of calling, rather incorrectly, ‘Chinese water’, for this pattern was already used in Sasanian ornaments. In fols. 377v378r (figure 15), scales, constituted by a succession  of concentric arcs drawn with a quill and red ink, are organized in opposite directions from one page to the other: in fol. 377v, they are directed towards the upper-right corner, while they face the lower-left corner in fol. 378r. Numerous parallels can be drawn with the Iranian ornament, as much within the Ilkhanid sphere (figure 11) as in the Inju one, both  important suppliers of this decorative solution.111 However, anecdotally, we can note the existence of at least one Mamluk manuscript using the Chinese water pattern, a remarkable work of art from the late 1360s.112 Of course, this does not prove any transmission from the Mamluk sphere to the Indo-Persian zone but only testifies to the favour this ornament received at  the end of the 14th century in all the Muslim East. Up to this time, it remained extremely rare in the Indo-Persian context, as Éloïse Brac de la Perrière has already noted,113 because only the Indian Anthology from the British Library (Or. 4110) dated 1401 makes use of it in fol. 152, and yet not for the background of the central area but as decoration for the heads of ʿunwān.114 The presence of this very pattern in this Indo-Persian Qurʾan copied in Gwalior may indeed attest to the role of Iranian models for the conception of the ornamentation of this manuscript. It should be noted that this pattern was also used in the Jain arts of  the  book  for  figuring  water  in  some  miniatures  just as in Arab or Persian paintings. A folio from a

110. The complete manuscript is available on http://gallica. bnf.fr/ark:/12148/btv1b8433296d. 111. The following manuscripts present numerous examples: the Maragha Qurʾan, dispersed between the collections of Dublin, Boston and Ankara: see James (1988), no. 61;  the Tabriz Qurʾan, housed at the BL and at the CBL of Dublin (Is. 1469);  the Demotte Shāhnāma (BnF Supplément Persan 1946; CBL Is. 1471); the Majmūʿa al-rashīdiyya of the BnF (Arabe 2324, fol. 9). 112. Dār al-Kutub, Cairo, MS 7 (Cairo, attr. 1367-1369): James  (1988), no. 29. 113. Brac de la Perrière (2009), 15. 114. See an illustration of this folio in Brac de la Perrière (2008), pl. 46.

Gujarati Kalpasūtra,115 which may be dated to the 15th century and now belongs to the collections of the National Museum in New Delhi, thus provides a meaningful example.116

Ornaments over backgrounds This first level of diversity, created by the accumulation of differing backgrounds, is heightened by their numerous fillings. Linear constructions: branches and arabesque foliage The first pattern, which occurs in 17 out of the 33 backgrounds, is made up of branches meandering through the lines of text (figure 9). Both extremities  have opposite orientations, one ascending and the other descending. This construction generates an impression of continuity between all these plant elements; this impression is increased by the alternating  directions of the leaves and flowers. Most of the time,  the branches are arranged regularly, but it is noteworthy that a few counterexamples occur, suggesting the possibility of several periods of completion. For instance, fols. 40v-41r and 415v-416r present a significant diversity in the patterns and the techniques used to decorate the facing folios. A rare but beautiful color gradient is introduced thanks to the botanical treatment  on  fols. 189v-190r  (figure 9).  But  the  greatest  surprise is provided by the diversity of techniques – such as black, blue or red inks, wash tint – and plant patterns, such as lotus blossoms, long petals flowers  or ‘fantasy’ flowers. These backgrounds filled with branches are unknown in Mamluk illuminations but take part in a tradition that seems to have begun in the Ilkhanid realm, in a Qurʾan copied in Tabriz in 733/1334,117 and continued in the Muzaffarid Shiraz (figure 16)118 and then, in a softened version, in Timurid

115. See the essay of Nalini Balbir in the present volume. This copy of the Kalpasūtra, which can be attributed to the 15th century, is considered to be one of the masterpieces of the Jain pictorial tradition in Western India. It was produced on paper, according to the palm-leaf format (potlī), and presents some pictorial innovations that characterize the influence of Persian culture at a turning  point in Indian painting. Most of its 201 folios are kept at the Library (bandhar) of the temple of Devasano Pado at Ahmedabad. Its colophon is now incomplete but however gives the identity of family members of a Deva minister who was certainly an influential Jain working for the administration: Britschgi, Guy (2011), 32. 116. This manuscript is illustrated in Doshi (1985), no. 8. 117. See n. 111. 118. Most of all, see the manuscripts ex-Christie’s, 24 April 1990, lot 161 (Shiraz, 1356-1358) and Khalili collection QUR182 (Shiraz, attr. 1336-1357).

42 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 15 – Gwalior Qurʾan, fol. 377v. [© By courtesy of the Aga Khan Museum]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 43

Figure 16 – Khalili collection, QUR182, fol. 27r (Shiraz, attr. 1336-1357). [© Nour Foundation. Courtesy of the Khalili Family Trust]

44 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

anthologies copied for Iskandar Sulṭān (BL, Add. 27261)  and for Bāysunghur (Museum für Islamische Kunst,  Berlin, I.4628) during the 15th century. Whilst the backgrounds  filled  with  arabesque  foliage  became  commonplaces in Mamluk illumination as early as the first half of the 14th century, only two double pages of the Qurʾan of Gwalior received this ornament. In addition, both frontispieces adopt a very distinct formal aspect. Thus, in fols. 251v-252r, the arabesque scrolls alternate with lines of text written in clouds that preserve a high level of readability. In contrast, in  fols. 452v-453r,  the  ornate  fields  are  completely  covered with arabesque foliages, constraining the writing to make its way with difficulty through the  scrolls. Quite surprisingly, a close analog has been found in a manuscript copied in 666/1267 for an Ayyubid Emir of the Syrian city of Hama.119 As no other example of such a pattern seems to be known in the Indo-Persian production, we can assume that it was not considered as really attractive by Indian illuminators.

background ornament coexist in this manuscript and we can presume that several artists took part in its creation. We can for instance notice that the semis does not always have the same density on two facing pages, which may suggest the intervention of two artists. But later additions may have been made here as already noted for other cases. The interesting point is that this ornamental pattern is never found outside of the Indo-Persian sphere and can be considered as typical of Qurʾans copied in bihārī script.121 Finally, in one case only (fols. 274v-275r), the background is decorated with little bunches drawn with a quill and black ink. They are made of grass tufts flanked by leaves. As far as we know, only two  Indo-Persian parallels exist: the Indian WAM Qurʾan and the manuscript ex-Christie’s, 1 May 2001, lot 21, attributed to fifteenth-century India. Likewise a single example can be found in the Near-Eastern production, a Qurʾan dated from the 1350s (Khalili collection, QUR187).122

Adventitious geometric elements: dots and crosses

structurinG the iLLuminated sPace

The second favorite option for the decoration of the backgrounds is used in 11 cases out of 33. It consists of a bed of small blue or red crosses and circles (figures 2, 15), sometimes associated with secondary  elements, which are quite homogeneously disposed but drawn roughly, without the help of any instrument. Chromatic composition plays an important role in generating harmony since the red and blue patterns systemically alternate – when they are both present. Even though this pattern seems closely associated with Indian productions, no precise parallel has been found, with the exception of manuscript BL Add. 5551 (figure 14) and ex-Christie’s, 9 October 2009, lot 249,  both attributed to the 16th century. Adventitious ‘vegetalized’ elements: ‘Grass-tufts’ and bunches Seven of the thirty-three backgrounds receive a dispersion of a typical pattern of Indo-Persian ornamentation, called ‘grass-tuft’120 because of its obvious likeness with this actual plant element (figure 2). They  are most of the time monochromatic, blue, red, black or bottle-green. As with the preceding case, the semis is generally heterogeneous and disposed in staggered rows. Such a layout generates an alternative between the blue and red colors, which denotes chromatic carefulness. Several types of construction for this

119. Süleymaniye Kütüphane, Ayasofya 3114. 120. See below for the origin of the pattern.

This Qurʾan in bihārī script does not avoid one of the characteristics of this group of manuscripts: its decoration results from a strong spatial division through the assistance of more or less thick but clearly recognizable ornaments. The diverse lines, ribbons, braids, pseudo-strapworks, borders, and vertical bands and banners create a succession of frames or pseudoframes and can be considered de facto as structuring patterns. Most of the time, Qurʾans in bihārī script rather attest to the outsourcing of these structuring elements in the marginal space of the folio.123 But this manuscript is characterized by the insourcing of this structuring in the space of the illuminated field.  By this means, the Qurʾan of Gwalior is connected with several other Qurʾans in bihārī script, all of them very beautiful and of high quality.124 These structuring elements, belonging to the geometrical vocabulary, therefore assure the distribution, development and spatial limitation of the botanical ornaments. Clearly freeing itself from Indian examples, the workshop that developed this decoration looked to Near and Middle Eastern models. Some pieces of these structuring elements are studied below.

121. Brac de la Perrière (2008), 163. This ‘grass-tuft’ pattern is however absent from the WAM Qurʾan. 122. James (1992), no. 44, 180-183. 123. See for instance the following Qurʾans: BnF Arabe 7260 and ex-Christie’s, 1 May 2001, lots 17 and 18. 124. For instance BnF Arabe 7260 (frontispieces) (see n. 113), Keir collection, Pl.1 (ca. 1400-1450) (see n. 37), Khalili collection QUR 237 (15th century), Walters Art Museum W563 (ca. 1420-1440).

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 45

Borders The borders found along all or part of the illuminated field  always  contain  a  filling,  whatever  their  format among the formats indicated above. Two main sorts of borders have been identified: stylized botanical or ‘vegetalized’ borders and borders consisting of a blue line associated with a semis of ornaments. Vegetalized borders Although it is once again based upon a surprising diversity,  this  filling  mostly  consists  of  vegetalized  foliage, made up of elements drawn from the botanical realm – curls of arabesques or garlands. They can be observed in ten double pages, i. e. about fifty-nine  percent of the items.125 As we have already noted, several types of vegetalized border can be distinguished. But even if the motifs used for background filling change from one double page to another, they  nevertheless use similar tones: the majority contain an ultramarine background on which there is a frieze of golden elements. The dullness of this decoration is interrupted by other motifs – such as vegetal medallions – which introduce new shapes and colors each time. The vegetalized borders filled with rows of vegetal  medallions on an ultramarine background are the most  frequent  within  this  manuscript;  the  vegetal  medallions are arranged in compositions in which they alternate with other motifs: floweret-like forms  (fols. 78v-79r: figure 2; 116v-117r), ‘eye of peacock’-like-  motifs (fols. 153v-154r), lotus (fols. 274v-275r; 358v-  359r) or even heart-like motifs (fols. 452v-453r). There is an obvious connection between some of these borders with Mamluk examples. The Qurʾan W563 and, in another style, the manuscript BnF Persan 36 (figure 4) both testify to the favour of such a model  in Indian territory. A very original solution is the use of vegetalized borders with a black background and garlands of flowers on three double pages (fols. 289v-  290r; 305v-306r; 433v-434r). No parallel is known in the  rest of the Arabic, Persian or Indo-Persian production. The vegetalized border filled with golden arabesques on an ultramarine background is present as such in fols. 1v-2r only (figure 1). It is frequently  found in the Mamluk context but its typology evolved throughout the 14th century in the Syro-Egyptian production. The kind of border that is used here is closer to the one that was fashionable in Egypt around the 1350s – quite an old model – in well-made manu

125. Fols. 1v-2r;  78v-79r;  274v-275r;  289v-290r;  305v-306r;  323v-324r; 358v-359r; 433v-434r; 452v-453r; 471v-472r.

scripts (figure 17). These borders are quite numerous  in the Indian126 and Persian productions. Beside these borders filled with vegetalized foliage,  the most frequently used models are those composed of lotus petals127 (figures 9, 15) and those consisting  of a simple blue line with scattered dots.128 These borders follow not only the illuminated fields but also some marginal vignettes and yokes. Though still rare in Mamluk illumination, the border of lotus petals is more usual in Persian manuscripts, in particular Ilkhanid and Injuid ones129. It is found alongside inaugural shamsas and sometimes even surrounding full page compositions; it also decorates marginal  medallions.130 This recurring motif reappears in Injuid illumination in the 1350s, certainly a source of inspiration for the Indian artists.131 Semis as a border The last type of border, which consists of a simple blue line and a semis of dots, is much more atypical;  in the Mamluk context of the 14th century, borders consisting of a simple blue line are either bare or decorated  with  fleurons  but  on  no  account  with  a  semis of dots. This also seems to be the case for the Persian production. This decorative solution may be shared by other Qurʾans in bihāri script, in particular that of the Library of Congress (1-84-154.21).132

Geometric framing lines Other elements, such as ‘pseudo-strapworks’, braids and diverse lines, generate the structuring patterns of the illuminated field and create rhythm for the illuminated decoration as well. Braids The diversity of these elements adds even here to the visual wealth of the illuminated decoration even if the smooth blue one is by far the most common. 126. See for instance BnF Persan 36 (figure 4) and Walters Art  Museum W563. 127. Fols. 59v-60r;  233v-234r;  251v-252r;  341v-342r;  377v-378r;  415v-416r; 490v-491r. 128. Fols. 3v-4r; 21v-22r; 97v-98r; 208v-209r; 289v-290r; 406v-  407r; 510v-511r. 129. They have been used before in manuscripts such as the Qurʾan attributed to Yāqūt al-Mustaʿṣīmī, dated from 1289,  and currently housed at the BnF (Arabe 6716): http:// gallica.bnf.fr/ark:/12148/btv1b8433294k. 130. See for instance: ex-Sotheby’s, 18 October 1995, lot 42 (Iran or Mesopotamia, 681/1282) and BnF Supplément Persan 95 (Shiraz?, 717/1317). 131. Wright (1996-1997). 132. See the manuscript on http://international.loc.gov/cgi-bin/ ampage?collId=ascs&fileName=070/ascs070.db&recNum=0.

46 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 17 – BnF Arabe 12, fol. 2v (Syria?, 753/1353). [© Bibliothèque nationale de France]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 47

It’s also present in another Indian Qurʾan (BL Add. 554851;  figure 14)  attributed  to  the  beginning  of  the  16th century and also in the bihārī Qurʾan of the Library of Congress. Note that such patterns are seldom found in both Mamluk and Ilkhanid productions.133 Some other types of braids present in the Qurʾan of Gwalior are pretty common in the Near and Middle Eastern productions: the golden braid and the white braid – sometimes decorated with pearls –134 and the uses of these ornaments echoes rather archaic practices in the other decorative traditions, going back sometimes to the very beginning of the 14th century. Brown, red and black braids, which all perform the function of underlining macrostructures – such as the heads of the sura and the central zones – seem almost endemic to the Indian subcontinent. Excepting the red braid, of which we find a very early use in  the Seljuk Qurʾan of Bust,135 no parallel can be found in the Near and Middle Eastern productions. Thus, although numerous motifs stemming from the Islamic tradition are adopted in the ornamentation of this Qurʾan, these more unexpected elements testify to the existence of other sources of inspiration for the Indian artists. Blue line

several manuscripts, such as that of the Walters Art Museum (W563), the ex-Sotheby’s, 6 April 2011, lot 190 or the Supplément Persan 1332 of the BnF.136 Other patterns, as original as this one, may punctuate the path of the blue line. All these elements form a very ethereal, slightly indented network, which contrasts with the prosperous character of the decoration associated with it. This former example concerning the framing lines shows how innovative the Indo-Persian art of illumination can be, far beyond the simple transfer of motifs developed in other places. Jain manuscripts are a little less original and take up once more the framing blue lines and their finery of fleurons, circles or dots,  from the end of the 15th century and especially at the beginning of the 16th century.137

amonG so many botanicaL Patterns The same creative inspiration presided over the production of the decorative speech of fillings; the  diverse elements, which mainly belong to the botanical register, testify to a highly developed freedom concerning the composition and the development of the artistic project. The diversity is still appropriate, both in the palette of the simple motifs that enter their composition – palmette, lotus and cotton or ‘fantastic’ flowers that we may suppose to have come simply from the imagination of one or several artists – and in the variety of the composite patterns – arabesques, branches and bunches – studied below.

According to a practice shared by the whole Islamic manuscript production, the illuminated field  of the double pages of the frontispieces is enclosed by a blue line that may or may not bear some ornamentation. It must not be confused with the blue line used as a border that is described above. This blue line goes along the main border, which is most of the time vegetalized. Beside the simple blue line devoid of any decoration and the blue line with a semis of dots, some folios present a most surprising ornamentation, with Indianlike accents and based on the chevron pattern and its various manifestations. We may find an eloquent  example on the first double page (fols. 1v-2r, figure 1):  the line supports chevrons whose summits are topped with fleurons with basal petals and a kind of quiff in the center. This kind of fleuron is foreign to the  Mamluk tradition, which prefers by far the simple fleuron and a lighter ornamentation for the line (figure 17). On the other hand it is rather common in the Indo-Persian context to decorate the margins of

As the ultimate pattern in the Islamic art of the book, the palmette remains in this context the key element of the ornamental programme of this manuscript. It is the preferred filling for vegetalized borders  and for the backgrounds of the cartouches and of the central zones, a classical topology that is practiced in the rest of Islamic manuscript production. Without notable originality, we also find it in some marginal  vignettes and yokes as well as in vertical banners. It generally adopts the ‘canonical’138 shape, which adorns Islamic works from the beginning of the 12th century. The palmettes can be uni-, bi-or trifid,  rolled up or not, ‘crossing’ or not. Yet next to this canonical palmette stand particular forms that, even

133. Nevertheless, this braid sets out the illuminated field of  the frontispiece of a Qurʾan produced in 1303 in Tabriz or Maragha (ex-Sotheby’s, 15 October 1998, lot 20, fol. 2). 134. Notice nevertheless the absence of the golden braid flanked with red braids, a real leitmotiv in the ornamentation of the other Qurʾans in bihārī script. 135. See n. 25.

136. See Brac de la Perrière (2008), pl. 50 for illustrations. 137. See especially the manuscripts housed at the Victoria and Albert Museum IM.7-193 and IS.46.39-159, both from the end of the 15th century. 138. This term refers here to the construction rules of the arabesque and to diverse motifs such as palmettes as they are described by Kühnel (1949), 7-8.

Palmettes

48 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

stylized, seem to be inspired by geometry more than by botanical patterns. They have no parallel in the Middle- and Near-Eastern arts of the book nor even in the Indo-Persian ones. They are used nevertheless quite frequently in this manuscript, mostly following the model shown below in a terminal position on the  stalk  with  a  semicircular  shape;  it  very  often  decorates the vegetalized borders in which it forms a scalloped line on the outside (figure 18).

Figure 18 – Palmettes, Gwalior Qurʾan, fol. 79r (detail). [© By courtesy of the Aga Khan Museum]

It can be compared to a model in the Indian WAM Qurʾan (figure 19), which, although a little different  from a formal point of view, shares its position and its function. Another ‘palmette’ can also be compared with its Mamluk counterpart (BnF Arabe 12, 753/1353;  figure 20).

two associated colors. As for the so-called palmettes ‘with a drop of water’ which are very rare in this manuscript (fols. 452v-453r), their surface is heightened with a colored water drop. The use of these last two categories of bicolored palmettes finds a certain echo in the Near- and Middle-  Eastern productions as well as in the Turkish one. Bicolored palmettes indeed are a solution that was particularly appreciated by Anatolian, Mamluk and Ilkhanid craftsmen, who used them for the decoration of the backgrounds of the central zones and the heads of ʿunwān. They remain nevertheless old-fashioned ornaments at this end of the 14th century, the other traditions having forgotten them for a very long time and generally preferring chiseled palmettes or other botanical forms over them. However no parallel was encountered in the Indo-Persian production. These palmettes perfectly illustrate the aesthetic paradox of this Gwalior manuscript, resting on a surprising diversity of patterns and models, the latter being able to be geographically contemporary and close but also very distant in time and space.

‘Grass-tufts’

Figure 20 – Palmettes, Gwalior Qurʾan, fol. 453 (detail). [© By courtesy of the Aga Khan Museum] (left) / BnF Arabe 12, fol. 2v (detail) [© Bibliothèque nationale de France] (right).

Other ornaments, such as the ‘grass-tufts’ that mark out the backgrounds of seven double pages (e.g., figure 2), have a much more easily recognizable  origin because they are characteristic of the IndoPersian production. This pattern is “constitué d’une touffe de tiges verticales à l’extrémité supérieure pansue et réfléchie, disposées en triangle sur un trait  horizontal”.139 Many models coexist in these manuscripts and some of them let us suppose that many artists have worked for the decoration of the Qurʾan of Gwalior. These grass-tufts constitute a frequent motive for the Qurʾan in bihārī script but can also be found in other works, as attested by the BnF Supplément Persan 1711140 or the Pseudo-Shāhnāma141 of the New York Public Library.142

This diversity – not to say extravagance – is also appropriate for the surface treatment of these palmettes; several solutions coexist in this manuscript,  sometimes on the same page, even in the same structure. Beside the ‘chiseled’ palmettes, the treatment of which recalls metalwork, and which go back to the 12th century, other palmettes exhibit a very simple surface treatment for which the effort has been focused on chromatism, whereas the centre of the unicolor palmettes, surrounded by a large golden line, is painted with a single flat tint (red, green or blue).  No element of comparison was found. As for bicolor palmettes, the surfaces are brought out with a single color painted with a pronounced gradation or with

139. Brac de la Perrière (2008), 163. 140. Ibid., pl. 49 for illustration. 141. In fact, this manuscript is an Intikhāb-i Shāhnāma, which is a selection of episodes extracted from the Firdawsī’s  Shāhnāma. This manuscript is dated 906/1501 and was copied in Jaunpur (New York Public Library, Spencer Coll., Indo-Pers. Ms. 1): Brac de la Perrière (2008), 279, no. 15 and pl. 24. 142. These patterns are characteristic of relatively late Indian productions but we may see here a manifestation of the grass-tufts that are sprinkled over the ground in Muzaffarid miniatures (see for instance Keir collection, Ham Richmond III, 13). This motif was transferred to the Mamluk art of the book as illustrated in a copy of the Kashf al-asrār produced ca. 1400 (Süleymaniye Kütüphane, Lala  Ismail 565): Haldane (1978), 53.

Figure 19 – Palmettes, Qurʾan W563, fol. 3v (detail). [© By courtesy of the Walters Art Museum]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 49

Denomination

Istance from the Qurʾan of Gwalior

Near-Eastern counterparts

Middle-Eastern counterparts

a

d

f

Jain counterparts

Lotus with central flower or petal

Lotus without central flower  or petal b

g

Lotus with crossed petal c

e

h

i

Figure 21 – Lotus flowers in the Gwalior Qurʾan and their counterparts. a: fol. 404v; b: fol. 233v; c: fol. 395r [© By courtesy of the Aga Khan Museum]; d: Dār al-kutub, ms. 8, fol. 3, Egypt ?, 1356 (Cairo: O’Kane [2006], 102-103); e: Detail from a Qurʾan box, Egypt or Syria, 1300-1350, MIA 183; f: QUR182, fol. 2r, Shiraz, ca. 1370 [© Nour Foundation, by courtesy of the Khalili Family Trust];  g: CBL 1494, fol. 13v, Iran beginning of 8th/14th  c.; h: CBL, Per 109, fol. 8v, Iran or Iraq, ca. 1300 [© The Trustees of the Chester Beatty Library, Dublin]; i: Kālakācāryakathā, ca. 1425, Gujarat, private coll., Doshi (1985), cat. 15.

Lotus flowers Another pattern weaves narrow decorative links with some cultural areas that are more or less contemporary with the Delhi Sultanate: the lotus flower.143 It appears mostly as an interlinear decoration, in compositions such as branches or bunches (figure 9);  but the lotus’s value can also be increased by its isolated use in structures framing the central zone. Here the floral variety is still appropriate, as we can  see on the board above (figure 21). The lotuses with a central floweret or petal found  in this very manuscript are also frequently found in the ornamentation of Iranian and Syro-Egyptian spheres from the 14th century (see also figure 22). Another form of lotus, devoid of any central element, also occurs in this manuscript but it is noteworthy that this variety is much rarer in the two other cultural

143. See Yves Porter’s article in this volume.

areas previously mentioned. Another type of lotus flower is structured around the intertwining of two  petals. This model remains very unusual in the rest of the Islamic production. It can be seen nevertheless from the years 1309-1310 on a glass lamp with the name of the Mamluk sultan Baybars II (r. 708-709/ 1309). Since an example is found in a manuscript of the Saʿdī’s Kulliyāt copied by 1300 but of unknown origin (CBL 109), this pattern could be of Persian origin. The open flower is not the only form of lotus that  can be found. The bud is very frequent, in particular in the  borders  of  the  illuminated  field, where they  constitute an alternative that is visually very close to the ‘eye of peacock’ motif (figures 9, 15). It is quite  interesting to notice that two forms coexist. The first, ‘classical’ in the Islamic context, is usually orange, lanceolate, with a peduncle and a blue fleuron (figures 9, 15). The second form, present on  three double pages (fols. 59v-60r, 341v-342r, 415v416r), seems to originate from the Jain vocabulary (figure 23); it presents a yellow centre, a red tip and  a hatched intermediate zone. Whereas this model is

50 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Figure 22 – BnF Arabe 2324, fol. 129r (Tabriz, ca. 1307). [© Bibliothèque nationale de France]

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 51

really rare in the rest of the Indo-Persian production,144 it is much more frequent in Jain miniatures, where it is used to portray petals of lotus flowers or leaves.145

constitution of these patterns seems to be largely indebted to the artist’s overflowing imagination while their treatment and their scenography participate in the creation of a convincing reality. Finding parallels for such patterns remains a delicate issue; only some examples from the Timurid period147 allow us to think that it is not simply a creation of Indian workshops but a re-using of Persian patterns revised in the light of Indian fashion.

Cotton flowers

Figure 23 – Lotus bud, Gwalior Qurʾan, fol. 59v (detail). [© By courtesy of the Aga Khan Museum]

In the Indo-Persian and Jain contexts of production, lotus buds not only follow the same chromatic scheme but also present a transitional zone between the red zone and the yellow one; this transitional zone is pictured with red and yellow hatching. This technique of hatching seems relatively old-fashioned, as attested by an example from a manuscript produced during the 11th century in Rajasthan.146

In contrast with these imaginary floral motifs, the cotton flowers have the capacity to suggest a real botanical variety (figure 24). It is quite a rare pattern in the Islamic arts of the book; nevertheless it can be found in folios of Iranian and Mamluk manuscripts: on the Persian side, it appears on tiles from the Takht-i Sulayman built during the 13th century, and it flourished in a Tabrizi Qurʾan dated 1334.148 Some decades later, it fills folios from the late Mamluk period.149 On the other hand, this cotton flower remains distant from the Indian, Indo-Persian and Jain productions. Let us note however that the Ilkhanid and Mamluk forms adopt a more simplistic aesthetic than the one in use in the Gwalior Qurʾan, from the models in order to follow certainly a more Indian aesthetic.

Floral and foliate patterns Another feature of the decoration of this manuscript is the very wide representation of floral and foliate patterns that differ aesthetically from the botanical elements traditionally used in the Islamic arts of the book. These floral patterns mainly fill the central illuminated field, sometimes closely associated with arabesques but most often in compositions such as branches. Finally, we may note that they are also used in bunches in the cartouches. In this manuscript a number of flowers, flowerets and other branches are juxtaposed, treated mostly with shading and shaping but nevertheless making reference to no existing, or recognisable, plant. The

Figure 24 – Cotton flower, Gwalior Qurʾan, fol. 97r (detail). [© By courtesy of the Aga Khan Museum]

144. The WAM Qurʾan W563 seems to be the only noticeable example (see fig. 11). 145. See for instance a copy of the Nemidarśana Iñānasālā Palitānā dated 1382: Nawab (1980), figs. 229 and 230 and some folios of the Devasano Pado Kalpasūtra, from the 15th century: Doshi (1985), no. 8, 50-51. 146. Nawab (1980), vol. 1, pls. 80 and 82.

147. See for instance one of the two folios from a Shāhnāma, now dispersed and probably produced in Herat ca. 14251450 (Lisbon, Calouste Gulbenkian Foundation, M.66A and M.66B): Lentz, Lowry (1989), nos. 44 a-b, 128-129. 148. Ex-Sotheby’s, 27 April 1982, lots 207 and 208. 149. TIEM 445, fol. 281v (Cairo, ca. 1370): Lings (2005), no. 110 and CBL 5193 (fol. 2r) (dated 853/1449).

52 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

comPosite veGetaL structures Except for grass-tufts, the last floral elements are  nearly never used separately but take part in the creation of much more complex patterns, usually used in the Arabic, Persian and Indo-Persian arts: arabesques, branches and bunches. Arabesques are certainly the most frequent composite form used in this very manuscript. Their different forms can be associated with a topology. For instance, the arabesques with medallions and other botanical alternating forms are localized in the borders of some double pages (e.g., fols. 1v-2r) (figure 1). In the Near-Eastern production, this kind  of  filling  flourishes  in  manuscripts  from  the  1350s  (figure 17)150 while parallels are less numerous in the Persian production, with Iranian artists seeming to prefer simple arabesques.151 Another example is provided by the centered symmetrical arabesques constructed with alternating medallions and botanical elements and which offer a solution for decorating circular structures such as marginal vignettes (e.g., figure 12, fol. 117r). Though  they are far away from the Mamluk ornamental vocabulary, they are used in prestigious Ilkhanid manuscripts and most of all in Injuid productions.152 This diversity, created by these various typologies, is increased by playing with variations on their chromatic format and with the variety of the incidental elements – leaves, flowers and flowerets, and so on.  Concerning some other arabesques, it is quite difficult  to establish a parallel with an Iranian or Egyptian inheritance and they may in fact be seen as later additions. This is the case for the purple arabesques in fols. 21v-22r. Careful observation showed that these are clearly later additions and may betray the intervention of Jain artists.153

150. See  for  instance  BnF Arabe 12  (753/1353);  Cairo,  Dār  al-Kutub 7 (attr. 1367-1369): James (1988), no. 29, BnF Arabe 5841-5846 (ca. 1382-1399): see Arabe 5844 on http:// gallica.bnf.fr/ark:/12148/btv1b84192299). 151. Notice however the following manuscripts: ex-Sotheby’s 15 October 1998, lot 20 (Tabriz or Maragha, 1303), TIEM 541 (Mossul, 1310) and Leipzig Albertina xxxvii, K1. 152. Ghāzān’s Qurʾan (Baghdad, 1302-1308), Öljeitü’s Qurʾan for Sultaniyya (Baghdad, ca. 1307), the Majmūʿa of Doha (Tabriz, 1311-1312) (Museum of Islamic Art, Doha, MA.006.98CH) and Khalili collection, QUR182 (Shiraz, 1336-1357). 153. This very sort of arabesque, purple on a gilded background, is found in a folio from a Kalpasūtra produced in the Gujarat ca. 1475 by the master of the Devasano Pado Kalpasūtra. This folio belongs to the Catherine and Ralph Benkaim collection, housed at the Metropolitan Museum of Art, New York.

Furthermore, the interlinear spaces and some elements of framing such as vertical bands and strips are filled with branches that, here too, demonstrate  a great deal of aesthetic meticulousness and attention to coloring and form. It is once again worth noting that this pattern and its use as an element for filling  in vertical bands and strips are frequently encountered in Iranian productions from the late 13th century and early 14th century.154 However, it is quite rare in the Mamluk manuscript production. Next to these composite plant patterns following a linear organization, the illuminated programme of this manuscript resorts a great deal to a plant motif with a centered and symmetrical organization: the bunch. In this Qurʾan, bunches generally consist of a  central  flower  with  long petals  flanked by minor  foliate or floral elements (figure 2). It is a very frequent  element of decoration because it occurs on twentytwo double pages as well as three simple pages (fols. 297v, 369v, 468r). Beside being used to fill in the  marginal vignettes and yokes, bunches are placed horizontally to decorate the heads of suras, which seems to be very frequent in the Indo-Persian context as shown by numerous examples.155 This decorative solution could have Injuid roots as illustrated by the Qurʾan QUR182 of the Khalili collection produced in Shiraz ca. 1336-1357 (figure 16). It can also be found in  the Muzaffarid context (BnF Persan 276). In the Iranian or Turkish manuscript Spencer Arab. 3 dated 1333 and housed in the New York Public Library,156 floral  bunches flank cartouches while remaining vertical.  It would be therefore interesting to explore Turkish productions to see if such patterns are also used.

a Pan-asiatic Pattern: the cLoud-coLLar The cloud-collar, which was originally a textile ornament in use throughout Asia, is composed of four fleurons  sometimes  alternating  with  four  smaller  ones disposed in gyration. It became an ornamental pattern that thereafter could be found on other supports such as architectural surfaces or paper, either in its entirety or in two, four or eight fragmented parts.

154. Examples of use can be found from the so-called ‘Qurʾan of Yaqūt’ (BnF Arabe 6716, Iraq, 1289) up to another Qurʾan produced in Iraq in 1307 (ex-Christie’s, 20-22 October 1992, lot 237). 155. See the Walters Art Museum Qurʾan W563, the Anthology BL Or. 4110 (Jaunpur, 1401): Brac de la Perrière (2008), pls. 46-47, the Qurʾan BL Add. 5548-51 (India, beginning of the 16th century; figure 14). 156. This manuscript is available on http://digitalcollections. nypl.org/collections/qurn-fragment-srat-al-naba#/ ?tab=about.

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 53

This latter pattern is used only once in this particular manuscript, as spandrels in the central fields on fols. 189v-190r (figure 9), but some antecedents among  the Persian manuscripts deserve to be mentioned: as an eighth part, i.e. in the very same way as in this frontispiece, we can observe such patterns on the dedication pages of the Mosul Qurʾan copied for Öljeitü or in a Majmūʿa al-rashīdiyya copied for Rashīd  al-Dīn.157 But the high symmetry of the cloud collar structure also easily explains that it can be used as a whole or as a half for illumination purposes from Iran to India, in Indo-Persian and Jain contexts. Finally, we would like to draw attention to the versatility of the cloud-collar pattern and its fragments as an architectural ornament throughout the whole Asiatic territory under Mongol domination, an ornament that can be seen on the lateral columns of Öljeitüs’  mihrāb in the Friday mosque of Isfahan or on roof tiles at the monastery of Zhwa-Lu (Tibet).158 To conclude, we have chosen not to point out the diverging aspects but rather to go beyond the prism of religious partition that is usually applied. For this purpose, we preferred to focus on the links between diverse cultural areas and religious communities that participated, directly or not, in the creation of the manuscripts copied in the Delhi Sultanate, by examining their structures and their fillings. The intricacy  of several periods of completion, influences, the obvious participation of several hands, among them a master and one or several assistants, all make it difficult to offer a correct or simple interpretation for  the ornamentation of this Qurʾan. This very profound hybridization occurs on many pages of this manuscript but has not prevented us from underlining some transferences of models, since obvious links between its illuminations and Mamluk and Iranian productions can be highlighted. The precise modes of transmission of these patterns still need to be fully elucidated. Two ways at least must be considered and may have acted in synergy: the first one implies a direct contact among  Indian craftsmen on the one side and Syro-Egyptian and Middle-Eastern productions on the other side. This mode of transmission could be based on embassy gifts but these can be considered as marginal because of their sporadic occurrences. In fact, it mostly implies the commercial circulation of goods such as textiles and ceramics159 and also manuscripts, for the similitudes between this Qurʾan and its potential analogs go much further than the simple quotations of isolated 157. Museum of Islamic Art, Doha, MA.006.98CH. 158. Chayet (1994), 214. 159. The influence of Persian tiles on Jain manuscript painting  was studied by Moti Chandra in 1949; he examined the  case of a Kalpasūtra from the Sarabhai Nawab Collection.

patterns but clearly imply the principles of composition that are typical of the Near- and Middle-Eastern arts of the book. These links are present both in the main principles of construction (e.g., double-page formats, framing formats), in the macro-structures (e.g., marginal vignettes and yokes, heads of the ajzāʾ) and in many details used in fillings (e.g., floral patterns  and structures, geometrical elements, and so on). Most of these transfers concern rather archaic elements, stemming from the Mamluk production from the years 1300-1330 and the 1360s as well as the Ilkhanid (654-754/1256-1353) and Muzaffarid (713-795/1314-1493) arts of the book.160 Concerning the Mamluk references, it is worth noting that they correspond to the two main economic and political peaks of the Sultanate, when the territory was respectively under the rules of Sultan al-Nāsir Muḥammad  b. Qalāwūn  and  al-Ashraf  Shaʿbān  (r. 764-778/1363-  1376). The reference to the first three decades of the  14th century might show how long artistic exchanges had existed between the Indian and Mamluk sultanates in the absence of direct proof. But it could also testify to the mediation of another means of transference that implies a crucial role for an extra-Islamic mediator, i.e. a group of Jain artists and craftsmen.161 From the 1350s onwards, Jain manuscripts copied in Gujarat and Rajasthan – belonging to the so-called ‘Western Indian style’ – show obvious signs of Islamization, or even better ‘Persianization’, in the way they are illustrated and illuminated. The Jain community may have therefore played the role of purveyor of this circulation of patterns thanks to its highly active patronage.162 Its painting, known for its strict conservatism, may have favoured the transmission of these archaisms from the Arabo-Persian elements towards the Indian painting of the 14th century. This community, whose main activity was trade, was strongly established in the Gujarat and was therefore in close contact with artifacts imported from the Middle East. Yet the ‘Western Indian style’ of Jain painting that flourished in the Gujarat and Rajasthan  gave birth to an endemic variation in the regions of Delhi and Gwalior, which can be called the ‘Northern style’. This observation allows us to postulate the existence of an ‘ascendant stream’ from the Gujarat to the region of Gwalior. Even more interesting is the fact that the style of Gujarati painting as practiced in

160. See Brac de la Perrière (2009). 161. See also Nalini Balbir’s article in this volume. 162. The most influential patrons belonged to the sect of Śvetāmbara, who considered that the patronage of illustrated or illuminated manuscripts and their donation to libraries (bandhars) linked to Jain temples were meritorious religious actions.

54 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

Delhi, Gwalior and their surroundings seems to have been very conservative at the turn of the 15th century;  in these areas it seems to have frozen in time the characteristics of earlier Gujarati painting, which had integrated Middle and Near-Eastern pictorial schemes, whereas this art followed its own evolution in the Gujarat. The milieu of the commissioners, or of reception, was therefore highly conservative and the attitudes of this milieu are clearly shown by the close study of the ornamentation of this manuscript. Thanks to this Qurʾan, we can gather more precise knowledge of how patrons considered their position within an Indo-Muslim continuum: they may have belonged to a group that simultaneously kept strong links with a pluri-secular Indian tradition and tried to keep in touch as much as possible with the great empires of the Near East, the Middle East and their immediate followers, such as the Mamluk Sultanate, the Ilkhanate, the Muzaffarids and other local Persian dynasties from the end of the 13th century or the beginning of the 14th century. By adapting some ornamental elements issued from Islamic art to Indian aesthetics, the Delhi Sultanate

demonstrates its role as an agent of hybridization of cultures and communities. This message was fully meaningful at the very end of the 14th century, when the territory of India was threatened by Tīmūr and,  more locally, Gwalior fell into the hands of the Rajputs. The ornamentation of this Qurʾan initiates a debate on the modalities of the integration and appropriation of Islamic aesthetics by communities growing through Islamic domination. The production of this manuscript implies a perfect assimilation of the aesthetic canons from many cultural areas, such as various Islamic lands and Indian regions. Its originality and its high quality necessarily required time for reflection and experimentation. To say the least, it would therefore be quite risky to affirm that this Qurʾan cannot be compared to any other one. As a matter of fact, Simon Rettig’s conclusions in his paper about the Qurʾan W563 of the Walters Art Museum and published in this volume confirm our point of view that the Gwalior  Qurʾan is by no means a starting point but the gorgeous completion of a process which still remains to be explored in detail.

bibLioGraPhy

References

Sources

aḥMad (Nazir) 1998 “Diplomatic relations between the sultans of Delhi and the Il-Khans of Iran”, in Nazir Aḥmad,  I. H. Siddiqui (dir.), Islamic Heritage in South Asian Subcontinent, Jaipur: Publication Scheme, 2 vols.

Ibn baṭṭūṭa [1958-2000] The Travels of Ibn Baṭṭūṭa, A.D. 1325-1354, translated by H. A. R. Gibb with revisions and notes from the Arabic text edited by C. Defremery, B. R. Sanguinetti, Cambridge: Published for the Hakluyt Society at the University Press, 5 vols. [1976] The Rihla of Ibn Battūta, translated by Mahdi Husain (Gaekwad’s Oriental Series CXXII). al‑Qūṣī (ʿAṭiya) [1976] Tijāra Miṣr fī al-baḥr al-Aḥmar munḏu fajr al-islām ḥattā suqūt al-khilāfa al-ʿAbbāssiyya, Cairo: Dār  al-nahḍa al-ʿarabiyya. al‑MaQrīzī [1958] Kitāb al-Sulūk li-maʿrifat duwal al-Mulūk, edited by Muḥammad Muṣṭafā Ziyādah, Cairo: Maṭbaʿa lajnat al-Taʾlīf wa al-tarjama wa al-nashr. rashīd al‑dīn al‑ṭabīb [1836] Histoire des Mongols de la Perse, texte persan, publié,  traduit  en  français ;  accompagnée  de  notes et d’un mémoire sur la vie et les ouvrages de l’auteur par Étienne Quatremère, Paris: Imprimerie Royale.

ahmad (Aziz) 1999 Studies in Islamic Culture in Indian Environment (Oxford India Paperbacks), Delhi: Oxford University Press. aker (Sābiha) 1978  “Ankara Etnografya Müzesinde Bulunan XIV.  Yüzyılca  Merāga’da  Yapılmış Vakif Kuran Güzleri”, in Vakıflar Dergisi XII, pp. 291-328. atil (Esin) 1981 Renaissance of Islam: Art of the Mamluks, exibition catalogue, Washington: Smithsonian Institution Press. Barrucand (Marianne) 1992 “Remarques sur le décor des manuscrits religieux hispano-maghrébins du Moyen-Âge”, in Afrique du Nord antique et médiévale: spectacles, vie portuaire, religions, Ve colloque international sur l’histoire et l’archéologie de l’Afrique du Nord, Paris: CTHS, pp. 235-248.

THE ORNAMENTATION OF THE GWALIOR QURʾAN • 55

1995 “Observaciones sobre les iluminaciones de Coranes hispano-magrebies”, in Jesus Bermudez Lopez (dir.), Arte islamico en Granada. Propuesta para un Museo de la Alhambra, exhibition catalogue, Granada: La Alhambra, pp. 164-172 and 194-195. Blair (Sheila) 2006 Islamic Calligraphy, Edinburgh: Edinburgh University Press.

doshi (Saryu) 1985 Masterpieces of Jain Painting, Bombay: Marg Publications. Flood (F. Barry) 2009 Objects of Translation: Material Culture and Medieval ‘Hindu-Muslim’ Encounter, Princeton: Princeton University Press. Fraser (Marcus), kwiatkowski (Will) 2006 Ink and Gold. Islamic Calligraphy, London: Sam Fogg.

Bosworth (Edmund C.) 2001 “Ghaznavids”, in Encyclopaedia Iranica X, no. 6, pp. 578-583 and http://www.iranicaonline.org/ articles/ghaznavids.

haldane (J. Duncan) 1978 Mamluk Painting, Westminster: Aris & Philips Ltd.

boyle (John a.) (dir.) 1968 The Cambridge History of Iran, vol. 5. The Saljuq and Mongol Periods, Cambridge: Cambridge University Press.

hamBly (Gavin R. G.), asher (Catherine B.) 1994 “Delhi Sultanate”, in Encyclopaedia Iranica VII, no. 3, pp. 242-250 and http://www.iranicaonline.org/articles/delhi-sultanate#pt1.

Brac de la Perrière (Éloïse) 2008 L’art du livre dans l’Inde des sultanats, Paris: Presses de l’Université Paris-Sorbonne. 2009 “Du Caire à Mandu: notes sur la transmission des modèles dans l’Inde des sultanats”, in F. Richard et M. Szuppe (dir.), Écrits et culture en Asie centrale et dans le monde turco-iranien, xe-xixe siècles / Writing and Culture in Central Asia and in the Turko-Iranian World, 10th-19th c., Paris: Association pour l’avancement des études iraniennes, pp. 333-358.

Jackson (Peter) 1975 “The Mongol and the Delhi Sultanate in the reign of Muḥammad Tughluq (1325-1351)”, in  Central Asiatic Journal 19, pp. 118-157. 2009 Studies on the Mongol Empire and Early Muslim India, Farnham/Burlington: Ashgate.

Brac de la Perrière (Éloïse), chaigne (Frantz), cruvelier (Mathilde) 2010 “The Qurʾan of Gwalior, kaleidoscope of the arts of the book”, in Treasures of the Aga Khan Museum – Arts of the Book & Calligraphy, Istanbul: Culture and Sakıp Sabancı Museum, pp. 114-123. brItschgI (Jorrit), guy (John) 2011 Wonder of the Age: Master Painters of India, 11001900, New York: Metropolitan Museum of Art. chandra (Moti) 1949 Jain Miniature Paintings from Western India, vol. 1, Ahmedabad: Sarabhai Manilal Nawab. chayet (Anne) 1994 Art et Archéologie du Tibet, Paris: Picard. clause-Peter (Haase) 2007 A Collector’s Fortune. Islamic Art from the Collection of Edmund de Unger, Chicago: University of Chicago Press.

James (David) 1980 Qurʾans and Bindings from the Chester Beatty Library. A Facsimile Exhibition, London: Al-Tajir World of Islam Trust. 1988 Qurʾans of the Mamluks, London: Alexandria Press/Thames & Hudson. 1992 The Master Scribes: Qur’ans of the 10th to the 14th century A.D., edited by Julian Raby (The Nasser D. Khalili Collection of Islamic Art, vol. II), London/Oxford: The Nour Foundation in association with Azimuth Editions and Oxford University Press. kühnel (Ernst) 1949 The Arabesque: Meaning and Transformation of an Ornament, Graz: Verlag für Sammler. laleh (Haeedeh) 1989 La structure fondamentale des arcs dans l’architecture saldjouḳīde de l’Iran, thèse sous la direction de Janine Sourdel-Thomine, Université ParisSorbonne, 3 vols. lentz (Thomas W.), lowry (Glenn D.) 1989 Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century, exhibition catalogue, Los Angeles: Los Angeles County Museum of Art.

56 • FRANTZ CHAIGNE / MATHILDE CRUVELIER

lings (Martin) 2005 Splendours of Qurʾān Calligraphy and Illuminations, Vaduz: Thesaurus Islamicus Foundation. lings (Martin), saFadi (Yasin H.) 1976 The Qurʾan, exhibition catalogue, London: British Library. Muzerelle (Denis) 1985 Vocabulaire codicologique: répertoire méthodique des termes français relatifs aux manuscrits, Paris: CEMI, 1985 et en ligne sur Institut de recherche et d’histoire des textes, CNRS, 2014, http:// codicologia.irht.cnrs.fr/accueil/vocabulaire. nawaB (Sarabhai M.), nawaB (Rajendra S.) 1980 Jaina Paintings, vol. 2. Paintings on Paper, Commencing from vs. 1403 to vs 1656 Only, Ahmedabad: Messrs Sarabhai Manilal Nawab. newman (Paul B.) 2001 Daily life in the Middle Ages, London: McFarland. rawson (Jessica) 1984 Chinese Ornament: The Lotus and the Dragon, London: British Museum Press. reeve (John) 2007 Sacred Books of the Three Faiths: Judaism, Christianity, Islam, London: The British Library. regourd (Anne) 2012 “Arabic documents from the Cairo Geniza in the David Kaufmann collection in the library of the Hungarian academy of sciences – Budapest”, in Journal of Islamic Manuscripts 3, pp. 1-19. richard (Francis) 1997 Splendeurs persanes: manuscrits du xiième au xviième siècles, Paris: Bibliothèque nationale de France.

salem (Sahar A. A.) 1993 “Commerce and one faith”, in Saryu Doshi, Mostafa El-Abbadi (dir.), India and Egypt: Thought, Art, and Trade, Bombay: Marg Publications XLV no. 2, pp. 92-111. Sam Fogg 2000 Islamic Manuscripts, cat. 22, London: Sam Fogg. siddiqui (Iqtidar H.) 1998 “The Turks and migration to Central Asia and India: An analysis of the historical information of the Turks and Turkestan in early Indo-Persian sources”, in N. Ahmed and I. H. Siddiqui (dir.), Islamic Heritage in South Asian Subcontinent, Jaipur: Publication Scheme. soustiel (Jean ), Porter (Yves) 2003 Tombeaux de paradis: le Shâh-e Zende de Samarcande et la céramique architecturale d’Asie centrale, Saint-Rémy-en-l’Eau: Monelle Hayot. vernay-nouri (Annie), guesdon (Marie-Geneviève) 2001 L’art du livre arabe. Du manuscrit au livre d’artiste, Paris: Bibliothèque nationale de France. wright (Elaine) 1996-1997 “An Indian Qurʾan and its fourteenth-century model”, in Oriental Art XLII, no. 4, pp. 8-12.

Websites http://www.jainology.org/projects/cataloging-of-jain-manuscripts/ http://digitalcollections.nypl.org/collections/ qurn-fragment-srat-al-naba#/?tab=about.

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION A preliminary study Nourane ben azzouna (University of Vienna, Department of Art History) Patricia roger-Puyo (IRAMAT, UMR 5060, CNRS)

Abstract The Gwalior Qurʾan is the first dated Qurʾan with a fālnāma, the first dated Qurʾan in bihārī and the first dated manuscript to be illuminated in Sultanate India. It is a milestone in  the history of the book in Sultanate India. This paper proposes a codicological investigation of the manuscript and a visual and spectrometric analysis of its decoration. Its purpose is to identify the major alterations brought about in the codex, in order to reconstruct as much as possible its original condition and formulate a few hypotheses about how it could have been produced. In addition to an important lacuna, the authors identify several types and most likely periods of intervention, ranging from restorations to additions and over- and re-paintings. They raise several questions. Is only a part of the decoration original? Why does this reveal pigments and dyes that are less typical of India than Iran? How many hands carried out the illumination, and how did the team collaborate?

Résumé Le coran de Gwalior: archéologie du manuscrit et de sa décoration. Étude préliminaire Le coran de Gwalior est le premier coran daté comportant un fālnāma, le premier coran daté en bihārī et le premier manuscrit enluminé daté de l’Inde des sultanats. Il constitue un jalon important de l’histoire du livre de l’Inde des sultanats. Cet article propose une étude codicologique du manuscrit et une analyse visuelle et spectrométrique de son décor. Notre objectif est d’identifier les transformations majeures apportées au codex afin de restituer,  dans la mesure du possible, son état d’origine et de formuler quelques hypothèses au sujet de sa production. Une lacune importante a été identifiée ainsi que différents types et, certainement, phases d’intervention, allant de la restauration au repeint et à l’ajout. Ces éléments soulèvent de multiples interrogations : le décor est-il seulement partiellement d’origine ? Pourquoi les pigments et les colorants qui sont utilisés sont-ils moins caractéristiques de l’Inde que de l’Iran ? Combien de mains ont-elles contribué à la réalisation des enluminures et comment l’équipe a-t-elle collaboré ?

Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 57-87

58 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

The Aga Khan Museum’s ‘Gwalior Qurʾan’1 is a muṣḥaf 2 with an interlinear translation and a fālnāma3 in Persian. It is provided with a colophon that states, also in Persian, that it was copied by a certain Maḥmūd  Shaʿbān, an inhabitant of the fortress of Gwalior, on  Monday Dhū al-Qaʿda 17, 801/July 21, 1399,4 which makes it the first known dated muṣḥaf with a fālnāma, but also the first known dated Qurʾan to be copied in the bihārī calligraphic style that was to become typical of Qurʾanic codices in Sultanate India, as well as the first  known  dated  manuscript  to  be  illuminated  in  Sultanate India.5 This illuminated decoration itself is also exceptional in comparison to other decorated Qurʾans from other regions at the same period, because of its impressive richness as much as its puzzling eclecticism. The Gwalior Qurʾan is, thus, a milestone in the history of Qurʾan and book production in Sultanate India. Besides the name of its calligrapher and the place and date of its copying, nothing, however, is known about the circumstances of its production. Gwalior, in 801/1399, was a small city dominated by a Rajput dynasty, but attached to the Tughluqid Sultanate of Delhi (1230-1414) that had been invaded, a few months earlier, by Tīmūr’s armies (in September 1398-January 1399). Thus, the colophon, and the quasi-absence of external sources about its exceptional context6 raise more questions than give answers about the manuscript, hence the need to turn to a kind of archaeological excavation of the codex and its decoration, in order to attempt to understand its history and its mode of production. Is it the work of an individual or a workshop, locals or foreigners, and what are the conditions of its patronage and its destination? The methodology used consisted of a codicological investigation of the manuscript and a visual and spectrometric analysis of its decoration. The visual analysis was carried out with the naked eye and with a binocular lens for the observation and high-magnification photography of the details of the pictorial  layer(s). The spectrometric analyses were performed with two techniques: Diffuse Reflectance Spectrometry 

1.

http://www.e-corpus.org/fre/notices/105296-TughluqQur-an-Gwalior.html, accessed May 30, 2013. 2. Muṣḥaf: “The name given to a complete text of the Ḳurʾān  considered as a physical object”: Burton, EI2 (1993), 668669. 3. Fālnāma: book of divination. 4.  The  colophon  indicates  “Monday  seventh  Dhū al-Qaʿda 801,” but the scribe probably mistakenly wrote seventh instead of seventeenth because Monday does not correspond to the seventh but to the seventeenth day of Dhū al-Qaʿda in 801. 5. Brac de la Perrière (2008), 133. 6.  Digby  (2004);  also  see  the  article  by  Johanna  Blayac  in  this volume.

(DRS)7  and  X-ray  fluorescence  spectrometry  (XRF),  two systems that do not require sampling, hence preserve the integrity of the manuscript, and have the further advantages of being portable, easy to implement in situ, and quick to use, which makes it possible to carry out a large number of measurements within a limited time.8 In the first phase of this project,9 our main objective was to identify the major alterations brought about in the manuscript, in order to go all the way back to its original state and formulate a few hypotheses about how it could have been produced. The results, still preliminary, are presented in this paper.

codicoLoGy

(appendices 1 and 2) In its current state, the Gwalior Qurʾan counts 550 numbered plus four unnumbered folios.10 In addition, it presents a gap corresponding to sura11 12 and the first and the beginning of the second verse of sura 13,  between fols. 221-222. The length of this lacuna, together with the type of quires (octonions or eightbifolio quires), and the fact that each page shows thirteen lines of text allow us to estimate the missing section at around thirteen folios. Thus, the original volume counted at least approximately 567 folios.

7.  In the visible, with an optic fiber spectrometer. 8. The analysis time is 100 seconds for DRS and 60-120 seconds for XRF. These are also proven analysis systems. DRS has been used since 1939: Barnes (1939), 208-214. XRF is a classical elementary analysis method. Both techniques have been used by the team of the Institut de recherche sur les archéomatériaux (CNRS), Orléans, for several years. A summary of the motivations and approach of this team is outlined in the introduction of Roger, Villela-Petit, Vandroy (2003), 155-170. The curves obtained by DRS in situ are compared to curves obtained on reference colored surfaces in the laboratory’s library. 9. Geneva, February 27 – March 2, 2012, and Orléans, April 1018, 2012. We would like to thank Heather Ecker, Head of Curatorial Affairs at the Aga Khan Museum; Benoît Junod,  representative of the Aga Khan Trust for Culture; Isabelle  Klinka-Ballesteros, Director of the Museums of Orléans;  Éloise Brac de la Perrière, Maître de conférence at ParisSorbonne University and Head of the Gwalior Qurʾan project; and Mathilde Cruvelier, PhD candidate at ParisSorbonne University, who made every effort to ensure that these two missions take place. We would also like to thank Le laboratoire Islam médieval, UMR 8167, for its financial support. 10. The four unnumbered folios appear after fols. 251, 289, 411, and 414 respectively. Similarly, the present numbering of the manuscript omits number 387, leapfrogging directly from 386 to 388. The numbering followed here corresponds to a correction of the numbers that appear on the manuscript and on its e-publication, integrating the unfolioed leaves and no 387. Thus, fol. 251bis becomes 252, etc. A mapping table is proposed in Appendix 1. 11. Sura: each of the 114 chapters of the Qurʾan.

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 59

It is copied on a good-quality, thin, transparent paper, which offers a good visibility of the paper stuff. This is very homogeneous and was moulded into very regular paper moulds.12 Most of the laid lines are vertical, which allows us to identify the format as a quarto. It is a medium format that measures 29 × 22 cm, but it seems to have been trimmed by at least 2 cm on each side,13 so that it must have been at least 33 × 24 cm originally. Finally, the paper was carefully sized and polished so that it shows a shiny and smooth finish. As with the paper, great care was taken to write and then decorate the manuscript. Although no traces of ruling are visible, the writing lines are regularly spaced. Most pages present thirteen lines, but the beginning of four suras – number 2 (al-Baqara) (fols. 3v4r), 7 (al-Aʿrāf) (fols. 143v-144r), 19 (Maryam) (fol. 274v275r), and 38 (Ṣād) (fols. 406v-407r), which roughly correspond to the four quarters of the Qurʾan –14 as well as the beginning of each juzʾ,15 are marked by a special layout, with five, seven, eight or nine lines.16 The Qurʾanic text is written in a neat bihārī, in gold, blue and red; the translation is in blue and red; and  both texts alternate according to an almost constant pattern. In the Qurʾanic text, the upper, middle and lower lines are in gold,17 the even lines (2-4...) in blue, and the odd lines (3-5...) in red,18 while in the 12. It is possible to distinguish between at least two different paper moulds, one with medium, the other with slightly thicker laid lines. The use of sheets originating from various moulds is commonly observed in medieval manuscripts. 13. The folios were trimmed by rounding the corners, which may be indicative of a particular place and/or time or fashion. 14. The Qurʾan is divided in sixty ḥizbs. The first ḥizb starts at the beginning of sura 2; the sixteenth at the beginning of  sura 7; the thirty-first in the middle (verse 75) of sura 18  – the third double illuminated page, at the beginning of sura 19, thus corresponds to the first sura following the  beginning of the third quarter. Likewise, the forty-sixth ḥizb begins at the end (verse 145) of sura 37. Thus, the fourth double illuminated page, at the beginning of sura 38, corresponds to the first sura after the start of the  fourth and last quarter of the Qurʾan. Several researchers have hypothesized that the choice of these suras could have been linked to a particular symbolic value, but this issue has to be further investigated. 15. Juzʾ: a division corresponding to one-thirtieth of the Qurʾan. 16.  Most of the special-layout pages present five lines, with  a few exceptions with seven (fols. 359v-360v) or eight lines (fols. 274v-275r and 305v-306r). Two double-pages stand out due to an asymmetrical layout with eight lines on the verso and nine lines on the facing recto (fol. 289v290r and 323v-324r). One more exception is fol. 116v, which shows no more than four lines, but this folio was largely refurbished. 17. The only exceptions are fols. 208v and 452v-453r, where gold was replaced by black ink. 18. This is the case in pages with thirteen, eight and nine lines and some pages with five lines (fols. 22 and 377v-  378r), whereas in most pages with five lines, line 2 is blue  and line 4 is red.

translation, it is the contrary so that each line of text in blue is associated with a line of translation in red and vice versa.19 The decoration itself is very lavish. It consists in one double-frontispiece at the beginning of the manuscript (fols. 1v-2r) (figure 1) and thirty-three or,  more probably, since there is a gap between fols. 221 and 222, originally thirty-four double illuminated pages with headings and frames at the beginning of suras 2, 7, 19, and 38, and at the beginning of juzʾs.20 The beginnings of the other suras are marked by illuminated headings.21 The verses are indicated by golden circles or rosettes, often adorned with blue and red dots. The written surfaces are also framed by golden, blue and red lines. Finally, other divisions of the Qurʾan, in one-quarter, one-half, and three-quarters of a juzʾ, and the ritual prostrations are indicated with a word (respectively rubʿ, niṣf, thalāthat arbāʿ, and sajda) in gold in the fore-edge margin, whereas all of the other decorations, i.e., vignettes and frames, that appear in the fore-edge margins seem to be later additions. As a matter of fact, observation with the naked eye and with a binocular lens and some analyses show that the manuscript underwent several types and most likely periods of intervention. Our goal, at this stage, is not to identify and date every phase of the history of the manuscript, but merely to uncover the major non-original elements and reconstruct as much as possible the original condition of the manuscript. We will, thus, begin by highlighting the most salient restorations, additions and over- and re-paintings, with a focus on the double-frontispiece and the thirty-three other extant double illuminated pages, in order to be able to formulate a few hypotheses about how they might have been achieved.

RestoRations, additions and overPaintinGs/rePaintinGs (appendix 3) Apart from the recent, twentieth-century binding and a few contemporary restorations and stabilization operations, such as applying a white film over  fragile decorations, plugging the holes, especially in the margins, or consolidating the margins, we notice several types as well as phases of older interventions on the illuminations.

19.  In  several  pages  with  five  lines,  the  gold  lines  are  not  translated (fols. 20v-21r, 40v-41r, 59v-60r, 78v-79r, then from 171v until the end), which is still to be explained. 20. The double illuminated page of juzʾ 13 (Qurʾan 12:53) has disappeared with the missing folios between fols. 221 and 222. 21. The headings of suras 12 and 13 have also disappeared.

60 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 61

Figure 1 – Gwalior Qurʾan, double-frontispiece, fols. 1v-2r. [© By courtesy of the Aga Khan Museum]

62 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

One particularly widespread type of intervention consists in a form of restoration. Numerous illuminations show holes that are quite extensive and regularly placed so that they seem to be caused by the reaction of a corrosive pigment, such as verdigris,22 rather than by other, external, factors. These holes are plugged using a piece of paper that is pasted over and painted so as to continue or at least complete the original decoration, but in a clumsier manner or even, often, different manners, which suggests successive phases of intervention. The most striking example of this type of restoration appears on folios 116r-v and 117r-v. (figures 2-4). Here,  we may assume that the original interlinear spaces were largely deteriorated, if not thoroughly punctured. In a first phase of intervention, due care was taken  in conserving as much as possible of the original lines of text that were cut out one by one, and pasted over new interlinear spaces that were, then, provided with new illuminations – these differ both in style and palette from the surrounding original decoration. In a second phase, and for an unknown reason only on fol. 116v, a larger piece of paper was stuck on the whole written surface, the text was rewritten (with only four lines), and the interlinear spaces re-decorated in a still different style from the two previous ones.23 A second type of intervention consists in a form of addition that is more blended in with the original manuscript, hence more difficult to detect with the  naked eye. In some cases, it is the fact that similar decorations appear on collages of the type discussed above, as well as on the original paper that makes it possible to determine that the latter are additions. For instance, on fols. 135v-136r (figures 5-6), similar  arabesques painted in yellow on a black background appear on hexagonal collages in the borders, as well as in the vignettes, which makes it possible to identify the latter as additions. Likewise, a purple color is often used on collages (e.g., on fols. 251, 396v...), as well as in marginal vignettes throughout the manuscript, which allows us to identify the latter as additions…

22. Verdigris is a basic copper acetate notably obtained by exposing copper to vinegar vapours: Guineau (2005), 739. 23. For similar, though less extensive, examples of this type of restoration, see the double-frontispiece and the double illuminated pages of folios 4, 40v-41r, 78v-79r, 135v-136r, 143v-144r, 171v-172r, 189v-190r, 208v-209r, 233v-234r, 252,  270v-271r,  274v;  289v,  305v-306r,  323v-324r,  341v-  342r, 377v-378r, 396, 415v-416r, 433v-434r, 452v, 490v-491r. Similar restorations also appear in sura headings (e.g., fols. 394, 400, 467, 473v, 478, 490, 506, 530, 539, 539v, 549, 549v, 551).

A third type of intervention consists in overpainting or repainting, which can be revealed both by observations and analyses. This was particularly the case in the double-frontispiece. For instance, in folio 1v, in the branches of the central rosette, a black, and maybe a blue layer seem to have been added over a first blue one (figure 7). In folio 2, the black background of the star and of the four octagons of the central field seems thick and pasty, which is probably also indicative of over-painting. Moreover, like the arabesques painted in yellow on a black background mentioned above, it is very likely that the decoration painted in yellow on a black background in the corners of the central field are also reworked.  An XRF measurement in the lower-left corner reveals, indeed, an important amount of mercury, the constituent molecule of cinnabar/vermilion,24 which suggests that the now-black layer was painted over a red or brown cinnabar/vermilion-based underlying one. Likewise, in the same folios, the visual appearance of gold differs between the wide ribbons and the small motifs, which is confirmed by XRF which reveals a  difference in composition between the two elements. In the ribbons, the alloy is composed of gold and a small amount of copper, whereas in the motifs, there are traces of silver, and of samarium,25 but no traces of copper, which may be indicative of repainting. As far as blue is concerned, the analyses first suggest that a synthetic ultramarine (produced since the 19th century) was used for touch-ups in various places.26 In addition, two different types of indigo27 are used both on collages and on parts of the original illumination, such as the blue vegetal motifs in the central field and probably the semi-six-pointed stars  on either side of the headings, on folio 2, which suggests that these areas were repainted or retouched,

24. ‘Cinnabar’ is used to refer to a natural mercury sulphide, and ‘vermilion’ to a synthetic mercury sulphide that is obtained from a mixture of mercury and sulphur: Guineau (2005), 226, 741. We deliberately use the two names because our analysis methods do not allow us to distinguish between them. 25. Samarium is a rare earth element. 26. Measured with DRS, synthetic ultramarine produces a larger signal than natural lapis-lazuli, but it is often very difficult to decide between the two options: Guineau (2005),  139-140, 424, 533. Natural lapis-lazuli is a semi-precious stone that, until the 19th century, was almost exclusively quarried from the mountainous region of Badakhshan, in present-day Afghanistan, and prepared through a complex and time-consuming process: Porter, Vesel (1993), 150. 27. Indigo is a vegetal blue pigment, as well as a dye, derived from various leguminous plants of the Indigofera family: Guineau (2005), 384-385. The second measurement is, however, difficult to interpret. It is an unreferenced blue  with a marked red component. It is probably an indigo of a different type or a different preparation, notably using plants to produce a red color.

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 63

unless indigo was applied purposely, in order to produce a difference of shade with the blue color that is dominant in the rest of the decoration: ultramarine In fact, this question of the added or original character of indigo arises and remains difficult to answer  throughout the manuscript, because if we are sure that this pigment is used in collages28 and in blatant additions – it is, for instance, mixed with cochinealkermes29 in order to obtain the purple color mentioned above – it is also found in apparently unaltered areas such as the blue lines of the Qurʾanic text, and some decorations. For instance, on folio 530v (figure 8), the blue arabesques on a light blue background and the light blue outline of the golden rosettes are ultramarine-based, whereas the grayish blue background of the triangular spaces between the elongated hexagons and other motifs show a mixture of ultramarine and indigo (?). Whether the appearance of indigo is due to a later intervention or to the original illuminator’s desire to play with the difference of hue between the two pigments remains an open question,30 which is also the case for two or three other colorants: verdigris on the one hand, and red and orange lead on the other hand.31 Besides later restorations and over- or repaintings, several decorations raise the question of whether only a part of the decoration was completed originally, while the rest was left unfinished and completed later.  For example, on folio 41 (figure 9), the different interlinear decorations seem to be original (ultramarine + white lead),32 though incomplete, because the shadings are only achieved in the third and fourth interlinear spaces, whereas the red elements seem to be added. As a matter of fact, not only do they overlap with the blue drawing in an illogical way, but also evoke blatant additions, such as the fifth interlinear  decoration of folio 59v. On folio 40v, on the contrary, the first interlinear decoration is probably original  or close to the original – the analyses reveal the use of ultramarine with a small amount of white lead – whereas the other interlinear decorations are most likely added because they show rather different motifs and are alternately blue and grayish blue – a mixture  of  ultramarine  and  an  unidentified  blue –  which is unique in the manuscript. Like on folio 41, 28. In addition to fols. 1v-2r, indigo is used on collages on fols. 116v, 143v, 306?, 377v, 433v… 29. Cochineal-kermes is an animal dyestuff principally derived from scale insects: Guineau (2005), 236, 610-611, 615-616. 30. For other problematic appearances of indigo, see fol. 143v (vignette); 171v (outer border). 31.  Red lead, also called ‘minium’, is an artificial lead oxide:  Guineau (2005), 474, 619-620. 32.  White  lead  is  an  artificial  white  lead-based  pigment:  ibid., 124-125.

on folio 154, the interlinear spaces are adorned with blue and red florets that are refined and may, thus,  be original. On the contrary, on folio 153v, only a few red  florets  between  lines 1  and  2,  and  3  and  4  are  probably original, while the other ones, much coarser, are most likely added.33 Once the most salient restorations, additions and over- and repaintings are set aside, or at least questioned, the original decorations seem to be consistent in some respects and inconsistent in others.

an iranian PaLette? (appendix 3)

The original palette seems to be limited to blue, red, yellow, green, brown, white, black, and gold. The pigments and dyes used are essentially the following: Blue: ultramarine, frequently mixed with more or less white lead; light blue is obtained by mixing ultramarine  with white lead; indigo (?); other undefined blue (?). Green: verdigris (?).34 Red: cinnabar/vermilion; pink is obtained by mixing  cinnabar-vermilion  with  white  lead;  cochineal-kermes;  red lead (?); orange lead (?) Yellow: orpiment.35 Dark brown is obtained by mixing cinnabar/vermilion with black (ink?); or cinnabar/vermilion or orange lead  with an organic substance. Light brown: unidentified organic substance. White: white lead. Black: carbon black.36 Gold: almost pure gold.

All of these pigments and dyes are well known,37 but their use in this manuscript is interesting because they evoke less late fourteenth-century India than medieval Iran. The few analyses of pre-Mughal Indian, 33.  Other apparently later florets appear on fols. 323v-324r  and 510v-511r. Likewise, cross patterns are probably original in some folios (21v-22r, 471v-472r) and added in some others (323v-324r, 341v-342r, 377v-378r). The crossand-circle and the scroll patterns also seem to be added (59v-60r). In other decorations, such as fols. 270v-271r, most of the interlinear decorations are probably added. 34. In one occurrence (fol. 510v), green is obtained by mixing indigo and orpiment, but this seems to correspond to a later addition (see n. 31 above). 35.  Orpiment  is  a  native  or  artificial  arsenic  trisulphide  (As2S3): Guineau (2005), 528-529. 36. Ibid., 498. 37. See for instance Purinton, Watters (1991), online edition: http://cool.conservation-us.org/jaic/articles/jaic30-02002.html, accessed Nov. 25, 2010 (this study was slightly modified and re-published in 2004 by H. F. Boroojeni under  the title “Materials Used by Medieval Persian Painters”;  Porter (1992) and for the english edition (1994); Guineau  (2005); Eastaugh et al. (2008).

64 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 65

Figure 2 – Gwalior Qurʾan, fols. 116v-117r. [© By courtesy of the Aga Khan Museum]

66 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

Figure 3 – Gwalior Qurʾan, fol. 117r, restoration of the interlinear spaces (detail). [© By courtesy of the Aga Khan Museum]

both Islamic and Jaina, paintings carried out so far38 show that these paintings are characterized in particular by the dominant use of azurite39 or indigo for blue – but also for green, which is usually composed of indigo and orpiment – and by the exclusive use of kaolin40 and/or calcic whites,41 whereas the corresponding pigments used in the Gwalior Qurʾan, ultramarine, verdigris and white lead, are hardly attested in India before the 15th and especially 16th centuries, which could be significant in respect of the origin of  the patron(s) and/or artist(s) of the manuscript. Contrary to the general consistency of the palette, several criteria allow us to divide the decorations into two main groups. The first group includes fols. 1v-2r,  2v-3r, 3v-4r, 21v-22r, 40v-41r, 97v-98r, 153v-154r, 171v-172r, 270v-271r, 358v-359r, 490v-491r, and 510v-511r; the second group, fols. 59v-60r, 78v-79r,  143v-144r, 233v-234r, 251v-252r, 323v-324r, 341v-342r, 377v-378r,  415v-416r,  433v-434r,  and  471v-472r;  whereas a number of other illuminations stand out with a more distinct character.

the decoration Process:

PreLiminary observations and hyPotheses

Two main groups As far as compositions are concerned, two main patterns are used in group 1: either a form of table is delimited by a succession of golden and colored stripes, and extended by a vignette on the axe of each heading

38.  Johnson (1972), 139-146; West FitzHugh (1988), 425-432;  Isacco (2008). 39. Azurite is a native blue basic copper carbonate: Guineau (2005), 94. 40.  Kaolin is a type of fine white clay rock: ibid., 411. 41. Calcic whites can be obtained from egg shells or calciumrich minerals such as gypsum.

and a blossomed or triangular projection in the middle of the fore-edge margin;42 or the headings are separated  from  the  written  surface  that  is  flanked  by  two vegetal bands.43 The classical pattern consisting in a table, surrounded with a vegetal border on three sides and accompanied by a vignette on the horizontal axis, is only used in two decorations,44 while it is largely dominant in group 2. In some decorations of this group, the written area is also flanked by vegetal  bands, but these, instead of being substituted for the frame, are inserted inside it.45 As far as headings are concerned, several differences, in spelling, paleography and decoration, appear between the two groups (table 1). First, in the numbering of juz’s (al-juzʾ[x] min ajzāʾ al-thalathīn [juzʾ no. [x] of 30]), the word thirty is usually mistakenly written with a long “ā” and in the nominative case  (thalāthūn) in group 1,46 whereas it is often correctly written, without a long “ā” and in the genitive case  (thalathīn), in group 2.47 Two calligraphic styles are used in group 1: the ‘New Style’ on the one hand, and a sort of mixture between muḥaqqaq and thulth on the other hand. It is, however, possible to draw parallels between the two scripts, notably in the use of a special type of hamza, with a very elongated base. A muḥaqqaq-thulth is also used in group 2, but it differs from the previous one, notably by the use of a specific form of hamza: in zigzag. Moreover, the style of group 2 seems less loose and round: the indentations are less inclined;  in the word al-juzʾ, lām never shows a concave curve;  nūn is also less concave.

42. 43. 44. 45.

Fols. 2v-3r, 97v-98r, 171v-172r. Fols. 3v-4r, 21v-22r, 40v-41r, 270v-271r, 510v-511r. Double-frontispiece and fols. 358v-359r. Fols. 59v-60r, 78v-79r, 233v-234r, 251v-252r, 323v-324r, 377v-378r, 415v-416r. 46. Fols. 3, 98, 154, 172, 511. 47. Fols. 60, 144, 324, 342, 416, 434, 472.

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 67

Figure 4 – Gwalior Qurʾan, fol. 117v. [© By courtesy of the Aga Khan Museum]

68 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 69

Figure 5 – Gwalior Qurʾan, fols. 135v-136r. [© By courtesy of the Aga Khan Museum]

70 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

Figure 6 – Gwalior Qurʾan, fol. 135v, collage, front and back sides. [© By courtesy of the Aga Khan Museum]

Figure 7 – Gwalior Qurʾan, fol. 1v: black and blue over-paints in the central rosette. [© By courtesy of the Aga Khan Museum]

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 71

Figure 8 – Gwalior Qurʾan, fol. 530v. [© By courtesy of the Aga Khan Museum]

72 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 73

Figure 9 – Gwalior Qurʾan, fols. 40v-41r. [© By courtesy of the Aga Khan Museum]

74 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

Finally, in group 1, the inscriptions often stand out against animated oblong or serrated leaves on a light-brown background, whereas in group 2, there are much stiffer sharp and curved leaves and webbed ones which are planted vertically and horizontally, in a much more static way, on a bright cochineal-red background. As far as the decorative vocabulary is concerned (table 2), some motifs seem characteristic of group 1, especially an oblong, rather round leaf, with a midrib; a wide variety of digitate palmettes; and especially  leaves  with  folded  leaflets,  which  illustrate  a  way  of suggesting volume and three-dimensionality that seems to be extremely rare, if not unique in this region at such an early date. These motifs never appear in group 2, which stands out by other patterns. Besides the sharp and curved leaves and the webbed ones mentioned above, other motifs are rounder, in particular: a digitate leaf with a fleshier and longer  central leaflet; a leaf where three long leaflets alternate with two heart-shaped ones; and lastly, a more  squashed leaf, with a large tri-lobed ‘head’, two long horizontal ‘arms’, and two small discs at the base. These motifs are often treated in monochrome shades. They are blue, red or gray in group 1; and red,  gray or yellow in group 2. But while in group 1, the shading is blurred until the point which is more pronounced, in group 2, it is almost the entire inner contour of the motif that is highlighted with a light line and the outer contour and/or the point with a dark line. This treatment echoes the fact that the inner contour of the same polychrome patterns is also, often, highlighted, with a golden or a yellow line, in this group. If there are, thus, enough links to group several decorations in coherent ensembles, there are, however, also observations that clearly raise the question of the intervention of other hands within a single group, and even, sometimes, in a single decoration, according to different patterns.

For instance, the ‘New Style’ used in the first three  double illuminated pages of group 1 (fols. 1v-2r, 2v-3r and 3v-4r) (type 1) is weak: unbalanced, with many discrepancies in the thickness of the lines and in the size and spacing of the letters.48 On the contrary, the ‘New Style’ used in the three following decorations of

the same group (fols. 21v-22r, 40v-41r and 510v-511r) (type 2) is better proportioned and formally more harmonious. This is particularly visible in final nūns. Lām-alifs  are  also  different,  with,  in  the  first  set,  a  lām (?) which is detached and extended by a slight downward curve, and, in the second set, two letters that are linked to form a closed loop (table 3). Some of these examples, but also other ones, such as the unfinished pairs mentioned above, raise the  question of the involvement of different hands in a single group, but also a single decoration. Thus, the interlinear decorations of folio 98 (of group 1) (figure 10) seem identical to other decorations of the  same group 1 (2v-3r, 3v-4r, 490v-491r), whereas those of 97v are less refined, show a different treatment of  shading and stand out by some motifs. An isolated decoration, fols. 208v-209r, also presents several asymmetries, especially in the headings, both in the inscriptions and in the backgrounds – on folio 208v itself, the borders of the upper and lower headings are also different. Moreover, the interlinear decorations of the two pages do not seem to be identical.49 In these decorations, one can ask whether the left-hand side, which seems to be often completed and of higher quality, is the work of a master, and the right-hand side, which is often unfinished or of lower quality, the work of an assistant. This has already been observed in other manuscripts such as a Qurʾan dated to 696/1296 by ‘The Cynosure of calligraphers’,  Yāqūt  al-Mustaʿṣimī  (d. 698/1298),50 or the oldest complete copy of the Majmūʿa al-rashīdiyya, a theological compendium dated to 707-710/1307-1311 by the most important Ilkhanid vizier, Rashīd al-Dīn  Faḍl Allāh al-Hamadhānī (d. 718/1318).51 In these cases, it has been hypothesized that the right-hand side might have been entrusted to an assistant because it appears on the back of the first page of the manuscript that is more fragile and, thus, more likely to be damaged or lost. This explanation obviously does not apply to the illuminations that appear throughout the Gwalior Qurʾan, but we can still notice that one of the two decorations mentioned above (97v-98r) appears at the beginning of a quire, while the second one (208v-209r) is divided between two quires.52 To sum up, several aspects of the Gwalior Qurʾan – its large size, high-quality paper, sophisticated calligraphy, and rich and elaborate decorative program – suggest that it was most likely made for an

48. The heads of jīm and kāf are often too thin (fols. 1v-2r: upper heading; 2v, upper and lower headings).The size  and spacing between letters, in particular at the ends of the lines, are often too small or, sometimes, on the contrary, too large, which even induces orthographic problems, for instance a sīn with two indentations (1v: lower heading);  ‘al-awwal’ with two lāms (3: upper heading).

49.  See also fols. 289v-290r; 305v-306r; 415v-416r; 452v-453r;  510v-511r. 50. Konya, Mevlana Museum, no. 15: cf. Ben Azzouna (2009), 96-98. 51. Paris, BnF, Arabe 2324: cf. Ben Azzouna, Roger-Puyo, forthcoming.

A multi-hand work

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 75

Group 1

Group 2

Table 1 – Comparison between the headings of groups 1 and 2.

Group 1

Group 2

Table 2 – Comparison between the vegetal motifs of groups 1 and 2.

Type 1

Table 3 – Comparison between ‘New Style’ types 1 and 2.

Type 2

76 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 77

Figure 10 – Gwalior Qurʾan, fols. 97v-98r. [© By courtesy of the Aga Khan Museum]

78 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

Fols. 116v-117r

Fols. 251v-252r

1

Table 4 - Comparison between the scroll patterns in fols. 116v-117r and 251v-252r

important patron, by a rather coherent team. At this stage of the research, it is impossible to determine how this team worked, but a few modes of distribution of tasks can be noted. First, two rather consistent groups of decorations emerge. The distribution of these groups throughout the manuscript suggests that some decorations, and sometimes only sections of decorations, may have been carried out in series. For instance, the first three illuminations of the manuscript bear inscriptions in New Style type 1, and are followed by two other decorations with inscriptions in New Style type 2. The headings of juzʾs 6 (97v-98r), 7 (116v-117r), 9 (153v-154r) and 10 (171v-172r) are written in muḥaqqaq-thulth. The interlinear spaces of juzʾs 9 and 10 are also adorned with  florets.  The  introductory  pages  of  sura 19  (274v-275r) and the two following juzʾs 17 (289v-290r) and 18 (305v-306r) show eight vs. nine lines of writing, as well as equivalent layouts and decorative elements, for instance lozenge-divided headings and gold scrolls on a black background (which also appear in the next juzʾ (323v-324r) etc.

52. See also fols. 289v-290r and 305v-306r. Another decoration is divided between two quires: fols. 530v (at the end of quire 34) and 531r (at the beginning of quire 35), but in this example, the pair seems very harmonious.

On the other hand, some patterns seem to come back in a cyclical way, every seven or eight decorations. Thus, scrolls punctuated by a particular motif of loops flanked by dots appear in juzʾ 7 (116v-117r), then only in juzʾ 15 (251v-252r) (table 4). Likewise, several particular features such as titles inscribed in cartouches with brace-shaped edges; large inner  borders associated with interlinear spaces decorated with scrolls with yellow contours; and lastly an outer  border of petals separated by jagged motifs appear in juzʾ 14 (233v-234r), then in juzʾ 22 (377v-378r). Similarly, titles flanked by orange bunches on a light-orange  background;  elongated  proportions;  and four-edge outer borders, appear in juzʾ 20 (341v342r), then in juzʾ 27 (471v-472r). Beyond groups, series or pairs, the left-hand(recto) vs. right-hand-side (verso) division seems to be recurrent, which may indicate that some illuminations were started on the recto, probably by a master, and continued on the verso by an assistant, or left unfinished and completed by a later contributor. These observations, thus, suggest the coexistence of several rhythms, which probably follow the order of writing (series) or maybe the division of the quires (pairs). It is, however, necessary to refine the definition of the individual styles and the attribution of the decorations, hence the identification of the subsequent phases of intervention, in order to better understand the contours of this organization.

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 79

bibLioGraPhy References (except Encyclopaedia of Islam) Barnes (Norman F.) 1939 “Color characteristic of artists’ pigments”, in Journal of the Optical Society of America 21, pp. 208-214. ben azzouna (Nourane) 2009 La Production de Manuscrits en Iraq et en Iran Occidental à l’époque des Dynasties Mongoles (Les Ilkhanides et les Djalayirides [658-814/1256-1411]), PhD dissertation, Paris, École pratique des hautes études. ben azzouna (Nourane), roger-Puyo (Patricia) forthcoming “The Question of the formation of manuscript production workshops in Iran according  to  Rashīd  al-Dīn  Faḍl  Allāh  al-Hamadhānī’s  Majmūʿa Rashīdiyya in the Bibliothèque nationale de France”, in Journal of Islamic Manuscripts. BorooJeni (Hamid Farahmand) 2004 “Materials used by medieval Persian painters”, in An Anthology of Iranian Studies 7, pp. 25-57. Brac de la Perrière (Éloïse) 2008 L’art du livre dans l’Inde des sultanats, Paris: Presses de l’université Paris-Sorbonne. digBy (Simon) 2004 “Before Timur came: provincialization of the Delhi Sultanate through the 14th century”, in Journal of the Economic and Social History of the Orient 47, no. 3, pp. 298-356. eastaugh (Nicolas) et al. 2008 Pigment Compendium. A Dictionary and Optical Microscopy of Historical Pigments, Oxford: Butterworth-Heinemann Ltd. guineau (Bernard) 2005 Glossaire des matériaux de la couleur et des termes techniques employés dans les recettes de couleur anciennes, Turnhout: Brepols. isacco (Enrico) 2008 Les pigments des miniatures indiennes, Paris: L’Asiathèque.

Johnson (Ben B.) 1972 “A Preliminary study of the technique of Indian miniature painting”, in Pratapaditya Pal (dir.), Aspects of Indian Art: Papers Presented in a Symposium at the Los Angeles County Museum of Art, October 1970, Leiden: Brill, pp. 139-146. Porter (Yves) 1992 Peinture et arts du livre : Essai sur la littérature technique indo-persane, Tehran/Paris: Institut français de recherche en Iran. 1994 Painters, Paintings and Books. An Essay on IndoPersian Technical Literature, 12-19th Centuries, translated by S. Butani, New Delhi: Manohar. Porter (Yves), vesel (Ziva) 1993 “La joaillerie et la peinture: approvisionnement en pierres et en pigments dans l’Iran médiéval”, in Res Orientales 5, pp. 141-157. Purinton (Nancy), watters (Mark) 1991 “A Study of the materials used by medieval Persian painters”, in Journal of the American Institute for Conservation 30, no 2, pp. 125-144 and http://cool.conservation-us.org/jaic/articles/ jaic30-02-002.html. roger (Patricia), villela-Petit (Inès), vandroy (Solène) 2003 “Les laques de Brésil dans l’enluminure médiévale : Reconstitution à partir de recettes anciennes”, in Studies in Conservation 48, pp. 155-170. West FItzhugh (Elisabeth) 1988 “Study of pigments on selected paintings from the Vever collection”, in Glenn D. Lowry, Milo Cleveland Beach (dir.), An Annotated and Illustrated Checklist of the Vever Collection, Washington (DC) / Seattle / London: Arthur M. Sackler Gallery/Smithsonian Institution in association with the University of Washington Press, pp. 425-432.

Article from Encyclopaedia of Islam Burton (John) “Muṣḥaf”, in EI2 VII, pp. 668-669 (for the french version).

80 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

aPPendix 1: maPPinG tabLe of the manuscriPt and its e-PubLication Decoration

Folio in the manuscript

Folio in e-corpus

Double-frontispiece

1v-2

1v-2

Juzʾ 1

2v-3

2v-3

Sura 2

3v-4

3v-4

Juzʾ 2

21v-22

21v-22

Juzʾ 3

40v-41

40v-41

Juzʾ 4

59v-60

59v-60

Juzʾ 5

78v-79

78v-79

Juzʾ 6

97v-98

97v-98

Juzʾ 7

116v-117

116v-117

Juzʾ 8

135v-136

135v-136

Sura 7

143v-144

143v-144

Juzʾ 9

153v-154

153v-154

Juzʾ 10

171v-172

171v-172

Juzʾ 11

189v-190

189v-190

Juzʾ 12

208v-209

208v-209

Juzʾ 14

233v-234

233v-234

Juzʾ 15

251v-252

251v-251bis

Juzʾ 16

270v-271

269v-270

Sura 19

274v-275

273v-274

Juzʾ 17 (mistakenly numbered 18)

289v-290

288v-289

Juzʾ 18

305v-306

303v-304

Juzʾ 19

323v-324

321v-322

Juzʾ 20

341v-342

339v-340

Juzʾ 21

358v-359

356v-357

Juzʾ 22 (mistakenly numbered 2)

377v-378

375v-376

Juzʾ 23

395v-396

394v-395

Sura 38

406v-407

404v-405

Juzʾ 24

415v-416

412v-413

Juzʾ 25 (mistakenly numbered 5)

433v-434

429v-430

Juzʾ 26

452v-453

448v-449

Juzʾ 27

471v-472

467v-468

Juzʾ 28 (unnumbered)

490v-491

486v-487

Juzʾ 29

510v-511

506v-507

Juzʾ 30

530v-531

526v-527

Juzʾ 13 (missing)

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 81

aPPendix 2: constitution of the quires and distribution

of the decorations throuGhout the manuscriPt

Key : *middle of the quire ; irregular quire ; decoration I : 1-2-3-4-5*(?)-6-7-8//9*-10-11-12-13-14-15-16 II: 17-18-19-20-21-22-23-24//25*-26-27-28-29-30-31-32 III : 33-34-35-36-37-38-39-40//41*-42-43-44-45-46-47-48 IV : 49-50-51-52-53-54-55-56//57*-58-59-60-61-62-63-64 V: 65-66-67-68-69-70-71-72//-73*-74-75-76-77-78-79-80 VI: 81-82-83-84-85-86-87-88//89*-90-91-92-93-94-95-96 VII : 97-98-99-100-101-102-103-104//105*-106-107-108-109-110-111-112 VIII : 113-114-115-116-117-118-119-120//121*-122-123-124-125-126-127-128 IX : 129-130-131-132-133-134-135-136//137*-138-139-140-141-142-143-144 X : 145-146-147-148-149-150-151-152//153-154-155-156-157-158-159-160 XI : 161-162-163-164-165-166-167-168//169*-170-171-172-173-174-175-176 XII : 177-178-179-180-181-182-183-184//185*-186-187-188-189-190-191-192 XIII : 193-194-195-196-197-198-199-200//201*-202-203-204-205-206-207-208 XIV : 209-210-211-212-213-214-215-216//217*-218-219-220-221-lacuna of three folios XV : Lacuna of ten folios- 222-223-224-225-226-227 XVI : 228-229-230-231-232-233-234-235//236*-237-238-239-240-241-242-243 XVII : 244-245-246-247-248-249-250-251//252*-253-254-255-256-257-258-259 XVIII : 260-261-262-263-264-265-266//267*-268-269-270-271-272-273 (septenion or seven-bifolio quire) XIX : 274-275-276-277-278-279-280-281//282*-283-284-285-286-287-288-289 XX : 290-291-292-293-294-295-296-297//298*-299-300-301-302-303-304-305 XXI : 306-307-308-309-310-311-312-313//314*-315-316-317-318-319-320-321 XXII : 322-323-324-325-326-327-328-329//330*-331-332-333-334-335-336-337 XXIII : 338-339-340-341-342-343-344-345//346*-347-348-349-350-351-352-353 XXIV : 354-355-356-357-358-359-360-361-362//363*-364-365-366-367-368-369-370 (f° 360 added?) XXV : 371-372-373-374-375-376-377-378//379*-380-381-382-383-384-385-386 XVI : 387-388-389-390-391-392-393-394//395*(?)-396-397-398-399-400-401-402 XXVII : 403-404-405-406-407-408-409-410//411*-412-413-414-415-416-417-418 XXVIII : 419-420-421-422-423-424-425-426//427*-428-429-430-431-432-433-434 XXIX : 435-436-437-438-439-440-441-442//443*-444-445-446-447-448-449-450 XXX : 451-452-453-454-455-456-457-458//459*-460-461-462-463-464-465-466 XXXI : 467-468-469-470-471-472-473-474//475*-476-477-478-479-480-481-482 XXXII : 483-484-485-486-487-488-489—490//491*-492-493-494-495-496-497-498 XXXIII : 499-500-501-502-503-504-505-506//507*-508-509-510-511-512-513-514 XXXIV : 515-516-517-518-519-520-521-522//523*-524-525-526-527-528-529-530 XXXV : 531-532-533-534-535-536-537-538//539*-540-541-542-543-544-545-546 XXXVI : 547-548-549-550//551*-552-553-554

82 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

aPPendix 3: nature of the coLor materiaLs used in the GwaLior qurʾan Color

Material

Illumination

Blue

Ultramarine

1v (heading, central field, border, vignette, outer frieze); 2 (vignette, outer frieze); 3 (interlinear decoration, frame, text); 21v (outer lateral border); 116v (inner & outer lateral borders, upper & lower borders, lower vignette); 171v (upper and lower vignettes); 377 (lateral border, frame); 395v (interlinear decoration between lines 1 and 2); 416 (outer lateral square in the heading); 426v (text); 433v (frame, vignette, floret); 530v (circle on line 1, text); 551v (text) Ultramarine + white lead

Indigo

2 (center); 116v (text); 143v (lateral squares in the heading [collage]); 306 (interlinear decoration); 339v (frame); 342 (frame); 377v (lateral squares in the heading [collage]); 415v (inner and outer borders); 416 (dark parts of the frame in the heading); 433v (lateral squares in the heading [collage]); 510v  (lateral  square  heading);  530v (triangular interstice in the upper border); 551v (frame) Ultramarine + indigo or + other (undefined  blue)*

Green

1v (center); 2 (center, border); 21v (heading, interlinear space); 40v (heading, interlinear decoration between lines 1 and 2, lateral border, frame, vignette); 41 (interlinear decoration between lines 1 and 2, lateral borders, frame); 59v (frame, border); 135v (lateral borders); 171v (triangular projection); 251v (heading); 304 (border); 342 (interlinear space between lines 3 and 4, border); 395v (frame?); 415v (frame); 416 (outer lateral border); 433v (heading?); 510v (heading, interlinear decoration?); 530v (circle on lateral border)

verdigris

1v (heading?, center, frieze?); 2 (heading, center, border, frieze); 21v (text*?); 40v (interlinear decoration between lines 4 and 5*, text*); 41 (text); 59v (interlinear decoration*, text*); 143v (dark circle in the vignette, [ultramarine + white lead: text, border]); 171v (interlinear space*, frame, text); 251v (ultramarine + white lead: interlinear decoration, lateral borders, frame, vignette); 252 (ultramarine + white lead: interlinear decoration); 305v (border); 306 (text, [ultramarine + white lead]): frame); 341v (border); 342v (text*?); 395v (ultramarine + white lead + small amount of indigo in the heading, triangular projection, text); 415v (text, [ultramarine + white lead + small amount of indigo in the lateral squares in the heading, border]); 416 (lateral square heading near the  frame, text); 510v (text); 530v (floret, text*?); 551v (text*) 1v  (with  white  lead:  heading,  center,  border);  2  (with  white lead: border); 116v (lateral square heading, text);  143v (border); 171v (border); 251v (frame of the heading  [collage]); 306 (heading, interlinear space [on older green]);  342  (border);  415v  (upper  heading  [collage],  lateral  border); 416 (frame)

Indigo + orpiment

510v (heading, interlinear space)

THE GWALIOR QURʾAN: ARCHAEOLOGY OF THE MANUSCRIPT AND OF ITS DECORATION • 83

Color

Material

Illumination

Red

Cinnabar-vermilion

1v (border, center [and + white lead]); 2 (center, border  [and + white lead]); 40v (outer lateral border); 41 (interlinear space, lateral border); 59v (text, interlinear space  decoration flowers, border); 143v (text); 171v (text);  251v  (text);  342  (text);  416  (border);  433v  (interlinear  decoration, vignette, text); 510v (heading [+ white lead?],  lateral border); 530v (and + white lead for pink, text); 551v  (interlinear decoration, [with white lead: text]); 554 (text) Cinnabar-vermilion 2 (border); 3 (text); 40v (text); 41 (text); 116v (text); 135v  + red lead (heading, border); 143v (border?, text); 250v (text); 251v  (heading?); 395v (heading?, vignette, text); 415v (text);  416  (text);  433v  (border);  509  (text);  510v  (interlinear  space, text)

Orange

Cochineal-kermes

1v (center [on light brown or gold?]); 2 (center, border:  on  cinnabar-vermilion  and  white  lead);  21v  (heading,  interlinear hatching, lateral border); 116v (heading),  143v  (heading,  interlinear  hatching);  171v  (interlinear  hatching);  172  (heading);  305v  (border);  342  (heading  [on  gold]);  395v  (heading  [preparatory  drawing]);  396  (heading);  416  (heading,  interlinear  decoration);  433v  (interlinear heading); 490v (lateral border); 510v (interlinear hatching)

Orange lead

116v (border: altered orange lead?, text); 234 (lateral border); 251v (interlinear decoration, border?); 415v  (border) Cinnabar-vermilion 41  (lateral  border);  59v  (lateral  border?);  416  (border,  on orpiment yellow floret); 510v (heading)

Purple

Yellow

Cochineal-kermes + ultramarine

510v (heading)

Cochineal-kermes + indigo

396 (triangular projection); 551v (marginal vignette)

Orpiment

1v (center); 2 (center); 40v (heading?); 41 (lateral border?);  59v  (border?);  135v  (lateral  border);  251v  (interlinear  decoration);  342  (text,  lateral  border?);  395v  (interlinear  decoration between lines 4 and 5 [on pink], triangular projection);  415v  (border);  416  (border,  floret);  433v  (interlinear  decoration,  border);  510v  (lateral  border);  530v (interlinear decoration) Orpiment + organic?

Brown

Ochre?

21v (heading); 251v (heading); 415v (border);  416 (lateral border) 1v (lateral square in the heading); 143v (border of the  lateral square in the heading [collage])

Cinnabar-vermilion 1v (center); 2 (center, border); 40v (lateral border); 171v  + red lead + black (florets);  342  (vignette);  395v  (triangular  projection);  510v (lateral border)

84 • NOURANE BEN AZZOUNA / PATRICIA ROGER-PUYO

Color

Material

Brown

Illumination Cinnabar-vermilion 1v (lateral square in the heading); 2 (heading?, border) + organic Organic?

143v (border of the heading);  251v (interlinear decoration)

Organic red on gold 116v (heading); 172 (heading); 342 (heading) Light brown Organic?

1v (center); 21v (heading); 251v (interlinear decoration); 415v (interlinear decoration)

Black

1v (center, border); 2 (center, lateral square in the heading); 21v (heading); 40v (heading, lateral border); 116v  (heading,  upper  border,  interlinear  space,  text);  143v  (border); 251v (heading, lateral border); 395v (heading,  triangular projection); 415v (lateral border); 416 (interlinear space, lateral border [black on old blue?]); 433v  (border); 510v (heading); 530v (interlinear decoration)

Carbon black

Indigo on cochineal-kermes

1v (lateral square in the heading)

White

White lead

1v (heading, center, border); 2 (center, border);  116v (heading); 415v (lateral border);  416 (heading); 433v (heading)

Gold

Gold

1v (+ traces of silver and samarium: heading, center; +  traces of samarium: border; + copper: frame); 2 (+ traces  of silver and samarium: border; + copper: frame); 116v  (heading); 251v (text); 342 (heading); 395v (heading,  text); 415v (border, frame, text); 416 (border, text); 510v  (heading); 530v (heading, text)

LES GLOSES MARGINALES ET LE FĀLNĀMA DU CORAN DE GWALIOR : témoignages des usages multiples du Coran dans l’Inde des sultanats* Sabrina alilouche (Université Paris 1 - Panthéon-Sorbonne) Ghazaleh esmailPour qouchani (Daneshgah Honar Shiraz - Shiraz University of Art)

Résumé Cet article a pour but de présenter les recherches menées dans le cadre de l’atelier dirigé par Éloïse Brac de la Perrière sur les gloses en persan et en arabe jouxtant le texte coranique ainsi que le livre de divination placé à la fin du coran de Gwalior de l’Aga Khan Museum. De par plusieurs éléments, ce manuscrit coranique s’inscrit dans une sphère indo-persane : le choix des titres de sourates, la traduction interlinéaire en persan, ainsi que la graphie bihārī. Mais sa particularité réside dans la diversité remarquable des textes qu’il présente : une copie du texte coranique accompagnée d’une version persane interlinéaire, des variantes de lecture canoniques, des faḍāʾil suwar al-qurʾān, des éléments liés à des pratiques particulières telles que les suggestions de répétition de duʿāʾs, ainsi qu’un énoncé allégorique et le plus ancien fālnāma coranique daté. L’article explore les différentes composantes de ce coran achevé quelques mois seulement après le sac de Delhi en décembre 1398 et qui semble fonctionner comme une sorte de manuel à usages multiples. Celles-ci fournissent de précieux renseignements sur le contexte dans lequel a été produit le manuscrit, peut-être lié aux milieux soufis établis à Gwalior à la fin  du xive siècle.

Abstract Marginal Glosses and the Fālnāma of the Gwalior Qurʾan: Witnesses to the Uses of the Qurʾan in Sultanate India This article presents research that was carried out in the context of Éloïse Brac de la Perrière’s workshop on the Persian and Arabic glosses to the Qurʾanic text in the Gwalior Qurʾan and on the book of divination at the end of the manuscript, conserved at the Aga Khan Museum in Toronto.

*

Nous remercions Éloïse Brac de la Perrière qui nous a accueillies au sein de l’atelier qu’elle dirigeait sur le coran de Gwalior et qui nous a invitées à participer au colloque international dédié à ce manuscrit en nous prodiguant de précieux conseils et en nous initiant à des recherches des plus passionnantes. Nous remercions également Christiane Gruber et Asma Hilali qui nous ont soutenues et aidées dans les recherches ici présentées. Merci encore à Frantz Chaigne qui a consacré un temps considérable à la relecture de cet article. Notre reconnaissance s’adresse également à Mathilde Cruvelier.

Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 85-110

86 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI This Qurʾanic manuscript fits into the Indo-Persian sphere based on several elements:  the choice of the titles of suras, the interlinear Persian translation and the use of bihārī calligraphy. But its particularity resides in the remarkable diversity of its elements: a copy of the Qurʾanic text with an interlinear Persian translation, variants of canonical readings, faḍāʾil suwar al-qurʾān, elements related to specific practices such as suggestions for the repetition of duʿāʾs, an allegorical narrative and the earliest dated Qurʾanic fālnāma. The article explores the different components of this Qurʾan, which was completed only a few months after the sac of Delhi in December 1398 and seems to function as a sort of manual with multiple uses. They provide precious information about the context in which the manuscript was produced, perhaps linked to the Sufi milieux that were established in  Gwalior in the late 14th century.

introduction Le coran de Gwalior est un manuscrit singulier et véritablement riche de par ses composantes hétéroclites. Il réunit en effet des éléments aussi inhabituels qu’un livre de divination (fālnāma) daté, des gloses marginales restituant des vertus ou les propriétés salutaires des sourates (faḍl al-sūra) selon des hadīths « apocryphes »1 qui côtoient des annotations relatives à des variantes de lecture canoniques (qirāʾāt) et des corrections didactiques. Il renferme également des prescriptions d’actes comme la prosternation et la répétition de versets, ainsi qu’une traduction interlinéaire persane2. Cet ensemble d’éléments nous informe de la multiplicité d’usages de ce muṣḥaf. Le coran de Gwalior semble donc remplir la fonction d’une sorte de manuel à usages multiples. Sa traduction interlinéaire ainsi que certaines notes marginales, liées à des corrections textuelles, témoignent d’un emploi pédagogique du manuscrit suivant l’expression de Sheila Blair dans son article « Uses and Functions of the Qurʾānic Text »3. D’autres annotations sont relatives à ces lectures canoniques et côtoient des prescriptions de récitations litaniques. Les premières sont relatives à une lecture en tant que tilāwa (simple lecture du texte coranique à haute voix) du Coran ;  les secondes ont une fonction performative et sont vraisemblablement liées à des pratiques exotériques. Dans la partie finale du manuscrit se trouve également  un fālnāma qui atteste de ce dernier usage. Il est remarquable car il s’agit du plus ancien fālnāma coranique daté, complet et annexé à un Coran qui soit connu.

1. 2. 3.

Hilali (2011), 163-174. Voir la récente publication de Zadeh (2012) sur l’histoire et les enjeux de la traduction persane du Coran. « Another type of Qurʾān  manuscript  designed  for  pedagogical purposes has interlingual translation and was used in proselytizing from at least the 12th century » : Blair (2006), 192.

Dans le présent article seront exposés, dans un premier temps, les différentes sortes de gloses marginales présentes dans le manuscrit coranique de Gwalior, à savoir : les annotations relatives à la récitation du Coran ainsi qu’aux qirāʾāt, puis les annotations liées aux pratiques performatives et enfin le cas d’une allégorie mystique. Nous terminerons notre présentation par l’analyse du fālnāma.

Les notes et Les GLoses marGinaLes Les gloses marginales constituent un des éléments récurrents du coran de Gwalior. Elles sont copiées dans différentes écritures et différentes couleurs (rouge, bleu, doré) et sont susceptibles d’être classées en trois catégories selon le référent auquel elles se rattachent : annotations liées à la lecture et la récitation ; mentions  en lien avec les qirāʾāt ;  enfin,  notes  relatives  à  des  actions performatives et à des pratiques vraisemblablement soufies.

Annotations relatives à la lecture du Coran Les annotations relatives à la récitation du Coran peuvent être destinées à une pratique dans une sphère privée ou publique, comme une madrasa ou une mosquée. Dans ce cas, elles sont en lien avec les divisions coraniques, les indications de prosternation (sujūd), ainsi que les corrections apportées à la copie du texte coranique. Les divisions du texte coranique et leur agencement dans le manuscrit La division opérée dès le début du recensement du Coran avait pour but l’apprentissage du Texte et sa récitation. Celle-ci est encadrée par des règles de prononciation et de lecture appelées tartīl et tajwīd (orthoépie). Le tartīl consiste en la récitation du Coran selon un rythme lent en vue de son étude, appelée aussi tadarrus, tandis que le tajwīd est la science

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 87

préconisant les règles de la prononciation psalmodiée du texte coranique, même si ce terme (tajwīd) est parfois identifié à la tilāwa, qui est la simple lecture du texte coranique à haute voix4. Juzʾ, rubʿ, thalāth arbāʿ, niṣf On relève dans le coran de Gwalior le système de division  tel  qu’il  était  en  usage  à  l’époque  où  il  fut  copié : il contient trente juzʾs, chacun d’entre eux comporte deux ḥizbs, et chaque ḥizb est divisé en quatre rubʿs. Les indications de ḥizb et de juzʾ sont données en graphie muḥaqqaq dans des bandeaux enluminés à l’intérieur de l’encadrement principal, le jadwal. En revanche, les indications des divisions internes à un juzʾ, à savoir le rubʿ (quart), le niṣf (moitié), et le thalāth arbāʿ (trois-quarts) sont inscrits dans la marge (figure 1). L’indication de prosternation (sajda) Parmi les notes marginales figure la sajda qui indique le moment de la prosternation rituelle lors de la lecture du texte coranique. Le mot est toujours copié à l’encre dorée en écriture thuluth. Un coran comporte habituellement quinze sajdas, mais dans le coran de Gwalior le terme n’est cité que quatorze fois, car la deuxième sajda de la sourate XXII au verset 77 est omise. La lettre ʿayn Parmi les indications marginales de divisions coraniques, une lettre apparait fréquemment dans le  manuscrit ;  il  s’agit  de  la  lettre ʿayn (‫)ﻉ‬. Elle est généralement notée à l’encre rouge, avec toutefois quelques occurrences à l’encre noire comme aux folios 143v et 257r. Elle est parfois couverte par un médaillon comme dans les folios 286r, 300r et 307v. Traditionnellement cette lettre, abréviation du mot ʿashara, indique le dixième verset selon une division dite « coufique » du texte coranique5. Cependant dans le coran de Gwalior, la lettre ʿayn est une abréviation du mot rukūʿ (inclination)6, une division coranique

pratiquée en Iraq, en Iran, en Inde et au Pakistan permettant d’apprendre et de mémoriser le Coran en deux ans. Suivant cette pratique, le manuscrit est divisé selon le contenu des versets et non en fonction de  leur  nombre ;  la  lettre  ʿayn détermine alors le moment de l’inclination au cours de la récitation du Coran7. Et en effet, à trois reprises dans le manuscrit de Gwalior, ce terme se trouve inscrit à l’encre rouge dans un énoncé en marge d’une sourate donnant au lecteur des indications précises sur le nombre de prosternations attendues durant la récitation de la sourate en question (figure 2). Ainsi, on peut lire  pour la sourate IV (fol. 73v) : « dar īn sūri bīst [u] -se rukūʿ ast » (dans cette sourate il y a vingt-trois rukūʿs) ;  dans les sourates VII (fol. 143v) et VIII (fol. 167r), sont annoncés de la même façon vingt-quatre et dix rukūʿs. Ces indications correspondent exactement au nombre de ʿayns que l’on trouve dans la marge des folios en question, ce qui établit de façon indubitable la signification de cette lettre dans notre manuscrit8. Les corrections On relève par ailleurs une autre catégorie de notes marginales qui soulignent des fautes ou des omissions commises par le copiste du texte coranique et qui sont corrigées et notées par endroits. Trois personnes pourraient avoir participé aux corrections : le copiste lui-même et deux mains ultérieures comme nous le montrerons plus loin. Ces corrections concernent principalement des omissions. Elles sont généralement inscrites dans la même graphie et les mêmes couleurs que le texte corrigé. Dans le cas d’une omission, un chevron indique l’endroit où le mot est manquant, et celui-ci  est donné dans la marge (figure 3)9.

7. 8. 4.  Gade (2004), 367 ; Denny (1989), 5-26. Concernant le tajwīd voir Nelson (1985), 14-31 où le chapitre 2 est consacré à  cet art primordial dans l’enseignement et la récitation du Coran. 5.  Al-Ṣabbāġ (1993), 88. 6. Il s’agit d’une inclinaison à angle droit avec les mains posées sur les genoux : Monnot, EI2 (1995), 961; voir également Gacek (2009), 315. J. Chelhod différencie les verbes présents dans le texte coranique rakaʿa (s’agenouiller) et sajada (se prosterner) bien que les deux semblent avoir été employés indifféremment comme synonymes de ṣallā

9.

(prier) dans les plus anciennes sourates. La différenciation entre ces deux termes n’est d’ailleurs pas explicite dans le texte coranique : Chelhod (1959), 166 ; voir également  Tottoli (2001), 254-255. Gacek (2001), 58. Notons que dans d’autres manuscrits coraniques proches du coran de Gwalior la lettre ʿayn est indiquée dans la marge de façon concomitante à l’indication « ʿashara » correspondant donc aux groupes de dix versets. Il s’agit du coran W563 du Walters Art Museum (Baltimore), ainsi que de deux autres corans indiens datables du xve siècle : Schimmel  (1970),  pl. XXII ;  Brac  de  La  Perrière  (2008),  pl. 43 ; et le coran dont deux feuillets sont conservés au  musée de l’Institut du monde arabe (Paris) sous le numéro d’inventaire AI 84.19. Ici une partie du verset, initialement omise, a été copiée dans la marge aux mêmes dimensions que l’inscription se trouvant à l’intérieur du jadwal.

88 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 1 – Indications de sajda, thalāth arbāʿ et ʿayn dans la marge, coran de Gwalior, fol. 166v. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 89

Figure 2 – La lettre ʿayn à côté de l’énoncé en persan « dar īn sūri bīst [u]‑se rukūʿ ast » inscrit au début de la sourate IV, coran de Gwalior, fol. 73v. [© By courtesy of the Aga Khan Museum]

90 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 3 – Un verset omis est inscrit verticalement dans la marge à l’encre dorée, coran de Gwalior, fol. 248r. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 91

Les corrections de seconde main et les réclames dont est doté ce manuscrit sont vraisemblablement de la même main. Elles sont notées à l’encre noire tandis que les énoncés corrigés à l’intérieur du jadwal sont raturés ou agrémentés de signes diacritiques à l’encre noire également (figure 4)10. Les corrections apportées par la troisième main sont inscrites à l’encre noire ou rouge dans les marges et sont parfois accompagnées d’un nom ou d’une lettre qui désigne l’un des qurrāʾ ou lecteurs, indiquant qu’elles ont été établies à partir de versions recensées11. Ainsi, l’un des atouts du coran de Gwalior est sa portée didactique ; il s’agit  principalement de la référence aux lectures canoniques conformes aux sept qurrāʾ et leurs transmetteurs dans l’établissement des variantes de lecture. Mentions relatives aux variantes de lecture et aux lecteurs (qirāʾāt et qurrāʾ) Le terme de qirāʾa est utilisé dans trois cas différents selon la définition de l’Encyclopédie de l’Islam12 : d’abord dans le sens d’une simple récitation de segments distincts du Coran, telle qu’elle est prescrite pour la prière rituelle (ṣalāt), puis dans le sens de la tilāwa ou récitation du Coran entier et enfin la recension et la transcription actives d’un mot ou d’un passage coranique, appelées « variantes de lecture »13.

10.  Ainsi par exemple, sur la figure 4, le mot wāsiʿ (le Vaste, un des noms de Dieu) est raturé et corrigé dans la marge par le mot samīʿ (l’Audient, un autre nom divin). 11. Relevons par exemple le cas du folio 117v qui présente la « correction d’une correction » : une première annotation rectificative à l’encre noire reprenait deux monèmes du  texte initial et a été par la suite raturée à l’encre rouge. 12. Paret, EI2 (1986), 129-132. 13. Ibid., 129. Les variantes de lectures ont été sélectionnées selon différents critères. Certaines expriment des variations de vocalisme et de diacritisme sans altération du sens de l’énoncé (phrase ou verset). D’autres lectures qui n’ont pas été retenues dans le corpus officiel des qirāʾāt affectent parfois l’ordre des mots, des sourates et se différencient nettement de la Vulgate ʿuthmānienne, qui  était dénuée de signes diacritiques. L’autorité sélective qui établira le choix des qirāʾāt acceptées divise les variantes de lecture en trois catégories : al-mutawātira (les plus connues et établies), al-ṣaḥīḥa (sûres) et les autres s’opposant à la recension de ʿUthmān,  considérées  comme  shāḏḏa « exceptionnelles » pour Blachère ou « déviantes » pour Leemhuis. Les plus célèbres de ses lectures rejetées et considérées comme shāḏḏa sont celles dites de ʿAlī,  d’Ubayy et d’Ibn Masʿūd ; concernant différents critères de  sélection des variantes de lecture, voir Blachère (1977), 116-117 et 200-210 ; Leemhuis (2004), 358-359. Il est intéressant de noter un exemple particulier de la variante d’Ibn Masʿūd qui considérait la Fātiḥa (première sourate de la Vulgate officielle du Coran) ainsi que les sourates  dites al-Muʿawwiḏatayn (deux dernières sourates) comme des duʿāʾs et non pas comme des versets : cf. Amir-Moezzi (2011), 78-79.

Les lectures selon cette dernière acception relèvent de la science des variantes, instaurée au début du xe siècle en réaction au développement foisonnant d’un nombre considérable de lectures qui menaçait l’intégrité du message initial14. Le principal personnage mentionné dans les sources médiévales de la plus importante réforme coranique depuis le recensement de ʿUthmān15 est Abū Bakr Ibn  Mujāhid al-Baghdādī (m. 324/936) qui sélectionna  sept lectures de qurrāʾ appartenant à la première et deuxième génération des compagnons du Prophète, leur conférant ainsi une autorité officielle. Ces qurrāʾ16 provenaient des cinq villes ayant reçu les cinq copies du codex ʿuthmānien et qui constituaient des centres intellectuels accueillant une grande tradition exégétique de l’époque : Nāfiʿ b. ʿAbd al-Raḥmān (m. 169/785) de  Médine, ʿAbdallāh Ibn Kathīr (m. 120/737) de La Mecque,  ʿAbd Allāh Ibn ʿĀmir (m. 118/736) de Damas, ʿĀṣim  b. Abī al-Najūd (m. 127/745), Ḥamza b. Ḥabīb al-Zayyāt  (m. 156/773) et ʿAlī b. Ḥamza al-Kisāʾī (m. 189/804) de  Kufa, et Abū ʿAmr b. al-ʿAlāʾ (m. 154/770) de Basra 17. 14. Cette réforme s’est imposée non sans réticence. Nombre de tentatives de sélection de variantes canoniques avaient été précédemment entreprises telle que celle de ʿUbayd al-Qāsim b. Salām (m. 224/838-839) qui inventoria 25 qurrāʾ, ou  celle  de  l’imam  Aḥmad  b. Jubayr  (m. 258/871-872)  intitulée Al-Khamsa : cf. Kafi (2008), 31. Viviane Comerro  qui a étudié les modalités de transmission des textes sur la constitution du muṣḥaf de ʿUthmān en confrontant  différents témoignages met en exergue la diversité des versions de la Tradition du ḥadīth pour une « légitimation » d’une variante de lecture et la survivance de nombre de qirāʾāt malgré la réforme officialisée : Comerro (2012). 15. Il convient de noter la disparité des récits conférant ce rôle de collecte du texte coranique au calife ʿUthmān. Cette  entreprise semble avoir débuté, selon certaines sources, sous Abū Bakr ou ʿUmar. Viviane Comerro parle d’une première étape d’enregistrement du Coran et de la constitution d’un « comité de rédaction » sous le califat d’Abū Bakr ; voir son chapitre 3, « Les motifs narratifs »  dans Comerro (2012). L’auteur note par ailleurs que dans le récit sur le jamʿ retenu par al-Bukhārī dans son Kitāb al-manāqib, l’entreprise de collecte du texte coranique a été effectuée par les trois califes : Abū Bakr, ʿUmar et ʿUthmān dans une volonté d’« unité littéraire » des traditions et d’une « légitimation » de l’acte de collecte du texte coranique : id., 94. 16. Dans un récent article sur les qurrāʾ, Mustafa Shah analyse les nouvelles lectures de l’historien M. Chaban (1975), 50-51. Il y relève la complexité de l’établissement de l’identité de certains lecteurs, cités dans la littérature classique, à cause de l’étanchéité des rôles joués par ces derniers lors des premières années de l’Islam et durant les  premiers  affrontements  armés :  Shah  (2005),  1-35 ;  voir également, Denny (1980). 17. Sur les lecteurs canoniques lire Melchert, Afsaruddin (2004), 367-392. L’article précise les listes des lecteurs et de leurs transmetteurs et indique des exemples de préférences d’adoption de certaines lectures selon les régions. Voir également l’ouvrage capital d’A. Jeffrey (1937) qui comprend les différentes variantes de plusieurs lecteurs de certains compagnons du Prophète et de ses épouses.

92 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 4 – Exemple de correction de seconde main, coran de Gwalior, fol. 38v. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 93

Cette sélection limita par conséquent le nombre de lectures canoniques, se fondant sur un ḥadīth affirmant  que le Coran a été révélé selon « sept aḥruf » (sing. ḥarf), qu’Ibn Mujāhid traduisit par « qirāʾāt » (lectures) bien qu’il y ait eu une quarantaine d’interprétations de ce terme18. Cette limitation des qirāʾāt par Ibn Mujāhid  ne fut acceptée par tous que progressivement, à partir du xe siècle19. Une fois fixées, ces lectures furent largement suivies dans les récitations publiques20. Il était convenu dès le viiie siècle qu’en s’en tenant au ductus consonantique  officiel,  le  lecteur  se  réservait  toujours  la  liberté de choisir parmi un certain nombre d’autorités. Dans certains cercles de lecture privés21, les qirāʾāt étaient régies par le principe de jamʿ (addition) qui consistait à segmenter les versets en unités et à les lire à chaque fois avec une variante canonique différente22. En procédant à une comparaison avec un coran timuride du Walters Art Museum23 qui renferme un nombre conséquent de gloses marginales, nous constatons que les ratures et les corrections du texte coranique sont plus fréquentes dans le coran de Gwalior. Dans les deux copies, certaines abréviations apparaissent également dans le corps du texte principal ;  elles peuvent être divisées en deux groupes. Il s’agit d’une part des lettres et des signes ponctuant et vocalisant le texte coranique ‫ ن‬،‫ م‬،‫ ال‬،‫ ق‬،‫ ط‬،‫ ص‬،‫ ز‬،‫ ج‬، ‫ت‬, et d’autre part des lettres liées : « ‫» قف‬, abréviation de yūqaf ʿalayh (pause à cet endroit) et « ‫» صلى‬, abréviation de al-waṣl awlā (la liaison est prioritaire)24.

Les abréviations en question sont inscrites dans l’espace du texte principal. D’autres abréviations figurent dans  les marges des deux corans : elles appartiennent au registre des qirāʾāt et font référence aux noms des lecteurs, les qurrāʾ. Ces dernières apparaissent dans les deux corans qui nous intéressent, mais dans des proportions sensiblement différentes. Dans le coran de Gwalior quelques folios présentent également des lettres isolées (figures 5-6) : fol. 195v  (alif, ḥāʾ), fol. 433v (wāw, ḥāʾ, ṣād, fāʾ, rāʾ), fol. 442r (rāʾ, kāf, ṣād, dāl, ḥāʾ). Ces lettres de l’abjad sont associées aux noms de qurrāʾ (lecteurs) reconnus, à savoir Nāfiʿ, Ibn Kathīr, Abū ʿAmrū, Ibn ʿĀmir ainsi que ceux d’Abū  Bakr et Ḥafṣ, transmetteurs (ruwwāt) d’un lecteur non cité qui est ʿĀṣim25. Le manuscrit W563 renferme lui aussi les noms de ces quatre qurrāʾ. Dans les marges mêmes des deux doubles frontispices initiaux figurent  les noms des sept qurrā’ ainsi que leurs ruwwāt26, inscrits en doré et accompagnés pour tous d’une lettre de l’abjad (figure 7). Les noms mentionnés dans les gloses du coran de Gwalior et de celui du Walters Art Museum sont ceux des sept qurrāʾ27 et de leurs transmetteurs, répertoriés et validés dans des traités dont le plus célèbre est celui de l’Andalou al-Dānī (m. 444/1053), al-Taysīr fī al-qirāʾāt al-sabʿ28. Celui-ci fut repris par le faqīh malékite andalou  Ibrāhīm  b. Mūsā  al-Shāṭibī  (m. 590/1194)29 dans son poème mnémotechnique (urjūza) intitulé : Ḥirz al-amānī wa wajh al-tahānī fī al-qirāʾāt al-sabʿ, plus connu sous le titre de Matn al-shāṭibiyya ou

18. Blachère (1977), 124. Sur le terme des « sept aḥruf » et les discussions concernant leur signification, voir Comerro  (2012), 119-136. 19. Ainsi le pouvoir de décision en matière des qirāʾāt fut simplement limité, sans être totalement rejeté. Ainsi la lecture d’Abū Jaʿfar b. al-Qaʿqāʿ (m. 130/747) n’a pas été admise parmi les sept lectures canoniques bien que le personnage fusse l’un des maîtres de Nāfiʿ. Cette variante fut par la suite ajoutée à la liste des Quatorze lectures canoniques  après  celle  des  Dix  d’Ibn  al-Jazarī  (m. 833/  1429), Al-Nashr fī al-qirāʾāt al-ʿashr, même si les lectures des Dix lecteurs ne connurent pas la même primauté que les Sept lectures canoniques : voir le chapitre « Primauté des Sept ‘Lecteurs’ » de R. Blachère (1977), 124-131 et n. 274. Concernant les critères de sélection des Dix lectures, voir Leemhuis (2004), 358. 20. Welch, Pearson, EI2 (1986), 410. 21. Rudi Paret considérait que ces cercles étaient formés principalement de lecteurs d’élite, car cette méthode d’enseignement et de lecture du Coran qui adoptait le principe de jamʿ est particulière et la grande partie des qurrāʾ se limitaient à une seule lecture : Paret, EI2 (1986), 131. 22. Blachère (1977), 112. 23. Coran W563 du Walters Art Museum de Baltimore, images numériques disponibles sur http://thedigitalwalters.org/ Data/WaltersManuscripts/html/W563/. 24. Abréviations de vocalisation et de ponctuation appelées ʿalāmāt al-waqf. Un glossaire exhaustif des termes techniques présents dans les manuscrits est proposé par Gacek (2001).

L’introduction des waqfs et des pauses a une portée scholastique essentielle, au même titre que l’interprétation du texte coranique, l’adoption de certains systèmes graphiques ou l’introduction de signes diacritiques ou de points de ponctuation. Un article récent d’Amr Osman paru en 2012, 90-109, est consacré aux différents débats qui accompagnèrent la détermination des waqfs dans la récitation du Coran. Il s’agit d’un domaine de recherche qui ne bénéficie  pas de l’intérêt qui lui est dû au vu des grands débats que ce domaine avait suscités et eu égard à son rapport aux traités des variantes de lecture. Dutton (1999) a étudié la présence des points de couleurs des manuscrits coufiques de la Boldeian Library d’Oxford,  de la British Library de Londres ainsi que de certaines reproductions de matériels de la collection Nasser David Khalili. Certains points de couleur désignent ainsi la référence à des variantes de lecture. Les noms correspondent en effet aux sept qurrāʾ, chacun possédant une tradition reconnue de deux riwāyas au sens de transmission autorisée des lectures canoniques. En plus de ces noms de qurrāʾ, le coran W563 offre des références à d’autres « lectures » et variantes, toujours mises en relief par une écriture en petits caractères à l’encre noire. Le traité a été édité notamment par Pretzl (1930). Neuwirth, EI2 (1998), 376-378.

25.

26. 27.

28. 29.

94 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 5 – Les abréviations alif et ḥāʾ ainsi que le nom du qāriʾ Ibn ʿĀmir, fol. 195v. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 95

Figure 6 – Les noms des qurrāʾ Nāfiʿ, Ibn Kathīr, Abū ʿAmrū inscrits dans la marge, coran de Gwalior, fol. 272v. [© By courtesy of the Aga Khan Museum]

96 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 97

Figure 7– Ce double frontispice renferme les noms des sept lecteurs canoniques et de leurs transmetteurs, accompagnés chacun d’une lettre de l’abjad, Coran W563, fol. 3v-4r. [© By courtesy of the Walters Art Museum]

98 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

al-Shāṭibiyya. Dans ce poème de 1173 vers, il qualifie  les lecteurs de « lunes éclairantes » (budūr) qui sont au nombre de sept, et leurs transmetteurs d’« astres » (shuhub)30. Il y annonce par ailleurs son intention de recourir à certaines lettres de l’alphabet pour se référer aux lecteurs et à leurs transmetteurs31 (figure 7). Il semblerait donc que le coran de Gwalior, qui est certes un objet d’apparat comme en atteste sa très riche ornementation, ait été produit dans un contexte érudit (et donc restreint). Les références aux qurrāʾ Nāfiʿ, Ibn Kathīr, Abū ʿAmrū, Ibn ʿĀmir, et aux  deux  transmetteurs  Abū  Bakr  et  Ḥafṣ,  le  prouvent  indubitablement. Aux côtés de ces annotations relatives aux qirāʾāt considérées comme canoniques, figurent quelques notes relatives concernant des actions performatives et des enseignements soufis.

Notes relatives à des actions performatives et à des pratiques soufies Outre ces mentions liées aux lectures canoniques, figurent dans le coran de Gwalior quelques notes relatives à des actions performatives : ces gloses en  arabe  − intitulées  faḍāʾil al-sūra (les vertus de la sourate)32 − ainsi que ces notes en persan proposent  la répétition de certains énoncés coraniques. Faḍl al-sūra33 Parmi les gloses marginales que recèle ce manuscrit figurent donc les faḍāʾil al-sūra, énoncés mentionnant les vertus d’une sourate et les bénéfices que peut en  rapporter la lecture34. Plusieurs recueils de hadiths et 30. Nous nous sommes référés pour la compréhension et la traduction de ces deux qualificatifs à l’ouvrage d’Abū al-Farah,  Hafiz (2003), 19 dont les auteurs expliquent que les sept  lecteurs étaient qualifiés de « lunes » parce qu’ils étaient  connus et reconnus et pour la complétude de leur savoir et leur prééminence. Les transmetteurs sont comparés à des « astres » car ils auraient emprunté la « lumière » des sept lecteurs et « illuminé » les ténèbres de l’ignorance. ً ِ‫ج َع ْلت أَبَا َجا ٍد َعلَى ُك ِّل َقارئ * َدل‬, َ ‫يال َعلَى ال َم ْن ُظوم أَو‬ َ ‫َّل أَو‬ 31. Dans ce vers ‫َّال‬ َ ٍِ ِ al-Shāṭibī souligne le classement de ces noms, à savoir :  noms du lecteur – noms de ses deux transmetteurs. 32. Ces formules se rattachant à la littérature des faḍāʾil ou Manāqib al-Qurʾān (les vertus du Coran) occupent une place prépondérante dans les sciences du ḥadīth : Sellheim, EI2 (1977) et Pellat, EI2 (1991). Ce genre a fait l’objet de peu d’études critiques, mais révèle pourtant des positions d’acteurs du milieu religieux et intellectuel concernant la fixation de la Vulgate avec ses différentes variantes de récitation-lecture (Afsaruddin a consacré un article à l’étude de ce genre littéraire ; cf. Afsaruddin (2002), 1-24). 33. Le texte recourt à ce terme à la place de « faḍāʾil » qu’on rencontre usuellement. 34. Par exemple selon le faḍl al-sūra de la sourate III (fol. 48r), quelqu’un qui lit cette sourate sera à l’abri de l’enfer un nombre de fois égal au nombre de versets de cette sourate.

d’exégèses comprennent ce genre de faḍāʾil mais ceux du coran de Gwalior ressemblent davantage aux ḥadīths cités dans le Kashf al-asrār wa ʿuddat al-abrār, exégèse datant du xiie siècle qui a pu servir de modèle pour certaines exégèses persanes35. L’ouvrage en question a été rédigé par Abū al-Faḍl Rashīd al-Dīn Maybudī,  un mystique sunnite de l’école chafiite. Le texte rédigé par ce dernier, est d’une importance primordiale car il s’agit du premier commentaire coranique soufi rédigé en langue persane permettant un emploi  pédagogique du livre saint. Ce texte rassemble à la fois la traduction, les commentaires exotériques ainsi que certains énoncés mystiques liés au Coran. Il est par ailleurs fondé sur le commentaire du même titre du célèbre soufi d’Hérat ʿAbd Allāh al-Anṣārī (m. 465/  1089) et enrichi par des sources classiques comme Ḥaqāʾiq al-tafsīr d’al-Sulamī (m. 412/1021) et Laṭāʾif al-ishārāt d’al-Qushayrī (m. 465/1072)36. Dans le coran de Gwalior, l’énoncé du faḍl al-sūra est rédigé tout à la fois en arabe et en persan. Il est inscrit dans la marge à l’encre rouge, accompagné de la mention en arabe « faḍl al-sūra » ainsi que, pour la traduction en persan, de « ḥāṣil-i maʿnī ānast ki … » (littéralement : le résultat de la traduction étant…) à l’encre bleue cette fois. Immédiatement après la mention « faḍl al-sūra », la sentence est introduite par ces mots : « ʿan rasūl Allāh » (selon le Prophète) puis se poursuit par l’évocation des bénéfices de la récitation  de la sourate en question. Les faḍāʾil al-sūra apparaissent immédiatement après le titre de chaque sourate, à l’exception des sourates I, XXIII, LXIV, CIX et de la dernière sourate où rien ne figure. La copie de ces  énoncés est contemporaine de celle du texte principal (figure 8). Les faḍāʾil al-sūra sont fondés sur un long ḥadīth – remis en question par les muḥaddithūn37 – attribué à Ubayy b. Kaʿb. Le Prophète y énumère les bienfaits

35. Keeler (1999), 119-123. 36.  Dans son travail Maybudī réalise une synthèse entre les  méthodes exotériques et ésotériques de la lecture du Coran et une fusion entre doctrines strictement traditionalistes et les interprétations métaphoriques. La singularité de l’ouvrage de Maybudī réside en sa structuration en trois  niveaux (nawbat). Il expose d’abord les explications littérales du texte coranique arabe en langue persane, ensuite les interprétations classiques exotériques qui comprennent différentes lectures, les asbāb al-nuzūl (les circonstances de la Révélation), les traditions doctrinales, les implications juridiques et des illustrations anecdotiques, et enfin les  interprétations ésotériques qui constituent la substance du commentaire soufi véritablement mystique et ésotérique :  Böwering (2006), XVI. 37.  Tels que Ibn al-Jawzī, al-Zarkashī et al-Suyūṭī, voir Afsaruddin (2002), 21. Les muḥaddithūn sont les spécialistes de la Tradition, « l’ensemble des transmetteurs (du ḥadīth) autorisés », Comerro (2012), 5 et Robson, « Ḥadīth », wEI2.

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 99

Figure 8 – Faḍl al‑sūra de la sourate XIX (Maryam), coran de Gwalior, fol. 275v. [© By courtesy of the Aga Khan Museum]

100 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 9 – Prière (duʿāʾ) de la sourate VI (Al-Anʿām), coran de Gwalior, fol. 132r. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 101

apportés par la récitation de chaque sourate38. La chaîne de transmission de ce ḥadīth trouve son origine chez Nūḥ b. Abī Maryam, l’un des disciples de l’imam  Jaʿfar al-Ṣādiq39. Quand Nūḥ b. Abī Maryam fut interrogé sur la source de ce ḥadīth, il répondit : « J’ai vu que les gens se détournaient du Coran et préféraient s’intéresser au fiqh d’Abū Ḥanifa et aux histoires (Maghāzī) d’Ibn Isḥāq ; j’ai donc inventé cette phrase avec l’intention de plaire à Dieu (ḥisbatan) afin d’inciter les gens  à revenir vers le Coran »40. Ainsi la présence des faḍāʾil dans les marges du coran de Gwalior indiquerait une volonté d’encourager la récitation du Coran par un public plus large. Suggestions de répétition

par diverses prières. Les informations ayant trait à ces pratiques et plus particulièrement à ce genre de prières et leurs origines demeurent fort peu nombreuses. Christiane Gruber42 et Alexandra Bain43 comptent parmi les chercheurs qui s’y sont intéressés. Leurs études de manuels de prières (Duʿāʾnāma) ottomans datant des xviiie et xixe siècles, contenant des prières de ce type liées à la sourate VI (Enʿam-ı Şerif), révèlent des indices de pratiques soufies (figure 9). Un récit métaphorique comme indice du contexte d’utilisation Une note singulière présentant un texte en langue persane apparaît dans la marge du folio 518v  (figure 10).  Il  s’agit  d’un  petit  commentaire  sous  forme de parabole du verset 19 de la sourate LXX. Sa traduction est la suivante :

Entre les folios 124v et 143r où prend place la  sourate VI, sept folios comportent des notes à l’encre noire en différents endroits de la marge, aussi bien en haut qu’en bas ou au milieu de la marge. Ces notes, rédigées en persan, suggèrent au lecteur de répéter un certain nombre de fois un duʿāʾ, invocation adressée à Dieu en arabe. La quasi-totalité de ces notes commencent par : « wa chun dar īn maḥal risad ki… » (quand il arrive à cet endroit où…) ou par « chun īnjā birisad ki… » (quand il arrive ici où…) et est suivi d’un mot appartenant à un  verset se trouvant dans le corps du texte de la même page. Ensuite, il est proposé au lecteur de répéter une, trois, onze, quinze ou soixante-dix-sept fois des invocations extraites pour la plupart de différents versets d’autres sourates41. Le verset 44 de la sourate XL (al-Ghāfir) « wa ufawwiḍu amrī ilā Allāh » (J’abandonne mon  sort  à  Dieu),  qui  fait  référence  à  la  confiance  absolue en Dieu, est quant à lui utilisé en guise de commentaire marginal au folio 124v. Cette pratique est aujourd’hui encore répandue en  Iran,  où  durant  une  cérémonie  appelée  Khatm-i Anʿām, la sourate VI (al-Anʿām) du Coran est entièrement récitée, entrecoupée à la fin de certains versets 

Une version arabe de cette anecdote se trouve dans l’ouvrage Kashf al-asrār dans l’exégèse consacrée à la sourate LXX45. Maybudī y affirme que la finalité d’un  tel commentaire dans un manuscrit coranique est de

38.  Al-Zarkashī [1990], 56 ; Goldziher (1967-1971), 149 ; la  traduction anglaise de la première partie de ce ḥadīth se trouve dans l’article d’Afsaruddin (2002), 21. Notons que dans l’article sus-cité les vertus prêtées aux sourates III et IV s’apparenteraient à ce qu’on rencontre dans le coran de Gwalior dans les fols. 48r et 73v. 39.  Al-Ḥusaynī al-Tifarishī, Naqd al-rijāl, vol. V. Āl al-bayt liʾiḥyā al-turāth, en ligne, http://rafed.net/booklib/view. php?type=c_fbook&b_id=136 ;  Afsaruddin  (2002),  21-22. 40.  Al-Zarkashī [1990], 60. 41. Notons que dans son œuvre fondamentale Biḥār al-anwār, Muḥammad Bāqir al-Majlisī (m. 1110/1698) cite un khabar du Prophète où des duʿāʾs similaires sont à répéter (onze et quarante-six fois) pendant la récitation de la même sourate VI, al-Majlisī [1983], 331-332 ; sur al-Majlisī, voir  Hairi, EI2 (1986), 1084-1085.

42. Gruber (2009), 117-153. 43. Bain (2001), 213-235. 44. Voici la version originale de l’énoncé persan : ‫[آورده اند] که جميع مخلوقات سه صنف اند که برای [رزق] سعی ميکنند‬ ‫] اندیشه ميکنند یکی بنی آدم و دوم‬...[ ‫و ذخيره می سازند و برای وقت‬ ‫] ذخيره ندارند [و] سعی نکنند‬...[ ‫موش و سيم [مور] چه و دیگر مخلوقات‬ ]‫و رزق خود را مفروغ گردانند [حکایت] آورده اند که نزدیک کوه قاف [؟‬ ‫دنياست آفریده ست و آنجا یکی خری [آ]فریدست که نام آن خر هلوعاست و‬ ‫در آن زمينی نيمی آب و نيمی کاه و حق تعالی آن خر را چندان [رزق] داده‬ ‫است که هر روز آن تمام آب و کاه [؟] می کند و فربه چنان می شود که از‬ ‫جای [نمی]تواند جنبيد چون شب در می آید آن صحرا را [خالی] می بيند‬ ‫پس اندوه ميکندکه چو بامداد [شود] چه خواهم خورد از این اندیشه ضعيف‬ ‫و نحيف[و] الغر می گردد باز چون بامداد می شود اهلل تعالی زمين را هم‬ ‫بدان صفت ميگرداند باز می خورد [؟] همبران حال هر روز حال هلوعا‬ ‫همين است [؟] اهلل[؟] ان االنسان خلق هلوعا ه ه ه ه ه‬ 45.  Maybudī [1344/1965], 288.

« On raconte que les créatures sont de trois types quant à leur recherche de nourriture et à la manière de l’entreposer : premièrement l’être humain, deuxièmement la souris et troisièmement la fourmi. Les autres créatures ne s’en préoccupent pas. À proximité de la montagne Qāf  se  trouve  un  âne  qui  s’appelle  Halūʿā  (c’est-à-dire avide). Il a un terrain à sa disposition à moitié rempli d’eau, l’autre moitié étant remplie de paille. Il a de la nourriture en quantité suffisante mais  il en mange tellement durant la journée qu’il grossit et qu’il ne peut plus bouger. Quand la nuit tombe, le terrain lui semble vide. Alors il s’inquiète de ce qu’il va manger le lendemain si bien qu’il devient maigre et faible alors que Dieu ne manquera pas de remplir à nouveau le terrain le lendemain. La nature de cet âne est semblable à celle de l’homme : ce sont [tous deux] des créatures avides »44.

102 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 10 – Récit métaphorique, coran de Gwalior, fol. 518v. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 103

faciliter la compréhension de ce passage coranique pour un large auditoire46. En employant un style narratif auquel participent plusieurs anecdotes et paraboles, cette catégorie de commentaires relève d’une méthode d’enseignement très populaire utilisée chez certains imams et se retrouve dans des traditions orales d’enseignement et d’exégèses soufis47.

Fālnāma Introduction Parmi les anciennes techniques divinatoires courantes en Orient musulman, notamment dans les zones indo-persanes, le fālnāma (livre augural)48 désigne une catégorie de textes divinatoires rattachée à un recueil de poésie (tel que le Diwān de Ḥāfiẓ)  ou à un manuscrit coranique49. Le fālnāma du coran de Gwalior, qui est pour l’heure le plus ancien des fālnāmas connus associé à un manuscrit coranique, est contemporain de la copie du texte principal et sans doute l’œuvre du même copiste. Un fāl-i muṣḥaf (fālnāma coranique) appartenant à un recueil persan d’œuvres littéraires et intitulé Safīna-yi Tabrīz50 est certes reconnu comme étant plus ancien, mais il n’est pas annexé à un coran51. Le fālnāma du coran de Gwalior, qui occupe six pages du manuscrit, est disposé à la fin du codex, du fol. 551v au fol. 554r.  Il est situé entre la prière d’achèvement de la lecture (duʿāʾ-yi khatm-i qurʾān) et le colophon. Avant d’aborder plus en détails le fālnāma du coran de Gwalior, il convient de s’attarder sur l’énoncé qui suit la prière d’achèvement de la lecture au folio 551r (figure 11). L’originalité de ce duʿāʾ repose sur l’énonciation de l’ensemble des lettres de l’alphabet arabe. Au cours de la prière, le requérant demande à se voir attribuer une suite de bénéfices, chacun ayant pour  initiale une des vingt-huit lettres de l’alphabet arabe : « Allāhumma arzaqnā bi-kulli ḥarfi min Qurʾān ḥalāwa wa bi-kulli kalimāt karāma […] Allāhumma arzaqnā bi-al-alif ulfa wa bi-al-bāʾ baraka wa bi-al-tāʾ tawba… » (Ô Allah, nourris nous grâce à toutes les lettres du Coran de la douceur et avec tous les mots de la générosité [...] avec alif de l’amitié et avec bāʾ du bénéfice et avec tāʾ

46. 47. 48. 49.

Böwering (2006), XVI. Keeler (2006), 8 et 93. Massé, EI2 (1977), 779. Dans ce cas il est appelé aussi fāl-i muṣḥaf (divination par le Codex) ou fāl-i Qurʾān (divination par le Coran), Gruber (2011), 33. Pour une étude récente sur les fālnāmas portant sur une période ultérieure, mais contenant des précisions sur les différents aspects des livres de divination, voir : Farhad, Bağcı (2009). 50. Ibn Masʿūd Tabrizī [2003], 397. 51. Voir l’article de Gruber (2011), 33.

de la repentance...). Le procédé d’attribution de valeurs omineuses aux lettres de l’abjad fait partie des principes de ʿilm al-ḥurūf (la science des lettres) comprise dans le jafr52, qui rassemblerait toutes les méthodes divinatoires fondées sur les lettres de l’alphabet, utilisées probablement dans le fonctionnement du fālnāma. Il s’agit d’une branche du jafr qui concernait l’onomatomancie qui, dans certaines sectes ésotériques, acquit le statut de pratique magique. Cette science repose sur les propriétés occultes des lettres de l’alphabet et des noms divins qu’elles composent. À chacune des vingt-huit lettres de l’alphabet arabe est attribuée une valeur numérique. Ces lettres sont aussi regroupées en quatre catégories, correspondant aux quatre éléments, c’est-à-dire l’eau, l’air, la terre et le feu. Par leurs associations avec ces derniers, les lettres seront alors en mesure d’agir sur la nature, aussi bien sur un plan physique qu’astrologique. L’équivalence numérique des lettres et leur coordination avec les quatre éléments sont les seules caractéristiques qui relèveraient d’une forme de logique, le hurufisme relevant dans les autres cas uniquement  d’expériences mystiques. Cette science nécessite l’intervention divine pour dévoiler les secrets du passé et du présent et prédire l’avenir. « Ainsi de part son objet, la « noble science des ḥurūf » occupe un rang privilégié parmi les techniques divinatoires en Islam, car elle est en rapport étroit avec les spiritualia (al-rūḥāniyyāt) et l’astrologie. Sa noblesse provient aussi de son rapport intime avec l’arithmétique, considérée par les anciens comme le principal pilier de la connaissance »53.

La description du texte Le fālnāma de ce coran, copié alternativement en rouge et en bleu, à l’instar du texte coranique luimême, se compose d’un guide de divination et d’interprétation de quelques versets du Coran (figure 12).  Il est constitué d’une notice expliquant la manière de consulter le fāl (augure) puis du texte divinatoire lui-même. La notice expose au consultant les conditions à respecter pour bénéficier d’une réponse à sa requête.  En voici la traduction : « Il  faut  être  purifié  (ba ṭahārat) et que le Coran soit complet. Il faut d’abord prier (deux rukʿa namāz), à chaque rukʿa il faut réciter une fois la sourate Fātiḥa, trois fois [la sourate] Ikhlāṣ. Après la salutation à la fin  de la prière (salām), il faut se prosterner trois fois et dire : Ô Seigneur accorde-moi du bonheur et non du malheur.

52. Fahd (1987), 221. 53. Id. (1990), 617.

104 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Figure 11 – Prière d’achèvement de la lecture du Coran, coran de Gwalior, fol. 551r. [© By courtesy of the Aga Khan Museum]

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 105

Figure 12 – Fālnāma du coran de Gwalior, fol. 551v-552r. [© By courtesy of the Aga Khan Museum]

Puis il faut réciter trois fois la sourate Fātiḥa et trois fois la sourate Ikhlāṣ, et puis réciter la prière suivante : « Il a les clefs de l’Invisible qui ne sont connues que de Lui. Il sait ce qui est sur la terre ferme et dans la mer. Nulle feuille ne tombe qu’Il ne le sache. [Il n’existe] ni graine dans les ténèbres de la terre ni [brin] desséché qui ne soit [consigné] dans un écrit explicite »54. Et puis il faut prier : « Ô Seigneur je recherche l’augure par Ton livre généreux, j’ai confiance en toi, alors préviens-moi de ce qui est caché dans Ton secret gardé (sirr-i maknūn) »55.

54. Coran, VI : 59. 55. Voici la version originale de l’énoncé persan : ‫چون خواهد که فال مصحف ببيند شرایط آنست که با طهارت باشد و‬ ‫مصحف جامع باشد اول دورکعت نماز بگزارد و بخواند در هر رکعتی‬ ‫فاتحه یکبار و اخالص سه بار چون سالم دهد سر بسجده سه بار بگوید الهم‬ ‫اعطنی خيرا و ال تعطنی شرا یا عزیر بعده سه بار سوره فاتحه و سه بار‬ ‫ بعده این آیت بخواند وعنده مفاتح الغيب ال یعلمها‬.‫سوره اخالص بخواند‬ ‫إال هو ویعلم ما في البر والبحر وما تسقط من ورقة إال یعلمها وال حبة في‬ ‫ بعده این دعا بخواند‬.‫ظلمات األرض وال رطب وال یابس إال في كتاب مبين‬ ‫الهم انی تفألت بکتابک الکریم و توکلت عليک فاخبرنی ما هو المکتوم فی‬ ‫سرک المکنون فی غيبک یا عالم غيب و الشهادة انک علی ما تشاء قدیر‬ ‫ بعد آن نيت کند‬.‫الهم انت الحق انزل علی الحق بحق محمد المبعوث بالحق‬ .‫مصحف بگشاید هفت ورق بگرداند و هفت خط بشمرد‬

Par la suite, il faut émettre un vœu avec une intention pure (niyya) et ouvrir le coran, feuilleter sept pages en arrière et compter sept lignes. À titre d’exemple, si le début du verset commence par la lettre alif, le fālnāma propose alors un verset débutant par un alif avec les interprétations associées à l’augure suivant : « Ô consultant de (ce) fāl tu es libéré du chagrin et de la souffrance ». Une étude comparative des fālnāmas Nous avons mené conjointement l’étude de trois autres textes similaires rattachés à des corans de l’Inde des sultanats. Le premier texte fait partie d’un coran conservé à la Bibliothèque nationale de France sous la cote Arabe 7260 (fols. 1276r-1278r)56. Il est écrit en langue

56. Brac de la Perrière (2008), 115. Les images du Ms. Arabe 7260 sont disponibles sur le site Gallica de la Bibliothèque nationale de France, à l’adresse suivante : http://gallica.bnf.fr/ark:/12148/btv1b84272593.

106 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

persane à l’encre noire, les lettres individualisées sont écrites en rouge. Le deuxième fālnāma persan appartient au Coran Pl.1 de la collection Keir et est inédit57 (figure 13) et le troisième appartient au coran  conservé au Walters Art Museum dont il a été question plus haut. Ce fālnāma qui se trouve au folio 2r, a probablement été ajouté ultérieurement. Il est copié à l’encre noire, en arabe, sur une seule page et sans aucune décoration58. Le texte du fālnāma de Gwalior s’adresse directement à l’utilisateur, comme en atteste l’emploi des pronoms personnels et l’utilisation de la deuxième personne du singulier, tandis que dans les fāls des trois autres corans l’adresse au lecteur est indirecte, rédigée à la troisième personne du singulier. Si le fāl du coran du Walters Art Museum se distingue par l’absence de décor et sa rédaction en langue arabe, celui de Gwalior manifeste son originalité par le contenu même de ses prédictions. D’un point de vue formel, le fālnāma de la collection Keir est le plus proche de celui de Gwalior. Dans les deux cas, le texte s’inscrit en diverses couleurs dans des cadres oblongs et les graphies sont très ressemblantes. Les lettres de divination copiées à l’encre dorée prennent place dans de petites cases rectangulaires positionnées au centre des bandeaux de texte. Du point de vue du contenu en revanche, seul le fālnāma du Walters Art Museum est dépourvu d’introduction. Les deux autres corans qui nous intéressent, comme le manuscrit de Gwalior, présentent une introduction expliquant le rituel à effectuer avant la consultation du livre divinatoire. Concernant la notice explicative, le manuscrit de Gwalior préconise, avant l’ouverture du livre, une prière, au sens de ṣalāt, suivie de la récitation de prières, duʿāʾs. Les duʿāʾs des textes de la Bibliothèque nationale de France et de la Keir collection constituent un groupe relativement homogène duquel s’éloigne le coran de Gwalior : dans les deux premiers, il faut réciter les versets 255 à 257 de la sourate II, Ayāt al-kursī, ce qui n’est pas indiqué dans le texte de Gwalior. Dans leurs introductions, les fāls des corans de la collection Aga Khan et de la collection Keir se rejoignent lorsqu’ils indiquent : « Si on veut consulter un fāl du Coran, il faut que le Coran soit complet, quand tu veux faire le fāl, il faut d’abord que tu sois pur ». Quant à leurs contenus, ces quatre fālnāmas se montrent parfois contradictoires, à l’instar de ce qu’on peut observer en comparant les textes du manuscrit 57. Nous remercions Christiane Gruber de nous avoir, la première, communiqué des photographies. 58. Disponible en ligne : http://thedigitalwalters.org/Data/ WaltersManuscripts/html/W563/, image no 6. Voir l’article de Simon Rettig dans le même volume.

Figure 13 – Fālnāma du Coran Pl. 1. [© The Keir Collection, cliché : É. Brac de la Perrière, by kind permission of Ranros Universal SA]

de Gwalior et celui de la Keir ; en revanche les contenus des deux fālnāmas de la Keir et de la Bibliothèque nationale de France s’avèrent plus proches (table 1).

concLusion Les gloses et le fālnāma annexés au coran de Gwalior révèlent un usage du manuscrit à des finalités autres  que liturgiques ou théologiques. Diverses « pratiques » du Coran, comme les faḍāʾil al-sūra, la répétition de certains duʿāʾs et le fāl sont d’ailleurs regroupées sous l’appellation « khawāṣ al-qurʾān »59, ou l’« art de tirer des pronostics des versets qurʾāniques auxquels on attribue des vertus salutaires »60. Ainsi le Coran est utilisé dans l’esprit de la rhapsodomancie (ʿilm al-qurʿa) et l’onomatomancie (voir ʿilm al-ḥurūf)61. Les pratiques évoquées par le coran de Gwalior et les faḍāʾil al-sūra souvent identiques à ceux du Kashf al-asrār, amènent à penser que le manuscrit s’inscrit

59. Voir l’article de Gruber (2011), 31. 60. Fahd, « K̲h̲awāṣ al-Ḳurʾān », wEI2. 61. Ibid.

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 107

Lettre

Coran de Gwalior

Coran du Walters Art Museum W563

Coran de la collection Keir Pl.1

Coran de la BnF Arabe 7260

Alif

‫الف‬

Ô consultant de (ce) fāl tu es libéré du chagrin et de la souffrance

Il sera (…) dans le bien

Le bénéfice est  dans cette affaire

Le bénéfice est  dans cette affaire

Bāʾ

‫ب‬

Il (Dieu) t’offrira la richesse éternelle

Il sera en contact avec une personne haut placée

Il bénéficiera de profits  de la part d’un riche et son vœu sera exaucé

Le consultant du fāl bénéficiera  de la richesse

‫خ‬

Bonne nouvelle : tu auras une grande richesse. Si tu as fait cette divination pour un malade, il sera guéri et sa souffrance sera finie

Il tombera dans la peur

Cela est risqué : il faudrait donner une aumône (ṣadaqa)

Il court un risque qu’il faut empêcher par une aumône

‫ض‬

Ô… tu recevras un choc qui te rendra triste, il ne faut pas te hâter dans cette affaire [… après quelques] jours ton vœu sera exaucé

Il sera chargé d’une mission (d’un poste)

‫ف‬

Tu es libéré du chagrin et de la souffrance. Dieu le plus haut t’aidera et si tu veux partir en voyage, fais-le c’est une bonne intention

Ses affaires dispersées seront rassemblées

‫ي‬

Bonne nouvelle : Dieu a fait [attention] à toi [il faut le remercier] et ton vœu sera exaucé

Il recevra la nouvelle qu’un absent est en vie

Khāʾ

Ḍāʾ

Fāʾ

Yāʾ

Il sera rassuré par un chef Il bénéficiera d’un  d’une communauté et ses soutien et sera rassuré souhaits seront exaucés

Grâce à Dieu, ses affaires dispersées seront rassemblées

Toutes ses affaires dispersées seront rassemblées et accomplies

(illisible)

On l’informera [il recevra] de personnes absentes d’heureuses nouvelles

Table 1 – Quelques exemples d’augures dans les quatre fālnāmas des corans de Gwalior, du Walters Art Museum, de la collection Keir et de la Bibliothèque nationale de France.

dans la tradition fondée par Maybudī. Celui-ci rassemble dans son écrit traduction, commentaires exotériques ainsi que divers énoncés mystiques. Maybudī affirme au demeurant que le but de son  ouvrage est de pouvoir servir de référence pour un prédicateur qui sera à même de s’en inspirer devant le grand public62. La variété des gloses marginales, la présence de ratures, de corrections, attestent d’un usage scrupuleux et répété du manuscrit63. La proposition des divers modes de lecture canonique qui sont associés aux qirāʾāt corrobore aussi le fait que le coran de 62. Böwering (2006), XVI. 63. Les réclames présentes en bas des pages à l’encre noire ainsi que les différentes annotations étudiées l’attestent, à l’instar d’un autre fait important, à savoir la présence au fol. 512v de l’énoncé « waqf munzal (‫ » )وقف منزل‬qui préciserait l’inscription de notes de relecture du texte coranique. Le terme waqf comme dans l’énoncé commun « waqaftu ʿalā » (« J’ai lu ») est une phrase d’ouverture de notes de lecture : Gacek (2001), 153.

Gwalior n’était pas seulement un bel ouvrage, mais qu’il était avant tout destiné à un usage savant. De cette étude s’est dégagé un faisceau de présomptions qui nous permet de proposer de très probables liens avec les milieux soufis64. On peut penser à l’anecdote ponctuelle de l’âne avide ou aux faḍāʾil al-suwar qui se trouvent aussi dans le célèbre commentaire  soufi  de  Kashf al-asrār wa ʿuddat al-abrār, ainsi qu’à la répétition de certains duʿāʾs qui n’est pas sans rappeler certaines pratiques de ḏikr des soufis. Enfin,  force  est  de  constater  que  ce  coran  offre  plusieurs niveaux de lecture, certains immédiats, d’autres plus savants et d’autres encore plus cachés, liés à des pratiques divinatoires à l’instar du fālnāma. Là encore, cette « stratigraphie de la connaissance » apparente ce manuscrit à la pensée soufie. 64. Parmi l’abondante littérature consacrée aux divers courants soufis en Inde médiévale, on pourra consulter  en guise d’introduction les travaux de Schimmel (1996), 423-492 et de Rizvi (1978).

108 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

Confirmant l’importance de la transmission orale  dans les usages et la finalité de tout manuscrit coranique, les éléments hétéroclites présents dans le coran de Gwalior révèlent une conception originelle du manuscrit en vue d’un emploi fréquent. Ses marges semblent avoir été conçues pour recevoir des gloses. Les différents maniements dans le tracé des lignes du jadwal (qui paraissent s’être adaptées à l’inscription de certaines indications de divisions et de corrections à la suite de relectures) attestent de ce fait65.

Ce manuscrit s’inscrit dans une sphère indo-persane par plusieurs aspects : les titres de sourates66, la traduction interlinéaire en persan ainsi que la copie en graphie bihārī et son colophon signé. Toutefois sa particularité majeure réside dans le fait de renfermer la plupart des éléments qu’un manuscrit coranique est susceptible de posséder. Ces fonctions multiples témoignent des usages pratiques du coran dans une communauté vraisemblablement exotérique de l’Inde des sultanats.

65. Il s’agit d’une pratique courante, voir Déroche (dir.) (2000), 188. Le tracé du jadwal se retrouve quelques fois interrompu juste à l’endroit où est inscrit une lettre ʿayn, un énoncé corrigé ou un faḍl al-sūra. L’intégration des rectifications à la suite de relectures, les suggestions de lecture  et l’adjonction des faḍāʾil al-sūra, etc., sont vraisemblablement préalablement prévues, certaines lignes du jadwal

ont été réalisées de ce fait a posteriori. Ceci nous conforte dans l’idée que le coran de Gwalior était destiné à un usage pratique et fréquent de récitation, de lecture, de mémorisation et de prière en plus d’être un manuscrit remarquablement enluminé. 66. Une liste complète des différents titres des sourates est consultable dans Paret (2005), 551-559.

bibLioGraPhie sources iBn masʿūd tabrIzī (Abū al-Majd Muḥammad) [2003] Safineh-ye Tabriz. A Treasury of Persian Literature and Islamic Philosophy, Mysticism, and Sciences, édition fac-similé d’un manuscrit compilé et copié en 721-723/1321-1323, Téhéran : Markaz-i nashr-i Dānishgāhī. al‑MajlIsī (Muḥammad B.) [1983] Biḥār al-anwār al-jāmiʿa li durar akhbār al-aʾimma al-aṭhār, Beyrouth : Muʾassasāt al-wafāʾ. Maybudī (Rashīd al-Dīn) [1344/1965] Kashf al-asrār wa ʿuddat al-Abrār, édité par ʿAli Aṣghar Ḥikmat, Téhéran : Ibn-i Sinā. al‑zarkashī (Abū ʿAbd Allāh Badr al-Dīn Muḥammad  b. ʿAbd Allāh b. Bahādur) [1990] Al-Burhān fī ʿulūm al-Qurʾān,  Muḥammad  Abū  al-Faḍl Ibrāhīm (dir.), Beyrouth : Dār al-maʿrifa, 4 vol.

Références (hors Encyclopédie de l’Islam) abū al‑Farah (Sayyid Lashin), haFIz (Khalid M.) 2003 Taqrīb al-maʿānī fī sharḥ ḥirz al-amānī fī al-qirāʾāt al-sabʿ, Médine : Maktaba dār al-zamān.

aFsaruddin (Asma) 2002 « The excellences of the Quran: textual sacrality and the organization of early islamic society », dans Journal of the American Oriental Society 122, no 1, p. 1-24. aMIr‑MoezzI (Mohammed Ali) 2011 Le Coran silencieux et le Coran parlant. Sources scripturaires de l’islam entre histoire et ferveur, Paris : CNRS éditions. Bain (Alexandra) 2001 « The Enʿam-ı Şerif. Sacred text and images in a late ottoman prayer book », dans Archivum Ottomanicum, Wiesbaden : Harrasowitz Verlag. Blachère (Régis) 1977 Introduction au Coran, Paris : Maisonneuve et Larose. Blair (Sheila) 2006 « Uses and functions of the qurʾānic  text »,  dans Mélanges de l’Université Saint-Joseph LIX, p. 183-201. Böwering (Gerhard) 2006 « Foreword », dans Annabel Keeler (dir.), Sufi Hermeneutics: The Qurʾan Commentary of Rashid al-Din Maybudi, Londres : Oxford University Press, p. XV-XX.

LES GLOSES MARGINALES ET LE FA¯LNA¯MA DU CORAN DE GWALIOR • 109

Brac de la Perrière (Éloïse) 2008 L’art du livre dans l’Inde des sultanats, Paris : Presses de l’Université Paris-Sorbonne.

2001 « Traductions, impressions et usages du Coran dans le sous-continent indien (1786-1975) », dans Revue de l’histoire des religions, 218, no 1, p. 97-111.

chaBan (Muhammad) 1975 Islamic History: A New Interpretation, Cambridge : Cambridge University Press.

gacek (Adam) 2001 The Arabic Manuscript Tradition. A Glossary of Technical Terms and Bibliography, Leyde/Boston/ Cologne : Brill. 2009 Arabic Manuscripts. A Vademecum for Reader, Leyde/Boston/Cologne : Brill.

chelhod (Joseph) 1959 « Les attitudes et les gestes de la prière rituelle dans l’Islam », dans Revue de l’histoire des religions 156, no 2, p. 161-188. comerro (Viviane) 2012 Les traditions sur la constitution du muṣḥaf de ʿUthmān (Beiruter Texte und Studien, 134), Beyrouth / Würzburg : Orient-Institut/ Erlon Verlag. denny (Frederick M.) 1980 Exegis and Recitation. Their Development at Classical of Qurʾanic Piety, Leyde : E. J. Brill. 1989 « Qurʾan recitation: a tradition of oral performance and transmission », dans Oral Tradition 4, nos 1-2, p. 5-26. dÉroche (François) (dir.) 2000 Manuel de codicologie des manuscrits en écriture arabe, Paris : Bibliothèque nationale de France. dutton (Yasin) 1999 « Red dots, green dots, yellow dots and blue: some reflections on the vocalisation of early  qurʾanic manuscripts. Part I », dans Journal of Qurʾānic Studies 1, no 1, p. 115-140. eFthymiou (Marie) 2006 « Un mode d’emploi du Coran. Note sur les gloses persanes d’un manuscrit coranique d’Asie centrale », dans Mélanges de l’Université Saint-Joseph LIX, p. 209-228. Fahd (Toufic) 1987 La divination arabe : études religieuses, sociologiques et folkloriques sur le milieu natif de l’Islam (coll. La bibliothèque arabe. Hommes et sociétés), Paris : Sindbad. Farhad (Massumeh), bağcI (Serpil) 2009 Falnama, The Book of Omens, Washington (DC) : Sackler Gallery. gaBorieau (Marc) 1996 « Le sous-continent indien », dans Alexandre Popovic, Gilles Veinstein (dir.), Les voies d’Allah : Les ordres mystiques dans l’islam des origines à aujourd’hui, Paris : Fayard, p. 285-295.

gade (Anna M.) 2004 « Recitation of the Qurʾān », dans Jane Dammen  McAuliffe (dir.), Encyclopaedia of the Qurʾān IV, Leyde : Brill, p. 367-392. goldzIher (Ignaz) 1967-1971 Muslim Studies, édité par S. M. Stern et traduit par C. R. Barber, Samuel Miklos Stern, Londres : Allen & Unwin, 2 vol. gruBer (Christiane) 2009 « A Pious cure-all: the Ottoman illustrated prayer manual in the Lilly Library », dans The Islamic Manuscript Tradition: Ten Centuries of Book Arts in Indiana University Collections, Bloomington: Indiana University Press, p. 117-153. 2011  « The ‘Restored’ Shīʿī muṣḥaf as divine guide? The practice of fāl-i Qurʾān in the Ṣafavid period », dans Journal of Qurʾānic Studies 13, no 2, p. 29-55. hilali (Asma) 2011 « Compiler, exclure, cacher. Les traditions dites « forgées » dans l’Islam sunnite (vie/xiie siècle) », dans Revue de l’histoire des religions 2, p. 163-174. JeFFrey (Arthur) 1937 Materials for the History of the Text of the Qurʾan: The Old Codices, Leyde : E. J. Brill. kaFi (Mansour) 2008 ʿIlm al-qirāʾāt. Mafhūmuhu, nashʾatuhu, maṣdaruhu, aqsāmuhu wa madārisuhu, Annaba : Dār al-ʿulūm. keeler (Annabel) 1999 « Exegesis in Persian », dans Encyclopaedia Iranica IX, no 2, p. 119-123 et en ligne sur Iranica online, 2014, http://www.iranicaonline.org/ articlexegesises/-iii. 2006 Sufi Hermeneutics, Londres : Oxford University Press. leemhuis (Frederick) 2004 « Readings of the Qurʾān », dans Jane Dammen  McAuliffe (dir.), Encyclopaedia of the Qurʾān IV, Leyde/Boston : Brill, p. 358-359.

110 • SABRINA ALILOUCHE / GHAZALEH ESMAILPOUR QOUCHANI

melchert (Christopher), aFsaruddin (Asma) 2004 « Reciters of the Qurʾān », dans Jane Dammen  McAuliffe (dir.), Encyclopaedia of the Qurʾān IV, Leyde/Boston : Brill, p. 386-392. nawya (Paul) 1970 Exégèse coranique et langage mystique, Beyrouth : Dar el-Mashreq. nelson (Kristina) 1985 The Art of Reciting the Qurʾan (Modern Middle East Serie, 11), Texas : University of Texas Press. osman (Amr) 2012 « Human intervention in divine speech: waqf rules and the redaction of the qurʾanic text », dans Journal of Qurʾānic Studies 14, no 2, p. 90-109. Paret (Rudi) 2005 Der Koran, Kommentar und Konkordanz von Rudi Paret, Stuttgart : Verlag W. Kohlhammer. Pretzl (von Otto) (dir.) 1930 « Das Lehrbuch der sieben Koranlesungen von Abū ʿAmr ʿUṯmān ibn Saʿīd ad-Danī », dans Bibliotheca Islamica II, Leipzig : Deutsche Morgenländische Gesellschaft. rIzvI (Sayyid Athar A.) 1978 A History of Sufism in India, vol. I: Early Sufism and its History in India to 1600 A.D., New Delhi : Munshiram Manoharlal Pub. Al-ṣabbāġ (ʿAbd Allāh Tawfīq) 1993 Fann al-tartīl fī aḥkām al-tajwīd, Dubaï :  Dār  al-qalam. schimmel (Annemarie) 1970 Islamic Calligraphy, Leyde : E. J. Brill. 1996  « Le soufisme en Inde et au Pakistan », dans Le Soufisme, ou les dimensions mystiques de l’Islam, Paris : Éditions du Cerf. shah (Mustafa) 2005 « The quest for the origins of the qurrā’ʾ in the classical islamic tradition », dans Journal of Qurʾānic Studies 7, no 2, p. 1-35. tottoli (Roberto) 2001 « Bowing and prostration », dans Jane Dammen McAuliffe (dir.), dans Encyclopaedia of the Qurʾān I, Leyde/Boston/Cologne : Brill, p. 254-255.

zadeh (Travis) 2012 The Vernacular Qurʾan. Translation and the Rise of Persian Exegesis, Londres : Oxford University Press in association with the Institute of Ismaili Studies.

articles de l’Encyclopédie de l’Islam EI2 wEI2

Encyclopédie de l’Islam, B. Lewis et al. (dir.), Leyde/Paris : E. J. Brill/G.-P. Maisonneuve & Larose S. A., 1977-2007. EI2 en ligne. Les articles ci-dessous ont été consultés sur : Brill Online (http://referencesworks.brillonline.com) [accès payant ou par bibliothèque universitaire]

Fahd (Toufic) 1990 « Ḥurūf », EI2 III, p. 616-617. « K̲h̲awāṣ al-Ḳurʾān », dans wEI2 [consulté en juillet 2014]. hairi (Abdul-Hadi) 1986 « Madjlisī », dans EI2 V, p. 1084-1085. massÉ (Henri) 1977 « Fāl-Nāma », dans EI2 II, p. 779-780. monnot (Guy) 1995 « Ṣalāt », dans EI2 VIII, p. 956-965. neuwirth (Angelika) 1998 « Al-Shāṭibī », dans EI2 IX, p. 376-378. Paret (Rudi) 1986 « Kirāʾa », dans EI2 V, p. 129-132. Pellat (Charles) 1991 « Manāḳib », dans EI2 VI, p. 333-341. roBson (James) « Ḥadīth », dans wEI2 [consulté en juillet 2014]. sellheim (Rudolf) 1977 « Faḍīla », dans EI2 II, p. 747-748. welch (Alford J.), Pearson (James D.) 1986 « Al-Kurʾān », dans EI2 V, p. 401-435.

site internet Al-Ḥusaynī al-Tifarishī, Naqd al-rijāl, vol. V. Āl al-bayt liʾiḥyā al-turāth, en ligne, 2014 http://rafed.net/ booklib/ view.php?type=c_fbook&b_id=136.

II

CONTEXTES CONTEXTS

CONTEXTUALIZING THE GWALIOR QURʾAN: Notes on Muslim Military, Commercial and Mystical Routes in Gwalior and India before the 16th century Johanna Blayac

(Centre d’études de l’Inde et de l’Asie du Sud - UMR 8564)

Abstract According to its colophon, the Gwalior Qurʾan was completed in the fort of the town by one of its inhabitants on 7 Dhū al-qaʿda 801 or 11 July 1399. On the other hand, the Persian chronicles state that the Hindu kings of the Tonwar dynasty ruled Gwalior between 800/1398 and 924/1518. Some Persian hagiographical texts indicate that the Chishti Sufi shaykh Gīsū  Darāz and a group of persons who fled from Tīmūr’s assault on Delhi on 7 Rabīʿ II 801/ 17 December 1398 to reach Gwalior were hosted there by his disciple ʿAlāʾ al-Dīn Gwāliyārī  before they again went away to the regions of Malwa, Gujarat and Deccan. While we know with certainty where and when the manuscript of this Qurʾan was completed, the political, social  and  cultural  contexts  in  Gwalior  at  this  time  remain  obscure,  and  it  is  difficult  to  determine who was or were the patron(s) of the Gwalior Qurʾan. The paper proposes to explore and question the different milieus and routes, whether military, mystical but also commercial, that could have played a part in the conception and production of the Gwalior Qurʾan. In all, the author argues this Qurʾan does appear to be a unique testimony to the composite medieval Indian context.

Résumé Le coran de Gwalior : au carrefour des voies militaires, commerciales et mystiques des musulmans à Gwalior et en Inde avant le xvie siècle Le colophon du coran de Gwalior indique que le manuscrit fut achevé dans la forteresse de la ville par l’un de ses habitants le 7 ḏū al-qaʿda 801 de l’hégire, soit le 11 juillet 1399 de l’ère chrétienne. Selon les chroniques persanes néanmoins, les rois hindous de la dynastie Tonwar auraient régné sur Gwalior de 800/1398 à 924/1518 et des textes hagiographiques persans rapportent que le shaykh soufi Chishti Gīsū Darāz se serait enfui à Gwalior avec plusieurs  personnes devant l’offensive de Tīmūr contre Delhi, le 7 rabīʿ II 801/17 décembre 1398. Leur séjour à Gwalior aurait été organisé par le disciple du shaykh, ʿAlāʾ al-Dīn Gwāliyārī, jusqu’à  leur départ vers les régions du Malwa, du Gujarat et du Deccan. Ainsi, si nous connaissons avec certitude l’endroit et la date à laquelle le manuscrit fut achevé, il demeure difficile de  préciser le contexte politique, social et culturel à Gwalior durant cette période, et d’identifier le,  ou les, mécène(s) du coran. Cet article propose d’explorer les voies militaires, commerciales et mystiques et les différents milieux susceptibles d’avoir joué un rôle dans la conception et la production du coran de Gwalior qui constitue, selon l’auteur, un témoignage exceptionnel du contexte composite de l’Inde médiévale. Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 113-125

114 • JOHANNA BLAYAC

Contextualizing the production of the Gwalior Qurʾan is not an easy task. While we know with certainty through the colophon that the manuscript was completed (tamām shud) in the fort of Gwalior (qalʿa-yi Kālyūr) by one of its inhabitants named simply as Maḥmūd Shaʿbān on Monday 7 Dhū al-Qaʿda 801 or 11 July 1399 (figure 1), the political, social and cultural contexts in Gwalior at this time (and in fact for the whole period of the Indian sultanates, from the beginning of the 13th to the middle of the 16th century) remain obscure, to say the least. This obscurity can be explained by the nature of the historical information available regarding this period. The Persian chronicles dedicated to the sultans, on the one hand, are not very much concerned with the description of the regions, towns and villages of the sultanate(s), unless these regions, towns and villages are visited for conquest or financial/fiscal matters by  the sultan(s).1 Beyond the report of historical data, they mainly intend to legitimize both the sultan of Islam and the author of the text. The Persian hagiographical texts, on the other hand, have not been studied much by historians, although they allow us to penetrate into a different milieu than the sultans’ court.2

1.

2.

The history of the Delhi sultanate is known through four contemporary Persian chronicles which supplement each other: the Ṭabaqāt-i-Nāṣirī of Minhāj-i Sirāj Jūzjānī  (d. 658-659/1260), the Tārīkh-i Fīrūz Shāhī of Ḍiyāʾ al-Dīn  Baranī  (d. 756/1357),  the  Tārīkh-i Fīrūz Shāhī of Shams al-Dīn  Sirāj  ʿAfīf  (who  wrote  at  the  beginning  of  the  15th century), the Tārīkh-i Mubārak Shāhī  of  Yaḥyā  b. Aḥmad Sīrhindī (who wrote during the reign of Sayyid  Mubārak  Shāh  (824-838/1421-1435).  The  numerous  writings of the poet Amīr Khusrau (d. 725/1325) are also  a significant source of information. Another important  chronicle, the Futūḥ al-Salāṭīn, was written in Persian verses  (on  the  pattern  of  Firdawsī’s  Shāhnāma) in 750751/1349-1350 by ʿIṣāmī, who was the court historian of  the first independent sultanate in the Deccan region, i.e.  the Bahmani sultanate (748-933/1347-1527). See the bibliography for the precise references of the editions and/or translations of these works. See also Hardy (1960), as well as Kumar (2007a), 20-45. Later Persian chronicles are also valuable and often quoted such as the Tārīkh-i Badāʾūnī  from  Badāʾūnī  (d. 1024/1615)  and  the  Tārīkh-i Firishta from Firishta (d. 1033/1624), or even the Ṭabaqāt-i Akbarī of Niẓām al-Dīn Aḥmad (d. 1030/1621). Other texts  including juridical treaties and “miroirs des princes” are not directly relevant here. The malfūẓāt (conversations of Sufi saints) and tadhkirāt (biographical dictionaries) are the main genres of the hagiographical literature. In addition, poetry and tales can also be considered. See Lawrence (1978). The most famous malfūẓāt of the Delhi sultanate are the Fawāʾid al-Fuʾād recorded by Amīr Ḥasan Sijzī, which gather the  conversations  of  his  Sufi  master  Niẓām  al-Dīn  Awliyāʾ Dihlawī  (d. 725/1325).  See  Niẓām al-Dīn  Awliyāʾ [1990] and [1991] for its English translation. Simon Digby, who passed away in January 2010, was one of the best connoisseurs of Indian Sufi and Sufi popular literature;  the present paper quotes several of his works.

Nonetheless, they contain many symbolic and allegorical features that should be considered very cautiously. Important historical information can also be found in Arabic geographical and travel literature developed in the central Islamic lands from the 9th century onwards and notably represented by the Moroccan traveler  Ibn  Baṭṭūṭa  during  the  14th century.3 This literature is particularly significant since it also highlights circulations and encounters that are not reported by Indo-Islamic chroniclers and writers and as such extend de facto the scope of what we must consider the medieval Indian subcontinent and the Indian sultanates. Next to strictly textual or literary sources, material sources (sometimes both objects and texts) such as coins and inscriptions as well as monuments and artifacts also constitute and provide fundamental data likely to complete, clarify and sometimes qualify literary information and representations. Several recent studies have shown the importance of these sources in bringing to light social and cultural encounters and collaborations that led to composite and hybrid productions.4 In this regard, the Gwalior Qurʾan, which displays so many stylistic influences  and could have been used both as an artistic and religious teaching book, is in itself an invaluable resource for historical knowledge. The present paper proposes to explore and question the different milieus and routes, whether military, commercial or mystical, that could have played a part in the conception and production of the Gwalior Qurʾan. It considers and uses both primary and secondary sources and does not intend to give any

3.

4.

Most of these sources, including also some historical texts such as annals and futūḥāt (literature of conquest), date back to the first centuries of Hijra prior to the 13th century. They document both the expansion of the caliphate (established in the region of Sind in modernday South-Eastern Pakistan in 92/711) and the networks and settlements of Muslim traders in various regions of the Indian Ocean. Among the geographers and travelers, we can cite for instance al-Masʿūdī,  Ibn  Ḥawqal  and  al-Muqaddasī,  all  writing  during  the  10th century. The anonymous Ḥudūd al-ʿĀlam are an example of geography written in Persian at the end of the 10th century. See, for these sources, Blayac (2005). Ibn Baṭṭūṭa, who is famous  for his Riḥla (or travels), was also a judge (qāḍī) in Delhi during  the  reign  of  the  Sultan  Muḥammad  b. Tughluq  (725-752/1325-1351) and included in his records a history of  the  Delhi  sultanate  and  of  the  reign  of  Muḥammad  b. Tughluq. Material culture studies have recently experienced an important development regarding the Indian subcontinent and the Indian sultanates, in particular the work of historians such as Sunil Kumar and the present author, as well as numerous art historians including Anthony Welch, Mehrdad and Nathalie Shokoohy, Finbarr Barry Flood, Elizabeth Lambourn, Alka Patel, Éloïse Brac de la Perrière.

CONTEXTUALIZING THE GWALIOR QURʾAN • 115

Figure 1 – Colophon of the Gwalior Qurʾan, fol. 554v. [© By courtesy of the Aga Khan Museum]

116 • JOHANNA BLAYAC

exhaustive and definitive answer but to hint at important issues and sketch the complex and moving background of what we can call Medieval India.5

GwaLior and the deLhi suLtans An obvious example of the bias of the Persian chronicles mentioned above can be given regarding Gwalior itself. Minhāj-i Sirāj Jūzjānī, the author the  main historical text for the first century of the Delhi  sultanate (the Ṭabaqāt-i Nāṣirī) was himself appointed as judge (qāḍī), preacher (khāṭib) and imām of Gwalior after its conquest by Sultan Iltutmish (r. 607-633/ 1210-1216) in 629-630/1232-1233. Nevertheless, he does not give any account of the history of Gwalior nor regarding his own life and his daily functions in the town (Gwāliyūr in his chronicle).6 Beyond the traditional bias of the chronicles however, it is also likely that Jūzjānī did not stay long  in Gwalior since the town was [probably] conquered again by the Tonwar (or Tomara) Rajput sovereigns as soon as 634/1236, during the reign of Iltutmish’s daughter Raḍiyya (r. 634-637/1236-1240).7 As suggested by Khaliq Ahmad Nizami in his article ‘Gwāliyār’ in  the second edition of the Encyclopaedia of Islam, the history of Muslim settlement in Gwalior truly resembles a chessboard. According to his review, Gwalior was raided eight times by Muslim troops between

5.

The periodization of Indian history covers several historiographical problems that are also linked with the nature of the sources and the use historians made of them until quite recently. On the one hand the coming of Muslims to India was long considered as marking the beginning of the ‘Middle Ages’ (with the connotation of ‘Dark Ages’);  on the other hand the emergence of the Delhi sultanate brought with it a new type of source, the historical chronicle (or tārīkh), which also diverted historians from other sources such as other literary (but ‘non-historical’) sources or archaeology. In this perspective, the ‘Muslim field’ and the ‘Indian/Hindu field’ were often considered  separately. This bias has unfortunately been artificially  maintained in academic research since few scholars are indeed able to combine the knowledge of the many languages that one should be able to read in order to be exhaustive. See Blayac (2012), 17-36. 6.  See Jūzjānī [1953-1964], vol. 1, 448-449. The Ṭabaqāt-i Akbarī of  Niẓām  al-Dīn  Aḥmad  records  the  text  of  a  commemorative inscription in Persian verses, celebrating the conquest of Gwalior by Iltutmish in 630/1232-1233. In the time of Niẓām al-Dīn, this inscription was situated  on the Urwahi Gate of the citadel but it is now lost and was not noted in the contemporary accounts of Iltutmish. Moreover, the use of Persian language in epigraphy is not attested under the Delhi sultans before the course of the 14th century. See Blayac (2012), note fiche 165. 7.  Jūzjānī  was  afterwards  appointed  in  Delhi  and  Raḍiyya  made him the head of the madrasa-i Nāṣirī as well as of the judges (quḍā) of Gwalior. See Jūzjānī [1953-1964], vol. 1,  448-449, 460, 466, vol. 2, 214, a career also discussed by Kumar (2007a), 224-225.

413/1022 and 924/1518, and the town was governed by the Tonwars between the end of 800/1398 (that is to say at the moment of Tīmūr or Tamerlan’s assaults  upon Northern India and Delhi) and 924/1518, when the sultan of Delhi, Ibrahim Lodi, conquered it.8 The Moroccan traveler and ʿālim Ibn Baṭṭūṭa visited  Gwalior in the 740s/1340s. He mentions in his Riḥla the name of an amīr, Aḥmad b. Sīrkhān, apparently a  Delhi officer of Afghan origin, and apart from the topological situation of the fort, evokes an important system of defense as well as the buildings and population of the town at the foot of the fort. Regarding the population he states that most of the tradesmen of the town were non-Muslims and that six hundred horsemen were stationed there and conducted an ongoing war because the fort was surrounded by nonMuslims, – very likely soldiers and kings, different from the tradesman of the town, although Ibn Baṭṭūṭa  does not specify the difference.9 This situation of quasi-permanent siege is most probably not particular to Gwalior.10 It globally reflects,  on the one hand, the mode of conquest of new places from the Ghurids’ settlement in Delhi and Hindustan from the end of the 12th century (after the victorious battle of Tarain in 588/1192), and on the other hand, the de facto discontinuity of the territory and authority of the sultanate. Military expeditions were indeed at first mainly intended to amass spoils and to gain  formal recognition for Islamic authority, symbolized by a tribute from other sovereigns without changing the local political structures, but as such, they had to be renewed and often led to the set-up of a new provincial Muslim government, which could in turn be  disputed  by  previous  sovereigns  and  officers  as  well as other Muslim officers and authorities. Apart  from Delhi and the region directly related to it as well as the capitals of the other sultanates, the history of the Muslim conquest and political control of all of India is indeed a puzzle to historians.11 What is truly particular to Gwalior, however, is the Qurʾan presented

8. See Nizami in EI2, as well as the French version. 9.  See Ibn Baṭṭūṭa [1958-2000], vol. 3, 645 and vol. 4, 787. 10. For Gwalior, see ibid., vol. 4, 786. Among other examples given in the chronicles regarding for instance the regions of the Himalayas or Bengal, several epitaphs from Bilram (in the Uttar Pradesh region near Aligarh) show a recurrent conflict  between  Muslims  and  local  Hindu  chiefs  from  the 1250s to the beginning of the 14th century. See Blayac (2012), fiches 146-149B. 11. See Jackson (1999), Blayac (2012), 141-176, and, regarding the Ghurid conquest of Gwalior and other places in Northern India, Kumar (2007a), 118-120. The very territorial and jurisdictional articulation or superposition of Muslim sultanates and Hindu kingdoms — which have never disappeared — needs to be explored further. Excavations would also be of interest in this respect.

CONTEXTUALIZING THE GWALIOR QURʾAN • 117

in this volume, completed in the fort during the summer of 801/1399, that is to say, according to the chronology based on the chronicles, during the Tonwar domination of the town. However, other accounts question this chronology.

Simon Digby was one of the first medievalists of  India to examine the hagiographical Persian texts in order to explore the history of the Delhi sultanate beyond the bias of the official chronicles compiled in  Delhi. In an article published in 2004, he questioned the relations between the regional establishments within Delhi sultanate territories and the capital during the 14th century.12 As Simon Digby states in this article, the 14th century represents a shift in the history of the Delhi sultanate as well as in the more general history of Muslim settlement in India. The politics  of  conquest  led  by  Sultan  Muḥammad  b. Tughluq in the 730s/1330s were intended to unify the whole Indian subcontinent under his centralized authority. This created a de facto link between the territories that since the first centuries of Islam have  hosted Muslim populations, stemming from the settlement of Muslim traders and missionaries involved in the Indian Ocean networks, and the territories conquered by Muslim armies in Hindustan at the end of the 12th century. Nonetheless, from this same period of centralization, during the 730-740s/13301340s, some provincial Muslim centers of power emerged and, at the very end of the century, in Rabīʿ I/ December 1398, while the Delhi sultans’ territory was reduced to the close region of Delhi, Tīmūr’s armies  penetrated into India through the northwestern regions.  On  7  Rabīʿ II 801/17 December 1398, after a short battle against the troops from Delhi, Tīmūr’s  troops entered the capital of the Indian sultanate and looted it. According to the texts examined by Digby, some inhabitants of the capital took flight as  soon as the battle began, anticipating the defeat of the Delhi army. Among these inhabitants were the Chishti  Sufi  shaykh  Sayyid  Muḥammad  Gīsū  Darāz,  who was then seventy-seven years old, accompanied by a group of seventy persons. They left Delhi to reach Gwalior (we may note here the symbolic and conventional repetition and declensions of the figure 7, which  is important in the esoteric areas of Islam, including Sufism, as well as Shiʿism and Ismaïlism). In Gwalior, a  disciple  of  Gīsū  Darāz,  ʿAlāʾ  al-Dīn  Gwāliyārī,  had 

been contacted and had to organize lodgings for them before they again took to the road. The saint would then follow the trade routes and get to Chanderi in Malwa, then to the region of Gujarat (in Baroda and then Cambay), where an independent sultanate was about to be born through the action of the governor Ẓafar  Khan,  and  then  to  the  region  of  the  Deccan  ruled by the Bahmani sultans since 748/1347, in Daulatabad (where Gīsū Darāz had grown up), and in  the new capital Gulbarga, where the shaykh  finally  settled and established his spiritual leadership.13 Gīsū Darāz and his group are said to have arrived in  Baroda on 1 Shawwal 801/6 June 1399 after a journey of three months and ten days from Gwalior. The group thus left Gwalior during the month of Jumādā II/at  the end of the month of February 1399, while they had taken flight from Delhi on (or around) 7 Rabīʿ II or 17 of December. According to the common historical sources, we know that the roads were well developed and organized, tree-lined and with regular stages during the 14th century. Ibn Baṭṭūṭa reports that one  needed forty days to reach Daulatabad, far South of Gwalior, from Delhi. We can deduce that Gīsū Darāz  and his group were able to reach Gwalior, not forgetting that they were also attacked along the way and rescued and escorted by Muslim troops from Gwalior, at the beginning of the year 801/1399. Simon Digby mentions the Qurʾan produced during that year, 801/1399, in Gwalior in a note to his article. He links its production with the arrival of notables from Delhi with Gīsū Darāz. To him, this magnificent  Qurʾan indeed proves that distinctive fugitives from Delhi did accompany the shaykh. Following this perspective, we might suppose that the Qurʾan was only completed in Gwalior, which corresponds with the term tamām shud in the colophon. However, it was completed after the shaykh and his relatives had left Gwalior,  during  the  month  of  Dhū  al-Qaʿda/July, and was thus most probably directly intended for Gwaliori people. Since the name of the last (?) copyist, Maḥmūd Shaʿbān, has not been identified as far as  we know, we cannot know if this person indeed arrived with Gīsū Darāz. Nonetheless, the hagiographical texts  explored by Simon Digby do indicate that ʿAlāʾ al-Dīn  Gwāliyārī, the Chishti disciple of shaykh Gīsū Darāz  in Gwalior, received his khilāfat (or succession) as thanks for his welcome in Gwalior.14 It is thus possible to imagine that the Qurʾan was produced or finished  in Gwalior for the Chishti disciples. Neither the historical and hagiographical sources, nor the Qurʾan which does not mention the name of any patron, provide any

12. Digby (2004), 298-356.

13. About this shaykh see for instance Hussaini (1986) and Askari (1952). 14. See Digby (2004), apparently principally based on Yazdī  [1379/2001].

the chishti sufi order in and across GwaLior

118 • JOHANNA BLAYAC

certainty on this point. However, the facts recorded by the hagiographical texts about Gīsū Darāz’s stay  in Gwalior with the support of both ʿAlāʾ  al-Dīn  Gwāliyārī  and  the  Muslim  army  of  the  town  could  question the so-called occupation of Gwalior by the Tonwar rajas by the beginning of 801/the end of 1398 – beginning of 1399. Let us admit this occupation, even though events are still to be more precisely dated. Could it be possible that the Qurʾan was completed during the Tonwars’ rule? Why should we exclude the possibility of a continuous coexistence of populations, both Muslim and non-Muslim (that is to say Hindus, Jains and/or Buddhists), beyond the ruling authority? This religious diversity and social fluidity was indeed  at work in several Indian regions, long before the conquest of Hindustan by the Ghurids at the end of the 12th century.

GwaLior and the ancient trade routes At the time of the first Arabo-Islamic conquests  in the 7th and 8th centuries, the various kingdoms of India (al-Hind) were both commercial partnerships (mainly in the southwestern regions) and political and territorial opponents (in the northwestern and northern regions). Furthermore, the Arab conquest of al-Sind (Sind, today southeast Pakistan) in 92/711 primarily aimed at securing maritime and land routes that linked the Middle East to India (at the time including Southeast Asia), Central Asia and China. Moreover, we know from the historical texts that describe the conquest of Sind that several delegations and expeditions were sent from al-Sind to al-Hind.15 One of these delegations was sent with a decree from the Umayyad caliph inviting people to adopt Islam, honor him and pay tribute to the treasury in 95/714. It moved forward in the direction of Kannauj, the former capital of the emperor Harsha (r. 606-647), but stopped in Udhepur/Udaipur. Another expedition was sent later on in the years 100-110/720-730 to the region of Malwa (Mālibah in al-Balādurī’s book), specifically  to the towns of Ujjain and Gwalior.16 Now, the geographer al-Muqaddasī in the 10th century, as well as a decorative piece of carved stone identified as a miḥrāb

15. These delegations and expeditions are reported on the one hand by the anonymous Shāhnāma (also known as the Fatḥnāma), a thirteenth-century Persian translation of an Arabic text said to date from the 9th century that narrates the history of ancient Sind and the Arab conquest, and on the other hand by the Futūḥ al-Buldān of the historian  al-Balāḏurī  (d. 302/892).  About  the  Shāhnāma see Friedmann (1984), 23-37. 16.  See, for the first expedition, anon. [1990], 192-193, and  for the second expedition, al-Balāḏurī [1969], 227. For all  this period and events see Blayac (2012), 42-68.

(or prayer niche), also respectively attest to the presence of Muslims, likely traders and probably religious scholars, in Kannauj and Gwalior.17 The miḥrāb (figure 2) has been studied by Michael  Willis from the British Museum at the beginning of the 1980s.18 It is standing today against a wall of the fort constructed by the Tonwars at the end of the 15th century,  near  the  elephant  gate  or  Hāthī  Por.  The shape of the arch of the niche is directly inspired by the pediment (or udgama) in Hindu religious architecture, the evolution of which allowed Michael Willis to date the carving of the stone not before the second half of the 8th century. Beyond the dating, which could be refined according to the development  of the use of the miḥrāb and movable miḥrāb in the Muslim world, the analysis of Willis makes it clear that the stone, because of its carving19 and because of its size (2.95 × 1.84 m),20 cannot be considered a reuse, but is indeed a miḥrāb built for a Muslim community by local craftsmen. As such, it illustrates the cultural and material adaptations, or translations to quote the art historian Finbarr Barry Flood, carried out in Islamic art in its regional declinations (figure 3).

cuLturaL hybridization and sociaL fLuidity in medievaL india Examples of cultural hybridization and social fluidity are not rare in medieval India, primarily in coastal kingdoms such as those of the Cheras in Malabar (Kerala) and the Rashtrakutas on the Konkan coasts of the 9th and 10th centuries, then in Gujarat and Maʿbar (Southern Tamil Nadu).21 The region of Gujarat (where Gīsū Darāz spent some  time after he went through Chanderi from Gwalior) is in this regard particularly well documented by epigraphic and architectural remains. The term ‘mosque’ is also transcribed in some medieval Jain architectural treaties by the term rahmanprasada, or literally

17.  Al-Muqaddasī reports the existence of a congregational  mosque in the suburbs of Kannauj. See Flood (2009), 23, as well as al-Muqaddasī (1967), 480. 18. Willis (1985): 227-246. 19. The carving shows no rearrangement (such as the removing of a divinity at the center of the niche) and is very different from the decorative motifs one can find  in Hindu temples. See ibid., 229-230. 20. Again according to Willis, this is uncommon in the elements of temple architecture. 21. See Blayac (2012), 114-140. Later on, starting from the end of the 13th century, examples are also known under the Delhi and regional sultans, who promoted Indo-Persian culture. About the Persian culture and its patronage see Delvoye (dir.) (1994), as well as Alam, Delvoye, Gaborieau (dir.) (2000), and Delvoye, Lefèvre, Sharma (dir.), forthcoming. And for instance Katre (1967), 357-367, as well as Welch, (1993), 311-322.

CONTEXTUALIZING THE GWALIOR QURʾAN • 119

Figure 2 – The Gwalior miḥrāb. [© By courtesy of the British Museum, Photograph: M. Willis, 2007]

Figure 3 – Niche from a ruined temple showing an example of what inspired the creation of the miḥrāb. This kind of material is used in temples architecture for instance in Gopaksetra (600-900). [© By courtesy of the British Museum, Photograph: M. Willis, Scindia School, Gwalior, 1980]

temple of the almighty.22 The study of the regional Islamic inscriptions dated from the 12th century to the beginning of the 14th century (i.e. the moment when the Delhi sultans undertook the conquest of Gujarat) allows us to note three interesting developments.23 1/  These  inscriptions  reflect  the  social  prestige  of  Muslim traders, most often coming from the Persian world (but also from more Western regions),24 sometimes Shiʿites and some also who had been brought over to the Fatimid Ismaili cause. 2/ They bear witness to the activities of religious men such as one mystic named Bābā Arjūn Shāh Damawī al-Akhsī, who gathered people of different faiths around him, and whose double nisba reveals links with Central India (Damoh, about 300 km South of Gwalior) and Central Asia (Akhs, for Akhsikat in Ferghana), one teacher (muʿallim) of Indian extraction (according to his name Golā b. Ḥasan 

Kambāyatī, i.e. from Cambay), one Sufi poet coming  from Yazd in central Iran, or even a jurist in the region of Cambay named “mufti of sects and groups”.25 3/ Finally, epigraphic data also show the difficult implementation of the Delhi sultans in the region, this due to the presence of other Islamic authorities,26 as well as the Hindu sovereign authorities that also engaged  Muslim  officers.  A  bilingual  inscription  in  Persian and Sanskrit carved in 704/1304 in the village of Sampla (in South Gujarat) reports in the name of

22. See Patel (2004). 23. See Blayac (2012), 121-134. 24. The Broach coin hoard for instance shows coins coming from Mamluk Egypt as well as from Italian Republics. See Digby (1980), 129-138.

25. All these persons are known through their epitaphs. Bābā Arjūn Shāh died in Petlad near the port of Cambay  in  633/1236;  Golā  b. Ḥasan  Kambāyatī  died  in  Rander  near the port of Surat (which became especially important in the 16th century)  during  the  same  year;  the  Sufi  poet, ʿAlī  b. Sālār  b. ʿAlī  al-Yazdī,  whose  Persian  verses  are also inscribed, died in Cambay in 685/1287; the muftī Muḥammad  b. ʿAlī  b. Yaḥyā  b. al-Jawzī  al-Jazrī  died  in  707/1307 in the same town. See Blayac (2012), fiches 53,  55, 37 and 43, and vol. 1, 125-129. 26. Apart from the religious and juridical authorities mentioned above, the sultan of Hormuz was also represented in the port of Somnath through the intermediary of ship captains (nākhudā). See Blayac (2012), 129-130 and fiche 57.

120 • JOHANNA BLAYAC

King Rāi Karn Dīv a donation to the great mosque of  Cambay  by  a  former  Delhi  officer  who  became  the  price minister of the said Hindu king.27 Although there is no concrete trace apart from the non-inscribed miḥrāb in Gwalior, it is still possible to transpose this situation: Muslims already settled in Gwalior do not appear to have contributed to the sustainable implementation of the Delhi sultans and on the contrary, by ‘tradition’, seem to have lived and worked with the Hindu authorities. In Gujarat, the authority of the sultanate was in fact only truly effective when a regional dynasty emerged in the early 15th century. In addition, it is likely that the diverse origins (even the various allegiances within Islam) of Muslims who settled before (and even after) the conquest of the Delhi sultans also made the situation difficult since one’s origins were more  important than one’s religion at the time. According to the texts consulted by Digby, the garrison of Delhi in Gwalior mainly consisted of Afghans at the end of the 14th century (and we can thus note a continuity with the account of Ibn Baṭṭūṭa). These are probably  the repeated difficulties encountered by the sultans  of Delhi as they attempted to assert their authority in Gwalior, difficulties that led them to appoint a Chishti disciple in the town. The  political  instrumentalization  of  Sufi  orders  was indeed enacted in the course of the 14th century by  Sultan  Muḥammad  b. Tughluq  (r. 725-752/13251351), who met regularly with the great and powerful Chishti saint Niẓām al-Dīn Awliyāʾ of Delhi (d. 725/ 1325) – that is to say the master of Gīsū Darāz, who  was himself the master of ʿAlāʾ al-Dīn Gwāliyarī –  before he became sultan.28 In this way, if the political authority of Delhi was not firmly recognized, the spiritual authority of Niẓām al-Dīn made it possible  to maintain a link with the capital and in addition to transcend the religious and dogmatic differences while formally remaining in the field of Sunni Islam,  promoted by the Delhi Muslim rulers. Nonetheless, the Deccani Muslim dynasties would soon adopt Shiʿism. These are points that seem to be important if we are to try to understand the possible context of production of the Gwalior Qurʾan, which is again itself a major and exceptional testimony to medieval 27. Ibid., 130-132 and fiche 66. One can also find several examples of rebel officers fleeing from their posts under  the Delhi sultans, notably under Muḥammad b. Tughluq  to seek refuge in some Hindu sovereign court or kingdom. See for instance Blayac (2004). 28.  Great Sufi shaykhs such as Niẓām al-Dīn Awliyāʾ were both allies and competitors of the sultans since they themselves represented the right path to God via the imitation of the prophet Muḥammad. See for instance Digby (1990),  71-81, as well as Kumar (2000), 37-65. See also for instance Eaton (1978), and the same (1996).

Indo-Islamic diversity and complexity. We might also ask if there may have been an attempt to found a new independent Muslim dynasty in Gwalior in 801/1399, but it seems unlikely since no patron is mentioned in the manuscript. Nevertheless, the two seals on the manuscript suggest that it had different owners, from different places, during the following centuries.

mysticaL Paths in medievaL southern india In this regard and to pursue our exploration of the Indo-Islamic context, it might also be interesting to consider religious circulations that were probably not attached to any Islamic political capital, nor, at first,  to any Sufi order or brotherhood (ṭarīqa), although oral traditions usually came to claim this kind of belonging. One example can be given through the story attached to the saint Shāh Fakhr al-Dīn Suhrawardī.  This saint is said to have been a renouncing king from Sistan in Iran who came to preach Islam in Penukonda (today in the southwest of the state of Andhra Pradesh, geographically on the Deccan plateau) in the 12th century. According to one tradition he was sent to this town by the saint Nathar Shāh Walī, who would himself have settled in Trichy (in Tamil Nadu) during the 10th or 11th century. Another saint, Shāh Hyder  Suhrawardī, said to have settled in Mulbagal (in South  East Karnataka) in the 12th century, is also claimed to have been a disciple of Nathar Shāh Walī. He seems to  be more surely than Shāh Fakhr al-Dīn affiliated with  the Suhrawardiyya29 the other important and influential Sufi order under the Delhi sultans. This order  collaborated closely with the political authorities and would also find refuge in Gujarat and Southern Indian  regions after Tīmūr’s assault in 801/1398-1399. In  addition, it is also linked with the Gwalior Qurʾan since one of the seals impressed on the manuscript (figure 1) dated 1037/1627-1628 can be read in Persian  as “I am a Suhrawardī hermit honored by [my] Master,  the beloved of The Lord, The Most Compassionate, The Most Merciful”. The other seal of the manuscript, which can be dated from the 19th century and shows a characteristic trefoil crown (figure 4), makes it possible to link the manuscript with the Deccan’s past, – more specifically with the Golconda court (934-1098/1528-1687) –, and with Iran.30

29.  It would seem that the heirs of Shāh Fakhr al-Dīn only  were affiliated with the Suhrawardī order. 30. These two seals have been examined and read by John Seyller  and  Manijeh  Bayani-Wolpert;  the  information  quoted here was reported to me by Éloïse Brac de la Perrière, whom I wish to thank.

CONTEXTUALIZING THE GWALIOR QURʾAN • 121

Figure 4 – Gwalior Qurʾan, trefoil crown seal on fol. 1r. [© By courtesy of the Aga Khan Museum]

language record the construction of the mosque by Fakhr al-Dīn in 693/1293-1294, and the death of the  saint in 694/1294-1295.34 In addition to Shāh Fakhr  al-Dīn’s tradition, some interesting material elements  are also to be noticed in a temple dedicated to Shiva dating from the 15th century within the ancient wall of the fortress. Two pillars within this temple present small stylized sculptures of two persons, one of them a man with a small beard and a pointed cap, which seems to be the way Turkish Muslim characters were represented, while the second looks rather feminine and is entirely covered with a mantle that lets only the face and one hand appear. This figure is making  the gesture of ajna (ajna mudra) or knowledge in Sanskrit (figures 5-6). These little characters carved  on a pillar are widely present on structures in HampiVijayanagar (the former capital of the Vijayanagar kings) from the 14th and 15th centuries, when the Vijayanagar kingdom flourished (figures 7-8). They can be  identified with performing artists, dancers, musicians,  and other visitors to the court, including wandering persons, which is to say, in a Sufi context, qalandars. The epithet of qalandar is used today to name both Nathar Shāh Walī and Shāh Fakhr al-Dīn. According to  tradition, it indicates that they both were not married, but it could also show the saint’s affiliation with the  Sufi order of the Qalandariyya, whose practices could  have been influenced by the earlier Malāmatiyya from  Khurasan, the people of self-blame. The presence of wandering Sufis as well as qalandars, or antinomian Sufis who considered social provocation and transgression to be a privileged mode of access to God, is well attested during the Delhi sultanate.35 Furthermore, this influence can possibly be detected in the  Gwalior Qurʾan through the uncommon interpretations given by the fālnāma.36

Penukonda is a small but strategic medieval town at the foot of a strongly fortified hill: after the battle  of Talikota in 972/1565 (and continuing until 1083/ 1672) the fortress became the capital of the Vijayanagar kingdom, which had been fighting the Indian  sultans of Delhi and then of the Deccan since 736/ 1336.31 According to other sources, written but traditional, recalling its story, a Muslim ascetic saint settled in the town around the middle of the 14th century, before it was integrated into the Vijayanagar kingdom in 761/1360. This saint, who is unfortunately not named but may be Shāh Fakhr al-Dīn or one of his  successors (khalīfa), is said to have protected the town against the attacks of the Deccani sultans during the 16th century.32 Two antedated inscriptions dating back at least to the 15th and 16th centuries also report the tradition of the saint. They are situated in the town, in the dargāh (shrine) of Shāh Fakhr al-Dīn, itself situated on the  site of an ancient temple said to have been won by the Muslim saint through a battle of miracles.33 Once again dates are different, which shows that the tradition surrounding the figure of Shāh Fakhr al-Dīn was  reformulated over the centuries: these inscriptions in naskhī and nastaʿlīq scripts and in the Persian

It is still difficult to determine who was or were  the patrons of the Gwalior Qurʾan, but we can note that the dating of this colophon, Monday 17 Dhū  al-Qaʿda 801 or 21 July 1399, does question the known political chronology of the fort of Gwalior at the very end of the 14th century. Additionally, the

31.  The Hindu kingdom of Vijāyanagar was founded by two  brothers who fled from Delhi after they had been captured,  converted and appointed high officers under the Sultan  Muḥammad  b. Tughluq.  Nonetheless,  there  is  evidence  that this kingdom, while fighting the sultanates, also imitated them. See Wagoner (1996), 851-880. 32. See Vasantha (2000), 18-19. 33. According to the tradition recorded in the Imperial Gazetteer of India 20, 104-105, available online at the Digital South Asia Library of the University of Chicago.

34. See Blayac (2012), 231-234. 35. See Digby (1984), 60-108, and id. (1976), 171-178. Other famous qalandars established themselves in Indian regions starting in the 12th century, such as Laʾl Shahbaz Qalandar in Sehwan Sharif, in Sind, whose membership is also claimed by the Suhrawardiyya, as well as by Ismaïlis. Regarding the Indian qalandars and Laʾl Shahbaz in particular, see Boivin (2012). 36. See also Alilouche and Esmailpour Qouchani’s article in this volume.

concLudinG remarks

122 • JOHANNA BLAYAC

Figure 5 – Turkish Muslim character carved on a pillar of a Shiva temple near the Gagan Mahall in Penukonda (15th century?). [© Photograph: J. Blayac, 2007]

Figure 6 – Feminine character making the ajna mudra carved on a pillar of a Shiva temple near the Gagan Mahall in Penukonda (15th century?). [© Photograph: J. Blayac, 2007]

Figure 7 – Example of character carved on pillars in HampiVijayanagar (14th-15th century). [© Photograph: É. Brac de la Perrière, 2008]

Figure 8 – Example of character carved on pillars in HampiVijayanagar (14th-15th century). [© Photograph: É. Brac de la Perrière, 2008]

CONTEXTUALIZING THE GWALIOR QURʾAN • 123

manuscript itself, through its location and the many stylistic influences it shows,37 surely reflects the global context of India during the medieval period. Other texts in Indian languages, as well as undiscovered Persian or Arabic texts or documents, could perhaps shed new light on the history of Gwalior, but it is globally important to keep in mind that the Indian context of the time was in perpetual motion and shaped by various circulations, in turn and at the same time political and military, commercial, religious and spiritual. Moreover, these circulations are not impervious to each other and to the environment, and weave a context that is both complex and composite. Furthermore, this context is not limited to the Indian towns or regions but is linked with others within the Muslim world and beyond, which involves thinking on multiple scales. The circulation of the manuscript itself illustrates these circulations and the permeability and fluidity that characterize them.  While it clearly shows different stages of production and/or conservation and could have been used as a workbook for artists as well as believers, it was touched by many influences and probably passed from hand to hand and through various milieus, including, at least, the Sufi Suhrawardiyya at the beginning of the 17th century. All in all, the Gwalior Qurʾan does appear to be a unique testimony to the medieval Indian context.

bibLioGraPhy Sources anon. [1990] The Chachnameh, an Ancient History of Sind, translated by M. K. Fredunbeg, Karachi, 1900, reprint. [1993] Ḥudūd al-ʿĀlam. ‘The Regions of the World’. A Persian Geography, 372 A.H.-982 A.D., translated and explained by V. Minorsky, with the preface of V. V. Barthold (d. 1930), London: Luzac & Co, 1937; reprint, Oxford: Oxford University Press. ʿaFīF (Shams al-Dīn Sirāj) [1891] Tārīkh-i Fīrūz Shāhī, edited by M. V. Husain, Calcutta: Bibliotheca Indica. [1996] Tārīkh-i Fīrūz Shāhī, partial English translation, edited by Henry Elliot, John Dowson in The History of India as told by its own Historians. vol. III, London: Trübner & co, 1871; reprint, Delhi:  Low Price Publications. al‑balādurī (Aḥmad b. Yaḥyā b. Jābir) [1866] Kitāb futūḥ al-buldān / taʾlīf al-imām Abī al-ʿAbbās Aḥmad bin Yaḥyā bin Ǧābir al-Balāḏurī. Liber expugnationis regionum, ed. by M. J. De Goeje, Leiden: Brill. 37. See also Cruvelier and Chaigne’s article in this volume.

[1969] The origins of the Islamic state: being a translation from the Arabic accompanied with annotations, geographic and historic notes, of the Kitāb Futūḥ al-Buldān of al-Imām Abū-l ʿAbbās Aḥmad ibn-Jābir al-Balādhurī, vol. 2, translated by F. C. Murgotten, New York: Columbia, 1924;  reprint, New York: AMS Press. baranī (Ḍiyāʾ al-Dīn) [1862] Tarīkh-i Fīrūz Shāhī, edited by Sayyid Ahmad Khan, Calcutta: Bibliotheca Indica. [1996] Tarīkh-i Fīrūz Shāhī, partial English translation, edited by Henry Elliot, John Dowson in The History of India as told by its own Historians, vol. III,  London:  Trübner  &  co,  1871;  reprint,  Delhi: Low Price Publications. Ibn baṭṭūṭa [1958-2000] The Travels of Ibn Baṭṭūṭa, A.D. 1325-1354, translated by H. A. R. Gibb with revisions and notes from the Arabic text edited by C. Defremery, B. R. Sanguinetti, Cambridge: Published for the Hakluyt Society at the University Press, 5 vols. ʿIṣāMī (Khwāja ʿAbd al-Malik) [1940] Futūḥ al-Salāṭīn, edited by A. S. Usha, Madras: University of Madras. [1967] Futūḥu’s Salāṭīn of ʿIṣāmī, or Shāh-Nāmah-i Hind: translation and commentary, ed. by Agha Mahdi Husain, New York: Asia Publishing House. jūzjānī (Minhāj-i Sirāj) [1863] Ṭabakāt-i-Nāṣirī, edited by W. N. Lees, Calcutta: Bibliotheca Indica. [1881-1897] Ṭabakāt-i-Nāṣirī: a General History of the Muhammadan Dynasties of Asia, including Hindustan; from A.H. 194 (810 A.D.) to A.H. 658 (1260 A.D.) and the Irruption of the Infidel Mughals into Islam, translated by Henry George Raverty, Calcutta: Asiatic Society of Bengal. [1953-1964] Ṭabakāt-i-Nāṣirī, edited by Abdul Hay Habibi, Kaboul: Anjuman-i Tarikh-i Afghanistan. al‑MuQaddasī [1967] Kitāb Aḥsan al-Taqāsīm fī Maʿrīfat al-aqālīm, edited by M. J. de Goeje, Leiden: Brill. nIẓāM al‑dīn aWlIyāʾ (Muḥammad) [1990] Fawāʾid al-Fuʾād, edited by Khwaja Hasan Thani Nizami Dihlawi, Delhi: Urdu Academy, 1990, English translation by Bruce B. Lawrence under the title: Morals for the Heart, New York: Paulist Press. [1991] Morals for the Heart: Conversations of Shaykh Nizam ad-Din Awliya Recorded by Amīr Ḥasan Sijzī and translated by Bruce B. Lawrence, New York: Paulist Press.

124 • JOHANNA BLAYAC

sīrhIndī (Yaḥyā b. Aḥmad) [1931] Tārīkh-i Mubārak Shāhī, edited by S. M. Hidayat Hosain, Calcutta: Bibliotheca Indica. [1996] Tārīkh-i Mubārak Shāhī, partial English translation in Henry Elliot, John Dowson (dir.),The History of India as told by its own Historians, London: Trübner & co; reprint, Delhi: Low Price  Publications, vol. IV, pp. 6-88. yazdī (Ghiyāth al-Dīn ʿAlī) [1379/2001] Saʿādat Nāma, edited by Īraj Afšār, Tehran/  Karachi: Institute of Central and West Asian Studies.

References (except Encyclopaedia of Islam) alam (Muzaffar), delvoye (Françoise ‘Nalini’), gaBorieau (Marc) (dir.) 2000 The making of Indo-Persian Culture. Indian and French Studies, New Delhi: Manohar. ali khan (Masood), ram (S.) 2003 Sufism and Suhrawardi Order, New Delhi: Anmol. anJum (Tanvir) 2011 Chishtī Sufis in the Sultanate of Delhi 1190-1400: from Restrained Indifference to Calculated Defiance, Karachi: Oxford University Press. askari (Syed Hasan) 1952 “Tazkira-i Murshidi. Rare malfuz of the 15th-century  sufi  saint  of  Gulbarga”,  in  Proceedings of the Indian Historical Records Commission. Blayac (Johanna) 2004 La dynastie tughluḳide de Delhi. État des recherches, M.A. diss., Paris, Université Paris-Sorbonne. 2005 Études des sources relatives aux premières communautés musulmanes du Gujarat, viie-xive siècles, MPhil Diss., Paris, Université Paris-Sorbonne. 2012 Formation et histoire des premières sociétés indomusulmanes et indo-islamiques à travers les inscriptions arabes et persanes (viie-xive siècles), PhD diss., Paris, EPHE, 2009, Lille: Atelier national de reproduction des thèses. Boivin (Michel) 2012 Le soufisme antinomien dans le sous-continent indien. Laʾl Shahbāz Qalandar et son héritage, xiiie-xxe siècles, Paris: Éditions du Cerf. delvoye (Françoise ‘Nalini’) (dir.) 1994 Confluences of Cultures: French Contributions to Indo-Persian Studies, New Delhi: Manohar.

delvoye (Françoise ‘Nalini’), leFèvre (Corinne), sharma (Sunil) (dir.) Forthcoming Patronage of Indo-Persian Culture, New Delhi: Manohar. desai (Ziaduddin A.) 1991 Malfuz Literature: As a source of Political, Social, and Cultural History of Gujarat and Rajasthan, Patna: Khuda Bakhsh Oriental Public Library. digBy (Simon) 1976  “Sufis and travelers in the early Delhi Sultanate:  the evidence of the Fawāʾīd al-Fuwād”, in A. Singh (dir.), Socio-cultural Impact of Islam on India, Chandigarh: Panjab University, pp. 171-178. 1980 “The Broach coin-hoard as evidence of the import of valuta across the arabian sea during the thirteenth and fourteenth centuries”, in Journal of the Royal Asiatic Society 2, pp. 129-138. 1984 “Qalandars and related groups: elements of social deviance in the religious life of the Delhi sultanate of the thirteenth and fourteenth centuries”, in Yohanan Friedmann (dir.), Islam in Asia, Jerusalem: The Magnes Press, pp. 60-108. 1986 “Tabarrukat and succession among the great Chishti Shaykhs”, in R. E. Frykenberg, Delhi Through the Ages: Essays in Urban History, Culture and Society, Delhi: Oxford University Press, pp. 63-103. 1990  “The Sufi Shaykh and the sultan: a conflict of  claims to authority in medieval India”, in Iran, Journal of the British Institute of Persian Studies 28, pp. 71-81. 2004 “Before Timur came: provincialization of the Delhi Sultanate through the fourteenth century”, in Journal of the Economic and Social History of the Orient 47, no. 3, pp. 298-356. eaton (Richard M.) 1978 Sufis of Bijapur, 1300-1700: Social Role of Sufis in Medieval India, Princeton: Princeton University Press. 1996 The Rise of Islam and the Bengal Frontier, 12041760, Berkeley: University of California Press. ernst (Carl W.) 1992 Eternal Garden: Mysticism, History, and Politics at a South Asian Sufi Center, Albany: State University of New York Press. Flood (F. Barry) 2009 Objects of Translation: Material Culture and Medieval ‘Hindu-Muslim’ Encounter, Princeton: Princeton University Press.

CONTEXTUALIZING THE GWALIOR QURʾAN • 125

Friedmann (Yohanan) 1984  “The origin and significance of the Chāch Nāma”, in Yohanan Friedmann (dir.), Islam in Asia. South Asia, Jerusalem: Magnes Press, pp. 23-37. haBiB (Mohammed), nIzaMI (Khaliq Ahmad) (dir.) 1970 A Comprehensive History of India, vol. 5: The Delhi Sultanate, A.D. 1206-1526, New Delhi: People’s Publishing House. hardy (Peter) 1960 Historians of Medieval India. Studies in Indo-Muslim Historical Writing, London: Luzac & Company Ltd. First Indian edition with new Preface, New Delhi: Munshiram Manoharlal Publishers, 1997.

1984 “The diffusion of Hindu/Muslim boundaries in South Asia: contrasting evidence from the literature and the tomb cults of selected IndoMuslim saints”, in Peter Gaeffke and David A. Utz (dir.), Identity and Division in Cults and Sects in South Asia, Philadelphia: Department of South Asia Regional Studies, pp. 125-132. Patel (Alka) 2004 Building Communities of Gujarat: Architecture and Society during the Twelfth through the Fourteenth century, Leiden: Brill. schimmel (Annemarie) 1980 Islam in the Indian Subcontinent, Leiden: Brill.

hussaini (Syed Shah Khusro) 1986 The Life, Works and Teachings of Khwajah Bandahnawaz Gisudiraz, Gulbarga: Sayyid Muhammad Gisudiraz Research Academy.

siddiqi (Muhammad Suleman) 1990  “Sufi-state relationship under the Bahmanids  (A.D. 1348-1538)”, in Rivista degli Studi Orientali LXIV, pp. 71-96.

Jackson (Peter) 1999 The Delhi Sultanate: a Political and Military History, Cambridge/New York: Cambridge University Press.

vasantha (Rangashar) 2000 Penugonda Fort. A defence Capital of the Vijayanagara Empire: History, Art and Culture, Delhi: Sharada Publishing House.

katre (Sadashiva L.) 1967 “Sultan Firuz Shah Tughluq: royal patron of a contemporary sanskrit work”, in Journal of Indian History 45, pp. 357-367. kumar (Sunil) 2000 “Assertions of authority: a study of the discursive statements of two sultans of Delhi”, in Muzaffar Alam, Françoise ‘Nalini’ Delvoye, Marc Gaborieau (dir.), The Making of Indo-Persian Culture, New Delhi: Manohar, pp. 37-65. 2005 “La communauté musulmane et les relations hindous-musulmans dans l’Inde du Nord au début du xiiie siècle”, in Annales Histoire, Sciences Sociales 60, pp. 239-264. 2007a The Emergence of the Delhi Sultanate, New Delhi: Permanent Black. 2007b “Balancing autonomy with service: frontier military commanders and their relations with the Delhi sultans in the 13th and the 14th centuries”, in Proceedings of the Punjab History Congress 39, pp. 86-100. lawrence (Bruce B.) 1978 Notes from a Distant Flute: Sufi Literature in Pre-Mughal Literature, Tehran: Imperial Iranian Academy of Philosophy. 1981 “The Chishtiya of sultanate India: a case study of biographical complexities in South Asian Islam”, in Michael A. Williams (dir.), Charisma and Sacred Biography, Chico (CA): Scholars Press, pp. 47-67.

wagoner (Phillip B.) 1996 “Sultan among Hindu kings: Dress, titles, and the islamicization of Hindu culture at Vijayanagara”, in Journal of Asian Studies 55, no. 4, pp. 851-880. welch (Anthony) 1993 “Architectural patronage and the past: the Tughluq Sultans of Indian”, in Muqarnas 10, pp. 311-322. willis (Michael D.) 1985 “An Eight century miḥrāb in Gwalior”, in Artibus Asiae 46, no. 3, pp. 227-246. wink (André) 1990-2004 Al-Hind. The Making of the Indo-Islamic World, Leiden: Brill, 3 vols. zarcone (Thierry) (dir.) 1996 Inde-Asie Centrale: routes du commerce et des idées (volume spécial des Cahiers d’Asie Centrale 1/2), Paris: Édisud.

article from Encyclopaedia of Islam nIzaMI (Khaliq Ahmad) 1977  “Gwāliyār”, in EI2 vol. II, pp. 1170-1171 for the French version.

KALPASŪTRAS ET CORANS : Réflexions sur l’écriture et la peinture de manuscrits jaina du Gujarat aux xive-xvie siècles Nalini BalBir

(Université Sorbonne Nouvelle - Paris 3, École pratique des hautes études)

Résumé Parmi les communautés jaina śvetāmbara du Gujarat, le Kalpasūtra est un texte qui revêt un rôle cérémoniel à nul autre pareil, et ce, en particulier, depuis le xive siècle. Livre-objet et livre-spectacle, il fait l’objet de copies manuscrites nombreuses, dont l’aspect matériel n’est pas anodin. L’environnement général du Gujarat et la présence de corans monumentaux dans la région n’ont pu rester sans impact. Cet essai s’attache à décrire les points de contact visuels qui rapprochent les Kalpasūtras et les corans : calligraphie, chrysographie, usage de papiers colorés, décor végétal des marges, imprimés des costumes. La recherche esthétique qui se concentre dans ces objets de commande suggère que les jaïns śvetāmbara de la région  ont cherché à faire du Kalpasūtra leur Coran, c’est-à-dire le signe visible de leur identité dans le registre du livre. Les peintures témoignent d’un savant équilibre entre traditions picturales communes et goûts spécifiques des commanditaires.

Abstract Kalpasūtras and Qurʾans: Reflections on the Writing and Painting of Jain Manuscripts from Gujarat, 14th-16th centuries In the Śvetāmbara Jain communities of Gujarat, the Kalpasūtra is a text that has played an unparalleled ceremonial role, especially since the 14th century. The book is both a material object and a visual performance in its own right, and was the source of numerous manuscript copies, whose material aspects are significant. The general environment of Gujarat and the presence of monumental Qurʾans in the region could not have been without impact. This article attempts to describe the visual contact points that bring together the Kalpasūtras and the Qurʾans: calligraphy, chrysography, the use of colored paper, the vegetal ornamentation  in  the  margins,  the  motives  painted  on  costumes.  The  esthetic  refinement  that  is  concentrated within these commissioned objects suggests that the Śvetāmbara Jains of  the region sought to make the Kalpasūtra into their Qurʾan, i.e. into the visible sign of their identity within the register of books. The paintings show a skillful balance between common pictorial traditions and the specific tastes of its patrons.

Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 127-138

128 • NALINI BALBIR

Ce bref essai veut faire entendre la voix jaïne dans le dialogue, encore naissant, avec les spécialistes de manuscrits non jaïns produits dans « l’Inde des sultanats »1. Il s’attachera plus particulièrement ici à la période comprise entre le xive et le tout début du xvie siècle. Le cadre géographique sera celui du Gujarat, terme ici préféré à l’expression plus large d’« Inde occidentale » utilisée par de nombreux historiens de l’art, puisque les objets étudiés sont exclusivement produits dans cette région. La période considérée correspond donc en partie à l’époque du sultanat du Gujarat. Marqué par le règne de la dynastie fondée par Aḥmad Shāh, il reste indépendant jusqu’en 1573.  Hindouisme, jaïnisme et islam sont présents dans des rapports complexes qui impliquent une certaine fluidité2. Il vaut sans doute la peine d’expliquer pourquoi, dans un ouvrage centré sur le coran de Gwalior, il peut être pertinent d’examiner de préférence les manuscrits du Kalpasūtra produits au Gujarat. Après avoir tenté de contextualiser leur production, on en décrira les caractéristiques visuelles. Réunir dans un même titre Kalpasūtra et Coran peut paraître hasardeux – ou même sacrilège ! Sans engager de comparaison systématique, on tentera de faire apparaître en  filigrane les points de contact qu’il y a dans les rôles  qu’on a voulu leur faire jouer au sein des cultures manuscrites, tout comme dans la facture des objets. Étant donné l’absence presque totale d’analyse scientifique des manuscrits produits en milieu jaïn au moyen  des techniques modernes, si l’on excepte Issaco3 qui utilise, notamment, un Kalpasūtra du Victoria and Albert Museum, l’observation reste la seule solution.

La tradition jaïne de L’oraL à L’écrit La transmission de la doctrine du jaïnisme, qui remonte à la parole des maîtres que sont les Jina, fut  d’abord  orale.  Mahāvīra,  qui  prêcha  dans  la  région  du Magadha au ve siècle avant notre ère, et fut donc un quasi-contemporain du Bouddha, est dit avoir enseigné à un groupe de disciples en contact direct avec lui. Les enseignements furent ensuite portés par des générations successives de religieux, soutenus par des fidèles laïcs. L’histoire du jaïnisme est marquée  par des séparations, dont la plus importante, qui distingue śvetāmbara et digambara, se produisit dans  les premiers siècles de notre ère. Elle était acquise lorsque, au ve siècle, la tradition des śvetāmbara fut  mise par écrit lors du concile tenu à Valabhī, au Gujarat et organisée en un « canon » réparti en plusieurs groupes. Mais c’est seulement du xie siècle que datent 1. 2. 3.

Brac de la Perrière (2008). Sheikh (2010). Issaco (2008).

les plus anciens manuscrits qui nous sont parvenus. Depuis, en revanche, la transmission manuscrite est ininterrompue, vaste et riche, en sorte que la tradition jaïne est probablement l’une des plus importantes « cultures manuscrites » de l’Inde.

Le KaLpasūtra Livre-objet, Livre-sPectacLe Le Kalpasūtra, objet de cet exposé, est un ouvrage de ce canon śvetāmbara.  Comme  tel, il est écrit en  langue prakrite, variété ardhamāgadhī. S’il n’est pas  comme les autres, c’est que, depuis le xive siècle en particulier, il occupe une place à part dans la tradition manuscrite tant les exemplaires en sont nombreux, et dans la tradition picturale, tant ils sont nombreux à être peints ou diversement décorés. Le Kalpasūtra fournit un exemple intéressant de livre perçu comme objet sacré, la matérialité s’ajoutant au contenu. Ce dernier regroupe des éléments centraux pour définir  l’identité des jaïns śvetāmbara. Le premier chapitre  (Jinacarita, « Vies des Jina ») leur fournit l’essentiel sur les fondateurs, le deuxième (Sthavirāvali « Lignée des Anciens ») donne au jaïnisme sa généalogie spirituelle en déroulant la succession de ses premiers maîtres, et le troisième (Sāmācārī « Conduite monastique appropriée ») est un exposé détaillé des règles particulières de vie que les religieux doivent observer durant la saison des pluies – période de « retraite », et non de pérégrinations comme le reste de l’année. Il se conclut en insistant sur la notion de pardon pour les transgressions éventuellement commises. Se trouve intimement associé au Kalpasūtra un autre texte, souvent copié à la suite dans les manuscrits. Il s’agit de l’une ou l’autre des versions de l’histoire du maître Kālaka (Kālakācāryakathā), tenu pour avoir modifié et fixé la date du dernier jour de la cérémonie  rituelle de Paryushan, lors de laquelle moines et laïcs sont réunis pour un repentir collectif. Par convention, dans le présent contexte, le terme « Kalpasūtra » vaut  pour l’ensemble ainsi formé. L’abondance d’exemplaires du Kalpasūtra, à partir du xive siècle, ne résulte pas seulement d’une créativité spontanée. Les manuels de conduite destinés à encadrer la vie religieuse quotidienne des fidèles laïcs  jaïns, écrits par des maîtres religieux souvent prestigieux, fournissent un cadre théorique et les encouragent à commanditer la copie de manuscrits. De l’un à l’autre, ils se répètent beaucoup. Les différences sont d’autant plus significatives. Au xiie siècle, au Gujarat, Hemacandra crée une nouvelle expression, parlant « des sept champs » dans lesquelles un bon jaïn doit « semer » sa richesse. Payer pour des statues et pour des temples sont les deux premiers. Le troisième, pour sa part, touche aux enseignements contenus dans les écritures et aux manuscrits (pustaka) qui les

KALPASU¯ TRAS ET CORANS • 129

transmettent. Hemacandra est assez bref : le laïc est invité « à donner respectueusement aux moines des manuscrits qui ont été copiés pour leurs prêches, et à les écouter avec dévotion quand ils prêchent »4. Au xive siècle, dans la même région, Ratnaśekharasūri,  un autre maître auteur d’un ouvrage comparable  destiné aux laïcs, est manifestement inspiré par son prédécesseur qu’il répète en partie. Mais il innove : « Faire copier des manuscrits relatifs aux vies des Jina, comme le Kalpasūtra, grâce à de l’argent gagné honnêtement, avec des feuillets non ordinaires et des lettres spéciales et parfaites, et les faire lire par des moines savants chaque jour à l’occasion de l’inauguration ou du cours des fêtes (…) doit être prescrit pour éveiller les différentes personnes qui peuvent l’être ».5

Les assertions de ces deux auteurs, à deux siècles d’intervalle, prennent la forme de prescriptions. En même temps, elles correspondent à deux points distincts dans l’évolution de la tradition manuscrite. Hemacandra s’exprime à une époque où les manuscrits  deviennent de plus en plus nombreux au Gujarat – son confrère du xive siècle en un temps où, justement, les manuscrits du Kalpasūtra commencent à faire l’objet d’une production de masse. Qu’il évoque « des feuillets non ordinaires et des lettres spéciales et parfaites » paraît aussi une référence on ne peut  plus claire à deux des caractéristiques matérielles de bien des manuscrits du Kalpasūtra sur papier : l’emploi de papiers à fonds colorés, et le recours à une forme de calligraphie qui tranche sur la cursive ordinaire. À leur manière donc, ces deux auteurs témoignent des modifications attestées par les documents eux-mêmes.  Au xive siècle, la copie de Kalpasūtras est entrée dans les mœurs et a la caution des autorités religieuses. L’acte de piété va de pair avec un objet qui se distingue. La diffusion de la culture écrite est une composante de l’éthique du fidèle laïque. Le Kalpasūtra en devient l’emblème. Ratnaśekhara souligne encore un autre trait, que  la pratique corrobore et soutient : le Kalpasūtra est associé à une fête et devient un « livre spectacle » ;  à cette occasion ses manuscrits font l’objet d’un usage public par les moines. Ici encore, l’époque de l’auteur paraît correspondre à un tournant. Le texte lui-même  4.

Hemacandra [1977], 572 sur III 119 : likhitānāṃ ca pustakānāṃ saṃvigna-gītārthebhyo bahumāna-pūrvakaṃ vyākhyānārthaṃ dānaṃ, vyākhyāyamānānāṃ ca pratidinaṃ pūjā-pūrvakaṃ śravaṇaṃ ceti. 5.  Ratnaśekharasūri [1960], section 6, 41b : pustakānāṃ śrīKalpādy-āgama-Jinacaritrādi-satkānāṃ nyāyārjita-vittena viśiṣṭapatra-viśiṣṭa-viśuddhākṣarādi-yuktyā lekhanaṃ tathā vācanaṃ saṃvigna-gītārthebhyaḥ proḍha-prārambhādy-utsavaiḥ pratyahaṃ pūjādi-bahumāna-pūrvakaṃ vyākhyāpanaṃ aneka-bhavyapratibodha-hetur vidheyam. upalakṣaṇatvāt tad-vācana-bhaṇanādi-kṛtāṃ vastrādibhir upaṣṭambha-pradānaṃ ca.

fait seulement état d’observances monastiques particulières commencées « un mois et vingt jours après le début de la saison des pluies ». L’assertion de l’auteur du xive siècle, au contraire, fait penser à la situation observable aujourd’hui. En clôture de la fête annuelle appelée  Paryushan,  qui  marque,  notamment,  la  fin  de la saison des pluies, le texte est lu dans l’original ou en traduction ou sert de support au sermon improvisé par les religieux dans les temples. Les pages illustrées sont montrées aux fidèles en accompagnement de la  parole ; le manuscrit est aussi porté solennellement  en procession dans les rues. Dans ce processus, l’objet lui-même a autant, sinon plus, d’importance que le contenu même. Même s’il n’y a pas d’objection explicite à réutiliser le même exemplaire d’une année sur l’autre, la pratique a encouragé la copie de « nouveaux » manuscrits, ou, de nos jours, l’impression de « nouveaux » exemplaires. La production de masse de manuscrits à peintures du Kalpasūtra à partir du xive siècle a donc un but concret et circonstanciel. Les commanditer devient un phénomène de mode, ou en tout cas, une activité de prestige parmi les riches communautés marchandes qui trouvaient là un moyen recommandé par la doctrine de dépenser leur argent ainsi qu’un moyen de reconnaissance sociale.

L’esthétique des Kalpasūtras sur PaPier : caLLiGraPhie, chrysoGraPhie, PoLychromie, marGes et bordures

Au Gujarat, les premiers manuscrits disponibles sont écrits sur feuille de palme (tāḍapatra). Durant une première période, qui va de 1100 à 1350, c’est le seul support utilisé. Durant la seconde, de 1350 à 1450, la feuille de palme est progressivement, puis définitivement, remplacée par le papier. Dans les deux cas, le format général reste le même : la page est rectangulaire, les feuillets ne sont pas reliés. Par ailleurs, le recours à l’un ou l’autre matériau et le passage de l’un à l’autre ne sont évidemment pas sans effet sur les styles. L’écriture nāgarī, ou nāgarī dite jaina, qui est celle  utilisée pour les textes jaïns, n’a pas de tradition calligraphique au sens strict, comme on parlerait de calligraphie arabe ou chinoise. Néanmoins, de nombreux manuscrits du Kalpasūtra donnent à voir des formes qui s’apparentent à des réalisations calligraphiques, et opèrent des choix qui les distinguent du tout venant. Le trait le plus superficiel est la taille,  importante, des caractères. Leur contour apparaît  dessiné avec un soin particulier, comme pour leur donner une épaisseur d’autant plus manifeste qu’ils sont fortement encrés et noirs. Leur forme est souvent angulaire, elle est volontiers renforcée par des petits traits qui sont autant de prolongements ornementaux, tandis que les points indiquant les voyelles nasalisées

130 • NALINI BALBIR

Figure 1 – Exemple de calligraphie dans un manuscrit du Kalpasūtra de 1464, Or. 5149, fol. 1v. [© British Library]

tendent à prendre la forme de petits cercles bien dessinés (figure 1). Ces caractéristiques sont présentes  dans les manuscrits sur palme, mais se déploient avec une grande ampleur dans ceux sur papier. D’un autre côté, le passage d’un support à l’autre implique des changements notables. L’usage du papier équivaut à l’introduction d’une nouvelle technologie et permet des innovations. La chrysographie est la première d’entre elles. À cette époque, au Gujarat, elle paraît limitée aux manuscrits du Kalpasūtra, ce qui n’est probablement pas le résultat du hasard (figure 2).  L’encre d’or pour écrire est considérée comme un produit rare et n’est donc utilisée que dans des circonstances choisies6. En outre, elle doit être maniée par un scribe compétent car elle se prête mal aux corrections. D’ailleurs, lorsque celles-ci existent, elles sont régulièrement faites à l’encre noire. La chrysographie convient évidemment à un livre-objet destiné à être regardé autant que lu, à un objet de prestige. Il n’est pas rare que les colophons des Kalpasūtras chrysographiques prennent acte de cette particularité,  comme  avec  fierté,  dans  un  processus  d’autoréférence. Il en va de même pour l’encre d’argent, également d’usage remarquable, comme le souligne le colophon du Kalpasūtra de 1427 : « Puisse ce manuscrit avec des lettres d’argent pareilles à de grosses perles être source de joie, manipulé par des gens intelligents »7. De fait, elle y est majoritairement employée, à côté de l’encre d’or (sur cinq faces), de la rouge (sur sept) et de la bleue (sur deux).

Rouge et noir alternent tantôt systématiquement, tantôt plus irrégulièrement. Les encres bleues et rouges (supra) ne sont utilisées que sur les fonds crèmes – elles seraient illisibles ailleurs. À côté du papier coloré en rouge, majoritairement employé dans le Kalpasūtra de 1512 conservé au Wellcome Trust à Londres, on compte huit feuillets à fond bleu, quatre à fond noir, et un bleu au recto, noir au verso (figure 3). En revanche,  les principes qui régissent l’alternance, s’ils existent, sont difficilement décelables : la polychromie est-elle  rationnelle ou purement esthétique ? Comme on sait, le papier coloré est un support fréquemment utilisé en Perse au xve siècle8. Le passage de la feuille de palme au papier ouvre de nouvelles perspectives à l’embellissement du manuscrit par la présence de décorations diverses. Bordures et marges sont gardées vides dans les

6. 7.

8.

Chandra (1949), 75. British Library I.O. San. 3177.

Un autre marqueur esthétique et distinctif consiste à délaisser le papier de couleur crème banal, en privilégiant un papier qui aura été peint, puis, une fois sec, utilisé pour écrire. La teinte la plus représentée est le rouge – du vermillon au carmin. Un pas supplémentaire  dans  le  raffinement  est  l’emploi,  dans  un  même manuscrit, de papiers de couleurs différentes. Dans le Kalpasūtra de 1427 la répartition est la suivante : - fond rouge : 151 faces - fond noir : 143 faces - fond crème : 11 faces

Déroche (dir.) (2000), 67-68.

KALPASU¯ TRAS ET CORANS • 131

Figure 2 – Exemple de chrysographie dans un manuscrit du Kalpasūtra (xve siècle ?), BL. Or. 14262, fol. 77v. [© British Library]

Figures 3a-b – Emploi de papiers à fonds peints dans les manuscrits du Kalpasūtra de 1512, Gamma 453, fol. 63v et 64r. [© Wellcome Trust]

132 • NALINI BALBIR

Figure 4 – Exemples de bandeaux médians décorés dans le Kalpasūtra de 1427, BL. I.O. San. 3177, fol. 6v, 19v, 26v, 41v, 70v. [© British Library]

manuscrits sur palme, et peuvent le rester dans les manuscrits sur papier. C’est là affaire de choix et de moyens financiers, non de chronologie : le Kalpasūtra de 1458 présente des marges vides contrairement à celui de 1427. Très souvent, cependant, les bordures externes sont les premiers espaces des manuscrits sur papier où, discrètement, apparaissent des nouveautés :  des frises de fleurons bleus ou des tiges, comme on en  trouve aussi dans les corans, forment l’un des principaux motifs d’enluminure ; les vignettes marginales  et centrales décorées en sont un autre. Les quatre marges des Kalpasūtras sur papier apparaissent comme l’espace de liberté et de créativité par excellence. Elles tendent à se peupler de formes et de motifs variés, où les arabesques, les entrelacs de fleurs  et tout un vocabulaire végétal décoratif ont la première place. De surcroît, les artistes jouent sur la mise en  page en sorte de ménager un espace de décoration supplémentaire : le feuillet est divisé en deux parties par un bandeau médian, de part et d’autre duquel le texte est réparti. Et ce bandeau est richement décoré de motifs géométriques ou floraux (figure 4). L’exemplaire qui aurait été exécuté vers 1475 aux environs de Broach et dont les feuillets sont malheureusement dispersés à travers le monde est un cas extrême de richesse décorative. Volutes foliées, motifs floraux et couples d’oiseaux y alternent. L’ensemble  du décor rappelle un tapis, et aucun espace n’est laissé

libre, tandis que, sur la droite, la scène représente des épisodes du légendaire jaïn, et que le texte déroule la généalogie des premiers maîtres. Chaque feuillet est  différent, mais tous dégagent la même impression d’un luxe qui veut se faire savoir. Alors que le programme iconographique des manuscrits sur palme est restreint, celui-ci déploie un large éventail de scènes. Loin de se contenter de fournir une contrepartie visuelle au texte copié, il attire à lui tous les épisodes qui s’y rattachent, de près ou de loin, dans une interprétation inclusive de la tradition. Les trois moines traversant une  rivière  de  ce  recto  (figure 5)  sont  une  mise  en  images de la situation décrite par le texte en regard et de la règle qu’il prescrit ; le verso9 montre des femmes supposées incarner les rāgas ou modes musicaux – sans rapport donc avec le texte et son atmosphère. Dans le Kalpasūtra daté de 1501 (figure 6), le bandeau médian conserve un motif floral décoratif, dans  l’esprit de l’art persan, mais les bordures ne sont plus décoratives. Elles représentent des scènes figurées qui  démultiplient et rappellent visuellement la peinture principale dont elles conservent le style. Le roi, père du futur Jina, se livre ici aux exercices physiques qui forment l’ordinaire de ses matinées ; des couples de  lutteurs y font écho en bordure. 9.

http://collections.lacma.org/node/170527.

KALPASU¯ TRAS ET CORANS • 133

Figure 5 – Scène illustrant le régulaire monastique de la saison des pluies dans un Kalpasūtra (ca. 1475). [© Los Angeles County Museum of Art]

Figure 6 – Exercices physiques du roi Siddhārtha avec démultiplication de la scène principale : Kalpasūtra de 1501. [© Parimoo, 2010, fig. 8]

Objet de la vie quotidienne et marqueur social, le vêtement est exposé à toutes les influences de l’environnement et aux modes. Sa représentation dans la peinture manuscrite jaïne ne fait pas exception : quadrillages noirs ou motifs à trèfles sur fonds roses,  voiles en transparence rappellent des éléments présents dans la peinture persane ou des motifs ornementaux dans les Corans. L’histoire du moine Kālaka, souvent relatée à la  suite du Kalpasūtra proprement dit, est le lieu principal où l’étrangeté est représentée comme frontalement.  Pour triompher du roi qui s’est rendu coupable de l’enlèvement de sa nonne de sœur, le moine Kālaka 

fait appel aux ressources humaines et financières du  fameux Sāhi dont le royaume s’étend au-delà de l’Inde.  Dans la représentation classique en vigueur depuis les manuscrits sur palme10, le Sāhi se distingue d’emblée  par son physique et sa tenue vestimentaire : visage plat, traits mongoloïdes, œil bridé, pose de face et non de profil11. L’influence persane est comme revendiquée  dans la représentation de ce personnage, incarnation par excellence de l’étranger. Le roi fait face au moine

10. Brac de la Perrière (2008), 243. 11. Ibid., 233.

134 • NALINI BALBIR

jaïn, tandis que ses subordonnés sont en pleine activité dans les plans inférieurs de la miniature principale (figure 7). Les bordures, divisées en cartouches, sont  peuplées d’individus appartenant à cette population vaquant à de multiples occupations, notamment belliqueuses. La profusion de ces personnages, qui plonge dans un univers différent, fait de ce manuscrit un cas singulier presque extrême. Dans l’immense majorité des exemplaires du Kalpasūtra, les traits distinctifs sont limités à la représentation de ce Sāhi : il est étranger,  et représenté d’une manière différente des personnages indiens, mais seul lui l’est ainsi. L’étrangeté demeure donc marginale12. Hors de cet épisode, le style pictural est classique et caractéristique : visage représenté de face uniquement pour les Jina en position de méditation, de profil pour les autres personnages, œil protubérant caractéristique du style dit de « l’Inde occidentale ».

déroulent selon une structure binaire. La première étape met au centre le commanditaire et sa famille, la seconde le maître religieux et sa généalogie spirituelle. Cette caractéristique formelle est une manière de souligner le rapport étroit entre les deux parties qui forment la communauté jaïne : elles ne peuvent agir que l’une avec l’autre. Les fidèles laïcs sont les  exécutants, les maîtres religieux les donneurs d’ordre  ou, au moins, des instigateurs convaincants. Dans la première moitié du xve siècle, l’un de ces religieux, Jinabhadrasūri, est très présent. C’est lui qu’on a  voulu représenter dans le Kalpasūtra de 1416 dont il a suscité la copie, face au texte donnant sa filiation  spirituelle, même s’il ne s’agit pas d’un portrait au sens strict. Le schéma syntaxique de chaque généalogie met en relief l’élément central – le fidèle laïc ou le  religieux –  qui  est  l’agent ;  les  autres  membres  de  la famille génétique ou spirituelle contribuent à

Figure 7 – Kālaka et le Sāhi : Kalpasūtra de 1501. [© Parimoo, 2010, fig. 9]

des Goûts et des commanditaires Que savons-nous de la conception et de la production de ces beaux objets ? Les informations viennent de leurs colophons. Souvent poèmes de style savant en sanskrit, ils font d’emblée du manuscrit qu’ils terminent un objet non anodin. C’est là qu’il faut chercher la présence, souvent ostentatoire, des partenaires humains impliqués dans le processus. Ces colophons se

12. Nos observations concordent en tous points avec celles de R. Parimoo (2010), 39: « We must also note that repeated checking of Kālaka story illustrations throughout the extent  illustrated corpus of the 15th century (nearly one hundred years) reveals that the Persian element is restricted only to the Sāhi king’s identity right from the 13th century manuscripts on palm-leaf. The influence is not spreading ».

inscrire le fidèle dans un réseau social, qui, de fait,  devient partie prenante de l’entreprise dans laquelle un individu est impliqué. Plus l’énoncé de la parentèle du fidèle est fourni, plus l’entreprise assume une  dimension collective, dans un système où la famille  élargie (joint family) joue son rôle. Il fallait certainement plus d’une bourse pour prendre en charge les frais impliqués par la production de tels manuscrits, dont tout concourt à faire des objets de luxe et de prestige. Mais les informations financières font cruellement défaut. S’il est occasionnellement fait état de patronages royaux dans des textes légendaires du xiie siècle, époque où le Gujarat est gouverné par le  roi Kumārapāla, hindou converti aux enseignements  du jaïnisme, il serait plus inattendu à l’époque du sultanat du Gujarat. Le processus ne fait intervenir que les religieux instigateurs, les notables commanditaires, et les copistes et artistes.

KALPASU¯ TRAS ET CORANS • 135

Les noms des peintres ne sont que rarement donnés. Un manuscrit de l’histoire de Kālaka daté de 1416 porte  celui d’un certain Deyiyā13, désigné comme le peintre, un autre de 1459 celui d’un certain Sārang. Rien de  plus. Mais on est au moins sûr de leur rôle d’illustrateur, car le nom du copiste est donné séparément. Dans la majorité des cas, on ne trouve qu’un seul nom dans l’expression « écrit par ». Le scribe est-il aussi le peintre ? Les indications présentes à côté des peintures fonctionnent probablement comme des directives destinées au peintre, plus que comme des légendes destinées au spectateur, le cadre réservé à la peinture étant parfois vide. La récurrence du même nom de copiste dans plusieurs exemplaires est un fait notable. Certains se sont spécialisés dans la copie, ou la copie et l’illustration, des manuscrits du Kalpasūtra. Dans une certaine mesure, pareille spécialisation va de pair avec la production de masse de ces manuscrits durant la période considérée, et avec la volonté d’en faire des objets différents des manuscrits ordinaires. On voit ainsi apparaître, sur une durée d’environ  25 ans pendant la seconde moitié du xve siècle, le nom d’un certain Mantri Vāchaka, habitant de la ville de  Patan au Gujarat, comme copiste de manuscrits du Kalpasūtra, dont plusieurs en lettres d’or ou d’argent. On lui en doit au moins une douzaine d’exemplaires, conservés dans des collections distinctes14. Il serait évidemment souhaitable de pouvoir les inspecter ensemble et complètement pour une étude stylistique comparée. Un  autre  scribe,  un  certain  Trikuṇṭha  Dvivedi,  a copié deux Kalpasūtras : celui en lettres d’argent de 1427 de Londres15, et, quelques mois auparavant, un exemplaire en lettres d’or conservé en Inde. Tous deux présentent des affinités dans la facture et le décor, 

pour autant qu’on puisse les comparer, car le second est pour l’instant inaccessible, mis à part des reproductions isolées16. Il est difficile de déterminer l’affiliation religieuse  de ces scribes, mais au moins le nom du dernier nommé suggère un hindou plutôt qu’un jaïn. Quoi qu’il en soit, ils ont œuvré en un même lieu, la ville de Patan, ou en sont des habitants. Ces données contribuent à distinguer Patan comme centre de production privilégié durant la période considérée17. Il est fort probable qu’il y existait des ateliers spécialisés dans la copie et l’illustration des Kalpasūtras, avec une tradition commune. Des manuscrits sur palme y ont été produits au xiiie siècle, des manuscrits sur papier y sont produits du xive au xvie siècle18. Centre particulièrement vivant de la culture jaïne depuis au moins le xie siècle, Patan le reste alors même qu’elle est la capitale du sultanat du Gujarat de 1391 à 1411 et aussi plus tard. Son architecture manifeste également des apports variés qui s’entrecroisent. La peinture des Kalpasūtras conserve ses propres fondamentaux. Mais pour le peintre au travail entrent en jeu deux éléments distincts. D’une part, il doit respecter une tradition picturale bien établie pour les sujets traditionnels et les personnages jaïns. On peut imaginer que c’est la condition nécessaire pour satisfaire les commanditaires des manuscrits, et l’autorité religieuse qui les cautionne. D’autre part, il faut compter avec la mobilité continuelle des techniques artistiques, les tendances picturales dans l’air du temps, les modes, auxquelles le peintre peut être sensible en tant qu’artiste ancré dans la société. L’essentiel est donc d’assurer un équilibre harmonieux des goûts. Le feuillet 16 verso du Kalpasūtra de 1427 (figure 8) associe ainsi une peinture  représentant l’épisode emblématique et identitaire

Figure 8 – Association de codes picturaux jaïns et d’éléments décoratifs indo-persans dans le Kalpasūtra de 1427, BL. I.O. San. 3177, fol. 16v. [© British Library]

13. Parimoo (2010), 82-85. 14. Ibid., 69-70 et 288-292, non exhaustif. 15. I.O. San. 3177.

16. Balbir et al. (2006), vol. I, 137-144. 17. Parimoo (2010), 288-292. 18. Ibid., 139.

136 • NALINI BALBIR

des jaïns śvetāmbara où, sur ordre du dieu Śakra, son  messager  Hariṇaigameṣin,  transfère  l’embryon  du  futur Mahāvīra de la matrice d’une brahmine à celle  d’une kṣatriya, et des bordures à décor géométrique  indo-persanes, tandis que la bordure médiane entrelace des arabesques sur fond bleu qui évoquent cet univers. Du manuscrit sur palme du xiiie siècle au manuscrit sur papier des xve-xvie siècles, les goûts évoluent. En plein sultanat du Gujarat, en pleine présence des manifestations de l’art islamique dans cette région, plutôt que « d’influences » persanes, souvent exagérées par  certains historiens de l’art, ou de « fusion »19, il vaut mieux  parler  de  « flux  quasi  circulaire  d’influences  et d’héritages entre l’art du livre jaïn et l’art du livre islamique »20. Le peintre doit négocier entre divers codes picturaux à sa disposition, sans perdre de vue la  finalité  de  l’objet  de  commande  qu’il  prépare  et  l’identité de ceux qui le commandent. L’hybridité contrôlée est donc de mise. C’est pourquoi les décors en marges et bordures, ou les imprimés des costumes, sont les lieux principaux de l’innovation, dans une iconographie jaïne caractérisée par des traits bien marqués, qu’elle conserve sans altération foncière en dépit de la présence de codes différents dans son entourage.

19. Ibid, 30. 20. Brac de la Perrière (2008), 243.

traditions PicturaLes

et identités reLiGieuses

L’environnement général du Gujarat au xive siècle a probablement joué un rôle, que les jaïns eux-mêmes ont tendance à minimiser. Ils n’ont pourtant pu faire abstraction des manuscrits arabo-persans. Les corans, et particulièrement les corans monumentaux, qui utilisaient à profusion l’or, l’argent et la calligraphie contribuant au raffinement et à la sophistication devaient être des références présentes qu’il leur était difficile d’ignorer. Tout se passe comme si les jaïns  śvetāmbara  du  Gujarat  avaient  cherché  à  faire  du  Kalpasūtra leur Coran, c’est-à-dire le signe visible de leur identité dans le registre du livre. La recherche esthétique qui se concentre dans les nombreux manuscrits de ce texte, ouvrages de commande exécutés avec attention et marqués par une cohérence stylistique certaine, y contribue largement. La calligraphie, la chrysographie, l’usage de papiers colorés, le décor végétal des marges21 ou les imprimés des costumes sont autant de signes de l’existence de traditions picturales communes. Mais elles donnent lieu à des réalisations diverses en fonction des textes illustrés (Corans ou Kalpasūtras) et des traditions religieuses qu’ils représentent dans un savant équilibre de nature à satisfaire leurs commanditaires respectifs.

21. Brown (1937), 235.

bibLioGraPhie Sources

Références

bhadrabāhu [1879] The Kalpasūtra of Bhadrabāhu, edited with an Introduction,  Notes  and  a  Prâkrit-Saṃskrit  Glossary by Hermann Jacobi, Leipzig : Abhandlungen für die Kunde des Morgenlandes.  hemacandra [1977] Yogaśāstram avec auto-commentaire, édité par Muni Jambuvijaya, Mumbai : Jain Sāhitya Vikāsa  Maṇḍala.

BalBir (Nalini) 1984 « Un manuscrit illustré du Kalpasūtra jaina conservé à la Bibliothèque nationale (cote : Sanscrit 1453) », dans Bulletin d’études indiennes 2, p. 17-39. 2010 « Is a manuscript an object or a living being? Jain views on the life and use of sacred texts », dans Kristina Myrvold (dir.), The Death of Sacred Texts: Ritual Disposal and Renovation of Texts in World Religions, Farnham : Ashgate, p. 107-124.

ratnaśekharasūrI [1960] Śrāddhavidhiprakaraṇa avec auto-commentaire, éd. revue par Paṃnyāsa Śrīvikramavijaya, Gaṇī  Muniśrībhāskaravijaya, Surate : Śreṣṭhi-Devacanda-Lālbhāī-Jaina Pustakoddhara Fund, 106.

BalBir (Nalini) et al. 2006 Catalogue of the Jaina Manuscripts at the British Library, the British Museum and the Victoria and Albert Museum, Londres : British Library and the Institute of Jainology.

KALPASU¯ TRAS ET CORANS • 137

Bednar (Michael) 2012 Illuminating Cultural History: Jains and Muslims in Late Miniature Paintings, communication présentée lors du Medieval Academy of America Meeting, St Louis University, États Unis, 22-24 mars 2012 (non vu). Brac de la Perrière (Éloïse) 2008 L’art du livre dans l’Inde des sultanats, Paris : Presses de l’université Paris-Sorbonne. Brown (William N.) 1933 The Story of Kālaka. Texts, History, Legends, and Miniature  Paintings  of  the  Śvetāmbara  Jain  Hagiographical  Work  the  Kālakācāryakathā,  (Oriental Studies 1), Washington : Smithsonian Institution Freer Gallery of Art. 1934 A Descriptive and Illustrated Catalogue of Miniature Paintings of the Jaina Kalpasūtra as exectued in the early western Indian style (Oriental Studies 2), Washington : Smithsonian Institution Freer Gallery of Art. 1937 « A Jaina manuscript from Gujarat illustrated in early Western Indian and Persian Styles », dans Ars Islamica 4, p. 154-172. chandra (Moti) 1949 Jain Miniature Paintings from Western India, vol. 1, Ahmedabad: Sarabhai Manilal Nawab. chandra (Moti ), shah (Umakant P.) 1975 New Documents of Jaina Painting, Bombay: Shri Mahavira Jaina Vidyalaya Publication. dÉroche (François) (dir.) 2000 Manuel de codicologie des manuscrits en écriture arabe, Paris : Bibliothèque nationale de France. dhavalikar (Madhukar K.), BaPat (Shreenand L.) (dir.) 2011 Paryuṣaṇā-Kalpasūtra, Pune: Bhandarkar Oriental Research Institute. doshi (Saryu) 1983 « Islamic elements in Jain manuscript illustration », dans Karl Khandalavala (dir.), An Age of Splendour. Islamic Art in India, Bombay : Marg, vol. 35/2, p. 114-121. Flood (Finbarr B.) 1983 « The Qur’an », dans Helena Evans (dir.), Byzantium and Islam: Age of Transition, catalogue d’exposition, New York : Metropolitan Museum of Art, p. 265-273.

isacco (Enrico) 2008 Les pigments des miniatures indiennes, Paris : L’Asiathèque. khosla (Preeti) en cours A Study of the Visual Language of the Indigenous Styles of Book Painting in North India during the Sultanate Period (1414-1525 A.D.), thèse de doctorat en cours à la School of Oriental and African Studies, Londres. masteller (Kimberley) 2012 « Multiple sources: contextualizing an early sixteenth-century manuscript of the Kalpasutra », communication présentée lors du Medieval Academy of America Meeting, St Louis University, États Unis, 22-24 mars 2012 (non vu). nawaB (Sarabhai M.), nawaB (Rajendra S.) 1985 Jaina Paintings, vol. 2. Paintings on Paper, Commencing from vs. 1403 to vs 1656 Only, Ahmedabad : Messrs Sarabhai Manilal Nawab. Parimoo (Ratan N. C.) 2010 Mehta Collection, vol. I. Gujarati School and Jaina Manuscript Paintings, Ahmedabad : Gujarat Museum Society. sheikh (Samira) 2010 Forging a Religion: Sultans, Traders, and Pilgrims in Gujarat 1200-1500, Oxford : Oxford University Press. simPson (Edward), kaPadia (Aparna) (dir.) 2010 The idea of Gujarat: History, ethnography and text, New Delhi : Orient Blackswan.

Références et sources des principaux manuscrits du Kalpasūtra étudiés Kalpasūtra de 1416 (1473 de l’ère Vikrama) : cote non spécifiée ; conservé à Jira (Penjab), cf. Parimoo (2010),  66 peint par Deyiyā, copié par Somasiṃha. Kalpasūtra de 1427 (1484 de l’ère Vikrama) : I.O. San. 3177 conservé à la British Library, Londres. Numérisé sur www.jainpedia.org et ici figures 4 et 8 ; colophon  transcrit, traduit et commenté dans Balbir et al. (2006), vol. I, 137-144. Kalpasūtra de 1427 (1484 de l’ère Vikrama) : frère du précédent, copié par le même scribe, conservé à Palitana, cf. Nawab, Nawab (1985), 26-27 et fig. 179,  qui  pourrait  représenter  le  maître  Jinabhadrasūri ;  reproduit dans Parimoo (2010), 65, fig. 7 avec légende  erronée.

138 • NALINI BALBIR

Kalpasūtra vers 1450 : feuillets conservés au Walters Art Museum, Baltimore, cf. http://art.thewalters.org/ detail/7912/two-illustrated-pages-from-a-kalpasutra-manuscript. Kalpasūtra de 1458 (1515 de l’ère Vikrama) : No 761/ 1899-1915 conservé au Bhandarkar Oriental Research Institute, Pune : reproduit intégralement dans Dhavalikar, Bapat (2011). Kalpasūtra de 1464 (1521 de l’ère Vikrama) : Or. 5149, British Library, Londres. Numérisé sur www.jainpedia.org. Voir ici figure 1. Kalpasūtra,  peut-être  vers  1475 :  Dayā  Vimalajī  Bhaṇḍāra, Devasano Pada, Ahmedabad : Brown (1978),  232-241 ;  Chandra  (1949),  fig. 106-136 ;  certains  feuillets sont au Los Angeles County Museum of Art, cf. http://collections.lacma.org/node/170527 ;  l’un  d’eux est reproduit ici figure 5. Kalpasūtra de 1492 (1549 de l’ère Vikrama) : Sanscrit 1453 conservé à la BnF. Numérisé sur Gallica : http:// gallica.bnf.fr/ark:/12148/btv1b6000521k/f1.image. Voir Balbir (1984).

Feuillets épars d’un Kalpasūtra qui pourrait dater du Or. 14262, British Library, Londres. Numérisé sur www.jainpedia.org. Voir ici figure 2.

xve siècle :

Kalpasūtra de la fin du xve siècle : IS 46-1959, Victoria & Albert Museum, Londres. Numérisé sur www.jainpedia.org. Utilisé pour l’analyse des pigments dans Issaco (2008). Kalpasūtra de 1501 conservé à l’Ancalagaccha Upashray, Jamnagar : figures 26-29 en noir et blanc dans Chandra,  Shah (1975) ; fig. 8 à 10, 88-89 dans Parimoo (2010) et  ici figures 6 et 7. Kalpasūtra de 1512 : Gamma 453, Wellcome Trust, Londres. Numérisé sur www.jainpedia.org et ici figures 3a-b.

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY* Preliminary Remarks Asma Hilali

(Institute of Ismaili Studies, London)

Abstract The aim of my contribution is to propose a number of methodological remarks related to the approach to Qurʾan manuscripts and their transmission history by raising the questions as to when, where and what the technicalities of transmission of the Qurʾan manuscripts are, and under what circumstances the Qurʾan manuscripts were read, dictated and copied by their contemporaries. In order to answer these questions, the circumstances of transmission of the Qurʾan and the circle of teaching will be addressed. In this paper, I firstly describe the various tendencies in contemporary Qurʾan manuscript studies and underline the methodological problems these raise. Secondly I reflect on the techniques of transmission of the Qurʾan manuscripts on the basis of three sources: An early Qurʾan manuscript dated to the 7th century, narratives about the seventh-century circles of transmission extracted from ṭabaqāt literature, and a tenth-century theoretical text about the rules of transmission. My choice of sources respects a progressive chronology and complementarity between empirical study (manuscript evidence) and a study of the representation of the textual transmission in historical narratives as well as in theoretical texts. My contribution proposes an overview of the transmission technicalities of Qurʾan manuscripts in early and medieval times and attempts to shed light on the annotations and glossaries in the Qurʾan manuscripts, strongly present in the Gwalior Qurʾan, and their contribution in echoing the dynamic link between oral and written transmission of the Qurʾan in teaching circles.

Résumé Manuscrits coraniques : Histoire et transmission Quand, où et par quels biais, quelles techniques, les manuscrits coraniques ont-ils été transmis ? Dans quelles circonstances ont-ils été lus, dictés et copiés ? L’article propose plusieurs pistes méthodologiques visant à mieux cerner l’ensemble de ces questions. Après avoir décrit les différents courants d’études contemporains ayant trait aux manuscrits

*

My thanks go to Éloïse Brac de la Perrière, Monique Burési, Michael Feener, Russell Harris, Stephen Burge and Sarah Novak for their feedback on an earlier version of this paper. I am also grateful to Sabrina Alilouche and Ghazaleh Ismailpour Qouchani for the rich discussions during and after the conference on the Gwalior Qurʾan. Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 139-150

140 • ASMA HILALI coraniques, cette analyse souligne les problèmes méthodologiques qui y sont liés. La transmission, telle qu’elle apparaît dans trois types de sources différents, est ici analysée :  tout d’abord dans les fragments coraniques les plus anciens, qui datent du viie siècle ; puis,  dans les récits sur la transmission des textes au début de l’Islam figurant dans les Ṭabaqāt ;  enfin, dans une source théorique du xe siècle établissant les bases de la transmission savante. S’appuyant sur une approche diachronique, cette étude repose tout à la fois sur l’analyse empirique du manuscrit et celle des sources biographiques et des écrits théoriques des viiie et xe siècles. Tenant compte des annotations et des glossaires omniprésents dans le coran de Gwalior et qui font écho à la dynamique existant entre transmission orale et transmission écrite du Coran au sein des cercles d’enseignement et d’étude, l’article propose de nouvelles perspectives pour l’histoire de la transmission des manuscrits coraniques.

introduction: the GwaLior qurʾan and contemPorary qurʾan manuscriPts studies Although the Gwalior Qurʾan contains a colophon that informs us about its date and place of completion (the Gwalior Fort on 17 Dhū al-Qaʿda 801/21 July 1399), numerous aspects remain mysterious in the manuscript such as the function of some annotations and marginal notes. There is almost no “unused space”1 in the Gwalior Qurʾan  and  it  is  difficult  to  assign  a  unique function to the totality of the textual and paratextual features in each folio. A quick glance at the Gwalior Qurʾan indicates that the manuscript contains an important amount of marginal notes, glossaries and a translation of the Qurʾanic text into Persian. I should like to highlight a few questions without thereby intending to express any critique of the article by Sabrina Alilouche and Ghazaleh Ismailpour Qouchani devoted to a deep analysis of the marginal notes and the fālnāma in the Gwalior Qurʾan. My main remarks concern the organization of the “extra-space” in the manuscript and more precisely in the margins. The Gwalior Qurʾan contains a multiplicity of layers of notes and corrections and their succession indicates the process of writing in the manuscript. The many layers of annotations show the many layers of production in the manuscript. The paratextual features and the historical context of production of the manuscript imply a variety of perspectives in the study of the Gwalior Qurʾan offered in this volume. My contribution proposes a number of methodological observations about contemporary studies of Qurʾan manuscripts and proposes an approach that takes into consideration the paradigmatic elements in the manuscripts related to the teaching context and to the written and oral transmission.

1.

Görke, Hirschler (dir.) (2011), 9.

The contemporary study of Qurʾan manuscripts is,  par  excellence,  a  field  of  varied  and  sometimes  isolated sub-fields. A quick glance at the website dedicated to Qurʾanic studies, ‘Coran et Science de l’Homme’ (Qurʾan and the humanities),2 illustrates the field’s clear separation between textual studies,  codicology, palaeography, epigraphy and art history. Most studies deal with Qurʾan manuscripts by considering them witnesses to the history of the written Qurʾan and they often compare the Qurʾanic manuscript evidence to information about the history of the Qurʾan disseminated in the Islamic sources of the Qurʾan sciences – in particular to the concept of Qurʾanic variants and readings (qirāʾa, pl. qirāʾāt).3 The identification of the concept of variants in the  textual issues in the early Qurʾan manuscripts is repeatedly raised by the scholars of Qurʾan as the major concern in the studies of Qurʾan manuscripts.4 Within this tendency that goes back to the beginning of the 20th century,5 connexions are made between the Islamic sources and the Qurʾan manuscripts. In particular evidence dated to the 7th and the 8th centuries are expected to evince the textual features reiterated in the qirāʾāt literature from the 8th to the 10th centuries.6 Nevertheless, the scholars who deal with the variants in Qurʾan manuscripts sometimes do not take into consideration the chronological gap between the early manuscript evidence (7th to 8th centuries) and the emergence of the concept (8th to 10th centuries).7 This remark does not intend to

2. 3.

http://www.mehdi-azaiez.org/?lang=fr. About the concept of Qurʾanic variants and readings, see Comerro (2012), 119-135 and Nasser (2012), 8-29. 4.  See for example, Puin, G. R. (1996), 107-111; Dutton (2004),  43-71; Fedeli (2005), 3-7; Powers (2009); Rabb (2006), 84-127;  Puin, E. (2007), 462. 5. Mingana, Lewis (1914). 6.  The  best  examples  are  Sadeghi,  Goudarzi  (2012), 116-121;  Sadeghi (2013), 21-41. 7. On the problem related to the application of the concept of Qurʾanic variants on the early manuscripts of Qurʾan, see Brokett (1990), 129-206, in particular 129.

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY • 141

cast doubt on the historicity of Qurʾanic variants but evokes the problem of the similarity between the “historical Qurʾanic variants” and the Qurʾanic variants transmitted, collected and progressively canonised by Muslim scholars between the 8th and 10th centuries.8 In the meanwhile, studies on the codicology of the Qurʾan manuscripts of various historical periods have occupied an important space in Qurʾanic studies over the last decade.9 Another important tendency that occupies a large part of Qurʾan manuscript studies consists of their examination as part of art history which extends the study of Qurʾan citations written on parchment or other material to the study of Qurʾanic fragments reproduced on fabrics and on buildings, as in the works of Sheila Blair, Éloïse Brac de la Perrière and Finbarr B. Flood.10 In addition, one has to consider the epigraphic discoveries of the Qurʾan graffiti by Frédéric Imbert.11 The Qurʾan inscriptions engraved on stones address textual features that show particular channels of transmission of the Qurʾan that Frédéric Imbert calls “Le Coran des Coeurs” (the Qurʾan of the hearts), expressing by this the particular knowledge of the Qurʾan text that the authors of the graffiti have strongly influenced  by their individualities12 meaning that the Qurʾan of the stones for the most part is different from the Standard Qurʾan.13 Within the different sub-fields in contemporary  Qurʾan manuscript studies, a number of assumptions occupy an important part of the research: a/ A separation of the analysis of the codicological aspects and the content of the manuscript into two disconnected entities; b/ The consideration of the Qurʾan book (muṣḥaf) as the form that encompasses the entire history of Qurʾan manuscript history; c/ The domination of the concept of Qurʾanic variants and readings used as a scaffold in the interpretation of the history of Qurʾan manuscripts. These three assumptions find their raison d’être in  one of the driving ideas in the field of Qurʾan manuscript studies – the quest for the origins of the Qurʾan. We witness book titles, conferences and articles that repeatedly include the expression “the Origins of the Qurʾan” and promise to provide an answer to the question as to what really happened with regard to the Muḥammadan revelation, the circumstances of  8. Comerro (2012), 137-158. 9.  See for example, Déroche (2009); Fedeli (website); Blair  (2008), 72-97 ; Georges (2011), 377-429. 10.  Blair  (2007),  271-284;  Brac  de  la  Perrière  (2009),  57-81;  Flood (2009), 91-118. 11. Imbert (2000), 381-390. 12. Ibid. 13. Ibid. The expression “Standard Qurʾan” refers to the popular version of the Qurʾan, most known to Muslims as “The Cairo edition”, published in Cairo in the 1925.

its being written down and collated into the Qurʾan.14 The haunting question about the origins of the Qurʾan has increased the interest of scholars in RC14 dating as a method of ascertaining an oldest date for a manuscript. The RC14 chemical analysis can provide a period of death for the animal whose skin was used for the parchment of the Qurʾan manuscript.15 RC14 results are sometimes used to arrive at conclusions about the content of a Qurʾan manuscript and its historical origin. It can be a method that influences  both the interpretation of the script and the philological analysis of the Qurʾanic text. The authors of  the  work  dedicated  to  manuscript  01.27-1  Dār  al-Makhṭūṭāt Ṣanʿāʾ, the so-called “Ṣanʿāʾ palimpsest”, posit that the manuscript was written 15 years after the death of the prophet Muḥammad.16 This conclusion allows the authors to identify the text as deriving from the codex of Ibn Masʿūd,  a  companion  of  the  prophet.17 In this respect, my critique does not address the identification of a number of Qurʾanic variants in the script but the over-interpretation of the text in a way that makes it correspond to the list of Qurʾanic variants known to us as part of the Ibn Masʿūd codex. The  identification  of  Qurʾanic variants in early Qurʾan manuscripts is moreover closely related to the representation of the history of the transmission of the Qurʾan as a book. Considering the history of Qurʾan manuscripts as mainly the history of the muṣḥaf results from studying the manuscripts within the parameters of the Arabic Islamic sources of the 8th to the 10th centuries, a period that coincides with the beginning of the Qurʾan sciences with their keyconcepts such as Qurʾanic variants and readings. The same perception gives prominence to the Qurʾan as a muṣḥaf and does not take into consideration the fragmentary form of its transmission. Most of the time, scholars of Qurʾan manuscripts do not dedicate much analysis to hypotheses regarding the fragmentary form of Qurʾan transmission.18 At the same time, the ambiguity surrounding the meaning of the term muṣḥaf is underlined by most scholars and draws attention to the polysemy of the same concept.19

14. See more recently, Déroche, Robin, Zink (dir.) (2015). 15.  Déroche (2014), 1314; on the example of the RC14 applied  on the Tübingen manuscript, see: http://www.uni-tuebingen.de/en/news/press-releases/newsfullview-pressemitteilungen/article/raritaet-entdeckt-koranhandschrift -stammt-aus-der-fruehzeit-des-islam.html; on the recent example of Birmingham manuscript, see: http://www. birmingham.ac.uk/news/latest/2015/07/quran-manuscript-22-07-15.aspx. 16. Sadeghi, Goudarzi (2012), 8-9. For a substantial critique of Sadeghi and Goudarzi’s work, see Hilali (2014), 12-27. 17. Sadeghi, Goudarzi (2012), 15-31. 18. On one of the exceptions, see Fedeli (2008), 25-48. 19. Déroche (2003), 254-275.

142 • ASMA HILALI

Muṣḥaf and its synonyms such as ṣaḥīfa (pl. ṣuḥuf) refer to a book as well as to disparate leaves.20 In the same vein, the term jamʿ (assembling) presupposes the disparate state of the leaves and suggests the circulation of the Qurʾan text in fragmentary way.21 In the current state of the field, a number of questions present themselves, including what are the fragmented written forms in which the Qurʾan was transmitted before and after the constitution of the manuscripts that are referred to as muṣḥafs? How can one identify the philological aspects that allow us to consider specific Qurʾan manuscripts as fragments rather than as part of a book? How should one conceptually approach narratives about the transmission of the Qurʾan in its fragmented form? And lastly, what are the manuscript features that refer to both the transmission and commentary of Qurʾan fragments?

the circLe of transmission in ibn saʿd’s Ṭabaqāt (8th century) I propose an overview of the narratives dated to the 8th century regarding the transmission of the Qurʾan. My study is based on a philological examination and a consideration of these narratives as reflecting a representation of the circle of transmission that has to be studied as such. The generally accepted image of teaching the Qurʾan in the two first  centuries  of  Islam  depicts  the  Qurʾan as being transmitted in teaching circles dedicated to various other subjects such as poetry, military narratives, etc.22 The recitation of the Qurʾan occurs in the middle of the transmission of various genres of material, and in this respect, Ibn Saʿd reports: “When the Companions of the Prophet gathered for the transmission of ḥadīth, their transmission (ḥadīthuhum) was about fiqh unless they ordered someone to recite in their presence (yaqraʾu ʿalayhim) a chapter of the Qurʾan or someone recites a chapter of the Qurʾan (sūratan mina al-Qurʾani) on his own initiative”.23

Following this representation of the circle of transmission, the sharing of the tasks of narration and recitation was structured in a way that Qurʾan recitation was often described as emerging from an order given by the Companions to one of the

20. Al-Azmeh (2014), 432-437. 21. Ibid. 22. On the heterogeneity of the materials transmitted in the first centuries, see Ibn Saʿd [2001], vol. I, 334-354; Hind  (1996), 188-198; Zaman (1996), 1-18, esp. 14. See an attempt  to a concrete reflexion on the heterogeneity of this material in Hilali (2012), 29-44. 23. Ibn Saʿd [1968], 374.

participants to recite fragments of the Qurʾan. The recitation of the Qurʾan could also result from the initiative of one of the participants to the circle. Qurʾan recitation is portrayed as a device that punctuated a session of transmission dedicated to various materials. The narrative underlines the recitation of the Qurʾan as a habit in the circle and refers to it as an improvisation. It emerges at the same time as an exception and as a permanent element. The narrative builds the picture of a circle dedicated fundamentally to fiqh wherein the recitation of the Qurʾan is introduced as an exception. The status of Qurʾan recitation in the circles of transmission as recorded in Ibn Saʿd’s Kitāb al-Ṭabaqāt has allowed me, in a previous study, to conclude that there is confusion between the genres of ḥadīth, Qurʾan, poetry and other material in the early teaching circles.24 This conclusion is founded on the observation of contradictory accounts concerning the content of the circles of transmission. In parallel with the previous account showing the recitation of the Qurʾan during a fiqh session, Ibn Saʿd repeatedly highlights the  separation  between  the  different  materials;  he quotes a number of narratives showing that the transmitters dedicated different days to different material.25 The maghāzī (accounts of the Prophet’s raids) are reported to have been taught on specific  days and fiqh is described as the topic of independent teaching sessions. The image of the early circles that emerges from these narratives is one where the differentiation between the various material is not as sharp as the compiler attempts to show. The Qurʾan and other texts may have been transmitted in the same sessions in an indistinct manner and some ‘in-between’ genres may have emerged from this lack of differentiation.26 The recitation of the Qurʾan in the early teaching circles was well known to the compiler of the 8th century and should be taken into consideration as one of the techniques of the oral text transmission including the Qurʾan. The impact of the oral transmission sessions has to be taken into consideration in both the examination of the narratives about the Qurʾan collection and the manuscript evidence. The Gwalior Qurʾan contains multiple literary genres (commentary, translation, fālnāma, etc.). It contains marginal annotations and corrections as well. The variety of texts written in the margins and between the lines of the Qurʾanic text in the manuscript express different approaches and refer to different contexts of Qurʾan study. Contextualizing

24. Hilali (2012), 29-44. 25. Ibid. 26. Ibid.

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY • 143

the different approaches to the text should take into consideration the scriptural differences between the various literary genres superimposed in the manuscript as well as the historical context of the production of the manuscript. For example, as I will show in the next section, the various techniques of marginal corrections show different conceptions of the error and its importance in the writing process;  they show also different conceptions of the manuscript and inform us about the different contexts of its production. The following paragraph reflects on the theoretical  framework of the dictation session in the 10th century, a century of the science of transmission, and represents an attempt to demonstrate the importance of the theory of transmission and the technicalities of corrections and emendations in highlighting the context of transmission of the manuscript.

the theory of transmission in aL-rāmhurmuzī’s Dirāya (10th century) The theory of transmission expresses an attempt to organize the history of transmission and the technicalities of the teaching circles in a systematic way. As I will show, every detail of the session of writing and reciting is submitted to a theoretical rule that establishes the best way to record the text. ʿAbd alRaḥmān al-Rāmhurmuzī (d. 360/971) was the author of the first systematic book on the science of transmission, al-Muḥaddith al-fāṣil bayn al-rāwī wa al-wāʿī (the transmitter who differentiates between the transmitter and the attentive receiver).27 In his book, he established methods of elaborating a school book and writing down transmitted material. He focuses on the technicalities of organizing the space in the manuscript, particularly, the techniques of correction and writing in the margins. The following example investigates the two techniques of 1/ correction, and 2/ writing in the margins. At the end of this chapter, I will refer to  specific  passages  of  the  manuscript  of  Gwalior  and  underlines  a  few  aspects  that  reflect  a  precise  concept of correction at the origin of the production of the manuscript.

Correction

(ḥakk) and crossing out (ḍarb). Al-Rāmhurmuzī gives  the following tradition: “Our companions said: ‘Erasing may be likened to suspicion and, so, the most skillful crossing-out is that which does not make the crossed-out [text] disappear, but that which marks the new text above a straight and clear line, thus showing the cancellation of its effect (ibṭāl). [Thus] one must read the text beneath it that replaces it.”28

This technique of correction is drawn from the circle of transmission and the common issues that occur to the scribe during the session of teaching that includes here the technique of dictation. The example under study shows the importance given by the author to writing as a process made of multiple layers of corrections. Any act of writing that is not in conformity with the rules is revisited. According to  al-Rāmhurmuzī,  writing  denotes  an  action  that,  after a long and complex process, recovers and preserves the texts from inevitable deterioration. Nevertheless, he prefers crossing out the errors rather than erasing them. By erasing part of the text, the accumulation of the multiple layers of corrections would be hidden. In a way, an error and its clear correction preserve the different layers of writing  of  the  text;  and  erasing  an  error  denies  the value of the same error. The error, thus, is part of the process of transmission and should be preserved as such. An error can also be a way of preserving knowledge by showing which fragment of the text has been readjusted or excluded. Erasing it completely makes the quest to establish the process of correction impossible, because  the  text  presents  only  the  final  result;  whereas,  for  al-Rāmhurmurzī,  the  important  thing  is the path that has been traveled in the process of multiple corrections and errors are integral to that process. Preserving the error helps to distinguish it from the corrected text and makes it possible to continue verifying the transmission. Verification and  correction define writing as a form of searching and  refer to the didactic context of transmission circles.

Writing in the margins

Correcting errors made during the process of copying the text from a written copy or in dictation session reestablishes a text that is not correct but corrected. The scribe can choose between two methods of cancelling text deemed erroneous: erasing

Another technique for correcting mistakes mentioned by the theoretician, and one that we can observe in the Gwalior Qurʾan, implies that there should be empty spaces in transmitted texts. The omission of words is at the origin of a technique that the author presents under the following title: al-takhrīj ʿalā al-ḥawāshī (noting in the margins).29

27.  Librande (1978), 267-280; Hilali (2006), 131-147.

28.  Al-Rāmhurmuzī [1971], 606. 29. Ibid., 606.

144 • ASMA HILALI

The omission of certain words is specific to the process  of oral transmission during which the student may often mishear a word during his master’s recitation. Far from being an error that would condemn the totality of the text, an omission is recoverable, writes al-Rāmhurmuzī, thanks to the possibility of using the  margin. The omitted word is written in the margin and linked to the passage from which it was left out by a line. The margin is the empty space on the page that exists both to protect the integrity of the text and to frame the omission in a separate space. Once again, the laws of writing are intended to facilitate the techniques of text verification and take into consideration the context of dictation in the circles of transmission. There is a complementariness between the laws of writing and the arrangement on the page that leaves room for correction, which is to say, the complete text is always the object of correction. Writing includes continuously correction and verification  that alternate throughout the process of transmission. Among numerous examples of marginal corrections in the Gwalior Qurʾan, I would like to mention two examples.30 The first one concerns a specific passage that joins a marginal note and a crossing-out technique in folio 117v, line 3 (Q. 5: 86) (figure 1). The marginal note seems to be simply the reproduction of the expression “wa al-laḏīna kafarū”. As the same expression is not missing in the Qurʾanic passage in the same folio, and nor does it contain any error in comparison with the marginal note, the marginal note seems useless. I would like to propose an interpretation of the passage based on the phenomenon of blank spaces in some ancient Arabic manuscripts where the blank space is inserted for words misheard during the dictation sessions. My hypothesis is that this first example shows this specific technique. The  expression “wa al-laḏīna kafarū” was most probably missed at some stage in the process of copying the text. The marginal note functions as the addition of the missing words in the process of filling in the  blank spaces in the manuscript. Subsequently to the filling in of the blank space, the marginal note has  no function anymore which explains the reason for its crossing out: the correction becomes in a later stage of production of the manuscript, an error. However, the technique of crossing out the error does not constitute a unique correction technique in the manuscript. My second example is related to  folio  44v,  line 6  (figure 2).  The  verb  “yaʿidukum”

30. On more developed analysis of the marginal notes in the Gwalior Qurʾan, see the contribution of Sabrina Alilouche and Ghazaleh Ismailpour Qouchani in the current volume.

(Q. 2: 268) is corrected in the margin. But the error remains unknown to the reader as the correction consists of copying and pasting a piece of paper on the space dedicated to the last two letters of the verb “yaʿidukum”. In constrast to the first example of correction based on the technique of crossing out, the second example shows another conception of error as well as of correction. The different conceptions of corrections imply in their turn different contexts of production in the manuscript. The two passages are addressed to two different readers. In the first example, the reader is involved in the process of writing and the error is maintained as such by the crossing out technique in order to keep the reader informed about the writing process. However, in the second example, the reader is excluded from the process of production because the correction technique is hidden by the mean of a copy/pasting technique. The first  example belongs to a stage of work in progress of the manuscript and involves a reader who is part of the production of the manuscript, in other words: the reader in the first example is the scribe himself.  The second example denotes a final form of the text or  a published form addressed to an unknown reader. The two categories of source material, narratives extracted from Ṭabaqāt Ibn Saʿd about the teaching circles of the 8th century and the theoretical writing of the 10th century, provide a comparison between history of transmission and theory of transmission. This comparison between theory and practice aims to bring some understanding to how textual transmission is portrayed in the two different sources in a way that takes into account the dynamic of the teaching circle. As Qurʾan fragments are part of the texts transmitted in teaching circles, they have been submitted to the same technicalities of transmission proper to the teaching circles. Do the theoretical writings of the 10th century  reflect  the  practice  of  teaching and transmitting text material in an earlier period? The answer to this demands a comparison between the theoretical setting and the early manuscripts themselves. In the last part of this contribution, I propose the examination of the Qurʾan manuscript as a manuscript submitted to the dynamic of the transmission circles and I endeavour to analyse the philological features that allow the consideration of the context of transmission in specific manuscripts,  in this case the lower text of manuscript 01-27-1, Ṣanʿāʾ Yemen.

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY • 145

Figure 1 – Gwalior Qurʾan, fol. 117v. [© By courtesy of the Aga Khan Museum]

146 • ASMA HILALI

Figure 2 – Gwalior Qurʾan, fol. 44v. [© By courtesy of the Aga Khan Museum]

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY • 147

teachinG circLes and manuscriPt evidence: the exampLe of the Ṣanʿāʾ PaLimPsest31 The last part of this contribution will be devoted to an example of an early manuscript: 01.27-1 DAM (figures 3a-b), Ṣanʿāʾ, the so-called “Ṣanʿāʾpalimpsest” that serves as an example of a Qurʾan manuscript bearing witness to the interaction between oral and written  transmission,  as  well  as  reflecting  by  this  the dynamic of the circle of transmission.32 Qurʾan manuscripts sometimes contain notes inserted, more or less intrusively, in the “unused” space, which indicates the way the manuscript is being used in the specific context of recitation, writing, commentary, translation and ritual usage. What follows is a preliminary discussion of the methods of transmission of the Qurʾan based on the Ṣanʿāʾ palimpsest dated to the 7th century.33 The transmission techniques are reflected in manuscripts in the shape of the script and  its various corrections, as well as in the organization of the space including the marginal notes, translations and glossaries. A manuscript with notes reveals a

process of interpretation as it is the case with the Gwalior Qurʾan. In the example under study in this contribution, the Ṣanʿāʾ palimpsest, the marginal note is individual inasmuch as it consists of a sentence written in the body of the text but having marginal status. Manuscript  01.27-1  Dār  al-Makhṭūṭāt  Ṣanʿāʾ is made of parchment with two layers of Qurʾan text: a lower layer, erased by the middle of the 7th century, over which an upper layer of text has been written;  the lower text is dated to the 7th century and the upper text to the 8th.34 The lower text has attracted significant attention as it contains differences from  the canonical version of Qurʾan.35 In previous published works, I have analysed specific passage in the  lower text found on folio 5r.36 The passage is situated between two Qurʾan chapters, the end of sura 8 (al-Anfāl) and the beginning of sura 9 (al-Tawba). This passage is interesting since the text includes the commentary or, more precisely, the instruction “lā taqul bism Allāh” (do not say bism Allāh) just after the formula of the basmala. From this inclusion of the basmala and the instruction not to say it before sura 9,

Figure 3a – Manuscript 01.27-1 Dār al-Makhṭūṭāt Ṣanʿāʾ, Yemen. Fol. 5r [first words of the lines 7, 8, 9].

Figure 3b – Manuscript 01.27-1 Dār al-Makhṭūṭāt Ṣanʿāʾ, Yemen. Fol. 5r [reconstruction of the lines 7, 8, 9].

31. A number of studies have been devoted to the manuscript 01.27-1 DAM, Ṣanʿāʾ since the 80s. Puin, G. R. (1996), 107111; Puin, E. (2007), 461-493; id. (2009), 523-681; id. (2010),  233-305; id. (2012), 311-402; id. (2014), 477-618; Hilali (2010),  443-448; Sadeghi, Goudarzi (2012), 1-129; Gurtman (website). 32. Hilali (2010), 443-448; id. (2012), 29-44; id. (2014), 24. 33. About the dating of the manuscript, see Sadeghi, Goudarzi (2010), 8-9; Déroche (2014), 13-14.

34.  On the lower text, see note 33; on the dating of the upper  text, see Puin, G. R. (1996), 107-111. 35. On studies on the manuscript and the role of the media, see Sadeghi, Goudarzi (2012), 31-36. 36.  Hilali (2010), 443-448 ; id. (2012), 29-44; id. (2014), 24; Puin,  E. (2009), 647-648; Sadeghi, Goudarzi (2012), 53 n. 157.

148 • ASMA HILALI

I have concluded that the reading instruction was most probably dictated by the teacher, indicating that at the beginning of sura 9, the recitation of the basmala is not required. The relevance or irrelevance of pronouncing the basmala at the beginning of this chapter is part of a discussion among the medieval scholars pertaining to the genre of sura 9 and its place in between the Qurʾan and invocation.37 The graphic above represents the reconstruction of the passage, where we can read at the beginning of three lines the end of sura 8, the basmala and the injunction not to pronounce the basmala.38 The injunction is not part of the standard Qurʾanic text, however, it is written in the space dedicated to sura 9. Its didactic dimension as a remark pertaining to the recitation of chapter 9 gives the injunction a marginal character vis-à-vis the core of the text. In this sense, it consists of a marginal note, though written in the text. The basmala followed by the injunction not to pronounce it could be considered a correction of the Qurʾanic reading or writing or both. This passage reveals the context of teaching the rules of reciting the Qurʾan. When we read this note we are in the presence of a manuscript note that refers to its oral recitation. However, it is important to emphasize that the basmala in itself at the beginning of sura 9 has not been erased, as might be expected from a dictation or copying session that was taking rules of reciting the Qurʾan into consideration. If we apply  al-Rāmhurmuzī’s  remark  cited  above  to  this  passage, it is noticeable that the error is not erased, but maintained. The writing down of the formula, and the commentary providing the instruction not to pronounce it that comes just after it, indicates the context of copying an original passage with the addition of new element, i.e. the injunction not to pronounce the basmala at the beginning of sura 9. Thus, the teaching context is delimited by the exercise of commenting on a passage of Qurʾan that may have being copied during the session and the instruction is introduced in order to correct and update it. I would like to emphasize that the correction shows that the teaching and copying the Qurʾan progresses in the light of the introduction of concurrent rules of reading and writing. In other words, the rule of pronouncing the basmala precedes chronologically the rule that the basmala should not be recited in this specific place i.e. the beginning of

37.  Al-Ṣuyūṭī [2003], vol. 1, 115. 38. Manuscript  01.27-1  Dār  al-Makhṭūṭāt  Ṣanʿāʾ, Yemen. Folio 5r, reconstruction of folio 5r. lines 7, 8, 9. Thanks to Russell Harris for preparing the graphic reconstructions of the scripts. See the image of this passage of the manuscript in: Hilali (2010), 445.

sura 9, and it is submitted in this didactic passage as an ‘update’ incorporating a newer practice. Another interpretation of this passage would be that the rule of not pronouncing the basmala at this specific passage of the Qurʾan was unknown to the scribe who wrote the basmala as well as to the reciter. The formula has been probably automatically pronounced by the reciter at the beginning of sura 9, then, the scribe corrected the error by writing down the didactic note. The second conclusion I would like to add to my former study of this passage is that the blank spaces that I noted in the lower text of the same manuscript may refer to a dictation session and could be empty spaces the scribe prepares in a way that allows him/her to fill them in with the words  he/she missed during the session in a later stage of the writing. The teaching circle is based on the copying of Qurʾan passages and is likely to be a dictation exercise. The recitation triggers specific questions and  the correction of the text is similar to the updating and rewriting of the text.

concLusion In their different forms and through a long period of time, the notes in a manuscript as seen in the two examples of Qurʾan manuscripts have the common aim of controlling the form of the Qurʾanic text and its interpretation. Despite the fact that they are connected to a complete book, the notes inserted in different ways in the manuscript of the Gwalior Qurʾan disconnect the pieces of the text from each other by the means of the marginal notes, the corrections and the additions that generate a superimposition of different texts that seem to belong to different historical periods. The two examples of Qurʾan manuscripts, the Ṣanʿāʾ palimpsest and the Gwalior Qurʾan, shed some light on the historical shift in teaching techniques from elementary manuscript notes such as reading instructions in the manuscript dated to the 7th century to the extensive use of the extra-space and the sophisticated organization of the layout including different colours in the manuscript of the 14th century. The common thread in the evolution of the manuscripts’ marginal corrections and annotations seems to be the accumulation of layers of texts that brings us back to the principle set by al-Rāmahurmuzī that  the error leads the way to a ‘new’ text. If we consider the palimpsesting technique as a metaphor for the ultimate correction, we may say that the history of the Qurʾan manuscripts is made up of an accumulation of palimpsests, and thus the Qurʾan manuscript practice is built on an infinite ‘discussion’ between  correctors and readers that expresses a perpetual updating of the understanding of the text in its literary form as well as its meaning.

QURʾAN MANUSCRIPTS AND THEIR TRANSMISSION HISTORY • 149

bibLioGraPhy Sources iBn saʿd (Muḥammad) [1968] Kitāb al-Ṭabaqāt al-kabīr, edited by Iḥsān ʿAbbās,  Beirut: Dār Ṣādir. [2001] Kitāb al-Ṭabaqāt al-Kabīr, edited by ʿAlī Muḥammad ʿUmar, Cairo: Maktaba Khānjī, 11 vols. al-rāMhurMuzī (ʿAbd al-Raḥmān b. Khallād) [1971] Al-Muḥaddiṯ al-fāṣil bayna al-rāwī wa al-wāʿī, edited by Muḥammad ʿAjjaj al-Khaṭīb, Beirut:  Dār al-Fikr. al‑ṣuyūṭī (Jalāl al-Dīn) [2003] Al-itqān fī ʿulūm al-Qurʾan, Beirut: Dār al-Fikr.

References al‑azMeh (Aziz) 2014 The Emergence of Islam In Late Antiquity. Allah and His People, Cambridge: Cambridge University Press. Blair (Sheila) 2007 “Written, spoken, envisioned: the many facets of the Qurʾan in art”, in Fahmida Suleman (dir.), The Qurʾan in Art, London: Oxford University Press, pp. 271-284. 2008 “Transcribing God’s word: Qurʾan codices in context”, in Journal of Qurʾanic Studies 10, pp. 72-97. Brac de la Perrière (Éloïse) 2009 “Les tuniques talismaniques indiennes d’époque pré-moghole et moghole à la lumière d’un groupe de corans en écriture bihāri”, in Journal Asiatique 297, no. 1, pp. 57-81. Brokett (Adrian) 1990 “Qurʾan readings in kitāb Sībawayhi”, in Occasional papers of the school of Abbasid studies, University of St. Andrews, no. 2 1988 (publ. 1990), pp. 129-206. comerro (Viviane) 2012 Les traditions sur la constitution du muṣḥaf de ʿUthmān (Beiruter Texte und Studien, 134), Beirut/Würzburg: Orient-Institut/Erlon Verlag. dÉroche (François) 2003 “Manuscripts of the Qurʾan”, in Jane Dammen McAuliffe (dir.), Encyclopaedia of the Qurʾan III, Leiden/Boston: Brill, pp. 254-275. 2009 La transmission écrite du coran dans les débuts de l’islam. Le codex parisino-petropolitanus, Leiden/ Boston: Brill.

2014 Qurʾan of the Umayyads (Leiden Studies in Islam & Society), Leiden: Brill. dÉroche (François), roBin (Christian), zInk (Michel) (dir.) 2015 Les origines du Coran, le Coran des origines, Paris: De Boccard. dutton (Yasin) 2004 “Some notes on the British Library’s ‘oldest Qurʾan manuscript’ (Or. 2165)”, in Journal of Qurʾanic Studies VI, no. 1, pp. 43-71. Fedeli (Alba) 2005 “Mingana and the manuscript of Mrs Agnes Smith Lewis, one century later”, in Manuscripta Orientalia 11, no. 3, pp. 3-7. 2008 “I manoscritti di Sanaa: fogli sparsi che diventano Corani”, in F. Aspesi et al. (dir.), Il mio ciore é Oriente, pp. 25-48. Flood (F. Barry) 2009 “Islamic identities and islamic art: inscribing the Qurʾan in twelfth-century Afghanistan”, in Elizabeth Cropper (dir.), Dialogues in Art History, from Mesopotamian to Modern: Readings for a New Century (Studies in the History of Art Series), Washington: National Gallery of Art, pp. 91-118. georges (Alain) 2011 “Le palimpseste Lewis-Mingana de Cambridge, témoin ancien de l’histoire du Coran”, in Comptes rendus de l’Académie des inscriptions & Belles-lettres, no. 1, pp. 377-429. görke (Andreas), hirschler (Konrad) (dir.) 2011 Manuscript notes as documentary sources (Beiruter Texte und Studien, 129), Beirut/Würzburg in  Kommission: Orient-Institut/Ergon Verlag. hilali (Asma) 2006 “Abd al-Raḥmān al-Rāmahurmuzī (m. 360/971) à  l’origine de la réflexion sur l’authenticité du hadīṯ. (Abd al-Raḥmān al-Rāmahurmuzī (d. 360/971)  at the origin of reflection on Hadith’s authenticity)”, in Annales Islamologiques 39, pp. 131-147. 2010  “Le  palimpseste  de  Ṣanʿāʾ et la canonisation du Coran : Nouveaux éléments”, in Cahiers du Centre Gustave Glotz 21, pp. 443-448. 2012 “Coran, hadith et textes intermédiaires: Le genre religieux aux débuts de l’Islam”, in Mélanges de l’Université Saint-Joseph 64, pp. 29-44. 2014  “Was  the  Ṣanʿāʾ Qurʾan palimpsest a work in progress?”, in David Hollenberg, Christoph Rauch, Sabine Schmidke (dir.), The Yemeni Manuscript Tradition, Leiden: Brill.

150 • ASMA HILALI

hind (Martin) 1996 Studies in Early Islamic History (Studies in Late Antiquity and Early Islam, 4), Princeton (NJ): Darwin Press. imBert (Frédéric) 2000  “Le Coran dans les graffiti des deux premiers  siècles de l’hégire”, in Arabica 47, nos. 3-4, pp. 381-390. liBrande (Leonard) 1978  “The categories high and low as reflections on  the Riḥlah and Kitābah in Islām”, in Der Islam 55, no. 2, pp. 267-280. mingana (Alphonse), lewis (Agnes Smith) 1914 Leaves from three ancient Qurʾans possibly pre-ʿOthmānic with a list of their Variants, Cambridge: Cambridge University Press. nasser (Shady H.) 2012 The transmission of the Variant Readings of the Qurʾan. The Problem of Tawātur and the Emergence of Shawādhdh, Leiden: Brill. Powers (David S.) 2009 Muḥammad is not the father of Any of Your Men. The Making of the Last prophet, Philadelphia: University of Pennsylvania Press. Puin (Elisabeth) 2007  “Ein früher Koranpalimpsest aus Ṣanʿāʾ (DAM 01-27.1)”, in Markus Gross, Karl-Heinz Ohlig (dir.), Schlaglichter. Die beiden ersten islamischen Jahrhunderte, Inārah  (Schriften  zur  frühen  Islamgeschichte und zum Koran, 3), Berlin: [Verlaƍ Hans Schiler], pp. 461-493. 2009  “Ein früher Koranpalimpsest aus Sanʿa (DAM 01-27.1): Teil II”, in Markus Groß and Karl-Heinz Ohlig (dir.), Vom Koran zum Islam, Berlin: [Verlaƍ Hans Schiler], pp. 523-681. 2010  “Ein früher Koranpalimpsest aus Ṣanʿā’ (DAM  01-27.1): Teil III. Ein nicht-ʿutmanischer Koran”, in Markus Groß and Karl-Heinz Ohlig (dir.), Die Entstehung einer Weltreligion I: Von der koranischen Bewegung zum Frühislam, Berlin: [Verlaƍ Hans Schiler], pp. 233-305. 2012  “Ein  früher  Koranpalimpsest  aus  Ṣanʿāʾ II (DAM 01-27.1): Teil IV. Die scriptio inferior auf den Blättern 17, 18 und 19 der Handschrift DAM 01-27.1 (sure 9:106-Ende, dann 19:1-67 und weiter)”, in Die Entstehung einer Weltreligion II: Von der koranischen Bewegung zum Frühislam, Berlin: [Verlaƍ Hans Schiler], pp. 311-402. 2014  “Eine  Früher  Koranpalimpsest  aus  Sanaa  (DAM 01.27-1) Teil V. Die scriptio inferior auf

den blätten 14 und 15 sowie. Auseinandersetzung mit den Thesen und der Edition des Koranpalimpsest von Behnam Sadeghi und Mohcen Goudarzi”, in Markus Groß and Karl-Heinz Ohlig (dir.), Die Entstehung einer Weltreligion III: Die heilige Stadt Mekka-eine literarische Fiktion, Berlin: [Verlaƍ Hans Schiler], pp. 477-618. Puin (Gerd R.) 1996 “Observations on early Qurʾan manuscripts in Sanʿaʾ”, in Stefan Wild (dir.) The Qurʾan as text, Leiden: Brill, pp. 107-111. raBB (Intisar) 2006 “Non-canonical readings of the Qurʾan: Recognition & authenticity”, in Journal of Qurʾanic Studies 8, pp. 84-127. sadeghi (Behnam) 2013 “Criteria for emending the text of the Qurʾan”, in Michael Cook et al. (dir.), Law and Tradition in Classical Islamic Thought, New York: Palgrave Macmillan, pp. 21-41. sadeghi (Behnam), goudarzI (Mohcen) 2012 “Sanʿaʾ I and the origins of the Qurʾan”, in Der Islam 87, pp. 11-29. zaMan (Muhammad Q.) 1996 “Maghāzī and the Muḥaddithūn: reconsidering the treatment of “Historical” materials in early collections of hadith”, in International Journal of Middle East Studies 28, no. 1, pp. 1-18.

Websites Fedeli (Alba), The Qurʾanic Manuscripts of the Mingana Collection and their Electronic Edition. International Qurʾanic Studies Association website. Electronic resource posted on 18 March (2013): http://iqsaweb. wordpress.com/2013/03/18/qmmc. gurtman (Hadiya), “A Qurʾan written over the Qurʾan Why making the effort?” in http://www.manu-scriptcultures.uni-hamburg.de/mom/2012_01_mom_e.html (last consultation 18/09/2015). http://www.birmingham.ac.uk/news/latest/2015/ 07/quran-manuscript-22-07-15.aspx. http://www.manuscript-cultures.uni-hamburg.de/ mom/2012_01_mom_e.html. http://www.mehdi-azaiez.org/?lang=fr. http://www.uni-tuebingen.de/en/news/pressreleases/newsfullview-pressemitteilungen/article/ raritaet-entdeckt-koranhandschrift-stammt-aus-derfruehzeit-des-islam.html.

III

LES MANUSCRITS INDO-PERSANS ET LE MONDE ISLAMIQUE INDO-PERSIAN MANUSCRIPTS AND THE ISLAMIC WORLD

ECLECTICISM AND REGIONALISM: The Gwalior Qurʾan and the Ghurid Legacy to Post-Mongol Art Finbarr Barry Flood

(Institute of Fine Arts & Department of Art History, New York University)

Abstract With its exuberant illuminations, the Gwalior Qurʾan, the first dated north Indian Qurʾan, is paradigmatic of the ‘global’ nature of the art of the Eastern Islamic world following the collapse of the Pax Mongolica around the middle of the 14th century. The phenomenon is especially well documented in architecture; in the north Indian milieu in which the Gwalior  Qurʾan was produced, the decoration of the monuments erected by the Tughluq sultans of Delhi during the course of the 14th century offers numerous points of comparison for the manuscript’s decorative eclecticism. Such tangible reminders of horizontal cultural flows  between India, Iran and regions to the West should not, however, obscure the existence of more enduring and geographically circumscribed modes of artistic production; to paraphrase  the anthropologist James Clifford, we need to consider both routes and roots. This paper locates the Gwalior Qurʾan at the intersection of contemporary ‘horizontal’ transregional circulations  and  the  ‘vertical’  axes  of  earlier,  rooted  and  regionally  specific  traditions  of  manuscript production. Based on certain structural idiosyncrasies of the Gwalior manuscript, it raises the possibility of continuities with Qurʾan manuscripts (maṣāḥif) produced two centuries earlier in the Ghurid sultanate of Afghanistan and north India (ca. 1150-1210). It speculates that the artistic patronage of the Kartid rulers of Herat (ca. 1245-1389) may provide a ‘missing link’ between the Gwalior Qurʾan and earlier Ghurid maṣāḥif.

Résumé Éclectisme et régionalisme : du coran de Gwalior et de l’héritage ghuride à l’art post-mongol Avec ses enluminures exubérantes, le manuscrit de Gwalior, premier coran daté en provenance de l’Inde du nord, est représentatif du caractère « global » des arts islamiques en contexte oriental suite à l’effondrement de la Pax Mongolica au milieu du xive siècle. Ce phénomène est particulièrement bien documenté dans l’architecture. Dans le nord de l’Inde où le coran de Gwalior a vu le jour, la décoration des monuments érigés par les sultans  tughluqs de Delhi au cours du xive siècle offre de nombreux points de comparaison avec les décors éclectiques du manuscrit. Toutefois ces rappels tangibles des flux culturels horizontaux qui existaient entre l’Inde, l’Iran et des régions situées plus à l’ouest ne devraient pas faire oublier l’existence d’autres modes de production artistique, plus durables et plus restreints géographiquement. Comme l’a écrit l’anthropologue James Clifford, les routes, mais aussi les racines, doivent être prises en compte. Cet article inscrit le coran de Gwalior

Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 153-169

154 • FINBARR BARRY FLOOD au croisement des circulations « horizontales », contemporaines et transrégionales, et des axes « verticaux » de traditions artistiques plus anciennes, spécifiques à différentes régions.  En s’appuyant sur certaines particularités structurelles du manuscrit de Gwalior, l’auteur propose d’y voir une possible continuité entre cette œuvre et des manuscrits coraniques (maṣāḥif) produits deux siècles plus tôt dans le sultanat ghuride, en Afghanistan et en Inde du Nord (ca. 1150-1210). Il soulève la possibilité que le mécénat des chefs karts de Hérat (ca. 1245-1389) puisse fournir le « chaînon manquant » entre le coran de Gwalior et ces  maṣāḥif ghurides plus anciens.

With its eclectic and exuberant illuminations, the Gwalior Qurʾan is in many respects paradigmatic of the ‘global’ nature of the art of the Eastern Islamic world in the 14th century. This quality is most obviously manifest in its juxtaposition of forms and motifs deeply rooted in regional (primarily Indic) traditions with those of non-indigenous origin that circulated widely across the Islamic world in the wake of the world empire of the Mongols, which had fostered the development of the late thirteenth- and early fourteenth-century world system so well documented by Janet Abu-Lughod.1 In attempting to sketch a broad cultural context for understanding the remarkable variety of the illuminations in the Gwalior Qurʾan, it is also important to acknowledge the paradoxical fact that, however successful the Pax Mongolica may have been in reconfiguring established  artistic and cultural geographies, the collapse of the Mongol system, and the decline of the Ilkhans of Iran around the 1330s in particular, may have provided a far greater stimulus than Ilkhanid patronage to the ‘global’ dissemination of certain artistic forms, practices and techniques over an arc ranging from Egypt in the West to India in the East. This is manifest on the one hand in the rise of regional courts and patterns of artistic patronage in the power vacuum that resulted from the collapse of the Ilkhanid state, whether the Injuids (1305-1357) or Muzaffarids of Shiraz (1335-1393), the Jalayirids of Iraq (1335-1442) or, ultimately, the Ottomans of Anatolia (1299 onwards). On the other, we have the reception of Ilkhanid forms even in areas that had never been under the political control of the Ilkhans. In the West, perhaps the most spectacular example of this phenomenon is the massive funerary complex built by the Mamluk sultan Ḥasan in Cairo  and completed in 1363. The entrance to the complex was originally planned with twin minarets probably modelled on Ilkhanid prototypes, as were those in the mosque of Amīr Qawṣūn (now destroyed) built  in Cairo three decades earlier by an architect from Tabriz,  probably  in  imitation  of  the  mosque  of  Tāj 

1.

Abu-Lughod (1991).

al-Dīn ʿAlī Shāh (ca. 1310) that stood in the Iranian  city. During the same period, a tile workshop from Tabriz was operating in Cairo, introducing Persianate modes of decoration to the monuments of the Mamluk elite, while Indian textiles may have provided sources of inspiration for the decoration of at least one contemporary Cairene mosque.2 Signifiers of both time and space were deployed in  the forms and ornament of Sultan Ḥasan’s complex,  in which the massive scale of the qibla iwān is said to have competed with the fabled iwān of the Sasanians at Ctesiphon in Iraq, one of a number of fourteenthcentury mosques from Egypt to India that competed with the same ancient model: others include the mosque of Tāj al-Dīn ʿAlī Shāh in Tabriz and the Adina  Mosque at Pandua in Bengal, built just a decade or so  after  Sultan  Ḥasan’s  mosque  was  completed.3 Eclecticism added to the ‘global’ filiations of sultan  Ḥasan’s complex, with an entrance porch in which  Crusader spolia depicting the sacred sites of Jerusalem were combined with newly made carvings of Chinese inspiration, which recur in the funerary chamber of the complex (figure 1).4 Both instances of chinoiserie were undoubtedly mediated by contact with Ilkhanid art, by virtue of which a wave of Sinicizing ornament broke over the architectural and minor arts of Mamluk Egypt, including manuscript illumination, from the 1350s onwards.5 Considering the eclecticism manifest in the decoration of Sultan Ḥasan’s funerary complex  as emblematic of a global moment in the history of fourteenth-century Islamic art, we might also remember that the economic resources to build the monument came at least in part from the estates of those who died intestate as a result of the Black Death, a disease whose pathways of circulation were in many cases the same as those along which fourteenthcentury artisans and artistic forms traveled.

2.

Meinecke (1976-1977), 85-144, esp. 89-97. For alternative views see Kahil (2008), 63-68. On the similarities between the ornament of the Altinbugha al-Maridani mosque (1340) and Gujarati textiles see Crowe (1989), 459-464. 3. O’Kane (1996), 499-522. For Pandua see Eaton (1996), 42-46. 4.  Kahil (2008), 79-84; Jacoby (1982), 121-138. 5. Rogers (1972), 385-403. For the wider contemporary phenomenon, see Kadoi (2009).

ECLECTICISM AND REGIONALISM • 155

Figure 1 – Funerary Complex of Sultan Ḥasan Cairo, detail of chinoiserie carving  with lotus ornament, funerary chamber, 1363. [© Photograph: F. B. Flood]

Sultan Ḥasan’s complex offers a particularly dramatic example of an artistic eclecticism that exploited ‘global’ cultural flows, but a similar trend is manifest  during the same period as far East as north India, which, like Egypt, had never been incorporated into the Mongol empire. This lack of incorporation was in itself conducive to the circulation of artistic forms, as refugees from the Mongol conquests brought their skills to Delhi in the course of the 13th century, leaving a palpable mark on the development of contemporary architectural decoration.6 It is, in fact, architecture that offers the most useful indicator of the transregional circulations and connections of the 13th and 14th centuries that are so palpable in the illuminations of the Gwalior Qurʾan. The architecture of the Tughluq dynasty that ruled north India from 1321 to the period of Tīmūr’s  invasion in 1399 (the year in which the Gwalior Qurʾan is dated) is particularly marked by the introduction to North India of forms and techniques from the wider Islamic world to the West. These include the precocious glazed tile ornaments in the tomb of Rukn-i ʿAlām of Multan (ca. 1320), itself a monument  possibly inspired by the tomb of the Ilkhanid sultan Öljeitü built at Sultaniyya in western Iran less than  a decade earlier.7  To  this  might  be  added  the  first  appearance of the four-iwān plan and carved stucco

6. 7.

Flood (2009a), 236. Hillenbrand (1992), 148-174.

ornament at Tughluqabad, the new Tughluqid capital near Delhi built by the Rumi (Anatolian) architect and vizier Aḥmad b. Ayāz,8 the tentative appearance of blue-glazed tiles in the rebuilt Friday Mosque of Badaʾun (roughly 150 miles from Gwalior) in 1326,9 and the engaged paired minarets, blue-glazed elements and carved stucco ornament in the Friday Mosque of Jahanpanah,  the  new  capital  built  by  Muḥammad  b. Tughluq around 1343 (figures 2-3).10 This receptivity to Persianate forms and motifs continued even in later Tughluq architecture, for the spectacular stucco ornament in the tomb of Firūz Shāh Tughluq (d. 1388)  at Haus Khas in Delhi (figure 4) shows affinities with  the illuminations of Ilkhanid manuscripts produced a few decades earlier, which have themselves been cited as comparanda for the illuminations in the Gwalior Qurʾan.11  In  view  of  this  filiation  between  north Indian architectural ornament and the illumination of highly portable manuscripts from both India and Iran, it is worth noting reports that the madrasa adjoining Firūz Shāh’s tomb was provided  with carpets from Shiraz, Yemen and Damascus.12 8. 9. 10. 11.

Shokoohy, Shokoohy (2007), 24, 113-122, pl. 7.29. Flood (2005), 178-180. Welch, Crane (1983), 130-133. See, for example, the illuminations in a copy of the Majmūʿa al-rashīdiyya produced in Tabriz between 1307 and 1310 (BnF Arabe 2324):  Chaigne  (2012),  255-265,  figs. 5b-5c.  For a discussion of these illuminations in relation to those in the Gwalior Qurʾan see Brac de la Perrière (2009), 346-347. 12. Welch (1996), 182.

156 • FINBARR BARRY FLOOD

Figure 2 – Friday Mosque of Jahanpanah, Delhi, general view of entrance to the prayer hall, ca. 1343. [© Photograph: F. B. Flood]

Figure 3 – Friday Mosque of Jahanpanah, Delhi, remains of blue-glazed lotus flower  in spandrels of exterior arches, ca. 1343. [© Photograph: F. B. Flood]

ECLECTICISM AND REGIONALISM • 157

Figure 4 – Tomb of Firūz Shāh Tughluq, Haus Khas, Delhi,  detail of painted stucco dome revetments, ca. 1388. [© Photograph: F. B. Flood]

In short, the architectural projects of north Indian sultans not only attest to the enduring mobility of artistic forms and techniques from the Persianate world to the West during the 13th and 14th centuries, but also indicate contemporary receptivity to western modes of ornament, perhaps even a vogue for eclectic combination and experimentation that has an obvious relevance to the kaleidscope of forms and motifs seen in the Qurʾan produced at Gwalior at the end of the same century. Gwalior may not seem to be the most obvious place for the production of such an eclectic manuscript, but it is worth noting that the city was not completely absent from the geographic imaginary of regions to the west: an ancient palace in the fortress of Gwalior is, for example, among the legendary monuments mentioned in a thirteenth- century Yemeni geography.13 The practicalities or pragmatics of artistic mobility between the central Islamic lands and India was undoubtedly tied to the role of mediators, both dynastic and individual, who sometimes facilitated the transmission of artifacts and artistic forms across remarkably long distances.14 Among them one might mention 13. Smith (2008), 193. 14.  Wagoner (1999), 241-64; Flood (2012), 131-142.

the Rasulids sultans of Yemen (r. 1229-1454), who reportedly sponsored the construction at least one Friday Mosque in China in which the khuṭba was read in the Rasulid sultan’s name, and also exchanged embassies  with  the  Yüan  and  Ming  dynasties,  contacts perhaps attested by reported finds of ‘re-gifted’  Mamluk enameled glass as far East as China.15 The chinoiserie of fourteenth-century Ilkhanid and Mamluk art even finds a counterpart in what might be termed  the occidentalism of Islamic architecture of this period in China. In the Sheng-Yu Si Mosque (Mosque of the Holy Friend) at Quanzhou on the Southern coast of China, forms from the central Islamic lands were mediated by the patronage of a Muslim from Shiraz, who renovated the mosque in 1310. It was presumably then that what is clearly a stone approximation or translation of a muqarnas semi-dome was set in place over the main entrance to the mosque (figure 5).16 15. See, for example, the enamelled glass vase now in the Freer Gallery of Art in Washington, DC: http://www.asia.si.edu/ collections/singleObject.cfm?ObjectNumber=F1934.19, accessed September 25, 2013. On the Rasulids and China see Vallet (2007), 158. 16. Luo (1994), 209-212, 216.

158 • FINBARR BARRY FLOOD

Figure 5 – Sheng-Yu Si Mosque (Mosque of the Holy Friend), Quanzhou, detail of entrance showing stone muqarnas semi-dome, 1310. [© Photograph: R. Lee]

Shiraz was of course a fulchrum for the circulation of things and persons along the maritime routes, a key nexus between the Gulf, India and the port cities of Southern China that lay beyond. This relationship is manifest in the close relationship between Shirazi manuscripts and some of the illuminations in the Gwalior Qurʾan.17 The possibility of an earlier relationship between Shirazi and Indian illustrated manuscript traditions is raised by the striking use of a red ground in Injuid manuscripts, a feature documented in earlier Jain manuscripts produced on the West coast of India, including Gujarat, a region long important to long-distance trade. Although it has sometimes

been assumed that the use of a similar red ground in the sultanate painting of India derives from Shirazi prototypes, the adoption of this feature in both sultanate painting and fourteenth-century manuscripts produced in Fars is more likely to reflect the  common impact of north-Indian, especially Jain, artistic traditions that can be documented a century or two earlier.18 Hinting at dimensions of culture contact whose significance awaits further investigation, this  common feature may serve as a reminder that artistic contacts between Iran and India during the 14th century were characterized by a multi-directionality or mutuality conducive to innovation in both regions.

17. See, for example, a Shirazi Qurʾan from the 1340s in the Nasser D. Khalili collection (Qur182, fols. 26v-27r), in which text blocks are framed by a wreath of foliage in a manner analogous to that of many folios in the Gwalior Qurʾan: Wright (2013), fig. 19.

18. Robinson (1991), 63. For the Jain material see Guy (1995), 30-41; id. (1994), 89-102.

ECLECTICISM AND REGIONALISM • 159

Perhaps more importantly for understanding the eclecticism of the Gwalior Qurʾan, the circulation of manuscripts between Shiraz and India during the 14th century is attested by a report in the Kitāb al-wāfī bi al-wafayāt of Ibn Aybak al-Ṣafadī (d. 1363), copied  from the Masālik al-abṣār of Ibn Faḍl Allāh al-ʿUmarī  (d. 1349). The report, attributed to the Tughluqid courtier and later ascetic shaykh Mubārak al-Anbayatī  (or  al-Anbatī),  refers  to  a  number  of  manuscripts,  including a copy of Ibn Sīnā’s Kitāb al-shifāʾ calligraphed by the celebrated thirteenth-century Iraqi calligrapher Yāqūt al-Mustaʿṣīmī, that was brought by the son of  the qāḍī  of  Shiraz  as  a  gift  to  sultan  Muḥammad  b. Tughluq of Delhi, at some point before 1340.19 If the breakdown of the Mongol empire and the fragmentation of the world system in the first half  of the 14th century not only enabled but enhanced artistic flows across long distances, my second point  relates to the question of origins and sources. Put simply, it is that global systems are never sui generis, they never simply spring into being spontaneously. The horizontal flow of images, forms and motifs across a large area of the Islamic world during and after the period of the Pax Mongolica should not obscure the fact that those artistic traditions in which new transcultural or transregional elements are clearly manifest were also inevitably informed by a second, vertical axis of cultural practice, the axis of established historical and/or regional tradition. While the incorporation of the Eastern Islamic lands into the Mongol empire undoubtedly encouraged the circulation of artistic forms over long distances, it is also important to recognize that the roots of some at least of this mobility had been laid in the second half of the 12th century and the beginning of the 13th, when a combination of political fragmentation, the revival of the authority of the Abbasid caliphate, and the restoration of Sunni hegemony, often as a result of military expansionism, proved particularly auspicious for the circulation of artistic forms and practices over an area extending from the Eastern Mediterranean to northern India. Whether looking at the bīmāristān of Nūr al-Dīn in Damascus  (1154) with its imported muqarnas dome and early Syrian manifestation of the four-iwān plan associated with the Persianate East, or the coins minted around 1200 by the Ghurid sultans of remote mountain Afghanistan based on Syrian or even North African models,20 one gets the impression that artisans, forms, and techniques were increasingly mobile in the century before the Mongol conquest.

19.  Al-Ṣafadī [1974], vol. 3, 173; Zaki (1981), 115. 20.  Tabbaa (2001), 119-124; Flood (2009a), 103-104.

The Ghurids may be especially germane to this broader context for understanding the ornamentation and structure of the Gwalior Qurʾan. The ephemeral nature of the Ghurid sultanate (ca. 1150-1210), and the paucity of manuscripts and other portable objects that can be securely attributed to Ghurid patronage make it difficult to evaluate its artistic legacy, but certain formal features introduced in the Qutb Mosque of Delhi (1192), the first Friday mosque built after the Ghurids conquered north India, were perpetuated in Delhi well into the 14th century, and possibly even carried to Samarqand as a result of Tīmūr’s invasion  of India.21 It is, therefore, possible that the openness to experimentation and innovation that characterizes much sultanate art may follow a precedent established in Ghurid art. Although it is easier to demonstrate this for architecture than the portable arts, one remnant of this pre-Mongol world that has not, as far as I am aware, been brought into discussions of the Gwalior Qur’ān is a four-volume leather-bound Qurʾan completed for the Ghurid sultan Ghiyāth al-Dīn Muḥammad b. Sām on 8 Rabīʿ II 584/6 June 1189 (figure 6).22 Unfortunately, the colophon does not seem to provide a place of production, Firuzkuh (the probable site of the Jam minaret) or Nishapur are possible but Herat is  likely;  in  the  second  half  of  a  major  metalwork  school was centered in the city, whose Friday Mosque was rebuilt in 1200, during a period when stylistic and epigraphic evidence attests the participation of its denizens in Ghurid architectural projects as far away as the Indus Valley and north India. Ghazna and Bust are less likely, since these fell within the territories  governed  by  Ghiyāth  al-Dīn’s  brother,  Muʿizz al-Dīn. The Qurʾan is a superlative example of the arts of binding, calligraphy and illumination, comprised of good quality large burnished sheets of paper, lavishly gilded and illuminated, with between six and seven lines of script on each folio; each volume consisted of  between roughly 170 and 195 folios. The illumination is most elaborate in the final volume of the manuscript,  and reaches a crescendo towards its end, where the short penultimate chapters possess the largest and most elaborate chapter headings, a foretaste of the dazzling, heavily gilded double finispiece of the final  21. The domes associated with the lateral entrances of the Qutb Mosque (1192) recur in the city’s Jahanpanah Mosque (mid. 14th century) and in the Bibi Khanum Mosque in Samarqand (1398-1405), where it has been suggested  that  their  presence  was  inspired  by  Tīmūr’s  familiarity with the Tughluqid mosque in Delhi: Golombek, Wilber (1988), vol. 1, 259. 22.  Anon.  [1949],  part 2,  nos. 30-33,  16-17;  Bahrami  (1949),  23, no. 52; Ettinghausen (1954), 470; London (1976), 320,  no. 509; Afrawand [1996], 4-14; Soucek (2000), 494, fig. 18;  Flood (2009b), 91-118.

160 • FINBARR BARRY FLOOD

volume. With its astonishingly high production values the Qurʾan of 1189 is perhaps the most spectacular Qurʾan manuscript to have survived from pre-Mongol Iran. It is also one of the largest: with a folio size of ca. 39 × 29 cm the Ghurid Qurʾan is approximately 30% larger than the Gwalior Qurʾan (29 × 22 cm), anticipating the scale of Qurʾans of the Ilkhanid and post-Ilkhanid period. The Ghurid Qurʾan manuscript is unusual in a number of respects. It possesses a lengthy colophon telling us that it took the scribe Muḥammad b. ʿĪsā  b. Muḥammad  b. ʿAlī  al-Nīshāpūrī  al-Laythī  (whose  nisba suggests he may have hailed from Nishapur) five years to complete. It also provides the most extensive and most bombastic rendition of the titles of  the  Ghurid  sultan  Ghiyāth  al-Dīn  Muḥammad  b. Sām  to  have  survived.  The  calligraphy  combines  a variety of angular and cursive scripts (including naskh, thuluth, and New Style), a combination that also characterizes the monumental epigraphy in Ghurid monuments.23 The presence of an interlinear Persian gloss on the Arabic text is not unique (although this must be among the earliest dated occurrences), but the inclusion of a popular Qurʾan commentary (tafsīr) by Abū Bakr al-Sūrābādī (d. ca. 495/1101) (although  not identified as such) at the end of each chapter is  unusual in a Qurʾan of this period.24 The importance of the manuscript can hardly be overstated. As a unique royal manuscript it fills a major lacuna in the history  of the material Qurʾan between the introduction of paper and cursive scripts in the 10th and 11th centuries and the celebrated Mamluk and Ilkhanid Qurʾans of the fourteenth. More importantly, some of the features of the Gwalior Qurʾan are anticipated in the Ghurid Qurʾan, although it shows none of the chinoiserie that is such a marked feature of the Gwalior Qurʾan, which was integrated into the repertoires of artists and artisans working in the Islamic world only after the Mongol invasions of the 13th century. Among the most obvious parallels between the Ghurid and Gwalior Qurʾans are a penchant for eclecticism in the illuminations, manifest in the variety of scripts used in sura headings, ranging in the Ghurid Qurʾan from kufic (in both foliated and non-foliated varieties),  to New Style and naskhī, while the sacred text itself is written in tawqīʿ. In the Gwalior Qurʾan, we find a  similar taste for variety in the use of foliated kufic 

and muḥaqqaq for juzʾ markers and thuluth for sura headings, with bihārī used for the sacred text.25 Like the Gwalior Qurʾan, the Ghurid Qurʾan is provided with  an  interlinear  Persian  gloss,  one  of  the  first  dated Qurʾans to include such a feature. In addition, the text of the opening folios of the first volume of  the Ghurid Qurʾan is surrounded by a cloud of vegetation that swirls around the interstices of the text, anticipating the dense elaboration of a similar feature in the Gwalior Qurʾan (figures 6-7). None of these features is of course unique. What does suggest a more specific relationship to the Gwalior Qurʾan is the unusual division of the Ghurid Qurʾan. Qurʾans of this period tend to be single volume or divided into thirty, seven (or more rarely, six or two) volumes, while the Ghurid manuscript is divided into four, each volume bearing some or all of its original tooled leather binding.26 Such a division is rare, although four-volume divisions of tafāsīr are known, and might conceivably have informed the division of the Qurʾanic text. Whatever the reason that this division of the revelation was favored, the four-fold division of the Ghurid Qurʾan may well have been employed in other Qurʾans produced in Afghanistan and north-western regions of South Asia in the 12th and 13th centuries. The persistence of this unusual division may help explain a peculiarity of the Gwalior Qurʾan that recurs in other sultanate period Indian Qurʾans (all undated). This is a distinction conferred on suras 1 (al-Fātiḥa), 7 (al-Aʿrāf), 19 (Maryam), and 38 (Ṣād) by the provision of heavily illuminated double-page frames that do not announce the opening of any other suras. The presence of elaborate double-page illuminations around these four suras (and no others) articulates, in effect, a four-fold division of the Qurʾanic text. Occurring within the single volume Gwalior Qurʾan, this division recapitulates the four-volume division of the Ghurid Qurʾan of 584/1189, each volume of which begins with one of the four suras highlighted in the Gwalior Qurʾan. The Gwalior Qurʾan is not unique in this respect, but can be located within a broader tradition attested to by other undated north Indian Qur’ans of the sultanate period in which the same four-fold division is articulated by the framing and illumination of the openings of the same suras. These include a single volume Qurʾan now in the Walters Art Museum in Baltimore (W563) and a double-volume Qurʾan in the

23. Hillenbrand (2000), 143. 24.  The popularity of al-Sūrābādī’s Tafsīr in twelfth-century Khurasan is evident from the fact that two earlier copies, dated 523/1129 and 535/1140-1141 survive: Lazard (1963), 91-94. For the intellectual and religious context see Flood (2009b); Zadeh (2012), 547-554.

25. Brac de la Perrière, Chaigne, Cruvelier (2010), 116-117. 26. Volume 1 (Iran Bastan Museum 3500) runs from sura 1 (al-Fātiḥa) to 6 (al-Anʿām); volume 2 (Iran Bastan Museum  3499) from sura 7 (al-Aʿrāf) to 18 (al-Kahf);  volume 3  (Iran Bastan Museum 3496) from sura 19 (Maryam) to 37 (al-Sāffāt); volume 4 (Iran Bastan Museum 3507): from  sura 38 (Ṣād) to 114 (al-Nās).

ECLECTICISM AND REGIONALISM • 161

Figure 6 – Ghurid Qurʾan of 1189, vol. 1, recto of double opening containing Sura 1, al-Fātiḥa (Iran Bastan Museum 3500, p. 5). [© Iran Bastan Museum, Photograph: F. B. Flood]

Figure 7 – Ghurid Qurʾan of 1189, vol. 3, double opening framing Sura 19, Maryam (Iran Bastan Museum 3496, folios unknown). [© Iran Bastan Museum]

162 • FINBARR BARRY FLOOD

Figure 8 – Ghurid Qurʾan of 1189, vol. 1, recto of double opening containing Sura 2, al-Baqara (Iran Bastan Museum 3500, p. 7) (detail). [© Iran Bastan Museum, Photograph: F. B. Flood]

Figure 9 – Ghurid Qurʾan of 1189, volume 4, double finispiece (Iran Bastan Museum 3507, fols. 195r-196v). [© Iran Bastan Museum, Photograph: F. B. Flood]

ECLECTICISM AND REGIONALISM • 163

Nasser D. Khalili Collection of Islamic Art (QUR237) in which a special distinction conferred on the beginning of suras 1, 7, 19 and 38 by the provision of elaborately illuminated double frontispieces. Neither bears a date, but both are tentatively ascribed to the late 14th or 15th centuries. A single volume sultanate Qurʾan in the al-Sabah Collection in Kuwait (LNS 278 MS) that distinguishes the same four suras in similar fashion is also undated, but certain stylistic features suggest a date in the 14th century, if not the 13th.27 The elaborate ornamentation used to distinguish the openings of suras 1, 7, 19, and 38 in these sultanateperiod Qurʾans, of which the Gwalior Qurʾan is the only dated example, suggests the persistence of a vestigial four-fold division even within single- or two-volume sultanate Qurʾans. This is likely to perpetuate the division favored in the four-volume structure of the Ghurid royal Qurʾan, and other twelfth-century Afghan Qurʾans that have not survived. In addition, in both the Ghurid and Gwalior Qurʾans (and also in MS W563), the double frame that opens suras 1, 7, 19 and 38 (figures 6 and 8), is more heavily illuminated and more richly gilded than those that open all other suras (compare, for example, figures 6 and 8 with figure 7).28 In the first  volume of the Ghurid Qurʾan, the double page richly illuminated jadwal containing sūrat al-Fātiḥa is, like their equivalents in the Gwalior Qurʾan, preceded by a pair of heavily gilded geometric carpet pages, which  are  repeated  in  the  finispiece  of  the  fourth  and final volume (figure 9). In the case of the Gwalior Qurʾan, it has been suggested that this use of a decorative double geometric frontispiece followed by double frontispieces framing the opening texts of each juzʾ was inspired by contemporary Mamluk Qurʾans.29 This is entirely possible: both direct and indirect contacts between fourteenthcentury India and Egypt are well-documented. Indeed, in its form and details, the double frontispiece that opens the Gwalior Qurʾan (fols. 1v-2r) is much closer to the frontispieces found in Ilkhanid and Mamluk Qurʾans than those found in pre-Mongol Qurʾans from Iran and Afghanistan. Nevertheless, it is entirely possible that the basic structure of the Gwalior Qurʾan – in particular, the quadrapartite division implied by the choice to distinguish suras 1 (fols. 2v-3r), 7

27. I am very grateful to Nahla Nasser of the Nasser D. Khalili Collection for drawing my attention to QUR237 and providing images of the manuscript, and to Sue Kaoukji of the al-Sabah Collection for permitting me to study LNS 278 MS during a visit to Kuwait in May 2015. 28. Brac de la Perrière, Chaigne and Cruvelier (2010), 117. 29. Brac de la Perrière (2015). I am grateful to Éloïse Brac de la Perrière for supplying me with a copy of her paper.

(fols. 143v-144r), 19 (fols. 274v-275r) and 38 (fols. 406v407r) by the elaborate framing of their opening verses – perpetuates a tradition favored, for as yet unknown reasons, in the Qurʾans produced for the Ghurids and their successors. If this is the case, we are dealing with a more complex scenario than the simple reception of Ilkhanid or Mamluk modes of manuscript illumination, a scenario in which the basic structure of the Gwalior Qurʾan perpetuates an earlier, perhaps even archaic, regional  tradition,  while  the  specific  forms  of its ornament are shaped by the transregional flow  of artistic forms and practices across a wide swath of the Islamic world from Egypt to India, during and after the period of Mongol hegemony in the 13th and 14th centuries. In other words, while the illuminations and ornament of the Gwalior Qurʾan are clearly informed by relatively contemporary traditions that flowed  horizontally  across  the  Islamic  work  in  the  14th century, the division of the text itself may be an archaism that perpetuates a less immediately visible inheritance from earlier regional traditions, documented in the Ghurid Qurʾan produced 200 years earlier. Such a perpetuation of earlier traditions would be very much in keeping with the archaisms in fourteenth-century sultanate painting that have been noted elsewhere by Éloïse Brac de la Perrière.30 The Ghurid Qurʾan never traveled to India, since it was endowed to the shrine of Shaykh Aḥmad b. Abū  al-Ḥasan  (d. 1141)  at  Turbat-i  Shaykh  Jam,  now  an  Iranian border town to the West of Herat, in 1256 and remained there until it was taken to Tehran in the early 20th century.31 Nevertheless, it seems highly unlikely that this was the only such Qurʾan ever made in Ghurid Afghanistan, especially when one considers that it was completed in 1189, a decade or two even before large amounts of Indian gold and booty started  flowing  into  the  Ghurid  sultanate  to  fund  major artistic projects, such as the rebuilding of the Friday Mosque of Herat in 1200. One possibility, therefore, is that other such four-volume Qurʾans existed and circulated eastward to India before or after the collapse of the Ghurid sultanate and the emergence of Delhi as the capital of an independent sultanate around 1210. There is, however, an alternative possibility that should be considered, one that relates to the revival of the legacy of the Ghurid sultanate in the 14th century under the Kart or Kartid dynasty of Herat, which ruled between roughly 1278 and 1383. The rise of the Kartids, who claimed descent from the Shansabanid clan of Ghur, ruled as Ilkhanid vassals, and were

30. Brac de la Perrière (2008). 31.  Golombek (1971), 27-44; Zadeh (2012), 550-554.

164 • FINBARR BARRY FLOOD

eventually defeated by Tīmūr, underlines the point  made earlier about the possibilities for realizing regional political ambitions after the collapse of Ilkhanid power in the 1330s. Although often overlooked, the Kartids were contemporaries of the Jalayirids, Muzaffarids, and Injuids, whose rule bookended the interregnum from the collapse of the Ilkhanids to the rise of the Timurids, at whose hands the Kartid dynasty met its end, the last of its scions put to death. The Kartids of Herat laid aggressive claim to the legacy of the Ghurids, claims apparent not only in the  name  of  Ghiyāth  al-Dīn  who  ruled  over  Herat  between 1307 and 1328, the Kartid namesake of the great Ghurid sultan Ghiyāth al-Dīn Muḥammad b. Sām  (r. 1163-1203);  contemporary  writers  often  conflated  and confused the two. The restoration of the Herat Mosque by the Kartid ruler in 1320 reenacted one of the central acts of the Ghurid sultan Ghiyāth al-Dīn’s  patronage, the rebuilding of the Friday Mosque of Herat around 1200. It was next to the Ghurid sultan in his tomb in the same mosque that the Kartid malik Ghiyāth al-Dīn was eventually laid to rest.32 More relevant to the broader context for the Gwalior Qurʾan is evidence for the continued role of Herat as a center for book production under the rule of the Kartids in the 14th century. A leather-bound Persian translation of the second rubʿa of al-Ghazālī’s  Iḥyāʾ ʿulūm al-dīn written by the scribe Nāṣir al-Ḥarāwī  in Herat in 1325 is now preserved in the Semyenov Collection of the Academy of Sciences in Dushanbe. The colophon invokes praise on Ghiyāth al-Dīn, the  Kartid ruler, and tells us that the manuscript was completed on the Southern platform of Herat’s Friday Mosque (dar masjid-i ādīna dar ṣuffa-yi janūbī), indicating that the Herat mosque functioned as a center for the production of religious texts.33 That mosques were a locus for manuscript production in both Afghanistan and India is confirmed by a fourteenth-century North Indian Ḥanafī ḥisba manual, which specifically  censures the activities of calligraphers and copyists (warrāq) in mosques.34 The Herat manuscript of the Iyḥāʾ ʿulūm al-dīn is executed in black ink with headings in red, and shows no sign of illuminations, but its scale is unusual: the paper is large in size, 30.5 × 25.5 cm, just slightly smaller than the folios of the Ghurid Qurʾan of 1189, which measure 39 × 29 cm. On the basis of the survival of this manuscript, Lola Dodkhudoeva has plausibly suggested that the Kartid rulers may have fostered the production of certain kinds of texts designed to enhance their appropriation of the Ghurid legacy.

Textual evidence indicates the production of illustrated manuscripts under the Kartids,35 and it seems probable that Qurʾans were also among the manuscripts that they commissioned, like the Ghurids before them. Given the instrumentalization of the Ghurid legacy by the Kartids, any such Qurʾans would, like so much of Kartid art and architecture, invariably have followed Ghurid precedents. Such a revival of Ghurid forms and practices offers a possible, if hypothetical, link between the manuscript traditions of the Ghurids, and the four-fold division of the Gwalior Qurʾan. This hypothesis is perhaps strengthened by the fact that Kartid artistic patronage not only reenacted that of the Ghurids, but mediated between the artistic legacy of the Ghurids and the innovations of the Timurids. The great brass basin that the Kartid Malik Pīr ʿAlī commissioned for the Friday Mosque  of Herat in 1374 provided, for example, the inspiration for that ordered by Tīmūr for the shrine of Aḥmad  Yasawī in Turkistan city in 1399.36 In addition, Dodkhudoeva raises the intriguing possibility of continuities between the patronage of the Kart rulers and the artistic  patronage  of  the  Timurids,  including  Shāh  Rukh in Herat (r. 1405- 1447), a likelihood signaled earlier by Terry Allen and Lawrence Potter.37 The suggestion that the structure of the Gwalior Qurʾan may perpetuate a tradition pioneered in earlier South Asian Qurʾan manuscripts in no way contradicts the evidence for a simultaneous relation to Ilkhanid, Injuid, Mamluk and Muzaffarid book production. What it does, however, remind us of is the need to be aware not only of spatial but also temporal dimensions of  artistic  patronage;  to  be  aware  not  only  of  the  horizontal flow of artistic forms across remarkable  distances, but also the need to consider the more vertical inheritance of regional traditions transmitted across time. In the case of the Gwalior Qurʾan, one might signal the need to be aware of both the potential legacy of earlier regional (Eastern Iranian, Afghan, and north Indian) traditions of manuscript production and the new artistic possibilities opened by the upheavals that led to the disappearance not only of the Ghurids, but also of a whole world order during the course of the 13th century. The  enhanced  cultural  flows  that  followed  provided the necessary conditions for the marked eclecticism of the illuminations in the Gwalior Qurʾan. These raise interesting questions about the extent to which this heterogeneity would have appeared as such to the late fourteenth-century users and viewers of the manuscript, and whether its geographic or spatial

32.  Glatzer (1980); Potter (1998), 55. 33. Dodkhudoeva (2009), 165-193. 34. Ibn ʿAwaḍ Sanāmī [1986], 163; Dien (1997), 46, 70.

35. Potter (1998), 158. 36.  Lentz, Lowry (1989), 29, fig. 4; Potter (1998), 159-160. 37.  Allen (1983), 46-49; Potter (1998), 153, 155-159.

ECLECTICISM AND REGIONALISM • 165

implications would have been apparent. Questions of cognition and perception relating to the circulation of ornament are perhaps better answered for the early modern period, highlighted, for example, by the use of the term khaṭā’ī to denote chinoiserie in Safavid Iran, one of seven ornamental modes, which also included geographic categories such as farangī or Frankish (i.e. European) ornament.38 This suggests at least a conventional memory of a relationship between ornamental form and geographic origins, but it is far from clear any late fourteenth-century viewers would have read the illuminations of the Gwalior Qurʾan in this way, even if its eclecticism was visible as such.39 Nevertheless, the suggestion that  the  manuscript  was  produced  in  a  Sufi  milieu  may be relevant,40 for in considering the visibility and potential meaning of the ‘global’ resonances of its illuminations, one might suggest that these were imbued with ideological or polemical resonances related to such an environment. Writing of the tensions between the authority exercised by sultans and Sufis in fourteenth-century India, Richard Eaton  has, for example, underlined a spatial dimension to the dialectic between royal sovereignty (ḥukūmat) and spiritual sovereignty (wilāyat) manifest in an opposition between region and transregion: “Whereas ḥukūmat, royal authority, was always limited in reach, and never coincided with the entire Muslim world – far less with the entire planet – the spiritual sovereignty of Sufis, wilāyat, was theoretically unlimited in territorial extent, and hence far greater than the worldly sovereignty of sultans”.41

Such a reading of wilāyat would fit well with the  transregional resonances of the Gwalior Qurʾan and its dazzling array of painted ornaments. By way of conclusion, I have to admit that on first  viewing, the vibrant visuality of the Gwalior Qurʾan, its dizzying combinations and variegated palette, invariably brought to mind the painted walls of a building far distant from Gwalior, and separated from our Qurʾan manuscript by almost century. In their eclecticism, luminosity, vivacity, and variety, the sheer brilliance and exuberance of the illuminations of the Gwalior Qurʾan immediately reminded 38.  O’Kane  (1992),  77-78;  Porter  (2000),  113-114;  Necipoğlu  (1995), 112-114; id. (2007), 12-13. 39. For general considerations of the relationship between identity or region and style in the perception of premodern Islamic ornament see Korn (2003), 237-260; Flood  (2009a), 200-205. Writing in a different context, Carolyn Dean and Dana Leibsohn have offered an insightful and provocative analysis of the relative visibility of hybridity in material culture to the modern viewer: Dean, Leibsohn (2003), 5-35. 40. Brac de la Perrière (2015). 41. Eaton (2000), 170.

me of the molded plaster ornament and paintings (kalem işi) executed in the interior of the mausoleum of Şehzade Mustafa and Cem Sultan in the Muradiye, the Ottoman royal cemetery at Bursa, far to the West, in the 1470s (figure 10). Like the illuminations  of the Gwalior Qurʾan, these also offer a kaleidoscopic repertoire of new and established techniques and forms drawn from over a wide geographic area. The Bursa paintings include calligraphic medallions, faux marble, flowers and vegetation, vases with floral sprays, mosque lamps, arabesques and knotted or interlace borders; many show the continuing impact  of earlier Timurid manuscript illuminations.42 Like the illuminations of the Gwalior Qurʾan, the Ottoman paintings constitute an “encyclopaedia of the kinds of  patterns  and  motifs”  in  contemporary  use;  as  Richard Turnbull has noted, they appear “as manuscript pages writ large”.43 The juxtaposition of the Gwalior Qurʾan illuminations with Anatolian tomb paintings executed seven decades later is admittedly superficial and shamelessly anachronistic, but Georges Didi-Huberman has argued persuasively that all artworks are necessarily anachronistic to the extent that they manifest diverse temporalities, brought into constellation in a single object.44 This is no less true of the Gwalior Qurʾan, which I have argued represents a point of intersection between older, regionally inflected traditions of  manuscript production and the dynamic transregional cultural flows that characterized much fourteenth-  century art. Moreover, even anachronisms can be useful heuristic devices, helping to provoke or stimulate our thoughts on particular artworks. What we might take away from the juxtaposition of illuminated Qurʾan and painted tomb, however anachronistic or seemingly arbitrary the comparison, is a way of thinking about the Gwalior manuscript. Not as an object or thing, so much as a space, the kind of space that Michel Foucault termed a heterotopia, a space in which geographically and temporally diverse artifacts, forms, and even languages co-exist, intermingle and inform their mutual reception.45 For Foucault, the museum was, of course, the heterotopia par excellence, a point of reference that may not be entirely irrelevant to the Gwalior Qurʾan, with its compendia of ornaments resembling a pattern-book of contemporary illumination. Thinking about the Gwalior Qurʾan as a space rather than a thing is perhaps more productive than invoking standard metaphors of hybridity or

42. Turnbull (2004), 119-145. The paintings are restorations or reproductions, believed to be faithful to the originals: 139. 43. Ibid., 139, 250. 44. Didi-Huberman (2003), 31-44. 45. Foucault (1986), 22-27.

166 • FINBARR BARRY FLOOD

Figure 10 – Tomb of Cem sultan, Muradiye, Bursa, detail of upper walls with painted ornament, 1470s. [© Photograph: F. B. Flood]

syncretism, which imply an uneasy and often a rather unstable juxtaposition of things that in some fundamental sense do not really belong together.46 Unlike the architectural spaces which Foucault imagined in developing his idea of the heterotopia, the Gwalior Qurʾan is of course highly mobile. In this sense also, it is also an appropriate place or space from which to acknowledge the extraordinary mobility that informed Islamic art and architecture during the 14th century, and the resulting eclecticism that lends so much fourteenth-century art and architecture of the Islamic world from Egypt to India its unusual vibrancy. However, if the importance of routes to the vibrancy of the Gwalior Qurʾan is abundantly clear, it would be a mistake to emphasize these at the expense of roots;47 indeed, the roots of these phenomena of mobility may already have been laid in Eastern Islamic art of the 12th and 13th centuries, before the emergence of the Mongol world system. In this sense, the Gwalior Qurʾan should be located at the intersection between the  synchronicity  of  horizontal  cultural  flows  – of  routes and their effects – and the diachronicity of vertical inheritances from earlier, rooted traditions. 46. See Shaw, Stewart (1994), 1-26 and Stewart, Ernst (2002), 586-588. 47. Clifford (1997).

It is also the case that both the Gwalior Qurʾan and the painted tomb at Bursa stand at the end of a series, on the cusp of an emerging world in which the hegemony of new world empires would ensure the dissemination of more canonical, more standardized and in many ways more homogeneous or uniform artistic forms and practices. What the Gwalior Qurʾan offers, perhaps, is a reminder that in many cases the great achievement of the new world empires of the Timurids and their Mughal, Ottoman, and Safavid successors lay in codifying and synthesizing artistic elements and forms that had existed earlier,48 and that had often been used in exuberant combinations where the cultural or geographic conditions were amenable to a multiplicity of receptions or, to use the rubric of this volume, to a reception characterized by polysemy. 48. Of course, this did not preclude mobility, even of the most literal and surprising kinds. In addition to Tīmūr’s  oft-cited removal of Indian stone-carvers and Iranian tile-workers to Samarqand, one might mention a 4-ft high tile panel in the Friday Mosque of Zabid, on the West coast of Yemen, which is strikingly similar to tile panels in the tomb of Shad-i Mulk Agha in Samarqand (1383) or in Yazd and appears to have been imported from Central Asia or Iran around the middle of the 15th century: Porter (1995), 66, fig. 60.

ECLECTICISM AND REGIONALISM • 167

bibLioGraPhy Sources anon. [1949] Rahnāma-yi ganjīnah-i Qurʾān dar mūzah-i Īrān-i Bāstān,  Tehran:  Chāpkhānah-i  Bānk-i  Millī-i  Īrān, 1328 solar. Ibn aybak al‑ṣaFadī (Ṣalāh al-Dīn Khalīl) [1974] Kitāb al-wāfī biʾl-wafayāt, vol. 3, Wiesbaden: Franz Steiner Verlag GMBH. iBn ʿaWaḍ sanāMī (Muḥammad) [1986] Niṣāb al-ihtisāb, edited by Murayzin Saʿīd Murayzin ʿAsiri, Mekka: Maktabat al-Ṭālib al-Jāmiʿī.

References aBu-lughod (Janet) 1991 Before European Hegemony: The World System A.D. 1250-1350, Oxford: Oxford University Press. aFrawand (Qadir) [1996] “Qurʾān-i mauqūfah-i Turbat-i shaykh Aḥmad  jāmī”, in Mīrāth-i Farhangī 15, 1375 solar, pp. 4-14. allen (Terry) 1983 Timurid Herat, Wiesbaden: Reichert. Bahrami (Mehdi) 1949 Iranian Art: Treasures from the Imperial Collections and Museums of Iran, New York: Metropolitan Museum of Art. Brac de la Perrière (Éloïse) 2008 L’art du livre dans l’Inde des sultanats, Paris: Presses de l’Université Paris-Sorbonne. 2009 “Du Caire à Mandu: notes sur la transmission des modèles dans l’Inde des sultanats”, in F. Richard et M. Szuppe (dir.), Écrits et culture en Asie centrale et dans le monde turco-iranien, xe-xixe siècles / Writing and Culture in Central Asia and in the Turko-Iranian World, 10th-19th c., Paris: Association pour l’avancement des études iraniennes, pp. 333-358. 2015 “Le Coran de Gwalior : nouvelles perspectives sur l’histoire des corans enluminés dans l’Inde pré-moghole”, in J. J. Witkam et N. Martinez de Castilla Muñoz (dir.), Codicology and history of the book in Arabic script. Journal of Islamic Manuscripts 6, Leiden: Brill Academic Publishers, pp. 219-238. Brac de la Perrière (Éloïse), chaigne (Frantz), cruvelier (Mathilde) 2010 “The Qurʾan of Gwalior, kaleidoscope of the arts of the book”, in Treasures of the Aga Khan Museum – Arts of the Book & Calligraphy, Istanbul: Culture and Sakıp Sabancı Museum, pp. 114-123.

chaigne (Frantz) 2012  “Le décor enluminé sous les Īl-khānides: entre  assimilation et innovation, de l’Iraq à la Chine”, in Beiträge zur Islamischen Kunst und Archäologie 3, pp. 253-265. cliFFord (James) 1997 Routes: Travel and Translation in the Late Twentieth Century, Cambridge (Mass.): Harvard University Press. crowe (Yolande) 1989  “Gujarat and Mamlūk Egypt: a pattern link”,  in Karen Frifelt (dir.), South Asian Archaeology, Papers from the eighth International conference of South Asian archaeologists in Western Europe, Moesgaard Museum, Denmark, 1-5 July 1985, London: Curzon Press/Riverdale. dean (Carolyn), leiBsohn (Dana) 2003 “Hybridity and its discontents: considering visual culture in colonial Spanish America”, in Colonial Latin American Review 12, no. 1, pp. 5-35. didi-huBerman (Georges) 2003 “Before the image, before time: the sovereignty of anachronism”, in Claire Farago, Robert Zwijnenberg (dir.), and translated by Peter Mason, Compelling Visuality: The Work of Art in and Out of History, Minneapolis: University of Minnesota Press, pp. 31-44. dien (M. Izzi) 1997 The Theory and the Practice of Market Law in Medieval Islam, London: E. J. Gibb Memorial Trust. dodkhudoeva (Lola) 2009 “Translating and copying in pre-Timurid Herat: A persian translation of the Iḥyāʾ ʿUlūm al-Dīn, 725726/1325”, in F. Richard, M. Szuppe (dir.), Écrit et culture en Asie centrale et dans le monde TurcoIranien, xe-xixe siècles, Paris: Association pour l’avancement des études iraniennes, pp. 165-193. eaton (Richard M.) 1996 The Rise of Islam and the Bengal Frontier, 12041760, Berkeley: University of California Press. 2000 “The articulation of islamic space in the medieval Deccan”, in Essays on Islam and Indian History, Delhi: Oxford University Press, pp. 159-175. ettinghausen (Richard) 1954 “The covers of the Morgan Manāfiʿ manuscript and other early Persian bookbindings”, in Dorothy Miner (dir.), Studies in Art and Literature for Belle da Costa Greene, Princeton: Princeton University Press, pp. 459-473.

168 • FINBARR BARRY FLOOD

Flood (F. Barry) 2005 “Persianate trends in sultanate architecture: the great mosque of Badaʾun”, in Bernard O’Kane (dir.), The Iconography of Islamic Art. Studies in Honour of Robert Hillenbrand, Edinburgh: Edinburgh University Press, pp. 178-180. 2009a Objects of Translation: Material Culture and Medieval ‘Hindu-Muslim’ Encounter, Princeton: Princeton University Press. 2009b “Islamic identities and islamic art: inscribing the Qurʾan in twelfth-century Afghanistan”, in Elizabeth Cropper (dir.), Dialogues in Art History, from Mesopotamian to Modern: Readings for a New Century (Studies in the History of Art Series), Washington (DC): National Gallery of Art, pp. 91-118. 2012 “Gilding, inlay and the mobility of metallurgy: A case of fraud in medieval Kashmir”, in Venetia Porter, Mariam Rosser-Owen (dir.), Metalwork and Material Culture in the Islamic World: Art, Craft and Text. Essays Presented to James W. Allan, London: I.B. Tauris, pp. 131-142.

JacoBy (Zehava) 1982 “Crusader sculpture in Cairo: Additional evidence on the temple area workshop of Jerusalem”, in Jaroslav Folda (dir.) Crusader Art in the Twelfth Century (British Archaeological Reports International Series, 152), Jerusalem: British School of Archaeology in Jerusalem, pp. 121-138.

Foucault (Michel) 1986 “Of other spaces”, in Diacritics 16, pp. 22-27.

lazard (Gilbert) 1963 La langue des plus anciens monuments de la prose persane, Paris: Librairie C. Klincksieck.

glatzer (Bernt) 1980 “Das Mausoleum und die Moschee des Ghoriden Ghiyāth ud-Dīn in Herat”, in Afghanistan Journal 7, no. 1, pp. 6-22. golomBek (Lisa) 1971  “The chronology of Turbat-i Shaikh Jām”, in  Iran, Journal of the British Institute of Persian Studies 9, pp. 27-44. golomBek (Lisa), wilBer (Donald) 1988 The Timurid Architecture of Iran and Turan, Princeton: Princeton University Press, 2 vols. guy (John) 1994 “Jain manuscript painting”, in Pratapaditya Pal (dir.), The Peaceful Liberators: Jain Art from India, New York: Thames and Hudson, pp. 89-102. 1995 “Early Jain painting: Sources and continuities”, in Oriental Art 41, no. 4, pp. 30-41. hillenBrand (Robert) 1992 “Turco-Iranian elements in the medieval architecture of Pakistan: the case of the tomb of Rukn-i ʿAlam at Multan”, in Muqarnas 9, pp. 148-174. 2000 “The Architecture of the Ghaznavids and Ghurids”, in Carole Hillenbrand (dir.), Studies in Honour of Clifford Edmund Bosworth, vol. II. The Sultan’s Turret: Studies in Persian and Turkish Culture, Leiden: E. J. Brill, pp. 124-206.

kadoi (Yuka) 2009 Islamic Chinoiserie: The Art of Mongol Iran, Edinburgh: Edinburgh University Press. kahil (Abdallah) 2008 The Sultan Ḥasan Complex in Cairo 1357-1364: A Case Study in the Formation of Mamluk Style (Beiruter Texte und Studien, 98), Würzburg/Beyrouth:  Ergon/Orient-Institut. korn (Lorenz) 2003 “Iranian style ‘out of place’? Some egyptian and syrian stuccos of the 5th-6th/11th-12th centuries”, in Annales Islamologiques 37, pp. 237-260.

lentz (Thomas W.), lowry (Glenn D.) 1989 Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century, exhibition catalogue, Los Angeles: Los Angeles County Museum of Art. London 1976 The Arts of Islam, exhibition catalogue, London: Hayward Gallery. luo (Xiaowei) 1994 “China”, in Martin Frishman, Hasan-Uddin Khan (dir.), The Mosque: History, Architectural Development and Regional Diversity, London: Thames and Hudson, pp. 208-223. meinecke (Michael) 1976-1977 “Die mamlukischen Fayencemosaikdekorationen: Eine Werkstätte aus Tabrīz in Kairo  (1330-1350)”, in Kunst des Orients 11, pp. 85-144. necIPoğlu (Gülru) 1995 The Topkapı Scroll: Geometry and Ornament in Islamic Architecture, Los Angeles: Getty Center for the History of Art and the Humanities. 2007 “L’idée de décor dans les régimes de visualité islamiques”, in Rémi Labrusse (dir.), Purs décors: Arts de l’Islam, regards du xixe siècle, Paris: Musée des arts décoratifs, pp. 10-23.

ECLECTICISM AND REGIONALISM • 169

o’kane (Bernard) 1992 “Poetry, geometry and the arabesque: notes on Timurid aesthetics”, in Annales islamologiques 26, pp. 63-78. 1996 “Monumentality in Mamluk and Mongol art and architecture”, in Art History 19, no. 4, pp. 499-522. Porter (Venetia) 1995 Islamic tiles, London: British Museum Press for the Trustees of the British Museum. Porter (Yves) 2000 “From the “theory of the two qalams” to the “seven principles of painting”: theory, terminology, and practice in Persian classical painting”, in Muqarnas 17, pp. 109-118. Potter (Lawrence G.) 1998 The Kart dynasty of Herat: religion and politics in medieval Iran, Ph.D. Thesis Columbia University. roBinson (Basil W.) 1991 Fifteenth-Century Persian Painting: Problems and Issues, New York: New York University Press. rogers (J. Michael) 1972  “Evidence  for  Mamlūk-Mongol  relations,  1260-1360”, in Colloque international sur l’histoire du Caire 27 mars-5 avril 1969, Cairo: General Egyptian Book Organisation, pp. 385-403. shaw (Rosalind), stewart (Charles) 1994 “Introduction: problematizing syncretism”, in Charles Stewart, Rosalind Shaw (dir.), Syncretism/Anti-syncretism: the Politics of Religious Synthesis, New York: Routledge, pp. 1-26. shokoohy (Mehrdad), shokoohy (Natalie H.) 2007 Tughluqabad: A Paradigm for Indo-Islamic Urban Planning and its Architectural Components, London: Araxus Books. smith (G. Rex), (translation) 2008 A Traveller in Thirteenth-Century Arabia, Ibn al-Mujāwir’s Tārīkh al-Mustabṣir, London: The Hakluyt Society. soucek (Priscilla P.) 2000 “The Development of calligraphy”, in C. E. Bosworth, M. S. Asimov (dir.), History of Civilizations of Central Asia. vol. 4, The Age of Achievement: A.D. 750 to the End of the Fifteenth Century, Part Two: The Achievements, Paris: UNESCO, pp. 492-511.

stewart (Tony K.), ernst (Carl W.) 2002 “Syncretism”, in Peter J. Claus, Margaret Mills (dir.), South Asian Folklore: An Encyclopedia, New York: Routledge, pp. 586-588. taBBaa (Yasser) 2001 The Transformation of Islamic Art During the Sunni Revival, Portland: University of Washington Press. turnBull (Richard H.) 2004 The Muradiye Tomb Complex in Bursa and the Development of the Ottoman Funerary Tradition, D. Phil. dissertation, Institute of Fine Arts, New York University, Graduate School of Arts and Science. vallet (Éric) 2007  “Les sultans rasūlides du Yémen, protecteurs  des communautés musulmanes de l’Inde (viieviiie/xiiie-xive siècles)”, in Annales Islamologiques 41, pp. 149-176. wagoner (Phillip B.) 1999 “Fortuitous convergences and essential ambiguities: Transcultural political elites in the medieval Deccan”, in International Journal of Hindu Studies 3, no. 3, pp. 241-264. welch (Anthony) 1996 “A Medieval center of learning in India: The Hauz khas madrasa in Delhi”, in Muqarnas 13, pp. 165-190. welch (Anthony), crane (Howard) 1983 “The Tughluqs: Master builders of the Delhi sultanate”, in Muqarnas 1, pp. 130-133. wright (Elaine) 2013 The Look of the Book. Manuscript Production in Shiraz, 1303-1452 (Occasional papers, new series, vol. 3), Washington (DC): Freer Gallery of Art/Arthur M. Sackler Gallery in association with University of Washington Press/ Chester Beatty Library. zadeh (Travis) 2012 The Vernacular Qurʾan. Translation and the Rise of Persian Exegesis, London: Oxford University Press in association with the Institute of Ismaili Studies. zakI (Muhammad) 1981 Arab Accounts of India during the Fourteenth Century, Delhi: Idarah-i Adabiyat-i Delli.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES Yves Porter

(Université Aix-Marseille, CNRS / UMR 7298 - LA3M*)

Shikar-shikan shawand hami tutīyān-i Hind Zin qand-i pārsī ki Bangāli mirawad “All the parrots of India are becoming sugar crunchers, Of this Persian sugar candy which circulates into Bengal” Ḥāfiẓ, ghazal no. 218

Abstract The lotus flower, issued from the decorative repertory of China, made a massive appearance  in Iranian art under the dominion of the Ilkhanids (late 13th – early 14th century); we find  it represented in manuscript illuminations, in so-called ‘Sultanabad’ ceramics and also in inlays on metalwork. In comparison, in Sultanate India, the Chinese-style lotus experienced a relatively limited flowering. Following the Delhi dynasties, where the lotus motif was mainly borrowed from  the local repertory, the flower in its Chinese form with a leaf was probably introduced in  two ways. One possibly direct source was ceramics (and other objects, possibly silks) from China; the findings of Firuz Shah Kutla illustrate this aspect brilliantly. The other source,  more indirect, seems to be Iranian or Mamluk art, in particular the field of manuscript illumination. However, although lotuses dominate the frontispieces of the Gwalior Qurʾan (but is this manuscript truly representative of the production of a period for which extremely few documents survive to bear witness?), they are very rarely represented in the architectural decor of the late 14th century. A notable exception appears in the Adina Masjid of Pandua (1374-1375); this monument, situated in the city that was for a time the capital  of the Bengal sultanate, is a rare and seductive milestone in the journey of this floral motif  on the subcontinent. The object of this paper is thus to examine the circulation of Chinese-style motifs such as lotus flowers and leaves, from their origin in Yüan China all the way to the arts of the  book and architectural decor of sultanate India.

*

This study was made possible thanks to the help of UMR 7298 (Laboratoire d’archéologie méditerranéenne médiévale et moderne, LA3M) at Aix Marseille University. I wish to address my deepest gratitude to Richard Castinel, not only for being with me in Bengal, but also for his constant help in writing and re-thinking this paper. Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 171-189

172 • YVES PORTER

Résumé Fleurs et feuilles de lotus, de la Chine et l’Iran à l’Inde des sultanats La fleur de lotus, issue du répertoire décoratif de la Chine, fait une apparition massive  dans les arts de l’Iran sous domination des Ilkhanides (fin xiiie - début xive siècle) ; on  la trouve représentée aussi bien dans le domaine des enluminures de manuscrits que dans la céramique dite de « Sultanabad » ou bien en incrustations sur des métaux. Dans l’Inde des sultanats, le lotus sinisant connaît une floraison relativement plus  limitée ; après les dynasties de Delhi, où le lotus est un motif majoritairement emprunté au  répertoire local, la forme sinisante de la fleur accompagnée de sa feuille est probablement  introduite par le biais de deux voies : l’une, possiblement directe, concerne les importations de céramiques (et autres objets, notamment les soieries ?) de Chine ; les trouvailles de Firuz  Shah Kutla illustrent brillamment cet aspect. L’autre faisceau, plus indirect, pointe vers des apports véhiculés par les arts de l’Iran ou du monde mamelouk, en particulier dans le domaine des manuscrits enluminés. Pourtant, si d’un côté, les lotus dominent les frontispices du coran de Gwalior (ce dernier est-il réellement représentatif des productions d’une époque pour laquelle nous possédons extrêmement peu de témoins ?), ils sont en revanche très peu représentés dans les décors architecturaux de la fin du  xive siècle. Une exception notable apparaît à l’Adina Masjid de Pandua (1374-1375) ; ce monument, situé dans la ville qui  fut un temps capitale du sultanat du Bengale, forme un jalon aussi rare que séduisant dans le cheminement de ce motif floral dans le sous-continent. Le but de cette communication est donc de s’interroger sur la circulation de motifs sinisants tels que fleurs et feuilles de lotus, depuis leur origine dans la Chine des Yüan  jusque dans les arts du livre et les décors architecturaux de l’Inde des sultanats.

introduction The lotus flower, coming from the decorative repertoire of China, made a massive appearance in the arts of Iran under Ilkhanid domination (late 13th - early 14th century);  it  is  found  on  many  kinds  of objects such as manuscript illuminations, on so-called ‘Sultanabad’ ceramics, and even incrusted on metalwork.1 In pre-Mughal India, the Chinesestyled lotus had a much more limited blossoming: after the period of the Delhi dynasties, during which the lotus motif was in majority borrowed from the local  repertoire,  the  Chinese  form  of  this  flower,  together with its characteristic leaf, was probably introduced in two different routes: the first one, possibly direct, had to do with the import from China of ceramics and other commodities such as silk fabrics;  the  porcelain  findings  at  Firuz  Shah  Kutla  brilliantly illustrate this aspect. The second route, more indirect, has to do with motifs brought by the arts of Iran or from the Mamluk area, particularly

1.

Several authors have written on the subject of the Chinese lotus in Islamic art; see for instance, Crowe (1999), 17-22.  Although less specific, see also Rawson (1984). On chinoiserie in Ilkhanid Iran, see Kadoi (2009).

in  the  field  of  illuminated  manuscripts.  However,  although on the one hand the lotus dominates the frontispieces of the Gwalior Qurʾanic manuscript – and one might still ask whether this work is really representative of the production from a period for which we possess extremely few witnesses – on the other hand the lotus flower is remarkably poorly represented in Indian architectural decors before the 15th century. A notable exception appears at Pandua’s  Adina  Masjid  (1374-1375,  West  Bengal);  this monument, situated in the city that once was the capital of the Bengal sultanate, forms a rare yet appealing milestone in the spreading of the lotus flower over the subcontinent. I will first go through  the origins and forms of the lotus in China, notably during the dynasties  preceding the Yüan; then the  Mongol period will be examined, in China as well as in  Ilkhanid  Iran;  finally,  this  survey  will  end  with  the arts of the book and architectural decors from the Indian sultanates before the 15th century. A second section will deal in greater detail with the architectural decoration of Pandua’s Adina Masjid, in particular with the meaning which can be attributed to its lotus flowers and leaf motifs, even if their localization in  the monument is indeed very limited. Both the absence of Bengali manuscripts dating from this period and the capital importance of this outstanding monument in fact justify a more detailed study.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 173

towards the oriGins of the Lotus motif Chinese lotuses It is certainly difficult to trace back the origins of  the lotus motif in the arts of China; indeed, at least  since the 5th century CE, more or less recognizable floral  motifs  were  used,  in  particular  in  Buddhist  architecture. Much later, from the 10th to 13th centuries onward, these patterns or others derived from them appeared on objects and artefacts. At this time we find  some  lotus  flowers  on  ceramics,  for  instance,2 and these types of wares were circulating well beyond the Middle Kingdom. However, we can note that whereas lotuses sometimes happen to be fully recognizable,  other  flowers,  often  drawn  in  a  schematic  way, might be taken for peonies.3 In other cases, such as on a Ting bowl with a moulded decor under a white glaze, we see on the contrary the design of a pond where lotuses are blooming in the midst of their nicely drawn high-stemmed leaves.4 As Yolande Crowe has clearly shown, lotus flowers  probably first penetrated into the Islamic sphere through Song textiles.5 An eloquent example of this kind of material is found in the trousseau of a young bride from Quanzhou (Fujian), buried in 1243 together with hundreds of textiles (both clothes and fabrics).6 It should be remembered that at this date the production of blue and white porcelain was still inexistent. With the appearance of this kind of ceramic – which can be dated towards the end of first quarter of the  14th century7 – the production of objects decorated with lotus motifs underwent a veritable explosion. Of course, a large number of these vessels were intended for export to the Middle East. Furthermore, during this period (from the late Song to the

2.

See for instance, Liao moulded earthenware dish, 10th-11th century, Medley (1998), fig. 105; while the flower follows the  usual pattern of a ‘lotus’, the leaves are, on the contrary, alien to this species. See also some Tz’u-chou (or Cizhou, Hopei province) stoneware painted in black from the 11th to 13th centuries (ibid., figs. 87, 90-91). 3.  Indeed, Rawson (1984) analyzes both flowers in the same  chapter: ‘Lotus and peony scrolls’, 63-88. 4. Ting bowl with moulded decoration of infants among flowers, 13th century, British Museum, Medley (1998), fig. 77. The  sacred lotus, Nelumbo nucifera, has rounded and pelted leaves,  measuring  up  to  50 cm  in  diameter;  they  are  either  floating,  even,  or  rise  up  to  75 cm  above  water,  becoming orbicular, in the shape of a cup. Flowers are whitish pink, measuring from 15 to 30 cm in diameter, and have about twenty petals. See Roguet (1999), 3-10. 5. See Crowe (1999), 17-18. 6. Ibid. 7.  Medley  (1998),  176;  Medley  points  out  that  very  few  dated pieces from the 14th century exist, besides the ‘David Vases’, dated 1351, kept at the Percival David Foundation, London; see id., 178 and fig. 131.

early 14th century), Chinese textiles continued to be largely imported by rich merchants and the upper classes of the Muslim courts. It is therefore through this double pathway (first textiles, then ceramics) that  the decorative arts of Islam came to be impregnated by the ‘Chinese’ lotus, starting in the late 13th century. The motifs on textiles are often more schematic than those on the 14th century blue and white ware; indeed,  this medium often uses a naturalistic idiom which can be appreciated in subjects such as the lotus pond with mandarin ducks. The delicate painting of these porcelains furnishes the details of distinct petals and leaves, often rendered in different shades of blue, even though, at the same time, the flowers painted  on rims and borders appear as less identifiable botanical species.

The Chinese lotus in Ilkhanid Iran The  first  representations  of  the  ‘Chinese’  lotus  probably appear in Iran at Takht-i Sulayman (ca. 1275).8 Straightaway, we can observe that, in what can be termed  its  ‘first  phase’  or  ‘first  form’,  the  Ilkhanid  lotus has nothing of the representation of a botanical species. On the contrary, it already reflects a simplification of its original shape, through the sieve of a ‘natural’ tendency toward schematisation. The medium allowing this transfer is most probably textiles; indeed, even in China there was a trend  of oversimplifying the rendering of various motifs. Lotus  flowers  principally  appeared  on  glazed  tiles,  both the so-called low-fired ‘lajvardina’, and those decorated with metallic lustre, on borders as well as in the central position on the objects; the leaves had  not yet appeared at this stage.9 At the beginning of the 14th century, the outer friezes of monumental miḥrābs still displayed lotus flowers on borders.10 As far as textiles are concerned, the collection kept at Verona’s Castello Vecchio, from the tomb of Can Grande I, exhibits several fragments of brocaded satins from the 14th century, probably from Ilkhanid workshops.11 In metalwork, a chandelier made of a

8.

See for instance the borders of a series of lustre tiles decorated with phoenixes or dragons, presumably from Takht-i Sulayman; ill. in Komaroff, Carboni (dir.) (2002),  figs. 97 and 100. 9. Examples on lajvardina display an unnatural shrub with lozenge  leaves  and  lotus  flowers  with  six  petals;  see  V. Porter (1995), fig. 27. 10. See the frieze kept at the Doris Duke Foundation, ShangriLa  (Honolulu),  signed  by  Yūsuf  b. ʿAlī  b. Muḥammad  b. Abī Ṭāhir; the tile with the date (710/1310) is kept at  the Museum of Islamic Art in Cairo: Pope, Ackerman (dir.) (1938-1939), vol. X, pl. 725, E. 11. Id., vol. XII, pl. 1005. A: Red, lotus with eight petals plus heart; B: Green, lotus with twelve petals plus heart.

174 • YVES PORTER

copper alloy and dated 708/1308 shows medallions inlaid with six silver lotus petals.12 This first phase of  the appearance of the lotus goes together with the massive irruption of the Chinese imperial bestiary, represented by the dragon-phoenix couple. Of course, while these fantastic animals embodied the imperial couple in the Chinese symbolic system, in the Iranian world they obviously had other references.13 A second phase or form, probably appearing at a later date (first quarter of the 14th century?) shows the lotus flower accompanied by its leaf. A few examples on the so-called ‘Sultanabad’ ceramic might be the first Iranian bearers of this form; but it is above all  on underglaze painted tiles, produced in Khwarazm, that the leaves are to be noticed, often adopting a navicular form with four or five finely dented lobes;  the first example is probably the double cenotaph in  the mausoleum of Sayyid ʿAlāʾ al-Dīn in Khiva (ca. 1305);  it has echoes in the tiled panels of the tomb of Najm al-Dīn Kubrā at Kunya-Urgench, ca. 1330.14 As we pointed out earlier, the motif of the ‘Chinese’ lotus is rapidly simplified in its Iranian representations.  Moreover, in the illuminations from the pre-Mongol period (11th - 13th centuries),  the  floral  patterns  are  frequently oversimplified, and thus frankly tending  towards abstraction (arabesque). Some of the fleurons  in Qurʾanic illuminations from this long period could easily be seen as a very schematic rendering of lotus flowers.15 This taste for stylized motifs is not limited to manuscript illumination, since some metal or ceramic objects also display designs evolving towards a certain degree of ‘denaturalization’.16 In the arts of the book, the oldest examples of ‘Chinese’ lotuses probably go back to the late 13th or early 14th century. However, some milestones are doubtless irremediably lost.17 An illustration of the

Marzubānnāma, copied at Baghdad in 1299, shows, in a throne scene, a kind of lotus ‘bush’.18 In the same way, a juzʾ (or part) of a Qurʾan in sixty parts, coming from Iran or Anatolia and loosely dated between 1280 and 1320, presents in its illuminated cartouches this kind of lotus ‘bush’.19 The Qurʾan of Öljeitü copied  at Baghdad in 706/1306 offers, for its part, in a border of its frontispiece, some vegetal scrolls punctuated by lotus flowers.20 In the 1330s, we can see numerous lotus flowers on curtains and other objects depicted  in the illustrations of the Great Mongol Shāhnāma.21 Later on, examples dated after 1336 contain some vegetal branches in the midst of which the lotus flower can be recognized, along with a very simplified  leaf; the flowers are sometimes so schematic however  that it is rash to identify them with lotuses.22

Obviously the lotus plays an eminent role in the arts and iconography of the Indian subcontinent. Whether be it in Buddhist contexts, in which it occupies a central position, or on the attributes of the Hindu gods (such as Lakshmi, for instance), the lotus (padma) seems quite ubiquitous. Therefore, when the first builders of the Delhi sultanate settled on an architectural decor, such as the screen-wall in front of the prayer hall of the Quwwat al-Islam Mosque (early 13th century), they were naturally led to carve full blooming rose-like lotuses, together with friezes of lotus buds (figures 1-2). The patterns for the architectural decor of the early 13th century appear mainly to  derive  from  local  repertoires;  the  remarkable  exception is of course Arab epigraphy; in any case,  the palaeographic particularities of these inscriptions,

12. Id., pl. 1355. H. 32.5 cm, now at the Museum of Fine Arts, Boston (55.106): ill. in Komaroff, Carboni (dir.) (2002), fig. 154. 13. See Masuya (2002), 96-97. 14. See Soustiel, Porter (2003), 40-44. 15. See for instance the finispiece of the Qurʾan by Ibn Bawwāb (Baghdad, dated 391/1000). The Chester Beatty  Library, Dublin, MS. 1431, fol. 285r: ill. Ettinghausen (1962), 171. 16. See Covered goblet, probably Iran, late 13th - early 14th century. Silver, punched, engraved. State Hermitage Museum (Kub-364): ill. in Komaroff, Carboni (dir.) (2002), fig. 12;  see also lustre-painted dish, Iran (Kashan), dated 667/ 1268-1269.  The  David  Collection,  Copenhagen  (Isl.  95);  ill. in ibid., fig. 3. 17. At least two fragments of paintings, dating back to the early 14th century, now kept in the Diez albums (Staatsbibliothek zu Berlin, Diez A fol. 71, S.55 and A fol. 72, S.25), represent princely funerals; we can see a coffin, protected  on the back by a screen, on the top of which emerges a lotus. The same motif is repeated at the head of the coffin.  It is probable that in this case, the lotus is used as the

symbol of eternal life inspired by the Buddhist background of the Mongols: ill. in Komaroff, Carboni (dir.) (2002), figs. 87 and 122. Archaeological Museum Library, Istanbul (ms. 216), fol. 7r; ill. in ibid., fig. 200. The lotus bush appears at the  upper  right  corner;  although  the  flower  is  fully  recognizable, it rises from a stem with various branches and trifoliate leaves. Khalili Collection, QUR228, Rogers (dir.) (2009), no. 163. Leipzig, Staatsbibliothek; Pope, Ackerman (1971), vol. IX,  pl. 937B. See for instance Shāh Zav enthroned, Arthur M. Sackler, Smithsonian Institution, Washington (S1986.107). The golden throne is decorated with lotuses of various shapes (from nine to twelve petals); the attendant in the  foreground wears a burgundy robe decorated with a cloud collar and a lower band, both ornamented with lotuses;  ill. in Komaroff, Carboni (dir.) (2002), fig. 37. See Khalili QUR182 and QUR242, probably Shiraz, 13361354: ibid., nos. 162 and 164. Of course, the examples coming from the Mamluk world should be added to this survey.

The Indian lotus, a motif from the local repertoire

18.

19. 20.  21.

22.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 175

Figure 1 – Delhi, Quwwat al-Islam Masjid. [© Photograph: Y. Porter, 2007]

Figure 2 – Delhi, Quwwat al-Islam Masjid, screen of Iltutmish. [© Photograph: Y. Porter, 2007]

176 • YVES PORTER

especially in the sections dating back to the reign of Iltutmish, deserve special attention.23 This floral vocabulary perpetuates itself, with some refinement, under  the Khaljis (in particular on the ʿAlāʾ-i Darwāza, 1305);  in the following Tughluq era, during the 14th century, the forms of the lotus often continue to appear in their autochthon shape, as can be seen at the Begumpuri Masjid (ca. 1345; figure 3). This rapid survey shows that  the decorative elements inspired by the Far Eastern arts are almost non-existent in the decoration of the architecture of the Delhi sultanate, at least up to the 15th century. Consequently, we may ask where the lotus flowers and leaves that adorn the multiple  frontispieces of the Gwalior Qurʾan come from. Were these floral motifs conveyed through the arts of the  book of Ilkhanid Iran (or the Mamluk world), or are they elements directly integrated from Chinese imported artefacts? Besides Delhi, do the Chinese lotuses figure in the decorative repertoires of other  Indian sultanates?

China and India The relations between China and India are ancient, notably concerning the diffusion of Buddhism, but at the same time, they appear as somewhat fluctuating.24 In the domain of the arts, in the time of the sultanates, the data concerning these exchanges are extremely fragmentary. Already in 1964, Basil Gray particularly noted the evidence of the export of porcelain from China to the subcontinent;25 though Gray based his study mainly on textual data (the Tao-i-chih-lio, 1349, and Ibn Baṭṭūṭa’s Riḥla, 1325-1354), he pointed out: “what has survived must represent a small fraction only of the great flow of imported Chinese porcelain  which reached India, especially in the 14th and early 15th centuries”. The discovery, in 1961, of seventy-two pieces of Chinese porcelain in the gardens of the Firuz Shah Kutla palace, has considerably enlarged the samples of Chinese ceramics imported by the sultans of Delhi. Incidentally, at the moment of their

Figure 3 – Delhi, Begumpuri Masjid, main entrance. [© Photograph: Y. Porter, 2007]

23. See Blayac (2012).

24. On this subject, see for instance Sen (2003). 25. Gray (1964-1966), 21-37.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 177

discovery these items were identified as “Mughal porcelain, mostly damaged”!26 In her short but dense study dedicated to these findings, Ellen Smart has tried to place these objects back in the context of Firuz Shah Tughluq’s palace.27 In addition, a systematic examination of the annual reports of the Archaeological Survey of India has made it possible to enumerate an elevated quantity of findings of Chinese  ceramics (see annex 1); not only are the coastal sites  concerned by these imports, but also many inland cities all over the subcontinent. Their import probably started even before the establishment of the Delhi sultanate; at the time of the Mughals, their importance  increased significantly. For all that, it can be noted  that no finding of Chinese silk is reported from these  excavations (which is not very surprising if we take the climate of the subcontinent into account), even though the importation of de luxe textiles is highly probable at least since the time of the Sultans. Concerning the trade routes for these commodities, Tansen Sen writes: “With the Muslim conquest of northern and eastern India in the 13th and 14th centuries, the international consumption of and, therefore, the trade in Chinese porcelain witnessed a rapid growth. Under the Sultanate of Bengal (1368-1576), Chinese ceramics started entering India in large quantities through the ports of Sonargaon and Satgaon. From Bengal, the Chinese product was distributed overland to the Delhi sultanate (1206-1526). The discovery of Chinese porcelain fragments in coastal Bengal and as far inland as Fatehpur Sikri are indications of the increasing local demand for the Chinese commodity subsequent to the establishment of Islamic empires in India”.28 Indeed, the Ganges delta region represents a doorway for trade coming from the Far East through maritime routes: the river port of Sonargaon, South of Dacca (Bangladesh), together with Satgaon, on the Hooghly (a few miles North of Chandernagor), certainly played important roles in this traffic. In the  early 15th century, the Muslim Chinese sailor Zhen He made several trips between the Middle Kingdom and far Western lands, even reaching the east coast of Africa. From one of his voyages, in 1414, he brought back a giraffe to the emperor, as a gift from the sultan of Bengal.29

26. “In the course of clearance a large quantity of ornamental Mughal porcelain, mostly damaged, was found in the grounds of the Kutla Firuz Shah by the Gardens Branch of the Survey”: ASI (1964), 97. 27. Smart (1975-1977), 199-230. 28. Sen (2003), 185. 29. Wilson (1999), 123-124.

A certain Ma Huan, an interpreter during a Chinese diplomatic mission to the Bengal sultan in 1406, briefly described the capital of the kingdom he named  Pang-ko la (Bengal).30 Furthermore, a tomb excavated at Sekta (Manipur, Imphal district), dating from the late 14th or early 15th century, has yielded several pieces of Far-Eastern ceramics (China, Vietnam or Thailand?),31 evidencing thus the penetration of these  goods  far  into  the  inland  regions;  of  course,  besides the maritime routes, trade through the mainland also remained possible. What impact did these importations have on the artistic production of the Indian sultanates? In the field of ceramics, Chinese products certainly did  generate local imitations; moreover, on a plate published by the ASI illustrating ‘Chinese porcelains’ excavated from Champaner, several fragments are undoubtedly local productions.32 The direct transmission of the motifs in the arts of the book – or in architectural decors – is more difficult to trace. In  any  event,  the  lotus  flowers  and leaves  were  adopted into the decorative idiom of manuscript illumination, and appear thus frequently in the Gwalior Qurʾan manuscript, but the use of these patterns on other media seldom appears before 1400. On the other hand, it should be underlined that though vegetal motifs were more or less adopted,33 other designs, and primarily the fantastic bestiary (dragons, phoenix, ki’ling), appear to be totally absent, based on what we know of the artistic production of the Indian sultanates; these animals would probably not appear  before the time of the Mughals and would follow tastes steeped in those of Timurid Central Asia.

30. Bhattasali (1922), appendix III, Ma Huan’s account of the Kingdom of Bengal; cited by Syed Mahmudul Hasan, 61. 31. ASI (1994-1995), 59. 32. ASI (1969-1970), pl. XIXA, fragment upper right. For other examples: Champaner, Panch Mahals District (Gujarat): “A glazed ware showing a white sandy friable fabric looks almost like an imitation of porcelain ware. Imported pottery is represented by Chinese porcelain and Celadon Wares”, ASI (1974-1975), 15. Lal-Kot, District South Delhi: “The significant feature of Sultanate period was the use  of plain and painted glazed ware, both of sandy-friable with whitish gritty core (…) Chinese celadon is represented by a few sherds from the upper layers and its indigenous variety seems to have become very popular having sandy friable brownish white core and greenish surface”, ASI (1991-1992), 14. 33. Besides the Gwalior Qurʾan, we can see lotuses in the so-called Jainesque Shāhnāma, for instance in ‘Siyāvush’s  ordeal’  (Rietberg  Museum,  Zurich;  see  the  spandrels  over the pyre and the king’s ‘cloud collar’).

178 • YVES PORTER

the ‘chinese’ Lotus in the indian suLtanates before 1400: Pandua’s adina masjid The profusion in the Gwalior manuscript of lotus flowers, both with crossing or radiating petals, but  also of their leaves is really a matter for amazement. This abundance is all the more disconcerting as, from a chronological point of view, this manuscript seems very isolated. In architectural decoration, Chinese lotuses are extremely rare before the 15th century. Their presence in the ornamentation of Pandua’s Adina Masjid is, therefore, particularly surprising.

The emergence of Ilyasid power

second campaign of the Delhi sultan against Sikandar Shāh (1358-1390, Shams al-Dīn’s successor), allow him  to take possession of the province. “[Sikandar Shāh’s]  success and sense of power at being able to turn back the huge Delhi army is projected in the size of the mosque that he built in Pandua, the largest ever built in India”.37 The moment we are focusing on thus corresponds to this time of affirmation, at last free from political  binds, of a sultanate whose economic development, in particular with its ties with China, guaranteed its prosperity.

Pandua, capital of the sultanate

The history of the Bengal sultanate (1336-1576) does not benefit from contemporary chronicles. The  available data for writing this history are limited to coins, monumental inscriptions and the random chronicles written in the neighboring kingdoms.34 The major historical texts concerning the Bengal sultanate were written during the 19th century and thus do not constitute primary sources.35 Starting with its Muslim conquest, the province of Bengal, very far from the heart of Delhi’s sultanate, repeatedly demonstrated its will for independence. Ghiyāth  al-Dīn  Tughluq  (1320-1325)  was  able,  for  a  short period, to cause this region to submit but after his death Bengal was divided into two states. However, in 1342, after seizing power, Shams al-Dīn Ilyās  Shāh (1345-1358) united all Bengal once again under  one crown.36 The city of Pandua – qualified as ‘Hazrat’,  perhaps to differentiate it from Chhota Pandua, in the Hooghly district, or because of the shrines that it sheltered – became in 1342 the capital of the sultanate, by the will of the founder of the Ilyasid lineage. It would remain the capital of the kingdom, probably in competition with Sonargaon, until the Ilyasid restoration (1437-1487), when the seat of power was transferred to Gawr, the former Lakhnawati. Shams al-Dīn  Ilyās  Shāh  extended  his  domain  in  every  direction: to the South, towards Jajnagar (Orissa), to the North and North-East up to Kathmandu (Nepal) and Kamrup (Assam), to the West towards Tirhut (Bihar) and up to Benares. Some of these territories were lost in 1353, to the benefit of Firūz Shāh Tughluq,  in his first attempt to reconquer Bengal. Nor did the 

Confident in its new political independence, and  willing to affirm its newly conquered influence, the  Ilyasid dynasty felt the need to found a new capital. This was intended to break free from anterior vassalages but also to get rid of the rivalries between Lakhnawati and Sonargaon by developing in western Bengal the specific Muslim tessitura which would allow the dynasty to moor itself in this cultural sphere. The city, with its belt of ramparts, was built on the Eastern side of the Mahananda, on an ancient alluvial plain; it covered approximately 38 square  kilometres. This immense site is punctuated by a multitude of ponds (dighi), some of which might have been dug out much earlier, during the Pala-Sena period; however, very few archaeological explorations  have taken place on this site. On the northern rampart wall, an opening was cut where the ancient road to Dinajpur passed by; this break corresponds to one  of the city gates. The vestiges of another gate were also observed on the Southern wall, towards Malda. Besides the Adina Masjid, Pandua probably sheltered several dargāhs;  one  of  them  is  known  as  Shaykh  Jalāl al-Dīn Tabrīzī’s (d. 1244-1245), otherwise called  Bari Dargah. A second one is Nūr Quṭb ‘Ālam’s (d. 1415)  or Chhoti Dargah.38 At a later date, the city was the abode of the so-called Eklakhi Mausoleum (ca. 1425);  although devoid of inscriptions, this monument is allegedly devoted to the lineage of Rajah Ganesh. Nearby  is  the  commemorative  mosque  of  Quṭb-i  ʿĀlam, or Quṭb Shahi Masjid, built in 1582 by a certain  Makhdūm Shaykh. Incidentally, we find, in the notes of Salim, a mention of the Satargarh palace at Pandua (‘The Seven-storied Palace’), unhappily without further

34.  Notably  the  chroniclers  of  the  Tughluqs:  Baranī  [1862]  and ʿAfīf [1891]. 35. This is the case with Salim’s Riyazu-s-Salatin [1902]. 36. Bosworth (1980), 194.

37.  Hasan (2007), 13. This affirmation is a bit overrated since  ʿAlāʾ al-Dīn Khalji’s enlargement of the Quwwat al-Islam  Mosque  measures  220 × 106 m;  although  much  later,  Awrangzeb’s Badshahi Mosque in Lahore, for its part, is 178 × 170 m. 38. Salim [1902], 45, 97 and 134.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 179

details; in the same foot-note, and still in an incidental  way, he mentions paper-making in the city of Pandua. This occasional citation is all the more important as, besides  the  brief  quotation  from  Simi  Nīshāpūrī  (kāghaz-i bangāliʾyī), it is almost the only indication of paper-making in Bengal I have come across for the sultanate period.39 The Adina Masjid The so-called ‘Adina Masjid’ occupies a central position in the urban perimeter (figure 4); on ancient  maps, it is situated along the East side of the road that  runs  across  the  city;  the  current  road  now  leaves the monument on its West side.40 The Persian

term ‘ādina’ (Friday), qualifying the ‘congregational mosque’, is not frequent in India; to some authors,  the  choice  of  this  name  itself  is  significant  of  the  monument’s exceptionalness. Indeed, several authors particularly underline the ‘Persian’ characteristics of the mosque; one of these features is its plan, in the  form of a huge, pillared courtyard mosque, probably the only attempt of this kind in Bengal.41 Hasan adds an interesting anecdote, throwing a ray of light on the relations between Bengal and Persia, which is more grist to the mill: the Sultan Ghiyāth al-Dīn ʿAzam Shāh  (1390-1410), son of Sikandar, sent an incomplete verse to Ḥāfiẓ of Shiraz, in order for the great poet to  complete it; Ḥāfiẓ is said to have accepted the sultan’s  request.42 Beyond the anecdote, this fact shows a

1: Access ramp; 2: Four-pillared entrance hall; 3: Foundation inscription; 4: Axial nave; 5: Hypothetical localization of Satargarh palace.

Figure 4 – Hazrat Pandua (West-Bengal): Plan of Adina Masjid area. [© Google earth, CNES/Astrium, 2015]

39. Ibid., 45 n. 2; regarding Bengal paper, see Y. Porter (1985),  187. Hunter (1978), mentions Bengal paper from more recent periods, notably at Autsahai, in the center of the state; he notes the peculiarities of the moulds made in this  region, the only ones in India to be made of bamboo, in the Chinese manner (107-109 and figs. 82, 83, 85); the place  where the reproduced Bengali mould is preserved is not mentioned, but could be the Dard Hunter Paper Museum, MIT. Neither Premchand (1995) nor Soteroiu (1999) mention paper-making in Bengal in the time of the sultanate.

40.  On the plans found in Hasan (2007), 63, fig. 3, drawn after  Beglar (1888), the road is indeed situated to the west of the mosque. 41. Ghosh (2006), 94. 42. Ibid., 62; see also Salim [1902], 40 n. 1. About this ghazal, see C.-H. de Fouchécour, ghazal  no. 218;  he  (wrongly)  attributes  the  dedication  to  Ghiyāth  al-Dīn  Tughluq  (d. 1325), 597; in all cases, it should be recalled that Ḥāfiẓ  probably died in 1390, i.e. the same year in which Ghiyāth  Shāh became sultan of Bengal.

180 • YVES PORTER

reciprocal knowledge of these two far distant regions (Fars and Bengal). This knowledge went along with a demand for recognition, since Sultan Sikandar Shāh  was almost the only sovereign in India to proclaim himself “the most perfect sultan of Arabs and Persians”.43 The eastern façade has forty-one identical arcades which provide its rhythm; the fact that no major axis  is enhanced here does not contribute to the monumental aspect of this side of the building. Moreover, the foundational inscription of the mosque is situated at the opposite side of the mosque, behind the main miḥrāb, on the exterior of the qibla wall, fully oriented towards the west (figure 5); the inscription provides  the name of the builder (Sikandar Shāh, 1358-1390)  and a date (Rajab 776/December-January 1374-1375), which is often considered as the beginning of the building works.44 The stone slab is framed, on three sides, by two rows of glazed bricks, turquoise and yellow;  this  example  is  probably  the  oldest  extant  case of glazed tiles in Bengal. Hasan gives a substantial yet critical synthesis of the various publications and readings, sometimes controversial, about this inscription. Indeed, the text is studded with several syntax and orthographical errors (frequent in Bengali monumental epigraphy), even though the calligraphy displays a high standard and is nicely carved. No signature of any calligrapher appears on the numerous inscriptions all over the monument, in spite of its unquestionable prestige. Nevertheless some authors attribute this epigraphy to Ghiyāth Zarrin Dast (whose qualification of bandi dargāh specifies that he formed part of the sultanian  workshop’s  staff);  this  artist  signed  a  foundation  inscription in a much more modest building, the Dargah of ʿAtāʾ Shāh at Dinajpur.45 The dimensions of the Adina Masjid are considerable: a huge rectangle whose long side (oriented northsouth) measures 157 m, while the east-west width is 95 m; besides these outstanding dimensions, which  make it the largest mosque in Bengal (for the record, in comparison the Begumpuri Masjid in Delhi is almost a square with sides of about 95 m), the building

solutions are not less impressive. The main axial arch (the vaulted section is now completely collapsed), at the centre of the prayer hall, distinguishes itself by its width; some authors even go so far as to compare  it with the great arch at Ctesiphon.46 On both sides of the axial nave, five rows of strong pillars form eighteen  aisles leaning against the qibla wall; each one has its  own miḥrāb; these secondary miḥrābs are embellished with terra-cotta decorations, each with different motifs and patterns, displaying an extraordinary variety in their composition. The axial nave leads to a monumental miḥrāb, made of a black basaltic stone, of an extremely fine grain, and beautifully carved  (figure 6).  Above  the  frame  of  the  niche,  a  large calligraphic frieze displays, in thulth-muḥaqqaq high-stemmed letters, Qurʾanic verses 18-19 from the 9th sura, widely used for inscriptions in mosques;  the main epigraph is doubled by another, written in small foliated kufic, which remains unread. Capping  the monumental inscription, a sculpted decor recalls the sarlowḥ (front-page ornament) of illuminated manuscripts; a trilobated central fleuron is flanked  by  two  unfinished  half  fleurons.  However,  we  can  observe that the stylistic unity, together with its cultural (and religious) references, are deliberately broken with the addition under the epigraphic band of re-used elements probably coming from a preIslamic building;47 these are very finely carved stone  pillars, which are here placed horizontally, forming a continuous band and curiously dedicated to the physical support of the inscriptions. To the right of the miḥrāb is a minbar;  it  is  also  made of the same material (black stone); some of the  sculpted elements in this specific quality of stone obviously  come  from  older  buildings;  among  these  re-used blocks, some, skilfully sculpted, were used to fill the raisers of the minbar stairs (figure 7). Although no trace of a previous building remains on the site, it is highly probable that Hindu (or Buddhist) temples  existed  in  its  vicinity;  thus  two  sculpted  basalt pillars are now displayed in front of Pandua’s Bari Dargah.48

43.  “[Sikandar Shāh] was the first Bengal sultan to style himself khalīfa (successor of the Prophet) in an inscription and the only one to emphasize his affiliation to Arabia  and  Persia  rather  than  to  India”,  Hasan  (2007),  63;  see  the transcription of the inscription in Annex 2. 44. Hasan offers a synthesis of the different readings of the date; it appears that this construction followed Firūz Shāh  Tughluq’s campaign in Bengal: ibid., 64-68. On this second Bengali campaign of Firūz Shāh, see ʿAfīf [1891], 144-160. 45. On the Dinajpur inscription, see Yazdani (1929-1930), 9-11. Although the town of Dinajpur is now included in Bangladesh, the South Dinajpur district, Gangarampur, where the Dargah of ʿAtāʾ Shāh is located, is for its part  on the Indian side of the frontier.

46. Hasan (2007), 69. See also Eaton (1996), 45. This author also affirms that the Adina is “the largest mosque ever  built in the Indian subcontinent”: id., 42. 47. Pika Ghosh points out the ambiguity of the equivalence in terms between ‘Islamic’ and ‘sultanate’, focused by the study of mosques, rather than the many unidentified  buildings; see Ghosh (2006), 95. 48. Reproduced in ibid., 98, figs. 9-11. Other pre-Islamic remnants did exist in the area: John Henry Ravenshaw took a photograph in the 1860s, at Pandua, of ‘Carved lintels’,  one  of  forty-five  prints  in  the  album  Gaur: Its Ruins and Inscriptions, British Library Online Gallery, 978 (31);  I  do  not  know  the  present  whereabouts  of  these  carved lintels.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 181

Figure 5 – Foundation inscription of Adina Masjid. [© Photograph: Y. Porter, 2010]

Figure 6 – Adina Masjid. Inscription on the main miḥrāb. [© Photograph: Y. Porter, 2010]

Figure 7 – Adina Masjid. Sculpted element re-used in the minbar. [© Photograph: Y. Porter, 2010]

182 • YVES PORTER

The lotus loggia Another architectural feature of the Adina Masjid is the tribune (or loggia), situated to the north of the axial nave; access to it is only possible from the outside  of the monument through a pair of doors cut into the qibla wall. From the exterior, a bending ramp rises from the street entry to the tribune; its entry is  monumentalized by a four-pillared vestibule.49 The importance given to this façade on the qibla wall will be particularly emphasized if we take into account that the old road used to skirt this specific wall of the building. We can assume that the royal palace of Pandua was situated just across the road, thus facing the qibla wall.50 Two small doors, one of them particularly  ornate  with  a  finely  chiselled  knotted-snake  pattern, were made with sculpted re-used materials in basalt and provide access to the tribune (figure 8).  This is sometimes termed as zenāna, or Takht-gāh-i sang-i namāzgāh-i bādshāhān wa Shāhzādigān (lit. ‘The stone-throne of the kings’ and princes’ prayer hall’).51 In fact, this tribune, which does not have direct access to the main prayer hall and was most likely closed by jālī, was certainly used for both functions; on the  other hand, it is clear that in addition to the axial nave with the main miḥrāb the place undeniably displays a serious decorative preoccupation (figure 9). The qibla wall on the loggia is adorned with three finely carved basalt miḥrābs; in 1810, William Francklin (1763-1839) drew several sketches and details of these.52 The central miḥrāb on the tribune opens with a  concave  niche  framed  by  a  seven-lobed  arch;  at  the centre of the niche a sculpted lamp can be seen;  the frame displays beautiful epigraphic bands (see quotations in annex 2). The whole is capped with a tympanum decorated with a frieze of trefoil-like fleurons  inserted  in  multi-lobed,  arched  compartments, over which stands a most unusual fan-shaped fleuron standing over a ground of floral scrolls. The  two side miḥrābs are also concave, but they open with a five-lobed arch; their interior is spread with  geometric interlace patterns; besides the epigraphic  frames, containing quotations from the Qurʾan in

an elegant script that hesitates between thulth and muḥaqqaq,53 they are both surmounted by tympanums (figures 10-12). The tantalizing motif composing their  ground is certainly unique for its time: a series of lotusflowered stems and leaves of Far-Eastern inspiration  radiate from the bottom of a central shamsa (circular motif, or ‘sun’), covering the entire remaining surface of these lunettes. Both lunettes are delicately carved, the motifs being drawn in a linear way, with very low relief over a perfectly flat ground; they offer similar  compositions, only slightly different in details, figuring  stems of multi-dented lobed petals and leaves in a variety of arrangements and shapes. These drawings seem to have been made with a point chisel; they call  to mind the ink sketches of illumination or embroidery motifs appearing, nervously yet firmly outlined  in black Indian ink, on the Istanbul ‘Saray albums’.54 The  presence  of  the  ‘Chinese-like’  lotus  flowers  and leaves in the tribune of the Adina Masjid is most surprising, since this motif, although very frequent in the arts of Ilkhanid Iran, is almost unknown in the decorative repertoire of the Indian sultanates before 1400. There is no doubt that this royal tribune (indifferently of the dedication to the male or to the female members of the ruling family) displays a very rich ornamentation. The idea for the Far-Eastern lotus composition could have been taken from Chinese silks but this inspiration could also come from manuscript  illumination;  the  shamsa form, in particular, seems derived from this domain of ornamentation. Unhappily, we don’t know anything about the codices produced or read in Bengal at that time. Besides, the quality of the epigraphs, together with the proximity of the ornamental motifs from the main miḥrāb and the royal loggia with illuminated manuscripts could be understood as a testimony to disappeared Bengali manuscripts, completely unknown for this period. The Adina Masjid is a spectacular monument; its  symbolic significance is probably still largely underestimated. Its considerable dimensions, together with its audacious architectonic solutions, probably indicate the will for affirmation or even the superiority of  the Ilyasid ruler to the detriment of the Delhi sultan.

49. Some authors interpret this entry pavilion as a mausoleum although no remnants of any funerary occupation are to be seen. See for instance the plans of Adina Masjid provided by Hasan (2007), fig. 31. 50. Concerning the Satargarh palace, see Salim [1902], 45 n. 2. 51. On this locution, see Prasad (1961), cited by Hasan (2007), 89. ʿAbid ʿAli Khan [1931] – mentioned by Hasan (2007), n. 16, 105 – qualifies this raised platform as “meant for  the ladies of the Harem”. 52. Some of these drawings are kept at the British Library (India  Office,  WD318-320-330)  and  can  be  seen  in  the  British Library Online Gallery.

53. The epigraphic program of the three miḥrābs is rather curious: the northern miḥrāb seems to combine short formulas with Qurʾanic  verses  (9:  20-22);  the  central  miḥrāb has a continuous yet truncated inscription (48: 27-29,  the  end  of  verse 29  is  missing);  the  southern  miḥrāb has the Ayāt al-Kursī (2:  255).  Given  the  significance of this verse, a central position could have been expected for this quotation (?). 54.  See for instance Roxburgh (2005), fig. 53: Topkapı Saray  Museum, H. 2152, fol. 86r, dated ca. 1400-1450.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 183

Figure 8 – Carved doorway to the loggia. [© Photograph: Y. Porter, 2010]

Figure 9 – The Royal Loggia. [© Photograph: R. Castinel, 2010]

184 • YVES PORTER

Figure 10 – Southern miḥrāb of the loggia, tympanum. [© Photograph: Y. Porter, 2010]

Figure 11 – Left part of tympanum in northern loggia miḥrāb. [© Photograph: Y. Porter, 2010]

Figure 12 – Right part of southern tympanum. [© Photograph: Y. Porter, 2010]

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 185

Thus, the luxuriousness of its decor, in which the engraved lotuses on the royal tribune participate, is the expression of the creative strength and the need for singularity that would enable the dynasty to materialize its power. In fact, this mosque crystallizes the variety and richness of cultural contributions, MiddleEastern, Far-Eastern, as well as local, at the confluence  of which it takes place. After the fall of Rajah Ganesh’s lineage (1414-1436), the restored Ilyasid dynasty decided to transfer the capital to Gawr, near the site of ancient Lakhnawati. This shift was accompanied by intense architectural activity. However, this re-founded capital lost part of the ambitions embodied by Pandua. The inspirations that had fashioned such remarkably elegant calligraphies, the delicacy of the decors or the nobility of the materials used at the Adina Masjid became at Gawr less audacious constructions, both in their dimensions and their architectural features. As if in counterpart,

bibLioGraPhy Abbreviations ASI

Indian Archaelogy. A review: An Annual Publication on Archaeological Reports of Archaeological Survey of India.

Sources ʿaBid ʿali khan (Sahib M.) [1931] Memoirs of Gaur and Pandua, edited and revised by H. E. Stapleton, Calcutta: Bengal secretariat Book Depot. ʿaFīF (Shams al-Dīn Sirāj) [1891] Tārīkh-i Fīrūz Shāhī, edited by M. V. Husain, Calcutta: Bibliotheca Indica. baranī (Ḍiyāʾ al-Dīn) [1862] Tārikh-e Firuz Shāh, edited by S. Ahmad Khan, Calcutta: Bibliotheca Indica. salim (Ghulām Husain) [1902] Riyazu-s-Salatin, translation by A Salam, Calcutta: Bibliotheca Indica.

References Beglar (Joseph D.) 1888 Archaeological Survey of Bengal Report, Calcutta: Bengal Secretariat Press.

the buildings were then adorned with remarkable polychrome glazed tiles. Henceforth, while we certainly can find among the floral motifs some elements  recalling the lotus flower, it is no longer in its ‘Chinese’  shape, nor it is now accompanied by its finely dented  leaf; on the contrary, the flower takes on an extremely  schematic design, almost tending to abstraction. Bengali architecture finds in its new Western capital  the path to an artistic personality, probably more ‘authentic’ and at the same time less torn by the hegemonic will that would lead to a quest for symbiosis between Iranian, Chinese and local influences, even  including their antagonisms. Meanwhile, in the Deccan, the diwān-i ʿām of Bidar palace (ca. 1463) displays instead some carpet-like dados over which lotuses blossom plentifully.55 55.  See Yazdani (1947); see also Crowe (1986), 86-91.

Bhattasali (Nalini K.) 1922 Coins and Chronology of the Early Independent Sultans of Bengal, Cambridge: Eng. W. Heffer Publication. Blayac (Johanna) 2012 Formation et histoire des premières sociétés indomusulmanes et indo-islamiques à travers les inscriptions arabes et persanes (viie-xive siècles), PhD diss., Paris, École pratique des hautes études, 2009, Lille: Atelier national de reproduction des thèses. Bosworth (Edmund C.) 1980 The Islamic Dynasties, Edinburgh: University Press, reprint. crowe (Yolande) 1986 “Coloured tilework”, in G. Mitchell (dir.), Islamic Heritage of the Deccan, Bombay: Marg Publications, pp. 86-91. 1999 “Jeux de lotus dans la céramique du monde musulman : apparition et adaptation d’un motif chinois”, in Autour du lotus, Geneva: Musée Ariana, pp. 17-22. eaton (Richard M.) 1996 The Rise of Islam and the Bengal Frontier, 1204-1760, Berkeley: University of California Press. ettinghausen (Richard) 1962 Arab Painting, Geneva: Skira.

186 • YVES PORTER

ghosh (Pika) 2006 “Problems of reconstructing Bengali architecture of the 14th-16th centuries”, in A. N. Lambah, A. Patel (dir.), The Architecture of the Indian Sultanates, Mumbai: Marg Publications, pp. 92-103. gray (Basil) 1964-1966 “The export of Chinese porcelain to India”, in Transactions of the Oriental Ceramic Society 36, pp. 21-37. hasan (Perween) 2007 Sultans and Mosques. The Early Muslim Architecture of Bangladesh, New York: I. B. Tauris. hunter (Dard) 1978 Papermaking. The History and Technique of an Ancient Craft, New  York:  A. A. Knopf,  1943;  reprint, New York: Dover Publications. kadoi (Yuka) 2009 Islamic Chinoiserie: The Art of Mongol Iran, Edinburgh: Edinburgh University Press. komaroFF (Linda), carBoni (Stefano) (dir.) 2002 The Legacy of Genghis Khan: Courtly Art and Culture in Western Asia, 1256-1353, exhibition catalogue, New York: Metropolitan Museum of Art.

Prasad (Munshi S.) 1961  “Ahwāl-e Gaur wa Pandua”, in Ahmad Hasan  Dani (dir.), Muslim Architecture in Bengal, Dacca: Asiatic Society of Pakistan, pp. 1-42. Premchand (Neeta S.) 1995 Off the Deckle Edge. A paper-making journey through India, Bombay: The Ankur Project. rawson (Jessica) 1984 Chinese Ornament: The Lotus and the Dragon, London: British Museum Press. rogers (J. Michael) (dir.) 2009 Arts de l’Islam, chefs-d’œuvre de la collection Khalili, exhibition catalogue, Paris: IMA/Hazan. roguet (Didier) 1999 “Le lotus botanique et ethnobotanique”, in Autour du lotus, Geneva: Musée Ariana, pp. 3-10. roxBurgh (David J.) 2005 The Persian Album, 1400-1600, from Dispersal to Collection, New Haven: Yale University Press. sen (Tansen) 2003 Buddhism, Diplomacy, and Trade: the realignment of Sino-Indian relations, 600-1400, Honolulu: Association for Asian Studies and University of Hawai’i Press.

masuya (Tomoko) 2002 “Ilkhanid courtly life”, in Linda Komaroff, Stefano Carboni (dir.), The Legacy of Genghis Khan: Courtly Art and Culture in Western Asia, 1256-1353, exhibition catalogue, New York: Metropolitan Museum of Art, pp. 74-103.

smart (Ellen S.) 1975-1977 “Fourteenth-century chinese porcelain from a Tughlaq palace in Delhi”, in Transactions of the Oriental Ceramics Society 41, pp. 199-230.

medley (Margaret) 1998 The Chinese Potter: A Practical History of Chinese Ceramics, 3rd edition, Oxford: Phaidon Press.

soteroiu (Alexandra) 1999 Gift of Conquerors. Hand Papermaking in India, Middletown (NJ)/Ahmedabad: Grantha and Mapin.

PoPe (Arthur U.), ackerman (Phyllis) (dir.) 1971 Survey of Persian Art from Prehistoric Times to the Present, 1938-1939, 3rd edition, London/ New York/Tokyo: Oxford University Press/ Meiji Shobo, 6 vols.

soustiel (Jean), Porter (Yves) 2003 Tombeaux de paradis: le Shâh-e Zende de Samarcande et la céramique architecturale d’Asie centrale, Saint-Rémy-en-l’Eau: Monelle Hayot.

Porter (Venetia) 1995 Islamic tiles, London: British Museum Press for the Trustees of the British Museum. Porter (Yves) 1985 “Un traité de Simi Neyshâpuri, artiste et polygraphe”, in Studia Iranica 14, no. 2, pp. 187-198.

wilson (Samuel M.) 1999 The emperor’s giraffe and other stories of cultures in contact, Boulder (CO): Westview Press. yazdanI (Ghulam) 1929-1930 “Some inscriptions of the musulman kings of Bengal”, in Epigraphia Indo-Moslemica 12-13, pp. 9-13. 1947 Bidar: Its History and Monuments, London: Oxford University Press.

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 187

annex 1: chinese ceramics found in india, from: inDian archaeology, a survey (1955-2001) andhra Pradesh

Amaravati, Guntur District - celadon, Period III, 9th-10th century: ASI 1958-1959, 5 Golconda - large number of Ming porcelain pieces: ASI 19701971, 1 - many dishes and cups of Ming porcelain: ASI 19711972, 2 - sherds of celadon, Qutb al-Mulk period (1494-1543): ASI 1972-1973, 1-2. - a rich collection of celadon ware: ASI 1973-1974, 5 Motupalli District, Prakasam District - early Ming celadon: ASI 1973-1974, 7

assam

Ambari, Gauhati, Kamrup District - celadon (12th-14th century): ASI 1968-1969, 3 - a few sherds of celadon, Phase III, 7th-13th century: ASI 1970-1971, 4 - celadon: ASI 1987-1988, 9 - celadon: ASI 1988-1989, 6

bihar

Rajmahal, Jhimjhimia-Kalisthan, Sahebganj District - ASI 1987-1988, 12-13 Jhimjhimia-Kalisthan - ASI 1988-1989, 9

deLhi

North and West - ASI 1991-1992, 16 Kutla Firuz Shah - large quantity of porcelain, Mughal – (sic!): ASI 19611962, 97 and pl. CXXXVIII B Lal Kot - ASI 1991-1992, 14-15 - ASI 1992-1993, 12 Purana Qila - celadon, Cheng Hua porcelain, Mughal, 15th century: ASI, 1969-1970, 5, pl. XVIII A -  a few Ming porcelains (with coins, one of Muḥammad III Tughluq, 1390-1393?): ASI 1970-1971, 11 - sherds of celadon and porcelain: ASI 1971-1972, 7 and pl. XX A Salimgarh - ASI 1994-1995, 6

Goa

Chandor, South Goa - ASI 2000-2001, 25

Velha Goa - ASI, 1983-1984, 17 - ASI 1992-1993, 14 St Augutin Church - “a ceremonial porcelain pot (bowl)”: ASI 1990-1991, 5 - blue and white porcelain, 16th-17th century: ASI 1991-1992, 18 - ASI 1995-1996, 7 - ASI 1996-1997, 11 - ASI 1997-1998, 19 - ASI 1998-1999, 4 - ASI 1999-2000, 17

Gujarat

Anhilwad Pathan, Mahesana District - Chinese porcelain and celadon ware, 12th-13th century?: ASI, 1978-1979, 68 Banas Kantha District, Kaira and Panch Mahals, around Champaner - blue and white porcelain dishes, celadon ware: ASI 1974-1975, 12 Champaner - Chinese porcelains, some with Chinese characters: ASI, 1969-1970, 7 and pl. XIXA-B - a few porcelains, some with inscriptions in Devanagari: ASI 1970-1971, 15 and pl. XXXV A-B-C - porcelain and celadon ware: ASI 1972-1973, 12 Residential area - Chinese porcelain and celadon ware: ASI, 19741975, 15 Daman, Moti Daman - celadon: ASI 1985-1986, 107 and 184 Vadapadra, Vadodara District - celadon, medieval: ASI 1975-1976, 16

haryana

Thanesar, Kurukshetra District - ASI 1987-1988, 30, pl. XIIIA Harsh ka-Tila, Thanesar - Mughal context?: ASI 1988-1989, 23 Harsh ka-Tila, Thanesar: ASI 1990-1991, 22

karnataka

Fort Mirjan, Uttara Kanara District - ASI 1999-2000, 76 and pl. 62 Hampi, District Bellary - porcelain: ASI 1975-1976, 20 - China ware bearing Chinese scripts: ASI 1978-1979, 45 and pl. XIV B - China ware: ASI 1979-1980, 33

188 • YVES PORTER

- ASI 1983-1984, 38 and pl. 34 - celadon: ASI 1983-1984, 39 - ASI 1984-1985, 28 - ASI 1985-1986, 40 - ASI 1987-1988, 36, pl. XXB - ASI 1991-1992, 41 - ASI 1992-1993, 41 - ASI 1995-1996, 27 - ASI 1996-1997, 39 - ASI 1997-1998, 71 - ASI 1998-1999, 62 Srirangapatna, District Mandya - ASI, 1999-2000, 69

keraLa

Cheraman Paranbu (Cranganur), Trichur District - a few porcelain pieces, 13th-16th century: ASI 19681969, 10 Fort Cochin - porcelain, Portuguese marks: ASI 1986-1987, 46 Cranganor, Trichur District - Chinese celadon and Chola coins: ASI 1969-1970, 15

madhya Pradesh

Bhojasala, Dhar District - porcelain fragments, pre-Mughal: ASI, 1972-1973, 14 Kotwar, Morena District - ASI 1996-1997, 65

maharashtra

Aurangabad - ASI 1984-1985, 149 Bahai, East Kandesh District - celadon, Period IV, 1300-1700 / Mughal? Bahmanid?: ASI 1956-1957, 18 Bhokardan, Aurangabad District - celadon: ASI 1973-1974, 20 Daulatabad, Aurangabad District - ASI 1984-1985, 51 - ASI 1985-1986, 57-58 - ASI 1986-1987, 60 - ASI 1988-1989, 46 Kolaba District: - porcelain, 16th-17th century: ASI 1977-1978, 39 Nevasa, Ahmadnagar District - celadon, Period VI, medieval: ASI 1955-1956, 11 Vetalwadi, Aurangabad District Chinese glazed ware: ASI 1988-1989, 56

maniPur

Sekta, Imphal District - ASI 1994-1995, 59

orissa

Barabati Fort, Cuttack District - ASI 1989-1990, 77

- ASI 1992-1993, 81 - ASI 1994-1995, 60 and pl. XXXIV B - ASI 1996-1997, 82 and pl. XXII B Ganjam District - ASI 1984-1985, 57 Kalkhatapatna, Puri District - ASI 1984-1985, 59 - celadon: ASI 1994-1995, 61-62 and pl. XXXV A Ratnagiri, Cuttack District - a few porcelain pieces, 8th-13th century: ASI 19581959, 34

tamiL nadu

Arikamedu, Pondicherry: - ASI 1991-1992, 90 Attivakam, Chengai Anna District - porcelain, 17th century?: ASI 1989-1990, 95 Gangaikondacholapuram, Tiruchirappalli District - ASI 1983-1984, 80 Idindakarai (Cape Comorin) - porcelain: ASI 1972-1973, 31 Kanchipuram, Chingleput District - a few celadon pieces, Period III, 9th-16th century: ASI 1969-1970, 35 Kovalam - ASI 1992-1993, 116 Kunattur, Chingleput District - porcelain, 12th century?: ASI 1956-1957, 34 Madras District - ASI 1992-1993, 116 Neman, Chengai Anna District - ASI 1989-1990, 96 Padavedu - ASI 1992-1993, 89 Tirunelveli-Kattaboman District - celadon: ASI 1988-1989, 81

uttar Pradesh

Agra Mahtab Bagh - Chinese and glazed Mughal pottery: ASI 19791980, 74 Atranjikhera, Etah District - celadon, Period VII: ASI 1979-1980, 76 Bamnipuri, Rampur District: - sherds of Chinese pottery, medieval: ASI 19711972, 80 Bhirchhabili Tila, Sikri District - ASI 1999-2000, 168 Fatehpur Sikri (FPS-II) - Chinese porcelain: ASI 1977-1978, 51 Structure 3 - Chinese porcelain: ASI 1978-1979, 56 - Chinese blue and white porcelain and celadon, Mint: ASI 1979-1980, 72 - ASI 1983-1984, 84, pl. 54 - ASI 1984-1985, 84

LOTUS FLOWERS AND LEAVES, FROM CHINA AND IRAN TO THE INDIAN SULTANATES • 189

- ASI 1985-1986, 73 and 75 - ASI 1986-1987, 74 Kannauj, Farrukhabad District - ASI 1986-1987, 123

west benGaL

Manoalkot, N. Burdwan District - porcelain, context East India Co.?: ASI 1987-1988, 114

maLdives

Landu -  celadon: ASI 1986-1987, 176, fig. 10.

annex 2: inscriPtions from the adina masjid Foundation inscriptions: ‫امربنيأ العمرة هذا المسجد الجامع في ایام الدولة السلطان ابن السلطان العظ[ي]م العالم العادل‬ ‫االكرم اكمل السالطين العرب و العجم الواثق بتائد الرحمن ابو المجاهد سلطان سكندر شاه‬ ‫ابن الياس شاه السلطان خلد خالفته الى اليوم الموعود كتبه في التاریخ رجب‬ ‫سنه ست [و] سبعه[ین] و سبعمایة‬ ‘This  Jāmaʿ Masjid was ordered to be built in the days of the reign of the great Sultān, the Wisest, the most Just, the most Liberal, the most Perfect of the Sultans of  the Arabs and Persians, who trusts in the assistance of the Merciful, Abū al-Mujāhid  Sikandar Shāh, the Sulṭān, son of Ilyās Shāh, the Sulṭān, may his reign (khilāfa) continue until the Day of Promise. Written in Rajab in the year seven hundred seven[ty] six (776/December-January, 1374-1375)’: figure 5. Inscriptions on the raised platform Northern miḥrāb: ‘Qāla Allāh taʿālā ʿizza min qāʾil wa jalla min mutakallim* aʿūḏu billāh min shayṭān al-rajīm* Inna Allāh huwa al-samīʿ al-ʿalīm* Bismillāh al-raḥmān al-raḥīm*’. Sura 9: 20-22: figure 11. Central miḥrāb: Sura 48: 27-29 (the ayāt 29 stops at riḍwānān, seeking the grace of God and His satisfaction; in the rectangle over the arch: Sura 33: 56. Southern miḥrāb: Ayāt al-Kursī (2: 255): figures 10 and 12.

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? Manuscript W563 at the Walters Art Museum, Baltimore Simon rettig

(Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian Institution, Washington, DC)

Abstract Long seen as a Timurid product, the sumptuous Qurʾan W563 from the Walters Art Museum actually originated in India, most probably in the Gujarat region around 1430-1450. Though it shares numerous stylistic and decorative features with the ‘Gwalior Qurʾan’, it differs nonetheless in its use of the cursive muḥaqqaq script as well as of some illuminated forms and motifs whose stylistic roots are to be found in Timurid Iran. Qurʾan W563 encapsulates the development of parallel artistic trends in manuscript production that occurred in Sultanate India during the 15th century. Alongside with the ‘Gwalior Qurʾan’ but also in opposition to it, it also demonstrates the complexity of the selection process for formulas that pinpoint local, regional, and extra-regional artistic productions. Due to its initial ‘Timurid-like’ appearance, the manuscript W563 was ultimately sent to Istanbul as a diplomatic gift a few decades later, as is indicated by the seal of the Ottoman sultan Bāyezīd II (r. 1481-1512).

Résumé Le coran W563 du Walters Art Museum de Baltimore : un manuscrit d’apparence timouride en réponse au coran de Gwalior ? Longtemps considéré comme une production timouride, le somptueux coran W563 du Walters Art Museum vit en réalité le jour en Inde, probablement dans le Gujarat, aux alentours de 1430-1450. Il partage de nombreuses caractéristiques stylistiques et iconographiques avec le coran de Gwalior, mais à la différence du manuscrit de Gwalior il est copié en muḥaqqaq et montre certaines formes et motifs enluminés dont les racines stylistiques sont à chercher dans l’Iran timouride. Le coran W563 synthétise le développement des tendances artistiques parallèles qui se sont développées en même temps dans la production manuscrite de l’Inde des sultanats au xve siècle. Comme le coran de Gwalior, mais d’une manière différente, il montre la complexité du processus de sélection des formules qui illustrent des productions artistiques locales, régionales et extra-régionales. Parce qu’il présente une « apparence timouride », le manuscrit W563 fut envoyé à Istanbul, quelques décennies après sa création, comme cadeau diplomatique, ce qui est clairement indiqué par l’existence du sceau du sultan ottoman Bāyezīd II (r. 1481-1512) dans ses pages.

Le coran de Gwalior. Polysémie d’un manuscrit à peintures, sous la direction d’Éloïse Brac de la Perrière et Monique Burési, 2016 — p. 191-205

192 • SIMON RETTIG

Despite recent research and publications, Sultanate Indian Qurʾan manuscripts remain little known. It still appears difficult to offer a general overview of the developments of Islamic book production in pre-Mughal India and to map them out with a degree of accuracy both on regional and local levels.1 This issue is further complicated by the scarcity of dated and located copies, of which the most famous and earliest known illuminated example is the ‘Gwalior Qurʾan’ now in the collections of the Aga Khan Museum.2 The latter may be considered as the ‘mother of Indian Qurʾans’ for it essentially includes the codicological, calligraphic, and decorative characteristics of pre-Mughal Indian Qurʾan manuscripts until the mid-sixteenth century: the main body of the text is copied in bihārī script with marginal gloss in naskhīdiwānī; two or three frames constitute the page layout;  the illuminations show a rich and peculiar palette, a repertoire of motifs and decorative layouts emanating from diverse sources, regional and extra-regional.3 However, a few Qurʾan manuscripts attributed to the 15th century show distinctive features, notably the use of a codified cursive script, wrongly defined as an Indian version of thuluth, for the copying of the Qurʾanic text.4 Among them is an undated copy kept in the Walters Art Museum’s collections that has been recently brought back to scholars’ attention.5

1. 2.

3. 4.

5.

For a recent study on manuscript production in preMughal India, see Brac de la Perrière (2008). Toronto, Aga Khan Museum, Accession number AKM00281. The manuscript in the present essay is denominated as the ‘Gwalior Qurʾan’ or the AKM Qurʾan. The completion date is provided in the colophon: 7 Dhū al-Qaʿda 801/ 11 July 1399 on fol. 550v; ibid., n. 31, 289 and Brac de la Perrière, Chaigne, Cruvelier (2010), 115. The sole other Qurʾan with an informative colophon is a copy dated 888/1483  now  in  the  Bijapur  Archaeological  Museum;  see Brac de la Perrière (2008), n. 32, 290. Brac de la Perrière established a useful list of known Qurʾans  from  the  Sultanate  period;  ibid, 81-86 and “Annexe 1 Groupe 5”, 289-296. Losty (1982), 38-39 and Brac de la Perrière (2008), 142. Brac de la Perrière convincingly demonstrates that Losty’s “Indian thuluth” is actually muḥaqqaq script. Two manuscripts from Sultanate India are copied in muḥaqqaq: one is in the British Library (Add.18163), the other was presented for auction by Christie’s in 2000; for complete  references, see Brac de la Perrière (2008), n. 47, 296 and n. 46, 295. On the British library copy, see also Wright (1996-1997), 8-12. Baltimore, The Walters Art Museum, Accession number W563. The manuscript has been entirely digitized and is now available online: http://www.thedigitalwalters.org/ Data/WaltersManuscripts/html/W563/. In the present article, the manuscript is referred as the WAM Qurʾan or the W563 Qurʾan. I wish to thank the DFG – Emmy-Noether Junior Research Group ‘Kosmos-Ornatus’ led by Very Beyer at the Freie Universität zu Berlin, which financially supported a trip to Baltimore in order to study the manu

Although it is also richly ornamented, this manuscript somehow differs visually from the illuminated “kaleidoscope” of the Qurʾan of Gwalior.6 Indeed, the WAM Qurʾan seemingly offers a striking unity in terms of formal structures as well as a decorative homogeneity to such an extent that the manuscript has long been thought to have originated in Timurid Iran.7 Through the analysis of the inner organization of the manuscript and its illuminated program, this essay seeks to demonstrate how the aesthetic conception of the W563 and Gwalior Qurʾans diverge radically from one another. Although these two copies undoubtedly benefited from princely patronage, they betoken  two distinctive visual conceptions of the Qurʾan manuscript in Sultanate India. Both the WAM and AKM Qurʾans may certainly illustrate the significant changes  subsequent to the fall of the Delhi Sultanate in 1398 and the rise of Muslim princely courts in India in the 15th century. The Timurid invasions provided a new impetus for the development of cultural and artistic patronage, whose models were mainly to be found in Iran. Nevertheless, India in the 15th century was more than a mere ‘cultural extension of Persia’. In fact, India found itself amid a complex network of commercial, diplomatic and artistic exchanges. As a result, commodities and luxury goods were disseminated on a wide scale, a circulation that the WAM Qurʾan instantiates: the presence of the seal of the Ottoman sultan Bāyezīd II (r. 1481-1512) surely indicates that this luxury copy of the Qurʾan was ultimately sent from India to Istanbul as a diplomatic gift. The aesthetic quality of the manuscript and its Timurid appearance in particular were more likely to appeal to the Ottoman ruler than an object like the Qurʾan of Gwalior. The Walters Art Museum Qurʾan is of relatively large dimensions: the folios in their current state measure 40 × 31 cm, which means that a bifolio has a format of 40 × 62 cm. The manuscript may have even been slightly larger initially, as cut marginal notes prove that the copy was trimmed, probably when the present Ottoman binding was added.8 The central zone containing the Qurʾanic text and delineated

6. 7. 8.

script in April 2012. I am also very indebted to Amy Landau and the Museum’s staff for their warm welcome and accommodation. Brac de la Perrière, Chaigne, Cruvelier (2010), 114-123. Simpson (2000), 98. The manuscript may have reached Istanbul unbound. The binding that still covers the manuscript can be attributed on stylistic grounds to the reign of Sultan Bāyezīd II (r. 1481-1512). On bindings in Sultanate India,  see Brac de la Perrière (2008), 109. On Ottoman bindings and for a comparison with the cover of W563, see Raby, Tanındı (1993), 81-104.

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 193

by the ruling lines (jadwal) has a homogenous size throughout the volume and shows only some slight variation: 26 to 26.5 cm tall by 19 to 19.7 cm wide. The British Library’s copy has the exact same dimensions. The W563 manuscript has a larger format than the Gwalior Qurʾan, but both copies show a similar ratio of approximately 1.40.9 On each page, nine lines of the Qurʾanic text are displayed.10 Lastly both manuscripts contain approximately the same number of folios: 554 for the W563 Qurʾan and 552 for the Gwalior Qurʾan. The WAM manuscript is made of a relatively thin, smooth, and generously finished cream paper.  Wire-lines are largely visible and twenty to thirty wire-lines extend over 30 mm. Moreover, an examination of each folio reveals that wire-lines are all arranged parallel to the spine of the codex. The apparent homogeneity of this material leaves little doubt about the fact that the paper used in the making of the WAM Qurʾan originates from a single place of production, although it is impossible in the present state of knowledge to determine it with more accuracy.11 Nevertheless an element so far unnoted might shed new light on the paper industry and book  production  in  fifteenth-century  India.  A  thin  ‘fold-like’ line is noticeable running horizontally through the middle of the bifolios (figure 1). Often one can count up to ten consecutive folios on which the line is visible, meaning that the manuscript may be – partially at least – constituted of quinions  (quires  of  five  bifolios).  Furthermore,  of 

the manuscript’s 277 bifolios, about a third – i.e. 90 bifolios or 180 folios – have this ‘fold-like’ line. It may in fact indicate that the original sheet of paper was initially composed of three superimposed bifolios, one located above and one beneath the median bifolio that bears the central horizontal “fold-like” line. It could thus be possible to determine approximately – since the manuscript has been trimmed - the original size of the sheets: approximately 120 × 62 cm, a format slightly different from the baghdādī format, traditionally given as one of the largest.12 A putative explanation for the presence of this fold line on large-sized manuscript leaves may be provided: the fold could actually be the trace left by the line upon which the sheet was spread until it dried. Because the paper was resized thereafter, it is now difficult  to identify the fold as a groove.13 The sheets of paper in the Gwalior Qurʾan do not show any trace of folding  but  the  format  is  also  smaller;  the  sheets  may  have completely dried on moulds or the operation may have been completed by placing the sheets upon plaster walls, following traditional techniques. This issue requires further research; nonetheless, it  indicates that the production of large sheets in the Eastern Islamic world in the 15th century may have not been as rare as we have thought. Although the Indian origin of the W563 Qurʾan’s paper cannot yet be formally ascertained, the existence of such large sheets testifies to the degree of technology reached  by papermakers after 1400. The drying line may ultimately point to a local Indian production.

Figure 1 – ‘Fold-like’ line, fol. 67r (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum]

9.

A ratio comprised between 1.386 and 1.442 corresponds to a rectangle of dimensions a × a √2. See Déroche (dir.) (2000),  181. For ratios and proportions in Sultanate Indian Qurʾans, see Brac de la Perrière (2008), 119. Interestingly, on the thirty-one examples Brac de la Perrière analyzed, only the Gwalior Qurʾan presents this ratio of ca. 1.40. 10. With the exception of the folio 148r that has ten lines of text. The first line opening the eighth juzʾ was actually added in the upper part. 11. The most recent overview on papermaking in Sultanate India is Brac de la Perrière (2008), 94-101. See also Porter (1994), 17-21.

12.  Al-Qalqashandī (d. 1418) gives the size of baghdādī sheet as corresponding to approximately 110 by 70 cm; cited  in Bloom (2001), 53. 13. I initially suggested that the fold could indicate how sheets of paper were kept – folded in two – in the papermaker’s storage or in the workshop before use, information about which sources are silent. I am indebted to Francis Richard for suggesting the “drying line” direction. On the Indian tradition of line drying, see Loveday (2001), 42.

194 • SIMON RETTIG

Figure 2 – Muḥaqqaq script, fol. 31r (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum]

Figure 3 – Muḥaqqaq script, fol. 477v (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum]

Figure 4 – Muḥaqqaq script, fol. 333r (detail), Add. 18163. [© British Library]

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 195

The most striking visual difference between the Gwalior Qurʾan and the WAM Qurʾan is first and foremost to be found in the calligraphic style used for transcribing the Qurʾanic text. The Gwalior Qurʾan was copied in bihārī script whereas the WAM manuscript presents a neat version of the muḥaqqaq style.14 Following the Gwalior copy’s prototype, most of the Qurʾans in India were written in bihārī script until late into the 16th century while only very few copies ascribed to India show other styles, namely one of the six so-called codified scripts (al-aqlām al-sitta).15 As long noted, muḥaqqaq is notably characterized by the pointed tails of sublinear letters (notably mīm, rāʾ, zāʾ and wāw) as well as by a strong linearity in the rhythm of the line counterbalanced by the verticality of the tall upstrokes of alif and lām (figure 2).16 An examination of the handwriting in the WAM manuscript tells us that only one scribe copied the manuscript for it shows a strong stylistic homogeneity in the shaping of letters. However, some particularities contribute to a more precise definition of the handwriting of the WAM Qurʾan’s unknown calligrapher and make it possible to differentiate his muḥaqqaq from other works. Vertical upstrokes often end with a thin and sharp serif and they are dissimilar to the ones in the British Library’s copy (figure 4). Moreover  the shaping of letters and the spacing on the page produce an unusual elegance and certain letters adopt peculiar trait. For instance, the initial and middle kāf has a horizontal upper stroke where it would normally present a downward inclination of a 25- to 45-degree angle from right to left. One may also note edginess and tension in the writing. Upstrokes are most of the time not perfectly straight and vertical, but rather trembling lines that

14. On the bihārī script, see Brac de la Perrière (2003), 86-89;  id. (2008), 132-137. Sheila Blair has also provided a more succinct  overview;  Blair  (2006),  386-389.  The  muḥaqqaq script was extensively used in the copying of Qurʾans in Iranian centers since the Ilkhanids (1256-1353) and in the Mamluk sphere. Many copies are also known for the Timurid period, although it was less much favored after 1450. On muḥaqqaq in the 14th century, see Roxburgh (2007), 41-51. For a reason that still needs to be determined, one illuminated folio (fol. 116v) in the Gwalior Qurʾan presents the text copied in muḥaqqaq. 15. Losty mentions a Qurʾan page copied in rayḥāni script that he attributes to fourteenth-century India but he omits to provide precise references; Losty (1982), 38. 16. The fact that the copyist added the vocalization and diacritics with a thinner reed pen is a distinctive feature of the muḥaqqaq style that seldom appears used for the thuluth style. On the characteristics of the muḥaqqaq and particularly the qualities that differentiate it from other styles, see the recent study by Mansour (2011), 139-149. Unfortunately, Mansour does not study in depth fifteenth-  century muḥaqqaq and notably the differences we might note in distinguishing the regional styles.

slightly slope down toward the left. As a result, the handwriting shows relative incongruity inasmuch as it strays from the formal stylistic qualities of the muḥaqqaq script. Finally, large parts of the text display improbable links between letters, a feature that is not usually inherent to the muḥaqqaq style (figure 3).  Although very elegant, these whirling linkages confer a unique character to the handwriting, a sense of fluidity in which letters and words appear fastened  to each other in a vibrant rhythm. Interesting parallels can actually be found in architectural inscriptions.17 Muḥaqqaq is virtually absent from epigraphic friezes that adorn buildings in Timurid and Turkman Iran as well as from the Mamluk architecture. But it was used to some extent in India as exemplified by an inscription dated 861/1467 in the  Deccan region of Khandesh.18 Two other inscriptions – one undated from Bidar made under the Bahmanid Aḥmad Shāh II (r. 1436-1458) and one dated 921/1515  originally from Raichur – show similarities with the WAM Qurʾan’s deviant script, though they are written in thuluth.19 The tails of the jīm, ḥāʾ and khāʾ, as well as of the ʿayn and ghayn are drawn up and inwards in the three cases. This treatment of letters may have led some scholars to confuse the WAM Qurʾan muḥaqqaq with the thuluth script. It may also be connected to chancery scripts, especially the so-called ṭughrā style used  in  Indian  courts.  Also  defined  as  convoluted  naskhī, it was used both in farmāns and royal edicts as well as for epigraphy in architectural patronage, notably under the Tughluqs (1320-1414) and contemporaneous local dynasties.20 The influence of various  styles on the muḥaqqaq script of the WAM Qurʾan makes it rather unique. The British Library and Christie’s Qurʾans, obviously from different hands, do not show any of the aforementioned peculiarities. The outline of the BL Add. 18163 muḥaqqaq in particular appears very balanced, firm and steady (figure 4).

17. Brac de la Perrière (2008), 129-132. While Islamic epigraphy in India has generally been rather well studied, the period between the fall of the Delhi Sultanate and the rise of the Mughal dynasty is still little known. On Indian epigraphy in Arabic scripts before 1400, see Blayac (2012). The exhibition catalogue Begley (1985) remains a useful reference, as do the issues of Epigraphia Indica. Arabic and Persian Supplements. 18. Rahim (1962), 67-68, pl. XXIa. 19. Begley (1985), 58-59 and 71. The two inscriptions are signed by émigré artists with Iranian nisba:  Mughīth  al-Qārī al-Shīrāzī and Ḥusayn b. Yūsuf al-Yazdī. For a list  of Persian artists’ signatures, see Brac de la Perrière (2008), Annexe III, 310-313. 20. Numerous Tughluq examples are given in Desai (1961), 26-34. See also another inscription from Bihar Sharif dated 807/1404-1405, produced under the reign of Ibrāhīm, a  Sharqi sultan of Jaunpur (r. 1402-1440); Begley (1985), 53.

196 • SIMON RETTIG

The calligraphic idiosyncrasy of the WAM Qurʾan places the manuscript on a similar level to the Gwalior copy: it firmly anchors these manuscripts in  Indian contexts of production. Another particularity of the Walters Art Museum Qurʾan lies in the extensive amount of marginal notes and inscriptions. The repetition of Qurʾanic phrases in the margins is enhanced by the use of different inks (blue, red, and gold) both for the letters and words and for the vocalization and reading signs. Parallel to the text in the central zone, the marginal Qurʾanic fragments are also copied in muḥaqqaq and as a result they visually expand the text displayed in the center of the page into the margin. Interestingly, the alternation of colors recalls the color scheme in the Gwalior Qurʾan. In the latter however this peculiar arrangement is restricted to the Qurʾanic text within the central zone whereas marginal notes are written in red and blue, while black ink is merely absent from the Gwalior Qurʾan.21 The marginal Qurʾanic excerpts in the WAM Qurʾan are also accompanied by extensive glosses concerning the various canonical readings (qirāʾāt).22 These comments in black ink as well as the red Persian interlinear translation are copied in a script called naskhī-diwānī, a peculiar style that was developed in India for noting marginal commentaries in Qurʾan manuscripts whose block of text is in bihārī.23 The handwriting here is particularly neat and shows a consistent homogeneity that indicates that the copyist had full mastery of the style. It may be also worth noting that the bihārī script seemingly absent from the WAM Qurʾan appears in fact only at the beginning of the volume on one of the illuminated opening double pages (fols. 7v-8r).24 It shows characteristics of the bihārī script, although some traits are further amplified, such as the angular  aspect of the letters and an apparent variation in the thickness of the stroke.

21. Only two lines are copied with black ink on fol. 208v. 22. The glosses emphasize the canonical readings of some of the seven qurraʾ and their disciples whose methods of recitation do not appear noted by a letter (the abjad system) above the Qurʾanic text in the central zone, namely: Nāfiʿ and Qālūn, Ibn Kāthīr, Abū ʿAmrū, Ḥamza, Al-Kisāʾī.  On the canonical readers and their followers, see Melchert, Asfaruddin (2004), 388-391. See also the essay by Alilouche and Esmailpour Qouchani in the present volume. 23. Brac de la Perrière (2003), 89-91 and id. (2008), 140. 24. Part of verse 6 of the second chapter (sūrat al-Baqara): ʿalayhim aʾanḏartahum am lam tunḏirhum; [As to those who reject faith,] it is the same to them whether thou warn them or do not warn them; [they will not believe] is copied in red ink eighteen times. Each verse fragment is followed by fī rawāyyati Qālūn on fol. 7v or Qālūn fī rawāyyati on fol. 8r (both meaning “in the way Qālūn says it”) noted in black  naskhī-diwānī.

Likewise great and meticulous care is given to the allographic components: the name of Allah is systematically written in gold with a black outline, and vocalization signs and the shadda are rendered in ultramarine blue, with the exception of the waṣla on the initial alif, which is in black ink. Similarly, the letters referring to the modes of reading and recitation are all executed in blue along with punctuation marks.25 The strong visual coherence in the overall aspect of the manuscript is further enhanced by the addition of gilded rosettes as verse markers and the ruling lines framing the zone of text, all identical through the volume. There is little doubt that the calligrapher and the illuminator of the WAM Qurʾan – if were they not one single artist – worked closely together in order to generate such a lavish work of constant uniformity. The even and balanced result is also perceptible in the rich decorative program. The W563 Qurʾan opens with a lengthy series of double framed illuminated folios. At least three double pages with uninscribed central roundels (shamsa) or almond-shaped medallions (toronj) adjoined with four round corner medallions were originally displayed at the beginning of the manuscript (on fols. 3r, 5v-6r, 6v-7r), though the first one is now missing its right  counterpart.26 Such purely ornamental opening compositions do not appear in the Gwalior Qurʾan. But they are to be found as early as the end of the 13th century in manuscripts whose production ranges from Turkey to India.27 Although they were initially meant to be functional above ornamental – they would either present the name of the patron or owner 25. The method of two canonical readers’ followers is extensively noted: al-Durī, follower of Abū ʿAmrū (given by the  letter ṭāʾ) and Warsh, follower of Nāfīʿ (given by the letter ḥāʾ). To a lesser extent are mentioned Ibn Ḏakwān follower  of Ibn ʿĀmir (letter mīm), Qunbul follower of Ibn Kathīr  (letter zāʾ),  Abū  Bakr  follower  of  ʿĀṣim  (letter  ṣād) and Khilād  follower  of  Ḥamza  (letter  qāf). Punctuation marks are noted by lām-alif (waqf mamnūʿ – the reader must not stop at the place indicated), ṣalā (waqf jāʾiz, al-waṣl awlā – the reader may or may not stop at the place indicated, but the first preference is not to stop), small jīm and the word qāf (waqf jāʾiz – the reader must stop at the place indicated). 26. A restoration of the manuscript probably occurred under the reign of the Ottoman sultan ʿUthmān II (r. 1618-1622),  as shown by an erased yet partially legible waqf statement on  fol. 3r.  On  that  occasion,  the  first  illuminated  folio  pendant to fol. 3r may have been removed and folios 3, 4 and 550 heavily mended. We can putatively further ascribe the insertion of fols. 1, 2 and 551 to this time. 27. Early Ilkhanid examples are provided in Chaigne (2012), 257-258. A copy of Arbaʿūn ḥadīth (Forty Sayings of the Prophet) attributed to fourteenth-century Ilkhanid Iran is an early example with corner-pieces instead of roundels (Chester Beatty Library, CBL Ar 4181); for an illustration,  see Wright (2009), 186, fig. 138. An undated copy ascribed  to Sultanate India and copied in bihārī script presents

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 197

of the work or bear the title and the name of the author(s) – in the case of the present manuscript, they convey a decorative meaning, yet with another fundamental function: they increase the lavishness and enhance the preciosity of the whole book. These pages surround two double pages marvelously introducing the names of the seven masters of canonic recitations and readings of the Qurʾan, each one associated with single letters according to the alphanumerical abjad system (fols. 3v to 5r). Lastly, the double-page frontispiece (fols. 7v-8r) that precedes the beginning of the Sacred Text offers a full-page decorative pattern with four ʿunwāns bearing the verses 77-80 of the fifty-sixth chapter, sūrat al-wāqiʿa.28 Like the Gwalior Qurʾan, the Walters Art Museum copy has a quadripartite inner division that is seldom seen in Qurʾan manuscripts outside India. It is characterized by illuminated sura openings on double pages.29

The more common division of the Qurʾan into thirty parts (juzʾ – pl. ajzāʾ) is marked as well. Each juzʾ has its first line written in gold muḥaqqaq (figure 5).30 Juzʾ beginnings are also highlighted in the margin: the word al-juzʾ is often written in white inside a small ʿunwān-like medallion with an upper framing border. It can be also occasionally displayed in a roundel or almond-shaped medallion, or simply written in muḥaqqaq in the margin with black, blue or gold ink.31 Sura titles are all inscribed inside an ʿunwān, the latter integrated in the zone of text with a marginal decorative medallion. Written in white thuluth or sometimes muḥaqqaq, they provide the name of the chapter and the number of verses it contains. The titles on the double pages additionally give the place of their revelation to the Prophet, either Mecca or  Medina.  One  should  note  that  the  angular  kufic  script often used for inscribing titles in illuminations

Figure 5 – Sura heading and juzʾ marginal, fol. 453r (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum]

a double-page frontispiece with a similar construction (Museum of Turkish and Islamic Art, TIEM 264); Istanbul  (2010), no. 92, 350-351. A slightly later Ottoman manuscript uses the composition as a finispiece, although the first  mandorla  conveys  the  colophon’s  text  (TIEM 224);  ibid. no. 100, 380. 28. Sura 56: 77-80: “That this is indeed a Qurʾan most honorable, in a Book well-guarded, which none shall touch but those who are clean: A Revelation from the Lord of the Worlds”. For an example from Mamluk Egypt dated 1440 and another one attributed to Timurid Herat around 1500, see James (1992), no. 13, 58-61 and no. 30, 116-117. 29.  The first chapter, sūrat al-Fātiḥa (fols. 8v-9r) and the first  verses of the second one (sūrat al-Baqara, fols. 9v-10r) constitute the first group. Then come the beginnings of  the seventh chapter (sūrat al-A‘rāf, fols. 137b-138a), the nineteenth chapter (sūra Maryam, fols. 274v-275r), the

thirty-eighth chapter (sūra Ṣād, fols. 408v-409r) and the last two ones facing each other (sūrat al-Falaq and sūrat al-Nās, fols. 549v-550r). See Brac de la Perrière, Chaigne, Cruvelier (2010), 117. 30. The conception of the Gwalior Qurʾan with one doublepage opening for each juzʾ in a single volume manuscript is rather unique; ibid., 115. The ‘Sam Fogg’ Qurʾan, albeit written in bihārī, displays the juzʾ opening in a similar fashion to the WAM Qurʾan: the first line is copied in thuluth, a change of calligraphic style that visually underlines the beginning of juzʾ. For an illustration, see Fraser, Kwiatkowski (2006), 108. 31. This may suggest that the calligrapher, after the completion of the copy and the making of the illuminations, went carefully through the text and noted each omission in the margin in muḥaqqaq script in such a way that it does not appear visually at odds with the rest of the marginal notes.

198 • SIMON RETTIG

is completely absent from the W563 copy.32 As for the  first  line  of  each  juzʾ, chrysography is used for writing the basmala at the beginning of each sura. In contrast to the AKM copy, the WAM Qurʾan is lavishly adorned with marginal inscribed vignettes. The function of these illuminated devices is mainly twofold: they can provide an indication of numerical divisions by number of verses, such as khams for groups of five verses or ʿashr for groups of ten.33 The vignettes can also indicate a sequential division. They mark parts within a juzʾ, like thulth (1/3), rubʿ (1/4), niṣf (1/2), thalatha arbaʿ (3/4), or of the whole work, such as ḥizb (1/60) and niṣf al-Qurʾān, the exact middle of the Qurʾanic text. Lastly a third category highlights bodily gestures to perform while reading the text: the most recurrent is rukūʿ (head bowing), always marked with an abbreviation by the letter ʿayn;  sajda (prostration) appears less frequently.34 Though all these marginal indications are usually inscribed within illuminated medallions, they are also occasionally just written in black or colored ink, without any surrounding illuminated element.35 Other Indian manuscripts copied in bihārī script share the marginal marker system as well. They are however less complete than the W563 Qurʾan. For instance, following the Gwalior Qurʾan’s pattern, the ‘Sam Fogg’ Qurʾan  does  not  have  any  five  and  ten  verse markers.36 The vignettes are only of two shapes: roundels  and  mandorlas  of  various  sizes  (figure 6).

Figure 6 – Marginal medallions, fol. 278r (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum]

32. This is one major difference with the Gwalior Qurʾan. Nevertheless, the latter might appear as a unique example of the use of kufic in an Indian Qurʾan for other copies in bihārī contain illuminated titles only written in various cursive scripts; Brac de la Perrière (2008), 145. 33. These numerical markers do not appear in the Gwalior Qurʾan. 34. The notation of rukūʿ is extensive – if not exhaustive – in comparison with the Gwalior Qurʾan. 35. As noted previously for the juzʾ markers and corrections of the Qurʾanic text itself, these marginal indications written in muḥaqqaq testify that the calligrapher cautiously revised the manuscript once completed. 36. Fraser, Kwiatkowski (2006), 109. The letter ʿayn is wrongly described as a ten-verse marker when it is in fact a rukūʿ notation.

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 199

The Gwalior Qurʾan has vignettes of five different  shapes, including the two selected for the WAM Qurʾan. Moreover, the pear-shaped medallion, which is numerically the most important in the AKM copy, is completely absent from the W563 manuscript. The vignettes’ outer border bears mainly two types of decoration. Roundels show a frieze alternating two half  palmettes  and  fleurons  or  buds  whereas  mandorlas have stylized garlands or sprays of leaves and flowers, formulae already used in the Gwalior Qurʾan, albeit much less frequently. In the latter, borders of round or pointy petals prevail, a decorative option for vignettes that seldom appears in the WAM Qurʾan.37 Despite these differences, the construction of some illuminated double pages both in the AKM and the WAM Qurʾans shows similar features: in the vertical lateral side of the margin, a central piece is surrounded by two smaller vignettes. This element in both instances can have three shapes: triangular, semi-circular,  or  ‘cloud  collar’-like  (figure 7).  The  central tripartite zone consists in a series of various borders and friezes with two ʿunwāns placed above and beneath the Qurʾanic text. The juxtaposition of plain lines, interlaces, strap-works, and bands of cartouches with stylized vegetal garlands offers a rich and intricate repertoire of decorative patterns. Their recurrence further provides a visual unity throughout the manuscript. Regarding the frames, one major difference between the WAM and Gwalior Qurʾans is noticeable: the broad use in the WAM copy – and the mere absence in the Gwalior manuscript – of a white or gilded frieze adorned with small black crosses.38 Within the ʿunwāns, the cartouches presenting titles most often have semicircular edges as in the Gwalior Qurʾan, although in the case of the W563 manuscript, a more elaborate scalloped and curvilinear outline seems to have been preferred.39

37. The design of marginals in the WAM Qurʾan seems to rely on models that were already adopted by Indian illuminators around 1400, as the Gwalior Qurʾan proves. They are in fact closely akin to marginals visible in late Injuid and Muzaffarid manuscripts from the 1340’s to the early 1380’s; Wright (2013), 52-62 and in particular  figs. 29-31. They are also very similar – if not identical –  to the vignettes in the British Library’s Qurʾan;  id.  (1996-1997), 9-10. 38. Extensively used in Persian illuminations since at least the 14th century, the stripe is totally integrated into the Iranian decorative vocabulary as well as in the Mamluk sphere. In the Gwalior Qurʾan, we find it only on fols. 116v-  117r, 171v-172r, 274v-275r and 356v-357r. 39. For an example in the Gwalior Qurʾan, see Brac de la Perrière, Chaigne, Cruvelier (2010), 120, fig. 11.

The palette used in the illuminations of the W563 copy is as varied as in the Qurʾan of Gwalior but the overall chromatic aspect that emerges from a comparison of the two manuscripts is rather dissimilar. The WAM Qurʾan’s double pages are dominated by blue and gold tones, a characteristic developed in the illuminations of Persian manuscripts under the Muzaffarids of Fars in the second half of the 14th century and then widely adopted under the Timurids but also in Mamluk Egypt in the 15th century.40 Pink hues and gradations of red color are used with more parsimony than in the Gwalior Qurʾan but mauve and purple are equally visible, notably serving as background colors. A shared feature, albeit less frequent in the WAM manuscript, is the use of black grounds, especially in the central zone of text. More generally, the vibrancy of colors displayed in the Gwalior Qurʾan appears to be softened in the WAM Qurʾan’s double-page illuminations, in which blue and gold seem to provide a chromatic unity to the whole. Still, the ‘kaleidoscopic’ palette seems to stand out in the marginal vignettes. Interestingly the association of some colors, whose presence albeit discreet is noticeable in the AKM Qurʾan, is conspicuous in almost every illumination in the WAM Qurʾan: red, yellow, green and white stripes. In the Gwalior manuscript, it discreetly appears in the shape of small buds comparable to ‘peacock-feather eyes’ arranged either in framing borders (fols. 50v-60r) or on two occasions at the extremities of a marginal mandorla vignette (fols. 429v-430r). In the WAM Qurʾan on the contrary, these buds with this peculiar palette of colors are ubiquitous, terminating almost every almond-shaped marginal. The color scheme has furthermore been applied  to  small  fleurons  that  are  displayed  in  the  outer borders of circular medallions, in the framing borders of the illuminated double pages as well as in their larger central compositions. Moreover numerous other stylized vegetal elements with lanceolate petals or leaves present these colors. In that sense, the association of red, green, yellow and white appears as important as the combination of blue and gold for it equally functions as a visual unifying tracer within the W563 manuscript. It is also a link with Indian nonMuslim traditions of painting, and notably Jain.41 At last, this palette appears in the British Library Qurʾan, both in finial buds and fleurons in borders (figure 8).

40. See Wright (2013), 48-80. 41. Gujarati Jain manuscripts contain paintings whose backgrounds are mainly yellow or red with details executed in white and green: for instance, the Kalpasūtra and Kālakāchāryakathā dated 1411 (Arthur M. Sackler Gallery, S1985.2) and the Vasanta Vilāsa made in Ahmedabad in 1451 (Freer Gallery of Art, F1932.24); Beach (2012), 45-47.  See also Brac de la Perrière (2008), 164 and 232-243.

200 • SIMON RETTIG

Figure 7 – Illuminated opening of the sura Maryam, fol. 274v, Qurʾan W563. [© By courtesy of the Walters Art Museum]

Some motifs appearing abundantly in the latter and widely used thereafter in other Indian copies in bihārī script are indeed virtually absent from W563, such as the stylized ‘tufts of grass’ (fols. 356v-357r, 500v-501r).42 Instead, one sees leafy branches and clusters painted in gold with scattered round flowers, 

42. Ibid., 163 and Brac de la Perrière, Chaigne, Cruvelier (2010), 120.

green or yellow and red colored fleurons mentioned  previously, and various stylized dentate leaves associated  with  feathery  floral  sprays.  All  of  these  are  already noticeable in the incredibly varied vegetal repertoire in the Gwalior Qurʾan. But the blossoms and leaf sprays painted in gold directly onto the plain paper in the illuminations of the WAM Qurʾan are hitherto unseen in Indian illuminations and probably originate from another source (figure 9). Not  only reserved to Qurʾanic illuminations, this type

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 201

Figure 8 – Marginal vignette, fol. 438r (detail). [© British Library, Add. 18163]

Figure 9 – Illuminated page with shamsa, fol. 6v (detail), Qurʾan W563. [© By courtesy of the Walters Art Museum].

becomes standard on the surface of decorated pages, either in full-page illuminations or for title cartouches and even in the colophon zone in the first half of the 15th century in Iranian centers, and in particular in Shiraz.43 Taken into consideration in parallel with the association of blue and gold, the decorative novelty perceptible in the illuminations contributes to placing the WAM Qurʾan within a scheme of production that goes beyond India. Stylistic analogies and visual parallels can be found first in Timurid Iran and thereafter  expanded to Mamluk Egypt, Ottoman Anatolia and Sultanate India. An encompassing aesthetic approach to the WAM Qurʾan definitely supports the idea that  the manuscript fits into a supra-regional scheme of  conceptualization for a certain category of Qurʾan manuscripts in the 15th century. 43. See for instance the double-page opening (fols. 1v-2r) of a Qurʾan dated 823/1420 and probably made in Shiraz (Khalili  Collection,  QUR212);  James  (1992),  no. 4,  26-27.  The opening of another undated Qurʾan, attributable to the 15th century, also presents this characteristic type of  Shirazi decoration (Khalili Collection, QUR123); ibid., no. 18, 70-71.

The Qurʾan of Gwalior as the oldest known dated Indian Qurʾan copy is generally considered as the Urmanuscript of the Sultanate period on the Indian subcontinent. Instead of a starting point that generated and epitomized all the visual aspects subsequently implemented in Sultanate Indian Qurʾans, this item should actually be regarded as a concluding experience, a visual compendium of the fourteenth-century artistic developments. Any manuscript produced thereafter would aggregate the manifold artistic and aesthetic experiments into one single volume as the Gwalior Qurʾan did. By comparison with the W563 Qurʾan, one can sense a chronological gap between the productions of the two manuscripts: the ornamental richness of the Gwalior Qurʾan has seemingly been ‘digested’ and assimilated. As a result some elements were fully integrated into the decorative vocabulary used in the WAM Qurʾan whereas, on the contrary, some others were seemingly precluded or least not selected. The lack of knowledge regarding the complex mechanisms of artistic and aesthetic transfers in preMughal India further complicates our comprehension of a selective and rejecting process of forms and motifs in the artistic field, and especially in the arts  of the Indo-Islamic book. The WAM manuscript partly

202 • SIMON RETTIG

Figure 10 – Illuminated heading of the sura al-Naml, fol. 457v (detail), Add. 18163. [© By courtesy of the British Library].

relies on models developed in the 14th century in Inju and Muzaffarid Shiraz and transmitted to India.44 But it also integrates new decorative trends developed under the Timurids. Hence, while it is still difficult to  offer a precise dating for the production of the Walters Art Museum Qurʾan, one may suppose that the adoption and the adaptation of illuminated filling motifs  and patterns and compositions would point toward a production in the second quarter of the 15th century. Indeed, the first half of the 15th century in Iran witnessed the exploration of new visual aesthetic systems in manuscript production and notably in Qurʾan copies. This occurs in fact beyond Iranian centers via the circulations of peripatetic artists and objects on a wide scale. It is also a time when luxury manuscript production not only burgeoned in the strengthening Ottoman Empire in Anatolia but flourished in the Mamluk sphere as well. A production  of small format and neat copies in naskhī developed after 1350.45 The manuscripts in bihārī and naskhīdiwānī may be seen as local responses to this trend. On the other hand, larger formats written in muḥaqqaq continued to be executed and it seems that an ‘international’ standardized version was very much favored, from Anatolia to India and from Egypt to Central Asia.46 44. Wright demonstrated how the conception of the BL Add.18163 Qurʾan is clearly based on a 30-volume copy made for Fārs Malik Khatūn, daughter of Maḥmūd  Shāh, founder of the Inju dynasty in the 1340’s. Wright  showed also that the illuminated program was started under the Injus and completed under the Muzaffarids. Wright (1996-1997), 8-12. 45. See for instance a Qurʾan  dated  Rabīʿ 767/NovemberDecember 1365; Sam Fogg (2000), 40-41. Numerous Timurid  examples are known, one of the most famous being the Qurʾan copied by Ibrāhīm Sulṭān in Shiraz and dated  7 Ramāḍan 830/29 June 1427; Lentz, Lowry (1989), 84 and  332-333. 46. For a Mamluk example ca. 1380, see TIEM 226 in Istanbul (2010), no. 60, 260-261; for a Timurid Qurʾan juzʾ ascribed to the first half of the 15th century (TIEM 564), see Lentz, Lowry (1989), 75-77, no. 19. For an Ottoman Qurʾan made in the second half of the 15th century (CBL Is 1492), see Wright (2009), 82, fig. 54.

The W563 Qurʾan  fits  very  much  into  this  group,  a  fifteenth-century  transregional  production  of  Qurʾan manuscripts that tend to present a uniform overall appearance (format, calligraphy, page layout and general composition of the illuminations) despite variations in the decorative modes. As a consequence, a putative dating of the WAM Qurʾan to the years 1420-1450 also relies on the illuminated decoration. The penetration into India of the Timurid manuscript tradition, be it mainly from Shiraz under Iskandar Sulṭān  b.  ʿUmar  Shaykh  (d. 1415),  Ibrāhīm  Sulṭān  b. Shāh Rukh (d. 1435) or from Herat with the making  of sumptuous copies in the princely kitābkhāna (library-cum-workshop) of Bāysunghur Mirzā (d. 1433)  is perceptible through the WAM Qurʾan. Motifs and colors traditionally ascribed to India and visible in the Gwalior Qurʾan were combined mainly with Iranian models along the lines of a synthesis operated within the Walters Art Museum Qurʾan. In other words, the W563 manuscript aggregates several elements that can be pigeonholed as local, regional, and extraregional. In that sense, the juxtaposition of diverse decorative components may indeed recall the Gwalior Qurʾan with its repertoire of forms and filling motifs  stemming from a multifarious range of regional and ‘foreign’ sources. The striking physical similitude with the British Library manuscript may also place the WAM Qurʾan within the context of a Gujarati production. Written in bold muḥaqqaq, the two copies also present similar illuminations, in both the layout and the filling motifs, as well as in the color palette,  notably  in  the  sura  headings  (figure 10).  However,  the British Library’s Qurʾan does not present Timurid features like the WAM manuscript, which may suggest an earlier date of execution, around 1400.47

47. Wright putatively placed the production of the BL Qurʾan in the early 15th century  as  well;  Wright  (1996-1997),  8-12.

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 203

Another Qurʾan provides further support for the hypothesis that the WAM Qurʾan may have been produced in Gujarat. It is a two-volume copy whose colophon indicates that it was completed in Batwa (modern Vatva) in the vicinity of Ahmedabad in 894/1488.48 Although the manuscript was copied in bihārī and has characteristics of Qurʾans in this script, the illuminations show a strong resemblance with the WAM Qurʾan ones, possibly indicating a fusion that occurred later in the century between the different physical conceptions of making Qurʾans in muḥaqqaq and bihārī. Despite an evident difference in treatment, the double page framing sūrat al-Nās and sūrat al-Falaq clearly follows the model of fols. 4v5r in the WAM Qurʾan, not only in the composition of  the  borders,  cartouches  and  marginals  but  first  and foremost in the choice and arrangement of colors as well as of vegetalized motifs.49 While the British Library Qurʾan was obviously part of a collection associated with the ruler of Gujarat,50 the WAM Qurʾan concurrently entered the imperial Ottoman collections in Istanbul no later than the end of Bāyezīd II’s reign (r. 1481-1512) as his seal affixed  on fol. 8r indicates. Thus these two sumptuous manuscripts not only share common characteristics that can be essentially defined as a strong effect of  structural, ornamental, and calligraphic homogeneity;  they  ultimately  address  royal  patrons  as  well,  both on regional and long-distance scales. In other words, the BL and WAM Qurʾans feature a stylistic quality that transcends local artistic productions and dynastic borders in the 15th century. As a consequence, one may speculate that Indian Qurʾans copied in bihārī oppose those in muḥaqqaq, in the sense that the use of a particular script would address a specific  category of patrons, commissioners, or persons for whom it was intended. However this issue has less to do with the levels of production and more with the destination of the visual layout of two radically distinct types of Qurʾanic text. In other words, in the fifteenth-century Indian context, both the W563 – with the British Library’s copy – and Gwalior Qurʾans represent two distinctive conceptions of the destination of the Qurʾan manuscript as an object and exemplify the mobility of its status. Qurʾans copied within the tradition of the ‘six scripts’ like the ‘Indian muḥaqqaq group may address persons of high status not only in the regional sphere but also beyond, whereas Qurʾans in bihārī script, whatever the quality of their production may have been, were made principally if not solely for an Indian audience, 48. Bombaywala (2011), 55. 49. For a reproduction of the illumination, see ibid., 56, fig. 4. 50. Although undated, the manuscript bears the ownership seal of Maḥmūd Shāh ‘Bigarha’ (r. 1458-1511).

for the script may have not appealed to foreign clients.51 In this context one may assume that the final geographic destination of the W563 Qurʾan was indeed the Ottoman capital and that the manuscript was sent as a diplomatic gift. Even so it does still raise the issue of the initial commissioner’s identity as well as the place of production. Furthermore the presence of the Qurʾan in Istanbul questions the relations between Ottoman Anatolia and Sultanate India before the reign of Selīm I (r. 1512-1519). Little  is still known for modern studies primarily focused on the 16th century, first with the rise of the Ottomans  as a maritime power and second with the expanding relationships – commercial, diplomatic, and cultural – between the Ottoman and Mughal empires.52 Nevertheless some historical accounts attest to the existence of diplomatic contacts between the two areas already in the 15th century. For instance the Ottoman chronicler Ṭūrsūn Beg mentions in the Tarih-i Ebüʾl-Fetḥ composed in 1490 the diplomatic relations between Mehmet II (r. 1452-1481) and the ‘Indian ruler’.53 Ṭūrsūn  Beg  reports  that  the  latter  sent  a  mission  with “rare and precious gifts” (tuḥaf-ı hedāya) to the Ottoman sultan in 1481. Traveling through the Red Sea, the convoy was stopped and retained by order of the Mamluk sultan in Jeddah. Meanwhile, Mehmet II passed away and Bāyezīd II, just enthroned, threatened the Mamluks, who finally let the mission depart  from Jeddah to Istanbul. One may speculate that a luxurious Qurʾan like the WAM manuscript may have been part of such a set of gifts.54 As often mentioned elsewhere, the completion date of the Gwalior Qurʾan coincides with the invasion of India by Tīmūr’s armies at the very end of the  14th century, a time traditionally considered as a turning point in the history of Northern India. The Gwalior Qurʾan is also a singular object, unequaled in

51. A notable exception would be south Arabia and Yemen, where Qurʾans in bihārī have been discovered; Blair (2006), 388-389. 52. Farooqi (1986). See also Casale (2010), 17 and 25-27. 53. The History of Mehmed the Conqueror by Tursun Beg [1978], 64-65 and fol. 167v. The Ottoman chronicler does not provide the name of the Indian ruler but Farooqi, according to diplomatic correspondence, argues that it is the Bahmanid sultan Muḥammad Shāh III (r. 1463-1482); Farooqi  (1986), 20-22. 54. Little is yet known about Qurʾans offered as gifts between rulers,  the  notable  exception  being  Shāh  Ṭahmasp’s  envoy to Selīm II with a manuscript allegedly from the  hand of ʿAlī b. Abī Ṭālib; Komaroff (2011), 17-19. Incidentally the Safavid ruler offered another Qurʾan, copied in nastaʿlīq  script  by  the  celebrated  calligrapher  Shāh  Maḥmūd  Nishāpurī,  to  the  Ottoman  sultan  Murad III  on  the occasion of his accession to the throne in 1574;  Blair (2006), 433-434.

204 • SIMON RETTIG

the richness and intricacy of its decorative program that epitomizes fourteenth-century artistic experiments and attests to the reception and adoption of manifold artistic traditions in India around 1400. Like a later echo, the WAM Qurʾan offers an original response with the absorption and assimilation of Timurid aesthetics and Iranian artistic formulae. The general layout, the selective choice of calligraphic scripts, and above all the decorative trends in illumination acknowledge the ties – political, cultural and artistic – between Timurid Iran and the various Muslim dynasties that rose in India in the 14th and 15th centuries, especially the Bahmanids in the Deccan, and the Aḥmad Shāhis in Gujarat. While also copied  in muḥaqqaq and showing similarity with the WAM

Qurʾan, the British Library’s copy, and to a lesser extent the Christie’s manuscript as well, rely further on pre-Timurid models. Consequently, like the Gwalior Qurʾan, the W563 manuscript uniquely encapsulates stylistic idiosyncrasies that deeply anchor the Indian subcontinent within an extended and complex network of exchanges rather than in isolation, an area located on the Eastern fringes of the dār al-Islām, as India was often thought to be before the rise of the Mughal dynasty. Finally it shows how aesthetics and artistic practices flow back and forth and circulate  within and outside of India and beyond Persia, as far as Ottoman Anatolia, in a constant flux of appropriation and transformation that lead ultimately to intricate visual hybridizations.

bibLioGraPhy Sources Tursun Beg [1978] The History of Mehmed the Conqueror by Tursun Beg, text published in facsimile with English translation by Halil Inalcik, Rhoads Murphey, Minneapolis/Chicago: Bibliotheca Islamica.

References Beach (Milo Cleveland) 2012 The Imperial Image. Paintings for the Mughal Court, revised and expanded edition, Washington (DC): Freer Gallery of Art, Arthur M. Sackler Gallery and Mapin Publishing. Begley (Wayne E.) 1985 Monumental Islamic Calligraphy from India, Villa Park (Ill.): Islamic Foundation. Blair (Sheila S.) 2006 Islamic Calligraphy, Edinburgh: Edinburgh University Press. Blayac (Johanna) 2012 Formation et histoire des premières sociétés indomusulmanes et indo-islamiques à travers les inscriptions arabes et persanes (viie-xive siècles), PhD diss., Paris, École pratique des hautes études, 2009, Lille: Atelier national de reproduction des thèses. Bloom (Jonathan) 2001 Paper before Print: The History and Impact of Paper in the Islamic World, New Haven/London: Yale University Press.

BomBaywala (Mohaiuddin) 2011 “Arabic-Persian scholarship: medieval manuscripts at the Hazrat Pir Muhammad Shah Dargah Sharif library”, in Suchitra Balasubrahmanyan, Sharmila Sagara (dir.), Ahmedabad 600. Portraits of a City, Mumbai: The Marg Foundation, pp. 50-63. Brac de la Perrière (Éloïse) 2003 “Bihāri et naskhī-diwāni: remarques sur deux calligraphies de l’Inde des sultanats”, in Studia Islamica XCVI, pp. 81-93. 2008 L’art du livre dans l’Inde des sultanats, Paris: Presses de l’université Paris-Sorbonne. Brac de la Perrière (Éloïse), chaigne (Frantz), cruvelier (Mathilde) 2010 “The Qurʾan of Gwalior, kaleidoscope of the arts of the book”, in Treasures of the Aga Khan Museum – Arts of the Book & Calligraphy, exhibition catalogue, Istanbul: Culture and Sakıp Sabancı  Museum, pp. 114-123. casale (Giancarlo) 2010 The Ottoman Age of Exploration, New York/ Oxford: Oxford University Press. chaigne (Frantz) 2012  “Le décor enluminé sous les Īl-khānides: entre  assimilation et innovation, de l’Iraq à la Chine”, in Beiträge zur Islamischen Kunst und Archäologie 3, pp. 253-265. dÉroche (François) (dir.) 2000 Manuel de codicologie des manuscrits en écriture arabe, Paris: Bibliothèque nationale de France.

A ‘TIMURID-LIKE RESPONSE’ TO THE QURʾAN OF GWALIOR? • 205

desai (Ziaduddin A.) 1961 “Some Tughluq inscriptions from Bihar”, in Epigraphia Indica – Arabic and Persian Supplement, pp. 26-34. Farooqi (Naimur R.) 1986 Mughal-Ottoman Relations: A Study of Political and Diplomatic Relations between Mughal India and the Ottoman Empire, 1556-1748, PhD. diss., Madison (WI): The University of Wisconsin. Fraser (Marcus), kwiatkowski (Will) 2006 Ink and Gold. Islamic Calligraphy, London: Sam Fogg. Istanbul 2010 The 1400th Anniversary of the Qurʾan. Museum of Turkish and Islamic Art Museum Qurʾan Collection, exhibition catalogue, Istanbul: Museum of Turkish and Islamic Art. James (David) 1992 After Timur: Qurʾans of the 15th and 16th centuries, edited by Julian Raby (The Nasser D. Khalili Collection of Islamic Art, vol. III), London/ Oxford: The Nour Foundation in association with Azimuth Editions and Oxford University Press. komaroFF (Linda) 2011 Gifts of the Sultan. The Arts of Giving at the Islamic Courts, New Haven / London: Yale University Press. lentz (Thomas W.), lowry (Glenn D.) 1989 Timur and the Princely Vision. Persian Art and Culture in the Fifteenth Century, exhibition catalogue, Los Angeles: Los Angeles County Museum of Art. losty (Jeremiah P.) 1982 The Art of the Book in India, London: The British Library. loveday (Helen) 2001 Islamic Paper. A Study of the Ancient Craft, [London]: Don Baker Memorial Fund.

mansour (Nassar) 2011 Sacred Script: Muhaqqaq in Islamic Calligraphy, London: I. B. Tauris. melchert (Christopher), aFsaruddin (Asma) 2004 “Reciters of the Qurʾān”,  in  Jane  Dammen  McAuliffe (dir.), Encyclopaedia of the Qurʾān IV, Leiden/Boston: Brill, pp. 386-392. Porter (Yves) 1994 Painters, Paintings and Books. An Essay on IndoPersian Technical Literature, 12-19th Centuries, translated by S. Butani, New Delhi: Manohar. raBy (Julian), tanindi (Zeren) 1993 Turkish Bookbinding in the 15th Century. The Foundation of an Ottoman Court Style, edited by Tim Stanley, London: Azimuth editions on behalf of l’Association internationale de bibliophilie. rahim (Syed Abdur) 1962 “Some more inscriptions from Khandesh”, in Epigraphia Indica – Arabic and Persian Supplement, pp. 62-86. roxBurgh (David J.) 2005 The Persian Album, 1400-1600, from Dispersal to Collection, New Haven: Yale University Press. Sam Fogg 2000 Islamic Manuscripts, cat. 22, London: Sam Fogg. simPson (Marianna S.) 2000 “‘A Gallant Era’: Henry Walters, Islamic Art, and the Kelekian Connection”, in Ars Orientalis XXX, pp. 91-112. wright (Elaine) 1996-1997 “An Indian Qurʾan and its fourteenth-century model”, in Oriental Art XLII, no. 4, pp. 8-12. 2009 Islam: Faith, Art, Culture, London: Scala Publishers. 2013 The Look of the Book. Manuscript Production in Shiraz, 1303-1452 (Occasional papers, new series, 3), Washington (DC): Freer Gallery of Art/Arthur M. Sackler Gallery in association with University of Washington Press/Chester Beatty Library.

INDEX Noms de lieux, monuments Œuvres, anonymes Personnes, dynastie Abbassides / Abbasid (califat, calife), 19, 159. al-Hākim II, 19. Abū ʿAmrū (= Abū ʿAmr b. al-ʿAlāʾ, qāriʾ, lecteur), 91, 93, 95, 98, 196n. Abū Bakr (calife), 91n. Abū Bakr (rāwī, transmetteur), 93, 98, 196n. Abū Ḥanifa, 101. Abū Jaʿfar b. al-Qaʿqāʿ (qāriʾ, lecteur), 93n. Abū Saʿīd (voir ilkhanide) Adina (mosquée, voir Pandua) Afghanistan, 62n, 159, 160, 163, 164. ʿAfīf, Shams al-Dīn Sirāj, 114n, 178n, 180n. Tārīkh‑i Fīrūz Shāhī, 114n. Afrique / Africa, 177. Aḥmad b. Ayāz (architecte et vizir anatolien), 155. Aḥmad b. Jubayr (imam), 91n. Al-Khamsa, 91n. Aḥmad b. Sīrkhān, 116. Aḥmad Shāh (voir muzaffarides) Aḥmad Shāh (voir sultans du Gujarat) Aḥmad Shāh II (voir bahmanides) Aḥmad Shāhi (voir sultans du Gujarat) Aḥmad Yasawī (voir Turkestan) Ahmedabad, 41n, 199n, 203. Devasano Pado, 41n. Akhsikat (Ferghana), 119. ʿAlāʾ al-Dīn Gwāliyārī, 117, 118, 120. ʿAlāʾ al-Dīn Khaljī (voir sultans de Delhi) ʿAlāʾ-i Darwāza (porte, voir Delhi) Alexandrie / Alexandria, 18. ʿAlī b. Abī Ṭālib, 91n, 203n. ʿAlī b. Sālār b. ʿAlī al-Yazdī (poète soufi), 119n. Aligarh, 116. Altinbugha al-Maridani (mosquée, voir Le Caire) Amid (voir Diyarbakır) Amīr Khusrau (poète), 114n. Amīr Qawṣūn (mosquée, voir Le Caire) al-Anṣārī, ʿAbd Allāh (soufi d’Hérat), 98. Anthology (BL Or. 4110), 41, 52n. Arbaʿūn ḥadīth (CBL Ar 4181), 196n. Asie centrale / Central Asia, 28n, 31n, 118, 119, 166n, 202 ; Timurid, 177. ʿĀṣim (= ʿĀṣim b. Abī al-Najūd, qāriʾ, lecteur), 91, 93, 196n. ʿAtāʾ Shāh (dargāh, voir Dinajpur) Anatolie / Anatolia, 28n, 30, 30n, 31n, 34n, 38, 154, 174, 202 ; Ottoman, 201, 203, 204. Autsahai (Bengale), 179n.

Ayyubides / Ayyubids, 18n, 34, 44. Awrangzeb (voir Moghols) Bābā Arjūn Shāh Damawī al-Akhsī (mystique), 119. Badaʾun, 155. Mosquée du Vendredi / Friday Mosque, 155. Badāʾūnī, 114n. Tārīkhi Badāʾūnī, 114n. Bagdad / Baghdad, 20n, 28n, 30n, 31, 34n, 39n, 52n, 174. Bahmanides / Bahmanids, 114n, 117, 188, 204. Aḥmad Shāh II (= ʿAlāʾ al-Dīn Aḥmad II), 195. Muḥammad Shāh III (= Muḥammad III Lashkarī), 203n. al-Balādurī, 118. Futūḥ al‑Buldān, 118n. Baranī, Ḍiyāʾ al-Dīn, 114n, 178n. Tārīkh‑i Fīrūz Shāhī, 114n. Bari Dargah (= Shaykh Jalāl al-Dīn Tabrīzī, voir Pandua) Baroda (Gujarat), 117. Barqūq (voir Mamelouks) Bassora / Basra, 91. Batwa (moderne Vatva), 203. Baybars (voir Mamelouks) ; Qurʾan (BL Add. 22406-13), 20, 28n. Baybars II (voir Mamelouks) Bāyezīd II (voir Ottomans) Bāysunghur Mirzā (= Bāysunghur, prince timouride), 44, 202. Begumpuri Masjid (voir Delhi) Benares (voir Varanasi) Bengale / Bengal, 116n, 154, 171, 172, 177, 178, 179, 180, 182. Voir aussi sultanat du Bengale. Bibi Khanum (mosquée, voir Samarcande) Bidar, 195. Diwān-i ʿām du palais de Bidar, 185. Bihar, 178, 187. Biḥār al‑anwār (voir al-Majlisī) Bihar Sharif, 195n. Bilram (Uttar Pradesh), 116n. Bouddha, 128. Brousse / Bursa (peinture), 165, 166. Muradiye (complexe) : Cem Sultan et Şehzade Mustafa (mausolées), 165, 166. al-Bukhārī, 91n. Kitāb al‑manāqib, 91n. Bust (ville), 159 ; (coran, BnF Arabe 6041), 20, 31, 41, 47. Cairo (voir Le Caire) Cambay (Gujarat), 117, 119. Grande mosquée / Great Mosque, 120. Cem Sultan (voir Brousse)

208 • INDEX

Central Asia (voir Asie centrale) Chaghatai, 19. Champaner (Gujarat), 177, 187. Chanderi (Malwa), 117, 118. Chandernagor (Bengale), 177. Chau-Ju-Kua (historien du xiiie siècle), 18. Chera (royaume), 118. Chhoti Dargah (= Nūr Quṭb ʿĀlam, voir Pandua) Chine / China, 18, 118, 157, 158, 172, 173, 176, 177, 178, 187. Mosquée du Vendredi / Friday Mosque, 157, 158. Chirāz / Shiraz, 30n, 31n, 34n, 41n, 43, 45n, 49, 52, 154, 155, 157, 158, 159, 174n, 179, 201, 202 ; Muzaffarid, 41, 202. Chishti, 117-118, 120. Chupan (amīr), 19n. Ctésiphon / Ctesiphon, 154, 180. Dacca (Bangladesh), 177. Damas / Damascus, 20n, 34n, 91, 155. Nūr al-Dīn (bīmāristān), 159. Damoh, 119. Al-Dānī, 93. al‑Taysīr fī al‑qirāʾāt al‑sabʿ, 93. Dawlatabad / Daulatabad, 117, 188. Deccan, 114n, 117, 120, 121, 185, 195, 204. Delhi (ville), 11, 18, 19, 53, 54, 114n, 116, 117, 120, 121n, 155, 159, 163, 176, 187 ; dynasties, 172 ; New Delhi, 41. Voir aussi sultanat de Delhi. ʿAlāʾi Darwāza (porte), 176. Begumpuri (mosquée / Masjid), 176, 180 Firuz Shah (= Firuz Shah Kutla, palais), 172, 176, 177, 187. Haus Khas : Firūz Shāh Tughluq (tombe), 155, 157. Mosquée du Vendredi / Friday Mosque of Jahanpanah, 156, 157, 159n. Qutb (mosquée / Mosque), 159. Quwwat al-Islam (mosquée), 174, 175, 178n. Demotte Shāhnāma, 41n. Devasano Pado (voir Ahmedabad) ; Kalpasūtra, 51n, 52n. Deyiyā (peintre), 135, 137. Dinajpur (Bengladesh), 178. ʿAtāʾ Shāh (dargāh), 180. Diyarbakır, 30. Diwān (voir Ḥāfiẓ) Diwān-i ʿām of Bidar Palace (voir Bidar) Djeddah / Jeddah, 203. Al-Durī (rāwī, transmetteur d’Abū ʿAmrū), 196n. Dushanbe, 164. Égypte / Egypt, 18, 19, 20, 21, 34n, 45, 49, 154, 155, 163, 166, 202 ; Mamluk, 119n, 197n, 199, 201. Eklakhi (mausolée, voir Pandua) Fars, 158, 180, 199.

Fārs Malik Khātūn (fille de Maḥmūd Shāh), 202n. Fatehpur Sikri, 177, 188. Fatḥnāma (= anonymous Shāhnāma), 118n. Fawāʾid al-Fuʾād (voir Niẓām al-Dīn Awliyāʾ) Ferghana, 119. Firdawsī, 48n, 114n. Shāhnāma, 48n, 114n. Firishta, 114n. Tārīkh‑i Firishta, 114n. Firuz Shah Kutla (palais, voir Delhi) Firūz Shāh Tughluq (voir sultans de Delhi ; tombe, voir Delhi) Firuzkuh, 159. Fustat, 18. Futūḥ al‑Buldān (voir al-Balādurī) Futūḥ al‑Salāṭīn (voir ʿIṣāmī) Gange, 177. Gawr, 178, 185. Genizeh / Genizah (voir Le Caire) al-Ghazālī, 164. Iḥyāʾ ʿulūm al‑dīn, 164. Ghāzān (= Maḥmūd Ghāzān , voir ilkhanides) ; Qurʾan, 28, 52n. Ghazna, 159. Ghaznavides / Ghaznavids, 18. Ghiyāth al-Dīn (gouverneur kart d’Hérat), 164. Ghiyāth al-Dīn Tughluq (voir sultans de Delhi) Ghiyāth al-Dīn Muḥammad b. Sam (voir ghurides) Ghiyāth al-Dīn ʿAzam Shāh (voir sultans du Bengale) Ghiyāth Zarrin Dast, 180. Ghur (Afghanistan), 163. Ghurides / Ghurids, 116, 118, 159, 163, 164 ; Qurʾan, 160, 161, 162, 163, 164. Ghiyāth al-Dīn Muḥammad b. Sam, 21, 159, 160, 164. Gīsū Darāz, Muḥammad (shaykh), 117-118, 120. Golā b. Ḥasan Kambāyatī (muʿallim), 119. Golconde / Golconda, 120, 187. Gujarat, 13, 18, 38n, 49, 52n, 53, 54, 117, 118, 119, 120, 128, 129, 130, 134, 135, 136, 158, 187, 203, 204. Voir aussi sultanat du Gujarat. Gulbarga, 117. Gwalior (ville), 11, 18, 53, 54, 58, 114, 116, 117,118, 119, 120, 121, 123, 140, 155, 157, 165. Hāthī Por (porte), 118. Madrasa-i Nāṣirī, 116n. Urwahi (porte de la citadelle), 116n. Ḥāfiẓ, 103, 171, 179. Diwān, 103. Ḥafṣ (rāwī, transmetteur de ʿĀṣim), 93, 98. al-Hākim II (voir Abbassides) Hama, 44. Hamadan, 20n. Hampi-Vijayanagar (capitale des rois Vijayanagar), 121, 122.

INDEX • 209

Ḥamza (= Ḥamza b. Ḥabīb al-Zayyāt, qāriʾ, lecteur), 91, 196n. Ḥaqāʾiq al‑tafsīr (voir al-Sulamī) Hariṇaigameṣin (messager du dieu Śakra), 136. Harsha (empereur), 118. Hāthī Por (voir Gwalior) Haus Khas (voir Delhi) Hemacandra, 128, 129. Hérat / Herat, 51n, 98, 159, 163, 164, 197, 202. Mosquée du Vendredi / Friday Mosque, 163, 164. Hindustan, 116, 117, 118. Ḥirz al‑amānī wa wajh al‑tahānī fī al‑qirāʾāt al‑sabʿ (voir al-Shāṭibī) Hooghly (rivière), 177, 178. Hormuz, 119n. Ḥudūd al‑ʿĀlam (anonyme), 114n. Ibn al-Jazarī, 93n. Al‑Nashr fī al‑qirāʾāt al‑ʿashr, 93n. Ibn ʿĀmir, ʿAbd Allāh (qāriʾ, lecteur), 91, 93, 94, 98, 196n. Ibn Aybak al-Ṣafadī, 159. Kitāb al‑wāfī bil‑wafayāt, 159. Ibn Baṭṭūṭa, 18, 19, 114, 116, 117, 120. Riḥla, 18n, 176. Ibn Bawwāb, 174n. Ibn Ḏakwān (rāwī, transmetteur d’Ibn ʿĀmir), 196n. Ibn Faḍl Allāh al-ʿUmarī, 159. Masālik al‑abṣār, 159. Ibn Ḥajar, 19. Ibn Ḥawqal, 114n. Ibn al-Jawzī, 98n. Ibn Kathīr, ʿAbd Allāh (= ʿAbd Allāh b. Kathīr, qāriʾ, lecteur), 91, 93, 95, 98, 196n. Ibn Masʿūd, 91n ; codex, 141. Safīna‑yi Tabrīz, 103. Ibn Mujāhid al-Baghdādī, Abū Bakr, 91, 93. Ibn Saʿd, 142-143. Kitāb al‑Ṭabaqāt, 142-143, 144. Ibn Sīnā, 159. Kitāb al‑shifāʾ, 159. Ibrāhīm (= Shams al-Dīn Ibrāhīm, sultan de Jaunpur), 195n. Ibrāhīm al-Amīdī, 30, 38. Ibrahim Lodi (voir sultans de Delhi) Ibrāhīm Sulṭān b. Shāh Rukh (prince timouride) Iḥyāʾ ʿulūm al‑dīn (voir al-Ghazālī) Ilkhanides / Ilkhanids, 19, 28, 30, 31, 34, 38, 41, 45, 47, 48, 51, 52, 53, 154, 155, 157, 160, 163, 164, 172, 173, 176, 182, 195n, 196n. Abū Saʿīd, 19. Ghāzān, Maḥmūd, 19, 28, 52n. Öljeitü, 19, 53, 155. Iltutmish (voir sultans de Delhi) Ilyasides / llyasids (= Ilyās Shāh), 178, 182, 185. Voir aussi sultans du Bengale.

Injus / Injuids, 34n, 41, 45, 52, 154, 158, 164, 199n, 202. Maḥmūd Shāh, 202n. Intikhāb‑i Shāhnāma (= Pseudo‑Shāhnāma), 48. Iran, 28n, 31n, 34n, 38n, 45n, 49, 53, 65, 87, 101, 119, 120, 155, 158, 160, 163, 166n, 172, 174, 202 ; Ilkhanid, 154, 172, 173, 176, 182, 198n ; mongol, 14 ; Safavid, 165 ; Timurid, 192, 201, 204 ; Turkman, 195. Iraq, 18, 20n, 28n, 31n, 49, 52n, 87, 154. ʿIṣāmī, Khwāja ʿAbd al-Malik, 114n. Futūḥ al‑Salāṭīn, 114n. Iskandar Sulṭān (= Iskandar Sulṭān b. ʿUmar Shaykh, prince timouride), 44, 202. Ispahan / Isfahan, 53. Mosquée du vendredi / Friday Mosque, 53. Istanbul, 174n, 192, 203 ; ‘Saray albums’, 182. Jaʿfar al-Ṣādiq (imam), 101. Jahanpanah (voir Delhi) Jaïn / Jain, Jaina, Jains, 13, 18, 19, 41, 47, 49, 51, 52, 53, 66, 118, 127-136, 158, 199. Jajnagar (Orissa), 178. Jalayirides / Jalayirids, 154, 164. Jam (minaret), 159. Jaunpur, 48n, 52n, 195n. Jeddah (voir Djeddah) Jérusalem / Jerusalem, 154. Jinabhadrasūri, 134. Jinacarita (= Vies des Jina), 128. Jūzjānī, Minhāji Sirāj, 114n, 116. Ṭabaqāti Nāṣirī, 114n, 116. Kālaka (moine), 128, 133, 134. Kālakācāryakathā (= Histoire du maître Kālaka), 49, 128, 134n, 135, 199n. Kalīla wa Dimna, 34n. Kalpasūtra, 13, 41, 51n, 52n, 53n, 127-138, 199n. Kamrup (Assam), 178, 187. Kannauj, 118, 189. Kart / Kartids, 163-164. Malik Pīr ʿAlī, 164. Kashf al‑asrār wa ʿuddat al‑abrār (voir Maybudī) Katmandou / Kathmandu (Népal), 178. Khaljis, 176. Voir aussi sultans de Delhi. Al-Khamsa (voir Aḥmad b. Jubayr) Khandesh, 195. Khiva, 174. Sayyid ʿAlāʾ al-Dīn (mausolée), 174. Khilād (rāwī, transmetteur de Ḥamza), 196n. Khorassan / Khurasan, 19, 121, 160n. Khwarezm / Khwarazm, 174. Kitāb al‑manāqib (voir al-Bukhārī) Kitāb al‑shifāʾ (voir Ibn Sīnā) Kitāb al‑Ṭabaqāt (voir Ibn Saʿd) Kitāb al‑wāfī bil‑wafayāt (voir al-Ṣafadī) Kitāb‑i mukātibāt‑i Rashīdī (voir Rashīd al-Dīn) Al-Kisāʾī (= ʿAlī b. Ḥamza al-Kisāʾī, qāriʾ, lecteur), 91, 196n.

210 • INDEX

Konkan (cotes de), 118. Kufa, 91. Kulliyāt (voir Saʿdī) Kumārapāla (roi), 134. Kunya-Urgench, 174. Najm al-Dīn Kubrā (mausolée), 174. La Mecque / Mecca, 91, 197. Lahore, 18. Badshahi Mosque, 178n. Lakhnawati, 178, 185. Voir aussi Gawr. Lakshmi, 174. Laʾl Shahbaz Qalandar, 121n. Laṭāʾif al‑ishārāt (voir al-Qushayrī) Le Caire / Cairo, 18n, 19, 20n, 30, 31n, 34n, 38n, 41n, 49, 51, 154, 173. Altinbugha al-Maridani (mosquée), 154n. Amīr Qawṣūn (mosquée), 154. Genizeh / Genizah, 18. Sultan Ḥasan (complexe funéraire), 154, 155. Ma Huan (interprète), 177. Maʿbar (Tamil Nadu), 118. Madrasa-i Nāṣirī (voir Gwalior) Magadha, 128. Mahananda (rivière), 178. Mahāvīra, 126, 136. Maḥmūd Shāh (voir Injus) Maḥmūd Shāh ‘Bigarha’ (voir sultans du Gujarat) Maḥmūd Shaʿbān, 11n, 12, 58, 114, 117. Al-Majlisī, Muḥammad Bāqir, 101n. Biḥār al‑anwār, 101n. Majmūʿa al‑rashīdiyya, 28, 31n, 41n, 53, 74, 155n. Makhdūm Shaykh, 178. Malabar (Kerala), 118. Malda, 178. Malik Pīr ʿAlī (voir Kart) Malwa (= Mālibah), 117, 118. Mamelouks / Mamluks, 18, 19, 20, 21, 28, 30, 31, 34, 38, 39, 41, 44, 45, 47, 48, 51, 52, 53, 54, 119n, 154, 157, 160, 163, 164, 172, 174n, 176, 195, 197n, 199, 201, 202, 203. Barqūq (= al-Ẓāhir Sayf al-Dīn Barqūq), 28n, 38. Baybars (= al-Ẓāhir Rukn al-Dīn Baybars Ier), 20. Baybars II (= al-Muẓaffar Rukn al-Dīn Baybars II al-Jāshankīr), 49. Ḥasan (= al-Nāsir al-Dīn al-Ḥasan), 154. al-Nāsir Muḥammad Qalāʿūn, 19, 53. Shaʿbān (= al-Ashraf Nāsir al-Dīn Shaʿbān II), 28n, 38. Mantri Vāchaka, 135. Al-Maqrīzī, 19. Maragha, 28n, 31n, 36, 41n, 47n, 52n. Marzubānnāma (Istanbul, ms. 216), 174. Masālik al‑abṣār (voir Ibn Faḍl Allāh al-ʿUmarī) Al-Masʿūdī, 114n.

Matn al‑shāṭibiyya fī al‑qirāʾāt al‑sabʿ (voir al-Shāṭibī) Maybudī, Abū al-Faḍl Rashīd al-Dīn, 98, 101, 107. Kashf al‑asrār wa ʿuddat al‑abrār, 48n, 98, 101, 106, 107. Mecca (voir La Mecque) Médine / Medina, 91, 197. Mehmet II (voir Ottomans) Ming (dynastie), 157, 187. Moghols / Mughals, 166, 177, 187, 188, 195, 203, 204 ; pre-Mughal India, 63, 172, 188, 192, 201. Awrangzeb, 178n. Mossoul / Mosul Qurʾan, 53. Mubārak al-Anbayatī (Shaykh), 159. Mubārak Shāh (voir sultans de Delhi) Mughīth al-Qārī al-Shīrāzī, 195n. Al‑Muḥaddith al‑fāṣil bayn al‑rāwī wa al‑wāʿī (voir al-Rāmhurmuzī) Muḥammad (Prophète), 120n, 141. Muḥammad Amīn, 12. Muḥammad b. Mubādir Abū Bakr (= Ṣandal), 20. Muḥammad b. Tughluq (voir sultans de Delhi) Muḥammad b. ʿAlī b. Yaḥyā b. al-Jauzī al-Jazrī (muftī), 119n. Muḥammad b. ʿĪsā b. Muḥammad b. ʿAlī al-Nīshāpūrī al-Laythī (scribe), 160. Muḥammad Shāh III (voir Bahmanides) Muḥammad III Tughluq (voir sultans de Delhi) Muʿizz al-Dīn (frère de Ghiyāth al-Dīn, ghuride), 159. Multan, 155. Rukn-i ʿAlām (tombe), 155. Al-Muqaddasī, 114n, 118. Muradiye (voir Brousse) Murad III (voir Ottomans) Muzaffarides / Muzafarids, 41, 48n, 52, 53, 54, 154, 164, 199, 202. Nāfiʿ (= Nāfiʿ b. ʿAbd al-Raḥmān, qāriʾ, lecteur), 91, 93, 95, 98, 196n. Najm al-Dīn Kubrā (mausolée, voir Kunya-Urgench) Al‑Nashr fī al‑qirāʾāt al‑ʿashr (voir Ibn al-Jazarī) Nāṣir al-Ḥarāwī (scribe), 164. Al-Nāṣir Muḥammad b. Qalāwūn (voir Mamelouks) Nathar Shāh Walī (saint), 120, 121. Nemidarśana Iñānasālā Palitānā, 51n. Nichapour / Nishapur, 159, 160. Niẓām al-Dīn Aḥmad, 114n, 116n. Ṭabaqāti Akbarī, 114n 116n. Niẓām al-Dīn Awliyāʾ Dihlawī (= Niẓām al-Dīn Awliyāʾ of Delhi, Chisti saint), 114n, 120. Fawāʾid al-Fuʾād (voir Sijzī) Nūḥ b. Abī Maryam, 101. Nūr al-Dīn (bīmāristān, voir Damas) Nūr Quṭb ʿĀlam (= Chhoti Dargah, voir Pandua) Öljeitü (voir Ilkhanides et Sultaniyya) ; Qurʾan, 28, 30n, 31, 34n, 52n, 53, 174. Orissa, 178, 188.

INDEX • 211

Ottomans, 101, 154, 165, 166, 192, 197, 201, 202, 203, 204. Bāyezīd II, 192. Mehmet II, 203. Murad III, 203n. Selīm I, 203. Selīm II, 203n. ʿUthmān II, 196n. Pakistan, 87, 114n, 118. Pala-Sena, 178. Pandua (Bengale), 154n, 178-179. Adina (mosquée), 154, 172, 178-185, 189. Bari Dargah (= Shaykh Jalāl al-Dīn Tabrīzī), 178, 180. Chhoti Dargah (= Nūr Quṭb ʿĀlam), 178. Eklakhi (mausolée), 178. Quṭbi ʿĀlam (= Quṭb Shahi Masjid), 178. Satargarh (palais), 178, 179, 182n. Pang-ko la (Bengale), 177. Patan (Gujarat), 135. Penukonda (Andhra Pradesh), 120, 121, 122. Perse / Persia, 31, 179, 180n, 192, 204. Petlad, 119n. Qalandariyya, 121. Al-Qalqashandī, 193n. Qālūn (rāwī, transmetteur de Nāfi), 196n. Quanzhou (Fujian, Chine), 173. Sheng-Yu Si (= Mosque of the Holy Friend), 157, 158. Qunbul (rāwī, transmetteur d’Ibn Kathīr), 196n. Al-Qushayrī, 98. Laṭāʾif al‑ishārāt, 98. Qutb (mosquée, voir Delhi) Quṭb-i ʿĀlam (= Quṭb Shahi Masjid, voir Pandua) Quwwat al-Islam (mosquée, voir Delhi) Raḍiyya (princesse ilkhanide, soeur d’Iltutmish), 116. Rāi Karn Dīv (roi), 120. Raichur (Karnataka), 195. Rajab al-Burquʿī (Shaykh), 19. Rājā Ganesa / Rajah Ganesh (lignée du Bengale), 178, 185. Rajasthan, 51, 53. Rajputs (dynastie), 11, 54, 58, 116. Al-Rāmhurmuzī, ʿAbd al-Raḥmān, 143-144, 148. Al‑Muḥaddith al‑fāṣil bayn al‑rāwī wa al‑wāʿī, 143. Rander, 119n. Rashīd al-Dīn (= Rashīd al-Dīn Faḍl Allāh al-Hamadhānī, vizir), 19, 28, 30n, 53, 74. Kitāb‑i mukātibāt‑i Rashīdī, 19n. Rashtrakutas (royaume), 118. Rasūlides / Rasulids, 157. Ratnaśekharasūri, 129. Riḥla (voir Ibn Baṭṭūṭa) Rukn-i ʿAlām (tombe, voir Multan) Saʿdī, 49. Kulliyāt (CBL 109), 49.

Safavides / Safavids, 165, 166. Shāh Ṭahmasp, 203n. Al-Sāfī al-Hindī Muḥammad b. ʿAbd al-Raḥmān, 19. Safīna‑yi Tabrīz (voir Ibn Masʿūd) Sāhi, 133, 134. Śakra (dieu), 136. Sāmācārī (= Conduite monastique appropriée), 128. Samarcande / Samarqand, 159, 166n. Bibi Khanum (mosquée), 159n. Shad-i Mulk Agha (tombe), 166n. Sampla (Gujarat), 119. Sanaa (Palimpseste / Ṣanʿāʾ palimpsest), 141, 144, 147-148. Ṣandal (voir Muḥammad b. Mubādir Abū Bakr) Sārang (peintre), 135. Al-Sarrāj al-Hindī ʿUmar b. Isḥāq, 19. Satargarh (palais, voir Pandua) Satgaon (port), 177. Sayyid ʿAlāʾ al-Dīn (mausolée, voir Khiva) Sehwan Sharif (Sind), 121n. Şehzade Mustafa (mausolée, voir Brousse) Sekta (Manipur), 177, 188. Seldjoukides / Seljuks, , 30n, 31; Qurʾan, 41, 47. Selīm I (voir Ottomans) Selīm II (voir Ottomans) Shaʿbān (voir Mamelouks) Shad-i Mulk Agha (voir Samarcande) Shāh Fakhr al-Dīn Suhrawardī, 120. Shāh Hyder Suhrawardī, 120. Shāh Maḥmūd Nishāpurī (calligraphe), 203. Shāh Rukh (voir Timourides) Shāh Ṭahmasp (voir Safavides) Shāhnāma (voir Firdawsī) ; Demotte Shāhnāma, 41n ; Great Mongol, 174 ; Jainesque, 177n ; Pseudo-, 48. Shams al-Dīn Ilyās Shāh (voir sultans du Bengale) Shansabanid, 163. Sharqi sultan, 195n. Al-Shāṭibī, Ibrāhīm b. Mūsā (faqīh malékite), 93, 98n. Ḥirz al‑amānī wa wajh al‑tahānī fī al‑qirāʾāt al‑sabʿ (= Matn al‑shāṭibiyya = al‑Shāṭibiyya), 93. Shaykh Aḥmad b. Abū al-Ḥasan (voir Turbat-i Shaykh Jam) Shaykh Jalāl al-Dīn Tabrīzī (= Bari Dargah, voir Pandua) Sheng-Yu Si (voir Quanzhou) Sikandar Shāh (voir sultans du Bengale) Simi Nīshāpūrī, 179. Shiraz (voir Chirāz). Shiva, 121, 122. Sind, 18, 114, 118, 121n. Sijzī, Amīr Ḥasan, 114n. Fawāʾid al-Fuʾād, 114n. Sīrhindī, Yaḥyā b. Aḥmad, 114n. Tārīkh‑i Mubārak Shāhī, 114n. Sistan, 120. Somasiṃha (copiste), 137.

212 • INDEX

Sonargaon (port), 177, 178. Song (dynastie), 173. Sthavirāvali (= Lignée des Anciens), 128. Suhrawardī (= Suhrawardiyya), 12, 120, 121n, 123. Al-Sulamī, 98. Ḥaqāʾiq al‑tafsīr, 98. Sultan Ḥasan (complexe, voir Le Caire) Sultanat du Bengale, 177, 178. Sultans : Ghiyāth al-Dīn ʿAzam Shāh, 179. Shams al-Dīn Ilyās Shāh, 178, 189. Sikandar Shāh, 189. Sultanat de Delhi / Delhi Sultanate (ou Sultans), 12, 18, 19, 49, 53, 54, 114n, 116, 117, 118n, 120, 121, 174, 176, 177, 178, 182, 192, 195n ; Tughluqid, 58. Sultans : ʿAlāʾ al-Dīn Khaljī [Muḥammad Shāh Ier], 19, 178n. Firūz Shāh Tughluq, 155, 177, 178, 180n. Ghiyāth al-Dīn Tughluq [Shāh Ier], 178, 179n. Ibrahim Lodi, 116. Iltutmish (= Shams al-Dīn Iltutmish), 116, 175, 176. Mubārak Shāh (= Muʿizz al-Dīn Mubārak Shāh II, Sayyids), 114n. Muḥammad b. Tughluq (= Ghiyāth al-Dīn Muḥammad Shāh II), 19, 114n, 117, 120, 121n, 155, 159. Muḥammad III Tughluq (= Nāṣir al-Dīn Muḥammad Shāh III), 187. Sultanat du Gujarat, 135, 136. Sultans (= Aḥmad Shāhis), 204. Aḥmad Shāh, 128. Maḥmūd Shāh ‘Bigarha’, 203n. Sultaniyya, 52n, 155. Öljeitü (mausolée), 155. Al-Sūrābādī, Abū Bakr, 160. Tafsīr, 160. Surat, 119n. Al-Suyūṭī, 98n, 148n. Ṭabaqāti Akbarī (voir Niẓām al-Dīn Aḥmad) Ṭabaqāti Nāṣirī (voir Jūzjānī) Tabriz, 28n, 30n, 31n, 41, 47, 50, 52n, 154, 155n. Tāj al-Dīn ʿAlī Shāh (mosquée), 154. Takht-i Sulayman, 51, 173. Talikota (bataille), 121. Tamarshīrīn (défaite), 19n. Tao-i-chih-lio, 176. Tarain (bataille), 116. Tarih‑i Ebü’l‑Fetḥ (voir Ṭūrsūn Beg) Tārīkhi Badāʾūnī (voir Badāʾūnī) Tārīkhi Firishta (voir Firishta) Tarīkhi Firūz Shāhī (voir Baranī) Tārīkhi Mubārak Shāhī (voir Sīrhindī) Al‑Taysir fī al‑qirāʾāt al‑sabʿ (voir al-Dānī) Téhéran / Tehran, 163. Thaïlande / Thailand, 177.

Timourides / Timurids, 18, 44, 51, 93, 164, 165, 166, 177, 192, 195n, 197n, 199, 201, 202, 204. Shāh Rukh, 164. Tamerlan / Tīmūr, 11, 19, 54, 58, 116, 117, 120, 155, 159, 164, 166n, 203. Ting (dynastie), 173. Tirhut (Bihar), 178. Tonwar (= Tomara, lignée rajpute), 116, 117, 118. Trichy (Tamil Nadu), 120. Trikuṇṭha Dvivedi (scribe), 135. Tughluqs, 58, 155, 159, 176, 178n, 195. Voir aussi sultans de Delhi. Tughluqabad, 155. Turbat-i Shaykh Jām (Iran), 163. Shaykh Aḥmad b. Abū al-Ḥasan (tombeau), 163. Turkestan / Turkistan (Kazakhstan), 164. Aḥmad Yasawī (tombeau), 164. Ṭūrsūn Beg, 203. Tarih‑i Ebü’‑‑Fetḥ (= The History of Mehmed the Conqueror), 203. Tz’u-chou (Cizhou), 173n. ʿUbayd al-Qāsim b. Salām, 91n. Ubayy b. Kaʿb, 91n, 98. Faḍāʾil al‑sūra, 98. Udaipur / Udhepur (Rajasthan), 118. Ujjain, 118. ʿUmar (calife), 91n. Umayyades / Umayyads, 18, 118. Urwahi (porte de la citadelle, voir Gwalior) ʿUthmān (calife), 91. ʿUthmān II (voir Ottomans) Valabhī (Gujarat), 128. Varanasi / Benares, 178. Vasanta Vilāsa, 199. Vietnam, 177. Vijayanagar (royaume, roi), 121. Voir aussi HampiVijayanagar. Warsh (rāwī, transmetteur de Nāfīʿ), 196n. Yāqūt al-Mustaʿṣimī, 45n, 52n, 74, 159. Yazd, 119, 166n. Yazdī, Ghiyāth al-Dīn ʿAlī, 117n. Al-Yazdī, Ḥusayn b. Yūsuf (poète soufi), 119n. Yémen / Yemen, 144, 147, 155, 157, 166, 203n. Yüan (dynastie), 157, 176. Zabid (Yémen), 166n. Mosquée du vendredi / Friday Mosque, 166n. Ẓafar Khan (gouverneur du Gujarat), 117. Al-Zarkashī, 98n, 101n. Zhen He, 177. Zhwa-Lu (Tibet), 53.

TABLE DES AUTEURS Sabrina aLiLouche, après des études en linguistique et en histoire de l’art, étudie dans le cadre de son Master les stèles funéraires hafsides d’Algérie, datant du xiiie au xvie siècle, puis les plafonds en bois peint d’Alger datant du xvie au xxe siècle sous la direction de JeanPierre Van Staëvel et d’Éloïse Brac de la Perrière (Université Paris-Sorbonne). Membre du projet consacré au coran de Gwalior entre 2008 et 2012 (UMR 8167-Islam médiéval), elle travaill aujourd’hui sur l’appropriation de palais algérois à la période française sous la direction de Mercedes Volait. Nalini baLbir, agrégée de grammaire, docteur d’état en études indiennes, est professeur d’indologie à l’Université Paris 3 et directeur d’études (philologie moyen-indienne) à l’École pratique des hautes études, Section sciences historiques et philologiques. Elle est membre de l’UMR 7528 Mondes iranien et indien (www.iran-inde.cnrs.fr avec liste complète des publications). Ses champs de recherche incluent le sanskrit, les langues moyen-indiennes (pali, prakrits) et les études jaïnes. Ses publications portent sur tous les aspects du jaïnisme, y compris contemporains. Ces dernières années elle a travaillé plus particulièrement sur la culture manuscrite des jaïns, à la faveur du catalogage de plusieurs collections qu’elle a effectué, s’intéressant aux modes de diffusion de cette culture, à l’utilisation du manuscrit comme objet de distinction sociale par les jaïns, à son illustration, à sa découverte par les chercheurs occidentaux.

Nourane ben azzouna a obtenu un doctorat en histoire de l’art islamique à l’École pratique des hautes études en 2009. Elle a collaboré en tant que chargée d’études pour les arts de l’Islam au projet du Louvre Abu Dhabi (2009-2013). Elle occupe depuis 2013 un poste d’Assistant Professor à l’Institut d’histoire de l’art de l’Université de Vienne. Ses recherches portent principalement sur la calligraphie et les arts du livre dans le monde islamique médiéval. Elle est l’auteur, avec Markus Ritter, de The Golden Qurʾan from the Age of the Seljuks and Atabegs. Cod. arab. 1112 of the Bayerische Staatsbibliothek in Munich, Adeva, 2015 et prépare un second ouvrage intitulé : Aux origines du classicisme : calligraphes et bibliophiles au temps des dynasties mongoles (Iraq et nord-ouest de Iran, 656-814 / 1258-1411). Johanna bLayac est docteur en histoire, spécialiste du monde musulman et de l’Asie du Sud. Elle a consacré sa thèse à l’étude des premières inscriptions arabes et persanes du sous-continent indien, dans le but de documenter la formation des sociétés indo-musulmanes, mettant en évidence des processus de transculturation. Elle a participé aux travaux de plusieurs équipes de recherche au sein du Centre d’Études de l’Inde et de l’Asie du Sud (CNRS-EHESS), notamment à ceux de la Mission interdisciplinaire française du Sindh (Pakistan), et a enseigné durant deux ans, en 2011-2012 et 2012-2013, l’histoire du monde musulman classique et du monde arabe moderne et contemporain au sein de la Licence de Langue et Civilisation Arabe de l’Université Paris 8 Saint-Denis. Elle a contribué au volume de référence Calligraphy and Architecture in the Muslim World, édité par Mohammad Gharipour et Irvin Cemil Schick, Edinburgh University Press, 2013. Éloïse brac de La Perrière est maître de conférences en histoire de l’art du monde  islamique à l’université Paris-Sorbonne. Après avoir dirigé durant quatre ans (2008-2012) le programme de recherche « Autour du coran de Gwalior » (UMR 8167 – Islam médiéval), elle est aujourd’hui co-responsable auprès d’Annie Vernay-Nouri (Bibliothèque nationale de  France)  d’un  projet  scientifique  consacré  aux  manuscrits  à  peintures  des  fables  de  Kalila wa Dimna (UMR 8167 / Bibliothèque nationale de France), projet qui a donné lieu, entre septembre 2015 et janvier 2016, à l’exposition « Paroles de bêtes (à l’usage des princes) » à l’Institut du monde arabe. Spécialiste des arts du livre, et plus particulièrement de la production livresque dans l’Inde islamique pré-moghole à laquelle elle a consacré un ouvrage, L’Art du livre dans l’Inde des sultanats (PUPS, 2008), elle mène parallèlement

214 • TABLE DES AUTEURS

des recherches sur la céramique indienne dans le cadre d’un projet scientifique dont elle  assure la direction auprès d’Yves Porter (Université Aix-Marseille / UMR 7298) et Hélène Renel (UMR 8167). Monique burési, après des études de russe et d’arabe, s’est tournée vers l’histoire de l’art. Documentaliste scientifique aux départements des Arts de l’Islam et des Antiquités  orientales du musée du Louvre, elle a participé aux catalogues des expositions « Chefsd’œuvre de la collection des arts de l’Islam du musée du Louvre » (Riyad, 2006) et « Chefsd’œuvre islamiques de l’Aga Khan Museum » (Paris, 2007), pour lequel elle a assuré la collaboration éditoriale. En 2014, pour le musée du Louvre, elle a été, avec Sophie Cluzan, commissaire de l’exposition « Voués à Ishtar. Syrie, janvier 1934, André Parrot découvre Mari » qui s’est tenue à l’Institut du monde arabe. Frantz chaiGne, ancien élève de l’École normale supérieure de Saint-Cloud, est agrégé de chimie. Après avoir enseigné à l’École normale supérieure de la rue d’Ulm, il dispense désormais des cours de sciences physiques appliqués au design et aux métiers d’art à l’École Boulle. Parallèlement, il a soutenu à l’université Paris-Sorbonne, une thèse en histoire de l’art islamique sur l’enluminure dans l’empire ilkhanide. Il a participé aux projets pluriannuels de l’UMR 8167 sous la direction d’Éloïse Brac de la Perrière, le premier portant sur l’étude d’un coran copié à Gwalior en 1399, l’autre sur les manuscrits à peintures de Kalila wa Dimna. Il est le co-auteur de deux articles consacrés au décor du coran de Gwalior, dont : Brac de la Perrière, Chaigne, Cruvelier, « The Qurʾan of Gwalior, Kaleidoscope of the Arts of the Book », dans Treasures of the Aga Khan Museum – Arts of the Book & Calligraphy, catalogue d’exposition, 2010, p. 114-123. Il présente aussi régulièrement des communications publiées aux colloques de la Ernst-Herzfeld-Gesellschaft : « Le décor enluminé sous les Īl-khānides : entre assimilation et innovation, de l’Iraq à  la Chine », dans Beiträge zur Islamischen Kunst und Archäologie, Band 3, 2012, p. 253-265. Il a pris part également à l’étude des céramiques islamiques du Musée de Lyon. Mathilde cruveLier est chercheuse indépendante, spécialiste des arts du livre chrétien arabe sous le sultanat mamluk. Chercheuse associée au département des manuscrits orientaux de la Bibliothèque nationale de France en 2008-2009, elle a été membre du projet portant sur le coran de Gwalior entre 2008 et 2012. Elle est l’auteur avec Éloïse Brac de la Perrière, Frantz Chaigne de « The Qurʾan of Gwalior, Kaleidoscope of the Arts of the Book », dans Treasures of the Aga Khan Museum – Arts of the Book & Calligraphy, catalogue d’exposition, 2010 ; elle a également publié en 2012 un article consacré aux « Emprunts post-ilkhanides dans l’enluminure mamluk du ixe / xve siècle », Chronos 26, 2012, p. 117-155. Ghazaleh esmaiLPour qouchani, après des études en arts appliqués à l’université Honar de Téhéran (Tehran University of Art), a obtenu une bourse d’études pour mener à terme un doctorat en France. Depuis 2008, elle participe aux programmes de recherches consacrés au coran de Gwalior et aux manuscrits de Kalila wa Dimna sous la direction d’Éloïse Brac de la Perrière. Sa thèse portant sur l’emploi de la couleur bleue dans la miniature persane du xvie siècle a été soutenue en 2014 à l’université Paris-Sorbonne. Elle est actuellement professeure assistante à l’Université Honar de Chiraz (Shiraz University of Art). Finbarr Barry fLood est professeur à l’Institute of Fine Arts, ainsi qu’au département d’histoire de l’art, New York University (NYU). Il travaille à l’heure actuelle sur un projet de recherche intitulé « Islam and Image: Polemics, Theology and Modernity » (Islam et image : polémiques, théologie et modernité). Parmi ses thèmes de recherche, les échanges interculturels dans la sphère islamique occupent une place importante, comme en attestent ses deux derniers ouvrages : Globalizing Cutures: Art and Mobility in the Eighteenth Century, publié en 2011 (avec Nebahat Avcıoğlu) et Objects of Translation: Material Culture and Medieval “Hindu-Muslim” Encounter, en 2009.

TABLE DES AUTEURS • 215

Asma hiLaLi est chercheuse associée au sein de l’Unité d’études coraniques (Qurʾānic  Studies Unit), département de recherches académiques et publications (Department of Academic Research and Publications), à l’Institut of Ismaili Studies à Londres. Elle a étudié l’arabe, sa littérature et la civilisation islamique à l’Université de Tunis 1. Boursière de la Chancellerie des Universités de Paris pour les Lettres et les Sciences humaines, elle intègre en 2001 l’École pratique des hautes études où elle soutient une thèse de doctorat, intitulée  « La théorie de l’authenticité dans les sciences du Hadith entre le premier et le sixième siècle de l’Islam » en 2004. Ses publications abordent le problème de la transmission des textes religieux aux débuts de l’Islam. Elle a entreprit récemment l’édition et l’étude du célèbre manuscrit 01.27-1 Dār al-Makhṭūṭāt, Ṣanʿāʾ Yémen, connu comme « Le palimpseste de Sanaa ». Yves Porter, est professeur d’art musulman au département d’Histoire de l’Art et d’Archéologie d’Aix Marseille Université (AMU) et membre du Laboratoire d’Archéologie méditerranéenne médiévale et moderne (LA3M). Spécialiste des arts et techniques dans les domaines de l’Iran, de l’Inde musulmane et de l’Asie centrale, il a publié de nombreux articles dans des revues à grande diffusion, comme dans des publications scientifiques, ainsi que des ouvrages, dont Peinture et arts du livre, Téhéran, 1992 ; Les Iraniens, Paris, 2006 ; L’Inde des sultans, Paris, 2009 ; Le prince, l’artiste et l’alchimiste. La céramique dans le monde iranien, xe-xviie siècle, Paris, 2011. Il a participé à des émissions de télévision (Des racines et des ailes), de radio, et collabore régulièrement avec des musées et collections d’art islamique dans le monde entier. Patricia roGer-Puyo est ingénieur de recherche à l’IRAMAT / UMR5060 du CNRS. Docteur en Sciences Physiques spécialité physique radiologique de l’Université Paul Sabatier de Toulouse. Après avoir exercé le métier de radiophysicienne dans le milieu hospitalier, une conversion professionnelle la dirige vers le Centre Ernest Babelon du CNRS où elle succède  à Bernard Guineau en 2002 afin de poursuivre les recherches sur les archéomatériaux à  l’aide de méthodes non destructrives sur le thème pigments et colorants. Le livre manuscrit médiéval constitue le principal sujet d’étude de ses recherches, qui s’effectuent au moyen de méthodes d’analyses spectrométriques non destructives in situ. Une partie de ces dernières concernent les manuscrits arabes. Simon rettiG est assistant conservateur à la Freer Gallery of Art et Arthur M. Sackler Gallery, le musée d’art asiatique de la Smithsonian Institution à Washington (DC). Il est titulaire d’un doctorat en histoire de l’art de l’Université de Provence Aix-Marseille (2011). Il a été commissaire de l’exposition « Nastaʿliq: The Genius of Persian Calligraphy » et prépare actuellement le catalogue des œuvres persanes sur papier de la Freer Gallery of Art ainsi qu’une monographie sur le manuscrit jalayiride Khusraw et Shirin de Niẓāmī, un des chefs-d’œuvre des collections du musée.

TABLE DES MATIÈRES

Tableau de translitération

5

Abréviations

6

Ressources en ligne (consultés le 01/08/2015)

6

Remerciements

7

L’Inde des sultanats à la fin du xive siècle (carte)

9

Éloïse Brac de la Perrière Introduction

11

I – LE CORAN DE GWALIOR Frantz chaigne, Mathilde cruvelier The ornamentation of the Gwalior Qurʾan, between diachronic legacies and geographic confluences

17

Nourane Ben azzouna, Patricia roger-Puyo The Gwalior Qurʾan: Archaeology of the manuscript and of its decoration: A preliminary study

57

Sabrina alilouche, Ghazaleh esmailPour Qouchani Les gloses marginales et le fālnāma du coran de Gwalior, témoignages des usages multiples du coran dans l’Inde des sultanats

85

II – CONTEXTES Johanna Blayac Contextualizing the Gwalior Qurʾan: Notes on Muslim Military, Commercial and Mystical Routes in Gwalior and India before the 16th century

113

Nalini BalBir Kalpasūtras et Corans : Réflexions sur l’écriture et la peinture de manuscrits jaina du Gujarat aux xive-xvie siècles

127

Asma hilali Qurʾan Manuscripts and their Transmission History. Preliminary Remarks

139

III- LES MANUSCRITS INDO-PERSANS ET LE MONDE ISLAMIQUE Finbarr Barry Flood Eclecticism and Regionalism: The Gwalior Qurʾan and the Ghurid Legacy to Post-Mongol Art

153

Yves Porter Lotus flowers and leaves, from China and Iran to the Indian Sultanates

171

Simon rettig A ‘Timurid-like Response’ to the Qurʾan of Gwalior? Manuscript W563 at the Walters Art Museum, Baltimore

191

INDEX

207

TABLE DES AUTEURS

213

Dans la même collection Volume 1 | 2007 Job, ses précurseurs et ses épigones, par Maria Gorea. Volume 2 | 2008 D’Ougarit à Jérusalem. Recueil d’études épigraphiques et archéologiques offert à Pierre Bordreuil, édité par Carole Roche. Volume 3 | 2008 L’Arabie à la veille de l’Islam. Bilan clinique (Actes de la table ronde tenue au Collège de France, Paris, 28-29 août 2006), édité par Jérémie Schiettecatte en collaboration avec Christian Julien Robin. Volume 4 | 2009 Sabaean Studies. Archaeological, epigraphical and historical studies, edited by Amida M. Sholan, Sabina Antonini, Mounir Arbach. Volume 5 | 2009 Les échanges à longue distance en Mésopotamie au Ier millénaire. Une approche économique, par Laetitia Graslin-Thomé. Volume 6 | 2011 D’Aden à Zafar, villes d’Arabie du sud préislamique, par Jérémie Schiettecatte. Volume 7 | 2012 Dieux et déesses d’Arabie : images et représentations (Actes de la table ronde tenue au Collège de France, Paris, 1er-2 octobre 2007), édité par Isabelle Sachet en collaboration avec Christian Julien Robin. Volume 8 | 2012 Alessandro de Maigret, Saba’, Ma’în et Qatabân. Contributions à l’archéologie et à l’histoire de l’Arabie ancienne,  choix d’articles scientifiques préparé par Sabina Antonini et Christian Julien Robin. Volume 9 | 2012 Scribes et érudits dans l’orbite de Babylone (travaux réalisés dans le cadre de l’ANR Mespériph 2007-2011), édité par Carole Roche-Hawley et Robert Hawley. Volume 10 | 2012 South Arabian Art. Art History in Pre-Islamic Yemen, par Sabina Antonini de Maigret. Volume 11 | 2012 L’Orient à la veille de l’Islam. Ruptures et continuités dans les civilisations du Proche-Orient, de l’Afrique orientale, de l’Arabie et de l’Inde à la veille de l’Islam (Actes de la table ronde tenue au Collège de France, Paris, 17-18 novembre 2008), édité par Jérémie Schiettecatte en collaboration avec Christian Julien Robin. Volume 12 | 2013 Entre Carthage et l’Arabie heureuse. Mélanges offerts à François Bron, édité par Françoise Briquel Chatonnet, Catherine Fauveaud et Iwona Gajda. Volume 13 | 2013 Bijoux carthaginois III. Les colliers. L’apport de trois décennies (1979-2009), par Brigitte Quillard. Volume 14 | 2013 Regards croisés d’Orient et d’Occident. Les barrages dans l’Antiquité tardive (Actes du colloque tenu à Paris, Fondation Simone et Cino del Duca, 7-8 janvier 2011, et organisé dans le cadre du programme ANR EauMaghreb), édité par François Baratte, Christian Julien Robin et Elsa Rocca.

Volume 15 | 2014 Paradeisos. Genèse et métamorphose de la notion de paradis dans l’Antiquité (Actes du colloque international), sous la direction d’Éric Morvillez. Volume 16 | 2015 Devins et lettrés dans l’orbite de Babylone, sous la direction de Carole Roche-Hawley et Robert Hawley. Volume 17 | 2015 Les Jafnides. Des rois arabes au service de Byzance (vie siècle de l’ère chrétienne) (Actes du colloque de Paris, 24-25 novembre 2008), sous la direction de Denis Genequand et Christian Julien Robin. Volume 18 | 2015 Figures de Moïse, sous la direction de Denise Aigle et Françoise Briquel Chatonnet.

ORIENT

MÉDITERRANÉE

19

UMR 8167, Orient et Méditerranée – Textes, Archéologie, Histoire CNRS, Université Paris-Sorbonne, Université Paris 1 Panthéon-Sorbonne, École pratique des hautes études, Collège de France

L

n September 24, 1398, the army of Timur Lang or Tamerlane took the city of Delhi by storm. This date marked the fall of the Delhi Sultanate; it had already been weakened by the emancipation of some of its territories, which had been transformed into small independent sultanates, and by endless conflicts with neighboring Hindu kingdoms. It would take nearly fifty years to recover from the ransacking of its capital and to free itself from Timurid control. It was precisely at the end of the 14th century, in this context of chaos, that a copy of the Qurʾan was completed in the fortress of Gwalior, a few hundred kilometers from Delhi. This manuscript, today kept in the collections of the Aga Khan Museum in Toronto, contains remarkable decorative elements and a hermeneutic system of commentaries. The glosses, which refer to various levels of reading, suggest the existence of complex memorization techniques and of multiple functions for the Qurʾanic manuscript. The flamboyant illuminations, which have no known equivalent in any other works, reveal a number of different influences. Between 2008 and 2014, a research program (UMR 8167- Medieval Islam) focused on this extraordinary work, the analysis of which required the skills of specialists in fields including Qurʾanic studies, history, art history, codicology, paleography and divinatory and magical practices. Their hypotheses and conclusions are gathered in this book.

ISBN 978-2-7018-0443-9

Polysémie d’un manuscrit à peintures

O

LE CORAN DE GWALIOR Polysémie d’un manuscrit à peintures sous la direction de

Éloïse Brac de la Perrière Monique Burési

LE CORAN DE GWALIOR –

e 24 septembre 1398, Delhi est prise d’assaut par les troupes de Tı¯mu¯r Lang, notre Tamerlan. Cette date marque la chute du sultanat de Delhi, déjà ébranlé par l’émancipation d’une partie de ses territoires, transformés dès lors en de petits sultanats indépendants et affaibli par d’incessants conflits avec les royaumes hindous voisins. Il lui faudra près de cinquante ans pour se relever du terrible sac de sa capitale et s’affranchir de la tutelle timouride. C’est précisément en cette même fin du xive siècle, dans ce contexte chaotique, qu’est achevée, dans la forteresse de Gwalior, à quelques centaines de kilomètres de Delhi, la copie d’un coran dont l’existence même soulève un grand nombre de questions. Ce manuscrit, aujourd’hui conservé dans les collections de l’Aga Khan Museum (Toronto), est doté d’extraordinaires décors et d’un savant système de commentaires herméneutiques. Ces gloses, qui font appel à différents niveaux de lecture, supposent des techniques de mémorisation complexes et des usages variés du manuscrit coranique. Les flamboyantes enluminures, dont on ne connaît aucun équivalent, renvoient à de multiples origines. Entre 2008 et 2014, un programme de recherche (UMR 8167- Islam Médiéval) a été consacré à cette œuvre hors du commun, dont l’analyse conjugue les compétences de spécialistes des études coraniques, de l’histoire, de l’histoire de l’art, de la codicologie, de la paléographie, des pratiques divinatoires et magiques. Leurs hypothèses et leurs conclusions sont réunies dans le présent ouvrage.

19

Éditions de Boccard