Les Sogdiens en Chine 2855396530

Table of contents :
Colophon
Sommaire
Introduction
粟特人在中国 Les Sogdiens en Chine : quelques réflexions de méthode _ Étienne DE LA VAISSIÈRE
Nouvelles découvertes de lits funéraires sogdiens de Chine
Carvings on the Stone Outer Coffin of Lord Shi of the Northern Zhou _ YANG Junkai 杨军凯
Investigations on the Chinese version of the Sino-Sogdian bilingual inscription of the tomb of Lord Shi _ SUN Fuxi 孫福喜
The Sogdian version of the new Xi’an inscription _ Yutaka YOSHIDA 吉田豊
Réflexions sur les pratiques religieuses désignées sous le nom de xian 祆 _ Pénélope RIBOUD
Études iconographiques
Hutengwu and Huxuanwu Sogdian dances in the Northern, Sui and Tang Dynasties _ ZHANG Qingjie 張慶捷
The Hejiacun Treasure and Sogdian Culture _ QI Dongfang 齐东方
The self-image of the Sogdians _ Frantz GRENET
From Stone to Silk : Intercultural Transformation of Funerary Fumishings Among Eastem Asian Peoples around 475-650 CE _ Angela SHENG
Marchands, soldats, vignerons
Towards a new édition of the Sogdian Ancient Letters : Ancient Letter 1 _ Nicholas SIMS-WILLIAMS
Caravans and Caravan Leaders in Palmyra _ Albert E. DIEN
Sabao or Sabo : Sogdian Caravan Leaders in the Wall-Paintings in Buddhist Caves _ RONG Xinjiang 榮新江
Sogdian merchants and Chinese Han merchants during the Tang Dynasty _ ARAKAWA Masaharu 荒川正晴
Military officers of Sogdian origin from the late Tang Dynasty to the period of Five Dynasties _ MORIBE Yutaka 森部豊
Čākar sogdiens en Chine _ Étienne DE LA VAISSIÈRE
Un vestige vivant de la présence sogdienne en Chine du Nord : le vignoble du Shanxi _ Éric TROMBERT
Le Bassin du Tarim
The Impact of the Silk Road Trade on a Local Community : The Turfan Oasis, 500-800 _ Valerie HANSEN
Documenting Sogdian society at Turfan in the seventh and eighth centuries : Tang Dynasty census records as a window on cultural distinction and change _ Jonathan Karam SKAFF
Non-Han ethnie groups and their settlements in Dunhuang during the Late Tang and Five Dynasties _ Zheng Binglin 郑炳林
Sogdians in Kucha, a study from archaeological and iconographical material _ Etsuko KAGEYAMA 影山悦子
Sogdiens, Chinois et Tures
A Survey of the Turkic Cemetery in Little Khonakhai _ LIN Meicun 林梅村
La monétarisation de la société türke (VIe -IXe siècle) Influence chinoise, influence sogdienne _ François THIERRY
Index
Abstracts

Citation preview

Les Sogdiens en Chine

ÉTUDES THÉMATIQUES 17

Les Sogdieiîis en Chine Sous la direction de

Étienne DE LA VAISSIÈRE et Éric TROMBERT

École française d’Extrême-Orient Paris, 2005

Avec la participation du Centre de recherche « Civilisation chinoise », UMR 8583 de l’École pratique des hautes études et du Centre national de la recherche scientifique, sis au Collège de France, 52 rue du Cardinal-Lemoine, 75005 Paris.

© 2005, École française d’Extrême-Orient 22, avenue du Président-Wilson, 75116, Paris Cedex 16, France. ISSN 1269-8067 ISBN 2-85539-653-0

Pour Boris Marshak et Zhang Guangda

SôWÆaIRE

Introduction Étienne Les Sogdiens en Chine : quelques réflexions de méthode.

DE LA VAISSIÈRE

11

Nouvelles découvertes de lits funéraires sogdiens de Chine YANG Junkai 楊 軍 凱 Carvings on the stone outer coffin of Lord Shi of the Northern Zhou.

21

SUN Fuxi 孫 ネ 喜 Investigations on the Chinese version of the Sino-Sogdian bilingual inscription of the tomb of Lord Shi.

47

Yoshida Yutaka 吉田 豊

The Sogdian version of the new Xi'an inscription. Riboud Pénélope

.

Réflexions sur les pratiques religieuses désignées sous le nom de xian 袄

57

73

Etudes iconographiques Qingjie 張 慶捷 Hutengwu and Huxuanw Sogdian dances in the Northern, Sui and Tang Dynasties.

Zhang

93

Ql Dongfang 齊 東方 The Hejiacun treasure and Sogdian culture.

107

GRENET Frantz The self-image of the Sogdians.

123

8

_

Les Sogdiens en Chine

SHENG Angela From stone to silk: mtercultural transformation of funerary fomishings among Eastem Asian peoples around 475-650 CE.

141

Marchands, soldats^ vignerons SIMS-WILLIAMS Nicholas Towards a new édition of the Sogdian Ancient Letters: Ancient Letter 1.

181

Dien Albert

Caravans and caravan leaders in Palmyra.

195

Rong

Xinjiang Sabao or Sabo: Sogdian caravan leaders in the wall-paintings in Buddhist caves.

[ IEW Arakawa Masaharu Sogdian merchants and Chinese Han merchants during the Tang Dynasty.

207

23 1

g MORIBE Yutaka Military offïcers of Sogdian origin from the late Tang Dynasty to the period of Five Dynasties. 243 Étienne Câkar sogdiens en Chine.

DE LA VAISSIÈRE

255

Trombert Éric Un vestige vivant de la présence sogdieime en Chine du Nord : le vignoble du Shanxi. 261

Le Bassin du Tarim Hansen Valerie The impact of the Silk Road trade on a local community: The Turfan oasis, 500-800.

283

SKAFF Jonathan Documenting Sogdian society at Turfan in the seventh and eighth centuries: Tang Dynasty census records as a window on cultural distinction and change. 311 ZHENG Binglin MP Non-Han ethnie groups and their settlements in Dunhuang during the Late Tang and Five Dynasties.

343

9

Sommaire

KAGEYAMA Etsuko 影山 悦子 Sogdians in Kucha, a study from archaeological and iconographical material. 363

Sogdiens, Chinois et Tures LIN Meicun 林梅村 A survey of the Turkic cemetery in Little Khonakhai.

377

THIERRY François La monétarisation de la société türke (Vïe-IXe siècle). Influence chinoise, influence sogdienne.

397

Index

419

ABSTRACTS

439

Les Sogdiens en Chine : quelques réflexions de méthode Étienne de la Vaissière

Figures extrêmement familières des villes de Chine sous les dynasties du Nord, les Sui et les Tang, les Sogdiens sont Fobjet de multiples travaux depuis quelques années. La multiplication des fouilles de sauvetage autour des grandes villes chinoises a donné un essor décisif à la recherche en Chine, qui s’est combiné aux travaux issus de l’historiographie soviétique, japonaise et occidentale sur l’Asie centrale. Ce contexte favorable a permis l’organisation du colloque de Pékin d'avril 2004 consacré aux Sog¬ diens en Chine, et rassemblant chercheurs chinois, japonais et occidentaux. Organisé à Paris par Éric Trombert (Centre National de la Recherche Scientifique) et moi-même, (Université de Pékin), Alain Arrault, Michela assistés à Pékin de Rong Xinjiang (Bibliothèque Bussotti (École Française d Extrême- Orient), et Zhang Zhiqing Nationale de Chine), ce colloque a rassemblé vingt-huit communications et plusieurs centaines de participants à la Bibliothèque Nationale de Chine du 23 au 25 avril 2004. Il a été financé grâce aux parrainages généreux de l’École Française d’Extrême-Orient, du Centre National de la Recherche Scientifique, de l’Ambassade de France, et surtout de la Chiang Ching-kuo Foundation for International Scholarly Exchange, et placé sous le et Chen Li bienveillant patronnage de Jacques Gemet, Zhang Guangda prendre recueil dans sa ce accepté de a Enfin, l’École Française d ’Extrême-Orient collection Études thématiques. Les articles ici publiés ne reprennent, pour des raisons diverses, qu’une partie des communications présentées au colloque, mais reflètent la diversité des questions actuellement posées. Ces quelques pages d'introduction ont pour seul but d’esquisser un cadre théorique dans lequel replacer ces interrogations. 5

La question est simple : qu’est-ce qu’un Sogdien en Chine ? Question qui peut être précisée ainsi : qu'est-ce que l'identité ethnique sogdienne ? comment survit-elle en contexte chinois ? comment sait-on qu'un personnage attesté dans les sources chinoises est un Sogdien ? Il n’y a pas lieu de douter qu'il y ait eu une identité géographique et ethnique sogdienne dans le regard des autres peuples. Ce n’est pas une création de l’historiographie contemporaine : les sources aussi bien chinoises qu’arabes ou le définit précisément : arméniennes identifient un pays sogdien. Xuanzang « Depuis la ville de la rivière Suye [Suyab] jusqu’au pays de Jieshuangna SSP [Sogdiane] et les gens portent ce même nom. [Kesh], le pays s’appelle Suli

12

Les Sogdiens en Chine

L’écriture et la langue parlée sont également appelées ainsi 1 . » Les Sogdiens eux-mêmes se définissent comme tels : le roi de Penjikent, Dewâstïc, prétend dans sa correspondance être roi de Sogdiane sywôyk MLK’2 Un caravanier sogdien du Gansu au début du IVe siècle parle à son correspondant de « nombreux Sogdiens syw6[y](k)3 ». Il y a bien une identité sogdienne, ce que confirme la découverte du terme Sogdikestan sywôyk-stn dans l’épitaphe de Wirkak publiée par Yutaka Yoshida dans ce recueil, quelqu’en soit la signification précise4. Cette identité transcende un cadre politique éclaté, constitué en Sogdiane d’autant d’Etats indépendants qu’il y a de grandes zones cultivables. Mais le transfert de cette identité dans l’émigration en Chine est un phénomène complexe, obéissant à de multiples facteurs. Le processus migratoire devrait renforcer l’identité commune face à celle du pays d’accueil. Pourtant, du Ve au VIIIe siècle, et contrairement à la période précédente, les Chinois donnent aux émigrants des noms de famille reflétant la ville d’origine — Kang pour Samarcande, An pour Boukhara... neuf noms en tout, les jiu xing Aft et non l’identité sogdienne. Lorsqu’ils veulent générali¬ ser, ils n’utilisent pas le mot Suie ren Sogdien, mais Hu ^|, vieux terme appli¬ qué initialement aux nomades du Nord, et progressivement restreint au sens d’Iranien oriental, qui inclut pêle-mêle les Khotanais, les Bactriens, les Khorezmiens et les Sog¬ diens, voire jiu xing hu Hu aux neufs noms de famille. Même si Samarcande est décrite par Xuanzang comme le cœur des pays Hu5 il faut souligner que les sources chinoises n’utilisent pas les mêmes catégories que nous. L’identité ethno-linguistique sogdienne en Chine n’est pas clairement différenciée. Minorité parmi tant d’autres dans une Chine majoritairement Han, les Sogdiens sont soumis aux courants très puissants de l’acculturation. Dès la deuxième génération pas¬ sée en Chine se posent les problèmes croisés du maintien de l’identité d’origine et de l’assimilation, étroitement dépendant de la force des structures communautaires, de la part des mariages mixtes, du maintien de la langue ou de la religion d’origine, etc. Or l’un des critères les plus utilisés pour identifier les Sogdiens dans les sources chinoises est de recourir aux noms de famille, c’est-à-dire à un aspect hérité de la personnalité sociale, qui ne signifie nullement une adhésion culturelle et ne dit strictement rien sur tous ces aspects identitaires. De surcroît, le critère du patronyme fait disparaître toutes les filiations en ligne maternelle : en tenant compte du seul nom de famille, Yu Hong 5A ne devrait pas être un Sogdien, alors que l’iconographie de son lit funéraire, sa fonction de sabao comme son prénom, Mo fan MS, transcription de Mâxfam6, prouvent sans ambiguïté que l’un de ses ascendants, peut-être sa mère, devait l’être et que Yu Hong l’est resté culturellement. Inversement, la famille Cao W qui dirige Dunhuang à partir de 914 n’a pour toute preuve de son éventuelle origine sogdienne que son seul nom, ce qui est insuffisant. 3.

1

Xuanzang, 1991, p. 8. Grenet, de la Vaissière, 2002, p. 156-7. 3 Grenet, Sims-Williams, de la Vaissière, 1998, p. 93. 4 Voir l’article de Yutaka Yoshida ci-dessous et l’intéressante remarque de Pavel Lur’e à ce propos. 5 Xuanzang, 1991, p. 87-8. 6 Yoshida, Kageyama, 2002, p. 44-9. 2

Étienne de la Vaissière

_

Introduction

13

Plutôt que de recourir au nom de famille, et si l’on n’a pas la chance de disposer d’une épitaphe ou d’une biographie explicite, il faut croiser les critères, comme l’a fait J. Skaff à partir des documents de Turfan : par exemple un prénom sogdien, montrant une pratique maintenue de la langue et des liens familiaux en milieu sogdien au moins à épouse une la génération précédente, un mariage en milieu sogdien — si un Shi il y a quelque probabilité, mais aucune certitude, qu’il s’agisse bien de SogKang diens , une activité économique ou sociale connue dans les sources pour être de celles que pratiquaient beaucoup les Sogdiens — un Kang se faisant représenter sur les reliefs de sa tombe échangeant de l’orfèvrerie occidentale contre de la soie avec des officiels chinois a de grandes chances d’être sogdien7 ; de même l’appartenance à une guilde spécialisée dans les produits précieux où la plupart des membres portent des noms sog¬ on diens, ou encore un témoignage archéologique de liens maintenus avec l’Ouest place rarement une pièce d’or byzantine dans la bouche d’un mort par pur hasard en Chine8. Cela n’assure pas de l’identité, mais permet au moins de proposer des hypothè¬ ses plausibles en ce sens. Etre sogdien n’est pas une donnée patronymique mais une pratique sociale. La question fondamentale est : avec qui s’associe-t-on ? Tout ce qui et le nom de famille n’en fait pas partie peut servir peut contribuer à y répondre d’indice. L’acculturation ne concerne pas uniquement les rapports avec la majorité chinoise. L’importance de l’immigration sogdienne en direction de la Chine à partir du Ve siècle est venue noyer dans son flux une composition beaucoup plus diverse. C’est particulièrement clair avec l’exemple du lit funéraire déposé au Musée Guimet et pré¬ senté lors du colloque par Pénélope Riboud : il s’agit à mon sens d’un exemple parfait des croyances non-sogdiennes, sans doute ici du Gandhara ou du Tokharistan, mais qui s’exprime dans le langage iconographique dominant dans les milieux émigrés en Chine, celui des Sogdiens9. De plus, l’immigration venue de l’ouest est numériquement et culturellement dominée par les Sogdiens, mais elle vient se mêler à d’autres flux, comme l’exemple de Yu Hong, cité ci-dessus, en témoigne : Lin Meicun a montré que son père est un Buluoji , une ethnie du Shanxi non-Han10. Éric Trombert a égale¬ ment souligné combien les Hu responsables de l’introduction de la vigne au Shanxi for¬ ment en fait un conglomérat ethnique sous forte influence sogdienne mais comportant une bonne part d’éléments steppiques. L’expression la plus exacte pour décrire ces milieux est donc bien celle Hu mêlés. L’appelation « Sogdien » qu’emploient les Chinois de l’époque : za hu est commode et correspond au fait démographique et culturel dominant. Elle doit être conservée. Mais elle ne doit pas conduire à sous-estimer la complexité des rapports sociaux et culturels enjeu. Paradoxalement, le problème méthodologique est moins aigu durant la première pé¬ riode de la présence sogdienne en Chine, jusqu’au IVe siècle, simplement parce que la rareté des sources disponibles pour reconstituer l’histoire de ces communautés en Chine

—

—

—

—

7

C’est le cas sur les reliefs de la tombe de Kang Ye Thierry de Crussol, 2000, p. 323-329. 9 Musée Guimet, Lit de pierre... 2004. 10 Lin Meicun, 2002, p. 70 et suiv. 8

découverte en 2004 à Xi’an.

14

Les Sogdiens en Chine

ne permet pas de descendre au niveau des individus. Les sources se contentent de désigner des groupes ethniques et rassemblent sous Fappelation Hu du Kangju 康居胡 aussi bien les Sogdiens sédentaires que les tribus nomades situées légèrement plus au nord, sur le Syr Darya. C'est ainsi que l'on peut identifier deux textes importants pour reconstituer l'histoire des Sogdiens en Chine : des ambassades du Kangju se présentent à la cour en 29 et 11 avant notre ère avec le mot de commerce à la bouche, et une communauté sogdienne importante est installée à Wuwei, au cœur du Gansu, en 227. Lorsque quelques personnages sont individualisés, tels Kang Senghui 康 僧 會, le fa¬ meux moine bouddhiste, elles précisent que « Kang Senghui avait pour ancêtres des gens originaires du Kangju11 ». Sans les Anciennes lettres sogdiennes qui démontrent la présence en 313 d5un vaste réseau marchand sogdien au Gansu et dans les capitales chinoises, ces maigres témoignages ne permettraient pas de dire grand-chose sur Pimportance de cette première période de présence sogdienne en Chine. Aussi est-il d5autant plus heureux que Nicholas Sims-Williams nous ait proposé lors du colloque une traduction commentée d'une de ces Anciennes Lettres, le seul texte permettant de voir de Fintérieur le fonctionnement d5une famille de marchands caravaniers sur les marches chinoises. C5est des dynasties du Nord aux Sui qu'a lieu la transition entre rappelation Hu du Kangju et les neufs noms de famille des Sogdiens, forgés en réutilisant d'anciens noms dévolus traditionnellement à des peuples occidentaux pour différencier les natifs des differentes villes sogdiennes : ainsi An 安,anciennement utilisé pour les Parthes arsacides et désormais appliqué aux Boukhariotes. On connaît mal ce passage, on a même contesté qu'il ait eu lieu, à tort, pour les familles An du moins, même s'il est certain que plusieurs vieilles familles occidentales en Chine remontent à des familles arsacides et que coexistent donc dans la Chine des Wei aux Tang des An iraniens et des An de Boukhara12. On connaît également fort mal les procédés de dénomination individuelle : il est trop simple d'imaginer que les migrants sogdiens recevaient toujours un nom de famille à la frontière de la part du premier officiel chinois qu'ils rencontraient, et que celui-ci le faisait en fonction de la ville d'origine, modèle tout théorique de r époque Tang. La Chine est politiquement éclatée, et les règles des Qi ou des Wei ne valent pas forcément pour les Liang, voire pour les Qu 麴 de Gaochang. On connaît de nombreux contre-exemples : la musique de Koutcha est ainsi introduite en Chine au Ve siècle par des Cao 曹, l'un des noms théoriquement sogdiens. Mais ce sont certainement des Koutchéens13, même si Rong Xinjiang et Etsuko Kageyama ont prouvé lors du colloque la présence de Sogdiens à Koutcha. Les changements de nom ne sont pas rares, et il semble bien qu'An Lushan 安祿山 ait été ainsi au départ Kang Lushan. La genèse du système des « Neufs noms des Sogdiens » mériterait certainement d'être étudiée en détail dans le double contexte des procédés de dénomination propres au monde chinois et des stratégies des migrants, indépendamment même de toutes les questions d'acculturation évoquées plus haut.

11

Hui Jiao, Gaoseng zhuan, 1927, p. 325. Forte, 1996, p. 645-652. 13Lévi, 1913,p. 349-52. 12

Étienne de la Vaissière

.

Introduction

15

Cette période, si mal connue, de l’histoire des communautés iranophones en Chine forme certainement le cœur des problèmes soulevés lors du colloque. C’est au Ve et au VIe siècle que se forment ces milieux za hu sous forte influence sogdienne. Parmi les nombreuses autres ethnies venues du Nord ou présentes depuis longtemps en Chine, les Sogdiens se mettent à jouer un rôle culturel majeur. On ignore toujours fondamentale¬ ment les raisons profonde de cette domination : éclipse de la Bactriane rivale qui avait fourni les modèles de la période précédente mais qui est alors ruinée par les grandes invasions ? implantation plus ancienne et plus importante des Sogdiens, notamment en direction de la steppe, qu’on ne peut en juger d’après les sources conservées ? J’avais dans Y Histoire des marchands sogdiens, tenté, après d’autres, notamment Pulleyblank, de proposer quelques hypothèses en ce sens. Il est particulièrement heureux qu’Éric Trombert soit parvenu à leur donner un commencement de base textuelle lors du collo¬ que. Des circulations steppiques de populations, de biens, d’idées ont eu lieu alors, au IVe et au Ve siècle et la grande époque des Sogdiens en Chine, entre le Ve et le VIIIe siè¬ cle, est tout autant le résultat des implantations commerciales des premiers siècles de notre ère que de ces échanges postérieurs à la lisière des mondes nomades et sédentai¬ res, qui trouvent leur aboutissement à l’époque turque, comme François Thierry et Lin Meicun l’ont montré. Les renouvellements les plus récents de la documentation viennent justement éclai¬ rer le VIe siècle. A cet égard, la tombe de Wirkak, sabao de Wuwei, mort en 579, telle que Sun Fuxi, Yang Junkai et Yutaka Yoshida nous l’ont présentée, est remarquable. Tout d'abord, l'épitaphe chinoise, habituelle pour des défunts de ce niveau social, a été doublée d'une version en sogdien : pourquoi le défunt a-t-il ressenti le besoin de faire connaître en sogdien sa carrière à ses compatriotes, de manière si inhabituelle pour un Sogdien, si chinoise ? le défunt était ensuite placé dans une maison funéraire de pierre, intégralement sculptée, dont l'un des panneaux donne la première représentation com¬ plète du jugement zoroastrien des âmes, le passage du pont du Cinwad14, mais aussi toute une iconographie syncrétique fortement teintée de manichéisme et conduisent à antédater de près d'un siècle et demi l'introduction de cette religion en Chine15. Angela Sheng a souligné combien l’adoption par les za hu de telles maisons funéraires trouvait un parallèle au Japon. La tombe de Wirkak s’ajoute au corpus déjà conséquent des tombes de sabao sogdiens en Chine, éclairant mieux cette institution, dont ont traité ici Rong Xinjiang et Albert Dien. La traduction de sabao dans le texte chinois de l’épitaphe de Wirkak, par s’rtp’w dans le texte sogdien, vient clore une vieille question qui intriguait la sinologie depuis Pelliot16. Les populations sogdiennes, tout en participant aux milieux za hu, restent alors largement encadrées dans des communautés dont le chef porte justement le titre de sabao. Les dynasties chinoises intègrent alors telle quelle la hiérarchie interne des commutautés sogdiennes à la pyramide mandarinale et se contentent d’un système de gestion indirecte des étrangers en Chine17.

14

Grenet, Riboud, Yang Junkai, 2004. de la Vaissière, 2005. 16 Pelliot, 1903, p. 665-671. 17 de la Vaissière, Trombert, 2004, p. 944-9.

15

16

Les Sogdiens en Chine

Il n’en va pas de même lors de la période suivante, Tang, certainement l’âge d’or, et des Sogdiens en Chine, et de notre documentation : à Turfan, dont ont traité Valérie Hansen et Jonathan Skaff, les sabao sont inconnus sous les Tang. En Chine intérieure, l’influence des milieux sogdiens est vaste, dans de nombreux domaines : Qi Dongfang et Zhang Qingjie en donnent des exemples iconographiques et artistiques. Masàharu Arakawa a montré le rôle de leurs marchands. Mais ces Sogdiens sont actifs dans des rôles plus inattendus : soldats, avec mes Càkar, paysans, avec la communication d’Eric Trombert. Frantz Grenet a rappelé l’écart entre nos a priori sur les Sogdiens — marchands et caravaniers — et l’image que ceux-ci donnaient d’eux-mêmes, diplomates et grands seigneurs, utilisant pour cela et les nouvelles tombes découvertes en Chine, et les données de l’iconographie sogdienne métropolitaine18. Le contexte institutionel de la présence sogdienne en Chine est alors profondément différent, aussi bien en raison de la conquête par les Tang de territoires assez massivement peuplés de Sogdiens à l’ouest, qu’en raison de la probable dissolution des communautés sogdiennes remplacées par un système de gestion directe Les sabao comme les sous la forme de cantons ralliés à la Chine, conghua du VIIe siècle. Les le milieu passé documentation la de communautés disparaissent 19 . Les processus autres les Sogdiens sont désormais des sujets chinois comme d’assimilation ont sans doute lieu à une allure redoublée, phénomène cependant alors compensé par les liens renforcés avec l’Ouest. Un milieu sogdien remplace les communautés sogdiennes, dont l’influence est peut-être plus diluée mais pas moins grande. Les Sogdiens intègrent massivement les rangs de l’administration, notamment l’armée20, d’où la révolte d’An Lushan, étape majeure dans la vie des Sogdiens en Chine, dont on peut regretter qu’elle n’ait pas fait l’objet d’une communication spécifique, alors que l’on peut désormais prouver que les Chinois l’ont vue à bon droit comme une révolte des za hu2'. Après la révolte, la nature de nos sources change radicalement, car désormais les Sogdiens cachent leurs origines étrangères. Les Sogdiens sont peut-être toujours en Chine, mais on ne peut le savoir, et il est probable qu’ils s’assimilent rapidement. Les Sogdiens ne restent identifiables comme tels que dans deux zones, toutes deux représentées dans ce volume : Dunhuang, bien sûr, avec la communication de Zheng Binglin, et, moins exploré, le Nord-est, la région d’origine de la rébellion, avec la communication de Yutaka Moribe. Pulleyblank avait là encore fait figure de pionnier en s’attachant au devenir des milieux sogdo-turcs des Ordos au sein des élites Shatuo des Cinq Dynasties22. Paradoxalement, même à une époque aussi tardive, où les liens avec l’Occident sont coupés, les milieux sogdiens conservent une partie de leur cohérence et les critères habituels peuvent continuer à être utilisés — essentiellement celui de l’association, qu’elle soit matrimoniale, de métier ou de résidence23. 18

Marshak, 2002, p. 227-264. Ikeda On, 1965, p. 49-92. 20 de la Vaissière, Trombert, 2004, p. 957-9. 21 Pulleyblank, 1955, et de la Vaissière, 2004, p. 194-8. 22 Pulleyblank, 1952, p. 317-356. 23 de la Vaissière, Trombert, 2004, p. 965-8.

19

Introduction

Étienne de la Vaissière

17

Au final il est fi'appant de voir à quel point l’ensemble des corpus et des méthodolo¬ gies disponibles pour l’histoire de la Chine du Nord ont été représentés lors du colloque. comme les recueils d’épitaphes, passion ancestrale Les plus classiques de ces corpus Chine en d’érudits reçoivent une nouvelle légitimité en raison des — générations de questions sociales qui leur sont adressées. Les épitaphes chinoises forment un matériau de choix pour la prosopographie, malgré tous les problèmes de fiabilité qu’elles posent : que faire des biographies extravagantes, des titres ronflants dont ces Sogdiens paraient leurs ancêtres sur leurs épitaphes ? Que faire aussi des topoi qui les font se rattacher trop souvent à la première vague d’immigration iranienne en Chine, sous les Han ? On voit l’importance de pouvoir croiser l’information et d’évaluer les déformations que la de Guyuan fierté familiale fait subir à Fhistoire : l’exemple de la famille Shi est à cet égard particulièrement instructif, car dans ce cimetière familial les épitaphes successives sur plusieurs générations permettent de voir à l’œuvre le travestissement des origines24. Les découvertes de Pelliot, de Stein ou des missions allemandes, russes et japonaises s’en trouvent également revivifiées ; à la lumière des fouilles archéologi¬ ques en Chine, et des critères de datation et de fiabilité qu’elles permettent comme la présence de représentations de coqs de Sros, mi-prêtres zoroasd’élaborer un certain nombre de triens à padâm, mi-oiseaux, dans les tombes des Sogdiens documents iconographiques et textuels trouvent sens25. Les corpus les plus nouveaux, comme bien sûr les découvertes archéologiques, parmi lesquelles les tombes de sabao, mais aussi les documents exhumés à Turfan, sont également représentés ici. Mais Astana était un cimetière chinois, et on manque cruellement de documents sogdiens pour les Ve-VIIIe siècles qui permettraient de corroborer ou de réinterpréter de l’intérieur les très riches données chinoises de Turfan. Il serait heureux qu’un sabao sogdien ait eu le bon goût, tel le préteur Zuo Chongxi de Turfan26, de se faire enterrer avec un certain nombre de documents sur les affaires pendantes, afin de les résoudre dans l’au-delà. .. L’épitaphe de Wirkak et le contrat d’esclave de Turfan27 seraient moins isolés. Notons du moins que d’autres corpus chinois sont en train d’être constitués, ainsi les milliers de planchettes Han récoltées ces dernières années au Xinjiang, au Gansu ou au Ningxia28. On peut espérer grâce à elles comprendre mieux la préhistoire des implantations sogdiennes en Chine, et proposer un jour une histoire complétée de la présence sogdienne en Chine.

—

—

24Luo Feng, 1996. 25

Kageyama, 2005. Hansen, 1995, p. 33 et suiv. 27 Yoshida, Moriyasu, 1988. 28 Zhang Defang, 2004 et Wang Su, 2004. 26

—

Les Sogdiens en Chine

18

Bibliographie

Forte, Antonino

« Kuwabara's misleading thesis on Bukhara and the family name An 安 », Journal of the American Oriental Society, 116/4, 1996, p. 645-52. Grenet, Frantz, Sims-Williams, Nicholas, DE LA VaissiÈRE, Étienne 1998 « The Sogdian Ancient Letter V », Bulletin of the Asia Institute, 12, 1998, p. 91/04 Grenet, Frantz, DE LA VaissiÈRE, Étienne « The Last Days of Panjikent »3 SilkRoad Art and Archaeology, 8, p, 155-96. 2002 Grenet, Frantz, Riboud, Pénélope, Yang Junkai « Zoroastrian scenes on a newly discovered Sogdian tomb in Xi^an, Northern 2004 China, The Southern and eastem walls of tlie sarcophagus of the sabao Wirk », Studia iranica, vol. 33/2, p= 273=285. Hansen, Valerie 1995 Negociating Daily Life in Traditional China; How Ordinary People Used Contracts 600-1400, Yale University Press, 285 p. HUI Jiao 慧皎 1927 Gaoseng zhuan 高僧 傳 dans le Taishö shinshû daizökyö 大正新修大藏 經 n° 2059,vol. 50. Ikeda On 池田温 « Hasseiki chüyô ni okeru Tonkô no Sogudo jin shüraku, 8世紀中葉に おける 1965 敦煌 の ソク'、ド 人 聚落», [Le canton sogdien de Dunhuang au milieu du huitième siècle ], Yürashia bunka kenkyü, ユーラシ ア文化研究, [Studies on the Eurasian culture], n° 1, p. 49-92. Kageyama, Etsuko 2005 « Quelques remarques sur des monuments funéraires de Sogdiens en Chine », à paraître, Studia Iranica. DE LA VAISSIÈRE, Étienne 2002/4 Histoire des marchands sogdiens, Paris, Mémoires de l'IHEC vol. 32, 2e éd.

1996

2004. DE LA VAISSIÈRE, Étienne 2005 « Mani en Chine au VIe siècle », Journal Asiatique, à paraître. DE LA VAISSIÈRE, Étienne, Trombert, Éric « Des Chinois et des Hu. Migrations et intégration des Iraniens orientaux en 2004 milieu chinois durant le Haut Moyen-Age », Annales. Histoire, Sciences

Sociales, p. 931-69. LÉvi, Sylvain

« ’Tokharien B' langue de Koutcha », Journal Asiatique^ s. XI, t. 2, p. 311-80. 1913 Lin Meicun 林梅村 « Jihu shi jikao. Taiyuan xin chu Suidai Yu Hong muzhi de jige wenti » 稽胡 2002 史跡考一太原新出隋代虞弘墓志的 機個問題 [Recherches sur Fhistoire des Jihu. Quelques questions à propos de la stèle funéraire de Yu Hong découverte récemment à Taiyuan], Zhongguoshi yaryiu, vol. 1, p. 71-84. LUO Feng 羅豐 Guyuan nan jiao Sui Tang mudi 固原南郊隋唐 墓地,[Graveyard of Sui and 1996

Étienne de la Vaissière

..

Introduction

：

于9

Tang Dynasties in the South Suburbs of Guyiian], Beijing : Wenwu chubanshe, 247 p.

Boris « La thématique sogdienne dans l'art de la Chine de la seconde moitié du VIe 2002 siècle », Comptes rendus des séances de l'Académie des Inscriptions e 力 Belles-Lettres, séances de Vannée 2001, Paris, p. 227-264.

Marshak,

Musée Guimet 2004

Lit de pierre, sommeil barbare. Présentation après restauration et remontage d "une banquette funéraire ayant appartenu à un aristocrate d'Asie centrale venu s établir en Chine au vf siècle, Paris : Musée Guimet

Pelliot, Paul

« Le sa-pao »5 Bulletin de l'École Française d^Extrême-Orient, 3, p. 665-671. Edwin PULLEYBLANK, « A Sogdian colony in inner Mongolia »3 Toung Pao, 41, 1952,pe 317-356. 1952 Background of the Rebellion of An Lu-shan, Oxford : University Press, The 1955 264 p. Thierry de Crussol, François « Obole à Charon et rite fanhan. À propos des monnaies déposées dans la 2000 bouche des morts de la nécropole d5Astana 凡 dans Cs. Bâlint (éd.), Kontakte zwischen Iran, Byzanz und der Steppe im 6.-7. Jahrhundert, (Varia Archaeologica Hungarica, X), Budapest-Naples-Rome, p. 323-329. Wang Su 王素 2004 « Xuanquan Hanjian suojian Kangju shiliao kaoshi 懸泉漢簡所見康居史料 考», [On the documents of Kangju in the inscribed slips from Xuanquan ], dans Rong Xinjiang 榮 新江 et Li Xiaocong 李孝聰 (éd)Zhongwai guanxi shi: Xin shiliao xin wenに 中外 關係 史： 新史 料 與 新 問題, [History of Relationship between China and fbreign countries, new sources and new Problems], Beijing, p. 149-161. XUANZANG 玄奘 1985 Da Tang Xiyu ji xiao zhu 大唐西域記校注 [Le Da TangXiyu ji^ texte critique et notes], Beijing, Zhonghua shuju. Yoshida, Yutaka 吉田豊,Moriyasu, Takao 森安孝夫,et al. « Kikushi Köshökoku jidai Sogudo bun onnna dorei monjo, 麴 氏高昌国時代 1988 ソ グ ド 文女奴隸売買文書», [A Sogdian sale contract of a female slave from the period of the Gaochang kingdom under the Qu clan], Nairiku Ajia gengo no kenkyû, 内陸 アシ 了百 語の研究, [Studies on the Inner Asian Languages] s vol. 4, p. 1-50. Yoshida Yutaka 吉田豊,Kageyama Etsuko 影山悦子 2002 « Sogudo jin, tenseki wo oginau saishin no shutsudo shiryö kara, ソク、、ド 人、 典籍を 補う最新 の出土資料 から », [Les Sogdiens, d'après de nouvelles données archéologiques qui complètent les sources historiques chinoises], Shinika, しに か,[Sz'Mca], 2002-9, p. 44-9. Zhang Defang 張德芳 2004 « Xuanquan Hanjian zhong ruogan Xiyu ziliao kaolun 懸泉漢簡中若干西域 資料考 論», [On some documents of Central Asia in the inscribed slips from Xuanquan], dans Rong Xinjiang 榮 新江 et Li Xiaocong 李 孝 聰 (6d) Zhongwai guanxi shi: Xin shiliao xin wenti, 中外關係史： 新史料 與新 問題, [History of Relationship between China and fbreign countries, New sources and new problems], Beijing, p. 147-129. 1903

1111

Carvings on the Stone Outer Coffin of Lord Shi of the Northern Zhou '

Yang Junkai

From June 12 through October 2& 2003, the Xi?an Municipal Bureau of Cultural Relies and Archaeology excavated the Northern Zhou era tomb of Lord Shi5 located in Xi’an’s Weiyang District or 5.7 km west of east of Jingshang Village from that km Chang’an. 2.2 the The site is of Northern Zhou An Qie the Han site of tomb (see fig. 1, p. 23). Several precious artifacts were unearthed, including a stone gate, stone inner- and outer-coffïns, a gold ring5 and a gold coin and omament. Stone reliefs were decorated with color paintings and gilding, and are rieh with thematic content ranging from Han culture to Zoroastrianism (see colour plates 1 to 7). According to epitaph on the outer coffin, the interred was of the sumame Shi £ and was the during the Northern Zhou. The discovery was another of Liangzhou sabao great contribution to the understanding of the interchange between Chinese and Western cultures, along with the excavations of the Yu Hong ^14 and An Qie tombs.1

The tomb

The tomb is in a deep, slanting pit, situated in the north and facing south oriented at 186 degrees. The tomb path, ceiling well, passageway compartments, entrance path and tomb chamber total 47.26 meters in length (see fig. 2 p. 23). The tomb path is 16.3 meters long, 1.70 meters wide at the south end and 1.48 at the north, and runs at a 15-degree angle. There are five passageway compartments and five ceiling wells. The entrance path has an arched covering, and is 2.8 meters long? 1.5 meters wide, and 1.9 meters high. Between the tomb path and the tomb chamber are two sealed doors, one of brick and one of stone. The stone door is 1.65 meters high and 1.66 meters wide. On the crossbeam above the door and the pillars on its sides are reliefs of grapevines, anthemion, apsaras, and protector gods. On the stone door are color paintings and gilding. Most of the paintings hâve wom away, but some pictures of apsaras and lotuses remain [see also supplementary remarks]. The tomb chamber runs 3.7 meters from east to west, and 3.5 meters from north to south. Due to tomb robbery and collapse, the characteristics of the chamber’s roof are unclear. The remains of the four walls only stand 0.5 meters high. In the middle of the i

Yang Junkai, 2003 a, and 2003 b.

22

_—

Les Sogdiens en Chine

chamber, to the north, a stone outer coffin has been excavated. The outer coffin is a “mountaintop” rest style structure situated in the north and facing south. It is five spaces across and three spaces deep. From east to west it is 246 cm, 155 cm north to south, and 158 cm high (see fig. 3 p. 24). The outer coffin is comprised of a foundation, four walls and a ceiling. The foundation is made of two pièces of stone, 250 cm east to west. The Southern piece is 88 cm north to south, and the northem piece 68 cm. The foundation extends out beyond the walls of the outer coffin on ail four sides, and is covered with relief patterns. The four walls of the outer coffin are constmcted of twelve pièces of stone (including the two pièces which comprise the door, the one piece which is the horizontal beam, and the one piece that is the threshold). The pièces at the four comers are ail whole pièces at right angles. Over the cracks between the pièces of stone is fastened an iron “slender waist” tenon and mortise arrangement, and each piece fits into a groove in the floor. On the fbur walls are reliefs with thèmes of four-armed protector gods, Zoroastrian gods, hunts, banquets, travel scenes, caravans, ceremonies, and ascension to the heavens. The images of people, along with their clothing and belongings, and the landscapes and structures around ther^, ail employ color painting and gilding and exhibit rather Western characteristics in their content and style. The roof of the inner coffin is comprised of five pièces of stone, fbur of which are above the outer coffin, and jut out on ail four sides in the form of eaves. These are covered with cinnabar paintings to form artificial wood. The roof itself is made of one carved piece of stone, the inside of which has obvious chisel marks. On the upper part of the outer coffin's Southern wall, epitaphs hâve been found in Chinese and Sogdian scripts. There are 33 columns of Sogdian script, and 18 of Chinese. Professor Yutaka Yoshida has translated the Sogdian script, which is generally the same as the Chinese but better preserved.2 From the Chinese we leam that the interred was “a man of the nation of Shi, originally from Western Lands, moved to Chang'an (...), and accepted the position of sabao of Liangzhou.” Fürther, “he passed away at home during the First Year of the Daxiang era (579) at the age of 86. His wife, sumamed Kang was buried alongside of him the twentieth day of the first month of the next year.” This is the earliest extant example of a side-by-side Sogdian-Chinese text, and is of great value as a resource in the study of the Sogdian language. A stone couch was excavated from inside the outer coffin. It measures 200 cm long, 93 cm wide, and 21 cm tall. Inside the outer coffin are also cinnabar powder wall paintings. Only leaf and grapevine patterns remain. On the roof of the inner coffin, cinnabar was employed to paint a building. A gold ring, a gold coin, and a gold omament were also discovered, as well as the remnants of a ceramic lamp. Because the tomb has been subjected to thorough plundering, skeletal remains are scattered inside and outside of the outer coffin. It has been confirmed that the skeletal remains are of one male and one female human, as well as animais. No signs of crémation hâve been found on the human remains.

2

See the papers by Sun Fuxi and Yoshida Yutaka in this volume.

Yang Junkai

Carvings of the Stone Outer Coffin of Lord Shi

23

Fig. 1: Map of the ancient limits of Chang'an with indication of recently discovered Sogdian tombs (An Oie, Kang Ye and Shi Jun).

Fig. 2: Plan of the tomb.

Les Sogdiens en Chine

24

Fig. 3: Inner and ou ter coffins.

Fig. 4: Southern side of the outer coffin (see also plate 1).

Yang Junkai

Carvings of the Stone Outer Coffin of Lord Shi

25

Contents ofthe reliefs

The information from this discovery is currently being analyzed. Here we will examine the reliefs from the four walls of the outer coffin. We can take the first step in analyzing the contents of these reliefs by Consulting literary documents and Zoroastrian classics, fairytales and legends, and taking into considération similar excavated images which are scattered throughout the world. The Southern WM of the Outer Coffin

Assembled from eight pièces of stone, 220 cm wide, 110 cm tall, and 8.5 cm thick, the southem wall is the front of the outer coffin with five open spaces (see fig. 4 p. 24, and plate 1 fig. 1). The central space is 68 cm wide, and the two to the east measure 33 cm and 41.7 cm in width. The two spaces to the west are 35.6 cm and 37.7 cm in width. The central space contains a stone double-door, a horizontal beam, and a threshold. The outer coffin door is situated in the center of the central space, measuring 68x90 cm, and is closed. It is a double-panel door. The door handle was made of iron, but has rusted away and now only two holes are extant. Each door panel is 43x98 cm. Each panel has four hinges, and each hinge has 5 nails for a total of 20 nails per door. There is a doorframe on either side of the door measuring three cm wide. Above the door is a long piece of stone, 88x23 cm. It is longer than the central space where it is situated, as to fit into grooves in the pillars at either side and act as a horizontal beam. There is an epitaph carved into this beam, in both Sogdian and Chinese scripts. Only part of this epitaph has survived plundering (see fig. 12 p. 45. The threshold is 86 cm long, 20 cm wide, and 12 cm high. It fits into the bottom parts of the adjacent pillars. It has six Steps. On the sides are carved a pair of lions and four children, a theme only seen a few times previously (see fig. 13 p. 45). In the center are four-armed protector gods stepping on démons, carved using a high relief style (plate 1 fig. 3). To the sides of the protector gods are mullion Windows in the center of the adjacent spaces. Each mullion window has eight 1.1 cm bars, and is 25 cm wide. The eastem window's frame is 29.5 cm high, while the western one's is 28.3 cm high. The bars are triangular, similar those carved in high Tang Buddhist tombs at Dunhuang.3 Above each window are four musicians, with two attendants to each side. Under each window are two winged hal&man half-bird priest figures wearing masks and carrying torches next to fire altars (plate 1 fig. 2). Western Wall of

Outer Coffin

This wall is assembled from two pièces of stone measuring 120 cm wide, 1.10 cm high, and 8.5 cm thick with a gap of 1 centimeter between the two pièces. The northem side is 61 cm wide, and the southem 58 cm. The four pendant posts are separated into three sections by relief carvings. We refer to them by number as, from south to north, Wl, W2, and W3 (see fig. 5, and colour plates 2 and 3). W1 is 83 cm tall, and 34.5 cm wide. Its picture depicts a god preaching the dharma 3

XiaNai, 1955, p. 28, plate 11, pp. 4-6.

26

Les Sogdiens en Chine

to ail sentient beings surrounding it. At the top center there is a god standing on a lotus throne with a halo behind. His hair is in a topknot, and he has a beard. His right arm is bent and raised up. The pinkie of his right hand is curled upward. The left arm is slightly bent, with the left hand over his ehest. His right shoulder is exposed, and his robe thrown over his left shoulder. He appears to be lecturing on the sutras and the dharma as well. Undemeath the god and to the right there is a braided rug laid out, on top of which there is kneeling a male and female couple. They are holding their hands up and out. To the left and right of the god, and undemeath him, there are three groups of three men, kneeling or sitting. On the bottom right of the image, there is another braided rug with five men kneeling on it and piously holding their hands out in front of their chests. Around these people are seven animais. From top to bottom, tliey are: a pair of lions, a male deer, an antelope, a sheep, and a wild boar. At the very bottom of the picture are a lotus leaf and a wave. Two water créatures are extending their necks upward from the wave. This picture is apparently Buddhist in its content, with its depiction of the Buddha preaching the Dharma for ail sentient beings. W2 is 83.5 cm tall, and 32 cm wide. Its imagery is centered around a wood and brick structure with a vaulted roof. A husband and wife at home hold a baby. The husband sits to the center left, wearing a jewelled hat and long tunic with a round neck and narrow sleeves. He has a sash round his waist, and sits cross-legged. His right arm is holding the baby, a bareheaded boy with his left arm bent and raised. The boy's right hand is placed on his stomach and wearing a similar tunic. He sits on his father's lap. The wife sits to the center right, wearing a jewelled hat and wrapped in a fur cloak. Two servants stand to the right of the couple. The one to the right wears a similar tunic and holds a bottle at his ehest. Some items hang from his sash. On the stairs are set two bottles, and at the bottom of the stairs lies a dog. At the bottom right of the image is a boulder and a wave, and to the left is a saddled horse. In front of the horse is a kneeling man, and next to the horse there stands a servant holding up a canopy. He wears a long tunic with a knife at the waist, and appears to be waiting on the husband of the image. At the bottom right corner of the image, there is a riderless horse under a canopy, with bearers of tribute in front of it. This composition is comparable to that said by the Miho Museum to represent the water god. This is perhaps the water god of the Sogdian

people.4 W3 is 83 cm tall, and 38.7 cm wide. It depicts Caravans traveling and hunting in rocky mountains and forests. Just left of the center there is an image of a man on a horse, shooting an arrow. He is wearing a Sogdian-style turban. His right hand holds his bow, with his left preparing to shoot an arrow. He is wearing a cross-folded tunic with narrow sleeves, and on his sash hangs a quiver. In front of his horse are five animais, one of which has already been struck by an arrow. The other four are running away. They are, in order, a male deer, an antelope, a wild boar, and a rabbit. Behind and to the left of the man's mount is another horse, extending its head out from behind a boulder. Behind and to the right there is a servant holding a falcon. Between the boulders and trees there are also four dogs, one of which is to the left of the master's horse, its head

4

See Rong Xinjiang, 2003a, p. 207-210,

Fig. 4: Eastem (left) and Western (right) sides of the outer coffin (see also plates 2, 3, 6, 7).

LShiord

of

Cof in

Stone Outer

the

of

Carvings

Junkai

Yang

Fig. 5: Northern side of the outer coffin (see also plates 4 and 5).

2

Chine

en

Sogdiens

Les

8

j

Yang Junkai

Carvings of the Stone Outer Coffin of Lord Shi

29

jUst visible from behind the boulder and chasing the rabbit in front of it. Two dogs are to the sides of the escaping animais on the right, crouching down as if encircling prey for their master. The fourth dog is at the bottom of the scene, to the front right of the caravan's horses, apparently searching for prey between the boulders. The bottom portion of the scene is a caravan made up of horses, camels, and donkeys. At the front of the caravan are two men riding horses, one of whom has a quiver of arrows at his waist. Behind these two men's horses are two camels carrying goods. Behind the camels is a horse ridden by a man with a "ship"-shaped hat. His right arm is bent and extended upwards, with an object in his right hand. He seems to be looking toward some distant place. To the right and above the two camels is a group of two horses and four donkeys carrying goods. The donkeys are between the horses. Behind the group is a man with a whip in his right hand urging them onward. Under the camels' feet is a wave-like pattern.

The Northern Wall of

Outer Coffin

This wall is constructed of four pièces of stone. It is 221 cm wide, 110 cm tall, and 8.5 cm thick. The widths of the four pièces are, from west to east, 28.8, 91, 74, and 32 cm. From the structure of the building, it is apparent that there should be six pendant posts, with five spaces between them. However, due to the limitations of the size of the stone materials, and the requirements of the subject matter of the images, the northem wall is less two posts. It only has four posts with relief carvings. However, the images are divided into five thèmes. We refer to them, from west to east, as N1-N5 (see fig. 6, and colour plates 4 and 5). NI is 82.5 cm tall, and 25 cm wide. The images depict a caravan encamping and doing business. It can be divided into upper and lower portions. The upper portion is centered around a tent. Its door is rolled up, and two birds perch on top. A male chief sits cross-legged inside. He wears a jewelled hat, and a long tunic with narrow sleeves and a tumdown collar. A sash is around his waist, and he wears high boots. In his right hand he holds a goblet. His left hand rests on his legs. Outside the tent is a lush forest. Two wild geese fly overhead. A braided rug is laid out in front and to the right of the tent, and a bearded elder wearing a feit hat kneels on it. He is dressed similarly, and a blade hangs from his sash. His right hand holds a goblet, and his left hand is slightly bent forward. Two people sit in front of him, drinking alcohol. On the sides of the tent are three servants, two to the left and one to the right. Between the door of the tent and the rug is a dog looking backward. Under the tent is a caravan led by four men and consisting of two camels, two horses, and one donkey. Two men are carrying on a discussion in the middle of the caravan. One has goods slung over his shoulder. The camels are carrying goods and are lying on the ground resting. This image corresponds exactly to the An Qie image of the sabao visiting the Turks. The top portion is the sabao and the Turk leader drinking in front of the tent, while the bottom portion is the caravan brought by the sabao.5 N2 is 82.7 cm tall, and 58.2 cm wide. It depicts a male and female drinking at 5

See Rong Xinjiang, 2001, p. 128.

30

_

Les Sogdiens en Chine

home. The entire picture is centered around a wood and brick structure with a vaulted piazza. The couple sits in the center, with the man at the left. The man wears a jewelled hat. His right arm is bent and raised upward, and his right hand holds a goblet. His left shoulder is leaning on a cushion (yinnang He wears a tunic with a round neck and narrow sleeves, and a blade hangs from his sash. The woman is to the right：She wears a hat and a long tunic. She holds in her right hand a small cup, and her left hand is placed in front of her stomach. Around this couple are three musicians. The one to the right plays a lyri (konghou and the other two play pipa Between the pipa players is a servant holding his arms up and out as if performing. In front of the master couple is a servant raising a cup in his left hand. Before him is a long-handled jar. In the piazza are four servants, three in front and one in back, holding out items. At the bottom and to the right of the stairs are three people. The one to the right is a musician playing a waist-drum. The one in the middle is standing, and the one to the left is dancing with long sleeves, her left leg up. This image is basically similar to ones fbund at the An Qie and Yu Hong sites of drinking at home. The house and bridge are similar to images from An Qie, and the musicians and dancers are similar to those from Yu Hong. N3 is 83.5 cm tall, and 52.2 cm wide. The image is of a male and a female master traveling on horseback, and is divisible into upper and lower portions. The upper portion depicts five men, one of whom is the master. Above their heads is a canopy held up by a male servant next to the horses. The master is wearing a Sogdian-style turban and a long tunic with a sash. He is on a horse, preparing to go forth. He is tuming around, right arm held high, to wave goodbye to two men, also on horseback, who are sending him off. The man in front is leaning forward, hands clasped together in front of him to bid farewell. The man in back is raising his right hand as a salute. In front of the master's horse is a man with a long sword hanging from his waist, who has already ridden ahead. The bottom portion concems the women leaving on their joumey. In front of the women is a man on horseback with a quiver at his waist. The women is comprised of three women on horseback wearing fur hats. In the middle is the female master under a canopy, wearing a fur cloak. A female servant to her horse's right holds the canopy for her. A dog follows to the left of her horse. N4 is 83.5 cm tall, and 41.5 cm wide. In the image, a male and a female master are depicted drinking in a vineyard. Amid the ripe grapes are two large braided rugs. Five men sit on the rug in the very center. Three large plates filled with fbod are set between the rugs. Each of the five men is leaning on a cushion (yinnang), and four of them are raising their cups. From left to right, they are holding a goblet, a cup, a rython cup, and a goblet. To the left, a man is reaching out for food with his right hand. Beside the rug, four male musicians are performing. To the right, two are playing a lyri and a pipa, and to the left one plays a flute and one plays a waist drum. Beneath the carpet, two servants are kneeling with their arms outstretched, serving alcohol from a pitcher. There are also five women on the rug at the bottom portion of the picture, and in its center are two large plates of food. The women are wearing long tunics with wide sleeves and tumdown collars. This is the only place in the outer cofïin's pictures where we see women dressed in Chinese style. The three on the right are holding cups and drinking, and the two on the left are reaching out to take food. Two female musicians are standing to the side of this rug as well. The one in the right is playing a flute, while the

Yang Junkai

Carvings_.of the Stone Outer Coffin of Lord Shi

31

one to the left is playing a pipa. Above them is a person playing a lyri. A female servant is attending to the women on the rug, holding a goblet out in her right hand. A pitcher sits in front of her. In front of the rug and to the left is a wave pattern. Among the waves are two vases containing flowers. This picture is comparable to the vineyard site, and therefore may be a scene depicting the Sogdian picture at the Anyang

|

• j

I | j

New Year festival, Nowruz.6 N5 is 83.7 cm tall, and 28.3 cm wide. The top portion is a scene of mountains and lush forest. At the top of one of the mountains is a cave with an old man who looks like a Buddhist ascetic sitting cross-legged and leaning forward inside. His right arm is stretched forward with thumb pointed upward and index and middle finger extended forward. He is wearing a loincloth, with ehest and legs exposed, and there is a bottle in front of him. Outside of the cave is a dog laying his head in front of the old man. The bottom portion of the image depicts two men in water. There are also lotus leaves, and two water créatures extending their heads out from the water. The two swimmers appear to be looking at the créatures in fear. They are holding their hands up to two flying apsaras coming to save them. Behind these two apsaras is flying a third, wearing a pearl necklace. In its right hand is a plate with several items, and in the left is a vase. It is flying toward the other two.

Eastern Wall of the Outer Coffin This wall is constructed of two pièces of stone. It is 120 cm wide, HO cm tall, and 8.5 cm thick. The width of the northem piece is 57.5 cm, while the Southern piece is 62.5 cm. Four pendant posts feature relief carvings separated into three portions, which we will refer to here, running north to south, as E1-E3 (see fig. 5, and colour plates 6 and 7). El is 84 cm tall, and 33.5 cm wide. In the center, it is divided into two portions, upper and lower, featuring mountains and clouds. The upper portion features a god with a stem expression on his wide face seated in the center of a circle. He wears a jewelled hat, and his right hand holds a trident up to the side of his face. His left hand is at his waist. He has bracelets on both wrists. He sits cross-legged, the right on top of the left. He wears a shirt with a round neck and narrow sleeves, shoulders exposed, and pants with a narrow waist. He has three mounts nearby, ail bulls. One is looking forward, flanked by the other two in profile. Above the circle is an arch, with a flying apsara at either side. Under the main god and to his right are three other gods wearing jewelled hats. The one to the right holds a cup to his ehest in his right hand. His left hand is extended upward. His waist is hidden by the mountains. The one in the middle holds his right hand to his ehest, and is picking a flower with his left. His body is also partially concealed. The one to the right wears a tunic with a round neck and narrow sleeves, tied at the waist. Wings sprout from his shoulders. His right hand is at his waist, and his left arm is bent. He is barefoot, with his legs crossed. He wears an anklet on his left ankle. Under the main god there is a male and female couple kneeling on a rug. The male is above, wearing a hat and a long tunic with a round neck and narrow sleeves, tied at the 6

Ibid., p. 135-144.

32

_

Les Sogdiens en Chine

waist. His right hand holds a long object, and his left is slightly raised. The female is combing her hair, and is wearing a long tunic with a tumdown collar and wide sleeves. The bottom portion of the g arment, below the ehest, is pleated. The bottom portion of the picture is a craggy mountain range. The tops of the mountains are dotted with turquoise green trees. On the side of a mountain are two dogs with bells tied to their necks. At the bottom of the mountains is an arched bridge with a railing and two columns. Two priests stand to the left of the bridge, wearing masks over their mouths, long tunics with round necks and narrow sleeves, high boots, and sashes. Some items hang from the left of their sashes. They are holding torches with both hands. Above the bridge and colunins are two lires. A group of animais is on the bridge, heading right. They are, from left to right, two sheep, a lamb, and two camels. There is a large load of goods on the last camel's back, and a pair of birds on top of the load. The railing of the bridge is decorated with five- or six-petaled flowers, and undemeath those are two rows of triangulär patterns. The bridge is supported underneath by four pillars, each with birds or animais in relief images. A swirling water pattern is undemeath the bridge. E2 is 84 cm tall, and 31.5 cm wide. Auspicious clouds and grasses divide the image into upper and lower portions, At the center, two winged horses run toward the left. The head of the bottom one is decorated with a sun and moon. Above them and to their left is a flying apsara wearing a jewelled hat and holding something in its left hand. Its right hand is slightly raised. It wears a long tunic with a round neck and narrow sleeves and a sash at the waist. It is barefoot. Its clothing is flying up in the air. Under it and to the right is an image of a lotus. To the left of the lower horse is the back of a person with hair pulled back, and ail four limbs in the air and clothing blowing upward as if falling from the sky. Something is in each hand. The bottom portion depicts an arched bridge connected to the one seen in El, similar in appearance and décoration of its railing and pillars. A procession crosses the bridge, heading right, in a similar männer to that in El. Three adults and a child are in the middle of the procession. The first person is only partially visible. In the center are a man and a woman. The man has a high nose and deep eyes, and holds his hands at his stomach. The woman follows close behind. The child is farthest to the left. Behind the four are two horses, and behind them are five donkeys and oxen. The oxen are looking back. In the water below the bridge are lotus flowers and leaves, and two créatures looking up with open mouths. E3 is 84 cm tall, and 38 cm wide. The upper portion, which appears to be a depiction of Heaven, takes up most of it. It is centered around a man and a woman riding winged horses. The man is in the front, wearing a jewelled hat. The hat's Streamers are blowing upward. His right index finger and pinkie are extended, and his left hand is open. The woman behind him holds her right index and middle Angers up, and her left hand is open. Her clothes trail behind in the wind. The two horses hâve sun and moon décorations on their heads, and are flying toward the right. Two apsaras wearing jewelled hats lead them in the upper right corner. The apsara on top has wings sprouting from its shoulders. It is holding a loop-shaped object in its right hand. Its left hand is raised up, palm upward. The apsara undemeath is raising its right pinkie, and holding out its left hand palm up. Its clothing is blowing upward. Around the two riders, there are also four winged musician apsaras with differing topknot hairstyles. Clockwise, from top left, they are playing: a pipa, a lyri, a flûte, and a panflute Çhengdi

Yang Junkai

Carvings of the Stone Outer Cojfin ofLord Shi

33

The one playing the panflute has a sun and moon décoration on its head. Under the musician apsaras are four running animais: from right to left, a lion, an ox, a camel, and a sheep. Their hindquarters hâve tumed into curlicues. Under the animais is a riverside cliff. The cliff is dotted with some végétation. The water is connected to that in E2. There are three créatures in the water. Looking at the contents of the reliefs of the outer coffin's eastem wall, we can see that the images are, at the same time, independent and closely connected to each other. El to E3 practically comprise one complété image, as the outer coffin is modeled on the wooden temple hall form and is split by four pillais into three sections. Although they are separated, together they show the whole process of the Sogdian ascent into Heaven. By looking at the direction in which the people and animais in the images are heading, it is easy for us to see that these images connect to form a storyline from El to E3. This is similar in fashion to the method of depiction in Contemporary scrolls, and leads us to suspect that the image existed in scroll form. El shows a god with a halo holding a trident in right hand, with three oxen steeds, which appears to be a Zoroastrian god. The people kneeling on the rug below are the couple interred in the tomb. At the lower right are three gods wearing jewelled hats. The apsaras and two winged horses going from right to left in E2 are moving the opposite direction from moving éléments in the rest of the image, as if going to receive the interred couple. At the bottom of El and E2 is an arched bridge with two priests at the side and two fires. There are two dogs on the mountainside. On the bridge is a large caravan of people and camels, etc. Under the bridge are two water créatures. In E3, the interred couple are riding winged horses into Heaven under the guidance of the two apsaras and the protection of the four musician apsaras.

Interpretation

This outer coffin is critical to the understanding of relief tombs of Sogdian people in China. We can see from this unearthed tomb that although the burial practices of Sogdian people in China maintained éléments of Zoroastrian beliefs but absorbed other éléments from burial practices of Han Chinese as they moved eastward. According to archaeological reports, Sogdian-style ossuaries hâve only been found in the Turfan Basin, in Xinjiang Province. As the Sogdians moved further east into China, they gradually adopted Chinese burial practices. From the Sui and Tang tomb unearthed south of Guyuan [SIJM,7 the Tianshui tomb,8 the Sui tomb of Yu Hong in Taiyuan Shanxi,9 the Shimaping Xi'an,10 and the Northern Zhou Northern Zhou tomb of An Qie at Kengdizhai tomb of Lord Shi at Jingshang Village, we can see that the burial practices of the Sogdian people in China did not differ from those of their Han contemporaries. But the 7

Luo Feng, 1996. Tianshui shi howuguan, 1992, p. 46-54. Shanxi sheng kaogu yanjiusuo, 2001, p. 27-52. 10 Shaanxi sheng kaogu yanjiusuo, 2000, p. 28-35; 2001, p. 4-26.

8

34

_

Les Sogdiens en Chine

tombs of Yu Hong, An Qie, and Lord Shi, etc., ail contain Sogdian-style stone foneral items with relief carvings, paintings, and gilding. Through these archaeological discoveries, we can see the transition from ossuaries to stone slabs and outer coffins. This is especially so in the An Qie tomb, where the bones are placed in a Buddhist-style gateway instead of on a stone slab. The body was also bumed inside the crypt before it was closed, and the four walls and the body show marks from fire and smoke. However, no signs of dog bites were found on the bones, and only a stone slab and an epitaph box were found buried with the body in the tomb. This seems to be related to Sogdian burial practices.11 These phenomena are not traditional Chinese methods, nor are they Sogdian. It seems to be the resuit of Sogdian people in China mixing Central Plain burial practices with Han-style stone coffins and tlieir own relief-carved ossuaries. The stone outer coffin of Lord Shi's tomb is not the only example of this type of temple-style funeral objects from Sogdian people in China. The ninth stone of the Fu Shandong Province depicts a scene in which a family tomb at Qingzhou 12 temple-style outer coffin is used in a fimeral (see fig. 7). In the image, four hors es are carrying a building forth, fbllowed by a dog. This building must be a funeral object, and the dog related to the Sogdian “Sagdid” (“dog's gaze”) rite.13 Another classic example is visible in a stone funerary couch now located in Japan's Miho Museum. On the third panel of stone on the back of the couch, there is a precious 14 image of a Sogdian funeral (see fig. 8). Judith Lemer's study on the Miho piece 15 carefully examines the contents of both the top and bottom portions of the image. The top is centered around a priest wearing a long tunic and, over his mouth, a white mask (the Zoroastrian padäm). Before him is an altar of fire. The priest stands before the altar to protect the sacred flame, which must be a Sogdian performance of the Zoroastrian outdoor dévotion ceremony (“Afrinagan”). To one side of the altar is a tray of bread or fruit, and to the other a bottle filled with liquid used in the ceremony. Behind the priest are four people, two kneeling and two standing, cutting their faces with knives. To the left of this group are seven people. In the front of this group are two women holding small bundles. The fïve men in the back are standing straight with arms joined, looking forward in sorrow. In front of the altar and the women is a pen containing three camels, of which only the hindquarters are visible. The bottom portion has two women and three men. Two men are in front, and one man and the two women are in back. Behind them ail are three horses. The group is ail walking toward some trees. Between the upper and lower portions is a small dog standing next to the priest and facing the altar. In Zoroastrianism, a dog gazes at a corpse in Sagdid. These images and the ossuaries p. 931 says “Burial in a coffin and chapter 54 of the Nanqishu The “Gu Huan Zhuan outer coffin is the way of the Chinese, and crémation and water burial are the customs of the Westemers.” use of the nation of Kang as an Xifan ji's quotes Wei Jie's Chapter 193 of the Tongdian in fanerais. There is also one specializing households are 200 city of the example: “Outside the walls building in which dogs are raised. Each time someone dies, the body is taken there, and the dogs eat ail of the flesh. The bones are collected and buried without a coffin.” 12 Taken from Zheng Yan, 2002, p. 224, illustration 173. 13 Zheng Yan, 2001, p. 73-109. 14 Taken from Zheng Y an, 2002, p. 261, illustration 188. 15 Lemer, 1995, p. 179-190. 11

Yang Junkai

Carvings- of the Stone Outer Coffin ofLord Shi

Fig. 7: Outer coffin in shape of a temple from the tomb of the Fu family (Shandong).

Fig. 8: Sogdian funeral from Miho’s funerary couch.

Fig. 9: Ossuary discovered in

Kashka Darya.

Fig. 10: Scene of facial mutilation from Panjikent.

35

36

Les Sogdiens en Chine

16 found in the Sogdian area of Kashka darya corroborate with each other (see fig. 9). Rong Xinjiang goes a step further, saying that it is possible that the bottom is a continuation of the top. After undergoing the “dog's gaze,” the dead is taken to the forest by the horse cart, with the men and women sending him off.17 Zheng Yan also agréés that it is an image full of classic Zoroastrian funeral iconography: the image is taking place outdoors, and has a masked priest tending to an altar of fire with four men performing facial mutilation with knives behind him, and others standing in mouming.18 As for Turk funerary customs, Cai Hongsheng believes that they can largely be separated into fiinerals and burials. Funerals featuring facial mutilation were common amongst northem and eastern minority tribes, a major characteristic of ancient Inner Asian funeral culture.19 Scenes of facial mutilation hâve also been uncovered among funeral images in the Sogdian ruins at Panjikent (see fig. 10).20 In combining the analyses of the images from Lord Shi's stone outer coffin's eastem wall and the funeral images from the Miho Museum pièces, we can see that although their contents are not the same, they certainly hâve some inherent connection. That is, they reflect Sogdian fiineral practices as well as their beliefs conceming departed's joumey to Heaven. Due to the outer coffin's limitations in size, the eastem wall of the tomb of Lord Shi, there are only two priests wearing long tunics and fiâmes symbolizing an altar at the edge of the bridge. In contrast, the Miho pièces' funeral image depicts the Zoroastrian outdoor dévotion ceremony in front of sacred fires which is held after believers passes away. In interpreting their contents together, we see that they complément each other, and from them we can piece together a more complété notion of the funeral customs of the time. Due to the limitations of the unearthed artifacts, the animal pen seen in the Miho pièces' funeral scene has not previously received from scholars the attention needed to be fully interpreted. This has affected the understanding of the entire image. Now, in light of the images from El and E2 in Lord Shi's tomb, we believe that the “animal pen” of the Miho image is actually a bridge's railing as seen in the relief images from Lord Shi's tomb's eastem wall.21 In this image too, we see a masked priest wearing a long tunic standing in front of an altar of fire and tending to the sacred fiâmes. There is also a small dog next to the priest facing the fiâmes. On the bridge are three camels with only hindquarters visible. When these two relief-carved funeral objects are viewed in light of one another, we can see that the priest on the bridge tending to the fire in the Miho image is actually performing the outdoor dévotion ceremony, and the wives and moumers in the back are performing self-mutilation rites in larnent. The dog nearby is for the “dog's gaze” rite.

16

Taken from Rong Xinjiang, 2001, p. 157, illustration 31. Ibid., p. 154-155. 18 Zheng Yan, 2002, p. 261. 19 Cai Hongsheng, 2001, p. 24-25. 20 From Rong Xinjiang, 2001, p. 156, illustration 30. 21 On October 22, 2003, at the author's dig site, Zheng Yan pointed out that the Miho piece's “pen” must also be a bridge, in connection with the Lord Shi tomb eastem wall image. 11

Yang

Junkai

Carvings- of the Stone Outer Coffin ofLord Shi

37

If we say that this image is realistic in its depiction of events, then we can view the images from Lord Shi's outer coffïn as its continuation. The priest and the ceremony take up only a small part of the image, and the moumers performing self-mutilation hâve been left out entirely. The fbcus of the image has tumed to the expression of the Reliefs conceming fanerais and burial, or, in other words, the hope that the interred will enter Heaven is the point of the image. The stone held by the Miho depicts a railing only symbolically, whereas the image from Lord Shi's outer coffin shows the entire bridge, It is possible that the people and animais on the bridge symbolize ones from the interreds' lives, heading with them to the afterworld. In the images from Lord Shi's tomb there are two priests standing on the bridgess edge, showing that the foneral ceremony is held and completed by Zoroastrian priests» Two dogs' heads appear from behind the rocks above them3 representing the "dog七 gaze55 rite in which it is determined whether the departed were good or evil. The interred couple brings a caravan of animais with them without incident. This bridge is the bridge connecting this world and the next world3 or the "Ginwad" bridge. The image in El is a depiction of the couple undergoing divine judgment and ascending into Heaven.22 The lord of Heaven is the god with the trident with the three bulls, or the Sogdian god Weshparkar. Weshparkar is équivalent to the Hindu god Shiva, and the Buddhist Maheshvara. It is recorded in the third chapter of the “New Record of Two Capitals,5 (Liangjing xinji 京新 记) that "the god of Heaven of the Western tribes [in the Zoroastrian temple in Chang'an's Buzheng District 布政坊] is the Maheshvara mentioned in Buddhist sutras.95 This god was accorded the position of the main god in Chang'an's Zoroastrian temple during the early Tang, and we can use this image for

reference.23 In the E2 image, we see in front of the two winged horses is falling a man with a topknot with his back to us, limbs upward, and holding something in his hands. Under the bridge wait two open-mouthed water créatures. This is in clear contrast with E3, in which the departed couple rides winged horses under the guidance of two apsaras, and are surrounded by an accompaniment of four musician apsaras and four heavenly créatures. The reliefs on the eastem wall of the Northern Zhou tomb of Lord Shi use imagery to depict the Sogdian belief of the soul's journey from the body to the edge of the Cinwad bridge where the Zoroastrian final judgment is carried out. If the soûl was 22

Wei Qingzheng, 1999, p. 448. Cinwad is a bridge in ancient Iranian myths which connects this world with the afterworld. According to Zoroastrianism, Zarathustra is the lord of destiny on this bridge. He meets the soûls (“anosharawan”) of the righteous on this bridge. According to later myths, the Cinwad is the "Fate-deciding Bridge” Mithra, Rashnu, and Sraosha judge the fate of the soûls of the dead. According to the Avesta, after death the soûl départs the body fbr the Cinwad Bridge to receive judgment, and is either allowed to ascend to Heaven or thrown into Hell. It is said that to the sinn er, the mouth of the bridge is as narrow as a knife's edge, but to the good it is as wide as “a fïeld of twenty-seven arrows." (Sad Dar 46, 1-2) Bridges are symbolic of the link between life and death in Han traditional culture as well. Professor Wu Hong has pointed out that the "Shang Wei 上卫 (ラ胃) Bridge" image from the tomb found in Cangshan 苍 山 (Shandong Province) dating to 151 CE is related to fiinerary practices. Xin Lixiang has also pointed out that bridges in Han dynasty stone images are a link between this world and the world of soûls, similar to later myths about the "Naihe Bridge 奈何 桥 See Wu Hung, 1994, p.100; Xin Lixiang, 2000, p. 331-334. This comparitive définition of Weshparkar was pointed out to the author by Professor Rong Xinjiang in December, 2003. See Rong Xinjiang, 2003b.

38

Les Sogdiens en Chine

good in life, it enters Heaven, and if it was evil, it falls into Hell.24 This image contributes to our understanding of the dualistic religious concept of good and evil in

Zoroastrianism. Besides this, there are still two problems which need to be examined. The fïrst is, when we examine the Sinifï cation of the fanerai practices of Sogdian people in China, we also must pay attention Sogdian influences on Chinese culture, as seen in those images with multicultural éléments from the Fu family tomb in Qingzhou, Shandong Province which hâve attracted the attention of scholars. This sort of multicultural quality is also seen to a dififerent degree in wall images and burial items from the recently discovered tomb of Xu Xianxiu in Taiyuan, Shanxi Province,25 which dates to 571, the second year of the Northern Qi Dynasty's Wuping era. Perhaps in contemporary or later burial sites, we can see more examples of this type which will provide deeper insight into the varions cultural, artistic, and stylistic changes seen during the Sui and Tang. The second problem is, these recently discovered tombs: such as those of Yu Hong, An Qie, and Lord Shi, are ail of Sogdian nobility. The interred are ail people of the upper classes, so the fiineral implements used include relief-carved and gilded stone inner and outer coffins. However, from the Wei, Jin, and Northern and Southern Dynasties on, and especially during the Sui and Tang, most Sogdian people in China and their descendants employed Sogdian fanerai methods. It is possible that these people's burials involved very few special funeral implements, or that they involved none and the tombs were already in the Han style. It is of great importance that these types of tombs be carefully excavated, and the remains inside collected and tested. This would further our understanding of the funeral practices and Zoroastrian religious beliefs of Sogdian people in China, and provide us with more material data.

According to Srosh Yasht Hadokht (i.e. Chapter 56 of the Yasnd), at the Cinwad bridge, the soûls of the good take on the appearance of beautifiil faiiy-lilce maidens and are safely taken across the wide bridge. They pass through the three passes of Good Thought, Good Speech, and Good Action, and enter Heaven. The soûls of evildoers take on the fonns of hideous monsters and are taken across a Cinwad tumed the width of a hair. They pass through three passes: Evil Thought, Evil Speech, and Evil Action, and in the end fal! into the terrifying dark forest ofHell. See Yuan Wenqi, 1997, p. 214-215. 25 Shanxi sheng Kaogu Yanjiusuo, Taiyuan shi Wenwu Kaogu Yanjiusuo, 2003, p. 4-40.

24

Carvin-gs of the Stone Outer Coffin ofLord Shi

Yang Junkai

39

Bibliography

Cai Hongsheng 蔡 鸿生

2001

Tangdaijiuxinghu yu Tujue wenhua 唐代九姓胡与 突厥文化 [The Culture of Nine-Sumames Hu and of the Turks in the Tang dynasty], Beijing: Zhonghua shtyu.

Lerner Judith 1995

LUO Feng 罗丰

1996 RONG Xinjiang 2001

"Central Asians in S ixth- Century China: A Zoroastrian Funerary Rite", Iranica Antiqua XXX, p.179-190.

Guyuan Nanjiao Sui Tang mudi 固 原南郊隋唐墓地 [A Sui-Tang cemetery in the Southern suburbs of Guyuan], Beijing: Wenwu chubanshe. 新江

Zhonggu Zhongguo yu wailai wenming 中古中国与 外来文明 [Médiéval China and foreign civilizations], Beijing: Sanlian shudian. 2003 a "Miho meishuguan Sute shiguan pingfeng de tuxiang ji qi zuhe” Miho 美术馆 粟特石棺屏风的 像及其 组合 [Images from the screening panels of the stone Sogdian funerary couch in the Miho muséum and their composition], Yishushi Yanjiu 艺术由研究,4, Zhongshan Daxue chubanshe, pp. 199-221. 2003 b "Foxiang haishi xianshen? 一 cong Yutian kan Silu zongjiao de huntong xingtain佛像 还是袄神？ - 从于阖 看路宗教的混同形 态 [Buddha or Xian spirit? Religious syncretismus on the Silk Road as seen through the Khotan case] , in Jiuzhou xuelin 九州学林 Vol. 1, No. 2, Shanghai Fudan daxue chubanshe, p. 103-110.

Shaanxi sheng kaogu yanjiusuo 陕西省考古 研究所 2000 "Xian beijiao Bei Zhou An Qie mu fajue jianbao” 西安北郊北周安伽墓 掘 [Preliminary report on the tomb of An Qie (Northern Zhou) discovered in the Northern suburbs of Xi'an], Kaogu yu wenwu 考古与文物,6, p. 28-35 2001 "XTan faxian de Bei Zhou An Qie mu" 西安 的北周安伽墓 [The tomb of Xi'an], discovered in Zhou) Wenwu 文物,1, p. 4-26. An Qie (Northern Shanxi sheng kaogu yanjiusuo 山西省考古 研究所 "Taiyuan Sui dai Yu Hong mu qingli jianbaoï, 太原隋代虞弘墓清理 2001 [Preliminary report on the inventory of the tomb of Yu Hong (Sui dynasty) in Taiyuan], Wenwu 文物,1, p. 27-52. Shanxi sheng kaogu yanjiusuo, Taiyuan shi wenwu kaogu yanjiusuo 太原市 文物考古 研究所 "Taiyuan Bei Qi Xu Xianxiu mu fajue jianbao55 太原北 徐显秀墓掘 2003 [Preliminary report on the tomb of Xu Xianxu ((Northern Qi) in Taiyuan], Wenwu 文物,10, p. 4-40. Tianshui shi bowuguan 天水市母物 隋唐 "Tianshui shi faxian Sui Tang pingfeng shiguan chuang mu" 天水市 1992 couch discovered funerary screen and with stone a tomb [The 屏 石棺床墓 in Tianshui], Kaogu 考古,1, p. 46-54. Wei Qingzheng 魏 征 1999 Gudai Yilang shenhua 古代伊朗神话 [Legends of Ancient Iran], Taiyuan, Beiyue wenyi chubanshe, Shanxi renmin chubanshe.

Les Sogdiens en Chine

40 Wu Hung 1994

XlANai 夏鼐 1955

"Beyond the Great Boundary: Funerary Narrative in the Cangshan Tomb", in J. Hay, ed.5 Boundaries 加 China. London, Reaktion Books, p. 81-104.

一 [Notes on the Archaeology of "Dunhuang kaogu manji 3" 敦煌考古漫 (三) Dunhuang, 3], Kaogu tongxun 考古通 , 3. Xin Lixiang 信立祥 Handai huaxiang shi zonghe yanjiu 代画像石 合研究 [General research 2000 on Stone engraving under the Han dynasty], Beijing: Wenwu chubanshe. Yang

Junkai 北周 族史君墓 2003 a "Xi'an you faxian Bei Zhou guizu Shi Jun mu" 西安又 [One more tomb of a Northern Dynasties noble discovered in Xi'an: the tomb of Lord Shi], Zhongguo Wenwu Bao 中国文物 , 26. 2003 b '*Guanyu xianjiao de disanci zhongda faxian —— Xi'an Bei Zhou sabao Shi jun mu" 笑于 藪的第三次重大 - 西安北周 宝史 君墓 [Third great discovery on mazdeism: the tomb of Lord Shi5 a Northern Zhou sabao, in Xi an], Wenwu tiandi 文物天地, 11, p. 26-29. Yuan Wenqi 元文琪 Eryuan shen lun — gu Bosizongjiao shenhua yanjiu 二元神 一 古波斯宗教 1997 ネ甲 研究 [A Dualistic Theology. Research on the religions legends of Ancient Persia], Beijing: Zhongguo shehui kexue chubanshe. Zheng Yan 岩 2001 "Qingzhou Bei Qi huaxiangshi yu ru Hua Sute ren meishu — Yu Hong deng kaogu xin faxian de qishi" 青州北 画像石与入 粟特人美 一 虞弘等考 古新 的 示 [Northern Qi Stone engravings from Qingzhou and the art of Sogdians in China : the importance of Yu Hong and others archaeological discoveries], in Wu Hong 巫 , ed.s Han Tang zhi jian wenhua yishu de 的互 与交融 [Between Han and hudong yu jiaorong 唐之 文化 Tang : Cultural and Artistic interaction in a Transfbrmative Period], Beijing: Wenwu chubanshe, p. 73-109. Wei Jin Nanbei chao bihua mu yanjiu 魏晋南北朝壁画墓研究 [Research on 2002 tombs with murais from the Wei, Jin, Northern Dynasties, Sui and Tang periods], Beijing: Wenwu chubanshe. 5

き

Yang Junkai

Carvings of the Stone Outer Coffin ofLord Shi

41

Supplementary remarks om the discovery of the stone sarcophagus of Lord Shi of the Northern Zhou.

Tn the March 2005 issue of Wenwu magazine, the Institute of Archaeology of the city of Xi5an published an extended report on the archaeological excavation of the tomb

ofLord Shi of the Northern Zhou?6 Although the Chinese version of Pr. Yang Junkai s contribution in this volume provided a large part of its content, this article gives also 5

precious information on the tomb itself and its contents. The editors hâve therefbre fbund it useful to present a brief summary of this newly published material, as a complément to Pr。Yang's article«

sonie

1. The tomb

The tomb of Lord Shi was excavated east of Jingshang village 井上村 in the Weiyang district 未央区 of Xi an, an area previously known as Longshouyuan 百丿京 5 which was an important necropolis located east of the Han dynasty city of Chang an from the Qin-Han period to the Northern Zhou. The tripartite structure of the tomb belongs to the tradition of Northern Zhou, and consists of a sloping tomb path prolonged by a succession of five vertical airshafts and five passageway compartments, a level corridor and a tomb chamber. The characteristics of these different éléments are as follow : 5

• Orientation : 186 ° on a north-south longitudinal axis, the entrance facing south. • Total length : 47,26 m. • The tomb path begins 0,9 m undemeath the actual ground level, beneath two sur自ce strata of soil (a 0,25-0,3 m deep ploughing level and a 0,6 m deep disturbed level). It is 16,3 m long, 1,7 m wide at its Southern end and 1,48 m wide at its northem end. The slope runs down 16,6 m at a 15° angle. It is 0 to 4,4 m deep. The vertical airshafts are square and slightly wider than the passageway compartments. Their dimensions are (from South to North), in meters: -compartment 1: Length 1,68, Width 1,4, Height of arch ceiling: 1,97 -compartment 2: 1,6ハ,3/1,9 -compartment 3: 1,65/1,32/1,8 -compartment 4: 1,6/1,32/1,76 -compartment 3: 1,56/1,28/1,46 -airshaft 1: Length 3, Width 1,45, Southern depth 5, Northem depth 5,75 -airshaft 2: 3,2/1,45/6,3/7,04 -airshaft 3 : 3,45/1,45/7,6/8,4 -airshaft 4: 3,4/1,44/9/9,7 専

26

Xfan city Institute of Archaeology and Cultural Relies Conservation 西安 市 文物 保 考古 所, "PreHminary report on the excavation in Xi'an of the tomb of Lord Shi, sabao in Liangzhou under the Northem Zhou dynasty” 西安北周凉州萨 保史君墓 掘 , Wenwu 3 (2005),pp. 4-33.

42

-

Les Sogdiens en Chine

-airshaft 5: 3,4/1,44/10,3/1 1 • The level corridor is vaulted and has a rectangular surface. It is 2,8 m long, 1,5 m wide, and 1,5 m high. It leads to a 1,54 m high and 1,46 m wide brick door, made of 22 layers of approximately 34-35x17-18x6-7 cm rectangular bricks. • Behind the brick door is another one made of stone, consisting of six pièces : a lintel, a threshold, two doorframes and two door leaves. • The corridor and tomb path, as well as the airshaft and chamber walls, were originally covered with paintings fixed on a white coating of which only faint traces hâve resisted the bad conservation conditions. • The chamber lies 12,3 m beneath the actual ground level, and is 3,7 m wide and 3,5 m long. Given the turmoil inside the chamber at the time of the excavation, the roof structure is not clear, and of the four walls, partly demolished, only 50 cm rernain. 2. The contents

a) The stone door (see fîg. 11, p. 44). • Total dimensions : 1,52 m high and 1,35 m wide. • The door leaves : both are 1,34 m high, 0,58 m wide and 0,08 m thick. Four rows of six 5,5 cm r〃-typenails ?LT are driven into the stone surface of each panel, which were both decorated with red, white and black designs of apsaras, lotus flowers and pearls. However, only very faint traces of these motives remain, however. • The lintel is 1,65 m long, 0,24 m high and 0,13 m thick, and has the shape of a pentagon. It is decorated on its main face by a light relief carving. The main figure at the centre of the composition is a four armed créature wearing a jewelled crown, curly hair, a nimbus and a fierce expression on its face. Sharp teeth emerge from its mouth, and it has pointed ears. The créature wears a round-necked suit and a pearl necklace. Its bottom hands are joined in front ofhis chest, and both upper arms are held up high, as if supporting the lintel’s superior border. A second créature appears at the eastem part of the composition, holding a wine-hom in his left hand. He wears a jewelled crown, has a moustache, and his head is slightly tumed towards the central figure. Both his arms are held upwards, and the bottom half ofhis body has been transformed into volutes of lush végétation. At the western end of the lintel, a bird headed créature as well as another monster with two homs, a curly tongue and a sinuous body look towards the central

scene. • The two doorframes stand at the east and west side of the door. The eastem pillar is 1,24 m high and 0,14 m wide, the western one measures 1,25 x 0,18 m. Apsaras and fierce looking guardian créatures standing in the midst of a luxuriant grapevine are carved in low relief on both frames. Ail the apsaras, except for one, wears their hair in a bun, is endowed with a nimbus, a pair of wings, a narrow sleeved garment and a long tunic. Part of them hold utensils (a wine hom, a jar, some cups), while the others play musical instruments (pipa, flûte, panpipes, etc.). The guardian créatures hâve flaming

Carvings of the Stone Outer Coffin of Lord Shi

Yang Junkai

43

hair, fierce eyes, a large mouth and sharp teeth0 They wear a pearl necklace crossed in of their ehest. One arm is held up, and the other rests on their hip. They stand on one foot while the other rests on a pile of rocks. The carving of the guardian on the eastem side is very detailed, so that bracelets were even depicted on his ankles and wrists. • The threshold is 1,04 m long, 0,18 mhigh and 0,125 m wide. fyont

b) Other artefacts

Due to the severe plundering5 very few artefacts were discovered in the tomb Nevertheless, a ring, an ear pendant and a coin - ail made of gold - , as well as a gilded bronze belt buckle and pièces of a terracotta lamp were fbund in the soil that filled the

sarcophagus. The ring s inner diameter is 1,9 cm, the outer one is 2,45 cm, and it weighs 13,37 g. It is made of gold and is set with a rectangular turquoise stone. On both sides of the golden brackets that holds the stone, a "V" type character is carved. • The gold ear pendant measures 3 cm and weighs 11,02 g. It is composed of a crescent-like upper section, under which are attached a bündle of minuscule gold pearls from which a white pearl hangs (diam. 0,85 cm). • The gold coin is most probably Byzantine, although its obverse and reverse motifs are hardly visible. Its diameter is 1,75 cm, it is 0,05 cm thick and weighs 1,75 g?7 • The bronze belt buckle is oval shaped and was gilded. It is 5 cm long and 3,2 cm wide. The fibula fits into a buttonhole plaque. The main piece of the buckle resembles a 触

5

horse saddle. • The terracotta lamp was made eut of argileous earth. Its mouth is wide, it has a square lip, and its profile reminds that of a slanting bow. There are traces of soot inside. It is 3,1 cm high. The diameter of the mouth is 9,4 cm, and that of the foot 4,6 cm. 27

Side A : An indistinct figure, apparently revealing its left profile, holding an orb toped by a cross in its hand. An inscription runs around the border from left to right, that is very difficult to read. Left side : right side : "HVLロXDXX”, under figure^ feet : "ONOB” Side A is very similar to the "VOT obverse of Byzantine coins produced under Theodosius II (408-450). A similar coin was discovered in Hebei in 1975-76 in the Northern Qi tomb of Li Xizong. The inscription read "VOTXX” "MVLTXXX”, which means "ceremony of taking an oath of the 20th year up to the 30th year". Lord Shi's coin probably mentions the ceremony that took place on the 40th year of the reign of Theodosius, i.e 438. CONOB stands for ^Confstantinople] Ob[rysum「,which means umint of Constantinople>,. Side B : In the centre one sees a standing figure. It has long wings on its left on top which is a star, and it holds a cross in its left hand. On the right part of the inscription, one reads "VI" and two or three indistinguishable letters that may form the word "Victoria". On the left side, one reads “VGGG", that stands for 4tAUGGG". Figures of the Goddess of Victory are frequent on Theodosius II coins. B side is an imitation of side A, and was probably made at the same time, or maybe slightly later, under the reigns of Marcian (450-457) or Leo I (457-474), nonetheless in the second half of the 5th Century. The détérioration of the coin shows that it must hâve been made long before it was placed in Lord Shi's tomb, although it may hâve been in the family for long. It is different from other Byzantine coins and imitation discovered in China : it was neither reemployed, nor has it a hole inside. It is also much lighter than other coins : authentic solidi found in China weigh between 4,4 and 4,5 g, whereas the one found in Lord Shi's grave weighs only 1,75. These informations are taken from Luo Feng 丰,北 周史君墓出土的拜占 庭 金 仿制 品析 Beizhou Shi Jun mu chutu de Bizhanting jin bi fangzhi pinxi "Some remarks on an imitation of the Byzantine Gold Coin from Shi's Tomb of the Northern Zhou”, in Wenwu, 2005.3, pp. 57-65.

1

44

__

Fig. 11: Door of the tomb, with its lintel and

doorframes.

Les Sogdiens en Chine

Yang

Junkai

Carvings of the Stone Outer Coffin of Lord Shi

Fig. 12: The broken epitaph on the lintel of the door of the outer coffïn.

Fig. 13: The threshold to the outer coffïn.

45

Investigations on the Chinese version of the Sino-Sogdian bilingual inscription of the tomb of Lord Shi

Sun Fuxi

From June 12 through October 28, 2003, while digging the foundations of the new building of the Xi’an Intermediate People's Court3 workers from the Institute for Archeology and the Préservation of Cultural Relies discovered an ancient tomb from Daminggong Township dynasty in Jingshang Village the Northern Zhou in the northem suburbs of Xi’an City. This tomb is in a deep pit with a slanting tomb passage; it has five airshafts and a vaulted chamber. Although the tomb had been robbed several times5 a stone sarcophagus imitating a wooden coffin, comprising five bays and a hipped gable roof, was unearthed. Furthermore, not only was the sarcophagus completely covered with colored reliefs decorated with gold fbil5 but even more important, it also bore a slate with a Sino-Sogdian bilingual inscription. of As my knowledge of Sogdian was insufficient, I asked Prof. Rong Xinjiang Peking University to contact Prof. Yutaka Yoshida E0ä of the Kobe Foreign Language University and requested that he translates the Sogdian text. On the basis of this translation, I began a tentative study of the inscription, which I hope will stimulate a response from experts in this fïeld.

The inscribed slate was originally laid on the upper part of the lintel of the sarcophagus. It is rectangular, 88 cm long, 23 cm wide, and 8.5 cm thick. The obverse side of the slate is engraved with a Chinese inscription on the left and the Sogdian inscription on the right. The Chinese inscription consists of 18 colums, partitioned by a vertical lines and written from right to left. The first column is inscribed in seal characters (zhuanshu while the rest in of the inscription is in the regulär script (kaishu The third and fifth characters in the first column, the second, third and fourth characters in the second column, the characters in the first half of the third column, the characters between the third and the fifth from the bottom in the fourth column, the characters between the fourth and the sixth character from the bottom in the fifth column, the eighth, ninth and eleventh characters in the eighth column, the characters between the seventh and the last one in the ninth column, the characters between the sixth and the third from the bottom in the eleventh column, and parts of the twelfth, thirteenth and fourteenth columns are all left blank, so that a total of 184 characters remain. There is a sideline on the right and left of the Chinese inscription, which is hence complété. The Sogdian inscription is engraved in the same type of column as the Chinese text, but it must be read horizontally. It begins immediately to the right of the first column of the Chinese text and continues, on that side for 32

Les Sogdiens en Chine

48

columns. A 33rd column lias been drawn but not inscribe, which leads us to believe that the Sogdian inscription is also complété. Due to the damage caused by the grave robbers, when the time came to excavate the tomb, the inscribed slate had been broken into over 20 pièces, which lay scattered here and there in the tomb chamber. Those in charge of the excavation tried hard tö find most of the pièces, but failed to find some of the smaller ones. Many characters were damaged and lost, and some were exfoliated. Some of the exfbliated fragments were restorable but a very small number were not. ï will begin by first transcribing the original Chinese text. In the following dénotés a blank space in the inscription, [ ] dénotés a character that is transcription, entirely lost, [-] dénotés a character that is partly damaged, illegible, and [?] dénotés a character that remains on the inscribed slate but bas not been definitely identified0

ロ

君石堂/ 國 人也 , 本居西域 , / 遠 遷 居長安 。/ , ロロロロ,水運應期,中/ 目他有 ロ,日百具德 。祖阿/ 京顯美,自 , 史磐陀 為本國薩 保 。父阿奴伽,並/ 懷瑾握踰 , 重規疊距 , 秀傑不/ 群。 立功立事 , 少挺ロ口,石父 僵英/ 聲,而君秉雲山,丘 ［?］初 閭擁護,其身為/ 薩 保判事曹主ロ,□□五［-］詔/ 授涼州薩保。 而天道［-］,ロ/ 。 芳永鹹 大象元年口 ［月或日］薨 於/ 薨 ,/ 家 , 年八十六 。妻康 氏 其 以其二年歲次庚子正月丁亥朔廿/ 長子/ 七已日 , 合葬永［?］縣 唄?］ , 雪 沙 ' 次 維摩 次 図多［?］ 並有孝行 , 乃為/ 父 造石口 區 ,刊碑墓道 , 永播/ 大周ロ州 君

ロロ

ロロ

ロロ

ロロ

ロ

A study of the writing reveals that the slate was probably inscribed by a literate Sogdian who was not very familiär with the Chinese language and culture. In his article Prof. Yoshida, makes clear that with its cursive script and early Sogdian style of writing, the Sogdian inscription on the slate is very similar to Contemporary Sodgian inscriptions both in its calligraphy and language, whereas the Chinese version is slipshod and unworthy of being praised. First, the inadequacies of the writer can be deduced from the number of blank spaces in the inscription. These are not due the requirement of a given literary style, but reflect the writer s clumsiness when attempting to reproduce the style of Contemporary Chinese epitaphs. Second, several Chinese characters are inscribed in a non-standard or distorted männer, such as the "國'' and "域'' in the second column "遠”and "遷”in the third column, and so on, are very numerous in the text. The above characters cannot be considered variants ? but are mistakes due to the 5

ざざ為;；工;： gun fuxi

Chinese version of the Sino-Sogdian bilingual inscription

49

writer's limited Icnowledge of Chinese.

His lack of knowledge prevented the writer from correctly writing the Hang 冰 in Liangzhou 涼州 and sa 薩 in sabao 薩 保 in seal characters, the style of calligraphy adopted in the title of the epitaph in the first column of the Chinese inscription. The fact that as of the second column, the writer began to use the regulär script (kaishu) which was the prévalent style of writing used in he Central Plain from the Six Dynasties to the Tang dynasty5 clearly demonstrates that the writer did his best to imitate the style used by the ethnie Han. Given the writer s ignorance, the large numbers of missing characters (almost as numerous as the one that are inscribed) understanding this text is particularly difficult. I thereföre seek the reader5s indulgence when reading my interprétation of the text ofthe epitaph.1 5

This is the stone house of Lord Shi 史君, [[sa]] bao ([薩]保) of [[Liang]] zhou 涼州 who lived during the Northern Zhou dynasty. His ancestors came from the Kingdom of Shi 史國. They originally lived in the western Régions (Xiyu 西域).[[6]] When they moved from the remote western Régions to Chang'an 長安, people believed he had [[2]].[[4]] During the period when the Merit of Water predominated,2 he enjoyed great famé at the capital of the Central Plain. Everyone believed [[3]], that he possessed all kinds of virtues. His grandfather A-shi-pan-tuo 阿史磐 陀 was formerly a sabao of his country of orig in. (Kingdom of Shi). His father A-nu-qie 阿奴伽 also had such moral strength that he was compared to good jade. As a man of outstanding (talent), he obeyed the rules, performed good deeds and was known fbr his achievements [[2]]. Even in his youth he enjoyed a great réputation. Lord Shi had inherited the integrity of his grandfather^s génération, which was loftier than mountains and clouds while his aspirations, were as broad as hills and valleys [[2]]. Supported by his fellow villagers, he was first and foremost the head of the Division of Supervision in the Office of the sabao (sabao fu panshi cao zhu [[ääz]] 薩 保府判事曹主 [事]); [[2]] Later, an imperial edict [?] appointed him sabao of Liangzhou. However, as [[just a the laws of nature condemn beautiful flowers]] to fade, no one can avoid the death.. At the age of 86, (Lord Shi) died at his home, on the [[2]] ofthe first year of the Daxiang 大 象 era (579). His wife, Lady Kang 康 氏, died at their [[home on the ... ]] day [[in the ... month]].3 On the 27th day in the first month of the second year of the Daxiang era (580), they were both buried in Yong [?] County (永 [?] 縣)」?] … the eldest son Pisha 沙, the second son Weimo 維摩 and the third son Xue [?][?] (雪 [?][?]) all showed filial respect. They built a stone [[house]] for their father in [[1]] in the(?) District4 and buried an

1

In the fbllowing text, words in square brackets hâve been added by the editors to facilitate the reader^ understanding. Double brackets indicate the number of blank spaces, when the translater h as not attempted any restoration. Question marks in brackets [?] indicate that a character has not been identifîed. (Editors' note). 2 Water, fîre, wood, métal and earth can be used to designate the natural cycle of years, or historié, inter or intra dynastie, periods. (Editors' note) Given that the entire sentence is composed of three blank spaces, any translation is bound to be problematic. (Editors5 note) Given Chinese usage when assigning place names, it is unusual to find only one character to designate a location. Qu (translated here as "district”) is the usual qualifier fbr buildings. Another, more grammatically valid interprétation, would read ” they built [[a]] stone [[house]] (zao shi \\tangyï\\ Furthermore, if,

50

_

Les Sogdiens en Chine

inscribed tombstone in the tomb’s passage, (in the hope that the merits and virtaes of their parents) would endure forever.

Except for a few complimentary formulas, this inscription mainly involves the names of persons, place names, official titles, and dates. The following is an analysis of these four features To begin with the names of persons, there are seven of them that can be listed in genealogical order: A-shi-pan-tuo (Lord Shi’s grandfather) ; A-nu-qie (Lord Shi’s father) ; Lord Shi, in Sogdian Wirkak ; Lady Kang (Lord Shi’s wife) ; Pi-sha (Lord Shi’s eldest son), Wei-mo (his second son), and Xue [?] [?] (his third son). Clearly, despite the title of the inscription’s which daims that the sarcophagus was the stone house of “Lord Shi”, in the text of the epitaph, each family member is not addressed by his given name preceded by his sumame as was customary among the Han people of the Central Plain, but only by his given name, which was the custom of the ethnie minorities at the time. This practice has been followed up to the génération of Lord Shi’s sons, which shows that the Shi clam was sinicized only to a limited degree. Next, the inscription mentions the following place names: the Kingdom of Shi, Chang’an, the western Régions, and the counties of Liangzhou, and Yong [?] In the following section I endeavor to explain some of these names. 1) The Kingdom of Shi: this is one of the small ancient States founded by the Jieshuangna Sogdians in central Asia, which is often translatée! as Qusha in ancient Chinese books and documents. It is located in or Keshi Qishi present-day Shahrisabz, which lies due south of Samarkand, Uzbekistan. Shahrisabz which, before the Sui dynasty, was was a dependency of the Kingdom of Kangju in books and records of the Han dynasty. It frequently translated as Suxian City took its place in history as a small semi-independent country, and after the Sui dominated the communications hub where the people from various places in the Amu and Syr River valleys congregated on their way to Afghanistan and India. Given that it on its possessed the dangerous and difficult Iron Gate Pass (Tiemen guan Southern border, it was a very important location on the Silk Road. It was here that Timur, the founder of Temurid in Central Asia, was bom in the fourteenth Century, and the city became the secondary capital of the kingdom. During the Sui and Tang dynasties, the Kingdom of Shi had dealings with China as one of the States of the Nine Cao W, Shi 5, An namely: Kang Clans of Zhaowu (Zhaowu jiuxing reign the of Emperor during (656-661) etc.). In the Xianqing era Mi Shi in this Gaozong of the Tang, the govemment established the Qusha Préfecture as its prefectural govemor (zhou kingdom, and designated Zhaowu Shi’ahe cishi In the Tianbao era (742-756) during the reign of Emperor Xuanzong of with the the Tang, the Kingdom of Shi was renamed the Kingdom of Laiwei as its dependency. It was located in Karshi, which lies to Small Kingdom of Shi the Southwest of present day Samarkand. these blanks are due to the limited knowledge of the engraver, it is curious that tang (house) has been left blank, whereas it has been written in seal characters in column 1 . (Editors’ note)

Sun Fuxi

Chinese version of the Sino-Sogdian bilingual inscription

51

2) Liangzhou: one of the thirteen préfectures set up during the Western Han dynasty. H (present-day The seat of its govemment was situated in Long County province) in the Hui Autonomous County of Gansu Zhangjiachuan City Guzang (present Wuwei Gansu and moved to dynasty, Eastem Han province) in the period of the Three Kingdoms. During the period of the Sixteen Kingdoms, the Former Liang, the Later Liang, and the Northern Liang all founded their States in this place. It was called Xiliang préfecture (Xiliang/i/ M)^W) during the Five Dynasties and the Xixia dynasty. In the Yuan x dynasty, it became the Xiliang région, and in the Ming dynasty it was laiown as the Liangzhou garrison (Xiliang zhou as well as the Liangzhou préfecture (Liangzhou fu in (Liangzhou wei the Qing dynasty. During the period of the Republic of China3 it was renamed Wuwei SM, which is the name by which it is still known today. 3) Yong [?] county: According to the inscription, it should be a county in the région of Chang’an during the Northern Zhou period3 but ancient records show there never was a county with “Yong” as the first character of its name ever since Chang’an was established. It is therefore difficult to know what it refers to. There are still two other place names. Although they do not appear in the Chinese inscription, they are mentioned in Yoshida’s translation of the Sogdian version. Consequently they deserve a brief explanation. Guzang was established as an administrative entity in the second 1) Guzang year of the Yuanshou period during the reign of Emperor Wudi of the Han dynasty (121 B.C.) where the modern Wuwei city is located today. It was the seat of the govemment of the Wuwei commandery (Wuwei jun Q5) in the Eastem Han dynasty. It then became, the capital of the Former Liang, the Later Liang and the Northern Liang during the period of the Sixteen Kingdoms. In the Tang dynasty, it became the seat of and the seat of the Military governmental of the Liang Region (Liang zhou Consequently, the Sogdian name refers Govemorship (jiedushi of Hexi to the same région as “Liangzhou” mentioned in the Chinese inscription. 2) Xiping was the name of a northwestem commandery or county in ancient times. In the Jian’an period (196-219) during the reign of Emperor Xiandi of the commandery, with the seat of its Eastem Han, it was separated from the Jincheng govemment located in Xidu Qinghai province). The (modem Xining City area under its jurisdiction corresponds to the present Huangshui River basin in the Huangyuan counties in Qinghai province. In the and Ledu County period of the Sixteen Kingdoms, the Southern Liang once established their capital in this place. During the Northern Wei dynasty, it was changed into the Shanshan garrison armies, Emperor Yangdi of the Sui After defeating the Tuyuhun dynasty reestablished it as the Xiping commandery. In the Tang dynasty, it was tumed into Shan préfecture (Shanzhou in the first years and then tumed back to Xiping commandery in the Tianbao era, but no changes were made in the area under its jurisdiction.

The official titles involved in this inscription include sabao of the Kingdom of Shi, the Chief of the Division of Supervision in the Office of sabao, and the sabao of Liangzhou. These three official positions are all related to the Office of sabao and contribute to our understanding of the Systems of sabao and the Office of sabao in the

Les Sogdiens en Chine

52

period from the Northern Zhou through to the Sui and Tang dynasties, means the same as safu jWW, and the term sabao also spelled 1) Sabao fbund in the historical records of the Northern Dynasties, Sui and Tang. The term originally referred to a Sogdian trade caravan leaders. However, as Sogdian communities were also the centers of the Zoroastrian cuit during this period, sabao were thus also Sogdian religious leaders. Hence, the words "sabao of his own country” mentioned in the inscription may be used in the latter sense. 2) Chief of the Division of Supervision in the Office of sabao (sabao fu panshi cao between the Han and the Tang dynasties, the Sogdian zhu [shi] WlSWWWWÏ [^D： people who inhabited the Valley of the Two Rivers drifted to the western Régions and the interior of China in large number. The Sogdian term sabao gradually evolved from designating caravan leader into a local administrative. According to varions documents discovered in recent years, the structure of the Office of sabao (sabao %) became increasingly complex due to the massive influx of Sogdians. The Division of Supervision (panshi cao) is probably the name of the section of the Office that was in actually in charge of handling the Office s business, while the Chief (zhu shi) was the person in overall charge. The sabao of Liangzhou was the political and religious leader of the Sogdians in “Liangzhou was Liangzhou. According to the biographers of the monk Xuanzang from all the Region. Merchants Hexi the meeting place of and the the metropolis the Congling east the of to and countries situated on the borders of Xifan Mountains, used to corne and go to this city in a ceaseless stream.”5 According to Kingdom ofWei “Biography of Cang Ci” (Cang Ci zhuan Ä'MW) in the History of p. 512-6), as (Weishu in Records the Three Kingdoms (Sanguo zhi flourished in Sogdian people (220-265), the early as the period of the Three Kingdoms the area of the Hexi Corridor around Liangzhou and had their own armed forces and leaders. Written and archaeological sources reveal, that as an important settlement of the Sogdian people, the Wuwei district was inhabited by diverse Sogdian communities clan, which from the Six Dynasties to the Sui and the Tang dynasties. While the An held the post of sabao for générations, and thus came from the Kingdom of An had different non-Han nationalities under their command, and controlled the political situation of Liangzhou, they were not the only clan to hold this position. Natives from such as Kang Bada who were sumamed Kang the Kingdom of Kang as well as Lord Shi from the Kingdom of Shi, who is mentioned in the inscription, also held the office of sabao. These facts produce a number of questions. For example, did the different Sogdian living in the Liangzhou area at communities of the Nine Clans of Zhaowu the time, each hâve their own leader (sabaoyi Or again, is it true that Lord Shi was appointed sabao of Liangzhou by an imperial edict from the emperor of the Northern Zhou? And was the Office of sabao to which he had been appointed, so large that it had been divided into sections? Were ail the Sogdian sabao of Liangzhou, whether 9

5

(The Life of the Da Ci fen si Sanzang fashi zhuan Huili ^201 and Yancong Grace), and Zhonghua Beijing, Great Kindness of Master of the Sanzang (Tripitaka) of the Monastery shuju, 2000, p. 11.

Sun Fuxi

Chinese version of the Sino-Sogdian bilingual inscription

53

designated by emperor or not, on good ternis with him? What conclusions can we draw from the above queries? During the period under discussion, the Liangzhou area was inhabited by various Sogdian communities from the States of the Nine Clans of Zhaowu, each with their own political/religious leader or sabao, who managed their internal affairs. However, in order to strengthen the govemment of the Sogdians in this area, the Northern Zhou imperial court set up an Office of sabao, which was administered by officiais directly appointed by the emperor. This Office was a complex Organization, divided into sections, each of which had its own fonctions to perform. It incorporated members from various Sogdian communities and was in charge of coordinating the relationship between different groups and the trade relations between the Sogdians and the Northern Zhou imperial court of the day. Finally, the inscription refers to four dates, three of which are accurate and one which is problematic. The accurate dates are those of Lord Shi’s and of his wife, Lady Kang’s deaths, and the date of the couple’s burial, but the date of Lord Shi ’s birth remains uncertain. 1) The date of Lord Shi’s death: “(Lord Shi) passed away in / on the (month / day) in the first Daxiang year”. Also considering Yoshida’s interprétation of the Sogdian version of the text, that “then, here in Khumtan (= Chang’an), he himself became god-souled (=died) in the year of the pig, one the seventh day of the fifth month (= 579 C.E. 16th June)”, therefore the exact date of Lord Shi’s death was June 16, 579. 2) For the date of Lord Shi’s wife, Lady Kang’s death: we only have the number of the day of her death. Yoshida’s interprétation of the Sogdian version goes as follows: “And again his wife became god-souled (= died) in the 6th month, on the 7th day, on a rabbit day, in this given (or blessed?) year, in this month, on this day (= 579 C.E. 15th July).” That is to say, the exact date of Lady Kang’s death was either July 15 or 25, 579. 3) The date of Lord Shi and Lady Kang’s burial: “on the 27th day, the day of Serpent (si EEL), in the first lunar month, with the Dinghai T day as its first day, in the next year, the Gengzi year”. Taking Yoshida’s interprétation of the Sogdian version into considération: “(It was in the period of) Great Chu Tay Zan (= Dazhou Daxiang) that is year two, of the year the mouse. The first month, on the 23rd (day) (= 580 C.E. 23rd February).” Hence the exact date Lord Shi and his wife’s joint burial was February 23, 580. 4) It now becomes possible to deduce the date of Lord Shi’s birth: “(Lord Shi passed away) at the age of 86, [[2]] in the first Daxiang year.” We can thus infer, that Lord Shi was bom in the 17th year of the Taihe era during the reign of Emperor Xiaowen of the Northern Wei, or 493.

Though the inscription records that Lord Shi’ sons “inscribed and buried a tombstone in the tomb passageway”, besides the inscribed slate, there is no tombstone either in front of the tomb, or in the tomb chamber. Nor was there an inscribed slate in one of the tomb’s passageways or an epitaph in this passageway as been the case of the tombs of Yuhong According to the archaeological material and of An Qie that has been recovered, an inscription such as the one found in the tomb of Lord Shi,

54

-

Les Sogdiens en Chine

inscribed on a slate placed on the lintel of the south door of a sarcophagus, and written in both Sogdian and Chinese, in imitation of the epitaphs of the Central Plain Han of the Northern Dynasties, is unusual. This, probably reflects the Sogdian custom of inscribing drawings and texts on ossuaries in which the remains of the deceased were placed. Based on the aforementioned reasons and Yutaka Yoshida’s studies, the following conclusions can be drawn: the clan of Lord Shi was very influential in the Kingdom of Shi, one of the States of the Nine Clans of Zhaowu. Shi’s grandfather was once a sabao of the Kingdom of Shi, and his father was also an eminent personality. Lord Shi was boni in the 17th year of the period of Taihe era during the reign of Emperor Xiaowen of the Northern Wei dynasty (493). He enjoyed a distinguished réputation in Chang’an soon after the family moved there. He held an important post in the Office of sabao prior to being appointée! sabao of Liangzhou by emperor of the Northern Zhou. His wife, a native of the Kingdom of Kang, was bom in Shanshan in Xiping Commandery. The couple married in 519 and gave birth to three sons. Influenced by the culture of the Central Plain Han, they displayed their filial piety by building a magnificent tomb for their parents. The tomb of Lord Shi was found in Jingshang Village, Dammggong Township in the northem suburbs of Xi’an City. It was located 2,250 meters west of the tomb of An whose occupant was buried in the sixth month of Qie, while the tomb of Kang Ye the sixth year of the Tianhe era of the Northem Zhou era (569), which was recently discovered by members of Xi’an Institute for Archeology and Préservation of Cultural Relies. This tomb was located a mere 150 meters north of the tomb of An Qie. Since Lord Shi was sabao of Liangzhou and An Qie was sabao of Tongzhou [q] »N during the Northem Zhou dynasty, and Kang Ye’s family, according to his epitaph, held the office for générations, this région was very of Great Heavenly Sovereign {datianzhu Chang’an in densely populated who inhabited the Sogdians cemetery of likely the communities during the Northem Zhou. Furthermore, at the time, the Northem Zhou capital was situated in a place about 3,000 - 5,000 meters west of this région. In view of the date that was previously suggested it seems likely that at the time, there was, within the city of Chang’an, a residential area of inhabited by Sogdians who had corne to China and their descendants. The outstanding members of this community were appointed officiais to other parts of the territory of the Northem Zhou mlers, but the families of these officiais continued to live in Chang’an, and that is also where they were buried. It would be of great value to our studies of Chang’an during the Northem Zhou dynasty, if we knew more about the condition of the Sogdians entering China at Chang’an.

Fig. 1: Chinese part of the inscription.

inscrption

bilngual

Sino-gdian

the

of

Chines version

Fuxi

Sun

The Sogdian version of the new Xi’an inscription Yutaka Yoshida

Tn view of the recent discoveries of several tombs of Sogdians by the Chinese archaeologists, it was more than likely that Sogdian epitaphs would eventually be unearthed in China. Yet, the news of the discovery of a Sino-Sogdian bilingual inscription gave us a great shock. Unlike ordinary Chinese epitaphs placed in tombs, the “lord Shi” was found newly discovered inscription of a Sogdian Chinese Shi Jun above the entrance door of a house-shaped sarcophagus built inside the tomb. The Sogdian text occupies the right half of an oblong slate measuring 88 cm by 23 cm. It comprises 32 short lines inscribed vertically. It is clear from the contents that the inscription is complété as it stands and this observation is also supported by the fact that one line is left blank at the right end of the slate. The Sogdian part was broken in the middle and one word of line 14 is damaged and lost. Furthermore, the surface of the upper part of lines 1-7 was exfoliated with the result that the first word of line 6 and of line 7 is lost. Fortunately, what has been exfoliated from lines 1-5 has been recovered with only a slight loss and one can read ahnost all the words.1 Before editing the Sogdian text I should like to express my sincere gratitude to Professor Sun Fuxi, Director of the Xi’an Institute for Archaeaology and Préservation of Cultural Remains and to Professor Rong Xinjiang for giving me an opportunity to work on the Sogdian inscription and for permitting me to publish my study. I also thank Professor N. Sims-Williams for reading the earlier version of this paper and for giving me valuable suggestions. Needless to say, I alone am responsible for any shortcomings. In this article I present the text of the inscription and its translation together with comments conceming the contents and linguistic problems. In the text [square brackets] are used to indicate completely damaged parts, while partly legible letters are surrounded by (round brackets). In the translation round brackets indicate words not found in the text but added to improve the English; doubtful translation are marked with an asterisk (*), which has been very sparingly used.2

1

For the archaeological details of the tomb itself see Y ang Junkai, 2003, 2004 and his paper in this volume. At the final stage of drafting the present article I was provided with a rubbing of the inscription which reproduces the late st state of the Stele after its final restoration. In the Sogdian version two small chips were added in lines 6 and 13 and I was able to improve upon my édition distributed in the Conference. I am gratefiil to Professor Rong Xinjiang and his colleagues who placed the rubbing at my disposai; they also permitted me to eite their late st readings of the Chinese version in this article.

2

Les Sogdiens en Chine

58

Text 1 mz'yxw [t’]y cw t’y z’nw 5w’ 2 sr8 pr m[w]s sröy prtmy 3 m’x ’wyh 23yh KZNH ZY 4 wm’ ’t (k)sy’n’k kwtr’k 5 kc”n (’)[..]’ntk cnn ßyp’wr 6 [ ]tßry[ ](.)rt kc’nc’n’k srtp’w sic 7 (.)[ ](t?) pr sywSyk-stn 8 ’st’m ö’r wyrk’k n’mt ZK 9 wn’wk BRY wn’wk ZK rstßntk 10 srtp’w BRY rty ZKh kty-”ßr synpyn 11 z’t(c)h wy'wsyh n’mt rty wyrk’k 12 sr(t)p’w k’w synpyn pr k’s srö 13 wxwswmy m’xy ’ßtsytyh (x)ryws’k 14 mySy kty”ßr pr’yßtw (5’)[rt ](r)ty 15 pytsrS my8’ xwmt’n xwty pr k’s 16 sr5 p(nc)my m’xw ’ßtsyth ßyr(w’n) 17 ’krty rty sy ms ZKh kty”ßr pr 18 wxwsw-my m’xy ’Btsytsytyh sic xr(yws’k) 19 myöy ßyrw’n ’krth pr ’yö pcßytk 20 srô ’yS m’xw ’yS myöy rty nyst w’tö’r 21 ’kytn ”z’y ZKZY ’myöry ’prtk L' ”y rt(y) 22 ms sk’wrö ZK zw’ntö’my zmnh ’nw(y)wth 23 cnn ’prth pyst ’wn’kw sk’wrtr 24 ’kyt pr mrt(x)m’k ô’my ’yö srö ’y§ m’x(y) 25 ’yS myô wyr' w8wh prwh L’ ptz’n 26 pr’’w wy-n’nt ms k’w ’xwstm’xw 27 pr ’yö zmn’ ’nywth pr’yw ”(zw)n 28 ß’y BLANK 29 ’krty ywn’k snkyn’k ßykt’k 30 ßr 'ysmnßntk örymtßntk 31 pr’wtßntkws”n ’BY’(?) 32 m’Sryh sc’w wy”kh 33 BLANK

Notes on the text line 6: [ ]tßry[ ](.)rt. r- of -rt is very uncertain and what I read -r- looks also like a flat ß. line 7: At the beginning of the line a short horizontal Stroke is visible which is not connected to the following letter. line 10: synpyn is squeezed at the end of the line but the reading is confirmed by its occurrence in line 12. line 11: Of what I read z’tch, only z- and -h are certain. line 12: t of srtp’w is not inscribed properly and one finds only the beginning and the end of the letter; it lacks a circle, which constitutes a main body of a letter t. line 16: srö looks more like syô, but there is hardly any doubt as to this reading; no

Yoshida Yutaka

The Sogdian version of the new Xi’an inscription

59

circle is found of p- of pncmy.

line 17: What I read ZKh is actually written as nph or zph, but neither of them

makes sense.

line 18: The yws’k of xryws’k does not look like anything in Sogdian script. However, the entire shape does somehow remind one of combination of the five letters. line 21: ”y may also be read z’y. line 22: y of ’nwywth does not look like any Sogdian letter. Apparently, the engraver first wrote w but noticing his mistake he tried to correct it by tuming it into y without success. For the correct spelling see line 27. line 24: % of mrtxm’k is not properly executed and looks like a letter c. line 25: The scribe at first wrote nöwh or zöwh but corrected the first letter by adding a circle to n/z-, prwh may also be read pr(y’)h. line 26: The last letter of pr”w may also be read r. w of what I read wy-n’nt seems to be corrected from ß or r. line 27: The last word of the line looks more like ”wzn but ”zwn seems to have been intended. line 31: On ’BY’ see my notes below.

Translation [1-3] (It was in the period) Tay Zan of Great Chu (= Dazhou Daxiang). The year two. In the first month of the year of the mouse. On the 23rd (day). [3-13] Thus, there was a man of a family from Kish [domiciled?] in (a city called) Kachan. He [obtained?] a [title?] sabao of Kachan from the Emperor (and was) a *grandee in the Sogdian land. [8] He is named Wirkak, Wanuk’s son. Wanuk, sabao Rashtvandak’s son. His wife was bom in Senpen and was named Wiyusi. [11] Wirkak married his wife in Senpen in the year of the pig, in the sixth month, on the seventh day (which was) the day of the rabbit. [15-17] Then, here in Khumtan (= Chang’an), he himself died in the year of the pig, in the fifth month, on the seventh day (= 16th June 579 CE). [17-20] And again his wife died in the 6th month, on the seventh day, on the day of the rabbit, in this *given year, in this month, on this day (= 15th July 579 CE). [20-28] There is no such living being who would be bom and would not be owing death (= Nobody cannot avoid the fate to die.) [22] It is also difficult to complete a period of time in the living world. (= It is difficult to live out one’s natural span of life.) However, this is even more difficult that in the world of men (i.e. life on earth) a husband and a wife see each other (i.e. live together?) without recognizing (it) (i.e. unintentionally or by accident) during these (same) years, these (same) months, and these (same) days, and that they would have life together during this (same) period of time also in the paradise. [29-32] This tomb (i.e. god-house) made of stone was constructed by Vreshmanvandak, Zhematvandak, and Protvantak (or Parotvandak) *for the sake of their father and mother in the suitable place.

Les Sogdiens en Chine

60 Commentary

General remarks

Although the inscription comprises as many as 32 lines, it is not very long because each line is short and contains an average of only four to six words. The content is largely clear except for lines 22-28, where the author of the inscription seems to insist how difflcult it is for a man to live out his natural life span and how rare it is for a husband and a wife to live their whole life together. Unfortunately, the Chinese version is not only damaged but also so poorly inscribed that some characters are hardly legible It is also to be noted that several places in the Chinese version are left blank without being filled with characters; furthermore3 the text differs considerably from that of the Sogdian version. Thus3 fbr example, the names of the couple buried in the tomb? Wirkak and Wiyusi, never appear in the Chinese; while Wirkak is simply called Shi Jun 史君 "lit. lord Shi\ ie a lord originating from the country of Kish, his wife is mentioned as having died on a certain day and is merely referred to as qi Kang shi 妻康 氏 ÉÊwife from the Kang fàmily". One does not find any part that corresponds to lines 22-28 of the Sogdian version either, where the exceptionally long married life and the coincidence of the year and date of death of the deceased couple seem to be the subjects The relationship between the Sogdian and the Chinese versions is reminiscent of another bilingual epitaph of a Zoroastrian woman unearthed in Xi'an in 1955. This inscription is dated to 874 CE and is inscribed in Chinese and Pahlavi; while the Middle Persian version ends with a typical Zoroastrian phrase wishing the deceased to stay in the heaven with Ahura Mazda, the Chinese inscribes nothing more than the dates of the woman's birth and death.3 Based on the inscription, one can establish the following family tree for Wirkak's family: Vreshmanvandak Rashtvandak—Wanuk— Wirkak Zhematvandak | P(a)rotvandak Wiyusi As far as one can see from the Chinese text, one finds at least fïve names transcribed in Chinese characters. They are as follows: Rashtvandak (rstßntk): e shi pan tuo 阿史盤陀 (lines 5-6) *-a si b uân Ja.4 Wanuk (wかw左)： e nu qie 阿奴伽 (line6) * nuo gi pia Vreshmanvandak (ßrysmnßntk): pi sha 毘 沙 (line 17)5 *bji mo 維摩 (line 17) ^im muâ^ Zhematvandak (Srymtßntk)\ 3

Humbach, 1988. Middle Chinese forms cited in this article are those reconstructed in Karlgren, 1957. 5 毘沙 is a Sinicized désignation of Skt. Vaisravana, of which the Sogdianized form is ßr ysmn. 6 Middle Chinese initial * シ or j- is not expected to represent Sogdian z-. One may be able to think of several explanations for this slight discrepancy. At fîrst, it can be a simple error in transcription; as noticed above, the Chinese text seems to hâve been executed by a stonemason who was not well versed in writing Chinese characters. Secondly, the comparison of yan mi pan tuo 炎蜜畔陀 ^jiäm miaët b uân d'â and she mi pan tuo 射 蜜畔陀 *dzja miet b "uân dd found in one and the same Turfan text (cf. Tulufan Chutu Wenshu, vol. 3, p. 317-325) seems to point to vacillation in pronunciation of the initial consonant of ôrymt 4

Yoshida Yutaka

The Sogdian version of the ne^v Xi "an inscription

61

(line 17)*7 % From the point of view of the transcription System, they are not accurate enough: for example, one does not find a character standing for the initial consonant of in the Rashtvandak. Particularly astonishing is the lack of the characters pan tuo last three names.8 Apparently the author of the Chinese version did not care much for exact or rigorous transcriptions and was satisfied with a broad and approximate are well-established Chinese désignations and rendition. On the other h and the reason why the two Buddhist respectively; Vimalalärti and of Skt. Vaisravana names were chosen for (transcriptions of) their names is not clear to me. The inscription is one of the most important sources for the history of Sogdians in China. First one must examine the fbllowing four dates mentioned in the inscription: 1) IL 1“3: The day when Wirkak and his wife were buried. That is the 23rd day of the first month of the second year of Daxiang (= 23rd February 580 CE). This [ HB J SW “First corresponds to lines 15-16 of the Chinese version: T month, of which the new moon (= first day) feil on ding-hai. On the twenty-third day (corresponding to) 2) 11. 12-14: The day when Wirkak married his wife? i.e. on the seventh day of the sixth month in the year of the pig. The day in question is described as the day of the rabbit and when one looks up the Chinese calendars of the sixth Century, the seventh day of the sixth month of the year of the pig can only fall in 519 CE (= 19th July 519 CE). In that year Shi Jun was 26 years old; when he died in 579 CE he was 86 years old, and one can calculate that he was bom in 494 CE. Needless to say, at the time,

Protvandak (pr \vtßntK)\ fo lu duo

= M. jymt; in other words z- of this name was sometimes pronounced with very weak affriction and sounded like semivowel^-. Thirdly, the pronunciation of Old Chinese may be considered: the Middle Chinese initial / of this character is known to originate firom Old Chinese voiced consonant which may hâve corne to be pronounced like z- before it finally became a semivowel. In this connection, Faxian’s (leaving China for India in 399 CE) désignation of Karashahr is also to be mentioned. He called the town Wuyi and this name has posed a considérable problem among the scholars. However, if one takes the shape of M in cursive script into considération, one can certainly ( > Middle Chinese Jän) can represent a emend wu S into yan of which the Old Chinese form closed syllable *är, cf. also the Old Chinese name of the oasis state yan qi standing for Argi, the ancient name of Karashahr, cf. Henning, 1938, p. 571. If one considers the Archaic pronunciation of *diar (> Middle Chinese *z), the initial consonant ofyi must hâve been a weak fricative Sound like z or d. Thus, the transcription is most likely to represent a Tocharian original form ^arzi which is spelled as ärsi in Brahmi script and is assumed to have been the old name of Karashahr ultimately derived from ^argi. On the Old Chinese initial ^d- standing for Sogdian fricative sound see also note 18 below. 7 I owe this superb reading of the second character to Professor T. Moriyasu, who suggested it when I imparted him my assumption that the character in question should correspond to -v'w- of pr’wtßntk. As he remarks, even if his reading is correct it was so badly inscribed that only its contour looks something like iS. It may be another instance which indicates the stonemason’s ignorance ofwhat he was inscribing. 8 This is an opportunity to draw attention to the fact that there used to be three ways of transcribing the Sogdian name element ßntk “slave, servant” in Chinese characters: (a) pan tuo (c) (b) pan tuo pan tuo The three are homophones in Middle Chinese: ’uan d’a. While (c) is restricted to the Chinese material discovered in Turfan, (b) is the form employed in the texts written in the Sui and Tang periods, such as Xuanzang’s biography and Shi Shewupantuo’s epitaph. (a) appears here in this inscription as well as in the Dunhuang census document of 751 CE studied in Ikeda, 1965. Whether a weak trace of the first letter is visible or not is not clear to me.

Les Sogdiens en Chine

62

when a baby was bom in China, it was considered to be one year old and every time the calendar was renewed a year was added. 3) 11. 15-16: The day when Wirkak died5 ie on the seventh day of the fifth month of the year of the pig (= 16th June 579 CE). 4) 11. 17-19: The day when his wife died, i.e. on the seventh day of the sixth month

(= 15th July 579 CE). The day in question is also said to be the day of the rabbit. However, the seventh day of the sixth month of the Chinese calendar in 579 CE tums out to be the day of the tiger. Since the day of the rabbit cornes just after that3 one can assume that the calendar of the author of the epitaph is one day behind the calendar restored in Chen Yuan, 1956, which I consulted. After establishing the chronology of the events described in the epitaph, one may wonder if there was any relationship between Shi Jun's family line and that of the Shi clan of Guyuan located between Lanzhou and Xi'an, where the tombs of some of the family members were excavated?0 The earliest tomb belongs to Shi Shewupantuo who died in 609 CE when he was 66 years old; thus he was bom in 544 (Middle CE. It is to be noted that the Chinese characters she ww pan tuo Chinese "dz'ia mimt b %ân d'â) stands for Zhematvantak. According to Shewupantuo5s epitaph his great grandfàther (as well as his grandfether) was a sabao in the home country, which coincides with what is reported in the Shi Jun’s epitaph. Moreover, in the tomb inscriptions of Shewupantuo’s two sons and a grandson, Shewupantuo’s father “great sabao" and cishi “govemor” is described as either mohe sabao i.e. Chang’an. of the préfecture of Shan or mohe sabao of the capital of Zhou Incidentally, the préfecture of Shan during the Tang dynasty designated Xiping, with which I propose to identify synpyn of the present Sogdian inscription (see below). However, Shewupantuo s father is called Duoxiduo or Renchou (on this name see grandfather Boboni and great grandfather Miaoni below); they bear no similarity in pronunciation to Wirkak, Wanuk, and Rashtvandak of our inscription. As a whole it is almost impossible to identify Wirkak’s son Zhematvantak with Shi Shewupantuo of Guyuan. Nevertheless, one may well hesitate to consider ail these corrélations as mere coincidences. From a philological point of view, the script and the language of the inscription are exactly what one expects on a monument from the end of the sixth Century, and are very similar to those found in the Bugut Inscription of the same period. ' (aleph) with only one hom is the style of writing of the early cursive script and one can sometimes distinguish % (cheth) from y (gimel). The shape of some instances of y (jod), which looks like a Roman letter s, is spécifie to this inscription and one sometimes fînds it difficult to distinguish y from ' (aleph). Another peculiarity of this inscription is that several letters are inscribed carelessly and do not look like any Sogdian letters at ail. One typical example is srtp 'w in line 12; as noted above the letter / of this word lacks a circle. This situation leads one to assume that the text was inscribed by a stonemason who was illiterate in Sogdian. (I owe this observation to Professor Sims-Williams, who saw the inscription in situ) What can be gathered from the rubbing of the Chinese version also créâtes the same impression. 5

10

Cf. Luo Feng, 1996; Yuanzhou Lianhe Kaogudui, 1999.

Yoshida Yutaka

The Sogdian version of the new Xi’an inscription

63

Finally some words on the relationship between the reliefs surrounding a sarcophagus and the inscription may be in order.11 Récognition that the tomb was constructed by the deceased couple’s three sons provides a valuable due to under¬ standing what is represented in the reliefs. When one sees the panels of western and northem sides one notices that the father and his three sons are depicted several times. For example, on the lower register of the left hand panel of the western side, four people are depicted as leading a caravan. The continuation of the scene seems to be found on the lower register of the right hand panel from the northem wall, that is to say just around the north-westem comer; on the upper register, four people are depicted as sitting around the entrance of a yurt and seem to be negotiating with a figure sitting within the tent. Apparently, the scene depicted on the panel took place in the destination of the caravan. When one compares this scene with similar scenes found on the panels of the fimeral couch of An Jia and that of the Miho collection12, the figure sitting in a domed tent of the Shi Jun panel differs from the other two in that in the former, the man wears a garland, while in the other two portrayals, they hâve long hair, which is believed to indicate their Turkic ethnicity. The figure of the Shi Jun panel may possibly be identified as a Hephthalite leader, since the Hephthalites were a nomad people and had close contacts with Sogdians. On the three panels on the left, one finds représentations of the family’s life. In the two central panels hiking scenes are depicted. The lower registers are occupied by the wife and her handmaidens, while on the upper registers male members of the family, that is the father and his three sons, are shown again being served by a male servants. When one considers the possibility that the scenes unfold clockwise, the panel on the left edge of the northem wall may perhaps be related to the death of the couple; they are depicted as being saved by two angels from a place that looks like a pond.

Linguistic and philological notes

In what follows I shall comment on difficult words and expressions. Ï t’y cw = Da Zhou *d’äi ts^du. t’yz’nw = Da Xiang *d’äi z^ang. It should be noted that w ofz’nw is silent and is an Orthographie device to avoid a long tail of the preceding n. On -n transcribing Middle Chinese -ng, cf. kc”n for Guzang *kuo tsäng. 3 23yh “on the 23rd day”. One can equally well read 23’h but I prefer 23yh and take it for the oblique case of the special ordinal number employed for denoting days of a

month.13 4 (k)sy’n’k “originating from Kish”. This adjective is attested in the Nafnamak.14 It is also found in a Mug document.15 Incidentally, Xuanzang’s désignation of Kish, jie shuang *kiat simg nâ, stands also for this adjective. 11

For the reliefs see now Xi’an shi wenwu baohu kaogu suo, 2005, p. 4-33. Miho Museum, 2004, Plates 5c and 9b. 13 Sims-Williams, 1990, p. 185-186. 14 Sims-Williams, 1992, p. 54. 15Livsic, 1962, p. 220-221. 12

64

Les Sogdiens en Chine

5 Foliowing Professor Sims-Williams’s suggestion I regard (’)[..]’ntk as a word meaning “resident, domiciling”. I venture to restore (’)[zw]’ntk “living”, an old fashioned present participle provided with the suffix - ’ntk.16 kc”n or kc’n is attested also in Ancient Letter II and was identified by H. B. Henning with Guzang 6 /" ]tßry[ ](.)rt. One may perhaps restore a preterite form [ ]tßry[ ô](’)rt meaning “he obtained ([a titel] sartpaw)”, though no suitable verb is known to me. In my first édition prepared when a piece containing ]tßry[ was not available, I proposed to restore the gap as [’ßy’rtô’]rt “ he obtained”. Curiously, what has remained comprises almost all the letters restored by me (t,ß, r, and y), which are alined in a totally different order. 6 srtp’w. Occurring also in lines 10 and 12 it cannot be an error but a spelling variant of s’rtp’w attested in Ancient Letter V. This spelling seems to represent a secondarily shortened form. 7 sywöyk-stn “Sogdiana”. This form has not been attested elsewhere. During the Conference Dr. P. Lur’e argued that sywôyk-stn literally means “Sogdians’ place”, and that it is not likely to dénoté Sogdiana itself but a certain place in China where many Sogdians were residents. However, I still believe that the expression refers to Sogdiana, because the Chinese version clearly States that Shi Jun had originated from xiyu “the Western Region”, which must refer to his ancestors’ home country. Moreover, the *siwokiak and Sute two old désignations of Sogdiana in Chinese chronicles, Suyi Nafnamak the form famous in the even swyôyk; adjective *siwok d’sk, stand for an designating Sogdiana can be restored as [swy]-ôyk.n 8 ’st’m ô’r. In principle one can translate the combination as either “(title) ’st’m-ô’r” or “he had/possessed ’st’m”. Since all the other finite verbs in this inscription are preterite forms, I prefer the former interprétation. So far ’st’m has been encountered only once in M 410, 16.19 But unfortunately, it lacks a context there and the editor of the text left the passage untranslated. If one compares the word with Skt. sthâma “power” (cf. also Khotanese stâma “exertion”), ’st’m-ô’r “power holder” would be a word denoting a nobleman or landlord, hence my translation “grandee”. Professor SimsWilliams wonders whether ’st’m is an error for *’wst’m, which could be a cognate of “estate”, a loanword from Middle Persian ’wst’m\ * ’wt’m-ô’r would Bactrian be translated as “landowner” and the translation also fits the context. wyrk’k. This is Shi Jun’s name and is most likely to be derived from wyrk “wolf’. 9 wn’wk. Many names containing an element wnw “victorious” are known20 and this is a name consisting of wnw- and a common suffix -(’)k. rstßntk. A typical Sogdian personal name consisting of an element ßntk “slave”. The first member is rst “(the goddess) Truth”.21 16

Cf. Gershevitch, 1954, (Grammar of Manichaean Sogdian, hereafter GMS) 1068. Henning, 1948, p. 609-610. 18 attested in the Hou Hanshu Yoshida, 1993, p. 152 and Weber, 1999, plate 180. The older form suyi is interesting in that yi (< Middle Chinese *jak < Archaic Chinese *diak) transcribes a syllable [8ik] of swyäyk. On the problem of Archaic Chinese initial *d- see also note 6 above. 19 Sundermann, 1981, line 2354. 20 Sims-Williams, 1992, p. 75. 21 Sims-Williams, 1992, p. 67 where rstö’yh, female counterpart of rstßntk, is referred to.

17

Yoshida Yutaka

The Sogdian version of the new Xi’an inscription

65

10 kty”ßr. This word is attested three times in this inscription. The context makes it aimost certain that it is a word referring to Wirkak’s wife. In my opinion the word is the female counterpart of kty’ßs’yws which is attested in Mug documents. Unfbrtunately, the contexts in the Mug documents (B1 and B17)22 are too obscure to find out the exact meaning. In contrast with the pair wyr- and wöw-, kty’ßs’yws and kty ' ’ßr seem to denote the male and female représentatives of a household, i.e. patriarch and matriarch. If this assumption is correct, kty ’ ’ßr is a synonym of x’n’kh p’txs’wnh yvôwh of a Mug document, Nov. 3, recto 10-1 1?3 10-11 synpyn z’t(c)h. synpyn (or syzpyn) is squeezed at the end of a line and is hard to read. But it must be the same word as that encountered in line 12, where the reading is fairly clear. According to the Chinese version, Wikak’s wife bears a sumame Kang which points to her Samarqand origin. If one takes ksy’n’k kwtr’k “from a family of the originating in Kish” into considération, corresponding to the sumame Shi Chinese version, a Sogdian expression meaning “(a feminine member of) a family originating from Samarqand”, i.e. *sm’rknöc kwtr’nch is to be expected. However, synpyn sounds more like a Chinese place name and I propose to identify it with Xiping (Middle Chinese *siei b’jwmg), although it is not easy to see why we hâve a nasal element in the first syllable. Xiping was an important city located to the Southwest of Kachan. For its history see Professor Sun Fuxi’s article in this volume. On synpyn z’tch cf. also twrkstny z’tcwh “bom in Turkestan” attested in the contract of 639 C.E. discovered in Astana, Turfan.24 11 wy’wsyh. This name of Witkak’s wife is derived from wy’ws “dawn”. wy’ws has been encountered in the Mug documents as the name of a man.25 1 should like to take this opportunity to mention that the above mentioned Shewupantuo’s great grandfather’s name Miaoni (Middle Chinese *miäu ni) is the male counterpart of mywn", the name of the woman who sent Ancient Letters I and HL 13-14 [x](r)yws’k myöy. At first, I took the combination for a désignation of a day in the Sogdian Zoroastrian calendar. However, no such day name has been known and one would also expect a West Iranian word rwc “day” to follow it. What has remained is aimost certainly to be restored as [x](r)yws’k “hare, rabbit”. In the traditional Chinese calendar each day was referred to also by the animal cycle. On the dating of the day, month, and year see my commentary above. 14 pr’yßtw (ô’)[rt], wôwh pryp is known to be a compound verb meaning “take a wife, get married”.26 1 take kty”ßr pryp to be a synonymous combination. 15 myô’: It is a word meaning “here” and derived from *imadâ.21 Professor SimsWilliams draws my attention to a doublet (’)mô’ attested in Ancient Letter I, line 4 and VII, line 2. 16-17, 19 ßyrw’n ’krty, ßyrw 'n ’krth. The context in this inscription makes it clear that ßyrw’n “lit. god-souled” is an euphonie expression meaning “dead”. This meaning

22Livsic, 1962, p. 177. 23

Livsic, 1962, p. 30. Published by Yoshida, Moriyasu, et al., 1988, p. 1-50, esp. p. 14-15. 25 Livsic, 1962, p. 29. 26 Livsic, 1962, p. 26. 21 GMS 136. 24

66

Les Sogdiens en Chine

suits all the contexts of this word known so far.28 It is to ne noted in passing that in this inscription the word ßy- is also employed in the context of death; the word for the house-shaped sarcophagus is called ßykt’k (line 29) which is again a compound consisting of ßy- “god” and kt’k “house, room”. In this

connection cf. also the Bugut Inscription, where the Qaghan's death is described with the following euphonie expressions: k’w ßyy s’r pwrsty (Bugut, B-2: 1) k’w ßyyst s’r pwrst (Bugut, B-2: 7) Although Livsic proposed to translate the expression as “asked the god/gods”29, the verb for “ask” has never been spelled pwrs- and pwrst(y) is to be understood as 3 sg. preterite of (’)pwrt “tum back”. An imperfect form of the same combination k’w ßyw s’r p’w’rtnt is found in the so far unpublished Mogolküre Inscription. On the inscription see Lin Meicun’s paper in this volume. A similar phrase is encountered in Ancient Letter II, 52: ’NWZK ßy’ ’pn’wt “he departed to the gods.” 18 ’ßtsytsytyh. It is a spelling error for ’ßtsytyh “on the seventh day”. At first I and assumed that it was an error for *’ßttssytyh “on the 17th day”; this form with error. to make an stonemason led a it that appeared so awkward -ss- would hâve However, since the day in question is described as a day of the rabbit, which corresponds to the seventh day of the sixth month of that particular year, it is most likely to be an error for *’ßtsytyh. On this point see above for my comments on the dates found in this inscription. 19 krth. The 3rd singulär feminine preterite form of verb a ßw-/’krt- “to become”. Special preterite forms for the feminine 3 sg. of the light stems are not common but hâve been encountered several times.30 19-20 pr ’yô pcßytk srô. As it stands, pcßytk is likely to be the past participle of *pcßxs-. pty-ßxs’ attested in Nov. 3, 6-731 is an imperfect form of a doublet ptßxs- “to give away, to give (someone as a wife)”, which may be a dialectal form of *pcßxs-. In the inscription pr ’yô pcßytk srô “in this delivered/given year” refers to the same year as her husband died, i.e. the year of the pig = 579 CE. 20 ’yô m’xw ’yô myôy. In this inscription ’yô seems to be used in a special sense “(in/on) this same”. 21 ’kytn. -n is a hypothetical particle recently discovered by Sims-Williams.32 21 ”z’y. It is either a 3rd sg. injunctive or optative of a verb ”z’y “to be bom”. If optative, it is an error for *”z’y’y. 21 ’myôry. It is a spelling variant of myôr- “death”. The combination of ’myôry and ’prt’k “guilty, indebted, owing” in this context would be understood as meaning “destined to die”. 21 L’ ”y. The reading of is not certain but seems possible. It can also be read as 28

Sundermann, 1984, p. 306. Livsic in Kljastomyj, Livsic, 1972, p. 88-89. 30 GMS 861A. On the neuter form ending with -w attested in Ancient Letter V see Sims-Williams, in Grenet, Sims-Williams, de la Vaissière, 2001, p. 96. More instances are found in Mug documents: xwrtw ,ßyrtw “it was obtained”, and Sßv^rtw ~ öß’rtw “it was given”. “it was 31 Livsic, 1962 ,p. 29. 32 Sims-Williams, 2001, p. 180. 29

Yoshida Yutaka

The Sogdian version of the new Xi’an inscription

67

as a 3rd which however does not make sense in the present context. I take Dhy. 211.” in copula, which is attested the of Optative sg. 22-28. This part of the inscription is the most difficult to understand. My translation is based on the assumption that the author of the epitaph wanted to place special emphasis on the fact that a man lived until the age of 86, and that husband and wife died on the same date of two consecutive months in one and the same year and were buried on one and the same day, i.e. they lived their whole life together. In this connection attention should also be drawn to the reliefs of the sarcophagus, where the couple are depicted in several scenes. For example on the eastem wall they are reproduced three times: first as they are Crossing the cinwad bridge, secondly as they are offered anaosa “ambrosia” (by Daëna?) and are drinking from cups, and thirdly as 34 they proceed towards the paradise (’xwstm xw) riding winged horses.33 22 sk’wrö ZK... cnn ’prth. The collocation of sk’v^rô with a prepositional phrase cnn plus the past infinitive is also encountered in Vessantara Jataka (hereafter VJ) 258-259": sk’wrö ZKh mrtxm’y ’ z’wn cnn ßyrt “it is difficult to get the existence of a human being, i.e. it is difficult to be bom as a human being”. zmnh ’nw(y)wth “period of time”, y does not look like any letter, but the same combination is attested in line 27 and there is little doubt as to this reading, ’nwywth or ’nywth seems to be the same word as ’nywty found in Mug V4, 10, where it means “period”.36 The past infinitive ’prth has not been known but the general meaning of the sentence appears to be certain: it is difficult to live out one’s natural life span. Therefore, I follow Professor Sims-Williams 's suggestion that ’prth is the past infinitive of a verb pr- “to fill”.37 According to the Chinese version Wirkak died when he was 86 years old and he can safely be said to have lived out his natural life span. In those days it must have been a very rare occurance and was worth mentioning in his Zyorn’y,

epitaph. ’k is a nom. sg. neuter of the demonstrative xwn 23 sk’wrtr. Not knowing an adjective *sk’wr, I assume it to be a comparative form of sk’wrö “difficult”, and that the author claimed that the following situation is even rarer than what is mentioned in the preceding lines. 24 ’kyt. As it stands it is a relative pronoun ’ky followed by a complementizer ’ty. However, the context leads one to take it as a subordinate conjunction meaning “that”, which was represented by kt in ordinary Sogdian. I cannot décidé whether ’kyt here is a relative pronoun employed as a conjunction or whether it is a simple error for ’PZY 33

Dhyana, ed. in MacKenzie, 1976. On the interprétation of the scenes see Yang Junkai, 2003, and Grenet, Riboud, Yang Junkai, 2004, p. 273-284. On the interprétation of the second scene Grenet et al. differ from mine in that they do not think the couple are drinking from the vessels they are holding. It is true that in the Zoroastrian tradition anaosa is offered by Ahura Mazda much later when a deceased soul goes up to the heaven where the supreme god is resident. However, in a Manichaean Sogdian text edited by Reck a drink is mentioned among those objects which a deceased soul will receive from a virgin goddess, cf. Reck, 2003, esp. p. 332-333. 35 Benveniste ed., 1946. 36 Livsic, 1962, p. 60. 37 See Sims-Williams, Hamilton, 1990, p. 26-27. 38 Sundermann, 1985, p. 42 s.v. ’www. 34

68

Les Sogdiens en Chine

“that”. It is also tme that ZKZY= ’kyty is sometimes used instead of (’P)ZYt ’yw xwnx ZKZY styw ZKw wrzrw w’xs w’ßt rty sy nyô’y L’ pyrt “One (rétribution) is that although

he speaks the truth, nobody believes him” (Sutra of Causes and Effects 455-6).39 One may compare this fonction of ’kyt(y) with Bactrian aKiôo.m mrt(x)m’k ô’my. mrtxm’k ô’mh is attested in VJ 1007 where it is contrasted with ßyyst’n “heaven, place where gods résidé”. Here it is opposed to ’xwstm’xw “paradise” of line 26. 24-25 ’yô srô ... ’yô myô. On the meaning of ’yô in this inscription see above. 25 prwh L ’ ptz’n may be compared with prw nw yrß’w “without laiowing, unconsciously” (VJ 381). If the comparison is valid the expression should be rendered as “without recognizing” and I assume that in the present context it means “without intention, by accident”. 26 pr”w wy-n’nt. Readings of the two words are not entirely certain. My translation “see each other” is based on the assumption that pr”w is a variant form of pr’yw. However, since the context is not absolutely clear to me, my translation remains hypothetical. Moreover, one finds the word meaning “together” speit in an ordinary way in the next line: pr’yw. 26 ’xwstm’xw. The later form of the word meaning “paradise” (cf. Av. vahistam ahüm) is wstm’x, etc. without the first -x-41, which seems to hâve been lost by dissimilation. See also ’rt’wxwst or ’rtxwst, name of the third day in the Sogdian wahista. calendar originating from Old Iranian 29 snkyn’k ßykt’k “tomb (lit. house of gods) made of stone”. In the present context the expression dénotés a house-shaped sarcophagus placed in the tomb. 30 ßr’ysmnßntk. The name of Wirkak’s fïrst son. ßr’ysmn is a Sogdian form of Skt. Vaisravana. One may be surprised to find out a Buddhist element in this otherwise totally Zoroastrian context of Shi Jun’s funerary construction which dates back to the latter half of the 6th Century. On the strength of Sogdian personal names transcribed in Chinese characters, I once argued that the Sogdians in China began to convert to Buddhism only in the seventh Century.42 Possibly Vaisravana had become an accepted naturalized syncretic deity in Sogdiana by the sixth Century due to the Bactrian influence; recently a Bactrian name ßpr]opavoKipöo has been found on a pot unearthed in Dal’verjin-tepe.43 30 örymtßntk. It is a populär name both in Sogdian script and in Chinese characters, cf. Sims-Williams, 1992, p. 49-50. 31 pr’wtßntk. pr’wt is attested on a wood inscription studied by Livsic, where he

39

Ed. in MacKenzie, 1970. Sims-Williams, 2001, p. 198. For an extensive study of the development of the conjunctions meaning “that” in Sogdian and Bactrian see I. Yakubovitch’s forthcoming article entitled “The syntactic évolution of Aramaic ZY in Sogdian”. 41 Cf. GW 113. 42 Cf. Yoshida, 1998, esp. 40-41. * Reading of N. Sims-Williams apnd Yamauchi, 1999, p. 161-162. On the appearance of Vaisravana in a wall painting discovered from Penjikent see Grenet, 1995/96. 40

Yoshida Yutaka

The Sogdian version of the ne^ Xi'an inscription

69

takes it as a place name.44 Here it serves as the first member of the -ßntk name and is likely to be a name of a deity, which has not been known. However, prwtßntk is attested among the Upper Indus inscriptions and may be compared with our name.45 31 ws '% has not been attested in Sogdian. I assume that it is a spelling variant of i.e. Old Iranian “on account of, fbr”. However, the etymology of very unlikely. assumption this makes Reading of this word is fairly hypothetical. Except fbr the first letter ' 31 (aleph), the others do not look like ordinary form of BY7ßy\ For example, the third letter looks more like ' (aleph). As a whole, the word gives the impression that it was not correctly inscribed. My reading is based on the fact that the first word of line 32 is an inflected form of m’th "mother'1 32 m'Bryh. It is the oblique form of m!th “mother55 which is also encountered in Ancient Letter III3 22. In Sogdian an ordinary word for “parents” is m'tptry (ßfc~stem) and when the two members are combined by a conjunction 'W “and”, m'th “mother” always précédés ’ptr- “fether”; this word order is a legacy of Old Iranian or even Proto [ndo-Iranian. If my reading and translation of the two words are correct, this order appears to reflect that of the Chinese counterpart: fu mu 5^® “father (and) mother”. 32 sc'w. Reading of c is not entirely certain. If it is an adjective modifying the next word wy,fkh "place”, it may be compared with Middle Persian sazäg “fitting” (< Old Iranian ^sacäka\ cf. also Sims-Williams ’s etymology of ptß "yscn "worthy of honour”.46 Bibliography

Benveniste, É.,

1946 Chen Yuan 1956

Vessantara Jataka, Paris. Ershishi shuorunbiao Official Historiés], Beijing.

[Table of Lunar Months for Twenty

Gershevitch, I. 1954 / Grammar of Manichean Sogdian (= G MS). Oxford: Basil Blackwell, Publications of the Philological Society. Grenet, F. 1995/6 “Vaisravana in Sogdiana About the orig ins of Bishamon-ten ”, in: Silk Road Art and Archaeology 4, p. 277-297. 2000 “Trois documents sogdiens d’Afrasiab”, in: Materialy mezdunarodnoj konferencii, Moscow, p. 196-202. Grenet, F., Sims-Williams, N., de la Vaissière, É., 2001 « The Sogdian Ancient Letter V », dans Alexander’s legacy in the East : Studies in Honor ofPaul Bernard, Bulletin of the Asia Institute, 12, p. 91-104.

—

44

——

Cf. Livsic, 1987, p. 53-62, esp. p. 55. On this inscription see also Grenet, 2000, where he follows Livsic.

Incidentally, pr’wt in Nov. 6, 9 has been proved to be a misreading of py’wt, cf. Sims-Williams, 1992, pp. 41-42.

45 46

On ’prwtßntk see Sims-Williams, 1992, p. 41-42. Sims-Williams, Halén, 1980, p. 7.

Les Sogdiens en Chine

70

GRENET5 E RIBOUD5「づ and YANG づ

2004

“Zoroastrian scenes on a newly discovered Sogdian tomb in Xi’an, Northern China' in: Studia Iranica 33, p. 273-284.

Henning, W. B.

1938 1948 Humbach, H. 1988

"Argi and the ‘Tolcharian’“, Bulletin of the School of Oriental Studies IX, p. 545-71. "The date of the Sogdian Ancient Letters,5. Bulletin of the School of Oriental and African Studies, 12/3, p. 601-15. "Die pahlavi-chinesische Bilingue von Xi'an”, in: J. Duchesne-Guillemin et al. (eds.), Barg-i sabz. / green leaf: Papers in honour of Professor Jes P. Asmussen, Acta Iranica 28, p. 73™82.

IKEDA 011 池田温 « Hasseilei chüyô ni okeru Tonkö no Sogudo jin shüraku, 8世&己中葉に お ける 1965 敦煌のソ グ ド 人聚落», [A Sogdian colony at Dunhuang in the mid-eighth Century], Yürashia bunka kenkyü, ユ ー ラシ ア文化 研究, [Studics on the Eurasian culture], n° 1, p. 49-92. Karlgren, B. Grammata Serica Recensa, Stockholm. 1957 Kljastornyi S.G., Livsic, VA "The Sogdian Inscription of Bugut Revised", Acta Orientalia Academiae 1972 Scientarum Hungaricae, T. XXVI-1, pp. 69-102. し1V810,VA. Sogdijskie dokumenty s gory Mug, [Sogdian Documents from Mount Mugh] II, 1962 1987

し00 Feng 羅豐 1996

Moscow. “Sogdijskij dokument iz drevnego Samarkanda", [A Sogdian document from Ancient Samarkand] in: Central’naja Azija, Moscow, p. 53-62.

[A Sui-Tang cemetery in Guyuan Nanjiao Sui Tang mudi chubanshe. Wenwu Beijing: Guyuan], of the Southern suburbs MacKenzie, D. N. The tlSutra of the causes and effects of actions " in Sogdian, London. 1970 The Buddhist Sogdian texts of the British Library, Acta Iranica 1 0, Leiden. 1976 Miho Museum Bulletin of the Miho Museum, IV. 2004 Reck Ch. 2003 "Die Beschreibung der Daenä in einem soghdischen manichäischen Text”, in: M. Maggi, E. Provasi (eds.), Religions thèmes and texts of pre-Islamic Iran and Central Asia, Wiesbaden, p. 323-338. N. Sims-Williams, "A Sogdian greetmg,\ R. E. Emmerick & D. Weber (ed.), Corolla Iranica. 1991 Papers in honour of Prof Dr. David Neil MacKenzie. Frankfurt am Main: Peter Lang, p. 176-87. Sogdian and other Iranian inscriptions of the Upper Indus, II (= UP 口), 1992 London. apudK. Y amauchi in: Bulletin of the Ancient Orient Museum 20, p. 161-162. 1999 Bactrian documents from Northern Afghanistan I, Oxford. 2001 Sims-Williams, N.5 Halén, H. The Middle Iranian fragments in Sogdian script from the Mannerheim 1 980 collection (Studia Orientalia, LI/13), Helsinki.

Yoshida Yutaka

The Sogdian version of the new Xi'an inscription

71

Sims- Williams, N., Hamilton, J. Documents turco-sogdiens d^u I戈-资 siècle de Touen-houang, London. 1990 Sundermann, W. Mitteliranische manichäische Texte kirchengeschichtlichen Inhalts, Berliner 1981 Turfantexte XI, Berlin. 1984 "Probleme der Interpretation manichäisch-soghdischer Briefe », in: J. Harmatta (ed.), From Hecataeus to Al-Huwârizmï, Budapest, (Collection of the Sources for the History of Pre-Islamic Central Asia, I, III), Budapest : Akadémiai Kiado, p. 28%316. manichäisch-soghdisches Parabelbuch, Berliner Turfantexte, XV 5 Berlin. Ein 1985 WENSHU TULUFAN CHUTU Tang Zhangru (ed) Tulufan chutu y^enshu, voL 3* Beijing. 1981 Weber, D. Iranian Manichaean Turfan texts in publications since 1934. Photo édition, 1999 London.

XI'AN SHI WENWU BAOHU KAOGU SUO "Xi'an Beizhou Liangzhou sabao Shi Jun mu fajue jianbao 西安北周涙 州薩 2005 保史君 墓發掘間幸長'' [Brief Report on the excavations in Xi an of the tomb of the Northern Zhou Lord Shi, sabao of Liangzhou], Wenwu, 3, p. 4-33 Yakubovitch, I. fbrth. "The syntactic évolution of Aramaic ZY in Sogdian”, forthcoming. J11111点 YANG 楊軍凱 "Beizhou Shijun mu shi guo dongbi fudiao tuxiang chutan” ゴ 匕周史君墓石椁B 2003 東壁 浮雕圖像初探 (Images from the Eastem Wall from the Stone Coffin of the Tomb of Shi Jun in Northern Zhou period), in: Yishu shiyanjiu (The study of art history) 5, p. 189-198. "Ru Hua Sute juluo shouling muzang de xin faxian — Beizhou Liangzhou 2004 Sabao Shi Jun mu shiguo tuxiang chushi” 入華粟特聚落首領墓葬的新發現 北周涼州薩保史君墓石槨圖像初釋 (Newly discovered burials of Sogdian community leaders in China: A preliminary decoding of the illustrations on the stone sarcophagus of Sabao Shi of Liangzhou dating from the Northern Zhou Dynasty) in: From Samarkand to Chang 'an: Cultural traces of the Sogdians in China, Beijing, p. 17-26, 59-65. Yoshida Yutaka 吉田豊 1993 "Sogudogo no Näfhämak tkokumeihyö, no 2-3 no yomi nitsuite” 「ソク” 「百吾 「 国名表 」 の NWfhämak の 二 •三 の読み について」[Some new readings of the Näfnämak in Sogdian] in: Oriento 『オリエント』(Bulletin of the Society for Near Eastern Studies in Japan) 36/1, pp. 151-153. and a paixiao 排 (a type of bamboo flute). The other instruments are not clearly identifiable (fig. 6). Some musicians on the octagonal gilded silver cup (fig. 2) are holding clappers (paiban 拍板), small cymbals (xiao nao 小 饒), a dong xiao 洞 (another type of bamboo flute), and a crooked-necked lute (qujing pipa 曲頸琵琶). Another figure holds a pot on one arm and a vessel in hand and two others dancers move their hands about. The octagonal golden cup is adomed with a pattern of eight standing human figures (fig. 3). Although the above mentioned instruments were populär during the Tang dynasty, they were originally from the western régions? Interestingly, many of the figures hâve deep set eyes and high bridged noses, wear clothes with tumed-up collars, and sport waleng 瓦 楞巾冒 caps or pointy bats with curled cdges. Figures with such features and dress do not appear in other Tang depictions of music and dance。

These cups portray Sogdian cultural characteristics, but many details indicate they were produced in China. First, the bodies of these three octa¬ gonal cups were cast in a bulky shape, which was not typical of Sogdian pièces.3 4 The casting technique created the effect of high relief, which was also rare among Sogdian pièces. Second, the figures on the cups in fîgs. 2 and 3 are set against a background of honeysuckle foliage,

Fig. 7: Musicians from the octagonal gilded cup (detail from fig. 2).

"Tinyue zhi" 音 志 (Treatise on Music), in the Suishu 随 (Records of the Sui dynasty) (Zhonghua shuju, 1913\juan 15, p. 378. 4 Part of the mouth of the octagonal golden cup adorned with a pattem of human figures is lost. From the nist on the frame, we can see that the piece was made of copper or a copper alloy and then gilt. See Shaanxi sheng bowuguan, et al., 2003, p. 74.

3

Les Sogdiens en Chine

110

mountain rocks, flying birds, butterflies, and fish scale pattem (fig. 7), which are ail characteristic of Tang handicrafts and are seldom seen in Sogdian artefacts. Tang people found these types of loop-handled, octagonal, gold and silver cups stränge. Such non-traditional forms did not become fashionable. The dimensions off these three octagonal cups are very small; they did not replace indigenous Chinese cups for drinking tea or wine. The handles of these Sogdian cups hâve a practical and clever design. When holding the cup, the thumb presses on the finger pad, creating a comfortable place for the hand to rest, which helps bear the weight and increases stability. However, the diameter of the looped-handle is very smail, so that most people are unable to inset their finger into it. Consequently, the cup has little practical value. But this faulty design is deliberate since it appears that the cups were produced as art objects, whose bulky shape gave them stability when on display. So far, no octagonal cups from after the 8th Century hâve been discovered, nor hâve large quantities of ceramics or replie as been unearthed. It is clear that although octago¬ nal cups became, for a time, Tang dynasty Utensils, and even exerted a cultural influ¬ ence, they were abandoned because they did not mesh well with the tastes, customs, and daily life of Tang people. However, the polyhedral handled cup did not entirely vanish. In 1992, a silver gilt, golden polyhedral cup was unearthed from the Liao dynasty tomb of Yelü Yuzhi HIW of Alukeerqin H décorative string of Banner in Chifeng beads frames all the edges of the piece. Unlike Tang pièces, it is heptagonal in shape. It has a finger pad in the shape of an oval flower on the edge of the mouth but no handle, only a fingerplate below the finger pad. Cups without handles are not found among Sogdian or Tang artefacts. Various figures are engraved on the belly of the cup seem to be Daoist priests or the Seven Worthies of the Bamboo Grove, both of which are likely Khitan innovations. In 2003, a loop-handled cup was discovered in a tomb from the Liao dynasty in Tueijishan This discovery confirmed that these types of Utensils were prévalent in the northem steppe for quite a while.

Losses through Innovation

Fig. 8: Sogdian cylindrical cup with a handle.

5

6

The process of innovation, commonly assimi¬ lâtes other cultures. Of ail Sogdian polygonal Uten¬ sils, only the Sogdian cylindrical cup was easily accepted by Tang people (fig. 8). When it became a major object of imitation, the design of the Sog¬ dian cup assimilated Tang concepts of design. The cup thereby embodies the trajectory of the évolu¬ tion of foreign cultures in China. An outstanding example is the golden cup with patterns of flower wreaths made of golden threads

Nei Menggu kaogu yanjiusuo et al., 1996, pp. 4-35. The material about Tuegishan Tomb hâve not been published.

Qi Dongfang

Hejiacun Treasure and Sogdian culture

111

(No. 71:100, fig. 9). This innovation, based on the Sogdian cylindrical cup, can bee Seen in two details. First, the finger pad above the handle is made in the shape of a leaf bud, whereas the fingerplate below the handle is omitted. These alterations changed the intrinsic style of the Sogdian piece, while echoing Han and Jin pièces. This minor alteration did not change the fonction of the object, it merely matched the tastes of Fig. 9: Golden cup with flower wreaths. Tang people. Second, gold and pearls were soldered on the smooth belly of the piece. Highly omamented, soldered clouds and flower wreaths emerged as raised, three- dimensional designs. Objects were also inlaid into the flower wreath design. An inventory of objects from the Faming Temple in Shaanxi mentions “an authentic golden coffin omamented with gems framed in gold threads and authentic pearls” (fig. 10, see colour plate No. 16). This golden box inlaid with gems, fine gold, and pearls, and decorated with a border of spun gold—was produced by the same techniques as the golden cup with pattems of flower wreaths (fig. 9).7 The An Lushan shiji (Traces of An Lushan) describes “baskets woven with spun silver,” “wicker-like baskets woven with silver,” and “vases with golden flowers and Strands of spun silver” that were included in These by Xuanzong the gold and silver gifts bestowed upon An Lushan objects were probably also produced by the same techniques as those discussed above. The flower shapes in the wreath design of these pièces are facing downwards, so one cannot identify which type of plant it supposedly reproduces. But the design is similar to a pattem on a gold coffin that was discovered at the same time as the stone reliquary from 694 in Jingchuan, Gansu.9 The design is also very similar to that of the tomb in Xi’an, which simple cylindrical silver cup unearthed from Yao Wupo’s was sealed in 697. Therefore it is likely that, this cup was produced during the last half of the 7th Century. Although Tang people liked handicrafts that used copper wire, solder, and inlays, pièces produced with these techniques may hâve been exquisitely wrought, but they were not necessarily used in daily life. Outside of religious ceremonies, these magnificent pièces were inconvénient to use and difficult to clean. Only one piece of this kind has been discovered to date. Although its beauty is unsurpassed, it was not a populär item.

—

7

8

Famen si bowuguan, 1994. Yao Runeng, 1983J. A, p. 7, 10. Gansu wenwu gongzuodui, 1966, pp. 8-14, 47.

Les Sogdiens en Chine

112

The gilt silver bowl with sea monster and wave pattern (No. 71: 216, fîg. 11, see colour plate No. 16) was also a novel piece. On the sides of the piece, fourteen curved lines were forged from the mouth to the base, an embellishment that deviated from Chinese tradition. The British Museum has a silver bowl with a wave pattern (fig. 12). It is probably Indian from the 4th or 5th Century or ir may corne from the eastem part of Iran.10 A similar silver bowl was unearthed from the tomb of Li Xizong’s wife in Zanhuang Hebei, which was sealed during the Eastem Wei dynasty (534-550).11 The bodies of Sogdian silver bowls from the 6* to the 7th Century, Fig. 12: Silver bowl with wave were usually forged into concave and convex pétals. designs. The gilt silver bowl with sea monster and wave pattern was probably influenced by this style of ornamentation. It displays rich embellishment: birds and beasts, flowers and plants, and hills and rocks were altemately engraved on the wave pattern. These embellishments also include deer, sheep, foxes, rabbits, and hoopoes, ail of which were considered auspicious in Chinese tradition. A gilt sea monster, flanked by mandarin ducks, adoms the inner base of the bowl, a Symbol of cultural communication between East and West. Just as the polyhedral cup pièces with a wave pattern disappeared after the Tang dynasty, a similar silver piece did appear again in the Liao dynasty, but its outer surface had become '

'

smooth.12

Included in the Hejiacun Treasure are a peach-shaped silver plate with a gilt tortoise designs, a sunflower-shaped silver plate with a bear design,

Fig. 13, 14, 15 : Silver plates. 13. Sunflower-shaped silver plate with a feilian design. 14. Sunflower-shaped plate with a phoenix design. 15. Silver plate in the shape of two peaches with a fox design. 10

Qi Dongfang, 1999, p. 67-68, 258-259, 310. Sun Ji, 1996, p. 146, 152.

Qi Dongfang

Hejiacun Treasure and Sogdian culture

113

a sunflower-shaped plate with a phoenix design, and a silver plate in the shape of two peaches with a fox design (No. 71: 42, 43, 44, 45,46, figss 13, 14, 15). The contours of these pièces are not round, the centers of the plates are adomed with a single-animal design, crafted to look like a relief, and the outer surfaces of these pièces are blank. The impressions on the back of the plates are relatively deep. The contours that form muscles and wings of the animais are distinct and clear. Single-animal designs were rare m traditional Chinese ornamentation, but were populär in the Sasanian and Sogdian régions. This simple style is contrary to the elaborate ornamentation favored during the Tang dynasty. Unlike Tang silver plates, the sides of these plates were thick5 the edges narrow3 and the bellies shallow. Some of the plates are shaped like a six-petal sunflower or a pe^ch? which contravenes the Sogdian preference for round silver plates, Realistic animal designs were characteristic of Sogdian art, but feilian, phoenixes, and tortoises were Chinese Symbols. These pièces were new créations during a time when the Chinese were drawing inspiration from western art. However, unified forms and high protrusions in the surfoce of the métal did not suit most Tang people. Therefore these styles did not become populär. Most scholars date this group of silver plates to the high Tang. It is diffïcult to détermine where the agate cup with an animal head (N° 71: 282. Fig. 16, see colour plate No. 16, 17) was produced, but it was certainly not produced in China.13 The Fig. 17: Agate cup with a gilded ^^31 raw material of the agate cup is supple head. and moist; red and white combine to give it its color. The bottom of the cup is engraved in the shape of an antelope's head, while the antelope's mouth forms the spout. The cup is capped with a décorative gold cover. The form and fonction of the piece suggests that it belongs to a dass of Utensils populär in the west called the "rhyton.” The rhyton first appeared in Assyria and in the Achaemenid dynasty in Persia; it became widely used prior to the Islamic era. There are two other elliptical agate cups in the Hejiacun Treasure (No. 71: 281, 282). They were delicately engraved and possess a smooth luster. Perhaps they were produced at about the same time. With the exception of the 4tear cup" (erbei 耳杯) of the Han and Jin dynasties, elliptical Utensils were not widely used in China, but they did enjoy a long history in central and western Asia. They were made of gold, silver, bronze, or a copper-nickel alloy.14 References to rhytons can occasionally be fbund in Shijiazhuang diqu geweihui wenhua ju wenwu fajue zu, 1977. Zhu Tianshu, 1998. See Shaanxi sheng bowuguan et alii, 2003. See also Sun Ji, 1991 . 4 Many wall paintings depicting revellers at a banquet holding this sort of cup have been found at the sites in Samarkand and Panjikent. See Azarpay, 1981. A long, poly-lobed silver cup that seems to be Sogdian was also excavated in Lijia yinzi 李家 子, in Aohan Banner 敖 旗, Inner Mongolia. See "Lijiayingzi chutu de yinqi yu sichou zhi lu shang de suteren,^ 李家 子 出土 的 器与 之路上 的 粟特人 (Silver Wares Excavated from Lijia yinzi and Sogdians of the Silk Road), in Qi Dongfang, 1999, pp, 321133 2, 12

Les Sogdiens en Chine

114

stone inscriptions from the Northern Dynasties. During the Sui and Tang, craftsmen were imitating the pièces that were entering from the west. Although the cup retained its overall hom shape and the animal head on the base, it lost its spout and the foot of the animal became a handle. These cups became impractical and were relegated to rôle of décorative pièces and funerary objects. Transformation and Amalgamation

Although transformation and amalgamation hâve a strong utilitarian core, they actively accept foreign cultures and create new knowledge while drawing from them. Tang dynasty inlays improved dramatically under the influence of Sogdian and Sasanian techniques. Thereafter; these new techniques were widely employed, and it became possible for the Chinese to pursue many different modes of Utensil production and décorative designs. For example, lotus-petal designs became populär in Northern Dynasties ceramics for a while. Due to a limited range of raw materials, only coarse lotus-petal designs could be engraved and moulded. However, forging and inlay techniques that worked on supple gold and silver supports, also enabled craftsmen to create three-dimensional reliefs on ceramics. These techniques also increased detail, which allowed a subtle union of the rugged and the délicate to emerge.

Fig. 18, 19: Eight-petal silver cup with gilt designs of maidens and hunters.

The eight-petal silver cup with a gilt design of maidens and hunters (No. 71: 213) is a highly acclaimed piece (figs. 18, 19). Its flower-shaped mould and lively scenes display a mastery of composition. The cup is embellished with four scenes of men hunting and four scenes of women at leisure. The kind of composition found here and the images of the figures and dress are also found in Tang wall paintings and other art forms. This suggests that this piece was produced during the Tang dynasty. Nonetheless, the piece still bears the traces of Sogdian silverwork. The body of the cup has eight petals instead of eight arris. A ring of beads is inlaid around the mouth and the base. A handle connects the edge of the mouth with the body. The outside of the handle is decorated with a string of beads. There is a finger pad above the handle and a fingerplate below it. Furthermore, a capricom is engraved in the center of the base. The

Qi Dongfang

Hejiacun Treasure and Sogdian culture

115

capricorn is a beast from Indian mythology and often appears in painting and sculpture fïom ancient India, westem Asia, and central Asia.15Thus remuants of a foreign origin are still visible in the dominant Tang style of the piece. Interestingly, the Sogdian characteristics of this piece were altered. On the surface of the cup, concave indentations in the shape of eight petals were replaced by raised edges. Around the edges, strings of beads were replaced by willow leaf patterns. A raised, eight-petal single lotus pattem was forged below the belly of the piece. The finger pad is triangular

Fig. 20, 21: Eight-petal silver cup with gilt designs of maidens and hunters (details).

and a deer and a simple flower is carved above it. These design alterations reflect the tastes of people in the Tang and embody the way they adopted or rejected foreign cultures. The eight-petalled silver cup with gilt designs of maidens and hunters deviates from the characteristic compositions of Sogdian pièces, yet it is not divorced from Sogdian culture. It is an interesting adaptation. The exquisite artistry of Sogdian cups permeates the new piece. The four hunting scenes and the four leisure scenes altemate on the outer body of the cup, reminiscent of the way scenes on a Tang screen painting unfbld one after another (figs. 20, 21). The base inside the cup is adomed with a ripple design.

l5CenRui, 1983.

Les Sogdiens en Chine

116

When the cup is filled with water or wine it seems as though fish were swimming and reeds floating in a rippled pond. This adds some pleasure to an ordinary, humdrum life. In each scene there is a definite design and shading scheme, the composition is clear, and the human figures , v Ihe fish are iirmly delmeated. . . scale designs in the tour hunting scenes are gilded, while the fish scale patterns in the four maiden scenes are not. This juxtaposition créâtes a brilliant contrast and a sense of movement in the images. The gilded silver reliefs, and engravings create a prosodie rhythm of movement and rest. The men in the hunt and the maidens at leisure elegantly contrast tense excitement and quiet relaxation. Not only is this piece a magnificent work of art, its clouds, beloved maidens, and rugged hunters also appealed to the Tang’s aesthetic sense. This highly original artistic innovation also meshed with Chinese notions of practicality and appréciation. In 1980, a similar piece was unearthed in Xi’an, but is base had been Fig. 22: Golden bowl with mandarin duck and lotus , , . " petal designs.

destroyed.16 As a resuit of the influence of Sogdian silverware, the Tang forms changed. The typical Sogdian style, which featured many detailed components, merged with the magnificent petal forms of Tang décorative arts. Naturally, this new form became populär. Among the Hejiacun Treasure, there are two golden bowls with mandarin duck and a lotus petal design. (No. 71: 275). They hâve the same mould and design as Sogdian bowls (fig. 22). The bowls were forged into shape and the bases display standard string-bead designs—ail of them characteristics of Sogdian silverware. However, the raised petals on the surface of the pièces are shaped Fig. 23: Sogdian silver bowl like double lotuses. The Sogdian flavor has become with a deer design (drawing). lighter. These pièces show how influences from foreign cultures permeated Tang society. Yet, traces of foreign cultures are not easily detected in pièces such as these. This sort of imitation followed by innovation and transformation, perfectly merges foreign artistic ideas with Tang aesthetic feelings. It became the dominant practice in Tang silverware production during the 801 Century and even influenced ceramic production. Fig. 24: Sogdian silver bowl with a deer design. 16

Tokyo kokuritsu hakubutsukan, 1998, p. 124-125.

Qi Dongfang

Hejiacm Treasure and Sogdian culture

117

Who wre the Craftsmen? Who are the craftsmen who produced these pièces? There are only two possibilities. Either they were Chinese trained by Sogdian craftsmen in China, or they were Sogdians who imitated Chinese craftsmen. Regardless off who was trained by whom9 they all needed the requisite skills, cultural background, and proper samples to produce these pièces. The silver bowl with a deer design (figs. 23, 24), which was unearthed in Shapocun in Xi’an, and the silver bowl with an interlocking flower design (figs. 25 see colour plate No. 16, 26), which was excavated in the western suburbs of Xi’an, are Sogdian pièces that were imported into China.17 Similar pièces with crooked petals fbrged on the sides can be seen among extant Sogdian silver pièces (fig. 27). This indicates that at the time samples did indeed exist from which copies could be made. There are three confirmed Sogdian silver pièces in the Hejiacun Treasure. Two are plain, jar-shaped silver cups with handles (No. 71: 66, 65, fig. 28). They have cloud-shaped finger pads, undemeath which are looped handles. These pièces exhibit the same style as two Sogdian pièces, the handled cup with a mountain goat design (fig. 29), now in Russia’s Hermitage Museum, and the handled cup with an interlocking flower design, unearthed from Kopeny chaatas Tomb Number 2 on the Yenisey River in Russia. Furthermore, these pièces are also similar to another Sogdian piece, the silver cup with an interlocking flower design Aohan excavated at Lijia yingzi 18 Banner, Inner Mongolia. Both plain, jar-shaped silver cups with handles are large and bulky (0.2 centimeters thick and 395 grams), and they are brown in color. These pièces are 17

Fig. 26: Silver bowl with interlocking flower design from the westem suburbs of Xi'an.

Fig. 27: Sogdian silver bowl with crooked petals.

Fig. 28: Sogdian jar-shaped silver cup with handle.

Fig. 29: Sogdian handled cup with mountain goat designs.

(Sogdian Silver Bowls Discovered in See “Zhongguo faxian de Sute yinwan,” U ^58 China), in Qi Dongfang, 1999, p. 333-344. 18 See “Lÿiayingzi chutu de yinqi yu sichou zhi lu shang de Suteren”, in Qi Dongfang, 1999, p, 321-332.

118

Les Sogdiens en Chine

somewhat different from the other Tang pièces that were buried in the same area. The alloy used may be composed of different metals. There is also a plain, oval-shaped silver bowl (No. 71: 215, fig. 30 see colour plate No. 16) that is also characteristically bulky and has the same color and luster. Large oval-shaped cups are, coincidentally, often seen in Sogdian régions. They are contemporaneous with Tang gold and silver pièces. They could hâve been Sogdian Figo 31: Cylindrical silver cup with pièces that were used as models for imitation. handle from the tomb of Yao Wupo. Direct evidence that Tang craftsmen produced silverware in the Sogdian style has been fbund» A cylindrical silver cup with handle (fig= 31) was unearthed in Xi an from the tomb of Yao Wupo5 which was sealed in 697 J' On the base of the cup the words, "Crafteman Zheng Qing 匠 卿”are engraved“ This indicates that the cup was produced by a Tang craftsman sumamed Zheng. The Hejiacun Treasure probably belonged to the national treasury, Gold and silver wares were mainly produced by official or even national workshops. Official craftsmen came from everywhere. Although there is no record as to the existence of Sogdian craftsmen in the official workshops, we know that central Asian craftsmen had come to China long before the Tang dynasty. The "Da Yuezhi zhuan55 大月氏 in the Beishi 史 (History of the Northern Dynasties) records that people of the Yuezhi arrived at the capital during the reign of Wei Tai Wudi. They claimed they could produce five-colored glass glazes, This suggests that glass was produced durinng the Northern Wei with the help of western craftsmen. During the Sui dynasty, a Sogdian named “He Chou 何 稠”100ked at many old pictures and knew a great deal about antiques. China had not produced glass for a long time; craftsmen did not dare produce it. He Chou produced a kind of glass using green porcelain, which was no different from real 5

glass.5520 Recently, tombs of three Sogdians from the Northern and Sui Dynasties——An Qie 21 安伽,Lord Shi 史君,and Yu Hong 虞弘——were discovered.19 Although the inscriptions and paintings from the tombs display rare Zoroastrian images, most of the pictures depict objects from the Sogdian's daily life. The inscriptions and paintings hâve a pronounced central Asian flavor. It seems that Sogdian craftsmen carved and painted them, but even if they did not, the craftsmen who did must have had preliminary Sketches to consult. Sogdians had been intermarrying with the Tang population for a 19

20

Xi^an shi wenwu baohu kaogusuo, 2002. SuishuJuan 68,p. 1596. An 〇生① example 1$:‘旧？ Yuanming,&〇〇 Dingzhou,was a丫©リ wealthy 111311.

11« managed three post-stations.Nea^ €3^ ^6111 he established inns whe^e merchants could stop,dealing especially with arrivmg 力〃. [1118] assets we^e enormous and in his house he had 行丫€hundred damask 。 巨万 起店停商 , 以袭胡 weavmg 11^011111©S"(定 州何明远大富 , 主官中三 , 每于 , 家有 机五百 ) Taiping guangji 太平广 (Shanghai目⑪ chubanshe),^¥1243,p. 1040-1570.II 18 presumed that He Mingyuan 何明近 was descended &0111 Sogdians. 21

Shaanxi sheng kaogusuo, 2001.

Qi Dongfang

Hejiacun Treasure and Sogdian culture

119

were passed down through the generations„ streams of Sogdians came to China. Not a endless open very and was The Tang dynasty few of them eamed their living as craftsmen, and common sense leads us to believe that long time. It is possible Sogdian techniques

of them were employed by officiais. It does not really matter whether the craftsmen were Sogdian or Tang, because they all understood Sogdian culture. They were able to blend eastem and western cultures and create a vibrant community of

Some

craftsmen.

The objects in the Hejiacun Treasure were not produced at the same time. For example, there are coins that date from the Spring and Autumn period to the Tang, and silver ingots which bear inscriptions of different dates. The Treasure’s exotic relies corne from many countries and régions. However5 nearly all the foreign Utensils date from the 7th Century or slightly later3 and represent the oldest pièces in the Hejiacun Treasure. The pièces that imitate foreign styles are still later, and the innovative pièces date from an even later Tang period. This temporal pattem is not coincidental; it reflects the step-by-step process of assimilation of fbreign cultures. When looking at these pièces, one can sense Tang people were curious about new fashions, and willing to absorb ideas from different cultures. The Hejiacun Treasure ia a concrète manifestations of social and cultural évolution. The Hejiacun Treasure also contains a crystal-lobed elliptical cup (No. 71: 283) and a white jade-lobed elliptical cup (No. 71: 283). This type of vessel probably originated in western Asia and was particularly populär during the Sasanid period, after which in the Xintang shu time it spread to central Asia. The “Kangguo zhuan” has fertile lands, and mountains, by enclosed Kang “is States that the Kingdom of produces calamine, brass, and ciystal.” The “Kangguo zhuan” in the Jiutang shu [HW 45 records that in the 6th year of the Kaiyuan reign (718), the Kingdom of Kang dispatched an envoy to offer chain armor, a crystal cup, an agate vase, and an ostrich egg. It is likely that these sorts of Utensils entered China from western and central Asia. Although the white jade-lobed elliptical cup has heavy Sogdian features, the jade texture of the cup suggests that it was produced by Chinese, who had a long tradition of cherishing jade. Sogdians, on the other hand, did not typically produce jade pièces. We cannot at the moment détermine whether the piece was an import or an imitation. Because Sogdians, Sasanians, and other foreign craftsmen worked in official workshops, Tang people could not readily distinguish the différence between westem Asian and central Asian cultures. Therefbre, it is difficult to affirm if a Tang imitation was simply and direetly influenced by a certain culture. Furthermore the characteristics of a spécifie culture are not necessarily simple.

120

Les Sogdiens en Chine

BibIiography Azarpay Guitty

1981 Cen Rui 岑蕊 1983

Sogdian Painting (Pictoral Epie in Oriental Art), Berkeley : University of California Press.

"Mojiewen kaolüe,, 摩 羯 考略 [Notes on the Capricorn Design], Wenwu, 10, p. 78-80, 85. Famen si bowuguan 法 寺 博物 Famen sむ去 寺 [Famen Temple], Shaanxi lüyou chubanshe. 1994 Gansu wenwu gongzuodui 甘肅支物工作隊 "Gansusheng jingchuanxian chutu de Tangdai sheli shihan55 甘 省 丿丨[ 出 1966 土的 唐代舍 禾！J 石函 [A Tang Stone Case Excavated in Jingchuan, Gansu Province], Wenwu, 3, pp. 8-14, 47. Marshak Boris L 1971 Sogdijskoe Serebro [Sogdian silverware], Moscow. 2001 Asia, ruta de las estepas. De Alejandro Magno a Gengis Kan, Barcelona: Fundacion "la Caixa". NE] MENGGUICAOGU YANJIUSUO内蒙古考古 研究所, CHIFENG SHI BOWUGUAN 赤峰 市 博物 , ALUKEERQIN QI WENWU GUANLISUO 阿 科尔沁旗文物管理所 « Liao Yelü Yuzhi mu fajue jianbao » 耶律羽之墓 掘 1996 [Excavation of Yelü Yuzhi's Tomb of the Liao Dynasty], Wenwu, 1, p. 4-35. Ql DONGFANG 方 1999 Tangdai jinyinqi yanjiu 唐代 金 器研究 [Studies of Tang Gold and Silver Wares]5 Beijing, Zhongguo shehui kexue chubanshe. Shaanxi sheng bowuguan 西 史博 物 , Beijing daxue kaogu wenwu xueyuan 北京 大学 考古文博学院,Beijing daxue zhendan gudai wenming yanjiu zhongxin 北京大学 震旦古代文明研究中心 2003 Hua wu da Tang chun 花舞大唐春 [Flowery Dances from de Great Tang Spring], Beijing, Wenwu chubanshe. Shaanxi sheng kaogusuo 唳西省考古所 2001 "XTan faxian de Bei Zhou An Qie mu" 西安 的北周安伽墓 [The Tomb of An Qie from the Northern Zhou Dynasty discovered in Xi'an], Wenwu, 1, p. 4-26. Shijiazhuang diqu geweihui wenhua ju wenwu fajue zu 石家 庄地区革 委会文化局文物 掘 "Hebei Zanhuangdong Wei Li Xizong mu” 河北 皇 魏李希宗 墓 [The Wei 1977 Dynasty Tomb of Li Xizong in Zanhuangdong, Hebei], Kaogu, 6, p. 382-390, 372. Sun Ji 机 1991 "Lun Xi?an Hejiacun chutu de manao shoushou bei" 西安 何 家村出土 的 瑙 首杯 [On the Animal-Headed Agate Cup from Hejiacun, Xi'an], Wenwu , 6, p 84-93. 1996 "Tubanwen yinqi yu shuibowen yinqi" 凸 瓣 器与水波 器 [Silver Pièces with Protubérant Pétal Designs and Silver Pièces with Wave designs], in his Zhongguo shenghuo: Zhonguo guwenwu yu dongxi wenhua jiaoliuzhong de ruogan wenti 中国圣 火：中国古文物与 西文化交流中 的若干

Hejiacun Treasure and Sogdian culture

Qi Dongfang

121

[Chinese Sacred Fire: Some Questions about East- West Cultural Communication], Shenyang, Liaoning jiaoyu chubanshe. TOKYO KOKURITSU HAKUBUTSUKAN 東京国立博物館 et als. (ed.) 1998 Kyütei no eiga: Tö no jyotei solaitenbukô to sono jidaiten 宮廷の栄華唐の女 帝 則天武后と その時代展 [The glory of the court : Tang Dynasty Empress Wu and her Times], Tokyo, Ötuka kögeisha 大塚巧 藝 社. baohukaogusuo 西安 市文物 保 考古 所 wenwu shi Xi'an “Tang Yao Wupo mu wajue jianbao" 唐 姚无陂墓 掘 2002 [Brief report on the Excavation of the Tomb of Yao Wupo of the Tang Dynasty], Wenwu, 12*

p. 72-81. YAO Runeng 姚汝能 1983 An Lushan shÿi 安禄山事迹 [Traces of An Lushan], Shanghai, Shanghai guji

chubanshe. ZHU Tianshu 朱 天舒 1998 Liaodai jinyinqi 代金 器 [Gold and Silver Wares of the Liao Dynasty], . Beijing: Wenwu chubanshe.

The self-image of the Sogdians Frantz Grenet

Written sources from diverse origins (mainly Sogdian, Chinese and Arab) show that Sogdiana and the Sogdian people had achieved, as early as the second or third Century AD, a central position in the conmiercial affairs of Central Asia, China, and to a lesser extent India and Southeast Asia; they maintained it continuously until the eighth Cen¬ tury1. Sogdian merchants, administrators and générais exerted an important influence in the Turkish and Chinese empires. They conducted missions in Byzantium, and attempted to do so (unsuccesfully) in the Sasanian empire. It has been demonstrated that on several occasions their initiatives betrayed a lucid perception of the Asian market as a whole ; the clearest example is the commercial approach to Byzantium in 567, disguised under the cover of an embassy sent by the Turkish qaghan, but in reality aiming at channeling towards the most prospective buyers the unprecedented mass of Chinese silk paid to the Turks by the Zhou and the Qi. This politico-commercial network was disrupted by the Arab conquest (in 712 at Samarkand), and subsequently by the closing up of China (after An Lushan’s rébellion in 755), but there were some partial revivais in the two following centuries. At the same time, the Sogdians hâve left a substantial legacy of literary records and an immense one of figured images : mural paintings in Sogdiana itself (mainly at Samarkand and Panjikent), carved reliefs (sometimes painted as well) which adomed graves of Sogdian community leaders in Northern China. The graves date from the second half of the sixth Century, the murais mostly from the period 600-750 ; these works of art are, therefore, exactly Contemporary with the heyday of Sogdian trading prosperity and international influence. It might seem natural to question the extent to which the unique position of the Sog¬ dians was reflected in these artistic créations. What self-image did they try to project? How did they stage their commercial activities? Did they ever express anything like a « Sogdocentrism »?

Expressions of centrality in the Sogdian language

A brief examination of the vocabulary used in Sogdian records of a non-literary na¬ ture shows an opposition between the centre and the periphery, but it is always determined by the particular situation of the speaker. In the Ancient Leiters, written in Gansu, 1

de la Vaissière, 2002/4.

Les Sogdiens en Chine

124

« inside » {cntr) means China beyond the Yellow River, « outside » (ßyK) the territories to the west of it (the opposition is also expressed in such terms in Ancient Letter II addressed to Samarkand); in the Mugh letter B-16, cyrts’r « to the inside » (< cntrs’r) probably refers to Panjikent2. «Up» (’sk and derived words) means «east» or «northeast», «down» (c’ôr and derived words) means «west» or «Southwest», a pair of words which reflects the same perception as in Iran (Abarshahr «Upper satrapies» ver¬ sus Ërak «Down land»).

Geopolitical Propaganda in Sogdiana

Two monuments standing in Sogdiana itself, commissioned or inspired by local rulers, conveyed an explicit spatial symbolism (other such monuments most probably existed)3. The earliest one we know about, situated near Kushaniya (a town between Samarkand and Bukhara), has not survived, but it is described in the Tangshu whose information probably derived from reports of Chinese envoys sent in the first decade of the seventh Century. It States that «to the east of the town there is a two-storeyed pavilion in which they painted, to the north, the ancient emperors of China ; to the east, the princes and kings of the Turks and of the Indians ; to the west, those of Persia and Byzantium ; every moming the prince of this country goes there and kneels.»4 The second monument is the famous Samarkand mural known as the «Ambassadors’ painting», today exposed in a muséum built near to where it was found on the Afrasiab plateau which corresponds to the site of the city befbre the Mongol invasion. The painting adomed not a free-standing pavilion, but the réception hall of a large aristocratie house ; although this was not the royal palace of the ikhshids (kings) of Samarkand, this house was most probably connected in some way with King Varkhuman, known in Chinese records to hâve been formally appointed «govemor» of his kingdom by the emperor Gaozong in 658. Varkhuman is mentioned in an inscription on the western wall as participating in the réception of ambassadors which is depicted on this wall, and as a whole the composition of the four sides appears to reflect directly his Propaganda. The identification of the scenes has been, however, the subject of varions and sometimes conflicting interprétations5. Drawings of the three best preserved walls

2

This interprétation is put forward by Livshits, 1962, pp. 173-175. For a possible depiction of four or more «kings of the world» in a private house in Panjikent, see Marshak, Raspopova, 1990, pp. 123-176, esp. pp. 156-176 and figs. 26-34. Here the attendants in the upper row are either kneeling in front of the king or standing on the sides, while those in the lower row are drinking seated on carpets, thus occupying a similar position to the more sober Turkish guards and their officers on the western wall of the Afrasiab painting. Khosrow Anoshirvan held his court above rows of grandees seated in hierarchical order (Ibn al-Balkhi, 1921, p. 97). 4 Chavannes, 1903, p. 145. The relevance of this description to the interprétation of the spatial layout of the Samarkand painting was first emphasized by Mode, 1993. 5 Al’baum, 1975; Silvi Antonini, 1989; Mode, 1993; Marshak, 1994; Kageyama, 2002 ; Grenet, 2003a. [At a round table held on 14 March 2005 at the Centro Studi e Documentazione della Cultura Armena, Venice the pre-Islamic paintings a/ Afrasiab (Samarkand)}, I hâve {Nawruz in Venice. New perspectives presented a shorter analysis, taking into account some calendar observations which came to my notice after 3

Illi' Frantz Grenet

The self-image of the Sogdians

125

are reproduced here (Figs. 1, 2, 3). On figs. 1 and 2 the upper drawing shows what is actually preserved, while the lower one is a reconstruction proposed by François Ory and myself, according to Al’baum’s previous proposai modified by our general interprétation of the subjects and also, conceming some details, by analogies found in Sogdian painting (on fig. 3 the reconstructed parts are indicated by a grey background). The action depicted on the southem wall (Fig. 1) clearly takes place at Samarkand : this is the New Y ear procession described in Chinese records pertaining to the kingdom of Chach (today Taslikent), but this ceremony surely took place in other Sogdian principalities as well. The procession is riding tov/ards the mausoleum of the kiiig’s parents, and is dominated at its rear by the oversized figure of the ruler ; between the latter and his wives riding in front, animais are brouglit for sacrifice, in a fashion typical of the «national» religion of the Sogdians, a spécifie form of Zoroastrianism with ample iconographie borrowings froni the Hindu religion (Chinese records applied the term xian jiao to that particular religious mixture). Here some attendants wear the ritual mask (padcun), others carry the maces prescribed for killing the animais ; the horse is most probably destined for Mithra (as also in Iran), the four geese, norm al ly the animal of Brahma, could be destined for Zurvan, the Iranian god of Time whom two Sogdian texts identify with Brahma. The spatial setting is strietly observed : in Chach the royal mausoleum is said to be located outside the town, to the east, and this corresponds precisely to the situation we can observe on the Afrasiab painting, where the procession moves away from the western wall, thus from Samarkand (see below), and heads eastwards. The northem wall (Fig. 2) is entirely occupied by Chinese characters, equally divided into two scenes. To the right, a group of riders are hunting panthers. The group of hunters is dominated by a huge figure whose size matches only the Samarkand king on the southem wall, and who therefore can only be the Chinese emperor Gaozong. To the left two pleasure boats are depicted ; one of them carries a group of Chinese ladies*6, including musicians and a slightly oversized figure. She is most probably the empress. Here Varkhuman’s Propaganda présents his Chinese allies (and nominally overlords) both as brave people, hence reliable, and as enjoying the good life as well, a feature which could only please the Sogdians7. The eastem wall was interpreted by Al’baum, the author of the first full publication the present paper had been written; the Proceedings will be edited by M. Compareti and É. de la Vaissière as a special issue oîRivista degli Studi Orientait, Rome], 6 The absence of close parallels to the known Chinese costumes from the Tang period caused Al’baum some difficulties and he preferred to identify the ladies (and the other characters on this wall) as coming from « one of the régions of Eastem Turkestan » (Al’baum, 1975, pp. 19, 70) ; but the painted grave of emperor Gaozang’s sister near Xi’an, dated from 663 and therefore Contemporary with the Afrasiab paintings, shows female costumes sharing many details with the latter (e.g. metallic wings attached to the hair, dresses with high yokes and narrow vertical bi-coloured Strips) : see Shaanxi Provincial Institute of Archaeology, 1997 (in Chinese). I owe this important information to Dr. Etsuko Kageyama, who will publish a detailed contribution. 7 These interprétations of the northem and southem wall are based upon those ofMarshak, 1994, with some additional observations.

Les Sogdiens en Chine

126

Fig 1: Afrasiab Painting, Southern Wall.

Fig 2: Afrasiab Painting, Northern Wall.

Fig. 3 (next page): Afrasiab painting, western wall, the réception of ambassadors : 1 - Turkish guards seated or standing (one of the latter carries a polo stick) ; 2 — Envoys from neighbouring principalities (the one in the middle is epigraphically identified as the chancellor of Chaganian); 3 - Kneeling envoy probably introduced by a chambellan (restored) and followed by Turkish offîcers (the one in the middle is identified as coming from Qarashahr; 4 - Chinese envoys bringing silk; 5 - Turkish guards walking up to their seating place ; 6 - Seated Turkish guards; 7 - Turkish offîcers ; 8 - Interpretor introducing envoys from a forest people, followed by Koreans ; 9 - Spears gathered in a bündle, shields with grotesque masks ; 10 - King Varkhuman ? (figure entirely missing).

Frantz Grenet

The self-image of the Sogdians

127

Les Sogdiens en Chine

128

Fig. 4: Afrasiab painting, detail of the eastern wall (below: Krishna and Yashoda, the attack by the crane ? above : Krishna and the horse-demon Keshin ?)

of the paintings, as an évocation of India ; this opinion was subsequently challenged8, but I hâve recently proposed additional reasons to support it9. Al’baum’s assumption had been based upon the unquestionably Indian costu¬ mes and hairstyle of some of the characters. In addition, I managed to show that the left end of the composition dérivés firom Roman images showing Urania, the Muse of astrology, teaching her art to the poet Aratus ; at Samarkand the costumes have been transformed in order to suggest a Greek teaching an Indian, in accor¬ dance with the way the Indians themselves viewed the origins of their astrology. The rest of the lower register is distributed into scenes set in a conventional marshy landscape. I have tentatively suggested the iden¬ tification of several épisodes from the legend of Krishna’s youth (Fig. 4: the baby comforted by his foster-mother Yashoda just before being attacked by the crane-demon, cf. Harivanisa, chap. 50, and the adolescent Krishna confronting the demon Keshin appearing as a rea¬ ring horse; Fig. 5: plump nude children derived

Fig. 5 : Afrasiab painting, right part of the eastern wall (the Kamas ?)

8

Mode proposed, unconvincingly, to recognize the myth of the origin of the Turks near Lake Etsin-gol (Mode, 1993, pp. 97-104), while Marshak admitted the Indian features but attributed them to the Indian influx in the Sogdian représentations of the heavenly world, a possible suject of the eastern wall according to him (Marshak, 1994, pp. 16-18). 9 F. Grenet, 2003a.

The self-image of the Sogdians

Frantz Grenet

；

129

fkom the Erotes archers of Greek and Roman art, sometimes reinterpreted in India art as Kama, the god of love; this could be an allusion to Krishna’s romances with the cowgirls10). At this stage of the présentation it appears that at least two sides of the Samarkand painting function as Windows opening to distant countries, which are placed roughly in the direction which they occupied from the Sogdian viewpoint : China is to the north, as the main roads leading to it first took a northeast direction ; India, which was mostly reached through Färsnäma (pertaining to the 6th c.) the passes of the Upper Indus (to the southeast of North: Hephtalites East: Chinese

West: Byzantines South: Persians

Kärnämag îArdashïr (8® c.) North: Turks West:

East: Indians (Kabul)

Byzantines South: Persians

Sogd. Manichean fragment (6th - early 7a c.) ?

East: Indians

West: Persians

South: Samarkand

Kushaniya pavilion (early 7th c.) North: Chinese West: Persians+

East: Turks+

Byzantines

South: Kushaniya Ruler Fig. 6: The “Kings of the World” according to varions sources. 10

Indians

Sogdiana), occupies the eastem wall. These orient¬ ations are consistent with those attested for the Kushaniya pavilion. One has now to tum to the sour¬ ces of these compositions in order to understand why at Samarkand (and, implicitly, at Kushaniya) the local Sog¬ dian ruler is on the south side, and what exactly did take place at Samarkand on the main, western wall, on which one can now see only the lower and middle registers by occupied converging foreign délé¬ gations. These sources of inspi¬ ration are clearly Sasanian (Fig. 6). The earliest known context is in the Färsnäma, a Persian political treatise of the xn" Century drawing upon information from the time of Khosrow Anoshirvan (r. 531-579)11. The Sasanian king is said to have, in his solemn courts, set an empty throne to his

1 owe this information to Dr. Madhuvanti Ghose (SOAS, London), who drew my attention to Rajput

miniatures showing Kama shooting at Krishna and Radha (see in particular the manuscript of the Gita Govinda in the Prince of Wales Museum, Bombay, c. 1525-1530). 11 Passage quoted above, fii. 3. For this and other cognate accounts see Grabar, 1954.

130

Les Sogdiens en Chine

left for the emperor of China, another one behind him for the king of the Hephtalites (or of the Khazars, in an anachronical variant) ; the third one, to his right, belonged to the Cesar of Rome (i.e. the Byzantine emperor). He himself faced his court, to the south, as one can deduce from the position of the other thrones. The Hephtalites appear in the position of northem neighbours of the Sasanians, which reflects the situation in the fïrst part of Khosrow’s reign. The second text is at the end of the Kârnâmag ï Ardashïr (XIV-19), a Pahlavi account of the deeds of the first Sasanian king12. Here no orientation is specifîed, but three foreign kings are said to hâve corne and paid hommage to Ardashir’s second successor Hormizd: the Cesar of Rome, the Tegin of Kabul who is the king of the Indians, and the Qaghan of the Turks. The scheme which can be determined corresponds to the one in the Fârsnâma, but the list has been adapted to the political realities of the post-Sasanian

period. The third text is a Manichean fragment in Sogdian (Leningrad 3549)13. Here again no orientation is specifîed, but the list of kings clearly moves in an anticlockwise direc¬ tion : first the Indian devarâja, then a king whose title is missing in a lacuna, then the shänshäi (Shähänshäh), and last the ikhshïd of Soghd. This list is of great interest because it proves that the Sasanian scheme had been adopted by the ikhsïds of Soghd, namely the kings of Samarkand. For the missing name of the northem king one can restore the Qaghan of the Turks, in conformity with the Sasanian model, or, altematively, the emperor of China, in conformity with the paintings at Kushaniya and Afrasiab. In either case the presence of a royal Samarkandian scene on the Southern wall of the Afrasiab room now appears perfectly consistent with an established tradition. We can now corne back to the western, rear wall (Fig. 3). The two other Sogdian examples known to us, the Kushaniya pavilion and the Manichean list, would lead us to expect this wall to be attributed to the Persian king, or to be shared between him and the Byzantine emperor. But between the time of these documents and that of the Afrasiab painting, the Sasanian empire has disappeared and Byzantium has great difficulties resisting the Arab attacks. The Arabs are hardly perceived as an empire in that period ; still sixty years later the king of Samarkand described them as a temporary nuisance in his letter to the Chinese emperor. The answer to this geopolitical vacuum was found in filling the western wall with an audience scene. In doing so the Samarkand painters perhaps retained something of the former model, as a possible way to depict the Sasa¬ nian king was to show him holding his court. It is most probable that the western wall, like the Southern one, was presided upon by the oversized figure of Varkhuman himself. This hypothesis is consistent with the contents of the inscription on the same wall, which explicitedly mentions both Samarkand and the ikhshîd Varkhuman. There are also many Turks on this wall, acting in varions capacities : some are guards, walking up to their alloted place; others are richly dressed dignitaries who seem to accompany the foreign delegates but do not carry gifts themselves. On the right-hand side two of these dignitaries are facing a plainly dressed character, whom Boris Marshak has convincingly identified as an interpréter. He himself introduces forest or 12 13

Grenet, 2003b, pp. 116-117, commentary pp. 44-45, 125. Ragoza, 1980, p. 43. 1 am grateful to Yutaka Yoshida for having drawn my attention to this text.

Ul' Frantz Grenet

The self-image of the Sogdians

131

mountain people (Tibetans, Tuyuhun ?) bringing a yak tail and a fur; they are followed

by two Koreans. The important role played by Turks in this composition, and the fact that they never carry gifts, has led Markus Mode to propose that the audience was presided not by the Samarkand king himself, but by a Turkish qaghan temporarily holding his court in his vassal’s résidence at Samarkand. Such an interprétation raises several difficulties. Firstly, it clearly goes against the inscription, which mentions only Varkhuman but no qaghan. Secondly, the direction allotted to the Turks in all the other examples we hâve examined, either pictural or textual, are the north or the east, never the west, which would be illogical in a Samarkand-centered vision of the space. Thirdly, one is at pains to find a suitable candidate in that period for the rôle of a qaghan enjoying equal prestige with both the Samarkandians and the Chinese. The political reality of the time is a steady disintegration of the Western Turkish empire, culminating in its surrender to the Tang in 65814. The last qaghan Icnown to hâve exerted direct control on Samarkand, Duolu, took the city by violence in 642 but was immediately disowned by the Chinese and was confined to Tukharistan. Of his two competitors the one who enjoyed Chinese favour, Shegui, ceased to play any political role after 649. This confused political context provides a likely background for the first Arab foray near Samarkand in 654, then the

ascent of King Varkhuman and his commited pro-Chinese diplomacy spectacularly displayed on the Afrasiab painting. At this point it is appropriate to quote two well-known texts. One is an edict of the Chinese emperor Gaozong in 657, in which he adresses the rallied Turkish nobles : On your own side, you will with a sincere heart go to your States ; you must, together with and his collègues, settle in a just männer in these hordes the official Lu Chengqing functions great and small, the insignia and the ranks high and low, distribute the functions of prefects and those below15.

As a matter of fact these instructions apply specifically to the Turkish tribes ; Tangshu mentions that of nevertheless, just after the reorganization of the tribes, the « the various kingdoms which were submitted to them, as far as Persia »16. The second text, already mentioned by Boris Marshak in connection to the Afrasiab painting17, is the fimerary inscription of the Turkish qaghan Kul-tegin : destruction of the Eastern qaghanate 630] the Turkish During fifty years [that is after nobles devoted their work and their forces to the Chinese emperor. In the direction of the rising sun they campaigned as far as Koguryo, in the other direction they campaigned as far as the Iron Gates.

Let us tum again to the figures on the western wall. To the extreme right we have 14

Chavannes, 1903, narrative pp. 265-268, translations of the relevant passages from the Jiu Tangshu pp. 28-40 and the Xin Tang shu pp. 56-67. 15 Chavannes, 1903, pp. 39-40. 16 Chavannes, 1903, p. 67. 17 Marshak, 1994, pp. 7-8 ; see Kliashtomyi, 1964, p. 23 and 2003, p. 96.

132

Les Sogdiens en Chine

_

two représentatives of Koguryo ; as Etsuko Kageyama has argued18, their image could in fact by stereotyped and not reflect an actual presence of Koreans at the Samarkand court, yet it was feit necessary to include it. To the extreme left, the inscription of the chief envoy of Chaganian (the chancellor) was painted on a white surface conveniently provided by the back of the dress of a Turkish guard ; the envoy himself, identified by an inscription19, is walking below, in the centre of a group of three. At that time? as confïrmed by recent excavations, the fortification at the Iron Gate of Derbent was functioning and meant to protect not Sogdiana from Tukharistan, but Tukharistan from Sogdiana, and it belonged to the kingdom of Chaganian20 As a resuit, the répartition of the foreign envoys on the western wall matches exactly the geographical span of Sino-Turkish coopération as perceived by the Turks themselves : « from Korea to the Iron Gates »a The Turks on the Afrasiab painting are not représentatives of a Turkish empire which no longer exists or5 to be more exact3 has been wholly instrumentalized by the Chinese power. They are go-betweens, commissioned by the Son of Heaven to supply his western allies with military assistance - hence the guards in the upper register - and to accompany envoys who hâve to travel through territories till recently controlled by the qaghans27. One of these territories can actually be identified thanks to fresh reexamination of an inscription, This inscription is carefolly painted on the wrist of a Turkish dignitary (the one in the centre of the group of three to the right of the envoys from Chaganian). It is read Ark (,»), which I hâve proposed to recognize as the already attested Sogdian rendering of Argi, today Qarashahr between Kucha and Turfan22. In 648 Qarashahr had 21.

i8

Kageyama, 2002. Quoted by A.M. Belenitskii, B.I. Marshak, in Azarpay, 1981, p. 62. In November 2004 Étienne de la Vaissière and I were able to check part of the unpublished tracings and photographs with Prof. V.A. Livshits. The alleged incription « Varkhuman » on the character preceding the chancelor of Chaganian is in fact illegible on the photographs. There is no reason to suppose that the inscription « a man from Tibet » was fancied by a later visitor. The fact that it is put on the face of a Chinese could be due to lack of coordination between the painter and the scribe : maybe Tibetans had been actually depicted somewhere (in front of the Koreans ?), but the scribe who had been provided with a list by Varkhuman’s Chancery was not able to recognize these people, then newcomers on the Central Asiatic stage, and he did what he could, picking up a Chinese to pass for a Tibetan. 20 Rakhmanov, Rapin, 1999. 21 According to AFbaum, 1975, and Yatsenko, 2004, some characters from the left part of the western wall, including the Chaganian envoy, reappear on the Southern wall, though dififerently dressed, besides Varkhuman and on top of the camel. No such link exists between the other walls. Presumably the events depicted on these two walls are meant to occur at the same place (Samarkand), at the same time (the New Year festival in June) and presided by the same figure (Varkhuman). Marshak is reluctant to restore him at the top of the western wall and instead proposes the image of a deity (preferably Nana), because at Panjikent the rear wall of a réception room is always occupied by a religious scene ; but, judging &om his own interprétation of what is still visible on the Afrasiab painting, the only wall here which présents a clear religious character is the Southern one. Another argument adduced by Marshak is the sentence « I am well aware of the gods and of the writing of Samarkand » in the inscription commissioned by the envoy of Chaganian. This does not necessarily refer to divine images presiding over the gathering of embassies, even less to the envoy renouncing to spread Buddhism and its sacred books in Zoroastrian Samarkand. I would tentatively suggest that he merely contributed to the New Year ceremony and made sure his friendly act was recorded in wrinting. As a matter of fact one goose on the southem wall bears an inscription « gift ». 22 Grenet, 2003a, pp. 123-4, fii. 3. 19

Frantz Grenet been

The selßimage of the Sogdians

133

snatched from the qaghan Shegui by a Turkish general in service of the Chinese,

with the local dynasty nominally maintained. We know from the paintings at Shorchuq that the aristocratie costume at Qarashahr was more Chinese than Turk, so it is perhaps préférable to identify the man from Qarashahr either as an officer of the Turkish garnison, or, altemately, as a leader of the tribe of the Shunishe, which controlled the pass just to the north of Qarashahr and which was tumed into a Chinese district in 65723. The overall situation can best be explained in the political context of the years just after 658, when China had dominated the Western Turkish empire and was using the rallied Turk aristocracy to help establish its political and military control far in the west. The western wall is largely a monument to the new pax sinica, celebrated through its two main instruments : Chinese silk and Turkish force. The Samarkand king Varkhuman poses as a loyal partner in this new situation. Nevertheless, in front of his own subjects, he intends to play his own tune and, in doing so, probably slightly offends the decency the Chinese would hâve required from a less distant client ruler. His own image, set on the Southern wall according to the established tradition, is a little taller than that of his Chinese overlord on the opposite wall. If, as I believe, he also présidés over the ceremony on the western wall, he is shown receiving silk directly from Chinese envoys, a privilège Chinese diplomacy had formerly reserved for the qaghans. Surrounded (only in the lower and middle rows ?) by seated Turkish soldiers and officers, he emulates the protocol of his former masters24. At the same time, in annexing this western wall which had always been reserved for the Shahanshahs, Varkhuman présents himself as their natural heir. Taken together, the Kushaniya and Samarkand monuments (as well as the Manichean fragment) prove the existence of a centered vision of the world, but the centring was on the Sogdian ruler rather than on Sogdiana itself, and the overall scheme was clearly borrowed from Sasanian royal Propaganda. Above ail, veiy few éléments in the depiction of foreign countries dérivé from the actual expérience the Sogdians had of them. The images of China are quite crédible, except for some details, but they are more probably inspired by Chinese scrolls rather than by the accounts of eye-winesses25. India is completely artificial, except for the costumes ; it is a collage of various pictures, some of recognizable Western origin, the purpose of which is to convey the idea of a fabled country endowed with abundant rivers, astrologers and pastoral legends - the same sort of image India had in Arab and western médiéval 23

Chavannes, 1903, p. 14 fh. 8, and pp. 110-4 on Qarashahr itself. In 662 the Shunishe rebelled against the puppet qaghan Puzheng but were again submitted by the intervention of the Chinese govemor (iH, p. 73). 24 There is at least one indication that éléments of the Turkish protocol had been adopted by the Samarkand rulers : according to the great inscription commissioned by the envoy from Chaganian, Varkhuman « approached » him and the envoy from Chach, which calls to mind Datang xiyu ji : « The khan came out of his tent about thirty paces to meet Xuanzang who after a courteous greeting entered the tent » (Xuanzang, 1904-1905, p. 74). According to Beishi (chap. 97) and Suishu (chap. 83), in the second half of the sixth c. the Samarkand rulers, though of local extraction, had become quite turcified : the king wore plaited hair and had married the qaghan’s daughter. There is no way to décidé when, if at ail, they discarded that part of their héritage. 25 The direct influence of Chinese pictural models on western and northem wall of the Samarkand painting was pointed out by E. Kageyama, 2002.

Les Sogdiens en Chine

134

literature. The real India known by the Sogdian traders, the India of towns and harbours, should be looked for in the paintings of the Buddhist monasteries in Xinjiang, not in Sogdian painting.

Trade ; the most notable absentee front Sogdian art These characteristics of the « exotic » panels in the Samarkand mural are well in line with the general lack of references to trade and far-away travels in the immense and diverse repertory of Sogdian painting as a whole, in which heroic subjects play an overwhelming rôle. Precious stones always appear in their final form (je weis, embroideries), silk always in clothes wom by distinguished people, with the one exception of the cocoons, hank and rolls carried by foreign envoys at Samarkand. Otherwise, the only raw material traded by Sogdian merchants which was reproduced in paintings is sandalwood, and only once, in the illustration of a Indian taie26. There is only one other hint of trade in the Southern Seas, the story of «The Spirit and the Merchant», but the illustration (Fig. 7) is reduced to the essentiel éléments : the merchant, his daughter, the Spirit of the Sea identified by two dolphins, nothing eise, not even the merchant’s «treasures» which are mentioned in the Sogdian written version of the story27. In the whole of Sogdian art there is not a single caravan, not a single ship, except the pleasure boats of the Chinese emperess. The few scenes including Sogdian merchants show them banqueting like good aristocrats28. The only scene celebrating the acquisition of wealth shows a harvest.29 Castles appear but no urban landscape, some

Fig. 7 : Panjikent, Room 41/VI, lower register : The Spirit and the Merchant

26

Marshak, 2002, pp. 64-65, fig. 32. Marshak, 2002, pp. 62-64, fig. 31 ; Henning, 1945, pp. 471-472. 28 See E. Kageyama’s paper in the present volume. 27

Frantz Grenet

The self-image of the Sogdians

135

merchants and craftsmen but qo bazar. Such an exclusion of trading activities from the

range of illustrated subjects appears too systematic to be attributed only to the hazards of documentation. 30 Examination of the Sogdian funerary reliefs discovered in Northern China29 can provide a valuable counter-appraisal, as most, if not all, were commissioned by a sabao, a community leader whose administrative position, or that of his close ancestors, derived from commercial activities (the Chinese word sabao reproduces Sogd. s’rtp’w « caravan leader »). Also, two of these sets of reliefs (the anonymous one in the Miho Museum and the one from the grave of the sabao An Jia) are particularly picturesque, aiming at conveying a vivid impression of scenes connected with the career of the deceased. What in fact appears is the following range of subjects : 1) Reference to the Sogdian national religion. They are quite developped on the Miho couch (the funerary ritual, Marshak fig. 4 ; the goddess Nana possibly appearing in the context of a joumey to or from the motherland, fig. 5 ; the cuit to a river-god, fig. 7). On the An Jia couch they are stricly restricted to the lintel (a place both central and discrète), with the depiction of the fire altar served by two «priest-birds» (/〃Jia tomb, pl. facing p. 16 ; this composition is found on some other Sino-Sogdian graves as well, also in an axial position, and is now proved to be originated from Sogdiana31). 2) Extremely sparse allusions to trade : on the Miho couch, loaded camels appear on two panels among thirteen (4, 8 c) ; on the An Jia couch, a loaded camel and donkeys appears only on one panel among twelve (pl. facing p. 24, middle). In both cases, the beasts do not necessarily carry traded goods, but perhaps rather yurts and supplies.32 3) The bulk of the repertory is devoted to social life. On the Miho couch, it consists both of hunting or riding with ethnically diverse partners (Hephtalites or Northern Indians, 9 a-b ; Turks, 8 a-b-c), and of holding a banquet with the consort (3 a) or with fèllow countrymen (3 b). On the An Jia couch one finds again the same subjects, but the theme of the banquet is quite overused, the aim being obviously to show the deceased (always recognizable by his white hat and short moustache) in every possible contact situation : with Turks (pp. 31, 34), with Sogdians (pl. facing p. 22, p. 37 left), with both (p. 35), variations being introduced in the setting (a Chinese pavilion, a yurt, a yurt fumished with a Chinese seat ). Three scenes break this monotony: a combined picnic and hunting scene copied on Sasanian models (p. 28, here Fig. 8), which explains the presence of a lion, an animal 29 Marshak, Raspopova, 1990, pp. 149-157, fig. 22. In the large house XXIV paintings in two rooms show grooms parading stallions in front of a Sogdian aristocrat, but this is not necessarily an indication of trade, let alone international trade (Belenitskii, Marshak, 1973, pp. 62-64). 30 Reference given in the texts are to Marshak, 2001, pp. 227-264 (the only comprehensive study to date), except for the An Jia grave for which page numbers refer to the b etter illustrated Shaanxi Provincial Institute of Archaeology, 2003. 31 Grenet, 2003c, pp. 6-7, fig. 10. 32 These pages were written before the sarcophagus from the tomb of the sabao Wirkak unearthed in Xi’an was published, see Yang Junkai’s paper in this volume. This set of reliefs stands apart in the whole repertory of Sogdian and Sino-Sogdian art for the place devoted to commercial activities, both in the narrative of the deceased's life and in the staging of his eschatological expectations (on which see also F. Grenet, P. Riboud, Yang Junkai, 2004, pp. 273-284).

Les Sogdiens en Chine

136

Sogdian hunters never came accross33 ; a banquet scene (p. 26, here Fig. 9) where the deceased entertains a mixed company of Turks and Sogdians by playing music (an art Iranian aristocrats were supposed to practice)34 ; a conversation between the deceased and a Turk (p. 33, here Fig. 10), inside a yurt besides which three Sogdians are watching, seated together with a character who wears the winged crown of Peroz presumably borrowed by the Hephtalites and surely by some later Sogdian kings.35 Clearly, the rôle in which An Jia wants to appear is that of a diplomat, or at least a trusted intermediary between his mother country, China and the Turks, knowing the aristocratie codes of ail three people.

Fig. 8, 9, 10 : Details from An Jia couch.

33

The same explanation can be proposed for the lion hunt on the recently appeared couch in the Vahid Kooros collection (Delacour, Riboud, 2004, p. 26, ill. 23). The impact of the Sasanian model is obvious also because one of the hunters wears a ribboned crown copied from a Sasanian type. 34 An Jia plays the harp, while a Turk plays the lute, like Ardashir, cf. Kârnâmag ï Ardashïr, II1.2 (Grenet, 2003b, pp. 64-65, commentary pp. 40, 122). 35 Rulers of Ustrushana (Smimova, 1981, pp. 324-335), date uncertain ; Dewashtich, ruler of Panjikent, as proclaimed ikhshïd of Soghd in 721-722 (Belenitskii, Marshak, in Azarpay, 1981, pl. 29-30, p. 64, fig. 30, pl-24).

Frantz Grenet

The self-image of the Sogdians

137

This last requirement could perhaps provide an explanation for an otherwise stränge peculiarity of these funerary reliefs : the defunct always appears in Sogdian costume but never his consort, who is always dressed like a Chinese person.36 Epitaphs of Sogdian sabaos prove, however, that they took wives in both communities, more often among Sogdians than among Chinese.37 The Turks used to show their wives, and in Chinese funerary art the image of the banqueting couple was a must, but this was not the case in Sogdiana where the wife appears only in the context of the domestic cuit38. In addition, in Chinese society the Sogdian female costume, similar to the male one, had probably become too associated with the underworld of dancing girls, either Sogdian or Chinese (who had adopted this fashion), to be considered respectable anymore. The only option for a distinguished Sogdian living in China was to show only his Chinese wife, or to hâve his Sogdian wife wear the Chinese costume.

Conclusion

The world the Sogdians chose to represent was very different from the world they were living in and which they undoubtedly knew very well. They were proud of their wealth but they were not keen on showing its sources, contrary to Chinese craftsmen who depicted Sogdians engaged in many commercial activities : caravaning, wine trade, hawking, horse dealing, trading of hunting animais. Ail of them are conspicuously absent from genuine Sogdian images. Partial explanations for this discrepancy can be sought in local circumstances. The mass of Sogdian painting cornes from Panjikent, a city which was little engaged in international trade and where the landed aristocracy was over-represented. In China, trade always had to be concealed behind a fiction of tribute. But, on a deeper level, Sogdian society appears to hâve remained merged in the warlike and aristocratie values inherited from the the period of « Sarmatian » hegemony which had immediately preceded the heyday of Sogdian trade. The ethics of Sasanian Iran, which exerted its fascination from the fifth Century onwards and even more after its disappearance, also played their part. Sogdiana was largely a trading society, but its art never celebrated the merchants. The originality of Sogdian art lies for a large part in this extreme tension between reality and dream ; this originality can be fully appreciated only when confronted with the Buddhist art of Gandhara and Serindia on the one hand, and Arab miniatures (in particular illustrations of al-Hariri’s Maqamat) on the other.

36

This is the case with the Miho grave (Marshak figs. 3 a, 4, 5, 6), the An Jia grave (p. 30, p. 37 right), and also the grave found near Anyang (Marshak fig. 1). The grave of Yu Hong is a possible exception: in the scene showing the couple banqueting, presumably in Paradise, the wife has a non-Chinese dress, and a crown (21) like in Sasanian banquet scenes, but these reliefs are the less Chinese of them ail; the deceased, whose ethnie origin remains mysterious, seems to hâve become strongly persianized as a resuit of his embassy in Iran. 37 de la Vaissière, Trombert, 2004. 38 Consorts are shown only in cuit scenes, except for one single image of a man and his wife seated on cushions, painted in their private sitting room : Belenitski, Marshak, 1973, fig. 9.

Les Sogdiens en Chine

138

Additional note

At the last minute before handing over this text for publication I came accross an article by S.A. Yatsenko in the online journal Transoxiana, 8, June 2004 (http://transoxiana.org), « The costume of foreign embassies and inhabitants of Samarkand on wall painting of the 7th c. in the ‘Hall of Ambassadors’ from Afrasiab as a historical source ». This article, which ignores Marshak' s and my last contributions, is vitiated by serions mistakes in the identification of some characters (the fur-bringing people are considered as the envoys from Chach, although the latter would surely not need an interpretor in Samarkand). Also, Yatsenko clings to Al’baum’s assumption in recognizing wedding cortèges both in the group of Chinese ladies (rather Khotanese according to him, but see above, fil. 6) on the northern wall, and in the procession on the Southern wall. In the latter case at least, Marshak ’s identification of the scene as the New Year ancestral sacrifice makes more sense of ail details than Yatsenko’s refined version (a marriage between Varkhuman and a Chaganian princess, celebrated in pure Turkish fashion and presided over by the gigantic figure of the bride’s uncle!). The article, nevertheless, contains important observations and comparisons of costumes, which in particular allow to recognize three hierarchical levels among the Turks depicted on the western wall, plus a higher dignitary dressed in blue and seated right in the axis of the composition. Yatsenko proposes to identify him as the tudun, chief représentative of the qaghan. If this identification is right, this official might hâve conveyed to Varkhuman some formai récognition which, however fictitious the qaghanate had become, was probably still feit as prestigious (cf. the Mugh document V-17 in which Dewashtich, on the move to proclaim himself ikhshîd of Soghd, boasts about having received « from the qaghan a high rank and honour »). The envoy preceding the chancellor of Chaganian carries two présents, a torque and a pearl necklace, the latter identical to those wom by women in the painting; Yatsenko considers them, convincingly, as destined for a man and a woman, who in his (and mine) interprétation can be no other than the Samarkand royal couple. As for the date of the painting, Yatsenko’s arguments pointing to the particular context of the year 662 are quite significant: the mission of the Chinese envoy Wang Minyuang to the Western countries, the réunification of the Western Qaghanate by the puppet qaghan Puzheng under Chinese military control, the diplomatie initiatives of the Koguryo kingdom in Order to escape growing Chinese pressure (provided the Korean envoys on the Afrasiab painting reflect a real event). The earlier mission of the envoy Dong Jisheng in 658, proposed by E. Kageyama (op. cit., p. 320, with reference to W. Anazawa and J. Manome, Chosen gakuho, 80, 1976, pp. 21 sqq.), remains a possible alternative ; contrary to Wang Minyuan this envoy is explicitly mentioned as having visited Varkhuman (Tang Huiyao, vol. 99).

Frantz Grenet

The self-image of the Sogdians

139

Bibliography

Al'baum, L.L

Zhivopis3 Afrasiaba, Tashkent. 1975 G. Sogdian Painting, Berkeley — Los Angeles - London. 1981 Belenitsku, A., Marshak, B. « Stennye rospisi, obnaruzhennye v 1970 godu na gorodishche drevnego 1973 Pendzhikenta » [Mural paintings discovered in 1970 on the site of ancient Panjikent], Soobshcheniia Gosudarstvennogo Ermitazha, 36, pp. 58-64. Chavannes, E. Documents sur les Tou-kiue (Turcs) occidentaux, Saint Pétersbourg. 1 903 Delacour, C. Riboud, P. Lit de pierre, sommeil barbare. Musée Guimet, Paris. 2004 Grabar, O. The painting of the six kings at Qüsayr 'Amrah, Ars Orientalis, 1, pp. 185-7. 1954 Grenet, F. 2003a « L’Inde des astrologues sur une peinture sogdienne du vile siècle », in Religious thèmes and texts in pre-Islamic Iran and Central Asia. Studies in honour of Professor Gherardo Gnoli, ed. C. Cereti, M. Maggi, E. Provasi; Wiesbaden, pp. 123-129, pl. 2-3. Geste d’Ardashir fils de Pâbag, Die. 2003b 2003c « Mithra, dieu iranien : nouvelles données », Topoi, 1 1, pp. 35-58. Grenet, F., Riboud, P., Yang Junlcai 2004 « Zoroastrian scenes on a newly discovered Sogdian tomb in Xi’an, Northern China », Studia Iranica, 33, pp. 273-284. Henning, W.B. « Sogdian taies », Bulletin of the School of Oriental and African Studies, 9, 1945 1945, pp. 465-487. IBN AL-BALKHI 1921 The Färsnäma oflbnu’l-Balkhï ed. G. le Strange, R.A. Nicholson, London. Kageyama, E. « A Chinese way of depicting foreign delegates discemed in the painting of 2002 Afrasiab », in Iran: Questions ef connaissances. Actes du IVe congrès européen des études iraniennes. Paris, 6-10 septembre 1999> voL I: Etudes l'Iran ancien, ed. Ph. Huyse, Paris, pp. 309-323. Kliashtornyi S.G. 1964 Drevn etiurkskie runicheskie pamiatniki kak istochnik po istorii Srednei Azii [Ancient turkic runic inscriptions as sources for the history of Central Asia], Moscow. Istoriia Tsentral’noi Azii / pamiatniki runicheskogo pis’ma [History of Central 2003 Asia and inscriptions in runic script], S aint-Peter sburg. DELA VAISSIÈRE, É., 2002/4 Histoire des marchands sogdiens, Paris, Mémoires de l’IHEC vol. 32, 2e éd. 2004. AZÂRPAY,

Les Sogdiens en Chine

140 DELA VAISSIÈRE,

2004

LIVSHITS, VA 1 962

É” TROMBERT, É” « Des Chinois et des Hu. Migrations et intégration des Iraniens orientaux en milieu, chinois durant le Haut Moyen-Âge », Annales. Histoire, Sciences Sociales, p. 931-69. Sogdijskie dokumenty s Gory Mug [Sogdian documents from Mount Mugh], H, Moskva,

Marshak, B.

« Le programme iconographique des peintures de la "Salle des ambassadeurs5 à A&asiab (Samarkand) », Arts Asiatiques, 49, pp. 5-20. 2001 « La thématique sogdienne dans Part de la Chine de la seconde moitié du Vïe siècle »3 Comptes rendus de F Académie des Inscriptions & Belles-lettres, pp. 227-264 Legends, taies, and fables in the art of Sogdiana, New York 2002 Marshak, B上, Raspopova, V.L « Wall paintings from a house with a granary », Silk Road Art and 1990 Archaeology, 1, pp. 123-176. Mode, M. Sogdien und die Herrscher der Welt, Frankfurt am Main. 1993 RAGOZA, AN Sogdiiskie fragmenty tsentraUno-aziatskogo sobraniia Instituta Vostoko1980 vedenia [Sogdian fragments in the central-asiatic collection of the Institute of Oriental Studies], Moskva. Rakhmanov, Sh., Rapin, C. « Les Tortes de Fer' près de Derbent (Ouzbékistan) », Dossiers 1999 d'Archéologie, 243, Paris, pp. 18-19. Shaanxi Provincial Institute of Archaeology et al. 陝西省考古研究所他 1997 唐 昭 陵 新城 長 公主 墓 発掘 簡 報 Tang Zhaoling Xincheng zhang Gongzhu mu fajue jianbo « A brief report on the excavation of the elder princess Xincheng's tomb of Tang Zhao Ling mausoleum », Kaogu yu wenwu, 1997 : 3, pp. 3-38 Xian Bei Zhou An Jia mu 西安北周安伽墓 [An Jia tomb of Northern Zhou at 2003 Xi5 an], Beijing. Silvi Antonini, Ch. 1989 « The paintings in the palace of Afrasiab (Samarkand) », Rivista degli Studi Orientalin 63, pp. 109-144; SMIRNOVA, O.I. Svodnyi katalog sogdiiskikh monet [General catalog of Sogdian coins], 1981 Moskva. Yatsenko, S.A. « The costume of fbreign embassies and inhabitants of Samarkand on wall 2004 painting of the 7th c. in the 'Hall of Ambassadors, from Afrasiab as a historical source » in the online journal Transoxiana, 8, June 2004 (http://transoxiana.org). Xuanzang 玄奘 1904-5 Da Tang Xiyuji transi. Th. Watters, On Yuan Chwang 's travels in India, 1904/905, London.

1994

From Stone to Silk: Intercultural Transformation of Funerary Fumishings Among Eastem Asian Peoples around 475-650 CE Angela Sheng'

From the collapse of tlie Han dynasty in 220 until the Sui-Tang unification in 589-618, China underwent turbulent political and socio-economic change. Despite repeated and massive destruction, this period of disunity also witnessed tremendous progress in the arts, ideas, and social customs —the conséquence of frequent and important intercultural exchange. Whether forced by the state or motivated by survival, more people moved around within China than previously and, often in large numbers. Besides domestic inter-regional migration and mingling, contact among peoples of different cultures also continued, even grew. Travel took three forms along three major routes, riding on horseback across steppes to pastures in the north, trekking with camel caravans across deserts to oases in the west, and sailing by ships in seas to lands further east. Diplomatie envoys, monks, merchants, and other travelers acted as agents of change, transmitting new ideology and praxis to areas far beyond the reach of each route. But, artifacts preserved along any of these routes and passed on as cultural héritage often lack complété contextual information, rendering scholarly investigation

inconclusive. Recent archaeological finds in China provide new evidence for understanding intercultural exchange not just along the east-west overland caravan route but also along the northward horseback route and the eastward seafaring route to Japan. A case in point is how Yu Hong’s house-shaped stone sarcophagus (Fig. 1), with Sogdian-inspired designs and dated 592 that was unearthed in Taiyuan in 1999, can shed light on some extant silk fragments of the Tenjukoku Shüchö (silk curtains embroidered with the motif of “heavenly-longevity-kingdom” hereafter 1 My thanks go to Rong Xinjiang with whom I first raised this topic in October 2002. For the révision of my paper on the relationship between textile art and the accumulation of Buddhist merit in which I cited the Tenjukoku curtains (Sheng, 2003), I was then reading Maria del Rosario Pradel's 1997 dissertation. Her argument against the prévalent interprétation of the curtains as Buddhist relies intrigued me. At the same time, Wu Hung’s grouping of eight house-shaped stone sarcophagi, including Yu Hong’s sarcophagus, also fascinated me (2002). I was able to explore the potential relationship between the latter and the curtains at the Conference on “Sogdians in China” held in April 2004, thanks to Étienne de la Vaissière and Éric Trombert who invited me and to l’École Française d’Extrême Orient for the funding of my participation. I am also grateful to Al Dien, Frantz Grenet, Valerie Hansen, Victor Mair, Qi Dongfong, Nicolas Sims-Williams, and Yoshida Yutaka for their helpful comments after my présentation in Beijing and to Sun Fuxi, Yang Junkai, and Wang Weikun after my présentation in Xi’an. Any mistakes that remain are mine.

142

Fig. 1: Yu Hong's sarcophagus, a reconstructed diagram. Taiyuan, Shanxi (Source: Shanxisheng, 2001, p. 29, fig. 4.)

Les Sogdiens en Chine abbreviated as the Tenjukoku curtains) commissioned by Princess Tachibana no Öiratsume to À S commemorate her husband, Prince Shötoku (b. 574), after his death in 622 and since then handed down as a national treasure in Japan see (Fig. 2 on colour plate n° 16). 2 Although many scholars hâve proposed varions ideas for the original composition of the curtains, it éludés us still The anomalous connection between these two seemingly unrelated artifacts is their similar. house“Shape5 a topic that has corne to scholarly attention of late? While obvions in Yu Hong’s sarcophagus, the house-shape is hard to discem from the scattered fragments of the Tenjukoku curtains. In Ôhashi Katsuaki s recon¬ struction, however, the original pair of embroidered curtains appears as closed over a house and by extension, as a soft and mobile house (Fig. 3).4 Assuming Ôhashi’s reconstruction to be correct, what was the signifïcance of the house-shape for the silk curtains? Was there a transmission of the house-shaped form for funerary use from its medium in stone in China to its medium in silk in Japan? If so, how did it get there? And how does Yu Hong’s sarcophagus help us better understand the Tenjukoku curtains? To answer these questions, I am formulating two hypothèses based on two theoretical frameworks. The fîrst concems how we view cultural exchange. Traditionally, ethnocentric historians view the encounter between peoples of two cultures asymmetrically as that between the bipolar3

Fig. 3: Tenjukoku curtains as a temporary house, a reconstructed diagram (Source: Ôhashi Katsuaki, 1995, p. 122, fig. 51.)

2

The terni, “house-shaped,” is coined by Wu Hung (2002). The sarcophagus is kept at the Shanxi Provincial Institute of Archaeology (Zhang Qingjie, 2001a). For color illustrations, see Zhang Qingjie, 2004. The large st group of embroidered fragments, sewn together as a hanging, is stored at the Nara National Museum (Matsumoto Kaneo, 1984, p. 111, no. 94). s Wu Hung, 2002 and Zheng Yan, 200 1. 4 Ôhashi Katsuaki, 1995, p. 122, fig. 51,

Angela Sheng

_

From Stone to Silk

143

ized dominant and the dominated. This adversarial view has given way to more fluid interprétations in the post-colonial context.*5 To understand the development of art, Ticio Escobar argues for an interculturality that allows us to see the encounter between two peoples as one of empowerment where each side chooses cultural practices that

resonate with their own needs and vision, incorporating significant transmission of the other’s culture leading to enriched outcomes.6 The second framework concems our attitude towards art, aesthetics, and rituals. Sociologists and art historians tend to look at art works only for their power to mark social distinctions and to treat them as iconographie objects. Instead, seeing art works as “concrète products of human ingenuity”, the late anthropologist Alfred Gell has considered art as the outcome of a technical process in which artists are skilled. In other words, art is viewed as a component of the technology of enchantaient, which itself is founded on the enchantment of technology, defined as “the power that technical processes hâve of casting a spell over us so that we see the real world in an enchanted form.”7 Seen in this light, art is also Propaganda of the status quo.8 Based on Escobar’s paradigm of interculturality and Gell’s theory of “the technology of enchantment and the enchantment of technology,” two related hypothèses underlie my analysis. Due to some unusual historical circumstances, the non-Chinese elite, both Xianbei and Sogdian, chose certain éléments of earlier Chinese funerary practices to meet their own needs. In so doing, they altered both the content and form of traditional mortuary fumishings and developed the house-shaped sarcophagus. This, in tum, freed the form for other interprétations by, not just the later Chinese imperial elite of the Tang, but also the Japanese Princess Tachibana no Öiratsume and her grandmother, Empress Suiko JB'É' (r. 592-628). These two imperial ladies used the house-shaped curtains that were embroidered with words and images to strengthen their family clan’s political power and ideology at a critical time in Japanese history. I will begin by presenting the Tenjukoku curtains as a historical document of the Asuka enlightenment period (587-645), discussing first, its 400-characters inscription that exceptionally included the full names of the three designers and one Since Edward Said first raised the issue of Orientalism (1978 and 1994), it is a view now largely disputed by most historians, see for example, Nicola Di Cosmo (2002). For a summary of more recent and revisionist views on the dynamics in frontier zones, see Jonathan Karam Skaff (2004, p. 117-19). His thesis that the early seventh-century China-Inner Asia frontier was a “highly dynamic zone of interaction -with overlapping political, économie, and cultural spheres where personal relations and self-interest, rather than idéologies of loyalty, generally determined political allegiance” can be equally stated for the earlier period under discussion here. I wish to thank him for sending me this important publication which appeared after the Conference. 6 See Escobar, 2003. His Statement was presented at the roundtable ‘Toward a Methodology for “Intercultural Art’”, held at Asia Society in New York on 20 February 2003, organized by Michel Oren and sponsored by the College Art Association Cultural Diversity Committee. In China, the issue of intercultural transfer has also attracted scholarly attention. Qi Dongfang proposes an interesting hypothesis based on his study of Toba Xianbei’s gold and silver omaments that the agricultural sedentary society absorbed the form and techniques of décorative arts from the pastoral nomads who, in contrast, adapted the ideological content of the former (2001, p. 571). See also Qi Dongfang, 2004. 7 Alfred Gell, 1992, p. 40-63. 8 Maurice Bloch, 1974 as cited in Alfred Gell, 1992, p. 43.

—

144

_

Les Sogdiens en Chine

Supervisor, and secondly, its style and iconography. I will next summarize how certain ideas and technologies of both the pastoral nomads and the sedentary farming societies came from the East Asian continent and affected the formation of elite groups in early Japan. Significantly, the immigrant clans rallied in the name of Buddhism in factional fighting for power at court and relied on artists, also of immigrant origin, for the material expression of their power. At this point, I will switch to examine the motifs on Yu Hong’s sarcophagus. In particular, the prominent banqueting scene on the sarcophagus can be traced to its actual Staging in Lady Dai’s tomb at Mawangdui in 168 BCE. The etemal banquet in life after death as a theme grew into the populär genre of tomb owners “sitting-under-curtains” often painted on tomb walls in the Eastem Han period (25- 220 CE). I will next show how funerary practice changed during the Period of Disunity (220-589) leading to the emergence of both the stone house-shaped sarcophagus and the stone bed/coucli with screens. The locus of pictorial représentation in tombs shifted back and forth from the surface of walls to that of coffins, screens, couch-beds, and sarcophagi. Because funerary fumishings with strong Sogdian content were all made in the late sixth Century, I will then summarize the special historical circumstances to suggest a few possible reasons. In conclusion, I will retum to discuss the house-shaped form and fonction of the embroidered curtains and propose a new interprétation for their original composition.

The Tenjukoku Curtains

400-characters Inscription The Tenjukoku Shüchô, dated 622, is the earliest embroidery preserved in Japan. Registered as a National Treasure, the original artifact survives in fragments. Most small pièces remain obscure to the public.9 Much more famous is the framed hanging of six disparate fragments sewn together into a 90 cm. square.10 Though belonging to the préfecture, for better conser¬ of Nara in Ikaruga Chügüji temple vation, it is now housed in the Nara National Museum. This hanging contains some fragments that were replicated in 1275 for the purposes of restoring the Buddhist nunnery at the Chügüji temple.” It was then considered a Buddhist relie and labeled as the Tenjukoku mandala. The later history of this artifact, that is, how it might hâve fimetioned as a mandala, does not pertain to this paper and will not be discussed. However, whether the original embroidery embodied Buddhist meaning or not was controversial from its inception in 622. (See next section on Style They are kept at the National Museum in Tokyo; the Chügüji temple, the Hôryüji temple, and the Shôsôin in Nara; the Kawashima Research Institute of Textiles in Kyoto; and a private co Hector. See Ôhashi Katsuaki, 1995, p. 1 and plate 15; Pradel, 1997, p. 46-48. 10 Matsumoto Kaneo, 1984, no. 94. 11 Only the purple silk gauze parts embroidered in the back-stitch date to the earlier Asuka period (552-645); the rest date to the latter Kamakura period (1185-1333). See Pradel, 1997, p. 29-40; Ôta Êizo 1948 as referenced in Pradel, 1997, p. 49-50; and Matsumoto Kaneo, 1984, p. 236-37.

9

Angela Sheng

From Stone to

左

145

and Iconography) The controversy underlies its signifîcance as a historical document for the period of Asuka 飛鳥 enlightenment that started in 587, leading to the Great Refbrms (Taika no kaishin 大化改新) from 645 to 701 when the Tang Chinese System of govemment was officially implemented. Extant fragments show twenty-five words. Four words on each carapace of the four turtles can be seen on the composite hanging; the turtles are interspersed among figures, flora, fauna, and buildings without any obvious connection?2 The first four words,［ K 間人公］, form part of Prince Shötoku's mother5s name: Anahofbe Hashihito no hime］ miko 孔部間人公主. Of particular interest is the word5 be 部, referring to "occupa。 tional group55 and here meaning that of the imperial in-law (gaiseki タト戚 )。 The important concept of occupational group probably came from Korea during the reign of Emperor Yüraku 雄略 (r。457-479).13 The original silk curtains featured an embroidered inscription of 400 Chinese characters bome on 100 turtles, Drawing primarily on the Jögü Shötoku höö teisetsu 上 宮聖德法王帝說 (compiledin the Eishö 永承 era5 1046-53) and other sources, Iida Mizuho has reconstructed the complété 400-characters inscription. 14 The first half details Prince Shôtoku's complex genealogy (Pradel5s translation): "The sumera mikoto or tennö (emperor) Ame-kuni Oshi-hiraki Hiro-niha no mikoto (emperor Kinmei, r. 539-571), who lived in the palace of Shikishima, took the daughter of the Soga Öomi, Iname no Sukune, as wife. Her name was Kitashi hime no mikoto and he gave her the title of ökisaki (empress). She gave birth to Tachibana no Toyoshi no mikoto (emperor Yömei, r. 585-587), and to his younger sister Toyo-mike Kashilci-ya-hime no mikoto (empress Suiko, r. 592-628). He also married a younger sister of Kitashi hime, named Oane no mikoto, who gave birth to Anahobe no Hashihoto no hime miko. A son of of the Shikishima palace sovereign (Kinme's son), named Nunakura no Futo-tama-shiki (emperor Bidatsu, r. 572-585) married his younger sister Toyo-mike Kashi-kiya-hime no mikoto, and gave her the title of ökisaki. She resided at the Osada palace and gave birth to a prince named Owari. Tachibana no Toyoshi married his half1sister princess Anahobe no Hashihito and gave her the title of ökisaki. She lived at the Ikebe palace and gave birth to Toyotomimi no mikoto［Prince Shötoku］, who married the daughter of the Owari prince, named Tachibana no Oiratsume."

The genealogy highlights the prominence of the Soga clan as the imperial in-law for several générations to the 29th Emperor Kinmei 欽明, the 31st Emperor Yömei 用明, and the 33rd Empress Suiko 推古天皇, especially in view of the Soga clan's support for Buddhism and the imperial patronage of Buddhist art. The genealogy also emphasizes the close relationship between Princess Tachibana no Öiratsume 多至波奈 大女良 B (altematively, 橘媛), the instigator of the embroidered curtains, and the ruling Ofthe other twelve words on the composite hanging,［千時 多 至］,［皇前日 ］ ,［佛是真玩］, the first three words of the last group meaning "Buddha is the Truth^ refers to Prince Shötoku5s belief in Buddhism. Nine other words are legible on some fragments kept at the Shösöin 正倉院： one turtle bearing the four words: ［利令者掠］and five small pièces with one word each 奈,甘,居,佛,娶 (Pradel 1997,p. 12). 13 Delmer M. Brown in Brown (ed.), 1993, p. 101 and 145. 14 Iida Mizuho 1966, p. 33-42 as cited in Pradel, 1997, p. 14-16. For the 400-characters text, see Öhashi Katsuaki, 1995, p. 145. 12

146

Les Sogdiens en Chine

clan headed by her patemal grandmothers Empress Suiko who was reigning when the princess5 husband, Prince Shötoku, died in 622 J5 Although Princess Tachibana ロ〇 Öiratsume was only one of four consorts to Prince Shötoku, the other three are not mentioned in the genealogy.15 16 The first half of the inscription, in fact3 ends with her

name.17

The second half of the 400-characters inscription documents the commissioning of the embroidery. To summarize (all italics below are mine), it States that Prince Shôtoku's mother passed away in the twelfth month of year 621 and himselg in the middle of the second month in 622. To have lost both in such a short time, Princess Tachibana no Öiratsume expressed her grief to her patemal grandmother, the reigning Empress Suiko, The princess remembered the words of her deceased husband: "This world is empty; the only truth is Buddha?518 Recalling how much he had valued the the princess feit sure that he was to be reboni into the heavenly kingdom of longevity, for which, however, she lacked an image. The princess wished to see him 加 the 川砂 he lived.19 Empress Suiko was deeply moved and ordered the court ladies to embroider a pair of curtains. The designs were drawn up by three men, Yamato no Aya no Maken 東漢末賢, Koma no Kas ei 高麗 加西 溢, and Aya no Nukakori 漢 奴加己利 and a fourth man, Kurahitobe no Hata no Kuma 掠 部 秦 久麻, supervised the whole production. The 400»characters inscription ends with these four names5 in effect, as signatures to the embroidered curtains.20 Three features of this account merit some discussion. First, Prince Shötoku's recitation of the Buddhist truth reveals his keen interest in this imported religion. He was femous for both having written commentaries on three Buddhist texts then populär in China: the Vimalakirti 維摩經 (Skt. Vimalakirti-nirdesa-sutra), the Shöman-gyö 勝 鬟 經 (Skt.Srimala-devi-simhanada-sutra) and the Lotus sutra 法 華 經 (Skt. Saddharma-pundarika-sutra) and having lectured on them at court.21 As we shall see further below, his court lectures might have inspired the design of the curtains. Second, the single word,% 法 (law),in the 400-characters inscription encapsulated his intention of emulating Chinese statecraft. Drawing inspiration from the Chinese administration,22 Prince Shötoku drafted two reforms in 603 and 604: respectively, the Twelve Caps and Ranks (冠位十二階) and the Seventeen Injunctions (憲法十七條). The first of latter prescribed ministerial behavior based on the Confucian concept of 15

Some historians consider that Princess Tachibana had married Prince Shötoku to strengthen the political alliance between the latter and Empress Suiko. See Pradel, 1997, p. 20 16 The other three were Ujinokaidalco no Himemiko, Tojiko no Iratsume, and Hokikimi no Iratsume who died the day befbre the Prince (Pradel, 1997,p. 20 and 192). 17 Line 11 in the 400-characters text, see Öhashi Katsuaki, 1995, p. 145. 18 Shijian xujia, weifo shizhen 世間虛假唯佛是真. 19 The text in Chinese words were 欲觀大王往生之狀. The ambiguous wangsheng 往生 can be translated as “going to the land of rebirth” in a Buddhist context as Pradel points out (1997,p. 17, n. 11). However, this seems far stretched as does Pradel's translation of the words as "the after life world. 20 The four names take up more than one out of 20 lines, see Öhashi Katsuaki, 1995, p. 145. Pradel, 1997, p. 15-17. 21 Pradel, 1997, p. 19. Stanley-Baker, 1984, p, 31. 22 After Empress Suiko sent a mission directly to the newly unified Sui court with its capital at Chang'an in 600, he gained much more information about China than previously (Tamura Enchö, 1980, p. 98). 5,

Angela Sheng

_

From Stone to Silk

147

obedience to the emperor.23 The second injunction commanded his subjects to revere Buddha as the supreme object of worship and his third, built above the first two, placed traditional imperial authority at the top, near Heaven.24 Both reforms held far-reaching implications for the Japanese court and govemment. Clearly, Prince Shötoku aimed to shift the imperial authority away from the kami-worshipping pro-Shintoism clans. Instead, he wanted their absolute obedience as preached by Confiicianism. The role of

Buddha was perhaps interpreted as more universal than the local kami spirits for this purpose and Buddhism seen as a way of encompassing varions idéologies.25 Third, the full citation of the three designers and one Supervisor by name, just like that of imperial family members, in the 400-characters inscription signais their exceptional social status. That and their intimacy with the imperial family can only be appreciated in the context of how the cultural transmission from the East Asian continent gradually changed tlie Japanese elite. But, their sources of inspiration become apparent with an analysis of the curtains’ extant imagery, to which we next tum.

Style

Iconography

The original composition of the Tenjukoku curtains remains elusive. Because the stylized lotus is clearly associated with Buddhism, its récurrence on the fragments has prejudiced some scholars to imagine the whole work as totally Buddhist in nature. In particular, the large lotus with the nascent figure has been viewed as a soul’s re-birth to {bÉfe flower-birth=transformation). This has led to one life after death (huasheng populär hypothesis that the central theme of the original Tenjukoku was Prince Shötoku’s re-birth into the Mahayanist paradise, the Pure Land of Amida (Skt. Amitabha) or Buddha of Boundless Light.26 Despite the limited evidence, Öhashi Katsuaki also subscribes to this view. Discounting this bias, his work based on thirty years of research serves as a valuable reference. Following his désignation, the six disparate fragments are labeled from the right to left as A and B in the top, C and D in the middle, and E and F in the bottom

" Delmar Brown in Brown (ed.), 1993, p. 180. " Sonoda Köyü in Brown (ed.), 1993, p. 387. " The Chinese Southern Liang Emperor Wu

(r. 502-548), passionately named Xiao Yan supported Buddhism that was imported from India as mediated through Central Asia. Significantly, he did so without ever dispelling the native Chinese Daoism (Zhou Yiliang, 1997, p. 357-63). Around 541, he sent Chinese monks, painters, and artisans with Buddhist sutras, practices, and paraphemalia to Paekche at King Söngmyöng’s request (Liangshu , juan 54, p. 805; Lou and Zhang, 1998, p. 90-92; Zhou Yiliang, 1997, p. 555. Stanley-Baker, 1984, p. 38) Soon after in 552, King Söngmyöng, for the help that he had received earlier from Japan to fend off threats from Koguryô and Silla, sent the 29“ Emperor Kinmei (r. 540-571) some Buddhist sutras, banners, and a gilt-bronze image of Buddha. These gifts marked the official arrivai of Buddhism in Japan. Perhaps Emperor Wu provided a role model for Prince Shötoku to contain both Buddhism and indigenous Shintoism? " Pradel, 1997, p. 73-75. Others have hypothesized that the Tenjukoku represented the Tusita Heaven, the Pure Land of the Bodhisattva Miroku (Skt. Maitreya), or that of Ashukubutsu (Skt. Aksobhya), or that of Ten Directions, or the Heavenly Realm (See Pradel, 1997, p. 76-105).

Les Sogdiens en Chine

148

register (see Fig. 2 on colour plate 16).27 Like the lotus, ail other extant motifs were either foreign to Japan when the curtains were commissioned in 622 or strongly influenced by cultural practices prévalent on the East Asian continent at an earlier time.

The Words-bearing Turtles The most signifîcant motif of the inscription-bearing turtles is clearly visible: two in B, one each in C and D. The ancient Chinese believed that the divine turtle supported on his back an island of immortality. It was understood as the island of Penglai 蓬來 in the eastem seas during the Eastem Han dynasty。Yet5 contemporaneous tomb carving also shows the divine turtle holding up Mount Kunlun 昆崙 in the west where the Queen Mother allegedly resided in her paradise?8 The ubiquitous coding of Taoist immortality in the turtle is further enhanced by another ancient Chinese belief that the divine turtle mediated between the humans and deities by carrying messages on his back Hence, the turtle was associated with words, evident in such artifacts as seals of the Han dynasty, inkwells of the Six Dynasties, and stelae (bei 碑) of the Tang

dynasty.29 The last type of stone artifacts, usually carved with inscriptions and sometimes further embellished with motifs, might hâve inspired the designers of the Tenjukoku curtains. Several extant examples that date to before 622 are found in and around Nanjing 南京, the capital of Southern Liang with which Japan was in communication since 522. These stelae are paired vertical stone slabs, each in the shape of a gui 圭 (flat scepter) standing atop a turtle, that flank the paths to the tombs of Southern Liang imperial family.30 Along the path leading to the tomb of Xiao Xiu 蕭秀 (d. 519),posthumously Prince Kang of Ancheng 南梁 安城康王, stand the remains of two sets of paired stelae supported by four turtles, paired sacred pillars (shendaozhu ネ申道柱), and paired male lions with wings carved in the round, all in stone.31 Though the stele's inscription of his biography is no longer legible, the History of Liang reveals that he was a brother of the famous Buddhist patron Emperor Wu 南梁武帝,named Xiao Yan 蕭衍 (r. 502-548).32 Similarly, two stone lions, two sacred pillars and two stelae atop two turtles guard the path to the tomb of another brother, Xiao Hong 蕭 宏 (d. 526),posthumously Prince Jing Hui of Linchuan 南梁臨川靖 蕙王ピ Again, the stele inscription is damaged beyond readability. However, still visible are the two intertwined dragons atop the west stele and on its north face, eight heavenly beings and beasts, including the red bird and

27

Ôhashi Katsuaki, 1995, p. 28, fîg. 3: ail six fragments together; p. 29, fig. 4: A; p. 35, fîg. 7: B; p. 40, fig. 9: C; p. 47, fig. 14: D; p. 54, fîg. 19: E; and p. 56, fig. 20: F. 28 For the Y inan tomb carving, see Wu Hung 1989, p. 120, fig. 49 as cited in Pradel 1997, p. 137-138. 29 Ôhashi Katsuaki, 1995, p. 110. 30 For an illustration, see Yao qian and Gu Bin 1981, col. pl, 3. 31 Yao Qian and Gu Bing (eds.); 1981, col. pl. 5 and nos. 58, 69 and 70. The lions measure about 3 m in length and width. The intact stelae measure about 4 m high and 1.5 m wide. 32 Liangshu, juan. 22, p. 342-45. 33 22, p. 339. Yao Qian and Gu Bing (eds.), 1981, col. pL 3 and nos. 55-57. Liangshu,

Angela Sheng

From Stone to Silk

149

blue dragon (of the four directional animals5 see below) each in a square?4 Bearing words5 the stone Stele atop the turtle was a critical element in the aboveground funerary design for the Southern Liang imperial family. Though Emperor Wu's stone stelae did not survive, their existence can be inferred from one extant stone sculpture of tianlu 天祿,an auspicious deer-like beast with wings, also male, in front of 35 his mausoleum.* Extracted from the Chinese aboveground funerary formula, the words-bearing turtles were flattened onto the surface of the curtains as an overlay atop another scheme of cohérent motifs“ (I will retum to discuss the meaning and placement of the 100 turtles on the Tenjukoku curtains in the final section.) Other Motifs Some motifs on the extant Teiijukoku fragments originated obviously in West Asia and South Asia. They arrived Japan having first passed through Central Asia and then China. For example, borders decorated with the Indian-Buddhist spade-shaped "treasure-gem" (baozhu^en 寶珠文) in A and E, and the Sasanian inspired "continuons ^Qdx^ilianzhuwen 連珠乂) motif in E and F. The seated figure in D is also foreign and syncretic: he has a large monstrous head5 two wings of a heavenly being, a naked and muscular body covered with a skirt like an Indian ya依q's loincloth, and holds a staff in his right hand that possibly identifies him as the Chinese thunderbolt-deity in the sixth

Century.36 Both the sartorial and architectural style in the embroidery suggest that these unusual motifs arrived Japan via the Korean peninsula. For example, the women's pleated skirts in C and D resemble those wom by female figures painted on walls in the Takamatsu tomb 高松塚 at Asuka village near Nara, dated to the end of the seventh Century or early eighth Century.37 These skirts recall those wom by female figures painted in the Koguryö 高句麗 tomb of Sunsari 修山里 dated to the fifth Century in South P'yong'an 平安 province in the middle of the Korean peninsula.38 Similarly, two men in E are depicted wearing pants with a cord tied around the knee so that they could walk without tripping. This use of the cord was just like that shown on a Japanese haniwa 直輪 figure excavated from Saitama 琦玉 préfecture dated to the Burial Mound (Kofun 古墳) period (300-600). It was a habit practiced by the Xianbei warriors of the Northern Wei (386-535) and Northern Zhou (557-581). 39 The transmission of this costume would be consistent with that of other Kofün artifacts, such as gilt-bronze and bronze accoutrements of the mounted archer. The latter were Yao qian and Gu nin, 1981, no. 56. Yao Qian and Gu Bing (eds.), 1981, col. pl. 4 and nos. 41-49. The stone sculpture measures about 3 m high and long. 36 For a related discussion on the Chinese représentation of winged deities, see Rosalind Bradford, 2002, p. 193-95; for a comparison with Indian yaksas, see Pradel, 1997, p. 175; and for the thunderbolt-deity, see Susan Bush on " Thunder Monsters” referenced in Pradel, 1997, p. 176. 37 Ohashi Katsuaki, 1995, p. 48, fig. 17. 38 ChuYonghong, 1985, p. 13-15, pis. 105, 107,and 108 as cited in Pradel, 1997, p. 117. 兆 Ohashi Katsuaki, 1995, p. 48, fïg. 16. See also a pottery figure (HM555:5) unearthed from a Northern Wei tomb in Luoyang in 2001 (dated 522),see Luoyang shi, 2002, p, 11, fig, 4. '

35

150

Les Sogdiens en Chine

similar to those of Contemporary Silla culture in the southwestem peninsula, stylistically related to the arts of Central Asian pastoral nomads.40 With regard to the architectural style, the hip-gabled roof of the bell pavilion in E resembles that of an offering shrine dated to about 650, the Tamamushi no zushi (beetle-wing shrine).41 Su Bai traces the hip-gabled roof on both the bell pavilion and on the beetle-wing shrine to a Chinese prototype found on pottery houses and pictorial bricks excavated from tombs dating to the Eastem Han dynasty (25-220). Many such roofs appear in mural paintings in the Buddhist caves of Mogao MS at Dunhuang dating to the Northern Zhou (557-581), for example, Cave N° 296.42 It kingdom (18 is also visible on a pictorial brick from Korea of the Paekche BCE-663 CE) dating to the seventh Century.43 In addition to the above motifs, several others also came from China througli Korea. The moon disk in fragment B, within which a hare stands beside a cassia tree and is pounding the élixir of longevity inside a tall jar, was commonly represented in opposition to the sun disk, within which usually stands a bird, on the ceilings of Koguryô tombs, such as the Ssangyongcheong (Twin Pillar tomb of the fifth ÄJgll and the the Century Jinpari sixth Century as well as in some tomb number one of Paekche tombs.44 The red bird, also in fragment B, is identified as the Red Bird of the South. One of the four directional animais (sishen 29^), it and the other three: the Blue Dragon of the East, the White Tiger of the West, and the Dark Warrior of the North assumed increasing importance in the mural décoration of Koguryô tombs in the fifth and sixth centuries and that of Paekche tombs in the late sixth Century.45 Stylistically, the red bird in fragment B is also related to the red bird carved on the north face of the stone stele of the Southern Liang Prince Jing Hui of Linchuan Xiao Hong (d. 526). This would not be surprising since the Southern Liang court dispatched artists to Paekche in the fifth and sixth centuries. Just like the cosmic symbols of the sun and moon disks, the four directional animais, as well as other standard décorative vocabulary populär since the Eastem Han, the stylized lotus was rampantly used to embellish the walls of tomb chambers in China. This fùnerary fashion spread to Koguryô by the early sixth centuiy.46 The lotus buds and blossoms in fragments A, B, C, and D could easily hâve connoted Buddhism without necessarily designating a Buddhist paradise.

* Stanley-Baker, 1984, p. 24. 41

Made of camphor and cypress wood indigenous to Japan, it was once inlaid with the iridescent wings of the tamamushi beetle, see Stanley-Baker, 1984, p. 34, figs. 18 and 19. Emest Fenollosa considered it Korean art of the Asuka period (Evelyn McCune, 1 967, p. 69-70). 12 Su Bai, 2002, p. 338-40. 43 For an illustration of the brick excavated from Puyo, South Ch’ungch’ong province, see Evelyn McCune, 1967, p. 102, fig. 21. For routes of transmission from the East Asian continent to the Korean peninsula and Japan, see Mizuno Yü's diagram (1972, p. 160). m Pradel, 1997, p. 128. See also Yang Hong, 2000, p. 166-185. 45 Pradel, 1997, p. 132-33. *HeXilin,2001,p. 196.

Angela Sheng

From Stone to Silk

151

Following those scholars who have studied Japanese funerary rituals/7 Maria del Rosario Pradel proposes that the Tenjukoku Shüchö not only constituted part of Prince Shötoku’s funerary paraphemalia but also recorded textually and visually the funerary rituals performed for Prince Shötoku and his mother, Empress Anahobe no Hashihito, after their death.*48 While Pradel’s first point makes sense, her second point is illogical. It would seem redundant to embroider a non-Buddhist funerary ritual on the curtains as part of the ritual. Were the ritual Buddhist, then at least the répétition could be understood as a way of increasing merit/9 To summarize the discussion on the 400-characters inscription and the style and iconography of the Tenjukoku curtains, all evidence point to the fact that the designs were heavily based on Chinese models. The formula of the inscription (content) and the vocabulary of the imagery (form) came to Japan either via the Korean peninsula, or direetly, or both. They were somehow made available to the three designers and one Koma no Kasei and Aya Supervisor:Yamato no Aya no Maken no Kuma their no Hata All Kurahitobe and Nukakori no immigrants.50 While have we names suggest that they descended from Korean-Chinese little information on these four individuals, their common background undoubtedly led to their Professional responsibilities fbr the Tenjukoku curtains and hence, the commissioner Princess Tachibana no Öiratsume5 but how? To answer this question, we next examine the impact of early cultural influences from the East Asian continent on the formation of elite groups in Japan.

Early Cultural Transmission from China to Japan via Korea51

Major cultural influences of the East Asian continent arrived in Japan in the early centuries through the Korean peninsula. 52 When and why the Japanese absorbed Continental cultures of both the pastoral nomads (Xiongnu and Xianbei) and the sedentary farmers (ethnie Han) reflected the interplay between Japan’s domestic conditions and Japan’s foreign relations with the nomads, Koreans and ethnie Han Chinese. Certainly, the continental cultures affected the formation of the Japanese elite groups in the fifth and sixth centuries, especially the immigrant clans, such as the Soga 1,7

Notably Ruch 1992, Saito 1987, and Wada 1995 as cited in Pradel, 1997, p. 9-10, p. 187-215. This is the central thesis of her dissertation submitted to the University of California at Los Angeles ” (1997). Sheng, 2003. 50 Pradel, 1997, p. 17-18 and 28. 51 Many scholars have published on this topic in both Asian and western languages. I have benefited greatly from several secondary sources, notably Zhou Yiliang’s Wei Jin Nanbeichao shilunji (1997), Tamura Enchö’s Kodai Chosen Bukkyö to Nihon Bukkyö (1980), The Cambridge History of Japan, vol. I5 Ancient Japan edited by Delmer M. Brown (1993) and Evelyn McCune’s The Arts of Korea, an Hlustrated History (1967). 52 (18 (37 BCE-668 CE) in the north, the Paekche The Three Kingdoms were the Koguryö BCE-663 CE) in the Southwest and the Silla ÿrO (57 BCE- 668 CE) in the southeast. In addition, there was the Kaya jQlß Federation (42-562) wedged between Paekche and Silla. Its Southern coast on the Korea Strait made it easy fbr Japan to dominate it until Silla absorbed it in 562.

Les Sogdiens en Chine

152

clan, and the hereditary craftsmen that served them, such as the Supervisor of the Tenjukoku curtains. Just as the Xiongnu and Xianbei descended southward to conquer northem China after the Han dynasty collapsed in 220, they also invaded the northem Korean peninsula and even Japan.53 Excavated artifacts show that their equestrian culture was much appreciated by the local elite. Some royal crowns unearthed from Kyongju (of Silla culture), dated to the fifth and sixth centuries, were worked from thin sheets of gold in techniques already used in the first Century in Bactria at Tillya Tepe (Shibarghan in northem Afghanistan today).54 Similarly worked gilt-bronze fittings also embellished saddles discovered in a tomb near Koryöng. 55 These metalworking techniques were later transferred to the making of Buddhist statues and paraphemalia in bronze. Immigrant artisans with this specialized laiowledge came to the attention of the rising immigrant power groups with an agricultural backgrotmd for reasons that we will see below. The second impact of the Chinese Han dynasty’s collapse in 220 was the southward spread of the Chinese agrarian culture on the Korean peninsula, eventually extending westward to Japan. When the Koguryô forces finally crushed the last Chinese colony at Lelang (near Pyongyang in 313, many descendents of ethnie Han Chinese who had first moved to the northeastem Korean peninsula since the first Century BCE fled south to Paekche. From there they precipitated the transmission of agricultural practices, ceramic and textile arts, the Chinese language, Confucianism, and Buddhism to Japan. In particular, the advanced irrigation System that the immigrants to Japan brought from Paekche enriched the Japanese landowning elite. Simultaneously, there emerged new powerful clans of immigrant origin, for example, the Soga clan.56 The increase in new power groups probably accounted for the 21st Emperor Yüraku (r. 457-479)'s greatly expanded use of kabana (clan title) appointments.57 These hereditary titles, along with the commensurate ranks (kurai marked the holders’ status and proximity to the ruler. The appointées were leaders of stratifïed groups, such as clans (uji [^,) that dominated the lands and people of entire régions, and occupational groups (6e who included the imperial in-laws and those in charge of painting, weaving, 58 and saddlery for the ruling elite.59 53 On the theory that Japanese kings descended from continental nomads, see Egami Namio, 1967 cited in Wang Weikun, 2002, p. 10. 54 Bailey, 1994 and Pak, 1988 referenced in Portai, 2000, p. 62. For the iconography of the tree-shape in the Crown, see Ishiwata Mie, 1987. 55 For an illustration of the saddle-fittings, see Kim Wong-Yong, 1986, p. 358, pl. 5-10. 56 The Soga clan issued from the Kawachi région that was enriched by the early irrigation projects, see Delmer M. Brown in Brown (ed_), 1993, p. 160. 57 Delmer E. Brown in Brown (ed,), 1993, p. 136-137. 58 Among them were the Hata weavers who wove fancy Chinese pattemed silks on complex looms. The pattemed polychrome silks (jin ) and silk twills (ling required patteming mechanisms not found on ordinary two-shed looms used for weaving plain cloths and silks. Because the luxury twill silks are called ling in Chinese but aya in Japanese and that the two designers of the Tenjukoku curtains had “aya” in their names, they might hâve descended from the se specialized weaving families. On the weavers, see Fujii Shuichi, 1986, p. 3-2. On the complex looms, see Sheng, 1994-95. On the name “Hata” as related to the people of “Qin,” see Wang Weikun, 2002, p. 2-3.

*

Angela Sheng

From Stone to Silk

153

The Supervisor of the Tenjukoku curtainシ Kurahitobe no Hata no Kuma イ京部秦久 麻, probably descended from the "Hata” weavers who supplied luxury silks to the ruling elite, while also belonging to the saddle-making occupational group, Kuratsukuribe 鞍部 F This clan name was awarded by the 26th Emperor Keitei 繼 體 (r. 507-531) to a Chinese named Sima Dadeng 司馬 達 等 (altematively 司馬達 止,Jap. Shiba Tatto or Tachito) who came from the Southern Liang state to set up a Buddhist temple in Nara despite his feilure to spread Buddhism immediately.5961 We will retum to discuss his descendents below. The pro-Buddhism Soga Clan

The contest for power between the new clans with the old elite sharpened ostensibly over a religions divide. In contrast to the locals who prayed to their Shinto kami (deity-spirits) for good crops, the immigrant clans relied on their advanced technology and more easily embraced the imported Buddhism after its official introduction in 552.62 As the practice of Japanese religions is syncretic in reality, the traditional and linear categorization of Shintoism versus Buddhism is best not to be considered as a clear marker of different belief Systems, but rather as a way of separating political power groups back then. During the fïerce struggle for temporal authority leading up to the Great Taika Reforms (645-701), the deities were identified in opposition to each other, mirroring their earthly supporters.63 The conservative pro-Shintoism Mononobe 物部 clan fought against the new pro-Buddhism Soga 蘇我 clan ever since 552 when Buddhism was ofïïcially introduced to the court of the 29th Emperor Kinmei 欽明 (r. 540-571). The Soga clan finally gained the upper hand in 585 when the 30th Emperor Bidatsu 厳達 (r. 572-584) died. At his temporary interment (mogari 殯), the two sides exchanged insults at first but then resorted to the use of military force. In the skirmish that ensued Mononobe Moriya 物 部守屋 the Great Chieftan died. Soga no Umako 蘇我馬子 the Top Minister, now the victor, propped to the throne Prince Anahobe as the 31st Emperor Yömei 用 明 (r. 586-87)ア The fourth son of Emperor Kinmei, Yömei had for his mother, a daughter of Soga no Iname 蘇我ネ 目 (Umako'sfather). Yömei himseげ married another of Iname's 59

Such holders also controlled royal estâtes (miyake 屯倉),provinces (kuni 國)and districts (agata 縣), see Delmer E. Brown in Brown (ed.), 1993,p. 28. Stanley-Baker, 1984, p. 30. 60 Pradel, 1997,p. 17, note 13. 61 Lou and Zhang 1998, p. 104. In 552, King Söngmyöng of Paekche, for the help that he had received earlier from Japan to fend off threats from Koguryö and Silla, sent the 29th Emperor Kinmei 欽明 (匕 540-571) some Buddhist sutras, banners, and a gilt-bronze image of Buddha. These gifts marked the official arrivai of Buddhism in Japan. They also reveal that the kind of Buddhism to arrive Japan initially came from the Chinese Southern Liang court. That the 400-characters inscription on the Tenjukoku curtains begins with a reference to Emperor , Kinmei has been interpreted to underscore the curtains Buddhist content (Pradel, 1997, p. 18). The date is 552 according to Nihon shoki but 538 according to Gangö-ji engi, see Sonoda Köyü in Brown (ed.), 1993, p. 372; see also Delmer E. Brown in Brown (ed.), 1993, p. 171-72. Stanley-Baker, 1984, p. 30. 63 George J. Tanabe, Jr., 1999, p. 154 andp. 168. 64 Delmer Brown in Brown (ed.), 1993, p. 30 and p. 161.

Les Sogdiens en Chine

154

daughter3 Anahobe no Hashihito, who gave birth to four sons5 the first being Prince Shôtoku (see the 400-characters inscription). After Yömei’s death in 587, Umako installed his nephew, Kinmei’s twelfth son with another Soga wo man, as the 32nd (r. 588-592).65 Umako s decision reflected his awareness of the Emperor Sushun need to abide by the single line of priestly kings’ entitlement to imperial authority. He continued to control the court as an impérial in-law,66 Umako’s support of Buddhism, however, irritated Emperor Sushun, who schemed to remove him. But, Umako heard of the plot and had Susun assassinated instead. Umako then set up Sushun ’s sister, his niece, as the 33rd Empress Suiko (r. 593-628), with the nineteen-year old Prince Shôtoku as regent. As we hâve seen above5 Prince Shôtoku had hoped to transfer imperial authority away from the pro-Shintoism faction through the adoption of Buddhism in order to establish the mandate of heaven as practiced in China. Although the Soga clan"s ultimate goal differed slightly from that of Prince Shôtoku, both the prince and the new powerfïil clan supported Buddhism for their own interests. 5

The Saddle-Making Occupational Group and Buddhist Art

This clan Retuming to the saddle-making occupational group, Kuratsukuribe While the Chin¬ (Jap. Shiba Tatto) in 522. name was given to Sima Dadeng ese sumame of Sima denoted “grooming horse,” his new occupational name, Kuratsukuribe, emphasized the making of saddles, which often had bronze accoutre¬ ment as embellishment. The skills required for the latter would be equally usefal in making Buddhist statues and paraphemalia in bronze.67 Shiba Tatto’s descendents played important rôles in support of the pro-Buddhism Soga clan and Prince Shôtoku. First, Shiba Tatto’s son, Tasuna, was assigned to make an image of Yakushi Nyorai (Skt. Bhaisajyaguru), the Healing Buddha to help Emperor Yômei recover from his illness as he had wished. But his death in 587 postponed the work. Tasuna finished it in 607 and set up the bronze statue in the Golden Hall of the temple of began Buddhist Hôryüji.68 Second, Shiba Tatto’s daughter, Shiba Shima studies at âge seventeen and was ordained nun Zenshin. She, in a délégation sent by Soga Umako, went to receive formal Buddhist studies in Paekche in 588.69 When they retumed in 590, they took up their Buddhist work and résidence in the newly built temple,70 Evidently one female member of the Kuratsukuribe clan was Toyora literate, well educated in the Buddhist literature and liturgy, and close to the court. Could she not hâve guided Princess Tachibana no Öiratsume later in the writing of the 400-characters Tenjukoku text? Third, Shiba Tatto’s grandson, Kuratsukuribe no Tori, or most famously known as 65

Delmer Brown in Brown (ed.), 1993, p. 161. Papinot, 1964, vol.2, p. 597. Delmer Brown in Brown (ed.), 1993, p. 29-30. 67 For a stylistic comparison between the gilt-bronze saddle and a gilt-bronze reliquary, see Kim Wong-Yong, 1986, p. 360, pl. 5-13. 68 Based on the inscription on the back of the statue’s halo, see Noma Seiroku, 1978, p. 40, fig. 24. 69 Tamura Enchö, 1980, p. 39-44. 70 Tamura Enchö, 1980, p. 44. 66

111 Angela Sheng

.一

From Stone to Silk

155

Tori Bushi 鳥 佛師, directed the making of a great Buddha that was set up in the temple of Gangö-ji 元興寺 in 606, as well as an embroidery. Unfbrtunately they did not survive and are only known in texts. However, both works so pleased Empress Suiko that she awarded the sculptor the prestigious fifth-grade rank with attendant lands. After Prince Shötoku died in 622, Tori made the bronze Shaka Triad in 623 to honor him.71 Kept at the Golden Hall of Höryüji, this is the most famous early Buddhist sculpture in Japan. That the prototype can be traced to a stone Buddha, dated to the late fifth Century, in a cave at the Xixiasi Temple in Nanjing confirms Tori's artistic héritage fiom his grandfather, Shiba Tattto.72 Since Tori, had already designed one embroidered hanging3 could he not have guided Kurahitobe no Hata no Kuma, the Supervisor of the Tenjukoku curtains, in their design? Certainly both could have received such news of the latest fashions in the Sui capital as funerary rituals and paraphemalia (as encoded in the Yu Hong sarcophagus) 利, who went with ambassador from another Kuratsukuribe clan member, Furi 鞍 Ono Imoko 小野 妹子 on diplomatie mission to China in 607 and 608.73

The Tenjukoku Supervisor and Designers

In summary, while we lack précisé records on the Tenjukoku Supervisor, Kurahitobe no Hata no Kuma, circumstantial evidence strongly suggests that he belonged to the elite saddle-making group, of which the most famous was the sculptor and embroidery-designer Tori. We can also surmise that the Tenjukoku Supervisor was either familiär with Buddhist practice and Chinese funerary rituals or had close contact with those who did. The three designers, Yamato no Aya no Maken 東漢末賢, Koma no Kasei 高麗 加西 溢, and Aya no Nukakori 漢奴加己利 were of immigrant origin, probably distant descendents of Lelang Han Chinese. Given the history of occupational groups, they would have been conversant with funerary art traditions in Paekche, if not of China. What imagery could they have designed for the Tenjukoku curtains? The 400 characters inscription States that Princess Tachibana no Ôiratsume had wanted to see the way in which Prince Shötoku lived. If so, why? In part, this depended on the fiinction of the curtains. The curtains had to be closed for a full reading of the text and imagery and they would if they were to serve as a house or a cofïïn-pavilion for the deceased at the temporary interment (mogari 殯). Because the pro-Shintoism and pro Buddhist clans fought at Emperor Bidatsu5s temporary interment in 587, a similar conflict might have risen at that of Prince Shötoku. Given that the latter was trying to shift the basis of imperial authority, what message could the embroidered imagery convey? Before answering that question, let us now study the motifs on Yu Hong's sarcophagus for some leads. 71

See Paine and Soper, 1981, p. 31, fig. 2. Seichi Mizuno has been compared to a similar stone triad dated 502-523 atthe middle Binyang cave at Longmen (See Swann, 1979, p. 41-42, figs. 12 and 13). 72 Su Bai 1989 referenced in Lena Kim (1998, p. 255, note 5). This style, initiated in China's south, was also transmitted northward to Yungang and Longmen in the late fïfth and early sixth centuries (Yang Hong, 1983 cited in Lena Kim, 1998, p. 256, note 7). 73 Chang Lan, 2001, p. 184. Tamura Enchô, 1980, p. 100-101.

Les Sogdiens en Chine

156

The Yu Hong Sarcophagus

The non-Chinese imagery carved on the Chinese house-shaped stone sarcophagus of Yu Hong and his wife hâve attracted the attention of many scholars since the tomb was excavated in 1999 near Taiyuan in Shanxi.74 Yu Hong’s epitaph in Chinese reveals an elite non-Chinese background and illustrions career. His grandfather was a tribal leader of the pastoral nomads sumamed Yu ^、. Related to the Ruanruan or the Avars who dominated East Central Asia from the fourth to the sixth Century, Yu Hong’s ancestors were familial with the Sogdians.75 When Yu Hong was bom in 534, his father was already an ambassador at the Toba Xianbei rulers ’ Northern Wei court (386-535). At âge thirteen, Yu Hong began his own diplomatie career in an auxiliary capacity on missions to Persia for the Eastem Wei (534-550). He then worked for the rulers of the Northern Qi (550-557), the Northern Zhou (557-581), and the Sui (589-612), acquiring fourteen titles in total by the time he died in 592. Not a Sogdian, he was in charge of the Sogdian Community late in life as Jianjiao Sabaofu His ancestral contact with the Sogdians in their homeland and his official administration of them in China underlie the unusual combination of the Sogdian iconography in content with the Chinese architectural style in form. The marble sarcophagus is shaped in imitation of a wooden palace. It features a hipped roof with a central ridge—a style emulated for the design of the bell pavilion on the embroidered Tenjukoku curtains (dated 622). As the tomb had been previously disturbed, probably sometime in the late Tang, no door was found at the time of the excavation.77 Altogether nine large slabs with carved designs stand above a rectangular base. Also carved on the exterior, it rests on eight lion-supports like two feet on each of the four sides. The whole structure measures 217 cm in height, 295 cm in length, and 137 cm in width.78 Jiang Boqin argues that the nine panels follow a cohérent pictorial program, outlining the process by which the deceased can attain immortality. Briefly, it begins with panel (1) showing the deceased riding a horse to see Mithra pictured in panel (9), 74

An excavation report appeared in 2001, see Shanxisheng 2001. For discussions on the imagery, see Zhang Qingjie, 2001a and 2004, Jiang Boqin, 2001 and 2004. Conceming the tomb epitaphs, see Zhang Qingjie, 2001b, Rong Xinjiang, 2001, Lin Meicun, 2002, Luo Feng, 2002, and Zhou Weizhou, 2004. 75 For details on the relationship between Yu Hong’s ancestors and the Ruanruan, see Luo Feng, 2002, p. 79, Mallory and Mair, 2000, p. 99 and p. 321. Zhou Weizhou, 2004, p. 248-50 argues that the Yu tribe was originally a brauch of the Da Yuezhi. On the relationship between the Da Yuezhi and the Sogdians, see Benjamin ’s “The Yuezhi Migration Craig on the website and Sogdia” posted http://www.transoxiana.com.ar/Eran/Articles/benjamin.html in March 2003, thanks to Victor Mair’s notice. The exact relationship between the Avars and the Sogdians is outside the limit ofthis paper. 76 The position of Sabao in charge of the Sogdian community was customarily assigned to a Sogdian (Rong Xinjiang 2004). Because Yu Hong was not a Sogdian, his official title included Jianjiao, meaning deputy, even though he was actually doing the work (Zhou Weizhou, 2004, p. 257). 77 The numbering follows those in the excavation report (Shanxisheng, 2001, p. 29-32). For an idea of what the doors might hâve looked like, see those of the stone sarcophagus of Shi Jun dated to 580, newly excavated in 2003. They are studded with round bosses. See Yang Junkai in this volume. 78 Zhang Qingjie 2001a, p. 3 and Shangsisheng 2001, p. 29-32.

Angela Sheng

From Stone to Silk

157

the Zoroastrian deity who présidés over the final judgment. Both these exterior panels visually flank the interior central panel (5) of the banqueting scene where the tomb owner and his wife, having passed the judgment, are in paradise. The other panels show complementary motifs of good overcoming evil, supporting the main theme of the deceased having successfully attained immortality.79

The Banqueting Scene The most significant banqueting scene on the sarcophagus is carved on the nearly

square slab prominently placed in the centre of the inner, re ar north wall, facing the south door opening.80 (Fig. 4) Depicted in the middle are the deceased and his wife seated on a raised platform under elaborate drapery. With cups in hand and a platter of food in front of them, they are shown as enjoying a feast while being entertained by haloed musicians and a swirling dancer.

Fig. 4: The Banqueting scene on Yu Hong’s sarcophagus, a line drawing. Source: Shanxisheng, 2001, p. 37, fig. 19.

79

For details, see Jiang Boqin, 2004, p. 147-54. Strictly spealdng, the Sogdians followed a polytheistic Version ofZoroastrianism, quite different from the Sasanian one (Foltz, 1999, p. 19). The nearly square slab measures 96 cm high and 110.15 cm wide whereas all other slabs are rectangular, being much narrower in width (Shanxisheng, 2001, p. 37).

二」

Les Sogdiens en Chine

158

Four deities of Zoroastrian origin are shown standing on either side of the deceased couple: to the left of the lady, Ameretat, the goddess of longevity who protects plants, and Hauvatats the goddess of health who protects water; and to the left of the man, Khshathra, the god of hope who guards the heaven, and Spenta Armaiti, the god of sacred heart who guards the earth. Below the feasting scene are shown two lions devouring the heads of two warriors. Since only the evil would suffer such an end at their final judgment according to Zoroastrian beliefs, the implication is that the deceased hâve been judged kind and admitted to the heavenly kingdom. Once there, they are drinking the white Haoma for gaining the immortality awarded by the suprême deity, Aliura Mazda. Clearly, this tableau idealizes life after death for the deceased?1 The material expression of this ideal in funerary practice was relatively recent to both sedentary Sogdians in their homeland and pastoral nomads who were settling into a sedentary life in northem China?2 However, it can be traced to ancient China。

Banqueting under Curtains and Immortality

Feasting in the etemal life after death was presented literally for aristocrats in the Western Han (206 BCE-CE 8). Moreover, silk curtains were used to set the stage for the etemal banquet, evidenced in the well~preserved tomb of Lady Dai that was unearthed in 1972. Besides the realistic setting, Chinese archaeologists also found on top of the coffin a painting that summarizes the conception of the tomb. The exceptional combination of such a "blue-prinピ and its actual construction afford a rare glimpse to the mentality and practice then that evolved with interesting variations for many subséquent centuries. Lady DaVs Banquet and Curtains in Her Tomb

When Lady Dai, wife of Marquis Dai (Li Cang 利倉), died at the âge of fifty in 168 BCE, her tomb burial at Mawangdui 馬王堆 in Changsha 長沙 was thoughtfully executed to ensure that both her hun 魂 soul would ascend to heaven and her po 魄 soul would remain under earth with the body, achieving immortality.8183 To this end, her wooden sarcophagus, guo 槨, functioned as her underground home. Her body rested inside the innermost of four polycentric coffins, surrounded by four compartments that represented different rooms in an aboveground house.84 81

For the iconographie identification, see Jiang Boqin, 2004, p, 140-41. For an analysis of Central Asian funerary practices betöre the arrivai of Islam, see Grenet, 1984. For a discussion of early funerary practices among pastoral nomads, see Rawson, 2001 and Su Bai, 1977a and b, and 1978. 83 For an analysis of Lady Dai's tomb in terms of the funerary ritual, see Wu Hung, 1992. For a detailed discussion on the ancient Chinese conception of the soul and life after death in relation to funerary art, see HeXilin,2003,p. 115-37. % The eastem compartment might have been where the accounts were kept (zhangfang 帳房) as a clothed figurine wearing a cap representing a chief male servant (Guanren 冠人) guarded the inventory list written on 312 bamboo slips and domestic objects in lacquer, pottery, and bamboo and some silks, with a retinue of 59 painted wooden figurines as household staff in attendance. Similarly, another figurine of capped chief 82

Angeia Sheng

_

From Stone to Silk

159

Of particular interest is the largest northem compartment that imitated the inner chamber (qin 31) in a real household.85 On ail four walls of this compartment hung real curtains made of beige silk gauze, monochromatically pattemed with triangles and diamonds and bordered in purple-red plain silk.86 In front of the western curtained wall stood a low screen, lacquered with patterns on both sides.87 The side facing the curtains shows a dragon; the other side, visible to the onlookers (those who set this stage), features interlocking rectangles and diamonds that intersect in the circular jade bi W dise. A ritual object, it signifies here the transformation from this to the next world. A reed-woven mat stretched in front of the screen from the south to the north wall.88 Because people in the Western Han still sat and slept on a mat on the floor, the mat fiinctioned like a bed.89 On top of this reed mat-bed were placed an embroidered pillow, some personal objects, a low lacquered table, and an embroidered robe.90 The last represented the deceased. Lady Dai’s feast in her afterlife was placed in front of this ensemble of screen-pillow-mat-bed (a precursor to the later couch-bed with screen). Lacquer dishes containing food were presented with bamboo chopsticks on a low lacquered tray.91 Facing the imagined Lady Dai in her afterlife were ten clothed female figurines standing in for personal servants ready to serve and eight other clothed figurines ready to entertain: four dancers and four singers, accompanied by fîve painted figurines who played musical instruments.92 Setting off this banqueting scene were the surrounding curtains. Similarly, curtains define a sacrificial feast on the silk painting that served as Lady Dai’s name banner (mingjing ^5®. (see Fig. 5, on colour plate n° 17) Described in the lower middle register of the painting, the sacrifice takes place above a ground level pattemed in diamonds and below a huge jade-heng suspended from the large jade èz-disc and through parted drapery pattemed with red and white triangles.93 In the foreground stand large vessels of food and wine; family and visitors bow to each other servant and 39 staff-figurines were found in the south compartment along with many pottery vessels containing food and some lacquer and bamboo ware. Textiles were conspicuously absent. This compartment might hâve represented the kitchen. The western compartment was clearly the storage where grains, fbod, medicine, money and clothing were kept (Based on an analysis of the content, type, medium, and location of the objects, see Hunansheng, 1973, vol. 1, p. 35-38, 95, 117-18, and 127). 85 Wu Hung, 1992, p. 135. 86 For details on the curtains, # 450, see Hunansheng, 1973, vol. 1, p. 73. For ail the numbered artifacts, see Hunansheng, 1973, vol. 1, p. 37. 87 # 447, see Hunansheng, 1973, vol. 1, p. 94-95; vol. 2, fîg. 8 and fîg. 192. 88 Woven of hemp warp and reed weft, the mat (# 438) is bordered with brocaded silk; see Hunansheng, 1973, vol. l,p. 121. a Cao annong, Y 1997. My thanks go to Wang Ningsheng for this reference. 90 The personal objects include an embroidered towel, numbered 439; a perfume bag: 442; the two layered lacquer boxes wrapped in embroidery: 441 and 443; the embroidered pillow was numbered 440; the low lacquer table 445 and the embroidered robe 437. 91 Ail items on the tray and the tray are numbered from 382 to 391. 92 Hunansheng, 1973, vol. 1, p. 97. 93 The ground level is shown upheld by a central figure of the world undemeath, possibly the human-faced and bird-bodied Yu Jiang a créature associated with the dark North of death; see He Xilin, 2003, p. 154.

160

Les Sogdiens en Chine

in the middle ground; the corpse wrapped in colorfal silk is placed on a couch in the center, in front of an altar with offerings.94 This remarkable painting summarizes the pictorial design of the entire tomb.95 Significantly, the ritual jade W-disc in the painting corresponds to the same painted on the lacquered screen found in front of the silk curtains in the northem compartment.96 Also, textiles in the painting are colored and pattemed differently for spécifie purposes. 97 Polychrome swirls embellish clothes wom by the tomb owner. Textiles with géométrie patterns serve to mark boundaries: the ground level beneath and the parted curtains above the sacrificial feast scene, and the ground level beneath the standing profile of Lady Dai. The latter ground level judiciously refers to the real silk that wrapped the innemiost coffin. This silk was uniquely pasted with feathers in a pattern of interlocking diamonds and further enhanced by an embroidered border forming the word, sun, ri H , a yang symbol. The salient feathers act as synecdoche for the birds that would assist Lady Dai’s hun soûl to fly to the sky, represented by the sun. On top of this feathered silk was placed the painting, that is, the name banner. In the painting, Lady Dai’s standing profile “names” her as the tomb owner and represents the name banner itself. Thus viewed, both painted and real curtains marked the feasting scene. In the painting, only the feasting scene is based on reality. To those living in the Western Han, the feast held paramount importance for their life after death.98 It symbolized the afterlife of plenty achieved only in immortality. Hence, the fonction of the curtains here extended to materially define this elusive realm. This concept was implicitly understood in ancient China, since it evolved from real uses of curtains in this life.

94

Wu Hung, 1992, p. 124 and He Xilin, 2003, p. 154-55. For the correspondence between the four registers of the painting and the four polycentric coffins, see Wu Hung 1992. 96 The same imagery of two dragons threading through the jade "-dise is also consistently painted on the foot side of the third, red lacquered coffin (Hunansheng, 1973, vol. 1, p. 26). 97 Only Lady Dai, the tomb owner, is portrayed wearing clothes made of silks with swirls, both as in her standing profile in the upper middle register and as in the corpse in the lower middle register. In reality, her corpse was wrapped in twenty-two layers of silk, which, if pattemed, were either woven with flying-bird motifs or printed and embroidered with métonymie swirls named as “longevity” (changshou xiu -ftW® They are so named and designed to facilitate her him soûl’s and “riding clouds” (chengyun as cent to heaven. In contrast, her servants and family members as well as guardians to the gâte of heaven are shown in the painting as wearing clothes made of plain fabrics (Sheng, 2002). 98 If the tomb providers omitted some items on the inventory list in the actual burial, they did their best to serve nearly all the listed food and drink, including ten kinds of grains, seven kinds of vegetables, five kinds of fish, seven kinds of birds with their eggs, seven main méats of beef, pork, lamb, horse, venison, dog, and rabbit and four kinds of wine. Ail could be cooked with six condiments in thirteen different ways. (Hunansheng, 1973, vol. 1, p. 154-55 and Wu Hung, 1992, p. 138). 95

Angela Sheng

From Stone to Silk

161

Sitting-under-Curtains as a Genre of Tomb Murals The ancient Chinese rulers used curtains to set up temporary palaces in the open fields.” Whether on hunting or warring expéditions, they could rest and grant audience within tented curtains. Some unearthed bronze implements reveal that the curtained tents could be either circular or rectangular and shaped like a house.100 That the Records compiled in the Western Han (206 BCE- 8CE), of the Rites of Zhou (Zhou Li strict rules for the use of curtains indicates the extent to which the wealthy prescribed the ruling elite. Some silk curtains were so extravagantly embroidered that emulated cautionary text, Yanteilun they even merited a mention in Huan Kuan’s (Discourse on Salt and Iron), dated 80 BCE. 101 The nobles also used curtained temporary houses to offer sacrifices and perform funerary rituals in the open. For these occasions, the curtains were made of purple or reddish black silks.102 Unlike the Tenjukoku curtains of 622, there was no embroidery. In addition to staging costly banquets in tombs viz. Lady Dai’s case, some elite gentry of the Eastem Han (25-220 CE) also began to have tomb walls painted with varions thèmes to further aggrandize extravagant burials.103 The practice of painting tomb walls probably derived from the earlier practice of painting murais in palaces and temples.104 Ubiquitously painted were the cosmic heavenly world above to where the hun-soul ascended and the talismanic figures and créatures that guided the hun-so\A there. More rarely painted were those stories of historical figures illustrating Buddhist and Daoist virtues that reflected well on the tomb owners and their descendants.105 Invariably, tomb owners were schematically portrayed sitting under curtains or zuozhang J°6 in contrast to real, floor-length curtains, curtains in painted murais were usually abbreviated as the overhead canopy above the tomb owners who sat on low bed-couches, often with screens behind them as well, while being feasted by dancers and musicians. Supporting this dominant theme of tomb owners sitting-undercurtains were other motifs that glorified their status, wealth, and achievement such as food préparation in the nearby kitchen, official résidence and grain storage, agricultural cultivation and harvesting on the estate, hunting, military tours, and cérémonial parades.107 '9

Lu Zhaoyin 1984. For an illustration of some circular tent implements made of bronze excavated from the tomb of King Zhongshan, dated to late fourth Century BCE, near Pingshan in Hebei in 1974, see Lu Zhaoyin, 1984, p. 455, figs. 1 and 2 and Zhongshan 1985, cat. nos.43 and 44. For the reconstruction of a house-shaped tent based on finds from Zhaoguqu Warring States Tomb no. 1 at Hui county: Lu Zhaoyin, 1984, p. 456, fig. 3. 101 Referenced in Lu Zhaoyin, 1984, p. 457. 102 Lu Zhaoyin, 1984, p. 463. 103 Yang Hong, 2000a, p. 2. 104HeXilin,2003,p. 139-42. 105HeXilin,2001,p. 169-71. 106 Lu Zhaoyin 1984, p. 458, fig. 4. Öhashi Katsuaki also includes représentations of Buddhist deities under draped canopies (1995, p. 96, fig. 32 and p. 97, fig. 97). 107 The latter two thèmes were often painted on the walls of adjacent corridors, cleverly extending the painted surface beyond that of the main chamber. For a séminal study on the tomb paintings of Han dynasty, see He Xilin, 2001. 100

Les Sogdiens en Chine

162

After the Han dynasty collapsed in 220, many people fled the war-tom Central Plains. Some moved south, others went northwestward to the oases around the Taklamakan desert and beyond, and still others migrated northeastward into the northem Korean peninsula. Along with these waves of migration also went the artistic vocabulary of tomb murais practiced in the Central Plains.108 For example, the genre of s itting-under-curtains was transmitted to Koguryô.109 The Chinese colonies in Koguryô further spread these imagery southward to Paekche in the fourth Century and from there, some reached Japan via the immigrant occupational groups in the fifth and sixth centuries. For example, the sun and the red bird of the four directional animais embroidered on extant fragments of the Tenjukoku curtains. (see section on Tenjukoku curtains’ Other Motifs above) Changes in Funerary Practice took over the wom-tom Central Plains and In China’s north, after Cao Cao ruled it as Emperor his son, Cao Pi (220-265), kingdom founded the Wei actively promoted frugal (r. 220-26). Following his father, Emperor Wendi Wendi burials.110 This was mostly targeted at the elite who observed the new policy during the Wei period. But, funerary customs changed for them in the subséquent Western Jin period (215-316). For example, the Wei prohibition against the érection of any aboveground vertical stone stele Qibei ïzW) led the noble families to hâve their généalogies inscribed onto much smaller stone epitaphs and buried them instead.11' This accounts for the origin of the stone epitaph found in Yu Hong’s tomb (dated 592). However, Southern China was not affected, viz. the aboveground stele near Nanjing, circa 520s, of the Southern Liang imperial family (see section on the iconography of Tenjukoku curtains above) Even more dramatically, the Wei policy of banning people from building above¬ ground family shrines © ^J) further encouraged the elite to aggrandize the under¬ ground tomb interior.112 During the Western Jin period aristocratie families revived lavish burials with a twist. Two significant changes in funerary practice ensued in the next Northem Wei period (386-535): the emergence of the stone bed-couch with screen and that of the stone sarcophagus shaped like a Chinese house in the fifth Century.

The Stone Bed-Couch with Screen in the The Western Han mat-bed had already evolved into a low couch-bed (ta a into substantially grew also single screen Eastem Han. By the Western Jin, the 108

Zheng Yan discusses in detail the impact on tomb paintings in the northeast, northwest, south, and the Central Plains (2002, p. 23-143). 109 sitting under curtains on a low The best laiown example is the frontal portrait of Dongshou He died in 357. His wife is painted in platform in front of a low screen, painted in his tomb at Anak profile, similarly sitting under curtains, to his side (Hong Qingyu, 1959). 110 Yang Hong, 2000a, p. 2-5. 111 Yang Hong, 2000a, p. 3, n. 36 and p. 5, n.68. 112 Yang Hong, 2000a, p. 4 n. 43 and p. 5.

Angela Sheng

From Stone to Silk

163

multi-paneled object often set above the now higher couch-bed. A beautiful example of the two combined was unearthed in 1965 from the previously disturbed tomb of Sima the Northern Wei5s capital of Ping¬ and his wife near Datong Jinlong of a platform resting on six legs. consists since 398. This stone couch-bed cheng The three front legs are carved in high relief, showing four musclemen as if supporting the bed, the central two of whom are animal-faced. Above them and along the front edge of the couch-bed are carved one dancer in the center3 surrounded by twelve musicians playing varions instruments among motifs of dragons and birds.113 The tomb is dated to 484 when Sima Jinlong was interred, ten years later than his wife. Howevcr, based on the scattered evidence inside the inner tomb chamber, the couple’s corpses were originally laid out on the stone couch-bed. Thus3 the stone couch-bed itself could also have been made earlier in or before 474. Above the corpses would have stood a parasol (instead of massive curtains) and behind them? a lacquered screen securely planted in stone stands. Only a few parasol sticks, four such carved stone stands, and five panels of the lacquered screens remain. The latter are painted with famous stories of virtuous rulers, brave générais, filial sons, and chaste ladies, all conferring Confucian values to the deceased.114 This iconographie content corresponds to Sima Jinling’s illustrions Han Chinese background, as he had descended from the imperial family of the conquered Jin dynasty. 115 In short, the combined stone couch-bed with screen continued an earlier Han Chi¬ nese funerary tradition (viz. Lady Dai’s tomb fumishings). Perhaps due to the screen ’s elaborate illustrations, there was no need to paint the walls in the tomb of Sima Jinlong and his wife. Here, more artistic attention was devoted to the lacquered screen than to the stone bed-couch. In contrast, a simpler stone couch-bed was found in another tomb dated to 474, also excavated from near Datong. Though carved, this stone couch-bed could not compete in splendor to the house-shaped stone sarcophagus that contained it.

The House-Shaped Stone Sarcophagus

This magnificent house-shaped stone sarcophagus was unearthed in 2000 near Datong. 116 The sixteen-square-metre sarcophagus resembled a traditional Chinese cérémonial hall. In its front stood four outer columns supporting the eaves of the roof above the veranda-colonnade. Obviously, the sarcophagus was much smaller than a real hall. All typical Chinese architectural features, such as wood beams and bracketing and ceramic tiles, are also replicated in stone. The façade and extemal walls are omamented with 22 animal’s faces pushou and 100 evenly spaced round bosses, paoding 113

The style is comparable to that fbund at Yungang Cave number 11 dated 483 (Shanxisheng, 1972). They include the famous story of Ban Zhao, the intellectual court lady who gave up her seat in the sedan to show propriety to the emperor. While the content can be compared to that of Admonitions of Court Instructress to Palace Ladies attributed to the legendary Western Jin artist, Gu Kaizhi, the style was a stiff imitation of Gu’s allegedly ûee-flowing brushwork (Wu Hung, 1997, p. 48 and p. 53, fig. 42). 115 Sima Jinlong’s epitaph reveals that he was the ninth-generation descendant of the Jin Emperor Xuandi’s brother. Though his father was a high-ranking official serving the Eastem Jin court, he submitted to the Xianbei in 419 (Shanxisheng, 1972, p. 27). 116 Shanxisheng, 2001 and Liu Junxi and Li Li, 2002. 114

164

Fig. 6: Song Shaozu’s house-shaped stone sarcophagus, dated 474, line drawings (Source: Shanxisheng, 2001, 7: 24-25, figs. 8a, 8b, and 11.)

117

Les Sogdiens en Chine The 100 bosses were unique among all nine excavated stone house-shaped sarcophagi dated before 592. This remarkable stone sarcophagus (Fig. 6) a Northern belonged to Song Shaozu Wei official who died in 474. Originally from Dunhuang, his family probably moved to Ping¬ cheng around 439 after the Northern Wei conquered the Northern Liang (398-439) in Gansu.117 He was then titled as Duke of Dun¬ huang and appointed as the regional inspecter (with its prefectoral seat loof Youzhou cated Southwest of today’s Datong near Beijing). In contrast to Sima Jinlong’s Han Chinese background, that Song Shaozu had come from Dunhuang where he was sure to have been in close contact with Central Asians might have meant that he was not stigmatized about adopting the house-shape of the Han Chinese aboveground funerary shrine for a sarcophagus. Many such small houses located near tombs were populär for ancestral worship in the earlier Eastem Han period.118 Some were still Standing in the fifth Century. Zheng Yan argues that the Xianbei people living during the Northern Wei times were inspired by these earlier shrines to have constructed the houseshaped sarcophagi. "° Although we do not know Song Shaozu’s ethnie origin, the fact that most figurines in Song Shaozu’s tomb sported Xianbei-styled military dress would intimate his familiarity with the Xianbei customs. Inside Song Shaozu’s houseshaped sar¬ cophagus was a low bed-couch. Made of stone, it is carved with motifs of stylized waves and

Based on an analysis of his epitaph and the inscription found on the stone sarcophagus, see Zhang Qingjie and Liu Junxi, 2001. 118 For an illustration of the stone mortu ary shrine at Xiaotangshan, Changqing, Shandong province and dated lst Century, see Wu Hung, 2002, p. 39, fig. 11. 119 Zheng Yan, 2001, p. 81-83. Wu Hung further attributes the vogue to four earlier kinds of funerary structures: the stone chambers found in the tombs of the Western Han prince Liu Sheng (d. 113 BCE) and his wife Dou Wan; the huaxiang shiguan (“pictorial coffin”) that first appeared in the Shandong-Jiangsu area in the first Century BCE but survived only in Sichuan after the first Century CE; the aboveground stone shrine mentioned above; and the “coffin pavilion” of the Northern Dynasties (2002).

From Stone to Silk

Angela Sheng

165

flora. It rests on legs, each carved with an animal’s face. The remains of two limestone pillows were found on top of the bed. Due to water damage, except fbr two musicians painted on the inner wall, most murais are no longer visible. Probably the tomb owners had been painted as sitting under curtains and enjoying a feast. This theme was painted on the inner north wall of another stone sarcophagus south of Datong in 1997.120 (Fig. 7) Dressed in Xianbei costumes, a man and lady are shown seated together—not customary for the Han Chinese in earlier times—on a low bed-couch under a draped canopy. Two men stand behind the tomb owner in attendance and to the right of the lady, probably his wife, two female servants are bringing food on a trav. The east and

unearthed from an anonymous tomb at Zhijiabao

Fig. 7 : “Tomb owners sitting-under-curtains ” mural on the inner north wall of the sarcophagus from Zhijiabao. Source: Wang Yintian and Liu Junxi, 2001, 7:43, fig. 6.

west walls are also painted with, respectively, four male and female attendants standing below winged créatures. On the south wall are painted ladies standing under a tree in the middle, with an ox-cart and attendant behind them on one side and on the other side a horse and groom. The simple pictorial composition describes a classic banqueting scene, with the horse and ox-cart suggestive of a joumey to the land beyond guided by the winged créatures above. Based on the style of the murais, Su Bai has dated the sarcophagus to the mid 480s.121 The few burial objects include some ceramic wares, a blue glass goblet, and some coins. There was no fumiture, as if ail the painted images totally replaced them.122 What is the relationship between the présentation and the pictorial représentation of banqueting scenes?

120

Wang Yintian and Liu Junxi, 2001. Wang Yintian and Liu Junxi, 2001, p. 51. 122 A point Zheng Yan also remarks (2001, p. 84).

121

F

166

Les Sogdiens en Chine Pictorial Representation on Tomb Walls, Screens, and Sarcophagi

Changes in funerary practices from the Western Han to the Northern Wei shifted the locus of pictorial représentation in the tomb. Initially, the locus was on the surface of the coffin and that of the screen as in Lady Dai’s tomb burial in 168 BCE. But, there was no painting on the tomb walls. During the Eastem Han, in addition to the staging, banqueting scene itself was also popularly painted on tomb walls as the genre of “sitting-under-curtains.” As the central focus, it anchored larger supporting tableaux that glorifïed the tomb owners’ status, wealth, and achievement. Other painted thèmes were the cosmic universe, talismanic créatures, and to a lesser extent, the stories of virtuous people. Thanks to the Wei policy of frugal burials, no tomb paintings of the Wei kingdom period hâve been found. 123 Similarly, very few survive from the subséquent period of Western Jin (215-316) and Sixteen Kingdoms (304-438). The stories of Confocian virtues painted on the lacquered screen above the stone bed (dated 474 or 484) in Sima Jinlong’s tomb departed from the usual theme of tomb owners sitting under curtains and banqueting. But these thèmes were painted on the inner walls of the house-shaped stone sarcophagi, viz. probably that of Song Shaozu dated 474 and that found at Zhijiabao dated mid 480s. After the Northern Wei court moved the capital from Datong to Luoyang in 493, the locals began to hâve their tomb walls painted again. One in particular reveals a dramatic change. The central east wall of General Wang Wen’s single-chamber tomb dated 532 shows the couple sitting under curtains in front of a multiple-paneled screen but inside a house with a window. (Fig. 8) Here the painted image replaced both the stone couch-bed and the house-shaped stone sarcophagus. It would seem that either a présentation of the real thing or its repré¬ sentation would suffice. Apparently there was no need for both. Zheng Yan also notes this phenomenon in the late Northern Qi (550-577) tombs: where the owners were painted as seated on screened couches, not only did the painted couches resemble those unearthed elsewhere but also no actual stone couches accompanied the tomb paintings. Indeed, in the Northern Qi tomb of Xu Xianxiu, dated 571 and unearthed from Taiyuan in 2000, the Fig. 8: “Tomb owners sitting under curtains screen-less couch-bed holding the wood inside a house”. Mural in General Wang coffin was made of brick, the same Wen’s tomb, dated 532 (Source: Zheng Yan material used for the construction of the 2002, fig. 66.) tomb. Also, the couch-bed, painted in the magnifïcent mural on the north wall, 123

Zheng Yan, 2002, p. 96.

Angela Sheng

From Stone to Silk

167

is covered with a carpet; it does not look as though made of stone.124 Conversely, Zheng Yan adds, in the later Tang tombs with stone-housed sarcophagi, the paintings in these tombs do not show the tomb owners’ portraiture.125 To understand why one surface or one medium might be chosen for représentation, it would be useful to study a group of unusual funerary fumishings. Stone Funerary Furnishings with Central Asian Motifs

The carved stone couch-beds of Sima Jinlong (dated 474 or 484) and Song Shaozu (dated 474) preceded at least five relatively intact stone beds with stone screens that hâve corne to scholarly attention in recent years. They all date to the second half of the sixth Century, feature some Central Asian motifs, and show variations of the theme populär since the Western Han, that is, tomb owners sitting on a bed under curtained canopy and banqueting. In the chronological Order of their extensive publications, they are the scattered pièces of the Anyang couch; 126 the Tianshui couch unearthed in 1982—the one with the most Han Chinese iconography;127 the incomplète Miho couch publicized in 1992; 128 An Qie’s couch excavated from Xi’an in 2000 and dated 579 the earliest of the group; 129 and the one Musée Guimet restored in 2003.130 Similarly, the two stone sarcophagi with the obvious Sogdian motifs also date to the late sixth Century: the one belonging to Yu Hong unearthed from Taiyuan in 1999 and dated to 592 the latest secure date of the both stone couch-beds and sarcophagi. The (dated 580) unearthed from Xi’an in 2003. other is the sarcophagus of Shi Jun Given that Zoroastrian-inspired Mazdeism prescribed funerary rituals without intemment,131 these unearthed stone artifacts reflect Central Asians’ adaptation of a long-standing Chinese tradition. But why were they made toward the end of the sixth Century, between 579 and 592, and not earlier or later? Let us tum to examine the historical circumstances at that time. By then the Sogdian presence in China had grown substantively, with Central Asian leaders—usually wealthy merchants in charge of the long-distance caravan trade—Controlling the

—

124

Shanxisheng, 2003, p. 6 and col. pl. 29. Zheng Yan, 2001, p. 84. 126 The publication of both the Tianshui and Miho couch recalled the three relief panels and two gateposts said to have come from Anyang but now scattered among three muséums: the Museum fur Ostasiastische Kunst in Cologne, the Museum of Fine Arts in Boston, and the Musée National des Arts-Asiatiques Guimet in Paris. See Scaglia, 1958. 127 A stone bed with five stone screened panels was unearthed from an anonymous tomb near Tianshui in 1982. Initially, it was dated to the Sui or early Tang dynasty, or late sixth century-early seventh Century; seeTianshuishi, 1992. 128 In 1992 appeared on the art market eleven marble panels and two gate towers originating from northem China and dating from the second half of the sixth Century. They form a couch now kept at Miho Museum in Japan with a base in the Levy White Collection in New York; see Juliano and Lerner, 1997, 2001, and 2004, p. 15 n. 5. 129 It belonged to the Sogdian leader, An Qie, who died in 579 and his tomb in Xi’an was unearthed in May 2000; see Rong Xinjiang, 2003 and Shaanxisheng, 2003. 130 Musée Guimet, 2004. 131 Grenet, 1984. 125

168

_

Les Sogdiens en Chine

community on behalf of the Chinese States.132 In his work on the économie history of the period, Han Guopan shows that not only did the imperial elite of the Northern Qi (550-557) and the Northern Zhou (557-581) favor foreign merchants from Central Asia, but also many high-ranking court officiais were engaged in trade, profïting from their association with them.133 This social proximity would tempt the Central Asian merchant-leaders to emulate the imperial elite in their conspicuous consumption of luxury, including that designed for the afterlife. More importantly, because Buddhist monasteries increasingly amassed more land and labor, depriving the state of revenue and corvée under the “equal-field” System of (r. taxation (juntianfa ES Ä), 134 the Northern Zhou Emperor Wu Æ a Xianbei— began his persécution of the Buddhist 561-579)—Yuwen Yong church with a command debate on the Three Teachings (Confucianism, Daoism, and Buddhism) in 569. The persécution peaked between 574 and 579.135 By 574, more than 200, 000 nuns and monks had been forced to leave monasteries and retumed to lay soci¬ ety.136 This coincided with Yuwen Yong’s campaign in 573 to actively recruit militia with which in large part he conquered the Northern Qi in eastem (fubingzhi 137 China in 577. Han Guopan attributes Yuwen Yong’s Buddhist persécution to his interest in promoting the “equal-field” System of taxation and the militia System. Until imperial patronage of Buddhism resumed in earnest in the Tang (618-906), possibly, the Northern Zhou Buddhist persécution movement also freed up talented stone carvers for non-Buddhist work? Wealthy Sogdians could hâve easily commissioned them to carve elaborate programs on commémorative stone couche-beds and sarcophagi.

Why Stone? Why were these bed-couches with screens or house-shaped sarcophagi made of stone? Basing on Alfred Gell’s theory on the enchantaient of technology,138 stone carvings that required more painstaking work than painting on tomb walls would hâve been more enchanting. This enchantaient was helpful to the deceased to reach the other world of immortality after death. Also, stone was a favored material for the Xianbei people with a pastoral nomadic past.139 They had a tradition of building stone rooms, their efforts to emulate the Chinese, the Xianbei chose to build stone shishi sarcophagi in the shape of a Chinese house, viz. Song Shaozu’s sarcophagus in the style of a Chinese ceremonial-hall (dated 474). It would appear that the idea of a stone house also held a special appeal to the 132

Rong Xinjiang, 200 1. Han Guopan, 1990, p. 319. 134 Gemet, 1995. 135 Han guopan, 1990, p. 374-78. 136 Fang Buhe, 2002, p. 254. 137 Han guopan, 1999, p. 155 andp. 335. 138 Gell, 1992. 139 SuBai, 1977a and b. 140 Mi Wenping, 198 land Steinhardt 2003. 133

Angela Sheng

From Stone to

左

169

Sogdians living in northem China in the late sixth Century. The tomb epitaph of Shi Jun 中君 (dated 580) provides an interesting word: "ßykt'k” for tomb. According to Yoshida Yutaka, it is a compound consisting of "ßッ”for “god" and ''kt'k" fbr "house, room.” Moreover, Yoshida indicates that this term was uniquely used in this context at this time. Also, in Shi Jun5s epitaph is the expression meaning "to die" which Yoshida gives a literal translation as "became god-soul,,? or in other words, achieving immortality.

Conclusion To

summarize3 Song Shaozu9s spectacular house-shaped stone sarcophagus

(dated 474, from Datong), modeled after a Chinese cérémonial hall, shows that the Xianbei people adapted the form of the aboveground Han Chinese family shrine for their own slightly different use. Shi Jun's stone sarcophagus (dated 580, from Xi, an) and Yu Hong s sarcophagus (dated 592,from Taiyuan) also show how Central Asians (mostly Sogdians) equally found the house-shaped sarcophagus appropriate as an expression of their social success in northem China at a level akin to that of the imperial elite (of both the Northem Zhou in the west and Northem Qi in the east). Once these non-Chinese elite had developed the house-shaped sarcophagus, the imperial elite of the Tang dynasties followed suit, but with iconographie content that suited them. Two examples suffice: Li Xiaohai s 李小孩 house-shaped stone sarcophagus dated 608 and excavated from near Xi'an in 1962, and Li Shou's 李壽 house-shaped stone sarcophagus dated 630 and unearthed from Sanyuan county north of Xi'an in 1972.141142 Both show ethnie Han Chinese style in architectural forms and incised motifs. Particularly pertinent as a possible source of inspiration for the design of the Tenjukoku curtains is the large stone turtle that carries on its carapace Li Shou's epitaph. (see Fig. 9). If the ruling elite in Tang China drew inspiration from these house-shaped stone sarcophagi, so could the Japanese. Why did Princess Tachibana no Oiratsume choose to have the curtains embroidered with text and imagery? Rather than made for any underground burial, the curtains were designed Fig. 9: Li Shou's house shaped to cover a temporary mouming pavilion (Jp. sarcophagus. 143 加oya; Ch. 56mgw〃喪 屋). Using modern 5

5

141

Last word in line 29 of Yoshida Yutaka's translation of the Sogdina inscription found in Shi Jun's tomb. I wish to thank him again for his clarification after the Conference. 142 For an illustration of the former, see Wu Hung, 2002,p. 41, fbg. 13 and Cheng Jianzheng, 2000, p. 84; for an illustration of the latter, see Cheng Jianzheng, 2000, p. 85-87. 143 Pradel, 1997, p. 194-210.

170

.

Les Sogdiens en Chine

anthropological analysis of the traditional death ritual,144 we can easily imagine the court gathered in front of the closed curtains that housed the corpse and, while the two opposing factions paid homage, they were also negotiating the héritage: access to

power. Given that Prince Shötoku (574-622) was trying to shift the basis of imperial authority and that Princess Tachibana no Oiratsume had wanted to see him the way in which he lived, that is, she endorsed his ideas; the design of the Tenjukoku curtains would hâve to convey both Prince Shötoku’s background and his ideological program. I propose that the curtains originally showed Prince Shötoku lecturing on the three sutras at court in the world after death, that is, etemally. Read this way, the monk-like figure was ringing the bell to announce the lectures. The lectures replaced the banquet as the key focus. The supplementary cosmic symbols merely embellished the “heavenly kingdom.” The immortality to be achieved was his wisli, which materialized with the Taika Reform beginning in 645 and ending in 701. The question for the designers then was how to combine both such imagery and the important 400-characters text? As it contained an edited version of Prince Shötoku’s genealogy, it was also a bold attempt on the part of Princess Tachibana no Oiratsume to ensure her family its continued hold on power. Possibly, the nun Zenshin, granddaughter of Shiba Tatto of the saddle-making occupational group, had guided the Princess in the drafting of the text? The placement of the 400 characters must hâve challenged the three designers and the Supervisor. With their immigrant background, they were no doubt familiar with the Southern Liang imperial family’s stone stelae, circa 520s, that lined the paths to their burial mounds. However, on the silk curtains made in 622, the words were embroidered four at a time on the carapace of 100 turtles interspersed among vivid motifs that described the pictorial setting of their true meaning. Since a fellow saddle-making clan member, Kuratsukuribe Furi, had accompanied the Japanese ambassador Ono Imoko on two diplomatie missions to China’s capital in Xi’an in 607 and 608, could Furi not hâve brought back the latest fashions, including funerary furnishings? Possibly, the three designers and the Supervisor, perhaps guided by Tori Bushi who had already made a design for another embroidery, might hâve benefited from two sources. Li Shou’s large epitaph-carrying turtle could hâve provided a model for the four-word version on the curtains. Second, the placement of the 100 round bosses on Song Shaozu’s stone sarcophagus might hâve been famous enough to hâve suggested the placement of the 100 turtles on the Tenjukoku curtains. We will never be sure if various stone sarcophagi inspired the création of the Tenjukoku curtains. Nonetheless, by examining the similarities and différences between the stone sarcophagi and silk curtains in the context of changing funerary practices and by focusing on the historical circumstances of intercultural transmission, we hâve been able to suggest probable answers to questions regarding both. In conclusion, Princess Tachibana no Oiratsume should be credited for having created a brilliant solution to ostensibly moum for her husband and his mother while subliminally supporting the pro-Buddhist Soga faction to which they ail belonged, 144

Watson and Rawski (eds.), 1990.

Angela Sheng

From Stone to Silk

171

including Empress Suiko. By petitioning Empress Suiko to hâve court ladies embroider silk curtains with an inscription and motifs representing Prince Shötoku's wish, the

princess also claimed her right to speak in the elite court dominated by men.145 Because her intention was never to give her dead husband a luxurious underground home, she would not hâve chosen to embellish a sarcophagus in the first place, nor a stone couch-bed. Besides, stone was not readily available in Japan for this purpose. Using wood to build a temporary mouming pavilion and then painting and writing on its exterior was out of the question。Women could not then build a wooden structure, let alone paint and write on its exterior, as the skills of both a carpenter and a painter werc controlled and passed on within Professional families?46 Also, elite ladies could not easily guard or visit a wooden structure. No doubt, Princess Tachibana no Oiratsume was aware of imperial women patronage of Buddhist art works in China3 including stone carvings and silk embroideriesJ47 Whereas ladies could not carve stones they could certainly embroider silk. Their work could even eam them merit.148 Moreover, this gendered access guaranteed control over the final product, invested with both memory and wish. Unlike Buddhist embroideries made fbr worship that could be used repeatedly, these silk curtains delivered their potent message most powerfolly once5 at the funeral ritual itself. However, after the ritual had imbued it with blessing by the dead as it were5 it could act as a talisman to fend off any evil spirits and protect one's staked out position. In this regard, though less durable than stone and so for misinterpreted, the Tenjukoku curtains made of silk constantly perpetuated an ideology of the past for the

present.

Bibliography

Primary Texts HouHanshu 後漢書 Xinjiao Hou Hanshu zhu 新校後漢書.6 vols. Taibei, Shijie shuju, 1973. Liangshu 梁書 Yang Jialuo 楊家駱 (ed.). Xinjiaoben Liangshu fu suoyin. 新校本梁書附索引 Taibei, Dingwen shuju, 1999. 郎血 u 魏書 Weishu 魏書. Beijing, Zhonghua shuju, 1974.

145

For an inspirational analysis of this issue at a later time in Japan, see Bargen 1997. Sheng, 2003b. 147 Wong, 2000 and Sheng, 2003a. 148 Sheng, 2003a.

146

Les Sogdiens en Chine

172

Secondary Sources Bailey, Lisa K.

"Crowning glory: Headdresses of the Three Kingdoms period", British Association for Korean Studies Papers, 1994, 5: 83-103. Bargen, Doris G. 1997 A Woman 's Weapon, Spirit Possession in the Tale of Genji. Honolulu, University of Hawai'i Press. Bloch, Maurice 1974 "Symbols, Song, Dance, and Features of Articulation: Religion , an Extreme Form of Traditional Authorityこ Archives Européennes de Sociologie, 15, 1 (1974):55-81.

1994

Bradford, Rosalind 2002 "The Guyuan Sarcophagus: Motifs from All Asia, A Study in Progress?5, Toronto Studies in Central and Inner Asia, 5 (2002): 177-99. Brown, Delmer M. (ed.) 1993 “Introduction" and "The Yamato Kingdom.” In Delmer M. Brown (ed.)0 Cambridge History of Japan, vol. 1 , Ancient Japan, 1-47, and 108-62. New York, Cambridge University Press. Bush, Susan 1975 "Thunder monsters, Auspicious Animais and Floral Ornament in Early Sixth Century China.", Ars Orientalis, 1975,10: 19-33. CAO, Yannong 曹硯農 1 997 "Zhongguo gudai zuofa kaolun" 中國古代坐法考論 [On Ways of Sitting in Ancient China], Zhongguo wenwubao, May 18, 1997. Chang, Lan 萇嵐 7U4 shiji Zhong Ri wenhua jiaoliu de kaoguxue yanjiu 7-14 世紀中日文 化交 2001 流的考古學 研究 [Archaeological Research on the Cultural Exchange between China and Japan from the 7th to the 14th Centuries]. Beijing, Zhongguo shehui kexue chubanshe. CHENG, Jianzheng 成建正 et al. (eds.) 2000 Xi'an Beilin Bowuguan 西安碑林博物館 [Xi'an Forest of Stone Tablets Museum] with English abstracts. Xi'an, Shaanxi renmin chubanshe. Chu, Yonghong 朱光憲 1985 Kökuri kofun hekiga 高句麗古墳 壁画 [Wall Paintings of Ancient Tombs in Koguryô]. Tokyo, Chosen Gahösha. Di Cosmo, Nicola 2002 Ancient China and Its Enemies: The Rise of Nomadic Power in East Asian History. Cambridge, Cambridge University Press. Ebrey, Patricia Buckley 1997 The Cambridge Illustrated History of China. Cambridge, Cambridge University Press, 1997 reprint of the 1996 édition. Egami, Namio 江上波夫 1967 Kiba minzoku kokka. 騎馬 民族国家 [States of Pastoral Nomads]. Tokyo:

Chüököron-sha. Escobar, Ticio 2003 "Thoughts about Hybrid or Intercultural Art.” Statement produced fbr the Roundtable Discussion "Toward a Methodology for "Intercultural Art” organized by Michel Oren (Laguna College of Art and Design, California) and

Angela Sheng

一

From Stone to

左

173

The College Art Association Diversity Committee held at the annual CAA meeting in New York. (ed.) Buhe 和 步 FANG, 方 Zhangyi shilüe 張掖史略 [A Brief history of Zhangyi]. Lanzhou, Gansu 2002 wenhua chubanshe. FOLTZ, Richard C. Religions of the Silk Road: Overland Trade and Cultural Exchange fro m 1999 Antiquity to the Fifteenth Century. New York5 St, Martini Griffin. FUJII, Shuichi 藤井守一 Senshoku no bunkashi 染織の文イ匕史 [The Cultural History of Dyeing and 1986 Weaving]. Tokyo, Rikögaku sha. Gell, Alfred

1992

1998

"The Technology of Enchantment and the Enchantment of Technology/5 In Jeremy Coote and Anthony Sheldon (eds.)5 Anthropology, Art, and Aesthetics , 40-63. Oxford, Clarendon Press. Art and Agency, an Anthropological Theory. Oxford: Clarendon Press.

Gernet, Jacques

1995

Transi. by Franciscus Verellen. Buddhism in Chinese Society, an Economie History from the Fifth to the Tenth Centuries. New York, Columbia University Press.

Grenet, Frantz 1984 Les pratiques funéraires dans l'Asie centrale sédentaire de la conquête grecque g V Islamisation. Paris, Editions du CNRS. HAN, Guopan 韓國 磐 1984 Beichao Sui Tang de juntian zhidu 北朝隋唐的均田制度 [The "Equal-field” System of Land Tenure during the Northern, Sui, and Tang dynasties]. Shanghai, Shanghai renmin chubanshe. Nanbeichao jingji shilüe 南北朝 經濟史略 [A Brief Economie History of the 1990 Northern and Southern Dynasties]. Xiamen, Xiamen daxue chubanshe. Han, Wei 韓偉 "Bei Zhou An Qie mu weiping shita zhi xiangguan wenti qiaiyian" ゴ 匕周安ロ 2001 墓圍屏 石 之相關 問題淺見 [A Brief View of Questions Related to the Stone Screen and Bed from the Northern Zhou Tomb of An Qie], Wenwu 2001, 1:90-101. He, Xilin 賀西林 Gumu danqing—Han dai mushi bihua de faxian yu yanjiu 古墓丹青̶ 2001 漢代 墓 室 壁畫的發現與研究 [Ancient Tomb Colors—the Discovery and research of Tomb Murals of the Han Dynasty]. Xi'an, Shaanxi renmin meishu chubanshe. 2003 "Cong Changsha Chu mu bohua dao Mawangdui yihao Hanmu qiguanhua yu bohua—zaoqi Zhongguo muzang huihua de tuxiang likT 從長沙楚墓帛畫至！] 馬王堆一號 漢墓漆棺畫與帛畫一-早期中 國 墓葬 繪畫的圖像理路 [From Silk Paintings of Changsha Chu tombs to the Lacquered Painting on the Coffins and the Silk Painting of Number One Han Tomb of Mawangdui―Pictorial Articulation in Early Chinese Tomb Paintings], Yishushu yanjiu 2003, 5: 143-68. Hong, Qingyu 洪晴玉 1959 "Guanyu Dong Shou mu de faxian yu yanjiu” 關 於冬壽墓的發現與 研究 [Conceming the discovery and Research of Dong Shou's Tomb], Kaogu, 1959, 1:27-35.

Les Sogdiens en Chine

174 Hunansheng Bowuguan :胡南 省博物館

and Zhongguo Kexueyuan Kaogu Yanjiusuo 中 國科學院考古 研究所 Changsha Mawangdui yihao Hanmu 長沙馬王 堆一 漢墓 [Number One Han 1973 tomb at Mawangdui, Changsha] (with English summary). Beijing, Wenwu. Huntington, John C.

-

i6The Iconography and Iconology of the cTan Yao5 Caves at Yungang,55 Oriental Art, N.S. (Summer 1986): 142-60. IIDA, Mizuho 飯田瑞穂 "Tenjukoku Shüchö mei no fukugen nitsuite^ 天寿国繍帳銘の復原に ついて 1966 [On the Reconstruction of the Inscription on the Tenjukoku Embroidered Curtains]55, Chüo daigaku bungakubu kiyö shigakuka 1966, 1 1 : 23-43. ISHIWATA, Mie 1987 ÉSGenealogy and Signifïcance of the Crown with Tree-shaped Ornament55 Bulletin of the Ancient Orient Museum (Tokyo), 1987,9: 45-69. JIANG, Boqin姜伯勤 "Sin jianjiao sabao Yu Hong mu shiguo huaxiangshi tuxiang chengxu shitan” 2001 隋 檢校薩 保虞弘墓石槨畫像石圖像程序試 探 [An Exploration ofthe Pictorial Program of the Relief Carvings on the Sui Dynasty Stone Sarcophagus of Yu Hong]. In Wu Hung, ed., Han Tang zhijian wenhua yishu de hudong yu jiaorong [Between Han and Tang: cultural and artistic interaction in a transfbrmative period], 29-50 (with English summary). Beying, Wenwu chubanshe. Zhongguo xianjiao yishu shi yanjiu 中 國 教 藝術 史研究 [A History of 2004 Chinese Zoroastrian Art]. Beijing, Sanlian shudian. Juliano, Annette L. and Lerner, Judith A. "Cultural Crossroads: Central Asian and Chinese Entertainers on the Miho 1997 Funerary couch.", Orientations, 28, 9 (1997): 72-78. "The Miho Couch Revisited in Light of Recent Discoveries.,,; Orientations, 2001 32,8 (2001): 54-61. 2004 "Stone Mortuary Fumishings of Northern China.'Tn Gisèle Croös (ed.) Ritual Objects and Early Buddhist Art, 14- 23. Bruxelles, Gisèle Croös SA Kim, Lena 1998 "Tradition and Transformation in Kore an Buddhist Sculpture. In Judith G. Smith (Coordinating Editor), Arts of Korea, 250-93. New York, The Metropolitan Museum of Art. Kim, Wong-Yong 1986 Art and Archaeology of Ancient Korea. Seoul, The Taekwang Publishing Company. Liu, Junxi and Li, Li. "The Recent Discovery of a Group of Northern Wei Tombs in Datong. 5,, 2002 Orientations, 34, 5 (2002): 42-47. Lou, Yulie 樓宇烈 and ZHANG, Zhigang 張志岡U (eds.) Zhongwai zongjiao jiaoliu shi 中外宗教交流史 [The History of Religions 1998 Exchange between China and Other Countries]. Changsha, Hunan jiaoyu chubanshe. Luo, Feng 羅豐 "Yijian guanyu Rouran minzude zhongyao shiliao" 一件關 於柔然 民族的重 2002 要史料 [An Important Text Conceming the Rouran People]”, Wenwu, 2002, 6: 78-83 and 93. 1986

5,

From Stone to Silk

Angela Sheng

175

LUOYANG BOWUGUAN 洛陽 博物館 Luoyang Hanmu か力我 臓 洛陽 漢墓 壁畫 [Tomb Paintings of the Han Dynasty in 1986 Luoyang]. Luoyang, Henan meishu chubanshe, DfER WENWU GONGZUODUI 洛陽 市第二文物工作隊 • LUOYANGSHI

“Luoyang shachang xilu Bei Wei HM555 fajue jianbao^ 洛陽紗廠西路北魏 HM555發掘 簡報 [Excavation of a Northern Wei Tomb at Luoyang], Wenwu, 2002, 9: 9-26 Zhaoyin 蔭 盧兆 LU, "Tüelun iiang Han Wei Jin de weizhang "略論兩漢 魏晉的帷帳 [A Brief 1984 Discussion on the Curtains of the Two Han, Wei5 and Jin Dynasties], Kaogu, 1984,5:45447. Mallory, J.-P., Mair5 V. H. The Tarim Mummes. New™ York, Thames and Hudson. 2000 Matsumoto, Kaneo 松本 包夫 Shösöin retsu to Asuka Tenpyö no senshoku 正倉院 裂と飛鳥天平の染織 1984 Jödai-Gire, 7也 and 8仍 Century Textiles in Japan from the Shösö-in and Höryü-ji . (in Japanese with English summaries). Kyoto: Shikösha Publishing Co. Ltd. McCune, Evelyn. 1967 The Arts of Korea, an Illustrated History. Rutland, Vermont and Tokyo, 2002

Charles E. Tuttle Company Publishers. Mi, Wenping 米文平

“Xianbei shishi de faxian yu chubu yanjiu^ 鮮卑石室的發現 與研究 [Discovery of and Preliminary Research on Xianbei Stone Rooms], Wenwu, 1981,2: 1-7. Mizuno, Yü 水野祐 Kökuri hekiga kofun to kikajin 高句麗 壁画古墳と帰化人 [Wall Paintings of 1972 Ancient Koguryö Tomb s and Immigrants] Tokyo, Yûzankaku. Musée Guimet de pierre, sommeil barbare; Présentation, après restauration e才 remontage, 2004 d'une banquette funéraire ayant appartenu à un aristocrate d'Asie centrale venu s 'établir en Chine au Vie siècle. Paris, Musée Guimet. NOMA Seiroku Translated from Japanese by John Rosenfeld. The Arts of Japan, vol. 1: 1978 Ancient and Médiéval. Tokyo, New York, and San Francisco, Kodansha International Ltd. ÖHASHI, Katsuaki 大橋一章 Tenjukoku Shüchö no kenkyü 天寿国繍帳の 研究 [Research on the 1995 Embroidered Curtains of Heavenly-Longevity-Kingdom]. Tokyo: Yoshikawa Köbukan. O 5 Riley, Michael Kämpen Art Beyond the West: The Arts of Africa, India and Southeast Asia, China, 2001 Americas. New York, Harry N. Japan and Korea, The Pacific, and Abrams, Inc., Publishers. ÖTA, Eizo 太田英蔵 "Tenjukoku Mandara no shügi to kenji shûri nitsuite” 天寿国曼茶羅の繍技 1948 と健治 修理 に ついて [On the Embroidery and Repair Construction of the Tenjukoku Mandala], Shiseki to bijutsu 史跡と美術 [Historie site and art], 188 (1948): 161-76.

1981

176

Les Sogdiens en Chine

Paine, Robert Treat and Soper* Alexander

1981 The Art and Architecture of Japan. New York, Penguin Books Ltd.. Youngsook Pak* 1988 “The Origins of Silla Metal work/\ Orientations 19, 9 (1988): 44-53, Papinot, E. 1 964 Historical and Geographie Dictionary of Japan. 2 vols. New York, Frederick Ungar Publishing Co. PENG, Feizhang 彭斐章 Zhongwai tushu jiaoliu shi 中外 圖書交流史 [The history of Book Exchanges 1998 between China and Other Countries]. Changsha, Hunan jiaoyu chubanshe. Portal, Jane 2000 Korea, Art and Archaeology. London, British Museum, 2000. Pradels Maria del Rosario 1997 "The Fragments of the Tenjukoku Shüchö Mandara: Reconstruction of the Iconography and the Historical Contexts.55 PhD. dissertation, University of California, Los Angeles. Qi, Dongfang 齊東方 2001 "Xianbei jinyinqi yanjiiT 鮮卑 金銀 器 研究 [A study of Xianbei Gold and Silver Objects]. In Wu Hung, ed. Han Tang zhijian wenhua yishu de hudongyu jiaorong [Between Han and Tang: cultural and artistic interaction in a transformative period], 559-73 (with English summary, 574-77). Beijing, Wenwu chubanshe. 2004 "Shuru, mofang, gaizao, chuangxin Sute qiwu yu Zhongguo wenhua" 輸入 • 模仿 • 改造 倉[I新-粟特器物與中國 文化 [Import, imitate, transform, create Sogdian artifacts and Chinese culture] . In Rong Xinjiang and Zhang Zhiqing (eds.), Cong Samargan dao Chang'an: Suteren zai Zhongguo Oe wenhua yiji [From Sarmakand to Changan: Cultural traces of the Sogdians in China] s 27-33. Beijing: Beijing Tushuanguan chubanshe. Rawson, Jessica 2001 “Creating Universes: Cultural Exchange as Seen in Tombs in Northern China between the Han and Tang Periods. In Wu Hung, ed., Han Tang zhijian wenhua yishu de hudongyu jiaorong [Between Han and Tang: cultural and artistic interaction in a transformative period], 113-52 (with Chinese summary). Beijing, Wenwu chubanshe. Rong, Xinjiang 榮 新江 2001 Zhonggu Zhongguo yu wailai wenming 中古中國與外來文明 [Médiéval China and foreign civilizations]. Beijing, Sanlian shudian. 2003 "The Illustrative Sequence on An Jia's Screen: A Depiction of the Daily Life of a Sabao^ Orientations, 34, 2 (2003): 32-35. 2004 « Cong Samargan dao Chang an—zhonggu shiqi Suteren de qianxi yu ruju" 從撒馬 爾干到長安―• 中古時期粟特人的 遷 徙與 入居 [From Sarmakand to Chang' an —Sogdians' Migration in the Middle Period]. In Rong Xinjiang and Zhang Zhiqing (eds.)? Cong Samargan dao Chang3an: Suteren zai Zhongguo de wenhua yiji [From Sarmakand to Changan: Cultural traces of the Sogdians in China], 3-8. Beijing, Beijing Tushuguan chubanshe. Ruch, Barbara 1992 "Coping with Death: Paradigms of Heaven and Hell and the Six Realms in Early Literature and Pamting” In James H. Sanford, William R. LaFleur, and Masatoshi Nagatomi (eds.), Flowering Traces: Buddhism in the Literary and Visual Arts of Japan, 93-130. Princeton, Princeton University Press. .

—

——

-

5

Front Stone to Silk

Angela Sheng

177

Said, Edward. W,

1978 Orientalism. New York, Random House,。 1994 Culture and Imperialism. New York, Random House. SAITÖ, Tadashi 斉藤忠 Higashi Ajia sö, bosei no kenkyü 東 ア ジア 葬,墓制の研究 [Research on East 1 987 Asian burials and tomb designs]. Tokyo, Daiichi Shobö. B. Japan, a Short Cultural History. London, The Cressett Press. 1962 SCAGLIA, Gustina 'sCentral Asian on a Northern Ch'i Gate Ârtibus Asiae,2\, 1 (1958): 1958

SANSOM, G.

9.280

SHANXI SHENG DATONG SHI BOWUGUAN 山西省大同市 博物館 SHANXI SHENG WENWU GONGZUO WEIYUANHUI 山西 省文物工作 委員 會 "Shanxi Datong Shijiazhai Bei Wei Sima Jinlong mu" 山西大同石家寨 北魏 1972 可 馬 金龍墓 [The Northern Wei tomb of Sima Jinlong at Shijiazhai, Datong in Shanxi], Wenwu 1972,3: 20-29, and 64. Shanxi sheng Kaogu Yanjiusuo 山西省考古 研究所,Taiyuan shi kaogu yanjiusuo 太原 考古研究所,Taiyuan shi Jinyuanqu wenwu luyouju 太原市晉 源區 文物旅遊 局 “Taiyuan Sui dai Yu Hong mu qingli jianbao,s 太原 隋代虞弘墓 清理簡報 [A 2001 brief report on the Sui dynasty tomb of Yu Hong], Wenwu, 2001, 1: 27-52. Shanxi sheng kaogu yanjiusuo 山西省考古研究所 Xi "an Bei Zhou 4% Qie mu 西安北周安伽墓 [Anjia Tomb of Northern Zhou at 2003 Xi'an] with an English abstract. Beijing, Wenwu chubanshe. Shaanxi sheng kaogu yanjiusuo 陝西省考古研究所 uXi'an beijao Bei Zhou An Qie mu fajue jianbao 西安北郊北周安伽墓發掘 2000 簡報.”[Excavation report of the Northern Zhou tomb of An Qie in Xi5 an], Kaogu yu wenwu, 2000, 6: 28-35. Shanxi sheng Kaogu Yanjiusuo 山西省考古研究所,Datong shi Kaogu Yanjiusuo 大同 市考古研究所. "Datongshi Bei Wei Song Shaozu mu fajue jianbao” 大同市北魏宋紹祖墓發 2001 掘簡幸長 [A Brief excavation report on the Northern Wei tomb of Song Shaozu], Wenwu, 7 (2001),19-39. Shanxi sheng Kaogu Yanjiusuo 山西省考古研究所,Taiyuan shi Wenwu Kaogu Yanjiusuo 太原市 文物考古 研究所 "Taiyuan Bei Qi Xu Xianxiu mu fajue jianbao” 太原北齊徐顯秀墓發掘 簡報 2003 [Excavation of Xu Xianxiu's Tomb of the Northern Qi in Taiyuan], Wenwu

2003,10:4-40. Sheng, Angela. 1994-5 "The Disappearance of Silk Weaves with Weft-effects in Early China.” Chinese Science, 12 (1994-95): 39-74. 2002 "Indexing Textile Art and Society in Ancient China: A (Re)viewing of the Mawangdui Textiles.” Paper presented at the annual meeting of the College of Art Association, Philadelphia, February 2002. 2003a "Textile Art, Technology, and Buddhist Merit Accumulation.55 In Sarah Fraser (ed.), Merit, Opulence, and The Buddhist Network of Wealth (in Chinese), 64-80. Shanghai: Fine Arts Publishing Company. 2003b "Of Tools and Access: Intercultural Transmission of Art and Technology along the Silk Road.” Paper presented to Crossing the Borders of China: 力 Conference on Cross-Cultural Interactions in Honour of Professor Victor H. Mair, December 5-7, 2003, Unniversity of Pennsylvania.

Les Sogdiens en Chine

178

Skaff3 Jonathan Karam 2004

"Survivai in the Frontier Zone: A comparative Perspectives on Identity and Political Allegiance in China’ s Inner Asian Borderlands during the Sui-Tang Dynastie Transition (617-630).” In Journal of World History, 15, 2 (2004): 117-53.

SONODA, Kôyü 1993 "Early Buddha Worship” (with Delmer M. Brown) In Delmer M. Brown (ed,). Cambridge History of Japan, vol. 1 , Ancient Japan, 359-414. New York, Cambridge University Press. Stanley-Baker, Joan 1984 Japanese Art. London, Thames and Hudson Ltd. Steinhardt, Nancy S. 2001 Koguryö to Gansu and Xinjiang: Funerary and Worship Space in North Asia 4th™7th Centuries.” In Wu Hung (ed) Han Tang zhÿian wenhua yishu de hudongyu jiaorong [Between Han and Tang: cultural and artistic interaction in a transformative period], 153-203 (with Chinese abstract). Beijing, Wenwu chubanshe. 2003 “Shishi55 (Stone Rooms). Paper presented to Crossing the Borders of China: A Conference on Cross-Cultural Interactions in Honour of Professor Victor H. Mair, December 5-7, 2003, University of Pennsylvania. Su, Bai S 1977a "Nei Menggu diqu de Xianbei yizhi" [Ancient Xianbei sites in Inner Mongolia], Wenwu 1977, 5: 42-52. 1977b "Shengle, Pingcheng yidai de Tuoba Xianbei—Bei Wei yizhi” [Ancient Xianbei sites of the Northern Wei in the Shengle and Pingcheng area], Wenwu, 1977, 11: 38-46. 1989 “Nanchao kanxiang yiji chutan” [A preliminary study of Buddhist niches: images of the Southern Dynasties], Kaogu xuebao, 1989, 4: 389-414. 2002 “Notes on Visits to the Hôryüji Temple in Nara, Japan.” Transi, by Feng Jiren Journal of East Asian Archaeology, 3, 1-2 (2002), 333-47. Swann, Peter C. / Concise History of Japanese Art. Tokyo, New York and San Francisco, 1979 Kodansha Internaional Ltd. Tamura, Enchö 1980 Kodai Chösen Bukkyö to Nihon Bukkyö [Ancient Korean Buddhism and Japanese Buddhism]. Tokyo, Yoshikawa Köbunkan. Tanabe, George J. Jr. 1 999 “Religions of Japan in Practice.” In Donald S. Lopez (ed.), Asian Religions in Practice, an Introduction, 154-75. Princeton, Princeton University Press. 「

Tianshuishi bowuguan 1992 “Tianshuishi faxian Sui Tang pingfeng shiguan chuang mu” [The Discovery of a Sui-Tang Tomb with Stone Screen and Couch in Tianshui], Kaogu 1992, 1 : 46-54. WADA, Atsumu SM Nihon kodai no girei to saishi, shinkô H 1995 ° fgfCP h [Ancient Japanese rituals, sacrifices and beliefs]. 3 vols. Tokyo, Hanawa

Shobô, 1995.

Angela Sheng

…一

From Stone to Silk

1.79

WANG, Weikun 王維 坤

Zhong R3 wenhua jiaoliu de kaoguxue yanjiu 中日文化交流的考古學 研究 [Archaeological research on the cultural exchange between China and Japan], Xi'an, Shannxi renminchubanshe. WANG, Yintian 王銀田 and LlU, Junxi 劉俊喜 "Datong Zhijiabao Bei Wei mu shiguo bihua” 大同智家 堡北魏墓石槨 壁畫 2001 [Murals inside the Stone Sarcophagus of the Northern Wei tomb at Zhijiabao, Datong]3 郎e施w% 2001, 7: 40-5 L WANG, Yong 王鐺,ed. Zhong wai meishu jiaoliu shi 中外 美術交流史 [The history of Artistic 1998 Exchange between China and Other Coun tries]. Changsha, Hunan jiaoyu . chubanshe. WATSON, James L. and Rawski Evelyn S. (edsj Death Ritual in Late Imperial and Modern China. Berkeley and Los Angeles, 1990 University of California Press, Wong, Dorothy C. “Women as Buddhist Art Patrons During the Northern and Southern 2000 Dynasties." In Wu Hung ed.s Between Han and Tang, Religious Art and Archaeology in a Transformative Period, 535-66 (with Chinese summary). Beijing, Wenwu chubanshe. WU, Hung The Wu Liang Shrine, The Ideology of Early Chinese Pictorial Art. Stanford, 1989 Stanford University Press "The Origins of Chinese Painting (Paleolithic Period to Tang Dynasty). In 1997 Richard M. Barnhart et. al. (eds.) Three Thousand Years of Chinese Painting, 15-85. Beÿing and New Haven: Yale University and Foreign Language Press. 2002 "A Case of Cultural Interaction: House-shaped sarcophagi of the Northern Dynasties. Orientations, 34, 5 (2002): 34-41. Yang, Hong 楊泓 "Shilun Nanbeichao qianqi foxiang fushi de zhuyao bianhua" 試論南ゴ 匕朝前 1983 期佛像服飾的主要變化 [A Study of the Major Changes in the Garments of Early Buddhist Sculptures of the Northern and Southern Dynasties], Kaogu, 6: 330-37. 2000a “Tan Zhongguo Han Tang zhijian zangsu de yanbian” 談中國 漢 唐之間葬 俗 的演 變 [A Discussion on the Change of Funerary Customs from Han to Tang dynasty in China] . In Yang Hong, Han Tang meishu kaogu he fuojiao yishu [Art, Archaeology and Buddhist Art from Han to Tang], 2-10. Beijing, Kexue chubanshe. 2000b "Gaojuli bihua shimu" 高句麗壁畫 石墓 [Painted stone tomb s from Korea] In Yang Hong, Han Tang meishu kaogu he fuojiao yishu [Art, Archaeology and Buddhist Art from Han to Tang], 164-85. Beijing, Kexue chubanshe. YAO, Qian 姚 遷 and Gu, Bing 古兵 (eds.) Liuchao yishu 六朝藝術 [The Arts of the Six dynasties]. Beijing, Wenwu 1981 chubanshe. Yl, Shui 易水 "Zhang he zhanggou” 帳和帳鉤 [Curtains and Curtain hooks], Wenwu, 1980, 1980 4: 85-88. Zhang, Qingjie 張慶捷 2001a "Taiyuan Suidai Yu Hong mu shiguo fudiao de chubu kaocha" 太原隋代 虞弓厶 墓石槨 浮雕的初步考 [Preliminary study of Reliefs from the outer stone

2002

'

5,

180

Les Sogdiens en Chine

coffin in the tomb of Yu Hong (Sui Dynasty, Taiyuan)], in Wu Hong, ed” Han Tang zhijian v^enhua yishu de hudong yu jiaorong [Between Han and Tang: Cultural and Artistic interaction in a Transformative Period], 3-28 (with English summary). Beijing, Wenwu chubanshe. 2001b '”YuHongmuzhi"zhongdejigewenti‘ 虞弘墓誌中的幾個問題 [Some Problems in Yu Hong's epitaph], 松也 w% 2001, 1: 102-8. "Ruxiang suisu yu nanwang gutu: ru Hua Suteren shizangju gaiguan” 入 隨 2004 俗與 難忘故土一人華粟特人石葬具 概觀 [Acculturation in a New Land and the Persistence in Homeland Memories: An OverView of the Funerary Use of Stone by Sogdians in China], in Rong Xinjiang and Zhang Zhiqing (eds.), Cong Samargan dao Chang1an: Suteren zai Zhongguo de wenhua yÿi [From Samarkand to Chang'an: Cultural traces of the Sogdians in China], 9-160 Beijing, Beijing Tushuguan chubanshe. Zhang, Qingjie 張慶捷 and LlU, Junxi 劉俊喜 "Bei Wei Song Shaozu mu liangchu mingji xi" 北魏宋紹祖墓兩 處 銘記 析 2001 [An Analysis of Two Funerary Inscriptions of Song Shaozu of Northern Wei], 2001,7: 58-61. Zheng, Yan 鄭 岩 "Qingzhou Bei Qi huaxiangshi yu ru hua Suteren meishu” 青州北 齊畫像石與 2001 入華粟特人美術 [Northern Qi Pictorial Carvings from Qingzhou and Sinicized Sogdian Art] In Wu Hong, ed., Han Tang zhijian -wenhua yishu de hudong yu jiaorong [Between Han and Tang: cultural and artistic interaction in a transformative period], 73-109 (with English summary). Beijing, Wenwu chubanshe. 2002 Wei Jin Nanbeichao bihua mu yanjiu 魏晉南北朝壁畫墓研究 [Research on mural paintings in tombs of the Wei, Jin, and Northern and Southern dynasties]. Beijing, Wenwu chubanshe. ZHONGSHAN Zhongshan: Tombes des rois oubliés, exposition archéologique chinoise du 1985 Royaume de Zhongshan. (aux Galeries nationales du Grand Palais) Paris, Association Française d'Action Artistique. ZHOU, Yiliang 周一 良 Wei Jin Nanbeichao shi lunji 魏晉南北朝史論集 [Articles on the History of 1997 the Wei, Jin and Northern and Southern Dynasties], Beijing, Beijing Daxu chubanshe. ZHOU, Faxiang 周發祥 and Li, Xiu 李岫 (eds.) Zhongwai wnxue jiaoliu shi 中外文學交流史 [History of relations between 1999 Chinese and Foreign Littératures] . Changsha, Hunan jiaoyu chubanshe. Zhou, Weizhou 周偉洲 "Sui Yu Hong muzhi shizheng55 隋 虞弘墓志釋證証 [An Explanation of Yu 2004 Hong 's Epitaph in the Sui Dynasty]. In Rong Xinjiang and Li Xiaocong (eds.), Zhongwai guanxi shi: xinshiliao yu sinwenti [The History of the Relations between China and Foreign Countries: New Sources and New Questions], 247-57. Beijing, Kexue chubanshe.

Towards a new édition of the Sogdian Ancient Letters: Ancient Letter 1 '

Nicholas Sims-Williams

As Étienne de la Vaissière has recently emphasized .in the second chapter of his Histoire des marchands sogdiens (2002/2004% the so-called “Ancient Letters”, now preserved in the British Library in London, are the oldest commercial documents in the Sogdian language and our single most important source of information on the early activity of the Sogdian merchants in China. The fact that we still have no reliable édition and translation of this unique corpus of texts is therefore a major obstacle to progress in understanding the history and dynamics of the Sogdian trading network.

The Ancient Letters represent the contents of a postbag lost in transit from China to the west. They were discovered in 1907 by Sir Aurel Stein at the site “T.XILa” to the west of Dunhuang, a guard-post on the wall guarding the western border of China, where they may have been confiscated by the Chinese authorities. The date of the letters has been much discussed. The clearest evidence is found in Letter No. 2, which contains references to Contemporary events in China: to a famine in Luoyang (Sogdian sry), as a resuit of which the emperor had fled from the capital, (xmi) and to the fighting between the Chinese (cyn) and the Huns or Xiongnu to ( ’nkp "). From these and other details, destruction of the cities of Luoyang and Ye W. B. Henning deduced that the letters must have been written shortly after the sack of Ye in 307 C.E. and of Luoyang in 311. Despite Janos Harmatta’s spirited attempt to prove that the letters belong to the late 2nd Century, a thorough re-examination of the evidence by Frantz Grenet and myself has completely vindicated Henning’s view.1

The letters were written by members of the Sogdian merchant colonies in western China, Nos. 1 and 3 in Dunhuang (kc n), No. 2 No. 5 in Guzang perhaps in Jincheng (kmzyn). There were Sogdians in Luoyang and Jiuquan (cwcn) too; Chang’an is also mentioned, but it is not specifically stated that Sogdian merchants were settled there. In some cities the Sogdians were apparently present in considérable numbers, since one passage refers to “a hundred freemen from Samarkand”. No. 2 was enclosed in an envelope addressed to Samarkand; some other (kr }wr 'n) or other places between letters may have been destined for Loulan Dunhuang and Samarkand.

1

Henning, 1948; Harmatta, 1979; Grenet, Sims-Williams, 1987; Sims-Williams, 1996, p. 47.

182

Les Sogdiens en Chine

Apart from two letters (Nos. 1 and 3) sent by a woman abandoned in Dunhuang by her husband, which will be the main topic of this paper, the letters are chiefly concemed with commercial matters. Values are calculated either in the silver staters used in Transoxiana or in “coppers”, a term which probably refers to the Chinese copper currency. Many commodities are named, including musk, gold, pepper, camphor, wheat and perhaps white lead, as well as cloth made of linen or of hair. It has been said that silk is not mentioned in the Ancient Letters. If that is so, it must be due to the limited extent of the texts, since there is no doubt that silk played an important rôle in the east-west trade at this period (even if the concept of the “Silk Road” had not yet been invented). But I suspect that silk is in fact refeixed to, not by one of the words attested in later Sogdian (pwth or cynh), but by the otherwise unlmown term pyrcyk, which occurs twice in Letter No. 6, lines 4-6: ’HRZYm m’S wywtw (S)[’r k](8)[ZY] ßy(k)[y ](s)w’ [’]t Icr’wr’n ’HRZY(m)sw pr pyrcyk xryny ’HRZY kS pyrcyk (L)[ ’ ... ’H](R)[ZYsw] ‘L kprwh xryn(y) ’HRZYmsw (”ßr)y [You] said to me: [Ifj you go out (from China) to Loulan you should buy pyrcyk for me (in exchange) for it, and if [you do not find(?) any] pyrcyk you should buy camphor (in exchange) for [it] and bring it to me.

P. O. Skjærvo, who was the first to make sense of this passage, plausibly suggested that pyrcyk may designate a type of fabric.2 If so, it seems likely that its first syllable is identical with Khotanese pira-, which means “worm”, especially “silkworm”.3 The addition of the common suffix -cïk would give in the first place an adjective meaning “derived from the silkworm”, hence, as a noun, “silk thread or cloth”.

The only complété translation of the letters is that accompanying the fïrst édition, published by Hans Reichelt in 1932. Reichelt was a fine scholar and his work was excellent for its time, but one must bear in mind that the study of Sogdian was then in its infancy and that the Ancient Letters are particularly difficult to understand because of their archaic language and their imperfect state of préservation. In the circumstances it is not surprising that Reichelt’s translation is frequently misleading, being based on guesswork which is sometimes inspired but sometimes demonstrably wrong. During the years following the publication of Reichelt’s édition many details were clarified by scholars such as Schaeder, Henning, Gershevitch, Skjærvo and Yoshida, but their remarks were usually limited to individual words or short passages. Even today, only two of the Letters hâve been re-edited and translated as a whole: No. 5, which I recently edited in collaboration with Frantz Grenet and Étienne de la Vaissière, and No. 2, of which we now hâve two complété éditions, one by Harmatta (unfortunately without any commentary justifying the many novel interprétations which he proposes) and one by myself.4 Obviously what is needed now is a complété new édition of ail the letters, with translation, glossary and linguistic and historical commentary. 2Skjærv0, 1976, p. 113-14. 3 4

See Duan Qing, 1991; [Emmerick], Skjærvo 1997, p. 91-3. Grenet, Sims-Williams, de la Vaissière, 2001; Hamatta, 1979; Sims-Williams, 2001.

Nicholas Sims-V/illiams

_

Ancient Letter 1

183

The éditions of Letters 2 and 5 which I have already mentioned constitute the first towards such an édition. Today, as a further step in the same direction, I would to like offer a new reading of Ancient Letter No. 1. (See Appendix I for the complété text with translation and philological notes.) To some extent I will also discuss No. 3, since the two letters are closely connected. Both letters describe the same situation and events and were written by the same person, a woman named Miwnay or Miwzay, No. 1 to her mother Chatis and No. 3 to her husband Nanai-dhat. No. 3 also contains as a postscript a message from their daughter Shayn.5 1 have described both letters in Whitfield, 2004, p. 248-9, which also includes plates. Digital images of Letter 1 (site-mark T.xii.a.ii.l, signature Or. 8212/92) and Letter 3 (T.xii.a.ii.3 = Or. 8212/98) can be accessed on the International Dunlmang database (http://idp.bl.uk). Like most of the other letters, Ancient Letter No. 1 was found folded and tied with string (as illustrated in Stein, 1921, IV, top left of PI. CLIII). On the outside of the folded letter a few words naming the sender and addressee are written in two columns separated by wavy lines (see fig. 1 and Reichelt, 1932, PI. VI). By comparing the wording of the letter itself, we can easily restore this text as follows: Steps

MN ôwyth mywn’yh ’ ’z’tch '(D) p(r)[yw] m’th [c’t’ysh] From (her) daughter, the free-woman Miwnay, to (her) d[ear] mother [Chatis].

The recto of the letter after opening is illustrated in Stein, 1921, IV, PI. CLV. In the course of a subséquent restoration, illustrated by Reichelt, 1932, PL I, a small fragment containing the second word of line 6 was displaced and a narrow strip of paper with words from the parts of lines 1-4 immediately above was moved down a line into the resulting gap. The manuscript has now been expertly restored to its former state in the British Library’s conservation laboratory, as may be seen from fig. 2.

The text on the recto begins with the usual formulaic greeting: (M)[N öw](Y)öryh m(y)[wn’yh ] ”z(’)tch 'D pryw [m’th] c’t(’)ysh [’](’)pryw nm’(c)[yw ](’)PZYs sy[r ]YWM (Z)[K](Z)Yß [HZY]H ZY ôrwch ’ncyh ’HRZY[m Z](K) YWM prtr ZY ’YKZYßn [x](w)ty wyn’ymn pr ôrwst’[n]wh [From her dau]ghter, the free-woman Mi [wnay], to (her) dear [mother] Chatis, blessing (and) homage. And (it would be) a good day for him who might [see] you healthy (and) rested; and [for me] that day (would be) the best when we ourselves might see you in (good) health” (lines 1-3).

5

For some general remarks on the content of Letter 1 and its relationship to Letter 3, see Henning, 1948, p. 615; cf. also Rosenberg, 1932, p. 760-61; Schaeder, 1942, p. 10 n. 1. On the date of Letter 3 see Henning, /oc. c". (21 April 313?), and Grenet, Sims-Williams, de la Vaissiere, 2001, p. 102 (21 April 314?). Since both letters describe the same Situation and events (as first recognized by Schaeder), Letter 1 must have been written at the same time.

184

_

Les Sogdiens en Chine

This and other greeting formulae were discussed in Sims-Williams, 1991. The one innovation in the text given here is a new restoration of the word for “health” as ôrw(see note in Appendix I below). st ’[n]wh rather than ôrwst

The substance of Miwnay’s message begins with the following words: (’)[HR](Z)Ym ’zm”nk k’ôyw ’PZYß[ w]yn’ym(n) ’HRZY ßxtk n(y)[s](t) ’HRZYm ptskw’t s’y’(r)[’](k) ß’nk(r’)m ’HRZY ß’nkr’m [m]’S w’ßt ’nyw ZY m8’ nny(St) pntrw nyst ’YKZ(Y) ’r(t)[yß](’)n I am very anxious to see you, (but) hâve no luck. I petitioned the councillor Sagha[ra]k(?), (but) the councillor says: Here there is no other (relative) doser to Nanai-dhat than Artivan (lines 3-5). '

The restoration of the name s ’y’(r)[’](k) “Sagha[ra]k” is just a guess based on the occurrence of the name s ’yr ’k. (referring to a different person) in Letter 2, line 32. On the title ß’nkr’m “councillor” (or possibly ß’zkr ’m “tax collector”?) see Appendix I

below. The councillor’s words “Here there is no other (relative) doser to Nanai-dhat than Artivan” must be compared with Letter 3, lines 12-13: ’HRZY ZK[y] ’yp(s)’r m’ô w’ß’nt( k)ôZY ’myô ô[rw’](n)y ’ny[w] pntrw y[wt]m nyst ’Y[K](ZY) ’rtyß’n “The leaders say: Here in Dunhuang there is no other relative doser than Artivan”. The wording is so similar that I think that these must be two accounts of the same conversation and that “the councillor” must be one of “the leaders” mentioned collectively in Letter 3.6 If I am right, we can get a clearer idea of events from Letter 3, which gives an account of certain exchanges preceding this one. In Letter 3 Miwnay addresses her husband with a mixture of ironical politeness and sharp anger, beginning as follows (lines 1-13): To the noble lord, (my) husband Nanai-dhat, blessing (and) homage on bended knee, as is offered to the gods. And (it would be) a good day for him who might see you healthy, happy (and) free from illness, together with everyone; and, sir, when I hear (news of) your (good) badly, not well, wretchedly, and I health, I consider myself immortal! Behold, I am living consider myself de ad! Again and again I send you a 1etter, (but) I do not receive a (single) letter from you, and I hâve become without hope towards you. My misfortune (is) this: it is three years that I hâve been in Dunhuang on Your Honour’s account, and there was a way out a first, a second, even a fifth time, (but) he(?) refused to bring me out. I requested the leaders that (financial) support (should be given) to Famkhund for me, so that he might take me to (my) husband and I would not be stuck in Dunhuang, (for) Famkhund says: I am not Nanai-dhat’s servant, nor do I hold his capital. I also requested thus: If he refuses to take me to (my) husband, then [provide](?) such support for me that he may take me to (my) mother.

° Comparison of the two passages also makes it clear that mô ' (=’mô

Letter 7, line 2), which Reichelt took to be an imperative, is the adverb “here” < *imada. The later form môy, left unexplained in the Grammar of Manichean Sogdian (hereafter GMS), §136, probably shows secondary addition of the oblique case-ending. The literal meaning of ’myô ô[rw'](n)y in Letter 3 may be “in this (city of) Dunhuang”, ’myô being an adjectival form rather than an adverb.

Nicholas Sims-Williams

Ancient Letter 1

185

The leaders say: Here in this (city of) Dunhuang there is no other relative closer than

Artivan ...7

If we combine the information in the two letters, we can reconstruct the fbllowing

situation. Miwnay, with her daughter Shayn, has been abandoned in Dunhuang for three years by her husband Nanai-dhat. She asks Famkhund to help her to rejoin her husband, but he disclaims responsibility. She then approaches Sagharak the councillor and the other leaders of the community, asking for help to go to her husband or her mother, but they refer her to Artivan, a relative of her husband. What happens next is recounted in the next words of both letters. In Letter 1 we read: ’HRZYm ptskw’t ’r(t)[yß]’n ’HRZY m’ö w’ßt prnxwnt ZY ’xsnyßnt ’HR(Z)[Y ]’zw L’ nm’w ’päßyr’t L(’) nm’w ptxwt And I petitioned Artivan, (but) he says: Famkhund (is) ’xsnyßnt, and I refuse to be hurried, I refuse to ... (lines 5-6).

Unfortunately the crucial word ’xsnyßnt does not occur elsewhere and its meaning is unknown. The corresponding passage in Letter 3 is no help, since virtually nothing is legible apart from the words “Artivan [say]s” followed immediately by the name of Famkhund. But it seems clear that Artivan’s reply is unhelpful and that he refers Miwnay back to Famkhund. Letter 1 does not quote Miwnay’s resulting request to Famkhund but continues immediately with his reply: ’HRZY prnxwnt m’S {m'3} w’ßt kSZY (n)ys kt’w’ntk ywtm L' ptsynt ’PZYtn sw’y pyts’r ’k’w m’th ’YKZYt ’zw slcr’n ’HRZY pcks’ ’YKZY nmcw’ s’n tys’t ’HRZY ’st kö nnyöt tys’t Famkhund says: If (your) husband’s relative (i.e. Artivan) does not consent that you should go back to your mother, how should I take you? Wait until ... comes; maybe Nanai-dhat will come (lines 6-9).

It is clear that Artivan is involved in this matter because he is a relative of Miwnay’s husband Nanai-dhat (see note to R7 in Appendix I below), but Letter 1 gives no hint who Famkhund is or why he should be expected to help Miwnay. From Letter 3 one gets the impression that he may be a business associate of her husband or at any rate that his Connection with the family is primarily commercial. I have already quoted the passage in which Miwnay asks the leaders to provide money so that Famkhund may take her to her husband and in which Famkhund is quoted as saying “I am not Nanai-dhat’s servant, nor do I hold his capital”. And at the very end of Letter 3, in the daughter’s postscript (lines 33-5), we read: “Famkhund has run away; the Chinese seek him but do not find him. Because of Famkhund ’s debts we have become the servants of the Chinese, I together with (my) mother”. If Famkhund is Nanai-dhat’s agent or business partner it is understandable that Nanai-dhat and his family might be held responsible for his debts. 7

The necessary philological commentary to justify my translation of this and other passages from Letter 3 will be provided in a forthcoming re- édition of that letter.

186

Les Sogdiens en Chine

Incidentally, the fact that Famkliund has run away and cannot be found makes it very unlikely that he should be identified with the man of the same name who wrote Letter No. 6 (as envisaged by de la Vaissière, 2002, p. 61-2 = 2004, p. 54-5). We should bear in mind that Famkhund or Famkhunde (which means “fortunate”) is quite a common name, attested also in the Mug documents and the Upper Indus inscriptions.8 Let us retum to Letter No. 1. Then, as now, charity from the church is the final refuge of the destitute. Her pleas rejected by both Famkhund and Artivan, as well as by the leaders of the Sogdian community, Miwnay tums to the priest of her local temple, as she explains in the last words of Letter 1: ’HRZY ’zw’m ßyzßr’k ’pw ny(’)w8n ’pw ’zy’mk ’HRZY x(wy)z’m p’(r)h ’HRZYm nySk mut 5ßrt ’HRZY ptwyS’m ’c ßynptw ZKw srôm’nytw < ’HRZYm m’5 wytw S’rt> ’HRZY kä swy ’HRZYtn ’zw ’xwstrw ößr’n ’HRZY mrt’xmk swy’ ’HRZY ’wy r’ôy(h) syr’klc ks’n ’HRZYm wytwr kwny ’ PZY Syyh prstyy I live wretchedly, without clothing, without money; I ask for a loan, (but) no-one consents to give (it) to me, (so) I depend(?) on charity(?) from the temple-priest. If you go, I will give you a camel, and a man should go (with you), and on the way I will look (after you) well. May he(?) do (so) for me until you send me a Ietter! (lines 9-12).

There are many uncertainties here; in particular the final sentence is quite obscure. But Miwnay’s desperate situation is clear enough. Letter 3 adds some further details. There it is mentioned two or three times that she and her daughter Shayn hâve become “servants of the Chinese” (lines 16(?), 20, 35), and in her postscript Shayn spécifiés that her task is to “watch over a flock of domestic animais” (line 30). The problem is basically one of money. “I know that you do not lack twenty staters(?) to send”, Shayn writes to her father (line 32), but nevertheless she and her mother remain “without clothing, without money” and no-one will give them a loan. No wonder Miwnay is angry with her husband, as she makes clear in Letter 3 (lines 21-4): “I obeyed your command and came to Dunhuang and I did not observe (my) mother’s bidding nor (my) brothers’. Surely(?) the gods were angry with me on the day when I did your bidding! I would rather be a dog’s or a pig’s wife than yours!”

8

Weber, 1972, p. 195 n. 18; Sims-Williams, 1992, p. 64.

Nicholas Sims=Williams

Ancient Letter 1

Appendix I: Ancient Letter No. 1. Text, translation

187

notes

Verso 1 2

3

MN 3wytha mywn’yh ”z ’tch

I '(D) p(r)[yw] | m’th [ I [c’t’ysh]

From (her) daughter, the free-woman Miwnay, to (her) d[ear] mother [Chatis].

Recto (M)[N 5w](y)5ryha m(y)[wn’yh ]”z(’)tcli 'D pryw [m’th] c’t(’)ysh [’](’)pryw nm’(c)[yw ](’)PZYs 2 sy[r ]YWM (Z)[K](Z)Yß [HZY]H ZY Srwch ’ncyh ’HRZY[mb Z](K) YWM prtr ZY ’YKZYßn [x](w)ty wyn’ymn ’zm”nk k’Syw ’PZYß[c w]yn’ym(n)d ’HRZY ßxtk 3 pr örwst’[n]wh (’)[HR](Z)Ym ' n(y)[s](t) ’HRZYm 4 ptskw’t s'y'(r)['](k)e ß’nk(r’)m ’HRZY ß’nkr’m [m]’ö w’ßt ’nyw ZY m3' nny(3t) pntrw 5 nyst ’YKZ(Y) ’r(t)[yß](’)n ’HRZYm ptskw’t ’r(t)[yß]’nd ’HRZY m’5 w’ßt pmxwnt ZY 6 ’xsnyßnt ’HR(Z)[Y ]’zw L’ nm’w ’pößyr’t L(’) nm’w ptxwt ’HRZY pmxwnt m’ö 7 m’5f w’ßt köZY (n)ys kt'w'ntkg ywtm L’ ptsynt ’PZYtn sw’y pyts’r ’k’w 8 m’th ’YKZYt11 ’zw skr’n ’HRZY pcks’ ’YKZY nmcw’ s’n tys’t ’HRZY ’st k8 9 nnyöt tys’t ’HRZY ’zw’m ßyzßr’k ’pw ny(’)wön ’pw ’zy’mk ’HRZY x(wy)z’mi p’(r)hJ 10 ’HRZYm nySk nmt 5ßrt ’HRZY ptwyS’m ’c ßynptw ZKw sröm’nytw k ’HRZY k5 11 swy ’HRZYtn ’zw ’xwstrw Sßr’n ’HRZY mrt’xmk swy’ ’HRZY ’wy r’SyCh)1 12 syr’kk ks’nm ’HRZYm wytwr kwny ’PZYn 3yyh° prstyy 1

[From (her) dau]ghter, the free-woman Mi[wnay], to (her) dear [mother] Chatis, blessing (and) homage. And (it would be) 2 a good day for him who might [see] you healthy (and) rested; and [for me] that day (would be) the best when we ourselves might see you 3 in (good) health. I am very anxious to see you, (but) hâve no luck. I 4 petitioned the councillor Sagha[ra]k(?), (but) the councillor says: Here there is no other (relative) doser to Nanai-dhat 5 than Artivan. And I petitioned Artivan, (but) he says: Famkhund (is) 6 and I refuse to hurry, I refuse to ... And Famkhund 7 says: If (your) husband’s relative does not consent that you should go back to your 8 mother, how should I take you? Wait until ... comes; perhaps 9 Nanai-dhat will corne. I live wretchedly, without clothing, without money; I ask for a Ioan, 10 (but) no-one consents to give me (one), (so) I depend(?) on charity(?) from the priest. If 11 you go, I will give you a camel, and a man should go (with you), and on the way 12 I will look (after you) well. May he(?) do (so) for me until you send me a letter!

1

188

Les Sogdiens en Chine

Notes on the reading aThe distinction between y and x is not strongly marked in this letter, but there is no justification for Reichelt’s reading ôwxth (VI) and [ôw]xôryh (RI). bThere is hardly room for ’HRZY[mn ]. cLess likely ’PZYß[n], dThus Rosenberg, 1932, p. 762. t-(r)[’](k) quite uncertain. fDittography? gWritten with a final as if the -k was an afterthought. hThe -t ends with a blot, but can hardly be read -tn (with Reichelt), bhe y is badly formed, so that the word looks like xwnz’m (or xp-, but hardly xps’m with Henning, 1948, p. 607 n. 2). jThus Henning, 1948, p. 607 n. 2. kAdded above the line. 'Final -h altered from =w? mThus Henning, 1945, p. 475 n. 4. D-m added below the line. °Sic.

Linguistic notes Rl-3. On the greeting formulae see Sims-Williams, 1991. In line 3 I now restore the word for “health” as ôrwst’[n]wh (= Bactrian Xpoyo(a)oTavo in several unpublished letters) rather than ôrwst / t]wh (for which a rather complicated explanation was offered ibid., 184). The address on the verso and the opening words of the main text are both damaged, but can be reconstructed on a mutual basis, the only différence being the two forms ôwyôryh (a correct but perhaps archaic ablative, also used in the same context in Letter 3) and the generalized nominative ôwyth (ibid. , 180 with n. 1). As pointed out by Henning (1948, p. 603 n. 3), the name of c’t’ysh, the mother, is also attested as Catisa in the Kharoshthi documents from Niya. It does not appear to be an Indian name, but may be Bactrian according to Sims-Williams, 1996, p. 52. The name of her daughter, which also occurs in Letter 3, can be read mywn’yh or mywz ’yh. It presumably contains myw “tiger”, a common element in Sogdian names (cf. Sims-Williams, 1992, p. 58), but the reading and etymology of the second component is unclear.

R3. ’zm’’nk (or ’nm’’nk, ’nm”zk etc.) is otherwise unknown. The construction (lit. “to me (there is) very much ’zm' ’mÆ”) indicates that it is a noun. Cf. perhaps Chor. (’)zmny- “to be concemed (about someone)”, which Samadi 1986, p. 264 tentatively dérivés from *uz-manya-, Hence ’zm’’nk “concem, anxiety”? A meaning such as “anxiety” may also suit Pth. ’wyzm”n, for which see Sundermann 1981, p. 154. — For ßxtk “luck, fate” cf. ßxtkw in Yoshida, 1979, p. 187, lines 3 and 4 (see ibid., p. 195, for the interprétation of the passage). In compounds the form without *-Æa-suffix is used: trßxtw “against fate” (GMS, §1144), Christian n’bxt “not fated” (Sims-Williams, 1985, p. 150aJ69V.28).

R4. The exact meaning of the title ß’nkr’m, which I hâve translated “councillor”, is unclear. It is attested again in the Mug document A13, line 1, a demand for payment issued by the “ß’nkr’m of Panjikent and the people”. Grenet, de la Vaissière, 2002, p. 187 n. 33 maintain the reading ß’zkr ’m and the translation “tax office” or “tax collecter" introduced by Frejman, 1962, p. 38 and elaborated by Bogoljubov, 1981, p. 106-7. However, the latter’s comparison of ß’zkr’m with Pth. ßa^typaßav

Illff Nicholas Sims-Williams

_

Ancient Letter 1

189

“custom-house” does not seem to provide a plausible explanation for the present passage, where the word clearly refers to a person rather than a place or office. Moreover, the reading ß’zkr ’m is hardly compatible with Gershevitch’s plausible comparison (1975, p. 200 n. 6) with ß’mkyr ’nt in the Buddhist text P6, line 167, which “court councillors” (as Yutaka Yoshida kindly informs translates Chinese fu xiang me). The etymology of this word is unknown —the suggestion of Henning, 1946, p. 733 gives undeserved primacy to the later form—but the metathesis of the nasals m and n would clearly imply a reading with n rather than z. — On the adverb mö’ “here” < *imada see above, p. 184 n. 6.

R6. As pointed out above, ’xsnyßnt does not occur elsewhere and its meaning is unknown. — For L’ nm- “to refuse” see Henning, 1948, p. 607 n. 2 and Benveniste, 1967; the negated injunctive L’ nm’w may fonction as a past tense, cf. Sims-Williams, 1996a, 182-3, but this need not be taken for granted. On the infinitive ’pößyr’t “to hurry” or “to be hurried” see Henning, 1937, p. 59 ad 506. The parallel infinitive ptxwt is inconclusively discussed by Henning, 1937, p. 76 ad 608; Gershevitch, 1975, p. 207. Such a form could most easily belong to a root xwan, but the meaning of Pth. pdxwn- “to Sound (a musical instrument)” does not suit. R7. kt ’w’ntk, lit. “householder” (*katâwant-aka^), seems to mean “husband”, at least in Letter 3, lines 9 and 11. Thus kt ’w’ntk ywtm here can be translated as “(your) husband’s relative”, referring of course to Artivan. Altematively, the phrase could be understood as “a relative (who is) a householder”.

R8. ’YKZYt ’zw skr “how should I take you?”. In general, the Ancient Letters distinguish clearly between the two forms of the enclitic pronoun of the second person singulär: acc./abl. -£ and gen. -/ (see Sims-Williams, 1985, p. 77 ad31V.31). So far as I can see, this is the only passage in the Letters where -t seems to be used as an accusative. (Henning, 1945, p. 475 n. 4 also takes -t as the object of skr-, although he construes difîerently: “if I shall lead you, then wait for me”.) — I do not understand nmcw 7zmcw’ s the usual s % “enemy” does not seem to fit. — The idiom ’st kô = Khot. aska “perhaps” (< *asti ka “es ist, wenn", i.e. Old Iranian *asti kadâ, Leumann, 1933-6, p. 392) has not previously been noticed in Sogdian.

R9-10. See Henning, 1948, p. 607 n. 2 (especially on p’(r)h “loan”) and Grenet, Sims-Williams, de la Vaissière, 2001, p. 97 (especially on ’zy ’mk “money” and the doubtful phrase ptwyô’m .. . ZKw srôm’nytw “I depend(?) on charity(?)”). — On the ßynptw “temple-priest” (Reichelt “sorcerer”) see Henning, 1948, p. 602 n. 3. RI 1. swy ' must be a third person form, perhaps an athematic Optative, cf. Khot. 3 sg. opt. act. -la < *-yät beside -ä < *-ait (Emmerick, 1968, p. 207). The 2 sg. optatives in -y’ which are well-attested in the Mug documents (Isxakov, 1977, p. 63-4, 180-81) may likewise contain the athematic ending *-yâh as in Khot. âye, 2 sg. opt. of “to be”. R12. The present stem ks-, cf. Oss. kæsyn, Pashto kas- “to look (at)”, Sogd. ’nxrks’y “star-gazer, astrologer”, was recognized by Henning, 1945, p. 475 n. 4 (though he

Les Sogdiens en Chine

190

oddly translates "i shall hâve a good time”). The past stem kst may be attested in Letter 3, line 22. — "HRZYm wytwr kwny ’PZY ôyyh prstyy “May he(?) do (so) for me (ie "look after me well'? or 'provide charity5?) until you send me a letter!” This sentence is quite obscure. The verb kwn- "to do” seems to lack the expected direct object. The subject of prstyy "send (a letter)” is presumably “you”, i.e. the addressee, but I cannot make sense of kwny as a second person form.

Bibliography

Benveniste, É. 1967 “Le verbe iranien nam- en sogdien55. Bulletin of »e School of Oriental and African Studies, 30/3, p. 505-1 L Bogoljubov, M. N. "Sogdijskie ètimologii” [Sogdian étymologies]. Iranskoe Jazykoznanie. 1981 Ezegodnik 1980. Moskva: Izdatel’stvo “Nauka”, p. 106-20. Duan Qing [The Khotanese “Yutian wen de can zi, jian zi, si zi” 1991 words for silkworm, cocoon and silk]. Ji Xianlin jiaoshou bashi huadan jinian [Collection of articles in lun^enji commémoration of the 80th birthday of Prof. Ji Xianlin]. Nanchang: Jiangxi People’s Publishing House, p. 45-50. Emmerick, R. E. Saka grammatical studies. London: Oxford University Press, London Oriental 1968 Sériés, 20. Emmerick, R. E. & Skjærvo, P. O. Studies in the vocabulary of Khotanese, III. Wien: Verlag der Österreichischen 1997 Akademie der Wissenschaften, Sitzungsberichte der phiL-hist. Klasse, 651. FREJMAN, A. A. Opisanie, publikacii i issledovanie dokumentov s gory Mug [Description, 1962 éditions and studies of documents from Mt. Mug]. Moskva: Izdatel’stvo Vostocnoj Literatury, Sogdijskie dokumenty s gory Mug, 1. Gershevitch, I. A Grammar of Manichean Sogdian (= G MS). Oxford: Basil Blackwell, 1954 Publications of the Philological Society. “Sogdians on a frogplain”. Mélanges linguistiques offerts à Émile Benveniste. 1975 Paris: Société de Linguistique de Paris, Collection Linguistique publiée par la Société de Linguistique de Paris, 70, p. 195-211. GRENET, F. & SlMS-WlLLIAMS, N. "The historical context of the Sogdian Ancient Letters”. Transition periods in 1987 Symposium de Fribourg-en-Brisgau (22-24 Mai Iranian history. Actes 1985), Leuven: Association pour l’avancement des Etudes Iraniennes, Cahiers de Studia Iranica, 5, p. 101-22. GRENET, F., SIMS-WILLIAMS, N. & DE LA VAISSIÈRE, É. “The Sogdian Ancient Letter V”. Alexander's Legacy in the East: Studies 2001 honor of Paul Bernard. Bulletin of the Asia Institute, 12, 1998 [2001],

+

p. 91-104.

Nicholas Sims=Williams

_

Ancient Letter 1

191

GRENET, F. & DELA VAÏSSIÈRE, É.

"The last days of Panjikent”. Silk Road Art and Archaeology, 8, p. 155-96. J. “Sogdian sources for the history of Pre-Islamic Central Asia”. J. Harmatta 1979 (ed.), Prolegomena to the sources on the history of Pre-Islamic Central Asia. Budapest: Alcadémiai Kiado, p. 153-65. HENNING, W. B. Ein manichäisches Bet- und Beichtbuch. Berlin: Preußische Akademie der 1937 Wissenschaften, Abhandlungen der phil.-hist Klasse, 1936, Nr. 10. School of Oriental and African Studies, 11/3, “Sogdian taies55. Bulletin of 1945 p, 465-87. School of Oriental and African "The Sogdian texts of Paris’5, Bulletin of 1946 Studies, ll/4?p. 7340. “The date of the Sogdian Ancient Letters5\ Bulletin of the School of Oriental 1948 and African Studies, 12/3, p. 601^15. Isxakov, M. M. Glagol v sogdijskom jazyke (dokumenty s gory Mug) [The Sogdian verb in the 1977 documents from Mt Mug]. Tashkent: IzdateFstvo “Fan” Uzbekskoj SSR. DE LA VAISSIÈRE, É. Histoire dCs marchands sogdiens. Paris: Collège de France, Bibliothèque de 2002 1’Institut des Hautes Études Chinoises, 32. Histoire des marchands sogdiens. Deuxième édition révisée et augmentée. 2004

2002

HARMATTA,

Leumann, E.

1933-6 Das nordarische (sakische) Lehrgedicht des Buddhismus. Leipzig: Abhandlungen für die Kunde des Morgenlandes, 20. Reichelt, H. Die soghdischen Handschriftenreste des Britischen Museums. II. Teil. 1931 Heidelberg: Carl Winters Univers itätsbuchhandlung . Rosenberg, F. “Zu Reichelts Ausgabe der soghdischen Handschriftenreste des Britischen 1932 Museums. II”. Orientalische Literaturzeitung, 1932, Nr. 12, p. 758-63. SAMADI, M. 1986 Das chwaresmische Verbum. Wiesbaden: Otto Harrassowitz. SCHAEDER, H. H. “Beiträge zur mitteliranischen Schrift- und Sprachgeschichte”. Zeitschrift der 1942 Deutschen Morgenlandischen Gesellschaft, 96, p. 1-22. Sims-Williams, N. 1985 The Christian Sogdian manuscript C2. Berlin: Akademie Verlag, Berliner Turfantexte, 12. "A Sogdian greeting”. R. E. Emmerick & D. Weber (ed.), Corolla Iranica. 1991 Papers in honour of Prof. Dr. David Neil MacKenzie. Frankfiirt am Main: Peter Lang, p. 176-87. 1992 Sogdian and other Iranian inscriptions of the Upper Indus II. London: School of Oriental and African Studies, Corpus Inscriptionum Iranicarum, Part II, Vol. III/2/2. “The Sogdian merchants in China and India”. A. Cadonna & L. Lanciotti 1996 (ed.), Cina e Iran da Alessandro Magno alla dinastia Tang. Firenze: Leo S. Olschki Editore, Orientalia Venetiana, 5, p. 45-67. 1996a “On the historié present and injunctive in Sogdian and Choresmian”. Münchener Studien zur Sprachwissenschaft, 56, p. 173-89.

Les Sogdiens en Chine

192 2001

“The Sogdiaïi Ancien! Letter IF1 M° G. Schmidt & W. Bisang (ed) Philologien et Linguistica. Historia, Pluralitas, Universitas. Festschrift für Helmut Humbach zum 80. Geburtstag am 4. Dezember 2001. Trier: Wissenschaftlicher Verlag Trier, p. 267-80.

SKJÆRV0, P. O.

1976

•

"Sogdian notes”. Acta Orientalia, 37, p. 1 1 1-16.

Stein, A.

1921

Serindia. 5 vols. Oxford: Oxford University Press.

Sundermann, W.

1981

Mitteliranische manichäische Texte kirchengeschichtlichen Inhalts . Berlin: Akademie Verlag, Berliner Turfantexte, 1 L

Weber, D»

“Zur sogdischen Personennamengebung5\ Indogermanische Forschungen, 77, p. 191-208. Whitfield, S. The Silk Road: Trades Travel, War and Faith. London: The British Library. 2004 Yoshida, Y. “On the Sogdian infinitives”. Journal of Aslan md African Studies , 18, 1979 p. 181-95.

1972

Plate 1. Ancient Letter No. 1 , verso. Courtesy British Library Board.

Nicholas Sims-Williams

Ancient Letter 1

Plate 2. Ancient Letter No. 1, recto. Courtesy British Library Board.

193

Caravans and Caravan Leaders in Palmyra Albert E. Dien

We know deplorably little about the Organization of the Sogdian communities in China and the “Western Régions,” and probably even less about the commerce in which they were engaged. What we have is anecdotal. The Caravans could be quite large; in (of Sogdian descent) seized an illegal 553 a govemor of Liangzhou, Shi Ning caravan consisting of 240 merchants, 600 camels, and 10,000 rolls of multicolored all names for the same place) was or Guzang silks.1 Liangzhou (or Wuwei so and attracted a large community of especially important in this international trade, Sogdians. In 439 when the city was captured by Northern Wei armies, a large number of Sogdian merchants were among the 30,000 households that were taken en masse to the Northern Wei capital at Datong A [°], and they were not ransomed by the Sogdian king until 452, some 23 years later!2 But this did not discourage a continuing migration of Sogdians, and Settlements flourished, as described in an early sixth Century, Contemporary description of Luoyang, the Luoyang Qielanji The recent archaeological discoveries of tombs of Sogdians who died in China has led to major advances in the study of these Sogdian communities, and a renewed interest in the term sabao (derived from the Sanskrit sârthavâha “caravan leader or merchant chief’) by which the leaders of these communities were known. In this context the so-called Sogdian “Ancient Letters” of the early fourth Century are especially important for the light they throw on the Sogdian mercantile activities.4 These merchants would appear to resemble closely those of another place and era who were described as “running and racing about as hungry folk,” or “in order to eam a little bit more they are willing to run to the end of the world, and they live so miserably that for the most part they only eat herbs.”5 Normally it would have been a prosperous enterprise plied by small peddlers buying and selling small quantifies as they travelled from market to market. The letters, written as they were in a particularly troubled time in Chinese history, are especially illustrative of the observation that “merchants’ letters do not contain much in the way of glad tidings.”6 Letter V, addressed to the s’rtp’w *

I wish to thank Christelle Fischer for her assistance with the Greek texts. Zhoushu — 50.913 (Beijing: Zhonghua shuju, 1971 édition). 2 Wei shu 4A.90 and 102.2270 (Beijing: Zhonghua shuju, 1973 édition). — 3 Fan Xiangyong 1978, p. 161; Jenner, 1981, p. 220; Wang Yi-t’ung, 1984, p. 150-51. 4 For a discussion of the dating and the significance of these letters, see Sims-Williams, 1985; and Grenet, Sims-Williams, 1987. See also Sims-Williams in this volume. 5 Steensgard, 1973, p. 28. 6Steensgard, 1973, p. 46. This observation can be confîrmed by Consulting the contents of Goitein, 1973. 1

196

—

Les Sogdiens en Chine

Aspandhât, who was perhaps at Loulan, is especially revealing of the writer’s desperate situation due to the current troubled conditions. It has been observed that the modest amounts involved in these dealings indicate that the s’rtp’w was probably no more than a chief caravaneer and that the letter would be typical of “petty merchants operating on a fairly extensive geographical scale.”7 This letter must hâve been only one of a sériés written almost daily to compensate for the non-transparency and unpredictability of the market, and so the amounts reported in any one letter would hâve been small.8 To be sure, the Sogdian merchants were petty merchants, for that was the nature of the commerce, but the meaning of the term “chief caraveneer” still remains rather unclear. It may be usefiil to compare this Sogdian case with other caravan cultures for any light that may shed on the problem. Instances of mention of caravan leaders or of head merchants are not diflïcult to find. In the Metropolitan Museum of Art there is a Sumerian calcite bowl dating to 2600-2500 BC that has an inscription which reads: Forlnanna Aka-Enlil, the chief merchant, the son of Heti,

dedicated this bowl.9

In Egypt of a similar period, (the VIth dynasty, ca. 2600 BC), there is reference to “caravan-conductors 10 The terms would indicate that a hierarchical status position was to be found among the merchants." This is no doubt to be expected, but the précisé rôle of the “merchant chief’ or “caravan conductor” is not laiown. The fonction of such a person would be affected by many social, économie and even political factors. Closer in time and place, there are Tibetan fragmentary materials from the Tarim Basin that list a caravan transporting barley, silk, silver, gold, oil, flowers, felt and other goods; the personnel included a leader (rad-mkham), banner-bearer (dar), quartermaster (Itha-bzï), bedding-man (mal-hphan), guide (sa-rig and sa-mkhan) and others.12 The term sârthavâha as caravan leader occurs in a Brâhmi inscription from Mathura in the Kushan period,13 and as sârtwâ (s’rtw 7srtw ’) “leader of a caravan” in a Sassanian Pahlavi text.14 7

Grenet, Nicholas Sims-Williams, de laVaissière, 1998, p. 101. Steensgard, 1973, p. 42-59. 9 The word for “merchant chief’ is dam-gar3 ga/, which I am told is composed of a loan from Akkadian tamkaru with gai, “big.” Personal communication from Prof. Niek Veldhuis, Mar. 5, 2002. 10 Breasted, 1906-07, 1.154: “I descended [from Yam, southem Nubia] with 300 asses laden with incense, ebony, heknu, grain, panthers, . . ., ivory throw-sticks, and every good product. . . . I was more vigilant than any caravan-conductor who had been sent to Yam before.” This was cited in Schoff, 1912, p. 61. Another reference to the caravan-conductor is in Breasted, 1906-7, p. 162-63. "Postgate, 1992, p. 221. 12 Thomas, 1951, p. 371. This was cited in Raschke, 1978, p. 789, note 620. 13 Konow, Venkayya, 1909-10, p. 6; Puri, 1945, p. 84, and Puri, 1940, p. 423. The inscription records a religions donation by a woman who is termed a sarttavâhinî (cited by Raschke, 1978, p. 811, note 790). 14 Tafazzoli, 1974, p. 195. This and the one above are in addition to the many other occurrences as cited in Dien, 1962, p. 337. 8

Albert E. Dien

Caravans and Caravan Leaders

197

At any rate, this ail leads to the question of what was the fonction of the sârthavâha as it appears in the various languages and, more especially, of the sabao of the Sogdian communities before or even during the period when the title was coopted by the Chinese. Might not a community largely composed of merchants and those connected with commercial dealings hâve had an elder to whom they could tum for arbitration and leadership?15 Were the sabao/s’rtp’w the heads of trading posts, of fonduqs, acting as middle-range managers of some larger commercial network ? Altematively, was the title one more narrowly used for the caravan leader, much like the wagon train leader of the American West, and then that prestigious title was utilized by the Chinese when they sought to impose some sort of administrative head over the Sogdian communities ? Or are there other ways in which to understand the rôle of these sabaol It may help to begin to answer some of these questions if we tum to the use of similar terminology on the other end of the Silk Road, on the western end, as it were, at Palmyra, during the flrst to the third centuries, about which there is relatively more information available conceming the caravan trade and the caravan leaders. A recent collection of the Aramaic inscriptions from Palmyra itself or related to Palmyra contains 2832 items, with Greek versions when such are available.16 While many of these were funerary in nature, there was also the practice in Palmyra of honoring certain individuals by installing statues with an accompanying dedicatory inscription.17 Among such inscriptions, numbering 181, there are 31 that deal directly with the Caravans, and another few that may be found in other sources.18 These inscriptions, dating from 19 CE to 267 CE, with two-thirds being from the second Century, are the primary source of information about the caravan trade at Palmyra. In any considération of these inscriptions there are two caveats to be kept in mind: those that survive may be a skewed sélection, and secondly, the conclusions to be drawn from the sparse data that they contain can only be highly spéculative.19 In the inscriptions, the following terms are used: merchants in Aramaic are tgry’ and in Greek empor oi-, caravans in Aramaic are syrt’ or the derivative bny syrt’ “members of a caravan,” and in Greek, synodia ; and the leader of the caravans in Greek is synodiarch, literally, “head of the caravan,” while in Aramaic the équivalent term is rb syrt ' or rs syrt ' "leader or head of the caravan.” The focus of my remarks will be on the terms for caravan leader: synodiarch and rb syrt ' or rs syrt In some of the inscriptions these pairs, the Greek and the Aramaic, are not used in parallel, a fact that has some signifïcance. 15

In this regard, the discussion by Goitein, 1973, p. 14-15, of the walàl in the médiéval Jewish trading communities and the qadi in that of the Moslems is especially interesting. This was a représentative of merchants who served to protect the interest of those who were away, kept a warehouse for storage, a bourse for business transactions, and a mail service, among his many services. See also référencé to the qadi in the Chinese port cities in the Tang; Dien, 1962, p. 340. 16 Hillers, Cussini, 1996. (Hereafter PAT) 17 As Will, 1992, p. 59, points out, only a few of the statues in stone hâve survived; those of bronze were probably melted down when the city was sacked in 273. 18 Gawlikowski, 1997, p. 142-43. 19 The PAT number by which the inscriptions are indicated here refer to the number in Hillers and Cussini, and G is the number in the Gawlikowski list, followed by the date in parenthèses.

198

Les Sogdiens en Chine

In the simplest of the inscriptions, no title is mentioned for tlie caravan leader rather saying that the person in question, in each case, had gone with the caravan, to or from Chasinou Charax, 20 Phorat,21 or Vologesias,22 and dedicating the statue in his honor. In two such cases no further details were given.23 In another the Greek added that the dedicatee had pleased the merchants while in the Aramaic it was said he had helped them.24 In still another,25 for which there is only the Aramaic version, the Statement is that Marcus Ulpius Yarhai had qm w’drnwn bkl sbw klh which is translated by J. Cantineau as “il s’est tenu (à sa tête) et qu’il l’a aidée en toute chose.”26 R. Drexhage has “er ihr vorstand und in allen Dingen geholfen hat,”27 where “vorstand” may be taken in the sense of “directed.” However, qm, tlie 3rd person masculine, singulär, perfective of qwm “to stand, to be in charge of, to erect, to establish,” is here perhaps better understood as the marker of ingressive aspect before a following verb joined by w, as in the example “he sei to work and . . . .”28 So it may well be that Marcus Ulpius Yarhai did not accompany the caravan. However, see another use of qm below. At any rate, none of the spécifie titles for caravan leader are found in these particular inscriptions. Marcus Ulpius Yarhai was an important person of the mid-second Century and there are three other dedications in which it is clear that he did not accompany the Caravans in question. In one of these, dated 155, the dedication is by one Zabdilla, the caravan leader (synodiarch in Greek and rb syrt’ in Aramaic).29 In another, dated 157,30 the dedication on behalf of the caravan is by Yaddaiou, son of Zabdilla, who in the Aramaic is termed bsyrt’ which may be an abbreviation or an error for brbnwt syrt’ “while he was caravan leader” or “during the term as caravan leader.”31 In the third, dated 159, the caravan that made the dedication is said to hâve come from Charax Spasinou under his son Abgar.32 Another prominant person of the period, associated with Vologesias, is Soados, who is also honored in two surviving dedications, dated 132 and 144. These specifically 20 Hansman, 1967; Mathews, 1984, p. 165; Fraser, 1996 p. 168-69. This was the capital of Mesene or Maishän, on which see Gawlikowski, 1994, p. 28-29. Founded by Hyspaosines, originally a Seleucid govemor of the area, the name of the capital, Charax Spasinou, means Palisade of Spasines or Hyspaosines; Hansman, 1967, p. 23-24. The usual name in the Aramaic texts is krk mysn. Maishän was mentioned as the “meeting-place of the merchants of the East” and “the haven of the merchants, That sitteth on the shore of the sea” by St. Thomas; see Bevan, 1897, p. 15 and 25. 21 Hansman, 1967, p. 46-53; Mathews, 1984, p. 165. 22 On the location of Vologesias, see Maricq, 1959; disagreeing with him are Chaumont, 1974, p. 77-81; and Gawlikowski, 1994, p. 29-30. See also Mathews, 1984, p. 165-66. 23 PAT 1366/G4 (28-88) and PAT 1376/G5 981). 24 PAT0279/G31 (247). 25 PAT 1411/G17 (156). 26 Cantineau, 1938, p, 75. 27 Drexhage, 1988, p. 43, note 86. 28 Hillers, Cussini, 1996, p. 405. See also Sokoloff, 2002, p. 480, sub qwm, #6, as an auxiliary verb of emphasis. 29 PAT 0274/G36 (155).

30PAT 1399/G19(157). 31 32

Drexhage, 1988, p. 43, note 88. PAT 1409/G21 (159).

Albert E. Dien

Caravans and Caravan Leaders

199

crédit him with saving the caravans from dangers that threatened it, most probably from attacks by the desert nomads. The dedicators in both cases, two for each caravan, carry the titles of synodiarch and brbnwt syrt’P In other instances the honors were extended by various merchant groups to officiais who had facilitated caravans in some other way, such as satrap,34 centurion,35 archon,36 customs contracter,37 or cavalry captain.38 In these, the only mention of caravan leader is when Nesa, brother of Soados, honored the customs contracter; he is entitled rs syrt’ in Aramaic, but in the Greek the verb proêgoumenou “one who goes before to show the way, a guide” is used. The merchant members of the caravan (emporoi or synodia in Greek, tgyr or bny syrt’ in Aramaic) might also show special gratitude to their leader. Nesa, in 142 and 150, on trips to and from Vologesias, was entitled synodiarch in the Greek while the Aramaic in one case simply said they had gone with him39 while in the other he is said to qm brshvm w’drnwn bkl sbw klh “to hâve stood at their head and helped them in every way.”40 This usage would need to be considered in understanding the sense of qm cited above. In 193 Taimarsu, the synodiarch and rb syrt was honored for having made a donation of 300 golden dinarii toward the expenses of the caravan.41 In 211 his son Yaddai, also a synodiarch (the Aramaic is too fragmented to read) “pleased” and “relieved” the caravan members, but no details are given.42 This may indicate that the position of caravan leader was hereditary in some fashion or other. The significant portion of the honorary inscriptions that relate to the caravan trade, 31 of 181, or some 16%, indicate that such commerce was well embedded in the Palmyrean economy, so it is no surprise that the city council démonstrated its support by honoring any individual who assisted in the commercial enterprise43 and did not stint his person and fonds.44 Julius Aurelius Nebonai, a synodiarch (the Aramaic is too fragmented) in an undated inscription was honored for bringing back the caravan.45 In two late pièces, of 25746 and 260+,47 the person honored is termed archemporos “chief merchant” but since the Aramaic in the 257 inscription has rb syrt’, it would indicate that archemporos and synodiarch are synonymous. In the other (for which the Aramaic 33

PAT 0197/G14 (132) and G15 (144). The latter is not included in Hillers and Cussini. PAT 1374/G7(131). 3' PAT 1397/G8(135). 36 PAT 1412/G10 (140) and Gl 1 (89-188), only in Greek; see Rostovtzefg 1935, p. 143, and Schlumberger, 1961, p. 256. 37PAT 1373/G25(161). 38 PAT 1422 (no date). This was sponsored by Marcus Ulpius Yarhai, showing his concem fbr the safety of the caravans. 39 PAT 1419/G13 (150). 34

4°PAT0262/G12(142). 41

PAT 0294/G28 (193).

42pAT0295/G30(210). *PAT1421/G6 (86). 44PAT1414/G9(138). 45

PAT 1360/G31 (3rd Century).

46PAT0282/G32. 47

PAT 0288/G34.

Les Sogdiens en Chine

200

is too fragmentary to read) the council expressed their approval of Septimus Worod, who lield high offices, for having funded the caravans on his own.48 Another inscription clearly indicates the high esteem in which the caravan leader was held, and again the question of the funding of these caravans is mentioned:

PAT 1378/G29 (199) Greek: By decree of the council and the people, the four tribes [are to erect statues to] ‘Ogeilu, [son of] Maqqai, son of ‘Ogeilu, son of Agegu, son of Shewira, because of his goodness and manly spirit, and for his having given satisfaction through continuée! commands (strategiaï) against the nomads, and having provided safety for the merchants and the caravans in ail his commands of caravans (synodiarchai) and having spent much from his own money, and in every tenure of office accomplished brightly and gloriousiy, in his honor.49

Aramaic: From a decree of the Senate and people, these four statues of ‘Ogeilu, son of Maqqai ‘Ogeilu Shewira, from the four tribes is erected in his honor, as he has been accommodating to them in many campaigns (strtgwn) and because of his support given to caravans with which. he has gone, and he has spent much gold of his own to help the merchants in many ways, and he conducted his public life in a praiseworthy and brilliant way.

The Aramaic thus follows closely the Greek, but makes no mention of his having served as a caravan leader. The rôle of the caravan leader must hâve involved the organization of the caravan itself, but unfortunately there is little information available. Early on, Michael Rostovtzeff suggested that the synodiarch was little more than a specialist, or a technician, one who was engaged for the joumey and in tum hired the animais, camels and horses, and the personnel to care for them, and who guided the party through the desert. In addition to doing all the preliminary tasks such as obtaining the necessary food and water, he also protected the party against attacks by nomads and carried out any diplomatie negotiations with the relevant authorities.50 But Rostovzeff believed that the members of the caravan were the merchant-princes who formed into a company for each joumey and chose from among themselves the archemporos who might also be the synodiarch, but not necessarily so. The caravan would simply disband at the end of each trip. Fergus Millar has gone further and claimed that there was a public position called synodiarch and another called archemporos 1 believe that the terminology in the inscriptions argues against this. The fact that at times when one language uses the term for caravan leader and the other does not, as noted above, this would argue against it being the title of a position, and further, we hâve seen that both synodiarch and archemporos are matched with the same Aramaic term would tend to equate them rather than distinguish them.52 48

See note 61 below.

'°50 Part of this translation is from Millar, 1998, p. 132.

Rostovtzeff, 1932a, p. 806. Millar, 1998, p. 133. 52 Drexhage, 1998, p. 94, agréés that the inscriptions do not make any significant hierarchical distinction 51

Albert E. Dien

Caravans and Caravan Leaders

201

Ernest Will, in 1957, in his discussion emphasized the complexity of the caravan’s Organization, for beside the caravan leaders and merchants, there were the funders or entrepeneurs, the fonduqs or trading communities outside of Palmyra, and the strategoi who provided for the caravan’s security, including any necessary diplomatie negotiations.*53 Will's emphasis on these other rôles in the caravan trade tended perhaps to reduce the importance of the synodiarch. This led Will to wonder a bit why such an important person as Marcus Ulpius Yarhai, who was mentioned above, may have led some caravans, but it may be that the evidence of his having done so is rather shaky.54 Rather, Marcus Ulpius Yarhai emerges as one of the major players in the Palmyrean commercial scene, quite likely one who supplied capital necessary to carry on the trade. Niels Steensgard, in his Carracks, Caravans, and Companies, admittedly which deals with the 17th Century, speaks of commenda, that is, capital extended to a merchant by a patron, in retum for a share in the anticipated profits.55 Possibly such a rôle would go to explain the contributions cited in the inscriptions that eamed the dedication of statues, that is to say, such displays of évergète or social benefaction might have been tempered by a desire to ensure the success of the investment. The other prominent Palmyrean cited above, Soados, as Will points out, seems rather to have been a patron, perhaps we should say head, of the fonduq at Vologesias.56 The Palmyrean merchants had established such trading communities widely; besides Vologesias there is mention of Dura-Europos, Spasinou Charax, Babylon, Seleucia on the Tigris, and beyond, even in Egypt and at Rome.57 These would have their counterpart in the Sogdian communities across Central Asia and into China, but unfbrtunately, we do not know of any spécifie term carried by the heads of those Palmyrean communities. Given the prominence of a man like Soados, he might be considered the équivalent of a Sogdian maha särthaväha^ J.F. Mathews, in an eloquent article, pictures such men as Marcus Ulpius Yarhai and Soados as having been Bédouin sheiks who brought to Palmyra their close connections with the desert nomads and so could “police” the desert and provide protection to the caravans; while not themselves merchant-princes, they could serve as

between the two ternis. Will, 1957, p. 268-69. 54 The two inscriptions cited by Will, 1957, p. 269, to show that Marcus Ulpius Yarhai led caravans are not unambiguous. In the first, PAT 0307, the words “went with him” (slqw mh) are restored, and in the second, PAT 1411, the verb qm “to be the head of,” as suggested above may, not necessarily indicate that he went with the caravan in person. For PAT 0307, Will cites Starcky, 1949, #89, p. 56-57. Starcky, in tum, cites Sobemheim, 1905, p. 29, but that citation is an error. 55 Steensgard, 1973, p. 22-24. 56 Wiü, 1957, p. 270. 57 Rostovtzeff, 1932b, p. 144-45, emphasizes the importance of those whom he caUs presidents of fondouqs, though he offers no spécifie term in the inscriptions with which he equates that position. On those Palmyrean communities at Charax and Vologesias, see Gawlikoski, 1994, p. 129. For Egypt and Rome, see Will, 1997, p. 127. 58 Gawlikowski, 1983, p. 65, suggests that Marcus Ulpius Yarhai himself may have been a résident of Charax Spasinou. 53

202

_

Les Sogdiens en Chine

protectors and patrons of the merchants.*59 Once Palmyra feil to Roman amies in 273, they simply merged back into the desert, having enjoyed, as Mathews says, “a phase of magnificent, but relatively short-lived, urban grandeur.”60 Such a vision may well be true, but the inscriptions do not demonstrate this, and who these men were, what was their rôle in that society, and much about the caravan trade itself, remains tantalizing vague. What one can say, then, is that the term synodiarch / rb syrt etc. was applied to the person in charge of a caravan for the duration of a trip. The duties could be shared by two who were both entitled caravan leaders. Since the person who had that responsibility was on occasion designated by a descriptive term such as “guide” it seems dear that synodiarch / rb syrt\ etc. was not the title of an official post with a spécifie term of office. During the late third Century, the term archemporos appears as an altemate appelation. There is at least one case of a father and son having both served as caravan leaders, indicating some possibility of a family metier. While we do not know the précisé status of those who undertook this responsibility, they were lionored, some were from prominent families, or held other, prestigious positions.61 Even if Marcus Ulpius Yarhai may not hâve personally led a caravan, his son did, as did the brother of Soados. Some were honored for defraying the expenses of the caravan, even to the extent of 300 golden dinar, and having facilitated the safe arrivai of the caravan in other ways, not specified, but which could well hâve required financial outlay. Not only were statues dedicated to them by gratefiil caravaneers or by “the council and the people,” but they themselves could dedicate such statues to benefactors and officiais who expedited the safe passage of the caravan. Finally, there were other men in that society to whom statues were dedicated for their rôle in the caravan trade and who may well hâve been the patrons, bankers, or merchant-princes, however we wish to identify them. When we compare what we know about Palmyra with the Sogdians in Central Asia and China, the différences may seem to outweigh any similarities. The Palmyrean evidence is largely limited to Palmyra itself while that for the Sogdians is to communities in a foreign land. There were caravan leaders in both societies, but whereas the title of caravan leader appears to hâve been carried by the heads of the Sogdian fonduqs, such was not the case, as far as we know, in the Palmyrean case. The Sogdian letter II, written back to the “home office” in Samarkand, indicates in the Sogdian case, too, that there were patrons of the trade and, in both, the relationship

"A few centuries later, the Syrian deserts remained a place known for its lawlessness, and the caravans found it necessary to pay for armed guards to obtain safe passage; Goitein, 1973, p. 10. 60 Mathews, 1994, p. 169. Crouch, 1972, p. 242, goes so far as to characterize the Palmyrene merchants as “semi-nomadic.” 61 “The council and the people, to Septimus Worod, a remarkable man, administrator of the emperor, dukinarian (pay grade of 200,000 sesterces), procurator of the mother colony, who hrought back a caravan with his own money, and who was approved by the testimony of the merchant leaders (archemporoi), who served splendidly as strategos and clerk of the market of the same mother colony, and who spent a great deal of his own money, and who pleased that same council and people, and is now president of a symposium (drinking party) of the temple of the god Bel, because of agneas (strict observance of religion?) and his honor.” PAT 0288/G34 (after 260). The Aramaic text is too fragmentary to be read.

Albert E. Dien

Caravans and Caravan Leaders

203

between the patron and the caravane er may hâve been on a commenda basis. That is, the bny of the bny syrt’ and the 260 Sogdian merchants intercepted by the govemor of Liangzhou may all hâve been independent businessmen, or merchants, operating on capital extended by banker-investors. But while the Sogdians were probably engaged in a peddling trade, Palmyra would seem to hâve been more of a transshipment center. In both cases, they relied on the caravan leader to bring them safely to their destination.

ADDENDUM On another level, is there possibly a linguistic connection between the terms sârtha and syrt’l The Sanskrit sârtha is explained as deriving froni the prefix sa “with” and artha “aim, goal,” resulting in sârtha “having meaning, having an aim or object, useftil, serviceable” and by extension, such other meanings as “merchant, caravan, and wealthy man.”62 Sârtha as “merchant” appears in the literature rather late. In discussing the terni as it appears in a Brahmi inscription, Puri cites two works, the Saundarananda and the Buddhacharita, both by Asvaghosa, a Buddhist poet who is thought to hâve lived at the time of Kaniska, sometime in the fïrst or second centuries CE.63 There is a possibility, on the face of it, that this is a folk etymology and sârtha as merchant was attached to an original sârtha to provide an explanation for a loanword into Sanskrit. The Aramaic syrt’ is listed as the emphatic form of syrh, the nominal form.64 In Palestinian Aramaic, the word is entered as syyr, “caravan leader” and syyrth, “caravan,” from a root swr, “to walk in a line,”65 which again may well be a folk etymology. It is the coincidence that the word fbr caravan in the two languages hâve the form *SRT that leads me to speculate that they may be related. Further, Sardis, that served as a Seleucid capital, and from earlier times was an important commercial and industrial center,66 possibly enters into this équation as another example of the name of the city from which foreign méchants hailed came to be the word for merchants. While the name of the city in Aramaic is Sparda, in the native Lydian it appears to hâve been “Sfard-” or perhaps “Sward-.”67 At any rate, it may be possible that the Indians heard and transcribed the initial syllable of the Lydian name in a fashion similar to that done by the Greeks. This topic is still in the process of study, and I would plan to present further discussion of it at a later time.

“Apte, 1978, p. 983. 63

Puri, 1945, p. 84; Mykius, 1988, p. 345-46; Keith, 2002, p. 55-59. Hillers, Cussini, 1996, p. 414. 65 Sokoloff, 2002, p. 547. 66 Sherwin-White, Kuhrt, 1993, p. 180-84; Hanfmann, 1983, p. 147. Pedley, 1968, p. 75, cites Herodotus who said the Lydians were the fïrst to mint coins and to become kapeloi, which Pedley says seems to mean “a merchant or dealer of some sort.” 67 Dusinberre, 2003, p. 7, n. 24. Gusmani, 1964, p. 202.; Bossert, 1944, p. 130. 64

Les Sogdiens en Chine

204 Bibliograpliy

Apte, Vaman Shivram

1978 The Practical Sanskrit-English Dictionary (Delhi: Motilal Banarsidass). Bevan, Anthony Ashley 1897 The Hymn of the Soul contained in the Syriac Acts of S/. Thomas (Cambridge: Cambridge University Press). Bossert, Helmuth Th. Geschichte und Entzifferung Ein hethitisches Konigssiegel: neue Beiträge 1944 der hethitischen Hieroglyphenschrift (Istanbuler Forschungen, 17) Berlin. Breasted, James H. 1906-7 Ancient Records of Egypt: historical documents from the earliest time to Persian conquest, collected, edited and translated into commentary (Chicago: University of Chicago). Cantineau, J. “Tadmorea (suite),” Syria 19, p. 72-82. 1938 CHAUMONT, Marie-Louise “Études d’histoire parthe, III: Les villes fondées par les Vologèse/5 Syria 51, 1974 p. 77-81. Crouch, Dora 1972 “A Note on the population and area of Palmyra,” Mélanges de l’Université Saint Joseph 47, p. 241-50. Dien, Albert E. “The Sa-Pao Problem Re-examined,” Journal of the American Oriental 1962 Society 82, p. 335-42. Drexhage, Raphaela Untersuchungen zum römischen osthandel (Bonn: Dr. Rudolf Habelt GMBA). 1998 Dusinberjœ, Eispeth R.M. 2003 Aspects of Empire in Achaemenid Sardis (Cambridge: Cambridge University Press), Fan, Xiangyong (Shanghai: Guji chubanshe). Luoyang qielanji jiaozhu 1978 Fraser, P.M. Cities of Alexander the Great (Oxford: Clarendon Press). 1996 Gawlikowski, Michael “Palmyre et l’Euphrate.” Syria 60, p. 53-68. 1983 1994 “Palmyra as a Trading Centre,” Iraq 56, p. 27-33. “Palmyra and its Caravan Trade,” Palmyra and the Silk Road (Special Issue of 1997 Annales Archéologiques Arabes Syriennes: Revue d’Archéologie et d’Histoire), (Damascus: Dire ctorateGeneral of Antiquities and Museums, Syrian Arab Republic), p. 139-44. Goitein, S.D. 1973 Leiters of Médiéval Jemsh Traders (Princeton: Princeton University Press). GRENET, Frantz, Sims-Williams Nicholas 1987 “The Historical Context of the Sogdian Ancient Letters,” in Transition Periods Symposium de Fribourg-en-Brisgau (22-24 Mai, in Iranian History: Actes 1985), p. 101-122. , Grenet Frantz, Sims-Williams Nicholas, DE LA VAISSIÈRE Étienne “The Sogdian Ancient Letter V,” Bulletin of the Asia Institute 12, p. 91-104. 1998 Gusmani, Roberto

Albert E. Dien

Caravans and Caravan Leaders

205

1 964

Lydisches Wörterbuch mit grammatischer Skizze und Inschriftensammlung (Heidelberg: Carl Winter, Universitätsverlag). Hanfmann, George M.A. Sardis from Prehistoric to Roman Times: Results of the Archaeological 1983 Exploration of Sardis 1958-1975 (Cambridge: Harvard University Press). Hansman John 1967 “Charax and the Karkheh,” Iranica Antiqua 7, p. 21-58. Hillers, Delbert R., CussiNl, Eleonora 1996 Palmyrene Aramaic Texts (Baltimore and London: The Johns Hopkins University Press). Jenner, W.J.f.

1981

Memories of Old Loyang: Yang Hsüan-chih and /osf capital (493-535) (Oxford: Clarendon Press). KEITH, A. Berriedale 2002 The Development and History of Sanskrit Literature (Delhi: Sanjay Prakashan). KONOW, Sten, Venkayya, V. 1909-10 Epigraphia Indica and Record of the Archaeological Survey of India, vol. 10, 1 (Calcutta: Superintendant Government Printing, India). MARICQ, A. 1959 “Vologésias, l’emporium de Ctesiphon,” Syria 36, p. 264-76. Mathews J.F. 1984 “The Tax Law of Palmyra: Evidence for Economie History in a City of the Roman East, Journal of Roman Studies 74. p. 157-80. Millar, Fergus 1998 “Caravan Cities: The Roman Near East and Long-distance Trade by Land,” in Michael Austin et al., eds., Modus Operandi: Essays in Honour of Geoffrey Rickman (London: University of London), p. 119-37. Mykius, Klaus

1988

Geschichte der altindischen Literatur (Leipzig: Wissenschaftliche Buch¬ gesellschaft).

Pedley, John G.

1968 Postgate, J.N.

1992 Puri, B.N. 1940 1945

Sardis in the Age of Croesus (Norman: University of Oklahoma). Early Mesopotamia: Society and Economy at the Dawn of Civilization (London: Routledge).

“Economie Guilds in the Kusäna Period,” Indian Culture 6.4. p. 421-428. “Some Aspects of Economie Life in the Kushana Period,” Indian Culture 12.1; p. 82-9. Raschke, Manfred G. 1978 “New Studies in Roman Commerce with the East,” in Hildegard Temporini and Wolfgang Haase, eds., Aufstieg und Niedergang der Römischen Welt: Geschichte und Kultur Roms im Spiegel der neueren Forschung (Berlin: Walter de Gruyter). Rostovtzeff, M. 1932a “Les inscriptions caravanières de Palmyre,” Mélanges Gustave Glotz, II. (Paris: Presses universitaire de France), p. 792-811. 1932b Caravan Cities, D. and T. Talbot, transi., (Oxford: Clarendon Press). 1935 “Une nouvelle inscription caravanière de Palmyre,” Berytus 2, p. 143-8. Schlumberger, Daniel 1961 “Palmyre et la Mésène,” Syria 38, p. 256-60. Schoff, W.H. 1912 The Periplus of Erythean Sea: Travel and Trade in Indian Ocean by a Merchant of the First Century (London: Longmans, Green, and Co.).

206

Les Sogdiens en Chine

Sherwin- White, Susan, Kuhrt, Amélie 1993 From Samarkand to Sardis: / new approach to the Seleucid empire (London:

Duckworth).

Sims-Williams5 Nicholas 1985

“Ancient Letters,” in Ehsan Yarshater5 ed.* Encyclopedia Iranïca (London: Routledge & Kegan Paul), II.l5p. 7-9.

SOBERNHEIM, Moritz 1905 “Palmyrenische Inschriften“, Mitteilungen der Vorderasiatischen Gesellschaft 10 (Berlin: Wolf Peiser Verlag), p. 1-56. Sokoloff, Michael / Dictionary of Jewish Palestinian Aramaic (Ramat-Gan: Bar Ilan University 2002 Press, 2nd édition). Starcky, J. 1949 Inventaire des inscriptions de Palmyre: vol. 10. L ’Agora (Damaseus: Publication de la Direction Générale des Antiquités de Syrie, #89), Steensgard, Niels 1973 Carracks, Caravans and Companies: The structural crisis in European-Äsian trade % the early 17 th Century (Odense: Studentlitteratur) Tafazzoli, A. “A List of Trades and Crafts in the Sassanian Period,” Archäologische 1974 Mitteilungen ms Iran 7. p. 191-6. Thomas, F.W. 1951 Tibetan Literary Texts and Documents conceming Chinese Turkestan (London: Luzac and Co., Ltd) Wang, Yi-t’ung 1984 A Record of Buddhist Monasteries in Lo-yang (Princeton: Princeton University Press). 0

WEÏSHU

1973

Wei shu

(Beijing: Zhonghua shuju).

Will Ernest

1957 1992 1997

“Marchands et chefs de caravanes à Palmyre,” Syria 34, p. 262-77, palmyréns: Venise des sables (Paris: Armand Colin). “Palmyre et les Routes de la Soie”, Palmyra and the Silk Road (Special Issue of Les Annales Archéologiques Arabes Syriennes: Revue d’Archéologie et d’Histoire), (Damascus: Directorate General of Antiquities and Museums, Syrian Arab Republic), p. 125-8.

Zhoushu 1 97 1

Zhou shu

(Beijing: Zhonghua shuju).

Sabao or Sabo: Sogdian Caravan Leaders in the Wall-Paintings in Buddhist Caves

RONG Xinjiang

榮 新江

Research on Sabao and Sabo In my paper "Sabao 薩 保 and Sabo 薩 薄: On the Problem of the Leader of the Sogdian Colonies during the Northern Dynasties of the Sui and Tang periods" that I presented during the Third Conference of Iranian Studies in China (organized by the Research Institute of Iranian Culture, Peking University, Nov. 13-14, 2002), I outlined what the result of previous research of Sabao and Sabo had yielded and reviewed the arguments on these two titles in detail. The work of re-examining relevant original data, enabled me to stress the significant distinction between Sabao and Sabo in ancient Chinese sources. Sabo in the Chinese Buddhist canon was a transcription of the Sanskrit särthaväha. It originally designated a caravan leader, and is transcribed as shangzhu 商主 "a caravan leader55, da shangzhu 大商主 "a great caravan leader55, daoshou 導首 "a guide", daoshi 導師 "an instructor", zhong zhi daoshi 衆之導師 "an instructor of the masses", guke 客 "a foreign merchant", and shangren 商人 ua merchanf9. Most sabo in Buddhist literature appear in legends. Basically, they are described as rieh merchants who outfït ships in order to lead merchants to seek treasures. A sabo may also be a businessman dedicated to the salvation of the masses when misfortune or disasters befall. These dedicated merchant leaders are, in fact, none other than a previous incarnation of the Buddha. Meanwhile sabao. the transcription of in Sogdian, has undergone a change of meaning from "caravan leader" to "Leader (chief) of a colony,J now translated as shouling 首領 "a leader” or da shouling 大首領 "a great leader55. The majority of sabao no matter whether they themselves held official titles in the Northern Dynasties of the Sui and the Tang, or whether their forefathers, when mentioned in epitaphs during the Tang dynasty, actually occupied such positions or were, in tact, Sogdians from the various countries in Sogdiana or other places in Central Asia. Consequently, although the two titles of sabo and sabao share some concrète features, they should by no means be confused with each other.1 1

Rong Xinjiang, 2003c. It has been translated into English: "Sabao and Sabo\ On the Problem of the Leader of Sogdian Colonies during the Northern Dynasties, Sui and Tang Period". Paper presented to: Crossing the Borders of China: A Conférence on Cross-cultural interactions in Honor of Professor Victor H. Mair, December 5-7, 2003, University of Pennsylvania, Philadelphia.

208

Les Sogdiens en Chine

In another paper titled “The Religions in Foreign Colonies and the Social fonctions of Zoroastrian temples in the Northern dynasties of the Sui and the Tang/’ (in Chinese), I attempted to clarify some of the erroneous conceptions regarding sabao and Buddhism. I discussed there terms such as sabao zi for the Buddhist disciple Yuwen Hu of the Northern Zhou dynasty, or mohe from the Sanskrit, in mohe da sabao found in Tang epitaphs5 or the fact that merchant He Yongkang registered as a secretary of the safu ( = sabao). paid to redeem a Buddhist temple, as well as the fact that sabao was used as a suffix by some Buddhists, such as Qi Sabao An Sabao Mi Sabao and so on. I furthermore demonstrated that it was none other than Zoroastrianism that was main religious belief in Sogdian colonies? However, since these matters were fully discussed in the above-mentioned paper, I shall not allude to îhem again here, so as not to deviate from my main subject of sabao and sabo I shall now tum to a question that has intrigued mes which is that of the images of sabo found in the murais in Buddhist caves. A Conference I attended proved enlightening on this point. The Conference in question was The International Conference on the History of Sino-Foreign Relationships: Investigation and Studies of New Materials (organized by the Center for the Study of Médiéval Chinese History, Peking University, Nov. 15-16, 2002). It immediately followed the Third Conference of Iranian Studies in China during which Zhang Qingjie presented his paper: “Foreign Merchant Figures, Pictures and Documents from the Northern Dynasties of the Sui and Tang Periods” (in Chinese). This paper proved to be a rieh source of material related to the question that intrigues me. Mr. Zhang pointed out that: “Stories of sabo and merchants frequently appear in wall paintings, for instance in caves Nos. 8, 17, 38, 114 and 184 of the Kizil Caves in Xinjiang. The paintings illustrate a number of stories among which one of the most famous is the taie of a sabo using his arms as a torches to provide light for the people. There is also the story of a Dragon King named Mabi rescuing merchants at sea, and another story of a sabo or caravan leader, also helping merchants who were in danger of drowning. These stories even appear several times in the same cave.” Mr. Zhang provided a detailed introduction to the basic content of these three stories and their surfacing in Chinese translations of Buddhist scriptures. In the same paper the author also discussed the appearance of foreign merchants in the wall paintings of the Dunhuang caves and refers to the illustrations of the Sutra of Seven Good Deeds in Cave No. 296 of the Northern Zhou as an example. Other examples are the paintings of merchants encountering bandits in Cave No. 420 of the Sui dynasty, and in Cave No. 45 of the High Tang period, as well as the painting of merchants and travellers in Cave No. 103.2 3 Not only did Mr. Zhang place emphasis on the images of foreign merchants in the wall paintings of Buddhist caves, but he also called on a great many artefacts such as pottery figures, sculptures, carved stones and tomb murais, in order to substantiate his Claims. •

2 3

Rong Xinjiang (ed.), 2003a, pp. 385-412, figs.1-9. For the paper please see Zhang Qingjie, 2004, especially pp. 199-201 and 189-92.

Rong Xinjiang

Sabao

Sabo

209

”

Sabo and Merchants in Murals in the Buddhist Caves of Kucha Paintings of stories featuring sabo have been preserved in the Kizil Caves, the Kumtura Caves, the Simsim Caves, the Kizil-Qargha 克孜 爾 哈Caves and the Taitai'er 台台 爾 Caves, all located in the régions today known as Kucha and Baicheng, once occupied by the ancient Kingdom of Kucha. We have an especially rich collection of such murais in the Kizil Caves3 and images there are comparatively well preserved。 Therefbre, in the present paper3 I will generally, rely on the Kizil material, and if necessary, provide a new introduction to similar paintings found elsewhere« I shall then quote from the Chinese versions of the stories found in the paintings, taken from the Chinese translations of Buddhist scriptures. But before dealing with these topics, I would like to call attention to two spécifie points, First, the dating of the Kizil Caves, especially the date when construction began and the last possible date when it ended,* 4 a subject that has long been a matter of controversy and still remain a problem for the author. My argument draws on the general conclusion of the Catalogue of the Contents of Kizil Caves but does not review other important arguments in detail, though they are also available in the catalogue 5. Second, in our material, there are usually no Chinese inscriptions in the Kizil Caves from the fburth to the seventh Century AD. Consequently, there are good reasons for believing that the stories in the paintings did not draw their subject matter from Chinese texts. Instead, they are likely to be récréations of Sanskrit or Tocharian texts circulating locally at the time. Or they might have developed from drawings. However, in comparison with their Chinese counterparts, the Sanskrit or Tocharian versions are not as well preserved. In order to fîll the gaps in these sources, I will also quote from the Chinese versions of the stories. We are able to trace almost ail the images mentioned in this article to the Xianyu jing 賢愚經 (the Sutra of the Wise and the Foolish). In accordance with Sengyou5s 僧祐 quotation of the Xianyu jing ji 賢愚經記 in Chu Sanzang jiji 出三 藏記集 (Collected Notes on the Formation of the Tripitaka, ca. 506-512), the sramana of Hexi ["West of the River", i.e. Gansu] Tanxue 曇學, Weide 威德, and others, formed a group of eight monks, who cooperated in the compilation of this text in Gao chang in 445. They relied on their memories of what they had heard from various scholars preaching the Tripitaka and the sutras, or lecturing on the vinaya in Khotan.6 This fact demonstrates that the stories recorded in the Xianyu jing were populär around the northem and southem routes along the Tarim basin. And this popularity explains the repeated appearances of these stories in the Kucha caves, sometimes even the répétition of the same story in a single cave. In fact, many other populär legends from Buddhist literatures are also '

' For the major critical révisions of dates for Kizil Caves, please see A. von Le Coq and E. Waldschmidt, 1933, pp. 24-9; Su Bai, 1989, pp. 10-22 (Also see Su Bai, 1996, pp. 21-38); Huo Xuchu, Wang Jianlin, 1993, (also see Huo Xuchu, 1994, pp. 1-30); Huo Xuchu, 1993, (also see Huo Xuchu, 2002); Liao Yang, 2001; and Li Chongfeng, 2002, pp. 160-76. 5 Xinjiang Qiu ci shiku yanjiusuo (ed.), 2000. 6 Taishö shinshü daizökyö, Vol. 55, No. 202, p. 67. For details of Xianyu jing, see S. Lévi, 1925; V. H. Mair, 1993; idem., 1999.

210

_

Les Sogdiens en Chine

preserved in Kucha Caves7. Now, given that the paintings illustrate stories from Buddhist literature, the figures in the paintings should, logically, be Indians. However, the facts do not appear to match this simple assumption.

Legend of the Pradïpapradyota Kingdom This story has been recorded in many scriptures. For the purposes of our discussion I will quote the two most detailed versions of the taie. Chap. 28 of Xianyu jing VoL 6 States:

I hâve heard that once when the Buddha stayed in the Jetavana Park in the Shrâvastï Ananda asked h im: "I wonder what you did to to repel darkiiess in the Kingdom, previous world?” The Buddha answered: “Long ago, on the continent of Jambudvîpa, fïve hundred merchants were trapped in a dark valley on their way through inhospitable surroundings. The merchants were afraid of losing their money and their goods, for robbers frequently haunted the valley. They wailed miserably and prayed to ail the deities of heaven and earth, and to those of the sun and the moon, as well as to the spirits of mountains and (caravan leader) felt compassion for their helplessness and told seas. The sabo zhu them: cBe not distress and fear not. I will be able to provide you with light.’ Then he wound white bandages around both of his arms, saturated them with ghee, lit them up as torches, and led the merchants out of the valley after seven days. The merchants greatly appreciate his kindness act of boundless mercy. They all feit relieved and joyful.” The Buddha then told Änanda: “Is the sabo then other than my previous incarnation? No. Are the five hundred merchants then other than the five hundred arhats? Neither.”8

If we now tum to in Chap. 4 on Cakravartin in Jinglü yixiang it States: and Baochang by the Liang monks Sengmin

compiled

There was once a Räk^asa who haunted merchants as if he were their shadow and frequently harassed them. One day this same Räk^asa raised a fierce wind that caused the merchants to lose their way and begin to wail miserably. They prayed to Mahesvara, and deities of water, earth, fire and wind and called upon their parents, their wives and sons as well as on other relatives fbr help. The Buddha told a good man: "As soon as I heard their calls, by pure divine ear and traced them to where they were floundering, I managed to comfbrt them with gentle words: ‘Do not be afraid. I will show you the way and deliver you back to Jambudvïpa safe and sound. Good man, I will wind white bandages around both my arms, saturated them with ghee, light them up as torches and talked the truth. My buming arms will last seven days which is long enough for you to retum to Jambudvïpa. ...Good man, I have done many things in the past. For example, I allowed various treasures to rain down on a caravan leader seven times a day in order to satisfy the needs of people.”

The two versions of the story are slightly different. In the Xianyu jing it is the dark valley that prevents the merchants from advancing. They fïnally escape from the danger 7 8

For more details please see Liang Liling, 2002, pp. 521-44 ('Xianyu jing Murais in Xinjiang Caves”). Taishô shinshü daizôkyô^ Vol. 4, No. 202, pp. 392-3.

Rong Xinjiang

Sabao or Sabo

211

with the help of the sabo. In the later version, in which Räkgasa causes a tierce wind to blow, the story does not directly identify the sabo leader as an incarnation of the Buddha Obviously, the Xianyu jing version is doser to the legend depicted in the Kucha Cave painting than is the Jinglü yixiang version. Murais with this theme exist in the following locations: In Kizil Cave No. 38 (see fig. 1, plate 18) the east slope of the ceiling shows Sabo raising his buming arms as torches, while a merchant looks into the distance with his right hand raised and his left hand holding a stick. Standing next to them are an ox and a pack donkey*9. The merchant has deep-set eyes and high-bridged nose. He is dressed in a tight-sleeved blue coat and wears a white, high-crowned, beret-like hat. The cave dates roughly from the fourth Century A.D. In Kizil Cave No. 17 (fig. 1) in a rhomboid niche of the east slope of the ceiling there is a painting of a Sabo raising his buming arms as torches. Next to him are two merchants who extend their arms towards the torches. Behind the merchants stands a pack camel.10 The cave dates firom the sixth Century AD. The merchant at the front is dressed in a white-spotted, tightfitting robe with brown lapels, and wears green trousers and high boots. The figure behind him is dressed in a brown-trimmed white skin-tight robe with a round collar, and wears white trousers. Both the robe with lapels and the one with a round collar are similar to Sogdian style clothes as depicted on An Jia’s screen. For Fig. 1: a Sabo raising his buming arms example, the two hunters in the lower section as torches from Kizil Cave No. 17. of the second panel on the left, and the dancer in the lower section of the sixth central panel, wear robes with lapels.11 The man with a knife in front of a buffalo drawn cart in the fïrst panel on the left, and the hunter in the second panel on the left, the one who holds a jar in the first central panel and one on the left side of the meeting scene in the central fourth panel are dressed in robes with round collars.12 The faces of the merchants in the mural are barely visible but the white high-crowned, beret-like hats they wear can be seen clearly. These hats look exactly like the one the sabao wears in the An Jia’s

screen.13

° Beijing daxue kaogu xi, et al. (ed.), 1989, Fig. 121, p. 244 (description). Beijing daxue kaogu xi, et al. (ed.), 1989, Fig. 62, p. 241 (description). For more details, see Ma Shichang, 1996, p. 184, fig. 41. 11 Shaanxi sheng kaogu yanjiusuo (ed.), 2003, p. 68, figs. 30, 32 and 65. Ibid., pp. 68-69, figs.31, 42 and 56. For the confirmation of the identity of sabao, see Rong Xinjiang, 2003d,pp. 32-35 + figs.1-7. 13 Shaanxi sheng kaogu yanjiusuo (ed.), 2003, figs. 49, 56 and 73.

Les Sogdiens en Chine

212

Fig. 2: Merchant and Bodhisattva with burning arms from Kizil Cave No. 114.

Legend of

In the central chamber of Kizil Cave No. 8, at the Southern end of the west slope of the ceiling, in the lower register of the lower section, there is a rhomboid niche whose wall painting depicts a sabo leading merchants out of the darkness by raising his burning arms.14 In this mural, the Sabo is dresses like a Bodhisattva and the merchants wear colored robes with lapels, and a white pointed hat as he pulls a donkey. The cave dates from the seventh Century. On a painting in the third tier on the left slope of the ceiling in the main chamber of Kizil Cave No. 178 (this painting is kept in the Museum of Indian Art in Berlin and numbered MIK III 8449a), merchants, dressed in fbreign-style robes with white hats and high boots, are following a sabo. A pack camel follows them.15 The cave also dates from the seventh Century A.D. The 15th painting on the lateral wall of the main chamber of Kizil Cave No. 184, (Now kept in the Museum of Indian Art and numbered MIK III 8888) shows a Bodhisattva raising his burning arms. On his left is a merchant who pulls a donkey. The donkey carries goods. On the left, near the edge of the picture is a cliff. The background is dark.”16 The cave dates from the seventh Century A.D. There are also similar murais in Kizil Caves Nos. 58, 63, 11417, and Taitai’er Cave No. 13.18 The merchant in Kizil Cave No. 114 wears a pointed white hat and skin-tight robe with lapels (fîg. 2).

Dragon King Mabi

In Chap. 4 on Cakravartin in the Jinglü yixiang vol. 24, it is written that: Once, on the continent of Jambudvîpa, there were five hundred merchants who went to sea to look for profitable commercial destinations. One merchant leader named Manyue having obtained what he sought, decided to leave. This displeased the Dragon King who decided to cause him trouble. There was another Dragon King called Ma Jian (Bi) MS(W> who was actually a great Bodhisattva and born in dragons because of his primordial vow. Feeling compassion for the merchant, he guided him to safety on the shore, and then retumed to where he had originally lived. 14

Beijing daxue kaogu xi, et al. (ed.), 1989, Fig. 33, p. 238 (description). H. Hartel, M. Yaldiz (ed.), 1982, pp. 100-101, pl. 34. 16 Beijing daxue kaogu xi, et al. (ed.), 1997b, fig. 207, p. 234 (description). 17 See Xinjiang Qiuci shiku yanjiusuo (ed.), 2000, pp. 69, 73, and 144. For the paintings in Cave No. 1 14, please see the Zhongguo bihua quanji, 1992, p. 134, fig. 145. 18 See Xu Wanyin, 1989, fig. 172, p. 228. 15

Rong Xinjiang

_

Sabao w Sabo

213

On the west slope of the ceiling in the main chamber in Kizil Cave No. 1 14, there is a painting depicting this story. Two merchants pull a camel and an ox respectively. They walk over a bridge supported by two dragon heads”. The merchants hâve neither deep-set eyes, nor high-bridged noses, and their faces resemble the characteristic plump faces of Kuchean donors. They are dressed in tight robes. The cave can roughly be dated to the fourth century AD. The same story is also displayed in the second panel in the third tier on the east slope of the ceiling in the main chamber of Cave No. 114 (The painting was peeled away by the German expédition, and is now kept in the Muséum of Indian Art in Berlin and numbered as MIK III 9103). The dragon is bulky and twisted into an S shape. On both ends the dragon opens its mouths wide. On the dragon a walking merchant pulls an ox and a donkey. The ox’s pack is composée! of three layers. It is bound with a rope through the middle. The pack looks exactly like the one on the camel on screen D of a Sogdian couch kept in the Miho muséum.19 20 The merchant wears a long belted robe, a white pointed hat and has inserted his trouser legs into his high boots.21 It has been said that the painting in the fourth rhomboid niche at the northem end on the east slope of the ceiling in Kizil Cave No. 14, also depicts the story above (see fïg. 2 on plate 18). Here, two men wearing coats with lapels, stand on a big dragon with two heads one at each end of his body. The man at the front end wears a pointed hat, a white coat, a waist belt and a pair of grey tight-fiiting trousers. The coat is trimmed with an orange border. The man has folded his hands on his chest, and looks back on the other individual. The latter has untied his hair, and wears a black coat and a pair of white tight-fitting trousers. He has also folded his hands on his chest thus echoing the posture of his companion. The dragon has a green body with white spots and white abdomen. On both dragon heads there are homs. The dragons’ eyes are wide-open. It twists around a mountain and bites into its own body. The head is painted khaki and outlined with ink lines. The mountain is blue.22 The man with the pointed hat seems to possess curly hair and deep-set eyes. The other man has a plump face that resembles that of Kuchean donors. The cave is dates from the sixth century A.D. The last mural in the second tier, on the west slope of the ceiling in the main chamber in Kizil Cave No. 17, also has the same theme. The painting is incomplète. Only the bridge supported by dragon heads and the legs of horses remain.23 The cave is dates from the sixth century A.D.

19

See Beijing daxue kaogu xi, et al. (ed.), 1996, the left lower part of fïg. 134. Also see p. 181, fig. 24. For more details, see Zhongguo bihua quanji, 1992, p. 139, fig. 150.

20

Miho Museum. Southwing, 1997, p. 251. Beijing daxue kaogu xi, et al. (ed.), 1997b, fig. 197. The description does not bear enough details on p. 231. 22 See Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 47, fig. 48 (for a close-up), p. 240 (description). Also see Beijing daxue kaogu xi, et al. (ed.), 1997a, Vol. 1, pp. 81-82, figs. 18 and 22. 23 See Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 66, p. 241 (description). Also see Ma Shichang, 1996, p. 181, fig. 24. 21

Les Sogdiens en Chine

214

Similar murais can also be found in Caves Nos. 7 and 13.24 In Cave No. 13 the merchant is dressed in a tight long robe and high boots (see fig. 3 on plate 18). He has a waist belt and sports a short dirk. On the lower sections of the left and right walls of Kizil-Qargha Cave No. 21, there are the same murais of Dragon King Mabi. Two merchants on the Dragon King are apparently dressed in Sogdian style, with typical Sogdian caps.25 In comparison with the pictorial material, the story recorded in the Jinglü yixiang are rather brief. Therefore, the textual original of the paintings is usually the longer, more complicated version. On the murais, in sum, Dragon king Mabi bridges over the océan with his own body to save the merchants. Legend of

sabo Shengyou providing milk

The foliowing story also appears in Chap. 34 of the Xianyu jing, vol. 8: The Buddha told Änanda: “Long ago, the Pratyeka Buddha lived in Väränasi on Jambudvipa. Once, when the Pratyeka Buddha feit ill, he went to consult Bhaisajyaguru who told him: ‘You are allergic to wind and should drink milk.’ At that time, there was a in Chinese, who lived in the same or Shengyou sabo named Aryamitra country. Pratyeka. Buddha went to see him, told him about his problem and asked him for milk. The sabo was delighted to help and provided the Buddha with a three months supply of milk. When the three months were up, the Pratyeka Buddha feit his that his health had greatly improved and decided to greatly benefit his host. He therefore undertook to transform himself eighteen times s urging into the air he proceeded to sit, lie, walk and stand there. When enlarged he suffused everything, but when contracted, he was no bigger than a tiny hair. The spectators were delighted by these feats. A few hours alter landing again on the ground, the people came to serve him but he entered into parinirvana. The sabo greatly grieved for him.”26

—

Fig. 3: Shengyou providing milk to Pratyeka Buddha, from Kizil Cave No. 14.

24

The story is depicted in the mural of the second rhomboid niche on the left of the upper tier on the west slope of the ceiling in Kizil Cave No. 14 (fig. 3) It depicts a kneeling Bodhisattva wearing a blue skirt of a foreign design, held by a knot on each side of his waist. His scarf floats behind his shoulders and hangs down in a loop on his ehest. Befbre the Bodhisattva lies a sick man in a blue cassock; his legs are crossed; his left hand rests on his knee, and his right hand is stretched as though waiting to receive something.

See Xinjiang Qiuci shiku yanjiusuo (ed.), 2000, pp. 13 and 20. For the pictures of Cave 13, see Zhongguo bihua quanji, Vol. 8, p. 156, fig. 168. 25 According to the author’s survey on Aug. 2, 2004. 26 Taishö shinshü daizökyö, Vol. 4, No. 202, p. 404.

Rong Xinjiang

Sabao or Sabo

215

The Bodhisattva leans slightly forward to feed the man. In his right hand he holds a wooden ladle and in his left hand a bowl.”27 In Chinese Caves: Kizil Caves the mural is captioned as a painting of “The legend of Shengyou providing milk to Pratyeka Buddha.”28 The cave dates from the sixth Century A.D. Similar murais can also be found on the west slope of the ceiling in Kizil Cave No. 178, which the German expédition dated to the seventh Century,29 and on the ceiling 30 of the main chamber of Simsim Caves No. 26. It is clear that these murais depict the story of sabo supplying the Pratyeka Buddha with milk. In these paintings, only the sabo and Pratyeka Buddha appear. The sabo is dressed like a Bodhisattva and he does not lead any merchants Legend of Mahätyägavan The story is told in Chap. 35 on the Mahätyägavan in the Xianyu jing vol. 8. Since it is too long to quote entirely I shall paraphrase it here. or Great Benefactor in Mahätyägavan, son of Nyagrodha, that is Dashi Chinese, under the guise of a sabo organised an expédition of five hundred merchants to look for treasures at to look for profitable destinations. The expédition was successful. Dashi took three pearls away from the dragon’s palace to be used to relieve the plight of the poor. While he was resting from his labours, some dragons seized the opportunity to steal the pearls. As soon as Dashi became aware of his loss, he drained the océan with a turtle shell. On his first attempt, he caused the waters to recede forty miles, on his second attempt the waters receded eighty miles, and on his third attempt they receded one hundred and twenty miles. The dragons were frightened and retumed the pearls tohim.31 The mural in the first tier of the west slope of the ceiling in the main chamber in Kizil Cave No.38, depicts this theme (see fig. 4 on plate 18). There are two figures in the mural. One is shown standing in the océan to drain the waters and the other holds a plate with both hand on which the pearls are displayed.”32 The sabo Dashi is dressed like a Bodhisattva. As he stands in the océan, the water does not reach his knees. In the other scene, when the dragon king is retuming the pearls to him, the water has reaches his waist. The cave dates from the fourth Century A.D. In the ninth rhomboid niche on the left of the lower tier of the west slope of the ceiling in Kizil Cave No. 14, there is a painting with the same theme. A Bodhisattva ties his blue hair above his head. He wears a pair of dark red trousers knotted on one side of 27

See Beijing daxue kaogu xi, et al. (ed,), 1997a, Vol. 1, p. 85, fig. 20. In this book the painting is considered to be about a story of feeding a sick monk with flesh. The author quotes the Chap. 21 in Xianyu jing vol. 4 to Support the conclusion. However, the sick monk does not lie in a grass cavem, and the person who feeds him is dressed as Bodhisattva instead of layman. It is diffïcult to décidé. 28 Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 50, p. 240 (description). 29 See Ma Shichang, 1996, p. 193, fig. 78. 30 See Ding Mingyi, 1993, p. 366. 31 Taishö shinshü daizökyö, Vol. 4, No. 202 ’ pp. 404-409. Cf. Chavannes, 1934, pp. 90-91. The story also appears in the Tocharian Manuscripts. 32 Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 129, p. 245 (description).

216

Les Sogdiens en Chine

his waist. He stands in a green océan. A blue scarf that hangs across his ehest is Imotted behind his back. The Bodhisattva stoops low to scoop up water with a turtle shell,”33 The cave dates from the sixth Century A.D. In the third mural, of the second tier of the west slope of the ceiling in the main chamber of Kizil Cave No. 17, we see Dashi scooping up water with a bowl. The person next to him, on bended his knees dedicates something to him.34 The person in question is certainly the dragon king seen retuming the pearls. The cave dates from the sixth Century A.D. A painting that once graced the second mural on the left of the middle pillar of Kizil Cave No. 1§6, is now in the Museum of Indian Art in Berlin and numbered as MIK III 8851. It shows the same story as the aforementioned examples. Daslii, dressed as a Bodhisattva is scooping up water with a bowl. On his both sides, there is a dragon king who is offering him the pearls. In the upper section there are two heavenly spirits who are scooping up water with a scarf5. The cave dates from seventh Century A.D. The same theme is also visible in the murais of Kizil Caves Nos. 91, 104, 178 and 206,36 and the right corridor ofNo. 14, and the corridor in Cave No. 21 in Kizil-Qargha Caves.37 It is also depicted on the ceiling of the main chamber in Kumtura Cave No. 2.38 In ail the painting of this story, sabo is always depicted as a Bodhisattva. To scoop up water he uses a turtle shell or a bowl, and the dragon king is seen offering him the pearls. The text and the illustrations are fairly well matched. Legend of Lenasheye

The story is recorded in Chap. 43 of the Xianyu jing vol.8. It reads: The Buddha told some monks: "Long ago, on the [continent of] Jambudvîpa, in this state of Benares, there lived a king called Brahmadatta. Also living in his kingdom was a great sabo whose name was Lenasheye ... Just at this time, a group of merchants came with a pétition to the sârthavâha because they wished to go with him on a sea voyage. He immediately replied saying: 'By my standards, one should begin (such a journey) with the construction and outfitting of a ship. [However] , I am today too poor to undertake such a task and hâve no possessions. How then could I agréé to this venture?” The petitioners replied saying: ‘There are five hundred men in our group. We ourselves will pay for the construction and outfitting of the ship.’ When he heard this statement, the sârthavâha agreed to go with them. No sooner had had they finished speaking that the men set to work to eut the necessary ropes. They continuée! working for seven days until ail the ropes were finished. The ship moved swiftly away. Half way through the voyage, the men encountered 33

See Beijing daxue kaogu xi, et al. (ed.), 1997a, vol. 1, p. 86, color-figs. 20 and 25; also see Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 50. p. 240 (description). 34 See Beijing daxue kaogu xi, et al. (ed.), 1989, fig. 66, p. 241 (description). Also see Beijing daxue kaogu xi, et al. (ed,), 1996, p. 181, fig. 21. 35 H. Hartel, M. Yaldiz (ed.), p. 104, pl. 36; Beijing daxue kaogu xi, et al. (ed.), 1997b, fig. 210, p. 234 (description). 36 See Xinjiang Qiuci shiku yanjiusuo (ed.), 2000, pp. 117, 132, 200 and 233. 37 See Ding Mingyi, 1993, pp. 368, 369, 371. 38 See Beijing daxue kaogu xi, et al. (ed.), 1992, fig. 6, p. 243 (description).

Rong Xinjiang

_

Sabao or Sabo

217

a fierce gale that smashed their ship to pièces. All the men cried out to be rescued but there was no place where they could find refuge. Some were able to find a board, a mast, or a floating leather bag with which to keep afloat. But others feil into the water and drowned. Five men then tumed to the särthaväha saying: ‘You are responsible for our p light. Now that this péril is upon us, we pray that we may be rescued.’ He replied: '1 have heard that the great sea will not abide the presence of a corpse. Each of you must hold on to me and for your sakes, I shall end the life of my body in Order to rescue you. I have vowed to become a Buddha. Later, when I have become a Buddha, I shall use the ship of the Unsurpassed Right Dharma to take you across the suffering of the great sea of birth and death.’ After these words the särthaväha took out a knife and inflicted cuts on himself. When his life had ended, the sea spirits generated gusts of wind that blew the ship to safety on the shore. Thus, the merchants succeeded in Crossing tlie great sea, and they all obtained peace and security.” The Buddha told the monks: “Should you wish to know who at that time was Lenasheye, it was me.»39

The story appears in the upper rliomboid niche on the left of the west slope of the ceiling in the main chamber in Kizil Cave No. 114 (see fig. 1 on plate 19): “A man is seen floating in the océan of round shape. Another four people, also in the water, are attempting to support him. In the upper section, another man is visible grasping a board.”*40 The sabo is dressed like a Bodhisattva. The cave dates from fburth Century. Similar murais can also be found in Kizil Cave No. 13,41 in the right passageway of cave No. 14 and in the passageway of Cave No. 21 of the Kizil-Qargha Caves.42 In the wall paintings, apart from the sabo, there are four people, while in the scripture there are five. This does not amount to a fundamental différence, but is probably due to the limits imposed by the rhomboid shape of the niche. There is no discrepancy between the basic content of the paintings and the scripture. And both the sabo and the fbur merchants have plump faces like Kuchean donors. As incarnations of the Buddha, most of the sabo in Kuchean murais are dressed like a Bodhisattvas. This is the case of the sabo in the murais of the legend of the Pradîpapradyota Kingdom, of the sabo Shengyou providing milk for the Buddha, of the legend of Mahätyägavan, and of the legend of Lenasheye. It is the Kuchean interprétation of traditional Indian images. Zhu Yingrong and Han Xiang ^^0, once believed that the merchants wearing Kuchean style garments, reproduced the appearance of real Kuchean merchants of that time.43 However, people’s nationality cannot be decided on their attire alone. The Sasanian-Persian style had exerted a strong influence on the dress of Kuchean people. For example, the portraits of the Kizil donors kept in the Museum of Indian Art in Berlin bear salient characteristics of that in murais in Sogdiana. Robes with lapels,

w See

Taishö shinshü daizökyö, Vol. 4, pp. 421-422. The story has been partly translatée! into French by É. Chavannes in his Cinq cent contes e/ apologues extraits du Tripitaka chinois (Chavannes, 1934, p. 129). Here the full English translation is based on Lenz, 2003, pp. 214-216. 40 See Beijing daxue kaogu xi, et al. (ed,), 1996, fig. 145, p. 257 (description). Also see pp. 190-1, fig. 71. 41 See Xinjiang Qiuci shiku yanjiusuo (ed.), 2000, p. 20. 42 See Ding Mingyi, 1993, p. 368-9 and 372. 43 See Han Xiang, Zhu Yingrong, 1990, pp. 285-6; also see Zhu Yingrong, 1993, pp. 20-1.

218

__

Les Sogdiens en Chine

brims and the Persian interlocked-pearl pattern.44 Li Mingwei recognised the existence of Kuchean merchants in the paintings, but he also linked these images to commercial activities on the Silk Road and he emphasizes that these were depictions of merchants from the Western Region, including Persian merchants, instead of Indians.45 Huo Xuchu pointed out that it was not until the mature period of the paintings, that is from the mid fourth to the end of the fifth Century, that merchants became part of the subjects that appeared in paintings in Kucha.46 This was due to the influence of merchants on the Silk Road. Zhang Qingjie once discussed the caravans on the Silk Road in the light of the Kizil Caves murais. Though he did not go into the question of nationality, he brought together large amounts of Iranian data including pottery figu¬ rines with traders and camels, wall paintings and sculptures.47 The images of merchants that appear in this collection of artefacts, share basic characteristics with Sogdian mer¬ chants, as people from the Northern Dynasties to the Sui and Tang described them. Of the merchants we hâve described above, we can conclude that most of them look like the Sogdians on An Jia’s screen, except for those who resemble Kuchean donors with plump faces. On this basis we can déterminé four basic categories. First, most of the merchants hâve deep-set eyes and high-bridged noses, which distinguish them from Buddhist figures and Kuchean donors with plump faces. Tn comparison with Indians and Kuchean merchants, Sogdians on An Jia’s or Miho's screen, match what we know about Sogdian clothing and appearance, thus providing a due to the nationality of the merchants represented in the paintings. Second, although clothing cannot be a décisive factors in determining a person’s nationality, Contemporary attire is usually characteristics of a given nation. Merchants in Kuchean murais frequently wear a high-crowned, beret-like hat, which has now become one of the basic criteria for identifying a Sogdian sabao as we saw on the An Jia’s screen. Therefore, it is reasonable to assume that hats offer a due to the nationality of a Sogdian merchant.48 Third, most of the merchants are dressed in tight robes with round collars or lapels, which was also a common type of Sogdian garment. Typical examples can be found on the Anyang and An Jia’s screens. Lastly, I once argued that because India is near the sea, the commercial activities of the groups led by Indian sabo frequently take place at sea. However, when it cornes to the Xianyu Jing, which was preached in Khotan and composed in Gaochang both in Central Asia, the location had to change. Groups of merchant in this book began to travel in valleys and in inhospitable surroundings. Geographical features must reflect the period and location of locations where the action takes place.49 Though in some Kuche¬ an murais, the océan appears as a background of some sabo stories such as those with the Dragon king Mabi, Mahâtyâgavan or Lenasheye, dark valleys and inhospitable sur¬ roundings which are doser to the landscape of the Western Region, are also depicted. 44

See H. Hartel, M. Yaldiz (ed.), 1982, p. 168, pl. 107. See Li Mingwei, 1995, pp. 425-7. 46 See Huo Xuchu, 1994, pp. 44-5. 47 See Zhang Qingjie, 2004, pp. 173-203. 48 For detailed discussion on the hat, see E. Kageyama's paper in this volume. 49 See Rong Xinjiang, 2003c, pp. 128-43. 45

Rong Xinjiang

_

Sabao or Sabo

219

Furthermore, whether in the legend of the Pradïpapradyota Kingdom which is set in a valley, or the legend of Dragon king Mabi that takes place at sea, horses, donkeys, oxen and camels, are the most common vehicles that Sogdian sabao and the merchants who hired them, use to cross deserts such as the Gobi. Thus camels, also known as desert ships, replace the real ships employed by native Indian sabo. In light of all the above data, I believe that my contention that the images of merchants led by Indian sabo were transformée! into images of merchants led by Sogdians sabao, is justified. Like all the other places in the Western Region, Kucha was a large urban Settlement that eastbound Sogdian immigrants could not avoid Crossing. 50 Although the local painters wanted to picture Indian merchants, in all likelihood, they consciously or unconsciously involved Sogdian subjects in their créative work. As a resuit, Indian sabo became Sogdian sabao.

Sabo and his Merchants in Murals in Dunhuang Caves

There are also many images of caravan leader and merchants in the Dunhuang caves. However, subjects such as Prince Sattva feeding himself to a hungry tigress, and the nine-colored deer saving a drowning man are the most numerous. To date, in Dunhuang murais, no stories featuring sabo as in the Kucha caves, have been found. Most of the Dunhuang merchants appear in the context of the illustrations of the Miaofa (Saddharmapundarika Sutra). lianhua jing The mural on the eastem section of the north slope of the ceiling of Mogao Caves No. 296, is said to illustrate the Futian jingbian ^03^5^, that is the Sutra of Seven Good Deeds)?' One of the seven ways of cultivating good deeds is to build bridges for the weak. In the corresponding wall-painting there is a tiny bridge crossed from the left to right by two pack horses, four merchants with typical Chinese features, fbllowed by two men on horseback and two on foot. At the other end of the bridge, there is a foreign merchant with deep-set eye and a high bridged-nose, wearing a pointed hat, pulling two camels. The camels carry big packs and are about to cross the bridge. Behind them there is another foreign merchant pulling two pack donkeys. In the section above this wall-painting there are a number of scenes, depicting a garden where merchants are resting, as well as merchants sitting on the ground with camels lying besides them. One person seems to be feeding the horses and mules, while two others are trying to give a camel some medicine.52 Sabo and his merchants are not specifically mentioned in the text, but Dunhuang painters of the Northern Zhou dynasty, delineated foreign merchants with Central Asian features instead of Indian ones, as a counterparts to the Chinese merchants across the bridge. It is not possible to clearly identify the somewhat sketchy figures of foreigners, yet it is obvious that they are non-Indians. Images of merchants on murais in the Mogao Caves are mainly illustrations of 50

See Rong Xinjiang, 1994, pp. 164-5. Also see Rong Xinjiang, 2001, pp. 32-4. See Shi Weixiang, 1980. 52 Dunhuang wenwu yanjiusuo (ed.), 1982, fig. 189. For clearer figures: Ma De (ed.), 2001, pp. 18-9, fig. 2.

51

220

_

Les Sogdiens en Chine

Chapter 7 and Chapter 25 of the Fahua jingbian (illustrations of Miaofa lianhua jing). The main story in the Avalokitesvara chapter (Chap. 25) is also recorded in the central colophon on the mural of Cave No.45 which reads: There was once a state in which many robbers were active. A caravan leader (sabo) leading merchants loaded with treasures, was forced to pass through a perilous route in the state. One of the men said: ‘Good men, do not be afraid. We will all call the name of Bodhisattva and he who will save us. If we all call upon him we will be able to drive off the robbers.’ So the merchants began to call: ‘Namo Avalokitesvara Bodhisattva!’ and because they called his name they finally obtained freedom and salvation.53

In Chapter 7, there is a story of merchants who went seeking for treasures and encountered dangerous paths and inhospitable surroundings. The merchants were completely wom out and intended to retreat. At that moment, a leamed leader of great wisdom brought them to a city with his magical powers. Later, the merchants were reluctant to leave the city, because they enjoyed their life of ease. The sabo then made the city disappear and urged the merchants to reach their intended destination. In the upper section of the east slope of the ceiling of Mogao Caves No. 420 there is a painting of Avalokitesvara from the Sui Dynasty (fig. 2, plate 19)." He Shizhe suggested that the panels should be viewed as a sequence running from right to left. The sequence begins before the departure of the caravan, with the leader praying to Avalokitesvara/ Guanyin for good fortune. Then we see the caravan tramping over mountains, with a team of pack camels and donkeys. When they corne down into a valley for a break, a band of robbers attacks them. The merchants lose the fight and ail were captured. At this moment, Avalokitesvara shows up and uses magical power to subjugate the robbers. The bandits are then shown lined up as two teams, with their hand folded in front of their chests.55 The mural covers grand scenes, but the figures themselves are small. Given their size and the dark colors of the painting, it is very difficult to détermine the nationality of the participants. However, the lines of camels still remind of Central Asian Caravans traveling in the desert of the Western Region. The illustrations from the Fahua jingbian demonstrate the transition from Early Tang to High Tang on the south wall of Cave No. 217 (fig. 4) by developing a new style in the flanking illustrations similar to the ones of the Pure Land in the Western World.56 In the mural that depicts a story from Chapter 7 of the of the Fahua jingbian, merchants are either walking slowly to rest their horses, or riding them towards a city produced by the magical power of their leader, or wandering along the banks of a river or in the multi-layers mountains.57 Illustrations of this theme, in the same style, painted during the Tianbao er a, (742755) also appear on the western side of the Southern wall in Cave No. 103 (fig. 5). In 53

See Yang Xiong (ed.), 1993, p. 81, fig. 83. See Dunhuang wenwu yanjiusuo (ed.), 1984, fig. 75. For clearer photographs see Ma De (ed.), 2001, pp. 26-7, fig. 8. 55 See He Shizhe (ed.), 2000, p. 33, fig. 18. 56 See He Shizhe (ed.), 2000, pp. 148-53. 57 See He Shizhe (ed.), 2000, pp. 64-5, fig. 49-50. 54

Rong Xinjiang

Sabao or Sabo

Fig. 4: Illustration from the Fahua jingbian on the Southern wall of cave No. 217.

Fig. 5: Illustration from the Fahua jingbian on the Southern wall of cave No. 103

221

222

Les Sogdiens en Chine

this section of Chapter 7, one merchant pulls an éléphant that carries a large pack, and behind them, stands a horse with no hamess. Two merchants follow the horse. In front of the caravan, a monk prays tumed towards a cliff. Another person prostrates himself on the ground to pray to Guanyin.58 The reason why the murais in Caves No. 217 and No. 103 share the same style, is according to He Shizhe’s analysis,59 because both were influenced by the style of painting then prévalent in the Chinese heartland. There is a major différence between the Central Asian Caravans depicted in the Kizil cave murais, and those of the Mogao caves of the Sui dynasty. For one thing, the merchants look different. It is very likely that Sui dynasty paintings are based on Chinese Originals from Central China. Each caravan is led by a monk, who replaces the {sabo) of the Fahua jing. The people pulling horses are slim figures and not dressed in the usual Sogdian tight robes. Women on horsebacks wear mili a fashion accessory of the early Tang.60 The appearance of this no longer fashionable item in High Tang art, indicates that the Dunhuang painters based their work on sketches rather than on real foreign caravans of the period. Another feature shared by the two murais is that both depict éléphants, which were certainly uncommon in Central Asian caravans. The appearance of these rare animais stresses the Indian flavor of the paintings. However, by contrast, the cities look very much like urban settlements in the Western Region. While éléments of Indian, Western Region and Chinese heartland culture ail feature in the paintings, they do not represent actual caravans from Chang’an or Dunhuang. The syncretic nature of the murais indicate the artists’ struggle to make real event coincide with literary descriptions. Attention must be focused on the illustration of Fahua jingbian on the northem side of the eastem wall gâte in Cave No. 217. A story of merchants’ encountering with bandits in a rough drawing is depicted on the central part. The merchants, who pull camels, are dressed in long robes with round collars, having tall caps and leather boots. Four armored bandits stand before them, with spears and shields in hands.61 These fig¬ ures of merchants on the Silk Road are different from the ones on the Southern wall, and they are typical Northem dynasties period. The tradition of the painting was not totally abandoned because of the introduction of the new style from China. It also appears as a transition factor to the images of the merchants displayed in Mogao Cave n° 45. The most famous painting among the illustrations of the Fahua jing is the one in which the merchants are confronted by bandits in Mogao Caves No. 45 (see fig. 3 on plate 19). It illustrâtes a scene from the Avalokitesvara Sutra during which a band of robbers block the merchants’ road, whereupon the merchants unload some of their goods and beg for mercy. Ail the foreign merchants hâve deep-set eyes, high-bridged noses and thick curly hairs. The caravan leader standing in front of them wears high felt-hat and a green robe. Some of the merchants following him wear hats, others do not, and some hâve wrapped long scarves around their heads. Most of the men are 58

See Dunhuang wenwu yanjiusuo (ed.), 1987, fig. 153. For clearer photographs see He Shizhe (ed.), 2000, p. 72, fig. 61. 59 See He Shizhe (ed.), 2000, p. 153. 60 SeeDuan Wenjie, 1986, pp. 231, 233, fig. 10. Also see Rong Xinjiang, 2003b, pp. 724-5. 61 According to the author’s survey in the Mogao Grottoes in July, 2004.

Rong Xinjiang

_

Sabao or Sabo

223

dressed in robes with round collars. The merchant bringing up the rear wears a red robe and pulls two pack donkeys. He appear very nervous and fold his hands as if praying to Guanyin for help. 62 Such a high felt-hat frequently appears in murais featuring foreigners. The wall-painting in Lou Rui’s tomb from the Northern Qi in Taiyuan, is an example of this.63 The robes with round collars are not the typical attire of foreigners. But the characteristics of deep-set eyes, high-bridged noses and thick curly hair prompt us to think again of Sogdian rather than the Indian merchants of Buddhist scriptures. The merchant standing at the front, looks older than the others and is likely to be their caravan leader. To sum up, the images of merchants painted by Dunhuang artists have undergone a transition from an Indian sabo to a Sogdian sabao.

The Transitions from Texts to Images

After our close scrutiny of the murais, let us now tum back to the question of sabao and sabo. In light of the data gleaned from the murais in Buddhist caves, which we presented above, we are able to conclude that sabo basically appear as Bodhisattvas in the stories. Sometimes, in illustrations of sutras, they appear as monks. Only in Mogao Cave No. 45 from the eighth Century A.D, is the leading merchant (if he is really meant to be identified as the sabo) depicted as a typical Sogdian sabao. The murais in Buddhist caves are either based on sketches or texts. Among the Dunhuang manuscripts found in the grottos, scriptures from every period occupy the lion’s share of the collection, while the number of sketches is relatively small and usually date from a later period. Therefore, early Buddhist murais are more likely to be based on texts. When a new design for a mural was created in a grotto, it could become a model to be copied repeatedly in neighboring caves of the same temple. The painters in other temples could also copy it when working on the murais in other places. Painters had to produce exact replicas of traditional models of portraits of Buddhas, Bodhisattvas, and devas, and they were not allowed to deviate from traditional Buddhist stories or historical events. However, they were more free in depicting details of stories around the margins of a picture. I believe that artists took advantage of this limited freedom, in order to introduce Sogdian faces among the merchants they were painting in the Kucha Caves’ murais because that is what they saw when watching actual Caravans. Chance plays no part in the appearance of Sogdians in Kizil murais as early as in the fourth Century. The Sogdian Ancient Letters which date from the beginning of the fourth Century, show that the Sogdians were active in an area roughly extending from the Tarim basin to the heartland China.64 Their trading activities encompassed trade between Sogdiana and China, trade with China and India, as well as trade with northem nomadic tribes and the Chinese empire. Sogdians ahnost dominate the trade on the Silk 62

See Dunhuang wenwu yanjiusuo (ed.), 1987, fig. 133. For clearer photographs see Ma De (ed.), 2001, pp. 28-9, fig. 9 and Yang Xiong (ed.), 1993, p. 81, fig. 82. 63 See Zhang Qingjie, 2004, pp. 188-9, fig. 29. He considérée! the person to resemble a Persian. 64 For the dating of the Ancient Letters, see Grenet, Sims-Williams, 1987. For the translations of the letters, see Sims-Williams, 2001; Grenet, Sims-Williams, de la Vaissière, 1998, and Sims-Williams in this volume.

224

_

Les Sogdiens en Chine

Road. As a resuit, Indian merchants are seldom seen on the road and even Persian merchants abandon land routes to concentrate on trade at sea.65 Consequently, the merchants on the Silk Road from the fourth to eighth Century in the Kucha or Dunhuang murais look like Sogdians rather than the Indians described in Buddhist text. In addition, while the sabo in Buddhist mural gradually lose their original appearance sabo in the Chinese texts of the Buddhist canons, are more and more frequently referred to as shangzhu (caravan leaders), daoshi (guides) or related tenus . In the most populär Miaofa lianhua jing translated by Kumârajïva there is no trace of the word sa¬ bo. For readers of Chinese translation of Buddhist scriptures, the original meaning of sabo, as the person who oversees the fitting out of a ship in order to lead merchants to profitable destinations, may hâve already escaped them. Consequently, they are likely to identify sabo with the Sogdian leader sabao. A case of a sabo in the shape of a sabao can be seen in Mogao Cave No. 45 the natural outcome of this confusion. Furthermore, the transitions from text to image generates a new misapprehension, namely that the sabo in Indian Buddhist stories is the same as sabao in real life. It is more than likely that people had no idea of what was meant by Sabo, and considered that the caravan leaders in texts or paintings were sabao. In several Uighur manuscripts of Chapter 25 of the Saddharmapundarika Sûtra of the Avalokitesvara Sütra, translated in the ninth Century and now in the Berlin Turf an collection, which purports to be based on the Chinese translation by Kumârajïva,66 shangzhu “a caravan leader” is translated as sartpau.61 Kuwabara Jitsuzô was the first to suggest that there should be a Sogdian word to correspondence to the Uighur one.68 Albert Dien also pointed out the more accurate Uighur spelling was s’rtp ’w.69 However, Haneda Akira believes that sabao was more likely to hâve developed from the Uighur sartpau, which itself may be traced back to a Sogdian origin.70 Finally Yoshida Yutaka Efi remarked in 1988, that the original Sogdian word s’rtp’w was discovered in Sogdian Ancient Letter V.71 Therefore, we are able to conclude by saying that sartpau or s’rtp ’w in the Uighur Saddharmapundarika Sütra is actually s’rtp’w which was borrowed from the Sogdian a long time ago, rather than from the Sanskrit särthaväha. The Uighurs used their own language to find a term to correspond to shangzhu (a caravan leader) in Chinese.72 Hence, the distinction between the Indian sabo and the Sogdian sabao has completely disappeared from late Uighur Buddhist canon. 65 For the Sogdians' domination of trade on the Silk Road, see Jiang Boqin, 1994, pp. 150-226. Also see Sims-Williams, 1996. For the fight for the Silk Road between Sogdians and Persians, see Rong Xinjiang, 2002, pp. 61-4. The paper has been translated into English and presented to “New Perspectives on the Tang. An International Conference”, Princeton University, April 18-20, 2002. 66 See Zhang Tieshan, 1991, p. 137. 67 See Radloff, 1911, p. 37. Haneda Susumu, 1957, pp. 144, 146. 68 Kuwabara Jitsuzô, 1926, p. 660; in his complété works, Kuwabara Jitsuzö, 1968, Vol. 2, pp. 292-3, 359-60, but the hypothesis of the Sogdian correspondence is removed in the volume. 69A.E. Dien, 1962. 70 Haneda Akira, 1971, pp. 426-7. 71 Yoshida Yutaka, 1988, pp. 168-71. 72 J. Elverskog thinks that the Uighur text is translated from a Sogdian original. If true, the Uighurs were the first to use s’rtp w to translate Chinese shangzhu. It could be possible, but we have not found any Sogdian fragments of Saddharmapundarika Sütra or Avalokitesvara Sütra. See Elverskog, 1997, p. 59.

Saba。or Sabo

Rong Xinjiang

225

Bibliography

Beijing daxue kaogu xi 北京大學考古系,et

1989 1992 1996

1997a

1997b

al (eds.) Zhongguo shiku: Kezi er shiku 中國 石窟 ® 克孜 爾 石窟 [Chinese Caves: Kizil Caves], Vol. 15 Beijing: Wenwu chubanshe. Zhongguo shiku: Kumutula shiku 中國 石窟 庫木吐喇石窟 [Chinese Caves: Kumtura Caves], Beijing: Wenwu chubanshe. Zhongguo shiku: Kezi er shiku 中國 石窟 克孜爾 石窟 [Chinese Caves: Kizil Caves], Vol. 2, Beijing: Wenwu chubanshe. Xinjiang Kezi'ey shiku kaogu baogao 新疆 克 孜 爾 石窟 考古 報告 [The Archaeological Report of Kizil Caves in Xinjiang], Beijing: Wenwu chubanshe. Zhongguo shiku: Kezi er shiku 中國 石窟 克孜 爾 石窟 [Chinese Caves: Kizil Caves], Vol. 3, Beijing: Wenwu chubanshe. 6

争

0

Chavannes, É. 1934 Cinq cent contes et apologues extraits du Tripitaka

Ernest Leroux. Dien, A. E. 1962

chinois. Vol. IV, Paris:

"The Sa-pao Problem Re-examined55, Journal of the American Oriental Society, 82.3, pp. 335-46. Ding Mingyi 丁明夷 1993 "Ji liangchu dianxing de Qiuci shiku: Senmusaimu yu Kezi'ergaha shiku" 記 兩處 典型的 龜茲石窟 一 森木塞姆與克孜 爾余哈 石窟,[Notes on two typical Caves in Kucha: Kizil-Qargha Caves and Sims im Caves] in Xinjiang Qiuci shiku yanjiusuo 新疆龜茲 石窟 研究所 (ed.)s Qiuci fojiao wenhua lunji 龜茲佛 教文 化論集 [Collection of Essays on Buddhism in Kucha], Urumqi: Xinjiang meishu sheying chubanshe, pp. 356-78. DuANWenjie 段文傑 1986 “Mogao ku Tangdai yishu zhong de fushi” 莫高窟 唐代藝術中的服飾,[Dress in Tang Art as Seen in Mogao Caves] in Xiang Da xiansheng jinian lunwenji 向 達 先生 紀念 論文 集 [Collection of Essays in Memory of Xiang Da], Urumqi: Xinjiang renmin chubanshe, pp. 220-275. Dunhuang wenwu yanjiusuo 敦煌文物研究所 (ed.) 1982 Zhongguo shiku: Dunhuang Mogao ku 中國石窟 °敦煌莫高窟 [Chinese Caves: Dunhuang Mogao Caves], Vol. 1, Beijing: Wenwu chubanshe. 1984 Zhongguo shiku: Dunhuang Mogao ku 中國石窟 。敦煌莫高窟 [Chinese Caves: Dunhuang Mogao Caves], Vol. 2, Beijing: Wenwu chubanshe. 1987 Zhongguo shiku: Dunhuang Mogao ku 中國石窟 。敦煌莫高窟 [Chinese Caves: Dunhuang Mogao Caves], Vol. 3, Beijing: Wenwu chubanshe. Elverskog, C. J. 1997 Uygur Buddhist Literature (Silk Road Studies I), Turnhout: Brepols.

Les Sogdiens en Chine

226

Grenet, H, Sims-Williams, N. "The Historical Context of the Sogdian Ancient Leiters”. Transition Periods in 1987 Iranian History (Studia Iranica, cahier 5),Leuven, pp. 101-22. Sims-Williams, N., de La Vaissière, É. Grenet, F., Sogdian Ancient Letter V”, Bulletin of the Asia Institute, XII, pp. 91-104. 1998 "The Yingrong 朱英榮 Han Xiang 韓翔,Zhu Qiuci shiku 龜幺幺石窟 [Kuchean Caves], Urumqi: Xinjiang daxue chubanshe. 1990 Haneda Akira 田明 Sogudo jin no Töhö katsudö ソ グ ド 人の東方 活動 [Sogdians* activities in the 1971 East], Iwanami köza sekai rekishi 岩波 講座 世界 歷史 [Iwanami world history], vol. 6, Tokyo: Iwanami shoten 岩波書店,pp. 409-434. Haneda Toni 田亨 Uiguru bun Hokke kyö fumon bon no danpen 回鷗文法華経普門品の断片 1957 [A fragment of the Uighur translation of the Avalokitesvara-b o dhisattvasamantamukha-parivarta of the S addharmapundarïka], in Haneda hakushi shigaku ronbunshü 羽田博士史学 論文 集 [Collection of potshumous works of Toru Haneda], Vol. 2: Gengo, shükyö hen 言語®宗教篇 [language, religion]”, Kyöto, pp. 143-147. Hartel, H., Yaldiz, M. (eds.) 1982 Along the Ancient Silk Routes: Central Asian Art from the West Berlin State Museums, New York: The Metropolitan Museum of Art. HE Shizhe 賀世哲 2000 "Dunhuang bihua zhong de Fahua jing bian” 敦煌 壁畫中的法華經變 [The Illustration of Fahua jing in Dunhuang Murais], in Dunhuang yanjiuyuan 敦煌 研究院 (ed.),Dunhuang yanjiu weryi: Dunhuang shiku jingbian pian 敦煌研 究 文集 藪 煌 石窟 經 變 篇 [Collection of Essays on Dunhuang Studies:

•

Volume of Illustration of Buddhist sutras in Dunhuang caves], Lanzhou: Gansu minzu chubanshe, pp. 148-53. He Shizhe 賀世哲 (ed.) 2000 "Fahua jing hua juan” 法華經畫卷 [Murais of Fahua jing] , Dunhuang shiku quanji 敦煌石窟 全集 [Complété Collection of Dunhuang Caves], Vol. 7, Shanghai: Shanghai renmin chubanshe. HUO Xuchu 霍旭初 1993 "Kezi'er shiku de fenqi wenti55 克孜 爾 石窟的分期問題 [The Periodization of Kizil Caves], Xiyu yanjiu 西域 研究,No. 2, pp. 58-70. Qiuci yishu yanjiu 龜茲 藝術 研究 [Studies on Kuchean Art], Urumqi: Xinjiang 1994 renmin chubanshe. 2002 Kaozheng yu bianxi: Xiyu fojiao wenhua lungao 考證與辨 析 - 西域 佛教 文 化論稿 [Explorations & Analysis: Essays on Buddhist Culture in the Western Region], Urumqi: Xinjiang meishu sheying chubanshe. Huo Xuchu 霍旭 初,Wang Jianlin 王嘘林 "Danqing banbo, qianqiu zhuangguan - Kezi'er shiku bihua yishu ji fenqi 1993 gaishu- 丹青 斑 駁 千秋 壯 觀 克 孜 爾 石窟 壁畫 藝術 及 分期 概述 [Variegated Brushes, Miracles in the Flowing Time: A General Introduction to the Art and Periodization of Murals in Kizil Caves], Qiuci fojiao wenhua lunji 龜茲 佛教文化論集 [Collection of Essays on Buddhism in Kucha], Urumqi: Xinjiang meishu sheying chubanshe, pp. 201-28. 一

-

Rong Xinjiang

Sabao or

227

JIANG Boqin 姜 伯勤 1994 Dunhuang Tulufan wenshu yu sichou zhi lu 敦煌吐 魯 番 文書 與 絲綢 之路 [Dunhuang-Turfan Documents and the Silk Road], Beijing: Wenwu chubanshe. KUWABARA JitSUZÖ 桑原鷲蔵 1926 Zui Tö jidai ni Shina ni raijü shita Saiiki jin ni tsuite 隋唐時代に 支那に 来住 し た西域人に 就いて [Türkistan people who settled in China during the Sui and Tang periods], in Haneda Töru 田 亨 (ed.), Naitö hakushi kanreki shukuga Shinagaku ronsö 內藤 博士還曆 祝賀支那学論叢 [Chinese studies dedicated to Dr. Naitö on the occasion of his sixtieth birthday], Kyoto, pp. 565-660. 1968 Kumbara Jitsuzö zenshü 桑原 鷺 蔵 全集 [Complété Works of Kuwabara Jitsuzô], 6 vols. Tokyo. von Le Coq, A., Waldschmidt, E. 1933 Die buddhistische Spätantike in Mittelasien. Berlin: Dietrich Reimer und Ernst Vohsen. Lenz, T. 2003 A New Version of the Gândhârî Dharmapada and a Collection of Previous-Birth Stories. British Library Kharosthi Fragments 16 十 25, Seattle and London: University of Washington Press. LÉvi, S. 1925 "Le sutra du sage e/ du fou dans la littérature de F Asie centrale5*, Journal Asiatique, pp. 304-32. Li Chongfeng 李崇 峰 2002 Zhong Yin fojiao shikusi hijiao yanjiu 中印 佛教 石窟 寺 比較 研究 [A Comparative Study of Chinese and Indian Cave Temples], Xinzhu: Juefeng fojiao yishu wenhua jijinhui. Li Mingwei 李明偉 1995 "Mogao ku he Kezi5er shiku bihua zhong de silu maoyi” 莫高窟和克孜 爾石 窟壁畫中的絲路貿易 [The Silk Road Trades in Mogao Grottoes and Kizil Cave Murals], in Duan Wenjie 段文傑 (ed.), (1990 nian) Dunhuangxue guoji yantaohui wenji: shiku kaogu bian ( 1990 年) 敦煌學國際研討 會文集：石 窟考古編 [Collection of Essays in the International Conference on Dunhuang Studies in 1990: Caves and Archaeology], Shenyang: Liaoning meishu chubanshe, pp. 414-430. Liang Liling 梁 麗玲 2002 Xianyu jing yanjiu 賢愚經研究 [Study ofXianyu Jing], Taipei: Fagu wenhua gongsi. Liao Y ang 廖日昜 2001 "Kezi'er shiku bihua fenqi yu niandai wenti yanjiu” 克孜 爾 石窟 壁畫分期與 年代問題研究 [Studies on the Chronology for the Kizil Mural Paintings], Yishushi yanjiu 藝術史研究 (刀/e Study of Art History). No. 3, 331-74. Ma Dazheng 馬大正 et al. (eds.) 1994 Xiyu kaocha yu yanjiu 西域考察與 研究 [Explorations and Studies of the Western Region], Urumqi: Xinjiang renmin chubanshe. Ma De 馬德 (ed.) 2001 "Jiaotong hua juan 交通畫卷" [Murais of Transportation], Dunhuang shiku quanji 敦煌石窟全集 [Complété Collection of Dunhuang Caves], Vol. 26, Shanghai: Shanghai renmin chubanshe.

228

-

…

Les Sogdiens en Chine

Ma Shichang 馬世 長 "Kezi'er zhongxin zhuku zhushi quanding yu houshi de bihua55 克 孜 爾 中心柱 1996 窟主室 券 頂 與 後室 的 壁畫 [The Ceiling in the Main Hall and Murals in Backrooms in Kizil Caves], in Beijing daxue kaogu xi 北京大學考古系,et al. (eds,), Zhongguo shiku: KezVer shiku 中國 石窟 。克孜 爾 石窟 [Chinese Caves: Kizil Caves], Vol. 2, Beijing: Wenwu chubanshe^ pp. 174-226. Mair, V. H. 1993 "The Linguistic and Textual Antécédents of The Sutra of the Wise and the Foolish"ゝ Sino-Platonic Papers, 38, April 1993, pp. 1-95. 1999 "The Khotanese antécédents of The Sutra of the Wise and the Foolish (Xianyu jing)", in E. Zürcher et al. (eds,), Collection of Essays 1993: Buddhism Across Western Régions, Sanchung (Taipei): Boundaries- Chinese Buddhism and 361-420. Foguang wenhua shiye gongsi, pp. Miho Museum

1997 Radloff, W. 1911

Southwing, Miho Museum,

türkische Übersetzung des XXXV. Kapitels der Kuan-si-im Pusar : chinesischen Ausgabe des Saddharmapundarika, St. Petersburg: Académie Impériale des sciences. RONG Xinjiang 榮 新江 1994 "Xiyu Sute yimin kao 西域 粟特移民考”[Sogdians Around the Ancient Tarim Basin], in Ma Dazheng 馬大正 et al. (eds.)? Xiyu kaocha yu yanjiu 西域考察 與研究 [Explorations and Studies of the Western Region], Urumqi: Xinjiang renmin chubanshe, pp. 157-172. Zhonggu Zhongguo yu wilai yvenming 中古 中國 與 外 來 文明 [Médiéval 2001 China & Foreign Civilizatious], Beijing: Sanlian shudian. 2002 “Bosi yu Zhongguo: Liangzhong wenhua zai Tangchao de jiaonmg” 波斯 與中 國：兩種文化在唐朝的交融 [Persia and China : Cultural Interaction under the Tang], Zhongguo xueshu 中國學術,No. 4, pp. 56-76. 2003a “Beichao Sui Tang huren juluo de zongjiao xinyang yu xianci de shehui gongneng" 北朝 隋唐 胡人 聚落 的 宗教 信仰 與 祠的 社 會 功能 [The Religions in Foreign Colonies and the Social functions of Zoroastrian temples in the Northern dynasties, Sui and Tang Period], in Rong Xinjiang (ed.), Tangdai zongjiao xinyang yu shehui 唐代 宗教 信仰 與 社 會 [Religions and Society in the Tang Dynasty], Shanghai: Shanghai cishu chubanshe, pp. 385-412, figs.1-9. 2003b ’Nü ban nanzhuang - Tangdai qianqi funü de xingbie yishi” 女扮男裝 —— 唐 代前期婦女的性別 意識 [Women dressed as men. Women gender awareness the first part of the Tang Period], in Deng Xiaonan 鄧小南 (ed.), Tang So?7g nüxing yu shehui 唐宋妥 性與 社 會 [Women and Society in Tang & Song Dynasties] , Shanghai: Shanghai cishu chubanshe, Vol. 2, pp. 723-50. 2003c "Sabao yu sabo: Beichao Sui Tang huren juluo shouling wenti de zhenglun yu bianxi" 薩保與萨薄：北朝隋唐胡人聚落首領問題的爭論與辨 析 [Sabao and Sabo: On the Problem of the Leader of Sogdian Colonies during the Northern Dynasties, Sui and Tang Period], in Ye Yiliang 葉奕 良 (ed.), Yilangxue zai Zhongguo lunwenji 伊 朗學 在中 國 論文 集 [Collection of Essays on Iranian Studies in China], Beijing: Beijing daxue chubanshe. Vol. 3, pp. 128-43. Illustrative Sequence on An Jia's Screen: A Depiction of the Daily Life "The 2003d of a Sabao'\ Orientations, Vol. 34, No. 2, pp. 32-5.

Rong Xinjiang

Sabao 〇ヂ Sabo

229

SHAANXI SHENG KAOGU YANJÏUSUO 陝西省考古 研究所 (ed.) Xi "an Bei Zhou An Qie mu 西安北周安 伽墓 [An Qie Tomb of Northern Zhou 2003 Found in Xi'an], Beijing: Wenwu chubanshe. Shi Weixiang 史葦湘 "Dunhuang Mogao ku zhong de Futian jingbian bihua” 敦煌莫高窟中的福田 1980 經 變 壁畫 [Murais of Futian jingbian in Mogao Caves], Wenwu, No. 9, pp. 44-8. SIMS-WILLIAMS, N. 1996 "The Sogdian Merchants in China and India”, in Cina e Iran da Alessandro Magno alla Dinastia Tang, ed. A. Cadonna and L. Lanciotti (ed.), Firenze: L S, Olschki, pp. 45-67。 "The Sogdian Ancient Letter IF', in M. G. Schmidt and W. Bisang (ed.), 2001 Philologica Linguistica: Historia, Pluralitas, Universitas. Festschrift für Helmut Humbach zum 80. Geburtstag am 4. Dezember 2001, Trier: Wissenschaftlicher Verlag, pp. 267-80. SU Bai 宿白 “Kezi'er bufen dongku jieduan huafen yu niandai deng wenti de chubu 1989 tansuo" 克孜 爾 部分洞窟階段劃分與 年代等問題的 初 步探索 [A Preliminary Exploration into the Dating and Periodization of Some Kizil Caves], in Beijing daxue kaogu xi5 et al. (eds.)5 Zhongguo shiku: Kezi 'er shiku 中國 石窟 克孜 爾 石窟 [Chinese Caves: Kizil Caves], Vol. 1, Beijing: Wenwu chubanshe, pp. 10-22. Zhongguo shikusi yanjiu 中國 石窟寺研究 [Study of Chinese Cave Temples], 1996 Beijing: Wenwu chubanshe. TAISHÖ SHINSHÜDAIZÖKYÖ 大正新修 大藏 經 1924-35 Taishö shinshü daizökyö 大正新修大藏 經,Tokyo. Xinjiang Qiuci shiku yanjiusuo 新疆龜茲 石窟 研究所 (ed.) 2000 KezVer shiku neirong zonglu 克 孜 爾 石窟內 容總錄 [Catalog of Contents of Kizil Caves], Urumqi: Xinjiang meishu sheying chubanshe. Xu Wanyin 許宛音 1989 "Taitai'er shiku tacha ji” 台台 爾 石窟踏 記 [The Exploration of Taitai'er Caves], Zhongguo shiku: Kezi "er shiku, Vol. 1, pp. 223-235. YANGXiong 楊雄 (ed.) 1993 Dunhuang shiku yishu: Mogaoku di siwu ku 敦煌石窟 藝術 。莫高窟第四五 窟 [Dunhuang Cave Art: Mogao Cave No. 45], Nanjing: Jiangsu meishu chubanshe. Yoshida Yutaka 吉田豊 Sogudo go zatsuroku (II) ソ グ ド 語 雑録 (II) [Sogdian Miscellany II], Oriento 1988 オ リェン b ,31.2, pp. 165-176. Zhang Qingjie 張慶 捷 "Beichao Sui Tang de hushang yong, hushang tu yu hushang wenshu" 北朝隋 2004 唐 胡商 俑 胡商 圖 與 胡商 文書 [Foreign Merchant Figures, Pictures and Documents in the Northern Dynasties, Sui and Tang Period], in Rong Xinjiang, Li Xiaocong 李 孝聰 (eds.), Zhongwai guanxi shi: xin shiliao yu xin wenti 中タト 關係 史： 新史 料與 新 問題 [History of Sino-Foreign Relationships: New Materials & New Questions], Beÿing: Kexue chubanshe, pp. 173-203. Zhang Tieshan 張 鐵山 1991 "Guowai shoucang kanbu de huihuwen fojiao wenxian jiqi yanjiu" 國 外收 藏 刊布的回鷗文佛教文獻及其研究 [Studies and Editing of Uighur Buddhist Literature in Foreign Collections] , Xiyu yanjiu 西域研究,No. 1, pp. 135-42.

•

ヽ

230

Les Sogdiens en Chine

ZHONGGUO BIHUA 03 中國壁畫全集 Zhongguo bihua quanji 中國壁畫全集 [Complété Collection of Chinese 1992 Murais], Vol. 8, Tianjin: Tianjin renmin meishu chubanshe and Urumqi: Xinjiang meishu sheying chubanshe. Zhu Yingrong 朱英榮 Qiucishiku yanjiu 龜茲石窟研究 [Study of Kuchean Caves]; Urumqi: 1993 Xinjiang meishu sheying chubanshe.

Sogdian merchants and Chinese Han merchants during the Tang Dynasty Arakawa Masaharu

Wim Judging from the contacts among peoples and the circulation of goods, the esta? blishment of the Tang Empire may be described as a process of taking over the admi“ nistrative power over communication and trade from the local govemment to the central govemment. Opening of post roads linking the capital city and all préfectures in the empire clearly illustrâtes this process.1 As the centralized administration of communication and trade was fbrmed, the com¬ munication System (consisting of post roads and canals) running all over the Tang Empire’s territory was also established, enabling direct contact between the capital city of the Tang Empire and the Central Asian région. The complété metropolitan grid within the territory of the Tang Empire, brought personal movements and the circulation of material goods within the empire to a new height. The activities of merchants in the circulation of Commodities had been brisk prior to the Tang dynasty. During the Tang dynasty, these activities produced a new and vigorous development. The région around the capital city Chang’an with its huge population deserved the name of the largest consuming metropolis of that time. With such enormous consumption needs, it attracted a flow of Commodities brought by merchants.2 Here I would like to call my reader’s attention to the fact that Chang’an was a starting point and terminal for the commercial activities of merchants from Central Asia, the north and north eastem régions of China and the areas south of Yangzi River which was linked to the north by post roads and Grand Canal.

One of the active merchant groups within the domain of the Tang Empire were the Sogdians. It was probably, during the Eastem Han Dynasty that Sogdian merchants began to visit China, and the Northern Wei Dynasty of the fifth Century saw an upsurge of their activities. They set up Settlements along the roads fbr their Caravans and in the destinations of their goods, and with these settlements as their bases, they engaged in trade. In the territory of the Tang Empire such settlements were fbund in major cities along the post roads, from capital cities to Central Asia and to Youzhou ® , and Hedong in the circuits (dao ü) of Hebei Yingzhou Taiyuan 1 ?

Arakawa, 2000, p. 214-220; 2002, p. 343. Cf. Hino, 1968, 1970.

Les Sogdiens en Chine

232

in the empire s north and northeast (Sec map).3 Their settlement bases, allowed the Sogdians to set up their own communication System. In Northern Dynasties, befbre the Tang era, these Sogdian settlements were indirectly ruled through the s'rtp'w 薩 寶,'but during the Tang empire, Sogdian people in the settlements had become ba 灰277g百姓 (permanent residents enrolled on general household registers), and the Sogdian settlements were incorporated in the administrative Organization of the Tang.5 This Sogdian communicational mechanism, which had been established long before, was reinforced by the public communication System, which supplied two kinds of traffic travel permits called guosuo 過所 and gongyan 公験, introduced and perfected by the Tang Empire. The examination of Turfan documents (see examples I5 II and III in the footnote6) reveals that gongyan guaranteed the passage within the territory of a préfecture and it was valid for a limited period that coincided with the lengtli of a given official joumey. By contrasta the guosuo passport had no spatial or temporal limits; it was issued for joumeys beyond the territory of a préfecture, and was valid throughout the territory of Tang empire.7 The Sogdian communicational mechanism combinée! with Tang travel permits enabled Sogdians to complété long-distance joumeys, allowing speedy and safe communication and transportation. 9

3

Aralcawa, 1999, p. 84; Rong, 2000, p. 119-151; 2001, p. 37-110; de la Vaissière, 2002, p. 148-149. Almost ail of these settlements in the Circuit of Hebei were located in a région where silk was collected as tax, in a word, main product région of silk in Tang empire. Cf. de la Vaissière, 2002, p. 150; Arakawa, 2004, p. 28-32. 4 Arakawa, 1998, p. 171-176; Rong, 2003, p. 128-143. 5 Arakawa, 1997, p. 192-196; 1998, p. 176-179. 6 ( I 「 ) 開元二十一(七三三)年西州都督 府 案巻 為勘給過所事」[The21s%fKaiyuan(733),anofficial document issued by the Government general of Xizhou, conceming the exammation of an application for a issue of gosuo passport.] (73TAM509: 8/21(a)之一,8/21(a)之二,8〃5(a)之一,8ハ 5(a)之二) [Jiang Huaming, a permanent resident enrolled on general household register 【北庭金満県百姓 '蒋 化明】 in Jinman County in the Protectorat© of Beiting] Yizhou 伊州 勺 Xizhou西州 与 Jimsa 北庭 一a v guosuo 過所 」 v gongyan 公験」 一mXizhou西州 * Jimsa北庭 (n) 「開元二十一 (七三三) 年唐 循忠 、薛光 泄;、康大之請給過所 案 巻」 [The 21stof Kaiyuan (733), an official document issued by the Government general of Xizhou, conceming applications for an issue of guosuo passport] (73TAM509: 8/4-1(a), 8/4-l(b), 8/23(a), 8/4-2(a)之一,8/4-2(a)之二, 8/4-2(a)之 三,8/4-2(a)之四) 【1•前長史 唐 循忠之 2•甘 州百姓 薛 光 泄; 3•康大之] [1 . A concubine of the ex-deputy prefect, Tang Xunzhong. 2. Xue Guangci, a permanent resident enrolled on general household register in Ganzhou. 3. Kang Dazhi] 1. Anxi 安西 ->Xizhou 西州 Yizhou 伊州 Guazhou 瓜 州 Ganzhou 甘州 一a v guosuo 過所」 …f Fuzhou ネ 州 v guosuo 過所」 一 2. Xizhou 西州％ Yizhou 伊州 夕 Guazhou 瓜州 も Ganzhou 甘州 v & 一* gongyan 公験 」 3 . [Xizhou 西州 Luntai 輪圍 [The 2 Ist of Kaiyuan (733),a record (m )「開元二十一(七三三)年染勿等保石染 典 往 伊州市易辯 辞」 of interrogation to the guarantors Ranwu etc., conceming the matter that Shi Randian goes to Yizhou and make a trade here] (73TAM509: 8/9(a)之一等) 【西州都督 府百姓 石 染 典】[Shi Randian, a permanent resident enrolled on general household register in the Govemment-general of Xizhou] r guosuo 過所」 Anxi 安西一… Xizhou 西州 & Y,zhou 伊州 * Shazhou 沙州 & Guazhou 瓜州 [kizhou 西州 * Yizhou 伊州| — 「 gongyan 公,験 j 7 Arakawa, 2000,p. 294-310; 2001,p. 6. 0

0

—

—

ヽ

一

Arakawa Masaharu

Sogdian .Merchants and Chinese Han Merchants

Iz^oiid Nuwber of regigtexeä hauw'hold.n Kaiyuen pariod 17“ and 18^ yeate of Kaiyiian) A (Prefoeiure of amall hcmIc) ui

-J9.999 Q

Sefecture "ftneiiiuto Beide) 20,000-39.999

O (Prefoeluro nf large aeale) 100.000-

—

Main paÄt'tuude in T«ng foswA

® l’rcfeclur« where Eho pre*,n«? of Sogdian coloniea

or osliniaf.ad

coiißxniod

AH«® Fraient Iiro whioh paid thn Ins in

kind (diao) and tiw œrvéo exemption inx (yong) in ailb (acpording s i)äj-1^ng-!Äu'[>iM 3, Ilie Board cf Finance)

Map 1: Main trade roads and towns in Tang China.

233

Les Sogdiens en Chine

234

In principle, the administration of the Tang Empire did not permit foreign merchants to frequent the Empire. However, Sogdian merchants were an exception. With a pass guosuo gained in Central Asia, a Sogdian merchant who had nothing to do with diplomatie missions, could easily corne to Chang 5an.8 Within the complété traffic network of the Tang Empire, Sogdian merchants, no matter whether their résidences were in China itself or not, could engage in trade while keeping close contacts with the authorities of different préfectures who were authorized to grant guosuo.

Also, in the latter half of the eighth Century at the latest, it seems that Persian (Bosi 波斯) and Arab (Dashi 大食) merchants, mainly carried on trade around cities along tiw Grand Canal south of Chang an and Luoyang, and Coastal metropolitan cities. That this 力我 who bought o treasure (Huren maibao was the case is evident from the Story of buys treasures such as jewels in China, merchant Hu a in tan 胡人買宝譚), which 胡 analysed by Seo Tatsuhiko^9 The stage for this story must have been one of the large Guang¬ cities such as Chang an, Yangzhou 揚州, Hongzhou 洪 州, Changzhou 吊 zhou 広州 and so on, and took place later than the Kaiyuan 開 兀 and the Tianbao 天寶 era, when most of non-Han people who bought treasures in China, were Persian and Arab merchants. When you locate these cities on a map, you will notice that they are on the major commercial routes which connected Chang an and the South China Sea at that time. Although the area for Persian and Arab merchants and that of the Sogdian merchants partly overlapped, the story indicates that in the latter half of the eighth Century at the latest, the main stage of Sogdian commercial activities was in northem China and that of Persian and Arab merchants took place in the South. 5

5

5

Hu 胡 (non-Han fbreigners) itinérant merchants existed alongside resident merchants (zuo 坐賈) during the Tang Dynasty. Some of them were wealthy merchants (hao shang 蒙商). We know from historical evidence, that the circulation of Uyghur (Huihu 回 骨島) and Persian money reflects the commercial activities of the Hu traders. The Uyghur money was the financial capital of Sogdian merchants,10 and the Persian money was the capital of Persian merchants.11 Despite their name of "money”, they were not currency, as we usually understand it. It was, instead, a general term fbr gold, silver, silk and other valuables used as capital for business and crédit. Apart from offering loans and goods/money conversions, as financiers in possession of huge amounts of capital, these wealthy merchants were also entrusted with the care of money deposited by other people. They carried on currency conversions and transmitted money, for which they delivered to the depositor invoices, checks and other certificates. They were also involved in investment. Judging from the geographica! distribution of the Sogdian and Persian merchants, the Sogdian merchants shared the foreign financial capital with the Persian merchants, the former in the north of China and the later in the south. 8

Arakawa, 1997, p. 171-184.

9Seo, 1991,p. 283-306. 10Moriyasu, 1997A,p. 93-119; 1997B,p. 24-28; Hino, 1965B,p. 38-49. nHino, 1965A,p. 367-381.

Arakawa Masaharu

Sogdian Merchants and Chinese Han Merchants

235

Needless to say, not only Hu marchants undertook commercial activities during the Tang Empire, but many ethnie Han Chinese were also involved in commercial acti¬ vities. One of them who has been repeatedly cited in the literature is Zou Fengzhi 鄒 鳳 熾, a merchant active in Chang5 an during the reign of Emperor Gaozong 高宗 (AD 650-683). While buying and selling mantou 饅頭 (a steamed pastry with stuffing), he found a large crock of gold in Shengye Fang ( 勝業 坊 )to the North of the city's Eastem market and thus became a millionaire. He built himself a large résidence in Huaide Fang 懷德坊 to the West of the Western market of Chang'an, a district inhabited by many Sogdian mer chants» He opened and managed dian 丿占 throughout the country, making huge profits by trading local products from different places.12 The dian or didian 氐 店 were set up throughout the territory of Tang Empire, and they functioned not only as lodgings, but also as storehouses and liubs of transportation and monetary

facilities.13 Apart from Zou Fengzhi, other noted wealthy Han Chinese merchants in Chang 'an included Wang Yuanbao 王元宝, Yang Chongyi 楊 崇義, and Guo Wanjin 郭万金." Naturally, among the Han Chinese merchants there were not only resident merchants, but also large numbers of itinérant merchants.

We should pay a great deal of attention to the description mentioned in the article of Lidai shengshuai hukou 歴代 盛衰戸口 in the Tongdian 通典 7, Shihuo 食貨 7. It infbrms us that, during the 13th year of the Kaiyuan 開兀 era (725), there were many diansi 店肆 that served the needs of itinérant merchants (shanglü 商旅) as far as Songzhou 宋 州 and Bianzhou 州 to the east, Shuchuan 蜀川, Qizhou 岐 州 and Liangfu 沢 府 to the west, Jingzhou 荊州 and Xiangzhou 襄州 to the south, and Taiyuan 太原 and Fanyang 范陽 to the north. "Diansi is a complex word composed with "dian" 店 and "si" 肆,both meaning "shop”. From this description, we can see that itinérant merchants called shanglü were greatly involved in commercial activities during the Kaiyuan era, basing themselves in "diansi" shops in the above areas of which Chang 'an and Luoyang were the centers. In addition, according to the Turfan documents, such "dian" were set up even in Turfan of Central Asia.15 Not only Han people but also Hu people managed the "dian" shops in Tang Empire. The Hu merchants had business dealings with the Han people, and Han merchants in their tum, traded with Hu people. We can see from the descriptions of the Tongdian and from Turfan documents, that during the Kaiyuan era in the above mentioned locations, many Han and non-Han mer¬ chants engaged in widespread commercial activities through the use of "dian". We should also notice that the location of Sogdian Settlements, mostly occurred in the 12

Taiping guangji 太平廣記,j. 495 Zalu 雜錄 3 Zou Fengzhi 鄒 鳳熾条参考 Hino, 1968; 1970 14 j- 1 (Saoxue yingbin 掃雪迎賓, Haoyou 豪友* Yingwu See the Kaiyuan tianbao yishi gaoshi 鸚鵡告事条),and Nanbu xinshu 南部新書,j. 8. 15 For exampïe, see the court file dated Tianbao 1 (762) (73TAM509:8/l(a),8/2(a), ed. in Wenshu 9, p. 128-134) which indicates "the shop of Zhang Youhe” (Zhang Youhe dian 張 遊鶴 信). 13

ヽ

Les Sogdiens en Chine

236

above mentioned régions, where "dian!! were set up, centered in Chang an and Luoyang. 5

While competing with each other, the ethnie Han Chinese merchants and the Sogdian merchants exchanged information with each other, lent money to each other and transferred Commodities from each other and even helped to build up each other 's business. Their close relationship is illustrated by documents unearthed in Turfan. These documents provide a case of coopération between Han and Sogdian itinérant merchants in trade in Central Asia, a phenomenon that emerged shortly after the Tang Empire5s sphere of influence reached Central Asia during the reign of Emperor Gao zong.

A mémo (draft) submitted to the Protectorate of Anxi (Anxi duhufii 安西都護府) by Gaochang County 高昌縣. Re£: record the interrogation to Cao Lushan 曹 禄山 the plaintiff and Li Shaojin 李 紹謹 the défendant. 16

Document 1 (66TAM61:17(b). (The first two lines that belong to a draft memorial to the emperor hâve been omitted.)

8 ロロ

牒上安西都護周 縣 山年世 (依檢？ ) 案内 〕 (牒得) 上件人辭稱 、向西州長史〔 〕 ロ在弓 月城有 京師漢名国〔 10 在弓月城擧取二百七十五疋絹、向圍 両) 圖相逐 、從弓月城向亀茲 。阿兄更有 (茲。 (馬) 届、馳両頭 、牛四頭 、驢一頭 、百疋絹價華 并椀 、別有百疋絹價財物及漢鞍衣裳 調度 。其李三両箇相共 從弓月城向亀茲 、 不達到亀茲 。其李三是漢 、有氣 力語同 、 身是胡 、不解漢語 。身了知此間 〔 〕 12 行恩澤於此間 、請一箇〔

3456789

1 2

ロロ ロロ ロロロ ロロ ロ ロ ロ

n

ロ

(後 欠)

Interprétation of the document: (Due to many missing parts, the document cannot be fully interpreted. The following interprétation only gives a rough draft of the meaning. Significant missing of words are indicated in brackets. The words among square brackets [] are added by the author of the présent article, to complément the text and make it intelligible.)

66TAM61:17(b), 23(b), 27/2, 27/l(b). Plates in Tuwen 3, p. 242-243; Transcriptions in Wenshu 6, p. 470-473. Huang, 1983,p. 344-361; Arakawa, 1997, p. 185-188.

16

Arakawa Masaharu

Sogdian .Merchants and Chinese Han Merchants

237

Memo Submitted by Gaochang County to the Protectorate of Anxi. Cao Lushan, at age 30 The said accuser Claims, "To the Governor of Xizhou (characters missing) a man named Li, ethnie Han from the capital city living in the Town of Gongyue 弓月 (characters missing) borrowed 275 bolts of silk from [my brother], then he went to Kucha (Qiuci). (characters missing) [My brother and Li] went to Kucha from Gongyue. My brother brought with him (characters missing) horses, two camels? four oxen, one donkey_and (character missing) and bowls worth one himdred bolts of silk“ Furthermore, he had property worth one hundred bolts of silk, as well as Han™ style saddles, clothes at his disposât [My brother] and Li the Third (i.e. e. Li Shaojin in Document 2) left Gongyue together fbr Kucha but [my brother] did not arrive in Kucha. Li the Third is a Han. He is Strong and speaks eloquently. [My brother] 17 is a Hu and he doesn5t know the Chinese language. I know that during this period... (the remaining part illegible).

Document 2 (66TAM61:23(b),27/2327/l(b)) (前

1

欠) 所歸、請乞禁身 、与謹對圍口(者)。

祿山

. 一

2 問得款 、李謹當時共兄同伴 、向弓回〔丨 曲

3

三并外牛 居者去〔

并 共 曹 果毅 及

〕

城

4 其曹果毅 及 曹二 留住 弓月 、其至三口ロ 5 兄邊取 練訖分明付、兄与 李三同n 6 西。 に見到、唯兄不來、既是 安西

7 西 。兄不至 、所以陳訴、更無口ロロロ 、図問風 8 山得款、別 兄已来 、經四年ロn (曹果) 毅 、曹二

ロロ ロ口多

9 胡輩處指的同擧 練ロロ(李) 三 。身及外

10 牛 回 逐李三後去 。其曹 果毅 、曹二是胡、 H 客 京師、有家口在。身當來日、留 住国 12 月城在、身亦不在豆口口(月城) 。當空三共 13 圃、 〔 〕 同 (後 欠)

Interpretation of the document (Like Document 1, this document is fragmented too. The following is a rough in¬ terprétation of the document. Signifïcant missing of characters are marked with "some 17

Tuwen 3, p. 242 note 1 believes a character xiong (elder brother) was left out befbre the character shen (one's self).

238

.

Les Sogdiens en Chine

characters missing here” in brackets. Phrases followed by a question mark indicate con¬ jectures. The words between square brackets [ ] are added by the author of this article, to complément the text and make it intelligible.) (characters missing) may return, I plead with you to detain [Li the Third], [so that I] could confront him in court. [The following is] Cao Lushan's complaint: “Li Jin went to the Town of Gongye with my brother. Along with them were Cao Guoyi Cao Er and his nephew. Cao Guoyi and Cao Er remained in the Town of Gongyue, while Li the Third (tow characters missing here) borrowed silk (Han from my brother, and then went to Anxi with him. (several characters missing here). Now Li the Third has arrived in Anxi but my brother alone has not arrived. Therefore I am lodging a complaint to the court. (Several characters missing), [The following is] Cao Lushan’s other statement: “Four years hâve passed since I left with my brother. Cao Guoyi, Cao Er and other Hu did lend silk to Li the Third. My nephew and I followed Li the Third there soon after. Cao Guoyi and Cao Er are Hu all lived in the capital city. Their families are there. (The remaining part is omitted)

The documents consist of ten fragments. We cannot examine the whole documents in this article. Instead, we will just make tentative analysis of the first two fragments. The earliest possible date of the documents is the fourth moon of the first year of ) era (AD 670) when the Anxi Protectorate was moved to Xizhou due Xianheng’s( to the capture of the four garrisons of Anxi by Tibetan troops. The latest possible date is the fourth year of Xianheng Reign (AD 673) when the occupant of the excavated tomb died. The basic content of the documents concems a lawsuit caused by a loan of silk between the Sogdian creditor Cao Lushan’s elder brother (Cao Yanyan and the Han debtor Li Shaojin (Li the Third). The party who lodged the complaints in the Xizhou court was Cao Lushan. At the beginning of Document 1, Cao Lushan’s accusation (c/ is quoted and corroborated by the first line of Document 2. The other parts of the fragmented documents are records of the case 's investigation and State¬ ments made by concemed persons (signature and oral testimonies). Document 2 is a re¬ cord of Cao Lushan’s statement. By analysing this documents, we leam that all the people involved in the law-suit were Sogdian and Chinese Han living in the capital Chang’an. They formed a group and went to Central Asia. The witnesses to the financial transaction were Cao Guoyi and Cao Er (Bi Suo ^^) from China’s capital city (see Document 2). This is a case of a commercial venture carried out jointly by Chinese Han and Sog¬ dian merchants (xinghu ÂÉ0) living in the capital Chang’an. They formed a trade Cara¬ van, went to the Anxi town of Gongyue, and further westward, and carried on com¬ mercial activities over a wide area. So far, we hâve not found any evidence of such commercial joint ventures in ré¬ gions other than Central Asia, but we can assume that such ventures existed even in the Chinese heartland, where many Han and non-Han merchants carried out a flourishing trade through the use of a large number of “dian”. In the latter half of the eighth Century, while an increasing number of Persian and Arab merchants made inroads mainly into Southern part of China, Sogdiana came under the control of Islam. However, as one can read in Akhbâr aï-Sïn wa al-Hind, there were

Arakawa Masaharu

Sogdian .Merchants and Chinese Han Merchants

239

still Sogdian merchants traveling overland to China’s heartland from Sogdiana in ninth Century.18 Also, just as Moribe Yutaka and Étienne de la Vaissière have already pointed

out.19 we have to bear in mind that the Sogdians coming from Mongolia like such as the Turkic Sogdian, were entering in China heartland. And in the middle of ninth Century, Uyghur began to move southward and west ward.

Under these circumstances, after tenth Century, the Uyghur merchants who arose after the Sogdian merchants, built their trading networks in northem China, Mongolia and Central Asia, as their predecessors had done. The “Uyghur” merchants in this article while referring to merchants from the (Uyghur) state of Huihu, in fact, included many Sogdian and Han people.20 Chinese Han merchants must still have preserved a close relationship of mutual assistance with Sogdian, condusive to the exchange of information, accommodation of fonds and merchandize, and business coopération in northem China, Mongolia and Central Asia. Sogdian and Uyghur texts of the tenth Century attest a word tym meaning “inn, It indicates that hotel” which is known to have been borrowed from Chinese dian “dian”, which were set up throughout Central Asia and China’s heartland during the Tang dynasty, played an important role not only in the commercial activities of Han people, but also in the activities of Sogdians over a long period of time. The trade network of the “Uyghur” merchants in northem China, Mongolia and Central Asia, was more or less based on the Sogdian-Han’s model that existed during the Tang dynasty.

18

Fujimoto, 1976,p. 56. de la Vaissière, 2002, pp. 196-221; Moribe, 2004. 20 Moriyasu, 1997B,p. 28-35,111-116. 21 Sims-Williams and Hamilton, 1990, p. 30; Yoshida, 1994, p. 379. 19

Les Sogdiens en Chine

240

Abbreviations

…

Wenshu 文書 °Tulufan Chutu Wenshu 吐 魯 番出土文書 1-10, Wenwu-chubanshe 文物出 版社,Beijing, 1987991. Tulufan Chutu Wenshu 吐 魯 番出土 文書 1-4, W enwu-chubanshe, Beijing, Tuwen 圖文 ° 1992/996. 3

り

Bibliography

Arakawa Masaharu 荒丿H 正晴 1997 "Töteikoku to Sogudojin no köeki katsudö 唐帝国とソ グ ド 人の交易活動" [Sogdians in the Tang Empire] In: Töyöshi kenkyü 東洋史硏究 [The Journal of Oriental Researches] 56-3, p. 171-204. "Hokucho Zui Tö Dai Ni Okeru Sappö No Seikaku Wo Megutte 北朝隋 唐代 1998 に おける「 薩 寶 」 の性格をめぐ つて”[Sか on the character ofs >rtp w 薩 寶 in the Northern, Sui and Tang dynasties} In: Töyöshien 50-51, p. 164-186. "Sogudojin No Ijü Shüraku To Töhö Köeki Katsudô ソ グ ド 人の移住聚落と 1999 東方交易>舌動''[Sogdian Migration如 Colonies and Their Commercial Activities] In: Iwanami Kôza Sekai Rekishi [Iwanami World History], Vol.15, Iwanami shoten, Tokyo, p. 81-103. "Tôchô No Kôtsü Shisutemu 唐朝の交通シス テ 厶''[Communication System 2000 in the Tang Period] In: Osaka Daigaku Daigakuin bungaku kenkyüka kiyö 大 阪大学大学院 文学研究科紀要 [Memoirs of the Graduate School of Letters Osaka University] 40, p. 199-335. "The Transit Permit System of the Tang Empire and the Passage of Merchants, 2001 In: The Memoirs of the Toyo Bunko, 59s p. 1-21. "Gi Shin Nanbokuchö Zui Tö Ki No Tsüka Köshö Seido To Shönin No Idö 魏 2002 晋南北朝隋唐期の通貨公證制度と商人の移動”[Official Assurance System of Passing and Travel of Merchants in the period of Wei, Jin, the Northern and Southern, Sui and Tang Dynasties] In: Chügoku No Rekishi Sekai, Togo No Shisutemu To Tagenteki Hatten 中国の歴史世界,統合のシス テムと多元的發展 [Historical World of China, System of Unification and Plural Development], Tokyo Toritsu Daigaku Shuppan, Tokyo, p. 337-349. "Tö Dai Ni Okeru Hito No Idö To Kahei No Ryütsü" 唐代に おける人の移動 2004 と貨幣の流通 [Personal Movements and Circulation of the Coins and Silk under the Tang Rule] In: Tödaishi Kenkyü 唐代史研究 [The Journal of Tang Historical Studies] vol.7, p. 17-59. Fujimoto Katsuji (translation with commentary) 1976 Akhbär al- Sm wa al-Hind, Kansai University Press. Hino Kaisaburö 日野 開三郎 1965A “Tödai No Hashi Sen Ni Tsuite” 唐代の波斯銭について [Persian Coins, during the Tang Dynasty] In: Ishida Hakushi Koju Kinen Töyöshi Ronsö 石田 博士頌壽記念 東洋史論叢 [Studies in Asian History dedicated to Prof. Dr. Mikinosuke Ikeda on the Occasion of his Seventieth Birthday], Tokyo, 9

a

Arakawa Masaharu

Sogdian Merchants and Chinese Han Merchants

241

pp. 367-38 L [Reprmted in: Hino Kaisaburö Töyö Shigaku Ronshü 日野開三 郎東洋史学論集 [Posthumous Collected Papers of Kaizaburö Hino on the Eastern Studies] 5,1982, pp. 231-243] 1965B “TOdai No Kaücotsu Sen Ni Tsuite!, 看代の回紇 銭に つし、て [Hui-he Ch '旭题 回紇 銭 山 the Time of the T'ang Dynasty] In: Töhögaku 東方学 [Eastern Studies] 30, pp. 38-49. (Reprinted in: Hino Kaisaburö Töyö Shigaku Ronshü [Posthumous Collected Papers of Kaizaburö Hino on the Eastern Studies] 5, 1982, pp. 246-259). 1968 Tödai Teilen No Kenkyü 唐代 邸店の研究 [Study of Dian-di in the Time of the Tang Dynasty], Fukuoka, Reprinted in: Hino Kaisaburö Töyö Shigaku Ronshü 日野開三郎 東洋史学 論集 [Posthumous Collected Papers of Kaisaburö Hino on the Eastern Studies] 17, 1991. 1970 Zoku Tödai Teiten No Kenkyü 続 唐代 邸 店の研究 [A Study of Dian-di in the Time of the Tang Dynasty, a Second Sériés] Fukuoka, Reprinted in: Hino Kaisaburö Töyö Shigaku Ronshü 日野開三郎東洋史学論集 [Posthumous Collected Papers of Kaizaburö Hino on the Eastern Studies] 18, 1992. HUANG Huixian 黃惠賢 "«TangXizhou Gaochangxian shang Anxi duhufu diegao wei lushang xunwen 1983 Cao Lushan su Li Shao-jin liangzao bianci shi) shi «唐西州高昌縣上安西都 護府牒 稿為錄上訊問曹祿山訴李紹謹兩造 辯 辭事»釋 (A Note on the Turfan documents 'Memo (draft) submitted to Protectorate of Anxi by Gaochang County')" In: Dunhuang Tulufan yvenshu chutan 敦煌吐 魯番文書 初探,Wuhan Daxue Chubanshe 武漢大學出版社,p. 344-363. 5

Étienne 2002 Histoire des Marchands Sogdiens, Paris, Collège de France. Moribe Yutaka 森 B 豊 2004 "Tömatsu Godai No Daihoku Ni Okeru Sogudokei Tokketsu To Sada" 唐末五 代の代北に おけるソ グド 系突厥と沙陀 (The Sogdian Turks and Shatou in Daibei during the Late Tang and Five Dynasties Periods) In: The Töyöshi- Kenkyü 東洋史 研究 (刀/e Journal o于 Oriental Researches) 62-4, p. 60-93. MORIYASU Takao 森安 孝夫 1997A "«Shiruku Rödo» No Uiguru Shönin, Sogudo Shönin To Orutoku Shönin No Aida” «シルクロード»のウイグル商人、ソグド商人とオル ト ク商人のあい だ [The Uighur Merchants on the Silk Road from the Sogdian Merchants to the Ortoq Merchants] In: Iwanami Kôza Sekai Rekishi 岩波講座世界歴史 [Iwanami World History], Vol.ll, Iwanami shoten, Tokyo, 1997, p. 93-119. 1997B "Orutoku to Uiguru shönin." オルト クとウイグル 商人 [Ortoq and the Uighur Merchants] In: T. Moriyasu (ed.), Kinsei Kindai Chügoku oyobi shühen chiiki ni okeru shominzoku no idö to chiiki kaihatsu 近世 °近代中国お よ び 周 辺地 域 に おける諸民族の移動と地域開発 Migration ofPeoples and Community Development in Pre-modern and Modern China and in Surrounding Areas, (Report of the Scientific Research Project Grant-in-aid, Japan Ministry of Education, No. 07451082),Toyonaka, Osaka, 1997 March, p. 1-48. Rong Xinjiang 榮 新江 "The Migrations and Settlements of the Sogdians in Northern Dynasties, Sui 2000 and Tang” In: China Archaeology and Art Digest^ vol. 4,No.l,p. 117-163. DE LA VAISSIÈRE,

Les Sogdiens en Chine

242

"Beichao Sui Tang Suteren zhi qianxi jiqi juluo 北朝隋唐粟特人之遷徙 及其 聚落”[The Migrations and Settlements of the Sogdians in Northern Dynasties, Sui and Tang] In: Zhonggu Zhongguo yu wailai wenming 中古中国 與 外 來文 明, Shenghuo * Dushu • Xinzhi sanlian shudian 生活 ©讀書 ® 新知三聯書店, p. 37-110. 4tSabao yu Sabo 薩 保與薩 薄 [s ' 2003 ァtp'w and sârthavâha]：Beichao Suitang huren juluo shouling wenti zhenglun yu bianxi 北朝隋唐胡 聚落首領問題争論 與 辨析”In: Yilanxuezai Chongguo lunwenji 伊 朗學 在中國 論文集 (第三 集), Beijing taxue chubanshe 北京大學出版社,p. 128-143. Seo Tatsuhiko 妹尾 達彦 "TGdai Chöan No Tenpo Ricchi To Gaisei No Chifütan" 唐代長安の店舗立 1990 地と街西の致富譚 (Th©Location of Shops and a Success Story of a Merchant in the West Market in T'ang Ch3ang-an) In: Numme Chöfü Hakushi Koki Kinen Ronsyü, Higashi Ajia No Hö To Shakai 布目潮風博士古稀記念論 集 ヽ 東 アジアの法と (Studies ofLaw and Society in East Asian History Honour ofD.Litt Nunome Chöfü in the Occasion ofhis Seventieth Birthday), Kyûko Shoin, Tokyo, 1990, pp. 191-243. "Toshi No Gaikoku Shönin - Hachi, Kyü Seiki No Chügoku Ni Okera Ijin 1991 Baihötan” 都市の外国商人 、8 、9世紀の中国に おける 異人買宝 譚 [Merchants and Commodities of non-Han people in China] In: Toshi to Kyödötai 都市と共同体 (City and community), Tokyo, meicho shuppan, 1991,pp. 283-306. Sims-Williams, N., Hamilton, J. 1990 Documents turco-sogdiens du IXe- Xe siècle de Touen-houang, (Corpus Inscriptionum Iranicarum, Part II /VoLW/IH), London. Yoshida Yutaka 吉田 豊 1994 “SogudoMojiDeHyokiSaretaKanjiOn” ソグド文字 で 表記された 漢字 音 [] In: The Töhö Gakuhö 東方学報 (Jo況mR of Oriental Studies Kyoto Number) 66, 1994, p. 271-380. 2001

入

Military officers of Sogdian origin from the late Tang Dynasty to the period of Five Dynasties Moribe Yutaka 森部豊

This paper clarifies the role played by the military officers of Sogdian origin in North China, using epitaphs published in recent years or newly discovered.

Armed Rebellion Led by An Lushan and Shi Siming and Military Officers of Sogdian Origin. When the armed rébellion led by An Lushan and Shi Siming 史思明 broke out in AD 755, that date acts as a dividing line between the early and late Tang periods. If we compare the ealier with the later period, we note that the political influence of the Tang during the later had decrease. However, by enfbrcing a sériés of financial reforms, such as introduction of a sait monopoly and the adoption of the two-tax System, the Tang dynasty entered into a relatively stable period that lasted, until the feil of the dynasty 150 years later. However, the disastrous conséquence of An Lushan5s armed rébellion was never eliminated. One conséquence was that the remnants of the armed rébellion set up semi-independent military régimes in the Hebei area (known as the Three Garrisons in Heshuo area 河 朔三鎮 一 "Heshuo” meaning "north of [the bend of] the Yellow River"), in défiance of the political power of the Tang court. Rulers of the Three Garrisons of the Heshuo area did not report their registered population or deliver collected taxes to the central govemment. Within the area, local officiais were selected and appointed by the local military lords. As a resuit, the Three Garrisons and the Tang central govemment had a tense and antagonistic relationship, not only politically but also militarily. From the outbreak of the armed rébellion led by An Lushan and Shi Siming in AD 755 to the establishment of the Three Garrisons in Heshuo area (AD 763) and later in the 8th Century up to the accession of Emperor Xianzong to the throne in AD 805, the period witnessed the greatest strain and the worst antagonism between Three Garrisons in Heshuo area and the Tang court. It has been confirmed that during this period many military officers of Sogdian origin were active in the military govemorship in Hebei. Besides the two leaders of the armed rébellion, An Lushan and Shi Siming, historical documents such as the Jiu Tang Shu and the Xin Tang Shu as well as the Zi Zhi Tong Jian (Historical Events Retold as a Mirror for Govemment), confirm that some subordinate officers in the rebellions army, such as He Qiannian 何千年,He Side 何 思 德, An Shouzhong 安守忠,General Cao 曹將軍,Cao Minzhi 曹閔之,Shi Dingfang 史

Les Sogdiens en Chine

244

定方, etc, were all of Sogdian origin. The article on "the twelfth moon of the second year of Zhide Era” (AD757) in Zi Zhi Tong Jian (chapter 220) daims that when An Qingxu 安慶 緒 was defeated by the allied forces of the Tang govemment and the Uyghur, he fled from Luoyang to north of Yellow River, leaving the troops under his command, who were an ethnie mix of Yeluohe 曳落河,Tongluo 同維 and Hu from the Six Préfectures (Liuzhou hu 六州胡), free to sadc Hebei on their way to Fanyang 范陽 (Youzhou 幽 州). On the instructions of Shi Siming, who was then Military Commissioner (jiedushî) of Fanyang5 Yeluohe and Hu soldiers from Six Préfectures were willing to surrender. The so-called "Hu from Six Préfectures^ mentioned here was a group of military offïcers of Sogdian origin? Furthermore, in the third moon of the second year of Shangyuan Era (AD 761), when Shi Chaoyi 史朝義 detained his father Shi Siming and sat on the throne, he caused chaos m Youzhou by slaughtering members of the opposition party led by his younger brother3 Shi Chaoqing 史朝)青 and his mother? née Xin 辛。In the article about the "Chaos in Jimen 薊門" quoted in ZZ Zhi TongJian Kaoyi there are records about "Hu" people with prominent noses being massacred in great numbers. Thus it can be seen that many Sogdians lived in Youzhou during the period of An Lushan and Shi Siming rébellion. After suppressing An and Shi s armed rébellion, four Military Commissioners installed in Lulong 廬龍, Chengde 成德, Weibo 魏博 and Xiangwei 相 衛 in the Hebei area by the Tang court, were ail former members of the military forces of An Lushan and Shi Siming. Among them, the régime of Military Commissioner of Xiangwei soon collapsed. The other three local administrations were later known as the "Three Garrisons of the Heshuo area", a synonym for anti-Tang forces. Many military officers of Sogdian origin were also active in the Three Garrisons of the Heshuo area. For example, the text on the reverse side of the "Eulogy of Li Baochen, Military Commissioner of Chengde" 成德節度使李寶臣碑 which is dated the second year of s Yongtai Era (AD 766)125 includes the names of 76 civil and military officiais of the Chengde régime. This is a first-hand historical document for the study of the early period of the Chengde régime. Among the 76 names on the list, excluding four illegible names due to damage on the stone, those of An Dutao 安者B滔, Kang Rizhi 康日矢口, Kang Ruzhen 康如 , He [xxx] 何ロ, An Zhongshi 安忠實, He Shanquan 何山泉, Kang Rizong 康日琮,Cao Minzhi 曹敏之,and Shi Zhaofu 史招福 can be identified as military officers of Sogdian origin. This provides crédible evidence that a large number of military officers of Sogdian origin were active during the period of the armed rébellion led by An Lushan and Shi Siming and the period of the Three Garrisons of the Heshuo area. What is noteworthy is that the military officers mentioned above were not pure Sogdians, but Sogdians who were influenced by Tujue (Turkish) people, and who led a semi-nomadic life. In the first Khanate of Eastem Tujue that dominated North Asia, there were many ethnicities 5

1

。。

1942;Pulleyblank,1952. 5 止01118IX P[è For early ,m山ès Qf Onogawa '匸加ロ且イ©jun jiedushi kaifu yitong 5308111 jiaIljia。 shangshu youpuye "血 yushi dafu llengzhou cishi chong guannei yingtianshi Qinghejun waIlg L^gong jigong zaizheng song bmg xu 成德軍節度使開府儀同 三 使檢 校尚書右仆射 兼禦 史大夫恒 州御史充管内營田使清河郡王李公紀功載政 頌並序”, M S113 1842"〃師 10,reprmted 〇 Shike shilia。xmbian,1977,pp.13324-13329. 2

Moribe Yutaka

Military officers of Sogdian origin

245

incJuding Sogdians following the Tujue people. Because these Sogdians in North Asia lived for a long time together with Tujue and other nomadic ethnicities, they naturally assimilated nomadic cultures and became Sogdians excelling in horsemanship and archery, skills of which these northem people were proud. In the present paper, Sogdians are referred to as “Turkicised Sogdians”. The first Khanate of Eastem Tujue was destroyed in the 4,h year of Zhenguan Era (AD 630). At that time, the Tang court ordered that Tujue and those ethnie groups that followed them, to move and settle along the Great Wall in North China. The Tang court also conferred on the chieftains of those tribes, official titles of the Tang court in order to exert an indirect control over them. Turkicized Sogdians also migrated and settled in (the great northward bend of the Yellow River). After that, the survivors of Hequ Tujue khanate, congregated around the ethnie Ashina group, began their struggle to break away from the Tang Empire. In the first year of Tiaolu Era (AD 676), the Tang (Retaining Préfectures) for Turkicised Sogdians and court set up six “Jimizhou” appointed officiais from the Tang court to practice direct rule, hence the name “Six The inhabitants there were called “Hu from Préfectures of Hu” (Liuhu zhou the Six Préfectures”, (Liuzhou hu). The Hu people (non-Han Chinese or foreigners) led a nomadic life and were a potential military force. I believe that the motive of the Tang court in establishing the “Six Préfectures of Hu” in AD 679 was to prevent a Tujue revival. On the other hand, Tujue, who were planning an independent state, created the second Khanate of Eastem Tujue with the Yinshan Mountains ^lL[, north of the région around the great bend of the Yellow River, as their stronghold. In the Khanate lived Turkicized Sogdians, who had not been allowed to live in “Six Préfectures of Hu.” This was the historical background against which An Lushan was bom. In the 29” year of Kaiyuan Era (AD 741), the second Khanate of Eastem Tujue was destroyed by a mutiny of the Huihu |l]f| (Uyghur), whom they had subjugated. Befbre and after this time, many ethnie groups following the Tujue had again migrated into the Tang Empire’s territory. In the first years of the Tianbao Era (AD 742-755), the Tang court near Youzhou, which was established a Retaining Préfecture in Lingzhou subjugated Hu) in the “record for (The Préfecture registered as “Xianghuzhou of geography” of Jiu Tang Shu and Xin Tang Shu. I believe that the Tang court moved Sogdians who had once lived in the second Khanate of Eastem Tujue to this Préfecture. Many military Sogdian officers in An Lushan’s army were Turkicized Sogdians from the Six Préfectures of Hu in Hequ or remuants from the second Khanate of Eastem Tujue. Below I will give a detailed description of some military officers by drawing on spécifie historical records.

Three Garrisons of the Heshuo Area and Turkicised Sogdians Cao Runguo Cao Runguo once served in the troops of An Lushan and Shi Siming during their armed rébellion. After the rébellion, he became an aide to Military Commissioner

liïls

Les Sogdiens en Chine

246

of Chengde. A record of his life can be fbund in the epitaph of Cao Runguo? According to the epitaph, Cao Runguo was from Hanzhou 含州 in Hequ area. Hanzhou was one of the "Six Préfectures of Hu." The family name Cao was a Han family name that had been adopted during their stay in China by the Sogdians who formerly lived in the Country of Cao. This proves that Cao Runguo was a "Hu” from the Six Préfectures of Hu. Cao Runguo came to the north of Hebei for an unknown reason, but he remained in Hebei because he was drawn into the armed rébellion led by An Lushan and Shi Siming. and joined their rebellions troops. While serving with these troops, he was promoted to the rank of a General of the Cloud-like Flags 雲麾 將軍 and Acting General of Left Impérial Insignia Guard 寸 左金吾 衛 大呼哥軍。After the armed rebellion3 he surrendered to the Tang court and was made Minister of Impérial Entertainment 靑式光ネ泰 卿. He servcd as Military Commissioner of Cliengde and was made Inspector- in-Chief of Cavalry 馬軍都 虞 侯.In the 11* year of Dali Era (AD 776)5 he died of illness at the âge of 47. One of his wives was née Shi 石. On the basis of the year of his death and his âge at the time of death5 we can détermine that Cao Runguo was bom in the fïrst year of Kaiyuan Era (AD 730),and that he was 26 when the armed rébellion broke out in 755= We can therefore deduce that it was during the era of Tianbao (AD 742-755) that Cao Runguo moved to Hebei However, the route of his joumey from Hanzhou to Hebei and the reason for his movements remain unclear. The highest position he attained in the army of Chengde was Inspector-in-chief of the Cavalry. The responsibility of the inspector-in-chief, a functionary under a military govemorship, was to enforce army régulations, ensure safety inside and outside of the barracks, post sentinels, and gather intelligence about enemy. The record of Cao Runguo has survived in the form of an ep辻aph. But other Turkicized Sogdians, such as Kang Rizhi, Shi Xiancheng and He Jintao hâve biographies in official historical books to which will now tum. s

Kang Rizhi According to “Kang Rizhi Zkua优 康日知 傳 (Biography of Kang Rizhi) in Xin Tang Shu (juan 148), Kang Rizhi's ancestral home was Lingzhou 靈 州. Shi Xiancheng and He Jintao, who will be discussed later in this article, were also from Lingzhou. Because Lingzhou bordered on the Six Préfectures of Hu, and the family names Kang, Shi and He were Han names adopted respectively by Sogdians from the countries Kang, Shi and He, which leads the author to believe that Kang, Shi and He were Sogdians who had belonged to the Hu from the Six Préfectures (i.e. Turkicised Sogdians). The so-called Six Préfectures of Hu were situated to the south of Lingzhou and Xiazhou 夏 州. During the Kaiyuan era (AD 713-741) due to an administrative merger of several divisions, the names of the Six Préfectures of Hu was abolished. Consequently, it may well be that the inhabitants of the Six Préfectures began to take Lingzhou or Xiazhou as their native places.

3

"Tang gu shiguangluqing Caofujun muzhi bing xu 唐故試光祿 卿 曹府君墓誌並序”, in Luo Zhenyu, juan II, reprinted in Shike shiliao xinbian, p. 13628.

Moribe Yutaka

Military officers of Sogdian origin

247

The "Biography of Kang Rizhi” only teils us that Kang Rizhi's grandfather Kang from the Six put down an armed rébellion led by Kang Daibin Zhi Kang youth 721). his In Rizhi was Préfectures of Hu in the 9th year of Kaiyuan era (AD a subordinate military officer of An Lushan and later the an aide to Li Weiyue Military Commissioner of Chengde. Recently, adopted son of Li Baochen was published,4 some until then unknown Kang Zhida of epitaph the when to historical facts came to light. According the epitaph, Kang Zhida was the son of was a Vanguard General of the Kang Rizhi, Kang Rizhi's father Kang Xiaoyi Hedong Circuit. And Kang Rizhi's in the of Jinzhou Garrison Wan’an Zhi) to Chang’an in Jingzhao native was a (Kang Yanqing Kang grandfather, Kang Yanqing, but in grandfather was Préfecture. The epitaph says Kang Rizhi's the “Biography of Kang Rizhi" he is known as Kang Zhi. Despite the discrepancy, we can attribute the two names to the same person. Comparing the biography and the epitaph, we have an outline of the migration to Hebei of Kang Rizhi’s family. Obviously betöre and after the armed rébellion of An and Shi in AD 755 they first moved from Lingzhou to Chang’an and then, by way of Jinzhou, to Hebei. In the afore mentioned text on the reverse side of the “Eulogy of Li Baochen", Kang Rizhi’s was placed 24th among the 76 names. The 22 names preceding that of Kang Rizhi were either names of army Supervisors, or names of relatives of Li Baochen, Military Commissioner of Chengde, or again names of civil assistants, and included only one name of a military officer. Therefbre, Kang Rizhi was the second military officer in the list of 76. Afterwards Kang Rizhi became Left Wing Infantry Command of Chengde The verso of and then Prefect of Zhaozhou Army the “Eulogy of Li Baochen” clearly introduces the structure of the Chengde army in its initial phase. It consisted of cavalry and infantry troops, each of which were divided into the left and right wings. The commander of each wing was called “Dushi” (Commander). Given that the Chengde army was composed of four wings we can infer that Kang Rizhi, the commander of the left wing infantry, was in a high position and clearly played an important role in Chengde army. The name of Kang Ricong also appears in the Eulogy. Considering Kang Rizhi and Kang Ricong might be brothers or cousins, we can infer from this that the family of Kang Rizhi was not the only one that had moved to Hebei. Although we cannot yet teil the scale of the migration, we now know for a fact that many families of Sogdian origin moved from the Six Préfectures of Hu to Hebei. This migration took place no later than the period of the armed rébellion of An Lushan and Shi Siming.

Xiancheng who later became We will now investigate the career of Shi Xiancheng Military Commissioner of Weibo, one of the Three Garrisons of the Heshuo area. 4

“Tang gu youzhou lulong jun jiedu yaqian bingmashi chaosan dafii jianjiao guangluqing jian jiancha

yushi zeng mozhou cishi guiji kanggong muzhiming bing xu

in Suitang wudai muzhi huibian: Shanxi juan,

1991, p. 85; also in Quan tang wen buyi, 1998, pp. 431-432.

248

Les Sogdiens en Chine

Records about Shi Xianclieng can be found in “Shi Xiancheng Zhuan" (Biography of Shi Xiancheng), volume 181 in Jiu Tang Shu (hereafter, “Old Biography of Shi Xiancheng”) and “Shi Xiancheng (Biography of Shi Xiancheng), volume 210 in the Xin Tang Shu (hereafter, “New Biography of Shi Xiancheng”). According to these two books, Shi Xiancheng’s ancestors belonged to the Xi M ethnie group. But could Shi Xiancheng possibly be a Sogdian? The Eastem Tujue once extended their influence to Northeast Asia, and consequently part of the Xi people in Northeast Asia followed the Eastem Tujue to North Asia. The inscription in the tomb of Shi Xiaozhang that is Shi Xiancheng’s son, States that they “he had been powerful in the North and dominated there”. The record indicates that part of Xi people had once migrated to North Asia. Presumably at that time, Shi Xiancheng’s ancestors and the Sogdians who had also moved to North Asia must hâve influenced each other. Given that later the Eastem Tujue sunendered and pledged allegiance to the Tang court, those ethnie groups who once followed them, such as the Xi people and the Sogdians, moved to and settled in China. I believe that the Xi people and the Sogdians were ordered to inhabit the Six Préfectures of Hu. Therefore, like Kang Rizhi, Shi Xiancheng took Lingzhou as his native home. But given that Shi Xiancheng’s ancestors were Xi people who had adopted the name Shi and taken Lingzhou as their native place, this fact indicates that they had close ties with the Turkicised Sogdians who had moved to the Hebei area. Therefore, we can consider these Xi people as Turkicised Sogdians. Now let us discover when Shi Xiancheng moved to Hebei. According to “Old Biography of Shi Xiancheng”, his grandfather Shi Daode was once Commander Unequalled in Honor, Probationary Chamberlain for Cérémonials, Suprême Pillar of State and Prince of Huaize Commandery his father Shi Zhouluo was once aide of Tian Ji’an the Military Commissioner of Weibo, and later he had been Commander, Grand Master of Imperial Entertainment with Silver Seal and Blue Ribbon, Acting Adviser to the Heir Apparent, Palace Aide to the Censor-in-chief, Pillar of State and Prince of Beihai Commandery The “New Biography of Shi Xiancheng” does not give his grandfather and his father’s social positions. So his grandfather’s official post in the army is unclear, and as for his father, the biography simply States he had been an aide to Tian Ji’an, the fourth Military Commissioner of Weibo, and once Grand Commander under Military Commissioner of Weibo. Tian Ji’an was Military Commissioner of Weibo from the 12* year of the Zhenyuan era (AD 796) to the 7th year of Yuanhe Era (AD 812). From this information we can infer that Shi Zhouluo must hâve lived in Weibo from the Zhenyuan Era of Emperor Dezong (AD 785-804) to the Yuanhe Era of Emperor Xianzong (AD 805-820). According to the “New Biography of Shi Xiancheng”, the three générations of the Shi family were aides to varions military officiais in Weibo. If this is crédible, Shi Xiancheng’s grandfather, Shi Daode should hâve been in Weibo from the Dali Era (AD 766-779) to Jianzhong Era (AD 780-783). However the biography has no record of Shi Daode’s army post. In my opinion, it is possible that the Shi family had not moved to Hebei in Shi Daode’s time, and it is therefore likely that he was not in the army at ail. It is certain that the family of Shi Xiancheng of Lingzhou did move to Weibo, south of Hebei, during the 5

Liu Yuxi, 1989, pp. 99-103.

Moribe Yutaka

Military officers of Sogdian origin

249

Zhenyuan Era (AD 785-805)s but the route and scale of the migration are unknown. After serving as Office Manager Commander of Middle Army and Vanguard Commander in Weibo, Shi Xiancheng became Military Commissioner of Weibo because he commanded the crack troops of Weibo.

He Jintao Finally we will tum our attention to He Jintao 何 進 滔 who after Shi Xiancheng became Military Commissioner of Weibo. He Jintao was also a native of Lingzhou^ I assume that He Jintao also belonged to the Hu people from the Six Préfectures because the fàmily name He was another Han name adopted by Sogdians. He Jintao s great-grandfather He Xiaowu 何 孝物, and his grandfather He Jun 何 俊 were military officers in Lingzhou, while his father He Mo 何 默 was Commander in Front of the Yamen of Xiazhou 夏 州衙削兵馬使. He Jintao came to Weibo as a young man and became an aide to Tian Hongzheng 田弘正,the Military Commissioner of Weibo from the 7th year (AD 812) to the 15th year (AD 820) of Yuanhe era. From this we can infer that He Jintao moved to Hebei during Yuanhe era (AD 806-820), and from Lingzhou, via Xiazhou to Weibo. But it is unclear where he crossed to Shanxi during his voyage from Xiazhou to Hebei. In recent years the epitaph of He Jintao' son He Hongjing イ可弓ム敬 was discovered65 and more facts about the family and the scale of its migration hâve emerged. Below is a record of He Jintao5s migration to Hebei and related facts. 5

The Lord named (He) Hongjing and styled Zisu 子粛 was a native of Lujiang 廬江,…six générations above him, (He) Lingsi 令思, ... commanded the cavalry and defeated Syr Tardush (Xueyantuo 薛延陀) at Shibao 石 堡 fortress. He argued with general Qiao Shuwang 喬叔望 and Zhishi Sili 執 失 思 力 about their outstanding service in battle, and was falsely accused by Shuwang. Then he with 800 of his subordinates (buqu 部曲八百人) moved to the Wei 魏, Xiang 相, and Bei 貝 préfectures. ...The Grand Guardian (He Mo) s son was the Grand Preceptor (He Jintao). Lord (He Hongjing) was the Grand Preceptor's successor and his mother, née Kang 康, was the Mistress of the State of Wei 衛國 太夫 人.… The Lord (He Hongjing) married a woman from Wuwei 武威 née An 安. 5

According to the "Epitaph of He Hongjing s family”, He moved to Hebei during the Zhenguan era (AD 627-649), because it was stated that He Lingsi, the ancestor of the He family, led 800 subordinates to move to three préfectures Wei, Xiang, and Bei. But that record contradicts the "biography of He Jintao" in Jiu Tang Shu according to which He Jintao moved to Hebei during the Yuanhe's era. According to the epitaph, He Jintao and He Hongjing came to Hebei later, but their clan must hâve arrived in Weizhou much earlier. It was therefore not difficult for a member of the He family to be promoted to the rank of Military Commissioner of Weizhou. It is also possible that the migration recorded in the epitaph was fabricated as evidence to justify He5s promotion. From the epitaph we also leam that He Jintao married a woman née Kang and He 5

6

“Weibo jiedushi He Hongjing muzhiming 魏博 節度使何弘敬墓誌銘”, in Suitang wudai muzhi huibian: Hebejuan 隋唐五代墓誌 匯編 河北 卷,p 123.See also Moribe Yutaka, 1997,pp. 125-147.

Les Sogdiens en Chine

250

Hongjing’s wife was née An and came from Wuwei. We can therefore surmise that Sogdians who had the family names of He, Kang and An must hâve moved to Hebei together. The saying that He Lingsi commanded 800 subordinates, suggests the scale of He Jintao’s migration to Weibo. Moreover, He Jintao was Office Manager Commander in Yamen of Weibo. The army in Yamen was the nucléus of the govemment army which is why He Jintao was finally promoted to be Military Commissioner of Weibo.

Turkicised Sogdians and Military Governorship of Weibo

Among the Three Garrisons in Heshuo, why was Weibo the only Military Govemorship where Turkicised Sogdians became Military Commissioners, and why did ail these Sogdians corne from Lingzhou? Before examining these questions, let us first consider the différences between Weibo on one side, and Chengde and Lulong on the other side. The strongholds of Chengde and Lulong were respectively Hengzhou '£ and Youzhou. They were strongholds of the rebellions army of An Lushan and Shi Siming during the armed rébellion of AD 755. After the armed rébellion, military offïcers from northem ethnicities who had followed An Lushan and Shi Siming took over political power. It was before the end of the armed rébellion that Tian Chengsi E0 surrendered to the Tang court and became Military Commissioner of Weibo. Tian built his army by enlisting peasants. Because of the imbalance between the infantry and the cavalry, building up the cavalry became a matter of first importance. Military offïcers who were good horsemen and archers were badly needed. This may possibly be the main reason why Turkicised Sogdians migrated from Hequ to Weibo. With this in mind, let us re-examine why Shi Xiancheng become Military Commissioner of Weibo. In the 7th year of Yuanhe era (AD 812), Tian Hongzheng, was Military Commissioner of Weibo, who had been selected by the govemment army, and surrendered to the Tang gave up the “former administration of Heshuo” court. Before long Tian Hongzheng became Military Commissioner of Chengde. Tian’s departure caused chaos in Weibo. To restore order, the Tang court made Tian Military Commissioner of Weibo. By that time Shi Hongzheng ’s son Tian Bu Xiancheng had gained control of Weibo’s crack army. Tian Bu who could not cope with his administrative duties, finally committed suicide. Taking advantage of the chaotic situation, Shi Xiancheng seized power and made himself Military Commissioner. However, one should remember that the military forces commanded by Shi Xiancheng were not the govemment army of Weibo, but armed forces from other fàctions. What différence is there between such factional forces and the govemment army? As quoted above, the epitaph of He Hongjing records that (He) led 800 subordinates in the migration to the préfectures Wei, Xiang and Bei. Although the date of the migration is not reliable, the record of “commanding 800 subordinates” does represent one stage of He Jintao’s migration to the military govemorship of Weibo. That implies that Shi Xiancheng’s military forces were composed of Turkicised Sogdians from Lingzhou, who had moved to Weibo with He Jintao. In the second year of Changqing era (AD 822), Shi Xiancheng from Lingzhou was chosen Military Commissioner of Weibo. In the third year of Taihe era (AD 829), He Jintao who had

Moribe Yutaka

Military officers of Sogdian origin

251

moved to Weibo in his youth was also chosen Military Commissioner of Weibo. Thereafter, for three générations, the He family brought forth Military Commissioners. What proportion of the Weibo’s army did the “800 subordinates” occupy? The Weibo army consisted of troops stationed at the capital and troops stationed elsewhere in the région. The troops stationed at the capital numbered about 10,000 men at the beginning of Tian Chengsi’s time; in Shi Xiancheng’s time, the number was down to time it was also 8,000, and at the end of the Tang Dynasty in Luo Shaowei about §,000. On the basis of these figures, the “800 subordinates" made up 10 percent of the govemment’s army. The usual version is that after Shi Xiancheng, Military Commissioners of Weibo were drawn from the govemment’s army composed mainly of local peasants. However, in the light of newly discovered historical material, it is necessary to réexamine the accepted version that since the Changqing era, the govemment army of Weibo constantly expanded its stength and became increasingly independent.

The Situation in North China During the Late Tang Dynasty As we consider the motives for the migration of Turkicised Sogdians in Hequ to Hebei, as outlined in the four aforementioned cases, it becomes possible to divide the migration to Hebei of the Turkicised Sogdians into three phases. The first stage took place befbre and during the armed rébellion led by An Lushan and Shi Siming and this was the period during which Cao Runguo and Kang Rizhi migrated. The second period lasted from the Jianzhong to Zhenyuan eras (AD 780-805), when Shi Xiancheng’s grandfather Shi Daode, or his father Shi Zhouluo, moved to Hebei. The third period was that of the Yuanhe era during which He Jintao arrived. First, let us study the background of the period during which of Cao Runguo and Kang Rizhi migrated to Hebei. At that time, Abusi of the Eastem Tujue in Hequ was defeated by Huihu and the remnants of his tribes were incorporated into An Lushan’s army and became the strongest part of An Lushan’s forces. The background for the migration to Hebei of Shi Xiancheng’s grandfather or father, was the conquest of the Hu people of the “Six Préfectures” by the Tang army in Shizhou and the Hu’s move to Yunzhou and (in the North of present-day Shanxi Province) in the second year of the Shuozhou Zhenyuan era. It is likely that part of the Turkicised Sogdians moved to Hebei, while the Hu from the Six Préfectures moved north. He Jintao’s migration in the third period tallies with the migration of the Shatuo ethnie group, in the 4* year of Yuanhe era (AD 809), from the Yinshan Garrison in Yanzhou to Taiyuan, Yunzhou and Shuozhou. The second and third phases of the migration took place during the second half of the 8th Century, a time when the Three Garrisons of the Heshuo area had just been established. During this period, the Tang court tried to put Hebei under its direct control, a move that the Three Garrisons of the Heshuo area strongly to resisted. The Three Garrisons, especially those of Chengde and Weibo, whose territory was adjacent to that controlled by the Tang court, launched a fierce military campaign against the Tang govemment. In Weibo, because local peasants were the main components of the military forces, strengthening the cavalry became of paramount concem. To meet this demand, many Turkicised Sogdians flowed into Hebei, and especially into Weibo.

Les Sogdiens en Chine

252

However, by Yuanhe era, the Three Garrisons in the Heshuo area, gave up their "former administration^ and surrendered to the Tang court. But after the Changqing era (AD 821-825), the Three Garrisons in Heshuo again area broke loose from the control of the Tang court. However5 they formed an interdependent political and military entity. Concurrently, the tense military confrontation eased and the Three Garrisons of the Heshuo area no longer attempted to maintain and strengthen their military forces as they used to do. As a resuit, measures to incorporate military ofïicers from northem ethnie groups were no longer vigorously pursued. By the time He Hongjing5s son He Quanhao govemed Weibo, a Han life style had been adopted and a civilian officiais took the post of Military Commissioner of Weibo。Consequently, horsemen in the military forces of Chengde became less combat effective* and the influence of Chengde and Weibo in the Three Garrisons in Heshuo area diminished.

Turkicised Sogdians During the Five Dynasties1 After the Yuanhe and Changqing eras5 there are no records of Turkicised Sogdians moving from Hequ to Hebei. The first reason for this was that after the Yuanhe and Changqing eras5 the relationship between the Tang court and the Three Garrisons of the Heshuo area became one of mutual co-existence? The second reason was that the Shatuo ethnie group moved to the northem part of Hedong where they settled. Consequently from the late Tang dynasty to the period of Five Dynasties, the military power of the military administration in Hebei, with the exception of the one in Lulong, declined, as those offices came under the influence of Zhu Quanzhong 朱 全忠 and Shatuo. At the end of the Tang Dynasty, Shatuo rapidly extended its influence in northem Hedong (Daibei 代ゴ匕) and established the Later Tang, the Later Jin, the Later Han5 and the Later Zhou dynasties which give the Five Dynasties its name. Turkicised Sogdians greatly influenced Shatuo. Many military officers who bore Sogdian family names were found in Shatuo. They were mostly offspring of the Hu from the Six Préfectures, who had moved there in the second year of the Zhenyuan era (AD 786). After those Hu moved to Daibei, Shatuo also moved there. When Huang Chao 貢巢 started his peasant uprising in AD 875, Turkicised Sogdians and other nomadic ethnie groups United into a powerful community centered on Shatuo. At that time, Turkicised Sogdians tribes submitted to Shatuo underthe name of Sage 薩 葛 (Suoge 索葛, Xuege 薛葛), Anqing 9 安慶 and Jitian 雞田. Through the analysis of historical documents and epitaphs,7 we can now see that even during the Five Dynasties, nomadic tribes existed in Daibei. To meet the demands of Shatuo dynasties, such as the Later Tang and Later Jin, some tribal chieftains occasionally mobilized members of their tribe for military activities undertaken by varions régimes. In my opinion, it was these nomadic groups that sustained the military forces of the Shatuo dynasties. 7

For a detailed study of the Tukicized Sogdians in the Five Dynasties, see Moribe Yutaka, 2004. On the relationship between the Tang court and the Three Garrisons in Heshuo area in the mid and late Tang times, see Moribe Yutaka, 2002b. 9 As fbr the epitaphs of the Turkicised Sogdians in the Five Dynasties, the available ones are those of the wife of An Wanjin 安萬金,He Junzheng 何 君政,Shi Jinjun 石金俊,etc. I hâve studied the epitaphs of the former two. See Moribe Yutaka, 2002a. 8

Moribe Yutaka

Military officers of Sogdian origin

253

Bibliography

Liu Yuxi 劉禹錫

"Tang gu Bin, Ning, Qing deng zhou jiedu guancha chuzhi shi chaosan dafu jianjiao hubu shangshu jian yushi dafo ci zi jinyu dai zeng youpuye shigong shendao bei 唐故 寧慶等 州節度觀察處置使朝散大夫 檢校戸部尚書 兼 禦 ,[The monument ofLord Shi, 史大夫 賜紫金 魚袋贈 右僕射史公 神道碑” Military, Surveillance and Supervisory Commissioner of Bin, Ning, Ging Préfectures, Grand Master for Closing Court, Ministry of Revenue, Censors-in-chief, given Pouch of Purplish goldfîsh, given Right Vice-Director in Tang Dynasty] in Liu Yuxi ji jianzheng 劉禹錫 集箋證,Shanghai: Shanghai guji chubanshe. Zhenyu LUO 羅 振玉 1977 [Collection of Epitaphs in Beijing area] Jingji zhongmu printed by Luo Zhenyu under the Republic, reprinted in Shike shiliao xinbian 石刻史 新編,Taipei: Xinwenfeng, 1977. Moribe Yutaka 森 B 豊 1997 "Gihaku setsudoshi Ka Kokei boshimei shisyaku 魏 博節度使何弘敬墓誌銘 試釋”[A Study of Epitaph of He Hongjing, the Military Commissioner of Wei-Bo], in Yoshida Tora sens ei koki kinen Azia shi ronshü 吉田寅先生古稀 記念 アジア史論 集,Tokyo, pp. 125-147. 2002a "Köshin An Mankin, Ka-shi Fusai Boshimei oyobi Ka Kunsei Boshimei 後晋 安万 金 • 何 氏 夫妻墓誌銘 および 何 君政墓誌銘”[Epitaphs ofThree Sogdian-Chinese from the Hou- Jin Period in the Tenth Century], in Nairiku Ajia Gengo no Kenkyü 内陸 アジア 言語の研究 2002b "Tö Takuro shögi gun setsudoshi kö 唐澤 潞昭 義軍節度使考" [A Study of the Military Commissioner of Ze-Lu, the army of Zhaoyi in Tang Dynasty] , in Noguchi Tetsuro koki kinen ronshü kankö iinkai hen 野口鐵郎古稀記念論集 刊行委員會 ed., Chüka sekai no rekishi teki 招か⑷'中華世界の歷史的展開, Tokyo: Kyüki shoin, pp. 97-131. 2004 "Tömatsu Godai no Daihoku ni okeru Sogudo kei Tokketsu to Sada 唐 末五代 の代北に おけるソグド系突厥と沙陀”[The Sogdian Turks and Shatuo in Daibei during the Late Tang and Five Dynasties Periods], Töyöshi kenkyü 東洋 史研究 624 pp. 60-93. Onogawa Hidemi 小野丿"秀美 1942 "Kakyoku rokkoshü no enkaku 河曲六胡州の沿革；'[The History of Six Préfectures of Hu in the northward bend of the Yellow River] Tôa jinbun gah/ う東亜 人文學報 14 193-226 Pulleyblank, E. G. 1952 "A Sogdian Colony in Inner Mongolia,” T'oung Pao 41: 317-356 QルNZ4ÆG 性N80}7 全 唐文補遺 Quan Tang wen buyi 全 唐文補遺 [The Filling Collection of Compositions 1998 during Tang Period], volume V, Xi'an: Sanqin chubanshe.

1989

254

Les Sogdiens en Chine

Shen Tao 沈 濤 ed. 1842 Changshan zhenshi zhi 常山貞石志 [Collection of Stone Monuments in Hebei area] reprinted in Shike shiliao xinbian 石亥リ史料 新編,Taipei: Xinwenfeng, 1977. Sui Tang 敗 4陵/W h切取4N 隋唐五代墓誌匯編 1991a Sm nwg wなdMがz/z初ル〃 M:Sレ劭x〃カ前 隋唐五代墓誌匯編 陝西卷 [Collection of Sui Fang and Five Dynasties Epitaphs. Shaanxi Volume] volume IV, Tianjin: Tianjin guji chubanshe. 1991b uwhi な2加 加/加a%: 次加レ 隋唐五代墓誌匯編 。河北卷 [Collection of Sui Fang and Five Dynasties Epitaphs. Hebei Volume] volume IV, Tianjin: Tianjin guji chubanshe.

•

Cäkar sogdiens en Chine

Étienne de la Vaissière

La diversité sociale des Sogdiens émigrés en Chine n?est pas douteuse. Les textes chinois, notamment ceux de Turfan3 montrent que les Sogdiens exercent en Chine toute sorte de métiers : marchands bien sûr, mais aussi paysans, artisans, fonctionnaires moines, soldats. Cfest sur cette dernière catégorie que je souhaiterais insister. En effet, la Sogdiane possède une institution militaire d!un type tout à fait particulier, que les sources chinoises sont les premières à révéler, les câkar. Ceux-ci forment un trait carac¬ téristique de la société sogdienne qui demande à être étudié aussi bien au regard des sources chinoises ou arabes que des pratiques militaires nomades. Mais ils sont éga¬ lement mentionnés en Chine même. . Le terme sogdien de câkar est connu des sources chinoises sous des transcriptions diverses, notamment Zhijie 柘羯 et Zhejie 赭芋曷, prononcés en Early Middle Chinese teiajk kiat et teia'ldat1. Chavannes avait déjà rassemblé un certain nombre de mentions, que je rappelle tout d'abord. Dès 629, Xuanzang 玄奘 les mentionne sous la forme Zhejie 赭羯： 3

[The king] had a splendid army the most of his soldiers being Zhejie men. These were men of ardent valour, who looked on death as a going back to their kindred, and against whom no foe could stand. 兵馬强盛, 多 諸赭羯 。赭羯之人,其性 勇烈, 視死如 歸,戰無前敵 。2

U Ancienne Histoire des Tang les signale dans sa description de Boukhara: Le royaume de An est appelé aussi Buhe (Boukhara) [...] on y enrôle les hommes braves et robustes pour former (le corps des) Zhejie ; l'expression Zhejie est Féquivalent de l'expression chinoise 'combattant'. 募勇健者為柘羯 • 柘羯,猶 中國言戰士也 。3

Par ailleurs, l'encyclopédie Cefu yuangui 冊丿付 兀龜,rédigée en 1013, qui donne des éléments des annales Tang, mentionne à deux reprises les câkar, sous l'année 740 : « Ce même mois, on ajouta aux titres de Sijinti 斯 謹提,roi des Zhejie, celui de 'spécialement promu*.4 » Sijinti est le roi de la ville sogdienne de Kesh. Puis elle signale, pour l'année 1

Les transcriptions en Early Middle Chinese sont faites d'après Pulleyblank, 1991. Xuanzang, trad. Watters, p. 94. 3 Jiu Tang shu, chap. 104, p. 3209. Trad. Chavannes, 1903, p. 137. 4 Chavannes, 1903, p. 63.

2

256

Les Sogdiens en Chine

759 : « le douzième mois [...] on offrit un banquet aux étrangers Zhejie (fan hu Zhejië) dans trois salles et on donna à chacun d'eux trente pièces de soie.5 » Les textes arabes et persans mentionnent également le terme sogdien rendu en chi¬ nois par Zhejie. Tabarï est la source principale ; le terme apparaît pour la première fois sous la forme Sâkiriyyah sous l'année 701-702 (AH 82) : les frères Hurayt et Tâbit b. Qutbah, deux grands nobles khorassaniens, se réfugient à Termez avec trois cents de leur Sâkiriyyah et de leurs suivants arabes et cette Sâkiriyyah un peu plus loin garde la tente de son maître. Le roi de Samarcande, Tarhün en dispose également (AH 85) : ses Sâkir s'occupent de sa tente, mais fuient à l'approche du danger. On voit un Khorassanien de Tûs partir à la chasse accompagné des membres de sa Sâkiriyyah (AH 98). Un Samarcandais et un Arabe de Bouldiara, puis des guerriers et nobles du Tokharestan, combattent avec les leurs. Un Sâkir prête à un Arabe de Sogdiane sa corne à boire tibé= taine. En 722 (AH 104) le neveu du chef des révoltés sogdiens, capturé par les troupes arabes à l'entrée du Ferghana, en dispose et tente en un combat desespéré de libérer son oncle avec leur aide. En 730-1 (AH 1 12) le roi de Samarcande se présente comme le Sâkir du KJaaqan des Turcs (« I am a slave of the Hâqân from his Sâkiriyyah » ). En 738-9 (AH 121), un envoyé du gouverneur du Khorassan Nasr b. Sayyâr auprès du roi du Ferghana déclare au roi : « I am a Sâkir and the amïr's deputy scribe ». Enfin en 750-1 (AH 133) le roi du Khuttal s'enfuit au Ferghana face à l'avancée des troupes arabes accompagné de ses dihqâns et de sa Sâkiriyyah6. Le terme est également utilisé pour l'Asie centrale pré-islamique dans VHistoire de Boukhara de Narsahî, texte arabe du dixième siècle dont on ne possède qu'une version persane postérieure, à propos de l'assassinat du souverain de Boukhara : Bukhâr Khudâh made a will and died in an hour. His servants (Câkirân) entered and removed his flesh and brought his bones to Bukhara.7

Il faut souligner enfin que le terme câkar est pour l'heure inconnu des textes sog¬ diens, qui sont, il est vrai, essentiellement religieux. Il a dû y exister cependant, puisqu'il a été transcrit du sogdien vers le chinois et l'arabe. Je dois à Ilya Yakubovich une suggestion d'étymologie, un dérivé de *c'5r-k'r "subalterne" (c'8r : sous, k'r: suffixe des groupes sociaux, on possède d'autres exemples d'abbréviation de c'ôr en c'), qui à mon sens doit rester hypothétique car le mot a pu être emprunté. Il se dégage de ces données un portrait fonctionnel et culturel assez cohérent : les câkar sont les soldats personnels des nobles et des souverains iranophones de la Sog¬ diane pré-islamique et, par propagation du phénomène, du Tokharestan et de l'Est du Khorassan. Ils sont bien une caractéristique de la Transoxiane et des régions limitro¬ phes. Le mot n'apparaît chez Tabarï que lorsque son récit atteint l'Amou Daria, constat également valable pour les sources chinoises. Les câkar assurent aussi bien la garde rapprochée que le service de leur maître et ne sont ni esclaves, ni nobles, quoique des 5

Chavannes, 1903, p. 95. Tabari, II, 1082, 1055, 1159-60, 1331, 1528, 1604, 1609, 1631, 1445, 1542, 1695, III, 74. Analyse plus détaillée des sources arabo-persanes dans de la Vaissière, 2005. 7 Narsahî, p. 85. Trad. Frye, p. 62. 6

Étienne de la Vaissière

__

Câkar sogdiens de Chine

257

grands nobles puissent métaphoriquement se présenter comme les câkar de leur suze¬ rain politique. Enfin les câkar sont nombreux : c'est ce qui les distingue des gardes rap¬ prochées (comitatus) fréquente dans le monde de la steppe comme chez les Germains, les Turcs ou chez les Hephtalites. Je voudrais insister sur ce point. Il faut soigneusement distinguer les câkar du modèle plus général des gardes rapprochées. On en possède une très bonne description pour l’Asie centrale hephtalite chez l’historien byzantin Procope : The wealthy citizens are in the habit of attachiiig to themselves friends to the number of twenty or more, as the case may be, and these become permanently their banquetcompanions, and hâve a share in ail their property, enjoying sortie kind of a common right in this matter. Th en, when the man who has gathered such a company together cornes to die, it is the custom that ail these men be borne alive into the tomb with him.8

Les Hephtalites furent les suzerains de la Sogdiane durant la première moitié du VIe siècle, et on pourrait envisager que cette institution sociale soit à l'origine du phéno¬ mène sogdien des câkar. Mais, s’il en est ainsi, ceux-ci ont largement divergés de leur modèle. Il ne s’agit plus de quelques happy few partageant tout avec leur maître, mais d’un corps de soldats d’élite. Les câkar vont avoir un grand avenir dans le monde musulman, et forment une partie de l’armée abbasside et l’entourage des plus hauts dignitaires dans l’Iraq du IXe siècle, à la suite de l'arrivée successive là de plusieurs vagues de soldats en provenance du Khorassan, que ce soit lors de la révolution abbasside en 750 ou lors de la prise du pouvoir par le caliphe al-Ma'mün en 813, ouvrant la voie au phénomène postérieur des mamluks9. Mais en Chine trois mentions intéressantes avaient échappé à Chavannes. La première se trouve dans le Tongdian et est contemporaine du témoignage de Xuanzang. Comme souvent, le Tongdian conserve un texte plus développé que les Histoires des Tang et signale qu’en 630, après la capture du kaghan des Turcs orientaux ^IJ, de nombreuses tribus se rendirent à la Chine : Xieli The chiefs and leaders who arrived to the court were ail appointée! générais [...] only the. Zhejie did not corne. It was decreed that envoys should be sent to persuade them to corne. » [-] ÎS fit

Le texte est suivi par un mémoire du préfet de Liangzhou sur le sujet10. D’où viennent les câkar ayant refusé de se rendre après la défaite de Xieli en 630 ? Il y a en fait deux façons de comprendre l’épisode. La première et la plus immédiate conduit à envisager l’existence de câkar dans l’entourage des khagans. Mais cette mention serait alors totalement isolée : il faut attendre près d’un demi-millénaire plus tard, pour que au XIIe siècle Marwazï mentionne de nouveau des câkar en contexte turc, chez le kaghan ouïghour11. Or les sources chinoises des VI-VIIe siècles décrivent assez précisément les 8

Procope, I, iii, 6-7. Trad. Dewing vol. I, p. 15. De la Vaissière, 2005. 10 Pulleyblank, 1952, p. 347-8.

9

11

Minorsky, 1942.

258

一̶

Les Sogdiens en Chine

cours turques et nous parlent d5un système de garde personnelle du souverain. Cependant cette dernière appartient à la vaste catégorie de groupes de comitatus, des germaniques. On en possède même le nom : bôri, les loups. Autrement dit les n5avaient nul besoin de câkar, étant engagés dans un système très turcs khagans différent. Mais il se pourrait que dans ce cadre Xieli fasse exception. On sait en effet par les sources chinoises que l'emprise sogdienne fut sous ce kaghan à son comble : Xieli confiait tout aux divers Hu et tenait son propre peuple à distance. Les hu sont avares et présomptueux et par nature peu sûrs et changeants. Les lois furent multipliées et F armée mise en branle chaque année. Les gens du peuple (les Turcs) n'aimaient pas cela et les tribus

désertaient?2 On peut effectivement envisager que parmi les reproches que les Turcs ont pu feire au kaghan, remploi de càkar ait pesé lourd. Pulleyblank a proposé une interprétation très différente de cet épisode et tenté de montrer que tout Fépisode était étroitement lié à la ville de Yiwu 伊吾 (Kami), soumise simultanément en 630 : le mémoire du préfet de Liangzhou qui suit immédiatement dans le texte du Tongdian souligne en effet que le coût de la conquête de cette ville serait supérieur aux bénéfices qu'on pourrait en attendre. Or on sait, par le Tongdian notamment, qu'à Yiwu: À partir des Sui, des marchands sogdiens s'y sont établis, mêlés (à la population). S'y trouvent un millier d excellents soldats, ils sont rattachés aux Tôlôs, ce sont des hommes très fiers et courageux. 至 隋 , 有 商 胡 雜 居 ,勝兵 千 餘 人 , 附 於 鐵 勒 , 人 甚 驕 悍. 5

Les câkar auraient pu être ces Sogdiens contrôlant Yiwu pour le kaghan turc, et effectivement, si « excellents soldats 月券兵 » est un topos appliqué a tous les pays d'Occident, il n'est pas impossible qu'il s'agisse ici des câkar mentionnés plus loin dans le même texte, le texte insistant sur le rattachement aux Turcs et sur la férocité, traits par¬ tagés avec les câkar. La reddition de Yiwu et de ses sept circonscriptions est d'autre part mise en œuvre en 630 par un personnage au nom sogdien, Shi Wannian 石萬年. Signalons au passage que Pulleyblank se trompe certainement lorsqu'il tente de relier par un raisonnement très élaboré cet épisode aux sept villes nées de la colonisation sog¬ dienne du Lob Nor : il souligne au départ de son raisonnement que jamais Yiwu n'est partagé en sept circonscriptions. Or ce point de départ est faux : la Géographie de Shazhou et Yizhou mentionne explicitement la présence à Yiwu de sept xiang , cantons, et c'est d'eux dont il s'agit ici, ici qualifiées de villes, là de tribus ou encore de cantons. Quoiqu'il en soit, à Yiwu ou au cœur du pouvoir kaghanal, l'épisode doit se lire dans le cadre de l'exceptionnelle influence sogdienne sous Xieli et ne correspond pas aux pratiques habituelles des Turcs. Les deux autres mentions de câkar les placent en Chine intérieure. Dans la Nouvelle Histoire des Tang, à propos d'un épisode de l'avance en 755-6 du général rebelle An 12

Jiu Tangshu, 194Ap. 5159.

H； Étienne de la Vaissière

.

Câkar sogdiens de Chine

259

Lushan 安ネ山,de père sogdien et de mère turque contre les armées des Tang, il est écrit du général impérial Feng Changqing : シ

Changqing envoya des cavaliers intrépides pour les repousser ; ceux-ci tuèrent des Zhijie au nombre de mille. 13 常清 使驍騎拒之 , 殺拓羯 數十百人 。*

Un an plus tard, alors que Femprise des rebelles s?est étendue au sud du Fleuve jaune, un groupe de leurs armées a mis le siège devant la ville de Suiyang 睢陽, située qui mène au riche bas-Yangtzé. Elle est le long du canal stratégique Bian Xun 弓長 巡 qui soutient un siège total de Zhang loyaliste énergiquement défendue par le quatre mois et se rend célèbre notamment pour avoir ordonné à ses troupes de recourir au cannibalisme pour tenir? Lors cTun épisode du siège : Il y eut un grand chef, couvert d'armure, qui amena des Zhijie 拓羯, des cavaliers au nombre de mille avec drapeaux et bannières chevaucher devant les murs et provoquer Xun.15

Ces mentions sont importantes : j'ai souligné dans mon Histoire des marchands sogdiens combien la révolte d'An Lushan pouvait être comprise, entre autres, comme une révolte hu. Mais avec le texte de VHistoire d'An Lushan le décrivant comme un chef de la communauté sogdienne utilisant les marchands comme agents, et la décou¬ verte assez récente des jades de cérémonies de Shi Siming 史思明 portant la mention de Zhaowu Huangdi 昭武皇帝, c'est à dire Jamuk-Empereur, associant le titre nobi¬ liaire sogdien au titre impérial chinois16, ces deux mentions de câkar sont un nouvel argument majeur permettant de souligner le caractère spécifiquement sogdien du soulè¬ vement : alors que le texte de VHistoire d'An Lushan. rédigé a posteriori par les vain¬ queurs, pouvait être mis en doute, l'inscription Zhaowu Huangdi et ces mentions de câkar, provenant des rebelles eux-mêmes, doivent être acceptées. An Lushan comme tout chef militaire sogdien, s'est constitué une armée de câkar. Des subordonnés en avaient également, ainsi à Suiyang. Sa conduite ne diffère en rien de celle des nobles sogdiens luttant contre les Arabes avec leurs câkar à la même époque. On peut d'ailleurs supposer que les câkar mentionnés par le Cefu yuangui sont probablement des anciens câkar d'An Lushan ralliés aux Tang : à cette date, en 759, le fils d'An Lushan, An Qingxu 安慶 緒,vient d'être tué par Shi Siming et le Cefu yuangui rapporte sans doute la reddition de quelques centaines de câkar, une partie de la garde familiale ayant refusé de suivre Shi Siming. On ne possède pas d'autres traces de câkar sogdiens ayant exercé en Chine : il est vrai que les trajectoires permettant à des émigrés d5accéder directement, dès la première génération, à des postes de responsabilité, et donc de devenir historiquement visibles, Xin Tangshu, chap. 117, p. 4581. Sur ies aspects sogdiens de la rébellion d'An Lushan voir de la Vaissière, 2002, p. 213-218. Voir également de la Vaissière, Trombert, 2004, p. 961-3. l’Gra建 1995. Xin Tangshu, chap. 117, p. 5537. 16 Voir Yoshida, 2004. Smimova, 1970, p. 24-38, avait déjà fait le lien entre les termes de Zhaowu et de Jamük, titre des nobles sogdiens dans VHistoire de Boukhara.

260

_

.

Les Sogdiens en Chine

sont rares et An Lushan, une exception. Il ne serait pas inenvisageable pourtant d5en trouver d’autres au Vïe ou au VIIe siècle, puisque Fon connaît des nobles sogdiens s'étant près du Lob Nor. Mais, installés sur les marges chinoises, tels Kang Yandian pratique nobiliaire, le phénomène câkar était tout entier dépendant d’une immigration de nobles sogdiens, immigration qui semble avoir été, contrairement à ce qui se passe durant la première moitié du IXe siècle entre la Sogdiane et l’Iraq abbasside, très limitée. Les sabao nouveaux riches qui se firent élever de si spectaculaires tombeaux n’avaient probablement pas droit à des câkar.

Bibliographie Chavannes, É. 1 903 Documents sur les Tou-kiue (T lires) occidentaux. Saint Pétersbourg, réimp『. A. Maisonneuve, Paris, 1973, 378 p. + 1 10 p. (notes additionnelles). GRAFF, A.D. « Meritorious Cannibal: Chang hsün's Defense of Sui-Yang and the Exaltation 1995 of Loyalty in an Age of Rebellion », Asia Major, VIII-1, p. 1-17. dela Vais sière, É., 2002/4 Histoire des marchands sogdiens, Paris, Mémoires de l’IHEC vol. 32. « Châkars d’Asie centrale. À propos d’ouvrages récents », Studia Iranica, 2005

34-1, p. 139-149. É., « Des Chinois et des Hu. Migrations et intégration des Iraniens orientaux en 2004 milieu chinois durant le Haut Moyen-Âge », Annales. Histoire, Sciences Sociales, p. 931-69. Minorsky, V. (trad.)5 Sharaf al-Zamân Tâhir Marvazï On China, the Turks and India Arabie Text 1942 (circa A.D. 1120) with an English translation and commentary, (James G, Forlong Fund, XXII), Londres : The Royal Asiatic Society. Narshakhï M. : Frye, R.N., (trad.) 1954 The History of Bukhara, Cambridge, 178 p. Procope : Dewing H.B. (éd., trad.) 1914 The History of »e Wars, I, Cambridge (Mass.) : Harvard University Press. PULLEYBLANK, EG 1952 « A Sogdian Colony in Inner Mongolia », T’oung Pao, XLI, p. 317-356. PULLEYBLANK, E.G. 1991 Lexicon of reconstructed pronunciation in Early Middle Chinese, Late Middle Chinese, and Early Mandarin, Vancouver : UBC Press. Smirnova, O. Ocerki iz istorii Sogda [Études d’histoire sogdienne], Moscou, 1970. 1970 Xuanzang : Watters, T. (trad.) On Yuan Chwang’s Travels in India 629-645, Londres. 1905 Yoshida, Y. and Related « On the Origin of the Sogdian Surname Zhaowu 0g 2004 Problems », Journal Asiatique, n° 291, p. 35-67. DE LA VAISSIÈRE, É., TROMBERT,

Un vestige vivant de la présence sogdienne en Chine du Nord : le vignoble du Shanxi

Éric Trombert

La vigne (vitis viniferd) est connue en Chine depuis plus de deux mille ans. Et

pourtant, à aucun moment elle n’a véritablement gagné les faveurs des agriculteurs. Jamais, en Chine intérieure, elle ne fat cultivée autrement que comme plante exotique (dans le parc impérial de Han Wudi ou dans celui de Tang Taizong)1, sauf dans une seule région, la province du Shanxi. Cette exception du Shanxi est d’autant plus remarquable qu’il s’agit d’une authentique tradition viticole - en termes d’ancienneté, de notoriété et de durée (voir fig. 1 et 2, planche couleur 20). Son intérêt dépasse largement le cadre de l’histoire de l’agriculture. On doit considérer en effet que la vigne n’est pas une plante ordinaire. N’étant pas une culture vivrière, son adoption ne reflète aucune nécessité alimentaire. C’est une culture « superflue » pour toute population vivant à la limite du seuil de subsistance, mais elle requiert néanmoins des connaissances techniques très spéciales qui sont nécessairement le fruit d’une longue expérience. Par conséquent, son implantation dans un nouveau terroir suppose à la fois l’intervention de cultivateurs qualifiés venus d’ailleurs, et l’existence d’une demande locale qui n’est pas d’ordre alimentaire mais culturel - l’attrait pour un produit de prestige. En d’autres termes, la création d’un vignoble est moins une affaire d’agriculture que de culture, et son essor n’est pas à rechercher dans l’étude des sols et du climat, mais dans l’histoire des hommes, de leurs mœurs, de leurs croyances et de leurs migrations. L’historiographie chinoise indique que la vigne fut introduite en Chine intérieure à trois reprises, chaque fois sans grand succès : sous le règne de Han Wudi (141-87 av. n. è.), après la reconquête de l’Ouest au début des Tang (vers 640), et à la fin du XIXe

Actuellement, on assiste à une siècle par un Chinois d’outre-mer, Zhang Bishi nouvelle « poussée » de la viticulture - notamment au Shanxi - qui s’inscrit dans le grand mouvement de modernisation et d’ouverture lancé en République populaire à partir des années 1980. On constate ainsi que chacune des avancées de la vigne correspond à un moment historique similaire marqué par une ouverture politique et culturelle en direction du monde occidental. Cette constante amène à s’interroger sur l’origine inexpliquée du vignoble du Shanxi en posant l’hypothèse selon laquelle la vigne aurait été apportée dans cette province chinoise Han par des populations venues de l’Ouest. Mais avant de développer cette hypothèse, il faut exposer ce que l’on sait sur l’histoire de ce vignoble en remontant aussi loin que les sources le permettent. 1

m

Shÿi,j. 123, p. 3113-3114 ; Tang huiyao,j. 100, p. 1796-1797.

262

一…

Les Sogdiens en Chine

Le vignoble du Shanxi

Marco Polo, décrit ainsi la région de Taianfu (Taiyuan fu 太原 府), la capitale du Shanxi actuel : Y a maintes belles vignes dont ils ont du vin en grande abondance. En toute la province du Catai, c'est ici seulement qu'on produit du vin : aussi, de cette ville, il en part pour toute la province. . . (Livre II, p. 154)

Cette observation du voyageur vénitien est corroborée par des sources chinoises de même époque. Dans son traité de diététique rédigé pendant Père Y anyou (1314-1320) des Yuan, Hu Sihui dénombre trois régions viticoleシ Celle de Turfan (appelée alors Qara khojas Halahuo ［zhuo］ 哈拉火［卓］), la meilleure ; celle du Xi Fan 西 番 (le Sud-Est du Gansu et F Est du Qinghai) ; et celle de Pingyang 平陽 et Taiyuan (la moyenne vallée de la fen 汾冽)を Un autre ouvrage bien documenté sur la fabrication du vin en Chine sous la dynastie mongole indique une seule région de production : le circuit de Jining 冀寧路 (capitale Taiyuan) et les monts Taihang 太行山 tout proches23. Il est certain que les Mongols avaient appris à apprécier le vin au cours de leurs conquêtes de F Asie intérieure, et qu'ils en répandirent Fusage en Chine sous les Yuan4. Mais pour ce qui est précisément du vignoble du Shanxi, on aurait tort de leur en attribuer la paternité. En effet, le vignoble du Shanxi est attesté sous les Song5 et même avant. C'est ainsi que les trois principaux traités de géographie administrative de la dynastie des Tang s'accordent pour indiquer qu'en dehors de Gaochang 咼昌 (actuel Turfan), une seule région de l'empire fournissait du raisin et du vin comme tribut annuel à la cour impériale, la circonscription de Taiyuan (appelée Taiyuan fu ou Taiyuan jun 君 selon l'époque)6. Cette région administrative comptait alors treize districts couvrant le centre du Shanxi, depuis les environs de Taiyuan jusqu5 aux districts de Wenshui 乂水 et de Qi ネロ au sud, jusqu'aux contreforts des monts Taihang à l'est. D'après une autre source, le tribut annuel de vin de raisin en provenance de la province du Hedong 河東 (le Shanxi) continua de parvenir à la capitale jusqu'en 837, date à laquelle il fut aboli par décret impérial7. Il va de soi que les produits constituant le tribut annuel étaient choisis parmi les spécialités locales de chaque région8.

。

Yinshan zhengyao力.3, p. 250. Caomu zi (j. 3 B, p. 68) composé vers 1378 par Ye Ziqi. 4 À propos de ce que les historiens actuels nomment la « seconde introduction » du vin en Chine, cf. Zhu Baoyong, 2000, p. 97. 5 Bencao tujing,). 16, p. 530-531. 6 Yuanhe junxian tuzhi,］. 13, p. 362 ; Tongdian, j. 6, p. 60-61 ; Xin Tang shu^y 39, p. 1003. Ces listes de tributs valent au moins pour une période allant de F ère Kaiyuan (713-741) à Fère Yuanhe (806-820). La liste des produits varie légèrement d'un ouvrage à P autre : raisin pour les deux premiers, vin pour le troisième. 7 Cefu yuangui,}. 168, p. 2028 a. 8 Selon le règlement, le tribut annuel de chaque commanderie devait être exclusivement des produits du terroir, pour une valeur totale qui ne devait pas excéder 50 coupons de soie standard ; Tongdian,}. 6, p. 59. 2

3

Éric Trombert

_

Ze vignoble du Shanxi

263

Ainsi, la tradition viticole du Shanxi signalée par Marco Polo était déjà une réalité bien établie au début des Tang. Malheureusement, son origine nous est inconnue, aucun texte antérieur à ceux mentionnés ci-dessus n’y faisant la moindre allusion. Parmi les hypothèses envisageables quant à la diffusion de la viticulture en Chine orientale, la piste la plus sérieuse est celle des Iraniens. Elle s’appuie sur une double constatation : d’une part, l’importance de la vigne dans le monde iranien. D’autre part, une forte présence sogdienne au Shanxi, et notamment dans la région de Taiyuan, présence désormais très bien documentée9. En somme, la viticulture n’aurait-elle pas suivi les mêmes voies que la coutume du déchamement des corps pratiquée à Taiyuan jusqu’au début du VIIIe siècle10 ? On explorera d’abord cette piste qui donne aux Iraniens, et plus précisément aux Sogdiens, le premier rôle. Puis, on tentera de l’enrichir en la couplant avec l’action menée par les Wei du Nord dans le Shanxi. Enfin, on affinera l’hypothèse de départ en évaluant le rôle qu’ont pu jouer, aux côtés des Sogdiens, certains peuples sans être iraniens, essentiellement de la steppe qualifiés eux aussi de « hu » nomades, mais fortement influencés par le monde iranien.

La piste sogdienne Il y a déjà un demi-siècle, Chen Yinque avait décelé, dans un conte d’époque Tang (au sud de Taiyuan), un lien entre une forte dont l’action se situe à Puzhou alcooliques et la présence de communautés de boissons fabrication locale de tradition sogdiennes11. D’autres indices, relatifs à cette même région et datant du tout début de la dynastie des Tang ou même antérieurs, permettent d’associer la production de vin et 12. D’une façon certains milieux hu, tavemiers-fabricants de boissons (jiujia hu goût pour le vin de prononcé un conjugué toujours ont Tang poètes des les générale, sogdiennes13. Mais bien sûr, ce raisin avec une véritable fascination pour les danseuses que suggère la littérature de fiction demande à être corroborés par des données

historiques. 9

Voir par exemple, Rong Xinjiang, 2001a. Jiu Tangshu, j. 112, p. 3335. Pour un rapprochement entre cette coutume et une pratique similaire au pays de Kang cf. Zhang Guangda, 1994, p. 204-209. 11 écrite par Yuan Zhen KÄ vers l’an 804 II s’agit de V Histoire de Yingying (Yingying zhuan vraisemblablement à partir de faits réels. Cette histoire d’amour inspira maintes œuvres littéraires dont la de son vrai nom L’héroïne, Cui Yingying célèbre pièce de théâtre des Yuan, ïeXixiangji que les poètes des Tang ont si souvent Cao Jiujiu WAÄ, serait l’une de ces courtisanes hu (huji fréquentées et chantées dans les tavernes tenues par des Iraniens. La piste ouverte par Chen Yinque dans les années 1950 (1978, p. 106-116) a été récemment reprise par Ge Chengyong, 2002. 12 (585-644), natif du sud Shanxi, cité par Ge Chengyong (2002, p. 67). On Voir le poème de Wang Ji de Liu Yuxi pourrait citer également d’autres poèmes, comme le Chant de la vigne (Putao ge (772-842) qui contient notamment ces vers : « Nous autres, hommes de J in # (la région de Taiyuan), travaillons (la vigne) comme jade précieux. Nous en tirons un vin si délectable que quiconque en boit ne peut s’en rassasier. » Quan Tang shi,]. 354, p. 3963. 13 Le même Liu Yuxi cité à la note précédente semble avoir été fasciné par les danseuses du Câc, comme en témoignent ses deux poèmes intitulés « En regardant la danse zhezhi » (Guan zhezhi wu er shou ^22^) ; Quan Tang shi,j. 354, p. 3972. 10

Les Sogdiens en Chine

264

La vigne e/ le vin, don du monde iranien

La culture de la vigne a son origine dans le monde iranien. La viticulture méditerranéenne, et notamment grecque, n’en fut que l’extension la plus spectaculaire. Cela est amplement démontré par l’archéologie et par les textes. L’iconographie en fournit aussi des preuves manifestes. Un seul exemple sera donné ici : cette carafe en argent d’époque sassanide qui représente les vendanges de façon très réaliste (voir fig- I)'4Le thème de la vigne et du vin est traité dans presque toutes les tombes sogdiennes découvertes en Chine. La vigne apparaît tantôt comme simple motif de décoration (des déroulés de sarments ornent la partie supérieure de plusieurs panneaux ; voir fig. 2), tantôt comme élément de décor : c’est sous une treille chargée de grappes que se déroulent plusieurs banquets représentés sur les reliefs des tombes d’Anyang (voir fig. 3), de Taiyuan (Yu Hong Ces scènes ne prouvent et de Xi’an (An Qie nullement la persistance de certaines pratiques ancestrales que ces immigrés auraient maintenues en terre chinoise. Elles renvoient plutôt aux lointains horizons de l’Asie centrale, tout comme les scènes de chasse au lion, nombreuses elles aussi dans les mêmes tombes et dont on a un exemple sur un panneau où figure également une scène dans les vignes (tombe d’An Qie, voir fig. 4) - on sait bien que le lion est un animal absent de Chine, mais présent dans le monde iranien1415. Mais elles témoignent de la place privilégiée que tenait la viticulture dans le mode de vie traditionnel des Sogdiens. On sait également que c’est par le même monde iranien que les Chinois connurent le vin et la vigne, à l’époque des expéditions de Han Wudi au-delà des Pamirs. Au récit bien connu de Zhang Qian sur l’usage du vin en Perse et au Ferghana vers la fin IIe du siècle avant notre ère, il faut ajouter les noms des autres pays viticoles cités dans V Histoire des Han antérieurs : Jibin BW, Nandou et Qiemo #5^. Leur localisation trace une voie de pénétration de la viticulture qui serait méridionale, indo-bactrienne, et qui déboucherait dans le sud du Xinjiang16. Notons encore que le

14

Conservée au British Muséum; repr. in Ghirshman, 1953, p. 57. Le fait que ces scènes sont peu indicatives quant au mode de vie des défunts dans leur lieu d’adoption est évident dans deux scènes des tombes de Tianshui et de Taiyuan (voir fig. 5 et 6) : dans les deux cas, l’artiste chinois n’a pas compris le sujet qu’il avait à reproduire, le foulage du raisin, un travail qui lui était totalement inconnu. Dans la tombe de Taiyuan, le fouloir est devenu une terrasse entourée d’une balustrade ouvragée, et les gargouilles par lesquelles le moût doit s’écouler ont été remplacées par des médaillons obturés. Sur cette terrasse, trois petits « danseurs » sont représentés à la place des fouleurs. Les porteurs de jarres figurés au pied de l’édifice nous rappellent toutefois le sens originel de cette scène. Ainsi, à en juger par ces reliefs, la fabrication du vin était peut-être, dans l’entourage de Yu Hong à la fin du VIe siècle, une activité aussi exotique que la chasse au lion pouvait l’être à Chang’an à l’époque d’An Qie. Le foulage du raisin est un thème d’inspiration dionysiaque fréquent dans l’art funéraire romain. Il est au centre du programme iconographique du mausolée de Sainte-Constance (Rome, IVe siècle), présent non seulement aux quatre orients sur la mosaïque qui orne la voûte de la galerie circulaire (noté par Marsh ak, 2001, p. 261), mais également sur le cénotaphe de la sainte (voir fig. 3 et 4, planche 20). 16 Cf. Trombert,2001,p. 301-304. 15

Éric Trombert

Fig. 1 : Carafe en argent sassanide. Londres, British Muséum. D’après Ghirshman, 1953, p. 57.

Fig. 3 : Lit mortuaire, Anyang. Museum of Fine Arts, Boston. D ’ après Osvald Sirén, La sculpture chinoise du au XIVe siècle, Paris/Bruxelles, G. Van Oest, 1925/1926, t. III5 pL 449.

vignoble du Shanxi

Fig. 2 : Lit mortuaire de Yu Hong. Taiyuan. D’après Wenwu, 2001, n° l,p. 37.

Figo 4 : Lit mortuaire d’Anqie. Xi’an. D’après Rong Xinjiang, 2001b, p. 127.

265

266

Les Sogdiens en Chine

Fig- 5 : Lit mortuaire.Tianshui. D’après Kaogu, 1992, n° 1, p. 48.

: Mahesvara. Yungang, grotte n° 8. D’après Yungang shiku, Pékin, Wenwu chubanshe, 1977, pi. 41.

Fig. 6 : Lit mortuaire de Yu Hong. Taiyuan. D’après Wenmi, 2001, n° 1, p. 35.

宜rie Trombert

一

Le vignoble du Shanxi

267

三:芝:W:三;ニ

seul pays d'Occident mentionné comme producteur de vin dans V Histoire des Han postérieurs (25-220) est un pays tributaire du Kangju 康居, le Suyi 粟弋,c'est-à-dire la Sogdiane17. Il n'y a donc rien d étonnant de constater que les plus anciennes traces écrites de la culture de la vigne sur le sol de Chine, datées du IIe siècle de notre ère3 5

concernent les marches occidentales de l'Empire : l'oasis de Yiwu 伊吾 (Hami)18 et la

province du Liangzhou 冰 州 (Gansu actuel)19. Les Sogdiens de Chine:des laboureurs et des vignerons

Que les Sogdiens5 les Bactriens ou d autres Iraniens aient popularisé la vigne en Chine par le commerce du vin et des raisins secs est chose très probable. Mais qii&n est-il de leur rôle dans la création de vignobles et dans la transmission des techniques viticoles ? Une telle action suppose en effet F implantation de migrants engagés dans l'agriculture. Précisément, il est désormais avéré que l'immigration sogdienne fut beaucoup plus orientée vers l'agriculture qu'on ne Fa cru pendant longtemps. L'image si souvent reproduite du marchand sogdien itinérant a longtemps occulté cette réalité jusqu'à ce que l'étude méthodique des communautés sogdiennes établies à Dunhuang et à Turfan apporte une autre vision. Les manuscrits retrouvés dans ces deux régions, registres de population et autres documents de la pratique datant des Ve-VIIIe siècles, montrent en effet qu'une forte proportion de la population sogdienne locale était alors des paysans. Plus surprenant encore, on a pu démontrer que l'orientation agricole concernait un grand nombre de ces immigrants dès la première génération20. Les ouvrages historiques sont moins propices que les documents de la pratique à cerner cette réalité. Ils offrent cependant un exemple d implantation sogdienne à vocation agricole et même viticole. C'est celui de Kang Yandian 康豔典, un « grand chef de Samarcande » (Kang guo da shouling 康國 大首 領) venu s'installer dans la ville de Shanshan 善ム善,au sud du Lob nor, à l'ère Zhenguan (627-649). Il était accompagné d'une population de Hu suffisamment nombreuse pour permettre la fondation ou la reconstruction de plusieurs cités autour de sa capitale. L'une d'elle fut appelée la Ville du raisin (Putao cheng 蒲MK 城)； on signale à son sujet qu'« on planta de la vigne au milieu de la ville ». Cette implantation d étrangers ressemble fort à une colonisation menée par des sédentaires : progression par étapes, fondation de villes, mise en culture 5

5

5

Hou Han shu,j. 88, p. 2922 ; cf. Trombert, 2001, p. 307. ” Cf. Trombert, 2002, p. 494-495. 18

19

Voir F anecdote sur Meng Tuo 孟佗 tirée du Sanfu juelu 三輔決錄 de Zhao Qi 趙岐 (mort en 201), cité en note in Sanguo z加； Weishu, j. 3, p. 92-93, et in Hou Han shu^ j. 78, p. 2534, commentée in Trombert, 2001,p. 304-305. Comme le nommé An Kuzhiyan 安苦 延 qui, en 640, reçoit de l'administration chinoise un champ et une vigne (TAM 78: 29, éd. in Tulufan chutu yvenshu^ vol. 4, p. 72-73). Comme beaucoup d'autres paysans sogdiens, son prénom est translittéré ; celui de la plus jeune de ses filles l'est également, ce qui est le signe d'une femille nouvellement arrivée. On constate en effet que, dès la seconde génération, les Sogdiens commencent à adopter des prénoms chinois. Sur ces questions, voir Ikeda, 1965 et 1993.

Les Sogdiens en Chine

268

de terres, caractère durable de l’implantation (au moins jusqu’en en 691)21. Etant donné l’importance de la communauté sogdienne de Turfan, concentrée principalement dans le canton de Chonghua ^"0* on doit bien sûr évoquer ici le vignoble de cette oasis qui était, et est resté, le plus important de Chine. La vigne est omniprésente dans les documents agraires de Turfan, et cela depuis les plus anciens, qui datent du Ve siècle. On observera d’abord que les textes les plus informatifs sur la viticulture locale proviennent de la même tombe qui contenait le fameux registre de la population de Chonghua pour l’année 707, celui-là même qui a permis de détecter l’importance de la colonie sogdienne turfanaise22 Avec les documents de cette tombe, on évolue dans un milieu où les Sogdiens sont omniprésents. Ceux qui nous concernent directement ici sont deux rapports administratifs relatifs à l’exploitation d’un important vignoble appartenant à l’Etat, désigné sous le nom de « Vigne des quatre coins » (Si jiao tao ES En 698, le responsable de cette vigne était l’un de ces Sogdiens déjà fortement sinisés, un nommé Shi Xuanzheng $^.^1 qui avait été, quelques années plus tôt, le chef d’un village du canton de Chonghua24. Toutefois, on ne doit pas sur-interpréter les documents de cette tombe, car dans leur ensemble, les manuscrits datant des VIIe-VIIIe siècles montrent que les Sogdiens étaient alors impliqués dans la viticulture ni plus, ni moins que les autres paysans de la région. La viticulture était devenue l’affaire de tous. C’est à une époque bien antérieure qu’il faut rechercher le rôle décisif qu’ont pu jouer les Sogdiens dans le développement du vignoble tourfanais. Dans le fonds de Turfan, les siècles anciens sont très mal documentés, mais il se trouve que le plus vieux document qui mentionne des activités viticoles nous oriente indirectement sur la piste sogdienne. Il concerne en effet un vigneron nommé Di Qiang 12^, actif vers 43625. Ce personnage n’était pas à proprement parler un Sogdien. Son patronyme le désigne en effet comme appartenant au groupe turc des Dingling ou Gaoju Mais les membres du clan Di sont bien connus dans l’histoire de Chine pour les liens étroits qu’ils entretenaient avec les Sogdiens. Ils portaient des prénoms iraniens, se mariaient souvent dans ce milieu. qui dirigeait le Certains furent même prêtres zoroastriens, comme ce Di Pantuo temple de Yiwu (Hami) peu avant 640 ; d’autres occupaient les fonctions de sabao 8 comme Di Suo dont nous reparlerons plus loin26. 23.

Il existe donc des liens bien attestés d’une part entre les Sogdiens et le Shanxi, d’autre part entre les Sogdiens et la viticulture. On dispose même de quelques informations précises démontrant l’action viticole des Sogdiens dans la région du Lob 21

D’après le Shazhou Yizhou dizhi (ms. de Dunhuang S. 367). Pour plus de détails, cf. de la Vaissière, Trombert, 2004, p. 950-951. 22 Tombe d’Astana n° 35, mss. TAM 35: 47 à 58/3, éd. in Tulufan chutu wenshu, vol. 7, p. 468-485. 23 Mss. TAM 35: 39 et 40, éd. in Tulufan chutu wenshu, vol. 7, p. 444-451. 24 Sur ce personnage et ses liens avec d’autres Sogdiens, cf. Trombert, 2002, p. 524 et 551. 25 TAM 62: 6/4 et 1, éd. in Tulufan chutu wenshu, vol. 1, p. 103-106 ; cf. Hu Rulei, 1978 ; Trombert, 2002, p. 492. 26 Sur Di Pantuo, cf. le manuscrit S. 367, cité in Rong Xmjiang, 2001b, p. 51. Pour d’autres exemples, qui abondent, se reporter à « Zhai » (autre prononciation de Di) dans l’index du livre de Rong, 2001b, p. 488.

Éric Trombert

_

Le vignoble du Shanxi

269

nor et à Turfan. Cependant, aucun fil conducteur n'est visible entre cette activité viticole des Sogdiens en Chine occidentale et la création du vignoble du Shanxi qui, rappelons-le, était déjà florissant au début des Tang. En effet, pour les siècles antérieurs, la documentation est pauvre, sur les Sogdiens aussi bien que sur un sujet aussi trivial que la vigne. C’est ici que certains domaines jugés plus nobles par les lettrés peuvent éclairer notre recherche. Prenons par exemple l’introduction de la musique occidentale5 très à la mode sous les Sui. Cette question est assez bien documentée. On sait que « la musique de Koutcha ftit acquise lorsque les Wei eurent pacifié la plaine du Nord » ; on connaît même le nom de musiciens célèbres ; ils sont manifestement d'origine sogdienne27. Ceci nous oriente vers une autre piste qui, au demeurant, ne s’oppose pas à celle des Sogdiens : Fessor économique et démographique de F empire des Wei du Nord. L'action des Wei nous intéresse en effet dans le nord du directement puisque, de 398 à 490, leur capitale était à Pingcheng

Shanxi. Du Liangzhou à Pingcheng avec les Wei

Nord

À partir de la fin du IVe siècle, la montée en puissance de l’État Tuoba Wei s’effectue en grande partie par l’installation forcée de diverses populations sur son territoire. Deux objectifs importaient en effet à cette jeune force dont la puissance reposait encore exclusivement sur le potentiel militaire, mais qui ambitionnait de devenir un véritable état calqué sur le modèle chinois. D’une part, capter, au besoin par la force, les élites intellectuelles. D’autre part, mettre en valeur les zones conquises en acquérant une main d’œuvre bon marché, voire servile, mais compétente dans les secteurs d’activité qui étaient étrangers aux pasteurs nomades, l’agriculture, l’artisanat textile, etc. Bien sûr, les sources historiques relatent beaucoup moins bien les transferts de populations laborieuses que ceux touchant le monde de la culture, des arts et de la religion28 . Cependant, les cas abondent dans le Weishu concernant des populations vaincues déportées pour mettre en culture la région de Pingcheng et pour « pourvoir en ：au premier mois 398, transfert de 360 suffisance la capitale » {chong jïngshi 000 personnes, des gens du commun parmi lesquels des Coréens et « plus de 100 000 artisans de tous corps de métier ». Le mois suivant, ces immigrants forcés reçoivent par décret impérial des bœufs de labour et des terres29. Autre exemple, en 446, ce sont 2 000 familles d’artisans de Chang’an qui sont transférés à Pingcheng30. L’un des transferts les plus massifs concerne une région qui nous intéresse tout spécialement pour la vigne, le Liangzhou (Gansu) où régnent alors les Liang du Nord (397-439). En 439, les armées Wei s’emparent de Guzang (Wuwei la capitale, et ramènent à Pingcheng une population de captifs évaluée entre 30 000 et 27

Cités in Jiang Boqin, 2000, p. 61. Voir cependant de la Vaissière, 2004, p. 110. La liste des lettrés et des moines bouddhistes transférés de gré ou de force à Pingcheng est longue. On peut parler de déportation en masse. 29 Weishu, j. 2, p. 32. 30 Weishu, j. 4 B, p. 100. 28

Les Sogdiens en Chine

270

100 000 familles de nobles et de gens du peuple31 à laquelle s’ajoutent 3 000 moines bouddhistes32. Même si les chiffres sont exagérés, ils sont énormes compte tenu de la faible densité de population du Shanxi, et ils révèlent de véritables bouleversements démographiques régionaux. Au cours des années suivantes, les Wei poussent en direction de l’ouest. En 441, ils s’emparent de Jiuquan d’où 4 000 jeunes gens, filles et garçons, sont déportés au Shanxi33. Cependant, aucun transfert de population n’est opéré à partir de Turfan. La raison en est simple : les Wei ne parvinrent jamais à s’emparer de cette oasis. Pour rétablir leurs communications avec l’Occident, ils poursuivirent pourtant leur marche vers l’ouest, mais en contournant Turfan par le sud. En 445, ils s’emparent de Shanshan, puis de Karashar (Yanqi Conquête éphémère, mais qui eut l’avantage de mettre pour la première fois en contact le pouvoir des Tuoba avec les pays du Xiyu. Selon la loi Wei, les paysans et artisans déportés au Shanxi restaient placés sous l’étroit contrôle de l’État ; ils ne pouvaient ni changer de profession, ni se marier hors de leur milieu. L’importance de leur contribution pour l’introduction de ce qu’on appelle aujourd’hui les nouvelles technologies est facile à imaginer, mais on en connaît très mal le contenu. Sur le plan agricole, on ne peut guère citer qu’un seul épisode, relatif à Helian Bobo le fondateur de la dynastie des Xia (407-431). Dans la ville dont il avait fait sa capitale, Lingzhou (Lingwu Helian fit planter un verger magnifique qui quatre siècles plus tard comptait encore plus d’un millier d’arbres fruitiers34. L’entreprise peut étonner quand on sait que ce souverain était un Xiongnu. On peut y voir une influence chinoise, mais aussi bien celle de sédentaires occidentaux. Il faut remarquer en effet que la région de Lingzhou, ainsi que l’ensemble de la boucle des Ordos qui constituait le royaume des Xia, était peuplée de Hu de diverses origines. On y reviendra plus loin. L’histoire ne nous dit pas si les arboriculteurs de Helian se trouvaient parmi les Hu que les Wei déportèrent au Shanxi lorsque ceux-ci s’emparèrent de Lingzhou en 426-427 ; mais la chose est envisageable. Le seul témoignage assuré de la contribution des gens du Liangzhou au développement technique et artistique de l’empire des Wei est apporté par les grottes de Yungang situées non loin de Pingcheng. A partir de 460, c’est à un moine originaire du Liangzhou, Tanyao que les Wei confient la construction des premières grottes (grottes 16 à 20). Il est bien connu que le style des grottes de Yungang fut inspiré, d’une manière générale, par l’art du Gandhara et de l’Asie centrale. Mais pour ce qui est précisément de l’architecture et de la décoration des premières grottes, les spécialistes y voient davantage l’influence directe de l’art développée au Liangzhou sous les Liang du Nord que celle, plus lointaine, du Gandhara35. Ces premières grottes seraient donc l’œuvre d’artisans déportés du Gansu en 439.

nos

R

Selon le Weishu, j . 4 A, p. 90 ; cf. également Jiang Liangfu, 1985, p. 81. Selon le Xu gaoseng zhuan,]. 25, p. 646 c 9. 33 Weishu,]. 30, p. 722. 34 Yuanhe junxian tuzhi,]. 4, p. 91. 35 Cf. Mizuno etNagahiro, 1951, p. 90-93.

32

Éric Trombert

Ze vignoble du Shanxi

271

Pour illustrer ce brassage d’influences dont Yungang fut le théâtre, voici un exemple qui est à ma connaissance la plus ancienne représentation de vigne en Chine une des manifestations de Siva, tient une grappe de orientale : Mahesvara l’une ses de n° 8 datée des années 470 ; cf. fig. 7). Il faut noter dans mains (grotte raisin à ce propos que Mahesvara a été confondu avec la divinité Xian des Iraniens (hu xian Cela est attesté pour le VIIIe siècle seulement36 ; mais la fusion de cultes shen rendus à des divinités iraniennes et indiennes est un phénomène bien antérieur. Si tel était le cas de la déité représentée dans cette grotte de Yungang, ce relief constituerait un témoignage probant de la contribution iranienne à la difîïision de la vigne en tant que thème iconographique dans cette région. Toujours est-il que la présence de Sogdiens parmi les déportés de 439 est clairement attestée dans le Weishu (j. 102, p. 2270) : Depuis longtemps, les marchands de ce pays (Sute guo la Sogdiane) venaient en grand nombre sur la terre de Liang (le Liangzhou) pour commercer. Lorsque Guzang fut prise, ils furent tous capturés. Au début du règne de Gaozong (452-465), le roi (de Sogdiane) envoya une ambassade pour les racheter ; sur ordre impérial, cela lui fut accordé.

Il est intéressant de noter que les Wei gardèrent ces Sogdiens pendant près de vingt ans bien que, depuis leur capture, deux ambassades se soient présentées à Pingcheng ; la première au 11e mois 439, quelques mois seulement après la prise de Guzang, la seconde en 44437. Une telle durée indique que ces marchands formèrent à Pingcheng une communauté, certes encadrée par le pouvoir tuoba, mais qui vivait certainement selon ses propres coutumes et qui menait une activité économique appréciée par les Wei. Ces Sogdiens installés au Shanxi en 439 n’étaient pas les premiers à venir là. On connaît un précédent individuel, celui de An Tong Pratiquant le commerce mais animé d’autres ambitions, ce supposé descendant d’An Shigao rallia le fondateur des Wei, Tuoba Gui, dès 386 quand sa capitale était encore à Shengle (au nord-est de la boucle des Ordos). Comme le note Antonino Forte (1995, p. 16), Shengle constituait alors “a very good position if he was engaged in commerce with Central Asia”. On doit signaler ici l’apparition de Wuyuan parmi les régions viticoles de Chine à une date inconnue comprise entre le IIIe et la fin du Ve siècle38. Wuyuan se situe sur le cours le plus septentrional du Fleuve jaune, à l’ouest de Shengle. L’endroit paraît peu propice à la culture de la vigne. Tout ce qu’on peut dire est que, dès le début de son 36

Selon Wei Shu (mort vers 757), le temple nommé Hu xian ci 40^^, fondé en la 4e année de Père Wude (621) à Chang’an, était consacré à « la divinité céleste des Hu du Xiyu que les textes bouddhiques nomment Moxishouluo (Mahesvara) » ; Liang jing xin ji, p. 7. Le propos fut repris par Du You dans son TongdianA. 40, p. 573. 37 Comme le note Enoki (1955, p. 44), l’ambassade « au début du règne de Gaozong » fiit certainement celle de 457 puisque c’est la seule qui eut lieu sous ce règne (Weishu,}. 5, p. 116). 38 Mingyi bielu,]. 1, p. 87. Cet ouvrage de pharmacopée et de botanique complète le premier ouvrage de ce genre, le (Shennong) Bencao jing des Han. Les informations nouvelles qu’il apporte doivent dater de l’époque indiquée ci-dessus. Cf. Needham, 1986, p. 249. Par rapport au Bencao jing qui se bornait à dire que la vigne pousse dans les vallées de montagne (shan gu LÜ@), le Bielu cite trois régions viticoles : le Longxi (sud-est du Gansu), Wuyuan et Dunhuang

272

一̶

Les Sogdiens en Chine

règne, Tuoba Gui (r= 386-409) avait implanté des colonies agricoles en ce lieu hautement stratégique situé au-delà de la grande muraille39, et qu'après 439 les Wei y déportèrent aussi des gens du Liangzhou40. L'expansion des Wei, qui ouvrirent ainsi une liaison directe entre le Liangzhou et le Shanxi à travers la zone de la steppe, donne une importance particulière à une voie de pénétration des influences occidentales plus septentrionale que la route classique menant du Gansu à Chang'an via Lanzhou et la vallée de la Wei. Cette orientation nordiste amène à envisager un processus d5 introduction de la vigne plus complexe, faisant intervenir à la fois les Sogdiens et certains peuples nomades de la steppe.

Dÿautres H%

nomades

« Les Hu apportent chaque année tribut de vin de raisin »41. Ce vers de Bao Fang 負包 防 est souvent cité pour preuve de F activité commerciale -des Sogdiens en matière de vin. Pour renforcer la démonstration, on le juxtapose de manière contestable avec deux autres citations : 1) Celle du Shÿi sur le vin du Ferghana» 2) Une mention de tribut de vin figurant dans les annales des Tang qui concerne une ambassade du pays de Shi 史, datée de 727 42. Le poème de Bao Fang mérite pourtant d 5 être cité en entier : Le peuple Han fait régner une paix durable dans tout l'empire. Les rois Rong des pays de F Ouest se prosternent tous jusqu5 à terre. Les chevaux célestes tiennent à la gueule des fleurs de luzerne. Les Hu offrent chaque année le tribut de vin de raisin. 漢家 海内承平久 , 萬國 戎王 皆稽首 。天馬常銜苜蓿草 , 胡人歲獻葡萄酒 。

Remis dans son contexte, le dernier vers ne permet nullement d1assimiler les Hu aux seuls Sogdiens. Au contraire, la valeur symbolique du poème oblige à élargir le sens de Hu à tous les Barbares de r Ouest et du Nord. Et ce n'est pas le fait que le seul tribut de vin jamais reçu à la cour fut apporté par une ambassade sogdienne qui peut donner un fondement historique au rôle des Sogdiens dans le commerce du vin sous les Tang. En effet, cette mention n'a pas grande valeur sur le plan économique puisqu'elle est unique parmi la trentaine d'ambassades sogdiennes dont on connaît les tributs depuis 618 jusqu'en 772. La plupart du temps, les Sogdiens apportent des chevaux, des fourrures et autres objets précieux beaucoup plus transportables que le vin. Et d'ailleurs, à cette époque, il n'est plus nécessaire d'importer du vin de si loin : le Liangzhou en produit d'excellents, et le meilleur vient de Turfan. C'est celui que boit Yang Guifei

39

Cf. Jiao Runming, 2003, p. 18. On connaît le cas de Duan Xin 段信 (biographie de Duan Rong 段榮, Bei Qi shu, j. 16, p. 207) et celui de SimaZiru 司馬子如 (ibid., j. 18, p. 238). 41 p. 3485. Quan Tang shi, 42 Cette ambassade amène également « des danseuses hu tourbillonnantes » 胡旋女十. Cefu yuangui, j. 971, p. 1 1408. Pour un raisonnement de ce type, cf. Cai Hongsheng, 1998, p. 68. 40

n Éric Trombert

_

Le vignoble du Shanxi

273

lors des réunions de dégustation de vin qui ont lieu dans les jardins impériaux43. Tous les textes anciens disent et répètent que c’est « à partir du moment où le royaume de Gaochang fut conquis » qu’on se mit à cultiver la vigne dans les parcs impériaux et que l’on acquit la technique de la vinification44. La conquête de Turfan eut probablement une certaine incidence sur la viticulture en Chine métropolitaine. Mais force est de constater qu’aucune des ambassades envoyées par le royaume de Gaochang jusqu’à sa chute en 640 n’apporta du vin, ni même du raisin45. Une seule offrande de raisin est enregistrée dans les annales, et elle fot le fait d’un personnage désigné comme un titre que portaient les kagan des Turcs occidentaux46. Même si, à mon yabgu n’est pas suffisant pour en déduire, avec Laufer, que les Turcs furent les avis, ce véritables artisans de l’introduction de la vigne et du vin en Chine orientale au début du VIIe siècle47, ce présent d’un yabgu ouvre une perspective qui était a priori peu vraisemblable pour de tels produits, celle du passage de la vigne et du goût du raisin à travers les steppes du Nord et le monde des nomades. La diffusion de la vigne par un peuple nomade est au premier abord difficile à concevoir en raison du mode de vie de ces populations et des conditions naturelles des régions qu’elles occupent généralement48. Mais précisément parce que la vigne n’est pas une culture vivrière mais un élément de civilisation, sa diffusion doit être envisagée différemment. Boire du vin sous une treille ombragée constitue un véritable luxe, un symbole de la douceur de vivre ; il est donc vraisemblable que l’usage, puis la fabrication du vin, entrent peu à peu dans les mœurs des nomades dès le moment où ils commencent à être séduits par le mode de vie des sédentaires. On en a un exemple éclatant avec les Ouïgours qui sont devenus les meilleurs vignerons de Chine une fois qu’ils se furent sédentarisés à Turfan. Avant eux, les Turcs avaient sans doute pris goût au vin au contact de leurs vassaux sédentaires - les Sogdiens - qui exerçaient sur eux une grande influence culturelle. Toutefois, pour une question de chronologie, ce ne sont pas les Turcs occidentaux qui ont pu apporter la vigne au Shanxi. Elle y était cultivée avant l’essor de leur empire. Pour résumer notre enquête au point où elle en est, la viticulture fut introduite au Shanxi avant le VIIe siècle par un ou (plusieurs) peuple « hu » (nomades du Nord ou de 43

Yueshi , cité par Cai Hongsheng, ibid. Ce vin vient du Liangzhou occidental ; le terme désigne manifestement la région de Turfan. 44 Pour les références in Tang huiyao, Taiping yulan, Bencao gangmu, etc., cf. Trombert, 2001 notes 55, 57. 45 Elles furent pourtant nombreuses dès le règne de Gaozu (618-626) à avoir été consignées dans le Cefu yuangui (j. 970, p. 11397-11400), et Qu Wentai^ÿ^, le roi de Gaochang, vint en personne à Chang’an en 630 (Xin Tang shu,]. 221 A, p. 6220). 46 Cefu yuangui,]. 970, p. 11400. Cf. Trombert, 2001 note 56. 47 « The Chinese. . . acquired the art of wine-making as late as the T’ang from a Turkish tribe of Turkestan » (Laufer, 1929, p. 233). Le lien de cause à effet qu’il établit entre l’offrande du yabgu et l’introduction de la viticulture à Chang’an après la conquête de Gaochang en se basant sur le Tang huiyao (j. 100, p. 1796) n’est pas aussi évident qu’il semble le dire. D’ailleurs, le Taiping yulan (j. 972, p. 4308 b) présente ces deux événements dans l’ordre inverse. La Bible en fournit un exemple pour la Palestine à propos des Réchabites, peuple nomade auquel il fut ordonné : « Vous ne boirez point de vin, et vous ne bâtirez point de maisons ..., vous ne planterez point de vignes et n’en posséderez point, mais vous habiterez sous des tentes pendant tous vos jours. » (Jér. XXXV, 6-7)

274

Les Sogdiens en Chine

r Ouest) sous F influence d'éléments sogdiens qui vivaient parmi eux„ C5est ici que le caractère mélangé de la communauté « sogdienne » du Shanxi doit retenir notre attention. Les études récentes sur la communauté « sogdienne » de Bingzhou (Taiyuan) donnent de cette communauté une image particulière : La plupart des personnages identifiables ayant exercé en son sein les hautes fonctions de sabao ne sont pas des Sogdiens de pure souche. Ce sont : Long Run ft/H (561-653), administrateur en chef au plus tard en 646 ; Yu Hong du bureau du sabao (sabaofu changshi père de Di (534-592), jianjiao sabaofu vers 580 ; et Di Suo 570. Les origines « hu » non vers da (546-615), Tusuo mohe sabao sogdiennes de ces trois sabao peuvent nous servir de fils conducteurs pour rechercher les introducteurs de la vigne dans cette région. Les Long étaient des ressortissants du royaume de Karashar„ Ils avaient essaimé vers l'est, notamment à Hami où une communauté de Long (Long buluo était présente sous les Tang49 à côté d'une forte population de Hu dont certains étaient Au début du qualifiés de « divers Hu des pays d'Occident » Xiyu za Hu VIIe siècle, Hami possédait un temple de la divinité du feu Xian (huo xian miao) dirigé par un prêtre originaire de Turfan nommé Di Pantuo. Comme Turfon et Hami, Karashar possédait également une tradition viticole ancienne51. Mais on ignore à quelle époque et dans quelles circonstances des membres du clan Long vinrent s’établir au Shanxi. un peuple non Han, ou Jihu Yu Hong serait apparenté aux Buluoji pour la première fois en dont clairement attestée la présence est peut-être turcophone, et Sizhou 524 (sous le nom de Buluojian hu 35^^) dans la région de Fenzhou au sud de Taiyuan52. Ces Hu avaient la particularité d’être des agriculteurs. (Nord et le Jinxi Peut-être étaient-ils un peuple autochtone dans le Shânbei Shanrong ; Chidi aux et c’est Shanxi), apparentés Ouest aux et Shânxi l’opinion de Pulleyblank53. Mais on en fait aussi un peuple inféodé aux Xiongnu de Liu (Liu Yuan Yuanhai avec lesquels ils auraient pénétré au Shanxi au début du IVe siècle54. Pour ce qui nous occupe ici, les deux thèses ne sont pas contradictoires. Pulleyblank évoque lui-même les multiples aller-retour effectués par ces peuples non Han entre la steppe et la Chine du Nord. Venant du Nord-Ouest, les Jihu ont donc très bien pu retourner au Shanxi, leur terre ancestrale, avec les troupes de Liu Yuan lorsque celui-ci, basé à Lishi s’empara de la région de Bingzhou en 304 où il se proclamera peu après empereur de la dynastie des Han. On notera que Liu Yuan établit dans la région de Wenshui (vallée de la Fen, au sud de son frère aîné, Yannian Taiyuan)55. Wenshui est resté jusqu’à aujourd’hui un des haut-lieux viticoles du Shanxi. dans Pour l’époque des Wei, on notera aussi l’invasion de Hu occidentaux (Xihu "9

Le Shazhou Yizhou dizhi (ms. S. 367) qui l’atteste, fut composé entre 850 et 886. Yuanhe junxian te/zz, j.40,p. 1029,Xin Tang shu, j. 4Q,p. 1643. 51 Weishu,]. 102, p. 2265 \Beishi.y. 97, p. 3216 \Da TangXiyu ji, j 1, p. 2. 52 Pour ce qui suit, les sources historiques sont données in Lin Meicun, 2002, p. 72 sq., sauf indication contraire. 53 Pulleyblank, 1994. 54 Zhoushu,j. 49, p. 896-897 ; Beishi.j. 96, p. 3194. 55 Yuanhe junxian tuzhi,j. 13, p. 372. 50

Éric Trombert

Le vignoble du Shanxi

le même district de

275

Wenshui5 et la présence de Jihu dans le district voisin de Xiaoyi

Une autre migration de Jihu vers le Shanxi est suggérée par la biographie de Yu , un contemporain de Yu Hong supposé, lui aussi, d’origine Jihu - tous Qingze deux portent le même patronyme, qui, d’après les biographes de Yu Qingze, s’écrivait à (poisson), et Yu Hong, d’après son épitaphe, serait originaire d’un pays l’origine Yu (le pays du poisson). Yu Qingze, qui orthographié de la même manière, Yu guo langue xianbei, commença sa carrière sous les Zhou au Shanxi où aimait s’exprimer en (un district il s’employa à discipliner les Jihu - il était alors duc de Qinyuan au sud de Wenshui)57. Mais ses ancêtres étaient de Lingwu où ils avaient servi les Xiongnu du clan HeliaOg donc à F époque où prospérait en ce lieu le verger planté par Helian Bobo. Il est possible que la famille de Yu Hong soit également originaire de Lingwu et qu’elle ait quitté cet endroit après que les Ruanruan en furent chassés par les Wei en 470. Cela expliquerait que Yu Hong et son père aient été tous deux au service des Ruanruan58. Il faut souligner ici l’importance de Lingzhou (Lingwu) - qui sera sous les Tang en tant que nœud de l’un des « Six départements Hu » (Liu Hu zhou communication, au bord du Fleuve jaune, reliant directement Liangzhou et Taiyuan sans passer par Lanzhou. C’était une étape habituelle sur le parcours migratoire de nombreuses familles sogdiennes venues du Gansu qui, après avoir transité parmi les tribus turques, finirent par se fixer après plusieurs générations dans le Hebei59. Considérons maintenant la piste des Dingling/Gaoju en imaginant un lien symbolique entre Di Qiang, le premier vigneron de Turfen connu, et Di Suo, sabao de Bingzhou. Jusqu’au IIIe siècle de notre ère, les Dingling X vivaient en Sibérie méridionale, « au nord du Kangju » dit le Weilü^. On connaît au moins deux vagues migratoires de Dingling vers la Chine du Nord. D’abord des guerriers Dingling, venus combattre dans les armées de Wang Mang (r. 9-23), sont fixés dans le nord-est du Shanxi, dans la commanderie de Dai La seconde vague a lieu en 330 : « Des Ding Ling conduits par Di Bin qui ont longtemps vécu au Kangju, émigrent en Chine du Nord où ils se rallient à la dynastie des Zhao » 62? c’est à dire à Shi Le dont nous reparlerons. Avec ce groupe, on a un cas précis d’immigration de gens de la steppe qui se sont fortement frottés aux coutumes iraniennes avant leur arrivée. Et comme les chefs Dingling actifs en Chine du Nord à cette époque portent tous le nom de Di, on comprend mieux pourquoi les Di vivant aux siècles suivants étaient si proches

—

56

Yuanhe junxian tuzhi^y. 13, p. 372 et 378. 73, p. 2516. Yu Qingze est actif au moins jusqu’en 597. SR Pour le détail de cette hypothèse, cf. Lin Meicun, 2002, p. 80 sq. 59 Lors de la conquête de la boucle des Ordos par les Tang au VIIe siècle, nombreux furent ces Sogdiens qui se soumirent avec les Turcs (par exemple, Kang Rizhi H Ces lignées achèveront leur parcours sur la frontière nord-est souvent comme gouverneur militaire ; par exemple, Shi Xiancheng et He Jintao qui fiirent tous deux jieshi de Weibo Cf. Rong, 2001b, p. 91-92. En fait, au nord-est du Kangju, ainsi que cet ouvrage le précise: « À 3 000 au nord-est de la frontière du Kangju et à 8 000 à l’est de la capitale de ce royaume. » Sanguo zhi, Weishu^]. 30, p. 862. 61 Hou Han shu,]. 90, p. 2981. 62 Zizhi tongjian,y 94, p. 2976. 57

〃

〃

276

Les Sogdiens en Chine

—

des Sogdiens, étaient prêtres mazdéens comme Di Pantuo, ou chefs de communauté comme Di Suo. Dans les récits des guerres qui embrasent la Chine du Nord pendant le reste du IVe siècle, on rencontre des groupes de Dingling en des lieux très divers, notamment dans le nord du Shanxi (à Dai), mais rien n’indique l’installation d’un groupe précisément dans la région viticole du Shanxi central63. Un peu plus tard, les Tuoba déportèrent à leur tour plusieurs groupes de Gaoju sur leurs terres du nord Shanxi64. Rien ne dit que ces Gaoju aient eu, avant leur arrivée au Shanxi, des rapports avec des Iraniens aussi étroits que les Dingling de Di Bin ; mais il est admis que les Gaoju entretenaient généralement des relations étroites avec les peuples d’Occident. En voici deux exemples. En 490, c’est un marchand hu portant un nom sogdien que le qagan des Gaoju de Beshbalik envoya négocier une alliance avec les Wei65. Et en 516, se réfugièrent en masse chez les Ephthalites après les Gaoju du roi Mi-e-tu avoir été écrasés par les Ruanruan. Ils y séjournèrent cinq ans avant de repartir vers l'est prendre une éclatante revanche sur leurs rivaux66 Lorsque Liu On ne peut manquer d’évoquer ici un quatrième peuple, les Jiehu dans région de la pouvoir son établit et Jin contre soulève les occidentaux Yuan se Bingzhou, il est assisté par Shi Le, un Jie qui, selon ses biographes, serait issu d’une On connaît les présomptions quant branche distincte des Xiongnu, les Qiangqu aux origines iraniennes des Jie68 ; elles sont en réalité fragiles. Mais ce qui est très probable c’est que des éléments sogdiens gravitaient autour et au sein même de cette nébuleuse constituée d’éléments divers (za Hu) que les Chinois ont regroupés sous le terme de Jie. Si j’évoque ici les Jie, c’est en raison de leur forte présence au Shanxi. Shi Le était originaire de Bingzhou, et la branche des Jie dont il était issu était bien implantée au sud-est de Taiyuan, sur les contreforts orientaux des monts Taihang et de Shangdang _h^). (commanderies de Wuxiang Ainsi, plus que d’autres régions de Chine du Nord, le Shanxi offre l’exemple d’un peuplement opéré par des vagues successives de populations mixtes, mêlant sédentaires occidentaux d’origine iranienne et nomades du Nord et de l’Ouest. Si la vigne, d’origine assurément iranienne, s’est développée dans cette province plutôt qu’ailleurs en Chine, c’est sans doute parce que, parmi les différentes voies de pénétration des Sogdiens en Chine, elle a suivi celle empruntée par les groupes sogdiens qui vivaient en étroit contact avec les gens de la steppe. La fusion culturelle sogdo-turque qui sera une réalité clairement observable à partir du VIe siècle - a été précédée d’une longue *.

—

63

homonyme parfait du chef du IVe siècle, qui vivait en 1739 dans un J’ai bien trouvé un Di Bin (Jinci Taiyuan village près de zhi, p. 862-3). Mais cela ne prouve évidemment rien. Seule une étude statistique des occurrences du patronyme Di dans les districts concernés pourrait être indicative. 64 En 363, le grand-père de Tuoba Gui amène sur ses terres 10 000 Gaoju vaincus et un million de tête de bétail (An Jiesheng, 1999, p. 106). En 399, ce sont plusieurs dizaines de milliers de Gaoju qui sont amenés du Nord-Ouest avec bêtes, outillage et charrettes. 65 Weishu,]. 103, p. 2310. Sur le nom de ce marchand, voir de la Vaissière, 2002, p. 202. 66 Zizhi tongian,}., 149, p. 4667. 61Jinshu,j. 13, p. 397 etj. 104, p. 2707. 68 Pour des arguments allant dans ce sens, cf. Tang Changru, 2002, p. 398-41 1 ; Pulleyblank, 1962, p. 246-247.

Éfic Trombert

_

Le vignoble du Shanxi

277

proto-histoire mal documentée 69 . On peut considérer que le vignoble du Shanxi constitue, du fait de son ancienneté, une trace de cette collaboration sogdo-nomade très

ancienne. Epilogue : Le vin, boisson de luxe, boisson des Hu Au cours de l’histoire, le tribut en vin de Taiyuan fot aboli sur ordre impérial à deux reprises dans des circonstances identiques : Sous les Tang, en janvier de l’année 837™, et au tout début de la dynastie des Ming, en 1373". Au tournant des VIIIe-IXe siècles, il se développe en Chine un courant de pensée prônant un retour aux sources de la tradition chinoise sensée avoir été corrompue par les influences étrangères (bouddhistes et occidentales) que la Chine a subies depuis la fin des Han. Sur le terrain littéraire et philosophique, ce courant s’exprime à travers le animé par des lettrés mouvement de retour au « style antique » (guwen Liu Zongyuan et orthodoxes et violemment anti-bouddhistes tels Han Yu qui ce fut à remontent ressenti comme une politiques prémices ses Mais Ao Li catastrophe nationale causée en grande partie par des Hu, la rébellion d’An Lushan. Le point culminant de cette poussée nationaliste fut la grande proscription des religions étrangères qui date des années 842-845. Elle fut précédée par des mesures xénophobes dont deux décrets, promulgués en 831 et en 836, interdisant aux Chinois tout rapport et autres « étrangers » (waiguo ren avec les « gens de couleur » (zhuse ren g A)72- On voit donc dans quel contexte intervint le décret de 837 abolissant le tribut de vin du Shanxi. L’abolition décidée par le fondateur des Ming, Taizu (r. 1368-1399), après qu’il eut reçu son premier tribut de vin de Taiyuan se situe, elle aussi, dans un contexte nationaliste : réaction à l’occupant mongol, à ses alliés (dont les Ouïgours), et rejet de leurs coutumes. On ne détaillera pas ici les mesures que les Yuan avaient prises en faveur de la viticulture, ni les témoignages de l’usage fréquent qu’ils faisaient du vin, mais il faut les signaler. La suppression du tribut par Taizu est d’ailleurs explicitement motivée par des raisons très chinoises : Elle vise à « réconforter le peuple », car « le rôle du souverain est de nourrir le peuple et non pas de le harasser pour le plaisir de sa bouche et de son ventre »73. En d’autres termes, les paysans ne doivent pas être détournés de leur vocation première qui est la culture des céréales. L’abolition décrétée par Taizu dut avoir un effet direct sur la production vinicole au Shanxi. La vigne continua à y être cultivée, mais exclusivement pour produire du raisin. On en a plusieurs témoignages. D’abord, pour le XVIIe siècle, celui des missionnaires

69

de la Vaissière, 2002, p. 196 sq. Décret du 12e mois de la première année de l’ère Kaicheng, cité in Cefu yuangui,]. 168, p. 2028 a. 71 Mingshi,}. 82, p. 1989 et Gangjian yizhi lu, p. 2609. C’est la même année que fut promulgué le code des Ming. 72 Cefityuangui,). 999, p. 11727 b-11728 a. 73 Gangjian yizhi lu, p. 2609. 70

278

_

Les Sogdiens en Chine

chrétiens74. Ensuite, le fait que durant l’ère Guangxu (1875-1909) le tribut envoyé à la cour consistait uniquement en raisins secs. C’est le district de Wenshui qui était chargé de cette contribution75. Enfin, aucun des vignerons des districts de Wenshui et de (entre Wenshui et Taiyuan) que j’ai pu interroger pendant l’été 2001 ne Qingxu savait faire du vin, et aucun n’avait le souvenir que leurs pères ou grands-pères en aient jamais fabriqué. Depuis les années 1980, les plantations de vignes ont été considérablement développées dans ces deux districts, et de nombreuses fabriques de vin ont été créées. Mais cette renaissance de l’art vinicole est l’œuvre exclusive de techniciens étrangers ou de Chinois formés à l’étranger. Les deux décisions impériales prises contre la production de vin au Shanxi démontrent a contrario les liens qui unissaient, dans l’esprit des élites chinoises, la viticulture et les « barbares occidentaux ». C’est bien en tant qu’élément de la culture hu - au sens large mêlant Sogdiens, Turcs, Ouïgours et Mongols - que l’usage du vin fut perçu, et proscrit, par les Han. Cette perception rejoint les hypothèses faites tout au long de cette étude sur l’origine du vignoble du Shanxi : Celui-ci peut être considéré comme l’œuvre commune de tous les Hu, témoignant en cela de la précocité de l’influence iranienne sur les peuples de la steppe.

74 « Cette province (le Shanxi) a des vignes, et le raisin qu’on en recueille est le meilleur qui croisse dans toute cette partie de l’Asie. Il seroit facile aux Chinois d’en fabriquer un très-bon vin ; mais ils préfèrent de faire sécher ces raisins, qu’on transporte et qu’on vend dans toutes les provinces de l’Empire. » Grosier, 1785, p. 63. 75 Shanxi tongzhi, ]. 100, p. 7086.

Éric Trombert

vignoble du Shanxi

279

Bibliographie Sources Les réferences aux Histoires dynastiques renvoient à Fédition Zhonghua shuju, Pékin, 1972 et suiv. (SHENNONG) Bencao jing(神農) 本草 經 (Classique de la matière médicale de Shennong) Shennong bencao jing Anonn (achevé au 2e siècle de ロ。èj. Ma Jixing, jizhu 神農本阜 經輯注 (Édition annotée du Shennong bencao jing), Pékin, Renmin weisheng chubanshe, 1995. BENCAO TUJING 本草圖 經 (Classiqueillustré de la matière médicale) Su Song 蘇 頌 (1020-1101), Shang Zhiyun, éd Hefei, Anhui kexue jishu chubanshe, 1994. Caomu zi 草木 子 Ye Ziqi 葉子奇 (actif vers 1387),Pékin, Zhonghua shuju, 1959, réimpr. 1997. Cefu YUANGUI 冊府元龜 (La grande tortue du magasin des documents) Wang Qinruo 王欽若 (962-1025),Hong Kong, Zhonghua shuju, 1960. Da Tang XlYU JI 大唐西域 記 (Notes sur les contrées occidentales à l'époque des Grands Tang) Xuanzang 玄奘 (600-664),Shanghai, Shanghai renmin chubanshe, 1977. Gangjian YIZHI LU 綱鑑 易知錄 (Recueil de choses simples compilées dans les Annales) Publié en 1711 par Wu Chengquan 吳乘權,Pékin, Zhonghua shuju, 1960. JlNCl ZHl 晉祠誌 (Monographie de Jinci) Liu Dapeng 崟［|大朋鳥 (1857-1942),Taiyuan,Shanxi renmin chubanshe, 1986. LIANGJING XINJI 兩 京新記 (Nouvelles notes sur les deux capitales) Wei Shu 韋述 (mort vers 757), Congshu jicheng jianbian, n° 793, Taiwan shangwu yinshuguan, 1965. MlNGYI BIELU 名醫別錄 (Autres écrits de médecins illustres) Tao Hongjing 陶弘景 (456-536), Shang Zhijun, éd., Mingyi bielu jÿiao ben 名 醫別錄輯校本 (Editionannotée du Mingyi bielu), Pékin, Renmin weisheng chubanshe, 1986. Quan Tang shi 全 唐詩 (Recueil intégral des poètes des Tang) Pékin, Zhonghua shuju, 1960. Shanxi tongzhi 山西通誌 (Monographie générale du Shanxi) Wang Xuan 王軒 (1823-1887),Pékin, Zhonghua shuju, 1990. Taiping YULAN 太平 御覽 (Encyclopédie impériale de Père Taiping) Li Fang 孝昉 (925-996), Pékin, Zhonghua shuju, 1960. Tang HUïYAO 唐 會要 (Documentsessentiels sur la dynastie des Tang) Wang Pu 王溥 (922-982), Pékin, Zhonghua shuju, 1955, réimpr. 1998. Tongdian 通典 (Histoire générale des institutions) Du You 杜佑 (735-812), Yan Pinzhong et al., éd., Changsha, Yuelu shushe, 1995. Tulufan CHUTU WENSHU 吐 魯 番出土文書 (Documents exhumés à Turfan) Guojia wenwu ju 國家 文物局,et al., éd., Pékin, Wenwu chubanshe, 10 vol. (1981-91) + 1 vol. d'index (1996). Xu gaoseng ZHUAN 續高僧 傳 (Suite aux Vies des moines illustres) Daoxuan 道宣 (596-667), in Taishö shinshü daizökyö 大正 新修 大藏 經, Tokyo, 1924-1935, n° 2060,vol. 50, p. 425-708. ソ

金；:武送;晓; およ三三三：三

280

Les Sogdiens en Chine

YiNSHAN ZHENGYâO, 飲膳正要 (Principes essentiels de diététique) Hu Sihui 忽 思慧 (actif déb. XIVe siècle), Li Chunfang, éd。, P6kin, Zhongguo shangye chubanshe, 1988. Yuanhe junxian TUZHI 元和 郡縣圖 誌 (Géographieillustrée des commanderies et des districts

àTèreYuanhe) Li JifLi 李吉甫 (758-814), He Cijun, éd.; Pékin, Zhonghua shuju, 1983,réimpr. 1995. ZlZHl TONGJIAN 資治通鑒 (Miroircomplet pour aider à gouverner) Sima Guang 司馬光 (1019-1086), Pékin, Zhonghua shuju, 1956.

Travaux An Jiesheng 安 介生 Shanxi yimin shi 山西移民 史 (Histoire des migrations au Shanxi), Taiyuan, 1999 Shanxi renmin chubanshe. CAI Hongsheng 蔡 鴻生 1998 Tangdai jiuxing Hu yu Tujue wenhua 唐代九姓胡與突厥文化 (La civilisation des Hu aux neuf noms et celle des Turcs sous les Tang), Pékin, Zhonghua shuju. Chen Yinque 陳寅恪 1978 Yuan Bai shi jianzheng gao 兀白詩 箋證 稿 (Commentaires provisoires sur les poèmes de Yuan Zhen et de Bai Juyi), Shanghai, Shanghai guji chubanshe. Enoki Kazuo 擾一雄 « Sogdiana and the Hsiung-nu », Central Asiatic Journal, vol. I, n° 1, p. 43-62. 1955 Forte, Antonino The Hostage An Shigao and his Offspring. An Iranian Family in China, Kyoto, 1 995 Italian School of East Asian Studies. Ge Chengyong 葛承雍 « Cui Yingying yu Tang Puzhou Sute yimin zongji » 崔鶯鶯與 唐蒲州粟特移 2002 民蹤跡 (Cui Yingying et les traces de l'immigration sogdienne à Puzhou sous les Tang), Zhongguo lishi -wenwu, n° 5, p. 60-68, 79. Ghirshman, R. 1953 « Notes iraniennes, V. Scènes de banquet sur r argenterie sassanide » Artibus Asiae, vol. XVI, n° 1/2, p. 51-76. Grosier (Abbé) 1785 Description générale de la Chine, ou Tableau de l'état actuel de cet empire. Paris, chez Moutard. Hu Rulei 胡如雷 « Jijian Xinjiang chutu wenshu zhong fanying de Shiliu guo shiqi zudian qiyue 1978 de guanxi » 幾 件 新彊 出土 文書 中 反映 的 十 六 國時 期 租 佃 契約 關係 (Quelques documents retrouvés au Xinjiang relatif au système de fermage à 1 époque des Seize royaumes), Wenwu, n° 6, p. 22-25. Ikeda On シ也田温 « Hasseiki chûyô ni okeru Tonkô no Sogudojin shûraku » 世紀中葉に おけ 1965 る敦煌のソ ク' ド 人聚落 (Les implantations sogdiennes à Dunhuang au milieu du VIIIe siècle), Yürashia bunka kenkyü, n° 1, p. 49-92. « Chu Tang Xizhou Gaochang xian shoutian bu kao » 初唐西州高昌縣授田 1993 シ専 考 (Étude sur les registres d'attribution de terres dans le district de Gaochang à Xizhou au début des Tang), 泊 Joseph Wong et Lau Kin-ming 5

ハ

Éric Trombert

Ee vignoble du Shanxi

281

(éd) Sui Tang shi lunji 隋唐史論集 (Recueild 5 articles sur Thistoire des Sui et des Tang), Hong Kong, The University of Hong Kong, p. 178-197.

JlANG Boqin 2000

"The Zoroastrian Art of the Sogdians in China”, China Archaeology and 4〃 Digest, "Zoroastrianism in China'ゝ vol. 4, n° 1, p. 35-71. Jiang Liang揖 姜亮夫 Mogao ku nianbiao 莫高 窟 年表 (Chronologie des grottes de Mogao), 1985 Shanghai, Shanghai guji chubanshe. JlAO Rumning 焦潤明 2003 « Guanyu lishi jiangyu guishu ruogan lilun wenti de yanjiu 關於歷史疆域 歸 屬 若干 理論問題的 研究 » (Recherches sur quelques questions théoriques 『histoiredes régions frontières), Zhongguo bianjiang shidi yanjiu, concernant '

nÔ25p. 148. DE LA Vaissière, Étienne 2002 Histoire des marchands sogdiens, Paris, Collège de France, Institut des Hautes

Études Chinoises (Bibliothèque de l’IHEC, n° XXXII), 2e éd., 2004. Vaissière, Étienne, Trombert, Éric3 2004 « Des Chinois et des Hu. Migrations et intégration des Iraniens orientaux en milieu chinois durant le Haut Moyen-Âge », Annales ESC, 59e année, n° 5-6, p. 931-69. Laufer, Berthold 1929 Sino-Iranica, Chicago, Field Museum of Natural History. Lin Meicun 林 梅村 2002 « Jihu shi jikao. T aiyuan xin chu Suidai Yu Hong muzhi de jige wenti » 稽胡 史迹考一太原 新出隋代虞弘墓志的幾個問題 (Recherches sur Fhistoire des Jihu. Quelques questions à propos de la stèle funéraire de Yu Hong découverte récemment à Faiyuan), Zhongguoshi yanjiu, vol. 1, p. 71-84. MARSHAK, Boris 2001 « La thématique sogdienne dans l'art de la Chine de la seconde moitié du VIe siècle », Comptes-rendus de VAcadémie des Inscriptions et Belles-Lettres, séances de l'année 2001,Paris, p. 227-264. Mizuno, S., etNAGAHiRO, T. 1951 Yün Kang, Detailed Report of Archaeological Survey Carried Out by the Kyoto, Kenkyusho, Tohobunko vol. V. DE la

Needham, Joseph, éd. 1986. Science and Civilisation in China, vol. 6, part I : Botany. Cambridge, Cambridge University Press. Pulleyblank, Edwin G. "The consonantal System of Old Chinese”, Asia Major, vol. IX, part 2, p. 1962 206-265 "Jihu: Indigenous Inhabitants of Shaanbei and Western Shanxi", in Opuscula 1994 Altaica: Essays Presented 比 Honor of Henry Schwarz, ed. E.H. Kaplan and D.W. Whisenhunt, Bellingham, WA, p. 499-531. RONG Xinjiang 榮 新江 2001a « Sui ji Tang chu Bingzhou de sabaofu yu Sute juluo » 隋及唐初并 州的薩保 府與粟特 聚落 (Lesimplantations sogdiennes et F institution des sabao à Bingzhou sous les Sui et au début des Tang), Wenwu, n° 4, p. 84-89. 2001b Zhonggu Zhongguo yu wailai wenming 中古中國 與 外來文明 (Les civilisations étrangères dans la Chine médiévale), Pékin, Sanlian chubanshe.

282

Les Sogdiens en Chine

Changru 唐長孺 Wei Jin Nanbeichao shi luncong 魏晉南北朝史論叢 (Recueil d articles sur 2002 rhistoire des Wei, des Jin et des dynasties du Nord et du Sud), Pékin, Sanlian shudian, 1955; rééd. Shijiazhuang, Hebei jiaoyu chubanshe5 2000, réimpr. 2002. Trombert, Éric « La vigne et le vin en Chine. Misères et succès d'une tradition allogène » 2001 Première partie : « De la vigne au vin - un chemin difficile », Journal Asiatique, tome 289, n° 2, p. 285-327. « La vigne et le vin en Chine. Misères et succès d'une tradition allogène » 2002 Deuxième partie : « Vins? vignes et vignerons de Tourfan », Journal Asiatique, tome 290, n° 2, p. 485-563. Zhang Guangda « Trois exemples d9influences mazdéennes dans la Chine des Tang Etudes 1994 chinoises, voL XIII, n° 1-2, p. 203-219. Zhu Baoyong 朱 寶 鐺 et Zhang Kechang 章竟昌 Zhongguo jiujing 中國 酒經 (Des vins et des alcools en Chine), Shanghai, 2000 Shanghai wenhua chubanshe.

Tang

5

The Impact of the Silk Road Trade on a Local Community: The Turfan Oasis9 500-800 Valerie HANSEN

What was life like for the people who lived along the Siik Road?1 Absorbed by the movement of people, religions^ and trade goods5 we rarely pause to consider how the long-distance overland caravan trade affected the communities through which it passed. Traders must hâve frequented inns5 religions sites, markets, and financial institutions, we suppose, but in fact we know surprisingly little about the day-to-day functioning of the oasis économies ringing the Taklamakan Desert. For those interested in understanding the Silk Road trade of the Tang dynasty, the Turfan oasis offers the best case study. The Century from 640 (the Tang conquest of the independent Gaochang kingdom) to 755 (the Tang withdrawal from Central Asia fbllowing the An Lushan rébellion) marks the apogee of the Tang dynasty's involvement in Central Asia.2 Nearly two thousand documents span the period befdre Tang rule and continue after 640, when Gaochang was renamed Xizhou 西州.コ 1

This paper rests on a foundation of collaborative work done by The Silk Road Project: Reuniting Turfan's Scattered Treasures, of which I was the principal investigator. The Silk Road project brought together a team of twenty-five Chinese and American scholars working in the disciplines of archeology, history, art history, and religions studi es. The notes to this paper will eite the work of individual collaborators (particularly Jonathan Karam Skaff), but I would particularly like to thank Wu Zhen 吳震, who wrote a paper for the 1998 Conference at Yale (Wu Zhen, 1999). Professor Zhang Guangda 張廣達 and I later translated this article into English: Wu Zhen, 2002. I leamed much from Prof. Zhang's patient explanations and first encountered many of the materials cited in this article in Prof. Wu's article. The pioneering work in this fïeld remains Jiang Boqin 姜伯勤,1994. The most recent and most thorough studies are: Rong Xinjiang 榮新江,2001 ; Étienne de la Vaissière, 2002 (second édition 2004). Some of the artifacts and documents mentioned in this paper were on display in an exhibition curated by Li Jian, who provided me with much help when I was writing entries for her catalog: Li Jian (ed.); 2003. Fumihiko Kobayashi 小林文彦 provided crucial research support that was funded by the Council of East Asian Studies at Yale. Professor Yoshida Yutaka 吉田豊 met with me in the spring of 2002 and answered many email quieries since on a variety of topics. I am particularly gratefiil to him for introducing me to the work of Arakawa Masaharu 荒丿丨I 正日青, who has published many fine articles about the Silk Road trade and Turfan, of which the notes below eite only the most direetly applicable. 2 For a detailed treatment of Turfan's history, see Zhang Guangda and Rong Xinjiang, 1998. The earliest document found so fàr dates to 367 while the last is from 769. Chinese rule Turfon came to a final end in 803 with the conquest of the oasis by the Uighurs. 3 1 am currently writing a book about the Tang capital of Chang'an and the six Silk Road sites that have produced written materials: Niya (modern Minfeng in Xinjiang, China), Kizil (near Kucha, Xinjiang),

284

Les Sogdiens en Chine

The residents of Turfan had an unusual custom: they outfîtted their dead with paper boots, hats, belts, and shoes. And because paper in the Silk Road oasis was scarce, they recycled govemment documents, contracts, and other texts to make these funeral vestments. Nearly two thousand documents hâve been found in 205 tombs at the Astana graveyard.* 4 After disassembling the paper shoes and hats, scholars painstakingly deciphered the handwritten documents. The fragmentary nature of these documents means that they can be frustrating: just as one begins to piece together the course of events, a crucial name is missing, and frequently large chunks of text hâve been eut away to make shoe soles. Even so, the surviving contracts, dépositions given in legal disputes, and travel passes offer a fleeting but informative glimpse of the overland trade of the sixth through eighth centuries. These documents make it very clear that, among the varions Central Asian merchants active in Turfan, the Sogdians were the most

numerous.5 To assess the impact of the Silk Road trade on Turfan, this paper will analyze the oasis’s residents in concentric rings. At the center stand the small group of Sogdian merchants who worked fiill-time as traders. Never long in Turfan, they were constantly on the move with their goods and their households from one oasis to the next. Local officiais referred to them as sojouming West Asian merchants (literally “Non-Chinese ‘Hu’ merchants conducting business” xingsheng hushang ÂÉÈ.Â0®, often shortened to xinghu).6 They must hâve coined this terni in response to local conditions because it occurs neither in the official historiés nor in The Tang Code. Surrounding this core of Silk Road merchants were the residents of Turfan who had frequent contact with them and whose livelihood depended on the Silk Road trade. This group included govemment officiais who supervised and taxed the trade, their interpreters, the people who worked at the inns where the travelers stayed, religious practitioners, and maybe even prostitutes. Many of the Sogdian traders were adhérents of Zoroastrianism, named fbr the legendary prophet Zoroaster, (also spelled Zara¬ thustra), also called Mazdeism, a label derived from the name of the suprême deity, Ahura Mazda. Finally, in the outermost orbital are perhaps the most interesting, yet least documented, group. These people, who comprised the bulk of the cultivators of Turfan, had occasional contact with the Silk Road traders but eamed their living independent of them. When they borrowed money or purchased goods (often animais or slaves) from Mount Mugh (outside Samarkand, in modem Tajikistan), Turfan (Xinjiang), Khotan (Xinjiang), and Dunhuang (Gansu, China). 4 The documents excavated by the Chinese from Turfan were published in a preliminary ten-volume set, and in a revised set of four volumes with photographie plates, ail of which are now available. Unfortunately, both sets are called Tulufan chutu wenshu niSStljiXW (Excavated documents from Turfan). This paper cites both the ten-volume set (TCWS-texts) and the four-volume set including photographs (TC WS-photos). From 1996 to 1998, Yale’s Silk Road project compiled a bilingual Chinese-English finding guide to over 3,000. artifacts and documents from Turfan. Web-site address: www.yale.edu:8084/turfan/ This URL is not always stable. An altemate route to the site is: www.yale.edu/ycias/ceas. Select "Research." Select "The Silk Road Database." 5 Skaff, 2003. 6 A computer search at Peking University showed that this terni occurs nowhere in the twenty-five historiés nor the Siku quanshu. Thanks to Deng Xiaonan and Gao Keli for their assistance.

The impact of the Silk Road trade

Valerie Hansen

285

the Silk Road traders, they usually drew up contacts.7 Sixth- and seventh-century contracts for even the smallest amount show the influence of Silk Road commercial consciousness: many prices are recorded in silver coins minted by the Sasanian dynasty

(reigned 224-651, in modem-day Iran), and contracts charge a ten percent penalty each

month.8 The Ethnie Composition of Turf an

By the sixth Century Turfan’s multi-cultural population was a mix of Chinese and

non-Chinese peoples who had settled there during the preceding centuries. The Silk Road merchants, those who worked fiill-time with them, and those who did not, drew their members from both the Chinese and non-Chinese residents of Turfan. The original inhabitants of Turfan were a semi-nomadic people who did not hâve an indigenous System of writing. The dynastie history of the Han reports that they “lived in felt-tents, kept moving in pursuit of water and grass for grazing, and had a fair knowledge of farming”.9 Artifacts found in early, non-Chinese style tombs at Jiaohe resemble those of the nomadic peoples to the west (whom archeologists conversant with Soviet terminology often refer to as Sarmatians).10 We do not know what language they spoke or what they called themselves. The Chinese called them the Jushi or peoples, and these peoples often adopted the sumame Ju $ when they took Gushi Chinese names. Starting in the third and fourth centuries, waves of Chinese migrants moved to Turfan and absorbed, displaced, or conquered the indigenous peoples. The Chinese historical record contains almost no mention of the oasis’s original residents. When the non-Chinese Qu family established the Gaochang Kingdom around 500,” Turfan’s population was largely Chinese. The Gaochang rulers followed Chinese models for almost every measure they imposed on their independent kingdom. Serving a govemment whose political structure mirrored the Chinese state, Gaochang officiais used Chinese as the language of administration. The capital city at Gaochang was surrounded by walls with named gates on four sides, just like other Chinese cities. Many of the city’s residents spoke Chinese at home, their children studied Confucian texts at school (possibly with glosses in local languages),12 and their king was a devout Buddhist. When the Tang armies conquered Turfan in 640, they took over a kingdom that was so culturally Chinese that its name in the Sogdian language (a language spoken in the 7

In an earlier study, I analyzed the Turfan contracts for what they showed about populär conceptions of law: Hansen, 1995. That study and this paper - dépend heavily on Yamamoto Tatsuro and Ikeda On’s usefiil édition that transcribes and annotâtes contracts from Turfan and Dunhuang: Yamamoto Tatsuro and Ikeda On, 1987. 8 Éric Trombert makes this point about the advanced commercial consciousness of the Turfan contracts by contrasting them with those from Dunhuang in his séminal study: Trombert, 1995, p. 190. 9 Zhang and Rong, 1’998, p. 14 n. 1, citing Hou Hanshu (Zhonghua shuju édition) 88:2928-2929. 10 Wang Binghua, 1 999, pp. 58-64. Personal communication, Frantz Grenet, July 2001. 11 499 or 502 are other possible start dates. l2Moriyasu, 1995.

—

286

Les Sogdiens en Chine

région of Samarkand that was distinct from, but closely related to, Middle Persian) was “Chinatown” or Town of the Chinese.13 After the conquest, the Tang govemment introduced the equal-fîeld System then in effect throughout the rest of the empire and renamed Turfan Xizhou Before redistributing the land, Tang officiais compiled household registers. Their census listed the population of the kingdom as 8,046 households with 37,738 residents and 4,300 horses in three préfectures, five sub-prefectures, and twenty-two cities.14 This figure recorded the population of the entire Gaochang kingdom, not just the capital at Gaochang city. Some ten years earlier, the Gaochang king had boasted to the pilgrim Xuanzang that several thousand clerics lived in his kingdom.15 (Unlike the Sui and the Tang dynasties, the Gaochang rulers did not exempt monks and nuns from taxes, a further indication that clerics formed a large sector of the population.) With a population of nearly forty thousand people, Gaochang city and its satellite Settlements would hâve been one of the largest, if not the largest, trading centers on the Silk Road. Interestingly, the Tang census did not record the ethnie identities or native languages of Turfan’s mixed population. Nor do the varions household registers that survive distinguish between the Chinese and non-Chinese residents of the oasis, an indi¬ cation that a black- and-white distinction to us was more variegated to the people of the

time.16 Historians of Turfan hâve devoted considérable energy to identifying the different ethnie groups resident in Turfan in spite of the limited social and cultural information and in spite of the very few pictorial représentations of non-Chinese. Of the various tomb figurines from Turfan that hâve been published, three male figurines17 and two detached heads are clearly non-Chinese. Two of these grooms [see fig. 1-2], whose hands hâve a hole for a rope, led ceramic 13

The 639 contact for the purchase of a slave (cited below in note 41) refers to Turfan as Chinatown. Other Sogdian texts also refer to Turfan in this way: Gershevitch, 1954, p. 158, item 1040. The anonymous tenth-century geography in Persian, Hudûd al- 'Âlam, calls Turfan Chïnanjkanth (trans. V. Minorsky, 1937, p. 271). (Both references provided by Yoshida Yutaka ’s article). 14 Tang huiyao (Zhonghua shuju édition) 95:1701-1702; Jiu Tangshu (Zhonghua shuju édition) 198: 5295, which rounds out the Tang huiyao figures. 15 Hansen, 1998, see page 58. The Chinese translation of this article, done by Huang Shih-shan, appeared in Dunhuang Tulufan yanjiu 4 (1999):17-38. 16 See, for example, the census register from 707 for Chonghua register (TWCS VII: 468-487.) Contrast with Rudelson, 1997, Table 4.1, “Population Growth in Turpan City, 1912-1985”, (p. 101), which lists Uyghurs, Hans, Tungans (Chinese Muslims), and others. 17 Wu Zhen, 2002, describes these figures. Two figurines of grooms were found in tomb 206 at Astana. One, with a green robe, is published in Xinjiang Weiwuer zizhiqu bowuguan, 1991, figure 122 [see fig. 2 p. 288 in this volume]. The other, with a brown robe, is published in Xinjiang chutu wenwu, 1975, figure 122 [see fig. 1 p. 288 in this volume]. The third non-Chinese groom stands a füll 1.10 meters tall and comes from tomb 216, and dates to the first half of the eighth Century. Xinjiang Weiwuer zizhiqu bowuguan, figure 116 [see fig. 4 p. 288 in this volume]. The two heads were found at Astana tomb 336 and date to 690-704. See Xinjiang chutu wenwu (Shanghai: Wenwu chubanshe, 1975), figure 127 [see fig. 3 p. 288]. The tomb guardian, 86 cm tall, is from Astana tomb 224 and dates to sometime during the Tang dynasty, Xinjiang Weiwuer zizhiqu bowuguan, figure 126 [see fig. 5 p. 288 in this volume]. Yoshida Yutaka and Kageyama Etsuko (personal communication) suggest that the groom excavated from Karaldioja in 1978 (75TKM98:7, shown in Wenwu 6 (1978): 13, figure 21) may also be Sogdian.

Valerie Hansen

The impact of the Silk Road trade

287

or mud figurines of camels or horses laden with goods for the next world.18 Standing 56 cm. high, they wear boots, colorful robes, and distinctive headgear, whether a tall pointed hat with a design or a rounded feit cap with the brim rolled up. Their facial features are exaggerated: dark eyebrows, big noses, and heavy beards mark them as non-Chinese. Both were buried in tomb 206 at Astana and date to either 633, when the husband was buried, or more likely to 689, when his wife was. Several of the female dancer figurines from this tomb hâve arms made of pawn tickets from Changan, a clear indication that they - and possibly these non-Chinese grooms - were also manufacturée! there. 19 The other Sogdian figures were probably made locally [see fig. 3-5], A tomb guardian provides the most extreme example of Stereotyping [see fig. 5]. The craftsman who made him topped his panther’s body and cloven hoofs with a noticeably hairy non-Chinese face. We should not assume that these figurines accurately depict the non-Chinese residents of Turfan. Not portraits of real-life Sogdian residents of Turfan, they are rather the créations of craftsmen deploying familiär stereotypes. With the exception of these figurines, the only clue to the ethnie identity of the Turfan population is their Chinese-language names. Tang-dynasty sources refer to people from Sogdiana (the région around Samarkand that straddles modem-day Tajikistan and Uzbekistan) as the nine jeweled clans Çzhaowu jiuxing even though most lists do not include exaetly nine different family names.20 The most common Sogdian sumames and their place of origin were:21 An Cao He Kang

Mi Shi Shi

18

w

»

*

Bukhara (modern Bukhara) Kabudhan, Gubdan (north of the Zerafshan River) Kushaniyah (between Samarkand and Bukhara) Samarkand (modern Samarkand) Maimurgh (either southeast of the Zerafshan River or

Panjikent)22 Kesh (modern Shahrisabz) Chach (modern Tashkent)

Knauer, 1998. It is also conceivable that the pawn tickets were shipped to Turfan from Chang’an and that these figures were made in Turfan. Angela Sheng, personal communication, April 2003. 20 Zhang Guangda, 1995. Cai Hongsheng, 1998, pp. 2-3. Étienne de la Vaissière explains that zhaowu was the Chinese transcription of the Sogdian word jamük, or “Jewel ”, (2002, p. 216). See now Yoshida, 2004. 21 Zhang Guangda and I drew up this table when we did our translation of Wu Zhen’s article (see note 1). 22 Scholarly opinion is divided about the location of Mi (Maimurgh) with most Russian scholars thinking it is to be identified with the present site of Kuldor-tepe, and Yoshida Yutaka convinced that it refers to Panjikent. See his detailed explanation in Kuwayama Shöshin ^LLlIEtt (ed.), 1992, pp. 163-166. 19

w

T

288

Les Sogdiens en Chine

Valerie Hansen

The impact of the Silk Road trade

289

Captions to the plates Fig. 1 (upper left): Turfan under the Tang dynasty, 689 or 633. Painted clay. Height 56 cm. Unearthed in 1972 from Tomb 206 at Astana, Turfan. This non-Chinese groom wears a white hat with a distinctive red diamond motif, a brown robe with red lining, and black boots. The figurine was fbund in a tomb that had been opened twice: once in 633, when the husband, Zhang Xiong, died, and then again in 689, with the death of his widow, who received a lavish funeral. Archeologists assume that this figurine was among the highquality goods made in Chang' an and placed in the tomb in 689. After Xinjiang Weiwuer zizhiqu bowuguan, 1991, fig. 122. Fig» 2 (upper center): Turfan under the Tang dynasty, 689 or 633. Painted clay. Height 56 cm. Unearthed in 1972 from Tomb 206 at Astana, Turfan. Excavated from the same tomb as that in Figure 1, and presumably also made in Chang’an, this figurine is very similar except that the figurine’s robe is green and his brimmed hat has no motif. The visible hole in his left hand indicates that he originally grasped a rope lead for either a camel

or a horse. After Xinjiang chutu wenwu, 1975, fig. 122. Fig. 3 (center and bottom left): Turfan under the Tang Dynasty, 690-704. Painted clay. Height 25,5 cm (above), 26,8 cm (below). Unearthed in 1960 from Tomb 336 at Astana, Turfan. These distinctive faces, all that remain from two clay figurines, show the range of head gear wom by Central Asians : the figure above wears a white round cap with an uptumed brim, while the bearbed man below wears a black kerchief, populär head covering wom by many Chinese men during the Tang. After Xinjiang chutu wenwu, 1975, fig. 127.

Fig. 4 (upper right): Turfan under the Tang dynasty, 713-755. Painted clay. Height 110 cm. Unearthed in 1972 from Tomb 216 at Astana, Turfan. Twice as tall as most of the tomb figurines fbund at Astana, this groom displays all the classic features of Central Asians : their clothing (a brown robe and black boots) and their facial features (heavy beard and long mustache, deep eye sockets, and large nose). After Xinjiang Weiwuer zizhiqu bowuguan, 1991, fig. 116.

Fig. 5 (bottom right) : Turfan under the Tang Dynasty, 640-907. Painted clay. Height 86 cm. Unearthed in 1972 from Tomb 224 at Astana, Turfan. This striking tomb guardian has the body of a panther but the head of a warrior with a distinctive helmet. The thick eyebrows, large eyes, and heavy beard all appear to be drawn from Central Asian prototypes ; After Xinjiang Weiwuer zizhiqu bowuguan 1991, fig. 126.

290

—

Les Sogdiens en Chine

In many cases people with zhaowu sumames also hâve first names that clearly hâve been transliterated from another language. Yoshida Yutaka has begun to work in this promising, but difficult field, and has identifïed some of the most common Sogdian names in Chinese and their possible Iranian reconstructions. See the Appendix by Yoshida Yutaka: Sogdian Names in Chinese characters, reconstructed Sogdian pronunciation, and English meanings (hereafter, p. 305). As is evident from Professor Yoshida’s chart, Sogdian children were often named guardian deity of the day of the month on which they were bom, and many of the for these deities were associated with Mazdean, or Zoroastrian, beliefs. (Scholars writing about the Sogdians in China tend to prefer the tenu “Mazdean”, derived from the name of the suprême deity Ahura Mazda, over the terni “Zoroastrian”, which they reserve for the less polytheistic form of the religion.) The Sogdian names in this list provide a snapshot of the deities thought most important by the Mazdeans resident in China. They are an eclectic group drawn from Iranian (Mithra and Räm), Mesopotamian (Nanai), and Indian (Buddha) traditions. (The name Jesus probably reflects Manichean - not Christian - influence because the prophet Mani acknowledged Jesus as one of his

predecessors.)23 Of those people whose identities we can surmise on the basis of their Chinese names, the vast majority hailed from Sogdiana. In this respect Turfan was no different than any other Chinese town, whether on the overland trade routes or in the interior. Every trading town on the Silk Road and in the interior of China had its own resident community of Sogdians.24

The Core: the Silk Road Traders

The fïrst detailed information that survives about the merchants moving along the Silk Road is a group of thirty-seven separate tax receipts recording payments made over the course of a year, probably around 600, at a single tax station outside Turfan.25 Cut out from shoe soles, the receipts contain gaps and are not continuous. Local officiais tallied the taxes they collected every fifteen days, and recorded the number of silver coins they had collected. On nine separate fifteen-day periods, spread out between the fïrst and twelfth months, they collected no coins at ail, an indication that the traffîc at this particular tax station fluctuated.

23

The name meaning “favor given by Jesus” appears on an unpublished document (Ch/U 6225) in Berlin. Professor Yoshida proposes the following date: “as for the date of Ch/U 6225, the Chinese text seems to be dated to the 8th Century, but the Sogdian text was written by a Manichaean scribe and is most likely to go back to the lOth Century” (email message, dated April 8, 2002). See also Yoshida, 1998, p. 47 n. 33. 24 Rong Xinjiang, 1999, expanded version in Rong, 2001, pp. 37-111. Professor Rong identifies twenty-eight Chinese cities with Sogdian districts. 25 TCWS-texts 3: 318-325; TCWS-photos I: 450-453.

Valerie Hansen

The impact of the SilkRoad trade

291

As Jonathan Karam Skaff has brilliantly explained, 26 the Gaochang kingdom collected a scale-fee tax every time someone bought a good that had to be weighed. For each transaction, tax officiais recorded the type of good purchased, its weight, the amount of the tax levied, the date of the transaction, and the names of the buyers and sellers. The merchants paid the tax with silver coins minted by the Sasanian empire. Famed for their purity, these silver coins enjoyed wide use in Turfan. Forty-one of the forty-nine people have Sogdian sumames, clear evidence of Sog¬ dian dominance of the Silk Road trade in this early period. Of the eight merchants who did not have characteristic Sogdian names, two (both named Bai |ÈI) were from the oasis of Kucha, which lay to the west of Turfan on the northeni Silk Route, three (two 27 one named Ju were descendants of the indigenous peoples of the named Di one was Chinese région, one was Turkish (Gongqin Daguan [=Tarqan] ^), H) (A one and identified.28 cannot be (Ning The scale-fee receipts reveal that most of the transactions involved five key Com¬ modities: spices (9 times), gold (6 times), silver (6 times), silk thread (5 times), and ammonium Chloride (6 times). Ammonium Chloride was used as a medicine or as a flux, both in the melting of metals and in dyes. The merchants traded other Com¬ medicine, copper, the spice turmeric, modities only once that year: brass "oushi and sugar. Quantifies ranged from quite large (800 jin fr, equal to perhaps 500 equal to 160 grams of gold), with kilograms of spice) to equally small (4 Hang roughly one-third over 100 jin. These documents offer one major surprise. Not one of the merchants bought silk cloth! Because silk was sold by length, and not by weight, it was not subject to the ‘scale fee' tax. The omission of silk provides a useful reminder. The scale-fee tax receipts do not cover sales of animais or slaves, two of the most frequently traded goods on the Silk Road. We should not leap to the conclusion that ail Silk Road traders were Sogdian. Other accounts, to be discussed below, reveal the participation of Chinese merchants in the Silk Road trade, and, by the eighth Century, more and more people with distinctly Turkic names appear, evidence of increasing Turkic influence in Central Asia in the seventh and eighth centuries. The Silk Road merchants readily formed partnerships with one another, we leam from a fascinating sériés of affidavits about a trade dispute between a Chinese merchant and his Sogdian partner that occurred around 670.29 Had the Chinese merchant Li 26

Jonathan Karam Skaff has done the most thorough study of these documents. See Skaff, 1998. Skaff has transferred the quantitative data from the documents to the very useful Table 5, entitled “Goods Traded in Seventh-centuiy Turfan”, on page 91. 27 Yoshida Yutaka (personal communication) notes that Di was the sumame used by one or two Turkish tribes. Some Sogdians may have adopted it as their sumame. "29 Skaff, 1998, p. 94. TCWS-texts, 6:470-479; TCWS-photos, III: 242-247. These documents are the subject of an entire article: Huang Huixian, 1983. Huang explains that this is probably a draft of a document because it lacks the seal of the sub-prefecture, those testifying did not give their fingermarks, and there are a few mi staken characters. The merchants first brought the case to the authorities in the Xizhou area headquarters, but they refèrred it to the Gaochang sub-prefectural authorities, who then submitted the case to the Anxi Protectorate General. The dépositions bear no date and can only be dated on the basis of other documents in the same tomb, which span 665-673. In 670, the Tibetans took Kucha, which had been the seat of the

292

Les Sogdiens en Chine

Shaojin borrowed 275 bolts of silk and failed to repay his Sogdian partner? Or, was Merchant Li telling the truth when he denied borrowing the silk? The Sogdian partner’s death made it even more diffïcult to détermine what had happened.

The case involved five Silk Road merchants, none of them resident in Turfan:

'

a Chinese merchant resident in (also called Li San 1. Li Shaojin Chang’an (jingshi Han 2. Cao Lushan/Rokhshan WWlM a thirty-year old Sogdian merchant, also resident in Chang’an; 3. Cao Yanyan W 3^.® , his deceased older brother, a non-resident Sogdian merchant (also called Cao Er W— ?), two Cao Bisuo 4 & 5. Cao Guoyi Sogdian, merchants, temporarily resident in Chang’an, where their families lived (hu, ke jingshi, you jiakou zai The géographie range of these merchants operations is impressive. Based in Chang’an, Li Shaojin and Cao Yanyan formed a partnership in Gongyuecheng (modem-day Almaligh), which lies some 2500 kilometers to the west in the Yili River basin close to modern China ’s border with Kazakstan.* 30 In Gongyuecheng, Merchant Li borrowed 275 bolts of silk from Cao Yanyan, Rokhshan testified, and the two men, who had no common language, agreed to meet in Kucha, some 300 kilometers to the south. At the time of their parting, the Sogdian was leading two camels, four cattle, and one donkey who carried his wares: silk, bowls, saddles, bows and arrows. The Sogdian merchant never arrived at his destination. One witness speculated that he might have died at the hands of Turkish bandits who wanted his cargo of weapons and saddles. Not surprisingly, the Chinese merchant never paid back the 275 bolts of silk he had

borrowed from the dead man. Then, sometime between 665 and 673, and probably before the Tibetan incursions of 670, Cao Rokhshan brought a complaint before the authorities in Turfan on behalf of his deceased brother. His name marks him clearly as a Sogdian; Cao was a sumame Chinese assigned to Sogdians who lived north of the Zerafshan River (which runs through Samarkand), and Lushan was the Chinese transcription of the Sogdian name Rokhshan “bright”, the masculine équivalent of the English “Roxanne”. Of course, this was An Lushan ’s name, too.31 In his affidavit the Merchant Li denied borrowing anything from his Sogdian partner. But then the court officiais confronted him with the testimony of two Sogdian Anxi Protectorate General since 649. From 670-679, Turfan was the seat of the Anxi Protectorate. I agree which could with Huang that Cao’s death probably occurred before the disturbances of 670. Anxi refer to either Kucha or Turfan, is used inconsistently - perhaps because those taking down the déposition were describing events before the change in name? Arakawa Masaharu, 1997 has also studied these documents. See a partial translation in this volume, p. 236. 30 Skaff, 2003 divides the Silk Road itineraries into long-, medium-, and short-haul routes and sees Cao and Li as traveling on long-haul routes (p. 507). 31 Cai Hongsheng, 1998, pp. 38-9 gives several examples of people named Rokhshan, all of whom he suggests came from low social strata.

Valerie Hansen

The impact of the Silk Road trade

293

merchants who had witnessed the original Ioan of the 275 bolts of silk at Gongyuecheng. Although the copy of the contract belonging to the deceased Sogdian partner had disappeared, and although the Chinese merchant denied it, the two Sogdian witnesses vouched that the Chinese merchant had indeed borrowed the silk. According to Tang law, their testimony had the same legal standing as a copy of the contract. Ruling in Rokhshan’s favor, the court ordered the Merchant Li to pay back 275 bolts of silk in addition to interest. We have no way of knowing whether the Chinese partner ever paid the younger brother back, because the documents come to an abrupt stop. Did other Silk Road merchants travel distances as great as those mentioned in this case? Absolutely. Map 1: Ruins of Gaochang city, with a Each time a merchant caravan crossed detailed plan of the market zone. through a pass, local officiais were required to check the caravan ’s travel pass, called a guosuo in Order to certify that each member of the caravan, whether human or animal, belonged to person whose name appeared on the pass. Family members could travel together, and several of the documents give the reason for travel as bringing family members to Chang’an.32 Some of the travelers are classed as zuoren ff A a dépendent laborer whose status was higher than that of a slave because zuoren could not be bought and sold.33 Slaves, both male or female, required a market certificate showing that they had been bought legally - unless they were bom to a slave already belonging to the family - as did all draft animais like cattle and horses that could be used by the military. In one case the travel document listed the age and color of eight different horses.34 Twelve travel passes found at Turfan confirm that the deceased Sogdian merchant’s party was indeed typical. Merchants often traveled with a dozen or so human companions and as many draft animais. The travel passes, unfbrtunately, do not list their Caravans cargo, but Cao Yanyan’s load of silk, bowls, saddles, bows and arrows gives some indication of the items a caravan might have carried. The distance from 32

TWCS-texts, 7: 88, 9: 31-69; TWCS-photos, III: 346-350; IV: 268-296. These documents have been throughly studied by Cheng Xilin, 2000 ; Arakawa Masaharu, 2001. 33 Sheng, 1998, p. 140. Arakawa, 2001, p. 17, spéculâtes that the zuoren whose names appear on the guosuo documents were not private traders but part of a “ military supply transport team”. 34TCWS-texts,9:33; TCWS-photos, IV: 268.

294

__

Les Sogdiens en Chine

Almaligh to Chang’an, great as it was, was equalled by other merchant’s itineraries: one wanted to travei all the way to Fuzhou, man with a Chinese name (Tang Yilian Fujian, from Turfan in 733 in the company of one male slave and two female slaves (thirteen and fourteen years old).35 One guosuo document, also dated 733, records the for why he twice visited one place on his reasons given by Wang Fengxian permitted itinerary: to pursue a debtor. The authorities drew up a new guosuo document for him since merchants were not permitted to diverge from their prearranged itineraries unless they obtained official permission. How big were the caravans traveling on the Silk Road? As with so many other basic questions about the Silk Road trade, the fragmentary information is subject to inter¬ prétation. Buddhist legends translatée! into Chinese in the fourth and fifth centuries often speak of groups of 500 merchants, clearly a stereotype, but possibly an infonued one. One or two caravans of several hundred people are mentioned in preTang sources, but caravans going to and from Turfan during the Tang seem to hâve been smaller, often a dozen people or so. The guosuo documents discussed above indicate that these small parties sometimes traveled together, forming groups of fifty people. Larger groups may hâve formed when Crossing greater obstacles such as the Pamir Mountains or long Stretches of desert. Smaller caravans may indicate greater security: because the Tang exercised greater control over the major routes than had its predecessors, merchants feit safe even in smaller convoys.36

First Orbital: Those Working Full-time with the Silk Road Merchants

As the travei passes show, govemment officiais, usually working for the Section of the People (Hu Cao p§37) or market Supervisors (shi yi rÇ®) watched over the movement of these caravans, drew up the market certificates required by The Tang Code each time an animal or slave was sold, and heard complaints like that brought by Cao Lushan. These officiais employed Interpreters when communication was im¬ possible. And they sent merchants to inns and doctors when they feil ill and even disposed of their corpses in the event of their deaths. Few documents describe the marketplace at Turfan where the Silk Road traders gathered (Xuanzang, who spent most of his time in the king’s palace and a monastery, 35

TWCS-texts, 9:31; TWCS-photos, IV: 268. Thanks to both Yoshida Yutaka and Arakawa Masaharu, who discussed the question of caravans with me during the Conference on April 22, 2004. Albert E. Dien discusses two early examples in his Conference paper, “Caravans and Caravan Leaders in Palmyra”, see above p. 195. The official history of the Wei dynasty mentions that in 439 ‘many’ merchants were among the 30,000 household forcibly moved to the Northern Wei capital of Datong (Wei Shu 4A: 90; 102: 2270. The official history of the Zhou reports that 240 merchants living in Liangzhou were captured in 439 {Zhou Shu 50: 913). These merchants could hâve been residents of Liangzhou and were not necessarily en route. Rong Xinjiang discusses the Buddhist legends in Rong, 2001, pp. 111-116. 37 www.yale.edu:8084/turfan/govemment.html. This URL is not always stable. An altemate route to the site is: www.yale.edu/ycias/ceas. Select "Research." Select "The Silk Road Database." Select "Govemment Structure." (This chart was drawn up by Zhang Guangda). 36

-

Valerie Hansen

The impact of the Silk Road trade

295

says nothing at all about a market.) The most likely location for the market was in the Southwest comer of Gaochang city, near the ruins of the monastery that still stand today (Map 1). Archeologists who have surveyed the ruins of Gaochang city have tentatively identified two open areas near to the ruins of the large monastery in the Southwest comer of the city as plausible candidates. Rows of houses flank the marketplaces, with sub-districts of the the remains of stalls (hang fï) and the walls dividing the fang Tang dynasty still visible (no fang gates have been found). Some of the walls are quite close together, suggesting these were originally workshops, which then supplied the merchants who sold similar goods in the same rows in the markets.38 The sole Sogdian-language contract found in the Astana graveyard indirectly sug= gests that a central market existed and that it was supervisée! by an official of the Gao¬ chang Kingdom. It records the sale of a girl from Samarkand to a Chinese sramana (usually translatée! as ‘monk’ but the term may simply indicate some kind of Buddhist adhèrent39) named Zhang for 120 silver coins in 639, just one year before the Tang conquest.40 The contract spells out the rights of the purchaser in grisly detail: the new owner can beat his slave, maltreat her, tie her up, seil her, hold her hostage, give her as a gift, or do whatever he likes with her.41 Interestingly the contract does not mention any sexual activity, since it omits any task a slave might normally be expected to perfbrm. In listing those who were bound to recognize the transaction - non-resident mer¬ chants, householders, the king, and high officiais - the text offers a snapshot of Turfan society on the eve of the Chinese invasion. The contract closes with the names of the witnesses as well as the recording official, an administrator of the Gaochang kingdom who oversaw the Sogdian Community. The contract does not say so explicitly, but it makes perfect sense that such an official would have an office in or near the market where the Sogdian traders were most active. The officiais of the Gaochang kingdom were responsible not just for monitoring private traders, like those described above, but also for hosting envoys from the different regional kingdoms of Central Asia as well as from Central China. One document from Turfan lists the Commodities sold to the Gaochang kingdom govemment officiais, which included a form of brass, feit rugs, Persian brocades, and gold. This list does not specify the sellers, names, but it seems most likely that they were the envoys themselves,42 for these envoys often engaged in private trade at the same time as they conducted official business. As a resuit, the distinctions among private trade and official tribute missions were blurred. 38

Yan Wenru, 1962. Reference provided by Sheng, 1998, particularly pp. 135-136. For a description of Tang cities, see Heng, 1999. Heng suggests that, even though Contemporary with the Song, the Liao city of Beijing adhered to the Tang layout with fang sub-district walls and gates (p. 208), and it is possible that Gaochang retained this layout even under the centuries of Uighur rule. 39 The term is used similarly in the Niya documents. See Hansen, 2004. 40 Xinjiang Uighur Autonomous Museum, 1988: Yoshida Yutaka EHs, Moriyasu Takao 1-50. Published in English: Yoshida Yutaka, 2003. In the interest of full disclosure, I confess that I do not read Sogdian but have depended on Yoshida's translation and copions notes. 41 Yoshida points out the similar wording of the Kharosthï contracts found at Niya, pp. 21-22. 42 Rong Xinjiang, 200 1, p. 187.

296

_

Les Sogdiens en Chine

After the Chinese conquest of 640, Turfan was no longer an independent kingdom but Xizhou, one of three hundred préfectures in Tang China. As such, it could no longer host envoys from different Central Asian States, and the Turfan merchants became subject to the provisions of The Tang Code. Ail sales of animais and slaves had to be registered on a market certificate, and only market officiais had the legal authority to issue such certificates. The market Supervisors inspected ail market stalls every ten days to make sure that accurate weights were in use and that merchants charged prices within the range stipulated by the govemment.43 Chinese officiais depended Interpreters to translate for them, but, because tliey did not always record their names or even mention them, we do not know whether or not Interpreters worked foll-time for the govemment.44 Unusually, one interpréter’s name appears on a travel pass, dated 685, issued to five men traveling east together. At the start of their trip, they had been unable to obtain a travel pass because no officiais had been available.45 The beginning of the document is tom, but the appearance of the date, the interpréter’s name, and the name of the reporting official (who signs only one as was the practice in the Tang bureaucracy) suggest character of his name [Heng that only a few lines, if that, are missing. The interpréter ’s name Di Nanipan is clearly non-Chinese: Di is a stan¬ dard last name for the descendants of the indigenous Gaoju peoples, and three syllables suggests a non-Chinese first name as well.46 Each of the travelers draws three finger lines below his name (the équivalent of signing one’s name with an X in Europe47) to show that he was present and was subject to the legal jurisdicti on of the officiais issuing the pass. And so too does the interpréter, probably after checking the written form of the document to make sure that it matches the oral testimony he translated. At the end of the document, the travelers give their füll names and the names of those traveling with them: two are named Kang, one He, and two identify themselves as Tokharians (Tuhuoluo), or residents of Tokharistan, or Bactria in northwest Afgha¬ nistan, which had been conquered by the Turks in the early seventh Century. Tang law required them to give the names of five guarantors, whose places of résidence cover a (Kami), Yanqi (Karashahr), large area: Tingzhou (Beshbalik), Yizhou places. these and Xizhou (Turfan itself). Surely these men traded in ail of Three of the meaning that their names five guarantors are identified as commoners (baixing 43

Twitchett, 1966. The Turfan database lists four documents that mention interpreters: very fragmentary dépositions about the purchase of silk on which the term yiyuren Win A appears (TCWS-texts, 6:70, 72; TCWS-photos, 111:38), the déposition of a Turkish maid servant who was asleep when her master's house was robbed (the interpréter is also named Di. TCWS-texts, 6:465; TCWS-photos, 111:239), the travel certificate described here (TCWS-texts, 7:88-94; TCWS-photos, III: 346-350), and a receipt written by Interpreter He Deli 'fôjig Turkic people. Duo Hai Da Gan for the sale of on behalf of the leader of the Tuqishi horses (TCWS-texts, 8:87; TCWS-photos, IV:41). 45 TCWS-texts, 7:88-94; TCWS-photos, III: 346-350. This document has been thoroughly studied by Cheng Xilin, 2000, pp. 62-80. 46 The name resembles those in Chart I whose meaning is ‘The glory of the female deity Nana'. 47 Hansen, 1995a, pp. 9-10. 44

Valerie Hansen

The impact of the Silk Road trade

297

were entered on the household registers of the towns listed and had the same legal status as other free people in the empire. Each man listed the slaves, the dépendent laborers (zuoren), women, and animais in his party. The interpréter Di Nanipan does not appear on these lists because he was not traveling with the caravan but was, instead, working for the Xizhou authorities in Turfan. Like interpréter s5 innkeepers provided a service crucial to the smooth fanctioning of the overland trade. Merchants often stored goods they could not carry with them in

inns5 which served as warehouses. Market officiais had close ties to local inn-keepers5 as we leam from an inquiry into the unexpected death of a visitor - maybe a merchant (his sumame is missing) in 643. 48 A Sogdian named He named Laifeng testified that local officiais had ordered him to provide Laifeng 's Shementuo meals and medical care, suggesting that He Shementuo was probably an innkeeper (The family name He dénotés those Sogdians from Kushaniyah, north of the Zerafshan River.) Even though the Sogdian He had called a doctor, Laifeng had died while under his care. The case was complicated by Shementuo’s failure to fill in the appropriate forms. At the end of his déposition, He gave the name of someone who could corroborate his testimony: a Mr. Kang (the most common Sogdian sumame). The case sub-district, feil under the jurisdiction of the official in charge of the Jieyi apparently the name of the fang where the market was located. The residents of Turfan definitely maintained inns, we leamed from a sériés of dépositions, dated 762, about a cart accident.49 Two eight-years-old children - a girl family - were playing in front of an from the Cao W family and a boy from the Shi lost control of when a driver Kang Shifen inn owned by Zhang Youhe his ox-cart. In his déposition Kang called himself a commoner of the Chumi tribe (Chumi buluo which lived in the northern Tianshan mountains and Sou¬ thern Jungarian Basin east and west of Beshbalik. His use of the sumame Kang is intriguing; it may indicated that some Sogdians had joined the Chumi tribe, or, more likely, it shows that some non-Sogdians used the sumame Kang.50 Fürther, he explained Jin Chennu that he had been hired by a temporary resident (xingke (whose name does not look Chinese either). Kang explained that the cart was not his own, that he had little experience in driving a cart, and that he had unintentionally wounded the two children when he lost control of the cart. The local authorities fbllowed the provisions of The Tang Code to the letter: Kang was ordered to take care of the children for fifty days. If, after the stipulated time, the children recovered, Kang could go free. If not, then he would receive the punishment appropriate for a murderer.51 What kind of burial would two non-Chinese children living in Turfan have received if they had died after the cart accident? What happened to the body of the merchant who died while in the innkeeper’s care? Given the recent spectacular finds of nonChinese tombs in Taiyuan and Chang’an, one cannot help wondering how the non-Chinese residents of Turfan disposed of their dead. The appearance of the corpses, 48

TCWS-texts, 6: 3-5; TCWS-photos, 111:2-4. TCWS-texts, 9:128-134 ; TCWS-photos, IV: 329-333. 5°Sk*2003,p. 481 n. 19. 51 Changsun Wuji, 1983, pp. 388-389. Johnson (trans.), 1997, II, pp. 333-334. 49

298

Les Sogdiens en Chine

the style of the tomb architecture, and the presence of Chinese-language documentation suggests that, with few exceptions, almost everyone buried in the Astana graveyards was Chinese.52 Several wooden slips hâve been found at Astana that hâve the Chinese (“substitute person” “in place of a person”) as well as something characters dairen difficult to read in Sogdian script.53 These slips suggest that the relatives of the deceased hoped to provide the dead with servants in the next world. Only bilingual people would hâve labelled them in both Chinese and Sogdian. Those commissioning the burial were probably Sogdians who had adopted many Chinese customs including Chinese-style burials. In the Sogdian heartland, in the centuries before the Islamic conquests of the eiglith Century, Mazdean believers feared that decaying human flesh would contaminate the earth so they buried only clean bones. The traditional Mazdean means of disposing of the dead was to expose corpses, to allow wild animais to eat the flesh from the bones, and then to place the bones in a container (an ossuary) for burial. Kageyama Etsuko has identified four ossuaries found in Xinjiang, two from Turfan.54 They were found at Toyok (Tuyugou) and their style suggests they date to the late seventh or early eighth centuries, the peak period of Sogdian-Chinese interaction. The strongest textual evidence for Mazdeism is also linked to Toyok. One of the most important Mazdean deities worshipped at Turfan was called the Heaven of Toyok (Dinggu tian T which Zhang Guangda thinks may be an altemate name for the god of victory, Verethraghna. The names of several Mazdean deities appear in a sériés of documents from the mid-sixth Century listing the dates on which animal sacrifices were made to them.55 The deities worshipped include the suprême deity Ahura Mazda (Dawu Amo and tree, rock, Weshparkar, the god of wind (Fengbo and mountain gods. The regularity of the sacrifices suggests that a group of Mazdean priests - working full-time? - lived in Turfan and conducted religious services for the Sogdian residents. The first mention of a Mazdean temple at Turfan dates to a Buddhist colophon from the year 430 or 490." The Sogdian expérience in other cities indicates that as soon as the Sogdian community reached a certain size - perhaps just a hundred households - it named someone as sabao, who served as both political and religious leader. This flexibility allowed Sogdians to sustain their patterns of worship and to 52

Not everyone would agréé. I was present at one discussion where all the Chinese archeologists emphatically agreed that ail those buried at Astana were ethnically Chinese, but a Uighur archeologist present vehemently protested. 53 Rong Xinjiang, 2001, p. 186, citing Wenwu 1 (1981): 63-64. The Chinese archeological report suggests that the Sogdian reads làsi (‘servant’ in Turkish), but Yoshida Yutaka says kisi is an impossible reading. He (meaning unknown) as a possible reading. suggests s 54 1997; Tulufan diqu wenguansuo, 1986, pp. 87-89 describes two caves Kageyama Etsuko containing ossuaries as Buddhist, but Kageyama argues persuasively that these were in fact Mazdean burials. See also her paper in this volume p. 365. 55 Zhang Guangda, 2000. Chinese version Zhang, 1999. TCWS-texts, 2: 39; TCWS-photos, 1: 132. 56 There was also a Chinese god of the same name who appears often alongside Yushi the master of rain, in Turfan and Dunhuang documents (Éric Trombert, of the C. N. R. S. in Paris, email dated April 23, 2002). 57 Rong Xinjiang prefers the date 430, 2001, pp. 200-201, while Éric Trombert thinks 490 more likely (email dated April 23, 2002).

Valerie Hansen

The impact of the Silk Road trade

299

maintain their identity as a group even in the first years they moved to a new site.58 The existence of a foll-time Manichean priesthood at Turfan during the seventh and eighth centuries seems much less likely. The Chinese-language texts from Astana say little about Manichean beliefs among the non-Chinese population, but the four German expéditions to Turfan found several Manichean libraries at the beginning of the twentieth Century.59 The many manuscripts are not dated, but some use archaic liturgical languages like Parthian and Middle Persian.60 If these manuscripts were created and then stored in Turfan, a Manichean community existed there as early as the seventh and eighth centuries. But in recent years the scholarly consensus has shifted to suggest that Manicheism became firmly established at Turfan only after 803, when the Uighur kingdom took control of the oasis. Carbon-14 dating has been done on the paper of some of the illustrated manuscripts, and produced dates consonant with early tenth to mid-eleventh centuries. The few cave paintings with undeniably Manichean subject matter also date to this late period.61 In sum, the people who lived in Turfan but worked full-time with Silk Road merchants included the officiais who regulated the trade, either by issuing market certifîcates or travel passes, their interpreters, the inn-keepers, and Mazdean priests. One cannot help wondering whether the inns at Turfan provided sex workers with an opportunity to service the Silk Road merchants since the official historiés report that there were markets in women at both Kucha and Khotan.62 The Sogdian-language contract buried at the Astana graveyard (discussed above) demonstrates that at least one Chinese man bought a young Sogdian girl in 639. One of the archeologists who excavated the Astana site, Wu Zhen, contends that, although many households along the Silk Road bought individual slaves, as we can see in the earlier documents from Niya, the Turfàn documents point to a massive escalation in the volume of the slave trade.63 A name register from Emperor Wu’s reign (690-705) lists seventy-nine people who had been omitted from earlier registers for two households.64 The seventy-nine names include 1 musician of inferior status (yueshi 9 male and female personal retainers, also of inferior status (buqu, kenü and 68 female and male (nubi slaves Musicians and personal retainers belonged to the lowest legal category of The Tang Code, the inferior classes (jianmin K), and unlike slaves, could not be bought and sold.65 One of the two households contained 61 people, with twenty individuals under the âge of fifteen. The slaves hâve no sumames, but the personal retainers do, and several 58

See Hansen, 2003. Zhang and Rong, 1998, pp. 24-28. 60 They have been translated in Klimkeit, 1993. 61 Moriyasu Takao, 1991; Gulacsi, 2001, pp. 9-10. 62 Xin Tangshu 221a:6230. In addition, Susan Whitfield offers a fictionalized account of a Kuchean courtesan’s expériences in the ninth Century without providing any sources, although she has clearly drawn on the description of the prostitutes’ quarter in Chang’ an in Beilizhi; Whitfield, 1999, pp. 138-154. s Wu Zhen, 2000 (p. 154 is a Chinese-language rendering based on Yoshida’s Japanese translation of the Sogdian contract of 639). 64 TWCS-photos, 3: 525-529. 65 Johnson, 1997, 1, pp. 28-29. 59

300

Les Sogdiens en Chine

_

are Central Asian; the slaves, names look as though they were transliterated into Chi= nese from another language, quite possibly Sogdian. The list is intriguing: what were ail these people doing in a single household? Wu Zhen offers a daring reading: “of course the male and female slaves could hâve undergone training of various types - like listening and speaking basic Chinese, becoming familiär with Chinese mann ers and customs, and even leaming some types of tasks - at the hands of the musicians and personal retainers. The goal was to increase the sale price of these slaves”. 66 It is also possible that the members of the inferior classes were also for sale (contrary to the provisions of The Tang Code) and not there simply to train the slaves. Still the large size of the household strongly supports Wu Zhen s hunch that this was a slave™ producing establishment. The few documented pairings of Chinese male owners with young Sogdian girls raise the question how often Sogdian and Chinese families intermarried. The historical record is largely silent on this topic, but Rong Xinjiang has found throughout Tang-dynasty China a total of twenty-one recorded marriages in the seventh Century in which one partner was Sogdian, and in eighteen cases, the spouse is also Sogdian. The only exceptions are very high-ranking Sogdian officiais who married Chinese wives.67 He concludes that most Sogdian men took Sogdian wives, and we may surmise that the pairings between Chinese men and Sogdian women were usually between a Chinese male master and a Sogdian female slave. 5

Outermost Orbital: Those Least Affected by the Silk Road Trade

Of the 212 Turfan contracts listed by Yamamoto Tatsuro and Ikeda On, only a handful can be linked to the long-distance exchanges of the Silk Road. Three of these were included with travel passes because Tang law stipulated that a caravan owner had to hâve documentary proof that he owned the slaves and animais traveling with him. These are not the actual market certifïcates required by law because they bear no official seals. These contracts expressly identify one of the parties to the contract as a non-Chinese merchant (xingsheng hu, see above discussion). Let us look at the three examples in chronological order. (1.) In 673 a company commander (duizheng ^TE) bought a camel for fourteen bolts of silk from Kang Wupoyan a non-resident merchant from Samar¬ kand (Kangzhou (2.) In 731 the Sogdian merchant Mi Lushan sold an eleven-year-old girl to a

66

Wu Zhen, 2000, p. 139. Rong Xinjiang, 2001, pp. 132-135. Of the twenty-one epitaphs, twelve are from Quan Tangwen buyi WXWÄ (Supplement to the complété writings of the Tang), fîve from Tangdai muzhi huibian (Collected epitaphs of the Tang), three were excavated at Guyuan, Ningxia, and one is from another site. 68 is a common ending for Sogdian fïrst names meaning ‘for the benefït of a certain deity. For Yan other examples, see Cai Hongsheng, 1998, p. 40. 69 Ikeda contract 29. 67

lllïlfn

Valerie Hansen

Thejmpact of the Silk Road trade

301

for forty bolts of silk.70 Five men served as resident of Chang’an, Tang Rong guarantors, vouching that she was not a free person who been enslaved Tang Code banned the enslavement of commoners.) Of the four with household registration in

Xizhou (Tang-dynasty Turfan), three had Sogdian last names (Shi, Cao, and Kang) and was probably a member of the Tuhuoïuo tribe, whose the one with the sumame Luo primary home was Tokharistan. The fifth guarantor, also with the sumame Kang, was designated a temporary resident {jizhu WtÈ) of Xizhou, an indication that he had not yet become a commoner and that his name was not yet entered on the household registers. (3.) In 733 a Sogdian commoner resident in Xizhou, Shi Randian bought a horse fbr eighteen bolts of silk from a Sogdian named Kang. Skaff has pieced together information from different documents to follow Shi Randian ’s route from Hami to Dunhuang, and concluded that he may have traveled the entire route specified by his travel permit from Guazhou, Gansu, to Kucha.71 When Shi Randian purchased the horse, three guarantors vouched that the horse was not stolen: one Tuhuoluo non¬ resident merchant, one non-resident merchant from Bukhara (sumamed An), and one commoner resident in Xizhou, a Sogdian named Shi.72 Although no document says so explicitly, it seems likely that the seller paid the guarantors a small fee to vouch for the legality of the goods being sold, since the guarantors were financially liable should the slave or animal in question tum out to be stolen. The presence of resident and non-resident Sogdian guarantors indicates that Sogdian trade networks included both non-resident merchants (xingsheng as well as those entered on the household registers of Xizhou and other localities as commoners (baixing). If it were not for the explicit labeling of the seller or guarantors as non-resident merchants, we would have no reason to dass these three transactions as part of the long-distance overland trade. After all, they simply record the sale of a single camel, slave, or horse. Evidence of the Silk Road trade is equally difïïcult to detect in the twenty-one labor contracts surviving from Turfen.73 Most specify the terms fbr hiring someone to perform someone else’s corvée labor obligations to perfbrm a watch on the beacon towers of the Tang, but two - unfortunately quite fragmentary - seventh-century contracts hire someone to transport Han silk to an unnamed destination.74 This type of silk, sometimes called “cooked silk” or “degummed silk” was ready to be dyed.75 Like Service on the beacon towers, the delivery of silk was probably a form of corvée labor - not a task performed for the benefit of private merchants active on the Silk Road. Two interest-free Ioans, made in 665, from a moneylender who usually exacted high interest hints - perhaps! - at a long-distance transaction.76 The contract was buried

—

70

Ikeda contract 31. Yoshida Yutaka and Arakawa Masaharu saw this document, which was clearly a copy of the original with space left for the places where the seals appeared. 71Skaff, 1998,p. 97. 72 Ikeda contract 32. 73 Ikeda contracts 190-211. 74 Ikeda contracts 209, 210. 75 Angela Sheng, email dated March 29, 2002. 76 Ikeda contract 74.

302

_

Les Sogdiens en Chine

(d. 673), whose tomb contains in the tomb of the moneylender, Zuo Chongxi fburteen other intact contracts, one of our best sources for understanding the Turfan economy.77 One loan, to a military guard named Zhang Haihuan was for forty-eight silver coins; the other, to Bai Huailuo was for twenty-four silver coins. If the two men failed to pay the money back within ten days, the contract authorizes the moneylender to confiscate Zhang Haihuan ’s house, possessions, or land as compensation. It seems likely that the three men formed a partnership with Zhang having twice as many shares as Bai (the moneylender’s share is not revealed). If the deal was successfully complétée!, then the moneylender eamed nothing from his partners. But if it was not, they had to pay him back at the prevailing interest rate in Turfan of 10% per month, four points higher than the maximum of six percent each month stipulated by The Tang Code?' Since Moneylender Zuo retained his copy of the contract, we can conclude that Zhang Haihuan and Bai Huailuo never paid back the money they had borrowed. The contract includes an extra line holding Zhang Haihuan’s mother, a female household head, responsible for his debt. It calls her a ‘big woman’ (danü a term that does not appear in the official historiés. Big women appear as the household head on govemment household registers because they bore responsibility for paying their family’s taxes when the male household heads were absent. A survey of seventy census documents from the period of Chinese occupation showed that between 16 and 30 percent of ail Turfan households were headed by such women. Several scholars hâve argued that households concealed the presence of men and gave the name of a woman as head because women paid lower taxes than men. It is equally likely that women managed these households because their husbands were genuinely away from home either to perform military service or to go on business trips.79 Even though they occurred at the peak of the Silk Road trade, the overwhelming majority of Turfan contracts document the day-to-day transactions of an agricultural community in which people buy, seil, and rent individual animais, slaves, or small plots of agricultural land or orchards. Many loans are for a small amount. Perhaps we should not be surprised. Even in today’s âge of high finance, most contracts are for the purchase of individual houses or cars.80 Even so, the Turfan contracts reflect the high degree of commercialization of the Turfan economy in the seventh and eighth centuries. I use the term commercialization to indicate that these transactions ail involved money - actually Sasanian silver coins and were not barter. In addition, the penalty for failure to fulfill the contract was 10% for each month, the same interest rate charged on commercial loans. In the third month of 668, a season when many who worked the land were short of money, the cultivator Zhang Shanxi signed a contract to borrow twenty silver coins from Zuo Chongxi. In 670 he went back to him again, this time to borrow forty 77

1 hâve discussed this tomb in Hansen, 1995a, pp. 33-39, and 1995b, pp. 59-66. Hansen, 1995a, p. 35. 79 Like the term xingsheng hu, the term for female household heads (danü documents, but not in the official historiés. See Deng Xiaonan, 1999: 85-103. 80 Personal communication from James Stepanek, March, 2002. 78

occurs in many Turfan

A Valerie Hansen

The impact of the Silk Road trade

303

silver coins. This was the same moneylender who lent Zhang Haihuan money for his ten-day business trip.81 Both men - one a long-distance trader, the other a farmer -

borrowed silver coins at the same interest rate of 10% per month.

Moneylenders like Chongxi bridge long-distance as served the trade a economy between and the local Zuo agricultural economy, and they charged the same high rate of interest to everyone, effectively pulling the local cultivators into the larger economy. If farmers like Zhang Haihuan wanted to borrow money, they had to do so at the same rate as Silk Road traders did. The Turfan contracts cl early document the shifts in the medium of exchange at Turfan.82 The earliest Turfan contract testify to the existence of a harter economy: in 273 a female household head (danü) bought a coffin for twenty bolts of degummed Han silk.83 Similar exchanges continue in the fourth and fifth centuries. The first mention of Sasanian silver coins in a real-life rental contract occurs in 584, when someone rented one sixth-acre (mu) for five silver coins.84 The use of Sasanian coins peaks in the Century from 550 to 650, but people continue to use both grain and silk along with silver coins throughout the seventh Century, in the years leading up to and following the Tang conquest of 640. Suddenly - just at the tum of the eighth Century - Turfan residents stop using Sasanian silver coins and adopt the use of Chinese bronze coins (that the contracts call copper coins and whose main constituent was indeed copper).85 Skaff suggests that several factors - the Tibetan occupation of the Tarim basin bet¬ ween 686 and 692, the Chinese infusion of monetary silk and bronze coins, and the growing Chinese preference for silver in ingot form - may account for the change.86 Although the reasons for the sudden shift continue to be debated, no one debates the immediacy of the change. Ail the residents of Turfan - both rich and poor - switched from silver to bronze coins over night, sure evidence of how embedded they are in the larger economy. Here, too, we can see the undocumented rôle of moneylenders as instrumental in introducing these changes. In the 670s people borrowed silver coins, ten or twenty at a time, but in 703 two different people borrow 320 bronze coins each, the équivalent of ten silver coins.87 (A tax receipt gives the exchange rate as 32 bronze «■■■BS 81

Hansen, 1995a, p. 36. This topic is throughly covered in several fine articles: Skaff, 1998 ; Thierry, 1995 and 2000 ; Zeymal, 1992. 83 Ikeda contract 1. 84 Ikeda 98; see Skaff, “Table of Dated Turfan Documents That Mention Silver Coins” in Skaff, 1998, pp. 108-109. François Thierry ingeniously draws on grave inventories excavated from Turfan to document the appearance of silver coins - he cites a grave inventory dated 543 TCWS-texts, 2:60 ; TCWS- photos, I: 143. But the grave inventory lists both gold and silver coins. It may in fact describe fictive currency - not actual coins in circulation at Turfan at the time - the living intended to send to the world of the dead. With exception of a few forged Byzantine gold coins, only silver coins hâve been fbund at Turfan. Peng Xinwei, 1994 explains: “The alloy used for [early Tang-dynasty] coins was then 83.32 percent 'Pper, 14.56 percent pewter and 2.12 percent black tin”, volume 1, 257. Skaff, 1998, 99-104. " Ikeda contracts 74-79, 89, 90. 82

Les Sogdiens en Chine

304

coins to 1 silver.88) And where cultivators had paid their rent in silver coins in the 670s they paid bronze coins in 703.89

Conclusion The stunning archeological finds of beautiful silks from Niya and the man with the gold mask from Yingpan (west of Lop Nor) in recent years hâve reinforced the conventional view that many rich merchants plied their wares along the Silk Road of the fïrst millennium of the Christian Era. But the excavated documents give an entirely different impression. The Kharosthï contracts from Niya90 and the loan contracts from Dunhuang (studied so thoroughly by Éric Trombert91) very clearly document the exis¬ tence of a large subsistence economy in which cultivators bartered for simple goods. Like the Turfan documents, the Niya and Dunhuang documents provide hardly any evidence of the fabled long-distance Silk Road trade in silk, gold, silver, jewels, and pearls. Turfan’s economy of the sixth to eighth centuries differed in important ways from Niya’s in the third and fourth Century and Dunhuang’s in the ninth and tenth centuries. It was much more commercialized - even in the subsistence transactions documented in surviving contracts. Éric Trombert has argued that the central govemment of the Tang played an enormous rôle in the rise of the Silk Road trade.92 Rather than see many low-level entrepreneurs crisscrossing the Tarim Basin, he suggests instead that we focus on govemment finance. When the Tang state decided to send large quantifies of silk to central Asia to pay its troops and occupying officiais, the Silk Road trade boomed. And when the state withdrew from Central Asia, as it did so suddenly after 755, it ended its massive subsidies (usually in the form of silk) to the région. While the Chinese govemment continued to buy horses in the northwest at a high price,93 the Silk Road trade dwindled to a small trickle. The Silk Road trade in the Century of Tang rule had clearly visible spill-over effects on the Turfan economy, which was highly monetized and in which ail transactions were subject to high interest rates. But even between 640 and 755, the Golden Age of the Tang on the Silk Road, more people eamed their livelihood working the land than did trading on the Silk Road. These cultivators had little to do with the Silk Road trade except when they borrowed silver coins from moneylenders or purchased animais and slaves from long-distance traders. Could it be that the Silk Road trade played a small rôle in Turfan’s overall economy? That is certainly what the limited numbers of surviving documents from Turfan suggest.

88

TCWS-texts, 7: 441; TC WS -photos, 111:517. Ikeda contracts 151-160, 175 90Buitow, 1940. 91 Trombert, 1995. 92 Trombert, 2000. 93 Trombert, 2000, p. 112. 89

Valerie Hansen

The. rnpact of the Silk Road trade

305

Appendix 1: Sogdian names in Chinese charactersÿ Pinyin, reconstructed Sogdian prononciation^ and English memings. by YOSHIDA Yutaka and KAGEYAMA Etsuko

An asterisk précédés those who did not possess sumames, while a square indicates that a sumame is damaged and lost.

(1) 称価銭文書 Scale-fee document 射蜜畔陀 shemipantuo Zhëmat-vandak 01

ロ

02 康

莫至

03 安

符夜門延 fuyemenyan

Avyäman-yän

04 安 05 康 06 何

09 康

符夜門遮 fuyemenzhe mopiduo 莫 毘多 阿倫(陵)遮 alun(ling)zhe dao 刀 那寧畔陀 naningpantuo 阿攬牛 延 alanniuyan

Avyämanch Mäkh-virt Rënchakk Täw Nanai-vandak Räman-yän

10 康

畢迦之

bijiazhi

Pêkach

u 何 卑尸屈

bishiqu

Pishkur

07 何

08 安

mozhi

Mäkhch

servant of the god Zhëmat (= llth month) related to the god Mäkh (= Moon) favour of the god Avyäman ( =Wahman?) related to the god Avyäman obtained from the god Mäkh small one (having) assets servant of the goddess Nanai favour of the god Räm (=Peace) relating to Pëk (meaning unknown) (meaning unknown)

(2) 訴訟文書 Document conceming the lawsuit between Li Shaojin and Cao Lushan 12 曹 禄山 lushan bright one Rokhshn 13 曹 畢娑 bisuo Pësakk coloured one 14 曹 炎延 yanyan Yam-yän (?) favour of the god Yima

(3) 接待文書 Documents recording the réception of West Turkic délégations 15 * juzhi 居職 Akuchïk Kuchean 16* mopantuo 莫畔陀 servant of the god Mäkh Mäkh-vandak 17* yanpantuo 炎畔陀 Yam-vandak servant of Yima 18* 脾娑 pisuo Pësakk coloured one 19* 陀 pantuo 畔 Vandak servant (of a certain deity) (4) 過所文書 Travel pass documents 20 翟 那你潘 nanifan Nanai-fam 21 康 阿了 Rëw aliao 22 曹 不那遮 Famch bunazhe

glory of the goddess Nanai rieh one related to Farn (= Glory)

し68Sogdiens en Chine

306 23 曹 24 曹

25 安 26 何

延那 野那 莫延 胡 数剌

yanna yena moyan

hushula

Yänakk Yänakk Mäkh-yän Ghösh-rät

favorite one favorite one favour of the god Mäkh given by the god Gösh (=14th day)

(5) 石染典関係 Documents conceming Shi Randian 27 石 染 典 randian Zhemat-yän 28 石 怒忿 nufen Nö-fam 29* ranwu 染勿 Zhëmat 30 安 達 漢 damo Tarldian yiduodi 31 * 移 多地 Yitädhi (?)

fevour of the god Zhemat new glory (of) Zhemat (title) ?

(6)契約文書 c011仃方義 wupoyan 32 康 烏破延 33 康 莫遮 mozhe 34 米 禄山 lushan 35 康 薄鼻 bobi 36 曹 娑堪 suokan 37 安 忽娑 husuo 38 安 不六多 buliuduo

Upä-yän Mäkhch Rokhshn Vagh-vïrt Askäm (?) Khers Parwëkht (?)

fevour of the god Upä (?) related to the god Mäkh bright one obtained from a god (=Mithra) ? ? ?

其他 0^618 39 史 ロ尸 番 40 何 祐所延 41 安 浮啣臺 42 何 浮口知延 43 石 羯槎 44 曹 提始潘

zhishifan yousuoyan fuzhitai fuzhiyan jiecha tishifan

Tîsh-fam Yishö-yän Buti-dhäy Buti-yän Karzh Dheshchi-fam

glory of a god Tishtriya fovor of Jesus female servant of the Buddha fevor of the Buddha miraculous one glory of the creator (= Ahura Mazda)

45 康

shiyan

Shyän

?

始延

で

Valerie Hansen

The impact of the Silk Road trade

307

BibliQg『題phy Arakawa Masaharu 荒丿丨［正晴 1997 "Töteikoku to sogudojin no köeki katsudö” 唐帝国とソ グ ド 人の交易活動 (The Tang Empire and Sogdian Commercial Activities), Töyöshi kenkyü 東)手 史研究 (The Journal of Oriental Researches) 56.3: 171-204 2001 "The Transit Permit System of the Tang Empire and the Passage of Merchants'5, The Memoirs of the Toyo Bunko 59: 1-21. BURROW, Thomas

1940

Translation of the Kharosthï Documents from Chinese Turkestan (London: The Royal Asiatic Society). Cai Hongsheng 蔡 鴻生 1998 Tangdai jiuxing hu yu Tujue ^enhua 唐代九姓胡與突厥文化 (The nine-surnames of non-Chinese in the Tang dynasty and Turkic culture) (Beijing: Zhonghua shuju). Changsun Wuji 晨孫無忌 1983 Tanglü shuyi 唐律 疏議 (Discussion and commenîary on The Tang Code) (Beijing: Zhonghua shuju). Cheng Xilin 程喜霖 2000 Tangdai guosuo yanjiu 唐代 過所 研究 (Studies on passports in Tang Dynasty) (Beijing: Zhonghua shuju). Deng Xiaonan. "Women in Turfan during the Sixth to Eighth Centuries: A Look at their 1999 Activities Outside the Home”, Journal of As ian Studies 58.1: 85-103. Gershevitch5 Ilya 1954 A Grammar of Manichean Sogdian (Oxford: Basil Blackwell). GULACSI, Zsuzsanna 2001 Corpus Fontium Manichaeorum Sériés Archaeologica et Iconographica I: Manichaean Art in Berlin Collections (Turnhout, Belgium: Brepols). Hansen, Valerie 1995a Negotiating Daily Life in Traditional China: How Ordinary People Used Contracts, (New Haven: Yale University Press). 1995b "Why Bury Contracts in Tombs", Cahiers d'Extrême-Asie 8: 59-66. 1 998 "The Path of Buddhism into China: the View from Turfan5*, Asia Major, Third Sériés, vol.ll, part 2:37-66. 2003 "New Work on the Sogdians, the Most Important Traders on the Silk Road, A.D. 500-1000", T^ungPao 89: 149-159. 2004 "Religions Life in a Silk Road Community: Niya During the Third and Fourth Centuries", in J. Lagerwey (ed.), Religion and Chinese Society vol. 1 Ancient and Médiéval China, (Hong Kong: Chinese University Press, EFEO): 279-315 Heng Chye Kiang 1999 Cities of aristocrats and bureaucrats: the development of médiéval Chinese cityscapes (Honolulu: University of Hawai'i Press). Huang Huixian 黃原賢 " 'Tang Xizhou Gaochangxian shang Anxi duhufu diegao wei lu shang xun 1983 wen Cao Lushan su Li Shaojin liang zao bianci shi' shi” 唐西州高昌縣上安 西都護府牒 稿為錄上訊問曹錄山訴李紹謹兩造 辯 辭事釋 (An Explication of the document entitled Draft die report from Gaochang sub-prefecture. Ä

308

.一

Les Sogdiens en Chine

Xizhou préfecture to Anxi Protectorate recording the testimony from both sides of those in the dispute in which Cao Lushan sued Li Shaojin, Tang dynasty), in Tang Zhangru 唐 長孺 (ed.), Dunhuang Tulufan ^enshu chutan 敦 煌吐 魯 番文書 初探 (Wuhan: Wuhan daxue chubanshe): 344-363. HUDÜDAL-'ÄLAM

The Régions of the World, trans. V. MlNORSKY, (London: Oxford Univ. Press). 1937 姜 伯勤 Dunhuang Tulufan wenshuyu sichou zhi lu 敦煌吐 魯 番文書與絲綢之路 1994 (Dunhuang and Turfan documents and the Silk Road), (Beijing: Wenwu

Jiang Boqin

chubanshe). Jiu Tângshu 舊 唐書 Jiu Tangshu 舊 唐書,(Beijing: Zhonghua shuju édition). 1975 Johnson, Wallace (trans.) 1997 The T'ang Code: Volume I, General Principles. Volume II, Spécifie Articles (Princeton: Princeton University Press). KAGEYAMA Etsuko 影山悦子 "Higashi Torukisutan shutsudo no ossuari (Zoroasutä kyöto no nökotsuki ni 1997 tsuite" 東トル キス タ ン出土の オッ スアリ (ゾロアスタ ー教 徒の納骨 器 ) について (The Ossuaries (Bone - réceptacles of Zoroastrians) Unearthed in Chinese Turkestan) Oriento オ リエ ン ト［Bulletin of the Society for Near Eastern Studies in Japan) 40-1 : 73-89. Klimkeit, Hans-Joachim Gnosis on the Silk Road: Gnostic texts from Central Asia (San Francisco: 1993 Harper). KNAUER, Elfriede Regina The CameVs Load in Life and Death: Iconography and Ideology of Chinese 1998 Pottery Figurines From Han to Tang and their Relevance to Trade Along ル e Silk Routes (Zürich: Akanthus). Kuwayama Shöshin 桑山正進 (ed.) Echoうgotenjikukoku den kenkyü 慧超往五天竺國 傳研究 ( Huichao'sWang 1992 Wu-Tianzhuguo zhuan Record of Travels in Five Indic Régions, translation and commentary) (Kyoto: Kyöto daigaku jinbun kagaku kenkyüjo). DE LA Vaissière, Étienne 2002/4 Histoire des Marchands Sogdiens (Paris: Collège de France, Institut des Hautes Études Chinoises, 2002 ; second édition 2004) Li Jian (ed.) The Glory of the SilkRoad: Art from Ancient China (Dayton: Art Institute). 2003 MORIYASU Takao 森安孝夫 Uiguru-ManikyÖ shi no kenkyü ウ イ グル ェ マニ教史の研究 (Researches in 1991 the history ofManicheism under the Uighurs). Osaka: Osaka daigaku. "Notes on Uighur Documents”, Memoirs of the Research Department of the 1995 Töyö Bunko 53: 67-108. Peng Xinwei / Monetary History of China (Zhongguo huobi shi), trans. Edward H. Kaplan 1994 (Bellingham, Washington: Western Washington University). Rong Xinjiang 榮 新江 "Beichao Sui Tang Sute ren de qianxi ji qi juluo” 北朝隋唐粟特人的遷 徙及 1999 其聚落 (The Sogdian Migration and Colonies in pre-Tang and Tang China) Guoxue yanjiu 6: 27-86.

Valerie Hansen 2001

The impact of the Silk Road trade

309

Zhonggu Zhongguo yu wailai wenming 中古中國 與 外來文明 (Middle period China and Outside Cultures), (Beijing: Sanlian chubanshe.)

RUDELSON, Jon

1997

Oasis Identities: Uyghur Nationalism along China 's Silk Road (New York: Columbia University Press).

SHENG, Angela

"Innovations in Textile Techniques on China's Northwest Frontier, 500-700 AD”, Asia Major 3rd sériés XL2: 117-160 SKAFF, Jonathan Karam uSasanian and Arab-Sasanian Silver Coins from Turfan: Their Relationship to 1998 International Trade and the Local Economy； Asia Major 3rd sériés XL2: 67415. "The Sogdian Trade Diaspora in East Turkestan During the Seventh and 2003 Eighth Centuries5^ Journal of the Economie and Social History of the Orient 464:475-524. Thierry, François "Sur les monnaies Sassanides trouvées en Chine”, Res Orientales V: 89-139. 1995 "Entre Iran et Chine, la circulation monétaire en Sérinde de Ier au IXe siècle,55 2000 in Sérinde, terre d'échanges: Arts religion, commerce du 尸 ピ siècle. XIVe5 Rencontres de l"Ecole du Louvre (Paris: La Documentation française): 121-147. , Trombert, Éric 1 995 Le crédit à Dunhuang: vie matérielle et société en Chine médiévale (Paris: Collège de France, Institut des Hautes Études Chinoises). "Textiles et tissus sur la Route de la soie: Éléments pour une géographie de la 2000 production et des échanges”, in Sérinde, terre d'échanges: Art, religion, commerce du 尸 au ブ siècle. XIVe5 Rencontres de l Ecole du Louvre (Paris: La Documentation française): 107-120. Tulufanchutu WENSHU 吐 魯 番出土文書 (abbreviated as TCWS) 1981-96 Tulufan chutu wenshu 吐 魯番出土文書 (Excavateddocuments from Turfan) ed. Guojia wenwuju wenxian yanjiushi 國家 文物局古文獻研究室 (Beijing: Wenwu chubanshe). Tulufan diqu wenguansuo 吐 魯番地區文管所 "Xinjiang Shanshanxian Tuyugou faxian taoguan zang” 新疆都善 縣吐峪溝發 1986 現陶棺葬 (The discovery of burials in clay coffins in Tuyugou, Shanshan county, Xinjiang) Kaogu 1:87-89 Twitchett, Denis 1966 Major 12: 202-248. "The Tang Market System”, Wang Binghua 1999 "New Finds in Turfan Archaeology”, Orientations 30.4: 58-64. Whitfield, Susan Life Along the Silk Road (London: John Murray, 1999). 1999 Wu Zhen 吳震 1999 "Asitana - Halahezhuo gumuqun kaogu ziliao zhong suojian de huren 阿斯塔 那 - 哈拉和卓古墓群考古資料中所見的胡人”( Non.Chinese 田ボ asseen in the archeological materials from the graveyards of Astana and Karakhoja), Dunhuang Tulufan yanjiu 敦煌吐 魯番 研究 4 (1999): 245-264. “Tangdai sichou zhilu yu hu nubi maimai” 唐礼絲嫡之路與胡奴婢買賣 (The 2000 Silk Road of the Tang dynasty and the sale and purchase of non-Chinese male and female slaves). In 1994 nian Dunhuangxue guoji yantaohui wenji 1994 年 1998

f

」

310

一

Les Sogdiens en Chine

敦煌學國際 研 討 會文集 (ed. Dunhuang yanjiuyuan) (Lanzhou: Gansu minzu chubanshe), 128-154. “'Hu' Non-Chinese as They Appear in the Materials from the Astana 2002 Graveyard at Turfan5', Sino-Platonic Papers #119. Xinjiang chutu wenwu 新疆 出土 文物 Xinjiang chutu wenwu 新疆出土文物｛Excavated artifacts from Xinjiang) 1975 (Shanghai: Wenwu chubanshe). Xinjiang Weiwuer zizhiqu bowuguan 新疆 羅吾爾自治區博物館 (ed.) Xinjiang Weiwuer zizhiqu みoiwgis附 新疆 維吾爾自治區博物館 (Xinjiang 1991 Uighur autonomous district muséum) (Beijing: Wenwu chubanshe). Yamamoto Tatsuro and Ikeda On Tun-huang and Turfan Documents Concerning Social and Economie History 1987 HI Contracts (A) Introduction and Texts (Tokyo: The Töyö Bunko). YAN Wenru 閻文孺 "Tulufan de Gaochang gucheng59 吐 魯番的高昌故城 (The ancient city of 1962 Gaochang, Tu命!！) Wenwu, 7-8 (1962): 28-32. Yoshida Yutaka 吉田豊 uSino-Iranican , in: Seinan Ajia kenkyü 『西南 アジア 研究』Bulletin of the 1998 Society for Western and Southern Asiatic Studies, Kyoto University, 48, p. 33-51. (in Japanese) "Appendix: Translation of the Contract for the Purchase of a Slave Girl Found 2003 at Turfan and Dated 639”, T'oung Pao 89: 159-161. « On the Orig in of the Sogdian Sumame Zhaowu 昭武 and Related 2004 Problems », Journal Asiatique, n° 291, p. 35-6 7. Yoshida Yutaka 吉田豊,Moriyasu Takao 森安 孝夫,Xinjiang Uighur Autonomous Museum 1988 "Kikushi Köshökoku jidai Sogudo bun onna dorei baibai monjo” 麹 氏高昌国 時代ソ グ ド 文女奴麻売買文書 (A Sogdian contract of a female slave from the period of the Gaochang kingdom under the rule of the Qu clan) Nairiku Ajia gengo no kenkyü 内陸 アジア 言語の研究 (Studies on the Inner Asian languages) 4 (1988): 1-50. Zeymal, Eugeny V. "Eastem (Chinese) Turkestan on the Silk Road, First millennium A.D.: 1 992 Numismatic Evidence" Silk Road Art and Archeology 2:137-177. Zhang Guangda 張廣 達 "Tangdai liu huzhou dengdi de zhaowu jiuxing" 唐代六胡州等地的昭武九姓 1 995 (The nine Zhaowu sumames of the six non-Chinese préfectures and other places during the Tang dynasty), in Xiyu shidi conggao chubian 西域史地叢 槁初編 (APreliminary version of collected drafts concerning the history and lands of the Western Régions) (Shanghai: Shanghai guji chubanshe): 249-279. "Tulufan chuta Hanyu wenshi zhong suo jian Yilangyu diqu zongjiao de 1999 zon 如”吐 魯番出土漢語文書中所見伊曲語地區宗教的褪跡 (Traces of Iranian religion as seen in the Chinese documents excavated from the Turfan région), Dunhuang T'ulufan yanjiu 4: 1-16 "Iranian Religions Evidence in Turfan Chinese Texts", in Zoroastrianism 比 2000 China, China Archaeology and Art Digest, IV- 15 December 2000, Hong Kong;

193-206. ZHANG Guangda and Rong Xinjiang "A Concise History of the Turfan Oasis and Its Exploration” , Asia Major 3rd 1998 sériés, XI.2: 13-36.

Documenting Sogdian society at Turfan in the seventh and eighth centuries: Tang Dynasty census records as a window on cultural distinction and change.1 Jonathan Karam Skaff

Although it has long been assumed that the Sogdians dominated Asîan Silk Road trade in the médiéval period, we only recently hâve begun to understand the Orga¬ nization and implémentation of their commerce. A key element of their mercantile activities was a diaspora of Sogdians living in cities and towns on trade routes lying between the Sogdian homeland and China. These Sogdians supported themselves as farmers, artisans and, of course, merchants. The settlements served as home bases for merchants and provided them with support as they visited cities and towns in the trade network to buy and seil wares. This paper will investigate some aspects of Sogdian (modem society and culture at the Tang Empire Silk Road oasis city of Xizhou Turfan) from the mid-seventh to mid-eighth centuries. It will demonstrate that Sogdians living in part of Turfan had a démographie profile that distinguished them from their mostly Han neighbors but, nonetheless, some Sogdians show signs of beginning to assimilate into local Han society.2 The interplay between Sogdian cultural distinction and assimilation at Turfan reveals that their diaspora society was dynamic, rather than a static réplication of an ancestral way of life. The Sources: Tang Dynasty Census Records

We can obtain a glimpse of Sogdian social structure in the diaspora thanks to rare surviving contemporary records that hâve been found in the Turfan oasis, presently located in northwestem China’s Xinjiang Uighur Autonomous Region. The so-called “Turfan documents” were unearthed in the late nineteenth and twentieth centuries by German, Russian, Japanese, British and Chinese excavators at the contiguous ancient and Karakhoja (Halahezhuo burial grounds of Astana (Asitana The three thousand tombs in the near the ruins of the ancient city of Gaochang burial grounds—used by the social elite among the ethnically Han settlers who came 1

Parts of this paper appeared previously in the Journal of the Social and Economie History of the Orient (Skaffj 2003). At the Sogdians in China Confèrence, this paper benefïted from the comments and présentations of other conférence participants. In particular I would like to thank Éric Trombert and Alain Arrault for their remarks that pushed me to clarify my methodology and arguments. 2 In this paper “Han” will be used to refer to the ethnicity of Chinese speaking people. “Chinese” will refer to the language.

Les Sogdiens en Chine

312

from China ’s interior to Turfan from the third Century onward —hâve been dated from the third to eighth centuries. Many of the tombs included documents placed there as waste paper that was eut and folded to make grave goods and clothing for the dead. The arid climate in the région helped to preserve the writings for more than a thousand years, but most manuscripts still are fragmentary because of the intentional mutilation prior to placement in tombs, natural détérioration over time, and disruption by tomb robbers.3 Despite the imperfect préservation of the Turfan documents, they can be of tremendous value to modem historians because they often are dated or at least approximately datable, and are extremely rare local records from this era in East Asia or any other part of the world. The spécifie documents that are the main sources for this study are census records, which Tang dynasty (618-907) officiais kept in order to allocate land and collect taxes under the equal field land tenure (juntian ±t]E0) System. This land System was imposed on Turfan after the Tang Empire conquered the formerly independent oasis city-state of Gaochang in 640, and renamed it Xizhou. The equal field System involved allocating state farmland to each household, which in retum owed taxes to the govemment in the forms of grain, cloth, and corvée labor or a labor exemption fee.4 The System required a great deal of paperwork, recording the names and âges of each household member and were compilée! trienexact descriptions of land allocations. Household registers (ji that noted changes in nially, based upon annual household déclarations (shoushi registration status for individual household member s. 5 Although Tang govemment offices originally would hâve been overflowing in this paperwork—copies of registers were stored in county, préfecture, and central govemment archives for periods of fifteen or twenty-seven years only fragments are extant, mostly from the arid oases of Turfan and Dunhuang. And unfortunately, only a minority of these surviving documents has

—

useful démographie data. Research for this paper included a survey of ail extant Tang Turfan census docu¬ ments, which yielded a total of ninety-one households with data useful to this study.6 In seventy-seven cases démographie data on ail members of a household survive. For these households, available data was gathered on each person’s name, gender, âge, marital status, and official social classification. Landholding, which was awarded to an entire household, also was recorded when available. Fourteen additional households with incompletely preserved data are included in this study because the sumames of both spouses survive. 3

Introduction to Turfan’s history and Turfan studies in Tang, 1982; Hansen, 1998; Zhang and Rong, 1998. On the Tang land and tax System, see Twitchett, 1970, 1-48. 5 The Tang household registration process and census documents are explained in TTDS-2: 6-8, 49-50. 6 There were two main sources that assisted in locating the documents. Yamamoto and Dohi compiled a total of sixty-eight Turfan census documents, dating from 640 to 757, from collections in Japan, China, Britain, Germany, and Russia, but most are so fragmentary that only four of them yielded information for this study on a total of fifteen households (TTDS-2: XVIII-LXXXVI). The data from these documents is compiled in Appendix I to this paper. Yamamoto and Dohi were not able to include many of the more recent Chinese finds that were still in the process of being edited and appeared in Documents Excavated at Turfan (TCWS-1 and TCWS-2). Dong Guodong identifies the relevant census records from TCWS-1 and TCWS-2, dating from 640 to 719, in his study of Tang demographics (1993: 379, 392, 402-406). Data from these documents appear in Tables 1a and 1b and Appendix I. 4

Jonathan K. Skaff

Documenting Sogdian society at Turfan

313

The single most important document is the 707 Chonghua Township (Chongwhich household register inspection record (dianji yang huaxiang and Han mostly summarized Sogdian households forty-seven on data yielded complété in Tables la and lb.7 Various other fragmentary documents—dating from 640 to about 750 from generally indeterminate locations within Turfan—provide data on forty-four mainly Han households compilée! in Appendix I. The 707 Chonghua Township house¬ hold register inspection record is uniquely valuable to social and économie historians, as Victor Xiong (1999: 358) points out, because “the spatial proximity of so many households and the temporal unity of their related information” cannot be found in the other more fragmentary Turfan census documents.8 The démographie information on the forty-seven households preserved in the Chonghua inspection record, although not randomly selected, probably represents about one-tenth of the township’s population.9 It can give a rough idea about Contemporary household structure in this part of Turfan. We can identify Sogdians in the census records because it was a standard practice when transliterating a foreigner’s name into Chinese to use the appellation for the na¬ tive country or tribe as the sumame.10 Thus, the homelands of foreigners or their descendants often can readily be identified. As is well known to readers of this book, there are seven sumames known to hâve been used by Sogdians in China: Kang U (Maimargh), (Bukhârâ), Shi Ç (Châch), Shih11 (Kish), Mi (Samarqand), An (Kushânîyah). However, these sumames are not a Cao W (Kabûdhanjakath), and He foolproof method of identification because during the Tang other ethnicities are known to hâve used most of them. Elsewhere, I discuss in detail the problems involved in using the sumames to distinguish Sogdians in Chinese language documents (Skaff 2003: 478481). Although a Sogdian given name or mention of birth in Sogdia are the safest stand¬ ards of identification, this paper will adopt less a rigorous criterion by assuming that in the Turfan documents people with these sumames generally can be presumed to be of Sogdian descent. We can suppose that people with these sumames who lived and passed through Turfan mostly had Sogdian backgrounds because the location of the oasis on the Silk Road made it a likely place for commercial activity and settlement. In addition, there are far more examples of people using these sumames in the documents who can be firmly recognized as Sogdian than can be proved not to be. I hâve found only one instance in the Turfan documents where a person with one of the seven sur7 64TAM 35:47(a) - 58/3(a) in TCWS-1: 7:468-485; TCWS-2: 3:533-544. These tables originally appeared in Skaff, 2003: 486-488. Complété démographie data derived from this document hâve been transcribed into English in Skaff, 2003: Appendix II. This particular document’s relationship to the equal field System h as been carefùlly analyzed in Ikeda, 1988; Xiong, 1999: 358-364, 368-374. 8 One caveat conceming the temporal unity of the document are six households listed as having absconded for ten years (see R9-R14 in Table la). I hâve chosen to include these households in the following analysis because they can still help us to understand the structure of Sogdian families in a wider time span. ° There are not any extant data on the populations of Xizhou’s townships, but Tang dynasty sources mention that Xizhou had a total of 11,647 households and 24 townships in the first half of the eighth Century (YHJX 40:1030-1032). This works out to an average of about 485 households per township. 10 Ikeda, 1965: 61; 1993: 155. It is not known who determined Chinese language names for people of non-Han ethnicity. Perhaps translators and scribes collaborated in creating the name. 11 In the Chinese pinyin romanization System, this sumame should be spelled Shi. I hâve adopted the alter¬ native spelling, Shih, to distinguish it from the homophonous sumame associated with Châch.

314

Les Sogdiens en Chine

names is associated with a non-Sogdian ethnicity.12 Finally, many of my identification of Sogdians using this criterion of identification in my previous publications subsequently hâve been verifïed by the philological research of Yutaka Yoshida.13 Nonetheless, we can présumé that the data in this paper are not perfect because on occasion non-Sogdians will hâve Sogdian sumames and some Sogdians may be overlooked because they might hâve adopted Han sumames. The beneflts of including the less certainly identifïed Sogdians, who mostly hâve Han-style given names, probably outweigh the potential risks because hints of cultural dynamisai and assimilation often can be glimpsed most vividly in the less firmly identified Sogdian households. Families, such as Kang Caibao’s that will be discussed at the end of the paper, assimilated to the point that it is difficult to recognize them as Sogdian, yet appear to retain traits that distinguished them from the Han majori ty. These cases are crucial to this study because they intimate how the ethnie désignation “Sogdian” evolved in new directions at Turfan. If this approach leads to mistakes that spark debate, it only will deepen our knowledge of Sogdian diaspora society. Sogdian Household Structure

The most important census document for the study of Sogdian social structure at Tang Dynasty Turfan is the afbrementioned 707 Chonghua Township household register inspection record. Out of the total of thirty-two extant Sogdian-headed households with füll démographie information appearing in census documents, twenty-fïve are preserved in the 707 inspection record. The other seven households, scattered in records covering a fifty-year time span, are less useful for démographie analysis because of their small number and temporal dispersai.14 Moreover, the inspection record provides an unusual opportunity to compare Sogdians and their Han neighbors, each of whom made up about half of the surviving registered households in the document. A large fragment of the inspection record referring to Anle “Peace and Happiness” Village which was located within the township, appears to hâve been a true Sogdian dias¬ pora community, having nineteen out of twenty-three households with Sogdian surnames (83%).15 A comparison of both groups can help us to see what was specifically “Sogdian” about the démographie profile of one cluster of people living in the diaspora beside a Han majority and a few other non-Han minorities.16 The data from the docu¬ ments is summarized in Tables la and 1b. 12

The Sogdian sumame attributed to a member of another ethnie group is in a court déposition where a is identified as a commoner of the Chumi tribe (73 TAM509:8/l(a), 8/2(a) in Kang Shifen TCWS-1 9:130; TCWS-2: 4:331; Also see Valerie Hansen ’s paper in this volume, p. 297). On the Chumi, see Skaff, 2003: 480 n. 19. 13 Skaff 1998; 2003. See the appendix to Valerie Hansen’s paper in this volume p. 305. 14 See Appendix I: C2, C16, C17, C28, C31, C32, C34. Appendix I also includes two incomplète Sogdian households: C23 and C33. 15 In the tables below household numbers R9 through R31 were residents of Anle. 16 A total of four non-Sogdian households out of twenty-two total hâve sumames that may be non-Han. Only two households with the Bai 0 sumame can be identified as non-Han with a high degree of certainty (R46, R47). The sumame Bai has been associated with the oasis city of Kucha (see Skaff, 1998: 94, n. 83) and the family heads appear to hâve non-Han given names. More speculatively, Jiang Boqin (1994: 173)

Documenting. Sogdian society at Turfan

Jonathan K„ Skaff

315

Table 1魏: Sogdians 加 Chonghm Towmhip, Demographic Breakdown Females

Males

ID

Household Head

Children

0

EM

ÂM

YM

TM

EF

AF

YF

TF

B

G

IM

IF

TC

R5

Kang Yÿi

2

0

0

0

0

0

0

0

0

0

1

1

0

2

R6

An Shengniang 安勝娘

1

0

0

0

0

0

1

0

1

0

0

0

0

0

R7

An Fuzhitai 安浮咖臺 Sog: Buti-dhây Cao Amianzi 曹阿 面子

1

0

0

0

0

0

0

0

0

0

0

0

1

1

2

0

0

0

0

0

0

0

0

0

2

0

0

2

9

0

1

1

2

0

2

1

3

1

1

1

1

4

7

0

1

0

1

0

1

0

1

2

1

2

0

5

R8

R9

R 10

Kang Lushan 康祿山 Sog:

Rokhshan Kang Tuoyan 康随延 Kang Enyi 康恩義

7

0

0

0

0

1

2

2

4

1

1

0

0

2

11 R 12

He Mopan 何莫潘

11

1

2

0

2

1

2

0

2

0

2

0

3

5

R 13

Kang Qiewei 康迦衛

1

0

1

0

1

0

0

0

0

0

0

0

0

0

R

Kang Azi 康啊子

11

1

2

0

2

2

3

0

3

0

3

0

0

3

14

R 19

An Dezhong 安德忠

7

0

0

0

0

0

3

1

4

1

0

2

0

3

R 20

Kang Waihe 康外何

3

1

0

0

0

0

2

0

2

0

0

0

0

0

R

Kang Nuoqian 康那虔

4

0

0

0

0

2

2

0

2

0

0

0

0

0

21

R 22

He Wuhezhi 何无賀口知

5

0

0

0

0

1

4

0

4

0

0

0

0

0

R 23

3

1

0

0

0

1

0

0

0

0

1

0

0

1

4

0

0

0

0

1

1

0

1

0

1

0

1

2

R

R

Shi Fuzhipen 石浮口知盆 Kang Achou

24

康阿醜

R 25

Shi Fuzhiman 石浮口知滿

4

0

1

0

1

0

1

0

1

1

1

0

0

2

R 26

Cao Fushi 曹伏食

8

2

1

1

2

0

1

1

2

0

2

0

0

2

suggests that households R15 and R16 may be of Indian descent, based on the hypothesis that their sumame Zhu 竹 (bamboo) may be a variation of Zhu 竺 (India), which would have been derived from the ancient Chinese name for India, Tianzhu 天竺.

316

Les Sogdiens en Chine

R 27

030 Mopen 曹莫盆

7

?

1

0

1

?

?

0

?

7

?

?

?

R 28

Kang Shougan 康薛 感

7

0

0

0

0

0

2

1

3

1

3

0

0

4

R 29

Kang YanpaIl 康演 潘

8

0

1

0

1

0

2

0

2

1

2

1

1-

5

R 30

An Yishi

8

0

1

0

1

0

3

0

3

2

2

0

0

4

R 31

An Shancai

8

0

C

0

1

4

0

3

0

3

0

0

0

1

]

超オ

R 44

Ca。 Xuanke 曹玄恪

5

0

1

0

1

0

1

0

1

2

0

0

1

3

R 45

仙 Shi^lu

4

0

0

0

0

?

?

0

?

1

?

?

?

1

Tot

25

137

6

16

3

19

9

36

6

42

13

23

7

9

52

5.5

0.3

0.6

0.1

0.8

0.4

1.6

0.2

1.8

0.5

1.0

0.3

0.4

2.2

17

44

16

48

30

78

22

78

4

12

2

14

7

26

4

31

安義師

ー

'

安師奴

Ave %

78

Ho %

9

17

5

7

38

Pop

Table 1b: Han Chinese and other Non-Sogdians in Chonghua Township, Demographie Breakdown Males 10

m R2 R3

R4 R 15 R 16 E 17 E 18 X 32 R 33 E 34 R 35

Household Head Zhang Cishan 張慈善 Wei Shuangwei 魏雙尾 Chen Sixiang 陳思香 Li Chounu 李醜 奴 Zhu Pande 竹畔德 Zhu Xiongzi 竹熊子 Yin Asun 陰阿孫 Xiao Wangxian 蕭望 仙 Zhao Duli 趙獨立 Xia Yunda 夏運達 Liu Shu 劉戍 Zheng Sishun 鄭思順

Females

Children

H

EM

EF

AF

B

0

0

0

0

0

1

0

G 1

IM 0

IF

0

YF 1

TF

2

0

TC 1

1

0

0

0

0

1

0

0

0

0

0

0

0

0

3

0

0

0

0

0

2

0

2

0

0

0

1

1

1

0

0

0

0

0

0

0

0

1

0

0

0

1

9

0

2

0

2

0

3

0

3

1

3

0

0

4

5

0

1

0

1

0

3

0

3

0

0

1

0

1

1

0

0

0

0

0

1

0

1

0

0

0

0

0

3

0

0

0

0

0

2

0

2

1

0

0

0

1

5

0

1

0

1

0

1

0

1

0

1

1

1

3

4

0

1

0

1

0

1

0

1

1

1

0

0

2

1

0

0

0

0

0

1

0

1

0

0

0

0

0

3

0

0

0

0

0

]

0

1

1

0

1

0

2

AM

YM TM

Documenting. Sogdian society ai Turfan

Jonathan K. Skaff

_ R

Guo Deren

36

郭德仁 Guo Taoye 郭桃葉 Guo Zhongmin 郭忠敏 Jiao Sengzhu 焦僧 住 Li Qingwu 李慶賦 Guo Junxing 郭君行 Zheng Longhu 鄭隆護 Zheng Huanjin 鄭歡 進 Bai Huren 白胡仁 Bai Mangzi 白盲子

E 37

E 38

R 39

E 40

E 41

E 42

E 43

R 46

E 47

TM

Ave %

22

317

6

0

1

0

1

0

2

1

3

0

0

0

2

2

2

0

0

0

0

0

1

1

2

0

0

0

0

0

5

0

0

0

0

0

3

0

3

2

0

0

0

2

8

0

1

1

2

0

1

1

2

0

4

0

0

4

2

0

1

0

1

0

1

0

1

0

0

0

0

0

8

0

1

0

1

0

2

0

2

2

?

?

7

5

1

0

1

0

1

0

0

0

0

0

0

0

0

0

3

0

1

0

1

0

1

0

1

?

?

?

?

1?

5

0

1

0

1

0

2

0

2

1

0

1

0

2

5

0

1

0

1

1

?

0

?

?

?

7

7

7

83

0

13

1

14

2

28

4

32

10

10

4

4

3.8

0.0

0.6

0.0

0.6

0.1

1.3

0.2

1.5

0.5

0.5

0.2

0.2

0

55

5

55

9

81

18

86

31 1.6 75

0

16

1

17

2

34

5

39

12

12

5

5

37

Ho %

Pop

Key: ID=ID Number of household in subséquent tables; H=Total Family Members in Household; EM=Elderly Males, 61+ years old; AM=Adult Males, 22-60; YM=Young Men, 16-21; TM=Total males, 16-60; EF=Elderly Females, 61+; AF=Aduit Females, 22-60; YF=Young Women, 16-21; TF=Total Females, 16-60; B=Boys, 3-15; G=Girl, 3-15; IM =Male Infant, 1-2; IF =Female Infant, 1-2; TOTotal Children, 1-15; Ave=Average per household; % Ho=Percentage of households containing a population category; % Pop=Percentage of Sogdian (Table la) or non-Sogdian (Table 1b) population in Chonghua township.

Notes: Age ranges are given according to Han Chinese practice with babies counted as being one year old at birth. Ages ranges of population categories differ slightly from those stipulated by Tang law because they have been adjusted to reflect local standards visible in the document. See Cao Shilou (R26) who is categorized as an "adult male," age 60, even though the "elderly" category begins at this age in the Tang Statutes. Zhang Cishan (RI) is categorized as a "young woman" age 21. According to Tang régulations, people were supposed to be categorized as “adults” when they turned twenty-one. Part 3, Xiao Wangxian (RI 8), age 3, is categorized as a "boy J rather than an "infant male,,s despite the stipulation that the boy and girl categories should begin at age four. See Xiong (1999: 355-6, 362-3).

A comparison of Tables la and 1b reveals some striking différences between the Sogdians and their mostly Han neighbors. Perhaps the most eye-catching contrast is the average size of households (H) with Sogdians having a mean of 5.5 family members versus 3.8 for Han and other non-Sogdians. Sogdians had a slight advantage in the average number of males (TM) and females (TF) per household âges 16-60, but other démographie categories account for more of the différence. Sogdians had more children, âges 1-15 (TC), with a mean of 2.2 per household versus 1.6 for non-Sogdians. The Sogdian figure most likely has been distorted downward and probably was doser to 2.6

318

Les Sogdiens en Chine

because, as will be noted below, their families tended to send away almost half of their male children. For the sake of comparison, another rare population register of mostly Han from Hexi in the middle of the eighth Century shows 2.6 children per household, so the figures for Chonghua Township Han can be considered low.17 Even more striking is the almost total lack of an older génération in the non-Sogdian community, which had no elderly males (EM), and elderly females (EF) only made up 2% of the population. On the other hand, seniors combined to constitute 11% of local Sogdians. In comparison, the above-mentioned population register from Hexi shows a more moderate rate of elderly in the community of 4%, which is evenly split between males and females.18 Overall, the Sogdian figure of 5.5 family members per household is in line with the Tang interior, which generally averaged around five or six people throughout the dynasty, but compares favorably with the average of 4.38 in Turfen in the mid-seventh Century.19 On the other hand the figure of 3.8 for non-Sogdians has to be considered low. What explains this discrepancy in household size and survival rates between Sogdians and non-Sogdians in Chonghua Township? Since Han Chinese had a long history of settlement in Turfen and are less likely to be recent immigrants with truncated households, a possible explanation is that the Sogdians had more wealth and consequently better nutrition than their non-Sogdian neighbors. We can support this hypothesis by comparing the agricultural resources in both communities. Fortunately, we hâve data that can allow us to make this comparison because this document also records the total amount of land that the Tang govemment granted to each household under the equal field System. Still, this comparison of landholding must be considered rough because the inspection record only lists the total acre âge in units of mu without specifying quantifies of the two categories of land quality known to hâve existed at

Turfan.20 The discussion

below assumes that the proportions of the different types of land were roughly equal. The data are tabulated in Tables 2a and 2b.

17

The figure for Hexi is based on Ikeda’s (1973: 137-139) data on 153 individuals from a total of 29 iàmilies. There were 40 males and 35 females in the 1-15 âge group, which, accord ing to my calculations, gives a rate of 2.6 children per household. 18 According to my calculations based upon Ikeda’s (1973: 137-139) data on 153 individuals, there were 3 males and 3 females in the 61-85 âge group, who together accounted for 4% of the population. 19Dong, 1993: 359-360, 379-380. Dong’s figures for Turfan are based on 78 households visible in 24 ftagmentary census documents dating from 640 to 671. Some of these households hâve incomplète records, so his figure for Turfan probably is low. 20 Turfan was unique during the Tang in specifying two types of equal field land: changtian ^03 and butian n|5EH. Chinese, Japanese and Western historians hâve had an extended debate on the meaning of these tenus. Eric Trombert (2002) argues persuasively that the availability of irrigation water determined that changtian was double- cropped and butian was single-cropped annually or left fallow for periods of one to three years. Individual plots of changtian and butian also varied in soil fertility. Families normally had a mixture of the different types of land.

Jonathan K. Skaff

Documenting, Sogdian society at Turfan

319

Table 2a: Sogdian Land Granta Chonghua Township L "

ID No. R5 R06 R07 R08 R09 RIO Rll R12 RI 3 R14 R19 R20 R21 R22 R23 R24 R25 R26 R27 R28

R29 R30 R31 R44 R45 Total Average/Household Median/Household

Family Size

2 1 1 2 9 7 7 11 1 11 7 3 4

5 3 4 4 8 7 7 8 8 8 5 4 137 5.5 5.0

Eligible for land

1 1 1 1 1 1 3 3 1 4 2 1 3 1 1 1 1 4 2 3 1 2

3 1 1? 43 1.8 1.0

Mw/person

Land

NY NY NY NY 9 10 8 25 ? 23 10 3 7 5 10 5 10 12 13 8 10

14 ? 10 ? 192 10.7 10.0

Afu/eligible person

1.0 1.4 1.1 2.3 ? 2.1 1.4 1.0 1.8 1.0 3.3 1.3 2.5 1.5 1.9 1.1 1.3 1.8 ? 2.0 ?

9.0 10.0 2.7 8.3 ? 5.8 5.0 3.0 2.3 5.0 10.0 5.0 10.0 3.0 6.5 2.7 10.0 7.0 ? 10.0 ?

1.6 1.5

6.4 6.1

Table 2b: Han Chinese and Other Non-Sogdian Land Grants, Chonghua Township ID No. RI R2 R3 R4 R15 R16 R17 RI 8 R32 R33 R34 R35 R36 R37 R38 R39 R40 R41 R42 R43 R46 R47 Total Average/Household

Median/Household

Family Size 2 1 3

1 9 5 1 3 5 4 1 3 6 2 5 8 2 8 1 3 5 5 83 3.8 3.0

Eligible for land

1 1 1 1 3 3 1 2 1 1 1 2 2 1 4 2 2 1 1 1 1 2 35 1.6 1.0

Land (Mu) NY NY NY

NY 17 9 5 5 9 7 2.5 5 ? 5 9 ? ? 11 10.5 ? 9 15 119 8.5 9.0

Mw/person

Ma/eligible person

1.9 1.8 5.0 1.7 1.8 1.8 2.5 1.7 ? 2.5 1.8 ? ? 1.4 10.5 ? 1.8 3.0

5.7 3.0 5.0 2.5 9.0 7.0 2.5 2.5 ? 5.0 2.3 ? ? 11.0 10.5 ? 9.0 7.5

2.8 1.8

5.3

5.9

Key: ID No.^Identification number in Tables la and b. NY=Land not yet distributed to these households.

320

Les Sogdiens en Chine

The tables demonstrate that aithough Sogdian households were granted median plots of land about one mu (0.13 acres) larger than non-Sogdian holdings, they had a median of 0.3 mu less land available per person than non-Sogdians because of their larger household sizes.21 In both cases, these land allocation figures of nine or ten mu (about 1 acre) per household were far below the standards stipulated in the Tang Statutes for restricted localities of sixty mu per adult male.22 Xiong (1999: 377-83) argues that a family of five in Turfan probably could survive on these small land allotments of ten mu (2 mu per person) thanks to high yields attributable to double cropping and the generally fertile soil, while Trombert (2002: 205; 2002a: 503-4) spéculâtes that people would hâve had to farm additional land that remained outside of state control to support themselves. In either case, based on the household sizes of Chonghua Township’s Han, it is likely that most were barely surviving on a marginal subsistence diet. Among the Han and other non-Sogdians, poor nutrition can explain the low number of children per household and almost total lack of elderly in the community. Consequently, it is startling to see that the Sogdians, who had less state-distributed land per person, could support more children and had higher survival rates. In Anle village the only non-absconded Sogdian household with adéquate land and labor to support a family entirely on subsistence farming may be Shi Fuzhiman (Table la: R25), who headed a nuclear family of four, farming about ten mu. The most likely conclusion that can be drawn is that most Sogdian households were pursuing occupations other than subsistence agriculture to supplément their incomes. The Turfan documents provide evidence that Sogdians practiced a number of non-farming professions. Long distance trade is the most frequently noted activity, but Sogdians in Turfan also are mentioned as government officiais, bronze-smiths,

ironsmiths, artists, a painter, leatherworkers, a veterinarian, and an innkeeper. 23 Leatherworking may hâve been a Sogdian specialty at Turfan because one Tang government list of craftsmen from circa 640 reveals that ail of the leatherworkers with visible sumames are Sogdian.24 Éric Trombert has advanced the hypothesis that Sogdians at Turfan were involved in entrepreneurial faming of grapes to produce raisins 21

1 compare median rather than average land holdings in this case because the latter are skewed by the vagaries of the Tang land distribution System. Since single person non-Sogdian households R17, R34, and R42 lacked other family members to share their land with, the per person averages hâve been distorted upward. On the size of a mu, see Twitchett (1970: xiii) 22 In the Tang System smaller parcels of land also were allocated to female and child household heads, elderly males, widows and the disabled (Xiong 1999: 364-374). 23 On merchants, see Sims-Williams 1996; la Vaissière 2002; Skaff 1998; Skaff 2003. For local officiais, see Table la: R12, R44 and Trombert 2002a: 519, 524-525. For the other professions, see Wu 1999: 258, 261. Angela Sheng (1998) has proposed the provocative hypothesis that new silk weaves appearing in Turfan were the products of Sogdian workshops. Aithough at Dunhuang Ikeda On only found direct evidence of farming among the Sogdians, this is probably a distortion caused by the nature of surviving documents, which for eighth Century Dunhuang mainly deal with issues of land registration and taxation (Ikeda 1965: 79-80; 1993: 188-190). 24 In the document six of the seventeen craftspeople or about one third with visible sumames are Sogdian. Five of the Sogdians are leatherworks and one is an artist. Among the other listed professions, there are not any Sogdian tailors, fùrriers, woodworkers, food oil makers, orpig slaughterers. See 64TKMl:28(b), 31 (b), 37/2(b) in TCWS-1: 4:15-17; TCWS-2: 2:11.

Jonathan K. Skaff

Documenting Sogdian society at Turfan

321

and wine for sale on the market. He argues that Sogdians would hâve been likely to possess the necessary capital, business acumen, and knowledge of viticulture to be prominent in this business.25 Although vineyards were common at Tang Turfan, only one census document provides spécifie information on the cropland of three Sogdian households. One of the three, the household of Kang Caibao, had more than five and one half mu (about half an acre) of vineyards, but the other two only grew grain and vegetables.26 This demonstrates that viticulture may hâve been an important income

source for some, but not all, of Turfan’s Sogdians. If we are to believe the letter of Tang law, only official service would hâve provided an allowable supplément to farming income because in areas with land shortages, like Turfan, merchants and artisans were prohibited froni receiving land allocations.27 Since most Sogdians living at Chonghua Township were not officiais, it is necessary to explain how they may hâve supplemented their income s. One possibility is that fanning, especially entrepreneurial viticulture, on additional rented land provided legal added revenue. A second possibility is that financial support came from male relatives who were craftsmen or itinérant traders, who were not listed on census registers. This would help to explain the lack of males in Sogdian households, which will be discussed below. A third possibility is that some of the resident Sogdian males may hâve found ways to circumvent the law to supplément their incomes with non-agricultural pursuits. For example, Kang Lushan (Table la: R9) supported a family of nine, including a concubine, on a mere nine mu of land. It is hard to believe that a relative’s contributions or farming about one acre, even if it was a vineyard, would provide sufficient fonds to affbrd the extravagance of a concubine. As will be discussed below, the fact that his household illegally absconded to another location shows that Kang was mobile, perhaps in pursuit of new business opportunities, and did not always play by the rules. 28 Another example proves that Sogdian merchants sometimes were legally registered taxpayers at Tang Turfan. In a certificate of ownership, listing goods that the merchant Shi Zhëmat-yân (Randian was transporting, guarantors vouched that Shi had a house and family and pledged to pay his taxes if he did not retum.29 Obviously, Tang law towards merchants was not being folly enforced at Turfan. These three possible ways that Sogdians at Turfan may hâve supplemented their income are not mutually exclusive and a single household could hâve practiced two or ail three of them. Together they may explain why Chonghua’s Sogdian households might hâve been more successfol than their immédiate Han neighbors in overcoming the financial constraints caused by shortages of state land available for distribution under the equal field System. 25

Trombert, 2002a: 553-554. Appendix I: C32. Unfortunately, the document breaks off before describing ail of the land allocated to Kang’s household, but we know that he also had over one mu of vegetables and half a mu of double-cropped grain land. Two other Sogdians in the same document, the unmarried male, Shih Gouren (C28), and the widow, Cao Duofii (C31), grew only grain and vegetables. 27 Merchants and artisans were eligible for half shares of land under normal circumstances, but were not give an allocation in “restricted” areas, such as Turfan (Twitchett, 1970: 4, 129). Perhaps Kang and some others used their wealth to bribe officiais in order to obtain special treatment, such as illicit land allocations. Another possibility is that local officiais did not fully enforce Tang law. 29 73TAM509:8/9(a) in TCWS-1: 9:44-47; TCWS-2: 4:277-278. 26

322

Les Sogdiens en Chine

-

Another fascinating aspect of the 707 Chonghua Township inspection record is the lack of adult males in the Sogdian and non-Sogdian communities. As Tables la and lb demonstrate, males 16-60 (TM) made up only 14% of the Sogdian and 17% of the non-Sogdian populations. In contrast women of the same âge group (TF) comprised 31% of Sogdians and 39% of non-Sogdians. What could be the cause of such skewed demographics? Clues begin to appear if we look at sex ratios of different âge groups in both communities that appear in Table 3.

Table 3; Chonghua Township Male-Fewtale Sex Ratios Age Category Infant (1-2) Child (3-15) Adult (16-60) Adult (16-60), adjusted for deceased males

Sogdians 1:1.3 1:1.8 1:2.2 1:1.6

Non-Sogdians 1:1 1:1 1:2.3 1:1.2

Among infants and children, non-Sogdians demonstrate a normal even split between males and females, but the Sogdians experienced a dearth of males from early in childhood. The 1:1.3 male-female ratio for infants could be distorted because the sample (7 males [IM] and 9 females [IF]) is so small, but the 1:1.8 ratio among children, based on a larger sample (13 boys [B] and 23 girls [G]), is firmer. It appears that almost half of the boys departed the Sogdian settlement in Chonghua Township before the âge of fïfteen, while neighboring non-Sogdian boys stayed put.30 The probable explanation is that many Sogdian boys were sent off to become apprentices in commerce and other trades. who visited This hypothesis gains support from the observations of Wei Jie Sogdia as an official Chinese ambassador in the early seventh Century. He noted that in Samarqand boys began to leam to read at âge five, and States explicitly that “when they attain rudimentary literacy, they are sent off {qian Je ) to study commerce.”31 Presumably this means that they apprenticed with itinérant traders on the road. Male dominance of Sogdian commerce is attested by the fact that ail of the long-distance Sogdian merchants and laborers in surviving Tang travel permit documents were males, and very few women are mentioned in the more than 650 Sogdian inscriptions found in the Upper Indus Valley (Sims-Williams 1996: 56; de la Vaissière 2002: 85-89; Skaff 2003: Appendix I). Tuming to adults, we can see that the two communities hâve almost identical sex ratio imbalances of more than two women for every man. This is far greater than the sex ratio of 1.25 seen in the previously mentioned document from Hexi.32 This wouid explain why Sogdians, despite their presumed wealth, averaged fewer children per household than the previously mentioned households in Hexi. If we correct for missing boys and male infants in Chonghua Township, the rate of children per family wouid be 2.6, which is about the same as the figure from Hexi. 31 is now lost, but parts hâve been preserved in Tong dian. For this Wei Jie’s book Xifan ji particular passage, see TD 193:5256; Chavannes, 1900: 133 n. 5. 32 According to my calculations based upon Ikeda’s (1973: 137-139) data on 153 individuals, there were 32 males and 40 females in the 16-60 âge group, giving a male-female ratio of 1.25. 30

Jonathan K. Skaff

Documenting Sogdian society at Turfan

323

Table 4: Sogdian and Han & other Non-Sogdian Widows in Chonghua Township ID R5 R06 R07 R08 R09 RIO Rll R12 R13 R14 R19

R20 R21 R22 R23 R24 R25 R26

R27 R28 R29 R30 R31 R44 R45 25

EW 0 0 0 0 0 0 1 0 0 1 0 0 2 1 0 1 0 0 ? 0 0 0 0 0

Ave %Ho

? 6 0.3 22

%Pop

4

Total

Sogdian W 0 1 0 0 0 0 1 0 0 0 1 0 1 0 0 0 0 0 ? 2 0 1 0 0 ? 7 0.3 26 5

TW 0 1 0 0 0 0 2 0 0

1 1 0 3 1 0 1 0 0 ? 2 0 1 0 0 ? 13 0.6

39 9

ID RI R2 R3 R4 R15 R16 RI 7 RI 8 R32 R33 R34 R35

R36 R37 R38 R39 R40 R41 R42 R43 R46 R47

22

Han & other Non-Sogdian EW W TW 0 0 0 0 1 1 1 0 1 0 0 0 1 0 1 2 0 2 1 0 1 ] 0 1 0 0 0 0 0 0 0 0 0 0 1 1 1 0 1 1 0 1 3 0 3 0 0 0 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 ? 1 1

13 0.6 48

2 0.1 10

15 0.7 55

16

2

18

Key: EW=Elderly Widows, 61+; W=Widows, 22-60; TW=Total Widows.

Among non-Sogdians5 their ratio of 2.3 females for every male represents a steep 130% increase in the imbalance between childhood and adulthood. In comparison the Sogdian imbalance increased more moderately by 22% to 2.2. Clues to explain the disappearance of adult males begin to emerge when we compare the rates of widowhood in both societies, which is tabulated in Table 4. Widows (W), âge 22-60, comprise 16% of the non-Sogdian population, but only 5% of Sogdians. Obviously, men were dying at a greater rate among the non-Sogdians. Assuming that each widow represents a dead adult male, we can adjust the sex ratios for both communities.33 We 33

_

It is valid to assume that each widow generally represents one dead man for two reasons. One, given the shortage of men, widow remarriage must hâve been uncommon. Two, concubines of deceased husbands

Les Sogdiens en Chine

324

then discover that non-Sogdian sex ratios retum to more normal levels of L2 women for every man. On the other hand, the Sogdian adult sex ratio is reduced more moderately to 1:1.6, which is a level slightly lower than that of the children in their community. The skewed adult Sogdian sex ratio in Chonghua Township is roughly corroborated by a smaller sample of data in Table 5 that includes adult Sogdians 34 appearing in census records transcribed in Appendix I.* Fifteen separate households, spanning the period from 640 to 750, contained a total of twelve women and eight men with Sogdian sumames. This works out to a sex ratio of 1:1.5.

Table 57 Households with Adult Sogdians ID Cl C2 C3 C3

C3 C5 C12 C16 C17 C23 C23 C28 C29 C29 C31 C32 C33 C34 C44

Date ca. 640 ca. 640 ca. 640

ca. 640 ca. 640 ca. 640 ca. 640 ca. 640 ca. 640 668 668 689 689 689 689 689 689 ca. 692 ca. 750

Total Sogdian

Husbstnd/Single Male Gao Shami He Mianren Xin Yanxi Xin [,..]ren Xin Huaizhen Meng Hairen Long Zhuzhu Cao Zaojini Shi Benning Kang Xianghuai Kang Haida Shih Gouren Di Jisheng Di

Kang Caibao Kang Ludu Wang Chuli 8

Age 37 55 64 ? 21 44 35 30 22 62 30 27 28 d. d. 40+ 40 d. 68

Wife/Widow

Age

Mi An Meng Qu Kang

22 42 64 39 18 27 20

Shih Kang An An Sun Tang

Concubine

Gao

28

Shih

26

25

16 63 30

An

22

Cao Duofu Gao Kan Shih Nubei Cao 11

78 30 34 36 57 1

Key: Sogdian names are in bold face type. ID=Identification number ofhousehold in Appendix I.

The persistent sex imbalance among Sogdians makes it appear that the boys who left during their childhood never came back as adults, or if they did retum, they may hâve stayed off census registers illicitly to avoid tax, labor and military obligations because they did not need govemment allocated land to support families.35 Anecdotal evidence were not considered widows. See the concubine, Shih, in the Di household (Appendix I: C29). 34 This does not include ail Sogdian adults extant in the census records, but only those for which a füll household record or at least the names of husband and wife survive. Even though this sample is smaller than if ail extant adults had been included, it is less biased. Some census records tend to preserve more men or women, depending on how the document was eut or decayed, because the name of the household head (usually male) was written higher and/or to the right of the rest of the family. For example, 64TAM15:24-25 (TCWS-1: 4:53-56; TCWS-2: 2:33-34) créâtes a bias toward males because in eight cases only names of husbands are preserved while names of wives survive alone in only six instances. 35 The elevated levels of unnatural deaths among males and conséquent high rate of widowhood in both

Jonathan K. Skaff

Documenting Sogdian society at Turfan

325

from Sogdian Ancient Letters 1 and 3 supports the possibility that Sogdian diaspora

communities at other times and places may hâve suffered from a lack of males. The letters testify to the anguish of the Sogdian wife Miwnay, abandoned by her husband in fourth Century Dunhuang.*36 Miwnay’s pain may hâve been shared by other Sogdian women, such as the widow Shih Nubei in late seventh Century Turfan. At âge thirty-six she lived alone with her daughter Gushi. Her son and another daughter had passed away in the previous year and her husband had died earlier from unknown causes (Appendix I: C34). Judging from the plights of Miwnay and Shih Nubei, Sogdian women may hâve frequently experienced hardships as they tried to manage households without the support of husbands, Nonetheless, not all wives may hâve been left beliind when their husbands moved to new places. At Chonghua Township, Sogdian households appear to hâve been more mobile as an entire unit than their Han neighbors. In the inspection record, even though newly arriving widows and orphans at Chonghua Township were evenly split between Sogdians and Han Chinese, who may hâve been fleeing warfare,37 the only households to abscond for more than ten years were six Sogdian ones, which constituted one quarter of Sogdian families in the document.38 Interestingly enough, ail but one of the absconding households were large, including the three largest households listed in the entire document.39 We can surmise that most of the absconders were entrepreneurs who left with their entire families in search of b etter business opportunities, even though abandoning their land in most cases would hâve been a violation of Tang law.40 Tang communities, but especially the Han, is an interesting phenomenon that is beyond the scope of this paper, but deserving of further study. Dong Guodong explains the overall sex imbalance in this document as a resuit of males absconding from their household registrations in order to avoid taxation and military service. He believes that many widows were “fakes” whose husbands probably were alive and hiding from tax officiais. Although this may explain some instances of widowhood, Dong’s aggregated census data of mostly Han at Turfan in the 640 to 671 period do not demonstrate any kind of sex imbalance, even though the same tax and military Systems were in place (Dong, 1993: 381, 402, 421-422). A more important factor to explain the sex imbalance probably was warfare. The Sogdian community in Chonghua Township was ahnost evenly divided between widows and elderly widows, whereas the vast majority of Han widows were elderly. A possible explanation for this phenomenon would be that the Sogdian community mostly arrived after a period of warfare approximately thirty years earlier during which a large proportion of Turfan ’s males had been killed. This would coincide with the Tang’s warfare with the Tibetans for control ofthe Tarim Basin from 670-692 (Beckwith, 1987: 37-75; Skaff, 1998: 100). A census document from the 690s mentions a household of at least fïve related Han women three of whom are relatively young widows at âges 32, 38, and 42 (Appendix I: C35). Two other census documents specifïcally mention Turfan men lost in battle (moluo during military campaigns in the Tarim Basin in 685-686. See Di Sijun (TTDS-2: XXXIII, p. 74, In 38) and three Sogdian brothers with Han-style given names Kang Anzhu Kang Anding and Kang Anyi (72TAM184:12/b(a) in TCWS-1: 8:280-1; TCWS-2: 4:127). See Nicholas Sims-Williams’ translations of Sogdian Ancient Letters 1 and 3 in this volume p. 183. 37 See note 35 above.

—

38Tablela:R9-R14.

39Tablela:R9,R12,R14. 40

It was only permissible to move from an area with a shortage of arable land, like Turfan, to places with a surplus (Twitchett, 1970: 128, art. 12; 145, art. 14, 15). It would hâve been difficult to arrange a legal transfer because land would hâve been in short supply in all of the East Turkestan oasis cities and many places in the interior (Twitchett, 1970, 7-11).

326

_

Les Sogdiens en Chine

authorities attempted to control this sort of migration. As mentioned above, the Sogdian merchant from Turfan, Shi Zhëmat-yân, traveled with a certificate of ownership in which guarantors vouched that Shi had a house and family and pledged to pay his taxes if he did not retum.41 Obviously, these précautions were not always effective. Perhaps absconding Sogdians bribed officiais to obtain paperwork that would allow them to move their households illicitly. Taken in its entirety, the démographie data in the 707 inspection record demonstrates that the Sogdians at Chonghua Township had a distinctive household structure in comparison to their mostly Han neighbors. Furthermore, the relatively large household sizes, lack of males, and tendency to be transient strongly suggest that the Sogdian households had wealth and traditions that distinguished them from others living in their section of Turfan. In particular, their greater mobility and propensity to send young males away from households presumably continued typical customs of their ancestral homeland.

Marriage and Naming Patterns Despite the distinctive éléments of Sogdian culture in the diaspora, we should not assume that the Turfan community was an exact replica of the home country. Analysis of marriage and child naming patterns recorded in Turfan census documents intimâtes that Turfan’s Sogdians probably wove a twisted path between cultural distinction and assimilation. Unfortunately, our most complété source of information on Sogdian society at Turfan, the 707 inspection record, is much less useful for a study of marital or naming practices because it only records the names of legal household heads, who mostly were male, and any other adult males in the household. This forces us to rely on the more limited and less reliable census document data compiled in Appendix I to leam the names of family members needed to study marriage and naming patterns. Consequently, the analysis in this portion of the paper will be more impressionistic than in the preceding section. Tuming first to marriages, there are a total of fourteen unions involving Sogdians in which the sumames of husband and wife or concubine are extant. The data are summarized in Tables 6a-6c. Based upon the male-female sex ratios drawn from the 707 inspection record showing that Sogdian women outnumbered men, we would expect to find Sogdian males normally marrying within their ethnie group because of the surplus of female Sogdians, while women would be much more likely than men to marry outside of their ethnie group. The limited data in Tables 6a-6c support the latter hypothesis but not the former. There are three purely Sogdian couples (Table 6a), four cases of Sogdian males marrying Han females (Table 6b), and six Sogdian wives and one concubine married to men with non-Sogdian sumames (Table 6c). The approximately equal numbers of Sogdian men marrying inside and outside of their ethnie group could demonstrate a general openness to ethnie interchange at a culturally diverse frontier trade city, or a distortion caused by the small size of the sample. It is instructive that the two households in Table 6b for which we hâve complété 41

See note 29 above.

Jonathan K. Skaff

Documenting Sogdian society at Turfan

327

démographie information (C23 and C32) are large and presumably wealthy. Perhaps these Sogdian households practiced a strategy of intermarrying with dite provincial Han households to increase their influence locally. Table 6a: Sogdian-Sogdian Marriages

Date ca. 640 ca. 640 ca. 640 Total Sogdians

ID C2 C16 C17

Husband

He Mianren Cao Zaojini SM Benning 3

Table 6b: Intermarriage: Sogdian Husband Date ID C23 668 C23 668 C32 689 C33 689 Total Sogdians

Wife An An An 3

Husband Kang Xianghuai Kang Haida Kang C”ib30 Kang Lndu 4

Wife Sun Tang Gao Kan 0

Table 6c: Intermarriage: Sogdian Wife or Concubine Date ID ca. 640 CI ca. 640 C3 ca. 640 C5 ca. 640 C12 689 C29 689 C29 ca. 750 C44 Total Sogdians

Husband Gao Shami Xin Huaizhen Meng Hairen Long Zhuzhu Di Jisheng Di Wang Chuli

Wife

Concubine

Mi Kang Shih Kang An

Gao

Shih

Cao 6

1

Key: Sogdian names are in bold face type. ID=Identification number of household in Appendix I.

For the most part the marriages are monogamous unions of husband and wife, but there are a few cases of concubinage. One is the marriage between a Sogdian woman sumamed Shih and her Han husband, Meng Hairen, who also took a concubine with the Han name of Gao (Table 6c: C5). Another case is the household headed by Di Jisheng, âge twenty-eight, that includes his wife, sumamed An, his deceased father’s concubine sumamed Shih, and a thirteen year old musician, He Liji (Appendix I: C29). The latter three sumames are clearly Sogdian. This household’s only child is a three-year-old girl, Nuosheng, whose name appears to be non-Han. Although the husband’s sumame, Di, is associated with the Gaoju people, the original inhabitants of Turfan, Rong Xinjiang and Valerie Hansen hâve shown that the Di sumame may hâve been used by Sogdians or other non-Han who lived and associated with Sogdians in Turfan and China’s interior.42 Even though we cannot be certain of Di Jisheng’s ethnicity, it is noteworthy that this young man, who only had ten mu of land for grain, was wealthy enough to maintain a musician and personal retainer in his household. He holds the title of “kinsman of an official” (pinzi pn“F)- Although this information is tantalizing, many interesting ques¬ tions remain unanswered. How much of the family’s wealth came from official service? Did income from business allow a family member to obtain the éducation in Chinese 42

See Rong, 2001: 131-133 and Hansen’s paper in this volume.

328

Les Sogdiens en Chine

needed to become an official? Why did Di Jisheng and his father favor unions with Sogdian women? There are two other known examples of concubinage besides the ones listed in the tables above. One case is from the 707 inspection record, involving a large and apparently wealthy Household headed by a Sogdian male, Kang Lushan âge forty-nine, whose wife and concubine’s sumames are unknown.43 The final example occurs in a census register dated 647 where a twenty-six year old Sogdian female, whose sumame was An, was listed as a concubine. The husband, [...] Gou 0^5 âge forty-one, whose sumame is lost, also had a wife, âge thirty-seven, with the non-Han sumame Linghu æ.44 This apparently was a fairly wealthy household because it also included three children and a young woman slave. The preference for monogamous unions with occasional cases of concubinage may be a continuation of Sogdian custom with some modification to suit Tang marital law. In Sogdia monogamy seems to hâve been preferred, but multiple wives may hâve been legally acceptable. Evidence for this cornes from the only existing Sogdian language marriage contract, dated to the eighth Century and found along with other documents at Mount Mugh in western Tajikistan, which penalizes the husband for taking additional wives or concubines. This implies that polygyny was legally permissible, but that there may hâve been a customary preference for a single female spouse.45 Han custom, which was codifiée! in Tang law, permitted taking one wife and multiple concubines. The pen¬ alty for having two wives was one year of penal servitude.46 Adhering to Tang law may hâve required that Sogdians in Turfan treat second wives as concubines, or at least give the appearance of doing so. We are fortunate to hâve an opportunity to compare the marriage patterns of Sogdi¬ ans at Turfan with some elsewhere in the East Asian diaspora because of Rong Xinjiang’s research on seventh-century elite Sogdians in China' s interior. In contrast to the apparent relative openness to intermarriage among Turfan’s Sogdians, Rong fïnds that those in the interior tended to marry among themselves. His slightly larger sample of data on twenty-one marriages mentioned in tomb inscriptions reveals that fourteen or two-thirds involve Sogdian males married to Sogdian females. Of the other seven mar¬ riages, three involved Sogdian females married to non-Han males, including one with the Di sumame, and four were Sogdian males married to non-Sogdian spouses (2 Han and 2 non-Han, including one Di). In both cases of wives with Han sumames, the Sog43

Table la: R9. Skaff (2003) is mistaken to claim that Kang Lushan may hâve had two wives. A young wife, âge sixteen to twenty-one, mentioned in his household must hâve been married to his son, âge sixteen to twenty-one, rather than to Kang. I had assumed that the son was too young to be married, but the example to be mentioned below of the brothers Afufen, 20, and Ningpi, 16, married to younger girls sumamed Cao shows that males and females could marry before the âge of twenty (see note 55 below). M 69TKM39:9/4(a) in TCWS-1: 6:101; TCWS-2: 3:53. 45 Frye, 1996: 195; Yakubovich, 2002; la Vaissière, 2002: 157. In the contract the bride is a Sogdian noblewoman and the groom carries the Turkic title of Tegin, which normally indicates a close relative of a Qaghan. These special circums tances could mean that this marriage was not typical, but unfortunately, it is ail the evidence that we hâve. For the Chinese description of Turkic and Western Turkic titles, see Chavannes, 1900: 164 n. 3. 46 TLSY 13:255-256, art. 177; Johnson, 1997, 154-5, art. 177. Ebrey, (1993: 47) describes the legal situation later under the Song Dynasty, but the fondamental legal principles were shared by the Tang.

Jonathan K. Skaff

Documenting Sogdian society at Turfan

329

dian husbands were high officiais serving at the capital. Rong do es not hâve any data on concubinage, but this may be due to the nature of the evidence. Concubines would hâve been less likely to hâve been accorded the honor of personal tomb inscriptions or men¬ tion in the epitaphs of their husbands. Rong concludes that elite Sogdians living in the interior generally preferred to marry among themselves or to other West Asian ethnicities, and only those whose lives took them away from the West Asian community, like high officiais, would marry Han.47 Interestingly, these Sogdian males and females married outside of their ethnicity at about equal rates, perhaps indicating that they did not suffer from a shortage of males. Could this be because the elite Sogdians, who could affbrd to build elaborate Han-style tombs, were mostly involved in governmental serv¬ ice and were less likely to send off sons to engage in business? Fürther research will be required to résolve the discrepancies between Sogdian marriage patterns in the interior and the frontier city of Turfan. If the data represent real social or regional distinctions, there are some possible explanations. Perhaps in a Silk Road oasis city, like Turfan, Han and other ethnicities were more likely to corne into contact with one another and barriers to interethnic marriages were more easily breached than in the interior.48 Another possibility is that the différences represent class-based distinctions in customs among Sogdians. The high Sogdian elite of the inte¬ rior, whose biographical data hâve been preserved because they could afford to build large tombs with funerary inscriptions, may hâve been more conservative about preserving traditions through intra-ethnic marriage. Hopefully more evidence can be located to résolve these problems. One thing that marriages at Turfan and in the interior had in common was that every known union involved husbands and wives of different sumames. For example, in Ta¬ ble 6a ail three wives hâve the An sumame, but the husbands are sumamed He, Cao, and Shi. Since the seven Chinese-language Sogdian sumames theoretically were assigned to people according to their place of origin, this implies that Sogdians who married within their diaspora communities would hâve had children with ancestral ties to at least two Sogdian cities. Consequently, we cannot assume that a Sogdian with a particular sumame had close ties with the associated city. Moreover, marriages and general social interaction within the Sogdian diaspora community would hâve tended to erase regional différences that might hâve existed in the ancestral homeland. A new Sogdian diaspora culture would hâve been created, blending influences from the cultures of multiple Sogdian cities and other ethnicities, including Han. However, we should not assume that this mixed culture was uniform across the broad expanse of the Tang Empire. As we can surmise from the differing patterns of Sogdian intermarriage at Turfan and the interior, regional variations may hâve developed among China’s Sog¬ dian diaspora communities, while dass différences may hâve existed within them. Another method of deepening our understanding of Sogdian assimilation and cul¬ tural change is to analyze naming patterns. In the Turfan census documents some peo¬ ple with Sogdian sumames hâve non-Han names phonetically transcribed into Chinese 47

Rong, 2001 : 132-135. See Skaff (2004), which demonstrates that identities and loyalties tended to be more flexible in premodem frontier zones.

48

330

_

Les Sogdiens en Chine

while others hâve Han-style given names.49 In the 707 inspection record most of the Sogdians hâve transcribed foreign given names, such as Kang Lushan (Roldishan) and An Fuzhitai (Buti-dhây), but a minority hâve typical Han ones, such as Kang Enyi and An Dezhong.50 In the census documents summarized in Appendix I, there appears to be a roughly even split between Sogdian households using transcribed Sogdian and true Han given names. There are even a few cases of Sogdian families adopting a typical Han naming custom of giving a common first syllable to the given names of sons. The Kang brothers Anzhu, Anding and Anyi—are one example.51 Surviving Turfan census documents, which will be discussed below can provide illuminating anecdotal evidence of Sogdian assimilation in naming practices, but the cases are too sporadic to discem any pattern over time. However, a Tang tax register from the oasis city of Dunhuang sheds more light on the problem because it contains a larger sample of names. Ikeda On has studied this document, dated 751, which shows the labor and military service adult males owed to the Tang govemment. The section of which under Tang rule the document dealing with “Bukhara Town” (Ancheng was renamed Conghua “Surrender to the Transformation [to Sinicization]” Township consisted of 257 households more than 90% of which were (Conghuaxiang headed by non-Han who mostly had Sogdian sumames.52 Among those with Sogdian sumames, Sogdian and Han-style given names occur in roughly equal numbers, but Ikeda found a tendency towards sinified names among the younger génération.53 Ikeda’s observation that later générations were more likely to be given Han-style names is not surprising. More problematic is Ikeda’s hypothesis that sinifïcation of given names was the resuit of Sogdian males intermarrying with Han females. Since only adult men appeared in the Dunhuang document that Ikeda studied, he could not supply evidence to support his idea. However, an examination of the Turfan census documents can help us to see that intermarriage only is a partial explanation for

—

49

1 hâve decided to include this section of the paper on naming practices, even though I am not a qualified linguist of the Chinese or Sogdian languages, because I hope to draw the attention of Sogdian philologists to names that are important for the understanding of Sogdian social history. The standards used to détermine the origins of given names are adapted from those established by Ikeda On in his study of Dunhuang’s Sogdians (1965: 62; 1993: 156). Han names are considered to be those that do not seem awkward in meaning or pronunciation. On the other h and, names with unfelicitous connotations or sounds are considered to be transcribed foreign words. In the Turfan census records, as in modem China, Han typically hâve nice-sounding given names with propitious meanings, but this standard of identification requires some caution because Han of low social status sometimes hâve simple or unpleasant names. For example, see the Han household of Song Bazi “eighth son” and his mother Gao Heimian “black face” (Appendix I: C43). Questionable non-Han names were cross-checked in the Turfen documents name index (Li and Wang 1996) to verify whether other Sogdians ever used the same appellations. Final say on the origins of given names will hâve to corne from qualified linguists. 50 Table la: R7, R9, RI 1, R19. Reconstructed Sogdian names are from the appendix to Valerie Hansen ’s paper in this volume p. 305. 51 See note 35 above. Another example are the sons of Kang Caibao (Appendix I: C32), discussed below. 52 Ikeda, 1965: 49-59; 1993: 140-154. The original document is P. 3559 (=3664 V°), P. 2657, P 3018 V° in TTDS-2: CI. Ikeda’s statistics on the percentage of Sogdians in Conghua Township may be inflated are Sogdian. The present article does not count and He because he assumes that the sumames Luo these sumames as Sogdian. 53 Ikeda, 1965, 62-70; Ikeda, 1993, 156-171.

æ

Jonathan K. Skaff

Documenting_ Sogdian society at Turfan

331

assimilation in name-giving practices. Some households listed in Table 6b demonstrate that there may hâve been a connection between intermarriage and a tendency for Sogdi-

ans to hâve Han-style given names. There is one clear case of a Sogdian male with a non-Han given name married to a Han female who gave their daughter a Han-style name. The husband, Kang Ludu—whose given name appears to be a transcribed foreign word that in Chinese means the very un-Confiician “unrestrained”—and his wife sumamed Kan had a daughter with a typical Han feminine appellation, Miaoying “beautiful blossom.” (Appendix I: C33) In another household the father, Kang Xianghuai, and his son, Kang Haida, represent two générations of men married to women with Han sumames. Xianghuai’s given name is of unclear origin, but Haida “attainments as great as the océan,” whose mother and wife had Han sumames, œrtainly had a Han-style given name. The children in the household also appear to hâve sinified given names. (Appendix I: C23) Both of these cases demonstrate that Han wives sometimes may hâve exerted influence over naming of children. On the other hand, there are several examples in the Turfan documents indicating that other factors may hâve been involved in choosing names for children. One case of intermarriage that provides insight into the extent of Han cultural influence over non-Han, appears in a household déclaration dated 640. The wife had a typical Han Although the husband’s name is lost, the children ail hâve sumame of Zhang non-Han given names, so he must hâve been non-Han and perhaps Sogdian because the two oldest sons living within the household hâve wives with Sogdian sumames.54 Afufen âge twenty, had a fifteen year old wife sumamed Cao. Ningpi âge sixteen, was married to a twelve year old girl also sumamed Cao.55 Regardless of whether their father is Sogdian, this example demonstrates that a non-Han male’s marriage to a Han female did not necessarily lead to a sudden “sinification” of a family. Despite the wife’s presumed Han ethnicity, the husband’s customs were followed in name giving and the sons were married to Sogdian women. The Han wife did not hâve an overriding cultural influence on this particular household.56 Further complicating the situation, Sogdian couples also might choose Han-style names for their children. For example, He Mianren, a Sogdian who appears to hâve a non-Han given name and his wife, a woman with the Sogdian sumame An, had a son with a Han-style given name, Hailong “eminence as great as the océan.” (Appendix I: Cl 2). In other cases, families appear to hâve adopted mixed naming patterns for their children. One example is He Mopan whose given name appears to be non-Han. He was an eighty-year-old former official whose wife’s sumame is unknown. His sons, bom “safe treasure,” âge only one year apart, may exhibit mixed name types. Anbao “baldy,” âge thirty-six, may hâve a thirty-five, has a Han-style name, while Tuzi transliterated foreign name given the fact that “baldy” is an odd appellation for an offi¬ cial’s son, and a number of other Sogdians in the Turfan documents share the name.57 In In accordance with normal Tang practice, the census document only records the given names of children. 67TAM78:21(a),22(a)in TCWS-1: 4:75-76; TCWS-2: 2:46. 56 Even if the wife, Zhang, had a non-Han mother, the conclusion still holds true that Han culture at Turfan was not overwhelmingly influential on naming practices. 57 Table la: R12. Some examples of Sogdians in the Turfan documents who hâve this same name are An Tuzi (72TAMI 50:49 in TCWS-1: 6:56; TCWS-2: 3:29), the ironsmith Cao Tuzi (64TAM61 :16(a) in

55

332

_

Les Sogdiens en Chine

the Dunhuang labor service register, Ikeda found ten out of twenty-eight sets of Sogdian brothers had mixed Han and non-Han names.* 58 A similar phenomenon can be seen in the généalogies of local elite Sogdian families that hâve been partially reconstructed from several tomb inscriptions found in Guyuan El IM in the modem Ningxia Hui (588-669) had two sons, one with Autonomous Region. For example, Shih Hedan and the other with a Han name, Huaiqing 'R®, “harbora foreign name, Huluo ing blessings”. 59 There is even one indication that Sogdian women married to Han men could exert influence on the naming of their children. In the household of a Han male, the county scribe Meng Hairen, who had a wife with the Sogdian sumame Shih and a mother and concubine carrying the Han sumames Zhang and Gao respectively, three children exhibit a stränge mixture of naming practices. A son Jiande “establish morality,” âge four, and daughter Guangying “brilliant flower,” âge five, hâve typical Han names, but the oldest son Heinu “black slave,” âge six, has a stränge name for a son of a county scribe. Even two female slaves in the household, Moxiang “tip of fragrance” and Shouxiang “preserving fragrance,” hâve more félicitons names in the Chinese language (Appendix I: C5). Could the boy hâve a transcribed Sogdian name chosen by the wife? Is this an example, ever so slight, of Sogdian influence on the dominant Han culture? The answers to these questions may be yes, based on the fact that there was another Sogdian with this name, Kang Heinu, who lived prior to 640 in Anle Village, the mainly Sog¬ dian settlement listed in the 707 inspection record.60 The above examples demonstrate that Sogdian-Han intermarriage was only one cause of assimilation in naming practices. The fact that Sogdian couples could give their children Han or a mixture of Han and Sogdian names shows that other factors were involved. Naming children in the Han style presumably became more common as Sogdians leamed the language and adapted to the customs of the dominant Han culture. Marital unions of Sogdian men and Han women were not required for this to happen, but probably accelerated the process. Perhaps local eûtes who were wealthier and better educated in the Chinese language, like the former official He Mopan, led the community in steps toward assimilation. Still, the case of He Mopan demonstrates that despite the tendency toward assimilation in the naming of his first son, he may hâve retained a parallel desire for cultural distinction in choosing a name for his second son. And even Sogdian women, such as Meng Hairen’s wife Shih, may hâve influenced the cultural practices of their adopted Han households. The household of Kang Caibao provides further insights into how Sogdians could TCWS-1: 6:467; TCWS-2: 3:240), and Kang Tuzi, âge 27, which is rather young to be bald (64TAM5:84 in TCWS-1: 6:361; TCWS-2: 3 :186). If philological study of the name reveals that Tuzi is not derived from a Sogdian word, we may hâve to assume that Turfan’s Sogdians were especially prone to baldness! 58 Ikeda, 1965: 69; Ikeda, 1993: 170. 59Luo, 1996: 68-72, 216. Shih Hedan was married to a Sogdian woman sumamed Kang until she passed away at the âge of thirty-nine. He remarried a Han woman, sumamed Zhang JJ. We do not know the identity of the wife who bore the sons. Could the variation in naming practices hâve resulted from the sons being bom to different mothers? 60 73TAM517:06/l(a) in TCWS-1: 4: appendix, p. 6; TCWS-2: 1:256. More definitive answers to these questions must await research by a qualified Sogdian linguist.

Jonathan K. Skaff

Documenting Sogdian society at Turfan

333

distinguish themselves culturally while taking steps towards assimilation via intemarriage and adoption of Han-style given names (Appendix I: C32). A household déclaration informs us that Kang was over forty years of âge and lived with his wife, two brothers, a sister-in-law, three daughters, four sons, one nephew, and a female slave. Kang and his brother Zhenbao, âge thirty-six, had Han-style given names, but another brother Fangyi, âge thirty-four, has a name of unclear origin that could be non-Han. Caibao’s wife had a Han sumame, Gao, and his four sons—Xuanying, Xuansu, Xuanshou, and Xuanzhong—had given names that seem Han and follow the typical Han naming practice of a common first syllable for names of brothers. Despite the fact that the children’s given names exhibit a high degree of assimilation, there are indications that this may be a family of Sogdian entrepreneurs. First, Kang Caibao’s given name ostensibly is Han, but its meaning is a double entendre that might be fevored by merchant parents. Literally, it signifies “talent at treasure” and is also an exact homophone in early and late middle Chinese for a common character combination (caibao MW) meaning “money and jewels.”61 At least one source of treasure for Kang’s household was grapes because the household déclaration reveals that the family farmed more than five and one-half mu of vineyards. Grapes were a crop used to produce wine or raisins, which were valuable commodities because they could be preserved and transported to sell in places where they commanded higher prices (Trombert, 2002a). In addition to the family’s commercial interests, the household déclaration alludes to the possibility that the family was mobile in a way that seems suspiciously Sogdian. The document notes that the six youngest children—the four sons, a daughter, and the nephew—were not entered in the previous census register, which would hâve been composed three years earlier in 686.62 There was no financial incentive to hide the chil¬ dren from Tang officiais because under the Tang taxation System the children would not hâve affected the household’s land allocation or tax bürden.63 Since ail of the children were over the âge of three, they must hâve lived elsewhere during the previous registra¬ tion period. It is unlikely that this was connected to the practice of sending away boys on apprenticeships because one child was a girl and four of the boys, who are from the âges of five to seven, are too young to hâve been apprenticed three years previously. The most likely possibility is that the older family members moved to Turfan from a previous place of résidence and the children came several years later. Perhaps the older members of the family moved to Turfan to pursue viticulture and other business opportunities and sent for the small children after becoming established. This implies that this family of mixed ethnicity, which exhibits a high degree of assimilation in nam¬ ing and marriage practices, may hâve retained a Sogdian cultural preference for commercial pursuits, and maintained contacts with a possibly Sogdian social network that extended outside of Turfan. 61

On the early and late middle Chinese pronunciations of the characters, see Pulleyblank, 1991: 44. During the Tang, census registers were compiled triennially on a schedule set according to the twelve branches of the traditional sixty-year cycle of dating. Until 731 registers were drafted in the chou ü, chen Jg, wei and xu brauch years (TTDS-2: 7). 686 was a xu year and 689 was a chou year. 63 Adult males were the main group eligible for land distribution and taxation. See note 22 above for others who might be given land and taxed. 62

334

Les Sogdiens en Chine Conclusion

The scattered remains of census records from Tang Turfan present a mixed blessing to the premodem historian. On the one hand, we are fortunate that these documents hâve survived for more than a millennium, allowing us to sketch a hazy outline of local Sogdian society. On the other hand, the relative paucity of census data leaves us hungering for more information to fïll in numerous details needed to sharpen our reconstructed picture of the past. Yet, most of Turfan ’s tombs remain unexcavated, so there is hope that more evidence will be put at our disposai in the future. The best preserved census document, the 707 Chonghua Township inspection record of registered households, demonstrates that Sogdian households had some features that distinguished them from their mostly Han neighbors. The Sogdians tended to hâve larger household sizes, lack males from childhood onward, and be more transitory. The hypothesis put forward to explain larger households is that the Sogdian males supplemented subsistence farming income by engaging in activities as itinérant merchants, artisans, and entrepreneurial farmers. Moreover, Sogdian propensity to carry out long distance trade and pursue business ventures in new locations would explain the mobility of some households, and lack of males, including boys, who presumably were away on business or involved in apprenticeships. Nonetheless, Turfan’s Sogdian households also show signs of assimilating into the dominant Han culture. Intermarriage was common and some families began to adopt Han naming practices for their children. Assimilation does not appear to hâve been a sudden or direct process. Some families show signs of simultaneous ethnie distinction and assimilation, such as giving one son a Sogdian name and another a Han name. This phenomenon of cultural assimilation combined with ethnie distinction has been noted elsewhere in the Sogdian diaspora in China, and was common among ethnie minorities in other contexts, such as the frontiers of the Roman empire.64 This leads to the conclu¬ sion that Sogdian culture in the diaspora was dynamic. Neither an exact replica of the homeland nor wholly assimilated into Han society, Sogdian culture at Turfan evolved in new directions as Sogdians interacted with each other, other minority ethnicities, and the Han majority.

64

Vivid examples of partial assimilation of Sogdians elsewhere in the diaspora are seventh Century tombs excavated in modem Ningxia that reveal spiritual practices and material culture that are neither purely Sogdian nor purely Han Chinese (Lemer, 2001; Luo, 2001). For comments on Sogdian ethnie assimilation and distinction, see Juliano and Lemer (2001: 296-299). Examples of ethnie groups in late Roman and post-Roman Europe can be found in Pohl (1998).

Jonathan K„ Skaff

Docimenting_Sogdian society at Turfan

335

Appendix I: Tang Turfan Census Records of Complété Households or Incomplète

Households that Preserve Surnames of Both Spouses65 References: 10 Date

C Household 0

head

m 01 ca.

640 02 ca.

640 03 ca. 640

04 ca.

640 e5 ca. 640

06 ca.

640 C7 ca. 640 08

ca. 640 e9 ca. 640 C

ca. 640 c ca. 11 640 10

65

? Y Gao Shami 高沙彌

Y He Mianren 何 免仁 Y Xin Yanxi 辛延熹

S A e g e X

M 37

M 55 M 64

Y Meng Huaix in M 24 孟懷 信 Y Meng Hairen M 44 孟 海仁 County scribe

N Jie Baoyou 解保祐 N Fan Xiangyan 氾相延 N Fan Zhide 氾致得 Y Linghu Longming 令 狐隆明 N Zhuan Huai*er 等懷 兒 Y Long Zhugen 龍朱 艮

M 52 M 35 M 55

M 42

TCW TCW ^08 Other family members & âges Arch. ID (Married adult females 8-1 8-2 v:p.ln 丫甲 designated by sumame, others by given name) 64TKM1:33 4:12.4 2:10 Wife 米 21, Daughter (b)5 34(b), Hanying 漢英 […] 32(a), 36(a) 4:12.6 2:10 Wife An 安 42, Son Hailong ラ每 Same 隆6 Sons: [...]ren 仁 age […レ Huaizhen 壊貞 21,Wife Meng 孟 64, Son […]ren's wife Qu 麴 39, Son Huaizhen 's wife Kang 康18, Grandsons: Hude 護德10, Chuanghu 幢護 8, Hulong 護隆 6,Hufeng 護豐 4, Granddaughter Miaojiang 妙姜 12,F Slave Nanu 捺女 âge [...] Mother Zhang 張 55

ロ

Same

Same

4:12.8 2:10

4:13.1 2:10 6

Mother Zhang 張 71, Wife Shîh 史 27, Concubine Gao 高 28, Son Jiande 建德 4, Daughter Guangying 光英 5, Son Heinu 黑奴 6, M Slave […] 63, F Slaves Moxiang 末香 18, Shouxiang 守香 âge [...] Wife Bai 白 50, Son [...], Daughter Debi 惠婢 4 Wife Suo 索 25, Son Qiuxu 秋敘 âge [...] Wife Dong 董 20, Daughter Yuanshi 元是 âge [...] Wife Jiang 姜 33

Same

4:13.1 2:10 7

64TAM15: 24 Same

4:53.7 2:33 4:54.1 2:33 0

Same

4:54.1 2:33

Same

4:54.1 2:33

1 2

M 31

Wife Zhang 張 20, Child [...]2

Same

4:54.1 2:33

M 38

Wife Linghu 令狐 23, Daughter Boyun 伯暈 3

Same

4:54.1 2:33

3

4

This appendix includes transcriptions from all extant Tang census records except the 707 Chonghua Township Inspection Record of Household Registers. Some of the information from the 707 inspection record appears in this article in Tables la and 1b and full transcriptions can be found in Skaff, 2003: Appendix IL

Les Sogdiens en Chine

336

1 c 13

ca. 640 ca. 640

0 14

ca. 640

ca. 15 640

C

0 16

ca. 640 C ca. 17 640 C ca. 18 640 0 ca. 19 640 0 640s 20 C 21 0 22

C

640s 640s 668

23

Y Long Zhuzhu 龍朱主 N Zhang Rongde 張 容得 N Zhang

M 35

Wife Kang 康 23

Same

4:54. 15

2:33

M 25

Mother Zhang 張 60, Wife Zhang 張 1 0+

831116

4:54. 16

2:33

M 40

Mother Song 宋 60, Wife Zuo 左 64TAM15:2 4:55.3 2:34 age[...] 5

M 35

Wife Zhao 趙 40, Son Shenhai 申海 5, Son Chenghai 成海 âge [-] Wife An 安 25, Daughter Yingnü 英女 5 Wife A寢 安 16

Qinglong 張慶隆 Y Feng Aguzi 馮阿谷子

Y Cao Zaojini 曹 僧居尼 Y Shi Benning 石本享 Y Gao Hailong 高海隆 Y Long Dexiang 龍德相 N Zhao Xixi 趙喜喜 N Zhang Juzi 張雛子 N Zuo Xiyuan 左喜願 N Kang Xianghuai 康 相懷

M 30 M 22 M 40 M 30 M 57 M 40+ M 40+

M 62

N Tian Longde 田隆德

M 20+

0 25

Late 600s

M 60+

C

689

N Wang Chenhuan 王 辰歡 Y Ning Hecai 甯和才

26

0 27

C

689

689

28

689 29

Y Wang Longhai 王隆海 Disabled Y Shih Gouren 史苟仁 Y Di Jisheng 翟 急生 Kinsman of official

4:55.4 2:34

Same

4:55.5 2:34

Same

4:55.6 2:34

Wife Ma 馬 20, Son Wu 武 âge Same [-] Mother Long 龍 60, Wife Suo 索 Same age […] Wife Wang 王 32 68TAM103: 18/11-1 Wife Ken 肯 30+,Son Fengzi 豐 68TAM103: 18/10 子 15, Daughter [...]nu 口女 9 Wife Zhang 張 âge [...] 68TAM103: 182-1 ) Wife Sun 孫 63, Son Haida 海達 72TAM179: 30 Guardsman, Haida' s Wife 16/4e)- 1 Tang 唐 30, Haida's daughter Dongshu 冬 鼠 3, Sons: […]zi 子 21,Huijun 惠 俊 13,Dazi 達 子 11 Mother Zhu 朱 50+,Sister 73TAM222: Shengren 勝 忍 30+,Older sister 54/9 (a) [-] Wife Li 李 60+,Son Shouqiu 守 66TAM67: 13 丘 30+,Shouqiu's wife Li 李 age[...] Mother Zhao 趙 52, Sister Heren 64TAM35: 和忍 13, Dead sister Hezhen 和 59 g) 貞 22, Dead sister Luosheng '羅 勝15 Brother Longzhu 隆住 41 64TAM35: guardsman, Longzhu' s wife Di 60(a) 翟 35

―

一

ー・

・

碎̶,

一

4:55.7 2:34 4:56.8 2:34 E

4:225- 2:129 6.1-2 4:228 2:130 4:229

=1 /

「エ! .

2:130

7:118- 3:3629 3

ロ

671

C 24

Sanle

M 14

M 51

M 27 M 28

Same

Wife An 安 22, Dead fäther's 64TAM35: concubine Shih 史 26, Daughter 61® Nuosheng 那勝 3, Musician He Liji 何 豊吉 13, Retainer Zan Atu 咎阿吐 2 ]

7:131

3:368

7:301- 3:441 2

7:414- 3:498 15

XXIII

7:416- 3:5002 8 1

7:419

3:503

7:420- 3:503- S 2 4

3

%

い」

Jonathan K. Skaff

c

689

30 C 689 31

"c

689

32

C 689 33 C c. 34 692 C 6905 35

C 714 36

C 716 37

C 716 38 C 716 39 C 716 40

C 716 41

Y Yang Zhixiang 楊 支香 Y Cao Duofu

Documenting Sogdian society at Turfan F 40

Son Mangnu 盲奴 4

F 78 曹 多富 Y Kang Caibao M 40+ Wife Gao 高 30, Brothers: Zhenbao 真寶 36, Fangyi 方 藝 康才寶 34,Younger brother^ wife […レ Sons: Xuanying 玄應 10+, Xuansu 玄素 7, Xuanshou 玄壽 4,and Xuanzhong 玄忠 5, Daughters: Shengjiang 勝姜 20+,Xingtan 行檀 12,Pujing 普 敬 6,Nephew Huaiwen 僮文 5S F slave Zhenzhu 真珠 52 N Kang Ludu M 40 Wife Kan 闕 34, Daughters: Miaoying 妙英 11, 10 康鹿獨 Y Shih Nubei F 36 Daughter Gushi 谷施14, Dead son Nanipen 那你盆 9, Dead 史女輩 daughter Qiele 迦勒 13 Widow N Zhao Xiaoshi F 38 Dead husband [...] Gouren 苟 趙小是 仁 44, Mother Zhang 張 66, Widow Sister Kangnu 康女 5 1, Widowed sister Zizhen 資真 42,

ロ

Y Linghu Boxiang 令狐伯香 Widow Y Jiang Yixuan 江義宣

Y Wang Xiaoshun 王孝順 Y Su Zhuluo 素住 洛 Y Bai Xiaoshang 白 小尚 Y Yin Poji 陰婆記 Y Zheng [...]

F 70

M 22

M 11

M 60

F 19

WidowedF[...]32 F Slave Anji 安吉 35

64TAM35:

7:422- 3:505

62

3

64TAM35:

7:423- 3:506

62 64TAM35: 64

4 7

64TAM35: 67 )

2

Same

Same

7:424- 3:507- Same 8

7:431- 3:511

Same

XXIII 66TAM67: 12

7:297

3:440

71TAM184: 8:284 12/4自),12/3( a),12/2(3)

4:129

Mother Zhang 張 41, Dead Brother Baoyi 抱義 1 5, Brother Yizhen 義珍 15, Sister Shouchi 壽持 13, Aunt Zuqu 俎渠 57 blind disabled Mother Su 蘇 36, Dead father Mangxiu 盲 秀,Brother Sizhong 思忠 4 Wife Linghu 令狐 58, Son Renhui 仁惠 4 Dead mother Li Xiaoniang 李小 娘 48 widow

Missing husband Di Sijun 翟祀 君 59 (disappeared 686) M [-■] Wife Song 宋 42, Son Wuji 无忌 64TAM27: 16, Daughter Wunian 无念 6 36-39(à) M 11 Mother Gao Heimian 高黑 面 Same 30+ M 68 Wife Cao 曹 57

F 48

337

XXXI II

Same

Same

Same

Same

C 716 8:314- 4:14542 鄭 315 6 C 716 N Song Bazi 8:316 4:146 43 宋捌子 C 0. N Wang Chuli LXX 44 750 王處立 VI Key: Sogdian names in bold face type; ID=Identifïcation number of households cited in paper; Com?=Is the households census record complété?; Y=yes; N=no; Arch. ID^Archaeological identification number; v=volume; p=page; Incline; M=male; F^female; [...]=Missing information because of gap in document; 0=Missing Chinese character.

ロ

Les Sogdiens en Chine

338

Bibliography

Traditionell Sources and Document Collections TCWS-1：Guojia wenwu ju gu wenxian yanjiu shi 國家文物局古文獻研究室, Xinjiang Weiwuer zizhiqu bowuguan 新疆 維吾爾自治區博物館, and Wuhan daxue lishi xi 武 漢大學歷史系,eds. 1981-91 Tulufan chutu wenshu 吐 魯番出土文書 [Documents excavated at Turfan]. 10 vols. Beijing: Wenwu chuban she^ Tang, Zliangra 唐 長孺,Guojia wenwu ju gu wenxian yanjiu shi 國家文物局古文 TCWS-2： 獻研究 室,Xinjiang Weiwuer zizhiqu bowuguan 新疆 維吾爾自治區博物館,and Wuhan daxue lishi xi 武漢大學歷史系,ed. 1992-96 Tulufan chutu wenshu 吐 魯番出土 文書 [Documents excavated at Turfan]5 4 vols. Beijing: Wenwu chuban she. TD: Du, You 杜 佑 [801] Tong dian 通典 [Comprehensive canons of administration]. Edited by Wang 1988 Wenjin 王文錦,et al. 5 vols. Beijing: Zhonghua shuju. TLSY: Zhangsun, Wuji 長孫無忌• [653] Tang lü shuyi 唐 律疏議 [The Tang code with subcommentary]. Edited by Liu 1983 Junwen 劉俊文. Beijing: Zhonghua shuju. TTDS-2: Yamamoto, Tatsuro and Dohi, Toshikazu, eds. 1 985 Tun-huang and Turfan Documents Concerning Social and Economie History II, Census Registers (A) Introduction & Texts. Tokyo: The Toyo Bunko. YHJX: U, Jifo 李吉甫.[813] Yuanhe junxian tuzhi 元/ロ君 縣圖志 [Geography of the commanderies and 1983 districts of the Yuanhe reign]. 2 vols. Beijing: Zhonghua shuju.

ロ

Secondary Sources Beckwith, Christopher

The Tibetan Empire in Central Asia. Princeton: Princeton University Press. 1987 Chavannes, Édouard. 1900 Documents sur les Tou-kiue (Turcs) Occidentaux, recueillis ef commentés suivi de notes additionnelles. Taibei: Ch'eng Wen Publishing, 1969. Dong, Guodong 凍國 棟

Tangdai renkou wenti yanjiu 唐代人口問題研究 [Problems in Tang dynasty demography]. Wuhan: Wuhan daxue chubanshe. Ebrey, Patricia Buckley Sung The Inner Quarters: Marriage and the Lives of Chinese Women in 1993 Period. Berkeley: University of California Press. FRYE, Richard N. The Heritage of Central Asia: From Antiquity to the Turkish Expansion. 1996 Princeton: Markus Wiener. 1993

:

Jonathan K. Skaff

Docimentin^Sogdian society at Turfan

339

HANSEN, Valerie

1998 Introduction: Turfan as a Silk Road Community. Asia Major 3rd Séries, 9: 1-6. ÏIŒDA, On 池田温 Ba shiji zhongye Dunhuang de Suteren juluo 世 紀中 葉敦煌的粟特人聚落 1993 [A Sogdian colony at Dunhuang in the mid-eighth Century]. In Riben xuezhe yanjiu Zhongguo shi lunzhu xuanyi, jiu juan, minzu jiaotong 日本學 者 研究 中國 史論著 選譯 第九卷 民族交通 [Selected works of Japanese scholarship on Chinese history, vol. 9, relations with nationalities], edited by Liu Junwen 崟!)俊文. Beijing: Zhonghua shuju. 1988 Shinryö sannen Köshö-ken Süka-kyö tensekiyö ni tsuite 神龍三年高昌県崇 化郷点籍様に ついて[The 707 inspection record of household registers from Gaochang County5 Chonghua Township]. In Chügoku kodai no hô to shakai: Kurihara Masuo Sensei koki kinen ronshü 中国古代の法と社会 栗原益男先 生古稀記念 [Studies of law and society in ancient China in honour of Prof. Masuo Kurihara 〇 the occasion of his seventieth birthday], edited by Hori Toshikazu 堀 敏ーヽ Tokyo: Kyüko shoin. 1981 Les marchands sogdiens dans les documents de Dunhuang et de Turfan. Journal Asiatique 269: 77-79. 1973 T'ang Household Registers and Related Documents. In Perspectives on the T'ang, eds. A. F. Wright and D. Twitchett. New Haven: Yale University Press. 1965 Hasseiki chüyô ni okeru Tonkô no Sogudo jin shüraku 8世紀中葉に おける敦 煌のソ グ ド 人聚落 [A Sogdian colony at Dunhuang in the mid-eighth Century]. Yürashia bunka kenkyü ユ ー ラシ ア文化 研究 (Studies of Eurasian culture) 1 : 49-92. JIANG, Boqin 姜 伯 勤 Dunhuang Tulufan wenshu yu sichou zhi lu 敦煌吐魯番 文書與絲綢 之路 1994 [Dunhuang and Turfan documents and the Silk Road]. Beÿing: Wenwu chuban she. Johnson, Wallace 1997 The T'ang Code: Volume II, Spécifie Articles. Princeton: Princeton University Press. Juliano, Annette L., Lerner, Judith A. 2001 Cosmopolitanism and the Tang. In Monks and Merchants: Silk Road Treasures from Northwest China, eds. A. L. Juliano and J. A. Lemer. New York: Harry N. Abrams. de LA VAISSIÈRE, Étienne 2002 Histoire des marchands sogdiens. Paris; Collège de France, Institut des hautes études chinoises. Lerner, Judith A. 2001 The Merchant Empire of the Sogdians. In Monks and Merchants: Silk Road Treasures from Northwest China, eds. A. L. Juliano and J. A. Lemer. New York: Harry N. Abrams. Li, Fang 李方,and WANG, Su 王素 1996 Tulufan chutu wenshu renming diming suoyin 吐 魯番出土文書人名地名索 弓 [ [Index of personal and place names in the documents excavated at Turfan]. Beijing: Wenwu chubanshe.

ハ

5

5

ヽ

Les Sogdiens en Chine

340

Luo5 Feng 羅豐 2001

1996

Sogdians in Northwest China. In Monks and Merchants: Silk Road Treasures from Northwest China, eds. A. L. Juliano and J. A. Lerner. New York: Harry N. Abrams. Guyuan nanjiao Sui-Tなng mudi 固 原南郊隋唐 墓地 [A Sui-Tang cemetery in the Southern suburbs of Guyuan]. Beijing: Wenwu chuban she.

Pohl, Walter, ed.

Strategies of Distinction: The Construction of Ethnie Communities, 300-800. Transformation of the Roman World, Vol. 2. Leiden: Brill. G. Edwiîl PULLEYBLANK, Lexicon of Reconstructed Pronunciation in Early Middle Chinese, Late Middle 1991 Chinese and Early Mandarin. Vancouver: UBC Press.

1998

Rong, Xinjiang 榮新江

2001

Zhonggu Zhongguo yu wailai wenming 中古 中 國 與 外 來 文明 [Médiéval China and foreign culture]. Beijing: Shenghuo, Dushu, Xinzhi Sanlian.

Sheng, Angela

Innovations in Textile Techniques on China!s Northwest Frontier, 500^700 AD. Asia Major 11: 117-60. Nicholas Sims- Williams, The Sogdian Merchants in China and India. In Cina e Iran da Alessandro 1996 Magno alla Dinastia Tang, eds. A. Cadonna and L. Lanciotti. Florence: Leo S. Olschki Editore. Skaff, Jonathan Karam Survival in the Frontier Zone: Comparative Perspectives on Identity and 2004 Political Allegiance in China's Inner Asian Borderlands during the Sui-Tang Dynastie Transition (617-630). Journal of World History 15 (2), pp. 117-154. 2003 The Sogdian Trade Diaspora in East Turkestan during the Seventh and Eighth Centuries. Journal of the Economie and Social History of the Orient 46 (4):475-524. The Sasanian and Arab-Sasanian Silver Coins from Turfan: Their Relationship 1998 to International Trade and the Local Economy. Asia Major 11: 67-116.

1998

Tang, Changru 唐 長孺

1982

Xinchu Tulufan wenshu fajue zhengli jingguo ji wenshu jianjie 新出吐 魯番文 書發掘 整理經過及文書簡介 [The expérience of excavating and arranging the documents newly unearthed at Turfan and a brief introduction to them]. Tôho gakuhö 54: 83-100.

Trombert, Éric

2002 2002a

Les cycles de culture et l'organisation des terroirs à Tourfan aux vie-Viile siècles. Bulletin de l'Eco le français e d!Extreme-Orient 89: 203-35. La vigne et le vin en Chine. Misères et succès d'une tradition allogène. II Vins, vignes et vignerons de Fourfan. Journal Asiatique 290: 485-563.

Twitchett, Denis

1970

Financial Administration under the T'ang Dynasty. 2nd ed. London: Cambridge University Press.

Jonathan K. Skaff

Documenting. Sogdian society at Turfan

341

WU, Zhen 吳震

1999

Asitana—Halahezhuo muqun kaogu ziliao zhong suo jian de huren 阿斯塔 哈拉 和卓古墓 群考古資料中所見的胡人 [Western barbarians as seen 那 in archaeological materials from the Astana—Karakhoja graveyard]. Dunhuang Tulufan yanjiu 敦煌吐 魯 番 研究 (Dunhuang and Turfan Studies) 4: ヽ

245-64. Victor Cunrui XlONG, The Land-Tenure System of Tang China—-A Study of the Equal-Field System 1999 and the Turfan Documents = T'oung Pao 85: 328-90. Ilya Yakubovich, Marriage Sogdian Style. Paper presented at the Conference Iranistik in Europa 2002 at Graz, Austria. Zhang Guangd% and Rong Xinjiang A Concise History of the Turfan Oasis and Its Exploration. Asia Major 3rd 1998 Séries, 9: 13-36.

Non-Han ethnie groups and their settlements in Dunhuang during the Late Tang and Five Dynasties Zheng Binglin

The publication of Ikeda On5s well-known article “Sogdian Settlements in Dunhuang during the mid-Eighth Century” has had an enormous influence on the study of the Sogdians in Dunhuang during the late Tang and Five Dynasties.1 Ikeda concluded that after the Tibetan occupation of Dunhuang, some Sogdians retumed to their homeland, some scattered throughout Uighur-controlled régions and other areas s while still others remained in Dunhuang and their households came under the ownership of monasteries. Due to the influence of Ikeda’s work, the questions of whether Sogdians lived in Dunhuang during the late Tang and Five Dynasties and whether they lived in their own settlements hâve gained currency, but hâve as yet not been proven by scholars. My research on the relationship between the Guiyi Military Regime (guiyÿun zhengquan [848-1036]* and non-Han peoples residing in Dunhuang, as well as on the relationship between Sogdians and Buddhism in Dunhuang, has shown that large numbers of Sogdians were indeed living in Dunhuang at that time. It is also true that Sogdians held varions posts in the Guiyi Military Regime which ranged from the office of vice military commissioner (jiedu fushi on down. They also served at every level of the hierarchy of the Buddhist samgha. I also believe that Sogdians founded associations (she j±) together with Buddhist monks in Dunhuang during this period.2 While my conclusions were confirmed by the recent research of various scholars in the field, they were also criticized by other scholars who thought that I had been too subjective. These scholars argued that it is difficult to confïrm whether the sources upon which I relied were indeed those of Sogdians.3 However, research has generally convinced scholars in the field that large numbers of Sogdians did indeed live in Dunhuang during the late Tang and Five Dynasties, yet many questions still remain unresolved. Did Sogdian settlements exist in Dunhuang during the Guiyi Military Regime? What was the nature and rôle of Zoroastrianism during that time? How did the Guiyi Military Regime govem the Sogdian population, etc.? The key question is still whether or not there were Sogdians living in Dunhuang during this period. Ikeda, 1965. Passages between brackets [...] were added by the editors. 2 Zheng, 1997. See also Zheng, 2003a and 1999. 3 Rong, 2001a and Rong, 2001b, pp. 239-240. Also Lu, 2003 and Feng, 2003.

.

——

Les Sogdiens en Chine

344

Scliolars understand the process by which the Tibetans came to occupy the Hexi 河 The Tibetans first took the area around the Tianshan 天山 Mountains and the région. 西 région govemed by the Longyou military commissioner (Longyou jiedushi 右 度 使). Then5 moving westward from Liangzhou 冰 州, they took the Hexi région, and who had finally besieged the Hexi military commissioner (Hexi jiedushi taken refuge in Shazhou 沙州 [Dunhuang] in 766 and they conquered this oasis [between 785-787]. Therefbre, the Sogdians had no reason and no means to flee to Sogdia or Uighur-controlled régions. In this article I argue that there was indeed a large number of Sogdian immigrants living in Dunhuang during the late Tang and Five Dynasties, and that Sogdian Settlements did indeed exist, Non-Han ethnie groups other Shanshan 善B善,Yanqi (Karashar) 焉耆,and Yutian (Khotan) 于 一 than Sogdians also lived in Dunhuang Given that a good deal of research has also been done on the Khotanese prince and his settlement3 this article will mainly concentrate on Sogdians and their Settlements.

—

s

Sogdians in Dunhuang during "e Guiyi Military Period (848-1036) During the Tibetan rule of Dunhuang [c. 785/787-848], many Sogdians lived in Liangzhou and Dunhuang.4 Although many Sogdians and other non-Han still lived in Dunhuang during the Guiyi Military period, due to the relatively high degree of sinisation among the non-Han during that time, it is difficult to differentiate between the ethnie Sogdian and the Han. Therefbre I have often been either unwilling or unable to say with any certainty, whether there were Sogdian settlements in the Dunhuang région, or even whether there were non-Han people in the Kang 康,Shi 史,and An 安 clans. In trying to define who was ethnically Sogdian and who was Han, the crucial question is the nature of their given names. In order to verify that Sogdian peoples were living in Dunhuang during the late Tang and Five Dynasties, I must examine the Dunhuang documents fbr persons having non-Han sumames who may not have completely adopted Han given names. Of course such persons are relatively few, but where they exist they prove that Sogdians did indeed live in Dunhuang during the period under discussion. Scholars engaged in the study of ethnie groups in Dunhuang believe that those sumamed Kang and Shi were Sogdians. This position can be verified by looking at the Dunhuang documents. Sogdians living in Dunhuang during the late Tang and Five Dynasties periods, cohabitated mainly with other non-Han. For example, in a list of Ioans in the Mogao sub-district 莫高 of Dunhuang there is a guarantor for a Ioan whose name is “Kang Xinding 康信定,föreign (fan 番), from the Mo[gao sub-district], living behind Shi Bolüe's 史 略house.555 The word "foreign” behind the name of Kang Xinding indicates the man's non-Han identity. The name Shi Bolüe is also who lived in Liangzhou, was almost certainly a For example, this man named Kang [?] zhilie 康 Sogdian; see S. 2214, a land tax mémorandum dated back to the Tibetan period; ed. in Tang Gung'ou and Lu Hongji, 1986-1990, [hereafter CKCL], vol. 2, p. 421. 5 Dx. 11080, ed. in CKCL, vol. 2,p.272. 4

Zheng Binglin

Non-Han ethnie groups in Dunhuang

345

obviously a transliterated name and so this man too was undoubtedly a Sogdian, In many other monastery accounts, we find occurrences of the names Kang Azhuzi 康阿 竹子, Kang Bolüe 康 略；and Kang Aduo 康阿 which confirm that Kang was a major sumame used by the Sogdians. During the Guiyi Military Regime, the Kang clan produced several eminent men, such as Kang Tongxin 康通 信, who served as the du Shandan zhen eshichong Liangzhou xijie youyifangcai duzhi bingmashi 都 丹 遏使充 凉 州西界游 防采都 知兵 使official, Kang Shijun 康 使君,who served as Governor of Guazhou {Guazhou cishi 瓜州朿 U 史), a préfecture bordering on Dunhuang, and Kang Xianzhao 康賢照 who assumed the ofßce of dusengtong 者ロ僧统 (General administrator of the samgha).6 8 The Shi clan also produced eminent men, such as the physician Shi Zaiying 史再盈フ who attained the secret mystical effïcacy of Qi Po 耆婆 and the sublime wondrous arts of Yu Fu 附.The Shi clan was also a powerfol family in Dunhuang. There has been much conjecture among scholars about whether the Cao 曹 clan was a non-Han, but no one has dared to give a conclusive answer. Two scholars5 Rong Xinjiang and Feng Peihong, argued that Cao Yijin 曹 金, the military commissioner (jiedushï) of the Guiyi Military Regime, was non-Han.9 That the sumame Cao was used by non-Hans in the late Tang and Five Dynasties in Dunhuang, can be seen from many documents. A list of commoners who received hempseeds (Dx. 1451, dated 973-979) mentions the name of Cao Chihu 曹赤胡；and another document (P. 2912) those of Cao Wuduo 曹勿 多, and Shi Huihu 石回鹘 [Shi the Uighur].10 Document P. 3745 records the name of Cao Huzi 曹胡子, who either received, or paid a certain amount of flour, bundles of tire wood, cakes and noodles, as well as bowls, plates, dishes, lacquer bowls, and one boit of cloth.11 Ail these given names clearly indicate that these men were non-Han and therefbre I can conclude that the sumame Cao was indeed a Sogdian name during this time. I also contend that intermarriage between non-Han minorities and the Han majori ty )朝 during the Guiyi Military Regime period was facilitated by Zhang Yichao's father, Zhang Qianyi 逸, who took a Sogdian woman sumamed An as his wife. The vice military commissioner (jiedu fushî) of the Guiyi Military Regime, An Jingmin, hailed from a Dunhuang Sogdian clan sumamed An. In the économie manuscripts from Dunhuang one finds the names of many Sogdians of the An clan: An Zulüe 安足略 and Shi Taxibin 石他悉 (P.5038); a cobbler named An Adan 安阿丹 (P. 4640) a

,

6 P. 2049 V° 1, Annual report of the Jingtu 土monastery, dated 925; ed. in CKCL, voL 3, pp. 347-366, see pp. 349, 354. 7 P. 3424 V°, Milling account (924 or 984); ed. in CKCL, vol. 3, p. 129. 8 About Kang Tongxin and Kang Shijun, see their Eulogies in P. 4660 R°, ed. in Zheng, 1992, pp. 114, 151. 9 See Rong, 200 1 a. 10 Dx. 1451 and P2912 are cited in Zheng, 1998, pp. 191-208. Also cited in Zheng, 2003a. 11 P. 3745 V°, Account of disbursements for meals, ed. in CKCL, vol. 4, pp. 19-22. 12 P. 5038, List of commoners owing hay (946?), ed. in CKCL, vol. 2, pp. 442-443. P. 4640 V°, Accounts of disbursements of paper and cloth paid by the Guiyi military govemment Court (899-901), ed. in CKCL, vol. 3, pp. 253-270, see p. 256.

346

-

Les Sogdiens en Chine

certain An Wu 安勿 who paid fifty-three piculs five bushels of wlieat (S. 6064);13 a certain An Taxilu 安他悉禄 who submitted one picul of wheat;14 another man called An Yanji 安衍鸡 and, fïnally, An Chouhu 安丑胡 who owed fïre wood (Dx. 2149). A list of monks and officiais who offered bolts of silk to a "Son of heaven" named He (He tianzi 天子) lists the name of several officiais: the dutou 度 An Kedun 安珂敦; a zhenshi 使An; a duzhi 都知 An, and a duya 者防吾亍 An,15 Document S. 4884 lists two other names: An Hu's 安胡 wife? and Ne'erximan 丿 L 悉曼* We can infer firom these sources that the social status of the Sogdians of the An clan was quîte high during this time. By the same token we can also note that the An clan was of Sogdian ethnicity. As for the Shi clan9 in addition to Shi Faxibin cited above5 document P. 2856 mentions a certain Shi Yanqianlübo 石炎欠律钵 of Liantai 台monastery. His name is also mentioned in document P。3638. He was undoubtedly also a Sogdian?7 The Luo and Mi 米 clans living in Dunhuang during this time were also non-Han. Document P. 3418 States that that Luo Taxibin 他悉 and Mi Naxiji 米 悉鸡who were commoners from the Dunhuang sub-district 敦煌 owed wood?8 Document P. 4906 mentions a certain Luo Xiji 悉ヌ鸟* Document P. 3234 gives the names of commoners who borrowed hempseeds; among them I note the zhihui 扌旨 official Kang who "lives in Xishuichi 西水)也, [near the house of] the dutou Kang," Mi Lijiu 米 20 These sources show that there were many 里久, and "the son of Mihu 米胡」 Sogdians sumamed Luo and Mi living in Dunhuang during the late Tang and Five Dynasties. A number of scholars believe that those sumamed Shao were also non-Han living in Dunhuang. Document P. 2049 V° 1 mentions a certain Shao Azhuduo 阿竹咄」' In a list of commoners who owed bundles of wood I notice the following names: Zhu Luyanxin 竹 研心 (ftomthe sub-district of Xiaogu 效谷 ), as well as Zhu Luhunu 竹 胡奴,Shao Lüedan 略丹,and Shi Lushan 石禄山 (ail from Cihui sub-district 慈 惠 )」2 The sumame Zhu is perhaps Indian; the surname Shi is that of Sogdian immigrants. I am at présent unable to détermine with any certainty whether the sumame Shao is Fibetan or that of another non-Han ethnicity, but the there can be no doubt that the name is definitely that of an ethnie minority, most probably a non-Han from central Asia. Was the Di (or Zhai) 翟 clan Sogdian or Han? According to document P. 464023 and other sources, the Di clan of Dunhuang seems to hâve been Han. The Labor and Tax r

r

13

S. 6064, Annual account of Bao'en 扌假恩 monastery (draft copy, early ninth Century?), ed. in CKCL, vol. 3, see p. 298. 14 P. 2049 V° 1 (see n 6), here see p. 349. 15 P. 3440, dated 936 or 996; ed in CKCL, vol. 4, pp. 16-17. 16 S. 4884 V° List of lenghts of wollen cloth borrowed (972),ed. in CKCL, vol. 2, p. 235. 17 P. 2856, List ofhay in a monastery account dated 893, ed. in CKCL, vol. 3, p. 114. P. 3638, Inventory of belongings ofthe Jingtu Monastery (911),ed. in CKCL, vol. 3,pp. 116-118. 18 P. 3418 V°, List ofhouseholds owing wood, ed. in CKCL, vol. 2, pp. 427-436, seep. 427. 19 P. 4906, fragment of a monastery account, ed. in CKCL, vol. 3, pp. 233-235. 20P. 3234 V°, List of Ioans granted by Jingtu monastery, dated 944, ed. in CKCL, vol. 2, pp. 212-215. 21 See n. 6, here see p. 353.. 22 P. 3418 V°, ed. in CKCL, vol. 2, p. 427-436. 23 Seen. 12.

Zheng Binglin

Non-Han ethnie groups in Dunhuang

347

Register of Conghua sub-district (Conghua xiang chake bu AM 匕 差矛斗 簿),dating from the mid-eighth Century [P. 3559 + 3664],reveals that there were Sogdians sumamed Di living in the Dunhuang région. There were also non-Han in the Di clan during the Guiyi Military period. Document P. 2629 States, "On the ペ day of the 8も month, Di Subudan 翟 理不丹 was given a jug of wine.5524 We can see from his name that he was not Han. Document S. 6452 records: On the 14th day of the 2nd moiitli, Di Hu 翟胡, who lived north of the canal, borrowed five jugs of wine to pay for his medicine..^ On the 9th day of the 3rd month, A Chaiyanhu 阿柴 胡 borrowed one jug of wine for medicine?5

Document S. 6981 records that a man named Di Buwu 翟不勿 is included in the list of people who were provided with wine. It is clear from his name that the récipient was also not a Han.24 26 Since this source clearly shows that the man sumamed Di, who traded medicine north of the canal was not a Han, I am entitled to assume that there were definitely non-Han sumamed Di living in Dunhuang during the late Tang and Five Dynasties. The Long clan was also a non-Han clan that had emigrated from Yanqi to Dunhuang. Long was the sumame of the King of Yanqi. After the fall of the Yanqi kingdom, the Yanqi people migrated to Xizhou 西州 (Turfan) and Yizhou 伊州 (Hami). Those Yanqi who lived in Dunhuang and other places in Hexi took Long as their sumame and called themselves the Long clan. This is well documented in the Dunhuang sources. The Long clan prospered as merchants and craftsmen. Those who immigrated to Dunhuang settled around Changle 常生 where they became famous for tending horses. Men sumamed Long occupied varions administrative levels in the Guiyi Military Regime and in the Buddhist samgha. However, the Long clan was highly sinisised, and judging by their given names, they were almost completely assimilated in the Han majority and it is therefdre impossible to detect vestiges of their ethnicity, Very few members of the Long clan preserved their ethnie character, although a certain Long Moguxiji 磨骨悉鸡 of the Tongxia 通 [tribe] is recorded as having borrowed two piculs of milleピ切 jn the late Tang and Five Dynasties period, most of the Long clan lived in Guazhou and were mainly involved in tending horses, and thereföre Guazhou became a center fbr the breeding of horses and livestock.28 As noted above, I attempted to list the Han sumames that were used to transliterate non-Han names in order to show that there were indeed large numbers of Sogdians living in tribes in Dunhuang during the late Tang and Five Dynasties. It has been argued by other scholars that after the Tibetans conquest of Dunhuang, a good percentage of Sogdians migrated to Sogdiana or to Uighur-controlled areas, and a small number became monastery households (sihu 寺戶). The argument that there were no Sogdians living in Dunhuang from that time onward can be challenged. For my part, I 24

List of disbursements of wine from the Guiyi military govemment Court, ed. in CKCL, vol. 3, p. 274. List of Ioans of oil and flour (981),ed. in CKCL, vol. 2, pp. 239-241. 26 Document ed. in Rong, 1994, p. 58. 27 P. 2680, List of Ioans of millet, ed. in CKLC, vol. 2, p. 234. 28 Zheng, 2003b, pp. 205-238. 25

348

_

Les Sogdiens en Chine

maintain that a large number of Sogdians did live in Dunhuang during the late Tang and Five Dynasties, and that Sogdian settlements certainly emerged and endured. Therefbre a sizeable non-Han population, mainly composed of Sogdians, had a definite influence on Dunhuang society and culture. For example, non-Han women participated in every aspect of life in Dunhuang society. The Dunhuang documents preserve many historical records of their activities.

Zoroastimism in Dunhuang during the Late Tang and Five Dynasties The Sogdian ethnicity's main characteristic was their belief in Zoroastrianism, which was populär in all areas in which Sogdians lived. Therefore in order to détermine whether Sogdians lived in Dunhuang during the late Tang and Five Dynasties, we must détermine whether Zoroastrianism was prévalent in that région during that time. Document P. 2005 States that in Dunhuang during the Kaiyuan era (713-741): Zoroastrian gods were located one li east of the préfecture, where a temple was built and portraits of deities were painted. In fact, some twenty shrines were built. Their courtyards measured about a hundred paces.29

The Zoroastrian temple became the social center fbr the Sogdians of the Conghua sub-district in Dunhuang. Moreover, there were also Zoroastrian temples in neighboring areas such as the Shicheng 石城 Garrison and Yizhou, which also became major centers of Sogdian life. Zoroastrian temples existed in Dunhuang during the late Tang and Five Dynasties and Zoroastrian sacrifices were very animated. At times even the local authorities of the Guiyi Military Regime took part in them. This demonstrates that the Dunhuang Sogdians were very active and influential during that time. The Dunhuang guji nianyong 敦煌古迹廿咏 (Twe 加y Odes to Historical Sites 比 Dunhuang) was the most widely circulated literary work in Dunhuang during the late Tang and Five Dynasties. Among them are odes about Zoroastrianism in “The city of An” (Ancheng 安城). At the time when the walls of Ancheng were built, a shrine to the deity was erected together with the city. People from ail the préfectures and districts prayed for prosperity and auspicious omens, since everything in the world relied on the good and bad fortune which the Zoroastrian deities could bestow. Many worshipers wearing fine clothes came to the temple since it seemed that ail its spirits had the [power] on which one could rely. Look at the place of sacrifice! Sacrificial wine flows continuously from moming to night.30

In ail there are six copies of the Dunhuang guji nianyong preserved in the Dunhuang collections. This work was composed during the early years of the Tibetan occupation, but most of these six copies were transcribed during the Tibetan occupation and the Guiyi Military Regime. This suggests that Zoroastrianism also existed in 29

30

"Shazhou dudufii tujing” 沙州都督 府 (Description ofthe General Govemorship of Shazhou). For entire text see Zheng, 1989, pp. 138-140.

II Zheng Binglin

Non-Han ethnie groups in Dunhuang

349

Ancheng during these periods. Mention of Zoroastrianism in Ancheng is preserved in

an exorcistic text: The sacred rites of exorcism have come down to us from antiquity. Tonight [we] purify and cleanse ourselves, in order to get rid of all bad omens and evils. The whole world celebrates the New Year. May the Governor (changshi 使(史)) live fbr ten thousand of years! May the population be prospérons and have an ample amounts of money. May we be govemed by the Great Tang, and never be fond of barbarian customs. May the Governor of Dunhuang reunite fathers and sons. May we be promoted to higher office and rank this year. May we find beautiftd wives to marry in the summer, May we receive precious edicts after submitting memorials. May our lands receive the title of Nanyang and be enfeoffed this year. This is the incantation and will of the gods3 not the words of humble folk. Tonight the band of exorcists is led by the Great Ormazd of Ancheng (Ancheng daxian 安城大 ) and followed by the Saints of Sanwei. We seek out and collect all the daggers and lances inside and outside our houses in order to repel the evil spirits of the past year, and welcome the Seven Worthies of Penglai. We also welcome the Four Luminous Brothers of Nanshan. Red lotuses of five shades will appear this fall. From now on there will be no war or disasters in Dunhuang. Long live the City God. 31

Up to the Five Dynasties period, the Dunhuang local authorities included the General of Ancheng (Ancheng jiangjun 安城将 ) in the panthéon of deities to which they sacrificed. Document P. 2555 States: On the 13th day of this month the changshi dafu 常侍大夫 official offered incense to 360 gods. Which gods were they? The General of Ancheng, the Fighting General (douzhan jiangjun 斗 将 ), the Marching General (xinglu jiangjun 行路将 ), etc.

The names "Great Ormazd of Ancheng55 and the "General of Ancheng^^ refer to the same Zoroastrian deity as the one worshiped in the Zoroastrian shrine in Ancheng. The fact that exorcistic rites mingled with Zoroastrian sacrifices suggests that a sizeable number of Sogdians lived among the Dunhuang population. Zoroastrian sacrifices became official activities during the Zhang Huaishen period [867-890]. Document P. 3569 States, "On the 14th day of the 4th month, summer, four )•"む Ma Sanniang was an jugs of wine for the Zoroastrian sacrifice (sai xian official wine maker. The raw materials that she used to make her wine were provided by the local govemment, and the wine that she produced was paid for by the govemment. Since the govemment provided the wine that was used in Zoroastrian sacrifices, it can be inferred that the govemment conducted the sacrifices. When Zhang Chengfeng was in power [894-910], Zoroastrian sacrifices became entirely a govemment affairs. Document P. 4640 records that the paper products used in many Zoroastrian sacrifices were paid for by the govemment out of the military co抵rs. This source records the sacrifices, but does not mention to which deities they were actually

31

P. 3552 "Erlang wei"丿L 良 B ,fbr en tire text see Huang, 1995, p. 594. P. 3569, Report of disbursements of wine5s grain by Ma Sanniang (887),ed. in CKCL, vol. 3,pp. 622-624. 32

三娘 and Long Fendui

粉堆

350

Les Sogdiens en Chine

sacrificed. For example, the text records that on the 14tb day of the 4th month of the jiwei year [899]5 there was a sacrifice east of the city... On the 19th day there was a sacrifice in Lujiaquan On the 23rd day of the 5th month there was a sacrifice in Baichixia On the 20th day of the 6* month there was a sacrifice in Pinghekou . On the 25th day of the 7th month a ritual space was established in Baichixia and: On the same day thirty sheets of paper were given for paintings of Zoroastrian deities (xian . (On the 5th day of the 10th month) another thirty sheets of paper were given for shen the same purpose,. . and on the 13" day (of the first month of the gengshen yeai [900]) thirty sheets of paper were again given for the paintings of Zoroastrian deities.

On the 3rd day of the 3rd month there were sacrifices in Sanshuichi Baichixia, and Fenliuquan “on the 8th day (of the 4th month) thirty sheets of paper for paintings of Zoroastrian deities were given?5 The same amounts of sheets of paper were given for the same purpose on the following dates: on the 16th day (month unknown)5 on the 9th day of the 10th month, and on the 1 lth day of the first month of the xinyou year [901]. On the 21st day of the 2nd month there was a sacrifice in Lujiaquan: On the same day thirty sheets of paper for paintings of Zoroastrian deities were given. . . On the 3rd day (of the 3rd month) one tie of coarse paper was used in the sacrifice held at Dongshuichi and at various other places. . . On the 1 3th day (of the 4th month) thirty sheets of paper for paintings were used in Zoroastrian sacrifices...33

Sanshuichi [the “three ponds”] refers to Dongshuichi [East Pond], Xishuichi [West Pond], and Beishuichi [North Pond]. Since it is known that the sacrifice at Dongshuichi was to Zoroastrian deities, then it fbllows that the sacrifices at the other places were also to Zoroastrian gods. Most of the Zoroastrian sacrifices used thirty sheets of paper, therefbre wc can infer that the places of sacrifice in Lujiaquan (also written Baichixia, and Fenliuquan, which used thirty sheets of paper, were ail Zoroastrian shrines. The fact that Zoroastrianism was flourishing in the late Tang and Five Dynasties and that its sacrifices were such lively events, allows us to assume that the Sogdian population in Dunhuang was fairly large. Furthermore, Zoroastrian sacrifices ail took place in vicinity of Dunhuang City, so it fbllows that Sogdians lived in these locations during that time. Zoroastrianism flourished in Dunhuang throughout the period of Cao’s control of the Guiyi Army [914-1036]. Document S. 1366 States: On the 17th day, permits were issued for fifty-seven fen of food, one sheng of lamp oil, and nine sheng of noodles with sausage for the Zoroastrian sacrifice east of the old City.34

Not only did local govemment officiais make frequent sacrifice in Zoroastrian 33

P. 4640 V° (seenl2). S. 1366, List of disbursements from the Guiyi military govemment Court (980-982), ed. in CKCL, vol. 3, pp. 281-286. 34

BI Bill Zheng Binglin

Non-Han ethnie groups in Dunhuang

351

shrines in Dunhuang, but traveling envoys of the Guiyi Military Regime also held sacrifices at Zoroastrian shrines in other areas of the Hexi région. This suggests that not only were there many Sogdians living in Dunhuang during the Cao's Guiyi Military Regime, but that there were also many Sogdians living in other parts of Hexi. A list of wine disbursements from the Guiyi Military Government Court (964 ?) records: On the 20th day (of the 4th month) for the Zoroastrian deities (xian shen) east of the city3 one jug of wine... On the 10th day (of the 7ぬ month) for a Zoroastrian sacrifice to the deities, two jugs of wine."む

Another list of disbursements from the same office records "[For] the Zoroastrian deities, east of the city3 two sheng of oil?336 From tliese records of disbursements of the Guiyi Military Regime, it is possible to see that the govemment took complété responsibility for providing all the ritual implements used in Zoroastrian sacrifices, from the paper used for liturgical paintings to the wine, lamp oil, sacrificial food, and noodles and sausage. The Dunhuang documents that record Zoroastrian temples and sacrifices, allow us to see that belief in Zoroastrianism flourished in Dunhuang under Guiyi Military Regime rule. Given that the Zoroastrian faith is mainly followed by Sogdians, the evidence shows that a large number of ethnie Sogdians lived in Dunhuang during the late Tang and Five Dynasties, and that they established many settlements in which Sogdians were the majority.

Non-Han Settlements in Dunhuang during the Late Tang and Five Dynasties During the Tianbao era (742-756), Sogdian settlements in Dunhuang were centered around the Conghua sub-district. The Labor and Tax Register of this sub-district [P. 3559 十 3664] records the names of many Sogdians who served corvées; among them: Kang Linqin 康令 , aged forty, zhuguozi 柱国ナ,community leader (lizheng 里 正)；Fengluan 奉 , son of Luo Ningning ,aged thirty-one, adult male (baiding 白丁), community leader; Tuhun 突 昏, son of An Tana 安他那, aged twenty-one, a young male (zhongnan 中男), village head {cunzheng 村正)； Baojin 抱金, son of He Nini 何 尼尼, aged eighteen, a young male, village head; Shuangli 双禾ル the younger brother of Luo Shuangliu 双流,aged twenty, a young male> village head; Luo Tejin 特勤, aged thirty-five3 adult male, village head. The second half of this document is fragmentary and incomplète, so it is impossible to know exactly how many communities (〃) and villages (c友匐) there were in the Conghua sub-district during the Tianbao era, but we do know from the above sources that this sub-district comprised at least two communities and four villages. In the following pages, on the basis of what 35

P. 2629, Dong Xiwen's 董希文 old collection's manuscript, and the Institute of Dunhuang Studies's manuscript n° 1, ed. in CKCL, vol. 3, pp. 271-276. 36 S. 2474, List of disbursements of oil and flour from the Guiyi military govemment Court (980-982), ed. in CKCL, vol. 3, pp. 278-280.

352

_

Les Sogdiens en Chine

we know about the organisation of rural areas during the late Tang and Five Dynasties in Dunhuang, I hope to shed some light one the question of Sogdian settlements and their names, through a systematic study of known villages and settlements.37

Cao Village (Cao jia zhuang Cao was a populär sumame in the late Tang and Five Dynasties in Dunhuang. At the time of Zhang's control of the Guiyi Army there lived a certain Sir Cao (Cao go〃g who held the title goudang sanku sengzheng and a certain Cao Fajing dusengzheng After the Cao clan took control of Dunhuang [in 914], the Cao name became well known in the oasis. In addition to Cao Yijin, who was the head of the Guiyi Military Regime, there was also a group of Cao clansmen, including Cao Liangcai who controlled important and Cao Yingda organisations within the régime. A Cao Village did exist during the late Tang and Five Dynasties, but it is recorded only in one document (P. 2032 V°, dated mid-tenth Century), which States, “Two bushels of flour, paid for the food rations of construction workers in Cao Village (Cao jia zhuang).” There are no other records among the Dunhuang documents about the geographical location of Cao Village. P. 3560 records that the Cao clan (Cao jia) owned a canal that belonged to the Qianqu waterway System. The Qianqu waterway System was probably located in the région east of Dunhuang City. It follows that Cao Village was also likely to be located east of Dunhuang City. A list of disbursements States that on the 17* day of the lst month of the xinmao year (931?), “the Cao clan delivered one and a half bushel of millet.” On the 2nd day of the 2nd month, “to brothers of the Cao clan, two bushels of millet, during the Hanshi Festival.”38 A list of vegetable gardens (guayuan MEI) mentions “the vegetable garden of the Cao clan” (P. 3396). Document S. 6981 mentions a certain “Alouyu fnfg of the Cao clan.” In order to détermine if the above records do indeed refer to a Cao Village in Dunhuang, I must do some further research. The Labor and Tax Register of the Conghua sub-district cited above, States that of the Sogdians who served corvées in this sub-district, there were four who served as village heads (cunzheng): An Tuhun, He Baojin, Luo Shuangli, and Luo Tejin. This suggests that there were at least four Sogdian villages at that time, and that Cao Village was one of these Sogdian settlements. Furthermore, not only was Cao Village one of the Sogdian villages during the late Tang and Five Dynasties, its status also rose during that time. An Village (An jia zhuang An Village was one of the villages established by Sogdians during the late Tang and Five Dynasties period. The An clan was very powerful and influential in Dunhuang during the Guiyi Military Regime. An Jingwen gave strong support to Zhang Yichao when the Regime was first established. During the Regime, An clansmen held varions offices from vice military commissioner (jiedu fiishï) on down, such as the The An office of zuodu yaya which was occupied by An Huai’en clan was one of the most powerful clans in the Guiyi Military Regime from the time the 37 38

See my article, Zheng, 2003b, pp. 122-162. P. 4907, ed. in CKCL, vol. 3, p. 205.

Zheng Binglin

Non-Han ethnie groups in Dunhuang

353

Regime was headed by Zhang to the time it was lead by Cao. It is therefore likely that An Village was one of the original villages within the Conghua sub-district during the Tang dynasty. One monastic account (P. 6002, dating from the end of ninth Century) States a disbursement of wheat, oil and flour in för the benifit of “the head of An Village” (An jia zhuang tou /£2“" Another account of the same period (S. 5049) records an amount of wheat collected from "members of the An clan” (An jia ren A)> A fragment of a list of Ioans of grain (P. 2161) mentions a certain An Village. These records show that An Village was probably one of the Settlements in the Conghua sub-district that was originally established by Sogdians and which endured through the Guiyi Military period. Luo Village (Luo jia zhuang Many of the Luo clan living in Dunhuang during the late Tang and Five Dynasties, were descendants of Dunhuang Sogdians. Although a Luo Village is not recorded in the Dunhuang documents, they do mention repeatedly the words “Luo clan” (Luo jia). Many villages in late the Tang and Five Dynasties periods in Dunhuang especially villages established by Sogdians took the sumame of the clan as the name of the village. Therefore it is possible that the “Luo clan” (Luo jia) mentioned in the sources, is shorthand for Luo Village (Luo jia zhuang). One monastery account dated 921 or 981 States a disbursement of oil for the benefit of several shami of the Luo clan.39 An annual report of the Jingtu monastery States: “Twenty piculs of millet were collected as land tax from the Luo clan”; “Twenty piculs of wheat and twenty piculs of millet were used to buy land from Luo clan”.40 From these sources it is diffïcult to discem precisely who these people were, but we can infer that the term “Luo clan” (Luo jia) is the name of a village during the late Tang and Five Dynasties in Dunhuang. The Labor and Tax Register of the Conghua sub-district also reveal that there were two Luo clansmen who served as village heads in this sub-district. Therefore it is probable that Luo Village had originally been established by Sogdians and survived through the time of the Guiyi Military Regime. A private contract also records that a certain Luo Zhuzi of Mogao sub-district asked for a loan from a man of the Luo clan named You [...] [-J- 41

—

—

Di Village (Di jia zhuang Many people sumamed Di in the late Tang and Five Dynasties in Dunhuang were discendants of Sogdians. The Di clan was an aristocratie family. According to the Di Clan Stele recorded in P. 4640, they migrated to Dunhuang from Di City (Di cheng ^) and were officiais for générations. By the Five Dynasties period the Di clan of Dunhuang had become the astrologers of Dunhuang, and therefore wielded great influence.42 A monastery account States, “Three piculs were paid for the clothing of a 39

P. 3490 V°. Document ed. in CKCL, vol. 3, pp. 186-191. P. 2032 V°. Document ed. in CKCL, vol. 3, pp. 455-509. 41 Dx. 11C, fragment of contract, repr. in S ankt-Petersburgskoe otdelenie Instituta vostokovedenija Rossijskoj Akademii Nauk, 1996, vol. 6, p. 14. 42p.4640V°(seen 12). 40

354

Les Sogdiens en Chine

certain Ashizi RW'T' of the Di clan.”43 An account from the Jingtu monastery records that “on the 22nd day one jug of wine was provided by the Di clan.”44 The words “Di clan” (Di jia) recorded in these sources perhaps refer to the village of the Di clan. As for the answer to the question of whether the ethnie makeup of the Di clan was Han, Sogdian, or Uighur, further research is needed. Kang Village (Kangjia zhuang The Kang clan was also an eminent family in Dunhuang during the Guiyi Military period. However the Dunhuang documents do not clearly mention a village called Kang Village (Kang jia zhuang); they only mention Kang jia. From an analysis of document S. 6981, which States that “seventeen piculs of wheat was received from the fields of Kang/M" we can see that the words Kang jia probably refer to the name of a village.45 Document S. 6829 V° records that: On the 22d day of the first month [of the year 806], four bolts of cloth, measuring a total of one hundred seventy-six feet, valued at the équivalent of ten piculs of wheat plus three piculs of millet, was paid to Kangjza for timber.46

A monastery account dated from the middle of the tenth Century, records that “four bushels of wheat, four bushels of millet, and eight feet of cloth were paid for the recitation and singing of sutras for a woman of Kang jia.'"7 We must wait for further research to leam more about the characteristics of the Kang village. The Kang clan enjoyed great réputation in Dunhuang during the late Tang and Five Dynasties. We hâve quoted above the offices held by Kang Shijun, Kang Tongxin, and Kang Xianzhao during Zhang’s control of the Guiyi Army. Therefore there were probably Sogdians sumamed Kang who established village settlements in Dunhuang during this period. Shan Village (Shan jia zhuang Many people sumamed Shan lived in Dunhuang during the late Tang and Five Dynasties. They were not Sogdians but the descendants of Shanshan people who had emigrated from Shicheng in the Lob nor région. There were three waves of Shanshan émigration, the first in the third year of the Taiping Zhenjun reign (442), the second when the Northern Wei kingdom destroyed Shanshan in 445, and the third at the end of the Northern Wei when the Tuyuhun conquered Shanshan.48 The term Shan jia refers to the village Settlement that was established by the Shanshan people who had immigrated to Dunhuang from Shanshan. This claim is supported by document S. 6981, which refers to a certain monastery that collected rent (six piculs of wheat) from the

43

P. 4694, ed. in CKCL, vol. 3, p. 565. m S. 6452 R°, List of disbursements of wine from the Jingtu monastery (982), ed. in CKCL, vol. 3, pp. 224-226. 45 Document ed. in CKCL, vol. 3, pp. 140-141. 46 S. 6829 V°, List of disbursements for construction, ed. in CKCL, vol. 3, pp. 146-147. 47 P. 2040 N°, ed. in CKCL, vol. 3, pp. 401-434. 48Zheng, 1995, pp. 272-288.

Zheng Binglin

Non-Han ethnie groups in Dunhuang

355

fields of “Shan jia south of the city.5î49 It also indicates the location of this village, south of Dunhuang City. Shi Village (Shi jia zhuang 史家 庄) Shi Village was established by Sogdians sumamed Shi who immigrated to Dunhuang and was probably one of the villages in the Conghua sub-district during the Tang Dynasty. Shi Village is not mentioned in the Dunhuang documents, but we can surmise that Shi Village was indeed a Sogdian village during that time. Document S. 1600 records that in the year 961y of all the channels, fields, and scattered buildings that were to be taxed, "two piculs of wheat were received from the Shi jia lands?350 Document S. 6981 States that éÉfive piculs of wheat were received from the Shi ノ必 lands?551 These two documents are records of the land tax imposed by the Lingxiu 修 Monastery on the households that lived on and farmed the fields of Shi Village. We can see from these sources that a Shi Village existed there during the Guiyi Military period.

Did non-Han establish a Mi Village (Mi jia zhuang 米家 庄), a Long Village (Long jia zhuang 家 庄), and a He Village (He jia zhuang 何 家 庄) in Dunhuang during the late Tang and Five Dynasties? The scarcity of sources makes it difficult at present to answer these questions. However, the study of available sources has yielded one conclusion: that in the late Tang and Five Dynasties in Dunhuang, many Sogdians and other non-Han lived in various sub-districts and among varions tribes (buluo 部落), other than the village Settlements discussed above. Sogdians left their mark in each of the scattered villages in which they lived. Document Dx. 1418 records a certain Shi Qingzi 石 子of the Shensha sub-district 神沙 , an An Shantong 安善通 and Di Mabu 翟 步of the Longle sub-district 勒 , a guanjian 官健 official named Di Qingde 翟 德, and a certain Shao Zaiding 再定of Yuguan sub-district 玉 . This source is a list of Ioans of beans granted by a monastery during the Guiyi Military period.49 52 It shows clearly that there were many Sogdians and Tibetans livmg in various sub-districts under the Guiyi Military Regime. Another monastery5s account (P. 4913) records the name of "An Baotong of Shouchang 寿昌安保通”who fumished half a bushel of oil in exchange for six bushels of hempseeds.53 Shouchang District was on the main thoroughfare between Dunhuang and the western lands of Khotan and beyond. All travelling envoys had to pass through Shoushang, and therefbre it could well be that this place was a commercial outpost. Therefbre Sogdians, being a people who lived by trade probably, also lived there. In these various sub-districts, it was easy fbr these non-Han to mix with the Han, and the rate and degree of sinification were very high. Sogdian women living in this environment were likely to accept Han culture and customs even more quickly. A request presented by a peasant from the Pingkang sub-district 平康 in 952 records: 49

List of accounts dated 2nd half of the tenth Century, ed. in CKCL, vol. 3, pp. 140-141. Document ed. in CKCL, vol. 3, p. 528. 51 Seen 49. 52 Document ed. in CKCL, vol. 2, p. 266. 53 Document ed. in CKCL, vol. 2, p. 256. 50

Les Sogdiens en Chine

356 Five mou 0116 bolt

land5 3〇1