Hansen v.-turfan as a Silk Road Community

7/29/2019 Hansen v.-turfan as a Silk Road Community

1/7

Yii Ying-shih

. s e m i - a ~ m u a l l y ) , . is published ?y the l n ~ t u t e . of History and. . T;tJwan. It IS a contmuation of MW Major, New Series,. 197.5, It co,,:":rs all periods of Chinese history, literature,ate the histOries and cultures of other East and CenChina. nstruc.tions axe in Asia Majors style sheet)

:;,uIlp.ei:i)J[nit is a!;iout 65 Pfges, total. Both main text and notes must. : " u . " , ~ ! U ! S , clear resolution, standard paper: We do not. acceptpages: All citations !).lust be complete, and allform only) and Japa nese di rectly in the text,

un l lUU I .cen. by compu.ter with clear "esoiution. Acceptednr.'''''r'',l : computer, .usmg standard programs (Word Per. not-enclose diskette with submittal, of time to our produ.ction, u'se only "smart",a separate file. containing text of footnotessheet), we require the standard Wade-Giles andpublication data are required at the first citation'used_ ~ r e q u ~ n t l y cited sources may be givenAsw Major uses standard abbreviations ofbelow..... ~ ; ; w H . _ e l " m a . t y reject as pellr "submittal require- 0 con orm lu y to our style sheet

t n p h ~ a t e , as well as any con' espondence about"Y!,iilIl"'glflgEdltor, Academi a Sinica, Institute ofHislor yR.O.C. Subscription inquiries go to the

'please inguire. Penn. residentspay 6% sales lax.)addressed to Asia Major,. P.O. Box 4(;5 Hanover

- ' ' 1 1 ~ ' ' ~ . ' .. 17) 632-8535; ask for Subscription S e ~ i c e s . Sub:E-mail: .. pubsvc@tsp .sheridan.com". Orders by?r MASTERCARD. Payments by checks and:. . ' U.S. funds, payable to "Asia Major.". ' ....EDU.TW/-KIESCH/ASIAMAJOR.HTMI.. . Items as the Stylesheet, titles of past articles and. and or dering back issues. . 1

V . ~ I . l l m . e - v e a r (two issues):$ 3 ( ~ students (to academic address, no 2-yr. discount);dlSCOun.t). Add $4 per vol.-yr. outside U.S. and Canada.Ser. (published since 1988), no minimum. $20 per

~ r ' : I e t \ f u ' j r , d ' : l m e s t i t : $8 for dam. express or overseas surface). 2()0f0 reduc- ." N ~ w Series" (1949-75), only spot issues, minimum fourAnmv. hard-bound issue. Add $8 per order domestic' $l4"Gomp' etteake several e e k ~ . ~ c l u d ~ a 1 t ~ m a t i v e t i t 1 e . l n q u i ~ e re.", . s e t ~ of both ongmal LeIpZIg senes and New Series are

~ ~ i . d d i ~ l j ; { h n b H D e t t e n d ~ ~ f . . ~ ~ 2 , l 2 ! ! l ~ F e ! ! r ; , b ~ ~ : . . e ~ r m a n Y ; ' ? f .

:.H I R D S E R I E S v 0 L U M E X I PAR T;1,",

Stable of Contents,tntroduction: Turfan as a Silk Road Community

VALERIE HANSEN

Concise History of th e Turfan Oasis and It s ExplorationZHANG GUANGDA and RONG XINJIANG

;0'{rfhe Path of Buddhism into China: Th e View from Turfan".""'":.'

VALERIE HANSEN',,-.

::!Sasanian an d Arab-Sasanian Silver Coins from Turfan: Their\.-t" Relationship to International Trade and the Local:: , Economy jONATHAN KARAM SKAFF

, \lnnovations in Textile Techniques on China's Northwest" Frontier, 500 - 700 AD ANGELA SHENG

I

13

37


2/7

V A L E R I E H A N S E N

Introduction: Turfan as a Silk Road Community

T he T'ang dynasty is justly famed for its openness to Indian, Ira niaCentral Asian cultures. Ch'ang-an, the T'ang capital, lay at the eterminus of a network of trade routes, now commonly known as the Silkthat linked China with civilizations farther to the west. The writings of fofficials and the surviving funeral figurines depicting non-Chinese tradehandlers of pack-animals provide vivid proo f of cultural exchange withsocieties, but they offer little information about the day-to-day existenon-Chinese living along the Silk Road. Turfan, an oasis on the northernaround the Taklamakan Desert, now in modern Sinkiang, commands tention because of its many primary materials on exactly that topic.

Unlike archeological sites in China, which often give us a gravindividual or a single family, the graveyards of Turfan contained ancommunity, in over 3 , 0 0 0 tombs. Lying outside the walled capital cityKao-ch'ang f.i.j I kingdom, the Astana and Karakhoja graveyards hegraves of the descendants of Chinese settlers who came to the oasisthird, fourth, and fifth centuries (see appended list of geographical term

Since 19590 Chinese archeolOgists have excavated 465 tombs in tana and Karakhoja graveyards, 20 5 of which yielded documents. TheALL TH E authors in this issue express their thanks to the Henry Luce Foundation, whichthe threeyear joi nt U,S"Chinese project, "The Silk Road Project: Re-uniting Turfan's STreasures," and supported the research appearing in this issue. The Foundation's fumade it possible for Professors Zhang Guangda SftJJiIl and Rong Xinjiang ~ w r r I tosemester each as visiting scholar at Yale University. The authors would also like to thaChinese institutions for making their collections available to project members during a in May. 1995: The Sinkiang Archeological Institute (Hsin-chiang k'auk u so ; f f i i j J ~ 1 iMuseum of the Autonomous Sinkiang Region (Hsin-chiang tzu-chih ch'ii po-wu kuan w

m : m ~ r g ) , and the Turfan Museum (T'u-lu-fan ti-ch'u wen-wu chunghsin lltHiJ:i!1.1-C,,). Finally, the authors would like to thank the Committee on East Asian Studies at Ythe Center for Middle Period History at Peking University (Pei.ching ta-hsueh Chung-kukushih yen-chiu chung-hsin ~ t * * * r : p W:p ~ 5 ! : l i f f ~ r:p 'C.') for co-sponsoring the projfour papers appearing here were first presented at the Silk Road Project inJuly, 1998.Conference proceedings of The Third Silk Road Conference at Yale University, ivolume photocopied packet, are available from RIS-Yale University. 155 Whitney AvenHaven. CT 06520. fax: 23-432-6274. Abstracts of the conference papers are on theweb-site: . Other papers from the projecpear in the follOWing journals: Orientations: '!he Maga:;ne for CoUectors and Connoisseurs of30 .4 (1999), Journal ofAsian Studies 58. I (1999), Tun-huang Tu-tu-fan yen-eMu t z j : l I ! 4 (1999), and T'oung Pao 85.2 (1999).

I


3/7

documen t dates to the third century AD; the latest, to the eighth. The Chinesesettlers in Turfan bu ried their dead with hats, belts, and shoes - unusually, allwere made from paper. Because writing materials were scarce, they recycledpaper that had writing on one or both sides, sometimes painting the charactersover. They cut up both discarded official documents, such as imperial ordersand household registers, and private records like contracts, family letters, andwriting exercises. Like pieces of a puzzle, these paper fragments must be joinedtogether before they can be read, and one of the greatest feats of Chinesescholarship in this century has been the transcription and publication of thesedocuments. The volume at hand demonstrates the value of linking these documents to objects that have been found in tombs and hoards - coins, figurines,and textiles. These materials from Turfan make it possible to write a grassroots history of an entire Silk Road community during the time when Chinalooked west, to India, to Central Asia, and to Iran.

From a Chinese point of view, Turfan was one of the farthest outposts ofthe T'ang empire. By the year 500, Turfan's population consisted largely ofChinese settlers. In the first article in this issue, Professors Zhang Guangda ~.ll nd Rong Xinjiang I U J f r r show that for much of the oasis's long historythe Chinese did not exercise direct rule. The one exception was the periodbetween 640, when T'ang-dynasty troops conquered the oasis, and 803, whenthe T'ang relinquished controL After 803 a series of different Uighur (Uygur)and Mongol dynasties governed the oasis until I756, when it once again cameunder direct Chinese rule. But it was during the seventh and eighth centuriesthat local officials performed the numerous bureaucratic duties reqUired bythe Chinese state. Turfanqs so far the only site in China to have produced somuch documentation about the fiscal system of the T'ang period.

Yet, from an Iranian point of view, Turfan was also the farthest outpost of Iranian civilization because it contained a large settlement of Sogdians. The homeland of the Sogdians, Sogdiana, was the region aroundSamarkand, in modern-day Uzbekistan, and the Sogdians spoke and wrote adialect of middle Persian. The Sogdians were heavily involved in commerce,and many traded with the Chinese oases on both sides of the TaklamakanDesert. Some s ~ t t l e d in Turfan, while others came on frequent business trips.The fall of the Sasanian empire to conquering Islamic forces in 651 and theexile bf the Sasanian Prince Firiiz to Ch'ang-an prompted further Sogdian settlement in T'ang China.

A corpus of fragments of official documents makes it possible to have aglimpse of the equal-field system in action. The provisions of that system applied to both the Chinese and to the non-Chinese residents (mostly Sogdian)

2

.:,:";. of Turfan.1 At the stipulated times local officials entered the names of non-

Chinese residents on household registers, assessed and allotted land to them,and recruited them to perform labor service. Some Sogdians received officialtitles and served in the T'ang bureaucracy.Sogdians not resident in Turfan also played in important role in lifethe oasis. They had access to the court system, by which they could brmg SUltsagainst their countrymen and also against Chinese, although they may havedepended on an interpreter to do so. An entire i n f r a s t r u c t u ~ e . d e v . e l o p e d tohost the merchants who came to the oasis to trade. Local offlClals Issued therequisite travel documents, while tradesmen prov ided places to stay. One legal dispute from Turfan concerns the death of a non-resident ~ r c h a n t whodied while staying with a Sogdian innkeeper. Th e list of occupations pursuedby sogdians illustrates the extent to which they participated in local life. Ascould be expected, there were Sogdian traders and farmers, but Turfan wasalso home to Sogdian coppersmiths, painters, leather-workers, iron-mongers,and even one veterinarian with the curious nickname "Ostrich."

The articles in this issue of Asia Major concentrate on the interaction between the Chinese and the non-Chinese in Turfan. The authors find very littleevidence of Indian influence at Turfan. What little there was dates to the fourthcentury AD, when the local rulers of Turfan patronized u d d h i ~ t s . The a r t i c ~ e sdo show, however, just how important and influential the Sogdlan commumtywas, especially in the years after 500 .Jonathan Skaffs paper skillfully combines an analysiS of the Chinese-lan-guage sources with a close examination of the numismatic e v i ~ e n c e order tosketch the economic activities of the Sogdians. Most of the SlIver coms foundat Turfan were minted either by the Sasanians or their Arab successors. Thesecoins have been found either in hordes, dating to the fourth century, or in themouths of the dead buried at the Astana and Karakhoja cemeteries.Aurel Stein, who arrived at the cemetery site onJanuary 18, I9 I 5, wasone of the first Westerners to documen t the custom of placing coins in themouths of the dead, which he attributed (wrongfully) to the influence of theGreeks, who had also placed coins in the mouths of the dead so they could paythe ferryman who took them over the river to Hades. Stein i r e ~ a man a m ~ dMashik, an experienced grave-robber, as his chief digger. Mashik assured Stemthat every single one of the tombs at Astana and Karakhoja had already been

1 Wu Chen [Wu Zhenl , "Assut'ana Ha-la-ho-cho kumu-ch'iin k'ao-ku tzu-liaohsien le hU'jen" i l " J J t f f J o g : ; J j ~ I l l t ! z : f D . ! j ~ : i 5 ~ l l f ; ; ; t i W M ~ f . l H t : J M . A . , in the proceedings of TheSilk Road Conference at Yale University. and forthcoming in Tun-huang Tu-Iu1an yen-clzzu 4(1999) .

3


4/7

disturbed, but that he could guide him to the tombs that were less disturbedthan the others. In addition, Mashik had developed his own ingenious measures for finding artifacts unnoticed by earlier intruders:

Mashik, our special cemetery assistant, whom long practice in searchingthe dead had relieved of all scruples, by breaking the jawbones of theskull recovered from the mouth cavity a thin gold coin .. Mashik claimedthe distinction of having been the first to learn by experience to look forcoins of gold or silver placed in the mouths of the dead, though his searchwas but rarely rewarded.2Archeologists working since Stein have continued to unearth these coins

from the mouths of the dead. The coins, interestingly, are not distributed evenlyover time. Coins first appear in the mouths of the dead late in the sixth century, peak in the 6 ~ ; w s through 650s, and disappear by about 710. The majorityof the coins found at Turfan were minted by the Sasanians or by the Islamicregime that conquered the Sasanians in 65!. T'ang-dynasty copper coins didnot circulate in the oasis until the 680s.

Several important historical events, Skaff suggests, led to the appearanceof the silver coins. At the end of the sixlh century, just as Turkic-speakingpeoples gained control over much of Central Asia, the Chinese empire wasreunified for the first time since the fall of the Han dynasty in 220 AD. Meanwhile, the Sasanian empire produced large quantities of high-quality coins withsilver content ranging between 85 and .go percent. These coins circulatedthroughout Central Asia. Both the independent kingdom of Kao-ch'ang andthe T'ang-dynasty governn.'lent accepted Sasanian silver coins for the paymentof taxes in that area.

The high point in the usage of Sasanian silver coins - the seventh century- coincides with the appearance of new types of textile. Angela Sheng's articleconcentrates on one group of.unusual textiles that used complex weaves forsimple designs. Although the simple deSigns have failed to attract the attentionof textile historians, some, such as the tree-leaf pattern, could only have comefrom Iran, and not from China. Sheng proposes that the group of simple-design textiles commissioned by Sogdians resident in Turfan. Some of thewealthiest had the necessary capital to invest in weaving workshops, and theChinese or Sogdian craftsmen they hired had the requisite knowledge of weav-ing to develop such new textiles. .

Clearly the Sogdians constituted an important community in Turfan. Yet2 Aurel Stein, Innermost Asia: Detailed Report ofExplorations in Central Asia, Kan-su, andEastern

Tran (Oxford: The Clarendo n Press, I928), p. 646.1-

the absence of evidence abou t their religious practices has long puzzled historians. Mazdeism, or Zoroastrianism, was the official religion of the Sasanianempire, yet no Mazdean temples have been found in Turfan. Zhang u a n g ~ a .has recently proposed a new interpretation of a little-understood text descnb

'.jng the sacrifice of sheep and oxen on certain festival days.3 These s a c r i ~ c e s'Were certainly not Buddhist, yet they were organized by the Board of Sacnfic-. of the independent kingdom of Kao-ch'ang, whose rulers consistently patronized Buddhism. According. to Zhang, though, the sacrifices were made toMazdean deities who included the supreme god Ahura Mazda and the god ofthe wind.Although little evidence of Iranian religion in Turfan survives,4 the m o ~ -ern tourist sees Buddhist remains everywhere. Turfan is justly famous for ItsBuddhist cave complexes at'Bezeklik and ne arby Sengim (Sheng-chin-k'ou WJ~ D ), and a towering Buddhist stupa rises above the earthen r u i n ~ ofch'ang city (see appended map 3)' In 630' the m ost celebr ated Chl.nese pIlgrim of all, the great monk Hsiian-tsang ~ ~ , staged a hunger stnke t h e ~ ebefore the king would allow him to continue his journey overland to Indla.One would expect that most of the Chinese residents were Buddhist e v ~ t e e s ,but the above-ground evidence of Buddhism everywhere does no t qmte fitwith the below-grourid evidence surviving in the Astana tombs.

This tomb evidence is the subject of my own article on the paths thatBuddhism took entering China. Many of the earliest records of Buddhism -dating to the late-fourth century - concern patronage of Buddhism bylocal rulers of Turfan. The latter built monasteries, commissioned the copymgof Buddhist texts, and in 382 sent a learned monk to China's northern dynastic capital, where he recited thousands of verses by heart. But the graves of thelocal residents of Turfan show few signs of Buddhist belief. The dead, the localresidents believed, traveled to another world (sometimes over a bridge of udgment like that crossed by Mazdean believers) where they could use many ofthe same objects they had used in life. They often included inventories of thesegoods in their tombs. .The first signs of Buddhist belief to be found in the Astana tombs date to

3 Zhang Guangda, "Some Iranian ReligiOUS Evidence in Turfan Chinese Tex.ts," the proceedings of The Third Silk Road Conference at Yale U n ~ v e ~ s i t y . A ~ h i n e s e verSion Will a ~ p e a rin Tun-huang T'u-lujan yen-eMu4 (1999) as "Tu-lu-fan ch u-t u Han-yu wen-shu chung so mien 1-lang-yii ti-ch'ii tsung-chiao te tsung-chi" j l : U l J r t : e ~ ~ ) < : . $ pfr # l " a t 3 ~ J j Q l i ! i i * ~ f . J ~ J l I J l i ..

4 Chao Huas han suggests that many of the cave-temples previously thought to be u d d h l s twere actually Manichean, although few would agree with all of his claims.; "New EVIdenceManichaeism in Asia: A Description of Some Recently Discovered Mamchaean Temples InTurfan," MS 44 (1996), pp. 267-3 1 5'

5


5/7

543 - the date of a tomb inventory that clearly uses Buddhist language: amonk reports the death of the tomb occupant, who is identified as upholdingthe five precepts of the Buddha. Such inventories are clearly based on a form(some are incorrectly filled in), and, as I show, they suggest that Buddhismcame to Turfan not from the west, as we might expect, but from the east, theancestral home of many of the Chi nese settlers there.

The above lack of Indian influences on Turfan after 500 reflects the situation of material culture, as presented by my colleagues in this volume. Weuse varied approaches and different bodies of evidence, but our findings overlap to a surprising extent. In short, we find that most of the evidence of anIndian presence in Turfan dates to the early period before the founding of theKao-ch'ang kingdom in 500. After 500 the non-Chinese community in Turfanconsisted almost entirely of Sogdians. These Sogdians pursued a host of occupations, used Sasanian silver coins to conduct trade, and commissioned newkinds of textiles with hybrid designs. The oasis of Turfan was able to thriveprecisely because it offered them a home where they could be Sogdian, evenas they lived among the Chinese.

A NDTE 10 READERSTwo tables are placed at the end ofthis introduction. The first gives the alterna

tive spellings of different places ana peoples associated with the Turfan area and withTurfan history - terms that the reader may have trouble recognizing or sorting intocorrect categories. The second gives the names and reign dates of the ten Ch'ii-familykings who ruled the Kao-ch'ang kingdom from 502 to its conquest by the T'ang dynas-tyin 640. .

Also at the end of this introduction are three maps, meant for use throughout theentire issue. Map I shows the trade routes connecting China with Central Asia andSogdiana (in modem-day Uzbekistan). Map 2 is a closer-scaled view of the historicalTurfan area. Map 3 is a detailed map of the city of Kao-ch'ang during the fifth to eighthcenturies, when it served first as the capital of the Kao-ch'ang kingdom ~ d then theprefectural center of Hsi-chou prefecture unde r the T'ang.

6

Table of Selected Geographical and Ethnopolitical NamesCHINESE(Tang admin. name)

ALTERNATIVEROMANIZATJON

(Geographic and administrative names)Astana A-ssu-t'a-na J l i i J W T J 3 = ; g ~

also Karakhoja Bmdn. TurfanKarashahr

SengimSinkiangTashkurganTokbaristan (Bactria)Tun-huangUrumchiYarkandYarkhoto(Ethnopolitical names)n/ aHephthalitesnl aKirghizn/anlaTurksUighurn/aWakhan

(Pei-t'ing :It!}!) BeshbalyqPai-tzu-k'o-li-k'0 m ~ ~ : R :An-kuo ~ ~ BukharaTa-yiian 7 : . ~ FarghanahHa-mi (I-chou {ll"')'I'I) QomulT'ien-chu ::K:?!

~ ~ A Qocho(Hsi-chou iffi 1-1'1)Ha-Ia-ho-cho ~ m ~ D . E j [ QarakhojaTu-Iu-fan o ~ *Yen-ch'i ~ ~Yii-t'ien T\jtJ, mdn. ~ D E BK'o-tzu-erh : R : ~ mK'u-ch'e J]j[]\[(An-hsi 3i:iffi ) cCh'iu-tz'u l i ~ 'Ho-kuo f i i J ~K'ang-chiialso K'ang-kuo ~ ~Sheng-chin-k'ou ~ ~ i z pHsin-chiang mIDiHo-p'an-t'o ~ J i ; J f t :

Qarashahr, or Agni

KushanlyahSamarqand

TukharistanT'u-huo-lo Q j ( ~Tun-huang(Sha-chou ~ : : H ' I ' I )Wu-lu-mu-ch'i , % , i ~ p ! \ : ~ { UrumqiChu-chii-po * l ~ i & :mdn. Yeh-ch'eng n-t:tnJ(Chiao-ho )i:}iiJ

C h u - s h i h l f - ~ l j i I Ku-shih ~ i ! i 8 f ! iI-ta ta'l:.Kao-chii iWi I!Chien-k'un Mre, Qyrqyzor Hsia-chia-ssu f a ~ W TShan-shan m ~T'ieh-leT'u-chiieh ~ f f i ! XHui-ho (-hu) @J*Z: ( ~ ) , Uygurmdn. Wei-wu-erh #.t:g:mJou-jan*f!.t Ouan-juan r r a ~ )Hu-mi ~ ~

A Indep. kingdom before T'ang conquest. " Na me used under Uighurs. C I.e An-h'i tu-hu-fu (protectorate).


6/7

Table ojCh'ii-Family ~ ~ Rulers in Kao-ch'angGEN'L.ORDER NAME REIGN NAME

I Ch'ii Chi a f t 1 ~ Ch'eng-p'ing :iJ:ifI-hsi

2 Ch'ii Kuang ~ * Kan-Iu 1 : t ~3 Ch'ii Chien ~ M Chang-ho4 Ch'ii Hsiian-hsi ~ ~ i } Yung-p'ing 7J


7/7

. .2t:no l ' - - -- ,\ _ ~ e;- - T'icn-shan_.i;;;--'-'- - *.. - -.:. -__ o : _ ~ . Kao-ch'ang (Qocho)TO KUCHA Hsi-chou Pre ftctur e. est. 64Cf(Ch'i1 Kao-ch' ang Kiogdom, 500-640)

Shao-shan

Qaidam Basin

Map 2 . The TUTfan Region under Tang Rule(Distances are not to scale)

; Cl) !>r ro ~rrr=; 'is-..., n (,,0. C)~ ' R -.i : I > 4 ~; "ifIo ~i:I at.. i=>~ ~iil ~~ ! ! ~(l) lJt!Il' 0i:I HlI! ,i,~ ~ ~2l Wr ~Rmm ~~ ~s . ~_."

: . ~ : ;;tti:I ~r::::x:: !:S~ " ~

. ~ ~~'"d Co. ...

"-==-="""" \... ;:,..

:> ....-:r "'"\l(\l ~ ~t:'"....o...,

\1j

NlA.

-z\1 - z..:j '"">

Documents

Hansen v.-turfan as a Silk Road Community