32
Eurasian Prehistory, 6 (1–2): 167–198. PREHISTORIC ARCHAEOLOGY IN THE ZHUNGE’ER (JUNGGAR) BASIN, XINJIANG, CHINA Peter Wei Ming Jia 1 , Alison V. G. Betts 1 and Xinhua Wu 2 1 Department of Archaeology, University of Sydney, NSW 2006, Australia; [email protected], [email protected] 2 The Archaeological Institute of Chinese Academy of Social Science, 27 Wangfujing Road, Beijing 100710, China; [email protected] Abstract Xinjiang is a vast area of mountains and desert basins that formed the main route for contact between early China and the West. The southern Talimu Basin is well known for the remarkable organic preservation there of burials and aban- doned settlements. Lesser known is the northern Zhunge’er Basin that lay across the route out to the Eurasian steppes. From at least as early as the Bronze Age it witnessed important transmissions of technologies and cultural influence, but at present the nature and timing of these is poorly understood. This paper is a review of research to date on the prehistoric pe- riods in the Zhunge’er Basin and the areas immediately surrounding it. The paper also provides a review of the problems that need to be addressed in the archaeology, and particularly the chronology, of the region. INTRODUCTION The Zhunge’er (Junggar) 1 Basin, together with the Talimu (Tarim) Basin to the south, make up the territory of Xinjiang, at the far western end of China. These vast arid inland basins have an extraordinarily rich cultural history that has so far only been documented in a very fragmentary and tantalizing manner. Archaeological work began in Xinjiang in the early 20 th century but, until a few decades ago, much of this related to basic explo- ration and not to systematic research and excava- tion. Of the two basins, the Talimu has been more extensively investigated. Its dry climate has per- mitted the most exceptional preservation of or- ganic material and from here researchers have re- covered a wealth of manuscripts, wooden arte- facts, fabrics and mummified human bodies. The Zhunge’er Basin to the north has a slightly moister climate and does not experience the same conditions of preservation. The main trade routes of the Silk Roads passed through the southern Talimu Basin. Because of this important historical focus and also possibly because of the unusual preservation, most archaeological work has been concentrated here. Yet, of the two basins, for the prehistoric periods, the Zhunge’er is arguably the more important, at least in the Bronze and Iron Ages. In this paper the authors seek to bring to a wider, and particularly a Western, audience a de- tailed review of what is known to date of the later prehistory of the Zhunge’er Basin, incorporating recently published Chinese research 2 . From the late 1980s onwards, several summaries of the available evidence for the later prehistory of Xinjiang have been published, each one refining that of previous works 3 . However, our knowledge of this important region is still very sketchy, based largely on relative ceramic chronologies from widely dispersed funerary contexts. The aim here is to discuss in more detail the particular evi- dence for the northern part of Xinjiang, the Zhunge’er Basin. The paper also reviews the problems that need to be addressed in the archae- ology, and particularly the chronology, of the re-

Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Embed Size (px)

DESCRIPTION

Eurasian Prehistory, 6 (1–2): 167–198.PREHISTORIC ARCHAEOLOGY IN THE ZHUNGE’ER (JUNGGAR) BASIN, XINJIANG, CHINAPeter Wei Ming Jia1, Alison V. G. Betts1 and Xinhua Wu21Department of Archaeology, University of Sydney, NSW 2006, Australia; [email protected], [email protected] 2 The Archaeological Institute of Chinese Academy of Social Science, 27 Wangfujing Road, Beijing 100710, China; [email protected] Xinjiang is a vast area of mountains and desert basi

Citation preview

Page 1: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Eura sian Pre his tory, 6 (1–2): 167–198.

PRE HIS TORIC AR CHAE OL OGY IN THE ZHUNGE’ER(JUNGGAR) BA SIN, XINJIANG, CHINA

Pe ter Wei Ming Jia1, Alison V. G. Betts1 and Xinhua Wu2

1 De part ment of Ar chae ol ogy, Uni ver sity of Syd ney, NSW 2006, Aus tra lia; pe [email protected],[email protected]

2 The Ar chae o log i cal In sti tute of Chi nese Acad emy of So cial Sci ence, 27 Wangfujing Road, Beijing 100710,China; wxh63114@ya hoo.com.cn

Ab stract

Xinjiang is a vast area of moun tains and desert bas ins that formed the main route for con tact be tween early China andthe West. The south ern Talimu Ba sin is well known for the re mark able or ganic pres er va tion there of buri als and aban -doned set tle ments. Lesser known is the north ern Zhunge’er Ba sin that lay across the route out to the Eur asian steppes.From at least as early as the Bronze Age it wit nessed im por tant trans mis sions of tech nol o gies and cul tural in flu ence, but atpres ent the na ture and tim ing of these is poorly un der stood. This pa per is a re view of re search to date on the pre his toric pe -ri ods in the Zhunge’er Ba sin and the ar eas im me di ately sur round ing it. The pa per also pro vides a re view of the prob lemsthat need to be ad dressed in the ar chae ol ogy, and par tic u larly the chro nol ogy, of the re gion.

IN TRO DUC TION

The Zhunge’er (Junggar)1 Ba sin, to getherwith the Talimu (Tarim) Ba sin to the south, makeup the ter ri tory of Xinjiang, at the far west ern endof China. These vast arid in land bas ins have anex traor di narily rich cul tural his tory that has so faronly been doc u mented in a very frag men tary andtan ta liz ing man ner. Ar chae o log i cal work be gan in Xinjiang in the early 20th cen tury but, un til a fewde cades ago, much of this re lated to ba sic ex plo -ra tion and not to sys tem atic re search and ex ca va -tion. Of the two bas ins, the Talimu has been moreex ten sively in ves ti gated. Its dry cli mate has per -mit ted the most ex cep tional pres er va tion of or -ganic ma te rial and from here re search ers have re -cov ered a wealth of manu scripts, wooden arte-facts, fab rics and mum mi fied hu man bod ies. TheZhunge’er Ba sin to the north has a slightlymoister cli mate and does not ex pe ri ence the samecon di tions of pres er va tion. The main trade routesof the Silk Roads passed through the south ernTalimu Ba sin. Be cause of this im por tant his tor i cal

fo cus and also pos si bly be cause of the un usualpres er va tion, most ar chae o log i cal work has beencon cen trated here. Yet, of the two bas ins, for thepre his toric pe ri ods, the Zhunge’er is ar gu ably themore im por tant, at least in the Bronze and IronAges.

In this pa per the au thors seek to bring to awider, and par tic u larly a West ern, au di ence a de -tailed re view of what is known to date of the laterpre his tory of the Zhunge’er Ba sin, in cor po rat ingre cently pub lished Chi nese re search2. From thelate 1980s on wards, sev eral sum ma ries of theavailable ev i dence for the later pre his tory ofXinjiang have been pub lished, each one re fin ingthat of pre vi ous works3. How ever, our knowl edge of this im por tant re gion is still very sketchy,based largely on rel a tive ce ramic chro nol o giesfrom widely dis persed fu ner ary con texts. The aim here is to dis cuss in more de tail the par tic u lar ev i -dence for the north ern part of Xinjiang, theZhunge’er Ba sin. The pa per also re views theprob lems that need to be ad dressed in the ar chae -ol ogy, and par tic u larly the chro nol ogy, of the re -

Page 2: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

gion, as the start ing point to a new field study thataims to fo cus on some of these is sues4. Key con -cerns with re gard to the an cient his tory of the re -gion are the or i gins and move ments of pop u la -tions and the trans mis sion of cul tural traits andin no va tions. The ques tion of the or i gins of theXinjiang pop u la tions is a vexed and com plex one. The most re cent ar gu ments are sum ma rized topro vide a back ground for in ter pre ta tion of the ar -chae o log i cal data. The spe cific ques tion of thespread of met al lur gi cal tech nol o gies is also ad -dressed as this is a fun da men tal com po nent of thepat terns of tech nol ogy trans fer be tween Eur asiaand China.

GEO GRAPH ICAL CON TEXT

Xinjiang is a large re gion cov er ing 165000square ki lo me ters of deserts, moun tains, grass -lands and oases. It has a sparse, scat tered and eth -ni cally mixed pop u la tion of no mads and oasisdwell ers. The larg est groups are the Turkic speak -

ing Uighur and Kazakhs, Mongols, and the Man -da rin speak ing Hui and Han Chi nese. No madic orsemi-no madic pastoralism – sheep, goat, cat tleand horse herd ing – forms the main econ omy formost Kazakhs and Mongols. The Uighur econ -omy is based on oasis ag ri cul ture com bined withstock rais ing. The Hui and Han have been mi grat -ing here from neigh bor ing ar eas to the south-eastsuch as Qinghai (Köknor) and Gansu since theQing Dy nasty around three hun dred years ago,bring ing with them a tra di tion of vil lage based ag -ri cul ture.5

Xinjiang is made up of two large desert bas -ins sur rounded by high moun tains. In the south isthe larger and more arid Talimu Ba sin; in thenorth is the Zhunge’er Ba sin, which has a veryslightly milder cli mate. The two are sep a rated bythe high ridges of the Tianshan (Tängri Tagh), asteep range of young, gla cier-capped, fold moun -tains. To the west, the way to the Eur asian steppeis blocked by a string of moun tains stretch ingfrom the Pamier (Pamir) range in the south up tothe Aletai (Altay) in the north. The main routes

168 Pe ter Wei Ming Jia et al.

Fig. 1. Map: Phys i cal fea tures of Xinjiang

Page 3: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

link ing East with West run from the Ferghana val -ley up over the dif fi cult Terekdavan pass acrossthe Pamier range into the Talimu Ba sin at Kashi(Kashgar), up the Yili (Ili) River val ley throughpasses across the Tianshan into the Tulufan(Turfan) De pres sion and the Zhunge’er Basin,and more eas ily into the Zhunge’er Ba sin at twopoints, across some low lands at Tacheng-Kela-mayi (Qöqek – Karamay), and up the E’erqisi(Irtish) river val ley. To the east, nar row cor ri dorslead out be tween the moun tains into the GobiDesert and In ner Mon go lia, pro vid ing the routeinto main land China. This spe cial geo graph icallo ca tion has placed Xinjiang in a unique po si tion,not only as a key node on the Silk Road in his tor i -cal times, but also as a crit i cal point for the meet -ing of cul tures and the trans mis sion of ideas andinnovations in the prehistoric periods.

The Zhunge’er Ba sin is a land of sharp con -trasts (Fig. 1). The cen tral Ku’erbantonggute(Gurbantunggut) Desert is low-ly ing, around 200me ters above sea level, while the sur round ingmoun tains rise rap idly to around 3000–4000 me -ters. The cli mate is hot and dry in the sum mer butcold in win ter. Ice caps form dur ing win ter snow -fall, and melt down into the streams and rivers inspring, feed ing the val leys and the oases. An cientset tle ments and buri als are usu ally dis trib utedalong the rivers and streams on the lower moun -tain slopes and in the oases cre ated as the riversrun out into the desert. Ver ti cal transhumance isthe most com mon pat tern of no madic move mentto day, as it al most cer tainly was also in the past(Frachetti, 2004: 146 ff.). Per ma nent set tle ment is only fea si ble in the small oases where lim ited ag -ri cul ture is pos si ble if sup ported by ir ri ga tion, and in the far west where rain fall in cer tain ar eas risesup to 600 mm.6 The an nual av er age tem per a tureto day is around 5–8°C in the low lands, val leysand desert, but 0–2°C in moun tain ous ar eas.There is about 40°C dif fer ence be tween sum merand win ter. Pre cip i ta tion is quite low, less than100mm in the desert but around 300mm in themoun tains, ris ing as high as 500mm on the peaks.A for est belt of Si be rian pines is well es tab lishedon the up per slopes of the moun tains. Be low thisis open grass land veg e ta tion, grad u ally de creas -ing in den sity to wards the foot hills. Stands of elmcan be found in the mid dle reaches of many of theriver val leys.

HIS TORY OF RE SEARCH

In ter est in the ar chae ol ogy of Xinjiang firstbe gan around the mid-nine teenth cen tury, with ase ries of Eu ro pean trav eler-ex plor ers, among themost fa mous of whom are Marc Aurel Stein, Sven Hedin and Al bert von Le Coq (Stein, 1907, 1912,1921, 1928; Hedin, 1899, 1903, 1931, 1933,1940, 1943; von Le Coq, 1928). These early ex -plor ers con cen trated mainly on the Talimu Ba sinin south ern Xinjiang. They ex plored the desertghost towns such as Loulan (Krorainia) andsearched for Bud dhist manu scripts among thecaves and tem ples on the desert rim. TheZhunge’er Ba sin was al most left un touched, ex -cept for some in ter est in tem ples and later walledcit ies (Chen Xingcan, 1997a: 46–48). Since thattime, ar chae o log i cal re search in Zhunge’er hasbeen car ried out by Chi nese ar chae ol o gists on alim ited scale. Work has been mainly re stricted tore gional sur veys and res cue ex ca va tions7 (Fig. 2).

Study of the pre his tory of Xinjiang be gan inthe early twen ti eth cen tury, but se ri ous anal y sis of cul tural con texts and chro nol ogy did not emergeun til the 1980s. In 1987 An Zhimin pub lished ananal y sis of ev i dence for the Neo lithic pe ri ods, in -clud ing a ba sic ty po logi cal se quence for the mi -crolithic stone tool tra di tions of the re gion, work -ing mainly from sur face col lec tions (An Zhimin,1987, see also 1992a). In his 1992 pa per on theBronze Age (An Zhimin, 1992b)8 he iden ti fiedtwo dis tinct cul tures within Xinjiang, one as so ci -ated with painted pot tery, the other with peb bletools and coarse-ware pot tery, some times with in -cised or im pressed dec o ra tion. By this time an in -crease in the num ber of ex ca va tions pro vided abroader data base for the ar chae ol ogy of the re -gion, and C14 dates linked to strati graphic se -quences be came avail able.

Most of the work up to this time was based on ev i dence from the Talimu Ba sin, and sites in andalong the foot hills of the Tianshan Moun tains.Still very lit tle at ten tion had been paid to theZhunge’er Ba sin. A few sites were ex ca vatedthere be fore the 1970s, most of them cem e ter ies.Based on the lim ited data avail able, ar chae ol o -gists at tempted to make fur ther anal y sis of theircul tural con texts. Chen Kwang-tzuu and Hiebert(1995) noted sim i lar i ties be tween ma te rial fromthe Qiemu’erqieke (Shamirshak cem e tery) and

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 169

Page 4: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

the Afanasievo/Okunevo cul tures of the Aletaiand south Si be ria.9 Based on some ran dom col lec -tions of Andronovo style bronze artefacts fromthis area, Mei Jianjun and Shell also dis cussed the pos si ble cul tural con nec tions be tween theZhunge’er re gion and the Eur asian steppe (MeiJianjun and Shell, 1999). As study of the ar chae -ol ogy has con tin ued over the past de cade, aclearer pic ture of the pre his tory of Xinjiang hasbe gun to emerge. Shui Tao iden ti fied eight cul -tural sub-re gions, three of which, Aletai-Tacheng, Balikun (Barköl) and Ulumuqi (Ürmüqi) arewithin the Zhunge’er area (Shui Tao, 1993). In alater anal y sis, An Zhimin di vided Xinjiang intoten cul tural zones, among them three, the moun -tain val leys of the mid dle Tianshan and the Bali-kun and Aletai grass lands, which are in or on themar gins of the Zhunge’er ba sin (An Zhimin,1996, 1998).

In the early 21st cen tury, study of pre his toricar chae ol ogy in Xinjiang has pro gres sively de vel -oped. There are now more re li able field data andre search re sults pub lished in Chi nese jour nals, as -so ci ated with a sig nif i cant num ber of pro fes sional ex ca va tions such as Xiaohe (Kiqik Müran) (Xin-

jiang In sti tute of Ar chae ol ogy, 2003) and Chawu- hugou cem e ter ies (Xinjiang In sti tute of Ar chae ol -ogy, 1999). Com pre hen sive pub li ca tion of theChawuhu burial grounds has pro vided sys tem aticstud ies of typology and chro nol ogy based on re li -able strati graphic ev i dence cov er ing all five cem -e ter ies. This is the first time that a ba sic chro nol -ogy of cul tural con texts has been es tab lished inone ex ca va tion site in Xinjiang. More re search ar -ti cles have been pub lished dis cuss ing new meth -ods and per spec tives. A case in point is the workof Mei Jianjun who has re worked ear lier ty po logi -cal anal y ses of of ten un strati fied metal artefactsby studying bronzes through chemical component analysis (Mei Jianjun, 2000).

Us ing the Chawuhu chro nol ogy, ar chae ol o -gists have at tempted to re vise the se quence of ar -chae o log i cal cul tures in Xinjiang. Build ing on the work of An Zhimin, Zhang Yuzhong (2002) hascre ated a frame work for Xinjiang pre his toric ar -chae ol ogy.10 He has placed the start of the BronzeAge around 2000 B.C. and the be gin ning of theEarly Iron Age around 1000 B.C. While this maybe broadly ac cu rate, given the size and di ver sityof the re gion there is un doubt edly lo cal ized vari -

170 Pe ter Wei Ming Jia et al.

Fig. 2. Map: Ar chae o log i cal sites in the Zhunge’er Ba sin

Page 5: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

abil ity. Guo Wu (2005) has un der taken doc toralre search on pre his toric cul tures in the Tianshan in the first mil len nium B.C., es tab lish ing a ba sicchro nol ogy and in ves ti gat ing cul tural in ter ac tions be tween Xinjiang and the out side world. Hiswork is based on a large num ber of ex ca va tions in Xinjiang in re cent years. An other scholar, HanJianye (2005), has also fol lowed on the work ofAn Zhimin, re de fin ing the dis tinct cul tural re -gions of Xinjiang. His di vi sions are al most thesame as An Zhimin’s, ex cept that he de fined onenew re gion in west ern Zhunge’er and com binedAn Zhimin’s dis tinct re gions of Balikun andHami (Kumul) into one.11 Han Jianye’s cul turaldi vi sions in clude six sub-re gions in or bor der ingon Zhunge’er. The two in the north, Aletai andTacheng (VIII, X), are not dis cussed in de tail dueto a lack of ar chae o log i cal data. There is muchmore in for ma tion on those in the south, Tulufanand the mid-north ern Tianshan slopes (II), theHami ba sin and the Balikun grass land (III), theYili Val ley (IV) and the Shihezi-Wusu (Xiho)area (VI).

To day, our un der stand ing of the ar chae ol ogyof the Zhunge’er Ba sin is still at a very ba sicstage. Ev i dence is based on a lim ited num ber ofex ca va tions and sur veys, most of them in thesouth. There is very lit tle in the way of re li ablestrati graphic data or ce ramic typologies linked toab so lute dates. The cul tural groups iden ti fied sofar are based in part on the dis tri bu tion of ar chae -o log i cal field work. Fur ther and more com pre hen -sive field re search is likely to re fine the cur rentpic ture sig nif i cantly. Based on very lim ited data,and us ing some par al lels from neigh bor ing ar eas,the aim of this pa per is to re view the cur rent stateof knowl edge and in ter pre ta tions of re gional cul -tures and chro nol ogy with a spe cial fo cus on theBronze and Early Iron Ages from the sec ond tothe late first mil len nia B.C. As a start ing point this study is fo cused on ce ramic ev i dence as this hasbeen the primary basis on which regional‘cultures’ have been identified.

ZHUNGE’ER BE FORE THE BRONZEAGE

Very lit tle is known about Xinjiang as awhole prior to the Bronze Age. Al most all avail -able ev i dence co mes from sur face col lec tions of

chipped stone artefacts.12 The Zhunge’er Ba sin,with even more lim ited ar chae o log i cal in ves ti ga -tion, of fers very lit tle ev i dence for the early pre -his toric pe ri ods.

Pre-Ho lo cene sites

While lit tle ev i dence ex ists for sites in Xin-jiang in pe ri ods prior to the Ho lo cene, finds fromsur round ing ar eas sug gest that in time more findswill be made in Xinjiang also.13 In China, Si be riaand parts of west ern Cen tral Asia there have beenfinds of stone tools dat ing back into the LowerPaleolithic. These ap pear to rep re sent two dis tinct tra di tions, a peb ble tool cul ture and a bifacial in -dus try (Derev’anko, 1998: 336). In south ernMon go lia there have been sev eral finds of LowerPaleolithic quartz ite peb ble tools. All are fromsurface col lec tions, oc cur ring on an cient river ter -races (Ranov et al., 1992: 57), while Okladnikovfound a bifacial in dus try in the foot hills of theYarkh Moun tains.14 An early peb ble tool com plex has been found at the site of Kul’daro in south ernTadjikistan; oth ers are known from Kazakhstan(Derev’anko, 1998: 338).

Mid dle Paleolithic sites char ac ter ised asMous terian oc cur widely across Eur asia but theyare very var ied in their tool typology and tech nol -ogy. Use of the Levallois tech nique oc curs atsome sites (Ranov and Da vis, 1979: 249). InUzbekistan the cave site of Teshik-Tash yielded aNeanderthal burial in as so ci a tion with a Leva-llois-Mous terian as sem blage of a spe cific lo caltype that in cludes more for mal tools such as scra-pers and points (Allchin, 1992: 83).15 Some sitesof Mous terian tra di tion have been found in theZhunge’er Ba sin. Bifaces have been found in theE’erqisi Val ley (Abdurrassul and Zhang Chuan,1997: figs 2.13, 3.5), and scrap ers found inQichengzi may ex hibit a form of Levallois tech -nol ogy (Fig. 3; Wang Bo et al., 1997: fig. 6.3, 4).A re cent sur face sur vey iden ti fied a large quarry-work shop with Levallois-like nu clei, large bladesand flake blades at Nongqishi in Tacheng dis trictand sur face scat ters of large flakes on the higherter races of the E’erqisi River near Fuyun (Olsen,2004: 3–4).

The Up per Paleolithic in Cen tral Asia is lesswell rep re sented that the Mid dle Paleolithic andap pears to de velop largely out of the pre ced ing

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 171

Page 6: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Mous terian in dus tries (Derev’anko and LüZun-E, 1992). Tool kits are char ac ter ized byblade tech nol ogy. Tools in clude a va ri ety of scra-pers and points. Sites have been lo cated acrosswest ern Cen tral Asia from Af ghan i stan up intoUzbekistan and Kazakhstan (Derev’anko and LüZun-E, 1992; Derev’anko, 1998; Da vis andRanov, 1999). In the east a num ber of sites havebeen found in Mon go lia (Derev’anko, 2000) andin China to the east of Xinjiang (Wang Bo et al.,1997; An Zhimin, 1992b; Xinjiang In sti tute ofArchaeology, 1995, 1989).

Early to Mid-Ho lo cene sites

The pe riod cor re spond ing to the Mesolithic,Neo lithic and Eneolithic of Eur asia and Si be ria isvery ill-de fined in Xinjiang and can not readily bedis cussed in terms of spe cific pe ri ods. Microlithic stone tool as sem blages have been found in theZhunge’er Ba sin but their ex act chro nol ogy is un -cer tain. In the north, six sites in Habahe Countywith mi cro-blade tech nol ogy were found dur ingsur vey in the up per E’erqisi val ley.16 An other sitewas found at Kelamayi in the west (Xinjiang In -sti tute of Ar chae ol ogy and Kelamayi Bu reau ofCul tural Rel ics, 1997) and two at Takeerbasitaoand Qichengzi in the east (Wang Bo et al., 1997).The sites of Xihonggou 1 and 2 (Xinjiang In sti -tute of Ar chae ol ogy and Shihezishi Mu seum,1998) and Yi’erkabake (Xingjiang In sti tute of Ar -chae ol ogy, 1995) in Mulei (Mori) are lo cated atthe south-east cor ner of the Zhunge’er Ba sin.Finds from these sites con sist mainly of stone

artefacts. There are rare oc cur rences of verycoarse red sherds and even rarer finds of metalfrag ments. The sites along the E’erqisi River oc -cur as ex ten sive shal low de pos its be low sanddunes im me di ately on the river banks. It is likelythat the sites con sist of pa limp sests of de flated oc -cu pa tion de pos its dat ing over long pe ri ods of time and rep re sent ing fre quent short-term use.Geometrics col lected from the E’erqisi val leymay be par al leled in Mesolithic as sem blages else -where in Cen tral Asia and Si be ria (Abdurrassuland Zhang Chuan, 1997: fig. 3-15). Leaf-shapedar row heads with semi-in va sive bifacial re touchfound on the same sites also oc cur widely acrossnorth ern China, Si be ria and the Rus sian Far East,where they are dated to the early Ho lo cene (Ab-durrassul and Zhang Chuan, 1997: fig. 3-3, 4).Later ar row heads are more elon gated with ex ten -sive in va sive re touch cov er ing both dor sal andven tral sur faces. Ar row heads of this type havebeen found at Yierkabake (Xinjiang In sti tute ofAr chae ol ogy, 1997: fig. 6-3) and in the Rus sianAletai (Fig. 4; Markin, 2000: fig. 3; Kungurova,2003: fig. 8).

In Eur asia and Mon go lia, the Neo lithic waschar ac ter ised by hunter-gath ers us ing microlithictech nol o gies. China to the south-east of Xinjiangsaw the rise of early sed en tary ag ri cul tural andstock-breed ing com mu ni ties in the area aroundthe Yel low River by at least 6000 B.C. (An Zhi-min, 1992a). The prin ci pal crops were mil let andrice; the main do mes ti cate pig. The ba sic NearEast ern Neo lithic pack age of sheep/goat/cat tleand wheat/bar ley was in tro duced into China

172 Pe ter Wei Ming Jia et al.

Fig. 3. 1. Hand-axe, E’erqisi Val ley (Abdurrassul and Zhang, 1997: fig. 2: 13); 2-3. Levallois core scrap ers,Qichengsi (Wang Bo et al., 1997: fig. 6: 3, 4). Not to scale

Page 7: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

either from the north through the Eur asian steppes or from the west across the Pamier range. Ce realcul ti va tion in Cen tral Asia was not wholly lim ited to the Bactrian oases. The Bronze Age peo ples ofthe Eur asian steppe prac ticed oc ca sional ag ri cul -ture and could have im ported ce re als into Xin-jiang along with do mes ti cated live stock.17 Theear li est ev i dence so far for wheat in Xinjiang isdated as late as 2000 B.C. at the Bronze Age siteof Xiaohe in the Talimu ba sin where it oc curs to -gether with mil let and do mes ti cated sheep, goatand cat tle (Xinjiang In sti tute of Ar chae ol ogy,2003).18

Ev i dence for the Neo lithic in Xinjiang is very sparse.19 Flint scat ters have been found in thefoothills of the Tianshan Moun tains and aroundthe Tulufan de pres sion. The econ omy of theseNeolithic com mu ni ties is un known al thoughdenticulated blades that may have been used assick les have been re cov ered from Asitana(Astana) in Tulufan.20 With the ex cep tion of ar eas

such as the Yili river val ley, Tacheng dis trict andparts of the north ern slopes of the Tianshan,Xinjiang lies out side the dry farm ing zone andearly ag ri cul ture may have been based on sim pleir ri ga tion. There is no ev i dence so far that thiswas prac ticed in the Neo lithic pe riod (An Zhimin, 1992a).

There is some ev i dence for the later pre-Bronze Age pe riod around the Zhunge’er Ba sin.Dur ing ex ca va tions in the Yili Val ley at the site of Jilintai, ar chae ol o gists found a de posit con tain ingmicrolithic cores sealed by a layer as so ci ated with Andronovo ma te rial (Ruan Qiurong, 2004).21

This is the first time that strati graphic ev i dencehas dem on strated the pres ence of pre-Andronovocul tural re mains in Xinjiang. More ev i dence ofthe peo ple who cre ated such stone tools may beseen from buri als in the Si be rian Aletai. Kami-nnaya Cave (Markin, 2000) is lo cated in the Kara- kol val ley, 1100 me ters above sea level. An in fant was bur ied to gether with an adult fe male in her

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 173

Fig. 4. Chipped stone ar row heads: 1–5. Grave 1, Solonstsy-5 (Kungurova, 2003: fig. 8); 6. Aletai (Abdurrassuland Zhang Chuan, 1997: fig. 3-3); 7. Yi’erkabake (Xinjiang In sti tute of Ar chae ol ogy, 1995: fig. 6). Not to scale

Page 8: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

mid-twen ties. A large num ber of artefacts werere cov ered from the burial, in clud ing stone ar row -heads, a com pos ite knife made us ing microlithicflakes, many small shell rings pos si bly once em -broi dered on fab ric, stone fish used as fish inglures and mar mot in ci sors. Based on C14 anal y -ses, the burial has been dated around 3400 B.C. Asim i lar burial was dis cov ered in the Solontsy-5cem e tery in the Up per Ob River val ley on thenorth ern slopes of the Aletai. Stone ar row heads, a com pos ite knife, a stone axe, a grind stone, stonefish hooks and har poons were found with BurialNo 1 (Fig. 5), as so ci ated with an adult body. C14anal y sis dates the cem e tery to around 3700 B.C.The ex ca va tors sug gest that the cem e tery be longsto a tran si tional pe riod when the lo cal Aletai cul -tural com plex merged with the Kitoi cul ture fromCis-Baikal, an event that prob a bly took place dur -ing the mid-fourth mil len nium B.C. in the Aletaiarea (Kungurova, 2003).

Both sites are close to the north ern part of theZhunge’er Ba sin. More ex ten sive re search in

Zhunge’er may well yield sim i lar sites on thesouth ern slopes of the Aletai range dat ing toaround 4000 B.C. or slightly later. Al though nosherds have been re cov ered from sites of this pe -riod, it is pos si ble that ce ramic pro duc tion mayhave be gun around this time. Ce ramic re mains oc -cur very rarely in the Kitoi cul ture from aroundthe early sixth mil len nium B.C. (Weber, 1995). In the south ern part of the ba sin, ar row heads sim i larto those from the Kitoi-Aletai cul ture have beenfound at Yi’erkabake in Mulei (Fig. 4.7; XinjiangIn sti tute of Ar chae ol ogy, 1997). Ar row headsfound at Shihezishi in the mid-Tianshan area (Fig. 6) may date the site as early as 3000 B.C. (Xin-jiang In sti tute of Ar chae ol ogy and Shihezishi Mu -seum, 1998). The unique asym met ric ar row head(Fig. 6 ex treme right) found with the more com -mon ar row head forms at Shihezi is of a typefound all over the Eur asian steppe and east Si be -ria dated around 3000 B.C. or ear lier. A metalform of this point continued its use into theBronze and Iron ages.

174 Pe ter Wei Ming Jia et al.

Fig. 5. Artefacts from Grave 1, Solonstsy-5 (Kungurova, 2003: figs 6, 7). Not to scale

Page 9: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

These post-Paleolithic sites are not clearlydated but, based on par al lels with finds from sur -round ing ar eas, they could be long to a pe riod justbe fore the Bronze Age proper, that is to some time be fore the third mil len nium B.C. Terms such asMesolithic and Neo lithic are dif fi cult to ap ply to a range of ill-de fined and poorly dated sites withmixed microlithic tra di tions that span a large body of time. This frag men tary ev i dence hints at aframe work for un der stand ing the pre his tory of the Zhunge’er Ba sin as more ev i dence co mes to light. The cul tural re mains from around the mid-Ho lo -cene to 3000 B.C. may prove sim i lar to the dis -cov er ies from Aletai and Cis-Baikal. The Kitoicul ture, and its lo cal vari ant which may beblended with the Aletai Mesolithic/Neo lithic,could also oc cur in the Zhunge’er Ba sin. This cul -ture may be as so ci ated with very lim ited ce ramicpro duc tion, pos si bly rep re sented by very coarsered sherds, but with quite ad vanced lithic tech nol -ogy, in clud ing microlithic com pos ite tools andpol ished stone tools. Fish ing may have played anim por tant role in the subsistence economy,supplementing hunting and gathering activities.

THE BRONZE AGE

It is likely that the tran si tion from theMesolithic/Neo lithic to the Bronze Age in theZhunge’er Ba sin, and in Xinjiang as a whole, isstrongly in flu enced by neigh bor ing ar eas (AnZhimin, 1992b; Mei Jianjun, 2003; Mei Jianjunand Shell, 1999; Chen and Hiebert, 1995). It ap -pears that pre-Bronze Age cul tures were re placedfairly rap idly by Bronze Age groups since a largenum ber of Bronze, and later, Iron Age sites werefound in this re gion. How this tran si tion oc curredand why it ap par ently hap pened abruptly is un -known. It is not clear as to whether it in volved ac -cul tur a tion and as sim i la tion, or com pe ti tion andcon flict. How ever, lack of ev i dence forpre-Bronze Age com mu ni ties in the Zhunge’erBa sin is not nec es sar ily in dic a tive of an emptyspace, a cul tural void sim ply filled in the BronzeAge by out sid ers. Ex ter nal cul tural in flu ence wasin deed very im por tant for the Bronze Age tran si -tion in Xinjiang but the pos si bil ity that the lo calpop u la tion adopted new tech nol o gies should notbe over looked. The sur round ing re gions that ex -

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 175

Fig. 6. Chipped stone artefacts from Shihezishi (Xinjiang In sti tute of Ar chae ol ogy and Shihezishi Mu seum,1998: fig. 1). Not to scale

Page 10: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

erted the great est in flu ence on the Zhunge’er Ba -sin were the Eur asian steppes and the Aletai/south Si be ria. In the south-east there is ev i dence forcon tact with Gansu, while in the west some dis -tant links to the oasis-based Bactrian-MargiananBronze Age cul ture of south-west ern Cen tral Asia can not be ruled out.22

To date anal y ses of the Xinjiang Bronze Agehave fo cused largely on iden ti fy ing cul tural com -plexes, pri mar ily on the ba sis of ce ram ics, ob -tained mostly from buri als rather than from strat i -fied con texts. Up un til re cently, it has beendif fi cult to tackle ques tions of chro nol ogy. Ab so -lute chro nol ogy has been ham pered by the lack ofC14 dates; given the lim ited ev i dence and a widedate range, rel a tive chro nol ogy can only pro vide a very gen eral frame work at best. The sum mary ofthe ev i dence for the Zhunge’er Ba sin be low out -lines a pre lim i nary chro no log i cal frame work andidentifies the evidence for external contacts.

Search ing for the Bronze Age

There are Bronze Age re mains around the pe -riph ery of the Zhunge’er Ba sin. These sites in -clude the Tianshanbeilu cem e tery in Hami city,the type site for a Bronze Age cul ture found along the Balikun grass lands in the south east cor ner ofthe Zhunge’er Ba sin (Lu Enguo et al., 2001), theQiemu’erqieke cem e tery in the south ern Aletaialong the north-west low lands of the E’erqisi Val -ley (Xinjiang In sti tute of Ar chae ol ogy, 1981a),Sazicun at Tacheng (Xinjiang In sti tute of Ar chae -ol ogy and Tacheng Cul tural Rel ics Bu reau, 1996) and Adunqiaolu in Boertala Mon go lian Au ton o -mous Re gion in the Tacheng-Kelamayi low lands(Li Jinguo and Lu Engguo, 2003), Xikanerzi inQitai in the north ern Tianshan (Qitai Wenhua-guan, 1982), Shuinichang at Shihezhishi, also inthe west ern part of the north ern Tianshan (Xinjia-ng In sti tute of Ar chae ol ogy and Shihezishi Mu -seum, 1998; Xinjiang In sti tute of Ar chae ol ogy etal., 1999), and the Banjiegou set tle ment sites onthe mid–north ern Tianshan slopes (Xinjiang In sti -tute of Ar chae ol ogy, 1981b). The lo ca tion ofthese sites and their cul tural char ac ter is tics maybe con nected to the three main routes into and outof the Zhunge’er Ba sin: the E’erqisi val ley, theTacheng–Kelamayi low lands and the Hami– Bali- kun cor ri dor. The best stud ied re gion is the south -

ern Hami–Balikun area where more ex ten sivework has been car ried out at Tianshanbeilu andNanwan. One key as pect of the Bronze Age re -mains from Xinjiang is that the great ma jor ity ofthem are de rived from tombs. Very few set tle -ment sites have been ex ca vated, none of them ex -ten sively.

The South-East

Tianshanbeilu and Nanwan23

Tianshanbeilu is a cem e tery lo cated in sideHami City. More than 700 buri als were ex ca vated there from 1988 to 1997 (Lu Enguo et al., 2001).Hand-made, painted red-ware ves sels were placed in the graves. The de signs were mainly ex e cutedin black with some use of a pur plish-red paint.Glob u lar jars with dou ble han dles are among themost com mon forms. There are also taller cy lin -dri cal jars, again with dou ble han dles and painteddec o ra tion (Fig. 7). The cy lin dri cal jars sug gestlinks be tween the east ern Siba cul ture found inGansu and the west ern steppic ce ramic traditions(Shui Tao, 1993).

The Nanwan cem e tery in the Balikun grass -lands is be lieved to be long to the same cul turalcom plex as Tianshanbeilu. Nanwan was ex ca -vated in 1981. As with Tianshanbeilu, a full re -port has not yet been pub lished, al though a num -ber of pre lim i nary re ports have ap peared (ChangEnxi, 1985; He Xin, 1987; Lu Enguo et al., 2001). Car bon dates have been ob tained from both sites,sev en teen from Nanwan and six from Tianshan-beilu (Ta ble 1). Chro no log i cal anal y sis of ma te -rial from both sites has been based mainly on thedates ob tained from the Nanwan cem e tery. Car -bon dates from Tianshanbeilu are re garded as less re li able as there are some anom a lous re sults andthe mean date range is ap par ently in con sis tentwith the ty po logi cal ev i dence. Dates de rived from the Nanwan buri als range from 1685–838 calB.C. The more con sis tent Tianshanbeilu datesrange be tween 1111 and 200 cal B.C. with out li -ers at ear lier than 6000 cal B.C. and as late as 200cal A.D. The Nanwan dates were all ob tainedfrom wood, while those from Tianshanbeilu werefrom hu man bone. Em ploy ing rel a tive chro nol -ogy, ar chae ol o gists have dated the Tianshanbeilucul ture to around the early sec ond mil len niumB.C., based on ty po logi cal com par i son with east -

176 Pe ter Wei Ming Jia et al.

Page 11: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

ern par al lels. These in clude the cy lin dri cal jarsmen tioned above with par al lels in the Siba cul ture of Gansu, while cer tain bronze artefacts such asmir rors can be com pared with ex am ples fromShang buri als in cen tral China (Lu Enguo et al.,2001).

It is con ceiv able that both the Tianshanbeiluand Nawan cem e ter ies were used for a long pe -riod. Based on strati graphic data, Lu Enguo et al.(2001) have cat e go rized the artefacts found inthese two sites into four phases. All four are rep -re sented at Tianshanbeilu, while the Nanwan se -quence has only three, the first one co in cid ingwith the sec ond phase at Tianshanbeilu. On theba sis of ce ramic par al lels, the third phase ofNanwan is equiv a lent to the early phase of theSidaogou cul ture, a sec ond early painted pot terytra di tion found at sites on the north-east ern flanksof the Tianshan range. The painted, dou ble-han -dled, round-bod ied jars from Tianshanbeilu (LuEnguo et al., 2001: 180, fig. 11-3, 4) par al leled inthe early Siba cul ture find their best anal o gies inves sels from the site of Huoshaogou (Shui Tao,2001: 245, fig. 23-M153: 3 or M208: 1) and

Donghuishan (Fig. 8; Jilin Uni ver sity, 1998: fig.53). Shui has sug gested that the early Siba cul turedates to around 1900 B.C. This would equate theear li est Tianshanbeilu phase with the sec ondphase at Huoshaogou, around or ear lier than 1900 B.C. (Shui Tao, 2001: 245).

How ever, still ear lier dates have been pro -posed. Sites of the Tianshanbeilu cul ture oc cur inthe south-east cor ner of the Zhunge’er Ba sin, thenorth-west end of the cor ri dor con nect ing cen tralChina and Eur asia. Han Jianye has noted ap par ent par al lels be tween pot tery from Tianshanbeilu andthat of the Machang cul ture found in Gansu.24

Machang is an Éneolithic cul ture dat ing to around 3800–2000 B.C., con sid er ably ear lier than theSiba cul ture. He has sug gested that the ear li estphase at Tianshanbeilu should be dated to ear lierthan 2000 B.C. (Han Jianye, 2005: 81). Thiswould push back the ear li est ev i dence for con tactbe tween north west China and Xinjiang into thethird mil len nium B.C. Based on cur rent ev i dencethe ear li est dated Bronze Age sites from aroundthe Zhunge’er Ba sin are those in the south eastcor ner, rep re sented by the ear li est phase at

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 177

Fig. 7. Ce ram ics from the early phase of the Tianshanbeilu cul ture (Lu Enguo et al., 2001: fig. 11). Not to scale

Page 12: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Tianshanbeilu, at around 2000 B.C. More pre cisedat ing awaits fur ther ev i dence, in par tic u lar ab so -lute rather than rel a tive dat ing.

Banjiegou and Sidaogou

A sec ond early painted pot tery tra di tion hasbeen found at Banjiegou in Qitai county (Fig. 9)

and Sidaogou (early phase) in Mulei county (Fig.10: 2). These two sites are both lo cated at thesouth ern end of the Zhunge’er Ba sin, on thenorth ern slopes of the Tianshan (Guo Wu, 2005:106–107; Han Jianye, 2005). The ce ram ics fromBanjiegou in clude red ware ves sels with sandy in -clu sions, found to gether with pol ished stone hoes, ham mers, mor tars and pes tles. The ves sels are

178 Pe ter Wei Ming Jia et al.

Ta ble 1Cal i brated C14 dates from Sidaogou, Nawan and Tianshanbeilu (Af ter Han Jianye 2005)

Page 13: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

com monly round-bot tomed with two small han -dles, dec o rated with dark red or pur ple paint onpink or yel low slip. In verted tri an gles and net pat -terns are the most com mon mo tifs (Xinjiang In sti -tute of Ar chae ol ogy, 1981b). Based on ce ramicpar al lels Han Jianye has sug gested that Banjiegou should be placed around 1500 B.C. (Han Jianye,2005). C14 sam ples from Sidaogou date the ear li -est phase to around 1000 cal B.C., while the laterphase has been dated at around 500 cal B.C. (GuoWu, 2005: 108; Xinjiang In sti tute of Ar chae ol -ogy, 1982). Pot tery found in the early phase atSidaogou is sim i lar to that from the third phase ofNanwan (Fig. 10: 1; Lu Enguo et al., 2001: 186,fig. 22), which has been dated to around 1100 calB.C. (Han Jianye, 2005). Pot tery sim i lar to thatfrom Banjiegou was found at Xintala on the

south ern slopes of the Tianshan, dated by C14anal y sis to around 1500 B.C. (Xinjiang In sti tuteof Ar chae ol ogy, 1988). Taken to gether theBanjiegou sites and Sidaogou could span a pe riodfrom around 1500–500 B.C. and may represent alocal development of Bronze Age culture.

The West and South-West

Sazicun and Adunqiaolu cem e ter ies

The sites of Sazicun and Adunqiaolu are lo -cated in the Tacheng – Kelamayi low lands on thewest side of the Zhunge’er Ba sin, near the Kaza-khstan bor der. Sazicun (Xinjiang In sti tute of Ar -chae ol ogy and Tacheng Cul tural Rel ics Bu reau,1996) and Adunqiaolu (Li Jinguo and Lu Engguo, 2003) cem e ter ies con tain no painted pot tery at all.

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 179

Fig. 8. Painted pot tery found at Tianshanbeilu (1, 2) and Siba cul ture (3, 4). Not to scale

Page 14: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

The ves sels are wide mouthed jars with in ciseddec o ra tion that find their best par al lels in theFedorovo, a late phase of the east ern Andronovocul ture of the Eur asian steppes (Figs 11, 12;Xinjiang In sti tute of Ar chae ol ogy and TachengCul tural Rel ics Bu reau, 1996; Li Jinguo and LuEngguo, 2003; Han Jianye, 2005). This Andro-novo sub-group is cen tered around the Aletai andthe up per Yenisei val ley and is dated around1500–800 B.C. (Koryakova, 1996; Kuzmina,1985).25 Some iron ob jects have been dis cov eredin the late Fedorovo phase in Kazakhstan, plac ingthe late phase of the Fedorovo cul ture in the earlyIron Age (Kuzmina, 1985). Qiongkeke, a Bronzeand early Iron Age cem e tery in the Yili val ley,was cut into the top of an Andronovo (Fedorovo)set tle ment site (Ruan Qiurong, 2004; Xinjiang In -sti tute of Ar chae ol ogy, 2002). C14 anal y sis datesthe cem e tery to as early as 1000 B.C. This sug -gests that Sazicun and Adunqiaolu may fall in thetime range of around 1500–1000 B.C., or pos si bly slightly ear lier based on the ev i dence from cal i -brated C14 dates for the Andronovo cul ture.26

Shuinichang

The Shuinichang cem e tery is at Shihezishi inthe south-west cor ner of the Zhunge’er Ba sin.The pot tery ves sels are short and full bod ied withflat bases, made of red or grey ware with sandy in -

clu sions. They are dec o rated with in cised geo met -ric tri an gu lar and net pat tern de signs. The shapeshave been com pared with Karasuk ce ram ics (Guo Wu, 2005: 127; Han Jianye, 2005), but they areshorter and wider than clas sic Karasuk ves sels(Fig. 13). The Karasuk cul ture fol lowed theAndronovo (Fedorovo) in the east ern steppes, theAletai and the up per Yenisei val ley and datesfrom c. 1500–800 cal B.C. (Frachetti, 2004: 192,fig. 5.1). Han Jianye has sug gested that there maybe a lo cal vari ant of the Karasuk cul ture in theZhunge’er Ba sin (Han Jianye, 2005). Based onthe gen eral sim i lar i ties be tween the Shuinichangas sem blage and Karasuk ves sels, the Shuinichang cem e tery has been re garded as con tem po rary with the Karasuk cul ture, around 1500–800 cal B.C.(Guo Wu, 2005), but the lack of di rect par al lelssug gest that this dat ing should be viewed withcaution until more reliable evidence is available.

Xiakalanggu’er

Xiakalanggu’er is a set tle ment site spreadalong the bank of the Kalanggu’er River, atErgongxiang in Tacheng Town ship. The site hasnot been ex ca vated but finds from the sur face in -cluded chipped stone tools, ground stone artefacts and im pressed and in cised pot tery. Sim i lar siteshave been re ported at other lo ca tions along thebanks of the same river (Yu Zhiyong, 1998). The

180 Pe ter Wei Ming Jia et al.

Fig. 9. Ce ram ics from Banjieguo (Guo Wu, 2005: fig. 2-4-4). Not to scale

Page 15: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 181

Fig. 10. Ce ram ics from 1. Late Nanwan (Xinjiang In sti tute of Ar chae ol ogy 1982, figs 7, 8); 2. Early Sidaogou (Lu Enguo et al., 2001: fig. 22). Not to scale

Page 16: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

pot tery from these sites has quite dis tinct char ac -ter is tics which do not find clear par al lels in otherre gional as sem blages. A par tic u lar fea ture is theap pli ca tion of a dec o rated band of clay along therim of some ves sels. While the unique forms ofthe ce ram ics from the site makes it dif fi cult to as -sign a date, the rel a tively sig nif i cant pres ence ofchipped stone artefacts suggests that it may befairly early.

The North-West

Qiemu’erqieke cem e ter ies

Finds from the Qiemu’erqieke cem e ter ies(Xinjiang In sti tute of Ar chae ol ogy, 1981a) areim por tant for the ar chae ol ogy of the Zhunge’erBa sin but are also prob lem atic. The sites aremulti-pe riod with a va ri ety of grave types in clud -ing pit buri als and buri als in stone cists, some sur -rounded by stone en clo sures and oc ca sion ally or -na mented with an thro po mor phic mono liths.Stone an thro po mor phic stat ues are widely con sid -ered to be Turkic, dated around the 6th cen turyA.D., but the ba sic form in Mon go lia and theAletai has its or i gins in older tra di tions (see, forex am ple, Wang Bo and Xiao-Shan Qi, 1995). The ex am ples in the Qiemu’erqieke cem e ter ies lack

the typ i cal cup held in the right hand that char ac -ter izes the later forms. In Eur asia, ear lier stonestat ues in a sim i lar tra di tion are as so ci ated withthe Saka/Scythians and have been found aroundthe Aral Sea and as far west as the shores of theBlack Sea (Olkhovskii and Evdokimov, 1994).27

The pot tery (Fig. 14) and meth ods of burial prac -tice are sim i lar to those of the Afanasievo cul turefound in the Aletai re gion of south ern Si be riaaround 3500–2500 cal B.C. (Frachetti, 2004: 192, fig. 5.1; pl. 198 ff.).28 These sim i lar i ties have ledto spec u la tion that there are con nec tions be tweenthe Afanasievo cul ture, Qiemu’erqieke and also,much less plau si bly, Xiaohe.29

In dis cuss ing the re la tion ship be tween theAfanasievo cul ture and the Qiemu’erqieke cem e -tery, Han Jianye has sug gested that what is gen er -ally rec og nized as “Qiemu’erqieke” is rep re -sented only by the ear li est graves from the sitesand also those from an other site, the ex ten sive set -tle ment of Xikan’erzi in Qitai County at thesouth ern end of the Zhunge’er Ba sin (Han Jianye,2005; Xinjiang Bu reau of Cul tural Rel ics, 1999:298). He ar gued that there are some sim i lar i tiesbe tween the Qiemu’erqieke buri als and theAfanasievo cul ture but some Qiemu’erqieke pot -

182 Pe ter Wei Ming Jia et al.

Fig. 11. Ce ram ics from the Adunqiaolu cem e tery (Li Jinguo and Lu Enguo 2003: figs 6, 7). Not to scale

Page 17: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

tery, spe cif i cally flat-based jars with punc tu ate in -ci sions on the rim, may in di cate that theQiemu’erqieke graves post date the Afanasievocul ture but are ear lier than Karasuk, that is latethird to early sec ond mil len nium B.C. This fitswith other ev i dence as par al lels have also beennoted be tween Qiemu’erqieke tomb types andthose of the Okunevo cul ture found in the Yenisei val ley (c. 2600–2000 cal B.C.) (Chen andHiebert, 1995: 269; Frachetti, 2004: 200).30 Han

Jianye’s dat ing of the Qiemu’erqieke cem e tery isear lier than that of pre vi ous schol ars such as ShuiTao (Shui Tao, 2001) who sug gested that theQiemu’erqieke re mains dated from around 1000B.C. be cause he felt that the Qiemu’erqieke ves -sels found closer par al lels in the Karasuk cul ture(Fig. 15). Re cently avail able C14 dates now datethe Karasuk cul ture slightly ear lier, at be tween1500 and 800 cal B.C. (Frachetti, 2004: 192, fig.5.1).

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 183

Fig. 12. Ce ram ics from Shuinichang cem e tery, Shihezishi (Guo Wu, 2005: fig. 3-1-5). Not to scale

Fig. 13. Sazicun: Fedorovo burial and pot tery (Xinjiang In sti tute of Ar chae ol ogy et al., 1996: figs 7, 8). Pot terynot to scale

Page 18: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

If the ce ram ics from all three groups, Afana-sievo, Qiemu’erqieke and Karasuk, are com paredit can be seen that some ves sels, such as thosewith whole body dec o ra tion found at Qiemu’er-qieke (Fig. 15: 3a, b) have gen eral Afanasievopar al lels, but plain or less dec o rated round orflat-bot tomed jars are closer to Karasuk ex am ples. One jar from the Qiemu’erqieke cem e tery (Fig.16) may be long to the Iron Age. This sug geststhat the Qiemu’erqieke cem e ter ies were in useover a long pe riod of time, and that their cul turalaf fin i ties are var ied. The early graves may be con -nected with the Afanasievo cul ture, but there arealso some par al lels with the first phase of theTianshanbeilu cul ture. The cy lin dri cal jars foundat Tianshanbeilu re flect the dec o rated round-bot -tomed jars found at Qiemu’erqieke (Li Shui-cheng, 2002).31 Some Qiemu’erqieke ves selsmight be late as the Karasuk cul ture and someburi als may be long to the Iron Age or even later.These ques tions can not be re solved with out fur -ther field work. The pot tery sim i lar to the ves selsof Afanasievo af fin ity found at Qiemu’erqiekethat has been re cov ered from Xikan’erzi (Fig. 14:3c) has caused ar chae ol o gists to as sume thatQiemu’erqieke re mains are dis trib uted across theZhunge’er Ba sin (Han Jianye, 2005; Lin Meicun,2002).

The Bronze Age: sum mary

The Bronze Age in the Zhunge’er Ba sin inthe pe riod from around 2000–1000 B.C. is var iedand re flects a va ri ety of dif fer ent in flu ences. Theear li est is the Tianshanbeilu cul ture at the south-east cor ner of the ba sin, which shows con nec tions with the east through the Siba cul ture in Gansu.The ear li est Qiemu’erqieke sites in the north-west cor ner of the ba sin may also date from the early2nd mil len nium B.C. A lit tle later in the south-west, on the north ern slopes of the Tianshan, arethe Banjiegou and early Sidaogou sites, start ing at around 1500 B.C. Dated to about the same pe riodare the Sazicun and Adunqiaolu sites in the westand Shuinichang in the south-west. The last two,to gether with the later Qiemu’erqieke, im ply thepos si ble ex pan sion of steppe cul tures into theZhunge’er Ba sin from Eur asia through the nat u ral low lands to the west and per haps through pat terns of no madic move ment across the Aletai. Pot teryfound at Shuinichang sug gests that ad ap ta tion byincomers to the lo cal en vi ron ment and in ter ac tionwith lo cal groups may have re sulted in new cul -tural vari ants of the steppic Bronze Age inXinjiang.

184 Pe ter Wei Ming Jia et al.

Fig. 14. Ce ram ics from Qiemu’erqieke (Xinjiang In sti tute of Ar chae ol ogy, 1981: figs 3, 4). Not to scale

Page 19: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 185

Fig. 15. Qiemu’erqieke and re lated ce ram ics com pared with Afanasievo and Karasuk ves sels: 1. Afanasievo: a, b(Artamanov, 1974: fig. 20); 2. Karasuk: a, c, e, f, g (Jettmar, 1950: 127, pl. 1: nos. 7–11), b, d (Artamanov, 1974:figs 31, 32); 3. Qiemu’erqieke: a, d, f, g, (Xinjiang In sti tute of Ar chae ol ogy, 1985: figs. 68, 73–75), b (Xinjiang Bu -reau of Cul tural Rel ics et al., 1999: fig. 0950), Xikanerzi: c (Xinjiang Bu reau of Cul tural Rel ics et al., 1999: fig.0811), Alepabulake, Aletai: e (Wang Bo et al., 2005: fig. 1). Not to scale

Page 20: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

THE EARLY IRON AGE

Only a small num ber of early Iron Age siteshave been found in the Zhunge’er Ba sin, pri mar -ily due to a lack of ar chae o log i cal field work.These in clude the Dalongkou cem e tery at Jimusar (Xinjiang Kaosuo, 1997) and Sidaogou (latephase) sites at Mulei (Xinjiang In sti tute of Ar -chae ol ogy, 1982) lo cated on the mid-north ernTianshan slopes, as well as Jijiandui (Yu Zhiyongand Yan Luncang, 1995). An other cem e tery hasbeen dis cov ered at Nashan in Shihezishi in thewest ern Tianshan (Xinjiang In sti tute of Ar chae ol -ogy et al., 1999).

Dalongkou, Sidaogou (late phase) andJijiandui

Han Jianye (2005) has placed Dalongkou,Up per Sidaogou and Jijiandui in the same cul tural tra di tion, but this may need to be re-ex am ined inthe fu ture as more ev i dence be comes avail able.Grave goods from Dalongkou cem e tery in cludedce ram ics, bronze, iron and sil ver ob jects. The pot -tery is made of a red paste with sand in clu sions. A few pieces are painted. The ves sels are short,full-bod ied jugs and dou ble-han dled jars with

round or small flat bases (Fig. 17). They may de -rive out of the Sidaogou (early phase) and Ban-jiegou tra di tion, and pos si bly also Tianshanbeilu(Guo Wu, 2005: 108–114). A C14 date fromSidaogou (late phase) dates this ma te rial toaround 300 B.C. (Han Jianye, 2005).

Shihezishi Nanshan

Nanshan cem e tery (Xinjiang In sti tute of Ar -chae ol ogy, 1999) lo cated in the west ern Tianshan, nat u rally has cul tural con nec tions to the Yili Val -ley (Guo Wu, 2005: 118). The graves con tainedce ramic ves sels, mostly tall jugs with small han -dles. The fab ric of the ves sels is red with sandy in -clu sions, and some are painted with red, mostcom monly net pat terns (Fig. 18). Small ironknives are also found in most buri als. No C14dates are avail able but, based on par al lels fromnearby ar eas, Guo Wu (2005: 118) has sug gesteda date around 400 B.C.

Other sites

Due to the lack of field work, there are no Iron Age sites known from other re gions of theZhunge’er Ba sin, but some Qiemu’erqieke buri als such as No. 4 may be rep re sen ta tive of the IronAge in the Aletai area. Artefacts found in thisgrave in clude one pot, some sherds and an un -iden ti fied iron ob ject (Xinjiang In sti tute of Ar -chae ol ogy, 1981a). The shape of the jar with itsround body and long, thin neck is quite dif fer entfrom the types of ves sels found in the early graves dated around 1500 B.C. or ear lier (Fig. 18). Thejar has no ob vi ous par al lels else where in this pe -riod but com pares well with ce ram ics from muchlater Saka buri als. This would date burial No. 4 toa pe riod around 300 B.C. or even later (MeiJianjun, 2000: 171, figs 32, 33). It is likely thatother sites of this pe riod re main to be found in thenorth of the ba sin.

The Iron Age: sum mary

Dur ing the early Iron Age, around 500 B.C.,ce ram ics re lated to those of the late phase atSidaogou spread along the east and mid dle slopesof the Tianshan. At the same time the Tianshanslopes to the west of Ulumuqi were oc cu pied bypeo ple us ing ves sels typ i cal of those fromNanshan. The late graves of Qiemu’erqieke cem e -

186 Pe ter Wei Ming Jia et al.

Fig. 16. Pot tery ves sel from Qiemu’erqieke (XinjiangIn sti tute of Ar chae ol ogy, 1981: fig. 3.3). Not to scale

Page 21: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

tery such as burial No. 4 may rep re sent the lateIron Age cul ture in the Aletai area. How ever, con -sid er ation of the over all ev i dence raises sev eralkey ques tions. Firstly, dur ing the Bronze Age, theslopes of the Tianshan west of Ulumuqi werepop u lated by peo ple with a ma te rial cul ture typ i -fied by Shuinichan (so-called Lo cal Karasuk).How this was re placed by the Nanshan ma te rialcul ture is un known. Sec ondly, early painted pot -tery and dou ble han dled jars were found at Tian-shanbeilu around 2000 B.C. Later in the BronzeAge this tra di tion spread west wards and north -wards, re plac ing cul tures with steppic af fin i tieslike Qiemu’erqieke, Fedorovo and Karasuk. Howthis oc curred is also un known. Thirdly, there is agap be tween the late Bronze Age / early Iron Agec. 1000 B.C. and the late Iron Age c. 500 B.C.Whether this ap par ent gap re lates to lack of field -work, poor chro no log i cal in ter pre ta tion or an ac -

tual re duc tion in oc cu pa tion is un known. Theseare all ques tions that need to be ad dressed in fu -ture re search.

DIS CUS SION

Key to fur ther prog ress in re con struct ing thelater pre his tory of the Dzhung’er Ba sin is a betterun der stand ing of chro nol ogy, trans mis sion of cul -tural traits and in no va tions, and the move ments of peo ples them selves.

Chro nol ogy

Ta ble 2 sum ma rizes in ba sic form the chro -nol ogy out lined above. Three ma jor re gions havebeen iden ti fied: West Tianshan (the north-west -ern slopes of the Tianshan); East Tianshan (thenorth-east ern slopes of the Tianshan, the Balikun

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 187

Fig. 17. Late Sidaogou ce ram ics from Jijiandui cem e tery (Guo Wu, 2005: fig. 2-4-13). Not to scale

Page 22: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

grass lands and the Hami Ba sin); north ernZhunge’er (the Aletai Moun tains and the E’erqisiVal ley). The first three pe ri ods, be fore 10,000B.P., around 10,000 and 5000 B.P., are very hy -po thet i cal with no strati graphi cal ev i dence at all.Dat ing is based solely on com par i son with sim i lar

dis cov er ies else where in sur round ing re gions.The di ag nos tic el e ments il lus trated are onlybroadly rep re sen ta tive of a much more com plexcul tural re al ity. This sum mary is pre sented sim ply as a start ing point for con struc tion of a more ro -bust chro no log i cal frame work for the re gion.

188 Pe ter Wei Ming Jia et al.

Fig. 18. Ce ram ics from Nanshan cem e tery, Shihezishi (Xinjiang In sti tute of Ar chae ol ogy, 1999: fig. 8). Not toscale

Page 23: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

This hy po thet i cal ta ble im plies cer tain im por -tant pos si bil i ties. Firstly, Tianshanbeilu and earlyQiemu’erqieke might be the ear li est Bronze Agecul tures in the Zhunge’er Ba sin at around 2000B.C., and could be rep re sen ta tive of ei ther in flu -ence or ac tual move ment of peo ples from east and west re spec tively. Qiemu’erqieke is ap par entlyin flu enced by the later pe riod of Afanasievo, andTianshanbeilu con nects to the Siba cul ture to the

east. At the same pe riod, the west ern part of theba sin might still be oc cu pied by lo cal groups withmicrolithic stone tool tech nol ogy of the Meso-lithic/Neo lithic tradition.

Sec ondly, Shuinichang prob a bly does notrep re sent the di rect ex pan sion of Karasuk fromthe far north, but seems more likely to in herit thelate Qiemu’erqieke from within the ba sin it self,pos si bly also in flu enced by the Fedorovo tra di tion

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 189

Ta ble 2Pre lim i nary cul tural chro nol ogy of the Zhunge’er Ba sin

Site namePinyin

Lab o ra tory codeDe ter mi na tion

(5568 B.P.)Cal i brate date

BCContext

Sidaogou

WB77-34 2800±70 1036 896 T2(5)

WB77-35 2510±80 796 432 H4

WB77-30 2400±65 756 399 T4(3)

WB77-32 2360±65 512 391 T41(4)

WB77-33 2320±65 409 379 T1(4)H44

WB77-29 2270±80 403 210 T3(3)

WB77-31 2260±80 401 208 T5(3)H50

Nanwan

WB84-06 3000±70 1389 1130 M083

WB84-25 2770±70 1003 838 M016

WB84-48 2860±65 1151 928 M003

WB84-29 2940±70 1266 1034 M117

WB84-35 2960±65 1306 1063 M003

WB84-26 2960±70 1310 1055 M358

WB84-52 2990±65 1380 1127 M37

WB84-50 2990±65 1380 1127 M084

WB84-24 3000±70 1389 1130 M088

WB84-54 3060±65 1418 1262 M65

WB84-28 3080±70 1428 1266 M091

WB84-51 3110±65 1442 1315 M086

WB84-53 3120±65 1448 1319 M62

WB84-49 3130±65 1500 1323 M48

WB84-27 3310±70 1685 1518 M057

WB81-40 2880±80 1253 932 M6

WB81-41 2870±80 1211 926 M2

WB81-42 2850±75 1157 917 M3

Tianshanbeilu

ZK-2788 3310±89 752 200 89M91

ZK-2789 3310±90 181 -235 89M128

ZK-2790 3310±91 6118 5887 89M198

ZK-2791 3310±92 807 430 89M214

ZK-2792 3310±93 1111 845 89M228

ZK-2794 3310±94 919 803 89M416

Page 24: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

in west Zhunge’er as well. Fedorovo re mains dis -cov ered in west ern Zhunge’er around 1500 B.C.may in di cate the east ward ex pan sion of the LateAndronovo. How ever, this ex pan sion seems notas ex ten sive as the wide dis tri bu tion of Andro-novo bronze artefacts, since the ex pan sionmarked by ce ramic pro duc tion is lim ited to west -ern Zhunge’er. Ac cord ing to ev i dence from theYili River re gion, the Fedorovo was quickly re -placed by the Qiongkeke before 1000 B.C. (HanJianye, 2005).

Fi nally, the Tianshanbeilu cul ture rep re sent -ing the east ern tra di tion played a sig nif i cant rolein de vel op ing lo cal cul tural tra di tions of ce ramicde sign and pro duc tion dur ing the Bronze and Iron Age pe riod. This tra di tion, rep re sented par tic u -larly by painted ves sels and dou ble han dled jars,is readily iden ti fi able in early Iron Age cultures in the Zhunge’er Basin.

Clearly, one of the key is sues is the prob lemof dat ing. Avail able dates range from rel a tive dat -ing on the ba sis of ce ramic par al lels throughuncalibrated C14 dates to cal i brated dates. Thefor mula used in cal i bra tion is not al ways pub -lished, al though this is of lesser sig nif i cance,given the broad timescale and gen eral na ture ofthe chro nol ogy at pres ent. There is a marked dis -crep ancy be tween uncalibrated and cal i brateddates, with the lat ter push ing back the timescaleby sev eral cen tu ries, but there is not such a greatdif fer ence be tween the re sults of the var i ous cal i -bra tion for mu lae used over the past few years. Indeal ing with rel a tive dat ing, most of the ce ram icsde rive from graves. This is ad van ta geous in thatthere is a good sup ply of whole ves sels, but whilethe cem e ter ies may have been in use for ex tendedpe ri ods of time, there is of ten lit tle ev i dence toiden tify the se quence of burial. The re gion wasused pri mar ily by no madic or transhumant pas to -ral ists but there are set tle ment sites in more fa -vored lo ca tions. Ex am i na tion of these sites to pro -vide ce ramic se quences tied to ab so lute dat ing iscrit i cal to fur ther ing our un der stand ing of the ar -chae ol ogy of the re gion. Fi nally, rel a tive dat ing isde pend ent on avail able data, and in and aroundthe Zhunge’er Ba sin lit tle sys tem atic ar chae o log i -cal field work has been car ried out. As moreresults become available, the spread of regionalstyles will become easier to pin down with greater accuracy.

Met al lurgyDe spite the im por tance of ques tions re lat ing

to early met al lurgy in west ern China, al most noan a lyt i cal work was car ried out on met als fromearly ar chae o log i cal con texts in Xinjiang be forethe 1990s. Stud ies of bronze artefacts were lim -ited to ty po logi cal clas si fi ca tion. Al though ty po -logi cal study is use ful and forms the ba sis of chro -no log i cal and re gional frame works, met al lur gi calanal y sis pro vides a greater depth of in ter pre ta tion, par tic u larly in re la tion to re gion al ism. In the pastde cade this sit u a tion has changed mark edly, par -tic u larly due to the work of Mei Jianjun who hasun der taken el e men tal anal y sis of a sub stan tialgroup of cop per and bronze sam ples de rivedmostly from the Zhunge’er Ba sin and sur round ing ar eas (Mei Jianjun, 2000: 37–49). He has also ex -am ined ore, slag and in got sam ples col lected from the Nulasai (Nurasay) cop per min ing site inNileke county, Yili Val ley in west ern Xinjiang(Mei Jianjun, 2000: 50–57).

Mei con cluded that both cop per and tinbronze were in use in Xinjiang by the early 2nd

mil len nium B.C. Ar sen i cal cop per ap peared in the later 2nd mil len nium B.C. Tin-bronze, as in theEur asian steppe, was the main al loy used in boththe Bronze and Iron Age. Use of tin-bronze atTacheng in the west ern Zhunge’er Ba sin sug geststhat the site has con nec tions with the Andronovocom plex in the 2nd mil len nium B.C. This isbacked up by ty po logi cal par al lels among theartefacts them selves. Tin bronze also dom i natedsam ples from Tianshanbeilu, sug gest ing thatsteppic in flu ence spread eastwards at least as faras Hami by quite an early date. Cop per ap pearsmore fre quently in the Zhunge’er Ba sin fromearly in the 1st mil len nium B.C., prob a bly as a re -sult of the open ing up of lo cal cop per mines inand around Ulumuqi (Mei Jianjun, 2000: 48,72–3).

Fol low ing Mei’s study of nine teen sam plesfrom Tianshanbeilu, a team in clud ing staff fromBeijing Sci ence and Tech nol ogy Uni ver sity,Xinjiang In sti tute of Ar chae ol ogy and Hami Cul -tural Rel ics Bu reau (Beijing Sci ence and Tech -nol ogy Uni ver sity, Xinjiang In sti tute of Ar chae -ol ogy and Hami Cul tural Rel ics Bu reau, 2001)ex am ined an ad di tional eighty-nine cop per andbronze sam ples from Tianshanbeilu cem e tery,pro duc ing sim i lar re sults to that of Mei. How ever,

190 Pe ter Wei Ming Jia et al.

Page 25: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

the au thors con cluded that the ev i dence sug gested that the Tianshanbeilu bronze tech nol ogy was notthe re sult of the east ward ex pan sion of theAndronovo cul ture from the Eur asian steppe, butwas de rived from tin bronze-us ing Bronze Agecul tures in Gansu to the east. This con clu sion hasbeen dis puted by Mei Jianjun and Gao Binxiu(Mei Jianjun and Gao Binxiu, 2003) who firmlybe lieve that most bronze artefacts found in Tian-shanbeilu, Hami, and even Gansu are sim i lar tomet als from the Eur asian steppe. They ar gue thatthe con nec tion be tween north-west China and thesteppe must have passed through the “Hexi Cor ri -dor” and Xinjiang. How ever, they con ceded thatthe or i gins of bronze tech nol ogy in north-westChina are still un clear. Fur ther met al lur gi cal ana-lysis is required and the results may vary indifferent periods and areas.

The avail abil ity of nat u ral re sources may beone of the rea sons for dif fer ences in composi-tional anal y sis from dif fer ent ar eas. North ernChina has lim ited ac cess to tin ore de pos its but isrich in lead, which might be one of the rea sons for the use of tin-lead cop per al loys in the earlyBronze Age in cen tral China (Liu Li and ChenXingcan, 2003: 37–44). Chase and Douglas haveex am ined 156 steppic style bronze artefacts col -lected from north ern China, south ern Mon go liaand south ern Si be ria, cov er ing the pe riod fromthe Bronze Age to the Han Dy nasty (Chase andDouglas, 1997: 313, ta ble 2). The re sults of thisanal y sis vary for dif fer ent pe ri ods and ar eas. Tingen er ally oc curs in high per cent ages dur ing theBronze Age in north-west China but the per cent -age falls during the Xiongnu (Hun) period around200 B.C.

Search ing for cop per min ing sites is an im -por tant step in find ing an an swer to the ques tionof met al lur gi cal or i gins. The cop per mines ofNulasai in the west ern Tianshan were used for thepro duc tion of cop per and bronze in the early tomid first mil len nium B.C. (Mei Jianjun, 2001:50–57). The re sults from Mei’s compositionalanal y sis in di cate that the bronze pro duced herewas ar senic cop per al loy rather than tin cop per al -loy, in con trast to the dom i nance of tin bronzeartefacts in the re gion at this time.32 The re sults ofthese two stud ies sug gest that compositional anal -y sis alone may be in suf fi cient to iden tify thesources of metal artefacts found in Xinjiang.

Stud ies of the or i gins of met al lurgy and metal artefacts in Xinjiang are nat u rally re lated to thesame topic in cen tral China. The an swer to theques tion as to whether Chi nese met al lurgy “wasan in dig e nous af fair, or was one that was sparkedby im pe tus from be yond the great wall and/or per -haps much fur ther west” (Linduff, 2000: 1) re -quires re sults from stud ies in neigh bor ing ar eas.West ern China, par tic u larly Gansu and Qinghai,and fur ther west into Xinjiang and the Eur asiansteppe, are crit i cal ar eas for this ques tion. Ar gu -ments about the or i gins of Chi nese met al lurgybased on lim ited sam ples of metal artefacts fromBanpo and Jiangzhai (Linduff, 2000) are sim i larto the dis cus sion of the or i gins of met al lurgy inXinjiang. All these ar gu ments are based on thesame re sults of compositional anal y sis. Con ven -tional compositional anal y sis pro vides im por tantin for ma tion, mak ing it pos si ble to dif fer en ti atetypes of met als, and add ing an ex tra di men sion toty po logi cal stud ies of metal artefacts. How ever,so far, it has proved in suf fi cient in an swer ing thelarger ques tions re lat ing to the de vel op ment ofmet al lurgy in China. One path for ward might bethe use of lead iso tope anal y sis which can iden tify the orig i nal ore source for in di vid ual artefacts.Each ore source pres ents a unique lead iso topesig na ture which is not al tered by smelt ing or cast -ing pro cesses (Meyers, 1988: 293). This tech -nique re quires test ing of sources, which are quiteplen ti ful in the steppe, but it of fers a more pre cisemeth od ol ogy for iden ti fy ing the prov e nance ofXinjiang bronzes.33

In sum mary, cur rent re search into early met -als and metal pro duc tion in China is still in a pre -lim i nary stage. The data avail able so far in di catechanges in com po si tion both re gion ally and chro -no log i cally, but, ac cord ing to cur rent re search,Bronze Age metal artefacts found in and im me di -ately around the Zhunge’er Ba sin are dom i natedby tin-bronzes. Tin bronze was widely usedacross the Eur asian steppe and its dom i nance inthe Zhunge’er Ba sin sug gests that the tech nol ogyspread into north ern Xinjiang from the west. Theadop tion of bronze tech nol ogy may have in volved some lo cal mod i fi ca tions within Xinjiang, per -haps due to vari abil ity in the locally availablenatural resources.

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 191

Page 26: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Mod els for the or i gins of Xinjiang pop u la tions

It is clear from an ex tremely wide va ri ety ofev i dence in clud ing met als, fab rics, bo tan i cal andfau nal data, burial cus toms, phys i cal an thro pol -ogy and lin guis tic anal y sis, that the early pop u la -tions of Xinjiang mi grated into west ern Chinafrom out side and/or re ceived a large amount ofcul tural in flu ence from be yond the re gion. A cer -tain amount came from con tact with the east, butby far the great est im pact came from the west;from the Aletai, the Eur asian steppes and thesouth Cen tral Asian oases. While there are manyhy poth e ses con cern ing the pop u lat ing of Xin-jiang, they gen er ally fall into two spe cific groups,those con cern ing move ments from the steppe andthose con cern ing in flu ence from the south ernoases.34 The ‘Steppe Hy poth e sis’ sug gests amove ment of Afanasievo-re lated steppic peo plessouth wards into Xinjiang, pos si bly in the early 2nd

mil len nium B.C.35 Later, around the late 2nd mil -len nium B.C., in no va tions in ma te rial cul turehave been in ter preted as sig nal ing a sec ond waveof mi gra tions, this time from the Eur asian steppe,con cern ing peo ples re lated to the Andronovo cul -ture. Pro po nents of the ‘Bactrian Oasis Hy poth e -sis’ note the sim i lar ity in en vi ron men tal con di -tions be tween the oases of the Talimu Ba sin andthose of Bactria and Margiania, and sug gest thatir ri ga tion ag ri cul ture, trade and the west ern ‘Neo -lithic pack age’ of do mes ti cated sheep/goat andwheat cul ti va tion may have been im portedthrough con tact with Bactrian pop u la tions, al -though there is no ev i dence of BMAC set tle ments in Xinjiang.36

The ma te rial cul ture may in di cate ei ther thespread of ideas and trade net works or whole salemi gra tions of peo ples. This ques tion might bebest re solved by phys i cal an thro pol o gists. Acraniometric study by Hemphill and Mallory(2004) has come up with data that of fers a quitenew pic ture of the pop u lat ing of Xinjiang in thepre his toric pe ri ods. They find lit tle sup port formi gra tion from ei ther the steppe or the Bactrianoases. It should be noted, how ever, that their sam -ples were taken only from Talimu Ba sin sites.Their find ings sug gest the pres ence of a pop u la -tion of un known or i gin in the Talimu Ba sin dur -ing the early Bronze Age. By the late 1st mil len -nium B.C., there was an in flux or ge netic

min gling with high land groups from the Pamierand Ferghana re gions. This should prompt ar -chae ol o gists to pay greater at ten tion to un der -stand ing the elu sive Neo lithic of Xinjiang. It may, how ever, not hold true for the Zhunge’er Ba sinwhich has closer con nec tions to the Eur asiansteppe and whose land scape is a more nat u ral ex -ten sion of the en vi ron ment fa mil iar to Afanasievo and Andronovo pop u la tions. It is fea si ble to sug -gest that there was some mi gra tion into thenorthern half of Xinjiang, allowing a spread ofideas and artefacts into the south.

SUM MARY

Ar chae o log i cal re search in west ern China has pro gressed a long way in re cent years, but our un -der stand ing of the cul tural and chro no log i calframe work for the Zhung’er Ba sin in pre his torictimes is still based largely on iso lated re gional ce -ramic as sem blages, ob tained pri mar ily from mor -tu ary con texts, and dated only through broad andsome times gen er al ized ty po logi cal com par i sons.In for ma tion about the Neo lithic is re mark able byits ab sence, al though the ex ist ing frag men tary ev -i dence for a prob a bly semi-no madic pop u la tion in this pe riod and the pres ence of pock ets of en vi -ron men tally at trac tive land scapes sug gest thatmore sites may be found if re search is spe cif i callydi rected to wards this prob lem. A key ques tion inre gard to the wider re gion is that of the in tro duc -tion of cul ti vated plants and do mes ti cated an i mals from both east and west. The West Asian “Neo -lithic pack age’ may have been un packed in itsjour ney eastwards. It is pos si ble that some itemsmay have been in tro duced into the Talimu Ba sinfrom Bronze Age Bactria, per haps via theFerghana val ley. How ever, it is im por tant to ad -dress this ques tion with re gard to the Zhung’erBa sin also. Here the an swers may be sought insites in the far west, per haps most likely aroundTacheng. Anal y sis of the ex ist ing ce ramic re cordshows that, over all, ex ter nal cul tural in flu enceson the Zhunge’er ba sin in the Bronze Age are asmight be ex pected. In the west, links with theAndronovo of the Eur asian steppes is ap par entwhile in the north, links with the Aletai, Yeniseiand south ern Si be ria can be iden ti fied. In thesouth-east the ap pear ance of painted pot tery andpar al lels with Gansu show east erly con nec tions.

192 Pe ter Wei Ming Jia et al.

Page 27: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

How ever, within this pic ture there are sub tle tieswhich are hard to de tect on the lim ited ev i denceavail able. Move ments of ideas and cul tural in flu -ences ap pear to fol low the rim of the ba sin, re -spect ing the dry in te rior of the hos tile Kuerban-tonggute desert. Yet Kazakh no mads in theAletai37 re port that they oc ca sion ally take theirflocks across the east ern part of the desert to graze in the Balikun grass lands when graz ing is poor inthe north. Balikun is also reg u larly vis ited by no -mads from Hami. There is mod ern transhumancefrom both the north ern Tianshan and the Yili Val -ley into the high pas tures of the west ern Tianshan. A more de tailed un der stand ing of the spread ofideas into and out of the Zhunge’er Ba sin must bebased at least in part on the practicalities of land -scape and eco nomic need.

An swers to the vexed ques tion of the or i ginsof the peo ples of the Zhung’er Ba sin must takeinto ac count the like li hood of a small but sig nif i -cant in dig e nous pop u la tion from ear lier pre his -toric times, whose ge netic stock would haveblended with that of incomers. All re search in thisarea to date has in volved study of Bronze Age and later pop u la tions. These in di vid u als most prob a -bly do not rep re sent first gen er a tion mi grants butrather peo ples who have al ready blended into thepre-ex ist ing ge netic mix of the re gion. To help inun tan gling this com plex is sue, fur ther re search isneeded into the Pre-Bronze Age pe ri ods. It islikely that sites, and prob a bly buri als, of the Neo -lithic and Eneolithic can be identified throughtargeted survey programs.

An other as pect of the pre his tory of theZhunge’er Ba sin is also re flected in mod ern sub -sis tence strat e gies. The pres ent non-in dus trial eco -nomy is based on a mix ture of no madic pasto-ralism and ag ri cul ture, mainly ir ri ga tion, but alsodry farm ing. The dis tri bu tion of no madic pop u la -tions is best doc u mented through ex ca va tion ofnu mer ous cem e ter ies, but de spite the like li hoodof set tle ment sites, very lit tle is known about thisother part of the eco nomic and cul tural sys tem.Some Bronze Age set tle ments have been foundalong river banks in the west, and some have been lo cated in the south-east ern oases. These sites rep -re sent the re mains of small vil lages and aremarked only by sur face scat ters of sherds andgrind ing stones. A fo cus on res cue ex ca va tion ofthreat ened cem e ter ies and a lack of prob lem based

ar chae o log i cal re search in the Zhunge’er ba sinmeans that al most none of these sites have beenex am ined, yet they have the po ten tial to pro videthe best and most cost ef fec tive method of es tab -lish ing re li able chro no log i cal sequences,environmental data, faunal and botanical remainsand detailed ceramic sequences.

REF ER ENCESABDURRASSUL Y., ZHANG CHUAN 1997. Stone

ar ti facts found in the E’erqisi Val ley and their ty po -logi cal study. In: Wang Binghua and Du Gengcheng (eds.) New achieve ments in Xinjiang ar chae ol ogy(1990-1996). Xinjiang Fine Arts and Pho tog ra phyPress, Ulumuqi, 91–101 (in Chi nese).

ABDURRASSUL I., ZHANG CHUAN HASYET1998. Stone ar ti facts found in the E’erqisi Val ley.Xinjiang Wenwu 1998(3), 45–53 (in Chi nese).

ALLCHIN, B. 1992. Mid dle Paleolithic Cul ture. In: A.H. Dani and V. M. Masson (eds.) His tory of Civ i li -za tions of Cen tral Asia. Vol. 1: The dawn of civ i li za -tion – ear li est times to 700 BC. UNESCO, Paris,65–88.

AN ZHIMIN 1987. The Neo lithic in West China.Kaogu Xuebao 2, 133–151 (in Chi nese).

AN ZHIMIN 1992a. The Bronze Age in the east partsof Cen tral Asia. In: A. H. Dani and V. M. Masson(eds.) His tory of Civ i li za tions of Cen tral Asia. Vol.1: The dawn of civ i li za tion – ear li est times to 700BC. UNESCO, Paris, 319–336.

AN ZHIMIN 1992b. Neo lithic com mu ni ties in east ernparts of Cen tral Asia. In: A. H. Dani and V. M.Masson (eds.) His tory of Civ i li za tions of Cen tralAsia. Vol. 1: The dawn of civ i li za tion – ear li esttimes to 700 BC. UNESCO, Paris, 153–168.

AN ZHIMIN 1996, The Bronze Age cul tures aroundthe Talimu Ba sin. Kaogu 12, 70–76 (in Chi nese).

AN ZHIMIN 1998. Cul tural com plexes of Bronze Agein the Tarim ba sin. In: V. H. Mair (ed.) The BronzeAge and early Iron Age peo ples of East ern Cen tralAsia. Vol. I. Uni ver sity of Penn syl va nia Mu seumPub li ca tions, Phil a del phia, 45–60.

ARTAMONOV M. (ed.) 1974. The Dawn of Art:Palaeo lithic, Neo lithic, Bronze Age and Iron Agere mains found in the ter ri tory of the So viet Un ion.Au rora, Le nin grad.

AT LAS 1998. At las of Chi nese Nat u ral Ge og ra phy(2nd edi tion). Chi nese Map Press, Beijing (in Chi -nese).

BEIJING SCI ENCE AND TECH NOL OGY UNI VER -SITY, XINJIANG IN STI TUTE OF AR CHAE OL -OGY AND HAMI CUL TURAL REL ICS BU -REAU 2001. Pre lim i nary study of the bronze ar ti -

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 193

Page 28: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

facts un earthed from Hami Tianshanbeilu, Xinjiang. Wenwu 2001(6), 79–88 (in Chi nese).

BRANTINGHAM P. J., OLSEN J. W., SCHALLERG. B. 2001. Lithic as sem blages from the ChangTang re gion, North ern Ti bet. An tiq uity 75(288),319–327.

CHANG ENXI 1985. Sum mary of the ex ca va tion ofburial 66 in Balikun Nanwan. Xinjiang Wenwu1985(1), 112–114 (in Chi nese).

CHASE W. T., DOUGLAS J. G. 1997. Tech ni cal stud -ies and metal compositional anal y ses of Bronze ofthe East ern Eur asian steppe from the Ar thur M.Sacker col lec tion. In: E. C. Bunker, An cient Bronzes of the East ern Eur asian Steppes: from the Ar thur M. Sackler Col lec tions. Harry N. Abrams, New York,306–312.

CHEN KWANG-TZUU, HIEBERT F. T. 1995. TheLate Pre his tory of Xinjiang in Re la tion to its Neigh -bours. Jour nal of World Pre his tory 9/2, 243–300.

CHEN XINGCAN 1997a. His tory of Chi nese Pre his -toric Ar chae ol ogy(1895 – 1949). Sanlian Shudian,Beijing (in Chi nese).

CHEN XINGCAN 1997b. The De vel op ment of Pre his -toric Chi nese Ar chae ol ogy. Shixue lilun yanjiu 4,54–60 (in Chi nese).

CRAWFORD G., UNDERHILL A., ZHIJUN ZHAO,GYOUNG-AH LEE, FEINMAN G., NICH O LASL., FENGSHUI LUAN, HAIGUANG YU, HUIFANG, FENGSHU CAI 2005. Late Neo lithic PlantRe mains from North ern China: Pre lim i nary re sultsfrom Liangchenzhen, Shandong. Cur rent An thro -pol ogy 46(2), 309–317.

DA VIS R. S., RANOV V. A. 1999. Re cent work on the Palaeo lithic of Cen tral Asia. Evo lu tion ary anthro-pology 8(5), 186–193.

DEBAINE-FRANCFORT C. 1988. Archéologie duXinjiang des Or i gins aux Han. Paléorient 14(1),5–29.

DEBAINE-FRANCFORT C. 1989. Archéologie duXinjiang des Or i gins aux Han. Paléorient 15(1),183–213.

DEREV’ANKO A. P, LÜ ZUN-E 1992. Up perPaleolithic Cul tures. In: A. H. Dani and V. M.Masson (eds.) His tory of Civ i li za tions of Cen tralAsia. Vol. 1: The dawn of civ i li za tion – ear li esttimes to 700 BC. UNESCO, Paris, 153–168.

DEREV’ANKO A. P. 1998. Hu man oc cu pa tion ofnearby re gions and the role of pop u la tion move -ments in the Palaeo lithic of Si be ria. In: A. P. Dere-v’anko, D. B. Shimkin, and W. R. Pow ers (eds.) The Palaeo lithic of Si be ria, new dis cov er ies and in ter -pre ta tions. Uni ver sity of Il li nois Press, Chi cago,336–351.

DEREV’ANKO A. P., OLSEN J. W., TSEVEENDORJ D., KRIVOSHAPKIN A. I., PETRIN V. T.,

BRANTINGHAM P. J. 2000. The strat i fied cavesite of Tsagaan Agui in the Gobi Altai (Mon go lia).Ar chae ol ogy, Eth nol ogy and An thro pol ogy of Eur -asia 1(1), 23–35.

DEREV’ANKO A. P., PETRIN V. T., TAIMAGA-MBETOV J. K. 2000. The phe nom e non of mi -cro-in dus trial com plexes in Eur asia. Ar chae ol ogy,Eth nol ogy and An thro pol ogy of Eur asia 4 (4), 2–16.

FRACHETTI M. D. 2004. Bronze Age Pas to ral Land -scapes of Eur asia and the Na ture of So cial In ter ac -tion in the Moun tain Steppe Zone of East ernKazakhstan. Un pub lished Ph.D. dis ser ta tion. Uni -ver sity of Penn syl va nia.

GANSU IN STI TUTE OF AR CHAE OL OGY ANDJILIN UNI VER SITY 1998. The Ar chae ol ogy ofDonghuishan in Minle County. Re search and In ter -pre ta tion of the Siba Cul ture Cem e tery. Sci encePress, Beijing (in Chi nese).

GUO WU 2005. Ar chae o log i cal stud ies of the Tian-shan re gion of Xinjiang around the first mil len niumBC. Ph.D. dis ser ta tion. Chi nese Acad emy of So cialSci ence, Beijing (in Chi nese).

HAN JIANYE 2005. Chro no log i cal and ty po logi calanal y sis of Xinjiang Bronze and early Iron Age.Xinjiang Wenwu, 2005(3), 57–91 (in Chi nese).

HEDIN S. 1899. Through Asia. Trans lated by J. T.Bealby. Harper, New York/Lon don.

HEDIN S. 1903. Cen tral Asia and Ti bet: To wards theHoly City of Lassa. 2 v. Trans lated by J. T. Bealby.Hurst and Black ett, Lon don.

HEDIN S. 1931. Across the Gobi Desert. Trans lated by H. J. Cant. Routledge, Lon don.

HEDIN S. 1933. Rid dles of the Gobi Desert. Trans lated by E. Sprigge and C. Napier, E. P. Dutton, NewYork.

HEDIN S. 1940. The Wan der ing Lake. Trans lated by F. H. Lyon. E. P. Dutton, New York.

HEDIN S. 1943. His tory of the Ex pe di tion in Asia1927-1935. Stock holm.

HEMPHILL B. E., MALLORY J. P. 2004. Horse-mounted in vad ers from the Russo-Kazakh steppe orag ri cul tural col o nists from West ern Cen tral Asia? A craniometric in ves ti ga tion of the Bronze Age set tle -ment of Xinjiang. Amer i can Jour nal of Phys i cal An -thro pol ogy 124(3), 199–222.

HE XIN 1987. Burial 95 in Balikun Nanwan, Xinjiang.Kaogu yu wenwu 1987(5), 7–8 (in Chi nese).

JETTMAR K. 1950. The Karasuk cul ture and itssouth-east ern af fin i ties. Bul le tin of the Mu seum ofFar East ern An tiq ui ties 22. Stock holm, 90–134.

JIN, ZHENGYAO, CHASE, W. T., YOSHIMITSU,HIRAO, HISAO, MABUCHI, YANG XIZHANG,KAROKU MIWA 1998. Re la tion ship be tweenBronze Civ i li za tions in Two River Val leys ofChina. In In sti tute of Ar chae ol ogy, Chi nese Acad -

194 Pe ter Wei Ming Jia et al.

Page 29: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

emy of So cial Sci ence Pro ceed ings of Shang Cul -ture In ter na tional Con fer ence. Chi nese En cy clo pe -dia Press, Beijing, 425–433 (in Chi nese).

KORYAKOVA L. 1996. So cial trends in tem per ateEur asia dur ing the sec ond and first mil len nia BC.Jour nal of Eu ro pean Aarchaeology 4, 243–280.

KORYAKOVA L., EPIMAKHOV A. V. 2007. TheUrals and West ern Si be ria in the Bronze and IronAges. Cam bridge Uni ver sity Press, Cam bridge.

KRAUSE J., OR LANDO L., SERRE D., VI OLA B.,PRÜFER K., RICH ARDS M., HUBLIN J.-J.,HÄNNI C., DEREVIANKO A. P., PÄÄBO S.2007. Neanderthals in Cen tral Asia and Si be ria. Na -ture 499, 902–904.

KUNGUROVA N. Y. 2003. Ar chae o log i cal ma te ri alsfrom Solontsy-5: Eth nic-cul tural re la tion shipswithin the Kuznetsk-Altai cul ture. Ar chae ol ogy,Eth nol ogy and An thro pol ogy of Eur asia 2(14), 30–39.

KUZMINA E. 1985. Clas si fi ca tion and periodisationof Andronovo cul tural com mu nity sites. In for ma -tion Bul le tin 9, 23–43.

LI JINGUO AND LU ENGGUO 2003. The sur vey atAdunqiaolu site in Wenquan county. XinjiangWenwu 2003(1), 20–26 (in Chi nese).

LI PAN 1989. New Dis cov er ies in Neo lithic site withag ri cul tural re mains at Donghuishan site, MinleCounty, Gansu. Ag ri cul tural Ar chae ol ogy 1989 (1),56–69.

LI SHUICHENG 2002. The in ter ac tion be tween NorthWest China and Cen tral Asia dur ing the sec ond mil -len nium BC: an ar chae o log i cal per spec tive. In: K.Boly, C. Refrew and M. Le vine (eds.) An cient In ter -ac tions: East and West in Eur asia. Mc Don ald In sti -tute for Ar chae o log i cal Re search. Cam bridge Uni -ver sity Press, Cam bridge, 171–181.

LI SHUICHENG 2004. Con sid er ing the dat ing ofwheat at the site of Donghuishan. Ar chae ol ogy andCul tural Rel ics 2004(6), 51–60.

LI XIAOQIANG, JOHN DODSON, XINYINGZHOU, HONGBIN ZHANG, RYO MASUTO-MOTO 2007. Early cul ti vated wheat and broad en -ing of ag ri cul ture in Neo lithic China. The Ho lo cene17(5), 555–560.

LIN MEICUN 2002. The or i gins of Tochar ian and itsmi gra tion. Xinjiang Wenwu 2002(3/4), 69–80 (inChi nese).

LINDUFF K. M. 2000. In tro duc tion: Met al lur gists inan cient East Asia: the Chi nese and who else. In: K.M. Linduff, Han Ru bin, Sun Shuyun (eds.) The Be -gin nings of Met al lurgy in China. The Edwin Mellen Press, New York, 1–22.

LIU LI AND CHEN XINGCAN 2003. State for ma tionin Early China. Duckworth, Lon don.

LU ENGUO 1995. Sev eral is sues in Xinjiang ar chae ol -

ogy. Xinjiang Wenwu 1995(2), 70–77 (in Chi nese).LU ENGUO, CHANG XIEN, WANG BINGHUA

2001. A pre lim i nary anal y sis of Bronze Age ar chae -ol ogy in Xinjiang. In: Su Bingqi and con tem po raryChi nese ar chae ol ogy. Sci ence Press, Beijing, 179–184 (in Chi nese).

MAIR V. (ed.) 1998. The Bronze Age Peo ples of East -ern Cen tral Asia. Uni ver sity of Penn syl va nia Mu -seum Pub li ca tions, Phil a del phia.

MARKIN, S. V. 2000. The Neo lithic burial of theNorth-West ern Altai. Ar chae ol ogy, Eth nol ogy &An thro pol ogy of Eur asia 2(2), 53–64.

MEI JIANJUN, SHELL, C. 1999. The ex is tence ofAndronovo cul tural in flu ence in Xinjiang dur ing the 2nd mil len nium BC. An tiq uity 73, 570–578.

MEI JIANJUN 2000. Cop per and Bronze Met al lurgyin Late Pre his toric Xinjiang: Its Cul tural Con textand Re la tion ship with Neigh bor ing Re gions. BARIn ter na tional Se ries 865. Archaeopress, Ox ford.

MEI JIANJUN 2002. The Iron Age cul tures in Xin-jiang and their steppe con nec tions. In: K. Boly, C.Refrew and M. Le vine (eds.) An cient in ter ac tions:east and west in Eur asia. Mac Don ald In sti tute forAr chae o log i cal Re search. Uni ver sity of Cam bridgePress, Cam bridge, 213–231.

MEI JIANJUN 2003. Cul tural in ter ac tion be tweenChina and Cen tral Asia dur ing the Bronze Age. Pro -ceed ings of the Brit ish Acad emy 121, 1–39.

MEI JIANJUN, GAO BINXIU 2003. Seima-Brodinaphe nom ena and the early Bronze Age in north westChina, dis cus sion of Marija Gimbutas’ Seima-Brodina phe nom ena and the or i gins of an i mal de -sign in Si be ria. Xinjiang Wenwu, 2003(1), 47–56 (in Chi nese).

MEYERS P. 1988. Char ac ter is tics of cast ing re vealedby the study of an cient Chi nese bronzes. In: R.Maddin (ed.) The be gin ning of the use of met als andal loys. Pa pers from the sec ond in ter na tional con fer -ence on the be gin ning of the use of met als and al -loys, Zhengzhou, China, 21-26 Oc to ber 1986. MITPress, Lon don, 288–295.

MU SHUNYING 1992. Com ments on the painted pot -tery in Xinjiang. In: Jarrige, C. (ed.) South Asian Ar -chae ol ogy 1989. Mono graphs in World Ar chae ol -ogy 14, Pre his tory Press, Mad i son, WI, 201–212.

OKLADNIKOV A. P. 1978. The Paleolithic of Mon -go lia. In: Ikawa-Smith (ed.) Early Paleolithic inSouth and East Asia. Mou ton, the Hague, 317–25.

OLSEN, J. 2004. The Paleolithic Ar chae ol ogy ofXinjiang, Mon go lia, and North ern Ti bet. 2004 Field Re port. Avail able at: http: //www.ic.ar i zona.edu/~mon go lia/Doc u ments/re port_2004.pdf .

QITAI WENHUAGUAN 1982. The lithic site andcem e tery found in Qitai, Xinjiang. Kaoguxue jikan1982(2), 22–24 (in Chi nese).

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 195

Page 30: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

RANOV V. A., DA VIS R. S. 1979. To wards a NewOut line of the So viet Cen tral Asian Paleolithic. Cur -rent An thro pol ogy 20(2), 249–270.

RANOV V. A., DORJ D., LÜ ZUN-E 1992. LowerPaleolithic Cul tures. In: Dani, A. H. and Masson, V. M. (eds.) His tory of Civ i li za tions of Cen tral Asia,Vol. 1: The dawn of civ i li za tion – ear li est times to700 BC. UNESCO, Paris, 45–64.

RUAN QIURONG 2004. The sig nif i cance of ex ca va -tion in the Jilintai site of Nileke, Xinjiang. XinjiangWenwu 2004(1), 80–82 (in Chi nese).

SHISHLINA N. I., HIEBERT F. T. 1998. The Steppeand the Sown: In ter ac tion be tween Bronze Age Eur -asian No mads and Ag ri cul tur al ists. In: Mair, V. H.(ed.) The Bronze Age and early Iron Age peo ples ofEast ern Cen tral Asia. Vol. I. Uni ver sity of Penn syl -va nia Mu seum Pub li ca tions, Phil a del phia, 222–237.

SHUI TAO 1993. Com par i son of Bronze Age cul turesin Xinjiang and the his toric prog ress in early cul -tural in ter ac tion be tween East and West. Guoxueyanjiu 1993(1), 447–490 (in Chi nese).

SHUI TAO 2001. The com par i son be tween BronzeAge cul tures found in Xinjiang: early prog ress ofcul tural in ter ac tion be tween China and West. In:Shui, Tao, Pa pers on the Bronze Age Ar chae ol ogyof North-West China. Sci ence Press, Beijing, 6–42.

SOKOLOVA L. A. 2007. Okunev Cul tural Tra di tion in the Strati graphic As pect. Ar chae ol ogy, Eth nog ra -phy and An thro pol ogy of Eur asia 30(2), 41–51.

STEIN A. M. 1907. An cient Khotan: De tailed re port of ar chae o log i cal ex plo ra tions in Chi nese Turkestan.Clar en don Press, Ox ford.

STEIN A. M. 1912. Ru ins of Desert Cathay: Per sonalNar ra tive of Ex plo ra tions in Cen tral Asia and Wes -tern most China. Macmillan & Co, Lon don.

STEIN A. M. 1921. Serindia: De tailed re port of ex plo -ra tions in Cen tral Asia and wes tern most China.Clar en don Press, Lon don & Ox ford.

STEIN A. M. 1928. In ner most Asia: De tailed Re port of Ex plo ra tions in Cen tral Asia, Kan-su and East ernIran. Clar en don Press, Ox ford.

TELEGIN D. YA., MALLORY J. P. 1994. The An thro -po mor phic Stelae of the Ukraine: the early ico nog -ra phy of the Indo-Eu ro pe ans. Jour nal of Indo-Eu ro -pean Stud ies Mono graph No. 11. In sti tute for theStudy of Man, Wash ing ton D.C.

VADETSKAYA E. B. 1986. Ar chae o log i cal mon u -ments in the steppes of the Mid dle Yenisei River ba -sin. Le nin grad (in Rus sian).

VON LE COQ A. 1928. Bur ied Trea sures of Chi neseTurkestan: An Ac count of the Ac tiv i ties and Ad ven -tures of the Sec ond and Third Ger man Turfan Ex pe -di tions. Trans lated by A. Barwell. George Allen &Unwin, Lon don.

WANG BO AND XIAO-SHAN QI 1995. Re search on

the Grass land Stone Fig ures along the Silk Road.Ulumuqi (in Chi nese).

WANG BO, TAN DAHAI, CHI WENJIE 1997. Twomicrolithic sites found in the east Junggar. In: Wang Binghua and Du Gengcheng (eds.) New achieve -ments in Xinjiang ar chae ol ogy (1990-1996). Xin-jiang Fine Arts and Pho tog ra phy Press, Ulumuqi,63–70 (in Chi nese).

WANG BO, WU DONGJUN, ZHENG JIE 2005. Eth -no log i cal anal y sis of skulls found in Altai buri als.Xinjiang Wenwu 2005(4), 38–41 (in Chi nese).

WEBER A. 1995. The Neo lithic and early Bronze Ageof the Lake Baikal re gion: a re view of re cent re -search. Jour nal of World Pre his tory 9(1), 99–158.

XINJIANG IN STI TUTE OF AR CHAE OL OGY1981a. Sum mary of ex ca va tion at Qiemu’erqiekecem e tery. Wenwu 1981(1), 176–186 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY1981b. Qitai Banjiegou Neo lithic site, Xinjiang.Kaogu 1981(6), 552–3 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 1982.Mulei Sidaogou site, Xinjiang. Kaogu 1982(2),113–120 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 1988.Sum mary of ex ca va tion at Heshuo Xintala site, Xin- jiang. Kaogu 1998(5), 399–407 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 1989.Sur vey re port of microlithic sites along the bank ofChaiwuopu Lake. Koagu yu Wenwu 1989(2), 12–19(in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY ANDSHIHEZISHI MU SEUM 1994. Shihezishi cem e -tery. Xinjiang Wenwu 1994(4), 12–19 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 1995.Sur vey and dis cus sion on the stone artefacts foundin Yerkabake, Mulei County. Xinjiang Wenwu1995 (1). Also in: Wang Binghua and Du Geng-cheng (eds.) New achieve ments in Xinjiang ar chae -ol ogy (1990- 1996). Xinjiang Fine Arts and Pho tog -ra phy Press, Ulumuqi, 71–80 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY ANDTACHENG BU REAU OF CUL TURAL REL ICS1996. Sazicun cem e tery of Tuolixian. XinjiangWenwu 1996(2), 14–22 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY ANDKELAMAYI BU REAU OF CUL TURAL REL ICS1997. Stone ar ti facts found at Kalamayi. In: WangBinghua and Du Gengcheng (eds.) New achieve -ments in Xinjiang ar chae ol ogy (1990-1996). Xin-jiang Fine Arts and Pho tog ra phy Press, Ulumuqi,81–90 (in Chi nese).

XIJIANG IN STI TUTE OF AR CHAE OL OGY 1997.Jimusar Dalongkou cem e tery, Xinjiang. Kaogu1997(9), 39–45 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY AND

196 Pe ter Wei Ming Jia et al.

Page 31: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

SHIHEZISHI MU SEUM 1998. Ar chae o log i calfield sur vey at Shihezishi. Xinjiang Wenwu 1998(4), 54–64 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY, SHI-HEZI MU SEUM AND XINJIANG UNI VER SITY1999. Nanshan cem e tery of Shihezi, Xinjiang.Kaogu, 1999(5), 38–46 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 1999.Xinjiang Chawuhu: ex ca va tion re port of large clancem e ter ies. Dongfang Press, Beijing (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 2002.Ex ca va tion re port at Qiongkeke cem e tery I, Nileke.Xinjiang Wenwu 2002(3/4), 13–53 (in Chi nese).

XINJIANG IN STI TUTE OF AR CHAE OL OGY 2003.Ex ca va tion re port of 2002 in Xiaohe cem e ter ies.Xinjiang Wenwu 2003(2), 8–46 (in Chi nese).

XINJIANG BU REAU OF CUL TURAL REL ICS,XINJIANG IN STI TUTE OF AR CHAE OL OGY,XINJIANG MU SEUM AND XIJIANG XINTIAN1999. A Grand View of Xinjiang’s Cul tural Rel icsand His toric Sites. Xinjiang Fine Arts Press,Ulumuqi (in Chi nese).

YI SEONBOK, CLARKE G. A. 1983. Ob ser va tions on the Lower Paleolithic of North east Asia. Cur rentAn thro pol ogy 24(2), 181–202.

YU ZHIYONG 1998. The Xiakalangguer site atErgongxiang, Tachen. Xinjiang Wenwu 1998(2),35–38 (in Chi nese).

YU ZHIYONG AND YAN LUNCANG 1995. Fieldsur vey at the Jijiandui site of Fubei Nongchang,Fukangshi. Xinjiang Wenwu 1995(1), 11–20 (inChi nese).

ZHANG PING 1999. The Bronze Age of Qiuci, ob -serv ing from Kezier site and buri als. XinjiangWenwu 1999(2), 59–64 (in Chi nese).

ZHANG YUZHONG 2002. Sum mary of Xinjiang ar -chae ol ogy. Kaogu 2002(6), 3–12 (in Chi nese).

Notes

1. Place names within the Peo ple’s Re pub lic of Chinaare given in Chi nese Pinyin with the com monTurkic or Mon gol spell ing in brack ets at the first oc -cur rence. Spell ing of lo cal names is not con sis tentthrough out Eng lish lan guage pub li ca tions and thenames here may vary from those pro vided else -where.

2. The pri mary re search for this pa per was un der takendur ing two vis its to Xinjiang by Betts and Jia. Thefirst, in 2006, was made pos si ble by a Seed Fund inggrant from the School of Philo soph i cal and His tor i -cal In quiry, Uni ver sity of Syd ney and a Uni ver sityof Syd ney Re search and De vel op ment Grant. The2007 visit took place as part of a ma jor re search pro -ject funded by the Aus tra lian Re search Coun cil(Grant No. 0770997) The au thors are grate ful to an

anon y mous re viewer for help ful ad vice and com -ments on the first draft of this ar ti cle.

3. Key Eng lish and French pub li ca tions are:Debaine-Francfort 1988, 1989; An Zhimin 1992a,b.; Chen Kwang-tzuu and Hiebert 1995; Mair (ed.)1998; Mei Jianjun 2000: 7–24.

4. Only a small num ber of C14 dates have been ob -tained for Xinjiang and most are pub lished inuncalibrated form with out stan dard de vi a tions.Where cal i brated dates are avail able these are des ig -nated cal B.C. in the text.

5. At las of Chi nese Nat u ral Ge og ra phy 1998.6. The rain fall in the Yili Val ley and around the town of

Tacheng reaches 600mm. In both ar eas dry farm ingis pos si ble al though ir ri ga tion ag ri cul ture is prac -ticed ex ten sively to day. Both ar eas also open di -rectly onto the Eur asian steppe and are only sep a -rated from it by the ac ci dent of mod ern po lit i calbound aries. The west ern mar gins of the Zhunge’erBa sin east of the hills en joy rel a tively high rain fallwhere dry farm ing is pos si ble, but ir ri ga tion is mostcom mon with mod ern ag ri cul tural prac tices. SeeFrachetti (2004: 146 ff.) for a dis cus sion of ag ri cul -tural po ten tial on the west ern slopes of the hills inKazakhstan.

7. The first gen eral sur veys be gan around the 1950’sun der the di rec tion of Mu Shunying. The re gion was di vided into three re search zones un der the re spon si -bil ity of Wang Mingzhe (west), Wang Binghua(east) and Hou Can (cen tral) (Debaine-Francfort,1989: 6).

8. See also Mu Shunying 1992 who cate gor ised painted pot tery into four groups.

9. The Afanasieveo and the Okunevo are broadly datedfrom the 3rd and the early 2nd mil len nium B.C. re -spec tively (Savinov, 1997: 9; Vadetskaya, 1986:15). These dates are based on rel a tive rather than ab -so lute chro nol ogy. For a use ful sum mary of rel a tivechro nol ogy see also Sokolova 2007. Cal i brateddates give an ear lier range of 3500–2500 cal B.C.and 2600–2000 cal B.C. re spec tively (Frachetti,2004: 192, fig. 5.1; P.198 ff.)

10. See also Mei Jianjun 2002.11. Han Jianye 2005, fig.1. Balikun and Hami are nos.

VI and II re spec tively in An’s anal y sis.12. For a list of early sites in Xinjiang see also

Debaine-Francfort 1988: 9.13. See for ex am ple Derev’anko 1998, Derev’anko et

al. 2000a, b.14. Okladnikov 1978. Okladnikov linked these finds to

Ach eul ian in dus tries in Eu rope, but this has notbeen widely ac cepted. See for ex am ple Yi andClarke 1983: 183; Ranov et al.1992: 59.

15. See also Krause et al. 2007.16. Sites 93AHE1, 2, 3 4, 5 6. Abdurrassul and Zhang

Prehistoric ar chae ol ogy in the Zhunge’er (Junggar) Ba sin, Xinjiang, China 197

Page 32: Archaeology in Zhuge'er - Tianshan Mt., Gobi Desert and Altai Mt

Chuan 1997; Abdurrassul et al. 1998.17. For a brief sum mary of the rather lim ited ev i dence

for ag ri cul ture in the Andronovo, see Frachetti2004: 242.

18. See also Li Xiaoqiang et al. 2007 for iden ti fi ca tionof cul ti vated wheat in Xishanping in mid-north west China by 4650 Cal. Yr. BP; Li Pan 1989, Gansu andJilin 1998, Li Shuicheng and Mo Duowen 2004 for4000 year old wheat seeds found in Donghuishansite, west ern Gansu; Crawford et al. 2005 for thedis cov ery of wheat in Liangchengzhen site,Shandong, east China dated to 2000 B.C. These sug -gest an ear lier, as yet un iden ti fied, pas sage throughXinjiang.

19. For the Xinjiang Neo lithic, see also Debaine-Francfort 1988: 7ff, but note that the sites de scribedthere as Neo lithic in clude some, such as Gumugou,which have now been re cog nised as be ing BronzeAge in date.

20. Chipped stone artefacts from Asitana are on dis play in the Ulumuqi Mu seum.

21. Based on C14 anal y sis, the Andronovo com plex inthis re gion has been dated to around 1500 B.C.

22. In ex ca va tions in the Koksu val ley on the west ernflanks of the Zhunge’er Moun tains (Kazakhstan),Frachetti iden ti fied a sherd from a wheel-made ves -sel that could be par al leled in BMAC forms.Frachetti: 2004: 370.

23. The cem e tery of Yanbulake is one of the bestknown sites in Hami dis trict. It has not been in -cluded here as the ce ramic styles are found onlysouth of the Tianshan and not in the north on theedge of the Zhunge’er Ba sin. Chen and Hiebert(1995) sug gest the painted de signs of Yanbulake are sim i lar to the pot tery in the later Chust cul turewhich in di cates pos si ble in flu ence from eastXinjiang into the Fergana Val ley in late pre his tory.

24. But see Chen and Hiebert (1995: 288) who see nocon vinc ing par al lels be tween Xinjiang painted pot -tery and that from Gansu, and spe cif i cally theMachang cul ture.

25. Frachetti (2004: 205, fig. 5.7) pro vides cal B.C.dates push ing back the east ern Andronovo daterange to c. 1800 – 1400 B.C.

26. See note 25.27. Some, in the Pontic re gion, have been as signed to

the Bronze Age (Telegin and Mallory, 1994).28. This date range is based on cal i brated C14 dates;

the ear lier rel a tive dat ing placed the Afanasievoaround 3000 to 2000 B.C. (Koryakova, 1996). Theclas sic Afanasievo is char ac ter ised as Éneolithicrather than Bronze Age.

29. Xiaohe is an aceramic cem e tery in the Taklamakan. It has been dated by C14 to around 2000 B.C. Thecon nec tions with the Afanasievo and Qiemu’er-

qieke cul tures have been made on the ba sis of theearly date, some par al lels be tween the shapes of thebas ketry and north ern ce ram ics, and the iden ti fi ca -tion of the Xiaohe pop u la tion as Caucasoid (Xin-jiang In sti tute of Ar chae ol ogy, 2003). Lin Meicun(2002) even claimed that the “Qiemu’erqieke cul -ture” rep re sents the traces of an cient Tocharianswho en tered Xinjiang across the Tacheng low lands,crossed the Zhunge’er Ba sin and the Tianshan, fi -nally set tling in the Talimu Ba sin and cre at ing theXiaohe cul ture. The con nec tions be tween Xiaoheand the Afanasievo are now even more ten u oussince cal i brated dates have pushed back the sug -gested date of the Afanasievo by sev eral cen tu ries.

30. Some Okunevo cem e ter ies such as Mokhov-6(Sokolova, 2007: 42) have a strati graphic se quencewhere earth-cut graves are over laid by a later burialin a stone cist.

31. Com pare Fig ure 7 bot tom left (this ar ti cle) with LiShuicheng, 2002, fig ure 14: 3a-d.

32. It is sig nif i cant that in the early to mid- 2nd mil len -nium B.C. there is a shift on the BMAC sites fromar sen i cal cop per bronzes typ i cal of the Kopet Daghto tin bronzes typ i cal of Andronovo sites. Shishlinaand Hiebert, 1998.

33. See for ex am ple Jin Zhengyao et al. 1998 who have ap plied this tech nique to Shang bronzes.

34. For a de tailed dis cus sion see Hemphill and Mallory2004: 213ff.

35. See, for ex am ple, Kuzmina 1998.36. See, for ex am ple, Chen and Hiebert 286ff.37. Per sonal com mu ni ca tions in 2006.

198 Pe ter Wei Ming Jia et al.