upper xiajiadian

128
Borrower: A-G:STWl:06AN Lending String: OCLC:GZM Patron: STWl-2073852 Journal Title: Monumenta Serica -- 1999 Volume: 47 Issue: MonthNear: 1999 Pages: pp viii-xiii, 1-126 Article Author: Psarras Article Title: "Upper Xiajiadian," OCLC Number: 568558052 ILL # - STWl-2073852 I llllllll lllll II lllllll Ill llllll lllll 111111111111111111111111111111111 Location: Online Call #: JSTOR Request Date: 6/6/2017 10:17:01 AM MaxCost: 0.00 Billing Category: Shipping Address: Appleton Public Library Interlibrary Loan 225 N Oneida St Appleton, WI 54911-4717 Borrowing Notes: [A-G:STWl:06AN] Note - LVIS Please send electronic copy to: [email protected]. Thank you. Copyright Compliance: eel Article Exchange This material may be protected by copyright law (Title 17 U.S. Code).

Upload: independent

Post on 06-Feb-2023

0 views

Category:

Documents


0 download

TRANSCRIPT

Borrower: A-G:STWl:06AN

Lending String: OCLC:GZM

Patron: STWl-2073852

Journal Title: Monumenta Serica --1999

Volume: 47 Issue:

MonthNear: 1999 Pages: pp viii-xiii, 1-126

Article Author: Psarras

Article Title: "Upper Xiajiadian,"

OCLC Number: 568558052

ILL # - STWl-2073852

I llllllll lllll II lllllll Ill llllll lllll 111111111111111111111111111111111

Location: Online Call #: JSTOR

Request Date: 6/6/2017 10:17:01 AM

MaxCost: 0.00

Billing Category:

Shipping Address: Appleton Public Library Interlibrary Loan 225 N Oneida St Appleton, WI 54911-4717

Borrowing Notes: [A-G:STWl:06AN] Note -LVIS Please send electronic copy to: [email protected]. Thank you.

Copyright Compliance: eel

Article Exchange

This material may be protected by copyright law (Title 17 U.S. Code).

UPPER XIAJIADIANAuthor(s): Sophia-Karin PsarrasSource: Monumenta Serica, Vol. 47 (1999), pp. 1-126Published by: Taylor & Francis, Ltd.Stable URL: http://www.jstor.org/stable/40727206Accessed: 06-06-2017 15:16 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted

digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about

JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

http://about.jstor.org/terms

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to MonumentaSerica

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Monumento, Serica

47 (1999): 1-126

UPPER XIAJIADIAN

Sophia-Karin Psarras

Contents

Acknowledgements 2 Abbreviations 2

Part I: Presentation of Sites 13

Xiajiadian K^/i' 13 Nanshan ' gen ]§ 1 1 1 ft| 16 Xiaoheishigou /'<^G^i 24 Zhoujiadi MläÖfe 30 Zhengjiawazi ^'^ííf 32 Shi'ertaiyingzi -{-Ufaj^-f- 34 Erdaohezi HjËfnJï1 36 Liangj iashan ^É ψ 1 1 1 37 Wujintang AáfeSí 37 Dadun -j^ 38 Houmuchengyi β$£ί$# and Tuchengzi iVMrf 39 Dahuofang :feRttf 39 Sunjiagou M7371 WMfà 39 Liangjiayingzi M8071 ΉΪΜ^Ψ 40 Tianjuquan M7301 ^EÄ 40 Wafangzhong M791 £0τ/' Ι ? 40 BeishanzuiM750Ut!ll«fa 41 Shuangfang Mils 42 Shangmashi ±ΜΈ 43 Menlian flDi 46 Nandonggou Sí["]?^J 46 Dongnangou '/}''^ykj 49 Sanguandian H'fTÊhJ 51 Comparative Site: Wudaohezi IuËÎïïÎÎ 56

Part II: The Objects 58 Ceramics 58

Shortswords 63

Knives 72

Ge Jc 73 Axeheads 76

Part III: Conclusions: Cultural Exchange 78 Addendum 83

Illustrations and Maps 84 Key to Sites 126

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

2 Sophia-Karin Psarras

Acknowledgements

Original photographs used in this article were taken during the International Academic Conference of Archaeological Cultures of the Northern Chinese Ancient Nations, Hohhot, August 1992. I would like to thank Emma C. Bunker, for inviting me to participate, and Guo Suxin and Jin Fengyi, as well as their colleagues, for making a wealth of material available for examination and photography by conference participants. I also thank Louisa G. Fitzgerald Huber, Corinne Debaine-Francfort, and Lothar von Falkenhausen for their help in providing photocopies of articles and offprints used in the preparation of this article. I have followed Lothar von Falkenhausen, "Issues in Western Zhou Studies," Early China 18 (1993), 225, in the use of capital letters to describe archaeological periods; translations of technical terms are based on the Han-Ying zonghe kexue jishu cihui ΥΧ^^^ίΨ^ίΑ'Λ^ fC (2nd ed.) (Beijing: Kexue, 1988). Special thanks are due Jim Schwaller and Todd Wilda, for facilitating computerization of this manuscript. The Chinese characters for this article have been computerized at the Institute Monumenta Serica (Sankt Augustin).

Abbreviations

Beishanzui, WWZLCK 9: Ningchengxian wenhuaguan T'Wi&SCikttt, Zhongguo shehui kexueyuan yanjiu shengyuan kaoguxi dongbei kaogu zhuanye ψΗ3|±^#^|ξτί:ίΐ^?ζΦ. PÄ%n^4yb^'i^lk, "Ningchengxian xin faxiande Xiajiadian shangceng wenhua muzang ji qi xiangguan yiwude yanjiu" Ttöi&ffiföMfft'ääUti. hü^Cifc-SSPÄ-TOti ¥aÈMïiftïi, WWZLCK 9 (1985), 25-58, PI. 3.

Dadun, KG 1964.12: Sun Shoudao #ypjS, Xu Bingkun ί£^ϊβ, "Liaoning Si'erbao deng di qingtong duanjian yu Daohuofang shiguan mu" il^^^M^JÉW^ffiâ'JA/^C^ClTÎÎ ffc£, KG 1964.12, 277-285, PI. 5-8.

Dahuofang, KG 1964.12: See Dadun.

Dajing, WWZLCK 7: Liaoningsheng bowuguan wenwu gongzuodui xL'y fâWPJJÏSSCtyj WYf- [>Λ, "Liaoning Linxixian Dajing gutongkuang 1976 nian shijue jianbao" ïï/v'W^-Wfc ttit^l976^iífifníffi, WWZLCK 1 (1983), 138-146.

Dapaozi, WW 1984.2: Jia Hong'en îfAS, "Wengniuteqi Dapaozi qingtong duanjian mu" t^WM±m-fííWímm,ww'9M.2, 50-54.

Dongnangou (Beidamian, Huangwozi), KG 1977.1: Hebeisheng bowuguan, wenwu guanlichu ^JJt^lrÇ^jtS^fé/^ííliA, "Hebei Pingquan Dongnangou Xiajiadian shangceng wen- hua muzang" ΗΖψ^^Η&^ΙΪ KS^1fcK#, KG 1977.1, 51-55, PI. 7-8.

Erdaohezi, KG 1977.5: Liaoyangshi wenwu guanlisuo iIlílrfíÍ%'Híí?!iryf» "Liaoyang Erdaohezi shiguan mu" illíl jfiM-f 5I&K-, KG 1977.5, 302-305, PI. 3.

KG Kaogu -*rù.

KGXB Kaogu xuebao vÇ-Jj^fli .

KGXJK Kaoguxuejikan -% ώ 'ψφ fij .

Kuntouyingzi, WWZLCK 9: See Beishanzui.

Liangjiashan, KG 1964.12: See Dadun.

Liangjiayingzi, WWZLCK 9: See Beishanzui.

Longtoushan, NMGWWKG 1992.1-2: Liu Bing ^iJfrK, "Shilun Xiajiadian shangceng wenhua- de qingtong duanjian" iÄifcX£'4Oi: hS SCittäftW''tiL$l NMGWWKG 1992.1-2, 26- 33.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 3

Menlian, KG 1981.2: Qingyuanxian wenhuaju ïWïÀ&SCÎtM], "Liaoning Qingyuanxian Menlian shiguan mu" ilTihïliKa-niÊStffSS, KG 1981.2, 189.

Nandonggou, KG 1977.6: Liaoningsheng bowuguan, Chaoyang diqu bowuguan UfT'^Wy'^] ffi · ÎOITUÉEW-fé/ffi, "Liaoning Kazuo Nandonggou shiguo mu" ]Tr$%J':MWM 5|$S, KG 1977.6, 373-375, PI. 3.

Nanshan'gen habitation site, KGXB 1975.1: Zhongguo kexueyuan kaogu yanjiusuo Nei- menggu gongzuodui »Ι'Ε^^^^ώ'Μ^βίΛ^'ώ'Ί'.ί^Ρλ, "Ningcheng Nanshan'gen yizhi fajuebaogao" TÄSlIltßiStlllSffiffiff, KGXB 1975.1, 117-140.

Nanshan'gen M101, KGXB 1973.2: Liaoningsheng Zhaowudameng wenwu gongzuozhan, Zhongguo kexueyuan kaogu yanjiusuo dongbei gongzuodui iX'T^Wn^jJi^iSC^J Γ. f^árft * M^^^Kc'^'iij'fiff^fiiï'^^Jt l".f^PA, "Ningchengxian Nanshan'gende shiguo mu" AftffilIlWflWfêE-, KGXB 1973.2, 27-39, PI. 1-12.

Nanshan'gen M102, KG 1981.4: Zhongguo shehui kexueyuan kaogu yanjiusuo dongbei gong- zuodui 'lîpattè^^&WiljW^rfr/jUbl^PÀ, "Neimenggu Ningchengxian Nan- shan'gen 102 hao shiguo mu" |^J'i< Λ'Τ-teííl^I Ι ltK102 y-^^^S, KG 1981.4, 304- 308, PI. 7.

NMGWWKG Neimenggu wenwu kaogu |Aj-J£ ifSCtyj^' is

Sanguandian, KG 1985.2: Liaoningsheng bowuguan xIrr^|1S4^/t6', "Liaoning Lingyuanxian Sanguandian qingtong duanjian mu" ilTWU&^^U^iÏÏL^]^, KG 1985.2, 125- 130, PI. 1.

Shangmashi, KG 1982.6: Liishun bowuguan, Liaoningsheng bowuguan MAl^#jtÈ ' 2T" SW^jffi, "Liaoning Changhaixian Shangmashi qingtong shidai muzang" ϊΣ'Τ':'χ)$ΦΆ ΪΆΈΜ^ΗΪΧΖΜ, KG 1982.6, 591-596, PI. 8.

Shi'ertaiyingzi, KGXB 1960.1: Zhu Gui T^rJI, "Liaoning Chaoyang Shi'ertaiyingzi qingtong duanjian mu" ilT^m+^ft^l^SêiJK, KGXB 1960.1,63-71, PI. 1-6.

Shuangfang, KG 1983.4: Xu Minggang iî-R/fêR, Xu Yulin iî-R#, "Liaoning Xinjinxian Shuangfang shigai shiguan mu" Π.Τ$&&&ΰ/Έ'ι&Έίί81&, KG 1983.4, 293-295, PI. 2.

Si'erbao, KG 1964.12: See Dadun.

Sunjiagou, WWZLCK9: See Beishanzui.

Tian/Guo, Ordos Tian Guangjin [ΙΙΓ^, Guo Suxin VfP^|)f, E'erduosishi qingtongqi TJfitf, ZWtâiffâffi· Beijing: Wenwu, 1986.

Tianjuquan, WWZLCK9: See Beishanzui.

Wafangzhong, WWZLCK9: See Beishanzui.

Wangyingzi, WWZLCK9: See Beishanzui.

Wudaohezi, WW 1989.2: Liaoningsheng wenwu kaogu yanjiusuo íL^V^XW^íiWílfíh "Liaoning Lingyuanxian Wudaohezi Zhanguo mu fajue jianbao" ίΧΤ'^νβί-Μ'^ΙίΕΜ^ ocffl££»fiïii, WW 1989.2, 52-61.

Wujintang, KG 1960.5: Jinzhoushi bowuguan íBttlrtfWí^jtS, "Liaoning Jinxixian Wujintang Dong Zhou mu diaochaji" ίΐ^^ΒΆ ^<feíef/j;J^SSiM^íin, KG 1960.5, 7-9.

WW Wenwu SC^h

WWZLCK Wenwu ziliao congkan ΧΡμΦΜΆ f'J.

Xiajiadian, KGXB 1974.1: Zhongguo kexueyuan kaogu yanjiusuo Neimenggu gongzuodui •Ι'ΕΒ^^'Ιδϊ^ώ'^ίεΓίίί-Ι^^ώ" r.f^PÀ, "Chifeng Yaowangmiao, Xiajiadian yizhi shi- juebaogao" *ili$|Ç-Hffi * Ä'^ÜiifiihiÄffiiliff, KGXB 1974.1, 111-144, PI. 1-14.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

4 Sophia-Karin Psarras

Xiaoheishigou, NMGWWKG 3 and WWZLCK 9: Xiang Chunsong ^#fö, "Xiaoheishigou faxian de qingtongqi" /hUSf^Sîffili^, NMGWWKG 3 (1984), 120-123. See also Beishanzui.

Zhengjiawazi, KGXB 1975.1: Shenyang gugong bowuguan, Shenyangshi wenwu guanli ban- gongshi itlilÖC'Ff If^ltS · tJcMUiX^MWl·^^ "Shenyang Zhengjiawazi de liangzuo qingtong shidai muzang" tW^^ii^ffíM&^^HiV^^ KGXB 1975.1, 141-156, PL 1-8.

Zhoujiadi, KG 7984.5: Zhongguo shehui kexueyuan kaogu yanjiusuo Neimenggu gongzuodui ^Hf^^^iij'W^rilflAl^iii'ir.f^PA, "Neimenggu Aohanqi Zhoujiadi mudi fajue jianbao" 'H'£ÙMWMmmïfcn.WM'k, KG 1984.5, 417-426, Pl. 5-6.

In "Exploring the North. Non-Chinese Cultures of the Late Warring States and Han" (Monumento, Serica 42 [1994], 1-125), I examined non-Chinese material from present-day northern China, proposing that much of it could be linked to specific cultural zones as well as specific dates, established through comparison with positively-dated sites. In this first study, however, I examined Upper Xiajia- dian JOs£0..hM onty m passing. The present work remedies this; its results further support the theory of the era-specificity of many small ornamental bronzes

found in the northeast: for example, in terms of my typology, "winged" ornaments, Northeastern zigzag ornaments, the zigzag pattern on semi-spherical appliqué ornaments, and bell types (e.g., type IV). Although these objects are associated with Upper Xiajiadian, they do not seem to occur there in large numbers. They will not be reexamined here.1

The study of Upper Xiajiadian is hampered by the small number of well- reported excavations, the conditions of the tombs themselves, and by confusion concerning the nature of the culture and its dating. In general, the term "Upper Xiajiadian" is considered to refer to a nomadic culture, frequently attributed to the Donghu 2KÎJ3. The culture, thought to be the descendant of Lower Xiajiadian, is defined as characterized by cist graves and shorts words with leaf-shaped blades.2

1 Upper Xiajiadian distribution of these small ornaments is outlined in brief in Part III of the present article. Because publication of sites is scattered and incomplete, this distribution (like all work on the non-Chinese north) cannot be taken as definitive.

2 The use of the term "shortsword" has been questioned by some scholars, who prefer the term "dagger." However, in the context of the non-Chinese north, I believe "shortsword" to be most appropriate. Technically, "shortsword" implies that both the edges and the point of the weapon were used in combat, for slashing as well as stabbing, while "dagger" designates a small arm used essentially for stabbing, although in European fencing it was also used, in combination with a full- length sword, to block the opponent's blade thrusts. The width of the weapons from the non- Chinese north and their often fairly rounded points suggest that the cutting edge was of great importance - hence, that these weapons functioned as short swords. The Upper Xiajiadian leafblade is well-suited for slashing and would possibly inflict more damage in such action than in stabbing, although the form of the blade may fulfill aesthetic rather than practical requirements. For nomadic warfare, the shortsword is a logical choice not only because of its more manageable

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 5

Little other material is advanced as diagnostic. Several problems arise from this definition.

The attribution of any non-Chinese culture to a name provided by early Chinese texts is risky. Unless material evidence appears to coincide with written evidence (assuming that there is indeed written evidence other than the notation of a name), we cannot be sure such attribution is justified. In the case of the Donghu, we have scant textual evidence. We know that this people occupied land east of their con- temporaries, the Xiongnu. We know that they were defeated by the Xiongnu hero Modun "gfíjSÍ, founder not of the Xiongnu state but of the Xiongnu Empire. The texts recount what Owen Lattimore was undoubtedly correct in terming a Xiongnu legend describing the outbreak of war between Modun and the Donghu and, more importantly, the inevitability of Modun' s victory.3 Nothing else remains. From this

alone, it is inconceivable that the term "Donghu" be applied to Upper Xiajiadian if, as is currently believed, Upper Xiajiadian dates to ca. 11-4 c. B.C. There is no indication in textual sources that the "Donghu" were known before ca. 4 c. B.C. It could be argued that, although the name is not recorded, the people did exist, known by a different name. This is possible, but with the multitude of names noted in the early texts with little or no further information, there is even slimmer

length, but because a slashing motion would reach the adversary more easily than a stabbing one. In the Western nomadic context, "shortsword" translates the Greek "akinakés," designating the Scythian weapon and, by extension, the steppe arm of similar type. Although the people of Upper Xiajiadian were not nomadic and, indeed, not a steppe culture, their material remains do demonstrate some affiliation with such cultures; the application of the nomadic term "shortsword" to Upper Xiajiadian is therefore appropriate.

3 The legend begins with the conclusion of a treaty between the Yuezhi ft Εξ and the Xiongnu &JÍK, when the young Modun was sent to the Yuezhi as a hostage guaranteeing the accord. His father Touman 3<l1k is said to have then attacked the Yuezhi, hoping to rid himself of Modun. Modun escaped and returned to his father's court, where he created a following of his own and then seized

power through patricide. His first confrontation as the new Chanyu -φ-Τ* (sovereign) was with the Donghu, who requested first his prized horse, then one of his wives, both of which were ceded. The Donghu then demanded control over a strip of unoccupied territory which lay between their lands and those of the Xiongnu. Modun refused and waged a war of conquest on the Donghu. (Sima Qian fj]-%ii, Shiji ÜC j> HO.) The narrative has several elements of legend: the murderous father (giving Modun justification for patricide, as well as legendary powers and divine protection); the repeated use of the number three (three tests to determine his followers' loyalty, three tests by the Donghu); and the repeated references in both the tests of loyalty and the demands of the Donghu to a prized horse and a favorite wife - followed by land seizure: Xiongnu land when Modun assumed sovereignty, Donghu land as he built his empire. While the story undoubtedly contains some factual elements (the treaty with the Yuezhi and the use of a hostage system to ensure treaties, the patricide, the founding of the empire), Lattimore's suggestion that Sima Qian recorded Xiongnu legend is undoubtedly correct (Owen Lattimore, Inner Asian Frontiers of China [New York: The American Geographical Society, 1940], 463-465). In this context, the term "Donghu" is highly ambiguous.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

6 Sophia-Karin Psarras

evidence for attributing Upper Xiajiadian to one of these other peoples.4 The term "Donghu" itself presents problems. "Hu" usually refers to horse or mounted nomads, often specifically to the Xiongnu or to peoples within the Xiongnu sphere of influence. It is a general collective term. "Donghu" could be rendered as "Eastern Mounted Nomads," but also as "the non-Chinese north of China and east

of the Xiongnu," referring to a specific group or coalition, or as a generic term distinct from earlier Yi ^. In any case, attributing Upper Xiajiadian to the Donghu compounds the problems the material remains themselves present. At this time, I see no benefit in making any specific attribution.

Another problem arising from the widely accepted definition of Upper Xiajiadian is the assumption that the culture was nomadic. The excavation report for Zhengjiawazi ^[5^c/É"P, one of only two sites where a nomadic presence is discernable, comes closest to my reading of the data when it concludes that the site

belongs to "members of a ... tribe who lived by farming and herding, which were occasionally supplemented by fishing and hunting."5 The question that arises is how to identify a nomadic culture. It is clear that nomad-type material culture (bronze ornaments, both plain and decorated, in appliqué form; shortswords; the representation of animals in art; the use of trousers and boots; horse-gear) may be and has been used by sedentary peoples. In the West, examples include, to greatly varying extents, the Thracians, Celts, Scandinavians, and Etruscans.6 In the East, China itself possessed many of those forms most frequently attributed to nomads:

semi-spherical appliqués, animal forms, horse-gear. Shortswords remain relatively rare, but the presence of the shortsword does not prove a site nomadic. In Upper Xiajiadian, the situation is complicated by the presence on some sites of Chinese ritual bronzes. Their presence, a mark of foreign sedentarism, implies that the vessels had some worth, in whatever capacity, for the people of Upper Xiajiadian. If the people of Upper Xiajiadian were nomadic, what worth did these vessels have

that offset their encumbrance? In combination with the horse-bits (relatively rare), arrowheads, and agricultural and fishing tools, the content of Upper Xiajiadian graves recalls Chinese burials more than Xiongnu tombs.

4 E.g., Liu Guanmin t'MK;, Xu Guangyi ^)ltM, "Neimenggu dongbu diqu qingtong shidaide liangzhong wenhua" l^^^jvÄEfflijnWWWfttit NMGWWKG 1 (1981), 5-14, which includes a confused discussion of Upper Xiajiadian, implying an attribution to the Hu ffl, translated by the authors as "Huns."

5 Zhengjiawazi, KGXB 1975.1, 156. In order to simplify footnotes, full references to Upper Xiajia- dian sites are listed under "Abbreviations."

6 The Thracians (west of the Black Sea) were in close contact with the Scythians (north of the Black Sea). See, inter alia, Gerda von Biilow, Treasures of Thrace (New York: St. Martin's Press, 1987); Lioudmila Galanina, L'Art des Scythes (Leningrad: Aurora, 1987); Mauro Cristofani, Marina Martelli, et al., L'Or des Etrusques (Paris: Editions Atlas, 1985).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 7

The sole element that I have been able to identify as distinguishing the nomad burial from the sedentary is the presence in the grave of animal bones, specifically of horses, sheep, or cattle. Everything else must be viewed in this context. If a given grave has no animal bones but occurs in a context of such ritual (i.e., a cemetery where such ritual is prevalent), the grave may be accepted as nomadic if there are no counter-indicators among the grave goods or burial structure. It is notable that sacrificed horses were buried in Chinese cemeteries as well, along with carriages and horse-gear. The context provided by the human graves, however, is heavily sedentary. Further, some nomadic groups, such as the Xianbei, sometimes bury horses in separate pits, as at Yushu tfoW, Juin;7 the human graves generally retain some indication of animal sacrifice, nonetheless.

Upper Xiajiadian graves seem to yield no assemblages. Although there are objects which seem unique to this culture, these objects do not appear on every, or even most, sites. Even the leaf-bladed shortsword occurs beyond Upper Xiajiadian, although it seems to originate there. A multiplicity of forms is used, many of which are shared with other cultures. Burial structures include not only the cist grave, but

pit graves in what seem to be equal numbers; burial rites include the use of wood coffins or biers, jar coffins, second as well as first inhumation, single or double burials, grave firing, and the burial of individuals in ash pits containing the refuse of habitation sites. Bodies are placed on their back or on their side; legs may be extended or crossed. Some graves contain remnants of birch bark, which may have lined the pit or cist. There are no discernable patterns which allow one type of burial to be identified as representing a specific, if unknowable, group. At this time, there also appear to be few cases in which burial rites may be determined as specific to the culture or to a larger zone, or as representing intrusions from other cultures. The exception is the presence of animal bones in the grave. With so much variety and a lack of assemblages, it is difficult to provide a de-

finition of Upper Xiajiadian. None of the forms which might characterize the culture appear uniformly in graves, except for those ceramic forms which belong not to Upper Xiajiadian specifically but to the northeast in general. The leaf-bladed shortsword, the double scabbard, the use of the triangle or L-triangle pattern as décor, do appear to be distinctive. Art distinguished by not only geometric designs, but an iconography centered around reproduction (mating animals; the nude female; the nude, erect male) also largely characterizes the culture. Cist graves and the use of shells or bronze semi-spherical plaques to cover the face or body of the dead do not. The latter practice, although often thought to be a mark of Upper Xiajiadian ritual, occurs only at Zhengjiawazi, Zhoujiadi j^^cife, and, allowing for

7 Jilinsheng wenwu kaogu yanjiusuo Yfffi&XW^k'ffiflffu Yushu Laoheshen Wl¥í&uíM (Bei- jing: Wenwu, 1987). Sometimes called "Yushu," which I have adopted, this site more properly should be referred to as Laoheshen. The site yielded several cultural layers.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

8 Sophia-Karin Psarras

minor variation, at Shi'ertaiyingzi -f-ziß^^. In contrast, it occurs in 30% of Yu- huangmiao -Κ^ρίΑ"? (Yanqing County 5ÊJ$c-ít, Beijing Municipality) tombs.8 Cist graves also occur in Yuhuangmiao îRâJÉ, and the Chinese states of Yan Sffi, Zhao M, Zhongshan ψΐΐ], Jin i^f, and Qi ^, as well as across Mongolia and non- Chinese sites in northwestern and southwestern China. Upper Xiajiadian cist graves are made of large slabs or of small stones, or a mixture of the two. There is no one method of grave construction, nor is any one specific type of stone used. The form is generally rectangular, often with the width at the top of the tomb greater than at the bottom. Some graves have stone walls, but no floor or lid; others have walls and a lid, but no floor; etc. Nor, indeed, is there any consistent link between the use of a cist tomb structure and the contents (in type or in quantity) of the tomb.

Unlike in the case of most Xiongnu tombs in present-day China, Upper Xiajia- dian does present some stratigraphy. Xiajiadian, Nanshan'gen ]§ll|^, Liang- jiashan ^ξψΐΐΐ, Dadun ^C^, Xiaoheishigou /J'^^f^, Zhengjiawazi, Zhoujiadi, and Shangmashi h-HS yielded ash pits, habitations sites, and cultural layers. Even though at Xiajiadian, Nanshan'gen, and Shangmashi, in particular, layers have been defined, the fact that there has obviously been much erosion throughout the region makes interpretation of this stratigraphy highly problematic. For instance, at Shangmashi, layers include: 10-28 cm of cultivated earth, 49-98 cm of cultural layer with shells and red-brown siliceous pottery shards, 8-28 cm of cultural layer with black-brown siliceous pottery. Graves, including some with leaf- blade shorts words, occur in the first cultural layer with red-brown pottery shards;

jar coffin graves intrude between the same layer and the black-brown pottery layer. The jar coffins include black siliceous ceramics, but also brown and grey. Both jar coffins and the so-called shorts word graves have been broken by ploughing. Both are therefore near the surface, regardless of the maximum depths of their respective layers, and do not, therefore, strictly lie one over the other. Soil erosion,

responsible for the exposure of many graves, has made uneven deposits of earth and of material remains. This means that we cannot assign a theoretical number of years to any layer, because, with shifting ground, the time taken to create the deposit cannot be estimated. Deposits have remained neither constant nor un- disturbed. On some sites, the cultural layer over the graves came from goods swept from the broken graves themselves (e.g., Wudaohezi WiEM"?» a site often con-

8 Beijingshi wenwu yanjiusuo Shanrong wenhua kaogudui '^tMih'XW&f^Í!J[''')^'Xít^'lS^K, "Beijing Yanqing Jundushan Dong Zhou Shanrong buluo mudi fajue jilue" JtJtC^^^ífSlIl^OnJ

llljJc^SSftÄfiffifn^ WW 1989.8, 17-35, 43, Color PI. 1, PI. 3-5; Beijingshi wenwu yanjiu- suo JfcM'rfJÍC#JW?£0f, Beijing kaogu sishinian Jt^YiW+^T- (Beijing: Yanshan j]k''', 1990), 79-86, PI. 19-22. I also use the term "Yuhuangmiao" to designate a culture which includes the site of Xiaobaiyang: Zhangjiakoushi wenwu shiye guanlisuo, Xuanhuaxian wenhuaguan Stc^ UWXPJJ-ltíY/SMJVí ' *mÀk-&-'£ikÎ5, "Hebei Xuanhuaxian Xiaobaiyang mudi fajue baogao" Wt^At&^Üm&i^mUfr. WW 1987.5, 41-51.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 9

sidered Upper Xiajiadian, but whose attribution I question). At yet other sites, the associated material (pottery shards, intruding graves) is Warring States or Han Chinese. There is no reason to believe that this material postdates the Upper Xiajiadian burials significantly. Indeed, given political history and material evidence, the Chinese material is likely to be contemporaneous with the graves or, in the case of intruding burials, largely contemporaneous. While Chinese tombs at Shi'ertaiyingzi mingle with Upper Xiajiadian burials, at Zhengjiawazi and Dadun, Han graves intrude on Upper Xiajiadian tombs. This may well reflect, not chronology, but political policy and the Han conquest or colonization of the region. While Yan respected the integrity of the burials (Shi'ertaiyingzi), the Han did not; this does not imply a much earlier date for Zhengjiawazi and Dadun, but a change in the political climate.

Although some Upper Xiajiadian tombs contain Chinese ritual bronzes dating from as early as ca. 11 c. B.C., Upper Xiajiadian sites have not always been dated according to these vessels. Indeed, if the presence of Chinese objects or Chinese- related material has sometimes been used to reinforce a date derived by other means, the Chinese material itself has not been extensively examined. A given site, such as Nanshan'gen M101 or Xiaoheishigou, may contain both Western Zhou bronzes and Late Warring States imports, a fact which has yet been ignored. Instead, a dating based on the shortsword analyses of Jin Fengyi $Jf fX |& has come to be widely accepted, further complicating analysis.9 Jin divides Upper Xiajiadian into three periods, the third further subdivided into

three: Dajing ^C#,10 from the Shang/Western Zhou boundary to the Early Springs and Autumns period (ca. 11-8 c. B.C.); Nanshan'gen, from the Early-Middle Western Zhou to the Early-Middle Springs-Autumns (ca. 11/10-8/7 c. B.C.); and Shi'ertaiyingzi-I, Late Western Zhou to Middle Springs-Autumns (ca. 8-7/6 c. B.C.); Shi'ertaiyingzi-II, Middle Springs-Autumns to Early Warring States (ca. 6- 5 c. B.C.); and Shi'ertaiyingzi-IH, from Early-Middle Warring States to Middle Warring States (ca. 5-4 c. B.C.).11 This periodization, largely overlapping, is

9 Cf. also the recent report, extraordinary in its analysis of dating: Maoxian Qiangzu bowuguan, Abazangzu Qiangzu zizhizhou wenwu guanlisuo ^ã-^^^MM^BM^^^à^X^J^' HlßJf "Sichuan Maoxian Moutuo yihao shiguan mu ji peizangkeng qingli jianbao" |5IJI|/!icJi4^ it- l^WXMn^Jif^lML WW 1994.3, 4-40. The site yielded Chinese bronzes from the Early Western Zhou and Late Springs-Autumns, but comparison with other non-Chinese sites in the region provided ceramic-based dates of the Middle-Late Warring States (op. cit., 32).

10 Dajing is a copper mine and production site with an associated habitation site. Material found was mostly lithic, but included "oracle bones," bone and antler tools, ceramics and ceramic shards

(including a // of my type II, see Part II of the present article), and a ceramic fitting ending in a horse-head (WWZLCK1, 142, f. 7:1, 10), like those of Sanguandian H'R'fej and Wangtu -HI;.

11 Jin Fengyi, "Lun Zhongguo dongbei diqu hanquren qingtong duanjiande wenhua yicun (shang)" ife1 1 ' Da/jCJtHfeE^lllJ^JW^MaÜniJjtiiÜSiT^-.h), KGXB 1982.4, 387-426; Jin Fengyi, "Xia-

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

10 Sophia-Karin Ps arras

based on shortsword form, which Jin defines as follows: for the Dajing era, leaf- shaped or straight- bladed shortswords (or swords) with conical hilt units (cf. Plates 62, 70); for Nanshan'gen, conical-hilt swords and shortswords, leaf-bladed short- swords with wood hilt units, which have since disintegrated, or with bronze hilt units in the form of crouching animals (Plate 68, far left), and straight-bladed shortswords with rectangular hilts and jutting guards which I term "notched" guards (PI. 68, third, fourth, and fifth from left; PI. 70, top); for Shi'ertaiyingzi, leaf-bladed shortswords with, in Period I, wood hilt units but iron ore counter-

weights of regular shape (in my terms, "regular crowns") (PI. 66, right); in Period II, leaf-shaped blades, bronze hilt units with horizontal pommel and regular crown (PI. 63); and in Period HI, leaf-shaped blades, bronze hilt units with thin, downward-curving pommels, and irregularly-shaped crowns (cf. PI. 65, right). This analysis is refuted by the examination of Zhengjiawazi, Shangmashi, and Xiaoheishigou in particular.

Jin's chronologies are further supported by reference to 14C dates, specifically for Dajing. Jin cites a series of six dates12: 2700 Before Present ("Present" being defined, as is standard, as 1950) ± 100, 2715 B.P. ± 85, 2720 B.P. ± 90, 2780 B.P. ± 100, 2795 B.P. ± 85, 2970 B.P. ± 115. Additional dates for the site have

been published which coincide in two cases with those given by Jin Fengyi.13 They are: 2720 B.P. ± ("physical date," using the corrected half-life of 5730 for 14C, corresponding to 2640 B.P. ± 90 "conventional date," with the Libby half-life of 5570 or 5568 for 14C, and 2810 B.P. ± 135, using a standard dendrochronological calibration formula) and 2970 B.P. ± 115 physical date (2885 B.P. ± 115 con- ventional date, 3120 B.P. ± 140 dendrochronologically-calibrated date). These dates yield, overall, a range of < 1290-600 B.C. > , which is close to the chrono- logical horizons proposed by Jin for Upper Xiajiadian as a whole. Other 14C dates have been published, although the sites from which they were taken seen not to have been: Jianzigou M64:2 jf^f y£j, in Luanping ^ψ, Hebei, 2370 B.P. ± 85 physical date, 2300 B.P. ± 85 conventional, 2390 B.P. ± 95 dendrochrono- logically-calibrated, yielding a span of < 535-265 B.O.14 For Tabu'aobao i'^MCül habitation site, in Balin Right Banner K#^Djft, Inner Mongolia, 2424 B.P. ± 70 (14C values not given) and 2355 B.P. ± 70, or < 544-335 B.C. > ; 2202 B.P. ± 69 and 2140 B.P. ± 69, or < 321-121 B.O; 2595 B.P. ± 76 and 2522

jiadian shangceng wenhua ji qi zushu wenti" UWk±MXÎk&j%ïfcMû]$â, KGXB 1987. 2, 177-208.

12 Jin Fengyi, "Xiajiadian shangceng wenhua," KGXB 1987.2, 189.

13 Zhongguo shehui kexueyuan kaogu yanjiusuo (ed.), ' | f[IItt^>#^K:-Ç" íítyfíífyi , Zhongguo kao- gwcuezhong tanshisi niandai shuju ji 1965-1981 '['Η^'ώ^ψ^+ΙΏ^^Μ'ίΑ 1965-1981 (Beijing: Wenwu, 1983), 25.

^ Zhongguo shehui kexueyuan, Zhongguo kaoguxuezhong tanshisi, 11.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 1

B.P. ± 76, or <721-496 B.O; and 2330 B.P. ± 56 and 2264 B.P. ± 56, or < 436-258 B.C. > ,15 Also in Inner Mongolia, at the habitation site of Dianjiangtai &Μ&, in Chifeng: 2598 B.P. ± 46, 2525 B.P. ± 46, or < 649-529 C.C. > ; 2589 B.P. ± 48, 2515 B.P. ± 48, or <687-517 B.C. > , again without indication of 14C values.16

Of these dates, all were derived from ash samples with the exception of Jianzigou, which was based on human bone. Although bone, because of its porosity, has been considered until very recent advances in radiocarbon techniques as an unreliable source for radiocarbon testing, in northern China, tests done on bone are often more reliable than those done on ash, even though many of these tests may not have been done using recent technological improvements. For example, at the site of Baimiao &f'§, in Zhangjiakou J^cP, Hebei, samples were taken from M64 for testing. Wood from the coffin yielded dates of 2995 B.P. ± 150 (physical), 2910 B.P. ± 150 (conventional), and 3155 B.P. ± 170 (dendro- chronologically-calibrated), or < 1195-895 B.O, < 1110-810 B.O, and < 1375-1035 B.O . Bone from the skeleton in M64 yielded dates of 2260 B.P. ± 140 (physical), 2195 B.P. ± 140 (conventional), and 2265 B.P. ± 145 (dendro- chronologically-calibrated), or <450-170 B.O, < 385-105 B.O, and < 460- 170 B.O. The tomb was dated in general to the Warring States period (475-221 B.C.) on the basis of the material it contained.17 The ash-derived dates are therefore far too high, while the bone-derived dates are within expected perimeters. A similar situation exists at the Houma ffcH, Shanxi site of Qiaocun fffâ, published in Wenwu 1972.1. Ash-derived dates from M19 read 2560 B.P. ± 115, 2485 B.P. ± 115, 2620 B.P. ± 155 (physical, conventional, dendrochrono- logically-calibrated, respectively), or < 725-495 B.O, < 650-420 B.O, < 825-515 B.O; while bone samples from M21 yielded 2325 B.P. ± 85, 2260 B.P. ± 85, and 2340 B.P. ± 95, or <460-290 B.O, <395-225 B.O, and < 485-295 B.O . Both M 19 and M21 were dated through object analysis to the Warring States.18 In Jilin, a Middle Warring States site at Houshishan $&HlL| yielded ash-based dates of 2990 B.P. ± 100, 2905 B.P. ± 100, and 3145 B.P. ± 130, or < 1040-940 B.C. > , < 1055-855 B.C. > , and < 1325-1065 B.C. > ; and 3030 B.P. ± 120, 2940 B.P. ± 120, 3195 B.P. ± 175, or < 1200-960 B.O, < 1110-870 B.O, and < 1420-1070 B.O. Bone samples (M34) read 2275

15 Zhongguo shehui kexueyuan kaogu yanjiusuo shiyanshi ΓΓτΕ3|±^#^Κ^-"ώ^3ξ0ρ^^^1, "Fangshexing tansu ceding niandai baogao (20)" WM^^WM&k-iWtrr (20), KG 1993.7, 645.

16 Zhongguo shehui kexueyuan, "Fanshexing tansu," KG 1993.7, 645.

17 Zhongguo shehui kexueyuan, Zhongguo kaoguxuezhong tanshisi, 1 1 .

18 Zhongguo shehui kexueyuan, Zliongguo kaoguxuezhong tanshisi, 17.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 2 Sophia-Karin Psarras

Β. Ρ. ± 70, 2210 Β. Ρ. ± 70, and 2280 B.P. ± 80, or < 395-255 B.O, <33O- 190 B.C. > , and < 410-250 B.C. > ,19

Clearly, bone samples are not uniformly reliable in northern Chinese testing any more than they are elsewhere in the world with the older technology, however in most cases where both ash and bone dates are provided, bone-derived dates are closer to the dates derived from object-analysis. It is not clear whether the obvious problems in many ash-based dates stem from sample association (which seems unlikely when the same problems are associated with wood taken from coffins), from laboratory preparations of the sample, or from some indeterminable factor (soil conditions, wood treatment before original use, reuse of wood, use of inner vs. outer portions of the trunk). Problematic dates may be much higher, or some- times much lower, than anticipated (object-derived) dates. Further, in those Chinese publications where no 14C values are cited, the dates given may be the result of various interpretations of one test, and not a spread of dates obtained from

multiple test-runs. In the north, only rarely is more than one test-run on a given sample reported.

In recent years, physicists have concluded that a single dendrochronology cannot be used worldwide, since trees absorb radiocarbon isotopes at varying rates according to local climatic conditions.20 Likewise, 13C, for which most Chinese

radiocarbon dates are not calibrated, varies according to the disintegration of plant matter on a given site.21 It is reasonable to assume, therefore, that historical changes in the amount of vegetation in a given locality may affect production of 13C, and hence the calibrated 14C date. In any case, at its best and most reliable, a radiocarbon date indicates only the possibility, usually calculated to 68%, that the site falls on a given date, plus or minus laboratory error (calculated in years). As such, it can influence our determination of date most in the absence of positively- dated references. In the case of Upper Xiajiadian, as with other Late Bronze Age cultures in northern China, a positive chronology can be established, and radio- carbon readings therefore assume only a minor role.

19 Zhongguo shehui kexueyuan, ZJiongguo kaoguxuezJiong tanshisi, 35.

20 Despite the contention that trends worldwide have proven remarkably similar, dendrochronologies have been built through the cumulative evidence of "floating" chronologies produced region by region, according to growth in that area. It is not at all evident that such chronologies have been produced for the various regions of China. Cf. M.G.L. Baillie, Tree-Ring Dating and Archaeology (Chicago: University of Chicago Press, 1982).

21 David C. Coleman, Brian Fry (eds.), Carbon Isotope Techniques (San Diego: Academic Press, Inc., 1991). Bone is also susceptible to new carbon contamination, yielding dates which are too low. However, given the data discussed here with dates established through object comparison, my conclusions on the preferability of bone-derived dates in China stand.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 13

PARTI

Presentation of Sites22

Xiajiadian SMfe

Located on a hilltop, some 30 m above the Yingjin River -^^ynj in Chifeng Municipality ff^, Inner Mongolia, the name-site of the culture presents little that is typical of Upper Xiajiadian as it is usually defined. Three sectors were excavated. Sector I, 115 square meters, yielded a semi-subterranean habitation F4, ash pits H23-27, and tombs M7, 9, 10. All were found just below the surface of the ground. In Sector II, 55 square meters, four layers were uncovered. One example, from trench 9 (T9), is published. According to the report, a layer of cultivated earth and a layer of spodosol (huitu M~t~) , 0.8-2. 25m thick at a depth of 0.2-0. 7m from the surface, covered nine ash pits (H6, 7, 14-20). Near the bottom of a lower layer of brown earth (hetu #gi) > 0.45-0. 7m thick and at a depth of 0.8- 2.3m from the surface, were found two tombs, Mil and 12. A fourth layer belonged to Lower Xiajiadian. Mil intrudes on the Lower Xiajiadian ash pits H13, 14, and is intruded upon by the Upper Xiajiadian ash pit H6. M 12 is intruded upon by H7 and H 16. H 19 and 20 and H 14 and 17 appear to intrude upon each other. Ceramics are of red clay, hand built, and fired at a relatively low temperature. No distinction in form can be made between the ceramics from the ash pits and those from the graves. Mil is a cist grave constructed of large rocks; H16 includes rocks in one wall.

22 In addition to Dajing, sites not discussed in this article include Dapaozi ~^zM~f, Wengniute Banner ΜΨ^ίύΐ, Inner Mongolia, whose most distinctive ceramics do not match those of Upper Xiajiadian, although the site yielded a type IIIc jar (PI. 40), a type I dou (PI. 48), as well as straight-bladed shortswords with conical hilts (Dapaozi, WW 1984.2, 52, f. 3:1, 3; f. 2:1, 2). The Dapaozi ceramic décor (e.g., WW 1984.2, 52, f. 4:2, 3; f. 5:4) is the same as that on ceramics from Baijinbao Ù&M, Zhaoyuan Co. Í&M, Heilongjiang: Heilongjiangsheng wenwu kaogudui ^'Mï-'&'XWj^ù'tik, "Heilongjiang Zhaoyuan Baijinbao yizhi di yi ci fajue" WZUMUÙ& ^i&itffr-YxUM, KG 1980.4, 318, f. 10:23 (Hl:16); 319, f. 11:4 (Hl:18). The relationship of these two sites is unclear; Baijinbao otherwise initially appears to be neolithic, judging from the predominance of ceramics and bone or stone implements (which need not imply date). Also omitted are: Nanshan'gen M102, which provided little data (KG 1981.4), and scattered finds summarized in Shenyangshi wenwu gongzuozu ffliï]ï'iJC$} T.fféR, "Shenyang diqu chutude

qingtong duanjian ziliao" YJc'n&Klh:h(fiftfifàMW¥i, KG 1964.1, 44-45, PI. 7; Jianpingxian wenhuaguan, Chaoyang diqu bowuguan jîf'a-XitÎS ' /^^FlÄEW/^jtß, "Liaoning Jianping- xian de qingtong shidai muzang ji xiangguan yiwu" αΙΙ^ϊΕ-Ί^-Μ-Ι^^^Π^-ΚΡ^^-^Μ^!/, KG 1983.8, 679-694, 713, PI. 2-3, which includes material from Shuiquan(chengzi) /KÄÄ-f not found elsewhere, but which also includes Shang-era non-Chinese knives, e.g., p. 690, f. 13:1; and Liu Bing ^Jfrjc, "Shilun Xiajiadian shangceng wenhuade qingtong duanjian" ÖafcJÖ^Jiv hSX iktitiifCnftL&L NMGWWKG 1992.1-2, 26-33, which includes material from Longtoushan jfc^lli and claims a dendrochronologically-calibrated radiocarbon date of 3240 B.P. ± 150 for this site, based on a sample from a wood bier or coffin.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 4 Sophia-Karin Psarras

Sector III, 90 square meters, also revealed Upper and Lower Xiajiadian layers. Beneath cultivated earth at a depth of 0.5-1. 05m from the surface, a 0.3-0. 85m thick layer consisting of thin layers of siliceous yellow-grey earth, spodosol, and loam {hangtu 3?ih),23 contained the Fl habitation. The next layer, 0.3-0. 55m thick at 1.75-2.0m from the surface, consisted of light grey earth (qianhuitu $ç$tih) and contained ash pits H5, 8, 9, 10, 22, a semi-subterranean habitat F5, and tombs M13-17. F5 and H10 intrude upon M15. Piles of rocks, found in what now seems random placement, indicate that M 13, 15, 16, and 17 were cist graves, and that some ash pits (H10 especially) may have been at least partially reinforced with rocks. A fourth layer, of yellow-grey earth, belonged to Lower Xiajiadian.

A Sector IV was also excavated, but belongs to Lower Xiajiadian. Within these sectors, no chronology can, to my eye, be distinguished. There

seem to be no traces in the ash pits of any bronze material, although tombs contained bronze arrow-heads (M12, two), knives (Ml, 12, 14, 17, one each), awls (Mil, 14, 17, one each), and ornaments (Ml, 7, 11, 14, 15, 17; Mil with 105 pieces; M17, with 39). No shortswords are reported, nor horse-gear, either in the tombs or the ash pits. Ash pits yielded high numbers of ceramics (including spinning whorls), and bone or antler and stone tools, as well as animal bones including dogs, pigs, sheep, cattle, horses, and birds. Bone and stone tools are also found in the tombs, as are a small number of ceramics (M7, 11, 12, 14, 15, one jar each). No animal bones were found in any tomb. Tomb structure includes rectangular pits framed with rocks, in the case of

Mil, 13, 15, 16, 17. These may be called cist graves if the term here is understood to mean only that stone figured in the grave structure in some way. The structure of these graves is not entirely clear, due to the shifting of the gorund, as

is evident from the stratigraphie measurements. All seem to be rectangular pit graves and all except for Mil and 12 are single burials. Mil held a woman and child; M12, a couple, placed side by side in opposite directions. M7, 9, and 13 each held a child. All burials were in extended dorsal position, except in M9 and 13, where the legs were somewhat bent. Grave orientation seems random, with Ml and 7 oriented to the north; M13-17, to the south; M9, 11, to the northwest; and M12, east-west. Ml and 11-17 contained some kind of wood inner-grave structure, either a coffin or a bier.

In addition, human beings (apparently adult) were buried in the ash pits H23- 27: one person in H24 and H25; two people in H23 and H26; three, in H27. There is no report of any special treatment of these bodies, nor any indication of how they died. Whatever social class or type of death (from disease? or sacrifice?) they represent, clearly these burials fulfill a social distinction which accords no more

23 Hangtu '£r:'' is rendered in geological terminology as "loam," but the phrase may also mean "tamped earth." In this context, "loam" seems the more likely meaning.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 5

respect to these dead than to the remains of the animals disposed of in the ash heaps. A body was also found in the fill of F4. The site is dated by the excavation report authors to the 8 - early 5 c. B.C.

with the claim that, had the site dated to a time after the early 5 c. B.C., the influence of Yan and Qin would have been visible. This deduction is questionable: it supposes that China had more influence on the non-Chinese in later periods (i.e., after the early 5 c. B.C.), and that Chinese influence on the non-Chinese was both inevitable and readily visible, without the reverse being true. Countering these assumptions, the site does yield a number of pieces which indirectly provide a positive date. In ceramics, these are: dou W. type I,24 with a broad, deep bowl and a short, flaring foot, is found on Xiajiadian (F4:l) and Erdaohezi ZljSM"? (PL 48). Dou II, with a shallow, flaring bowl which may be triangular in cross-section, and a long, heavy stem barely widening to a foot, is found on Xiajiadian (H9:4, T3-4:37) as well as, in similar form, at the Upper Xiajiadian site of Tianjuquan p^lii^ (Ningcheng County τίδζ-ϋ» Inner Mongolia), and Dajing (PI. 49). The Xiajiadian bowl fragment H9:4 is similar in form to the bowl of a dou from Fushun $tejl|j/j (Liaoning),25 which is dated to the Late Warring States/Early Western Han by virtue of its ceramic forms (PI. 49). Ding II ^}, with a deep body, vertically looped handles extending downward from the rim, and squat, thick legs, is found on Xiajiadian (H5:15, T7-l:2) and Yuhuangmiao (M52:l) (PI. 54, right). Yuhuangmiao is positively dated to the Late Western Han through the presence on the site of a Chinese garment hook (PI. 113, second row, far left) like that found on the Late Western Han site of Qianping Hp^, in Yichang County S lü -It, Hubei (PI. 112, top).26 This date is further reinforced by the relation of two Yuhuangmiao shortswords to the Mancheng ffâifò (Hebei) shortsword Ml:5109, ca. 113 B.C.27 Clearly the hilt of the Mancheng weapon (PI. 73, far right) is patterned after the Yuhuangmiao example seen second from right on Plate 72, while its pommel takes its form from the Yuhuangmiao piece second from left on the same plate. A // g with a highly vertical shape, full legs, and small lugs placed just below the neck, is

24 Typologies, unless otherwise credited, are my own. Object references for material pictured in the present article will be given in the list of illustrations. Other material will be footnoted.

25 Wang Zengxin '-fir^f, "Liaoning Fushunshi Lianhuabao yizhi fajue jianbao" üT'KJ'ICrfi mimmia:%:Mm'i, kg i964. 12, 286-293.

26 Yichang diqu bowuguan ÎLmffiEft^JÎB, "1978nian Yichang Qianping Hanmu fajue jianbao" l97S^XT.^mffUW&WM'i, KG 1985.5, 413, f. 4:11 (M108:9). See also Psarras, "Explor- ing the North: Non-Chinese Cultures of the Late Warring States and Han," Monumento, Serica 42 (1994), 1-125.

27 Zhongguo shehui kexueyuan kaogu yanjiusuo, Hebeisheng wenwu guanlichu, ΐ[ΐ[1Ι|±^4^1%>ξ-' 'iSWWft ' Hll'&WfãM!é, Mancheng Hanmu fajue baogao W&UUfcMUYf (Beijing: Wenwu, 1980), vol. 2, PI. 44:4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 6 Sophia-Karin Psarras

found on Xiajiadian (Tl-2:7). Closely similar pieces were excavated from Dajing28 and, without lugs, at the Yuhuangmiao-culture site of Xiaobaiyang /Jn^PH, Xuan- hua County jfEJ-fcJl·, Hebei (H3:15),29 as well as, in bronze, from Nanshan'gen M101 (PI. 53, top row). Xiaobaiyang may be dated to the Late Western Han, since its material is largely identical to that of Yuhuangmiao. Nanshan'gen M101 proves to date to the Western Han as well, as will be discussed below. A short, squat pot with a slightly flaring neck, sometimes a slight angle-edge around the mid-belly, and sometimes with two lugs attached on that edge, is termed bo |ψ in the Xiao- baiyang report. The Xiajiadian example, M12:7,30 slightly different from Xiaobai- yang M 15: 1,31 is nonetheless apparently identical to a piece from Longtoushan jt^kih (Chifeng ^ili$, Inner Mongolia) (PI. 47), 32 which may also be of red clay. Zhoujiadi M4:l33 is similar. Two other Upper Xiajiadian sites yielded essentially the same form: Nandonggou ]^J7|p]j^J and Kuntouyingzi i^-^^-f (Ningcheng County TÍÃ, Inner Mongolia, object number 30, coarse siliceous red earthen- ware).34 Xiaobaiyang provides a Late Western Han date.

In the absence of horse-gear and, more importantly, of horse, cattle, or sheep bones in the graves, there is no basis for considering the site nomadic. In addition,

the numerous ash pits and semi-subterranean dwellings, all constructed close toge- ther, indicate heavy settlement. Although nomadic peoples establish base camps, this site appears to have been occupied over a lengthy period of time. The types of animal bones found in the ash pits include pig, which is common to sedentary societies. The use of small appliqué ornaments as seen in Mil, where they seem to have covered a headdress and the chest of the upper garment, is often interpreted as indicative of nomadism. However, in the absence of evidence, this use of orna-

ment must be considered the result of cultural choice independent of the economic base.

Nanshan'gen ]^n I |fl|

Located in Ningcheng County, Inner Mongolia, Nanshan'gen M101 in particular has come to represent Upper Xiajiadian in the minds of many scholars. The tomb is part of a habitation site in the mountains near the village of Nanshan'gen, at an altitude of 525m. Although excavations were made in an area approximately 700m

28 Jin Fengyi, "Xiaiiadian shangceng wenhua," KGXB 1987.2, 183, f. 2:1.

29 Zhangj iakoushi, "Xiaobaiyang," WW 1987.5, 43, f. 3:13.

30 Xiajiadian, KGXB 1974.1, 131, f. 23:8.

31 Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:19.

32 For reference to Longtoushan, see Liu Bing, "Shilun Xiajiadian shangceng," NMGWWKG 1992.1-2,26-33.

33 Zhoujiadi, KG 1984.5, 420, f. 8:1.

34 Nandonggou, KG 1977.6, 373, f. 1:4; Kuntouyingzi, WWZLCK9, 39, f. 39.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 7

χ 200m (from approximately 505m to 535m in altitude), only a small number of disparate points were selected for digging, thereby separating the tombs, ash pits, and dwellings found by, at times, 400m or more from one another. This means that further excavation may possibly reveal settlement across the entire mountainside south of Nanshan'gen village. This same area yielded a total of 71 bronzes in the eastern sector of the habitation site, discovered in 1958. The Upper Xiajiadian remains here again cover Lower Xiajiadian layers; on the surface were found shards of Warring States Chinese ceramics. Nanshan'gen M101 is an unusually large cist tomb: 3.8m long on the surface;

1.8m wide at one end, 2.23m at the other; 2.4m deep, with a slab floor and roof, and walls of large, irregularly-shaped rocks. A single burial, only fragments of the skull remain. Over the bottom of the grave, an ash layer indicates the presence originally of a wood inner-grave structure. The tomb contained more than 500 objects, including 284 semi-spherical appliqué ornaments, ranging in size from 7.5cm to 0.85cm. The smallest pieces are the most numerous (115 measured 0.85cm). Most are plain; some have a central convex bosse. Prominent among the remaining grave goods are Chinese ritual bronzes, many of which are Late Western Zhou/Early Springs and Autumns (i.e, ca. 8 c. B.C.) in date: the ding M101:5, 6, the gui U. M101:ll, and iht fu Ä M101:10.35 The Chinese ge X halberd M101:15-1736 may also date from ca. 8 c. B.C., although the form of ge found at Shangcunling Μψΐφ (Sanmenxia Municipality Ξ^Π^, Henan)37 continued to be in use through the 4 c. B.C.38 These Chinese imports nonetheless provide the ca. 9-8 c. B.C. date assigned to the site by the excavators, after an initial debate in which a date of ca. 3-2 c. B.C. was also advanced. The presence of M 10 1:29, a Late Shang or Western Zhou-type helmet39 at first seems to confirm this date, as does a gold ring M101:126.40 This piece is in the form of a slim, cylindrical length of gold bent into a circle (5. 1-5. 5cm in diameter); the ends broaden and flatten into symmetrical, nearly triangular shapes. From the published photograph of M101:26, it is impossible to tell whether the circle is closed or, as in

the case of identical forms from other sites, open. One such comparative piece

35 Nanshan'gen M101, KGXB 1973.2, 29, f. 4, 3.

36 Nanshan'gen M101, KGXB 1973.2, PI. 7:1-3.

37 Zhongguo kexueyuan kaogu yanjiusuo ^JM^[^Jc'^'''ífW^íf}Ji'' Shangcunling Guoguo mudi IMWfMMÍL. Beijing: Kexue, 1959.

38 See Part II of the present article. I am researching the development of the ge in China for a future article.

39 Nanshan'gen M101, KGXB 1973.2, PI. 5:1.

40 Nanshan'gen M101, KGXB 1973.2, PI. 5:5.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 8 Sophia-Karin Psarras

comes from the Upper Xiajiadian site of Beishanzui JfcULlifô M7501;41 also of gold, it measures 4.6cm in diameter. The Lower Xiajiadian site of Dadianzi ^cfSJ"?" (Chifeng, Inner Mongolia) yielded at least three identical rings, of approximately the same diameter and unspecified metal composition (PI. 10). Two pieces of the same form, in gold, were found in the Shang Chinese tomb of Liujiahe ^J^c/nJ (Pinggu County ^ÍJ-íl·» Beijing Municipality).42 The latter, however, measure 12.5cm in diameter (with a total weight of 173.5 gr), and are identified as bracelets. The small size of the Lower and Upper Xiajiadian pieces precludes such use. Although the rarity of the form initially suggests that the Nanshan'gen and Beishanzui examples must be contemporaneous with the Dadianzi pieces, other evidence in Nanshan'gen M101 yielded a much later date which, coincidentally, is only slightly lower than the discredited 3-2 c. B.C. date originally proposed by some of the excavators.

The hilt of Nanshan'gen M101:35 (PI. 68, second from left), a bronze "notch-

ed" shortsword, closely adapts the décor of a bronze ding from a Late Warring States tomb at Fenshuiling ^>/JC^ (Changzhi County txínM^ Shanxi) M36:2 (PI. 69). 43 The bands of interlocked serpents spanning the ding have been placed vertically over the hilt of the Nanshan'gen shortsword, while the animal masks

which hold the rings on either side of Fenshuiling M36:2 are placed over the guard and pommel of Nanshan'gen M101:35. A similar pattern in less intricate form ap- pears on a bronze shortsword of Xiongnu type I (i.e., a thick, upward-curving guard) at Yuhuangmiao (PI. 71, second from top). The intertwined dragon motif, of course, appears often on much earlier Chinese sites, such as Late Western Zhou/Early Springs and Autumns Shangcunling. However, many of these earlier examples use heavier, wider lines; the form of the animal mask does not seem to be uniform for a given period, particularly in the later Warring States and Han, when several forms coexist. The animal mask on Nanshan'gen M101:35 therefore does not in itself indicate date.44

41 Beishanzui, WWZLCK9, 26, f. 6.

42 Beijingshi, Beijing kaogu sishinian, 37; PI. 8:2. See also Beijingshi wenwu guanlichu J't^' '''i'$C$) ifBlSAh, "Beijingshi Pingguxian faxian Shangdai muzang" db^CrfíT-íí ]A!&M'u}iV£W, WW 1977.11,6, f. 13.

43 Shanxisheng wenwu guanli weiyuanhui, Shanxisheng kaogu yanjiusuo ''']&'&[X$J^Í'M$£jr£ ' lUW&^ilSffiííffr, "Shanxi Changzhi Fenshuiling Zhanguo mu di er ci fajue" ! I iHixín^jvJClIlÊ ktM$k%5~-^k%:M, KG 1964.3, 126, f. 15:1, 2. The same pattern of fine, intertwined dragons is found on a bronze ding inscribed with the characters "Hu Zhi" $/jh from Xunyi County iFJß-M·, Shaanxi, where it is dated simply to the Warring States. Lu Jianguo ^5^(33, "Shaanxi Tongchuan faxian Zhanguo tongqi" ^täWHßiSlffiMftiffi, WW 1985.5, 44-45, f. 1-3.

44 This portion of the present article was presented as a paper ("Chinese Bronzes in the Non-Chinese North") at the annual meeting of the Association for Asian Studies, Boston, 1994. My thanks to David Goodrich and Lothar von Falkenhausen for their questions and comments, which helped me further clarify my thoughts on this research.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 9

Two bronze semi-spherical appliqué ornaments, M101:67, 68 (6.5 and 6.3 cm in diameter, respectively) (PI. 108) are close in form and décor to examples from Xiaobaiyang (PI. 109, left column). The Nanshan'gen pieces have a central, closed disc surrounded by a series of curves like the outline of petals. The interior of the curves is open. A double border decorated with fine radiating lines encircles the whole and is in turn encircled by a wider border of widely-spaced radiating lines. Xiaobaiyang M13:3 (PI. 109, upper left) clearly has an outer border with a décor of widely-spaced radiating lines, but the forms within this frame are essentially indistinguishable. Xiaobaiyang M 13:5 (PI. 109, lower left) also has a border of radiating lines, framing openwork spirals which surround a central, closed space corresponding to the central disc of Nanshan'gen M101:67 and 68. Although the comparison is imperfect, it nonetheless suggests that the pieces from the two sites are related; and, hence, implies a Late Western Han date for Nanshan'gen, since Xiaobaiyang M13:5 corresponds exactly to a Yuhuangmiao piece (PI. 110, right).45 A small, ovoid bronze bell on a long stem (type IV in my terminology),

Nanshan'gen M101:79 (PI. 29) is nearly identical to a piece from Pingyang -ψ^ (Tailai County #5^ä, Heilongjiang), e.g., M126:l, M211:52.46 Another bell

45 Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 31, f. 24:5 (M102:10).

46 Heilongjiangsheng wenwu kaogu yanjiusuo, Yang Zhijun, She Side, Li Chenqi ^jfctl^JCtyjJÇ' iMîHWr ' fäffi' ' A©fë ' ΦΙ&^Τ, Pingyang muzang Τ#Κ#· Beijing: Wenwu, 1990), PI. 39:14, 59:5. This type of bell also ocurs on the so-called Shajing tWt* culture site of San- juecheng Ξΐ/ΟΜ, Yongchang Co. yjCpVU-, Gansu, whose burial structures and rituals are closely linked to those of the non-Xiongnu site of Daodunzi {SJläH^, Tongxin Co. [ü^k-ö-, Ningxia: Ningxia wenwu kaogu yanjiusuo, Zhongguo shehui kexueyuan kaogusuo Ningxia kaoguzu, Tongxinxian wenwu guanlisuo rÄ>C#J^i'WÄW ' *':*Mα£fà¥$ï%'-uff[TM:%lkM^ 'b&XtyffiMIÍU "Ningxia Tongxin Daodunzi Xiongnu mudi" ^'aM*tMÚ^HWLWt, KGXB 1988.3, 333-356, PI. 9-20. The excavators of the Sanyuecheng report a 14C reading of 2680 B.P. ± 115 for the site, on the basis of which they propose a Late Western Zhou to Warring States date for the site while maintaining a Yuezhi attribution. The Yuezhi are not attested in the written record until the Western Han. Objects on Sanjuecheng are related to southwestern non- Chinese, Saka, Maoqinggou, and Xiongnu material. Oracle bones are also attested; ceramics are distinctive. Although the radiocarbon date is evidently disputable, the excavators' attribution may prove accurate. See Psarras, International Relations: Cultural Exchanges between Early China and the "Other Asias" (in preparation). I cannot at this time explain the appearance of type IV bells in this region. See Gansusheng wenwu kaogu yanjiusuo Vnfàfâ'XtyJï^k'ffiWÎU "Yongchang San- juecheng yu Gemadun Shajing wenhua yicun" /Ki^^ití^-J^Mtt^Xitm.^, KGXB 1990.2, 215, f. 10:7. For a discussion of bell types, see Psarras, "Exploring the North." Note that earlier, related forms were found in the Fu Hao tomb (Zhongguo shehui kexueyuan kaogu yan- jiusuo ^lltt^ft^'KYi'W^W, Yinxu Fu Hao mu ßg^aiHS. 2nd ed. [Beijing: Wenwu, 1984], PI. 68:1, 3), as well as at Linzheyu (Baode, Shaanxi) and Cirnokovo (Krasnojarsk, Russia), as cited by Lin Yun, "A Reexamination of the Relationship between Bronzes of the Shang Culture and of the Northern Zone," in: K. C. Chang (ed.), Studies of Shang Archaeology (New Haven: Yale University Press, 1986), 252, f. 51:6, 10, 12. Lin Yun's discussion claims an early date for Upper Xiajiadian on the basis of mirror types, again using Fu Hao material for comparison (Zhongguo shehui kexueyuan, Fu Hao, PI. 68:4, 5), claiming that the non-Chinese mirrors of Fu

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

20 Sophia-Karin Psarras

from Pingyang, e.g., M211:32, M219:75,47 is essentially identical to a spherical bronze bell from Xiaobaiyang M41:5 (PI. 109, second from right), also found on Yuhuangmiao (type IX in my typology). Because of their presence on Yuhuang- miao and Xiaobaiyang, type IX is clearly Late Western Han in date; by extension, it dates type IV to the same period. Because these bells appear to be era-specific, the presence of Nanshan'gen M101:79 suggests a Late Western Han date.

Nanshan'gen M101:70 and 69 are small, bronze decorative plaques in the form of a half-oval and rectangle, respectively, with a central semi-spherical bosse (PI. 28). Both carry a décor consisting solely of a double row of fine radiating lines which serves as a border. The same border, typical of Xiongnu plaques, is found on pieces from the tomb of the king of Nanyue ]§i| (Guangzhou Γ^Η, Guang- dong), d. 122 B.C.,48 including many which are identical to Xiongnu work from the non-Xiongnu site of Daodunzi ffiJJtJ^ (Tongxin County IW^fr-gL, Ningxia),49 dated by the presence of wushu Jî$)f coins to no earlier than 118 B.C.; the same border in a more pronounced version is seen on gold plaques from the Xiongnu site of Aluchaideng MQ^^t (Hangjin Banner Wi&M, Inner Mongolia),50 dating to the Western Han. This border treatment on small plaques appears to be Xiongnu- related and to be specific in terms both of origin and of era.

A Late Western Han date for Nanshan'gen M 101 is further reinforced by com- parisons established through Xiaoheishigou, much of whose material is identical to that of Nanshan'gen M101, including bronze double scabbards with openwork

Hao are identical to Upper Xiajiadian examples. The herringbone décor found on one Fu Hao mirror does occur in the Upper Xiajiadian decorative vocabulary, and mirrors of this type have been found. In these cases, the mirrors in question may have less to do with Upper Xiajiadian than with Jilin cultures. See Zhang Xiying 5feiSÎ5ï» "Shilun dongbei diqu xian-Qin tongjing" iÄifc/folt H!lE;5t^#fêfc, KG 1986.2, 163-172 and Kong Xiangxing ?L#S, Zhongguo tongjing tudian ]']M^^t :''Á (Beijing: Wenwu, 1992), 162, 163. Since the form of the mirror remained constant, our only key to dating is décor. Given the Upper Xiajiadian predilection for the herring- bone pattern, it is not surprising that Upper Xiajiadian mirrors might carry such a décor. In this instance, identical pieces are not contemporaneous, given their context. Note that most Upper Xia- jiadian mirrors seem to carry the L-triangle décor (as at Zhengjiawazi, KGXB 1975.1, 149, f. 10:5; PI. 6:2).

47 Heilongjiangsheng, Pingyang, PI. 58:11, 16.

4° Guangzhoushi wenwu guanli weiyuanhui, Zhongguo shehui kexueyuan kaogu yanjiusuo, Guang- dong bowuguan Γ^Μΰ'8$ΜΆ& ' >'m±^WU^W%Wí ' Γ/ί«ΐβ, Χι Han Nanyue wang mu BfíXS/lSíK- (Beijing: Wenwu, 1991), ν. 2, PI. 134:3 (E21, 22); PI. 123:1 (D164, 71).

49 Ningxia, "Daodunzi," KGXB 1988.3, 333-356, PI. 1-20.

50 Tian Guangjin |1]ΓΊΕ:, Guo Suxin %M:?ffu "Neimenggu Aluchaideng faxiande Xiongnu yiwu" iWÄWß -%&£M(ftte)iUfâ$J, KG 1980.4, 333-338, 364, 368, PI. 10-12; reprinted in Tian/ Guo, Ordos, 342-350. Many of the Xiongnu excavation reports which have been collected in Tian/ Guo, Ordos are more complete in their original publication.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 2 1

triangle décor (PI. 9); bronze ladles (PI. 8); a socketed rectangular bronze axe with a mushroom-shaped counterbalance (chui ti or zhuochui I@cHI) (PI. 99, lower right); bronze helmets; bronze footed cups with a second, inner cup (PI. 7); as well as notched shorts words, leaf-bladed shorts words, and swords or shorts words with a

conical hilt (PI. 70, 85). These objects establish firm contemporaneity for the two sites and, thus, the Late Western Han date of Xiaoheishigou reaffirms that of Nanshan'gen M101. Curiously, this analysis allows a similar date to be suggested for a non-Chinese

site in Liangcheng County W^M3c Inner Mongolia.51 Distinct from other Liang- cheng County non-Chinese cultures, this find includes a bronze decorative plaque depicting a tiger, its head turned toward its hindquarters, surrounded by three goat heads: one is placed between the tiger's head and forefeet; the second, between the tiger's muzzle and hindquarters; and the third, next to the tiger's hindquarters (PI. 22). The goats' muzzles and horns are marked by vertical lines; the tiger's feet, by a series of four spirals radiating from a central point. There appear to be circular markings on the tiger's shoulder and haunch. This piece is not identical to Nan- shan'gen M101:66 (PI. 21), but enough similarities are visible that the Nanshan'gen plaque may be better read, and the Liangcheng piece may be dated. Nanshan'gen M101:66 appears to depict a goat with something below its muzzle extending to its forefeet. The goat appears to have its legs drawn beneath it, in a pose like that of the Liangcheng tiger. The goat's horns are drawn, like those of the

Liangcheng goats, with vertical markings. The goat's head is comparable to those on the Liangcheng piece.

The Nanshan'gen plaque is corroded and difficult to read. Various explanations for the creature beneath the goat's muzzle might be advanced, but iconographically it should be the head of prey. Similar pieces exist at Guoxianyaozi ili?J^ïS also in Liangcheng County (PI. 26)52 and at Pingyang (Heilongjiang) (PI. 25). On Ping- yang M 150: 5, the victim of a winged, recumbent tiger is represented only by a head placed between the tiger's mouth and forelegs, as on the Liangcheng piece above. On Guoxianyaozi M12:2-l, 2, the victim (a goat) is also represented by its head, this time placed between the fore and hind legs of the tiger. The tiger's feet and lower tail appear to be marked with vertical lines, just as are the lower legs or feet of the Nanshan'gen goat. Note that the feet and lower legs of the deer (?) on a small bronze plaque from Xiaoheishigou are also marked in this way (PL 6, left).

51 This find was made at Guoxianyaozi ilíí-u-^-f·. To avoid confusion with the cemetery of the same name (see note 52), I have used the general provenance of Liangcheng Co. to identify this isolated find. "Yao" is written both as ;£ and as #?. Gai Shanlin ftll|#, "Neimenggu Wumeng nanbu fa- xian de qingtongqi he tongyin" 'H^ti &&M$fëWliîWè^W'l , KG 1986.2, 185-187.

52 Neimenggu wenwu kaogu yanjiusuo NMtuXtyj^iiíftfÍffí', "Liangcheng Guoxianyaozi mudi" StWV-ü^fSSÄ, KGXB 1989.1, 57-81, PI. 9-15.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

22 Sophia-Karin Psarras

We may reasonably assume that Nanshan'gen M101:66 is part of the iconography represented at Liangcheng, Guoxianyaozi, and Pingyang, although the tiger seems to have been replaced by a goat, while leaving the head of the tiger's prey intact and in position. Images of death do not seem common to Upper Xiajiadian, where depictions of humans and animals are most often reproductive: copulating animals (e.g., PI. 6, left); animals with pronounced genitália (e.g., PI. 20); a nude woman with enlarged vulva and a man with an erect penis, whose placement on opposite sides of the same shorts word hilt effectively mates them (PI. 19). If no element of the Nanshan'gen plaque has disappeared through corrosion, it is possible to deduce that when the image of predator and prey in its tiger + goat head form reached Up- per Xiajiadian, it was changed to eliminate not the prey's head, or perhaps even the depiction of the tiger's feet, but the tiger's head, which was replaced by a goat head. A similar piece, executed in a different style, occurs at Maoqinggou ^$^ y£j,53 a non-Xiongnu, sedentary culture also located in Liangcheng County. This tin-coated bronze depicts a tiger standing with one foreleg apparently raised to its muzzle (PI. 27). The connection between the foreleg and the tiger's lower jaw echoes the representation of the head of the prey between the tiger's jaw and forelegs. The tiger's feet are marked by vertical lines. Nanshan'gen M101:66 remains indicative of external cultural influences.

The Liangcheng piece is pertinent also because of its context. It was found together with a bronze hu $'? (PI. 23) and a leaf-shaped bronze chariot ornament.

The hu carries a décor reminiscent of that found on the Marquis Yi of Zeng ^&Δ (Sui County ßg-H-, Hubei) bronze gui C-108 (ca. 433 B.C.) (PI. 24).54 In the second band of décor are depicted human figures, walking. Their heads, arms, and torsoes are shown frontally, while their legs are seen from the side. This manner of depiction recalls figures on Chu lacquerware from Changtaiguan ixnN^ Ml (Xinyang Municipality {j=fßH, Henan), dating to the 4 c. B.C.55 This suggests a similar date for the Liangcheng hu, but a Western Han date seems most plausible for the ornamental plaque. The report of the discovery of the Liangcheng objects gives no indication of their geographic context. If they were meant as a single unit, then it seems probable that here, as in Upper Xiajiadian, Chinese bronze vessels were buried by the non-Chinese much later than their date of production.

Although the numerous semi-spherical appliqué ornaments (whose position in the grave is not recorded) in Nanshan'gen M 101 recall Zhengjiawazi M6512, the

53 Maoqinggou is published only in Tian/Guo, Ordos, 227-315. I do not consider it Xiongnu or nomadic, although much material is related to Xiongnu work, particularly the shortswords.

54 Hubeisheng bowuguan MltfäWÜM, Zeng Hou Yi tnu fîfèZM (Beijing: Wenwu, 1989), PI. 58:1.

55 Teng Rensheng Jj^TÍrít· , Chu qiqi yanjiu Σ$>S)BÏ (Hong Kong: The Woods, 1991), 67, f. 72- 73.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 23

other material in the former presents no evidence of nomadism. No animal bones are mentioned in the published report. The sole horse bit (M101:80), a bronze snaffle bit with loops on the outer edge for reins,56 has four points on each end, facing inward toward the horse's muzzle. Such a bit may have been used for training the horse,57 but is at variance with the uniformly smooth bits found in Xiongnu and Xianbei contexts. In addition, various bronze hoes and mattocks were found in the tomb, recalling the importance of agriculture even in the context of a wealthy burial.

A further find of interest is M 10 1:104, 58 a bone fragment into which numerous

holes have been drilled, reported as an "oracle bone." A similar piece reported from the Nanshan'gen habitation site59 is termed a "bone tool." The implement is pierced with a hole at one end, while the surface is covered with a series of small circles with a dot in the center, as if drilling had begun but never been completed. There is no sign of cracking, as from firing.

The site of Nanshan'gen is assumed to be contemporaneous with M101, jud- ging from the presence of a type I ding (H3:25) fragment, supposed to be identical to a Nanshan'gen M101 bronze ding (PI. 54, left). Other material is identical or nearly so to pieces from Zhoujiadi: a bronze bird-shaped ornament with diagonal crossing lines marking its body (Nanshan'gen M4:28) (PI. 18, Zhoujiadi scattered find #3), a bronze "winged" ornament (Nanshan'gen M3:8, Zhoujiadi M2:3).60 The site consists of M2-6, tombs dug in virgin soil which had no overlying cultural layer, but whose fill included Upper Xiajiadian ceramic shards. M7 was covered with 30 cm of earth including Warring States and Han grey ware; M10 and 12 ap- parently had no such cultural layer. Again, as with Xiajiadian, the ground has shifted from erosion as well as from excavation and refill made at the time of the

grave, making stratigraphy uncertain. Two habitation sites were excavated, the second being 10 m distant from Nanshan'gen M101.

Habitation site I, 214 square meters, yielded three layers above a Lower Xiajia- dian stratum (in spodosol) including two ash pits. The surface area of the latter was small, but the layer itself 80 cm deep, at a depth of 50-70 cm from the earth's surface. Above this lowest layer, 10-30 cm from the surface, a 30-60 cm thick layer of brown earth belonging to Upper Xiajiadian included Ml and ash pits Hl-3, 8, 14, 16, 22, 24, 25, and 27. Above this, also 10-30 cm from the surface, was a

56 Nanshan'gen M101, KGXB 1973.2, PI. 11:13.

57 Katheryn Linduff, personal communication, 4.94.

58 Nanshan'gen M101, KGXB 1973.2, PI. 12:3; cf. Nanshan'gen, KGXB 1975.1, 131, f. 14:11.

59 Nanshan'gen, KGXB 1975.1, 137, f. 19:5.

60 Nanshan'gen, KGXB 1975.1, 137, f. 19:11; Zhoujiadi, KG 1984.5, 421, f. 9:9.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

24 Sophia-Karin Psarras

recent layer up to 2.1 m thick in places where gullies had been filled. Above this was cultivated earth.

Habitation site II measured 22 square meters. Beneath the cultivated earth at a depth of 0-20 cm from the surface, a layer of light grey earth, 10-70 cm thick, was assigned to Upper Xiajiadian. It contained some objects (?) and, at bottom, two ash pits, H10, 21. At a depth of 10-60 cm below the surface, a 20-70 cm thick layer of yellow-grey earth was also attributed to Upper Xiajiadian and contained ash pits H15, 23. Beneath this, a layer of loam containing brown earth and spodosol 0.2- 1.0 m thick, at a depth of 0.2-0.8 m below the surface, belonged to Lower Xiajiadian.

The ash pits contained ceramics, both shards and complete pieces in red, brown, or black earthenware, as well as stone tools; animal bones, some with fire

marks from cooking (pig, dog, cattle, sheep, deer, horse, rabbit, fox, birds). All graves were cist structures, some with floor and lid in stone (M3-5, 7), others with a lid but no floor (M10), or a floor and no lid (Ml). The use of stone in M2, 6 is indistinct. M 12 was covered with a stone tumulus, and was the only grave with a shelf. M3-7 and 12 included traces of wood inner-grave structures. The bodies in

M3, 5 were placed in an extended dorsal position; the body in M7 had its legs crossed. All were single burials. Sex was determinable in some cases: M3 and 4 were women ("adult" and "middle-aged," respectively); M5 and 10 were old men, while M 12 was a middle-aged man. M2 contained a child. A greater number of objects seems to have been placed in the larger graves particularly Ml (2.32 m x 0.5-0.75 m), M3 (2.42 m χ 0.8 m), M4 (2.28 m χ 0.64 m), M4 (2.28 m χ 0.64 m), M12 (2.3 m χ 0.75 m). Nonetheless, M5 measured 2.25 m χ 0.7 m, yet con- tained only two shells and a bone bead. There may therefore be some intentional connection between the size of the grave and the wealth of the burial, but there is no evidence that this was systematic. Spinning whorls are associated with women's graves, but knives figure in both sexes'. Grave orientation seems random: Ml, 4, 5, and 7 faced south; Ml and 12, west; M3 and 6, east; and M10, north.

M4 contained a number of small bronze appliqué ornaments, including 160 connected semi-spheres, 17 bird-shaped ornaments, 42 bell-shaped ornaments, and 20 square appliqué ornaments. All were found strewn throughout the grave. None of the graves contained shortswords or, indeed, arrowheads. Evidence points to a closely settled, sedentary population.

Xiaoheishigou /J'M;g7ig

Located in Ningcheng County fÄä» Inner Mongolia, on the flank of the Nan- shan l§ll[, the site of Xiaoheishigou has been published in Neimenggu wenwu kao- gu 3 (1984) as a group of more than 40 objects and in Wenwu ziliao congkan 9 (1985) as tomb M8061, containing 48 objects. The second publication mentions discoveries made in 1973 and 1980 on the site, but lists in each case M8061 as the

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 25

source.61 At exhibits given by the Archaeological Institute and by the Municipal Museum of Hohhot in 1992, I was able to photograph objects from Xiaoheishigou, totalling at least 75 pieces and, in addition, two pieces of turquoise and a long necklace of black and white (bone or ceramic?) and turquoise beads (cf. PI. 124, 125). Despite this doubt as to the specific provenance of the material: tomb(s) or habitation site, we have no reason to suppose that the material does not constitute a contemporaneous whole. Chinese archaeologists date the site with reference to Nanshan'gen M101, thus producing a Late Western Zhou/Early Springs and Autumns date (ca. 8 c. B.C.). Such a date at first appears substantiated by the site's Chinese ritual bronzes, including a zun j^l (PI. 15), gui J£ (PI. 3, 4), and lei U, as well as other apparently Chinese forms such as helmets (PI. 123), yue f$ axes (PI. 91, upper row, center), and chui f|§ or zhuochui I$jcf|| axes (PI. 99, lower right). The latter was also found in Nanshan'gen M101 and will be discussed, with the yue, in Part II, below. The zun (PI. 15), with a widely flared mouth, a slightly flared foot, and a broad

central band slightly wider than the rest of the body, with two, thin raised bands above and below it, is nearly without visible décor. An example with the same shape and same broad band framed by thin bands was excavated from the Baoji ^i% (Shaanxi) cemetery of the state of Yu ?ft, BZM8:7.62 The Baoji piece (PI. 16) carries bands of décor within the upper and lower edges of the broad central band;

these are lacking on the Xiaoheishigou zun, but in the corresponding areas, cor- rosion notwithstanding, there is some suggestion of similar patterning. The remnants of décor on the Xiaoheishigou zun now appear as if incised. The piece should date to the same period as BZM8:7, Early Western Zhou, or ca. late lie. B.C. A related piece, a ceramic zun from the Yan 5πΕ site of Liulihe 3E^ï^ynJ (Fang- shan County J^ÜJ-M-, Beijing Municipality) (PI. 17), also dates to the Early Western Zhou.63 Contemporaneously, elaborate versions of zun with four equally- spaced flanges extending from the lip to the foot are known from both Baoji and Liulihe.64 The style of the Xiaoheishigou zun therefore does not indicate a precise geographic origin within China.

61 Xiaoheishigou, WWZLCK 9, 44, 52. For additional material, see Liu Bing, "Shilun Xiajiadian shangceng," NMGWWKG 1992.1-2, 26-33.

62 Lu Liancheng /^j£lÄ, Hu Zhisheng #$?#., Baoji Yuguo mudi Sâl^HISÉ (Beijing: Wenwu, 1988). BZM refers to the Zhuyuangou ttlM/^J group of tombs; BZM8 is dated by the excavation report to the late Chengwang )& ^H/early Kangwang Jj&l period (Baoji, vol. 1, 267).

63 Zhongguo shehui kexueyuan kaogu yanjiusuo, Beijingshi wenwu gongzuodui ψΗίΐ^ίψ^' $WkW?Zß'' ' ^tt''ib'XW:''MK, "1981-1983nian Liulihe Xi Zhou Yanguo mudi fajue jian- bao" 1981-1983^íjfLW"M^!!tE^iÈAff,fa"í[í, KG 1984.5, 405-416, 404, PI. 1-4.

64 Wen Fong (ed.), The Great Bronze Age of China (New York: Alfred A. Knopf, Inc., 1980), 216, PI. 42; 223, PI. 51.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

26 Sophia-Karin Psarras

A lei found on Xiaoheishigou is comparable to an example from the Early Western Zhou (ca. late 11 /early 10 c. B.C.) Beidongcun (storage pit 2), Liao- ning.65 Although the lid is missing, the Xiaoheishigou lei has a relatively high base and angular handles ornamented with equally angular animal heads; rings hang from the handles. These animal heads appear deer- or goat-like, with two distinct, tubular horns, rather than a single flat flange behind the head, as on the Beidongcun lei handles. However, on the Xiaoheishigou lei, the angle from neck to shoulder is flatter than on the Beidongcun piece and more closely comparable to a late 10/early 9 c. B.C. ling ft& said to be from Qishan jfcfcii], Shaanxi.66 This piece carries a décor of horizontal scales about the shoulder, as does the Xiaoheishigou lei, although the latter retains a band of elongated animal mask décor on the lower neck which is not present on the Qishan vessel. Since the scale pattern is so common in the Late Western Zhou, and since the form of the Xiaoheishigou lei is mid- way between the Early and the Late Western Zhou versions, an approximate date of ca. 10 c. B.C. would therefore seem reasonable for the Xiaoheishigou lei. As with the zun, no specific geographic provenance may be advanced.

Xiaoheishigou also yielded two gui dating to the Late Western Zhou or Early Springs and Autumns. One (with no lid), which seems to be identical to a gui from Beishanzui (M7501:19), also without a lid,67 has handles in the form of animal heads with upright coiled ears (PI. 3). The handles of the second gui (PI. 4) depict an animal head with leaf-shaped ears which lay flat; beside that portion of the handle which descends from the animal's mouth there curls a tongue-like spiral. Both types occur at Shangcunling (e.g., M 1 8 10: 6 and M 1052: 150). 68 It is note- worthy that Xiaoheishigou also yielded a hollow bronze fitting, broken along one edge, in the form of a beast with erect ears in the form of coils (PI. 6, right). The animal's features are generally flatter than those of the gui handle, but the piece is nonetheless Chinese in form.

These Chinese bronzes have been used to justify a date of ca. 8 c. B.C. for Xiaoheishigou, which at least one of the site's two reported helmets (PI. 123) would substantiate. Bronze helmets with a square loop at their crest and one (type I) or two (type II) loops at their base were found not only on Xiaoheishigou (I), but also Wafangzhong Κ&ή* M791:188 (II), Beishanzui M7501:13 (II), Wujintang -Öj<n£ iM M3, and Nanshan'gen M101.69 Late Shang examples may be ornamented

65 Fong, Great Bronze Age, 21 1 , 229, PI. 55.

66 Fong, Great Bronze Age, 236, PI. 61; 247.

67 Beishanzui, WWZLCK 9, 25, f. 3: 1 .

68 E.g., Zhongguo kexueyuan, Shangcunling, PI . 43 : 3 .

69 Wafangzhong, WWZLCK 9, 27, f. 5:3; Beishanzui, WWZLCK 9, 25, f. 3:2; Wujintang, KG 1960.5, 7, f. 1:2; Nanshan'gen M101, KGXB 1973.2, PI. 5:1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 27

with a cast décor over the helmet surface;70 one from Gaohong fijeX, Liulin County $P#H-, Shanxi, like the Upper Xiajiadian finds, has no décor and a square loop at the crest, but the base is perforated with a series of circular holes.71 One Western Zhou example from Changping p^, Beijing Municipality, has no square loop, but an openwork flange over the crest; it has neither loops nor perforations at the base, and the sides of the helmet flare slightly outward.72 A final example from an unspecified site in Zhaowuda Banner H^i^üjjtt, Inner Mongolia, is without décor, with straight sides, and a square loop on the crest. The presence of loops on the base is not ascertainable.73 This piece is dated by Chinese archaeologists to the Eastern Zhou (probably ca. mid-8 c. B.C.); it appears closely comparable, if not identical, to the Xiaoheishigou, Beishanzui, and Wafangzhong examples, and may in fact come from an Upper Xiajiadian site. As with Xiajiadian and Nanshan'gen, however, other material on Xiaoheishi-

gou provides a much later date. This material incudes a burnished black ceramic dou with a long stem and shallow bowl (type HI in my typology), a form known in the state of Yan through the Late Warring States period (e.g., Xiadu T*^, Yi County ^VH-, Hebei, site 22, Tl:3:15; site 13, H8:10, T2-2:7),74 although it is also known in Middle Warring States Zhongshan (e.g., Sanji Ξ'ϊΚ, Pingshan County -flilß-, Hebei, habitation site 2).75 The form also occurs in Middle Warring States Zhao (e.g., Baijiacun 'ftlicJt, Handan tß$[5, Hebei, M40:4),76 and in Yuhuangmiao M61:4 (PI. 50). At earliest, the Xiaoheishigou piece therefore probably dates to the second half of the Warring States (ca. mid-4 c. -3c. B.C.). A bronze dou (no type number assigned) with a short, somewhat broad stem and a wide, moderately deep bowl, found on Xiaoheishigou, is closely comparable to a ceramic dou from a culture in Jilin, extending from the central plain to the south-

70 Cheng Dong JjJò/jí, Zhong Shaoyi W/Pzl, Zhongguo gudai bingqi tuji Ί'Ξώ^Α-^Η^ί (Bei- jing: Liberation Army Press, 1990), 40-41. This book provides good photographs for a number of Upper Xiajiadian shortswords.

71 Cheng/Zhong, Gudai bingqi, 42, no. 2-121; reference: KG 1981.3.

72 Cheng/Zhong, Gudai bingqi, Color PI. 5, upper left, no. 3-62.

73 Cheng/Zhong, Gudai bingqi, Color PI. 12, no. 4-157.

74 Hebeisheng wenhuaju wenwu gongzuodui W-'t'ÊXÎkM)Xtyj T.fËPA, "Yan Xiadu di 22hao yizhi fajue baogao" $ |> "Wl^22 JlW±ÍtMf'íff* KG 1965.11, 562-570, PI. 5-6; Hebeisheng wenwu yanjiusuo MJt'à'XWWiïW, "Hebei Yixian Yan Xiadu di 13hao yizhi di yi ci fajue" ínJJt^-M- m^fFAWi&itflÍ-lkfòm, KG 1987.5, 414-428.

75 Hebeisheng wenwu yanjiusuo MJté'&WW?E0f» "Hebei Pingshan Sanji gucheng diaocha yu muzang fajue" H1W-'U=WlUW$&±j1&W!&M, KGXJK 5 (1987), 157-193, PI. 19-22.

76 Hebeisheng wenhuaju wenhua gongzuodui WM'&XitMjXit T.ff^A, "Hebei Handan Baijiacun Zhanguo mu" HMW^ti'^UtM^ KG 1962.12, 613-634, PI. 1-4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

28 Sophia-Karin Psarras

central portion of the Dishan-Qiuling fäll] / £;|&.77 Most information pertaining to this Jilin culture is currently unpublished, but burials are said to include single burial first inhumations, single or multiple burial second inhumations, and a small number of fired graves (located on the Jilin plain), and the same burial customs in

cist graves (on the mountainsides). The single radiocarbon date reported is 2160 B.P. ±70 (210 B.C. ±70). The types of burial rituals used here, as well as perhaps other ceramic forms, provide important comparisons for Upper Xiajiadian. Xiaoheishigou also yielded a notched shortsword with a double spiral pommel

essentially identical to hilt and pommel examples from Yuhuangmiao, although the Yuhuangmiao guard and blade forms are Xiongnu, not notched (PI. 70, upper right; PL 74, 4th from left and 2nd from right; PL 72, 3rd from left). Since Yuhuangmiao dates positively to the Late Western Han and because this type of shortsword cannot be shown to occur over a broad range of time within the context of present-day China, the Xiaoheishigou notched shortsword must date to approximately the Late Western Han. In addition, a Xiaoheishigou straight-bladed shortsword with a straight bar guard, a broad rectangular hilt, and a straight bar pommel provides an even later date (PL 81, left). The form is simple but occurs only rarely in what is now northern China. Although the Xiaoheishigou example is bronze and carries a filigree-like raised line décor of standing deer, with haunches and shoulders marked by spirals, on its hilt, the form is the same as that of three

other shorts words found in other contexts. These are: Yinniugou tfc^H&J, Liang- cheng County W^ÂM:, Inner Mongolia, Ml:2 (iron, no décor) (PL 84);78 the Xiongnu site of Hulusitai Pf ^ÄTyfc, Center-Rear United Banner of Wulate ELjlfc

77 Xingjiadian Μ'Μΐίΐ, Nong'an County ^£-£1-, Jilin, see Wenwu X$J (ed.), Wenwu kaogu gongzuo shinian 1979-1989 'XW%iiT*W·^' 1979-1989 (Beijing: Wenwu, 1990), 70; PI. 70, center of center row, dated to the Warring States with a radiocarbon date of 2160 B.P. ± 70 (210 B.C. ± 70) Cf. M2:2, Ml:2 of Jilinsheng wenwu kaogu yanjiusuo pí#^ÍC#J^^}f^0f, "Jilin Dehuixian Beiling mudi diaocha yu fajue" YfW&M&MvMEWiä^j!kM, KG 1993.7, 600, f. 4:2, 3, dated to the transition of the Late Springs-Autumns and Early Warring States. The term Xituanshan H03ll|, which covers both these sites, is both the general name and the name of the earliest phase of the culture. I have followed the practice of using the term in the general sense. Chinese archaeologists divide the culture into three periods: a) Xituanshan and Xingxingshao SSnB; b) Houshishan %%fi''' and Changsheshan ixfâlj; c) Tuchengzi ±ftB1 and Dahaimeng ^fftínL, supposedly dating from the Early Western Zhou to the Qin. Burials are cist graves built from slabs and are said to include single burial first inhumations, single or multiple second inhumations, and a small number of fired graves (on the Jilin plain). The same customs, with an emphasis on multiple burial second inhumations and grave firing in cist graves, occur on mountainside sites. See Dong Xuezeng 35^a, "Shilun Jilin diqu Xituanshan wenhua" iÄifc rf# ÄEHEIlll^fc, KGXB 1983.4, 407-426. At least some of these dates are too high, but will be examined in Psarras, International Relations.

78 Neimenggu zizhiqu wenwu gongzuodui l^MÎiïa'iùK'XW^ff^k, "Liangcheng Yinniugou mu- zang qingli jianbao" WMtX^BWWMÍniM, NMGWWKG 3 (1984), 26-31, 25, PI. 7-8.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 29

ffîtpJntffcrîM, Inner Mongolia, Ml:3 (bronze, with hilt décor) (PL 80, center);79 and habitation site 7 at the Chang'an armory, 7:2:2, dated through coins found on the site to no earlier than the reign of Wang Mang (including regency, ca. 6-25 A.D.) (PL 83, right).80 The last example is of iron, with no décor. In the case of each of the three comparative examples, the center of both hilt and blade is marked

by a low ridge; in each case, the guard, rather than being straight across, angles downward from the center of the hilt. This is the Xiongnu type I guard, in my terminology; the Xiaoheishigou piece conforms to the Xiongnu type II (straight) guard. Differences in material (iron vs. bronze) are not of great significance in dating these weapons, since shorts words may occur in iron on sites where they also occur in bronze. In the non-Chinese north, iron pieces are not likely to be highly decorated. It must be noted that Hulusitai Ml:3 is highly decorated, with vertical banding over the pommel, and two rows of horizontal banding over the hilt, separated by a vertical ridge. This division of the hilt into three vertical rows is common on Yuhuangmiao and the Yuhuangmiao culture site of Xiaobaiyang. The relative rarity of the T-form shortsword with Xiongnu guard allows for the use of

the Chang'an example as a marker of positive date. These shortswords may therefore be considered to date to the very late Western Han and Wang Mang eras. Shortswords will be discussed in greater detail in Part II, below. Although no good external comparisons presently can be found for the Xiaohei-

shigou bronze double-cupped dou (PL 7) or bronze ladle (PL 8), these pieces demonstrate contemporaneity with Nanshan'gen M101,81 further substantiated by the bronze double scabbards of Xiaoheishigou (PL 9), Nanshan'gen M101, and Beishanzui (M7501:12). The Xiaoheishigou piece, with an openwork triangle décor, is essentially identical to Nanshan'gen M101:27, while the décor of open- work diamond shapes on Beishanzui M7501:12 is echoed on a single scabbard flanked on each side by openwork triangles on Nanshan'gen M101:26.82 A W- shaped bronze object from Xiaoheishigou, the outer ends and lower curves of which are accentuated by spherical bells, with two attachment holes on the upper

79 Ta La Jgfö, Liang Jingming ^ï?M, "Hulusitai Xiongnu mu" Wf^WfX^M'g,, WW 1980.7, 11-12; Tian/Guo, Ordos, 223-226. Note that I consider the first appearance of the Xiongnu in present-day northern China to date to no earlier than the very late 4 c. B.C. or ca. 300 B.C. See Psarras, Han and Xiongnu: A Reexamination of Cultural and Political Relations (forthcoming).

80 Zhongguo shehui kexueyuan kaogu yanjiusuo Hancheng gongzuodui Ψ03}±^#^|5^#τ!}*§ί ^WiXíCT-fWÀ, "Han Chang'ancheng wuku yizhi fajuede chubu shouhuo" 'iXixïcÎAA"i¥iS±ll: UMffiW&i&tR, KG 1978.4, 261-269, PI. 9-11.

81 See Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:3, 4.

82 Double scabbards existed in Ba-Shu during the Warring States, but in a much different form. Sichuansheng bowuguan Hjl|^M#JtÈ, Ba-Shu qingtongqi ESuW^S (Chengdu: Chengdu Publishing House, n.d.), 85; PI. 97. See also Li Xueqin φ'-^ÜJ, Qingtongqi (xia) ^^f§ ( Τ ) (Beijing: Wenwu, 1986), 138; PI. 159, dated to the Middle-Late Warring States.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

30 Sophia-Karin Psarras

curve (PI. 30, top), is probably a carriage component used as a harness guide. It may be compared to a piece from Wudaohezi TSjUffl-Jr, Ml: 15, of similar size and shape (PI. 30, center). The Xiaoheishigou piece measures 13.4 cm χ 14.0 cm, while Wudaohezi Ml: 15 is 15 cm in height. A double-headed snake carriage ornament from Liangjiayingzi ^MM1^ M807 1:220 is very similar in form to the Xiaoheishigou piece, but considerably smaller (6.2 cm χ 8.8 cm) and, judging from the five attachment loops on the reverse side, probably had a purely ornamental function (PI. 30, bottom). Two cheekpieces (components of a horse bit), Xiao- heishigou M8061:210, also bronze, are in the form of thin rods terminating in a horse's head at one end (PI. 11). Four loops extend from one side of the rod. On one cheekpiece, one double spherical bell (rattle) remains; on the other, two re- main. These pieces may be compared, if somewhat distantly, to Han examples. While Eastern Han work such as Shaogou ]£$] (Luoyang ?&K, He'nan) M1038: 11, in bronze, has changed the rod into a long spiral to which Xiongnu birdheads (i.e., with a curved beak and an ear) have been added (PI. 12), 83 some Western Han cheekpieces are more comparable to Xiaoheishigou M8061:210. These Western Han pieces are symmetrical and have a more pronounced curve than Xiao- heishigou M8061:210, lending them something of an S-shape; the openwork flange at each end of the piece bears some affinity to the rattles on Xiaoheishigou M8061:210 (PI. 13). 84 Although these similarities are too distant to allow for cross- dating, they may indicate cultural influences from the same, presently unidentifi- able source.

Xiaoheishigou is one of the relatively small number of Upper Xiajiadian sites to yield horse-gear: two cheekpieces and at least one bronze snaffle bit (PI. 120, top). A large bronze fitting suggests the presence of a carriage or chariot, as does a bow- shaped bronze pipe with two loops on its upper edge, which is probably a yoke (PI. 122). There is, however, no indication in the published reports of animal bones in the grave(s). Located on the flank of the Nanshan, the site's context is that of Nanshan'gen. Nomadism is therefore not in evidence.

Zhoujiadi j^^cife

Located in the Aohan Banner ffiiXM, Inner Mongolia, the site of Zhoujiadi con- sists of 54 graves in an area measuring 77 m north-south and 18 m east- west.

83 Luoyangqu kaogu fajuedui ^ΠΕ#^βΛ, Luoyang Shaogou Hanmu fôPH^&fêlîXSS (Bei- jing: Kexue, 1959), PI. 49:3 (M1038:ll), dated in the report to the Late Eastern Han.

84 Luoyangqu, Shaogou, PI. 49:5, dated by the report to the Late Western Han/ Wang Mang; Da- baotai Hanmu fajuezu jçMÉVlXMlfcWM* Beijing Dabaotai Hanmu jfcîfcfcfë&SUS (Beijing: Wenwu, 1989), PI. 89:9, ca. 45 B.C.; Guangzhoushi wenwu guanli weiyuanhui, Guangzhoushi bowuguan r^W'SCWû'M^^Ù · Γ^Τ|ί|Α/ί&, Guangzhou Hanmu Γ'ΜΫ&Μ (Beijing: Wenwu, 1981), vol. 2, PI. 66:11, Middle Western Han.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 3 1

Because of its location on a promontory on the northern bank of the Laoha River ^nft, much mountain runoff passes through the cemetery and has disturbed the soil. Beneath a surface layer of yellow sand, a 0. 10-0.51 m thick layer of dark grey earth was uncovered in which were found shards of Upper Xiajiadian type ceramics. Below this layer was virgin soil. The tombs opened into the dark grey earth layer, and also broke open on the surface of the ground. The extent of erosion is therefore clear. All graves were rectangular pits, oriented northwest- southeast, and placed in rows from the southwest to the northwest. Most had traces of inner-grave wood structures which are reconstructed as a plank coffin placed on two logs, which rest on the bare ground. There are no cist graves. Graves measured between 1.25-3.10 m in length, 0.6-1.5 m in width, and 0.52-2.05 m in depth. M45 included birchbark to the right of the head of the deceased. M16 was a joint burial second inhumation. The grave was fired, then filled in with soil. In M45, M2, and M43, a hemp cloth to which bronze semi-spherical appliqué ornaments had been sewn covered the face of the dead. In M45 and M2, a large oyster shell covered this cloth. Oyster shell ornaments were present in M 16, M2, and M45. Analysis has determined the age and gender of the dead in M45 (a 12-13 year old boy), M43 (a six year old child), M2 (a 16-17 year old female), and M 16 (two 6-7 year old children). The boy in M45 was apparently killed by an arrow shot into his right eye, where the arrowhead remained lodged. The girl in M2 was buried with a spinning whorl, which characterizes female burials on the site. Male

burials, even those of the children in M 16, are characterized by the presence of arrowheads, often one in bone and one in bronze (see PI. 2).

According to the preliminary report, most graves contained one piece of hand- built ceramics of various colors, and bronze semi-spherical appliqué ornaments. In some cases, traces of clothing and face coverings were preserved. Most tombs apparently included the bones of sacrificed animals: dog, cattle, and horse skulls, as well as horses' hooves. M45, for instance, included a horse skull and four hooves placed in the grave, outside of the coffin. Here, then, is evidence of nomadism, at least in burial ritual. Does Zhoujiadi indeed belong to Upper Xiajiadian? Of all the sites which may belong to this culture, only Zhoujiadi and Zhengjiawazi include the burial of animal bones. Both sites also follow a ritual of placing bronze plaques or oyster shells over the face or body of the deceased. Both include only pit graves. Does this point to a separate, although related, culture?

Zhoujiadi is dated by Chinese archaeologists to the Western Zhou or Springs and Autumns period. Though there is little comparative material, the bird-shaped ornaments #2, 3 link the site to Nanshan'gen M4:28 (PI. 18); the winged ornaments such as Zhoujiadi M2:3 and the knives with a serrated lower edge near the handle (M45:53, M2:l) provide links with other sites including Xiaoheishigou and suggest that Zhoujiadi may reasonably be assigned to approximately the same period (PL 89, 90).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

32 Sophia-Karin Psarras

Zhengjiawazi ^ic/ÉÍ

Located 5 km southwest of Shenyang tfcßB, Liaoning, the site consists of 14 tombs, called Area 3, which totals approximately 20,000 square meters. Area 2,500 m to the northeast, was the location of a single shorts word fine. On Area 1, 500 m to the

north, a group of 27 bronzes, including shortswords, was found. The 14 tombs of Area 3 are separated into two groups: a cluster of 12 burials (relatively small) and, 80 m to the southwest, two larger tombs. All graves are rectangular pits. Only the two larger tombs, M6512 and M659, are discussed in the published report. M6512, oriented to the west (280°), was a large rectangular pit grave, 5 m

long, 3 m wide, and 1.4 m deep. The disintegrated remains of both an inner and outer coffin were discerned, but no reconstruction of their shape presented. The floor of the inner coffin was covered with hemp matting. Grave goods (797 objects including horse bits and cheekpieces) were placed between the inner and outer coffin; cattle bones were placed on the east side of the pit, outside the coffins. The tomb belonged to an elderly male. He was buried in extended dorsal position with bronze mirrors placed above his head, chest, pelvis, knees, ankles, and feet (PI. 1). The seven large bronze discs placed to the right (south) side of the body lie with various objects assumed to be horse-gear: conical bronze pieces with openwork triangle décor thought to have held plumes on a horse's head, horse bits and cheek-

pieces, tubular ornaments, passants (strap guides), semi-spherical bronze appliqué ornaments. There is gear for a team of four horses. The dead man was also equipped with three shortswords, which contradict Jin Fengyi's theory of Upper Xiajiadian shortsword development: leaf-bladed, M6512:l and M6512:33 both had wood hilt units and regular crowns (iron-ore counterweights), while M6512:2 had a wood hilt unit and an irregular crown (PI. 66). All were found with short, rectanglar bronze pieces which are the foot of the scabbard, originally made of wood. All three carry a décor of triangles, rectangles, and L-shapes (L-triangle décor). A broad, U-shaped bronze scabbard with border design of heavy diagonal lines which was originally fitted with a leather backing, housed a knife; a bronze plaque with a décor of symmetrical curves, probably also attached to leather, was designed as an axe-scabbard. In addition, to the left (north) of the deceased were found bone ends of a bow estimated to have measured 1.0 m in length; the bow casing, decorated with small semi-spherical bronze appliqué ornaments placed in three rows on one side, two rows on the reverse (130 plaques total), can be recon- structed. The casing may have held two bows. With the bow and casing were found 71 trilobed bronze arrowheads. 98 bilobed bronze arrowheads were found

next to the lower left leg of the deceased. The grave also contained three earthenware jars made of grey argilaceous clay, hand-built, but with the mouth smoothed on a wheel. The pieces were fired at a relatively high temperature, and have a black or grey exterior.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 33

M659, also the grave of an elderly man buried in extended dorsal position, his head turned slightly to the left, measures 1.75 m in length, 0.5 m in width, and 0.95 m in depth. The tomb is oriented to the west (298°). On the south side of the grave, the leg(s) of a sacrificed cow have been cut to fit on a shelf. There are apparently few grave goods, including a bone shorts word (?), a bone ring, and a grey earthenware jar like the pieces in Zhengjiawazi M6512. The northwest corner of the grave was broken into by an ash pit of late date; a layer of ashes covered the grave.

The site is dated by the Chinese report to ca. 6-5 c. B.C. (Late Springs- Autumns/Early Warring States) on the basis of three Western Han jar coffin burials which break into the site, although it is not clear if they actually break into Upper Xiajiadian graves, and the Western Han cultural layer covering the site. An early date is thought to be corroborated by the wood-hilted shorts words in Zhengjiawazi M6512. There is, however, no reason to assume such an early date. Since the shorts words are like those of other Upper Xiajiadian sites, as is the L-triangle pattern used on the scabbards and mirrors, the site may be indirectly dated to the same time as the general dates of Upper Xiajiadian, derived from positively-dated sites of the culture. The Western Han layer may be nearly contemporaneous to the graves. It will be seen that Late Middle Warring States graves of the state of Yan coexisted with Upper Xiajiadian tombs without intruding on them at Shi'ertaiyingzi. If the Western Han tombs at Zhengjiawazi cut into non-Chinese graves it may be that the Han deliberately sought to destroy these tombs. The Han presence may be the result of a colonization program intended to secure non- Chinese territory or territory formerly of mixed settlement.

At first glance, Zhengjiawazi M6512 appears nomadic: there is much horse- gear and a multitude of nomad- type objects (semi-spherical appliqué ornaments, arrows, shorts words). There seems to be an absence of agricultural tools, and animal bones (cattle) are indeed included in the grave. However, cattle are not diagnostic of nomadism since they are common to sedentary cultures, particularly when those portions of the animal included in the tomb seem to constitute a meal for the dead, not the reconstruction of animal-based wealth, as represented by the skull or hooves of the beast; further, the horse-gear in the tomb belongs not to horses used for riding, but to those driven as a team. The man buried in M6512 wore boots ornamented with semi-spherical bronze appliqué ornaments and carried tools housed in similarly ornamented casings. Much of this material belongs in a nomadic context, but is not exclusive to it. There is, in short, no conclusive evidence in the tomb that points to nomadism, nor is there conclusive evidence pointing to a culture distinct from that of Upper Xiajiadian.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

34 Sophia-Karin Psarras

Shi'ertaiyingzi hUft^^

The site of Shi'ertaiyingzi, Chaoyang J^OßH, Liaoning, was originally thought to consist of three graves discovered in 1958 during an irrigation project. The graves were built on a height at the foot of the Bo mountains ffilil, overlooking a dry riverbed; the river Daling ^CÎ^M presently runs near the site. Ml was excavated, the objects from M2 were salvaged, and two pieces from M3 retrieved before the site was destroyed. Material from M2 is said to be like that of Ml. From M3, only an iron-ore shorts word crown and a bronze mirror were recovered.

Ml was a cist grave formed of unworked rock and pebbles with a slab lid, pebble floor, and a "door" of an upright slab measuring 0.7 m by 0.6 m (0.9 m thick). The interior measured 1.18 m long, 1.0 m wide, with a height of 1.2 m; the top of the tomb was 2.0 m below ground level. The structure was oriented to the west and contained a wood inner-grave structure and rush matting (as at Zheng- jiawazi). The grave contained a man, about 50 years of age, and a woman, 40-50 years old, both in extended dorsal position with their hands crossed on their chest. The report notes that all their teeth were intact, although their bones were dis- integrating. Bronze mirrors were found above the head and below the feet of the man. Two leaf-bladed shorts words with regular crowns (originally with wood hilt units), two bronze arrowheads, and a bronze axehead also lay about the man. A clay spinning whorl lay near the woman, as did horse bits, cheekpieces, and harness ornaments. A bronze fishing hook and stone sinker were found between the man and woman. The material included wide use of the L- triangle décor and a bronze ornament in the form of a human face.85 This face is essentially the same as that which appears on the Nanshan'gen eastern sector habitation site shortsword with the nude made-nude female pommel (PI. 19) and on a bronze fitting from Xiaoheishigou (PI. 14). The Xiaoheishigou face has strongly horizontally oval eyes; the Nanshan'gen faces have round eyes set in a somewhat ovoid eye socket; the Shi'ertaiyingzi piece has round eyes in round sockets. All, however, have a deeply marked nose and a thin, broad mouth. Although this face does not occur often in published Upper Xiajiadian material, it appears to be characteristic of the culture, as does the common L-triangle pattern.

The Shi'ertaiyingzi horse bits and cheekpieces indicate the use of the horse (we may assume for riding), but there is no other sign of nomadism. The fishing gear is conspicuous and, while not incompatible with nomadism per se, must be seen as forming part of the context within which the horse-gear is to be evaluated, lending weight to a sedentary attribution. The excavation report dates the site to the Late Springs- Autumns/Early Warring States period. Because of its similarities to other Upper Xiajiadian sites, Shi'ertaiyingzi must be assumed to be essentially contem-

85 Shi'ertaiyingzi, KGXB 1960.1, 67, f. 4:2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 35

poraneous with other sites. Other excavations at Shi'ertaiyingzi support this idea and help provide an upper horizon for the culture. A brief notice in Wenwu kaogu gongzuo shinian 1979-1989 announces the

excavation of a cemetery of more than 100 graves.86 Of these, 20 belonged to Middle to Late Warring States era Yan; more than 50 were "like the Yan tombs in structure" but contained red siliceous hand-built pottery (1-2 pieces per tomb), rather than Warring States Chinese pottery assemblages, as well as leaf-bladed shorts words with crowns, bilobed arrowheads, bronze semi-spherical appliqué ornaments, passants, and other bronze or stone ornaments and implements. Dog and goat skulls were also present. These graves therefore seem to correspond to Upper Xiajiadian at least because of the characteristic leaf-bladed shorts words, although these weapons could have been used by neighboring cultures in some instances. The dog and goat skulls are a mark of nomadic burial rites, but could indicate the influence of nomadic cultures rather than actual nomadism; they could likewise point to a nomadic fringe of Upper Xiajiadian. These tombs, because they neither intrude upon the Yan burials nor are cut into by the latter, may be presumed to be contemporaneous with the Shi'ertaiyingzi Yan graves: that is, dating to the Middle to Late Warring States. The only object from the site illustrated in Wenwu kaogu gongzuo shinian

1979-1989 is a ceramic hu whose décor matches that of the ceramic assemblage of King Cuo of Zhongshan ^iJj-H«, ca. 309 B.C. (PI. 31). 87 The Cuo assemblage is black pottery with décor created through the burnishing or roughening of the surface: therefore, polished black lines on a matte ground, arranged in bands over a polished black surface. The Shi'ertaiyingzi Yan piece is not described in the text and the photograph published is unclear; the décor may be incised. Much Yan pottery is grey ware. Although the form of the Zhongshan hu is distinct (a small base forming a stem beneath the belly, and a narrow neck flaring out to a much wider mouth), this form seems to be due to regional and not chronological differences. The décor consists of bands of spirals and diamond or triangular shapes filled with zigzag bands. These or closely related patterns are found on other Middle Warring States pieces from Chu and Yan: Yun County ßß-M-, Hubei pieces88 have a red and white painted décor, while Yan examples from north of Huairou |>;F^, Beijing Municipality,89 are greyware (no further description is pro-

86 Wenwu, Wenwu kaogu gongzuo shinian, 63-64; PI. 6, lower left.

87 Zhongshan: tombes des rois oubliés (Paris: Association française d'action artistique, 1984), no. 78-88.

88 Zhongguo shehui kexueyuan kaogu yanjiusuo Changjiang gongzuodui ψΕίΐ^ίΨ^'&ΐΐ^'ώ^ί ÏÎÏÏÏixïC''MK, "Hubei Yunxian Dong Zhou Xi Han mu" MM$-&'fcMW)Ü&, KGXJK 6 (1989), 143-174, PL 18-26.

89 Beijingshi, Beijing kaogu sishinian, 53-56, Pl. 16 (state of Yan ^t).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

36 Sophia-Karin Psarras

vided) (PL 32). All date from ca. 4 c. B.C. This, combined with the Zhongshan example, suggests that this décor is datable and that, therefore, a late 4 c. B.C., or 309 B.C., date is an appropriate upper horizon for Shi'ertaiyingzi. Wenwu kaogu gongzuo shinian stresses that, while Yan graves do not intrude on the non-Chinese burials, Western Han graves do. The authors conclude that there probably is not much difference in date between the Yan and Western Han burials. I will therefore

take ca. 309 B.C. as the approximate upper horizon not only for Shi'ertaiyingzi, but for Upper Xiajiadian in general.

Erdaohezi HW^Íl1

Located in the Xuchagou Mountains i^p/^Jli], on the east bank of the Yang River Wffi, 40 km southeast of Liaoyang jTßH, Liaoning, the site of Erdaohezi comprises more than 20 tombs in an area measuring approximately 20 m east-west and 100 m north-south. Only two burials were excavated, both cist graves cut into a layer of graphite. The longer tomb walls were formed of layers of rock filled in with pebbles, while the shorter walls were made of large slabs. A slab formed the slid and floor, although in M2 much of the floor was missing or had never been laid. Both Ml and M2 were covered with a 200 cm thick layer of graphite. Ml measured 2.4 m χ 0.5-0.58 m and 0.64 m in depth; M2, 1.9 m x 0.44 m, and 0.44 m in depth. Both were oriented with the head to the west. Ml contained only a skull and scattered bone fragments; M2, one adult in extended dorsal position, with legs crossed, the arms perhaps crossed above the pelvis, and the head turned to the

right (north). A child lay on its left side atop (between) the adult's legs. Ml yielded a leaf-bladed bronze shorts word (no remaining hilt unit or crown), a bronze axe

and awl, coarse, siliceous red-brown and black-brown earthenware bowls and ajar (broken), all hand-built. The bowl and base of the bowl were made separately and joined. Also found in the tomb was a two-part talc mould for production of bronze axes. The presence of the mould could reasonably indicate that the deceased in Ml had been a bronze caster, but it could also signify control over the casting, hence a limitless number of axeheads belonging to the deceased. Although not incompatible with nomadism, which engaged in bronze production, the presence of the mould as such, emphasizing production rather than use, seems more indicative of sedent- arism.

M2 yielded only pottery. Various jars were recovered from other graves on the site, as well as from what excavators presumed to be a habitation site about a kilometer away. Both the other graves at Erdaohezi and those found near the habitation site were cist graves, apparently of slabs; many had broken open. The ceramics found include jars of the general type Me, which I have identified as existing at Pingyang, Yushu, Jreim UM^SÎ (Inner Mongolia), Yuhuangmiao,

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 37

Xiaobaiyang, and Zhoujiadi (PI. 40). 9° Others are also found on Shangmashi, as will be seen below. This, in addition to the axehead and shortsword blade, provide basis for considering Erdaohezi as part of Upper Xiajiadian and datable within its perimeters.

LiangjiashanícH1!!!

Located 2.5 km to the south of Erdaohezi near Liaoyang, Liaoning, Liangjiashan lies on the west bank of the Yang River yj&ynf. The site consists of seven tombs, of

which M5 and 7 were excavated, the others salvaged, and a habitation site. Nearby were found the site of a Han city and a Han-Wei cemetery. The Upper Xiajiadian habitation site included fire pits and ash pits, into which M7 intrudes, with stone tools and pottery shards, as well as a Warring States grey ceramic dou. For this reason, the site is dated by the excavators to no later than the Early Western Han. No information on the Han sites is included in the excavation report.

M5 and 7 were rectangular pit graves oriented to the north and east, respective- ly. Both were single burials. In M5, the deceased was placed in extended dorsal position, hands crossed over the pelvis; in M7, the deceased was placed on his or her right side, legs bent. Both tombs contained two pots (one hu, one guan) cor- responding to my types Ilia and XIV (PL 39, 44). Agricultural workers claimed to have seen cist graves on the site; from their information, the body in M2, oriented to the northeast, was laid on its right side, legs straight; M3 was also oriented to the northeast, while M6 was oriented to the north. Ml, 2, and 3 contained shortswords of a modified leaf-blade form: without the prominent points midway down the blade and with narrow shoulders at the upper part of the blade, which I will call Upper Xiajiadian I(b)-elongated (cf. PL 65, 66 left). None had hilt units or crowns. No animal bones or horse-gear are reported from the graves.

Wujintang JHr^ít

Located on a mountainside in Jinxi County f$WJi, Liaoning, this site consists of three rectangular pit graves, all oriented to the east (Ml at 80°; M2, 90°; M3, 70°). The dimensions of Ml and 3 were 0.8 m long, 0.35-0.40 m wide at the level

90 Type IIIc is defined primarily as a wide-mouthed jar with an undercut lower belly; jars classified under this type may be only loosely related. Shangmashi JBMllil, Xiaobaiyang M30:14, Yu- huangmiao M190:l (and particularly the latter two examples) are more closely related. The Wangtu piece (PI. 41), although shorter, is also more closely related to Xiaobaiyang and Yuhuang- miao examples. Other pieces in this general category include Pingyang M135:38, M145:5, M153:8: Heilongjiangsheng, Pingyang, PI. 37:4; p. 163, f. 92, second column from right, bottom row; PI. 22:3; Yushu M27:4, Jilinsheng, Yushu Laoheshen, PI. 11:4; Jreim no. 2181, Zhang Bo- zhong ïfer?ji&, "Zhelimumeng faxiande Xianbei yicun" ^M^Ä^SCnW-^itTf-, WW 1981.2, 11, f. 6; Xiaobaiyang M16:3, M18:4, M37:3, M38:l, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:10, 6, 11, 12; Zhoujiadi M31:l, M42:8, KG 1984.5, 420, f. 7:4, 1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

38 Sophia-Karin Ps arras

of the floor, and 1.0 m deep; and 2.2 m long, 0.6-0.7 m wide, respectively. The skeletons had largely disintegrated before excavation, but the deceased in Ml was determined to have been placed in an extended dorsal position. M3 yielded the greatest number of grave goods, including two leaf-bladed shortswords (originally with wood handles), one apparently bronze shortsword crown (nearly regular in shape, but with cannelation across the width of the piece, rather than along its length, as at Zhengjiawazi), a helmet, 15 bronze arrowheads, and six bronze armor fragments. The excavators date the site to the Warring States period, although the reasons for this are not clear. The L- triangle pattern which seems characteristic of Upper Xiajiadian occurs on much material from the site (ornaments, an axehead, semi-spherical appliqués). The reverse side of a square bronze ornament (8.0 cm χ 7.0 cm) with two attachment loops on one end matches the openwork triangle décor on a similar piece from Xiaoheishigou (PI. 14, bottom); the Wujintang pieces car- ries an L- triangle décor over the face, while the Xiaoheishigou example apparently does not.91 Nonetheless, the openwork triangle décor on two ornaments of similar type links Wujintang to Xiaoheishigou. Finally, although I have been unable to find

an example of Chinese platelet armor made of bronze, platelet armor in general is thought to have been in common use from the Eastern Zhou onward. Until the Han, the most common material used seems to have been leather: sometimes

lacquered, and sometimes fitted over a wood backing.92 During the Han, and prob- ably before, iron was used as well.93 If the ge Jc halberd and helmet at Wujintang are in forms dating to the Western Zhou, the armor probably is not. Because the decorative patterns are consistent with those at Xiaoheishigou, for example, Wujin- tang must date to a similar period.

Also in Jinxi County, the site of Sierbao ^ JLJÉ yielded a leaf-bladed bronze shortsword with a bronze hilt unit and regular iron-ore crown. No tomb was found on the site.

Dadun ^ΐ

Located in Haicheng County ?§ίδ£Μ-, Liaoning, in mountainous terrain, the site is comprised of an unspecified but apparently large number of Han tombs around which were found objects characteristic of Upper Xiajiadian, including the blade of a bronze leaf-shaped shortsword (in a Han tomb pit) and two pieces of a pommel meant for holding the crown (found in grave fill). The crown itself was never found. It was supposed that the Han tomb had destroyed a pre-existing Upper Xiajiadian tomb. Clearly the Han tomb with the shortsword, at least, postdates the

91 Wujintang, KG 1960.5, 8, f. 1-2.

92 Cheng/Zhong, Gudai bingqi, 96.

93 For an overview, see Cheng/Zhong, Gudai bingqi, 150-151.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 39

Upper Xiajiadian site; by how much is open to speculation. Because of other evidence, gleaned particularly from Xiaoheishigou, it is reasonable to assume that little time had elapsed; indeed, the Upper Xiajiadian site at Dadun may have been early Han in date. The Chinese report dates the site arbitrarily to the Middle to Late Warring States.

Houmuchengyi JuttííacP and Tuchengzi ;''Μψ

Both sites are located in Lüshun County Äfejlj^-M-, Liaoning94 and are said to have shown signs of grave firing (but no description of the graves is provided), and to have been built over Lower Xiajiadian sites. A shortsword on the site is said to be like that of Dadun (i.e., an elongated leaf-blade, bronze hilt unit, and irregular crown). The shortsword does not conform to Jin Fengyi's categorizations: the pommel is thick and straight, not thin and down-curving, despite the irregular crown. Tuchengzi yielded mingdao RJjTJ coins and thus a terminus post quern of the Warring States, to the late part of which the excavators assign the site.

Dahuofang MkJj}

Located on the southern slope of the Beishan :|fcll], overlooking the Hun River yip ynf in Fushun County J£ Jl|i/J M-, Liaoning, Dahuofang consists of two small cist graves built with finished slabs of regional purple stone. Ml included a lid and floor. The graves were about 30 cm below present ground level. The skeletons had disintegrated, and no grave goods were found in M2. Ml, measuring approximate- ly 2. OmxO. 8m and oriented to 196°, included a stone axe, a bronze axe, and 14 pottery shards. The report dates the site to the Early to Middle Warring States through comparison with Wujintang and Sierbao. Further discussion of redating will be included in Part II, below.

Sunjiagou M7371 ΐ>Μ^/£)

Located in Ningcheng County τ^-Μ*» Inner Mongolia, some 20 m from the eastern bank of the Laoha River 4?nft, the cist tomb of Sunjiagou stands on a plateau on the mountainside, bordered by a seasonally flowing river. The di- mensions and specific structure of the tomb are not indicated. It contained six items

including a bronze leaf-bladed shortsword with a bronze hilt unit and regular crown, a bronze axehead, a bronze awl, knife, and two niuzhong %$f -shaped ling $c bells.95 The latter, with no décor, measure 16.4 cm in height. The attachment

94 At this time, I have read only the reference in Sun Shoudao, "Si'erbao," KG 1964.12, 283, having been unable to locate Lüshun bowuguan M^WWitti * "Liishunkou diqu Houmuchengyi Zhanguo mu qingli" Μ$ν&&)η&%&9£18ΜΜ?ΪΜ, KG 1960.8.

95 See also Lothar von Falkenhausen, "Ritual Music in Bronze Age China." Ph.D. dissertation, Harvard University, 1988, 469-470.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

40 Sophia-Karin Psarras

loop nearly spans the bells' top; two elongated notches were apparently cast in the lower edge of each bell. They are listed in the excavation report as horse/carriage equipment, an attribution which raises a number of questions, since the grave con- tained no readily-identifiable horse-gear. Because of Jin Fengyi's shortsword typo- logies, the site is dated to the Middle to Late Springs and Autumns period (i.e., the later 7 - early 5 c. B.C.). Again, shorts words and their chronology will be discus- sed below.

Liangjiayingzi M8071 ί^Μ^Ψ

This cist tomb, containing 24 objects, was part of a cemetery located on a plateau of a mountain range near a small, seasonally flowing stream, not far from the Laoha River ^?nfí, in Ningcheng County, Inner Mongolia. The tomb, measuring 2.0 m χ 0.5 m, was oriented to the southeast. The walls, floor, and lid were con- structed of unworked slabs. It was a single burial, with the deceased in an extended dorsal position. Grave goods included a bilobed arrowhead, a trilobed arrowhead, an axehead, a semi-spherical appliqué ornament, and a W-shaped ornament (6.2 cm χ 8.8 cm) in the form of a bicephalic snake, all executed in bronze. The site is

arbitrarily assigned by the excavators to the Late Western Zhou/ Early Springs and Autumns (i.e., ca. 8-7 c. B.C.).

Tianjuquan M7301 ^E7^

Also located in Ningcheng County, Inner Mongolia, near the Laoha River ^^ , some 4 km southeast of Nanshan'gen, Tianjuquan M7301 was a cist grave containing 11 objects. No further description of tomb structure is available. Grave goods include semi-spherical appliqué ornaments, a snaffle bit, and a notched shortsword with a hilt décor of repeated small birds standing erect on thin legs, their wings folded; all of these objects were in bronze. The site is dated in the report to the Late Western Zhou/Early Springs and Autumns (i.e., 8-7 c. B.C.).

Wafangzhong M791 ΤΟτ/Ψ

Located on the western bank of the Kuntou River iff^k, a branch of the Laoha 3ξ nfr, in Ningcheng County, Inner Mongolia, Wafangzhong M791 was a cist tomb, containing ten items. These include, in bronze, a helmet, a ge Jc, semi-spherical appliqué ornaments, and an undecorated notched shortsword. Probably because of the type of helmet (like those from Nanshan'gen M 101 and Xiaoheishigou), the tomb is dated by its excavators to the Late Western Zhou/Early Springs and Autumns (i.e., ca. 8-7 c. B.C.).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 4 1

Beishanzui M7501 JfclJlBffi

Located in the northern region of Ningcheng County, Inner Mongolia, the cist tomb of Beishanzui M7501 was built on a habitation site on top of a loess plateau approximately 1 km from the Gala Mountains ίο^ίϊΐΐ]. The habitation site, covering an area of 5000 square meters, lies on the west bank of the Laoha River ^nft, over a Lower Xiajiadian site. The Upper Xiajiadian cultural layer ranges from 1.5 to 2.5 meters thick. Scattered finds on the site include a "fishbone "-décor

pen íâ'Ph$t, which would probably date to the Late Western Zhou/Early Springs and Autumns, if it is related to such pen as Shangcunling M 1744: 1.96 No illustration of the Beishanzui pen is provided in the report. The site also yielded Warring States grey ware dou fragments. Despite the presence of the latter, Beishanzui M7501 is dated by the report to the Late Western Zhou/Early Springs and Autumns, probably because its material is similar to that of Nanshan'gen M101 and Xiaoheishigou, according to Jin Fengyi's chronologies. Instead, it is these similarities which provide a Late Western Han date for the grave.

Beishanzui M7501, oriented to 30°, was constructed of unworked stones, with

no floor, a slab lid, and a tumulus of stone; it lay at a depth of approximately 1.0 m below the present ground level. A single burial, the deceased was placed in an extended dorsal position, and was accompanied by 49 objects. These included (in bronze) a helmet, a double scabbard with openwork décor of a central row of diamond-shapes flanked on each side by triangles, a yue |$ axe, a leaf-bladed shortsword with no remaining hilt unit, a shortsword with an irregularly shaped blade and a hilt in the form of two crouching animals, 40 trilobed arrowheads, what may be a bar bit (one end is broken off), a gui ££ like one from Xiaoheishigou, hence dating to the Late Western Zhou/Early Springs- Autumns, and a small gold ring with flaring ends. The small gold ring has formal counterparts in Lower Xiajiadian (PI. 10, upper right, metal composition not indicated) and in the Shang,97 the latter also in gold, but measuring 12.5 cm to Beishanzui's 4.6 cm diameter. The Lower Xiajiadian pieces are of approximately 5.0 cm in diameter, therefore essentially the same size as the Beishanzui ornament.

Does this imply an early date (but no earlier than ca. 8-7 c. B.C., given the gui) for Beishanzui; was the form of the circlet used over a long period of time in the northeast; or, as is conceivable, was the circlet found by the people of Upper Xiajiadian? At present, the problem is unanswerable. It is clear, however, that because of its demonstrable contemporaneity with Nanshan'gen M 101 and Xiaoheishigou, as seen in the double scabbard, the leaf-bladed shortsword, and the

96 Zhongguo kexueyuan, Sliangcunling , PI. 19.

97 Beijingshi, Beijing kaogu sishinian, PI. 8:2; Beijingshi wenwu guanlichu 'Atsp^'X^J^n'Milh, "Beijingshi Pingguxian faxian Shangdai muzang" ^t^i'íí''^&&M'u]i'MW, WW 1977.11, 6, f. 13.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

42 Sophia-Karin Psarras

animal-hilt shortsword, Beishanzui must be Late Western Han in date. Habitation site finds of Warring States dou fragments make an upper horizon of the later Warring States possible for the site as a whole. No other information about the site is currently available, to my knowledge.

Shuangfang Mfjs

Located on the southern flank of the Xi Mountains ]Hj lL| in Xinjin County iff^fí-, Liaoning, Shuangfang consists of a group of tombs described as six dolmens ^fÜ and three cist tombs with slab walls and lids but no floor, since the tombs, if we may judge according to M6, had been built on stone. Granite was used in construction. The distinction intended by the excavation report authors through the use of this structural terminology is unclear; all tombs had been damaged and were open at ground level, and only M6 is described. M6 was covered by a lid con- sisting of a round stone with a diameter of some 1.7 m and a thickness of ap- proximately 0.3 m, which was felt to have originally been above ground (hence, dolmen). The tomb measured 1.55 m χ 0.50 m (width at surface, vs. 0.60 m width

at bottom) χ 0.73 m deep. M6 contained four ceramic jars, a bronze leaf-bladed shortsword with no remaining hilt unit or crown, and a talc (with a small amount of

graphite) mould for the production of axeheads. The pottery, of coarse, siliceous black-brown earthenware, was hand built; two of the pieces are closely related to pieces from Erdaohezi (PI. 46). Xu Minggang i^pR/j^ and Xu Yulin #3t#, the excavation report authors,

stating their disagreement with Wu En J5j^ and others, date the site not to the Late

Western Zhou/Early Springs-Autumns (i.e., 8-7 c. B.C.), but to the Late Shang/ Early Western Zhou (i.e., 11-10 c. B.C.). They base this dating on radiocarbon readings from a stone tumulus tomb at Yujiacun ^M^t and the upper level of Shangmashi (discussed below), both of which are variously identified as being located in Lüda jjfcfc or Lüshun County JjfêJllj/íH-, Liaoning. These radiocarbon readings are 3280 B.P. ± 85 and 3130 B.P. ± 100, respectively, with no further information supplied.98 This would yield dates of 1330 B.C. ± 5 and 1280 B.C. ± 100, which are even higher than the date adopted by Xu and Xu. The context from which the radiocarbon readings were obtained is unclear, and no source references are given by the report. A site at Shangmashi has been reported elsewhere as yielding dates ranging from 4045 B.P. ± 100, 3930 B.P. ± 100, and 4460 B.P. ± 150, calculating in conventional, physical, and dendrochronologically calibrated terms, based on oyster shells from T5-4; and dates of 3130 B.P. ± 100, 3040 B.P.

± 100, and 3320 B.P. ± 160 (conventional, physical, and dendrochronologically calibrated, respectively) from oyster shells in T5-2. Dates of 3170 B.P. ± 150,

98 Shuangfang, KG 1983.4, 295.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 43

3080 Β. P. ± 150, and 3365 B.P. ± 195, based on wood ash, were derived from T5-2 as well." However, these dates, while corresponding in general to those cited by Xu and Xu, are reported by the radiocarbon laboratories as proceeding from a neolithic site located at Shangmashi. This site and its dates, whatever their ac- curacy, therefore have nothing to do with the Upper Xiajiadian graves at Shang- mashi. Radiocarbon dates for Yujiacun have been published elsewhere as 2275 B.P. ± 110, 2210 B.P. ± 110, and 2280 ± 115 (conventional, physical, and dendro- chronologically calibrated) based on human bone samples.100 This yields dates, not of 1330 B.C. ±85, but of <435-215B.C>, < 370-150 B.O , and <445-215 B.O . Beyond all problems of radiocarbon dating, there would therefore seem to be confusion concerning the sites to which Shuangfang should be compared. Because of the ceramics, shorts word, and the form of axehead produced from the mould, this site should date to the same period as Erdaohezi and Xiaoheishigou.

Shangmashi h^í/íí

Located in Changhai County ix)§H-, Liaoning, no description of the terrain in which seventeen jar burials (JBM1-17) and ten pit graves were found has been published. The site was first discovered during ploughing in the winter of 1974; the

implication is that many graves lay near the surface of the earth. A stratigraphie section has been published, but without indication of where the graves lie in relation to one another in cross-section. As described, the stratigraphy is: I) a layer, 10-28 cm deep, of cultivated earth into which the pit graves intrude; II) a layer of oyster shells, 48-98 cm deep, which also contained shards of siliceous red- brown earthenware; jar coffin burials intruded into this layer as well; and III) virgin soil consisting of loess and gravel. It is not necessarily true, as the stratigraphie report implies, that red-brown pottery preceded black-brown pottery chronologically, since it is clear from Erdaohezi that the two may coexist. Given the layer of oyster shells and ceramic debris, Shangmashi would seem to have been

a habitation site. The jar coffin burials are reported to lie approximately 15 m apart, horizontally, from the pit graves; their relative vertical distances are not reported and do not seem too significant, given that both types of burial were near enough to the surface of the ground to be disturbed by routine agricultural work.

The jar coffin burials are largely of children (the bones in JBM6 and 8 had disintegrated too much for analysis), with the exception of JBM3 and 4, which were diagnosed as belonging to adolescents or "pre-adults." These are the only burials identified as second inhumations. The jars were buried with the mouth facing upward (JBM1, 2, 8, 11, 15, 17) or downward (JBM3-7, 9, 10, 12-14, 16).

99 Zhongguo shehui kexueyuan, ZJiongguo kaoguxuezhong tanshisi, 29, 31 .

100 Zhongguo shehui kexueyuan, Zhongguo kaoguxuezhong tanshisi, 28; cf. idem, 27, 30.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

44 Sophia-Karin Psarras

Approximately half are said to have been placed in circular pits. JBM1, 5, 6, 8, 9, 11-13, and 15-17 included funerary goods: JBM17, two bowls; each of the other tombs, one jar. JBM6's jar was found outside the coffin. The ceramics were of siliceous black earthenware, wheel thrown, fired high, and polished. The ceramics used as funerary goods are characterized as mingqi HJj#fr, miniature copies made for funerary purpose, because of their small dimensions: between 10.2 cm and 16.0 cm high, for the jars. This, however, is not necessarily smaller than the dimensions of other Upper Xiajiadian pottery, such as that from Erdaohezi (e.g., 16 cm high) and Shuangfang (e.g., 16.7 cm high). JBM1 also contained a shell bead; the grave fill yielded a stone arrowhead, as did the fill for JBM3. JBM11 included pig bones in the jar coffin. Given the paucity of grave goods and the type of burial, can the jar coffins be attributed to Upper Xiajiadian, regardless of the presence nearby of pit tombs with, in some cases, leaf-bladed shortswords?

An answer may be found in the form of ceramics, including those of the jar coffins themselves. Although one, JBM12:1101 is comparable in general terms to Early Western Zhou ceramic forms from Liulihe M 1008 (PI. 55, 56), all the other comparisons which I have been able to identify are with Late Warring States and Han era sites. For example, using the typology I advanced in "Exploring the North": Shangmashi JBM11:1 is type IIIc, found also at Zhoujiadi M31:l, M42:8, as well as Xiaobaiyang M30:14 (as well as several other examples on this site), Yuhuangmiao M190:l, Yushu M27:4, and Pingyang M135:38, M145:5, and M153:8 (PI. 40). 102 Shangmashi JBM1:1 and JBM9:1 are both type XIV (PI. 44), which occurs on Liangjiashan M5, but also at Budonggou YHPI^J, Yikezhao League ^jËBinaL Inner Mongolia, Ml:l (Xiongnu), Zhalainuoer ÍLJfií^, Hu- lunbaier League n?fÊJJ!^j'iïï, Inner Mongolia, and Menggen (Menggenchulu néíS^rÍ), Hulunbaier League, Inner Mongolia, both Xianbei.103 The jar coffin of JBM14 is also type XIV. The jar coffin of JBM11 is of type Vb, which includes Daodunzi M26:l and Xiaobaiyang M21:2 and M41:8, as well as the non-Xiongnu site of Maoqinggou ^Jjcyíj, Liangcheng County SjV^-M- Inner Mongolia, M23:l (PI. 42). 104 Shangmashi JBM13:1, type XXII, also occurs at Pingyang M104:9 (PI.

101 Shangmashi, KG 1982.6, 592, f. 2:9.

102 See note 90.

103 Neimenggu wenwu gongzuodui iM^tÎÏ'JC'-tyj T.fî^À, "Neimenggu Zhalainuoer gumuqun fajue jianbao" N&tâl$ffifoÍiUffi!£WfâU, KG 1961.12, 673-679; Zheng Long ML· "Nei- menggu Zhalainuoer gumuqun diaocha ji" lAl^lWL^iS^ÎTÎSîffi^ÎiiC, WW 1961.9, 16-19; Cheng Daohong iMiË^, "Yiminhe diqu de Xianbei mu" ffiUÏNi&KfâfyW^, NMGWWKG 2 (1982), 18-23; Yimeng wenwu gongzuozhan iP!StSC$j T.ffcifi, "Yikezhaomeng Budonggou Xiongnumu qingli jianbao" WWii^M^fàteMWvMÍniU, NMGWWKG 1 (1981), 27-32, 34; Tian/Guo, Ordos, 394-402.

104 Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:7; Maoqinggou, Tian/Guo, Ordos, 256, f. 29:4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 45

45). 105 In each case, the comparative sites have directly or indirectly positively- established dates of Late Warring States and Han. These comparisons, because they occur in such a limited context, both suggest that an Upper Xiajiadian attribution is acceptable for the jar coffin burials, and provide a late date for these burials (i.e., generally very late 4 c. B.C. through at least 1 c. B.C. or even later, given Xianbei formal comparisons). Although the majority of the jar coffin burials are of children, the pit graves

also contained children in at least two instances (M6, 8). The pit graves made no use of stone, according to the excavation report. Grave dimensions vary, but frequently exceed one meter in length: Ml, 1.90 m χ 0.70 m; M2, 1.88 m x 0.80 m; M3, 1.70 m χ 0.84 m; M4, 1.88 m χ 0.78 m; M5, unclear; M6, 1.12 m χ 0.56 m; M7, 1.44 m χ 0.44 m; M8, 0.64 m χ 0.30 m; M9, 1.64 m χ 0.66 m; M10, 1.98 m χ 0.64 m. M8, exceptionally small, is a child's grave; since M6, also a child's grave, is nearly twice the size, it is not clear if grave size reflects the age or status of the deceased. M2 was oriented to 325°; M3, 145°; M4 and M9, 360°. The deceased in M4 and 8 were buried in an extended position on their stomachs. The hands of the body in M4 were placed in front of the pelvis. The bodies in M5 and 9 were placed on their (left) side, with one or both legs bent. The hands of the deceased in M9 were crossed over the chest. All other burials were in extended

dorsal position. There were few burial goods: one ceramic jar in M3 and M4, and a bronze

leaf-bladed shortsword in M2 and M3, although this group of pit graves was named "bronze shortsword graves" in the excavation report. There is, therefore, little evidence on which to base dating. The ceramics are of types Ilia (Shangmashi M3:2, PI. 39) and XII (Shangmashi M4:l, PI. 43), which provide comparisons to Zhoujiadi Ml:4, Zhengjiawazi M6512:63, Liangjiashan M5, Xiaobaiyang M12:5, M35:2, and the Xianbei sites of Wangong y£X, Chenba'erhu Banner [^ti^g, Inner Mongolia, M1B:72, Tiejianggou |£E$J, Aohan Banner j$£YJ£, Inner Mongolia, M2:l, M2:2, AM3:1, and Yushu M92:2 (type Ilia)106 and to the non- Xiongnu Inner Mongolian sites of Yinniugou tfc^N^J M9:l, Guoxianyaozi ΐ1$Μ- ^C-jP M25:l, and the Xianbei sites of Wangong M1B:62, Sanjiazi H^c-f, Qiqi- ha'er 5fcfipnft^K, Heilongjiang, scattered find #5, and Jreim fëMft find #2047

*05 Heilongjiangsheng, Pingyang, PI. 20:2.

106 Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:16; Neimenggu zizhiqu wenwu gongzuodui 'H^tdSäiüS.'XW-''.YffA, "Neimenggu Chenbaerhuqi Wangong gumu qingli jianbao" μ%<ώ' ^E^lAÍK^:i;i!rKfffSliffiíílí, KG 1965.6, 273-282; Shao Guotian «I® LU, "Aohanqi Tiejiang- gou Zhanguo mudi diaocha jianbao" MUM^l&I^Í&MMtM^Milí, NMGWWKG 1992.1-2, 86, f. 5:1, 4, 5; Jilinsheng, Yushu Laoheshen, PI. 11:1. I have elsewhere listed Yushu M42:l and M105:2 (idem, PI. 11:2, 4) as type Ilia, but while they are related to Zhoujiadi Ml:4, they are a subgroup of the type, whereas Yushu M92:2 more generally fits the classification. I have allowed considerable variation within types.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

46 Sophia-Karin Psarras

(type XII).107 Because of their narrow distribution, these forms may be considered as dating to a fairly narrow time period: the Western Han, in general. Shangmashi M3:2, 21.5 cm tall, of brown argilaceous earthenware, was built by hand and finished on a wheel, then fired at high temperature. Shangmashi M4:l, 15 cm tall, of grey siliceous earthenware, was also hand-built and wheel-finished. No in- formation on firing temperature is provided. The shorts words of M2 and M3 con- form to types known on other Upper Xiajiadian sites, including the decorative L- triangle patterning on the Shangmashi M3:l hilt unit (hilt and pommel). M3:l consists of a leaf-blade, a bronze hilt unit with pommel, and an irregular bronze crown, as well as a regular iron-ore crown.108 This provides both evidence of the site's contemporaneity with other Upper Xiajiadian sites, and evidence of the con- temporaneity of all variants of leaf-bladed shorts word forms. The excavation report dates the site to the Early Warring States (i.e., ca. 5 c. B.C.) through comparison to other Upper Xiajiadian sites.

Menlian Π)&

Located 35 m up the south flank of the foothills of the Tuanshan [ï[ Ü4 in Qingyuan County Yr'JjK-M-» Liaoning, Menlian consists of two cist tombs 20 m apart. The tombs, including a floor and lid, were built of regularly-shaped, worked slabs. Dimensions of one tomb are given as 1.85 m long, 0.45 m wide, and 0.55 m in depth. Some grave goods are described, without specification of tomb number. These include a bronze leaf-shaped shortsword blade, with indications of wear; a

stone bilobed arrowhead; a bronze arrowhead; a red-brown siliceous (added grog) earthenware jar, wheel thrown, with two "bridge-shaped" handles like those on pieces from Erdaohezi and Shuangfang (PI. 46), whose dimensions are no greater than pieces from the Shangmashi jar coffin burials (13.7 cm high).109 Red-brown siliceous earthenware pottery shard were also found. The site is therefore consistent with other Upper Xiajiadian sites. No date is proposed in the report.

Nandonggou ffln|$J

Discovered during ploughing in Chaoyang Municipality U]ßH, in the left wing of the Kalaqin Banner n^iifij}^, Liaoning, the site consists of a single cist grave located on a slope near a stream running to the Daling River ^ζΐ^. Oriented to 80°, the

107 Neimenggu wenwu, "Guoxianyaozi," KGXB 1989.1, 64, f. 10:10; Zhang Bozhong, "Zhelimu," WW 1981.2, 9, f. 1:6; Heilongjiangsheng bowuguan, Qiqihaershi wenguanzhan ^JiLfS^WY^J Í& ' iï^Ùfcïb'X'S'ïk, "Qiqihaershi Dadao Sanjiazi muzang qingli" ^jfVftfolfijZiE^^i^f SSPfêfSIL KG 1988.12, 1092, f. 4:7. The drawing of the latter is not symmetrical; I assume that the right side is more accurate.

108 Shangmashi, KG 1982.6, 594, f. 6:1, 2 (Ml:3), 3 (M2:l).

109 Menlian, KG 1981.2, 189 description only.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 47

tomb measured 2.9 m χ 2.0 m, and 1.1 m in depth; its walls were made of stones and included a 20 cm wide shelf on the north and south sides. Burial goods included a leaf-bladed bronze shortsword with a bronze hilt unit and a regular iron- ore crown. The hilt and pommel carried the characteristic Upper Xiajiadian L- triangle décor. The tomb also contained a bronze snaffle bit and a single bone cheekpiece, as well as a stone axehead and one brown siliceous, polished, hand- built earthenware jar. The tomb yielded two bronze appliqué ornaments (15.7 cm χ 13.1 cm) in the form of manna rays, as well as eight smaller (9.5 cm χ 9.3 cm) passants in the same form (PI. 38, top).110 These compare to an appliqué ornament of a somewhat different form, although also depicting a manna ray, from Wu- daohezi Ml: 18 (see below) (PI. 38, bottom). The Chinese objects on the site have, logically, been responsible for the Late

Springs- Autumns/Early Warring States (ca. 5 c. B.C.) date assigned by the excava- tors to Nandonggou. The bronze garment hook, 8.8 cm long, with a cannelated body ending in a small birdhead,111 cannot be precisely dated, to my mind, since close comparisons may be established throughout the Warring States and into the Western Han.112 The bronze, rounded-point ge is of a form known in the second half of the Warring States and Western Han (PI. 59). In particular, comparison may be made with Fenshuiling jí'/}<Av (Changzhi, Shanxi) M53:12113 and with Yunxian f^Jl· (Hubei) M132:16,114 both dated to the Middle Warring States (i.e., ca. 4 c. B.C.; PI. 95, center, bottom). A different date is provided by the site's gui (measuring 11.0 cm high, 9.8 cm girth, 12.9 cm diameter of the mouth), which is

Late Springs and Autumns in date (i.e., ca. late 6/5 c. B.C.). The gui carries a band of braid décor around the foot and a band of two lines of rectangles encom- passed by a rope décor about the upper body.115 The two vertical ring handles are decorated with a feline head on the upper edge, facing the gui body. The lid is missing (PI. 36). A nearly identical piece, said to come from an Upper Xiajiadian

110 See also, Li Xueqin, Qingtongqi (xia), 42, PI. 51; 82, PI. 95. The same work, 41, PI. 50, includes a good photograph of the Shi'ertaiyingzi shortswords.

111 Nandonggou, KG 1977.6, 375, f. 3:4.

112 Osaka shiritsu bijutsukan ^l>j£i'i M'MWiWi (ed.), Osaka shiritsu bijutsukan kiyõ: Chügoku taikõ mokuroku oyobi shutsudo shiryõhyõ ^RW^'MiitJÎSîîl (Osaka, 1980), 62, no. 21; cf. 62-63, no. 22 (sacrificial burial no. 1 at Langjiazhuang, Linzi Co., Shandong: Shandongsheng bowuguan lIl/jvSW^JÍÈ, "Linzi Langjiazhuang yihao Dong Zhou xunrenmu" ifSïmffl^JÎ- v/jCj^fU ÀSC-, KGXB 1977.1, 77, f. 5, but also cf. Zhongguo shehui kexueyuan, Mancheng, v. 2, PI. 43:2 (Ml:5040, ca. 113 B.C.).

113 Shanxisheng, "Shanxi Changzhi Fenshuiling," KG 1964.3, PI. 6:3.

114 Zhongguo shehui kexueyuan, "Yunxian," KGXJK6, 143-174, PI. 18-26.

115 The terms "braid" and "rope" décor follow George W. Weber, Jr., Vie Ornaments of Late Chou Bronzes (New Brunswick, N.J.: Rutgers University Press, 1973).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

48 Sophia-Karin Psarras

site, is displayed in the Museum of History, Beijing (PL 5). The handles on the latter lack the feline heads; the lid is present, with a décor matching that of the body, a central vertical ring handle, and three deer heads rising from the edge of the lid. A nearly identical piece was found on a site attributed to the state of Yan at Nanyang "j^ßH, Rongcheng County §J$cll·, Hebei.116 Essentially the same form and décor (with the lid and ring handle) was also found in Jiagezhuang 'Sf^rJEË, Tangshan Municipality HfLll, Hebei, M18:3, originally dated to the Early Warring States, but more likely to date to the Late Springs-Autumns period.117

Nandonggou also yielded two bronze Chinese axle caps, which may also date to the Late Springs and Autumns period. The two axle caps differ slightly from one another: although each has two feline head loops just below the rim, the décor on one (PL 34, bottom) is of a loose interweave of thin snakes, while on the other (PL 35, bottom) the interweave is more compact and the snakes are somewhat wider. Each variant has closely related parallels at Beixinbao Jt$M, Huailai County 'R^-M-, Hebei,118 a site which I believe dates from the Late Warring States or Western Han, although its Chinese bronzes are clearly Early Warring States (PL 34, top; PL 35, top). Other comparisons for the Nandonggou axle caps may be established. An example nearly identical to that on PL 34 (bottom) was found in Yuhuangmiao Ml:5 (PL 33); also at the state of Zhao site of Handan #[$$[$ in Baijiacun 'fo'MH, Hebei,119 M57:28, dated to the Middle Warring States. A nearly identical piece from the state of Yan site of Longwandun jèM^Í in Shunyi County ΜΧ-Ά, Beijing Municipality, is dated to the Middle to Late Warring States.120 The Yan piece lacks the feline face on the loops near the rim, which in this case take the form of rectangles extending from the axle cap body. Similar décor is used on a pair of axle caps from the 4 c. B.C. Chu site of Baoshan ^li|, Jingmen Municipality MÍ I, Hubei; here, the feline face is replaced by the entire feline depicted in the round and placed at the outer edge of the cap (M2:329, ca. 316 B.C.).121 At the same time, the loose interweave of snakes occurs on an axle cap from the 491 B.C. tomb of the Marquis of Cai '$£{% Shou County 7?rä, Anhui,

116 Sun Ji'an #£|1i£-, "Hebei Rongchengxian Nanyang yizhi diaocha" MJt^ií-flAlíliSíltÍ^a, KG 1993.3, PI. 2:1.

117 An Zhimin ·&-;£$[, "Hebei Tangshanshi Jiagezhuang fajue baogao" MJbÄflllTfiW4fJEÄ:iffi ífêff.JTGXB 1953.1-2, PI. 9.

118 Hebeisheng wenhuaju wenwu gongzuodui M^t^'XitMj'X^n^YP^ "Hebei Huailai Beixinbao Zhanguomu" mt'W^t^Mi&M^ KG 1966.5, 231-242, PI. 1-4.

119 Hebeisheng, "Handan Baijiacun," KG 1962.12, 626, f. 21:2 (M57:28).

120 Beijingshi, Beijing kaogu sishinian, PI. 16:3.

121 Hubeisheng Jingsha tielu kaogudui ÎfflJt^^Jt^ttS&^'ÎbîA, Baoshan Chumu fölil^SS (Beijing: Wenwu, 1991), PI. 74:1 (M2:329).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 49

M46:2, with the addition of a band of braid décor along the outer edge of the cap.122 Such braid is used on axle caps of the Early Warring States at Fenshuiling M26:21,123 the tomb of the Marquis Yi of Zeng N-146 (ca. 433 B.C.),124 and at Middle Warring States Handan (Baijiacun) M3:48.125 This type of axle cap décor was obviously in use at least from 491-316 B.C., if not longer. It therefore provides only a broad chronological horizon for the piece itself, without reference to the site on which it was found. Its presence at Yuhuangmiao, for example, provides evidence for continued use at a late date of earlier productions. Therefore, despite the Nandonggou gui, the site itself may be provided with a ca. 4 c. B.C. upper horizon by the ge and the axle caps. Internal comparisons of Upper Xiajia- dian shorts words demonstrate a very late 4 c. B.C. upper horizon and a Wang Man era lower horizon, with the latter being the most probable date (Late Western Han/

Wang Mang).

Dongnangou tfcMfà

Located in Pingquan County ^^-M-, Hebei, the site termed Dongnangou encom- passes two separate burial grounds: Huang wozishan j£^^pll| and Beidamianshan JfCJBJli], each located on the side of the mountain bearing that name. Huangwozi is known to have at least 46 graves, of which 26 were cleared. Beidamian yielded more than 50 graves, of which only one was excavated. The excavation report gives data only on Huangwozi Ml-10 and Beidamian Ml. The tombs had broken open at ground level because of rain and runoff, but most, if not all, were originally covered by a tumulus either of earth or stone. Grave structure was pit (Huangwozi M6, 7) or cist (slab: Huangwozi M4, 8, 9; Beidamian Ml). Some pit graves had slab lids (Huangwozi M4, 8, 9). Huangwozi M 1-3, 5, and 10 were oriented to between 110° and 125°; Beidamian Ml, to 186°. Huangwozi Ml and 5 yielded the ash remains of a coffin or bier (1.88 m χ 0.42 m in Ml). All were single burials with the skeletons largely disintegrated. Huangwozi M8 and 9 contained children. There were few grave goods. Bone beads figured in several graves (Huangwozi M3 contained 114 placed next to the skull; Huangwozi M10, 166 on the chest; Huangwozi M4, 17 near the head; Huangwozi M6, 29). Huang- wozi M6 yielded a ge and a notched bronze shorts word (M6:5) with a repeated animal design in raised line relief over the hilt (PL 82, right). The hilt ends with a rectangular band ornamented with fine vertical lines. The form of the hilt unit

122 Anhuisheng bowuguan ^Wî'&WWHÎS , Shouxian Cai Hou mu chutu yiwu ^ïa^fë^lMiM^J (Beijing: Kexue, 1956), PI. 61:2 (M46:2).

123 Shanxisheng, "Fenshuiling," KG 1964.3, 128, f. 18:1 (M26:21).

124 Hubeishens. Zens Hou Yi, v. 2, PI. 105:4 (N-146).

125 Hebeisheng, "Handan Baijiacun," KG 1962.12, 626, f. 21:1 (M3:48).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

50 Sophia-Karin Psarras

belongs to the category seen at Xiaoheishigou, Yinniugou, Hulusitai, and the Wang Mang-era Chang'an weapons depot (PI. 80, center; PI. 81, left). The hilt décor belongs to a type found not only on the Xiaoheishigou bar-pommel notched shortsword (arranged vertically rather than horizontally, as on Huang wozi M6:5, see PI. 81, left), on a knife also from Xiaoheishigou (PL 81, second from left), and on notched shortsword hilts from Nanshan'gen M101 (PI. 68, right). This raised line depiction of animals is also found on Yuhuangmiao (PI. 74, fifth and sixth from left). The animals on the hilt of Huang wozi M6:5 are not highly differentiated, but could be horses or deer. This places Huangwozi well within the chronological perimeters of Upper Xiajiadian in general, particularly of Nanshan'gen and Xiaoheishigou. Note that the form of the Huangwozi M6 ge dates not only from the Western Zhou and the Springs and Autumns period (e.g., Middle/Late Springs- Autumns Shangmacun _':MH, Houma fë£j, Shanxi, M13),126 but from Middle Warring States sites such as Baijiacun (Handan) in the state of Zhao, as well.127 The problem of ge will be discussed below, in Part II. Because ge forms were frequently maintained over long periods of time, although several forms might coexist, they often are not good indicators of era.

Also present on the site were two semi-spherical bronze appliqué ornaments from Huangwozi M6, with a border of fine radiating lines (5.4 cm and 4.7 cm in diameter). This particular décor seems to be era-specific, even though its distribu- tion is geographically broad, occurring in Upper Xiajiadian not only at Huangwozi, but also Dabannanshan ^jsu^lll, Linxi County ^M-M-, Liaoning; in Yuhuang- miao at Xibozi Ffíê-p-, Wudaohezi, and Xiaobaiyang; in Xiongnu sites such as Xi- gouban M?íj#, Yikezhao League #*JnLH£{äL Inner Mongolia; and into the Qinghai culture of Kayue -fr^J, Huangyuan County ^ΒΜ-Ά, Qinghai.128 In each case, the

126 Shanxisheng wenwu guanli weiyuanhui Houma gongzuozhan ÚlW^SC^f&W^^uiW^iVÁ^- jrfí, "Shanxi Houma Shangmacun Dong Zhou muzang" ΐΙΐΜΜ-^..Κ-Π;ίίί^ΪξΡ, KG 1963.5, PI. 4:3.

127 Hebeisheng, "Handan Baijiacun," KG 1962.12, 624, f. 15:12 (M57:24).

128 Also in Upper Xiajiadian, Wafangzhong M791, WWZLCK9, 55, f. 50; Dongnangou, KG 1977 .1 , 54, f. 7:1, 2; Jin Fengyi, "Xiajiadian shangceng," KGXB 1987.2, 184, f. 3:9 (Dabannanshan); Beijingshi wenwu guanlichu Jitt^ïb' SC^J^S'Milh, "Beijingshi Yanqingxian Xibozicun jiaocang

tongqi" Jt^^^äM^^i^'ÄÄtff KG 1979.3, 229, f. 4:6, 7, included in Beijingshi, Bei- jing kaogu sishinian, 70, f. 34:5 (the cache belongs to Yuhuangmiao); Wudaohezi, WW 1989.2, 56, f. 8:26; Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 48, f. 15:18; Qinghaisheng Huangyuan- xian bowuguan, Qinghaisheng wenwu kaogudui, Qinghaisheng shehui kexueyuan lishi yanjiushi ff U'&mmmm · ttm^xw^^x · w^m^w^m^m:^ -Qmghai Huang- yuanxian Dahuazhongzhuang Kayue wenhua mudi fajue jianbao" i^W&W^^ftViïr^î^X k^ife&Mfíílí, Kaogu yu wenwu ^"^jX^J 1985.5, 22, f. 17:13 (variant form); Yikezhao- meng wenwu gongzuozhan Neimenggu wenwu gongzuodui Í^yíWnlMX^jy.Yf-íf^'H^t'h'XW^ f^PA, "Xigouban Xiongnu mu" WM^HWM, WW 1980.7, 4, f. 6:11. This analysis depends on the cultural context in which this décor is found.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 5 1

tombs where this pattern is found date to no earlier than the Late Warring States. The décor, not the form, therefore seems to date Huangwozi to no earlier than the Late Warring States, again consistent with Upper Xiajiadian horizons. Four ovoid bronze plaques with a hole pierced through the broader end were

found in Huangwozi MIO (M10:l). Believed to be parts of earrings, the décor of these small plaques (2.1 cm-2.3 cm in length) becomes important because of its rarity. On the plaques is depicted a plant with a central stalk and four pairs of op- posite leaves.129 This motif occurs in the round in bronze on the non-Xiongnu site of Maoqinggou M58:3, dating to the Late Western Han.130 The balance of comparative evidence for Huangwozi therefore points to a Late Western Han date, countering the Chinese archaeologists' assumption of a Late Western Zhou/Springs and Autumns date (ca. 8-5 c. B.C.). The Chinese date is based on comparisons with Nanshan'gen M101, but also on the assumption that from the Warring States period onward, non-Chinese states must reflect Chinese influence. Given the fact that Chinese sites along the northern border of China proper do not uniformly evidence contact with the non-Chinese, this assumption cannot be sustained. No mention of any animal bones in the Dongnangou burials is made in the

excavation report, nor is there mention of horse-gear. Huangwozi MIO is marked by the use of shells as one part of burial ritual: ten cowries (1.5 cm long) were placed over the head and mouth of the deceased. A plain bronze semi-spherical appliqué ornament was found next to the head, as well. Huangwozi MIO was a fairly wealthy burial, including as it did the double plaque bronze earrings and 166 bone beads of various shapes on the chest of the deceased.

Sanguandian HTT^J

The cemetery of Sanguandian is located on a mountainside on the west bank of a

gully in Lingyuan County 10iM&, Liaoning. The gully carries water only during the rainy season; this water and the runoff from the mountains has broken the Sanguandian graves, making it impossible to reconstruct their layout. The number of graves is not indicated and only two are discussed in the preliminary excavation report. A large number of burials were small, with a stone floor. Some were large cist graves made of irregularly shaped stone. The larger graves generally had a greater number of burial goods. All burials are said to have been single, with the head of the deceased pointing east. Grave goods included items associated with Upper Xiajiadian: a bronze shortsword hilt unit with the characteristic L-triangle décor, a broad rectangular pommel, and a regular iron-ore crown; but also a siliceous red earthenware tubular fitting, one end of which terminates in a horse's

129 Dongnangou, KG 1977.1, 54, f. 8:8.

130 Maoqinggou M58:3, Tian/Guo, Ordos, PI. 113:6.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

52 Sophia-Karin Psarras

head; and a two-part talc mould for the production of bronze axeheads. Such moulds are known from the site of Wangtu Ξν+Ο Chifeng Municipality T^llif , Inner

Mongolia (PI. 37), a site considered Upper Xiajiadian; from Erdaohezi, Shuang- fang, and Dajing.131 The shortsword hilt (PI. 63) provides data on the structure of Upper Xiajiadian leaf-bladed shortswords. Within this bronze hilt was found a two-

piece wood "sleeve" which had originally been fitted over the bronze tang of the blade unit and bound with hemp. This demonstrates not only how the hilt was fastened to the blade in the case of bronze hilt units, but also the fact that wood re-

mained an essential element of the hilt unit even though it was not visible. This means that there is very little structural difference between leaf-bladed shortswords with (now disintegrated) wood hilts and those with bronze hilts.

Important material both for dating and for intercultural relations is provided by a series of gold and bronze plaques and passants . The bronze passants in the form of frogs, or of frogs at whose hind legs two intertwined snakes are biting, provide a general link to Upper Xiajiadian sites such as Nandonggou and the non-Upper Xia- jiadian site of Wudaohezi with their bronze manna ray ornaments, and to Upper Xiajiadian Huangwozi, with its bronze duck plaque.132 The depiction of felines and, in particular, of a feline with prey (identified in the excavation report as a rabbit) on other Sanguandian plaques and ornamental pieces raise several questions. One gold plaque from Sanguandian (PI. 116, bottom) depicts a crouched feline

with its raised head turned to face its hindquarters. The animal's shoulder and hip are marked by an oval outlined with small circles. At first glance, this feline seems

related to a bronze feline garment hook from Fenshuiling ^/jCdi^, Changzhi ixfn, Shanxi, M49:6, dated by the excavators only generally to the Warring States period.133 The contents of Fenshuiling M49 are largely unpublished, but com- parisons made by the authors of the excavation report of various bronzes within the

Fenshuiling cemetery make a Middle to Late Warring States date likely for M49. Nonetheless, closer inspection verifies that, whatever its absolute origin, in the context of Fenshuiling and the Warring States period in general, the feline on Fenshuiling M49:6 is the descendant of earlier Chinese felines such as those depicted on ceramics such as Xiadu T^, Yi County ÜJl·, Hebei, M29:14,134 from the state of Yan, dated to the Early Warring States (visible on PI. 32, left) and on Chinese bronzes of the Late Springs and Autumns and Early Warring States

131 Jin Fengyi, "Xiajiadian shangceng," KGXB 1987.2, 183, f. 2:19; Dajing, WWZLCK7, 142, f. 7:10; PI. 13.

132 Dongnangou, KG 1977.1, 54, f. 8:11.

133 Shanxisheng, "Fenshuiling," KG 1964.3, 134, f. 26:1.

134 Hebeisheng wenhuaju wenwu gongzuodui HJ{tfè3CikM}X'$rLYF-VA, "1964-1965nian Yan Xia- du muzang fajue baogao" l964-'965^^tfU^W!&WM^r, KG 1965.11, 552, f. 6:7.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 53

periods.135 The Sanguandian gold plaque feline, however, is depicted in a crouching position with its hindquarters arched high, but its hind legs fully on the ground. This posture places it in relation to the animals of Wudaohezi136 and, by extension, to those of the Yuhuangmiao culture, on which the marking of haunch and shoulder is also present in various forms (PI. 113 provides various examples, including horses and felines). The question of the nature of intercultural relations is further exemplified by

two bronze ornaments (measuring 11.0 cm in length, 4.1 cm in width, and 5.0 cm in height) from Sanguandian: executed in the round, these pieces show a crouching feline holding a small prey (rabbit? fawn?) between its forefeet.137 The feline's muzzle rests against the prey but, judging from the excavation report illustration, without actually biting the prey. This composition has two immediate parallels: that of Late Springs-Autumns and Warring States Zhongshan (and, hence, Jin i^f)138 and that of the Xiongnu. Each of these two cultures has a distinct manner of presenting closely related prédation scenes. Perhaps because the motif occurs very rarely in Upper Xiajiadian and, in fact, in Yuhuangmiao, the Sanguandian pieces are not entirely like either the Zhongshan or the Xiongnu versions.

The Xiongnu version of the steppe theme of sacrifice involves a feline (tiger) carrying its prey (horse, ram, or deer). In Xiongnu pieces such as the one from Yanglang %%, Guyuan County @MJl·, Ningxia (PI. 118),139 the prey hangs limply from the tiger's mouth, its body thrown over the tiger's shoulders. The tiger is in motion, its movement indicated by the placement of its legs (fully extended, each leg visible) and tail, which has swung forward between the two hind legs. The

135 Bronze vessels with inlaid copper décor, including Fong, Great Bronze Age, 280; PI. 70 (late 6-5 c. B.C., Liyu $ii!ift, Hunyuan Co. Í#7j1Í|$, Shanxi); Li Xueqin, Qingtongqi (xia), 24, PI. 30 (Late Springs- Autumns, Gushihou [frlíní^, He 'nan); 45, PI. 54 (Early Warring States, Liyu, Hunyuan, Shanxi); and, different animals executed in similar style, 57, PI. 69 (Early Warring States, Liulige 5ííW#J, Hui Co. )ff-M-, He'nan). In highly abstracted form, this style seems to me to be carried over in several bronzes of the Marquis Yi of Zeng, e.g., Hubeisheng, Zeng Hou Yi, vol. 2, PI. 58:1 (C-108); PI. 59:2 (C-194).

136 In addition, Sanguandian yielded a small gold plaque in the form of a recumbent deer (?), head up, legs tucked beneath its body, forelegs above hind legs in the steppe manner (KG 1985.2, PI. 1:7). This piece is very close to a bronze appliqué plaque from Yuhuangmiao which differs in the form of the antlers (the Yuhuangmiao piece has branching stag antlers; the Sanguandian piece, goat-like horns hanging back over the head) and in the addition of a small Xiongnu birdhead in the place of a tail. The Sanguandian deer is essentially identical to that on the Wudaohezi M9:14 bronze garment hook (WW 1989.2, 56, f. 8:12; 60, f. 17). These pieces evidence steppe influence and link these three sites together.

137 Sanguandian, KG 1985.2, 127, f. 5:1-4.

1J8 Fong, Great Bronze Age, 277-279, PI. 69-70; Zhongguo gu qingtongqi xuan tyMlïfflftffiifc (Beijing: Wenwu, 1976), PI. 59 (Liyu, Hunyuan, Shanxi).

139 "Ningxia Guyuanxian chutu wenwu" ^'aMJU-ScllirhX^J, WW 1978.12, 86-87.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

54 Sophia-Karin Psarras

Chinese version, known from ca. the early 5 c. B.C., follows the usual steppe iconography of a crouching feline gripping the neck of its prey, which is alive and, although caught, appears to be moving away from the feline. This imagery appears for example on a bronze hu from Liyu ^ill§, Hunyuan County y?p $£ ]$, Shanxi, whose second, fourth, and sixth registers carry repeated designs of reclining animals alternating with the feline-and-prey motif (PI. 117, detail). This same prédation imagery is used on the pedestals from the tomb of King Cuo of Zhong- shan, ca. 309 B.C.140 The commonality of this specific image in Jin and Zhongshan suggests to me a strong link between the two cultures, beyond that suggested by other, inscribed bronzes which seem to demonstrate the existence of trade relations between the Xianyun 5ϋ$ΐ, later rulers of Zhongshan, and Jin.141 This image is one of the only non-Chinese elements visible in the material from Cuo' s tomb; the fact that Jin bronzes carry the same image must mean that the same culture influenced both Jin and Zhongshan, whether that culture is attributable

to the Xianyun or to some other steppe people. When related scenes occur in Late Warring States or Han work, such as on a

bronze plaque from the Pingshuo ^$JM- cemetery (Shanxi) 6M50:8 (PI. 119),142 either of these two distinct traditions may have inspired it. In the Pingshuo piece, the tiger's prey is indistinct, but the tiger bears a small birdhead on its back, executed in the Xiongnu style (with ears), which points to a Xiongnu origin for the image as a whole. In other cases, such as Pingyang M 150: 5 (PI. 25) or Guo- xianyaozi M 12:2-1, 2 (PI. 26), where the prey is represented only by its head, the source of the image is less clear. Because the decorative plaque, whether rect-

*4^ Zhongshan: tombes des rois oubliés, no. 17 (bronze with silver and gold inlay).

141 Since Jin was divided into Wei $J, Zhao jg, and Han t$ in 403 B.C., by the time of King Cuo of Zhongshan, it is of course no longer possible to speak of contemporary Jin. However, because the motif of animal prédation with live prey occurs on Jin and, later, Zhongshan work, I term the image a Jin-Zhongshan variant and postulate earlier cultural exchange between these two states. Jenny F. So, "New Departures in Eastern Zhou Bronze Designs: The Spring and Autumn Period," in: Fong, Great Bronze Age, 268, no. 70, mentions a bronze hu with an inlaid copper décor of animals (hunting scenes) in the collection of the Museum fur Völkerkunde, Staatliche Museen Preussische Kulturbesitz, Berlin, whose inscription claims that the vessel was received from the Xianyun. Since this type of décor is related in China to Jin, a further Jin-Zhongshan link

seems clear, regardless of the absolute origin of such inlaid copper scenes. A close relationship between the Jin-successor state of Wei and Zhongshan by conquest, cooperation, or eventually by shared culture is noted by Li Xueqin, Eastern Zhou and Qin Civilizations (New Haven: Yale Uni- versity Press, 1985), 93, 101-103. Note that K.C. Chang, translating the above work, transcribes ψ as "Xi" (idem, 101), while Zhongshan: tombes des rois oubliés gives "Cuo," which I have fol- lowed.

142 Pingshuo kaogudui fWYiVPÀ, "Shanxi Shuoxian Qin Han mu fajue jianbao" '''0$i-&^)L£k £ΑίΙί, HW 1987.6, PI. 3:5 (6M50:8).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 55

angular or irregular in shape, is associated with the Xiongnu in the Far East,143 it is tempting to see plaque depictions of the prédation scene as Xiongnu-derived. It is, however, possible that Zhongshan may have produced plaques, although I have seen none reported; or that to the Xiongnu plaque form was added an element of Zhongshan imagery, already familiar at least to the Chinese. The plaque form is not, in fact, solely Xiongnu, but occurs as well in Saka cultures. The image of the tiger with the dead prey, in contrast, appears to be genuinely Xiongnu. The Xiong- nu plaques therefore provide an immediate example of both the plaque form and the prédation iconography together, and would thus seem the likely source of inspi- ration for the Pingshuo, Pingyang, Guoxianyaozi, and other non-Xiongnu plaques. The fact remains that the Jin-Zhongshan prédation image was already familiar to China before the arrival of the Xiongnu and that it may have continued to inform both Chinese and non-Chinese art. It is the Jin-Zhongshan tradition that seems to have inspired the bronze feline in the round from Sanguandian. Not only is the Sanguandian prey clearly alive, but the tiger, in its recumbant position, seems friendly, playful, qualities more akin to those of the Zhongshan pedestal tigers than the Xiongnu stalking tiger. In addition, the Sanguandian tiger bears on its forehead a cross with two horizontal strokes 41» which recalls the contemporary custom of marking the tiger's forehead with the character wang j£ more than any other, earlier, non-Chinese system of representation.144

Note that, although Sanguandian is one of the few Upper Xiajiadian sites to have yielded horse-gear, no mention of animal bones is made in the preliminary excavation report. The site may be accepted as Upper Xiajiadian because of its hilt

143 Cf. Anatoly I. Martynov, The Ancient Art of Northern Asia (Urbana/Chicago: University of Illinois Press, 1991), 80, where rectangular bronze plaques and belt-buckles as a form are attribu- ted to the Xiongnu. The presence of the form in Siberia (including Minusinsk) and Mongolia is therefore attributed to Xiongnu expansion.

144 The frontispiece of Karl Jettmar, Art of the Steppes (New York: Crown, 1967), shows a bronze plaque of the Yanglang type, said to come from the Ordos region. The tiger's tail is curled up along its back; the head, in high relief, is turned toward the viewer. A live deer lies between the tiger's fore and hind legs. The composition therefore appears akin to the Jin-Zhongshan version of animal prédation, yet we have at present no Jin or Zhongshan plaques of this image, to my knowledge. Closer to the Jettmar Ordos piece are the gold plaques of a tiger, head in high relief, without prey, from Nalin'gaotu âWiSj^, Shenmu Co. tt;fcJä-, Shaanxi: Dai Yingxin |$ß/|Tr, Sun Jiaxiang #Ϊ£#, "Shaanxi Shenmuxian chutu Xiongnu wenwu" KMtt^ítíi}il*Aj&3t#J, WW 1983.12, 24, f. 4-5; 29, f. 11. In "Exploring the North," I advanced a possible Yuezhi at- tribution for Nalin'gaotu, taking the Yuezhi to be a branch of the Saka. Note that although the Nalin'gaotu plaques do not show animal prédation, which seems in fact to be largely absent from Saka art, the form of the tiger is very close to that of the Jettmar Ordos piece and distinct from the form of the Xiongnu tigers. The Jettmar Ordos piece seems to me to draw on several traditions: Xiongnu by iconography, Yuezhi by stylistic treatment, and with a possible Zhongshan reference in the variant iconographie image.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

56 Sophia-Karin Psarras

and red clay tubular fitting, but it gives evidence of many non-Upper Xiajiadian ties, particularly to Wudaohezi, and through it, Yuhuangmiao.145

Comparative Site: Wudaohezi ZEiEMi^

Described as being part of a Warring States cemetery, Wudaohezi is located on the rocky bank of the upper reaches of the Qinglong River |f;ÍÈ, below the Houshan ^lLj, in Lingyuan County l^M-Mc* Liaoning. The site's eleven graves occupy an area measuring 100 m from east to west and 70 m from north to south. A few objects from tombs which had broken open at ground level were found scattered on the site, including bone and turquoise beads, and red pottery shards with a pattern of fine lines. Graves were oriented to between 330° and 350°. All were rectangular pit graves except M5, a cist grave constructed of small stones from the river. M5 included a shelf running along all four sides, and the disintegrated remains of a wood coffin or bier, as did MIO. All graves were lined with birch bark and some (e.g., M10) contained horse teeth placed in a pile or in rows on the grave bottom or in the grave fill. The tombs measured between 1.6-2.1 m in length, 0.5-0.8 m in width, opening at a depth of 1.5-2.0 m below ground level. Burials were single with, at least in M5 and 10, the body of the deceased placed in extended dorsal position.

Here, then, is evidence for nomadism, provided by the inclusion in the graves of horse teeth. Horse bits were also found. The use of horse teeth and the position in which they were buried recalls the practice at Xichagou M^/^J, a mixed Xiongnu-Xianbei-Wuhuan site in Xifeng W^, Liaoning.146 There are few con- nections between Wudaohezi material and that of Upper Xiajiadian, and those con- nections which are evident belong to a zone wider than the specific culture of Upper Xiajiadian. That is, the site's bronze semi-spherical appliqué ornaments with a border décor of fine radiating lines (Wudaohezi M7:14)147 also exist in Upper Xiajiadian (Dabannanshan), Yuhuangmiao (including Xiaobaiyang), Xiongnu sites

145 The Sanguandian ding (KG 1985.2, PI. 1:1) is closely comparable to a Late Warring States ding from Xinzheng $Y(£, Beijing Municipality: Beijingshi wenwu guanlichu dfcJjCrti^'^î'fflÎL "Beijingshi Xinzheng ji de Shang Zhou qingtongqi" JfcSC l'i§iföMfäftM1tfflftffi, WWZLCK 2 (1978), 66; PI. 6:3-4. For the form, and hence the date, cf. Li Ling/Lothar von Falkenhausen, "On the Typology of Chu Bronzes," Beiträge zur Allgemeinen und Vergleichenden Archäologie, Band 11 (1991), Pl. 2:13 (from the tomb of King You |^| of Chu at Zhujiaji 9f.Wk, Changfeng Co. ix-l1, Anhui, r. 237-228 B.C.). This vessel provides a terminus post quern of the late 3 c. B.C. for the leaf-bladed bronze shortsword with bronze hilt unit and regular crown, which form occurs on Sanguandian as well.

146 Sun Shoudao #£pjË, "'Xiongnu Xichagou wenhua' gumuqun de faxian" « iíijfyLVMlnfàSCik » ώ'Κ-ïÎWftAÎÏ, WW 1960.8-9, 25-35; Sun Shoudao, "Xichagou gumuqun beijue shijian de jiaoxun" WtifàU&fflÈkfflNfttftffîil WWCKZL 1957.1, 53-56.

147 Wudaohezi, WW 1989.2, 56, f. 8:26.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 57

(Xigouban M2:l), and into Late Warring States/Early Western Han-dated tombs of the Kayue culture, Qinghai;148 bronze semi-spherical appliqué ornaments with a zigzag border décor (e.g., Wudaohezi M10:17)149 are also found at Xiaobaiyang (M36:2) and the Xianbei site of Yushu (Jilin; M123:8).150 One exception is the large bronze appliqué ornament (12.0 cm χ 1.5 cm) depicting a manna ray, which is related, though not identical to the Nandonggou manna rays (PI. 38). All com- parisons point, at present, to a date of the Late Warring States or later, but do not provide for a clear cultural attribution. Thin, rod-shaped bronze ornaments from Wudaohezi M7151 are pierced at one end for suspension, and carry a décor of horizontal incised lines. They appear similar to examples from Xiaobaiyang (e.g., M6:l)152 and related to pieces from Yuhuangmiao (cf. PI. 110, left). The gold foil half-moon ornaments (Wudaohezi Ml:38),153 the garment buckle in the shape of a recumbent deer (Wudaohezi M9:14),154 and the appliqué ornaments in the form of recumbent horses (e.g., M8:3, M9:7, M 10:4, see PI. 114) all match pieces from Yuhuangmiao (PI. 113) or Xiaobaiyang (PI. 115). Because of this close association and in the absence of typically Upper Xiajiadian material, I attribute Wudaohezi to the Yuhuangmiao culture despite its geographic location. It should be borne in mind that 30% of Yuhuangmiao tombs are cist graves, although they are not pictured in published reports.155 It may be, therefore, that the single cist grave at Wudaohezi is comparable to those of Yuhuangmiao. Regardless of cultural attribution, comparisons with Yuhuangmiao material

provide a Late Western Han date for Wudaohezi despite this site's Middle to Late Warring States Chinese material. The latter includes a bronze ge Ml:43,156 small niuzhong iftf^-type bells with two registers of décor, e.g. Ml:41,157 and four bronze swords with cap pommels, hilts round or ovoid in cross-section, narrow rectangular guards, and blades rhombic in cross-section with or without a pro- nounced median ridge.158 The marked presence of Chinese material culture in the

148 See note 128.

149 Wudaohezi, WW 1989.2, 56, f. 8:25.

150 Jilinsheng, Yushu Laoheshen, 69, f. 62:2.

151 Wudaohezi, WW 1989.2, 55, f. 6:7.

152 Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 48, f. 15:19.

153 Wudaohezi, WW 1989.2, 60, f. 18.

154 Wudaohezi, WW 1989.2, 60, f. 17.

155 Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 21-22.

156 Wudaohezi, WW 1989.2, 57, f. 10:3.

157 Wudaohezi, WW 1989.2, 57, f. 10:1.

158 Wudaohezi, WW 1989.2, 56, f. 8:9, 10.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

58 Sophia-Karin Ps arras

form of these swords is a prominent divergence from both Upper Xiajiadian and Yuhuangmiao sites, where, apart from ge, Chinese material seems limited to luxury items not of practical necessity.

Despite the difficulty of presenting a definition of Upper Xiajiadian, a co- hesiveness in the material from various sites attributed to the culture is readily apparent. This establishes not only attribution, but general contemporaneity for the material.

PART II

The Objects Ceramics

The majority of Upper Xiajiadian ceramic forms may in general be divided into three types: northeastern forms, found not only on Upper Xiajiadian sites but also on Xianbei and Yuhuangmiao culture sites as well as, at times, Pingyang; forms which now appear limited to Upper Xiajiadian; and forms which belong to the Liangcheng-Northeastern zone.159 What I have, for convenience, termed jar type Ilia, a jar with a rounded belly, long neck, and relatively wide mouth, occurs on Shangmashi M3:2, Zhoujiadi Ml:4, Zhengjiawazi M65 12:63, and Liangjiashan M5. It also occurs on the Xianbei sites of Tiejianggou AM3:1, M2:l, M2:2, Wangong M1B:72, and Yushu M42:l, M105:2, as well as on the Yuhuangmiao- culture site of Xiaobaiyang M12:5 and M35:2 (PI. 39).160 Type IIIc, a jar with a wide mouth, short everted neck, and a lower body which curves inward somewhat sharply to the base, is found on Shangmashi JBM11:1, Zhoujiadi M31:l and M42: 8, and, in a shorter form, at the Upper Xiajiadian site of Longtoushan. The form occurs frequently at Xiaobaiyang (e.g., M30:14), and on Yuhuangmiao (M190:l): it also appears on the Xianbei site of Yushu M27:4, and in Pingyang M135:38, M145:5, M153:8 (PI. 40, 41). 161 Type Vb, ajar with a wide mouth, moderately long neck, wide shoulders, and a narrow base, has a broad geographic distribution: the Shangmashi JBM11 coffin jar, the only Upper Xiajiadian example; Xiaobaiyang M21:l, M41:8; Maoqinggou M23:l; and Daodunzi M26:l (PI. 42)162 Type χπ, a wide-mouthed jar with a long neck and a moderately curved body, is found on Shangmashi M4:l, again the only Upper Xiajiadian site; the

159 In "Exploring the North," I used the term "Liangcheng-Pingyang" to designate the area extending in a belt from Liangcheng Co., Inner Mongolia to the site of Pingyang, Tailai Co., Heilongjiang. I here use an alternate term, "Liangcheng-Northeast," as a simplification.

160 See note 106.

161 Zhoujiadi, KG 1984.5, 420, f. 7:4, 1; Jilinsheng, Yushu Laoheshen, PI. 11:4; Heilongjiangsheng, Pingyang, PI. 37:4; p. 163, f. 92, second column from right, bottom row; PI. 22:3.

162 Maoqinggou M23:l, Tian/Guo, Ordos, 256, f. 19:4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 59

Xianbei site of Wangong M1B:62, Sanjiazi scattered find #5, and Jreim #2047; as well as the Liangcheng County S^tfilc-H·» Inner Mongolia sites of Yinniugou M9:l and Guoxianyaozi M25:l (PI. 43). 163 Type XIV, a wide-mouthed jar with a short flaring neck, sharply defined but only moderately wide shoulders, and a narrow base, is also found on Shangmashi JBM1:1, JBM9:1, JBM14 (the coffin jar itself), as well as at Liangjiashan M5. It is also present on the Xianbei sites of Zhalainuoer %UM i^^K and Menggen ^T.ff|, as well as at Budonggou Ml:l, a site whose Xiong- nu attribution I accept, but which may bear modification given its Xianbei affinities

(PI. 44). 164 Type XXII, a wide-mouthed jar with a relatively long neck, and up- ward-curving shoulders, is found on Shangmashi JBM13:1 and at Pingyang M104:9 (PI. 45). Type XXIII is a squat-bellied jar with a small foot, lugs attached to the widest part of the belly, and a cup-like neck. It occurs on Shuangfang and Erdaohezi (PI. 46). A short bowl-like pot with lugs, which I identify as bo %$ according to the

terminology of the Xiaobaiyang excavation report, occurs in Xiajiadian M12:7, Zhoujiadi M41:l, Longtoushan (PI. 47), and perhaps Nandonggou.165 The form is essentially the same as a "Λ#" w. from the Upper Xiajiadian site of Kuntouyingzi i^^k^-f' and is close to Xiaobaiyang M15:l.166 Footed bowls include dou-l _&, a deep, curved bowl set on a broad, low foot (Xiajiadian F4:l; Erdaohezi, two examples; PI. 48); dou-ll, long-stemmed with a moderately shallow bowl with fairly straight lower angles (Xiajiadian fragments H9:4, T3-4:37, and Tianjuquan ^11%. in Ningcheng County fÄä, Inner Mongolia; PI. 49, left);167 dou-lll, very long-stemmed with a shallow bowl that is somewhat square when viewed in cross-section (Xiaoheishigou, not pictured); and dou-YV, with a deep, well-rounded bowl set on a moderately short stem which curves inward from the foot to the bowl, unlike the stem of dou-l, which curves outward (PI. 51). Dou-TV is a common Chinese form, but it is similar in shape to a piece from the Upper Xiajia- dian site of Kuntouyingzi.168 The date for dou-TV is uncertain. Reports attribute it to the Late Springs-Autumns/Early Warring States (Yan Xiadu M31:7),169 to the

163 Heilongjiangsheng, "Qiqihaershi Dadao Sanjiazi," KG 1988.12, 1092, f. 4:7; Zhang Bozhong, "Zhelimu," WW 1981.2, 9, f. 1:6; Neimenggu, "Guoxianyaozi," KGXB 1989.1, 64, f. 10:10.

164 Cheng Daohong, "Yiminhe," NMGWWKG2 (1982), 19, f. 3:1-3.

165 Xiajiadian, KGXB 1974.1, 131, f. 23:8; Zhoujiadi, KG 1984.5, 420, f. 7:5; Nandonggou, KG 1977.6, 373, f. 1:4.

166 Kuntouyingzi, WWZLCK9, 39, f. 39; Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:19.

167 Tianjuquan, see Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:38.

168 Kuntouyingzi, WWZLCK9, 39, f. 36:1.

169 Hebeisheng, "1964-1965nian Yan Xiadu," KG 1965.4, 550, f. 3:4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

60 Sophia-Karin Psarras

Middle Warring States on the Zhongshan habitation site 2,170 and to the Late Warring States on Yan Xiadu ~fU site 22, T4:3:22.171 It is found in the Late Warring States/Early Western Han period at the site of Fushun ^cM (PI. 51). 172 While dou-' is limited to Upper Xiajiadian, the presence at Fushun of a form similar to the Xiajiadian M9:4 fragment classified with dou-ll infers a Late Warring States/Early Western Han date for the Xiajiadian habitation site. Dou-lll, also common in China in Middle and Late Warring States Yan,173 Middle Warring States Zhongshan,174 and Middle Warring States Zhao,175 as well as in Yuhuang- miao M61:4 (PI. 50). 176 An untyped dou was found on Longtoushan, with a moderately deep, well-curved bowl set on a nearly straight foot. The bowl has two horizontal lugs (PI. 52). Horizontal lugs also appear on a // üj from Xiajiadian Τ 1-2: 7 whose form is

strongly vertical (// type II, PL 53, upper right). Another form of li (li-l) occurs at Nanshan'gen M101 in bronze, with a pronounced outward curve over the body (PI. 53, upper left). The Nanshan'gen M 101 li raises issues of intercultural influence because of the form, not of the body, but of the handles. These are in the shape of felines with raised tails, hind legs braced on the body of the //, forelegs on the rim. Such handles occur in China over a wide range of time and, in the absence of a similar body form, may not provide conclusive evidence of contact as early as the Late Springs- Autumns/Early Warring States. A ceramic piece with a similar body

170 Hebeisheng, "Hebei Pingshan Sanji," KGXJK5, 165, f. 9:4.

1/1 Hebeisheng, "Yan Xiadu di 22hao," KG 1965.11, 564, f. 2:2.

172 Wang Zengxin, "Liaoning Fushunshi," KG 1964.12, 292, f. 6:19.

173 Hebeisheng, "Yan Xiadu di 22hao," KG 1965.11, 564, f. 2:4 (Tl:3:15); Zhangjiakoushi wenwu shiye guanlisuo SjS^U 'ih'XW'l^/S'Mßf, "Zhangjiakoushi Baimiao yizhi qingli jianbao" S£^c WitiÙJiSTâikffiMtniU, W1985.10, 24, f. 4:9, 11 (fragments T8-2:l, T2-2:2).

174 Hebeisheng, "Hebei Pingshan Sanji," KGXJK5, 165, f. 9:2 (habitation site 2, no number).

175 Hebeisheng, "Handan Baijiacun," KG 1962.12, 620, f. 8:5 (M40:4). Zhu Yonggang 5fe/KWJ, "Xiajiadian shangceng wenhuade chubu yanjiu" SÍMJÍd-tMyíítU^JJ^Wtíí, in: Su Bingqi iftSfcífi (ed.), Kaoguxue wenhua lunji ^ÎS'^XikYëia, vol. 1 (Beijing: Wenwu, 1987), 114, f. 10, upper row, compares a globular-bowl dou from late-era Hongshan iX''' (Chifeng Muni- cipality, Inner Mongolia) to one from Jiagezhuang 5ϊ#Ίΐ (Tangshan )íf][|, Hebei). This is probably Jiagezhuang M23:12, An Zhimin, "Tangshanshi Jiagezhuang," KGXB 1953.1-2, PI. 5:3, which has a rounded lid not pictured by Zhu Yonggang. The form without a lid occurs at Yushu, especially M22:5, Jilinsheng, Yushu Laoheshen, PI. 12:4, with a Late Western Han/Eastern Han date.

176 Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 26, f. 16:6 (Yuhuangmiao M61:4).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper XiAJiADiAN 61

form does exist at Xiaobaiyang H3:15.177 L/-II is closely related to a ceramic piece from Dapaozi ^cy{3~P·178

While li-l and II are not both attested in bronze and ceramic versions, ding-l ^' yields examples in both. I define the type as having a broadly curved bottom leading to a mouth wider than the immediately lower body, two vertical circular handles attached at the rim, and long, slim legs (PI. 54, left). An unattributed rim fragment published by Jin Fengyi179 matches an example from Zhoujiadi M45:l (approximately 18 cm in height), while Nanshan'gen M101 yielded the bronze of the same type. Differences in the handle-rim area of the Nanshan'gen M 101 piece are not great enough to warrant creation of a new type at present. Another ding with similar features will, however, be classified with ding-ll: that is, Xiajiadian H5:15, whose belly is deeper and more vertical than that of ding-l, and whose legs are likewise heavier (PI. 54, right). Xiajiadian T7-l:2, with a smaller mouth and lower belly, is also classified as ding-ll for the purposes of the present work. A similar form exists at Yuhuangmiao M52:l. This comparison of ceramic forms yields some evidence for dating:

Jar type Ilia: through Wangong, Yushu, Xiaobaiyang, a date of Late Western or Eastern Han.

Jar type IIIc: through Yushu, Yuhuangmiao, Xiaobaiyang, a date of Late Western or Eastern Han.

Jar type Vb: through Daodunzi, Xiaobaiyang, a date of Middle to Late Western Han. Jar type XII: through Wangong, Sanjiazi, Jreim, a date of Late Western or Eastern Han. Jar type XIV: through Zhalainuoer, Menggen, Budonggou, a date of Late Western or

Eastern Han.

Bo: through Xiaobaiyang, a date of Late Western Han. Dou-ll: through Fushun, Late Warring States/Early Western Han. Dou-lll: through Chinese sites, Middle to Late Warring States, but through Yuhuangmiao,

Late Western Han.

Ding-Ill: through Yuhuangmiao, Late Western Han. Zi-II: through Yuhuangmiao, Late Western Han.

This proposed dating is viable because closely similar shapes are not, apparently, found at sites positively datable to an earlier period; at the same time, those sites where these forms exist are datable as given above. Where comparisons are approximate, error may be expected. However, the dates proposed above are borne out by other material, as will be seen below. Negative evidence has importance as well. Absent from Upper Xiajiadian are not only Lower Xiajiadian but also Western Zhou ceramic forms. One exception is Shangmashi JBM12:1, ajar with a

177 Zhanâjiakoushi, "Xiaobaiyang," WW 1987.5, 43, f. 3:13.

178 Jin Fengyi, "Xiajiadian shangceng," KGXB 1987.2, 183, f. 2:1; Dapaozi, WW 1984.2, 52, f. 3:3.

179 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:33.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

62 Sophia-Karin Psarras

broad, slightly flared mouth, a moderately long neck, a sharp downward slope to the greatest width of the belly, and a fairly broad base.180 This shape, essentially that of Shangmashi JBM13:1 (type XXII, PI. 45), is related to vessels (one with a lid) from the Early Western Zhou site of Liulihe ïj&WM, Fangshan County J^lll-M-, Beijing Municipality (PI. 55, 56). The Liulihe version tends to be broader and squatter, however. Ding with stubby legs exist in Lower Xiajiadian (e.g., Dadianzi ^íãj-f1, Chifeng Municipality τ}ρί%, Inner Mongolia),181 but the body form is distinct from the Upper Xiajiadian ding-U. Likewise, the Lower Xiajiadian- Zhukaigou %:jffä (Yijinhuoluo Banner $'&Wi$M, Inner Mongolia)182 zone li with voluminous legs flaring low from a narrow mouth, and terminating in small points without forming distinct legs (PI. 57), does not exist in Upper Xiajiadian. (Its distribution in both Lower Xiajiadian and Zhukaigou follows the Liangcheng County-Northeast zone discussed in "Exploring the North.") Other forms characteristic of Lower Xiajiadian, such as the elongated // (PI. 58), the short flaring jar (PI. 59), and the broad-bodies jar with a small mouth and foot (PI. 60), found both in painted and undecorated versions, are equally absent from Upper Xiajiadian, as is the footed bowl attributed to Lower Xiajiadian but also found on Late Shang/Early Western Zhou sites such as Liulihe (PI. 61). 183 Upper Xiajiadian ceramic comparisons do demonstrate the close inter-

relationship among the various sites of the culture and, in many instances, with the

Yuhuangmiao culture. Ties with Liangcheng County, Inner Mongolia, are present, as are ties with other northeastern cultures. These relationships are important not only in dating Upper Xiajiadian, but in identifying and defining the culture.

Specifics on clay type or production modes are often lacking in the archaeo- logical reports. Although siliceous red-brown earthenware is often associated with Upper Xiajiadian, several types of clay are used: black-brown and red-brown siliceous clay at Erdaohezi; black at Shuangfang; brown at Nandonggou; black, brown, and grey at Shangmashi; red at Liangjiayingzi M8071; red-brown at Men- lian; red and brown at Xiajiadian; apparently red-brown at Xiaoheishigou; red, brown, with grey-black engobe décor at Zhoujiadi. Most of the clay is siliceous.

180 Shangmashi, KG 1982.6, 592, f. 2:9.

181 Cf. also Neimenggu wenwu gongzuodui 'H'Mk'SC$jS-W$ki "Aohanqi Fanzhangzi gumuqun fa- juejianbao" ΜΚΨΜϋΨ^Μ^^^Μ^ NMGWWKG 3 (1984), 22-25.

182 Neimenggu wenwu yanjiusuo |^jla< U*^WW^0í' "Neimenggu Zhukaigou yizhi" fy'Milh'i^Jffà iSih, KGXB 1988.3, 301-332, PI. 1-8.

183 Zhongguo shehui kexueyuan, "1981-1983nian Liulihe," KG 1984.5, 411, f. 6:7 (M1126:2). This same form is given in Beijingshi, Beijing kaogu sishinian, 32, f. 12:3, 4 as coming from a Lower Xiajiadian tomb in Liulihe (Ml). An Early Western Zhou example is on display at the Beijing University Sackler Museum with no listed provenance. The form has a reasonably broad distribution.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 63

There are indications of wheel-thrown pieces at Shangmashi and Menlian, and of wheel-finishing of the rim at Shangmashi. Xiajiadian pieces were hand-built and fired low. Only at Shangmashi does there appear to be a correlation between clay type and production: black siliceous earthenware is said to be thrown; brown, to be hand-built, polished, and fired high; and grey, to have a hand-built body and a wheel-thrown or wheel-finished mouth. The Shangmashi pieces raise the question of their production as burial rather than utilitarian objects. The excavators suggest that because of their size, all Shangmashi ceramics were made as mingqi RJj^#. Their measurements do tend to be small: M4:l (type XIII) at 15 cm high, 8.6 cm in diameter at the mouth, 6.0 cm in diameter at the base; M3:2 (Ilia) at 21.5 cm χ 9 cm or 7.4 cm; JBM9:1 (XIV) at 14 cm χ 9 cm or 5.2 cm. However, these measurements are comparable overall to those of other sites: Erdaohezi, with ajar measuring 16 cm high and 14 cm in diameter at the widest point; a Shuangfang jar measures 16.7 cm high, 14.7 cm in girth, 8.5 cm at the mouth, and 6.4 cm at the base; Zhoujiadi jars range from 16 cm (Ml:4) to 8.1 cm (M4:l) in height, and the Zhoujiadi ding M45:01 measures 12.8 cm high, though the legs are broken. Dimensions of pieces from the Nanshan'gen habitation site are not given in the reports. If the scale used in the report illustrations is correct, the pieces are extremely small (e.g., bowls of some 10 cm in height). This seems to indicate that small-sized ceramics had utilitarian function in this culture and that, therefore, small pieces in graves are not necessarily mingqi.

Shortswords

Jin Fengyi184 argues that Upper Xiajiadian shortswords are divisible typologically, and therefore chronologically in a linear evolution, according to: 1) the form of the

184 Especially Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 384-426; and Jin Fengyi, "Xiajiadian shangceng," KGXB 1987.2, 177-208. Lin Yun has made the same arguments with nearly the same range of dates, although Lin is willing to extend the lower horizon into the Han period. Lin includes in his study non-Chinese shortswords which have nothing in common, in form or distribution, with Upper Xiajiadian types. He rightly rejects a Donghu attribution for Upper Xiajiadian, but suggests an equally unascertainable attribution to "the ancestors of the Huimo ... Zhenfan and Chaoxian." See Lin Yun #£, "Zhongguo dongbeixi tongjian chulun" ''xW£ÀtãMM$]'b, KGXB 1980.2, 139-161. Other articles on the subject include Zhai Defang ffifëî?, "Zhongguo beifang diqu qingtong duanjian fenqun yanjiu" 'hESJb^/iÈEW^^Iffiâ'J

ΙΥψβϊίΙ, KGXB 1988.3, 277-299. Clearer than Jin Fengyi, Zhai nonetheless considers chrono- logically and culturally distinct shortswords as representing a single type and still posits a linear evolution. Chi Lei }IA%' "Guanyu quren qingtong duanjiande ruogan wenti" ^TlilJyUWÍM &Mf]^]r''i^Sâ, KG 1982.1, 54-59, like Lin Yun, accepts a lower horizon "slightly later than the Late Warring States" for Upper Xiajiadian, but only for (my) type I(b)-elongated shortswords.

See also, with the same shortcomings: Liu Guanmin >fiJlK, Xu Guangyi '&%%> "Neimenggu dongbu diqu qingtong shidaide liangzhong wenhua" í^^^v^itfeEWfín^WMW^fc, NMGWWKG 1 (1981), 5-14; Tao Zongzhi l^-fê, "Shilun Zhangjiakou diqu Zhanguo yiqiande kaogu wenhua yicun" Kifc&*UJâEAffi^^ Beifang wenwu 't)TXfy]

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

64 Sophia-Karin Psarras

blade, 2) the form of the crown (counterweight), 3) the form and especially angle of the pommel. It is impossible to accept Jin's highly detailed schema. For instance, as to the form of the blade: straight blades and leaf-shaped blades coexist in the same tomb (e.g., Nanshan'gen M101 and Xiaoheishigou); elongated leaf- blades and the more rounded variation coexist (Nanshan'gen M101, Zhengjiawazi, Shangmashi). These forms therefore do not supercede one another and do not provide a linear succession of dates. As to the form of the crown, what Jin categorizes as "early" forms, with a regular, pronounced double curve marked by deeply concave horizontal ridges (Jin type II),185 coexist with completely irregular forms classified by Jin as "late" (e.g., Zhengjiawazi M6512, PI. 66). 186 Concern- ing the form and angle of the pommel, a regular crown with no remaining hilt unit coexists with a Jin type T-IV (hence, very late) hilt unit with an irregular crown (Shangmashi M3).187 According to Jin's own analyses, the regular crown with vertical, rather than horizontal ridges (e.g., Wujintang), cannot appear late in the development of the crown, as Jin proposes,188 since it occurs with a leaf-blade with

no extant hilt unit. To be a late development, as it is according to Jin's theories, this crown should occur with a bronze hilt unit. In addition, shortswords found in

Hebei189 carry an Upper Xiajiadian regular leaf-blade combined with a Xiongnu- type rectangular hilt, double-spiral pommel, and straight rectangular guard (Xiongnu type II) (PI. 77, 1-4 from left), found commonly in the Yuhuangmiao culture (cf. PI. 77, right, PI. 71, 72). This demonstrates the contemporaneity of the leaf-blade with Yuhuangmiao, since the Xiongnu-type shortsword is not attested in

northern China prior to the Xiongnu and, indeed, in the northeast, not prior to Yu- huangmiao. The hilt décor is more common at Yuhuangmiao than on Xiongnu sites. This provides, broadly speaking, a date of the very late 4 c. B.C. (earliest appearance of the Xiongnu) as an upper horizon for the hilt unit and, more specifically, a probable date of Late Western Han (via Yuhuangmiao). Jin applies his typology-chronology to all sites, regardless of the fact that other

material on those same sites is identical. He thereby eliminates all variants, claiming that only one mould was used for shortsword production at any given

1994.2, 14-22; Liu Bing, "Shilun Xiajiadian shangceng," NMGWWKG 1992.1-2, 26-33; Wu En AS, "Guanyu woguo beifangde qingtong duanjian" ^^WAt1ïttii5Wï%L&l KG 1978.5, 324-360 (as well as Wu En, "Woguo beifang gudai dongwu wenshi" 'íkMlfâlkV^tyJkkfâfi , KGXB 1981.1, 45-61, PI. 3-6, which includes some references to Upper Xiajiadian).

185 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 418, f. 14:2.

186 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 418, f. 14:15.

187 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 417, f. 13.

188 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 418, f. 14:14.

189 Zheng Shaozong %ï&%, "Hebeisheng faxiande qingtong duanjian" Î"ttt'Ë!fcMfâiÎfflÂL&], KG 1975.4, 226-227, 248, PI. 2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 65

time. At the same time, other material (particularly ceramics, ornaments, decora- tive patterns) remained unchanged over long periods. He thereby creates artificial chronologies and a false developmental schema. Unfortunately, perhaps because his work is presented in overwhelming detail, his system has not been closely analyzed, and has become widely adopted both in China and in the West. Given the coexistence of essentially all shorts word forms, any typology must not be seen as chronological. Some terms are needed, however, to discuss the interrelationship of shortswords. I had once thought to find less cultural variation in swords than in any other form of material culture because swords represent technology, and I expected technology to spread quickly among interacting cultures. I have found instead that, among the non-Chinese peoples of present-day northern China, shortsword forms or décors are remarkably culture-specific, with a distribution far narrower than that

of ceramics or ornaments. Therefore, shortsword forms, including decorative motifs, demonstrate intercultural influences far more clearly than most other material. I propose the following typology:

Type I(a): Regular leaf-blade (generously curved shoulders narrowing at approximately one- third of the length of the blade, before flaring out into sharp horizontal points, narrowing again before a minor flare, narrowing finally to the point; the blade has a prominent median ridge) with a bronze hilt unit sometimes seemingly cast as part of the blade rather than fitted separately. This may take the form of a hilt with a large semi-ovoid pommel whose upper half is marked by a series of raised bands (Xiaoheishigou, PI. 91, second row, right) or the hilt, with no discernable pommel, may be an irregular shape with an animal or other décor (e.g., Nanshan'gen M101:36, with mirror-image crouching felines, PI. 68, left, or Nan- shan'gen eastern sector habitation site, with a nude man and woman on opposite sides of the hilt, PI. 19. A shortsword with a hilt décor similar to that of Nanshan'gen M101:36 was found on Beishanzui M7501:20; its blade, however, is irregular, without the curves typical of Upper Xiajiadian).

Type I(a)-smooth: refers to the modified leaf-blade, without side points. The blade is otherwise generally leaf-shaped, and the hilt unit seems to have been cast with the blade. The hilt is conical, with or without a hole through the end. Examples: Xiaoheishigou (PI. 62, left), Nanshan'gen M 101 (no hole visible, PI. 62), Dapaozi (no hole, PI. 62), and a Liaoning stray find (no hole). 190

Type I(b): regular leaf-blade with separate hilt unit, with or without crown. The hilt unit may be of wood, in which case it has disintegrated. The iron-ore (or sometimes bronze) crown may be combined with a wood or with a bronze hilt unit. Both types of hilt rely on wood and probably hemp strings for their attachment to the blade tang. Crowns may be attached by various means to the pommel. Most fit into concavities and are held in place by small projecting "pins" on the pommel. Hemp twine may have reinforced this with exterior binding over the crown and pommel. The side of some pommels bear a décor of narrow crossing bands at the center of the pommel, as if recalling twine.191 This is, however,

190 WWZLCK9, 37, f. 32:1.

191 E.g., Sunjiagou M7371:15, WWZLCK9, 34, f. 28.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

66 Sophia-Karin Psarras

speculation on my part. Nonetheless, it is to be supposed tht the crown was bound with twine to the hilt unit in those cases where crowns were found with blades whose hilt units have

disintegrated. The use of bronze or wood for the hilt seems determined, not by chronology (the two coexist in Nanshan'gen M101, for example), but by less tangible factors, perhaps including cost. We have no way of knowing how the wood hilts appeared. Décor on the bronze hilts is frequently what I term "L- triangle": a series of triangles, often combined with or made up of rectangles, squares, and L-shapes (see, for example, PI. 63). Another common décor is a herringbone pattern: a series of fine diagonal lines, slanting toward the left and toward the right in alternate registers (e.g., PI. 65, right). These patterns also occur on mirrors, axes, ornaments, and other objects. Examples: Nanshan'gen M101, Menlian, Zhengjiawazi M6512:l, Beishanzui, Wujintang, Erdaohezi, Shi'ertaiyingzi (where Zhengjiawazi M6512: 1 and Shi'ertaiyingzi examples had a crown; see PI. 64. )192

Type I(b)-elongated: in this case, the angles of the blade are less pronounced than in type I(b). All other type I(b) characteristics apply. Examples: Sunjiagou M7371:15 (with crown

and bronze hilt unit),193 Shangmashi M3:l (crown, bronze hilt unit), Shangmashi M2:l, Wangyingzi -f.^^f (Ningcheng County, Inner Mongolia; with crown, bronze hilt unit),194 Nanshan'gen M101, Nanshan'gen eastern zone habitation site, Zhengjiawazi M6512:2 (ir- regular crown), Zhengjiawazi M6512:33 (regular crown), Liangjiashan Ml, M3, Dadun ^4î (crown, bronze hilt unit), Nandonggou (crown, bronze hilt unit), Sierbao ^fJLJÉ (Jinxi

County #ííHil·' Liaoning; crown, bronze hilt unit; see PI. 65, 66. )195

Type I(b)-low: indicates that the points on the blade occur near the tip, rather than near the middle of the blade (near one-quarter of the length of the blade up from the point, rather than

approximately one-third). Other I(b) characteristics seem to apply. Examples: Shuangfang (PI. 67). 196

Type II (a): a full-length sword, straight-bladed, with a protruding, downward-angled guard, and a conical hilt. Examples: Xiaoheishigou (PI. 91, bottom, right).

Type II (b): a full-length sword with a straight blade capped by a thin guard, and a conical hilt (with a hole near the end). Examples: Xiaoheishigou (PI. 70, bottom), Longtoushan (blade broken, PI. 85).

192 Nanshan'gen M101:37, KGXB 1973.2, PI. 6:1; Zhengjiawazi M6512:l, KGXB 1975.1, 144, f. 4:1; PI. 4:1 (far left); Beishanzui, WWZLCK9, 54, f. 44; Wujintang, KG 1960.5, 7, f. 1:3; 8, f. 2:3; Erdaohezi, G 1977.5, 303, f. 1, f. 2:1; Shi'ertaiyingzi, KGXB 1960.1, 65, f. 2; PI. 1:1.

193 Sunjiagou M7371:15, WWZLCK9, 34, f. 28.

194 Wangyingzi, WWZLCK9, 38, f. 34.

195 Si'erbao, KG 1964.12, 277, f. 1:1.

iyo Type I shortswords occur in Korea as well, but without access to Korean reports, I have been unable to ascertain their distribution, the context in which they occur, or their relationship to Upper Xiajiadian (e.g., as the source of the form? as imports? as the local product of a form borrowed from Upper Xiajiadian?). Sarah Milledge Nelson, The Archaeology of Korea (Cam- bridge: Cambridge University Press, 1993), 133-135, mentions the leaf-bladed shortsword (which she calls a "dagger"), with no discussion of context or of the date assigned to the sites on which the shortswords were found. For a map of Korean distribution, see Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 242.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 67

Type III: the "notched" shortsword, with a straight blade usually with a prominent median ridge, a rectangular hilt, and a small guard jutting out just above the blade so that a small space is formed between the blade and guard (the "notch"). The hilt may be decorated or plain, and may or may not have a pommel. Examples: Nanshan'gen M101 (PI. 68), Beidamian M6:5 (Dongnangou; PI. 82, right), Xiaoheishigou (PL 70, upper right), Wafangzhong M791:190, Tianjuquan M7301:4, and in scattered Liaoning finds.197

Type IV: the "T-pommel" shortsword, with a broad, straight blade; an essentially Xiongnu- type guard, whether broad and straight (type II) or downcurving (type I); a broad flat hilt with or without décor; and a horizontal bar pommel extending ony slightly beyond the hilt. Examples: Xiaoheishigou (bronze, with a raised-line décor of a row of deer facing the blade, PI. 81, left), Hulusitai Ml:3 (bronze, ridged décor over hilt and pommel, hilt with central vertical ridge, PL 80, center), Yinniugou Ml:2 (iron, no décor, PL 84), Chang'an weapons depot 7:2:2 (iron, no visible décor, PL 83, right). As discussed above ("Xiaoheishigou"), the Chang'an example of this shortsword dates to no earlier than the era of Wang Mang, including his regency (ca. 6-25 A.D.). This provides, conservatively, a general date of the very late Warring States as the possible upper horizon for the form, and ca. 25 A.D. as the lower horizon. This is presently the only instance I have found of a Xiongnu shortsword form within Upper Xiajiadian. The Xiaoheishigou piece carries a northeastern décor, how- ever.

Type V: Not occurring in Upper Xiajiadian, but important for comparative purposes, this type refers to a shortsword with a straight blade with multiple median ridges; spurs jut out at the shoulders of the blade; there is no guard. The hilt is flat and thin, and has the appearance of having been folded forward to meet at a central seam, opening into the median ridges as the hilt meets the blade. The pommel is flat and circular, like a nail head. This type of pommel is common to swords in northern China during the Warring States and Western Han. Examples: Xiaobaiyang M39:5, Beixinbao Ml:83 (PL 86; for a Han example of the pommel, see PL 73, left).198

While it is rare in Upper Xiajiadian to find shortsword forms from other cultures, even those used by Yuhuangmiao, shortsword décor does show cross-cultural influence. Type II provides the most information, together with some examples of types I(a) and IV. A number of décors used in Yuhuangmiao are shared with other cultures. One of the most prominent of these is the double-spiral pommel. The term, as I use it, refers to both the form of the pommel, which consists of two circles rising from each corner of the hilt end, and to their décor, one of concentric

circles (e.g., PL 71, third from right; PL 72, third from right; PL 74, fourth from

left and second from right). In some cases, concentric circles are in openwork (e.g., PL 71, second from right; PL 72, third from left). This pommel décor is often combined with a linear hilt décor consisting of two vertical bands framing an openwork pattern of alternating triangles or squares, for example, as on PL 72,

197 Dongnangou, KG 1977.1, 53, f. 6:1; PL 7:2; Wafangzhong M791:190, WWZLCK9, 55, f. 56; Tianjuquan M7301:4, WWZLCK9, 57, f. 67; Liaoning finds, WWZLCK9, 37, f. 32:4.

198 Hebeisheng, "Beixinbao," KG 1966.5, 235, f. 6:2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

68 Sophia-Karin Psarras

third from left, and PI. 74, fourth from left and second from right. The outer vertical bands on the hilt may include or be replaced by vertical zigzags as on Yuhuangmiao M22:6, M70 (no number), and M257:4 (PI. 77, right).199 These hilt and pommel décors occur on Xiongnu-type shorts words from Xiaobaiyang (e.g., M 12:4, PI. 75, left) and as the hilt unit coupled with the Upper Xiajiadian leaf- blade on finds from Hebei (PI. 77, left). They occur on a type III notched shortsword from Xiaoheishigou (PI. 70, upper right). A related variant occurs at Beixinbao. Beixinbao Ml:76 (PI. 78), Xiongnu in form, has a cannelated hilt, and a cannelated double loop pommel. Each spiral of the pommel emerges from a circular shape ornamented with concentric circles. Although the cannelated hilt itself does not occur with the double-spiral pommel in Yuhuangmiao, as far as I know, it does occur with an openwork circular pommel within which symmetrical loops curve to each side (e.g., PI. 71, right, also visible on PI. 72, left). This can- nelated hilt/circular openwork pommel combination occurs on Xiaobaiyang as well (e.g.,M30:13, PI. 75, right). The openwork circular pommel also exists in the form of a curled animal or of

two animals facing one another, common in Yuhuangmiao (e.g., PI. 74, second and third from left, far right; PI. 71, left and second from left). Variants whose inner-pommel patterns lie between the animal and the vegetal, as on PI. 71, third from left (Yuhuangmiao), and PI. 75, second from right (Xiaobaiyang M37:l), are also found on the Hebei- Yuhuangmiao-Upper Xiajiadian combination shorts words (PI. 77, left and second from left) and on a curved knife from Xiaoheishigou (PI. 89, left).

Beyond the northeast, the cannelated hilt décor appears most often at Maoqinggou. Although the Maoqinggou shorts words are Xiongnu in type, only a modified cannelated hilt occurs on presently-known Xiongnu sites within modern China, namely on the type IV Hulusitai Ml:3 bronze shortsword. Maoqinggou cannelated hilts may be combined with an openwork double-spiral pommel (M60:6, PI. 80, right); in this example, the outer vertical columns of the hilt are marked

with a pattern of fine lines, while the central column carries a herringbone-type pattern of nested v-shapes. This pattern occurs in the broad central band of Yuhuangmiao M 164 (no number) (PI. 76, left). It occurs on type III shorts words from Nanshan'gen M101:33, 32 (PI. 68, third and fourth from left), as well as on a curved knife from Xiaoheishigou (PI. 81, third from left). Other cannelated-hilt

shorts words from Maoqinggou carry a double birdhead pommel, a form occurring on Xiongnu (Gongsuhao y£^të§>, Yijinhuoluo Banner, Inner Mongolia, Ml:5; PI. 79, second from left)200 and Saka (e.g., Tombs of the Bronze Shorts words,

199 Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 28, f. 17:2, 3.

200 Tian Guangjin ΙϋΓ<&, "Taohongbalade Xiongnu mu" ^fâlEíufKjíâJÍK-E-, KGXB 1976.1, 131- 144, PI. 1-4; Tian/Guo, Ordos, 203-219.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 69

Zhongning County ψτΜ-, Ningxia, Ml:3) pieces.201 The Maoqinggou examples bear more resemblance to Xiongnu rather than to Saka pieces (PI. 79, left and second from right, Maoqinggou M58:4, M45:3, of which M58:4 shows some Saka influence in the form of the birdheads of the pommel). The double birdhead pommel, while it does not seem to occur within the Yuhuangmiao culture (i.e., at Yuhuangmiao or Xiaobaiyang), does occur on a curved knife from Xiaoheishigou (PI. 81, third from left).

Through these comparisons, patterns of contact emerge. Because these décors are, surprisingly, not widespread in the non-Chinese areas of present-day northern China and seem therefore to be characteristic of specific cultures, comparisons of these décors suggest patterns of intercultural contact. The cannelated hilt, which may have had an antecedent at the Early Western Zhou site of Baifucun Élí?4t, Changping County ^f^JL5 Beijing Municipality,202 occurs typically on Maoqing- gou and Beixinbao, with variants more rarely among the Xiongnu sites, but frequently in Yuhuangmiao/Xiaobaiyang. The double spiral pommel occurs at Maoqinggou, but is common in Yuhuangmiao and reaches Xiaoheishigou. The double birdhead pommel is a Saka-Xiongnu trait, not used in Yuhuangmiao (although a double loop pommel in the form of a snake does occur both on Yu- huangmiao M 164 and Xiaobaiyang M44:l; PI. 76), but does occur on Xiaohei- shigou. The herringbone hilt patterning occurs at Maoqinggou, but more often in

Yuhuangmiao, and reaches Upper Xiajiadian at Xiaoheishigou and Nanshan'gen M101. Zones therefore become clear: the occurrence of the double birdhead in

Upper Xiajiadian may be seen as Xiongnu influence, particularly since the form of the birdheads is comparable to Xiongnu work. This influence may have reached Upper Xiajiadian directly or through Maoqinggou. The double-spiral pommel seems to be a Yuhuangmiao device, which penetrated Upper Xiajiadian. The animal-form pommel may be another example of Yuhuangmiao influence. None of

these motifs or forms appears to have been widely adopted in Upper Xiajiadian, which fact only underscores their significance as indicators of outside contact.

It is the cannelated hilt of Beixinbao Ml:76 as well as the type V sword form which calls the date of that site into question. There is no doubt that bronzes from

Beixinbao are of Late Springs- Autumns/Early Warring States date (e.g., Ml: 88 guan j£H, Ml:85 ding ${).203 Comparative examples for form and/or décor are

201 Ningxia Huizu zizhiqu bowuguan kaogudui ΤΙίΘ^ίΞ)ίρΕΜ-#/ΐ5#~ώ'ΡΛ, "Ningxia Zhong- ningxian qingtong duanjianmu qingli jianbao" TS1TT'íl-#fflMâiJSSySaifaííK, KG 1987.9, 773-777.

202 Beijingshi wenwu guanlichu Jbjl rfr^^J^Hl^h "Beijing diqude you yi zhongyao kaogu shou- huo" ltâJ&KffyJL-1g&%l!Mfà, KG 1976.4, 246-258, 228.

203 Hebeisheng, "Beixinbao," KG 1966.5, 234, f. 5:1, 2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

70 Sophia-Karin Psarras

easily found at Fenshuiling (e.g., ding M25:37)204 and Xiadu (e.g., M31:l)205 for the dart décor on the Beixinbao ding; the Beixinbao Ml:91 jian M206 compares well to Fenshuiling M269:33;207 Beixinbao Ml:87, a hu,20S compares well to a Late Springs- Autumns piece from Nanyang.209 Other pieces are comparable to Zeng Hou Yi material, such as the form of axle caps with their long extension (Ml:48, cf. Zeng Hou Yi N-146).210 The lacquer boxes (120 cm χ 94 cm χ 28 cm, and 51 cm χ 28 cm χ 12 cm; PI. 87) found on top of the Beixinbao Ml coffin carry a geometric décor comparable to lacquer patterns from the late 5/4 c. B.C. from sites such as Changtaiguan ixéi^c, Xinyang Municipality fgPH, He'nan,211 the Zeng Hou Yi M2 coffin,212 as well as other Zeng Hou Yi material (PI. 88). Similar interlocking triangular spiral patterns are found on bronzes of the Late Springs- Autumns tomb M8004:15, in Zhongshan.213 We know little of lacquer patterns in northern China during the Warring States period because of the inherent problem of the disintegration of the material. Therefore, the Beixinbao boxes are, in effect, out of context.

The ge J£, Beixinbao Ml:56,214 belongs to my ge type V (see below), with a high curve and a narrow neck on the blade. This type incudes the patterned bronze ge Wudaohezi Ml:43, as well as pieces from Zhao (Middle Warring States), Zeng Hou Yi (ca. 433 B.C.), and Jin (Early and Middle Warring States) (PI. 96, 97).215 The shape is therefore current in the Middle Warring States, as well as slightly earlier. The ge Beixinbao M2:12 (PI. 95, bottom) belongs to type IV, which I define as having a short point, a nearly straight upper edge, and a curved median

204 Shanxisheng, "Fenshuiling," KG 1964.3, 122, f. 11:3.

205 Hebeisheng, "1964-1965nian Yan Xiadu," KG 1965.11, 550, f. 2:2.

206 Hebeisheng, "Beixinbao," KG 1966.5, PI. 2:5.

207 Shanxisheng wenwu gongzuo weiyuanhui Jin dongnan gongzuozu, Shanxisheng Changzhishi bowuguan lIlHS^^ "Changzhi Fenshui- ling 269, 270hao Dong Zhou mu" -£íétf>/JC^269 » 270'^/j-j^E , KGXB 1974.2, 67, f. 4:5.

208 Hebeisheng, "Beixinbao," KG 1966.5, PI. 2:1.

209 Sun Ji'an, "Hebei Rongchengxian Nanyang," KG 1993.3, 237, f. 2:2.

210 Hubeisheng, Zeng Hou Yi, vol. 2, PI. 105:4.

211 He'nansheng wenwu yanjiusuo W^fâXPjjfflfÍfyí, Xinyang Chumu 'W''nl-%k (Beijing: Wenwu, 1986), PI. 31:1 (Changtaiguan -fcfr^ Ml: 136).

212 Hubeisheng, Zeng Hou Yi, vol. 2, PI. 16:2.

213 Hebeisheng, "Hebei Pingshan Sanji," KGXJK5, 174, f. 22:2.

214 Hebeisheng, "Beixinbao," KG 1966.5, 235, f. 6:1 (Ml:56).

215 Wudaohezi, WW 1989.2, 57, f. 10:3 (Ml:43); Hebeisheng, "Handan Baijiacun," KG 1962.12, 624, f. 17:2 (M20:15, M57:37); Shanxisheng, "Fenshuiling," KG 1964.3, 132, f. 21:3 (M53:3); Hubeisheng, Zeng Hou Yi, v. 2, PI. 89:3 (E-121).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 7 1

ridge. The type includes Wudaohezi Ml:7 (see PI. 95, top), and examples from Jin (Middle to Late Springs- Autumns), but also from Zeng Hou Yi.216 The form was therefore in use for an extended period in China. Similar pieces occur in the Western Han (ca. 120 B.C.) in the Guangzhou Γ^Η, Guangdong tomb of the King of Nanyue ]§iÄ.217 Therefore, Beixinbao Ml:56 in all likelihood dates to a period later than the Beixinbao Ml bronze vessels. Late Springs- Autumns/Early Warring States pottery shards found in the fill of Ml were clearly contained in earth which had been filled with refuse. These shards imply only that there had been a Late Springs- Autumns/Early Warring States, probably Chinese, settlement in the vicinity. This unknown settlement may even have been the source for the Beixinbao Chinese bronzes. The settlement provides only a terminus post quern. The Beixinbao excavation report compares some material (such as the

shortswords) to non-Chinese material known at the time the report was written, and

expresses some surprise at the nomadic burial rites followed on the site. Although a double coffin appears to have been used in the principal burial of Ml, the coffin was covered with the bones of sacrificed horses and the remains of a cart or

chariot, as well as cattle and sheep bones. Despite this, the grave was attributed to the state of Yan because of the presence in the tomb of some Chinese objects and because of geographic proximity to the state of Yan. However, if the burial rites in evidence here do not conform to those of Yan, then the burial cannot be attributed

to that state nor, indeed, to the Chinese in general. The substantial amount of non- Chinese objects in these tombs further underscores a non-Chinese attribution.

I do not believe that Beixinbao can be, at this time, attributed to any specific non-Chinese culture. The site's importance is not one of attribution, however, but of date. Beixinbao sums up in itself all the problems of dating Upper Xiajiadian and Yuhuangmiao, and represents the problems of dating most other non-Chinese sites, particularly in the north (including northwest) of present-day China. Beixinbao incudes two distinct bodies of material: Chinese and non-Chinese. Either may be used for dating. If the Chinese material is used, and a date of ca. 5 or even 4 c. B.C. is accepted, then it follows that non-Chinese material remained static in form and décor over many centuries. That is, identical non-Chinese objects are not con- sidered contemporaneous, in direct violation of a central archaeological theorem. Alternately, non-Chinese material may be used to date the site. Through the short- swords, identical to examples from Yuhuangmiao and Maoqinggou, a date of the Western Han is obtained, with a probable date of Late Western Han. This con- clusion conforms to archaeological theorems and is therefore sound. It leaves us

216 Hebeisheng, "Beixinbao, KG 1966.5, 240, f. 12:2 (M2:12); Wudaohezi, WW 1989.2, 56, f. 8:1; 57, f. 9 (Ml:7); Shanxisheng, "Shangmacun," KG 1963.5, PI. 4:1 (M13); Hubeisheng, Zeng Hou Yi, v. 2, PI. 89:1 (N-37).

217 Guangzhoushi, Nanyue, v. 2, PI. 23:1a (B103).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

72 Sophia-Karin Psarras

with unanswerable questions regarding, not so much the preservation of the Chinese material, but the role and transmission of that material. Both are political questions concerning international relations, and we have no way of resolving them at this time.

Knives

A single example of a moderately curved knife claimed as belonging to Upper Xiajiadian merits discussion. Found on the site of Shilishan M741 ïfï&lli, Jianning County lETJär, Liaoning,218 the knife handle ends in a deer (?) head. This would seem to make the piece comparable to the Early Western Zhou Baifucun fzf/?4t (Changping County H^-Jt, Beijing Municipality)219 shortswords. The pommels of the latter, however, are in the for of animal heads (eagle, horse) executed in the

manner of the Middle Western Zhou, as can be seen for instance at the Baoji 3?J%, Shaanxi, cemetery of the state of Yu $§ (e.g., BRMl-a24, BRMl-a:27).220 The Shilishan piece is executed in a less severe, less rigid style. It is, however, the handle décor which places the Shilishan knife in the Western Han era. The handle

carries a décor of triangles very similar to that found on a differently-shaped knife (with, therefore, a different function) from Yuhuangmiao (PI. 102, lower left). The triangle décor often occurs in Upper Xiajiadian work.

Knives with a few deep serrations at the heel of the blade have been found extensively in present-day northeastern China, from the Jilin culture of (late-era) Xituanshan M[î|li|221 where they occur in tombs dated in Chinese reports to the Late Warring States, to Xiaobaiyang (e.g., Ml:l, Mll:l, M31:6, Μ32.Ί1; PI. 90, first through fourth from left), as well as in Upper Xiajiadian at Xiajiadian M 17:6

218 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:24.

219 Beijingshi, "Beijing diqude you yi zhongyao kaogu shouhuo," KG 1976.4, 246-258, 228.

220 Lu/Hu, Baoji Yugo mudi, vol. 2, Color PI. 19. "BRM" refers to the Rujiazhuang ffiMlï. group of tombs. These similarities, apparently occurring at different times (Early vs. Middle Western Zhou) raise several questions: not only the relationship between Baifucun and Baoji, but the number and distribution of objects related to the Baifucun shortswords. How widespread was the culture possessing these arms? Where was it centered? Note that another early form of non- Chinese knife, curved, the handle terminating in a slim, highly stylized, highly elongated animal head, such as Fu Hao no. 690 (Zhongguo shehui kexueyuan, Fu Hao, PI. 66:1) is also found in association with Lower Xiajiadian material: Jilinsheng wenwu gongzuodui T^frW^^O-fí^À, Li Dianfu ^Ku/f0, "Jilinsheng Kulun, Naiman liangqi Xiajiadian xiaceng wenhua yizhi fenbu yu nei- han" YfWèlM^ ' ^^VmU^r^TBXÍtMWjy^j^Mu WWZLCK 1 (1983), PI. 7:4. The latter article is incorrect in concluding that "the culture of Lower Xiajiadian belongs to (the sphere of) Shang culture" (idem, p. 106), but it is clear that Lower Xiajiadian coexisted with the Late Shang and Early Western Zhou (see also ceramic comparisons, above). Quite extraordinarily, Li Dianfu gives the Early Warring States as a possible lower horizon for Lower Xiajiadian.

221 Dong Xuezeng ^ψϊ^, "Shilun Jilin diqu Xituanshan wenhua" AifetnttíltEHHlIlXík, KGXB 1983.4, 342, f. 31:7-9 (79W-M47:4, 79W-M18:!, 79W-M43:3).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 73

(PL 90, second from right), Xiaoheishigou (PL 89, third and fourth from left), Zhengjiawazi (PL 90, two views on the far right), Zhoujiadi M2:l, M45:53,222 and the site of Shuiquan yJC^R, Jianping County, Liaoning (PL 90, fifth from left).223 The serrations seem to play the role of facilitating the attachment of a handle, since

they occur at the extreme end of the blade, beyond the cutting edge. This hypothesis is borne out both by the Zhengjiawazi and Zhoujiadi examples, which are still encased in a bone (Zhengjiawai) or wood (Zhoujiadi) handle. Zhoujiadi M45:53 also retains the hemp twine binding the handle to the blade. Other examples may have had wood handles, which have since disintegrated.

Ge #24

Type I: with a broad blade, rounded point, straight upper edge, pronounced median ridge; a dewlap with one hole extends into a short "tail"; the piece is attached to the handle through a socket rather than a tang, but the socket is formed

by the overlap of a band of metal extending from the blade, over the handle, and beyond to form a rhombic counterweight. The form occurs on Xiaoheishigou (PL 99, top) and, without a dewlap, on Yuhuangmiao (PL 102, upper left). Lin Yun225 suggests Yinxu comparisons for this form, but the Yinxu examples have no dewlap, or a highly abbreviated form. There are, however, compelling Western Zhou comparisons, such as one from the Zhangjiapo ^%jÜ$L site at Chang 'an (Shaanxi).226 In the Eastern Zhou, there are some Jin ff forms which could perhaps be distantly related, but in these examples, the median ridge is a concavity, and the loop socket is a tubular socket.227

Type II: with a straight upper edge, a median ridge, usually three holes in the dewlap, a tang with a rectangular hole, and a sharply triangular point. Although there is not enough difference in the form overall to create a separate category, ge of this type occur with (a) a short point, and (b) a longer point. Both variants generally coexist. Type Η-a is often what is meant in Chinese texts which refer to a "Shangcunling-type ge," even though more than one form of ge occurs on Shang- cunling. Examples include Shangcunling M1052:53, 54, 163; M1605:3; M1747:

222 Zhoujiadi, KG 1984.5, 421, f. 9:12, 11.

223 Zhai Defang, "Zhongguo beifang," KGXB 1988.3, 288, f. 6:3, 4.

224 Beijingshi, Beijing kaogu sishinian, 70, f. 34:8, from the Yuhuangmiao culture hoard of Xibozi.

225 Lin Yun #fc;, "Shang wenhua qingtongqi yu beifang diqu qingtongqi guanxi zhi yanjiu" 'inj'SCib M^^tMtKM^yc^ZWÍl, in: Su Bingqi (ed.), Kaoguxue wenhua lunji, vol. 1, 139.

226 Cheng/Zhong, Gudai bingqi, 49, no. 3-8; reference: "1967nian Chang'an Zhangjiapo Xi Zhou muzangde fajue" 1967^-fe^c**«cH^IS#n^Ä:ffi, KGXB 1980.4.

227 Cheng/Zhong, Gudai bingqi, 71, no. 4-9.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

74 Sophia-Karin Psarras

10;228 and Liulihe e.g. M1055:ll,229 as well as the Western Zhou site of the Zeng H state cemetery in Shuyang County j^ßH-M-» Hubei.230 These examples suggest a Western Zhou/Early Springs and Autumns date (ca. 11-8 c. B.C.), yet other examples of essentially the same form occur not only in the Early Springs and Autumns period (the tomb of Huang Jun Meng j^ff^, Guangshan County ^fcll] -H-, He'nan, ca. 648 B.C.,231 type Il-b), in the Late Springs and Autumns (Fen- shuiling M269:73, type Il-a, and M269:68, type II-b),232 but in the Warring States period, as at Zeng Hou Yi E-151,233 type II-b, ca. 433 B.C., and the Middle Warring States site of Songzui fòlífô M27:l, Fang County J^g-, Hubei,234 type II- a. Therefore, ths type has a broad chronological distribution and cannot be ascribed with certainty to one specific period. In Upper Xiajiadian, the type occurs at Nanshan'gen M101:15, Xiaoheishigou (PI. 91, upper right), Wujintang (point broken), Wafangzhong M791:192, and Beidamian M6:4 (Dongnangou).235 (See also PL 93.)

Type III: with a rounded point, a tang (with or without attachment hole), usually two holes in the dewlap, and the absence of a median ridge, the blade oval in cross- section. Chinese examples appear much more restricted in chronological distri- bution than type II, occurring from the Middle Warring States through Western Han. Examples include the Middle Warring States sites of Zeng Hou Yi N-138, ca. 433 B.C.;236 Fenshuiling M53:12, 13;237 Chengbei ftfclfc, Liu'an County Tv^câ,

228 Zhongguo kexueyuan, Shangcunling, PI. 35:2, 3, 6; PI. 20:1, 2.

229 Zhongguo shehui kexueyuan, "1981-1983nian Liulihe," KG 1984.5, 413, f. 10:3.

230 Hubeisheng bowuguan ÎfflJtS'hSfé/W» "Hubei Shuyangxian faxian Zengguo muzang" íífUtiSWl &^WM^W, KG 1975.4, PI. 1:1.

231 He'nan Xinyang diqu wenguanhui, Guangshanxian wenguanhui ínJ^fglínJ&K^IB^ » jfclllíl· 'Xfâ^è;, "Chunqiu zaoqi Huang Jun Meng fufu mu fajue jianbao" #Í^WMí©7?íí^fâJ£^ff, ffSÍIÍ, KG 1984.4,312 f. 13:3 (G1:A13).

232 Shanxisheng, "Fenshuiling 269, 270hao," KGXB 1974.2, 72, f. 11:2; PI. 4:1-6.

233 Hubeisheng, Zeng Hou Yiy vol. 2, PI. 88:2.

234 Hubeisheng wenwu kaogu yanjiusuo, Yunyang diqu bowuguan, Fangxian bowuguan ftjytíííC ^Ίίί^^Γ'ΒΙ^^ "1986-1987nian Hubei Fangxian Songzui Zhanguo Liang Han mu fajue baogao" 1986-1987^}Md^a^%OÍHMOiKAÍBÍKi!f, KGXB 1992.2, PI. 9:3.

235 Nanshan'gen M101:15, KGXB 1973.2, PI. 7:1; Wujintang KG 1960.5, 7, f. 1:1; Wafangzhong, WWZLCK9, 55, f. 53; Dongnangou, KG 1977.1, PL 7:1.

236 Hubeisheng, Zeng Hou Yi, vol. 2, PI. 85:2.

101 Shanxisheng, "Fenshuiling," KG 1964.3, PI. 6:3.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 75

Anhui;238 Xiongjialing BW^v, Hanyang County ÖlßöÄ, Wuhan, M18:5;239 Songzui fö&fö, Fang County B&, Hubei, M25:l;240 Yun County fi[SJL, Hubei, M136:16;241 Baijiacun "ft^t, Handan tf|$|S, Hebei, M3:64;242 and the tomb of King Cuo of Zhongshan, ca. 309 B.C., Pingshan ^LÜ, Hebei;243 and for the Western Han, the site of Yun County, Hubei, M301:9A.244 In Upper Xiajiadian, the form occurs at Nandonggou and Sanguandian; in Yuhuangmiao, at Wudaohezi M8:4 (PI. 94).

Type IV: with a long point, a nearly straight upper edge, a curved median ridge, and a tang. The blade tends to widen slightly just before the point. The form occurs in China in the later part of the Springs and Autumns period, as at the Jin state site

of Shangmacun h^rft, Houma fëSj, Shanxi, M13,245 but also in a closely related form at Zeng Hou Yi, e.g., N-16 and N-37.246 Examples also include Beixinbao M2:12 and Wudaohezi Ml:7. The form is similar to a Western Han piece from the tomb of the King of Nanyue, B103, ca. 120 B.C. (see PL 95).247

Type V: with a high curve, a narrow neck flaring to a wider point, and a tang. This type appears in China during the Warring States, as at Fenshuiling M53:3 (Early Warring States),248 Zeng Hou Yi N-14 and cf. E-121,249 Handan (Baijiacun) M20:15, M57:37 (Middle Warring States).250 Examples are found on Beixinbao Ml:56 and Wudaohezi Ml:43 (PI. 96, 97).

In the case of the ge, the Chinese dates of the forms provide only an upper horizon for the non-Chinese sites in which those forms occur. Even if specific types were

238 Liu'anxian wenwu guanlisuo, Chu Jinhua ^'J^&'X'PjJ^BMffi ' í#á£-fr "Anhuisheng Liu'anxian Chengbei Chumu" ^it^/N^cüÄJt^E-, WW 1993.1, PI. 5:3.

239 Wuhanshi kaogudui, Hanyangxian bowuguan ïÇiXîff^Ç" ώ'ΡΛ ' Y3ÍIÍIÍtM#JÍ& "Wuhanshi Han- yangxian Xiongjialing Dong Zhou mu fajue" &UWZW'uMMfà?iiMM%M, WW 1993.6, 73, f. 27:1.

240 Hubeisheng, "Songzui," KGXB 1992.2, PI. 9:4.

241 Zhongguo shehui kexueyuan, "Yunxian," KGXJK6, 158, f. 16:2.

242 Hebeisheng, "Handan Baijiacun," KG 1962.12, 624, f. 15:11.

243 Zhongshan: tombes des rois oubliés, no. 27.

244 Zhongguo shehui kexueyuan, "Yunxian," KGXJK6, 167, f. 21:1.

245 Shanxisheng, "Shangmashi," KG 1963.5, PI. 4:1.

246 Hubeisheng, Zeng Hou Yi, v. 2, PI. 88:3; PI. 89:1.

247 Guangzhoushi, Nanyue, v. 2, PI. 23:1a.

248 Shanxisheng, "Fenshuiling," KG 1964.3, 132, f. 21:3.

249 Hubeisheng, Zeng Hou Yi, v. 2, PI. 89:2, 3.

250 Hebeisheng, "Handan Baijiacun," KG 1962.12, 624, f. 17:2, 1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

76 Sophia-Karin Psarras

in production only for a limited time, regardless of how broad that time was, forms which apparently could not coexist in China are indeed found together on non- Chinese sites. For example, Xiaoheishigou yielded ge of types I and II; Yuhuangmiao, types I and III. The ge therefore present an unresolvable question: were these pieces Chinese imports which were transmitted over relatively long periods of time, as we may reasonably conclude? Were they local productions which, by Chinese standards, might be considered conservative in form? How extensively did they occur throughout the non-Chinese northeast?

Axeheads

Chui WZhuochui '%fäfe: a straight, rectangular blade seems to wrap itself over a tubular socket; a mushroom-shaped counterweight protrudes from the opposite side of the socket. Examples are to be found on Xiaoheishigou (PI. 99, lower right) and Nanshan'gen M101:47.251 Although the term chui (zhuochui) is used for Shang weapons, this particular form does not appear to be attested at that time. Western Zhou sites have yielded related, but not identical, pieces such as an unprovenanced weapon from the Palace Museum, Beijing, with a median ridge bearing geometric décor and a hole in the center of the blade (PI. 100).252 Baicaopo f^ïpCJ^, Lingtai County JTcfí a, Gansu, also Western Zhou, yielded a similar piece, also with pro- nounced median ridge, a very large mushroom-shaped counterweight, and a blade which narrows to a rounded point (PI. 10 1).253

Yue í$: In Upper Xiajiadian, the yue axe has a wide, curved cutting edge; a circular hole in the blade; a long, tubular socket; and a small, tang-like projection on the opposite side of the socket. Both examples, from Xiaoheishigou (PI. 91, second from right, top row) and Beishanzui M7501:22254 are undecorated. Chinese yue, of essentially the same blade form, are all tanged rather than socketed; examples exist from the Shang through the later Warring States, often heavily decorated.255 The tomb of King Cuo of Zhongshan (ca. 309 B.C.) yielded a less ornate example (PI. 98).256

251 Nanshan'gen M101:47, KGXB 1973.2, PI. 8:8.

252 Cheng/Zhong, Gudai bingqi, 30, no. 2-49.

253 Cheng/Zhong, Gudai bingqi, 65, no. 3-73; reference: "Gansu Lingtai Caopo Xi Zhou mu" ttílí i/itWiSlWm·, KGXB 1977.2.

254 Beishanzui, WWZLCK9, 54, f. 48.

255 Cheng/Zhong, Gudai bingqi, 27, no. 2-33; 29, no. 2-43; ff, no. 3-34; Color PI. 2, no. 2-34; 3, no. 2-41; 8, no. 4-61.

256 Cheng/Zhong, Gudai bingqi, Color PI. 8, no. 4-61; Zhongshan: tornbes des rois oubliés, no. 26.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 77

Axehead type I: with a long, narrow body curving outward to a flaring but fairly narrow cutting edge; the form frequently bears a décor of two horizontal ridges around the neck; the upper neck may be pierced with a small hole, undoubtedly for reinforcing attachment. The axe is socketed. Examples include Wujintang (with an L-triangle décor, PI. 103, left), Erdaohezi, Xiaoheishigou (PI. 99, lower left), Liangjiayingzi M8071:175 (listed in the report as a hoe), as well as Xiaobaiyang M30:12, M37:2, and a Yuhuangmiao piece from a cache at Xibozi ifSÎ, Yan- qing County JiJEfcM-, Beijing Municipality (PI. 102, right).257 The form also occurs, with the cross-hatch décor seen on both Xiaobaiyang M30:12 and the Zhengjiawazi example, in the Xituanshan Wfflllj culture of Jilin province (PI. 105).258 The Xituanshan piece, 79W-M88:1 from the habitation site of Houshishan, is associated with a 14C reading of ca. 325 B.C. ± 70, taken from a human bone sample. Forms similar to this are found in the Wang Mang era tomb Jin-Ml:12 in the Shaogou ^y£j cemetery, Luoyang, He'nan, in iron; in this case, the form is considered a hoe.259 The identification of tools can pose a problem, and pieces with identical form and dimensions are sometimes termed "axes" and sometimes "hoes." The specific dimensions of a piece do not seem to determine the category into which it falls; most type-I and related axeheads measure between 8.0 and 12.6 cm in length.

Type II: is characterized by a moderately broad body with a pronounced curve from the waist to the end of the cutting edge, as well as over the cutting edge itself. The latter is considerably wider than the neck of the axe. The axe is socketed. Type II examples include the axe moulds from Erdaohezi (PI. 106, top), axeheads from Shi'ertaiyingzi Ml, Dahuofang j^i'XBr, and Menlian (PI. 104, right column, second and third from top).260 Chinese examples include a Middle Warring States piece from the Chu tombs at Chengbei MJL Liu'an County 7^3$?^ Anhui.261

Type III: with a long, narrow neck, flaring suddenly in nearly straight lines to the outer point of the cutting edge, which is moderately curved. The axe is socketed. The form is attested by a mould from Shuangfang (PI. 106, bottom).

Type IV: is essentially rectangular, with a slightly curved cutting edge. The axe is socketed. The form is attested by moulds from Wangtu _:£_+. (PI. 107, upper right),

257 Sanguandian, KG 1985.2, 126, f. 2:7, 8; Sunjiagou, WWZLCK9, 35, f. 30:3.

258 Jilinsheng wenwu kaogu yanjiusuo, Jilinshi bowuguan ffW^XW^ii&f^Jï ' nr#rfi|TS-/#jtß' "Jilinshi Houshishan yizhi di er ci fajue" τπ#Μ^5!ΐΐϋί±^-ΐΛ^#,, KGXB 1993.3, PI. 6:1.

259 "Jin-M" refers to a Jinguyuan <MJ@ tomb excavated and reported with the Shaogou tombs. Luoyangqu, Shaogou, PI. 54:5.

260 Menlian, KG 1981.2, 189, f. 1:3; f. 2:2.

261 Liu'anxian, "Anhuisheng Liu'anxian," WW1993.1, 34, f. 14.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

78 Sophia-Karin Psarras

and an axehead from Wafangzhong M791:189.262 Chinese examples may be found on Late Warring States sites such as Xiadu "f|f[5 Tl:3:6 (in iron)263 and Fushun iJtM (in bronze), dating to the Late Warring States/Early Western Han.264

It is not clear to me how specific, culturally or chronologically, these axe forms are. It is evident that, for those forms with Chinese counterparts, such as type IV, distribution covers the Late Warring States and Western Han.

PART III

Conclusions: Cultural Exchange

Although I have excluded the areas of Jilin and Heilongjiang from the present work (pending separate investigation), these regions are integral to an understanding of Upper Xiajiadian. It is clear that there is significant interplay between Jilin and Upper Xiajiadian, whether the instances of closely similar or identical objects in each culture indicate a common heritage or cultural exchange. Leaf-bladed shorts words have been found in the heart of Jilin province, including a type I(b)- elongated example found in 1984 in Xiushui fjrVjC, Shuangliao County M jT J^·265 The blade may have been cast together with the hilt unit (?), which carries a highly irregular pommel but, apparently, no separate crown. The hilt décor is of triangles on a stippled ground. Type I(b)-elongated shorts words with bronze hilt units and symmetrical crowns with a central bulge are attested by Jin Fengyi at the Jilin site

of Xihuangshandun M^îIjlË.266 Also found on that site, according to Jin, are Xichagou type III swords with antennae pommels.267 At Xichagou, the only other site on which I have yet found this form, the swords have an iron blade and a bronze hilt unit. The form does not occur on Xianbei sites such as Yushu, despite the coexistence there of much Xichagou material. Xichagou is provided with a positive date of no earlier than 118 B.C. through wushu coins. The Xichagou type III sword therefore dates Xihuangshandun to the same era, and thus provides a late 2 c. B.C. terminus post quern date for Upper Xiajiadian type I(b)-elongated short- swords. Jin Fengyi cites the Xichagou example and adds Jilin radiocarbon dates from another associated site, Yangdun t^Î, of 2165 B.P. ± 75 and 2105 B.P. ± 75, or 215 B.C. and 155 B.C., each ± 75. Without knowing the accuracy of these

262 Wafangzhong, WWZLCK9, 55, f. 54.

263 Hebeisheng, "Yan Xiadu di 22hao," KG 1965.11, PI. 6:5.

264 Wang Zengxin, "Liaoning Fushunshi," KG 1964.12, 288, f. 2:1.

265 Jilinsheng bowuguan rftt^ltÇ^/íff, Jilinsheng bowuguan ~n#fëW$JÎS (Beijing: Wenwu, 1992), 161; PI. 1.

266 Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 411, f. 9:44.

267 See Psarras, "Exploring the North"; Sun Shoudao, "'Xiongnu Xichagou wenhua'," WW 1960.8- 9, 26, f. 2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 79

dates as such, they are earlier than the Xichagou date but still within the general era pertaining to Xichagou (i.e., Western Han). Jin nonetheless concludes that this provides a Late Warring States/Early Western Han date for the material. The appearance of Upper Xiajiadian shorts words in Jilin with late Xituanshan

type ceramics and Xichagou type swords raises the question of the interaction of three cultures: not only Upper Xiajiadian and Xituanshan, but Xianbei as well. A zone represented mainly by ceramic forms seems to extend from Upper Xiajiadian to the Xianbei, while another such zone extends from Xituanshan to the Xianbei,

but connections between Xituanshan and Upper Xiajiadian appear limited to type I axes and Upper Xiajiadian type shortswords. Upper Xiajiadian's relationship with the Xiongnu is similarly limited. The presence of the animal plaque and the square or semi-ovoid plaques with a border décor of fine radiating lines, all seen at Nanshan'gen M101, may be accepted as representing Xiongnu influence, yet the distribution of such pieces in Upper Xiajiadian appears limited to Nanshan'gen M101 and Zhoujiadi (Ml: 11, M45:59).268 The winged ornament (a semi-spherical disc from which two wing-like shapes extend) occurs in Upper Xiajiadian (Nan- shan'gen M3:8, Zhoujiadi M2:3, Xiajiadian Ml 1:1) and in Inner Mongolia but, rather than at a Xiongnu site, at Yinniugou Mil: I.269 (The form also occurs in Siberia, which will be treated in a later article.) Semi-spherical appliqué ornaments with a border design of fine radiating lines have a broader distribution, including Dabannanshan j&iSMlh, Linxi County ^"pfJl·, Liaoning; Xiajiadian M 17:2; Bei- damian (Dongnangou); and probably Longtoushan (PI. 121), all in Upper Xiajia- dian, and Xigouban M2:l for the Xiongnu, as well as Yuhuangmiao, including Wudaohezi M7:17.270 Except for these few forms, there is no commonality of material between the Xiongnu and Upper Xiajiadian. Yet, Xiongnu presence may be felt in burial rites. When animal bones are buried at Zhoujiadi, the use of horse and cattle skulls and horse hooves echoes not Yuhuangmiao tradition, but Xiongnu practice. (Zhoujiadi also included dog skulls.) The placement of cattle leg bones in the Zhengjiawazi graves would seem to indicate a funerary feast, not animal sacrifice in its own right. The sacrifice of animals on the grave occurring so rarely in Upper Xiajiadian, it may well have been an intruding custom. Even when such nomad-related ritual as the sacrifice of a horse does occur, the horse nonetheless

268 Nanshan'gen M101, KGXB 1973.2, PI. 11:3, 4 (M101:70, 69); Zhoujiadi, KG 1984.5, 421, f. 9:8, 1O(M1:11,M45:59).

269 Nanshan'gen, KGXB 1975.1, 137, f. 19:11 (M3:8); Zhoujiadi, KG 1984.5, 421, f. 9:9 (Mll:l); Neimenggu, "Yinniugou," NMGWWKG 3 (1984), 30, f. 8:1.

270 Jin Fengyi, "Xiajiadian shangceng," KGXB 1987.2, 184, f. 3:9; Xiajiadian, KGXB 1974.1, 140, f. 30:18; Dongnangou, KG 1977.1, 54, f. 7:1, 2; Yikezhaomeng, "Xigouban," WW 1980.7, 4, f. 6:11; Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 48, f. 15:18 (M41:ll); Wudaohezi, WW 1989.2, 56, f. 8:26.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

80 Sophia-Karin Psarras

never assumes the central role that it holds in Xiongnu culture, not just as part of the economy, but as myth and ritual. One instance when both Yuhuangmiao and Upper Xiajiadian may have

admitted Xiongnu decorative influence may be the circular plaques with openwork spiral motifs, Nanshan'gen M101:67, 68 (PI. 108), compared to Xiaobaiyang M13: 3, M31:5 (PI. 109, left). These pieces are also comparable to examples from Yu- huangmiao (PI. 110, far right). A Xiongnu variant is to be found on Kesheng- zhuang M140:5 %-fëJÎ. (Chang'an, Shaanxi), with more open space in the com- position, and the pattern of fine radiating lines replaced by ovoid shapes.271 Ke- shengzhuang M 140 is often dated to the Late Warring States, but since this periodization is based on the assumption that the grave would have held wushu coins had it been later, the date is not reliable; stylistic comparisons with Xichagou point to a Middle to Late Western Han date. The Han site of Shaogou yielded a related piece, in gold, which combines the very open Keshengzhuang M140:5 form with Xiongnu birdheads (i.e., with ears) as the central spirals, and the fine line pattern of Yuhuangmiao in the form of beading (Shaogou M 1040: 5, ca. 132 A.D., PI. 111). Because Nanshan'gen M101:67, 68 appear to be anomalies in Upper Xia- jiadian while being in context in Yuhuangmiao and because both appear related to Xiongnu forms and motifs, both may reflect Xiongnu influence. The date of Shaogou M 1040 is noteworthy, adding as it does to the evidence for a late date for Upper Xiajiadian.

Yuhuangmiao and Pingyang or Xianbei influence may be visible on a bronze feline from Xiaoheishigou (PI. 120, bottom). The feline's striated tail, positioned below the body, compares to Yuhuangmiao practice (PI. 113); the concentric circles marking the haunch and shoulder recall Xianbei use of wings on felines and horses, marks perhaps echoed at Pingyang on tiger prédation plaques derived from Xiongnu Yanglang-type prédation scenes (Pingyang M150:5, M188:26, see PI. 25). 272 The treatment of the Xiaoheishigou feline's muzzle is closely related to Pingyang M 188: 26. Upper Xiajiadian had significant exchange with Yuhuangmiao (ceramics,

double-spiral pommel, shortsword hilt décor, small ornaments in isolated cases), yet Xiongnu forms such as tubular hanging ornaments, "flower discs," zigzag ornaments, occur at Yuhuangmiao but not in Upper Xiajiadian. The same is true with regard to Pingyang, with which Upper Xiajiadian (Nanshan'gen M101:79) shares the type IV bell (spherical with a stem). The type III bell is closest to Nan- shan'gen M4:5, 9, 41, 47, but has a wide distribution across Pingyang, the Xian-

271 Zhongguo kexueyuan kaogu yanjiusuo ^P3í4^'&1q^Ç'Í*W?lJW» Fengxi fajue baogao i^lffiSl ffiílíff (Beijing: Wenwu, 1963), PI. 103:1 (M140:5).

272 Heilongjiang, Pingyang, PI. 40:2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 8 1

bei, Yuhuangmiao, as well as at the Transbaikalian Xiongnu site of Dyrestuj,273 Daodunzi, and the Kayue culture (Qinghai). At the same time, Upper Xiajiadian and Yuhuangmiao share the funerary rite of covering the face of the deceased with shell or bronze semi-spheres, although this rite is used in 30% of Yuhuangmiao graves and only on Zhoujiadi and Zhengjiawazi in Upper Xiajiadian. There is no association of the practice with any other visible burial structure or rite, or with any type or style of grave goods. Lining the grave with birchbark occurs not only in Upper Xiajiadian, but at Pingyang. Grave firing was practiced as well by the Xianbei, as at Yushu. Upper Xiajiadian coffins may have resembled Xianbei cof- fins, as does Zhoujiadi M43 (PI. 2). Yuhuangmiao was more open to Xiongnu in- fluence than was Upper Xiajiadian, but never absorbed Xiongnu burial rites.

Chinese authors have written that the absence of material influence from China

in Upper Xiajiadian proves the latter to predate the Warring States,274 yet the reverse theory would never be advanced. It is evident from the examples of Upper Xiajiadian and Yuhuangmiao that foreign elements absorbed through intercultural contact are specifically, individually selected. This is not to say that such choices were the result of conscious deliberation, but certainly choices were made. Given the data presented above, I date Upper Xiajiadian as a whole to no earlier than the 4 c. B.C. (through the geographic association of Shi'ertaiyingzi with ca. 309 B.C. Chinese graves, based on Zhongshan ceramic comparisons). Much of the published material must date to the Late Western Han through comparisons with Yuhuang- miao and such Xiongnu-influenced sites as Nanyue and Daodunzi (ca. 120 B.C.). Upper Xiajiadian, Yuhuangmiao, and the Xiongnu are therefore all contemporary. They all had contact, traceable through the data of material culture. Nonetheless, the image of animal prédation was excluded from the lower northeast (that is, excluding Pingyang) despite its adoption in China and the préexistence in China of a related form.

The role of China in Upper Xiajiadian seems to have been limited. Shi'ertai- yingzi was intermingled with Chinese graves dating from ca. 309 B.C. through the Western Han; Dadun and Liangjiashan, with Western Han and/or Wei tombs, as well as some Late Warring States ceramics; Zhengjiawazi, with Western Han tombs; Nanshan'gen M101 and Xiaoheishigou with Late Warring States pottery shards; Nanshan'gen M7, Late Warring States and Han greyware. The use of shards as fill supposes that the shards predate the site. Shards over the site may have been washed there by erosion or may have fallen as debris: they may therefore predate or postdate the site. The intrusion of Western Han tombs indicates the earlier date of the Upper Xiajiadian remains, but given present data,

273 Sergei I. Rudenko, Die Kultur der Hsiung-nu und die Hügelgräber von Noin Ula (Bonn: Rudolf Habelt, 1969), 142, f. 41 :j.

274 Xiajiadian, KGXB 1974.1, 144.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

82 Sophia-Karin Psarras

those intrusive Western Han graves may have been only slightly later than (hence, essentially contemporaneous with) the sites they destroyed. It is clear from the dynastic histories that Han foreign policy was not concerned, officially, with coexistence or accomodation. The goal of the government was the eradication of foreign culture and, by this means, of foreign power.275 The destruction by the Han of non-Chinese graves may reflect this attitude. At the same time, intercultural contact did occur regardless of government policy, both from China toward the non-Chinese, and from the non-Chinese toward China. It is notable that, in the case

of the Xiongnu, one of the greatest examples of the influence of Xiongnu material culture on China (the numerous Xiongnu or Xiongnu-type plaques in the tomb of the King of Nanyue, ca. 120 B.C.) occurred during the time of Wudi's wars against the Xiongnu, initiated in 133 B.C. Government policy was then at its most anti-Xiongnu, yet on a less official level, cultural exchange was unimpeded. The occurrence of Chinese bronze vessels in Upper Xiajiadian as well as at

Yuhuangmiao raises a number of questions: why did these vessels find a place in the northeast, but almost never in Xiongnu territory?276 Where were these Chinese vessels between the time of their production and their burial? How did the non- Chinese obtain them? Which non-Chinese originally obtained them? Clearly the vessels had value outside of China, but the bronzes do not seem to have exercised

275 See Psarras, Han and Xiongnu. Lu Jing H® (Liu Jing MWO formulated the heqin $'M policy for Gaozu iff?]!, first emperor of the Western Han, in ca. 200 B.C. Although specifically directed toward the Xiongnu, this policy characterizes the Han attitude toward the non-Chinese in general, advocating political conquest through cultural subversion when force of arms proved inefficient. The Xiongnu or other non-Chinese would not only be led into a theoretically subordinate relationship with the Emperor (i.e., as imperial sons-in-law), but would face an influx of Chinese goods. Dependency on these goods would destroy indigenous culture, it was believed, and Chinese political domination would follow economic/cultural dominance. See Sima Qian, Shiji (Beijing: Zhonghua shuju, 1982), j. 99, 110; p. 2719 (for the text of Lu Jing's proposal); Ban Gu gíg], Hanshu ΪΧ+λ/ 94. In practice, Han policy was determined more pragmatically and, in any case, local conditions clearly did not necessarily reflect central government rhetoric. Nonetheless, the intent to destroy non-Chinese cultures and political forces is evident throughout the Han dynastic histories. Those peoples most affected by this intent were those with the least political power of their own: for example, in the north, the Oiang ^ and Wusun ST/k

276 The only Xiongnu site with Chinese ritual bronzes of which I am aware is Nalin íftWs Zhungar Banner Ki'$$5j'JM> Inner Mongolia, apparently still unpublished, whose finds I was able to view in 1992. The site may be positively dated through comparison of its bronze decorative plaques with Xiongnu pieces: the bronze Yanglang tiger prédation plaque ("Ningxia Guyuanxian," WW 1978.12, 87, f. 1) and the Aluchaideng gold plaques of tigers (Tian/Guo, "Aluchaideng," KG 1980.4, 334, f. 2; PI. 11:1; Tian/Guo, Ordos, PI. 5). Aluchaideng is positively dated to the Western Han, in my analysis, by pottery fragments with Han décor, published only in the original excavation report (KG 1980.4, 337, f. 4) and not in the Tian/Guo, Ordos reprint. The Nalin Chi- nese bronze vessel, however, dates to ca. early 5 c. B.C. through comparison with Houma moulds ("Shanxi Houma Dong Zhou yizhi faxiande taofan" iWBÍ^j^M^Ml&M^MÍL· WW 1960.8- 9, 7, f. 5).

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 83

more than a limited, stylistic influence in the northeast, if the Nanshan'gen M101:35 shortsword is an Upper Xiajiadian product and not a Chinese import. Whatever the role of the Chinese ritual vessel may have been in these non-Chinese contexts, the vessels do not serve to date the cultures; because of their date, they do not elucidate contemporary Chinese-foreign cultural or political relations. We are left, therefore, with unanswerable questions about the preservation and transmission of the Chinese vessels, and the necessity of accepting the use of non- Chinese material as the most accurate determinant of chronology.

Addendum

Tadashi Sekino Hfliril, et al, Rakurõgun jidai iseki W&W&ÍXtSM [Archaeological Researches on the Ancient Lolang District] (Government-General of Chosen, 1925), vol. 3, 180, PL 846, is identified as an "oddly-shaped bronze object," found in association with straight-bladed bronze shorts words. Because this piece may reasonably be identified as an Upper Xiajiadian-type shortsword pommel crown, it provides an additional positive date of the Later Han for Upper Xiajiadian material.

Since the preparation of this article in 1994, the final report on Xiaoheishigou has been published as: Chifengshi bowuguan, Xiang Chunsong ^i^flíM^ÍÈ' ' ^M^à' Ning- chengxian wenwu guanlisuo, Li Yi ^riMMc'SC^^WMBf ' ^SL, "Ningcheng Xiaoheishigou

shiguo mu diaocha qingli baogao" yiA^W^WMM^WS.U-^, Wenwu 1995.5, 4- 22, Color PL 1-2, journal cover, inside cover. In addition, a recent report from Liaoning which presents the excavation of habitation sites and graves in Anzhangzi fSrÍJ^J1, Lingyuan County §£?MJÏ-, attributes the lowest level on the site to Upper Xiajiadian, partially covered by Warring States and Western Han habitations and tombs. While the stratigraphy appears convincing, the artefacts from the Upper Xiajiadian layer consist mostly of ceramic forms (such as dou) which may be Middle-Late Warring States and are not diagnostic of Upper Xiajiadian itself. The attribution of the site is therefore open to question; the site could conceivably provide indications of local cultures in the immediately pre-Upper Xiajiadian era. See Liaoningsheng wenwu kaogu yanjiusuo UJr^'X^Ji^'tSMtWx, "Liaoning Lingyuan Anzhangzi guchengzhi fajue baogao" K^WiU^t^tiW^^WM^, Kaogu xuebao 1996.2, 199-236, PL 9-14.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

84 Sophia-Karin Psarras

Illustrations and Maps

l.ZhengjiawaziM6512,#GX5 1975.1, 142, f. 3.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 85

II ι 1 ' ι Κφ^ϊ^

0 50«»

2. Zhoujiadi M45, KG 1984.5, 418, f. 2.

3. Xiaoheishigou. Photograph® S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

86 Sophia-Karin Psarras

4. Xiaoheishigou. Photograph © S.-K. Psarras.

5. Upper Xiajiadian (no site specified). Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 87

6. Xiaoheishigou. Photograph © S.-K. Psarras. 7. Xiaoheishigou. Photograph ® S.-K. Psarras.

8. Xiaoheishigou. Photograph ® S.-K. Psarras. 9. Xiaoheishigou. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

88 Sophia-Karin Psarras

10. Dadianzi. Photograph © S.-K. Psarras. 11. Xiaoheishigou M8061 :210, WWZLCK9, 29, f. 1 1:3.

12. Shaogou M1038: 1 1, Luoyang, Shaogou, PL 49:3 (bronze). 13. Shaogou Mill :46, M632: 142, Luoyang, Shaogou, PL 49:4, 5 (bronze).

14. Xiaoheishigou. Photograph ® S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 89

15. Xiaoheishigou. Photograph ® S.-K. Psarras. 16. Baoji BZM8:7, Lu Liancheng, Hu Zhisheng, Baoji Yuguo mudi, v. 2, PL 93:3.

17. Liulihe. Photograph © S.-K. Psarras. 18. Zhoujiadi scattered find no. 3, KG 1984.5, 422, f. 10:2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

90 Sophia-Karin Psarras

19. Bronze shorts word, Nanshan'gen eastern sector habitation site cist tomb, WWZLCK9, 36, f. 31.

20. Longtoushan. Photograph © S.-K. Psarras. 21.Nanshan'genM101:66,#GX£ 1973.2, PL 11:1.

22. Liangcheng find, Gai Shanlin, "Neimenggu Wumeng nanbu faxiande qingtongqi he qingyin," KG 1986.2, 185, f. 2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Χι aji adi an 9 1

23. Liangcheng find, Gai Shanlin, "Neimenggu Wumeng nanbu," KG 1986.2, 186, f. 4. 24. Zeng Hou Yi C- 108, Hubeisheng, Zeng Hou Yi, v. 2, PL 58:1.

25. Pingyang M 150:5, Heilongjiangsheng, Pingyang, PL 40: 1 .

26. Guoxianyaozi M12:2-l, Neimenggu, "Guoxianyaozi," KGXB 1989.1, PL 16:4. 27. Maoqinggou. Photograph ® S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

92 Sophia-Karin Psarras

28. Nanshan'genM101:70, 69,^X5 1973.2, PL 11:3,4. 29.Nanshan'genM101:79,^GXß 1973.2, PL 11:5.

30.XiaoheishigouM8061:209-l, WWZLCK 9,28,f. 10:2; Liangjiayingzi M807 1:220, WWZLCK9, 31, f. 14:8;WudaoheziMl:15,WW 1989.2, 56, f. 8:19.

31. Shi'ertaiyingzi (state of Yan), Wenwu, Wenwu kaogu gongzuo shinian 1979-1989, PL 6, lower left.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 93

32. Xiadu M29: 14, Hebeisheng, "1964-1965nian Yan Xiadu," KG 1965. 1 1, 552, f. 6:5; Yunxian M417:3 (top), M131:2 (bottom), Zhongguo shehui kexueyuan, "Yunxian," KGXJK 6, 149, f. 10:2, 13; all others: state of Yan, Huairou, Beijingshi, Beijing kaogu sishinian , 56, f. 30:6, 5, 4, 3.

33. YuhuangmiaoMl:5, Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 31, f. 25:2. 34. Beixinbao Ml :48, Hebeisheng, "Beixinbao," KG 1966.5, 237, f. 8: 1 ;

Nandonggou,£G 1977.6, 375, f. 3:5.

35. Beixinbao Ml:20, Hebeisheng, "Beixinbao," KG 1966.5, 237, f. 8:6; Nandonggou,#G 1977.6, 375, f. 3:6.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

94 Sophia-Karin Psarras

36. Nandonggou, KG 1977.6, 373, f. 1 : 1 . 37. Wangtu. Photograph © S.-K. Psarras.

38. Nandonggou, KG 1977.6, 375, f. 3:1; Wudaohezi, WW 1989.2, 57, f. 10:5. 39. Shangmashi M3:2, KG 1982.6, 592, f. 2:11; ZhoujiadiMl:4, KG 1984.5, 420, f. 7:2; ZhengjiawaziM6512:63,^GZi? 1975.1, 151, f. 14:l;XiaobaiyangM12:5,Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6: 16; Liangjiashan M5, KG 1982.6, 592, f. 2:3; Wangong

M1B:72, Neimenggu, "Wangong," KG 1965.6, 276, f. 4:7.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 95

40. Shangmashi JBM1 1 : 1 , KG 1982.6, 592, f. 2:2; Xiaobaiyang M30: 14, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6:2; Yuhuangmiao M190: 1, Beijingshi, "Beijing Yanqing

Jundushan,"WW 1989.8, 26, f. 16:1.

41. Longtoushan. Photograph © S.-K. Psarras.

42. Shangmashi JBM11, KG 1982.6, 593, f. 3:2; Xiaobaiyang M21 : 1, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 44, f. 6: 18; Daodunzi M26: 1, Ningxia, "Daodunzi," KGXB 1988.3, 342, f. 8:5. 43. Shangmashi M4: 1, KG 1982.6, 592, f. 2: 10; Yinniugou M9: 1, Neimenggu, "Yinniugou,"

NMGWWKG 3 (1984), 31, f. 9:4; Wangong M1B:62, Neimenggu, "Wangong," KG 1965.6, 276, f. 4:1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

96 Sophia-Karin Psarras

44. Shangmashi JBM1:1, JBM9:1, JBM14, KG 1982.6, 592, f. 2:3, 8; 593, f. 3:1; Zhalainuoer M28, Neimenggu, "Zhalainuoer," KG 1961.12, 676, f. 5:2; Budonggou Ml:l, Yimeng, "Budonggou,"

NMGWWKG 1 (1981), 31, f. 4:2; Liangjiashan M5, KG 1964.12, 281, f. 5:1. 45. Shangmashi JBM13: 1, KG 1982.6, 592, f. 2:6; Pingyang M104:9,

Heilongjiangsheng, Pingyang, 61, f. 40:3.

46. Shuangfang, KG 1983.4, 294, f. 3:1; Erdaohezi, KG 1977.5, 305, f. 8. 47. Longtoushan. Photograph ® S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 97

^L_l

48. Xiajiadian F4:l, KGXB 1974.1, 131, f. 23:11; Erdaohezi,tfG 1977.5, 304, f. 6:3, 2. 49. Xiajiadian fragments H9:4,Ts-4:37,^GX5 1974.1, 131,f. 23:12, 13;Tianjuquan, JinFengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:38; Fushun, Wang Zengxin, "Liaoning

Fushunshi," KG 1964.12, 292, f. 6:20.

ψ six

50. Left: Zhongshan habitation site 2, Hebeisheng, "Hebei Pingshan Sanji," KGXJK5 (1987), 165, f. 9:2; Xiadu site 22, Tl :3: 15, Hebeisheng, "Yan Xiadu di 22hao," KG 1965. 1 1, 564, f. 2:4;

Yuhuangmiao M61 :4, Beijingshi, "Beijing Yanqing Jundushan," WW 1989.8, 26, f. 16:6; Right: Baijiacun (Zhao) M40:4, Hebeisheng, "Handan Baijiacun," KG 1962.12, 620, f. 8:5.

51. Fushun, Wang Zengxin, "Liaoning Fushunshi," KG 1964. 12, 292, f. 6: 19; Xiadu site 22 T4:3:22, Hebeisheng, "Yan Xiadu di 22hao," KG 1965. 1 1, 564, f. 2:2; Zhongshan habitation site 2,

Hebeisheng, "Hebei Pingshan Sanji," KGXJK5 (1987), 165, f. 9:4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

98 Sophia-Karin Psarras

52. Longtoushan. Photograph © S.-K. Psarras. 53. Nanshan'gen M101 (bronze), Jin Fengyi, "Lim Zhongguo dongbei," KGXB 1982.4, 390, f. 1 :2; Xiajiadian (ceramic) Tl-2:7, KGXB 1974.1, 131, f. 23:3; Bottom: Fenshuiling (bronze) M269:62,

Shanxisheng, "Fenshuiling 269, 270hao," KGXB 1974.2, 67, f. 4:7.

54. No attribution, Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:33; Zhoujiadi M45:01, KG 1984.5, 420, f. 7:3; Nanshan'gen M 101 (bronze), Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:1; Right: Xiajiadian H5: 15, KGXB 1974.1, 131, f. 23:5; Xiajiadian T7- 1:2, GXB 1974.1, 131, f. 23:4; Yuhuangmiao M52:l, Beijingshi, "Beijing Yanqing Jundushan," WW

1989.8, 26, f. 16:2.

55. Liulihe. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 99

56. Liulihe. Photograph © S.-K. Psarras.

57. Zhukaigou. Photograph ® S.-K. Psarras. 58. Lower Xiajiadian: Dadianzi ~Xh]-f- . Photograph φ S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 00 Sophia-Karin Ps arras

59. Lower Xiajiadian: Dadianzi. Photograph © S.-K. Psarras. 60. Lower Xiajiadian: Dadianzi. Photograph © S.-K. Psarras.

61. Liulihe. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 101

ο

Ι θ

62. Xiaoheishigou M8061 :208, WWZLCK 9, 29, f. 1 1 : 1 ; Nanshan'gen M101 , Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1:5.

63. Sanguandian, KG 1985.2, 128, f. 6:1.

64. Erdaohezi, ÄTG 1977.5, 303, f. 2: 1, f. 1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 02 Sophia-Karin Ps arras

65. Liangjiashan Ml, M3, KG 1964.12, 281, f. 6; Dadun, KG 1964.12, 278, f. 2.

66. Nanshan'gen habitation site, KGXB 1975.1, 138, f. 1:1; Zhengjiawazi M6512:2, M65 12:33, KGXB 1975.1, 144, f. 4:2, 3.

67. Shuanefans. KG 1983.4, 294. f. 3:5.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 03

68. Nanshan'gen M101 :36, 35, 33, 34, 34, KGXB 1973.2, 33, f. 5. 69. Fenshuiling M36:2, Shanxisheng, "Fenshuiling," KG 1964.3, 126, f. 15:1, 2.

70. Xiaoheishigou. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 04 Sophia-Karin Psarras

71. Yuhuangmiao. Photograph © S.-K. Psarras.

72. Yuhuangmiao. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiaji adian 1 05

73. Mancheng M 1:5046, 5024, 5109, Zhongguo shehui kexueyuan, Mancheng, vol. 2, PL 44:2-4.

74. Yuhuangmiao. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 06 Sophia-Karin Ps arras

75.XiaobaiyangM12:4,M16:l,M37:l,M30:13, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 46, f. 13:3, 6, 7, 9.

76. Yuhuangmiao M164, Beijingshi, Beijing kaogu sishinian, 83, f. 36:3; Xiaobaiyang M44: 1 , Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 46, f. 13:1.

77. Luotuoliang ïjftii'i:®, Longhua Co. RHtll·, Hebei; Xiadianzi KfuJ-f , Longhua Co., Hebei;

Luotuoliang, Longhua, Hebei; Fengning 4^fi"!·, Hebei; all from Zhai Defang, "Zhongguo beifang," KGXB 1988.3, 284, f. 4:3-6; Yuhuangmiao M70, Beijingshi, Beijing kaogu sishinian, 83, f. 36:1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 07

78. Beixinbao Ml:76, Hebeisheng, "Beixinbao," KG 1966.5, 235, f. 6:7. 79. Maoqinggou M58:4, Gongsuhao Ml :5, Maoqinggou M45:3, Tian/Guo, Ordos, PL 27: 1-3.

80. Hulusitai Ml :3, Maoqinggou M60:6, Tian/Guo, Ordos, PL 28: 3, 4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 08 Sophia-Karin Ps arras

81. Xiaoheishigou. Photograph © S.-K. Psarras.

82. Dongnangou, KG 1977.1, 53, f. 6:2, 1. 83. Chang 'an weapons depot H7:2:2, Zhongguo shehui kexueyuan,

"Han Chang'an cheng wuku yizhi fajue de chubu shouhuo," KG 1978.4, PI. 1 1 :4 (iron). 84. Yinniugou M 1:2 (iron), Neimenggu, "Yinniugou," NMGWWKG 3 (1984), 29, f. 7: 1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 09

85. Longtoushan. Photograph ® S.-K. Psarras.

86. Xiaobaiyang M39:5, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 46, f. 13:4; BeixinbaoMl:83, Hebeisheng, "Beixinbao," KG 1966.5, 235, f. 6:2.

87. Beixinbao lacquerware, Hebeisheng, "Beixinbao," KG 1966.5, 233, f. 4.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

110 Sophia-Karin Ps arras

88. Changtaiguan Ml: 136, He'nansheng, Xinyang Chumu, PL 31:1.

89. Xiaoheishigou. Photograph © S.-K. Psarras.

90. Xiaobaiyang M32:ll, Mll:l, M31:6, Ml:l, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 46, f. 13:16-19; Shuiquan, Jin Fengyi, "Lun Zhongguo dongbei," KGXB 1982.4, 390, f. 1 :25; Xiajiadian

M17:6, KGXB 1974.1, 140, f. 30: 1 ; Zhengjiawazi, KGXB 1975.1, 147, f. 1:1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 111

91. Xiaoheishigou. Photograph © S.-K. Psarras.

92. Zhangjiapo, Chang'an Co., Shaanxi, Cheng/Zhong, Gudai Bingqi, 49, no. 3-8. 93. Zeng Hou Yi E-15 1, N-16, Hubeisheng, Zeng Hou K, vol. 2, PL 88:2, 3.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

112 Sophia-Karin Psarras

94. Nandonggou, KG1911.6, 375, f. 3:2; Yunxian M132: 16, Zhongguo shehui kexueyuan, "Yunxian," KGXJK 6 (1989), 158, f. 16:2; WudaoheziM8:4, WW 1989.2, 56, f. 8:2.

95. Wudaohezi Ml :7, WW 1989.2, 56, f. 8: 1 ; Shangmacun M13, Shanxisheng, "Shangmacun," KG 1963.5, PL 4: 1 ; Beixinbao M2: 12, Hebeisheng, "Beixinbao," KG 1966.5, 240, f. 12:2.

96. Zeng Hou Yi N-14, Hubeisheng, Zeng Hou K, v. 2, PL 89:2.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xi ajiadian 113

97. Fenshuiling M53:3, Shanxisheng, "Fenshuiling," KG 1964.3, 132, f. 21:3.

98. Zhongshan King Cuo, Cheng/Zhong, Gudai bingqi, Color PL 8, no. 4-61. 99. Xiaoheishigou. Photograph © S.-K. Psarras.

100. Beijing Palace Museum collection, Cheng/Zhong, Gudai bingqi, 30, no. 2-49.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

114 Sophia-Karin Ps arras

101. Baicaopo, Lingtai Co., Gansu, Cheng/Zhong, Gudai bingqi, 65, no. 3-73.

^53fc3 CD

102. Xibozi cache (Yuhuangmiao), Beijingshi, Beijing kaogu sishinian, 70, f. 34:8, 9; Right: idem, f. 34:11.

*

103. Wujintang, KG 1960.5, 8, f. 2:5; Erdaohezi, KG 1977.5, 303, f. 3: 1 .

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 115

104. Xiaobaiyang M30: 12, M37:2, Zhangjiakoushi, "Xiaobaiyang," WW 1987.5, 46, f. 13:22, 23; Liangjiayingzi M8071:213, WWZLCK9, 31, f. 14:5; Dahuofang, KG 1964.12, 282, f. 7:1; Zhengjiawazi, KGXB 1975.1, 147, f. 7:2; Shi'ertaiyingzi, KGXB 1960.1, 67, f. 4:4.

105. Houshishan 79W-M88: 1 (bronze), Jilinsheng, "Houshishan," KGXB 1993.3, PI. 6: 1 . 106. Erdaohezi, KG 1977.5, 303, f. 2:2, 3.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

116 Sophia-Karin Ps arras

107. Wangtu. Photograph © S.-K. Psarras.

108. Nanshan'genM101:68, 67 (bronze), KGXB 1973.2, PL 10:9, 13. 109.XiaobaiyangM13:3,M31:5,M41:5,M41:10,

Zhangjiakoushi, "Xiaobaiyang," WW 1987.5,. 48, f. 15:24, 21, 17, 28.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiaji adi an 117

110. Yuhuangmiao. Photograph © S.-K. Psarras. 111. Shaogou M1040:5, Luoyangqu, Shaogou, PL 59:7.

112. Qianping M108:9, Yichang diqu bowuguan, "1978nian Yichang Qianping Hanmu fajue jianbao," KG 1985.5, 413, f. 4: 1 1 ; Tiejianggou, Shao Guotian, "Aohanqi Tiejianggou Zhanguo mudi diaocha

jianbao," NMGWWKG 1992.1-2, 88, f. 8:2. 113. Yuhuangmiao. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

118 Sophia-Karin Psarras

114. Wudaohezi M8:3, MIO: 17, M10:4, WW 1989.2, 56, f. 8:24, 25, 21. 115. Xiaobaiyang M 18:2, M31:4, M22: 1, Zhangjiakoushi, "Xiaobaiyang,"

WW 1987.5, 48, f. 15:5,8, 15. 116. Fenshuiling M49:6 (bronze), Shanxisheng, "Fenshuiling," KG 1964.3, 134, f. 26:1;

Sanguandian (gold), KG 1985.2, 129, f. 7:4.

117. Liyu, Hunyuan Co., Shanxi, detail, Zhongguo gu qingtongqi xuan, PL 59.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiaji adi an 119

118. Yanglang, "Ningxia Guyuanxian," WW 1978. 12, 87, f. 1 .

119. Pingshuo 6M50:8, Pingshuo kaogudui, "Shanxi Shuoxian Qin Han mu fajue jianbao," WW 1987.6, PL 3:5.

120. Xiaoheishigou. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 20 Sophia-Karin Psarras

121. Longtoushan. Photograph © S.-K. Psarras.

122. Xiaoheishigou. Photograph © S.-K. Psarras.

123. Xiaoheishigou. Photograph © S.-K. Psarras. 124. Xiaoheishigou. Photograph © S.-K. Psarras.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadi an 121

125. Xiaoheishigou. Photograph© S.-K.Psarras.Cf.Nanshan'genM101:64,Ji:GXß 1973.2, PL 12:1.

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

122 Sophia-Karin Psarras

Maps

Both maps drawn by S.-K. Psarras, based on Tan Qixiang ji.'UHl, Zlwngguo lishi dituji 'hOUj'ilJtk l^lili vol. 2: Qin, Xi Han, Dong Han shiqi 4> · '}l'U · /j;f3ill jlW (Beijing: Ditu, 1982). Map 2 is not to scale. For a detailed map of Upper Xiajiadian sites, see WWZLCK9, 24, f. 1.

Map 1: General overview

BACTRIA Λ ^ ° '' ^ ^ ο ' Ά Λ WESTERN REGIONS ^*-*<^-^-^

V^ Λ "

X^V^, ν Μ.·

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 123

Lake / m s' 'L· '' ' Baikal //" Λ 1? ' /

XIONGNÜ EMPIRE /' Λ Λ ^J/

^Λ^ί}_Λ ^ /^ l/l 7 Λ.

Λ ^^ ^ ΛΛ η HAN EMPIRE ^v .

ΛΛ ΛΛ λ Λ ^

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

1 24 Sophia-Karin Psarras

Map 2: Archaeological sites in the Northeast

r' Λ ο 1

Λ 2 o O, Íia0X1

• ^ ° Λ

Λ Λ Λ · ' Bohai Λ Λ Zhong- zhuo ' - ' ' shan ' ^^ _

Λ / Α Λ jT

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

Upper Xiajiadian 1 25

«27 I; " J

r~ Lelang '^

Lelang Western Han Commanderies '«^^^^ ^^^f 1-27 Modern archaeological sites *" ? ^^' f ^

Map is not to scale. *

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms

126 Sophia-Karin Psarras

Key to Sites

1. Wengniute Banner H^t^ífe, Inner Mongolia. 2. Chifeng Municipality J^il$, Inner Mongolia. 3. Aohan Banner JÜCStJjft, Inner Mongolia.

4. Ningcheng xftt Inner Mongolia. 5. Jianping ^ψ, Liaoning. 6. Lingyuan H^M, Liaoning. 7. Pingquan ψ^ Hebei. 8. Luanping %$ψ, Hebei. 9. Zhangjiakou 3^^P, Hebei. 10. Xuanhua ^ft,, Hebei. 11. Huailai 'ff^, Hebei. 12. Yanqing Jijf^, Beijing. 13. Changping ^ψ, Beijing. 14. Beijing :|t^. 15. Shunyi Jl^, Beijing. 16. Pinggu ψ£τ, Beijing. 17. Fangshan j^lll, Beijing. 18. Yixian jyL, Hebei. 19. Lüshun SJR» Liaoning. 20. Lüda g^c, Liaoning. 21. Changhai )£$$, Liaoning. 22. Jinxi fijJH, Liaoning. 23. Chaoyang ^(]Κ, Liaoning. 24. Liaoyang UK, Liaoning. 25. Shenyang ttK, Liaoning. 26. Fushun JEJIK, Liaoning. 27. Qingyuan if J^, Liaoning.

Han dynasty commanderies

Zhongshan (kingdom) ψΐ!| ( Ü )

Daif^CglO Zhuo gc Bohai igjíÊÊ Shanggu _h^ Yuyang t&ffi Youbeiping ^Jb^ Liaoxi jXH Liaodong jl]fc Lelang ^?g

This content downloaded from 128.104.46.196 on Tue, 06 Jun 2017 15:16:34 UTCAll use subject to http://about.jstor.org/terms