Art and Archaeology of the Sichuan Basin

© Brill, Leiden 2003 JEAA 5, 1–4

CONTENTS

Li Yung-ti 李永迪, On the Function of Cowries in Shang and Western Zhou China .......................................................................... 1

Shing Müller, Chin-Straps of the Early Northern Wei: New Perspectives on the Trans-Asiatic Diffusion of Funerary Practices . ........................ 27

Donald F. McCallum, Review Article: The Emergence of Japanese Kingship, by Joan R. Piggott . .............................................................................. 73

Special Section: Art and Archaeology of the Sichuan Basin

Jay Xu, Guest Editor’s Preface ................................................................. 101Wang Yi, Prehistoric Walled Settlements in the Chengdu Plain ... 109Jay Xu, Defining the Archaeological Cultures at the Sanxingdui Site ..... 149Lothar von Falkenhausen, The External Connections of Sanxingdui ... 191Zhu Zhangyi 朱章義, Zhang Qing 張擎, and Wang Fang 王方,

The Jinsha Site: An Introduction ....................................................... 247Sun Hua, The Zhuwajie Bronzes ............................................................ 277Alain Thote, Lacquer Craftsmanship in the Qin and Chu Kingdoms:

Two Contrasting Traditions (Late 4th to Late 3rd Century BC) ........ 337Michael Nylan, Ordinary Mysteries: Interpreting the Archaeological

Record of Han Sichuan ...................................................................... 375Susan N. Erickson, Eastern Han Dynasty Cliff Tombs of Santai Xian,

Sichuan Province ................................................................................ 401

Reviews

Mark Aldenderfer: John Vincent Bellezza’s Antiquities of Northern Tibet: Pre-Buddhist Archaeological Discoveries on the High, and Antiquities of Upper Tibet: Pre-Buddhist Archaeological Sites on the High Plateau ............... 471

Shing Müller: Susanne Greiff and Shenping Yin’s Das Grab des Bin Wang. Wandmalereien der Östlichen Han-Zeit in China ......................... 479

Fumiko Ikawa-Smith: Yaroslav V. Kuzmin, ed. Radiocarbon Chronology of the Stone Age of Northeast Asia .............................................................. 483


ON THE FUNCTION OF COWRIES IN SHANG AND WESTERN ZHOU CHINA

BY

LI YUNG-TI 李永迪(Institute of History and Philology, Academia Sinica, Taiwan)

AbstractCowrie shells are often found in Bronze Age sites in China. The commonly accepted explanation for their function is that they were used as money or currency during the Shang and the Zhou periods, if not as early as the Neolithic. References to cowrie shells in Shang and Zhou bronze inscriptions and in received classical texts are often regarded as evidence for such an interpretation. This paper reviews the hypothesis that cowries were money and examines textual evidence commonly cited in support of the hypoth-esis. It argues that a number of different concepts, such as wealth, value, and money, are often misleadingly conflated in the discussion of “cowrie money,” and that some of the textual references to cowries have been misinterpreted. The paper suggests that, on present evidence, cowries began to assume the role of a standard of value only during the Middle Western Zhou period. The main function of cowries in the Shang and Western Zhou periods is more likely to have been ornamental, funerary, or ritual.

Introduction: the issue

The lustrous color and the gem-like durability of cowrie shells attracted human interest long before the dawn of civilization. All over the world, they are found hundreds of miles away from their marine habitats, used as grave furnishings, personal adornments, charms and amulets, and sometimes, currency for exchange (e.g., Jackson 1917; Anders-son 1934; Quiggin 1949; Einzig 1966; Polanyi 1968b; Bronson 1976; Safer and Gill 1982; Hogendorn and Johnson 1986). Cowrie shells are commonly found at archaeological sites within the boundaries of modern day China. They have been unearthed in Gansu, Henan, Jilin, Ningxia, Qinghai, Shandong, Shanxi, Shaanxi, Sichuan, Yunnan, and as far inland as Xinjiang (e.g.,Wang Qingzheng; Huang Xiquan 2001; for an extensive survey of archaeological sites with cowries from the Neolithic to the Bronze Age, see Peng Ke and Zhu Yanshi 1995). The history of cowrie use in China goes back to the Neolithic. In the late

2 LI YUNG-TI

third millennium, cowrie shells were interred in burials of the Machang phase of the Majiayao culture (Qinghai 1976; Qinghai 1984: 167–8; cf. Peng Ke and Zhu Yanshi 1995). They were included in early second- millennium burials at Erlitou and are frequently encountered at Shang and Zhou period sites (for reviews of cowrie use in the Shang and Zhou periods, see Egami 1974; Peng Ke and Zhu Yanshi 1995; Kondō 1998, 1999; Yang Shouchuan 2000; Wu Danmin 2002). At a much later time, Marco Polo recorded witnessing tax payments made in cowrie shells by the lords of Yunnan to the Yuan court (Marco Polo 1929, quoted in Hogendorn and Johnson 1986). In fact, cowrie shells were the means of Yunnan’s tax payments to the central government until the Qing dynasty (Li Jiarui 1956; Fang Guoyu 1981; Wang Dadao and Zhao Zhensheng 1989; Vogel 1993; Yang Shouchuan 2000). As recently as the first half of the twentieth century, Andersson recorded the use of cowrie shells as amulets in Gansu (Andersson 1934). The earliest metal coins in China are now dated to around the middle of the Eastern Zhou period (Wang Yu-chuan 1951; Peng Xinwei 1965; Wang Xiantang 1979; Zhu Huo 1984; Peng Xinwei 1994). Many scholars have argued that before the minting of coins in China, other forms of money were already in use, and cowries are often considered the earliest money in China (e.g., Wang Yü-chüan 1951; Kao Chü-hsün 1954; Wang Xiantang 1979; Xiao Qing 1984; Zhao Ningfu and Ji Jin 1984; Zhu Huo 1984; Wang Qingzheng 1988; Zhao Shande 1988; Yang Shengnan 1992; Peng Xinwei 1994; Wei Si 1998; Yang Shouchuan 2000; Huang Xiquan 2001; cf. Egami 1974; Kondō 1998; Wu Rongz-eng in Zhongguo Qianbi 2001: 34). The same argument has also been widely accepted in the West. A quick survey of Western publications, especially the secondary literature on the history of money, shows an almost unanimous acceptance of the idea of cowrie currency in Bronze Age China (e.g., Gibson 1940; Quiggin 1949; Yang Lien-sheng 1952; Einzig 1966; Davies 1994). Yet despite the acceptance and propagation of such claims, existing discussions of the use of “cowrie money” are often ambiguous, contra-dictory, and unsubstantiated. Many archaeologists in China, including pioneering scholars like Li Chi (= Li Ji) 李濟 and Tung Tso-pin (=Dong Zuobin) 董作賓, have stated in their publications that cowries found at early Bronze Age sites were used as money, but most have provided no specific evidence (e.g., Li Chi 1933: 575). Sweeping speculations are often made about the circulation of cowrie currency; it is even suggested that imitation cowries in media such as bone, jade, and stone represent different denominations or local versions of cowrie currency (e.g., Zheng Jiaxiang 1959; Xiao Qing 1984; Zhao Shande 1988; Yang Shengnan 1992). Such statements have usually ignored the political, economic, and social contexts of the supposed use (Wang Yu-chuan 1951).

ON THE FUNCTION OF COWRIES IN SHANG and W. ZHOU 3

In this paper, I would like to argue that the hypothesis that cowries were used as money in Bronze Age China is not based on solid evidence. Key concepts such as value, treasure, and money have been confused; the textual and inscriptional data have been dubiously interpreted. Once we weed out the problematic evidence, we have little basis for arguing that cowries were used as currency in Shang and Western Zhou China. In fact, the present evidence suggests that even as late as Middle Western Zhou, cowries had not yet assumed full monetary functions.

The biology of cowries

About 200 species of cowries are known (Burgess 1985; Hogendorn and Johnson 1986; Lorenz and Hubert 1993). The most commonly identified cowries at archaeological sites in China are Cypraea moneta, or money cowrie, and Cypraea annulus, or ring cowrie (Zheng Jiaxiang 1959; Zhao Shande 1988; Jung Bor-sheng 1993). The two species occur widely in warm sea waters and shallow lagoons in the basins of the Indian and Pacific Oceans. They are often found together, and their habitats range from the Red Sea to Mozambique in the west, and to Japan, Hawaii, New Zealand, and the Galapagos in the east. Concentrations of Cypraea moneta exist in the Sulu Islands of the Philippines; in Indonesia; and most importantly in the Maldive Islands of the Indian Ocean, the primary source of cowries for many parts of the world, including Southeast Asia, India, and Africa (Hogendorn and Johnson 1986). Scholars disagree about the distribution of cowries in the coastal regions of China during the Bronze Age. While cowries can be found nowadays in the southeastern coastal provinces, and especially in Taiwan (Jung Bor-sheng 1993), it is not clear whether they reached further north to the Yellow Sea during the Bronze Age because of the slightly warmer climate (see Jung Bor-sheng 1993 for a review of the debate). Peng Ke and Zhu Yanshi (1995) in a recent article argue that in the Bronze Age cowries had not even reached the South China Sea and that cowries found in archaeological contexts must have been imported from west-ern sources through the northwestern corridor of China. Nonetheless, if we consider both the warmer climate of Bronze Age China and the modern distribution, the possibility that cowries were present at least in the southeastern coastal area and that sources there were exploited during the Shang and Zhou times seems higher than Peng and Zhu have argued (for hypotheses about ancient routes through which cowries were transferred into inland China, see Kinoshita 2003: 29–30).1

1 An ongoing project by Kinoshita (2003) investigates the possibility that the Ryukyu Islands chain was one of the sources of cowries found in Bronze Age sites in China.

4 LI YUNG-TI

Cowries as Money

Economists and philosophers of money argue that a suspension of reasoning must occur in the acceptance of money (see Einzig 1966; Pryor 1977; Crump 1981; Orléan 1992 for a review of the issue). For something to circulate as money and to be accepted in every transac-tion, there has to be universal agreement among individual agents as to the function and value of the item, agreement that entails a certain confidence or blindness. The confidence is based on faith, habit, or, in complex societies, strong institutional support. Once the agreement is in place, money as an object disappears. It becomes a universally recognized symbol, independent of what the object actually is (Polanyi 1968a; Orléan 1992; Parmentier 2002: 51–53). This means that the choice of object to be used for money is often arbitrary. Various ethno-graphic accounts have established that the physical attributes of cowries make them an appealing form of money (e.g., Quiggin 1949; Einzig 1966; Polanyi 1968b; Safer and Gill 1982; Evans 1985; Hogendorn and Johnson 1986). They are durable; their luster is appealing; their uniformity in shape and size make them easy to count; and they are impossible to counterfeit.2 However, these physical attributes do not by themselves make cowries money. Though cowries are geographically the most widely used currency in the world (Hogendorn and Johnson 1986), it does not follow that they are money in every context in which we encounter them. The existence of currency cannot be inferred from the presence of objects that are suitable for performing monetary func-tions. To decide whether cowries were used as money in Bronze Age China, we must examine the evidence.

The argument that cowries were money in Bronze Age China

Archaeological data, unfortunately, do not provide firm answers to our question. Without a behavioral or social context, cowries themselves, unlike minted coins, do not disclose any meaning. The hypothesis that cowries were money, as put forth in the literature, is therefore based mainly on textual and inscriptional data. What follows is a brief summary of the evidence commonly quoted in support of the hypothesis:

2 A different opinion was offered by Xie Zhaozhi 謝肇淛, who served as an official in Yunnan from 1615 to 1622. In comparing cowries, paper money, and copper coins, Xie contended in Wu za zu 五雜組 (volume 12, p. 33; Xie Zhaozhi 2000: 606) that cowrie shells were difficult to carry in large quantities and broke easily. Despite Xie’s observation, however, cowries circulated for a long time in Yunnan. Although they are not indestructible, their shells are fairly thick and able to withstand a fair amount of weight and force.


1. Cowrie-granting inscriptions are one of the most common types of bronze inscription in the Shang and early Western Zhou periods (see Huang Ranwei 1978 for a comprehensive study of gift and cowrie-granting inscriptions). Such inscriptions record the bestowal of cowries by the king or a high court official upon subordinates. The cowries granted are usually counted in peng 朋, or strands, and the phrase yong zuo bao zun yi 用作寶尊彝 or yong zuo zun yi 用作尊彝 (“used to make this precious vessel;” see discussion below)3 often follows immediately after the number of strands. The character yong in this sentence is usu-ally interpreted as “used for,” and yong zuo is therefore taken to mean “used to make” (e.g., Ma Chengyuan 1988: vol. 3, p. 4, note 4). In other words, it is assumed that the cowries were used, or spent, by the receiver of the gift to cast the vessel. From this it is inferred that cowries were used as either a means of payment for the casting of the vessel or as the medium of exchange used to purchase the necessary material for casting it (e.g., Wang Yü-chüan 1951; Wang Xiantang 1979; Zhu Huo 1984; Yang Shengnan 1992).

2. A supposedly Western Zhou bronze vessel, the Ju Bo Huan gui 遽伯還簋, has an inscription that reads Ju Bo Huan zuo bao zun yi, yong bei shi peng you si peng 遽伯還作寶尊彝、用貝十朋又四朋; or “Duke Huan of Ju made this precious vessel, using fourteen strands of cowries” (see Sun Zhichu 1981: 129, entry no. 2201; Ferguson 1991: 561, 676, for references to the inscription in various bronze inscription catalogues). The fourteen strands of cowries mentioned in the inscription have been interpreted as the price of the vessel. In other words, it is assumed that the cowries were used to pay for the vessel’s casting (e.g., Wang Yü-chüan 1951; Kao Chü-hsün 1954; Zhu Huo 1983; Cai Yunzhang 1989; Yang Shengnan 1992; Peng Xinwei 1994).

3. The inscription on the Qiu Wei he 裘衛盉, a Western Zhou ves-sel unearthed in Qishan 岐山 , Shaanxi in 1975 (see below), records a transaction between Qiu Wei and a feudal lord in which land was exchanged for ceremonial jades and attire. The value of the jades and perhaps the value of the lands as well were measured in strands of cow-ries. Many have argued that valuation in terms of strands of cowries implies a well-established monetary system (e.g., Zhu Huo 1983; Cai Yunzhang 1989; Yang Shengnan 1992).

4. Characters related to economic transactions or to value and wealth in oracle bone and bronze inscriptions frequently include the radical bei

3 There are several variations of this phrase. The earlier form is yong zuo + ancestor name + vessel name, e. g. yong zuo Fu Ji bao yi 用作父己寶彝 (“used to make this pre-cious vessel dedicated to Father Yi”) in the Xiaozi Sheng hu inscription (see discussion below).

6 LI YUNG-TI

4 For discussions of these inscriptions and issues related to land exchange in Western Zhou, see Lin Ganquan 1976; Pang Huaiqing 1976; Tang Lan 1976a, 1976b; Zhao

貝, which is often interpreted as depicting and signifying a cowrie. This is taken as evidence that cowries were used as a medium of exchange (e.g., Gibson 1940; Kao Chü-hsün 1954; Wang Xiantang 1979; Wan Xianchu 2002).

5. Various historical texts, such as the Shi 詩, the Book of Songs, and the Shu 書, the Book of Documents, have passages describing the granting and receipt of cowries. In these passages, cowries are often described as valuables. For many scholars, this implies a monetary function for cowries (e.g., Kao Chü-hsün 1954; Wang Yü-chüan 1951; Wang Xian-tang 1979; Zhu Huo 1984).

Review of the Argument

Historical and etymological evidence

As Peng Xinwei 彭信威 has argued in his book A Monetary History of China (Peng Xinwei 1994), scholars who quote historical texts in support of use of cowrie money in early Bronze Age China regularly confuse the concepts of “value” (or “treasure”) and “money.” The passages in the Shu and the Shi which say that cowries were treasured for their value cannot be taken to imply that they had an economic or monetary func-tion. Like jades, turtle shells, and other luxuries, cowries were considered exotic and valuable in Bronze Age China. Like diamonds today, they were treasures, but not necessarily money. The same can be said of the etymological evidence. Peng made it clear that there is no need to invoke the meaning “money” in explaining the occurrence of the cowrie radi-cal in characters related to economic activities. The concept of “value” or “treasure” is enough to explain the presence of cowries in both the historical texts and the composition of Chinese characters.

Qiu Wei he and Kang ding

In 1975, 37 vessels, most of them inscribed, were found in a cache in Dongjia village, Qishan county, Shaanxi (Pang 1976). Four of the 37 vessels were made by Wei; three of the four—the Qiu Wei he, the Fifth Year Wei ding, and the Ninth Year Wei ding—bear inscriptions that record exchanges between Wei and members of the Zhou elite.4 The exchange described in the Qiu Wei he inscription (Figure 1) is between the Duke of


Ju 矩 and Qiu Wei. The Duke, in order to attend a ritual performed by the Zhou king, gave up a few pieces of land in exchange for ceremonial attire and jades from Qiu Wei. The value of the jades, as measured in strands of cowries, was 80 peng (cai bashi peng 才八十朋), for which 10 units of land were given up (she tian shi tian 舍田十田). The value of the ceremonial attire was 20 peng (cai ershi peng), for which 3 units of land were given up (she tian san tian). The exchange was supervised and witnessed by an array of Zhou court officials. The Qiu Wei he inscription has been frequently cited to support the existence of an established Western Zhou monetary system (e.g., Zhu Huo 1983; Cai Yunzhang 1989; Yang Shengnan 1992). Certainly it is clear that during the exchange between Duke Ju and Qiu Wei, cowries were used as a standard of value for the items exchanged. However, it is not at all clear that cowries changed hands during the transaction, serving as a medium of exchange or a means of payment. Although economists disagree among themselves as to which of the two functions of money comes first, standard of value or medium of exchange (see von Glahn 1996: 16, note 1, for references), it is commonly recognized that when something is used as a medium of exchange, the function of standard of value is invoked, but that the reverse is not the case (e.g., Einzig 1966; Polanyi 1968a; Grierson 1977; Pryor 1977; Crump 1981; Davies 1994; cf. Melitz 1970; Melitz 1974). A medium of exchange changes hands during a transaction, but something used as a standard of value need not do so. Examples illustrating the func-tion of a standard of value can be found in the Iliad and Odyssey (for references, see Einzig 1966: 220–223, footnotes): when goods or slaves are exchanged or purchased, their value is measured in units of oxen. The ox was used as a “notional standard” (Grierson 1977: 16), but the actual payments were made using other means (Einzig 1966: 221). Another often quoted example is the weight measure for copper, deben, in ancient Egypt. Textual materials have shown that the value of the exchanged items was measured in terms of deben, but the transaction did not involve the actual trading of copper (Einzig 1966: 193–202; see also Kemp 1989: 248–250). In the case of the Qiu Wei he inscription, first the value of the jades and ceremonial attire was measured in terms of cowries (cai X peng), then units of land, presumably of that same value, were given up (she tian X tian), but there is no mention that cowries were obtained (qu 取) by anyone or that cowries changed hands. In other words, the inscrip-tion shows only that cowries were used to measure the value of certain

Guangxian 1979; Huang Shengzhang 1981; Matsui 1984; Itō 1987; Ma Chengyuan 1990; Li Ling 1992; Takeuchi 1993; Shaughnessy 1997, 1999.

Figure 1. Qiu Wei he (Zhongguo Qingtongqi 1996: 107, no. 112) and the inscription (Shanghai 1986: 92, no. 193).


items; it does not establish that cowries were the medium of exchange or an acceptable means of payment. Cowries were the notional standard; they need not have been money. The same can be said for the quantifier peng. Peng is a unit of mea-sure used to count cowries, and should not be confused with the unit of account or standard of value of a currency. Although the mention of certain numbers of peng of cowries can be found in oracle bone and bronze inscriptions, such mentions should be seen as referring to the quantity of cowries themselves, not to amounts of cowrie money. It should be noted that there is no consensus among scholars in the field as to how many cowries one peng contained (e.g., Guo Baojun 1936: 194; Kao Chü-hsün 1954; Wang Guowei 1959; Wang Xiantang 1979; Guo Moruo 1982). It is often ignored that the Qiu Wei he inscription’s use of cowries as a standard of value is very unusual. Inscriptions which describe the exchange of lands or gifts by feudal lords form only a small fraction of the corpus of Western Zhou bronze inscriptions, and among them the Qiu Wei he inscription is one of only a few (see below) that mention strands of cowries in acts of exchange. On neither of the two Qiu Wei ding from the 1975 find does the inscription specify the value of the exchanged items in terms of cowries. The inscription on the Ninth Year Wei ding records another exchange between Duke Ju and Qiu Wei. It describes how Duke Ju gave Qiu Wei a piece of forested land and a village in exchange for an elaborately decorated chariot. Cowries are not mentioned. Nor are they mentioned in the inscription of the Fifth Year Wei ding, which records a dispute over an exchange of land between Qiu Wei and Lord Li 厲 (for an English translation of the inscription, see Shaughnessy 1997: 79–81). Another Western Zhou inscription recording a similar exchange between feudal lords appears on a bronze ding 鼎 recently acquired by the Shanghai Museum (Figure 2). The vessel, the Kang ding 亢鼎, is dated to Early Western Zhou on stylistic grounds (Ma Chengyuan 2000). It bears an inscription of 49 characters recording an exchange between Gong Dabao 公大保 and X Ya 亞. The following discussion is based on Ma’s interpretation of the inscription in Ma Chengyuan (2000). The exchange involved a large piece of jade valued at 50 peng of cowries (cai wushi peng 才五十朋 ) owned by X Ya. The recipient of the jade, Gong Dabao, gave X Ya 50 peng of cowries along with alcoholic beverages, an ox, and other items yet to be identified. The person representing Gong Dabao in the exchange, Kang 亢, was granted a red ox and two units of copper by X Ya. The character mai 買 [or ] is used to refer to the act of exchange; this is the earliest known use of the character mai in such a context (Ma Chengyuan 2000: 121).

10 LI YUNG-TI

Figure 2. Kang ding, now in the Shanghai Museum (Ma Chengyuan 2000: 120–121).

Though it is evident that cowries changed hands, the exact nature of the transaction recorded in the Kang ding inscription remains unclear. It was among members of the Zhou elite and involved items that may have been meant for ceremonial use. The transaction was performed not by Gong Dabao and X Ya directly, but through a third person, Kang, who represented Gong Dabao. Considering that Gong Dabao’s side of the transaction included other items besides cowries, and that


X Ya made a substantial gift to the intermediary, the exchange would seem not to have been a purchase paid for with cowries, but something more like bartering (Ma Chengyuan 2000). In other words, the cowries seem to be items exchanged, not the means of payment. The value of the exchanged items was probably negotiated between the two sides, and additional items were appended by Gong Dabao, perhaps as a gesture of gratitude to X Ya for having allowed such an exchange to take place. Although in later times the character mai invariably referred to a monetary transaction, we cannot assume that it had the meaning of “purchase” from the moment of its first appearance in a Western Zhou inscription; in the Kang ding inscription it may only signify the act of exchange. The few bronze inscriptions in which cowries are referred to in connection with acts of exchange do not establish the use of cowrie money or currency in Early and Middle Western Zhou. What we see instead is the process in which cowries gradually assumed monetary functions because of their role as a standard of value in the exchange of ritual and ceremonial items among the Zhou elite.

Ju Bo Huan gui

The Ju Bo Huan gui inscription (Figure 3), which mentions the casting of a vessel and the number of strands of cowries “used,” may seem the most convincing evidence for the cowrie-as-money argument, and it has been widely cited. Without exception, the scholars who use it as evidence seem unaware of the vessel from which it comes. The Ju Bo Huan gui, now in the Nelson-Atkins Museum of Art at Kansas City,5 is almost certainly not ancient; it is an archaistic bronze in Western Zhou style made in Ming or Qing times (Watson 1973).

Yong zuo bao zun yi

A large fraction of Shang and Zhou bronze inscriptions record gift- granting events. Although cowries are the gift most frequently mentioned in late Shang inscriptions, other items, such as jades and ceremonial beverages, do occur. The gift list became more and more elaborate over time, and by the Middle Western Zhou period, jades, ceremonial attire, and even horses and chariots were among the gifts bestowed (see Kaizuka 1977: 102–166; Huang Ranwei 1978; Shaughnessy 1991, 1997). Few scholars have argued that all the items included in Middle West-ern Zhou gift lists were used to pay the cost of casting the bronze vessel

5 I am grateful to Robert Bagley for identifying the vessel for me.

Figure 3. Ju Bo Huan gui, now in the Nelson-Atkins Museum of Art at Kansas City. Photo courtesy of the Nelson-Atkins Museum of Art at Kansas City.


(cf. Cai Yunzhang 1989). The inscriptions of this period often include the sentence dui yang wang/tianzi xiu 對揚王/天子休, “in response extols the king’s/the Son of Heaven’s beneficence” (translation by Shaugh-nessy 1991: 83–84), inserted after the gift list and immediately before the phrase yong zuo bao zun yi. In this context, separated from the gifts, “yong zuo” has been translated as “herewith making” (Shaughnessy 1991: 84) or “using [this occasion I] made” (Mattos 1997: 86), and on this interpretation, the phrase dui yang wang xiu yong zuo bao zun yi can be translated as follows: “To extol the king’s beneficence, on [this occasion of this gift I] made this treasured vessel” (see Shaughnessy 1997: 73–74 for an example of the phrase and a more detailed translation). The Shang and Early Western Zhou cowrie-granting inscriptions are in fact gift-granting inscriptions in which cowries were the only gift recorded. Even though the phrase yong zuo bao zun yi does not appear together with dui yang wang xiu in cowrie-granting inscriptions, it should not be interpreted differently from the same phrase in the later Middle Western Zhou gift-granting inscriptions. In other words, yong zuo bao zun yi in cowrie-granting inscriptions should also be understood to mean “for that reason, a vessel is cast” (Egami 1974: 20). On this interpreta-tion, the inscription records the casting of a vessel to commemorate an event, the king’s gift. It is not a record of bookkeeping for the cost of the bronze vessel. Three inscriptions from the Shang period can be used to illustrate the point. The inscription on the Sixth Year Bi Qi you 六祀G其卣 (Figure 4), which mentions that two ceremonial jades were bestowed upon Bi Qi, is one of a very few Shang inscriptions that record gifts other than cowries.6 The jades in this inscription are similar to the gifts listed in Western Zhou bronze inscriptions and, like the cowries, should not be interpreted as the cost of casting the vessel. The Shanghai Museum’s Xiaozi Sheng hu 小子省壺 (Figure 5) has an-other interesting Shang inscription. Within the usual sequence of date, gift ceremony (mentioning bestower, recipient, and gift), and dedication, this inscription inserts a fourth part. Before the dedicatory phrase yong zuo Fu Ji bao yi 用作父己寶彝, we read Sheng yang jun shang 省揚君賞, “Sheng, extolling the lord’s beneficence.” This is clearly the equivalent of dui yang wang xiu in later inscriptions. On the Xiaozi Sheng hu, we see the standardized formula of Western Zhou dedications beginning to take shape.

6 The authenticity of the so-called Second and Fourth Year Bi Qi you inscriptions is in doubt (several papers published in Gugong Bowuyuan Yuankan 1998, no.4, pp. 1–16 argue in favor of the authenticity without, however, answering all the questions raised in Zhang Zhenglang 1998 and Bagley 1987: 525–528). The authenticity of the Sixth Year Bi Qi you inscription has not been challenged.

Figure 4. Sixth Year Bi Qi you, now in the National Palace Museum, Beijing (Zhong-guo Qingtongqi 1997: 129, no. 128), and the inscription on the vessel (Shanghai

1986: 9, no. 14).

Figure 5. Xiaozi Sheng hu, now in the Shanghai Museum (Zhongguo Qing-tongqi 1998: 147, no. 151), and the inscription inside the lid (Shanghai

1986: 10, no. 17).

Figure 6. Xiaozi Feng you, now in the Hakutsuru Fine Art Museum, Japan (Hakutsuru 2000: 22, no. 20), and the inscription on the vessel (Shanghai

1986: 4, no. 5).


The inscription on the Hakutsuru Bijutsukan’s 白鶴美術館 Xiaozi Feng you 小子卣 (Figure 6) is one of the longest Shang bronze inscriptions so far known. It records in some detail the granting of two strands of cowries by Zi 子 to Xiaozi Feng, saying that Zi gave the cowries and then announced the reason for the award: bei wei mie ru li 貝唯蔑汝歷, “These cowries are to reward you for your merits.” “Merits” in this case refers to Xiaozi Feng’s success in a military campaign. Since the phrase yong zuo Mu Xin yi 用作母辛彝 comes not after the number of cowries but after Zi’s praising of Xiaozi Feng, it would be difficult to argue that the cowries were spent to cast the bronze vessel. The phrase “mie li” also became a standardized expression in Western Zhou gift-granting inscriptions (for discussions, see, for instance, Tang Lan 1979; Shirakawa 1980). In summary, it can be argued that cowrie-granting inscriptions such as the three late Shang examples just discussed are simply a precursor form of the later gift-granting inscriptions. The phrase yong zuo bao zun yi in cowrie-granting inscriptions should, in other words, be understood as “on this occasion I made this treasured vessel,” or “for this reason, a vessel is cast.” Even though the Shang inscriptions do not follow the sort of fixed formula we encounter in Western Zhou inscriptions, they do refer to similar events in a similar social and political context. The vessel and inscription were cast to commemorate an honor received by the owner. Like jades, chariots, and ceremonial attire, cowries are mentioned by way of specifying the king’s gift. They are a measure of the king’s esteem, not of the value of the bronze. They do not have a monetary function.

Conclusion

According to the Shi Ji 史記, bei, or sea shells, were widely circulated before the Qin empire established a unified monetary system. Yet, tempt-ing though it may be to make a connection between the circulation of bei in Eastern Zhou and the presence of cowries at Shang and Western Zhou archaeological sites, the available textual and inscriptional data do not confirm the use of cowries as money in the Shang and Western Zhou periods. Cowries mentioned in Shang and Early and Middle Western Zhou bronze inscriptions were prestige items given by the king or other members of the elite to their subordinates. They were an established part of elite ritual exchange. They were treasured probably because their maritime origin made them exotic items procured at some cost through long-distance trade networks, and they were often used as ornaments. In archaeological contexts, they are typically found as

18 LI YUNG-TI

components of personal adornments and horse trappings (e.g., Egami 1974; Kondō 1998, 1999; Wu Danmin 2002). Imitations made in other materials may also have been for decorative use. Cowries may even have had religious functions, for oracle bone inscriptions document inquiries about the proper quantities of cowries to be used in rituals. Most of all, cowries seemed to have been used as mortuary objects. Their placement in Shang and Zhou burials in locations such as the mouth, the hands, and areas around the knees or ankles of the deceased suggests that they were the early form of the han 含 and the wo 握 of Zhou mortuary custom (Kao Chü-hsün 1954; Wu Danmin 2002). It is possible that the widespread use of cowries for ornamental, religious, and mortuary purposes among the Shang and Zhou elite made them suitable for use as royal gifts. By the Middle Western Zhou period, these standard gifts were perhaps beginning to be thought of as a standard of value. This, in turn, may have eventually led to the use of cowries as a medium of exchange and means of payment. But on present evidence, it seems unlikely that cowries were used as money until some time after the Middle Western Zhou period.

Acknowledgements

I would like to thank Lothar von Falkenhausen and especially Robert Bagley for making valuable suggestions to the earlier drafts of this paper. I would also like to thank Chen Kwang-tzu 陳光祖 for gener-ously sharing his knowledge on the subject. Hwang Ming-chorng 黃銘崇, Richard Parmentier, and Robert Hunt all kindly offered their comments, making improvement of the paper possible. I am solely responsible for any mistakes.

References cited

Andersson, J. Gunnar (1997). Children of the Yellow Earth: Studies in Prehistoric China (reprint of 1934 edition). Cambridge, Mass.: MIT Press.

Bagley, Robert W. (1987). Shang Ritual Bronzes in the Arthur M. Sackler Collections. Washington D.C.: Arthur M. Sackler Foundation.

Bronson, Bennet (1976). “Cash, cannon, and cowrie shells: the non-modern moneys of the world.” Field Museum of Natural History Bulletin, volume 47, no. 10: 3–15.

Burgess, Clarence M. (1985). Cowries of the World. Cape Town: G. Verhoef, Seacomber Publications.

Cai Yunzhang 蔡運章 (1989). “Xi Zhou huobi goumaili qianlun—jian-tan Xi Zhou wujia de ruogan wenti 西周貨幣購買力淺論—–兼談西周物價的若干問題 (The purchasing power of Western Zhou cur-


rency—also issues related to commodity prices in Western Zhou).” Zhongguo qianbi 中國錢幣 1989.1: 31–37.

Crump, Thomas (1981). The Phenomenon of Money. London: Routledge & Kegan Paul.

Davies, Glyn (1994). A History of Money: From Ancient Times to the Present Day. Cardiff: University of Wales Press.

Egami, Namio (1974). “Migration of the cowrie-shell culture in East Asia.” Acta Asiatica: Bulletin of the Institute of Eastern Culture 26: 1–52.

Einzig, Paul (1966). Primitive Money: In Its Ethnological, Historical and Eco-nomic Aspects. Oxford: Pergamon Press.

Evans, Philip (1985). “The cowrie economy and the maintenance of social boundaries in northern Ghana.” Cambridge Anthropology 10(2): 29–40.

Fang Guoyu 方國瑜 (1981). “Yunnan yong bei zuo huobi de shidai ji bei de laiyuan 雲南用貝作貨幣的時代及貝的來源 (The time periods when cowries were used as currency and the sources of cowries).” Yunnan shehui kexue 雲南社會科學 1981.1: 24–40.

Ferguson, John Calvin 福開森 (1991). Lidai zhulu jijin mu 歷代著錄吉金目 (Citation index of published bronze inscriptions). Beijing: Zhong-guo Shudian (reprint of the 1939 original edition).

Gibson, Harry E. (1940). “The use of cowries as money during the Shang and Chou periods.” Journal of the North China Branch, Royal Asiatic Society 71: 33–45.

Grierson, Philip (1977). The Origins of Money. London: The Athlone Press.

Guo Baojun 郭寶鈞 (1936). “Xunxian Xincun gu can mu zhi qingli 濬縣辛村古殘墓之清理 (The cleaning of a partially preserved ancient tomb at Xincun, Xunxian).” Tianye kaogu baogao 田野考古報告 1: 167–200 .

Guo Moruo 郭沫若 (1982). “Shi peng 釋朋 (Interpreting peng).” In Jiagu wenzi yanjiu 甲骨文字研究 (Studies of oracle bone inscription), pp. 107–114. Beijing: Kexue Chubanshe.

Hakutsuru 2000: Hakutsuru Bijutsukan 白鶴美術館 (2000). Hakutsuru bijutsukan meihinsen 白鶴美術館名品選 (Art collection of the Haku-tsuru Fine Art Museum). Kōbe: Hakutsuru Bijutsukan.

Hogendorn, Jan and Marion Johnson (1986). The Shell Money of the Slave Trade. Cambridge: Cambridge University Press.

Huang Ranwei 黃然偉 (1978). Yin Zhou qingtongqi shangci mingwen yanjiu 殷周青銅器賞賜銘文研究 (Study of the granting inscriptions on Shang and Zhou bronzes). Hong Kong: Longmen Shudian.

Huang Shengzhang 黃盛璋 (1981). “Wei he, ding zhong ‘zhu’ yu ‘zhu tian’ ji qi qianshe de Xi Zhou tianzhi wenti 衛盉、鼎中 ‘貯’ 與 ‘貯田’ 及其牽涉的西周田制問題 (On ‘zhu’, ‘zhu tian’ in the inscriptions of

20 LI YUNG-TI

Wei he and ding, and the issues of Western Zhou land tenure system).” Wenwu 文物 1981.9: 79–82.

Huang Xiquan 黃錫全 (2001). Xianqin huobi tonglun 先秦貨幣通論 (A general survey of pre-Qin money). Beijing: Forbidden City Publish-ing House.

Itō Michiharu 伊藤道治 (1987). Chūgoku kodai kokka no shihai kōzō 中国古代国家の支配構造 (The control structure of ancient Chinese states). Tokyo: Chū’ō Koronsha.

Jackson, John Wilfred (1917). Shells as Evidence of the Migrations of Early Culture. Manchester: University Press.

Jung Bor-sheng (= Zhong Bosheng) 鍾柏生 (1993). “ Shiyusuo cang Yinxu haibei ji qi xiangguan wenti chutan 史語所藏殷墟海貝及其相關問題初探 (A preliminary investigation of sea shells from Yinxu in the collection of the Institute of History and Philology). Bulletin of the Institute of History and Philology Academia Sinica 中央研究院歷史語言研究所集刊 64(3): 687–737.

Kaizuka Shigeki 貝塚茂樹 (1977). Chūgoku kodai shigaku no hatten 中国古代史学の発展 (The development of the study of ancient Chinese History). Tokyo: Chū’ō Koronsha.

Kao Chü-hsün (= Gao Quxun) 高去尋 (1954). “Yin li de han bei wo bei 殷禮的含貝握貝 (The ritual use of cowrie shells in the Yin Period).” Annals of the Academia Sinica 中央研究院院刊 1: 373–401.

Kemp, Barry J. (1989). Ancient Egypt: Anatomy of a Civilization. Routledge: London and New York.

Kinoshita Naoko 木下尚子 (2003). “Cong gudai zhongguo kan liuqiu liedao de baobei 從古代中國看琉球列島的寶貝 (Ancient China and cowries of the Ryukyu island chain.” Sichuan wenwu 四川文物 2003.1: 29–34.

Kondō Takaichi 近藤喬一 (1998). “Shangdai haibei zhi yanjiu 商代海貝之研究 (A study of sea shells of the Shang period).” In Institute of Archaeology, CASS (editor), Zhongguo Shang wenhua guoji xueshu tao-lunhui lunwenchi 中國商文化國際學術討論會論文集 (Papers of the international symposium on the Shang culture in ancient China), pp. 388–412. Beijing: Kexue Chubanshe.

Kondō Takaichi 近藤喬一 (1999). “Xi Zhou haibei zhi yanjiu 西周海貝之研究 (A study of sea shells of the Western Zhou period).” In Zhong-guo Yinshang Wenhua Xuehui et al. 中國殷商文化學會等 (editors), Sandai wenming yanjiu 1 三代文明研究 (1) (Studies of the Three Dynas-ties, volume 1), pp. 412–416. Beijing: Kexue Chubanshe.

Li Chi (= Li Ji) 李濟 (1933). “Anyang zuijin fajue baogao ji liu ci gong-zuo zhi zong guji 安陽最近發掘報告及六次工作之總估計 (Report of recent excavations at Anyang and summary of the six excavation seasons).” Anyang fajue baogao 安陽發掘報告 4: 559–577.

Li Jiarui 李家瑞 (1956). “Gudai yunnan yong beibi de dagai qingxing


古代雲南用貝幣的大概情形 (Overview of the use of cowrie money in ancient Yunnan).” Lishi Yanjiu 歷史研究 1956.9: 85–100.

Li Ling 李零 (1992). “Xi Zhou jinwen zhong de tudi zhidu 西周金文中的土地制度 (The land tenure system as seen in Western Zhou bronze inscriptions).” Xueren 學人 2: 224–256.

Li Xueqin 李學勤 (1989). “Chongxin gujia Zhongguo gudai wenming 重新估價中國古代文明 (Re-evaluating ancient Chinese civilization).” In Li Xueqin 李學勤, Li Xueqin ji 李學勤集 (Collected essays by Li Xueqin), pp. 15–27. Haerbin: Helongjiang Jiaoyu Chubanshe.

Lin Ganquan 林甘泉 (1976). “Dui Xi Zhou tudi guanxi de jidian xin renshi 對西周土地關係的幾點新認識 (New understanding of the Western Zhou land tenure system).” Wenwu 文物 1976.5: 45–49.

Lorenz, Felix and Alex Hubert (1993). A Guide to Worldwide Cowries. Wiesbaden, Germany: Verlag Christa Hemmen.

Ma Chengyuan 馬承源 (ed.) (1988). Shang Zhou qingtongqi mingwen xuan 商周青銅器銘文選 (A selection of Shang and Zhou bronze inscrip-tions). Beijing: Wenwu Chubanshe.

Ma Chengyuan 馬承源 (1990). “Xi Zhou jinwen zhong youguan zhu zi ciyu de ruogan jieshi 西周金文中有關貯字辭語的若干解釋 (Interpretations for the character zhu and other related terms in Western Zhou bronze inscriptions).” Shanghai Bowuguan jikan 上海博物館集刊 5: 82–91.

Ma Chengyuan 馬承源 (2000). “Kang ding mingwen—Xi Zhou zaoqi yong beibi jiaoyi yuqi de jilu 亢鼎銘文—–西周早期用貝幣交易玉器的紀錄 (Inscription on Kang ding: an Early Western Zhou record of exchanging jades with cowrie money). Shanghai Bowuguan jikan 上海博物館集刊 8: 120–123.

Matsui Yoshinori 松井嘉德 (1984). “Sei-Shū tochi ijō kinbun no ichi kosatsu 西周土地移讓金文の一考察 (A study of land transfer on Western Zhou bronze inscriptions).” Tōyōshi kenkyū 東洋史研究 43 (2): 1–30.

Mattos, Gilbert L. (1997). “Eastern Zhou bronze inscriptions.” In E. L. Shaughnessy (editor), New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts, pp. 85–123. Berkeley: The Society for the Study of Early China and the Institute of East Asian Studies, University of California, Berkeley.

Melitz, Jacques (1970). “The Polanyi school of anthropology on money: an economist’s view.” American Anthropologist 72:1020–1040.

Melitz, Jacques (1974). Primitive and Modern Money: An Interdisciplinary Approach. Reading, Mass.: Addison-Wesley Publishing Company.

Orléan, André (1992). “The Origin of Money.” In Francisco J. Varela and Jean-Pierre Dupuy (editors), Understanding Origins: Contemporary Views on the Origin of Life, Mind, and Society, pp. 113–143. Boston: Kluwer Academic Publishers.

22 LI YUNG-TI

Pang Huaiqing 龐懷清 (1976). “Shaanxi sheng Qishan xian Dongjia cun Xi Zhou tongqi jiaoxue fajue jianbao 陝西省岐山縣董家村西周銅器窖穴發掘簡報 (Preliminary excavation report of a Western Zhou bronze cache at Dongjiacun, Qinshan county, Shaanxi province).” Wenwu 文物 1976.5: 26–44.

Parmentier, Richard J. (2002). “Money walks, people talk: systemic and transactional dimensions of Palauan exchange.” L’Homme 162: 49–79.

Peng Ke and Zhu Yanshi (1995). New Research on the Origins of Cowries Used in Ancient China. Sino-Platonic Papers no. 68. Philadelphia: Department of Asian and Middle Eastern Studies, University of Pennsylvania. The Chinese version is published in 1999 in Kaoguxue jikan 考古學集刊 12: 119–147.

Peng Xinwei 彭信威 (1965). Zhongguo Huobi Shi 中國貨幣史 (A Monetary History of China). Shanghai: Renmin Chubanshe.

Peng Xinwei 彭信威 (1994). A Monetary History of China, Volume 1. Belling-ham: Western Washington University. English translation of Peng Xinwei 1965 by Edward H. Kaplan.

Polanyi, Karl (1968a). “The semantics of money-uses.” In George Dalton (editor), Primitive, Archaic, and Modern Economies: Essays of Karl Polanyi, pp. 175–203. Boston: Beacon Press.

Polanyi, Karl (1968b). “Archaic economic institutions: cowrie money.” In George Dalton (editor), Primitive, Archaic, and Modern Economies: Essays of Karl Polanyi, pp. 280–305. Boston: Beacon Press.

Polo, Marco (1929). The Book of Ser Marco Polo the Venetian Concerning the Kingdoms and Marvels of the East. Edited and translated by Sir Henry Yule. London: J. Murray.

Pryor, Frederic L. (1977). The Origins of the Economy: A Comparative Study of Distribution in Primitive and Peasant Economies. New York: Academic Press.

Qinghai 1976: Qinghai Sheng Wenwu Guanlichu Kaogudui 青海省文物管理處考古隊 and Beijing Daxue Lishi Kaogu Zhuanye 北京大學歷史系考古專業 (1976). “Ledu Liuwan yuanshi shehui muzang di yici fajue de chubu shouhuo 青海樂都柳灣原始社會墓葬第一次發掘的初步收穫 (Preliminary results of the first season of excava-tions at the primitive society cemeteries at Liuwan, Ledu, Qinghai).” Wenwu 文物 1976.1: 67–78.

Qinghai 1984: Qinghai Sheng Wenwu Guanlichu Kaogudui 青海省文物管理處考古隊 and Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (1984).Qinghai Liuwan: Ledu Liuwan Yuanshi Shehui Mudi 青海柳灣: 樂都柳灣原始社會墓地 (Primitive society cemeteries at Liuwan, Qinghai). Beijing: Wenwu Chubanshe.


Quiggin, Alison Hingston (1949). A Survey of Primitive Money: The Begin-nings of Currency. London: Methuen & Co. Ltd.

Safer, Jane Fearer and Frances McLaughlin Gill (1982). Spirals from the Sea: An Anthropological Look at Shells. New York: Clarkson N. Potter, Inc.

Shanghai 1986: Shanghai Bowuguan 上海博物館 (1986). Shang Zhou qing-tongqi mingwen xuan 1 商周青銅器銘文選 1 (Bronze inscriptions of the Shang and Zhou periods, volume 1). Beijing: Wenwu Chubanshe.

Shaughnessy, Edward L. (1991). Sources of Western Zhou History: Inscribed Bronze Vessels. Berkeley: University of California Press.

Shaughnessy, Edward L. (1997). “Western Zhou bronze inscriptions.” In E. L. Shaughnessy (editor), New Sources of Early Chinese History: An Introduction to the Reading of Inscriptions and Manuscripts, pp. 57–84. Berkeley: The Society for the Study of Early China and the Institute of East Asian Studies, University of California, Berkeley.

Shaughnessy, Edward L. (1999). “Western Zhou history.” In M. Loewe and E. L. Shaughnessy (eds.), The Cambridge History of Ancient China: From the Origins of Civilization to 221 B.C., pp. 292–351. Cambridge: Cambridge University Press.

Shirakawa Shizuka 白川靜 (1980). “Zai lun mieli 再論蔑曆 (More thoughts on the phrase mieli).” Bulletin of the Institute of History and Philology, Academia Sinica 歷史語言研究所集刊, volume 51, part 2: 337–347.

Sun Zhichu 孫稚雛 (1981). “Jinwen zhulu jianmu 金文著錄簡目 (Con-cise catalogue of bronze inscriptions).” Beijing: Zhonghua Shuju.

Takeuchi Yasuhiro 竹內康浩 (1993). “Kyu Ei shoki meibun kōshaku 裘衛諸器銘文考釋 (Notes on the inscriptions of the four bronze vessels made by Qiu Wei).” Tōyō bunka kenkyūjō kiyō 東洋文化研究所紀要 120: 1–41.

Tang Lan 唐蘭 (1976a). “Shaanxi sheng Qishan xian Dongjia cun xinchu Xi Zhou zhongyao tongqi mingci de yiwen he zhushi 陝西省岐山縣董家村新出西周重要銅器銘辭的譯文和注釋 (Translations and annotations of the bronze inscriptions on some important West-ern Zhou bronzes newly unearthed at Dongjiacun, Qishan county, Shaanxi province).” Wenwu 文物 1976.5: 55–59, 63.

Tang Lan 唐蘭 (1976b). “Yong qingtongqi mingwen lai yanjiu Xi Zhou shi 用青銅器銘文來研究西周史 (Studying Western Zhou history via bronze inscriptions).” Wenwu 文物 1976.6: 31–39.

Tang Lan 唐蘭 (1979). “ ‘Mie li’ xin gu ‘蔑歷’ 新詁 (A new interpreta-tion of the term ‘mie li’). Wenwu 文物 1979.5: 36–42.

Vogel, Hans Ulrich (1993). “Cowrie trade and its role in the economy of Yünnan: from the ninth to the mid-seventeenth century.” Journal of the Economic and Social History of the Orient 36.2: 211–252; 36.3: 309–353.

24 LI YUNG-TI

von Glahn, Richard (1996). Fountain of Fortune: Money and Monetary Policy in China, 1000–1700. Berkeley: University of California Press.

Wan Xianchu 萬獻初 (2002). “Shuowen bei bu wushijiu zi tongkao—tui-lun Shangzhou shiqi de huobi xingtai, liutong shouduan ji qi ying-xiang 《說文》貝部五十九字通考—–推論商周時期的貨幣型態、流通手段及其影響 (Survey of 59 characters under the bei-radical section in the Shuowen dictionary: reconstructing forms, circulation, and influence of currencies of the Shang and Zhou periods).” In Feng Tianyu (editor), Renwen luncong 2001 人文論叢 2001 年卷, pp. 423–437. Wuhan: Wuhan Daxue Chubanshe.

Wang Dadao 王大道 and Zhao Zhensheng 趙震生 (1989). “Di er zhang: beibi 第二章: 貝幣 (Chapter two: shell money).” In Tang Guoyan 湯國彥, Lei Jiaming 雷加明, and Luo Bingjun 駱秉鈞 (editors), Yun-nan Lishi Huobi 雲南歷史貨幣 (Historical currencies of Yunnan), pp. 11–15. Kunming: Yunnan Renmin Chubanshe.

Wang Guowei 王國維 (1959). “Shuo yu peng 說玨朋 (Interpreting yu and peng).” In Guantang Jilin 觀堂集林 (Collected essays of Wang Guowei), Volume 3, pp. 161–163. Beijing: Zhonghua Shuju.

Wang Qingzheng 汪慶正 (ed.) (1988). Zhongguo lidai huobi daxi (1): Xian Qin huobi 中國歷代貨幣大系 (1): 先秦貨幣 (Catalogue of Chinese currencies, volume 1: Pre-Qin periods). Shanghai: Shanghai Renmin Chubanshe.

Wang, Yü-chüan (1951). Early Chinese Coinage. New York: The American Numismatic Society.

Wang Wenchang 王文昶 (1983). “Tong you bianwei 銅卣辨偽 (Authen-ticating bronze you vessels).” Gugong Bowuyuan yuankan 故宮博物院院刊 1983.2: 49–52, 59.

Wang Xiantang 王獻唐 (1979). Zhongguo gudai huobi tongkao 中國古代貨幣通考 (Comprehensive study of ancient Chinese currencies). Ji’nan: Qilu Shushe.

Watson, William (1973). “On Some Categories of Archaism in Chinese Bronze.” Ars Orientalis 1973 (IX):1–13.

Wei Si 衛斯 (1998). “Shilun beibi de zhineng yu Yin Shang shiqi de shangpin jingji 試論貝幣的職能與殷商時期的商品經濟 (The function of cowrie money and the commodity economy of the Yin Shang Period).” In Wei Si kaogu lunwen ji 衛斯考古論文集 (Collected essays on archaeology by Wei Si), pp. 106–110. Taiyuan: Shanxi Guji Chubanshe.

Wu Danmin 吳旦敏 (2002). “Chutu bei xianxiang fenxi yanjiu 出土貝現象分析研究 (Analysis and Study of Excavated Cowry).” Shanghai Bowuguan jikan 上海博物館集刊 9: 102–133.

Xiao Qing 蕭清 (1984). Zhongguo gudai huobi shi 中國古代貨幣史 (History of currencies in ancient China). Beijing: Renmin Chubanshe.

Xie Zhaozhi 謝肇淛 (2000). Wu za zu 五雜組. Ming dynasty edition in


Siku Jinhuishu Congkan Bianzuan Weiyuanhui 四庫禁燬書叢刊編纂委員會 (editor), Siku Jinhuishu Congkan, Zibu 37 四庫禁燬書叢刊子部 37. Beijing: Beijing Chubanshe.

Yang, Lien-sheng (1952). Money and Credit in China: a Short History. Cam-bridge, Mass.: Harvard University Press.

Yang Shengnan 楊升南 (1992). Shangdai jing ji shi 商代經濟史 (Eco-nomic history of the Shang dynasty). Guizhou: Guizhou Renmin Chubanshe.

Yang Shouchuan 楊壽川 (2000). “Beibi yanjiu—Zhongyuan yu Yunnan yong haibei zuo huobi de lishi kaocha 貝幣研究—–中原與雲南用海貝作貨幣的歷史考察 (The study of cowrie money: a historical review of the use of cowrie as money in Zhongyuan and Yunnan).” Gugong xueshu jikan 故宮學術季刊 17 (3): 35–55.

Yin Zhou Jinwen (1984): Zhongguo Shehui Kexueyuan Kaogu Yanjiu-suo (1984). Yinzhou jinwen jicheng 殷周金文集成 (Corpus of Yin and Zhou bronze inscriptions). 18 volumes. Beijing: Zhonghua Shuju.

Zhang, Shanxi 張善熙 and Chen Xiandan 陳顯丹 (1989). “Sichuan Sanxingdui wenhua de beibi shitan 四川三星堆文化的貝幣試探 (Preliminary study of cowrie money of the Sanxingdui culture in Sichuan).” Zhongguo qianbi 中國錢幣 1989.3: 48, 31.

Zhao Guangxian 趙光賢 (1979). “Cong qiu wei zhu qi ming kan Xi Zhou de tudi jiaoyi 從裘衛諸器銘看西周的土地交易 (Western Zhou land exchange seen in the inscriptions of the Qiu Wei vessels).” Beijing shifan daxue xuebao (shehui kexue ban) 北京師範大學學報（社會科學版）1979.6: 16–23.

Zhao Ningfu 趙寧夫 and Ji Jin 汲津 (1984). “Shilun jianguo yilai Henan chutu qianbi de xueshu jiazhi 試論建國以來河南出土錢幣的學術價值 (The academic value of coins unearthed in Henan since the founding of P.R.C.).” Zhongyuan wenwu 中原文物 1984.2: 99–103, 109.

Zhao Shande 趙善德 (1988). “Shangzhou shidai de ‘beihuo’ 商周時代的 ‘貝貨’ (The ‘cowrie money’ of the Shang and Zhou periods).” Wenbo 文博 1988.1: 80–83.

Zheng Jiaxiang 鄭家相 (1959). “Gudai de bei hua 古代的貝化 (Ancient cowrie money).” Wenwu 文物 1959.3: 65–66.

Zhongguo Qianbi 2001: Zhongguo Qianbixuehui Huobishi Weiyuan-hui 中國錢幣學會貨幣史委員會 (2001). “Huobi qiuyuan wenti zuo-tanhui jiyao 貨幣起源問題座談會紀要 (Minutes of the symposium on the origin of money).” Zhongguo qianbi 2001.4: 30–36.

Zhongguo Qingtongqi 1996: Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅器全集編輯委員會 (1996). Zhongguo qing-tongqi quanji 5: Xi Zhou (1) 中國青銅器全集 5: 西周 (1) (Corpus of Chinese bronzes, volume 5: Western Zhou (1)). Beijing: Wenwu Chubanshe.

26 LI YUNG-TI

Zhongguo Qingtongqi 1997: Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅器全集編輯委員會 (1997). Zhongguo qingtongqi quanji 3: Shang (3) 中國青銅器全集 3 : 商 (3) (Corpus of Chinese bronzes, volume 3: Shang (3)). Beijing: Wenwu Chubanshe.

Zhongguo Qingtongqi 1998: Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅器全集編輯委員會 (1998). Zhongguo qingtong-qi quanji 4: Shang (4) 中國青銅器全集 4 : 商 (4) (Corpus of Chinese bronzes, volume 4: Shang (4)). Beijing: Wenwu Chubanshe.

Zhu Huo 朱活 (1983). “Gu bi san tan—tan wo guo xian Qin huobi de gui bei, zhu yu, jin yin 古幣三談—–談我國先秦貨幣的龜貝、珠玉、金銀 (Three essays on ancient money: turtle shells, cowries, beads and jades, and gold and silver in Pre-Qin China). Zhongguo qianbi 中國錢幣 1983.1:10–17.

Zhu Huo 朱活 (1984). Gu qian xintan 古錢新探 (New studies on ancient money). Ji’nan: Qilu Chubanshe.

Address:

Institute of History and PhilologyAcademia SinicaTaipei, Taiwan [email protected]


CHIN-STRAPS OF THE EARLY NORTHERN WEI: NEW PERSPECTIVES ON THE

TRANS-ASIATIC DIFFUSION OF FUNERARY PRACTICES

BY

SHING MÜLLER(Ludwig-Maximilans-Universität, München)

AbstractBased on an excavated bronze chin-strap from the Northern Wei tomb M107 at the cemetery south of Datong, the author examines the custom of using such a metal device to hold the jaw of the deceased in China. The starting point of the custom in the Asian part of the Eurasian continent was probably the Tarim Basin in the 8th century BC, but there are significant earlier parallels in West Asia and the Mediterranean world. The custom was brought by tribal members of the Tuoba-Xianbei no later than the 5th century AD to the Northern Wei capital, Pingcheng (modern Datong), where the use of chin-straps was first evidenced in China proper. At that time, the use of metal chin-straps was apparently restricted to certain non-Chinese members of Northern Wei society. There is no archaeological or written evidence that the Chinese applied such a device onto the face of the deceased before the Tang. In Tang times some Chinese up-per class members also adopted the custom of using metal chin-straps. After the Tang, the users all seem to have been Han Chinese, and by then metal chin-straps were only found in southern China, where the custom eventually died out.

Keywords: chin-straps, face-covers, Northern Wei, Pingcheng period, Tarim-Basin, Tuoba

Introduction

The archaeological study of transcontinental cultural transfer between the Mediterranean and East Asia in ancient and medieval times presents great challenges to scholarship. One promising avenue of investigation consists of tracing individual cultural traits that are specific in nature and whose occurrence in different places can be dated with some precision. The present article attempts to do this for a hitherto little noticed material component of early burial practices: chin-straps that served to prevent the jaw of the deceased from opening in the course of decomposition.

28 SHING MÜLLER

The point of departure for this discussion is 167 earthen tombs that were excavated in a cemetery near the Datong Electric Welding Equipment Works south of Datong 大同 City (hereafter referred to as Datong 1988 cemetery), Shanxi Province (Shanxi 1992). Based on the typological sequence of the burial ceramics, the excavators —Mr. Wang Yintian 王銀田 and his colleagues —concluded that the cemetery had been continuously occupied during the entire Pingcheng 平城 period (AD 398–494), named after the early capital of the Northern Wei 北魏 Dynasty (AD 386–534) near modern Datong. One of the most unusual findings among these highly interesting but poorly published materials was the use of bronze chin-straps (xiahe tuo 下頜托; lit. “jaw-support”). Altogether 15 specimens, all similar to one another, were found in 14 burials of children and adults of both sexes (one was so badly corroded that it was not collected) (Wang and Wang 1999: 160). To date, these are the oldest metal chin-straps known in China. Metal chin-straps, some made of pure gold, had in fact already been sighted in burials from earlier excavations in other regions, but these caused little notice. Before their archaeological context at Datong clarified their true purpose beyond any doubt, a variety of now obso-lete interpretations had been proposed: “torture instruments” (Zhang 1957: 62), medical equipment for curing some chin disease or ailment, or ornaments worn in life and taken to the grave after death (Michael-son 1999: 74). Some collected specimens were interpreted as funerary adornments or headdresses (see below). Since most of the chin-strap finds, when reported at all, have been inadequately described, and the data of the Datong 1988 cemetery have not yet been fully published, it is impossible to discuss any cultic or religious associations of chin-straps at this stage. Until now, few efforts have been made to trace the origin and the tradition of the usage of chin-straps. The problem has been tackled by Benko (1992/1993), Lerner (Juliano and Lerner 2001: 264–66), and Luo Feng (1998: 103–05). But there seems to be a general confusion of chin-straps with other kinds of face-covers. Based on the finds from the Datong 1988 cemetery, this article will survey the forms and the development of chin-straps according to recent finds in China, and place them in their wider context through comparison with specimens from Central Asia and areas further west. The analysis will show that chin-straps are alien to the indigenous burial traditions of mainland East Asia, but were adopted there for a time. Their history as part of Chinese material culture may be seen as symptomatic, at least in some respects, for the complexities of cultural contact and cultural transfer along the ancient Silk Roads.

CHIN-STRAPS OF THE EARLY NORTHERN WEI 29

The Construction of Early Metal Chin-straps in China based on the Specimens from Tomb No. 107 at the Datong Cemetery

The construction of the oldest metal chin-straps found in China can be clearly demonstrated by the specimens from tomb no. 107 from the Datong 1988 cemetery. The tomb is dated to the period before the reign of the Emperor Xiaowen 孝文 (AD 477–499) by Wang and Wang (1999: 160) based on its grave form and the typology of the funerary ceramics. The material of the chin-strap from M107 is either copper, bronze, or brass; the short report refers only to tong 銅, but gives no metallur-gical analysis for the copper alloy. In the following, the general term “bronze” will be adopted for this and other finds that were also given only a tong designation, but note that “bronze” is used only as a matter of convenience—a definitive description of tong must await proper analysis. The chin-strap, found disjointed near the skull, consists of three major components (Wang and Wang 1999: 156; Fig. 1):

1) a bowl-like jaw-guard with two elongated ends (L. 13 cm; W. 9 cm; H. 3.5 cm) (Fig. 1c);

2) two cheek-girdles, each consisting of a medallion (D. 1.8 cm) and a V-shaped strap (W. 1 cm; L. unknown) (Fig. 1b);

3) a headband (L. 64 cm; W. 1.4 cm) with a buckle at the end (Fig. 1a).

The elongated ends of the jaw-guard are attached to the medallions by means of “hinges” (heye 合頁; Fig. 1e). The cheek-girdles are inserted into the headband through two horizontal slots on each side (Fig. 1f). The jaw-guard has a plain surface and, according to the excava-tors, 15 small holes along its rims. Remains of silk fabric and silk floss were detected inside and outside of the jaw-guard. The outer silk lin-ing fragments show T-formed or cross-formed ornament bands with embroidered medallions (Fig. 1d), indicating that the bronze chin-strap was originally carefully wrapped in fine decorated silk and padded. The report does not mention the color of the silk, the type of the stitches, or the precise motifs of the embroidery. Altogether 47 objects, located either close to the head or to the feet of the deceased, were recovered from this unrobbed earthen chamber tomb with a short ramp (Fig. 2).1 In addition to seven ceramic vessels (Fig. 3.3–5, 18, 20–21), the following objects were found: a partially gilded silver bowl decorated with medallions with busts in profile and

1 The descriptions of tomb no. 107 in Shanxi 1992 differ considerably from those in Wang and Wang 1999. The following account of tomb no. 107 is based on the more detailed report of Wang and Wang 1999.

30 SHING MÜLLER

acanthus leaves (D. 10.2 cm; H. 4.6 cm) on the south wall (Fig. 2, no. 16; Fig. 3.16); an undamaged transparent glass bowl in Sasanian style with cut oval facets in slightly yellowish green color (D. 10.3 cm; H. 7.5 cm), located close to the head of the coffin (Fig. 2, no. 17; Fig. 3.17); and an unusual small silver jug (H. 14.4 cm) with a short straight neck, and soldered in the middle of the belly (Fig. 2, no. 1; Fig. 3.1). The fragment of a presumed golden earring (D. 1.5–1.6 cm; Fig. 3.11), a pair of badly corroded iron scissors, and a small iron mirror (D. 11.26 cm), originally wrapped in silk, lay close to the head (both unillustrated). These last

Figure 1. The chin-strap found in M107; from Wang and Wang 1999.1: 156, fig. 15; a. headband; b. cheek-girdle; c. jaw-guard; d. embroidery fragment; e. hinge;

f. headband slots for cheek-girdles.


two items, and the lack of a hair pin as a symbol of maturity, might have led the excavators to the conclusion that the deceased could have been a young girl. The scarce remains of the skeleton, however, were not anthropologically examined. The richness of the tomb inventory, especially the deposition of the precious silver and glass vessels, indicates that an upper class member was buried here.

Parallels to the Chin-Straps from Tomb No. 107

A golden apparatus, described as a “headdress” (Kopfschmuck), but simi-lar to the chin-straps from tomb no. 107, appears in the Pierre Uldry collection (Fig. 4). It is decorated with ornaments of vine scrolls, lions and chimeras in repoussé, and has been dated to the first half of the sixth century AD according to the catalogue Chinesisches Gold und Silber of the Museum Rietberg of Zürich (Uldry 1994: cat. no. 121), and to the second half of the fifth to the early sixth century according to Louis (1994: 91–93). This apparatus measures 10.3 cm in length, and 6.9 cm in width, and the two cheek strips are 22.0 cm long. Along the rim of the “cap” there are over 60 tiny holes, indicating that this part was sewn onto fabric. But since the outside is beautifully decorated, it is likely that only the inside was lined. Each cheek-girdle splits from a

Figure 2. Ground plan and distribution of the grave goods of M107; from Datong 1992.8: 3 (above), and Wang and Wang 1999.1: 144 (below).

32 SHING MÜLLER

2 I am indebted to Dr. François Louis, the scientific consultant of the exhibition, for kindly providing me with the detailed information of the chin-strap from Pierre Uldry’s collection, since no medallions can be recognized from the photo of the Rietberg Mu-seum catalogue.

medallion-like circular section and thus forms a V-shape on each side.2 Hinges connecting the cheek-girdles and medallions are not clearly rec-ognizable on the photo. A headband was not included in the purchase of this set. While the measurements of the “cap” are definitely too small

Figure 3. Other artifacts found in M107: 1: silver jar; 3–5, 18, 20–21: clay jars; 9: bronze tray; 10: bronze pectoral; 11: golden earring; 16: silver bowl; 17: glass bowl; 19: clay basin; 22: stone lamp. (The drawings do not have the same scale. The numbers

correspond to the grave good numbers in Fig. 2.)


Figure 4. “Headdress” of the Pierre Uldry collection, courtesy of Museum Rietberg in Zürich; the “cap” L. 10.3 cm, W. 6.9 cm; the cheek-girdles: L. 22 cm.

for a headdress, even for a child, the device possesses two of the basic components of the chin-strap from tomb no. 107: the “jaw-guard” and the cheek-girdles. Also, the hammered ornamental bands with scrolls on the jaw-guard recall the embroidery bands of the set from tomb no. 107. The device can therefore be identified with great certainty as a chin-strap. The length of the cheek-girdles makes it reasonable to sug-gest that originally a headband was needed in order to fix the jaw-guard properly. Another magnificent piece, attributed to the Old Turkic (Tujue) culture of the sixth to the eighth centuries, from a private collection in Belgium, has been described as “funerary adornment” (parure funéraire). According to the description, it consists of a “cap,” several bands, a pectoral, a belt, and several other parts (Desroches 2000: 165–66, no. 153). Most of the components are covered with motifs of mythical figures and animals and scrolls of half-palmette, some even with pasted glass and/or precious stones (Fig. 5a). Although no measurements are provided, the essential components of a chin-strap of the type from tomb no. 107 such as a bowl-shaped jaw-guard, i.e. the “cap” (Fig. 5b), two V-shaped straps con-nected to medallions (Fig. 5c), and a headband (Fig. 5a) are all present. The jaw-guard has perforations on its rims, indicating that this part, too, was lined with fabrics. Each elongated end of the guard is joined to a short strip leading to a medallion through a hinged fastening (Fig. 5b,

34 SHING MÜLLER

Figure 5. Golden “parure funéraire” of the private collection in Belgium; after Desroches 2000: cat. no. 153; a. the whole set as exhibited in Paris; b. jaw-guard (upside down, note the hinged fastening); c. cheek-girdle (also with a hinged fastening); d. pectoral;

e. plaque.

ba

c

d

e


5c), which functions exactly the same way as the bronze chin-strap from tomb no. 107 described above. François Louis (1994) pointed out that the ornaments of the set in the Uldry collection consist of typical motifs unique in the repertoire of the early Yungang Grottoes. This argument can also be applied to the style of the ornaments of the Belgian private collection. The ornamen-tal style thus indicates that these two pieces without provenance are of the early Northern Wei, that is, the Pingcheng period. The structural similarity with the set from tomb no. 107 strengthens such a presump-tion of a Pingcheng period origin. Another indication is provided by the rectangular golden plaque (Fig. 5e) accompanying the chin-strap of the Belgian collection. Though unique for its kind, the plaque displays a rider in unmistakable Xianbei 鮮卑 garb typical of the Pingcheng period (Dien 1991: 44–46). Seen in this light, not only should the al-leged purpose of the devices in the Belgian and Zürich collections be reappraised, but the dates should probably also be pushed backwards to the last quarter of the fifth century, when the Yungang style became prevalent in the art of the North.

Other Evidence for Metal Chin-Straps in China

From the Luoyang 洛陽 Period (AD 495–534) of the Northern Wei Dynasty there are only two metal chin-straps documented through excavations. Neither drawings nor photos of these have been published. One example comes from a medium-sized unlooted brick tomb in Xi’an 西安, housing the body of the 99 year-old Shao Zhen 邵真 (died AD 520). This piece, found with three teeth in it, is described as “[being composed of] several silver strips which wrapped the jaw and were fastened at the top of the skull” with “[the strips] under the jaw form-ing the shape of [the bowl of] a spoon” (Zou 1955: 62). The other find was reported from a severely looted tomb YDIIM926 at Xingyuancun 杏園村, in the vicinity of Yanshi 偃師, Henan (Wang 1991: 828–30). The silver chin-strap was still attached to the skull when the tomb was opened, but was too corroded to be collected. The sex and age of the dead were not determined. However, the brick construction of the tomb and a rare gilded bronze jar of the type tuohu 唾壺 with a lid indicate that the deceased belonged to the wealthy upper class during the last years of the Northern Wei. Without providing any sources, Zou Jingbi (1955: 62) claimed that several metal chin-straps were found in tombs dating from the North-ern Dynasties 北朝 (AD 439–581) and Tang 唐 (AD 618–907) periods in Xi’an and vicinity. If this information is reliable, it can be assumed

36 SHING MÜLLER

that metal chin-straps were continuously in use in burials at least in the Xi’an region after the Northern Wei and did not disappear completely before the Tang. For the Tang era, the use of metal chin-straps is better documented than the previous periods. Wang and Wang (1999) made a list of finds and paved the first step for a further investigation of this unusual fu-nerary apparatus. Their results, which I have supplemented with some additional finds, are shown in Table 1. As the table shows, the finds from the Northern Wei (nos. 1–4) were made at both of the successive capitals, Pingcheng and Luoyang. Simi-larly, five (nos. 5, 9, 10, 14, 15) of the eleven Tang tombs (nos. 5–15) with chin-strap finds were located in the vicinity of modern Xi’an close to the then capital of Tang. The deceased associated with nos. 7 and 8 were the wife and son of Li Tai 李泰, the third son of Taizong 太宗 of Tang and the author of the geographical work Kuodizhi 括地志. Excluded from the line of succession, Tai was sent to Yun Xian 鄖縣 in modern Hubei, where he spent the rest of his life. His son, Li Hui 李徽 (no. 7), died and was buried there with his father; the wife of Li Tai, Yan Wan 閻婉 (no. 8), was first buried near Luoyang and then conveyed to the tomb of her husband in Hubei. Zheng Xun 鄭洵 (no. 12), otherwise unknown, belonged to the eminent Zheng family of Xingyang 滎陽 (modern Zhengzhou 鄭州 in Henan) according to his epitaph. He was banned to Baling 巴陵 where he died. Thus, the findings in Yun Xian (Hubei) and Baling (Hunan) actually represent a sepulchral custom transplanted from the Central Plain. Only one find (no. 6) was located at the northwestern frontier. To summarize, the geographical distribution of the chin-straps in pre-Tang and Tang times was confined to north and northwestern China, if not only to the capitals. No excavation reports for nos. 14 and 15 are available, and thus no precise dates can be attributed to these finds. But seven out of the nine datable Tang specimens were made before the end of the Tianbao 天寶 era (AD 742–756). Thus, the Middle Tang seems to be a turning point for the use of chin-straps in China. All the known metal chin-straps from Xi’an and its vicinity (nos. 5, 9–10, 14–15)—regardless of their material—are of simple construc-tion. They run from the lower jaws to the skull in one single strip that may split up into two or more strings towards the end (Fig. 6, 8–12). It is unclear whether they were wrapped with fabrics: the golden Tang chin-strap from Wangsan3 near Xi’an–as shown in the catalogue Gilded Dragons (Michaelson 1999: cat. no. 41; Fig. 12)–is not perforated on the rim of the bowl, indicating that this one was probably fitted onto the

3 No excavation report of Wangsan is thus far available.

Tab

le 1

: Fin

ds o

f Met

al C

hin-

Stra

ps in

Chi

na

No.

Dat

eSi

teSe

x an

d ag

eTo

mb

Occ

upan

t(s)

Mat

eria

lC

ompo

nent

sSo

urce

sFi

g. no

.

Nort

hern

Wei

1a

c. 4

70D

aton

g, S

hanx

iG

irl

tom

b no

. 107

bron

ze

jaw

gua

rd, c

heek

gir

dles

with

m

edal

lions

and

hea

dban

dSh

anxi

199

2: 3

; W

ang

and

Wan

g (1

999)

: 156

1

1b

c. 4

70–

494

Dat

ong,

Sha

nxi

fem

ale,

adu

ltto

mb

no. 1

09br

onze

ja

w g

uard

, che

ek g

irdl

es w

ith

med

allio

ns a

nd h

eadb

and

Shan

xi 1

992:

2-

1c

398–

494

Dat

ong,

Sha

nxi

unkn

own

from

the

sam

e ce

met

ery

as 1

a an

d 1b

bron

ze;1

2 se

tsno

des

crip

tions

; pro

babl

y si

mila

r to

1a

and

1bW

ang

and

Wan

g (1

999)

: 10

249

4–53

4X

i’an,

Sha

anxi

??

lead

?

Zha

ng 1

957:

62

- 3

520

Xi’a

n, S

haan

xim

ale

(99

sui)

Shao

Zhe

n 邵

真, G

over

nor

of

And

ing 安

定太

守

silv

er

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece

Zou

195

5: 6

2-

449

4–53

4Ya

nshi

偃師

, Hen

an?

YD

IIM

926

silv

er

no d

escr

iptio

nsW

ang

1991

: 830

-

Tang

5di

ed 6

21X

iany

ang 咸

陽,

Shaa

nxi

fem

ale

(53

sui)

Lad

y H

eruo

賀若

gold

ja

w g

uard

and

che

ek b

ands

in

one

piec

e: c

heek

ban

ds s

plit

into

sev

eral

str

ips

at th

e en

ds,

jaw

gua

rd w

ith a

n op

enin

g in

th

e m

iddl

e

Yun

1993

: 50;

KG

YWW

20

00.4

, fro

nt

cove

r, in

side

6

6di

ed 6

78G

uyua

n 固

原, N

ingx

iam

ale

(66

sui)

Shi D

aode

史道

德go

ld

jaw

gua

rd w

ith a

n op

enin

g in

th

e m

iddl

e; c

heek

gir

dles

an

d he

adba

nd w

ere

rive

ted

toge

ther

Luo

199

8: 8

27

Tab

le 1

(con

t.): F

inds

of M

etal

Chi

n-St

raps

in C

hina

No.

Dat

eSi

teSe

x an

d ag

eTo

mb

Occ

upan

t(s)

Mat

eria

lC

ompo

nent

sSo

urce

sFi

g. no

.

7di

ed 6

83

Yun

Xia

n 鄖

縣, H

ubei

; bu

ried

toge

ther

with

hi

s fa

ther

Li T

ai

mal

e (4

0 su

i)L

i Hui

李徽

, sec

ond

son

of

Li T

ai 李

泰 a

nd Y

an W

an

(no.

8);

Li T

ai =

thir

d so

n of

T

aizo

ng

not m

entio

ned

no d

escr

iptio

n, b

ut w

as il

lust

rat-

ed in

the

floor

pla

n; p

roba

bly

sim

ilar

to n

o. 8

Hub

ei 1

987:

fig.

1,

no. 5

0-

8di

ed 6

90

first

bur

ied

nort

h of

L

ongm

en 龍

門 o

n R

iver

Luo

洛川

; in

724

conv

eyed

to h

er

husb

and’

s to

mb

in

Yunx

ian

fem

ale

(69

sui)

Yan

Wan

閻婉

, dau

ghte

r of

Ya

n L

ide 閻

立德

and

wife

of

Li T

ai

silv

er

a st

rip

wid

ened

at t

he ja

w; t

he

ends

of t

he c

heek

str

aps

stuc

k un

der

the

head

dres

s

Hub

ei 1

987:

fig

.16.

18

9di

ed 7

24su

burb

of X

i’an,

Sh

aanx

i; a

doub

le

buri

al, b

ut th

e gr

ave

good

s al

l bel

onge

d to

th

e pr

ince

ss

fem

ale

(71

sui)

Prin

cess

Jin

xian

g 金

鄉縣

主,

gran

ddau

ghte

r of

Gao

zu

of T

ang,

Li Y

uan 李

淵,

daug

hter

of L

i Yua

nyin

李元

嬰, a

nd w

ife o

f Yu

Yin

于隱

(d

ied

in 6

90)

bron

ze

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece:

the

ends

of t

he

stri

p do

not

see

m to

spl

it

Xi’a

n 19

97: 1

4,

fig.3

49

10di

ed 7

45H

anse

nzha

i 韓森

寨 in

X

i’an,

Sha

anxi

fem

ale

wife

of P

alac

e A

tten

dant

Lei

內侍

雷府

君夫

人, n

ée S

ong

宋氏

gold

ja

w g

uard

and

che

ek b

ands

in

one

piec

e: c

heek

ban

ds s

plit

into

2 s

trip

s at

the

ends

Zha

ng 1

957:

60

10

11di

ed 7

50L

uoya

ng, H

enan

fem

ale

Lad

y né

e L

u 盧

氏br

onze

no

des

crip

tion;

but

mos

t lik

ely

sim

ilar

to n

o.12

acc

ordi

ng to

th

e flo

or p

lan

of th

e to

mb

Luo

yang

198

0:

382,

fig.

111

12di

ed 7

68

(m)

776

(f) Z

heng

firs

t bur

ied

at

Bal

ing 巴

陵, H

unan

; W

ang

first

bur

ied

at W

ansh

an 萬

山,

Hen

an; i

n 77

8 bo

th

conv

eyed

to Y

ansh

i, H

enan

mal

e (5

3 su

i)

and

fem

ale

(age

un-

know

n)

Zhe

ng X

un 鄭

洵 a

nd L

ady

née

Wan

g 王

氏

2 ch

in-s

trap

s:

bron

ze,

ham

mer

ed

chin

-str

ap fo

r Z

heng

: jaw

gua

rd

and

chee

k ba

nds

in o

ne p

iece

: th

e en

ds o

f the

che

ek b

ands

st

uck

unde

r th

e ca

pe;

chin

-str

ap fo

r W

ang:

no

desc

rip-

tion

Xu

1996

: 11

-

Tab

le 1

(con

t.): F

inds

of M

etal

Chi

n-St

raps

in C

hina

No.

Dat

eSi

teSe

x an

d ag

eTo

mb

Occ

upan

t(s)

Mat

eria

lC

ompo

nent

sSo

urce

sFi

g. no

.

13E

nd o

f T

ang

Hua

ngdi

gang

皇帝

崗,

Gua

ngzh

oufe

mal

e (a

ge

unkn

own)

?gi

lded

silv

ersi

mila

r to

no.

10

Gua

ngzh

ou 1

959:

66

8, fi

g.1

(no.

4)

-

14T

ang

Wan

gsan

vill

age

in e

ast-

ern

subu

rbs

of X

i’an

??

gold

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece:

wid

ened

to a

bow

l sh

ape

at th

e ja

w

Mic

hael

son

1999

: ca

t. no

. 41

12

15T

ang

Cha

ng’a

n co

unty

, Sh

aanx

i?

?br

onze

sim

ilar

to 1

4M

icha

elso

n 19

99:

cat.

no. 4

1; 7

4-

Shig

uo

16di

ed 9

18C

heng

du 成

都m

ale

(72

sui)

Wan

g Ji

an 王

建, K

ing

of th

e Fo

rmer

Shu

前蜀

silv

er

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece:

wid

ened

at t

he ja

w Fe

ng 1

964:

33,

pl.

37.5

13

Song

17E

nd o

f N.

Song

Lei

yang

耒陽

, Hun

an?

M60

silv

er

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece:

the

chee

k ba

nds

hook

ed o

nto

the

ears

Hen

gyan

g 19

96:

267,

fig.

34.

1214

18E

nd o

f N.

Song

Lei

yang

耒陽

, Hun

an?

M27

7si

lver

si

mila

r to

no.

16

Hen

gyan

g 19

96:

259,

fig.

22

-

19E

nd o

f N.

Song

Hen

gyan

g 衡

陽, H

unan

mal

e?

silv

er

jaw

gua

rd a

nd c

heek

ban

ds in

on

e pi

ece:

the

chee

k ba

nds

hook

ed o

nto

the

ears

Che

n an

d Fe

ng

1984

: 74,

fig.

315

20di

ed 1

274

De’

an 德

安, J

iang

xife

mal

eL

ady

née

Zho

u 周

氏si

lver

ja

w g

uard

and

che

ek b

ands

in

one

piec

e Ji

angx

i 199

0: 1

1-

Figure 6. Chin-strap of Lady Heruo found in Xianyang; from Kaogu yu wenwu 2000.4, front cover, inside;

no measurements available.

Figure 7. Chin-strap of Shi Daode; from Luo Feng 1998: 90, fig. 65.

Figure 8. Silver chin-strap of Yan Wan (right) and parts of her bronze diadem with glass beads found in Yun Xian, Hubei (left). Maximum W. 5.2 cm; from Hubei 1987:

38, figs. 15, 16.1.

Figure 9. Bronze chin-strap (right) and diadem of gilded bronze, glass, medallions with gold filigree works (left) of Princess Jinxiang from the suburb of Xi’an, Shaanxi. No

measurements available, from Xi’an 1997: 14, 15, figs. 34, 35.


Figure 10. Chin-strap of Madam Song from Hansenzhai, after

Zhang 1957: 60.

Figure 11. Bronze chin-strap (no. 12) of tomb 70IIIM109 in Guanlin near Luoyang, Henan (no detailed photo available). The remaining L. 46 cm;

from Luoyang 1980: 382, fig. 1.

Figure 12. Golden chin-strap of the Tang dynasty from the Wangsan village, eastern suburbs of Xi’an; excavated 1979; H. 9 cm; W. 5.4 cm; total L. 66 cm; after Michaelson

1999: 73, cat. no. 41.

42 SHING MÜLLER

Figure 13. Silver chin-strap of Wang Jian from Chengdu, Sichuan. No measurements avail-

able; from Feng 1964, pl. 37.5.

Figure 14. Silver chin-strap of M60 from Chengguan near Leiyang, Hunan; the bowl L. 8.5 cm, W. 3.1 cm, 1.4 cm deep; the cheek-girdle L. 16 cm. The ends of the cheek-girdles form into a loop and were fixed onto the ear shells; from Hengyang 1996:

267, fig. 34.12.

Figure 15. Silver chin-strap from a Northern Song tomb in Hengyang. L. 39 cm, max. W. 5

cm; after Chen and Feng 1984: 74, fig. 3.

face of the deceased without padding. This whole group of Xi’an finds is structurally homogenous with the Northern Wei chin-strap from the tomb of Shao Zhen 邵真. According to their structural characteristics, three major types of metal chin-straps can be categorized, as I have summarized in Table 2.


The chin-straps from the Datong 1988 cemetery of the late Pingcheng Period (Type I) have hinged cheek-girdles and are padded, and need a headband to fasten the jaw-guard, whereas the Xi’an specimens (Type II) are more simplified, with the jaw-guard and cheek-girdles fused to-gether. The cheek-girdles would be bound together on top of the head, resulting in the disappearance of the headband. No padding seems to

Table 2 Types of Metal Chin-Straps found in China

��

N. Wei

Tang

Song

No. 5 (621)

No. 9 (724)

No. 10 (745)

No. 14

No. 16 (918)

No. 1a (c. 470)

No. 8 (690)

No. 17

No. 19

(chin-strap of Shan Zhen, died 520; no illustration)

44 SHING MÜLLER

exist anymore. This type persisted for a long time; by the end of the Tang, chin-straps of this type (see Table 1, nos. 13 and 16) were found only in the South. There is an indication of the emergence of a third group (Type III) as early as the end of the seventh century. These were found in the Henan area, such as the examples of Lady Yan Wan (no. 8) and of Zheng Xun and his consort née Wang (no. 12). The chin-straps were even more structurally degenerated: cheek-girdles became so short that their ends reached only the ears and could merely be fixed onto a headgear. Most of the post-Tang specimens seem to fall under this type (nos. 17–20), which has been found, like those of Type II, only in south China according to present evidence. The Song silver chin-straps from Leiyang 耒陽 and Hengyang 衡陽, Hunan (Table 1, nos. 17, 19) for example, have short cheek bands ending with loops which were to be hooked onto the ears. With this construction, they obviously could not carry out the primary function of keeping the mouth tightly closed and were used only symbolically or decoratively. The golden chin-strap of Shi Daode 史道德 (d. AD 678) (no. 6) is reminiscent of Type I from Datong. It consists of a jaw-guard with an opening in the middle (L. 8.4 cm; W. 4.5 cm), two cheek bands (L. 17.3 cm; W. 1.4 cm) linking to the jaw-guard, and a fragment of a headband (remaining L. 5 cm; W. 1.2 cm). Each cheek band consists of three seg-ments, which are connected by four petal ornaments (Luo 1998: 90–91; Juliano and Lerner 2001: 264–66, no. 90). Luo Feng (1998: 103–04) argues that the golden covers for eyes, eyebrows, nose and lips, together with the chin-strap pieces, constitute a face-cover set for the dead, in keeping with an age-old Chinese tradition of using jade pieces in the shape of eyes, brows, nose, lips, and ears. Face masks in a different style and material are seen in the tombs of the highest nobles of the Jin 晉 state from the late Western Zhou to the Chunqiu period, and of the Guo 虢 state of the Eastern Zhou period (for an example of such a mask, see Rawson 1996: 122). A detailed examination of the find from the tomb of Shi Daode reveals that the golden covers for eyes, eyebrows, nose and lips indeed possess small holes on the edges or fringes, indicating that they were sewn onto a fabric covering the face. The chin-strap, on the other side, does not feature any perforations. Only the end of each cheek band has a small hole for the attachment onto the headband by means of a rivet, suggesting that the chin-strap, unlike the golden facial features, was not sewn onto a piece of fabric, and that both were merely used together, but did not necessarily constitute a functional set. Despite its resemblance with Type I, the chin-strap of Shi Daode was subject to the overall development during the Tang, as indicated by the absence of any perforations on the rims of the jaw-guard, which means that the jaw-guard was not padded.


Chin-Straps before the Northern Wei

Metal chin-straps dating from earlier than the Northern Wei have been discovered only outside of the modern borders of China, with the most famous examples coming from the royal tombs at Tillya-tepe in Shi-barghan, northern Afghanistan, discovered in 1978 by Viktor Sarianidi (1985).4 The tombs were built between the first century BC and the first century AD. The chin-straps are simple, long, golden strips widened in the middle and tapered towards the ends (Fig. 16). They were found in situ under the jaw with both ends extending to temples. The tiny perfora-tions at the ends of the golden chin-straps indicate that they were sewn onto fabrics which most likely would be fastened on the skull.

Finds in Xinjiang

Recent archaeological discoveries have pushed the chronology of using chin-straps for the deceased further back into the past. The earliest chin-straps–made of fabrics and well-preserved under the arid conditions of Xinjiang–can be attested by finds at the Zaghunluk (Zhagunluke 扎滾魯克) cemetery near Jumo 且末 (Cherchen). The mummies of a man and a woman from the multiple burial 85M2 dated to the eighth century BC5 wore woollen chin-straps of dark red color (Xinjiang 1998a: 5–6;

4 Sarianidi 1985, photo on p.12, and cat. no. 3.23. no. 4.23, and no. 5.5. The chin-straps were found in tomb 2 (female, p. 22), tomb 3 (female, p. 28), tomb 4 (male, p. 35), tomb 5 (female, p. 45), and tomb 6 (female, p. 47). Tomb 4 yielded two chin-straps. The lengths of these chin-straps range from 31.7 to 36.3 cm. 5 While the radiocarbon dating of a timber sample from 85M2 gives the result of 2960± 115 BP (Xinjiang 1998a: 16), this date becomes obviously too early when compared

Figure 16. Chin-straps found at Tillya-tepe. Right: deceased female with a diadem and a golden chin-strap, Site 6; left: Chin-strap of a thin sheet of gold from Site 3 (left). The ends have three and four perforations each; after Sarianidi 1985: 12, and 239,

cat. no. 3.23.

46 SHING MÜLLER

Barber 1999: 24, 28–19, fig. 2.2, pl. 3A). Two further interred women were too decayed to allow any examination. A well-preserved baby from the same tomb is not reported to have such a device, although a one-year old boy from another tomb 89M2 (excavated 1989) in the vicinity, did wear an ochre-colored chin-strap (Mallory and Mair 2000: 194). The head of a man found in the same cemetery, who was prob-ably contemporary with those deceased in 85M2, also wore a bright red woollen chin-strap (Wang Binghua 2001: photo on p. 87). This fact indicates that there are more finds of chin-straps at Zaghunluk, which are not mentioned, however, in any publication. While burials earlier than 85M2 at Zaghunluk (and also those finds of Loulan and Hami) testify neither to the use of face-covers nor of chin-straps, similar chin-straps were found in burials at Jumbulak Kum (Francfort et al 2001: 140), a recently excavated site on the upper reach of the Keriya River in the middle of the Taklamakan desert with a walled settlement and cemeteries (Debaine-Francfort and Idriss 2001b). The burial practices at Jumbulak Kum are reportedly comparable to those at Zaghunluk, and the cemeteries are dated to around the middle of the 1st millennium BC. Most likely, the Zaghunluk and Jumbulak Kum sites witnessed this custom being introduced into the region and becom-ing widespread in the first half of the 1st millennium BC. However, it was not yet a common phenomenon among these early inhabitants of Xinjiang. Even though only a fraction of the earliest chin-straps were preserved, it is true that not every corpse with well-preserved clothing and headdresses wore a chin-strap. Another site, Jiawaairike 加瓦艾日克, near Jumo, whose sepulchral culture is related to that of Zaghunluk, however, reveals another custom of using mouth-covers or lip-bands of gold. In a burial ground with twelve tombs excavated in 1995, two collective tombs of the eighth to the fourth (tomb no. 1) and of the fourth to the second centuries BC (tomb no. 3) each yielded a rectangular, thin, golden foil with rounded corners (L. 7.1 cm, W. 3.1–3.2 cm) over the mouth. The one from tomb no. 3 (with 17 individuals) is perforated on two ends and thus can be fixed onto the face (Zhongguo Xinjiang 1997: 24, 28). These golden lip-bands, also attested in M94 at nearby Zaghunluk (with a perforation

with the radiocarbon values of other samples from the same phase (Phase II) of the Zaghunluk cemetery (Xinjiang 1998c: 50), since 85M2 is classified into the Phase II according to the assemblage of grave goods. It was thus suggested to place 85M2 in the early stage of Phase II set at roughly the 8th century BC. This date is confirmed by the radiocarbon dating of several other samples from the same phase (Xinjiang 1998a: 16; Xinjiang 1998c: 50). Mallory and Mair (2000: 189) speak of a date later than 600 BC, but do not give their argument for this dating, although this dating fits rather well with the dating of Jumbulak Kum (see below). At present, the more general date of 1000 BC for the 85M2 is still preferred in various publications, see e.g. Barber 1999.


at each end; L. 10 cm; B. 3.5 cm) (Xinjiang 1998c: 37)6 and at Jumbulak Kum, appear to have arrived in the Tarim Basin simultaneously with the earliest chin-straps of Zaghunluk. On the northern edge of the Tarim Basin, a later use of chin-straps can be witnessed in the burial ground at Yingpan 營盤 on the Könch-darya (Kongque River 孔雀河) near Yuli 尉犁, dated from the second to the third centuries AD. Tomb no. 15, excavated in 1995, yielded a male corpse with a moulded face-mask of hemp (ma 麻). Underneath the mask, the deceased wore a chin-strap of silk, which had been put on before his head was wrapped with silk floss (Xinjiang 1999: 7). Recent excavations at the same graveyard brought more evidence to light. Dur-ing the excavations in 1995, chin-straps of silk or cotton were found in tomb nos. 14, 18, 19, 22, 26 (Xinjiang 2002a), and the excavations in 1999 again located specimens in tomb nos. 6, 8, 13, 59 (Xinjiang 2002c). According to the excavators (Xinjiang 2002a: fig. 8 and 18), a chin-strap was made of a single wide fabric strip, most likely white-colored or undyed, and was usually accompanied by a headband that reminds one of the metal headband of the Northern Wei tomb no. 107 from the Datong 1988 cemetery. Some of the headbands were of red silk and were decorated with small round bronze, silver, or gold plates or foils. Most of the deceased were treated in the same way: after the jaw had been fastened with a fabric strip and a headband had been placed around the forehead, the entire head was wrapped with a layer of silk floss, which was again packed into a cotton kerchief. At last one square piece of cloth was laid over the face. Chin-straps are a rather common phenomenon at the Yingpan cem-etery, and they were not gender or age restricted. They were also used across two major burial forms, earthen pits and catacomb tombs, that coexisted at the cemetery over a long period, albeit in different zones of the cemetery. The reason for this spatial segregation has not been studied, but does not seem to be determined by status or wealth, for the grave goods found in both kinds of tombs are rather similar. Dif-ferences in body disposal (with vs. without coffins, single vs. multiple burials) and in clothing (textile vs. leather) might, however, suggest that the burials belonged to different tribal groups living in the same cultural and political environment. In the more eastern regions of Xinjiang, at the burial grounds at Subeshi (Subeixi 蘇貝希) near Shanshan, roughly from the 5th to the

6 Most of the tombs at Zaghunluk were plundered and the excavators maintained that it was gold on the heads of the deceased that the tomb robbers were after. Therefore, it is likely that more burials at Zaghunluk contained golden lip-bands. The tomb M94 belongs to the middle stage of the Zaghunluk Phase II, roughly from “Chunqiu to Han times” (Xinjiang 1998c: 51).

48 SHING MÜLLER

3rd century BC (Xinjiang 2002b), no use of chin-straps has thus far been evidenced. The faces of the dead were either veiled with leather or uncovered. The catacomb tombs at Subeshi (e.g. M25; Xinjiang 2002b: 47–48) are structurally identical to those found at Yingpan (e.g. M59; Xinjiang 2002c: 72). Also, the garments from catacomb tombs at both sites are stylistically similar, with differences only in materials: Both deceased wore a belted knee-long jacket with tight sleeves; the one of the male of M25 at Subeshi is of sheepskin with fur inside, and the one of the female of M59 at Yingpan is of silk. Besides, the male of M25 at Subeshi wore a square breast-patch of sheepskin directly on the breast while the breast-patch of the female of M59 at Yingpan is of silk and is sewn onto her pullover-shirt. These similarities (tomb structures and garments) suggest an intrinsic link of the occupants of the catacomb tombs at both sites. It should be noted that, as in the case of Yingpan, the catacomb tombs and earthen pits at Subeshi graveyards are spa-tially segregated. The excavators of Subeshi date the catacomb tombs to be somewhat later than the earthen pits in the whole Subeshi burial grounds according to the style of ceramics (Xinjiang 1997: 168). This preliminary observation suggests that the occupants of the catacomb tombs at Yingpan might have been ethnically or tribally related to those of Subeshi but were different from those buried in earthen pits at Ying-pan, where the catacomb tomb occupants took up the local funerary rite of binding the chin as a sign of assimilation. The burial site of Niya 尼雅 near Minfeng 民豐, located halfway between Jumo and Hetian (Khotan) on the southern edge of the Takla-makan, attests to the use of chin-straps in tombs almost contemporary with Yingpan. The female occupants of 95MNI tomb nos. 5 and 6 and the male occupants of tomb nos. 3, 4 and 8, as well as the male occupant of MNI tomb no.2, all dated no later than the third century AD (for 95 MNI tomb nos. 5 and 8, see Xinjiang 1998b: 25, Xinjiang 2000: 7; for 95 MNI tomb nos. 3, 4, 6 and 97 MNI tomb no.2, see Chū-Nichi 1999 II: 91, 110; III: figs. 105.2–3), were found with their jaws bound with silk bands, which, however, are more complex than those at Yingpan. The chins of the deceased individuals in 95 MNI tomb no. 8 and 97 MNI tomb no. 2 were fastened with two separate bands run-ning under the jaw and under the lower lip, forming jaw-guards (Fig. 17). The bands met at the cheeks and ran together to the top of the head, where they were tightly bound. The well preserved head from 97 MNI tomb no.2 (Chū-Nichi 1999 III: fig. 105.3) makes it clear that, in this case, the beard was stretched intentionally through the opening of the jaw-guard and thus retains its natural form and length.7 The same

7 To this find only a black and white photo is available. The tomb find is not reported otherwise.


Figure 17. Textile chin-strap of M8 in Niya, Xinjiang; from Xinjiang 2000: 30, fig. 63.

concern of demonstrating the beard is also observed in Sampula (see below). Reminiscent of the old-fashioned wool straps from Zaghunluk, but so much improved that the jaw was indeed tightly secured to the cranium, these chin-straps from Niya show similarities in structure to the jaw-guard with an opening of the later golden chin-straps of Lady Heruo, and of Shi Daode of the Tang dynasty (Table 1, nos. 6, 7), as Wang and Wang (1999) have noticed. More evidence has been coming to light in the regions further west along the southern edge of the Taklamakan. From the burial ground LS I excavated in 1984 at Sampula (Shanpula 山普拉) near Hetian 和闐 (Khotan), altogether 5 woollen and 19 silk chin-straps, termed hu he zhao 護頜罩 (jaw-guards) by the excavators (Xinjiang 2001: 37, 39), were recovered from at least three multiple burials (M01, M24, M49). These jaw-guards were fastened onto the head with two textile straps, and thus could be treated as a special type of chin-strap. (Fig. 18). The forms of the jaw-guards are gender-dependent. The only one published for males comes from M01 (which contained 103 individuals) and is made of silk; it is rectangular in shape and has an aperture roughly 5 cm long in the middle which, as in the case of the Niya 97 MNI M2 find, allows the wearer’s beard to be visible (Xinjiang 2001: 39, pl. 426). Those for females are rhomboidal and were sewn to form a bow-shape. Two woollen and two silk ones were found in M24 (which contained 8 individuals), and five silk specimens came from M49 (16 individuals) (Xinjiang 2001: pls. 316, 317, 427, 428, 433). Again, more face-covers and clothes were found in this cemetery than chin-straps or jaw-guards, indicating that the latter were used only within limited groups. However, no information is available about the remaining chin-straps/jaw-guards,

50 SHING MÜLLER

nor are the data of the assemblages of burial goods or body ornaments of those burials completely published. It is therefore unknown how “jaw-guards” were used in the context of funerary rites at this cemetery. Tomb M01, from which at least one jaw-guard for a male was found, is dated from the first century BC to the first century AD and belongs to the early phase of the cemetery, while M24 and M49 with jaw-guards for females are dated to the third to fourth centuries AD (Xinjiang 2001: 43). The dating is based on log samples from the wooden tomb furniture. The 103 individuals in M01 were, however, interred over a considerable span of time, which means that some of the dead could have been buried much later than the construction of the tomb. This is, however, not the case for M24 and M49. In each tomb, the eight and sixteen deceased, respectively, were packed into one single wooden coffin. Both burials were sealed afterward. Thus the time of the construction of M24 and M49 was close to the time of the burials. The similar embroidery decors and the similar hemmed edges suggest that the male jaw-guard from M01 is contemporary to those female ones. Obviously, the form evolved locally, and deviates from the earlier chin-straps of textile bands. These bowl-shaped jaw-guards with their beautiful embroidery, including lit-tle medallions in chain stitches similar to those embossed on the metal specimens from tomb no. 107 at the Datong 1988 cemetery remind us of the Northern Wei bronze chin-straps discussed at the beginning of this article. The latest evidence in the Tarim Basin comes from two Buddhist burials of women at Imammusakazim in the region of Khotan, dated to the Northern Dynasties (439–581). Both deceased wore silk bands as chin-straps (Mallory and Mair 2000: 200).

Figure 18. “Jaw-guard” from 84LS 1 M49 (female burial); Xinjiang 2001: 227, fig. 427.


Evidence from Europe

On the other side of the Eurasian continent, golden lip-bands and mouth-pieces have been found in Mycenaean and Geometric period burials in Greece (c. 1200–800 BC). In the Near East, they were already in use for the deceased by 2000 BC and they spread throughout Western Asia as well as along the north coast of the Black Sea during the Hel-lenistic period (Curtis 1976: 59). Early evidence for chin-straps in the eastern Mediterranean includes lead devices serving to hold the lower jaw from chamber tombs of the Mycenaean period in Attica and a well-preserved golden chin-strap, combined with what was most likely a lip band, found in 1885 in a tomb probably dating to the Geometric period close to Athens (von Salis 1957: 97). From the Archaic and Clas-sical periods (8th–4th centuries BC), only a few chin-straps of gold sheet have been found in tombs in Attica (Ohly 1953: 68–9). In most inter-ments since the Geometric period, chin-straps were most likely made of fabrics, for they are mentioned in the literature (the Greek word is othone, pl. othonai ), and they are also depicted on black-figure and red-figure vases of the 6th to the 5th centuries BC in scenes of prothesis (the ceremonial laying out of the body) (Kurtz and Boardman 1971: 144, 211; Garland 1985: 23). The depictions on Greek vases provide a clear picture of these chin-straps: they consist either of a single textile band of the same width running under the lower jaw and knotted on top of the head, or of a band widened at the chin (Fig. 19). Both types are sometimes combined

Figure 19. left: a laid out young man on a red-figure loutrophoros in the Louvre. He wears a chin-strap which crosses the headband with decorations above the ears; right: a laid out girl on a red-figure loutrophoros in Athens who wears a chin-strap widened

in the middle and a diadem for brides. Both after von Salis 1957, fig. 4–5.

52 SHING MÜLLER

with a headband. Interestingly, as mentioned by von Salis (1957: 96), a deep red color was adopted for the depiction of the chin-straps on figure vases, which may symbolize the color of blood. The same color is used on the earliest known chin-straps found at Zaghunluk. One of the latest chin-straps of gold foil was found in Campania in southern Italy, in a tomb from the end of the Roman period, c. the 3rd century AD (von Salis 1957). However, the practice of tying up the jaw seems to have been widespread in Europe up to the present. Among the Slavs, the custom is practiced for fear that a corpse with an open mouth could become a vampire. And Greeks in rural areas still tie the mouth of the corpse shut, as well as its hands and feet. These are untied again later, in the belief that the corpse would otherwise not decompose and instead turn into a vampire (Barber 1988: 36, 49). It is believed that Greek chin-straps of the Bronze to the Iron Age were used for the dead not only to avoid an unsightly sagging of the jaw, but also, more importantly, out of fear that the soul would leave the body through open body orifices such as the eyes and the mouth and do harm on the living. Not all prothesis scenes depict the use of a chin-strap. Often the head of the dead is raised very high and inclined forward, which could also have the effect of keeping the mouth shut (Boardman 1955: 57). In European antiquity, sometimes a piece of turf was put under the chin to achieve the same effect. Interestingly enough, two parallels to this were found in prehistoric Xinjiang: in M14 with 19 individuals at Zaghunluk (8th–2nd centuries BC), wooden blocks were laid under the jaws of two corpses (Xinjiang 1998c: 14), and in M4 at Subeshi a stone pestle was positioned likewise (Xinjiang 1997: 153). These addi-tional findings suggest a transmission not only of the superficial use of chin-straps as a device to prevent the mouth from opening, but more fundamentally of the fear of the vicious dead. By contrast, contempo-raneous inhabitants of the Chinese world do not generally seem to have associated their fear of the dead with the gaping mouth of a corpse. The origin of the metal chin-straps of the Northern Wei at Ping-cheng seems to be indebted to the fabric chin-straps traditions of the Tarim Basin. The use of headbands suggests particular resemblance to Yingpan, whereas the presence of jaw-guards harks back to Sampula. However, no metal chin-straps have yet been found in the Tarim Basin, in contrast to chin-fastening devices made of gold foil used in the Helle-nistic world, as shown by the finds in Campania and at Tillya-tepe. This suggests that the Northern Wei chin-straps continued certain traditions from the Tarim Basin, but were also inspired by new ideas from beyond East Turkestan that influenced the jewellery techniques used (such as hinges) and the choice of metals. In any case, the Han Chinese do not seem to have been the originators of chin-straps.


Chinese Ways of Treating the Face of the Dead— Textual and Archaeological Evidence

In China proper, there is no evidence for the use of chin-straps of any kind before Northern Wei times. According to the “Book of Etiquette and Ceremonial” (Yili 儀禮), which contains texts from the Zhanguo period and is frequently cited in matters concerning funerary ritual, the eyes of the dead should be covered with a black padded silk fabric with a red lining, the ears plugged with a jade stick, and the head wrapped with a fumian 覆面 (lit.: face-cover) of degummed plain silk (Couvreur 1951: 444–45; Riboud 1977: 440–41). There is no mention of any special cover over the mouth, probably because the important final step—the ritual feeding—had to be done before the corpse was dressed and bound up for the coffin. In order to do this, the same text prescribes that the mouth should stay open by forcing a horn spoon (jiaosi 角柶) between the teeth. By the time of the ritual feeding (fan 飯), the mouth is to be opened with this same horn spoon (Yili “Shi sangli” 士喪禮; cf. Yili zheng-yi 儀禮正義 26: 5a; Couvreur 1951: 440, 449–50) and filled with han 唅, mouth fillings such as rice grains, shells, pearls, or jade according to the rank of the deceased (Yili zhengyi 26: 16b–17a; Loewe 1999: 34–36). After the ritual feeding, the face of the deceased was to be veiled again with the face-cover, but if the deceased was a nobleman, the face-cover, instead of being lifted, should be cut in the position of the mouth, and the dead was fed through this hole. Given such a preoccupation with filling the mouth, there was clearly no concern to keep the mouth of the dead shut, and it is unlikely that the mouth would have been tightly bound either before or after the feeding. Archaeological evidence in China proper confirms this. In the inner coffin from the tomb of the Marquis Yi of Zeng 曾侯乙 (died between 433–400 BC) at Sui Xian 隨縣, Hubei, archaeologists found 21 small jade mouth pieces (yu han 玉唅) in the form of oxen, goats, pigs, dogs, ducks, and fish in the place where the mouth cavity would have been located (Hubei 1989: 426). It should be noted that the first three rep-resent the conventional sacrificial animals for ancestral ceremonies of Zhou times, while the last two cover the animals of the sky and water. Thus, these little jade figures literally fulfilled the demand of “filling [the mouth of the dead] with food,” as mentioned by Xunzi (Loewe 1999: 34). In addition to these, the Marquis was also supplied with a “mouth plug” (kousai 口塞) of oval shape and bulging in the middle (L. 10.8 cm, max. W. in the middle 4.1 cm, and max. thickness 1.3 cm; Fig. 20). The two longitudinal ends are perforated. The excavators postulate that the piece was used to cover the mouth like a mask (Hubei 1989: 427), and was fixed on to the head with two silk straps attached on each

54 SHING MÜLLER

end. Functionally, it resembles the lip-bands found at Jiawaairike and in preclassical Greece. The usage of face-covers in the Chu area is demonstrated by the find on the face of the deceased lady from tomb no. 1 at Mashan 馬山 in Jiangling 江陵, Hubei, dated to the late fourth or early third century BC. This covering of plain silk has a slit for the eyes and a triangular opening for the mouth (Hubei 1985: 16–17), testifying to the above-mentioned custom of ritual feeding for the nobility. The wife of the Marquis Dai 軑侯夫人 (died 168 BC) from tomb no. 1 at Mawangdui 馬王堆 gives another demonstration of this Chinese tradition of covering the face of the deceased: her eyes and forehead as well as her nose were covered with two padded silk strips (Riboud 1977: 444; Hunan 1973, vol. 1: 28, vol. 2: pl. 70). Her mouth, however, was wide open, yet it did not contain jade pieces.8 Perhaps she was fed with cooked rice, such as was mentioned by Zheng Xuan’s 鄭玄 (AD 127–200) in his commentary to Liji (Loewe 1999: 34). In any case, there is no sign of any binding of the jaw. The well-preserved corpse of Sui Shaoyan 遂少言 (died 167 BC) from tomb no. 168 at Fenghuangshan 鳳凰山 in Jiangling, Hubei, cannot give either positive or negative evidence for the use of chin-straps, since all silk fabrics had decomposed (Hubei 1993: 507). However, his mouth was found open at the time of the excavation. The tongue was found rolled to the palate, and underneath the tongue was a jade seal. The movement of the tongue indicates that the process of decomposition had begun to set in before burial, which implies that there was no attempt to bind the jaw to prevent the mouth from gaping (Ji’nan 1975: 3). For Liu Sheng 劉勝 (died 113 BC), King of Zhongshan, and his

8 What Loewe (1999: 36) identified as a “sizable stopper” in the mouth of Lady Dai is actually the well-preserved protruding tongue. It was pushed outwards as the gases produced by the microorganisms inside the corpse immensely increased their pressure in the course of decomposition, which is well described by Barber 1988: 102–19.

Figure 20. Stone mouth-piece of Marquis Yi of Zeng; from Zhong-

guo 1989: 428, fig. 251,1.

Figure 21. Jade mouth-piece of Liu Sheng, King of Zhongshan; from

Zhongguo 1980: 139, fig. 97.


Consort Dou Wan 竇綰 (after 113 BC), whose cliff-faced tombs were found at Mancheng 滿城, Hebei, jade “mouth fillers” were found at the position where the mouths were once located. The skeletons were not preserved. Liu Sheng’s “mouth filler” is crescent-shaped with a protuberance in the middle of the convex side (Fig. 21), and measures 7.2 cm long, 3 cm wide, and 1.7–3.5 cm thick, and the one for Dou Wan 4.75 cm long, 2.3 cm wide, and 0.5 cm thick (Zhongguo 1980: 139–40). When introduced into the mouth, it functions like a bit with the upper and lower teeth resting on it. Only the edges and the outer fringe area are visible from outside (Xia Nai 1983: 135). Thus the mouths were in fact kept partially open. As implied by the lack of evidence for chin-straps in written sources and archaeology, it seems that preventing the mouth from gaping in the course of decomposition was not a major concern of the Han Chinese at least from the Zhanguo period to the Han times. How to fill the mouth, and with what kind of symbolic food, engaged more discussion than how to prevent the face of the deceased from becoming unpleasant-looking. Thus the principal concern of the preparation of the corpse was different between the Chinese and the non-Han population in the Tarim Basin up to Han times. But what happened in China proper after the Han? In the Southern Dynasties (AD 420–589), there are no apparent signs for the use of metals to keep the jaw from falling apart, even in tombs richly outfitted with jade, gold, and silver jewellery, and bronze finds such as those of Wang Yi 王廙 (died 322; Nanjing 1972: 27–32) and of Gao Song 高崧 (died 366; Nanjing 2001). One might argue that, possibly, the southern Chinese did bind the jaws like their northern contemporaries, but with textiles that have not survived. But one observes that during the Southern Dynasties, jade pieces of different shapes were often deposited in the mouth of the dead of the nobility, which was not done in the contemporary North. This indicates that the Chinese of the Southern Dynasties preserved the old Han tradition for the preparation of the bodies. In the meantime, the first metal chin-straps appeared in the realm of the Tuoba-Xianbei in Northern China. It is thus very likely that the use of chin-straps was indeed associated with some foreign burial traditions introduced into North China during the Tuoba occupation.

Chin-Straps and Their Users as Read from Archaeological Contexts

In their discussion of the chin-strap from tomb no. 107 of the Datong 1988 cemetery, Wang and Wang (1999: 160) maintained that all of the 14 chin-straps collected were retrieved from tombs with relatively

56 SHING MÜLLER

rich inventories and suggested that they were associated with status. However, the structure of the cemetery—with four major tomb types (Shanxi 1992; Müller 2000: 117–20) and an occupation of roughly one hundred years—seems to be complex and the burials heterogeneous. With the complete data of the cemetery still unpublished, it is difficult to examine the assumed correlation between the use of chin-straps and the social status of the deceased in this cemetery. While the wealth of the grave goods may have been connected to high social status and been a prerequisite for the use of chin-straps, some factors other than wealth and/or status could also have been involved. For example, there might also be rich burials in the same cemetery without chin-straps. Also, at least for later times, the metals used for chin-straps were not absolutely correlated with the richness of the burial. For Northern Wei times, instances of gold, silver, and bronze have been found, but lead was also allegedly adopted (Table 1, no. 2). Although Shao Zhen’s chin-strap from the later Northern Wei (Table 1, no. 3) is of silver, his burial was moderate in terms of the amount and quality of the grave goods as compared to other Northern Wei burials in Luoyang, and his position was not a high one (see below), indicating that a moderate burial and official rank could still be associated with the usage of chin-straps made of a costly material. For Tang times, where data have been more care-fully retrieved and the identification of the deceased is less problematic because of tomb inscriptions, no absolute relationship between the status / rank and the metals of chin-straps can be demonstrated. Gold was used for the wife of a palace attendant, while bronze was used in the chin-strap for a princess and granddaughter of the founder of the Tang Dynasty (Table 1, nos. 9, 10).

Cultural affinities

One of the factors that could have been involved in the use of chin-straps is the “tribal” affiliations of the deceased. As suggested above, the binding of the jaw does not seem to have been practiced among the Han Chinese because their preferences in caring for the dead were dif-ferent, and early circumstances of chin-straps all came from burials on the fringes of the Tarim Basin with foreign burial types and habits. No finds have ever been made east of the Taihang Range 太行山. Without giving explicit arguments, such as associating the finds of chin-straps with a certain burial custom, Wang Yintian has proposed that the occupants of the Datong 1988 cemetery belonged to one group of tribal members under the leadership of the Tuoba (Shanxi 1992: 11). However, the heterogeneity of the burials suggests that the cemetery was occupied by several different ethnic components within the Tuoba confederation. It


should be pointed out that grave forms, inventories, and burial habits in this cemetery are distinct from those of the contemporary nobility of Chinese origin at Pingcheng who were buried to the southeast of the city in their own burial grounds, and who tended to follow a Northwestern pattern of burials at this time (Soong 2002: 281–83). Thus, the Datong 1988 cemetery was obviously occupied by non-Chinese tribal members. If the excavation data were fully published, an analysis of the intrinsic structure of the cemetery could give clues as to why only 14 complete chin-straps out of a total of 167 burials were found and for what kinds of burials these devices are characteristic. One later tomb find of Shedi9 Huiluo 庫狄廻洛 (died 562) and his two wives, née Hulü 斛律氏 (died 545) and née Yu 尉氏 (died 559) in Shouyang 壽陽, Shanxi (Wang Kelin 1979), may illustrate the ethnic factors in the use of chin-straps. Despite the fact that this is one of the few great aristocratic tombs of the Northern Qi that has not been plundered, there is no sign of using chin-straps in this tomb. Lady Yu was, according to her epitaph, obviously a descendant of the Yuchi 尉遲 (Khotanese Saka: Visya) group,10 while the Shedi and the Hulü were both clans/tribes originally belonging to the “High Carts” (the Gao-che 高車, also known in China as the Chile 敕勒, the Tiele 鐵勒, the Tele 特勒, and the Dingling 丁零) from southern Siberia (Pulleyblank 1990; Pulleyblank 2000: 85–88) who joined the Tuoba confederation in Northern Wei times. This example implies that chin-straps were not widely adopted even among the non-Chinese aristocrats, such as the High Carts, under the reign of the Tuoba-Xianbei. In the case of Shao Zhen (Table 1, no. 3), one cannot be certain to which tribal group he or his ancestors belonged. His epitaph, recorded in Chinese, states that he was commandant (ling 令) of Eyang 阿陽 (probably in what was then Qinzhou 秦州, modern Tianshui 天水, Gansu; see Zhao Chao 1992: 115), deputy Governor of Anding 假安定太守 (in modern Jingchuan 涇川 in Ningxia), and descendant of an “Inspector of Yongzhou” 雍州刺史. It is intriguing that the inscription does not state that his ancestors were of Yongzhou origin, nor does it boast a genealogy as epitaphs of members of the great Chinese families usually did at the time. The fact that Shao Zhen was buried at Yongzhou (which, during the fifth century, included modern Xi’an), rather than

9 On the pronounciation of Shedi see Yao Weiyuan 1962: 182–83. 10 Yao Weiyuan considered that the Yuchi originally belonged to the Tuyuhun 吐谷渾 and lived in what is today northeastern Qinghai, before some of the Yuchi members took hold of Khotan in the 4th or the 5th century, and others were integrated into the Tuoba confederation in the 5th century (Yao Weiyuan 1962: 189–198; Pulleyblank 2000: 83–84). An interesting description in Weishu, which states that the rulers in Khotan being rather “Chinese-looking” (Weishu 102: 2263), may indicate that the Yuchi indeed were of eastern origin.

58 SHING MÜLLER

at the place where he was serving in office, may indicate that he rec-ognized the place where his ancestors first became famous as his place of origin.11 This would imply that Shao Zhen was not a Han Chinese. This is not surprising considering the extreme ethnic complexity in the Guanzhong area since the fourth century (see Ma Changshou 1985). Shi Daode (Table 1, no. 6) of early Tang times was a Sogdian descend-ant (Daode of “Kesh”, Shahr-i-Sabz in Uzbekistan of today). However, his epitaph states that his ancestors lived in Jiankang 建康 (southwest of modern Gaotai 高臺, western Gansu) for many generations, before the family, most likely in Northern Wei times, moved to Guyuan (see Ningxia 1985: 30; Luo Feng 1998: 213–16; Luo Feng 2001: 238–245). The surname Heruo (Table 1, no. 5) goes back to the Tuoba Wei era. The Heruo group stemmed from “north of the desert” (mobei 漠北) and joined the Tuoba no earlier than the 460s (Beishi 北史 68: 2378). Boodberg (1936: 178), who understood Heruo as title, suggested that this designation contains Turkic elements and those so designated might have been of the Ruanruan or the Tuyuhun groups. Except for a few famous personalities, nothing is known about the Heruo, not to mention their customs. The same is true for the ancestors of Lady Yan Wan (see Table 1, no. 8), who, according to Yuanhe xingzuan 元和姓纂 (juan 5: 771), were tribal members of the Tuoba confederation. The earliest known personality of this group, Yan Man 閻滿, served the Northern Wei emperor Taizu Tuoba Gui 拓跋硅 (r. 386–409), but nothing more is known about the early Yan group. The background of the other individuals buried with chin-straps is even more difficult to trace. Madame Song’s (Table 1, no. 10) identity is unknown, although the burial practice—grave goods extremely rich in golden jewellery and the small, trapezoidal earthen chamber (L. 3.45–3.47 m, W. 2.85–3 m, H. 1.75–1.85 m)—is reminiscent of those of the non-Chinese at Pingcheng during the Northern Wei.12 The ancestors of the royal Li family of the

11 A parallel of this mode of recognizing the first place of a career establishment in a Chinese dynasty as the place of origin inside China proper by a non-Chinese immigrant can be drawn from the epitaph of Bo Minzhong 白敏中 (?–863), chancellor under Tang Xuanzong 唐宣宗 (847–858). The inscription traced the very origin of the Bo family, from which the famous poet Bo Juyi 白居易 (772–846) probably also stemmed, back to Chu 楚 (probably the area along the River Chu to the Lake Issyk-kul), and says “at the beginning of Yuan Wei 元魏 (i.e. Tuoba Wei, 386–534), [Bo] Bao 包, the Marquis of Yangyi 陽邑侯, became the governor of Taiyuan (Taiyuan taishou 太原太守). Since then the offspring recognized this place as their homeland. Until today they claim themselves being of Taiyuan origin.” (Ma Changshou 1985: 64). 12 One possibility to associate her with China’s northwest is her surname. The Song family was one of the most prominent families in the Hexi / Dunhuang region during the Nanbeichao era. One well-known burial from the members of this family is that of Song Shaozu 宋紹祖 (died 477), in which the use of a large number of ceramic figu-rines and a stone sarcophagus in the shape of a house––one of the earliest such forms known–– show innovations that are not observed in the contemporary burials among


Tang from Longxi 隴西 intermarried widely with the non-Chinese tribal aristocracy (Wechsler 1979: 150–51), which explains that some Li members adhered heavily to northwestern traditions. However, every descendant from the imperial Li family did not necessarily wear a chin-strap, as the unlooted tomb of little Li Jingxun 李靜訓 (Finsterbusch 1976: 23) attested. The wearing of a chin-strap may have depended on who they had intermarried with. The situation in Xinjiang is even more obscure. Recent studies give a picture of settlements of different peoples from the northern steppes and from Central Asia. Depending on periods and locations, these peoples have been categorized as Indo-Iranians (early Bronze Age at Qäwri-ghul), Indo-Europeans/Tocharians (Bronze Age at Hami), Europoids of “Mediterranean” type (late Bronze to early Iron Age at Zaghunluk), and Indo-Afghan or Pamir-Fergana types of Iranians (on the southern edge of the Taklamakan) (cf. Mallory and Mair 2000). Beside these anthropological traits that differentiate the settlers in Xinjiang, written sources retained in the Chinese historical records from the 2nd century BC onwards manifest the flux and reflux of different peoples, the Saka, the Xiongnu, the Wusun, Bactrians, Persians, Indians, Sogdians, and Chinese. All of these exerted certain political, military, and cultural influences on the locals, and some even became part of the inhabitants of the oases states. Without detailed studies of the burial habits and the related social structures, it is impossible to relate the users of chin-straps found in the Tarim Basin to certain ethnic/tribal groups. Concerning the two best-known female burials (tomb nos. 107 and 109) with chin-straps from the Datong 1988 cemetery, at this stage it is only possible to point out that both burial customs are strongly western-related. The deposit of hazelnuts, walnuts, and jujubes on a tray in these two tombs have their parallels in female burials in Xinjiang from the second century onwards. In China proper, except for a short period dur-ing the early Western Han in aristocratic tombs such as in the Kingdom of Changsha (e.g. Mawangdui tomb no. 1), in the tombs of princes in Xuzhou and in Nanyang, the deposits of such fruits remained rare until the second half of the sixth century. Only a few burials, such as that of little Li Jingxun of the Tang dynasty, contain walnuts. Conversely, wal-nuts and hazelnuts as deposits for females had been common in Europe since the Iron Age, and were considered symbols of fertility (Moschkau 1957). Such fruits were still deposited in some burials in the northern

the high members of the Chinese society (Shanxi 2001; Soong 2002: 290). On the other hand, the tomb form of Madame Song is totally different from that of Song Shaozu, but similar to those of the non-Chinese members at the Datong cemetery. Therefore, the relationship between Madame Song and the Songs from Dunhuang remains specula-tive, as long as her tomb inscription is not fully published.

60 SHING MÜLLER

Caucasus region such as Moscevaja Balka until the eighth and ninth centuries (Ierusalimskaja 1996: 30–31). The presence of the pectoral in the Belgian collection (Fig. 5d) pro-vides a clue that the “crescent-shaped bronze strip” found in tomb no. 107 (Fig. 3.10) could have served as such as well. Pectorals, otherwise unknown in the so-called Xianbei and Xiongnu finds in China, may also imply that the deceased of the tomb no. 107 belonged to a west-ern group. The western link of the early chin-straps of the Pingcheng period is further illustrated by the hinged fastenings connecting the jaw-holders with the cheek-girdles, which are described for the one from tomb no. 107 and attested to by the one in the Belgian collection; the other specimens from the Datong 1988 cemetery may have been of the same construction. These hinges make the cheek-girdles movable. They bring to mind the same methods for connecting the cheek guards to the helmets of the Roman legionaries during the first to the second centuries. Hinges are also seen in belt-buckles and jewelry in Europe up to the Period of Migration from the late 4th to the 7th century AD. The golden torque and bracelet from the Sarmatian burial (Barrow 10) at Kobyakov on the eastern outskirts of Rostov-on-the-Don, dated to the first century AD (Guguev 1996: 53–59), give a good example for this handicraft. Hinges were used in China for furniture and architectural constructions since Zhou times, if not earlier, but hinged fastenings have not yet been evidenced for Chinese garments and armor. On jewellery, at least on bracelets, the technique can be traced back to the last centuries BC on the lower reaches of the Syr-Darya (Guguev 1996: 59). This implies that the chin-straps were constructed by craftsmen with knowledge of jewellery traditions from Bactria at that time. One of the major changes after the move of the Northern Wei capital from Pingcheng to Luoyang in 494 was the discard of the hinges. Does it in-dicate a loss of the technique because the craftsmen with this knowledge did not move with the royal Tuoba members? Or was there a change in taste? It is unclear why the chin-straps were simplified in this way. What is intriguing is the lack of any indication for the use of metal chin-straps in the burials of the Oxus culture of the Bronze Age,13 al-though chin-straps of textile and gold, as well as lip-bands, may have been widely distributed in Europe and especially in and around Greece, as mentioned above. The golden chin-straps used by the royal members of the Yuezhi-Kushan Kingdom found in the Tillya-tepe burials at Shi-barghan could have been introduced by Alexander the Great during his last campaign to Bactria. Aside from chin-straps, the Tillya-tepe burials attest to other Hellenistic influences as well: coins in the mouth serving

13 I am indebted to Dr. Dietrich Huff of the German Archaeological Institute for this information.


as the fee to be paid to the ferryman Charon over the river Styx and shoe models to aid the journey to the Netherworld (Kurtz and Board-man 1971: 210–13; Sarianidi 1985: 58). With the exception of this one single piece of evidence, there is a blank in the distribution of chin-straps in West Turkestan. In fact, there are few tombs in this region from the time between ca. 300 BC and AD 500. One major reason for this could have been that corpses were seldom inhumed due to the influence of Zoroastrianism. The more usual way of treating the remains of the dead may have been the use of ossuaries for bones after the disposal of the corpses for excarnation.

Concluding Remarks

The lack of literary and archaeological evidence for chin-straps in China proper on the one hand, and their relatively abundant and continuous presence in East Turkestan on the other, makes it apparent that the custom of using chin-straps was introduced from this area into China during Northern Wei times. In East Turkestan, the chronology of the use of chin-straps can now be traced back to the eighth century BC. Before then, there is no evidence that the inhabitants in East Turkestan used chin-straps. Hsü (1998) suggests that several episodes of global climatic cooling around 2000 BC, 800 BC, and later initiated waves of migrations of Indo-Europeans from the southern Siberian steppes into the Tarim Basin. The second wave, around 800 BC, corresponds relatively well with the first occurrence of chin-straps in Zaghunluk. The similar finds with chin-straps at Jumulak Kum testify to the spread of these immigrants in the less harsh climate of the Tarim Basin. Not only the sudden occurrence of these early chin-straps at Zaghunluk, but also the “dark red color” and the simultaneous occurrence of lip-bands, and also, in a few cases, the blocks underneath the chin to stop the dropping of the lower jaw suggest that the beliefs on afterlife of these early immigrants stemmed directly from, or were at least strongly influenced by, the people of the area around the eastern Mediterranean. The use of chin-straps in the Tarim Basin persisted from the time of Zaghunluk down to the period of Yingpan, Niya, and Sampula, whose linguistic and anthropological characteristics differed from those of the inhabitants at Zaghunluk and from one another. While the European counterparts remained relatively simple and uniform, and were occa-sionally made of gold, the chin-straps from Yingpan to Sampula of the first centuries AD varied from one another in their structure and all dif-fered from those European ones, including those from Tillya-tepe, which

62 SHING MÜLLER

follow the Greek tradition. This suggests that the custom of binding the chin of the deceased was passed down from Zaghunluk and Jumbulak Kum to later newcomers. Meanwhile, the archaeological findings point to a broader spread of the use of chin-straps in Tarim during the 2nd and the 3rd centuries AD, as witnessed by the finds at Yingpan, Niya, and Sampula. One of the factors could have been the increasing wealth through trading, which intensified the practices of local funeral rites. The inhabitants of the oases states at Yingpan, Niya and Sampula then developed the new forms of chin-straps locally by using the material they valued most—silk. While the chin-straps found in Yingpan keep the relatively traditional and simple form of a textile band, Sampula in Khotan, however, seems to have been relatively isolated from the east-ern Tarim, developed a unique type of chin-strap/jaw-guard. Such a device is not known from Europe or the Near East in the historic era. In some cases, the custom was presumably also accepted by other foreign tribal/ethnic groups who lived in those states. The tradition became so deeply rooted among some members that even the arrival of new religions did not affect it, as demonstrated by the two Buddhist burials from Khotan. The reason for the impact of the Xiongnu/Saka culture seen in many burials in the Tarim Basin of the centuries around 0 AD can be explained by the fact that the Xiongnu and Saka controlled the Silk Roads before the Han entered the area. Until this day, however, the so-called Xiongnu burials found in China and in Mongolia, even those with rich gold finds, have not yielded anything similar to chin-straps or lip-bands. This suggests that such devices did not belong to the funerary rites of the Xiongnu or, for that matter, other eastern Siberian groups such as the Gaoche. The use of chin-straps in the oases states of the Tarim Basin must have been limited to the sedentary inhabitants and was not practiced by the nomadic rulers. The chin-straps found at Zaghunluk were not combined with face-covers. According to my preliminary survey, face-covers, on the other hand, were frequently encountered since the 3rd century BC in cem-eteries in Tarim. The use of face-covers can probably be regarded as a sign of the advance of the old Han-Chinese custom which could be traced back to the 8th or the 9th centuries BC into this area. The spread of the custom of face-covers overlapped and even exceeded the use of chin-straps in Tarim. What seems to be clear is the relationship between these early textile chin-straps and the first metal chin-straps of Northern Wei times. The Northern Wei metal chin-straps must have been an innovation out of the jaw-guards exemplified by the finds from Sampula combined with headbands as seen at Yingpan. These typological traits suggest that some


members of the Tuoba confederacy were related to, or descended from, the inhabitants of these oases states. On the other hand, the metalwork as evidenced on the finds from the Datong 1988 cemetery and the two specimens now in private collections must have gained impulses from elsewhere. This was most likely through Central Asia, where the tech-niques of using hinges to connect jewellery and to mount glass paste in cells had been practiced for a long time. The ethnic/tribal identity of the occupants of Northern Wei tomb nos. 107 and 109 remains, however, obscure. No identical burial customs have yet been found in the contem-porary Tarim Basin. By the 470s, the Tuoba tribal groups already had a rather well-established centralized political system, which manifested itself on the cultural pursuits of all their members by giving them certain “Tuoba” traits such as the Xianbei garb and trapezoidal-shaped coffins (Dien 1991). The structural similarity of the metal chin-straps found in the Datong 1988 cemetery and the two golden specimens in private collections (see above) suggests that chin-straps in the early Northern Wei had also been standardized. The use of chin-straps also demonstrates influences from the oases states on the northwestern part of the Northern Wei realm, i.e. the regions of modern Gansu and Ningxia. Although no chin-straps have yet been found in that area, various later users of chin-straps (Shao Zhen, Shi Daode, Madame Song, and the Li family of the Tang) originated from there, and their ancestors had played a role in the formation of the Tuoba culture during the Pingcheng period. Most chin-strap users during the early Tang belonged to ethnic groups of non-Chinese ancestral origin. The use of chin-straps during the Tang was first confined to the vicinity of the capitals of Chang’an and Luoyang, and the members of high society, as indicated by epitaphs or exclusive grave goods. After the rebellion of An Lushan in 755, China became xenophobic, and the use of chin-straps declined. Meanwhile, since the middle Tang there was a tendency for the use of chin-straps to move southwards. As seen in the cases of Li Hui and Zheng Xun, high society members banned to the south may have contributed to the distribution of chin-straps. In addition, the use of chin-straps was most likely also accepted by certain Han Chinese. Wang Jian, for instance, was, according to his biography, a native of Henan and had been an outlaw before his military career moved him to the power center. He assisted Emperor Xizong in fleeing to Sichuan, where he later established his own kingdom of the Great Shu 蜀. The continuous revolts and unrest in north China during the last days of the Tang were possibly a reason that the custom of using chin-straps became obsolete in that region, surviving only in the Yangzi area, where it finally died out as well. The evidence mentioned in this article is far from exhaustive, and thus the described scenario of the movements of chin-straps in time and

64 SHING MÜLLER

space remains speculative because of insufficient archaeological data. The problem remains twofold: 1) Excavators are not always aware of the existence of chin-straps and thus fail to mention them; 2) The burials are often not thoroughly published, if at all, so that the burial customs cannot be studied. More work must be done in order to clarify who the users of chin-straps were, as well as their possible religious meanings.

Acknowledgements

I am grateful to Professor Annette Juliano and Professor Thomas O. Höllmann for their suggestions and advice. I would also like to express my great thanks to Professor Lothar von Falkenhausen and Professor Robert Murowchick for their valuable comments and many editorial suggestions. My thanks also goes to the reviewers for their comments on an earlier draft of this article. Any errors remain, however, my own responsibility.

References Cited:

Barber, Elizabeth W. (1999). The Mummies of Ürümchi. New York: W.W. Norton and Company.

Barber, Paul (1988). Vampires, Burial, and Death: Folklore and Reality. New Haven and London: Yale University Press.

Beishi 北史, compiled by Li Yanshou 李延壽 (Tang). Beijing: Zhonghua Shuju, 1987.

Benko, Mihály (1992/93). “Burial Masks of Eurasian Mounted Nomad Peoples in the Migration Period (1st Millennium A.D.),” Acta Orientalia Academiae Scientiarum Hungaricae XLVI, 2–3: 113–31. Budapest.

Boardman, John (1955). “Painted Funerary Plaques and Some Remarks on Prothesis.” The Annual of the British School at Athens 50 (1955): 51–66.

Boodberg, Peter A. (1936). “The Language of the T’o-pa Wei,” Harvard Journal of Asiatic Studies 1: 167–85.

Chen and Feng 1984: Chen Guo’an 陳國安 and Feng Yuhui 馮玉輝 (1984 “Hengyang xian Hejiazao Bei Song mu 衡陽縣何家皂北宋墓 (A Northern Song Tomb at Hejiazao, Hengyang County).” Wenwu 文物 1984.12: 73–74.

Chū-Nichi 1999: Chū-Nichi Nitchū kyōdō Niya iseki gakujutsu chōsatai 中日日中共同尼雅遺跡學術調查隊 (1999). Chū-Nichi Nitchū kyōdō Niya iseki gakujutsu chōsa hōkokusho 中日日中共同尼雅遺跡學術調查報告書 (Report on the Sino-Japanese excavations at the Niya sites). 3 vol. Kyōto: Chū-Nichi Nitchū kyōdō Niya iseki gakujutsu chōsatai.

Couvreur, Seraphin (1951). I-li. Cérémonial. Leiden: Cathasia (first pub-lished 1916).


Curtis, John (1976). “Parthian Gold from Nineveh.” The British Museum Yearbook 1, The Classical Tradition, pp. 47–66. London: British Museum.

Debaine-Francfort, Corinne, and Abduressul Idriss (editors) (2001a) Keriya, mémoires d’un fleuve. Archéologie et civilisation des oasis du Taklamakan. Paris: Editions Findakly.

Debaine-Francfort, Corinne, and Abduressul Idriss (2001b). “Djoum-boulak Koum, une cité fortifiée.” In Debaine-Francfort and Idriss (eds.) 2001a:120–136.

Desroches, Jean-Paul (2000). L’Asie des Steppes d’Alexandre le Grand à Gengis Khân. Paris: Réunion des Musées Nationaux.

Dien, Albert (1991). “A New Look at the Xianbei and Their Impact on Chinese Culture.” In George Kuwayama (editor), Ancient Mortu-ary Traditions of China, pp. 40–59. Los Angeles: Los Angeles County Museum of Art.

Feng Hanji 馮漢驥 (1964). Qian Shu Wang Jian mu fajue baogao 前蜀王建墓發掘報告 (Report on the Excavation of the Royal Tomb of Wang Jian of Former Shu). Beijing: Wenwu Chubanshe.

Finsterbusch, Käte (1976). Zur Archäologie der Pei-Ch’i (550–577) und Sui-Zeit. Münchener Ostasiatische Studien vol.1. Wiesbaden: Franz Steiner Verlag.

Francfort et al. 2001: Francfort, Henri-Paul, Abduressul Idriss and Zhang Yuzhong (2001). “Des morts dans les sables. Pratiques funé-raires.” In Debaine-Francfort and Idriss 2001a:137–143.

Garland, Robert (1985). The Greek Way of Death. Ithaca: Cornell Uni-versity Press.

Guangzhou 1959: Guangzhou Shi Wenwu Guanli Weiyuanhui 廣州市文物管理委員會 “Guangzhou Huangdigang Tang muguo mu qingli jianbao 廣州皇帝崗唐木槨墓清理簡報 (Preliminary report on the excavation of the Tang wooden coffin tomb at Huangdigang in Guangzhou).” Kaogu tongxun 考古通訊 1959.12: 668–70, 678.

Guguev, Vladimir (1996). “The Gold Jewelry Complex from the Kobyakov Pit-Burial.” In Adriana Calinescu (editor), Ancient Jewelry and Archaeology, pp. 51–61. Bloomington: Indiana University Press.

Hengyang 1996: Hengyang Shi Wenwu Gongzuodui 衡陽市文物工作隊 “Hunan Leiyang Chengguan Liuchao Tang Song mu 湖南耒陽城關六朝唐宋墓 (Six Dynasties, Tang, and Song tombs at Chengguan, Leiyang, Hunan).” Kaogu xuebao 考古學報 1996.2: 237–77.

Hsü, Kenneth J. (1998) “Did the Xinjiang Indo-Europeans Leave Their Home Because of Global Cooling?” In Victor H. Mair (editor), The Bronze Age and Early Iron Age Peoples of Eastern Central Asia. 2 vol., pp. 683–96. Philadelphia: The University of Pennsylvania Museum Publications.

66 SHING MÜLLER

Hubei 1985: Hubei Sheng Jingzhou Diqu Bowuguan 湖北省荊州地區博物館 Jiangling Mashan yihao Chu mu 江陵馬山一號楚墓 (Chu tomb no. 1 at Mashan, Jiangling). Beijing: Wenwu Chubanshe.

Hubei 1987: Hubei Sheng Bowuguan 湖北省博物館 and Yun Xian Bowuguan 鄖縣博物館 “Hubei Yun Xian Tang Li Hui, Yan Wan mu fajue jianbao 湖北鄖縣唐李徽閻婉墓發掘簡報 (Preliminary report on the excavation of the the Tang dynasty tomb of Li Hui and Yan Wan in Yun County, Hubei).” Wenwu 文物 1987.8: 30–42, 51.

Hubei 1989: Hubei Sheng Bowuguan 湖北省博物館. Zeng Hou Yi mu 曾侯乙墓 (Tomb of Marquis Yi of the State Zeng). 2 vol. Beijing: Wenwu Chubanshe.

Hubei 1993: Hubei Sheng Wenwu Kaogu Yanjiusuo 湖北省文物考古研究所 “Jiangling Fenghuangshan 168 hao Han mu 江陵鳳凰山168號漢墓 (Han tomb no. 168 at Fenghuangshan, Jiangling).” Kaogu xuebao 考古學報 1993.4: 455–513.

Hunan 1973: Hunan Sheng Bowuguan 湖南省博物館 and Zhongguo Kexueyuan Kaogu Yanjiusuo 中國科學院考古研究所 (editors). Ma-wangdui yihao Han mu 馬王堆一號漢墓 (Han tomb no. 1 at Mawang-dui). 2 volumes. Beijing: Wenwu Chubanshe.

Ierusalimskaja, Anna A. (1996). Die Gräber der Moscevaja Balka: Frühmittel-alterliche Funde an der Nordkaukasischen Seidenstraße. München: Editio Maris.

Ji’nan 1975: Ji’nan cheng Fenghuangshan 168 hao Han mu fajue zhengli zu 紀南城鳳凰山 168 號漢墓發掘整理組 “Hubei Jiangling Fenghuangshan 168 hao Han mu fajue jianbao 湖北江陵鳳凰山168號漢墓發掘簡報 (Preliminary report on the excavation of the Han tomb no. 168 at Fenghuangshan in Jiangling, Hubei).” Wenwu 文物 1975.9: 1–8, 22.

Jiangxi 1990: Jiangxi Sheng Kaogu Yanjiusuo 江西省考古研究所 “Jiang-xi De’an Nan Song Zhoushi mu qingli jianbao 江西德安南宋周氏墓清理簡報 (Preliminary report on the excavation of the Southern Song tomb of Lady Zhou at De’an, Jiangxi).” Wenwu 文物 1990.9: 1–21.

Juliano and Lerner 2001: Juliano, Annette, L. and Judith A. Lerner (editors) (2001). Monks and Merchants: Silk Road Treasures from Northwest China. Gansu and Ningxia Provinces, 4th–7th Century. New York: The Asia Society Museum.

Kurtz, D.C. and J. Boardman (1971). Greek Burial Customs. London: Thames and Hudson.

Loewe, Michael (1999 [2001]). “State Funerals of the Han Empire.” Bulletin of the Museum of Far Eastern Antiquities 71: 5–72.

Louis, François (1994). “Gold and Silver from Ancient China. The Pierre Uldry Collection on Show at the Museum Rietberg Zürich.” Arts of Asia September/October: 88–96.


Luo Feng 羅豐 (1998). Guyuan nanjiao Sui Tang mudi 固原南郊隋唐墓地 (A Cemetery from the Sui and Tang period in the southern suburbs of Guyuan). Beijing: Wenwu Chubanshe.

Luo Feng (2001). “Sogdians in Northwest China.” In Juliano and Lerner 2001: 238–245.

Luoyang 1980: Luoyang Bowuguan 洛陽博物館 “Luoyang Guanlin Tang mu 洛陽關林唐墓 (Tang tombs at Guanlin, Luoyang).” Kaogu 考古 1980.4: 382–383, 379.

Ma Changshou 馬長壽 (1985). Beiming suojian Qian Qin zhi Sui chu de Guanzhong buzu 碑銘所見前秦至隋初的關中部族 (Tribal groups in Guanzhong from the Former Qin to the beginning of the Sui as seen from stelae inscriptions). Beijing: Zhonghua Shuju.

Mallory and Mair 2000: Mallory, J. P. and Mair, Victor H. (2000). The Tarim Mummies. Ancient China and the Mystery of the Earliest Peoples from the West. London: Thames & Hudson.

Michaelson, Carol (1999). Gilded Dragons. Buried Treasures from China’s Golden Ages. London: British Museum Press.

Moschkau, Rudolf (1957). “Nußbeigaben in vorgeschichtlichen Gräbern im Lichte volkskundlicher Überlieferungen.” Forschungen zur Vor- und Frühgeschichte 2: 137–53.

Müller, Shing (2000). Die Gräber der Nördlichen Wei-Dynastie (386–534). Ph.D. Dissertation. Ludwig-Maximilians-Universität, München.

Nanjing 1972: Nanjing Shi Bowuguan 南京市博物館 (1972). “Nanjing Xiangshan 5 hao, 6 hao, 7 hao mu qingli jianbao 南京象山 5 號, 6 號, 7 號墓清理簡報 (Preliminary report on the excavation of tombs 5, 6, and 7 at Xiangshan in Nanjing).” Wenwu 文物 1972.11: 23–41.

Nanjing 2001: Nanjing Shi Bowuguan 南京市博物館 (2001). “Jiangsu Nanjing Xianheguan Dongjin mu 江蘇南京仙鶴觀東晉墓 (Eastern Jin tombs at Xianheguan, Nanjing, Jiangsu).” Wenwu 文物 2001.3: 4–40, 91.

Ningxia 1985: Ningxia Guyuan Bowuguan 寧夏固原博物館 (1985). “Ningxia Guyuan Tang Shi Daode mu qingli jianbao 寧夏固原唐史道德墓清理簡報 (Preliminary report on the excavation of the Tang dynasty tomb of Shi Daode in Guyuan, Ningxia).” Wenwu 文物 1985.11: 21–30, 20.

Ohly, Dieter (1953). Griechische Goldbleche des 8. Jahrhunderts v. Chr. Berlin: Gebr. Mann.

Pulleyblank, E.G. (1990). “The ‘High Carts’: A Turkish-Speaking People Before the Türks.” Asia Major 3.1 (1990): 21–26.

Pulleyblank, E.G. (2000). “The Nomads in China and Central Asia in the Post-Han Period.” In Hans Obert Roemer (editor), History of the Turkic Peoples in the Pre-Islamic Period, pp. 76–94. Berlin: Klaus Schwarz Verlag.

68 SHING MÜLLER

Rawson, Jessica (editor) (1996). Mysteries of Ancient China: New Discoveries from the Early Dynasties. New York: George Braziller.

Riboud, Krishna (1977). “Some Remarks on the Face-Covers (Fu-mien) discovered in the Tombs of Astana.” Oriental Art, new series XXIII 4: 438–454.

von Salis, Arnold (1957). “Antiker Bestattungsbrauch.” Museum Helveticum 14 (1957): 89–99.

Sarianidi, Viktor (1985). The Golden Hoard of Bactria. From the Tillya-tepe Excavations in Northern Afghanistan. Leningrad: Aurora.

Shanxi 1992: Shanxi Sheng Kaogu Yanjiusuo 山西省考古研究所 (1992). “Datong nanjiao Beiwei muqun fajue jianbao 大同南郊北魏墓群發掘簡報 (Preliminary report on the excavation of a North-ern Wei cemetery at the southern suburb of Datong).” Wenwu 文物 1992.8: 1–10.

Shanxi 2001: Shanxi Sheng Kaogu Yanjiusuo 山西省考古研究所, Da-tong Shi Kaogu Yanjiusuo 大同市考古研究所 (2001). “Datong shi Bei Wei Song Shaozu mu fajue jianbao 大同市北魏宋紹祖墓發掘簡報 (Preliminary report on the excavation of the Northern Wei tomb of Song Shaozu in Datong City).” Wenwu 文物 2001.7: 19–39.

Soong, Shing 宋馨 [Shing Müller] (2002). “Bei Wei Sima Jinlong mu-zang de chongxin pinggu 北魏司馬金龍墓葬的重新評估 (The tomb of Sima Jinlong of the Northern Wei reconsidered).” Journal of Chinese Studies 中國文化研究所學報 11: 273–98. Hong Kong.

Uldry, Pierre (1994). Chinesisches Gold und Silber. Die Sammlung Pierre Uldry. Zürich: Museum Rietberg.

Wang and Wang 1999: Wang Yintian 王銀田 and Wang Yanqing 王雁卿 (1999). “Datong nanjiao Bei Wei muqun M107 fajue baogao 大同南郊北魏墓群M107發掘報告 (Report on the excavation of tomb M107 in the Northern Wei cemetery in the southern suburbs of Datong).” Beichao yanjiu 北朝研究 1999.1: 143–62. Beijing.

Wang Binghua 王炳華 (editor) (2001). Xinjiang gushi: Gudai Xinjiang jumin jiqi wenhua 新疆古屍：古代新疆居民及其文化 (The ancient corpses of Xinjiang: The peoples of ancient Xinjiang and their culture). Urumchi: Xinjiang Renmin Chubanshe.

Wang Kelin 王克林 (1979). “Bei Qi Shedihuiluo mu 北齊庫狄迴洛墓 (The tomb of Shedihuiluo of Northern Qi).” Kaogu xuebao 考古學報 1979.3: 377–401.

Wang Zhulin 王竹林 (1991). “Henan Yanshi Nancaizhuang Bei Wei mu 河南偃師南蔡莊北魏墓 (A Northern Wei tomb at Nancaizhuang, Yanshi County, Henan).” Kaogu 考古 1991.9: 818–31.

Wechsler, Howard J. (1979). “The Founding of the T’ang Dynasty: Kao-tsu (Reigned 618–26).” In Denis Twitchett (editor), The Cam-bridge History of China. Vol. 3 Sui and T’ang China, 589–906, Part I, pp. 150–241. Cambridge: Cambridge University Press.


Weishu 魏書, compiled by Wei Shou 魏收 (Northern Qi). Beijing: Zhong-hua Shuju, 1987.

Xi’an 1997: Xi’an Shi Wenwu Guanliju 西安市文物管理局. “Xi’an Tang Jinxiang xianzhu mu qingli jianbao 西安唐金鄉縣主墓清理簡報 (Preliminary report on the excavation of the tomb of the Tang Princess Jinxiang in Xi’an).” Wenwu 文物 1997.1: 4–19.

Xia Nai 夏鼐 (1983). “Handai de yuqi—Handai yuqi zhong chuantong de yanxu he bianhua 漢代的玉器—漢代玉器中傳統的延續和變化 (Han Jades: Confirmation and transformation of tradition).” Kaogu xuebao 考古學報 1983.2: 124–45.

Xinjiang 1997: Xinjiang Wenwu Kaogu Yanjiusuo 新疆文物考古研究所 and Tulufan Diqu Bowuguan 吐魯番地區博物館. “Shanshan Xian Subeixi muqun san hao mudi 鄯善縣蘇貝希墓群三號墓地 (Cemetery no. 3 at Subeshi in Shanshan County).” In Wang Binghua 王炳華 and Du Gencheng 杜根成 (editors), Xinjiang wenwu kaogu xin shouhuo (xu) 1990–1996 新疆文物考古新收穫 (續) 1990–1996 (New archaeological achievements from Xinjiang 1990–1996), pp. 150–70. Urumchi: Xinjiang Meishu Sheying Chubanshe.

Xinjiang 1998a: Xinjiang Bowuguan Wenwudui 新疆博物館文物隊 (1998a). “Jumo Xian Zhagunluke wu zuo muzang fajue baogao 且末縣扎滾魯克五座墓葬發掘報告 (Report on the excavation of five tombs at the Zaghunluk site in Jumo County).” Xinjiang wenwu 新疆文物 1998.2: 2–18.

Xinjiang 1998b: Xinjiang Wenwu Kaogu Yanjiusuo (1998b) “95 nian Minfeng Niya yizhi yihao mudi chuanguanmu fajue jianbao 95 年民豐尼雅一號墓地船棺墓發掘簡報 (Preliminary report on the 1995 excavation of Niya site I at Minfeng).” Xinjiang wenwu 新疆文物 1998.2: 21–34.

Xinjiang 1998c: Xinjiang Bowuguan 新疆博物館 et al. (1998c). “Xin-jiang Jumo Zhagunluke yihao mudi 新疆且末扎滾魯克一號墓地 (Graveyard no. 1 at the Zaghunluk site in Jumo, Xinjiang).” Xinjiang wenwu 新疆文物 1998.4: 1–53.

Xinjiang 1999: Xinjiang Wenwu Kaogu Yanjiusuo (1999) “Xinjiang Yuli Xian Yingpan mudi 15 hao fajue jianbao 新疆尉犁縣營盤墓地15號發掘簡報 (Excavation of tomb no. 15 of the Yingpan site in Yuli County, Xinjiang).” Wenwu 文物 1999.1: 4–16.

Xinjiang 2000: Xinjiang Wenwu Kaogu Yanjiusuo (2000). “Xinjiang Minfeng Xian Niya yizhi 95MN1 hao mudi M8 fajue jianbao 新疆民豐縣尼雅遺址 95MN1 號墓地 M8 發掘簡報 (Preliminary report on the excavation of M8 in Cemetery 95MN1 of the Niya site in Minfeng County, Xinjiang).” Wenwu 文物 2000.1: 4–40.

Xinjiang 2001: Xinjiang Weiwu’er Zizhiqu Bowuguan 新疆維吾爾自治區博物館 and Xinjiang Wenwu Kaogu Yanjiusuo 新疆文物考古研究所 (editors). Zhongguo Xinjiang Shanpula—Gudai Yutian wenming de jieshi

70 SHING MÜLLER

yu yanjiu 中國新疆山普拉—古代于闐文明的揭示與研究 (Sampula in Xinjiang of China—Revelation and study of the ancient Khotan civilization). Urumchi: Xinjiang Renmin Chubanshe.

Xinjiang 2002a: Xinjiang Wenwu Kaogu Yanjiusuo 新疆文物考古研究所 (2002a). “Xinjiang Yuli xian Yingpan mudi 1995 nian fajue jianbao 新疆尉犁縣營盤墓地 1995 年發掘簡報 (Preliminary report on the 1995 excavation of the Yingpan cemetery in Yuli County, Xinjiang).” Wenwu 文物 2002.6: 4–45.

Xinjiang 2002b: Xinjiang Wenwu Kaogu Yanjiusuo 新疆文物考古研究所, Tulufan Diqu Bowuguan 吐魯番地區博物館 (2002b). “Xinjiang Shanshan Xian Subeixi yizhi ji mudi 新疆鄯善縣蘇貝希遺址及墓地 (The site and cemeteries at Subeixi in Shanshan County).” Kaogu 考古 2002.6: 42–57.

Xinjiang 2002c: Xinjiang Wenwu Kaogu Yanjiusuo 新疆文物考古研究所 (2002c). “Xinjiang Yuli Yingpan mudi 1999 nian fajue jianbao 新疆尉犁營盤墓地1999年發掘簡報 (Preliminary report on the 1999 excavation of the Yingpan Cemetery in Yuli County, Xinjiang).” Kaogu 考古 2002.6: 58–74.

Xu Diankui 徐殿魁 (1996). “Henan Yanshi Shi Xingyuancun Tang mu de fajue 河南偃師市杏園村唐墓的發掘 (Excavation of Tang tombs at Xingyuancun, Yanshi, Henan).” Kaogu 考古 1996.12: 1–24.

Yao Weiyuan 姚薇元 (1962). Beichao huxing kao 北朝胡姓考 (Foreign names of the Northern Dynasties). Beijing: Zhonghua Shuju, 1962.

Yili zhengyi 儀禮正義, commentary by Hu Peihui 胡培翬 (Qing). Sibu beiyao 四部備要 edition, Shanghai: Zhonghua Shuju, 1936.

Yuanhe xingzuan 元和姓纂 , compiled by Lin Bao 林寶 (Tang), with commentary by Cen Zhongmian 岑仲勉. Beijing: Zhonghua Shuju, 1994.

Yun Anzhi 員安志 (1993). “Shaanxi Chang’an Xian Nanliwangcun yu Xianyang feijichang chutu daliang Sui Tang zhengui wenwu 陝西長安南里王村與咸陽飛機場出土大量隋唐珍貴文物 (Sui and Tang treasures excavated from the Xianyang Airport and Nanli Wangcun in Chang’an County, Shaanxi).” Kaogu yu wenwu 考古與文物 1993.6: 45–52, 24.

Zhang Zhengling 張正嶺 (1957). “Xi’an Hansenzhai Tang mu qingli ji 西安韓森寨唐墓清理記 (Excavation of a Tang tomb at Hansenzhai, Xi’an).” Kaogu 考古 1957.5: 57–62.

Zhao Chao 趙超 (1992). Han Wei Nanbeichao muzhi huibian 漢魏南北朝墓誌匯編 (Collected epitaph inscriptions of the Han, Wei, and Southern and Northern Dynasties). Tianjin: Tianjin Guji Chubanshe.

Zhongguo 1980: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所. Mancheng Han mu fajue baogao 滿城漢墓發掘報告 (Report on the excavation of the Han tombs at Mancheng). Beijing: Wenwu Chubanshe.


Zhongguo Xinjiang 1997: Zhongguo Shehui Kexueyuan Kaogu Yan-jiusuo Xinjiangdui 中國社會科學院考古研究所新疆隊, Xinjiang Bayin’guoleng Menggu Zizhizhou Wenguansuo 新疆巴音郭楞蒙古自治州文管所 (1997). “Xinjiang Jumo Xian Jiawaairike mudi de fajue 新疆且末縣加瓦艾日克墓地的發掘 (Excavation of the cemetery at Jawaairike in the Jumo County, Xinjiang).” Kaogu 考古 1997.9: 21–32.

Zou Jingbi 鄒景璧 (1955). “Xi’an Renjiakou M229 hao Bei Wei mu qingli jianbao 西安任家口M229號北魏墓清理簡報 (Preliminary re-port on the excavation of the Northern Wei tomb M229 at Renjiakou, Xi’an).” Wenwu cankao ziliao 文物參考資料 1955.12: 59–65.

Adress:

Department für Asienstudien, Institut für SinologieLudwig-Maximilans-Universität, MünchenKaulbachstraße 51a80539 München, GermanyE-mail: [email protected]


REVIEW ARTICLE

The Emergence of Japanese Kingship. By Joan R. Piggott. Stanford: Stan-ford University Press, 1997. 434 pp.

Reviewed by Donald F. McCallum

This book, the most comprehensive treatment of early Japanese history yet to appear in English, is required reading for all scholars and students concerned with the formative time span comprising the Yayoi 弥生, Kofun 古墳, Asuka 飛鳥, Hakuhō 白鳳, and Nara 奈良periods. Joan Piggott has mastered a very extensive secondary literature in modern Japanese, carefully studied the primary sources and archaeological material, and read deeply in theoretical discussions dealing with the “origin of the state,” broadly conceived. More than just summarizing the conclusions of our Japanese colleagues, the author consistently advances her own hypotheses and interpretations. Occasionally these may seem daring; in any case this review will reflect a more skeptical point of view. In the “Introduction” we are informed that the author’s methodol-ogy is what she terms an “archaeology of kingship” (p. 3). Extending this metaphor, Piggott tells us that she will “cut temporal cross sections” (trenches) through seven historical epochs. These seven, covering some six centuries, are the periods of: (1) Himiko 卑彌呼 (third century), (2) Yūryaku 雄略 (fifth century), (3) Suiko 推古 (late sixth century–early seventh century), (4) Tenji 天智 (ca. third quarter of the seventh century), (5) Tenmu 天武 and Jitō 持統 (fourth quarter of the seventh century), (6) the period of the development of the Ritsuryō 律令 system (early eighth century), and (7) Shōmu 聖武 (second quarter of the eighth century), with a chapter devoted to each. In a sense, the work consists of seven separate monographs, although considerable material is included in individual chapters to bridge the gaps lying between the main areas of focus. Much of the “Introduction” is devoted to a presentation of a variety of theoretical perspectives, which the author proposes to incorporate into her project, including the work of numerous scholars not dealing specifically with Japan. This material is difficult to assimilate, but many

74 Piggott’s Emergence of Japanese Kingship

readers will undoubtedly find it useful to see it applied to the Japanese case. Naturally, Piggott also acknowledges and assesses the work of scholars dealing directly with early Japan, both foreign and Japanese. Piggott offers a stimulating assessment of the nature of the “mon-arch” in early Japan; from the outset she commendably eschews the inappropriate “emperor” as a translation of the term tennō 天皇 (p. 8). In dealing with the transition from chieftain to “Great King” (Ōkimi 大王), she focuses on the sacral and diplomatic dimensions of rule, stat-ing: “Kings were apical rather than autocratic, and their major focus was creation and maintenance of a hierarchy of status and prestige over which they officiated as ‘king of kings’” (p. 6). Much of the book is devoted to developing the implications of this insight, probing the sacral status of the paramounts and a major shift Piggott sees occurring around 600. More specifically, the author tells us that she has developed a model which she calls “Japan as a secondary state formation within the Chinese sphere of influence,” and directs our attention to Fig. 1, “A logical repre-sentation of Japanese state origins.” This diagram is rather complicated, and while it may be helpful in clarifying her approach, many readers will be struck by the absence of Three Kingdoms Korea. Although the text refers to Korea from time to time, much greater emphasis is placed on China, the Chinese monarchy, and Chinese ideology throughout the book. While no scholar would deny the deep influence of China on both the Korean peninsula and the Japanese islands, I believe this study does not pay adequate attention to the northeast Asian cultural sphere of the early first millennium AD that was so crucial for develop-ments in both the peninsula and the islands. Instead, what it espouses is a rather traditional, Sino-centric approach, somewhat surprisingly considering that Japanese specialists are just now striving for a more balanced assessment of peninsular/insular relations. Piggott concludes the “Introduction” by stating: “Japanese colleagues are bemused by the scope of this book. Why, they ask, should a single scholar undertake a study spanning six centuries?” (p. 13). Although Pig-gott is certainly courageous in tackling such a long period, one fraught with difficult problems, I am not entirely convinced that her approach is the most satisfactory. A narrative history, summarizing what is known about the period, would be a significant contribution; similarly, an in depth monograph on any of the seven chapter-topics would be very welcome. The problem, of course, is whether it is possible to combine the two approaches in one book at this stage of research. The account begins in Chapter 1, “Himiko, Paramount of Wa,” with the Yayoi 弥生 period, here dated as 300 BC to AD 300, but now often as 400 BC to AD 250 (in all cases, circa). Given the author’s aims,

Donald F. McCallum 75

no consideration is needed concerning the transition or transformation from Jōmon 縄文 (ca. 10,000 to 400 BC), but there is also no general discussion of the characteristics of the Yayoi period, a period which might be characterized as the “Origins of Japanese Culture,” defined by the introduction of rice agriculture and metallurgy (bronze and iron), and probably of the early forms of the Japanese language as well as an ethnic group related to the historic population of the islands. Since the focus is on the years surrounding the putative rule of Himiko (mid third century), the alternatives for the end of Yayoi (ca. AD 250 or 300) are of crucial importance, for while the latter would place her as a purely Yayoi figure, the former would situate her at the transition from Yayoi to Kofun 古墳; in fact, the latter alternative would seem quite congru-ent with Piggott’s viewpoint. Counterfactual historical narrative is of dubious value, but one can-not help wondering what would happen if the Chinese sources on the Japanese islands during the third century did not exist. The names of (Queen) Himiko and (the country) Yamataikoku 邪馬台国 have so dominated research that it might be salutary to draw back from them for a moment and treat the third century in a manner similar to how one treats the second or fourth, for which no Chinese records exist. This strategy would remove from our project a discouragingly extensive body of publication, material ranging from scholarship at the highest level through general presentations down to depths as low as the proposi-tion that Yamatai was located in Egypt. (To the best of my knowledge, Himiko has not [yet] been equated with Cleopatra!) Such a strategy, while perhaps appealing to some archaeologists and art historians, would probably be of limited interest to historians since, for obvious reasons, their very practice is usually rooted in documentary sources. Himiko and Yamatai appear in unimpeachable Chinese histories, and thus must be dealt with, but we should be conscious of the degree to which their presence has overdetermined scholarly discourse. In this chapter, Piggott is basically concerned with the development of political, economic, religious, and cultural forms at the beginning of the proto-historic period. Clearly, there was a significant shift from the Jōmon hunter-gatherer life-style to that of the Yayoi period; how rapidly this shift took place is a difficult question, but it is apparent that fundamental changes occurred in the course of the fourth century BC.The author devotes most of her energy to an examination of the development of chieftainship in early Japan, marshalling both theoreti-cal approaches and comparative data. Many formulations have been advanced, ranging from “tribal alliance” to “early state” (p. 19), and in order to better understand such institutional structures, Piggott probes a variety of possibilities, especially Sherry Ortner’s idea of “prestige


societies” (Ortner 1981). In this regard, the analysis of the important Yayoi site at Yoshinogari 吉野ケ里 (Saga 佐賀 Prefecture) provides extensive archaeological data against which the theoretical concep-tualizations can be tested. Since Yoshinogari was fortified, one can be certain that warfare to some degree was prevalent at this time, an opinion validated by numerous fortified sites elsewhere in Kyūshū 九州 and in Honshū 本州. While Piggott connects this warfare to the years prior to the “reign” of Himiko, as described in the “Woren zhuan” 倭人傳 section of Wei zhi 魏志, she follows the Chinese text in suggesting that Himiko’s period was generally peaceful. While there is no space to test this hypothesis in detail here, it is evident that we should carefully examine the archaeological evidence without any preconceived ideas based on the “Woren zhuan” account; I personally suspect that the third century, like following centuries, was somewhat less peaceful than Piggott suggests. As a precaution against distorting the author’s viewpoint, a direct quote might be useful:

By Himiko’s time peace had been secured among most of the warring polities by the elevation of a paramount chief responsible for coordinating the common affairs of several hamlet hierarchies (p. 23).

Piggott accepts, then, a quite literal application of the “Woren zhuan” account of third-century political organization. To her credit, however, she declines to see Himiko presiding over a large-scale “state,” and she points to similar chieftains elsewhere in northern Kyūshū, in Kibi 吉備 (Okayama 岡山 Prefecture), in the Kinai 畿内 region, and along the East Sea/Japan Sea. In a certain sense, Himiko and Yamatai constitute one paradigm for third century chieftains throughout Japan, and thus there is no problem in seeing the two as instances of broader phenomena. Problematical is the presentation of Himiko as Queen (political power) and Enchantress (religious role), for this viewpoint necessitates direct acceptance of the “Woren zhuan” account. And even if this account is applicable to Himiko and Yamatai, to what degree does it also apply to all the other “countries” (kuni 国) of the period? A matter of compelling interest in the study of ancient Japan is, to use Piggott’s terminology, “international diplomacy” between Wa and Han 漢 China (p. 25). Perhaps some care is required here, lest we fall into the trap of conceiving of this early period in terms more appropri-ate to later centuries. Although the history of Han China is quite well charted, the same cannot be said for the peninsula and the islands; for that reason, interpretations of two Han Dynasty objects—a gold seal (dated to AD 57, based on an entry in Hou Hanshu) with the inscription 漢委奴國王 found on Shikanoshima 志賀島 in 1784, and a sword (dated


to AD 184–189) excavated at the site of Tōdaijiyama 東大寺山, in Tenri City, Nara— are notoriously difficult, especially in terms of reciprocal diplomatic relations. Furthermore, careful attention must always be paid to the mediating role of the Chinese colonies on the Korean peninsula, rather than simply assuming a direct Han-Wa connection. All of this serves as background for our understanding of the “inter-national diplomacy” seen in the “Woren zhuan” narrative. If Himiko presided over a relatively small political unit, why did she and not others establish connections with Han? (Of course, if Himiko ruled a larger “state,” including northern Kyūshū and the Kinai, this question would be unnecessary, but this is not Piggott’s position, nor mine.) Might the Wei zhi account tell us more about the ideological needs of the Chinese empire than it does about the political organization of third-century Wa? Is it not possible that the Chinese court would have been quite satisfied with a mission from the “king” (王) of any insular country (国)? Here I am simply suggesting that we be very careful in interpreting the Chinese data in the insular context. (Although from a much later period, I am reminded of the enthusiastic reception by the Song 宋 court of the Heian 平安 priest, Chōnen 奝然, who happened to arrive in the capital near the beginning of the Song dynasty, thus fulfilling the dynasty’s need to have such a visitor.) Much of the meat of Piggott’s narrative is found in her account of the development of tombs in early Japan; naturally, this account be-gins from the “Woren zhuan” account of Himiko’s large tomb, said to be a hundred paces in diameter (p. 28). Too ready acceptance of this measurement must be tempered in view of the further assertion that more than one hundred male and female attendants were buried with her; while this practice is known in China, there is no evidence for hu-man sacrifice in early Japan. Is Chen Shou 陳壽, the author of the Wei zhi, perhaps telling his readers what they might expect of a barbarous monarch in a distant land? Nevertheless, there is growing archaeologi-cal evidence for an increase in tomb sizes during the third century, so there is no a priori reason why Himiko –and her peers– could not have been interred in substantial tombs. The chronology of insular tombs is extremely complex, and Piggott provides a useful analysis of the developments leading up to the large-scale tombs of the third century. I personally would have liked to see more on contemporaneous peninsular developments, but there is only so much space, and one cannot do everything. Since her book appeared, the placement of the Hokenoyama Kofun ホケノ山古墳 tomb in the mid-third century indicates an earlier inception for the Kofun period than previously thought (Kashihara 2001) . This revised chronology strongly supports some of the author’s basic arguments.


Considerable attention is paid to the famous Hashihaka Kofun 箸墓古墳 in the Miwa 三輪 area of Nara 奈良 Prefecture, with the author suggesting that its size (275 m. in length) may signify “resources on a supraregional scale” (p. 29); while this is certainly plausible, the further suggestion that such tombs reflect practices recorded in the Book of Rites (Li ji 禮記) seems less so. In exploring the idea of a “supraregional league of chieftains,” Piggott points out the wide distribution of tombs of related configuration, and focuses particularly on the later third cen-tury Gōdo Yongō Kofun 神門 4 号古墳 (Chiba 千葉 Prefecture) as a smaller keyhole tomb which may reflect center/periphery connections. Although she is understandably cautious in putting forth strong claims about the degree of subordination reflected by tombs such as Gōdo Yongō, Piggott suggests that by the end of the fourth century both the shapes and grave deposits of these regional Kofun do express recognition of the Miwa paramounts. (Whether there were other “paramounts” in the Nara basin cannot be taken up here.) In this section, the author is striving to close the gap between the Himiko/Yamatai third century and the more fully defined fifth century, which she discusses in the next chapter. To this end, there is considerable discussion of both the shapes and sizes of fourth-century tombs throughout the islands, and Piggott suggests that these relate directly to status hierarchy, especially between outlying areas and the Yamato core. Japanese scholars readily assume that shape and size express a superior/subordinate relationship, but this should be tested in further research; while Piggott partially endorses this interpretation, she presents a more nuanced analysis based, in part, on her study of Izumo 出雲. A working hypothesis willing to see consider-able independence in the various regions certainly seems desirable; after all, if it does not withstand forceful critiques, it can be thrown out and we can then all cheerfully accept a Yamato-centric approach. No problem of third-century historical archaeology is of greater signif-icance—and complexity—than the mirrors found in tombs throughout the islands (pp. 31, 34). These mirrors, especially the “triangular ridge, deity and animal” (sankakubuchi shinjūkyō 三角縁神獣鏡) category, appear to have been made in China, so it is only natural to associate them with the gift of the Wei ruler to Himiko in 239. A further assumption is that Himiko’s court then distributed the mirrors to allies in areas ranging from northern Kyūshū to the Kantō 関東 plain. The most compelling aspect of this scenario is the existence of duplicates, made from the same mold, discovered in widely separated locales. This hypothesis, largely developed by Kobayashi Yukio 小林行雄 on the basis of research on some 32 examples found in the Tsubai Ōtsukayama Kofun 椿井大塚古墳 (Kyōto-fu 京都府, but very close to the border with Nara), has been reinforced by the relatively recent discovery of more than 30 “triangular


ridge, deity and animal” mirrors in the Kurozuka Kofun 黒塚古墳 near Tenri 天理 (Nara Prefecture) (Kashihara 1999), which tends to reinforce some of Piggott’s basic arguments. (Perhaps even more im-portant for her argument is the discovery of a single mirror of this type in the Hokenoyama Kofun, cited above, which presumably dates very close to Himiko’s lifetime [Kashihara 2001, pls. 35–39].) Nevertheless, it remains a problem whether the “triangle-ridge, deity and animal” mirrors can be taken as definitive proof that Yamatai was in Yamato. Particularly puzzling is the fact that apparently identical examples of this type of mirror have not yet been found in China, even though the names, dates, and iconography are all Chinese. Where they were made, by whom, and why are still largely open questions. Chapter 2, “Great King Yūryaku,” contains much interesting mate-rial about the mid-fifth century, although the relatively limited treatment in this book of the fourth century produces somewhat of a gap. The fact that there are no Chinese sources from Himiko/Yamatai until the “Five Kings of Wa” (倭之五王) mentioned in the historical records of the Liu Song 劉宋 dynasty (420–479), makes it difficult for the historian to establish a documentary foundation for the crucial period from ca. 280–420. One must hope that archaeologists will someday be able to trace the development of state formation during this time. While numerous problems remain in associating the Chinese names of the “Five Kings of Wa” with specific Yamato monarchs, it is clear that there was important diplomatic interchange between a polity in Japan and the Liu Song dynasty of southern China. Piggott has little to say about the first four (San 賛, Chin 珍, Sei 済, and Kō 興), instead concentrating on the fifth, Bu 武, who is generally thought to have been the monarch later called Yūryaku. Basically, she sees a transformation from Himiko the enchantress to Yūryaku the warrior. Fundamental to the study of this period is the assessment of the large tombs (kofun) found throughout the islands, for they certainly manifest the great authority of the individuals buried in them. To this end, Piggott suggests that the tombs must be interpreted in terms of size, distribution, shape, and burial goods (if known). Very complex problems arise in this interpretive project, since it is not even clear if the center of political authority was in Kawachi 河内 (modern Ōsaka 大阪 Prefecture) or in the Nara basin; moreover, there are also very large tombs in Okayama, Ibaragi 茨木, Tottori 鳥取, and Kyūshū. Broadly speaking there are two basic interpretations: the first argues that large tombs in Osaka and Nara are those of rulers who exercised significant authority over the chieftains buried in tombs in the other areas; the second sees the occupants of all (or most) of the large tombs as largely independent figures not subservient to a central authority. There is


evidence for both positions, although on balance Piggott seems to lean toward the second, acknowledging that “ . . . chiefly hierarchies across the archipelago continued to display distinct signs of autonomy” (p. 54). The inscribed swords from the Inariyama Kofun 稲荷山古墳 and the Eta Funayama Kofun 江田船山古墳 are a second source of crucial data. The former, a relatively recent discovery, has an inscription identifying its owner, Owake no omi 乎獲居臣, who served Great King Wakatak-eru 獲加多支歯大王, thought to be Yūryaku; the date, equivalent to 471, is within the reign of that ruler. The other sword also refers to Great King Wakatakeru and the owner is identified as Murite 旡利弓, an official who served the Great King. Since the Inariyama sword was excavated in Saitama 埼玉 Prefecture and the Eta Funayama sword in Kumamoto 熊本 Prefecture, it appears that the Great King Wakatakeru had vassals throughout the islands who protected his interests. Piggott suggests that men such as Owake no omi and Murite received lavish gifts in recognition of their services. Although further research is required, it certainly looks like we are seeing here a nationwide network of a type not apparent earlier. Another aspect of the political organization of the fifth century is the status of titles such as “country chieftain” (kuni no miyatsuko 國造) and occupational group manager” (tomo no miyatsuko 伴造) and the sig-nificance of social units like “lineage” or “clan” (uji 氏), “occupational group”(be 部), and various “titles of rank” (kabane 姓). The early formu-lation of such titles and groups would indicate an early consolidation of royal power, although at present it is difficult to determine when the terms actually came into use. Piggott investigates the contributions of “immigrants” to the establishment of Yūryaku’s polity, suggesting that they worked directly for the monarch in a variety of roles, including various crafts, writing, and management. While this may have been par-tially the case, I have long believed that the “immigrants” were a good deal more autonomous that the orthodox Japanese viewpoint holds; it seems likely that as bearers of advanced continental civilization they were members of the elite rather than occupying a subordinate position in society. In examining the sources of Yūryaku’s power, Piggott points to four factors: (1) control of diplomacy, (2) dispatch of mercenary forces to the peninsula, (3) the iron trade, and (4) marital alliances. The first is confirmed by the Chinese dynastic histories and seems likely. However, great debates have raged on the second factor, since this relates to the whole problem of peninsular/insular relations during the fifth century. A previous generation of Japanese historians saw the insular control of “Mimana” 任那 as direct evidence for a powerful state in the islands,


but more recent research has demonstrated that the situation is a good deal more complex. Certainly the iron trade was a key factor in peninsular/insular relations, especially between the Kaya 伽耶 region and Yamato 大和 or Wa 倭. Lastly, quite a lot of evidence exists for Yūryaku’s consorts, who hailed from various places, and Piggott sug-gests that these marriages led to mutually beneficial relations between center and periphery. Yūryaku was characterized in Nihon shoki 日本書紀 ( compiled AD 720) as a “wicked ruler,” apparently based on the constant strife between him and other powerful leaders that resulted in considerable bloodshed. The degree to which Yūryaku succeeded in unifying his domain is uncertain, although it is evident that other powerful clans and provincial rulers retained a great deal of power. Piggott makes it very clear that whatever polity Yūryaku had forged soon collapsed after his death (479?), leading to a period of disunity. Chapter 3, “Suiko, Heavenly Heir and Polestar Monarch,” covers the period from the late fifth century until the end of Suiko’s reign in the 620s. Out of the chaos following Yūryaku’s death, a time when various individuals and groups were contending for power, emerged a new monarch, Great King Keitai 継体大王, who was active during the first decades of the sixth century. Keitai, an individual born outside of the central region, was supported by provincial chieftains, especially of the Okinaga clan 息長氏. Although efforts were made by the compilers of Nihon shoki to provide Keitai with an imperial lineage, one assumes these efforts were motivated by the ideological necessity of having a single “imperial” line; in actuality, it seems unlikely that Keitai really had any specific connections with his predecessors on the “throne.” More-over, as Piggott points out, there were other contenders for supremacy throughout the islands, so that if one of these had prevailed he or she would have constituted the royal lineage. In what sense was Keitai the Great King? One’s answer to this ques-tion fundamentally determines one’s general conceptualization of this phase of Japanese history. In my opinion, there were a number of clans and individuals contending for power, and it was only later that certain of these contenders were retrospectively claimed to be “Great Kings,” primarily in order to preserve the fiction of an unbroken imperial line. Following the usual pattern, Keitai was succeeded by two mysterious figures, Ankan 安閑 and Senka 宣化, about whom virtually nothing is known, and then by Kinmei 欽明, a crucial figure for standard accounts of early Japanese history, especially in the context of the introduction of Buddhism. Piggott offers the dates 539–571 for his reign; while the final year seems reasonably accurate, there is considerable uncertainty about the beginning year.


I believe that it is possible to conceptualize this period in quite differ-ent terms, focusing more on the Soga 蘇我 clan than on the “imperial” line. Piggott consistently sees the Soga as subservient to the Great King. But it may well have been, in fact, the Soga who were the dominant political and economic faction from the mid-sixth until the mid-seventh century. Soga no Iname 稲目 (d. 570?), whom Piggott has serving Great King Kinmei, was the central leader in Kinmei’s time, and was respon-sible for the bringing of new cultural forms from the Korean peninsula, primarily from Paekche 百済. We are told that Kinmei had two Soga wives, daughters of Iname, which is not surprising, since we are sup-posed to see the Soga as offering their daughters to the emperor in the same manner that the Fujiwara did in later centuries. More likely, in my view, is a reciprocal exchange of daughters among the various great families; the reason why we know that Iname’s daughters, Kitashi hime 堅塩媛 and Oane no kimi 小姉君, became royal consorts of Kinmei is that it was necessary to construct a royal lineage connecting with that of the later seventh century. Piggott assumes literacy and the study of Chinese texts at a very early time in Japanese history. In tracing what she sees as new rituals at Kinmei’s court, texts such as the Book of Rites 禮記, the Rites of Zhou 周禮, and Sima Qian’s 司馬遷 Shi ji 史記 (Records of the Historian) are cited as “no doubt” contributing to the new developments (p. 75). But I wonder if a Nihon shoki entry for 513 stating that a master of Chinese classics came from Paekche is adequate evidence to support these claims? Although Piggott postulates “significant progress toward royal sta-bility and reproducibility” (p. 78) in the forging of Kinmei’s polity, she acknowledges that as usual there was a succession dispute following his demise. Three factors are brought forward to explain this: (1) a changing situation in foreign policy; (2) the Soga-Mononobe 物部 conflict; and (3) an excess of heirs. With regard to the first factor, it is obvious that the reunification of China under the Sui 隋 and the taking of Kaya by Silla 新羅 had an impact on the Yamato state. The second factor is far more problematic, as Piggott conceives of the triumph of the Soga over the Mononobe in 587 as leading to “Soga ambitions [which] were unchecked” (p. 78). This interpretation emerges out of the orthodox view that somehow the Soga were interlopers, challenging the legiti-mate rulers rather than being, as I believe, a comparable family with great economic and political power. Concerning (3), Piggott argues that because of the serious problems with the succession a compromise was reached with the enthronement of Suiko as Great King. Piggott’s discussion of “Suiko, Heavenly Heir and Polestar Monarch,” is internally consistent; unfortunately, however, in my view some of her interpretations are incorrect or, at the very least, misleading. Neverthe-


less, I would like to stress that if one accepts her premises, the conclusions normally follow in a logical manner. All of this places a heavy burden on the reader of this review, as he or she is now required to weigh assertions suggesting radically different conceptions of Asuka 飛鳥 period Japan. The author argues that Suiko played a significant role in government, and criticizes those scholars (eg. George Sansom) who, in her view, underestimate Suiko’s position while overstressing the position of the Soga. As should be clear from my previous comments, I believe that the Soga leader during Suiko’s reign, Soga no Umako 馬子, was the central figure of the time and Suiko more of a shaman than ruler. The crux of the problem lies in the conceptualization of Prince Stable Door (Umayado no ōji 厩戸皇子 – also known as Shōtoku Taishi 聖徳太子). The image of Stable Door, as is recognized by practically all historians, was very much mythologized in the years following his death, leading to claims that he could speak at birth, settle ten suits simultaneously, and predict the future. If this is dismissed as legend, why is it that so many other extravagant assertions are believed? Can one accept statements that Stable Door formulated the Cap Rank system (in 603), wrote the Seventeen Articles (604), and produced lengthy commentaries on three Buddhist sutras written in Chinese? In fact, what is the evidence for Stable Door’s literacy? Ōyama Seiichi 大山誠一 (1999) has proposed that Stable Door is a fictitious character. Although it is not necessary to go that far, since presumably there was a princeling of that name, it is important to analyze the reasons why the compilers of Nihon shoki felt it necessary to create a “Crown Prince of Saintly Virtue” (= Shōtoku Taishi). Stated briefly, the motivation seems to have been the necessity of transferring the accomplishments in government, culture, and religion from the Soga, where they rightly belonged, to a member of the “imperial” line. Stable Door got the credit and lives on in Japan as a combination of Constantine the Great and Asoka. Piggott apparently accepts the traditional account of Stable Door’s career, asserting that he was “co-ruler” with his aunt, Suiko (p. 82). Earlier I traced the activities of Soga no Iname and here it should be noted that Iname’s son, Umako, inherited his father’s position and was perhaps even more powerful. For example, Japan’s first full-scale temple, Asukadera 飛鳥寺, a lavish palace, and a great tomb can all be associated with Umako, indicating that he had the wealth and authority to build on the most extravagant level. Suiko was said to have lived in the Toyura Palace 豊浦宮 for the first decade of her reign; this palace, located on Soga land, has been archaeologically discovered below the later temple, Toyuradera 豊浦寺. In 603 she moved to Oharida Palace 小墾田宮, also on Soga land.


Piggott wrongly states that this palace occupied the same site as Tenmu’s Kiyomihara Palace 浄御原宮; current thinking suggests that it was to the north of that palace, on the south-eastern slope of Ikazuchi Hill 雷岡, at a place where pottery inscribed with the characters “Oharida” 小治田宮 (variant character) was discovered (Asuka 2002, p. 45, fig. 46). Earlier, it was believed that the location was on the other (west) side of the Asuka river, quite close to the Toyura Palace. Piggott is determined to make a strong case for Suiko’s support of Buddhism, and in that effort she utilizes a variety of texts. Unfortunately, at times her usage of these texts is somewhat uncritical. Let us look carefully at one instance, her citation of Gangōji engi 元興寺縁起 (AD 747) as evidence for Suiko’s support of Buddhism prior to the time she became “Great King.” Piggott states:

If we accept the narrative presented in the Gangōji Chronicle, Suiko and Prince Shōtoku were strong proponents of Buddhism long before Suiko became Great King (p. 93).

As historians, we must first determine if there are grounds for what Piggott refers to as accepting the “narrative presented in the Gangōji Chronicle.” The analysis of this text is one of the most complex problems in early Japanese history and although she cites Fukuyama Toshio’s 福山敏男 fundamental article of 1934, she does not seem to have completely grasped Fukuyama’s argument. In her note 104 (p. 351), Piggott writes:

Fukuyama concluded that the Gangōji text combines the historical materials of more than one temple; that conclusion does not invalidate the sections of the text that witness Suiko’s energetic patronage of Buddhism (italics added).

At the very least, Piggott’s understanding that Gangōji engi as a composite text should have made her realize that it does not make sense to speak of accepting its narrative in general terms; instead, one must analyze each section of the text individually, weighing the degree to which a specific section appears to be plausible within the broader context of early Japanese history. I stress this here in order to cast further light on Piggott’s next sentence:

Researchers value this temple chronicle highly because it contains many detailed anecdotes concerning the early history of Buddhism in sixth-century Yamato not included in Nihon shoki (p. 93).

Unidentified “researchers” apparently “value” the text “highly,” but we are given no indication that there may be problems with the text, nor are we told of other researchers who may have serious reservations about the reliability of certain aspects of Gangōji engi. In fact, there is


significant slippage from one sentence to the next that may lead us to accept the following:

We are told, for example, that Suiko and her brother, the prince who ruled as Yōmei, personally fended off rebellious ministers bent on throwing Buddhist texts and icons into the harbor in 570.

Presumably the “for example” indicates this event as just one of a group in Gangōji engi that refer to Suiko’s support of Buddhism prior to the time she is thought to have become monarch. Be that as it may, one assumes that a cautious historian would try to determine what factors underlaid and motivated the construction of this narrative. For example, how old were Suiko and Yōmei 用明 in 570? What, if anything, is known about their activities at this period? Who were the “rebellious ministers”? What was the situation of Buddhism in Japan around 570? Why would the compilers of Gangōji engi want to suggest that Suiko and Yōmei took this heroic action? I suggest that certain answers to these questions might cast consider-able doubt on the veracity of the story. Suiko, hardly more than a girl in 570, seems unlikely to have had the power to personally fend off any “rebellious ministers.” Other details are similarly implausible, and the most logical conclusion is that the 570 account is a myth concocted by the compilers of Gangōji engi once again to transfer credit for the further-ance of Buddhism to Suiko and Yōmei. This very specific detail is followed by a much more general state-ment:

Then during Suiko’s reign construction of temples by royals and courtiers got under way, either as memorials to dead parents or in gratitude for boons received (p. 93).

To what years exactly this sentence refers is not entirely clear, but the general context, and the next sentence, to be discussed in a moment, suggest that it indicates the earlier phase of Suiko’s reign. Two questions might be asked: 1, who were the “royals” who constructed temples?; and 2, what is the evidence that these temples were vowed for the reasons suggested by Piggott? There is, to my knowledge, no convincing evidence for “royals” building temples during the Suiko reign, especially during its early years. With regard to my second question, as to the motivation for the building of these temples, Piggott seems to confront the issue in the next sentence:

An early inscription on an aureole dated 594, the second year of Suiko’s reign, indicates that the offspring of deceased parents ordered the cast-ing of a bronze icon to effect their parent’s rebirth in Amida’s Western Paradise (p. 93).


Piggott here refers to an aureole now included in the “Forty-eight Buddhist Deities” (四十八体仏) group in the Tokyo National Museum, originally from Yamato Province but “donated” to the museum at the end of the nineteenth century by Hōryūji (法隆寺). Can this be consid-ered an adequate analysis of the inscription? As was the case with our discussion of Piggott’s utilization of the Gangōji engi text, here too a certain amount of background material is necessary in assessing the nature of the evidence (Asuka Shiryōkan, pls. pp. 51–56, pp. 148–9, 173–175). To what degree were Buddhist icons made in Japan towards the end of the sixth century? This is a difficult question indeed, but it is unlikely that a great many were produced there at that time (McCallum 2001). In the case of the monument under consideration, the date of 594 is the most likely interpretation of the cyclical characters, but this is a hypothesis; however, there is nothing about the “second year of Suiko.” Most scholars believe that this icon actually was made on the Korean peninsula, and only subsequently brought to Japan. If so, it hardly seems appropriate to use it as a source for insular attitudes. One would also like to know what the evidence is for the worship of Amida 阿弥陀 in late sixth century Japan, especially as the inscription refers to Shaka (釋迦), not Amida. Chapter 4, “Tenji, Fortress Monarch,” covers the period from the assassination of Soga no Iruka 入鹿 (645) until the death of Tenji (671), during which four monarchs were on the throne: Kōgyoku 皇極, Kōtoku 孝徳, Saimei 齊明, and Tenji. Since more than half of the chapter deals with the Kōtoku and Saimei reigns, the chapter title is perhaps not entirely felicitous. Central to Piggott’s argument is her interpretation of the coup against the Soga, whom she sees as opposing the legitimate royal line. Nakatomi Kamatari 中臣鎌足 is the hero of this account as the leader of the anti-Soga coup, although she finds it “curious” that Kamatari then disappears from the record of the fol-lowing major events. It is not clear to me why it is stated that Emishi 蝦夷, the last leader of the main line of the Soga, committed suicide in Hōkōji 法興寺 (=Asukadera), as Nihon shoki states that Prince Naka 中大兄 occupied the temple. The suicide probably took place elsewhere, presumably at the Soga Palace. The extremely interesting detail concerning the alleged attempt by Emishi and his followers to burn the two national histories they possessed is not treated by Piggott (Nihon shoki 645.06.13).1 This wanton act of de-struction certainly relates to the general image of Soga evilness, although a skeptical interpretation might suggest that Fune no Fuhito Esaka’s

1 The format employed here is to give the year in Western form, followed by the month and then the day. (Occasionally the month and/or day is not given in the text.


船史恵尺 heroic rescue from the flames of part of the manuscript was limited to those sections favorable to the royal line. In the Nihon shoki entry for the same day we are told that Emishi and Kuratsukuri 鞍作 (=Iruka) were allowed proper burial and the appropriate mourning rites, hardly the treatment one might expect for such heinous individuals. Piggott offers a quite detailed analysis of the so-called “summer oath” (Nihon shoki 645.06.19) of Kōtoku, Kōgyoku, Prince Naka, and the Great Ministers shortly after the assassination of Emishi, suggesting that its ideology reflects various Chinese classics as well as the Buddhist concept of the “degenerate age” (mappō 末法). While the sources may, indeed, provide the wording, it seems to me more likely that such prose is a product of the eighth century compilation of Nihon shoki than of the year 646. There can be no doubt, however, that significant changes occurred around the middle of the seventh century as a result of the displacement of the main Soga line. Could it not be the case of a family that had been very progressive in the sixth century when it first achieved power becoming increasingly conservative and old-fashioned after ruling for about one hundred years? Be that as it may, the second half of the seventh century is quite different in character from the first. In conventional history, this difference is marked by the Taika Reform (Taika kaishin 大化改新) edict said to have been promulgated on the first day of 646 at the Nagaratoyosaki Palace 長柄豊崎宮 in Naniwa 難波 by the new monarch, Kōtoku. Although Piggott states that there was “a throne hall roofed with tile” (p. 107), this was not the case of the palace during Kōtoku’s reign. Piggott surveys the great scholarly controversy surrounding the Taika Reform, choosing Naoki Kōjirō 直木孝次郎 as representing the positive interpretations and Hara Hidesaburō 原日秀三郎 the negative; I would like to have also seen a citation of Kadowaki Teiji (Kadowaki 1991) here, as he has been one of the principal critics of the traditional posi-tion. Although she acknowledges the force of some of Hara’s critique, Piggott generally tends to accept the historicity of the edict, striving to place it in the context of the anti-Soga coup. The Saimei reign is one of the most interesting of the seventh century. Piggott is concerned first with the development of governing structures, informing us that key positions were occupied by scholars such as the priest Min 僧旻 and Takamuko Genri 高向玄理 after “Abe Kurahashi Maro 阿部倉梯麻呂 and Soga Yamada Ishikawa Maro 蘇我山田石川麻呂 died in 649.” Died hardly captures their demise: the Abe minister died on 649.3.17, probably murdered, and Ishikawa Maro was accused of treason a week later, committing suicide with his family on 649.3.25 at Yamadadera 山田寺. With the decks conveniently cleared it was possible to get on with the “reforms,” but we should not ignore the


accompanying violence. (Equally conveniently, Prince Naka realized a few days after the suicides that Ishikawa Maro was, in fact, innocent!) Saimei returned with Prince Naka to Asuka in 653, assuming the throne for a second time in 655 upon the death of her brother, Kōtoku. Nihon shoki includes a strong denunciation of her extravagant building projects, and excavations in recent years appear to confirm that she undertook such projects. Particularly interesting is a whole complex surrounding the Sakabune Rock (Sakabune ishi 酒船石 ), including a masonry wall around the hill on which the rock stands and, at the base of the hill, the quite extraordinary Turtle Rock (Kame ishi 亀石), appar-ently some type of ritual site (Asuka 2002: 60–63). These happy years ended abruptly in 660 with the fall of Yamato’s ally Paekche and the subsequent move to Kyūshū in the following year by Saimei and her court to supervise a rescue operation. Saimei died that year (661.7), and Prince Naka assumed control. Preparations were made for the invasion and although the records are somewhat contradictory, we know that the Japanese armada was destroyed in 663, thereby ending many years of meddling on the peninsula, until Hideyoshi’s 秀吉 invasions at the end of the sixteenth century. Prince Naka continued in the position of interim regent (shōsei 稱制) from 661 until 668, an unparalleled length of time, prior to ascending the throne. The reasons for this policy are not explored in depth by Piggott and remain somewhat of a mystery, although she does point out that Naka was deeply involved in coastal defense preparations, un-doubtedly easier to carry out as regent than as monarch. In the spring of 667 Naka and his followers moved to Ōtsu 大津, perhaps because it was seen as more easily defended than either Naniwa or Asuka. A grand new capital was constructed, and Naka assumed the royal dignity there as Tenji in 668. Piggott suggests that Ōtsu was an attractive location in part because of the convenience of access to various regions and also because of the nearby large settlements of Korean “immigrants” whose “relationship with the throne was particularly subordinate” (p. 118). As indicated, I would suggest some caution with this sort of characteriza-tion of the “immigrants.” In the few short years at Ōtsu much seems to have been accom-plished in the organization of government; particularly significant was the appointment of Tenji’s eldest son, Prince Ōtomo 大友皇子 to the newly established position of prime minister (Daijōdaijin 太上大臣) in 671. Various ministries were organized and a bureaucratic system put in place to govern the state more effectively. It is possible that Tenji promulgated a legal code, the Ōmi ryō 近江令, although the evidence on this is not clear and the author does not take a strong stand one way or the other.


One of the most striking sections of Piggott’s narrative is her explora-tion of the nature of the Ōtsu court. She mentions that the priest Dōshō 道昭 brought back many texts from China, including material dealing with Taizong 太宗, the founder of the Tang 唐 dynasty. Piggott suggests that Tenji may actually have modeled himself on the Tang emperor. Tenji followed Tang Taizong’s precedent by focusing on civilian arts (bun 文), including establishing a university and encouraging literature and scholarship. Particularly important for Piggott’s analysis is the text Fond Recollections of Poetry (Kaifūsō 懐風藻), from which she quotes a lengthy passage characterizing Tenji and the Ōtsu court. This passage offers a wonderfully idealized image of Tenji and his achievements, although the fact that it was composed after the middle of the eighth century might suggest the need for some caution. In particular, one wants to know why such a characterization suited the needs of a Nara-period author. A similarly idealized view is found in a passage quoted from Kamatari’s biography, and although here Piggott does state that the text is “certainly late and self-interested in composition” she concludes that “this portrait of Tenji and his court is both credible and revealing” (p. 122). In reading this analysis, one might have the feeling of a parallel with Prince Shōtoku, although I would assume that the characterization of Tenji is somewhat more strongly anchored in historical reality than that of Shōtoku. The treatment of Tenji’s relationship with Buddhism begins with a curious phrase: “Even had he wished to, Tenji could not have ignored Buddhism” (p. 122). One wonders what conceivable reason Tenji could have had for ignoring Buddhism. Piggott’s discussion of Buddhism here is not especially revealing. For example, she cites the Asuka Shiryōkan 飛鳥資料館 compilation of inscriptions on Asuka and Hakuhō 白鳳 images as evidence for religious practices of this time, even though only one of these date to the Tenji reign (Asuka Shiryōkan, pls. pp 10, 22–25, 77–79, pp. 155–156, 180–181). (This is the Yachūji Miroku 野中寺彌勒 of 666, actually two years before the official beginning of that reign. On p. 224 the same image is incorrectly referred to as “an Amida from Nonakadera.”) There is confusion as to the Nihon shoki entries concerning Tenji’s final illness, and I was unable to locate the passage describing the “first mass admission of novices to the Buddhist order.” Similarly, the statement that “Hōkōji [= Asukadera] and Kawaradera 川原寺 continued to function as royal temples” after the capital moved to Ōtsu in 667 is not entirely accurate: Asukadera may not have had this status prior to 680, Kawaradera certainly was not complete in 667, and Kudara Ōdera 百濟大寺, definitely a royal temple, is omitted. The situation regarding Sūfukuji 崇福寺 is more complicated than indicated, and an-other key temple, Minami Shiga haiji 南滋賀廢寺 is not mentioned.


With regard to the administration and control of the Buddhist order, the Nihon shoki 645.8.8 entry concerning the achievements of the Soga in furthering Buddhism and Kōtoku’s appointment of the “Ten Teachers” (jisshi 十師) is cited, presumably as background for the measures taken by Tenji. The author asserts that these measures were influenced by those laid down by Tang Taizong in his campaign to control the clergy. Nevertheless, she correctly points out that the status of the Buddhist clergy was higher in Japan than in China, since they had skills not widely available elsewhere. A final issue dealt with in this chapter is the ongoing and vexing problem of the succession. While Tenji’s brother, Prince Ōama 大海人皇子, had a strong claim to the throne, Tenji decided that his son, Prince Ōtomo, should be heir. The outcome was the important Jinshin Disturbance (Jinshin no ran 壬申の乱) in which Ōama triumphed over Ōtomo, gaining the monarchy and reigning as Tenmu. Prior to this war, Ōtomo and his ministers had sworn an oath to follow Tenji’s command (ie, Ōtomo as successor). At this ceremony the Minister of the Left, Soga Akae no omi 蘇我赤兄臣, played an especially important role; Piggott finds “the Soga presence here surprising.” Perhaps throughout the period covered by chapters 3 and 4, it might be profitable to reassess the actual role and status of the Soga. Chapter 5, “Tenmu and Jitō, Stem Dynasts and Divine Kings,” deals with three decades during which fundamental transformations were made in the Yamato polity. The more extensive documentary record allows a somewhat fuller treatment of the important issues than was possible in the previous chapters. Piggott draws attention to (1) the elaboration of the Kiyomihara and Fujiwara 藤原 courts, (2) the first Chinese-style law codes, (3) the nature of the Fujiwara capital, and (4) a new royal iconography of fully divine kingship; all of which developed during the reign of the erstwhile Prince Ōama as Tenmu tennō. The basic events of the Jinshin Disturbance are outlined, with atten-tion paid to the background of Tenmu’s supporters. Piggott sees Tenmu as modeling himself on the founder of the Han dynasty, Emperor Gaozu 高祖, since the latter collaborated with his senior consort, Empress Lü 呂, just as Tenmu did with his consort, Princess Uno 鸕野皇女 (= Jitō tennō). And also like Gaozu, Tenmu had to unify his country after the battle, and he did so by emphasizing the civilian arts. One factor that might warrant further analysis is the unification of the Korean Peninsula under the Unified Silla as a motivation for greater efforts at political organization in Yamato. Certainly the Jinshin Disturbance was a suc-cession dispute, but I would suggest that it, and related developments of these years, cannot be fully understood without greater attention to the Korean situation. Parallels with Tang Taizong and Han Gaozu


are well and good, but there is another level that must be rooted in the northeast cultural sphere. Just as his mother Saimei had moved the capital back from Naniwa to Asuka, so did Tenmu return there from Ōtsu. He resided in the Kiyomihara Palace, but from the beginning of his reign he was plan-ning on building a grand new palace and capital. Piggott sees the Kiyomihara Palace as somewhat grander than it may have been: for example, she states that three “official temples” were located nearby, but fails to understand that Kudara Ōdera, Takechi Ōdera 高市大寺, and Daikandaiji present three steps in the development of a single temple institution. Kudara Ōdera was never on the banks of the Asuka River, but on the Kudara River; when it moved in 673 to Takechi it became Takechi Ōdera, and four years later was renamed Daikandaiji 大官大寺. Where the latter was located is uncertain. (It is important to keep the Kudara Ōdera–Takechi Ōdera–Daikandaiji sequence separate from the second Daikandaiji built by Monmu 文武 around 700.) Piggott stresses the necessity for the Tenmu regime to work closely with supporters in the provinces; as she notes, Tenmu did not have a standing army that could enforce his will, and thus had to employ other means to gain support. These included gifts and titles that recognized the power of the provincial elite while at the same time bringing them into the central administrative structure. Continuing efforts were made to develop an ideology of universal sovereignty to justify the “imperial” rule of Tenmu and his line. Perhaps the most spectacular project of Tenmu’s reign was the conceptualization of a great capital, Fujiwarakyō 藤原京, based on Chinese models. After some false starts work began in the later 670s, primarily in the laying out of a geometric grid of streets and avenues that divided the city into regular wards. This plan may be related to the ideal Chinese capital as described in the Rites of Zhou. Excavations in recent years have indicated that Fujiwarakyō was, quite likely, larger than its successor, Heijōkyō 平城京, and thus was an extraordinarily ambitious project. Archaeological research has also shown the sources for lumber, masonry, and roof tiles, gathered from a rather wide area. Piggott notes that Nihon shoki does not contain entries concerning levies for materials and conscript labor to build city and palace, an important factor for evaluating the nature of that text; if Nihon shoki does not offer details as to the construction of the most expensive and time-consuming project of the century, one must wonder what the compilers’ criteria for inclusion were. Work continued after the death of Tenmu through the early years of Jitō’s reign, with the court finally moving into the palace in 694. A second important project was the production of a formal legal code. Piggott suggests that the culmination of this project, the Kiyomihara


Code of 689, was actually based on policies developed by Tenmu and Jitō through the years of their reigns, rather than being an entirely new formulation. Since the code is now lost, it is difficult to analyze fully, although data can be found in other sources. For instance, information is available concerning the organization of government ministries dur-ing this period, institutions which must correspond with the buildings excavated at the Fujiwara Palace site. Additionally, the court rank system was substantially altered in a manner that enhanced the monarch’s authority. Piggott sees a combination of courtly ceremonial activities and sacred ritual as the key factors in developing a unifying ideology for the new polity; the former consisted of things like festivals associated with the New Year, enthronement ceremonies, diplomatic receptions, and the like. The utilization of the name “Nihon” 日本 for the country and “Tennō” (“Heavenly Sovereign”) for the monarch also date from this period; Piggott suggests that tennō was first used during the Jitō reign, but a recently discovered wooden slip (mokkan 木簡) from Asuka Pond, stratigraphically datable to the mid-670s, indicates that it was already in use by the reign of Tenmu (Asuka 2002, p. 91, fig. 68.) Sacred ritual was conducted at shrines and temples; several shrines, such as Hirose 広瀬, Tatsuta 竜田, and Ise 伊勢 were of importance, although as pointed out by Piggott, most were relatively close to the capital, suggesting that the court’s relationship with distant shrines was still quite limited. In the case of temples, she again cites the “three official temples” near the palace, Daikandaiji, Gufukuji 弘福寺, and Gangōji. As pointed out above, we do not know the location of the Tenmu Daikandaiji, and it should also be noted that the Buddhist names “Gufukuji” and “Gangōji” are unlikely to have been in use at this early date, despite what Piggott states in note 68. Although the very important fourth of the “Four Great Temples,” Yakushiji 藥師寺, is indicated on Map 7, p. 136, it seems not to be discussed in the text; in my view, Yakushiji is the first of the early temples with a proper Buddhist name. Its placement within Fujiwarakyō is of key symbolic importance, as it protected one side of the capital. (The Daikandaiji shown on the same map is that built during the Monmu era, not the earlier one.) Piggott’s understanding of the famous Nihon shoki edict of 680.4 (not 25th day, but “this month”) concerning the administration of Buddhist temples is incomplete. Although it might seem that the quote on pp. 147–148 is a literal translation of the edict, in fact important details are missed. Piggott apparently assumes that the edict concerns only the number and resources of the Daiji (=Ōdera) whereas it is actually much broader. The “two or three Great Temples” that will continue to be administered by officials, are distinguished from a much large group


of temples that had received government support over the years. It is the members of this latter group that will have their sustenance households limited to thirty years, not the Great Temples. This fact is clear from the 679.4.5 edict that required a survey of the history of all temples that had sustenance households. The utilization of the term “official temple” is inappropriate, especially in the case of Asukadera, since that temple was a Soga temple, not one vowed by the Great King, and thus is more correctly referred to as a Great Temple. (Since Piggott translates Ōkimi 大王 as “Great King” there should be no problem with rendering Ōdera 大寺 as Great Temple.) Piggott does not comment on the ambiguous phrase “two or three,” which has evoked considerable speculation. The most convincing explanation is that of Ōhashi Katsuaki, who argues that the “two” refers to Kudara Ōdera and Kawaradera, both of which were in place by the date of the edict, while the additional temple in the “three” would be Yakushiji, which was not vowed until 680.11.12, seven months after the promulgation of the edict (Ōhashi 1995, pp. 45–46). Since the compilers of Nihon shoki would have known this fact, it seems that they rendered it by the ambiguous phrase. The slightly peculiar treatment of Asukadera in this entry was necessary because of the auspices under which it was commissioned and built, but with its inclusion we arrive at the standard group of Four Great Temples 四大寺, a group that continued to be important in the following Nara period. In addition to the planning for Fujiwarakyō, there are indications of plans to build other capitals elsewhere, including intriguing mentions of Shinano 信濃 (modern Nagano 長野 Prefecture), presumably to assert the court’s authority in the periphery. While Piggott’s argument con-cerning the relative weakness of the courts of Tenmu and Jitō is largely convincing, I still wonder how we can account for the vast resources spent on building the enormous Fujiwara Palace and Capital. A central theme in this book, but especially of the present chapter, is the importance of literacy in extending the court’s authority. Piggott sees the elite’s possession of literacy as a crucial tool in the achieving of its goals, and she focuses on the development of literacy in both center and periphery citing, for example, the production of things like a na-tive dictionary (Niina 新字). In her view, the governors (mikotomochi 宰) dispatched to the provinces were not so much administrators as cultural ambassadors, striving to encourage local rulers to affiliate more strongly with the court. An equally important tool for the spread of new culture was Buddhism. Piggott deals with the famous 685 edict concerning the worshiping of Buddhist images throughout the country, with her translation of the Nihon shoki text stating that this should be carried out in “every government office” (p. 154); in fact, the phrase so translated is 毎家 which might be rendered in this way, but is more naturally read


as “every household.” It is evident that this would not refer literally to every household in the land, but only to those of the elite. Piggott includes important material on the spread of temples throughout the country, but while acknowledging the contribution of Korean “immi-grants,” she tends to stress the efforts of the court. In my view, this is a Yamato-centric conception that fails to take fully into account the fact that many early temples were established by groups coming directly across the East Sea/Japan Sea from the peninsula, with no stopover in Yamato. Naturally, there was a strong likelihood of a succession dispute upon the death of Tenmu, since as yet no definite formulation of the mode of succession had been achieved. Such disputes usually took place during the lengthy period of mourning for the departed monarch. After Tenmu died (686.9.9), numerous ceremonies were performed, beginning with the temporary interment in a special building (mogari no miya 殯宮) and including rituals at various Buddhist temples. Piggott harshly criticizes Ebersole’s (1989) treatment of the mogari as “incomplete and flawed” (note 118, p. 365), although it might be noted that her own understand-ing of Musha daie 無遮大会 is not entirely accurate either; it seems that Piggott does not really try to come to terms with Ebersole’s innovative study. Nevertheless, she is right to point out the very significant partici-pation of Buddhism in Tenmu’s final rites. Tenmu was finally buried in the splendid Hinokuma 檜隈 Tomb, about which Piggott might have said more. For example, its placement some distance to the south on the central north-south axis that ran through the Fujiwara capital obvi-ously had symbolic meaning. His consort, Jitō, was also buried there, but in distinction to her husband’s sarcophagus, her ashes were placed in an urn; it would seem that Jitō was the first monarch to be cremated, something that further reinforces the notion of the great importance of Buddhist thought at this time. Piggott is especially concerned with the image of monarchy that can be seen in poems of the Man’yōshō 萬葉集, particularly those associated with Tenmu’s mourning period. She suggests that a transformation from sacral to divine kingship can be observed in these poems, and believes that this crucial process was being elaborated during these specific years (ie. 686–688). Many of the developments discussed under Tenmu con-tinued with his successor, Jitō. When Prince Kusakabe 草壁皇子 died, Princess Uno, in order to protect her and her husband’s line, ascended the throne as Jitō, following in the footsteps of Suiko and Kōgyoku/Saimei, presumably for similar reasons. With Kusakabe dead, the suc-cession fell to his son, Prince Karu 軽皇子, who ruled as Monmu tennō from 697, although Jitō probably remained the power behind the throne until her death in 702. Evidence of this is her selection of appropriate


consorts in order to assure the succession in their line. Piggott brings up the theory that the famous Takamatsuzuka Tomb 高松塚 was for Jitō’s great ally, Prince Takechi 高市皇子, but this remains unproven. Chapter 6, “Great Kings and Ritsuryō Law,” dealing with the first two decades of the eighth century, concentrates on the formation of the ritsuryō state. Less emphasis is placed on individual monarchs than on broader ideological and political developments. An important feature of this chapter is an organizational approach that does not see a strong break in the historical process marked by the move to Heijōkyō (Nara) in 710; rather, Piggott sees continuity in the growth of the ritsuryō state from the later seventh century (Chapter 5) through the years of the pres-ent chapter to the period of Shōmu, the subject of the final chapter. In Piggott’s view, the ritsuryō state should be seen not so much as an efficiently running administrative system, but more as an ideological construct designed to unite various elites, both of the center and pe-riphery, under the symbolical authority of the tennō. Consequently, three hierarchies —bureaucratic, geographical, and sacral— are analyzed at length in order to show how they were woven together into one complex entity and united in the person of the Heavenly Sovereign. Fundamental to her approach is the realization that the provincial elites maintained considerable autonomy and thus could only be incorporated into the broader ritsuryō polity through persuasion, and not by coercion. Throughout the book Piggott consistently stresses the great signifi-cance of Chinese texts, especially those of the Han Dynasty and earlier, to the Yamato leaders in their formulation of an ideology of kingship. Perhaps, as mentioned above, the notion of very early familiarity with these texts needs further examination, but with the period under con-sideration in this chapter there can be no doubt about their central role. Now Yamato thinkers utilized canonical works such as the Book of Rites and the Rites of Zhou, crucial in the formulation of Chinese ideas of sovereignty. Piggott clarifies both those aspects on which the Japanese drew as well as the ways in which their formulation differed from the Chinese prototypes. There is an extensive discussion of those sections of the Yōrō 養老 Code that can be reconstructed, focusing primarily on the structure of the ritsuryō state. Important units, such as the Council of State (Daijōkan 太政官 ), are examined and placed in context; Piggott emphasizes the essentially conservative nature of these units, which she sees as perpetuating the power and prestige of already existing elites, rather than constituting novel relationships between, for example, capital and province. Nevertheless, in the awarding of ranks to local chieftains, the new system rewarded the individual, not the uji as had earlier been the case. In her examination of these matters, Piggott shows the way in


which the chiefly elites were incorporated into the officialdom while generally retaining most of their traditional authority. Although at first one might be inclined to see the system of provin-cial roads and administrative facilities as only ideas adumbrated by the codes, archaeological research has shown their widespread distribution throughout the polity. Of course, the most extensive excavations have been at the site of Heijōkyō, capital from 710, and they have yielded quite extraordinary results. Piggott states that Heijōkyō was twice as large as Fujiwarakyō, although as noted above, recent investigations have demonstrated that the latter was much larger than previously believed and may even have been larger than the new capital. If so, a recon-sideration of Fujiwarakyō in terms of the initial formulation of ideas about a “proper” capital would be in order. Fascinating as Heijōkyō is, greater attention to Fujiwarakyō might lead to some revisions of tradi-tional concepts, especially those that over emphasize the “Golden Age” of Nara Japan. In her treatment of the provinces, Piggott once again suggests that the governors dispatched from the capital served more as “cultural emissaries” than as political leaders, since clearly they lacked the military resources needed to enforce their wishes. They relied on persuasion and symbolic authority. Needless to say, the monarch and court utilized religion to enhance their authority. As Piggott shows, great reliance was placed on kami 神 worship, but the official Buddhist cult was particularly important in constructing an image of the tennō. Of course, in dealing with Bud-dhism controls were essential, and the ritsuryō court was quite success-ful in incorporating the Prelates’ Office (Sōgō 僧綱) into the overall bureaucratic structure. Piggott contrasts Gien 義淵 and Dōji 道慈, two priests well integrated into the government bureaucracy, with Gyōki 行基, a priest outside of that structure. While priests of the official church were primarily concerned with the welfare of tennō and state, Gyōki represented somewhat of a challenge to them in his activities among the people at large. In tying together her analysis of the development of ritsuryō, Piggott criticizes earlier views that saw it as a highly centralized system (e.g. George Sansom), finding that : “My own view shares much with that of Ishimoda Shō” (p. 232). This viewpoint is summarized in her statement:

Ritsuryō officialdom was an archipelago-wide alliance of elites unified by mutual acceptance of elements of the worldview promulgated by Yamato court society (p. 233).

Piggott rejects the theory that the ritsuryō system aimed primarily to serve the interests of the Kinai elite, arguing that if that had been the case it would probably have failed. Naturally, Piggott considers various


tensions that developed, including struggles among various segments of the polity, central and provincial, over issues such as taxation; those resulting from sheer change that threatened traditional practice; and internal struggles at court. Chapter 7, “Shōmu Tennō, Servant of the Buddha,” which concludes the book, covers material the author has long investigated and on which consequently she is an authority. She sees the reign of Shōmu and the construction of the Great Buddha at Tōdaiji 東大寺 as the culmination of the development of ritsuyō kingship. Nevertheless, as Piggott points out, these decades witnessed one crisis after another, beginning with succession problems after the death of Monmu. As had been the case in the preceding century, female monarchs ruled, first Genmei 元明 (707–715) and then Genshō 元正 (715–724), before Prince Obito 首皇子 occupied the throne as Shōmu. At court there was constant battling between various cliques, particularly those associated with the impe-rial line and those of the Fujiwara family. Thus Prince Nagaya 長屋王 opposed the Fujiwara during the early years of the reign, but failed and was forced to commit suicide in 729. The elevation of Princess Asukabe 安宿媛, daughter of the leader of the Fujiwara, Fujiwara no Fuhito 藤原不比等, to the position of consort was a triumph for their interests, although the deaths by smallpox of Fuhito’s four sons in the 735–737 epidemic was a devastating blow to Fujiwara power. In the end, however, the princess, known to us as Kōmyō 光明, was able to achieve her goals and those of her husband by having their daughter, Princess Abe 阿部, appointed as Crown Princess. She ascended the throne twice, first as Kōken 孝謙 and later as Shōtoku 称徳; however, these very interesting reigns are beyond the scope of the book. Signifying the unrest during Shōmu’s reign was the frequent transfer of capital, first to Kuni 恭仁 (741) and then to Shigaraki 紫香楽 (743). These moves, related to the struggle for power among the various fac-tions, must have been very expensive and it would have been a great relief when the court finally returned to Nara in 745. In addition, Shōmu initiated a system of provincial temples (kokubunji 國分寺 and kokubun-niji 国分尼寺) to spread Buddhism throughout the realm; at the apex of this system was Tōdaiji, the largest temple built in Japan, housing a monumental gilt-bronze image of Rushana 廬舎那佛, the Cosmic Buddha. Piggott offers a good deal of information about this enormous project, indicating that Shōmu saw himself as a bodhisattva-king. In this context, she disputes the common assumption that there was some sort of opposition between court and clergy, arguing that these two institu-tions were closely allied and had many common interests. In the “Epilogue” the author presents her ideas concerning what she sees as the maximum development of a ritsuryō type government under


Shōmu and the tensions that developed after his death. Problems of succession, infighting at court, and economic difficulties combined to work against an ideal polity rooted in Chinese thought and practice; when the succession passed from the Tenmu-Jitō line to that of Tenji, substantial changes occurred, especially in connection with the abortive move to Nagaoka 長岡 (784) and the definitive move to Heiankyō 平安京 in 794. With the latter we are in a new historical epoch beyond the scope of the present book. In addition to the narrative account, we are provided with a detailed appendix, “A Catalogue of Major Source Documents and Collec-tions” (pp. 287–303), a “Glossary” (pp. 305–329), and an extensive “Bibliography” (pp. 393–423), all of which will be of use to students of this period. Also, much of the argumentation of the book will be found in the “Notes” (pp. 331–392). Stanford University Press is to be congratulated for the fine production of the text.

References Cited

Asuka (2002). Nara Bunkazai Kenkyūjo 奈良文化財研究所, ed., Asuka Fujiwarakyō ten: Kodai ritsuryō kokka no sōzō 飛鳥ー藤原京展: 古代律令国家の創造 (The Capitals of Asuka and Fujiwara: The Formation of the Ancient Ritsuryō State). Nara: Asahi Shinbun.

Asuka Shiryōkan (1979). Nara Kokuritsu Bunkazai Kenkyūjo, Asuka Shiryōkan 奈良国立文化財研究所, 飛鳥資料館, ed., Asuka-Hakuhō no zaimei kondōbutsu 飛鳥ー白鳳の在銘金堂仏 (Inscribed Bronze Sculptures of the Asuka and Hakuhō Periods). Kyōto: Dōmeisha.

Ebersole, Gary L. (1989). Ritual Poetry and the Politics of Death in Early Japan. Princeton: Princeton University Press.

Kadowaki Teiji 門脇禎二 (1991). Taika kaishin shiron 大化改新史論 (Historical Treatise on the Taika Reform), two volumes. Kyōto: Shinbunkaku shuppan

Kashihara (1999). Nara Kenritsu Kashihara Kōgaku Kenkyūjo 奈良県立橿原考古学研究所 ed., Kurozuka Kofun 黒塚古墳 (The Kurozuka Tomb). Tōkyō: Gakuseisha.

Kashihara (2001). Nara Kenritsu Kashihara Kōkogaku Kenkyūjo 奈良県立橿原考古学研究所 ed., Hokenoyama kofun ホケノ山古墳 (The Kokenoyama Tomb). Tōkyō: Gakuseisha.

McCallum, Donald F. (2001). “The Earliest Buddhist Statues in Japan,” Artibus Asiae LXI.2: 149–188.

Ōhashi Katsuaki 大橋一章 (1995). “Chokuganji to kokka kanji no zōei soshiki 勅願寺と国家官寺の造営組織 (The Construction Organiza-tions of Imperial Vowed Temples and State Temples).” Bukkyō geijutsu 佛教藝術 222: 41–61.


Ortner, Sherry (1981). “The Case of Polynesia and Some Comparative Implications,” in Sherry Ortner and Harriet Whitehead (eds.), Sexual Meanings: The Cultural Construction of Gender and Sexuality, pp. 359–409. Cambridge: Cambridge University Press.

Ōyama Seiichi 大山誠一 (1999). “Shōtoku Taishi” no Tanjō 〈聖徳太子〉の誕生 (The Birth of “Shōtoku Taishi). Tōkyō: Yoshikawa Kōbunkan.

Address:

Department of Art HistoryUniversity of California,Los Angeles, CA 90095E-mail: <[email protected]>


GUEST EDITOR’S PREFACE

BY

JAY XU

The present volume of the Journal of East Asian Archaeology includes a special section on the art and archaeology of the Sichuan Basin, par-ticularly of the Chengdu Plain in the western part of the basin. The eight papers that comprise it span three millennia from the late Neo-lithic period (third millennium BC) through the Eastern Han period (AD 25–220). They introduce recent results of field archaeology or new interpretations of debated issues. In assembling this collection of papers, we follow a long-standing scholarly tradition of inquiry into a border region with a complex his-tory of interaction between Chinese culture and the cultures of other ethnic groups. As early as 1922, an English-language journal focusing on Sichuan and surrounding regions—the Journal of the West China Border Research Society—began to publish, the society being an international and multidisciplinary association headquartered in the West China Union University Museum of Art, Archaeology and Ethnology in Chengdu 成都, the capital of Sichuan province. In print until 1946, the journal was largely devoted to ethnographic and environmental studies, but also published field reports and research on ancient remains, includ-ing the first ever report of an excavation at the Sanxingdui 三星堆 site in the Chengdu Plain (see Jay Xu, this volume). In 1942, the Shuowen yuekan 說文月刊 (Shuowen monthly) devoted a special issue to ancient Sichuan (vol. 3 no. 7). In 1946, Zheng Dekun (=Cheng Te-k’un) 鄭德坤 produced the first monograph on ancient Sichuan, Sichuan gudai wenhua shi 四川古代文化史 (A Cultural History of Ancient Sichuan) (Zheng Dekun 1946). In 1957, Zheng published Archaeological Studies in Szechwan (Cheng 1957) to “trace the development of culture in this western part of China and to find how this province was linked up with the rest of the world” (p. xiii), a goal still shared by scholars today. In 1979, Tong Enzheng 童恩正 published his monograph, Gudai de Ba Shu 古代的巴蜀 (Ancient Ba and Shu) (Tong Enzheng 1979), a simpler sketch similar in nature to Zheng’s Sichuan gudai wenhua shi.

102 JAY XU

These publications attest to a history of strong scholarly interest in ancient Sichuan and present considerable results achieved in fieldwork and research. The finds accumulated had apparently made it possible for Zheng to “describe the development of culture in Szechwan (Si-chuan) purely from the archaeological point of view” (Cheng 1957: xiii) in Archaeological Studies in Szechwan. However, compared to the large increase of archaeological data that was to take place beginning in the late 1970s, the materials available at Zheng’s time (and in the follow-ing two decades) were still very limited: for the prehistoric section of Archaeological Studies(a period that Zheng estimated to last from about 5000 BC to 700 BC), although no fewer than 90 sites had been investi-gated, few had actually been excavated, and the major portion of the material consisted of surface finds of lithic implements and potsherds (Cheng 1957: 3–4). Few artifacts of elite material culture were known. This factor must have contributed to the rather uninspiring assessment that Zheng provided for ancient Sichuan: as he put it, “Szechwan is fundamentally a marginal area, and the culture of this province has never been a product of an independent development. It has always been under the influence of some neighboring cultures.” (Cheng 1957: xix; Cheng 1982: 31). In 1986, the picture of ancient Sichuan changed dramatically, when brickyard workers at Sanxingdui accidentally exposed two pits filled with hundreds of bronzes, among them strange sculptures never seen before, alongside stone and jade implements, gold objects, and elephant tusks. The pits had preserved in staggering abundance an elite mate-rial culture startlingly different from any other in Bronze Age China. Following the discovery, archaeologists surveyed and mapped the site, including its walls, revealing it to be among the largest cities known from early Bronze Age China. Sanxingdui proves that the Chengdu Plain was home to a major indigenous civilization. The 1986 Sanxingdui discovery belongs to a great wave of activ-ity in field archaeology in China since the end of the 1970s. Largely caused by extensive capital construction following economic reforms, the momentum has remained unabated. In Sichuan, particularly in the Chengdu Plain, a series of striking discoveries has turned the region into a major focus of Chinese archaeology. Some of the most notable finds up to 2003, besides the 1986 Sanxingdui discovery, are listed below, ar-ranged chronologically by the periods to which they belong. Fieldwork in the Three Gorges area, a vast salvage operation that has been going on since 1994 in response to construction of the Three Gorges Dam on the Yangzi River, calls for a separate treatment and is not included here.

GUEST EDITOR’S PREFACE 103

1. Neolithic period. The fourth-millennium BC Yingpanshan 營盤山 site in Maoxian 茂縣 County in the western edge of the Sichuan Basin, discovered in 2000 (Chengdu 2002: 1–77).

2. Late Neolithic period. A cluster of six walled settlements of the third-millennium BC Baodun 寶墩 Culture discovered in the Chengdu Plain since 1995 (see Wang Yi, this volume). A seventh walled settlement, the Yandian 鹽店 site, was discov-ered in 2003.

3. Late Shang and early Western Zhou period. The Shi’erqiao 十二橋 site in the city of Chengdu, discovered in 1985 (see Jay Xu, this volume for references), and related sites, including Jin-sha 金沙, discovered in 2001 in the city’s west suburb (see Zhu Zhangyi et al., this volume).

4. Western Zhou period. Two hoards of bronzes discovered in 1959 and 1980 at Zhuwajie 竹瓦街 in Pengxian 彭縣 County in the Chengdu Plain (see Sun Hua, this volume).

5. Warring States period. Numerous tombs and cemeteries, the most remarkable of which includes those discovered at Moutuo 牟托 in Maoxian County, at Hejiaping 郝家坪 in Qingchuan 青川 County in the northwestern edge of the Sichuan Basin, and at Majia 馬家 in Xindu 新都 County in the Chengdu Plain (lacquer wares from those tombs are discussed in Alain Thote, this volume). The latest finds include a cemetery discovered in 2000 at Shangyejie 商業街 in Chengdu (Chengdu 2002: 78–141), and a large site partly excavated in 1999 and 2003 at Luojiaba 羅家壩 in Xuanhan 宣漢 County towards the eastern edge of the Sichuan Basin (Sichuan 2004).

6. Qin and Han periods. A large number of tombs, many of which were clustered on cliffs, including those at Santai 三台 County (see Susan Erickson, this volume).

In tandem with the great wave of fieldwork and finds, publication of field reports and research has surged, though it understandably lags behind the fieldwork, in some cases considerably so. Besides individual reports and papers published in a large number of journals and collec-tions of papers, three journals devoted to the archaeology of Sichuan and surrounding regions have been established in Chengdu: Chengdu wenwu 成都文物 (1983– ), Sichuan wenwu 四川文物 (1984– ), and Nanfang minzu kaogu 南方民族考古 (1987– ). Book-length excavation reports and collections of research papers on ancient Sichuan have appeared since 1986, and a number of scholars have published personal anthologies of research papers. To meet the growing need for space to publish field reports, the Sichuan Provincial Institute of Archaeology published

104 JAY XU

Sichuan kaogu baogaoji 四川考古報告集 (Collection of Sichuan archaeo-logical reports) (Sichuan 1998), and in 2001, the Chengdu Municipal Institute of Archaeology began publishing an annual collection of field reports, Chengdu kaogu faxian 成都考古發現 (Archaeological discoveries in Chengdu) (Chengdu 2001, 2002, 2003, 2004), which has proven to be an indispensable source of information. The Sanxingdui site naturally has become a subject for intensive investigation; the literature published on it is among the most voluminous in Chinese archaeology, though the quality varies enormously. Several scholars have attempted new syntheses of ancient Sichuan: Ancient Sichuan and the Unification of China by Steven F. Sage (Sage 1992), Shu wenhua yu Ba wenhua 蜀文化與巴文化 (The Shu culture and the Ba culture) by Song Zhimin 宋治民 (Song Zhimin 1998), and Sichuan pendi de qingtong shidai 四川盆地的青銅時代 (The Bronze Age of the Sichuan Basin) by Sun Hua 孫華 (Sun Hua 2000). The region has also been the subject of an exhibition that toured North America in 2001–2002, Treasures from a Lost Civilization: Ancient Chinese Art from Sichuan, organized by the Seattle Art Museum, which featured a large selection of artifacts from the Sanxingdui pits and from later periods through the Han dy-nasty. Its accompanying catalogue, Ancient Sichuan: Treasures from a Lost Civilization (Bagley 2001), provides a survey of ancient Sichuan’s material culture. The present collection of research papers originated in a symposium held at the Seattle Art Museum at the time of the exhibition in 2001. The purpose of the symposium was to review the latest archaeological finds, discuss the unique characteristics of the art and material culture of ancient Sichuan, and explore relations between Sichuan and other regions. Seven papers from that symposium were eventually scheduled for publication in this volume of the Journal of East Asian Archaeology. The news of the Jinsha site, which had just been discovered in the begin-ning of the year, was reported to the symposium by Wang Yi 王毅, who heads the Chengdu Municipal Institute of Archaeology, and during the planning of the present volume, it was decided that a paper introducing the Jinsha site and its finds should be added. Although the papers in this section are arranged chronologically, they do not attempt to provide a continuous narrative about ancient Sichuan. Given large gaps still existing in the archaeological record, this would hardly be possible. The Spring and Autumn period at the moment is an archaeological blank. Moreover, beyond this practical consideration, we wished to give room to individual focuses and perspectives. In the first paper, “Prehistoric Walled Settlements in the Chengdu Plain,” Wang Yi introduces the six walled settlements of the Baodun (or Baoduncun 寶墩村) Culture and characterizes this recently identified


Neolithic culture, whose distribution concentrates in the Chengdu Plain. In so doing, Wang shows that the Sichuan Basin, particularly its western part, was a place of origin and development of ancient civilization in the upper Yangzi River valley. In my essay “Defining the Archaeological Cultures at the Sanxingdui Site,” I review the fieldwork at the Sanxingdui site, discussing the stra-tigraphy, pottery typology, periodization, and absolute dating of the site, and re-evaluating major periodization schemes and different definitions of the Sanxingdui Culture. In conclusion, I offer my understanding of the spatial and temporal dimensions of that culture. The striking finds from the two pits at Sanxingdui are analyzed and discussed by Lothar von Falkenhausen in “The External Connections of Sanxingdui.” Falkenhausen elucidates the relationships of the San-xingdui culture to archaeological cultures in the surrounding regions. Connections are recognized between artifacts from the Sanxingdui pits (as well as from Jinsha) and finds from other parts of China. He demonstrates that Bronze Age Sichuan was not isolated in the cultural interaction sphere of mainland East Asia. In their essay “The Jinsha Site: An Introduction,” Zhu Zhangyi, Zhang Qing, and Wang Fang, the archaeologists chiefly responsible for the fieldwork at Jinsha, introduce the site in considerable detail, particularly the rich array of artifacts unearthed. The authors describe the primary loci and associated finds, attempt temporal and cultural definition of the site, and evaluate the archaeological significance of this major discovery. This paper and Wang Yi’s paper on the Baodun sites are the first English introductions of these latest results from the field. In “The Zhuwajie Bronzes,” Sun Hua revisits the two hoards of bronzes, which have been the subject of much controversy. The date and circumstances of deposition have long been debated, as have the styles, dates, and likely place of manufacture of the bronzes. Sun carefully examines these questions and also pays close attention to bronze casting technology. He discusses the hoards in the context of local cultures and of Western Zhou history. With the recent discovery at Jinsha, we should expect further inquiries into a possible relationship between Zhuwajie and Jinsha. Alain Thote’s “Lacquer Craftsmanship in the Qin and Chu King-doms: Two Contrasting Traditions (Late 4th to Late 3rd Century B.C.)” differs from the other papers by dealing primarily with cultural traditions outside the Sichuan Basin in relation to the recent discoveries from Sich-uan. Thote points out that, in the later Warring States period, Sichuan was a key region in the development of the lacquer art of Qin. Michael Nylan’s paper, “Ordinary Mysteries: Interpreting the Archaeological Record of Han Sichuan,” brings complex, layered

106 JAY XU

readings to seemingly familiar themes of Han art from Sichuan. Nylan challenges her reader to consider the possibility of multiple metaphori-cal meanings encoded in images and demonstrates how such meanings may be deciphered. She also discusses features of the Sichuan figurines and tomb tiles that are unique to the area. In “Eastern Han Dynasty Cliff Tombs of Santai County, Sichuan Province,” Susan Erickson addresses a particular type of burial con-struction in Sichuan: rock-cut tombs popular during the Eastern Han period. She concentrates on a cluster of sites in Santai County. By discussing in detail the basic characteristics of those tombs, and their architectural elements and decorative designs, Erickson sheds light on the real architecture that provided models for the tombs and on the meaning of the decorative images seen in the tombs. In reading the papers, one may observe a recurring contradiction: the art and material culture of ancient Sichuan often appear distinctive, sometimes startlingly so, even in the Eastern Han period when Sichuan had long become part of a unified empire; yet at the same time, con-nections with the outside world are extensive. Factors contributing to this unique cultural pattern should be among questions for continuing fieldwork and research. Here, we may tentatively identify one possible factor for consideration: Sichuan’s geographical position. Situated in the upper reaches of the Yangzi River in southwestern China, the Sichuan Basin’s geography may be characterized as “quali-fied isolation.” The land is marked off from the outside world by land barriers, with mountains or high plateaux on all sides. Against this for-bidding topography, however, waterways afforded by the Yangzi River and its tributaries, and trails along mountainsides, often precipitous, offered routes of communication. Thus while the land barriers isolated the Sichuan Basin, severely limiting communication, rivers brought communication from all directions. This qualified isolation probably helped shape the course of Sichuan’s cultural development. The rise of distinctive Neolithic and bronze-using cultures in the Chengdu Plain may well have depended crucially on the combination of geographical isolation and far-flung contacts. The plain was isolated enough to foster the development of native cultures and escape being overwhelmed by any one outside influence, but open enough to be stimulated by outside contacts (Xu 2001a: 21–22). In preparing the Seattle symposium and this collection of papers, many colleagues and organizations made invaluable contributions. I would first like to thank the speakers at the symposium and the authors whose work is included here. I am indebted to Mimi Gardner Gates, Director of the Seattle Art Museum, for her unfailing support and pa-tience through the process. Generous support for the symposium was


provided by the Blakemore Foundation, the Katherine Agen Baillargeon Endowment, and the Asian Art Council of the Seattle Art Museum. I am extremely grateful to the Journal of East Asian Archaeology and its edi-tors for providing a forum for the views of the authors. Warmest thanks are due to the translators of the papers originally written in Chinese: Wenbin Peng, Mark Pitner, and Adam Smith. To everyone else who has helped in the symposium and the publication of this collection of papers, I extend my heartfelt gratitude.

References Cited

Bagley, Robert (editor) (2001). Ancient Sichuan: Treasures from a Lost Civi-lization. Seattle: Seattle Art Museum in association with Princeton University Press.

Cheng, Te-k’un (1957). Archaeological Studies in Szechwan. Cambridge: Cambridge University Press 1957.

Cheng, Te-k’un (1982). “An Introduction to Szechwan Archaeology.” In Cheng, Te-k’un, Studies in Chinese Archaeology, pp. 27–33. Hong Kong: The Chinese University Press.

Chengdu 2001: Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所 (editor) (2001). Chengdu kaogu faxian 1999 成都考古發現 1999 (Archaeological discoveries in Chengdu 1999). Beijing: Kexue Chubanshe.

Chengdu 2002: Chengdu Shi Wenwu Kaogu Yanjiusuo (editor) (2002). Chengdu kaogu faxian 2000. Beijing: Kexue Chubanshe.



Sage, Steven F. (1992). Ancient Sichuan and the Unification of China. Albany: SUNY Press, 1992.

Sichuan 1998: Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 (editor) (1998). Sichuan kaogu baogaoji 四川考古報告集 (Collec-tion of Sichuan excavation reports). Beijing: Wenwu Chubanshe.

Sichuan 2004: Sichuan Sheng Wenwu Kaogu Yanjiusuo, Dazhou Diqu Wenwu Guanlisuo 達州地區文物管理所, and Xuanhan Xian Wenwu Guanlisuo 宣漢縣文物管理所 (2004), “Sichuan Xuanhan Luojiaba yizhi 2003 nian fajue jianbao 四川宣漢羅家壩遺址 2003 年發掘簡報 (Preliminary report of the 2003 excavation of the Luojiaba site in Xuanhan County, Sichuan).” Wenwu 2004.9: 34–47.

Song Zhimin (1998). Shu wenhua yu Ba wenhua 蜀文化與巴文化 (The Shu culture and the Ba culture). Chengdu: Sichuan Daxue Chubanshe.

108 JAY XU

Sun Hua (2000). Sichuan pendi de qingtong shidai 四川盆地的青銅時代 (The Bronze Age of the Sichuan Basin). Beijing: Kexue Chubanshe.

Tong Enzheng (1979). Gudai de Ba Shu 古代的巴蜀 (Ancient Ba and Shu). Chengdu: Sichuan Renmin Chubanshe. Chongqing: Chongq-ing Chubanshe, 1998 reprint.

Xu, Jay (2001). “Sichuan before the Warring States Period.” In Robert Bagley ed., Ancient Sichuan: Treasures from a Lost Civilization. Seattle: Seattle Art Museum in association with Princeton University Press. pp. 21–37.

Zheng Dekun (1946). Sichuan gudai wenhua shi 四川古代文化史 (A cultural history of ancient Sichuan). Chengdu: West China Union University Museum Monograph Series No. 1; Chengdu: Ba Shu Shushe, 2004 reprint.

Address:

Department of Asian ArtThe Art Institute of Chicago111 South Michigan AvenueChicago, IL 60603-6110USAEmail: [email protected]


PREHISTORIC WALLED SETTLEMENTS IN THE CHENGDU PLAIN

BY

WANG YI 王毅(Chengdu Municipal Institute of Cultural Relics and Archaeology)

Translated by Wenbin Peng 彭文彬 and Mark Pittner

Abstract The Baoduncun Culture, named after prehistoric walled settlements in the Chengdu Plain, was a regional archaeological culture distributed in the western part of the Sichuan Basin along the upper reaches of the Yangzi River and centered in the Chengdu Plain. Dating from the Late Neolithic (or Chalcolithic) and contemporaneous with the Long-shan cultures of northern China, its recent discovery constitutes a major achievement in the clarification of the pre-Qin archaeological sequences in southwestern China. The Baoduncun Culture, like the Sanxingdui Culture that followed it, is highly instructive demonstrating the special status that the Sichuan Basin, particularly its western por-tion, holds as a place of origin and a center of development of ancient civilization in the upper Yangzi River valley.

History of Baoduncun Culture discoveries

Major fieldwork relating to the Neolithic period in the Sichuan Basin began in 1953, when the Bianduishan 邊堆山 site was discovered in Mianyang 綿陽 County (now City), Sichuan Province, and provision-ally dated to the Neolithic period (Xi’nan Bowuyuan Choubeichu 1954: 11).1 In the autumn of 1963, Sichuan University and the Sichuan Provincial Commission of Cultural Relics Management undertook a joint excavation at the Yueliangwan 月亮灣 locus of the Sanxingdui 三星堆 site in Guanghan 廣漢 County (now City) (site locations are

1 The Editors thank Wenbin Peng and Mark Pitter for their draft translation of the paper presented here in revised form. The Editors have provided additional references, including those for a number of field results and preliminary excavation reports that have been published since the submission of the manuscript in Chinese.

110 WANG YI

shown in Figure 1),2 yielding features and artifacts that ranged in date from the late Neolithic to the early Bronze Age periods (Ma Jixian 1993). Between 1980 and 1981, a large-scale excavation at Sanxingdui carried out by the Sichuan Provincial Commission of Cultural Relics Management and other archeological institutions identified cultural remains predating the Bronze Age; these late Neolithic remains have since become known as Sanxingdui Phase I remains (Sichuan Sheng Wenwu Guanli Weiyuanhui et al. 1987). From 1989 to 1991, the Si-chuan Team of the Institute of Archaeology, Chinese Academy of Social Sciences conducted a series of field surveys and excavations along the northern edge of the Sichuan Basin, at Bianduishan in Mianyang, as well as at Zhangjiapo 張家坡, Dengjiaping 鄧家坪, and Zhongzipu 中子鋪 in Guangyuan 廣元 City, adding much to our knowledge of the Neolithic cultural history of the region (Wang Renxiang and Ye Maolin 1993). Since the 1990s, however, all these previous results have been dwarfed in importance by successive discoveries of prehistoric walled settlements in the Chengdu 成都 Plain. Six of these recently identified walled settlements are described in this paper.3

The Baodun site in Xinjin

The Baodun 寶墩 site (30° 26' N, 103° 45' E, elevation 472–474 m above sea level) is a walled settlement located in Longma 龍馬 Town-ship, about 5 km northwest of Wujin 五津, the Xinjin 新津 County seat. Four km to the northeast of Baodun is the Xihe 西河 River, and 500 m to the southwest is the Tiexihe 鐵溪河 River, which flows from the northwest towards the southeast. A branch of the Sijiangyan 泗江堰 irrigation system flows from the northeast to the southwest and cuts across the northeastern section of the site, where mounds are still clearly visible on the surface. The Baodun site first caught the serious attention of archaeologists in the winter of 1984, when Luo Weixian 羅偉先, Wang Yi, and other archaeologists from the Archaeological Team of the Chengdu Munici-pal Museum conducted a survey of the area and collected fragments of brown coarse fabric pottery sherds and small stone axes (Wang Yi et al. 1997). In 1995, survey and preliminary excavations were conducted principally by the Chengdu Municipal Cultural Relics and Archaeology Team at Baodun (Chengdu Shi Wenwu Kaogu Gongzuodui et al. 1997).

2 Unless otherwise stated, all figures are based on line drawings provided courtesy of the Chengdu Municipal Institute of Cultural Relics and Archaeology. 3 In addition to this essay, a summary paper in Chinese on the Baoduncun Culture sites has been recently published. See Jiang Zhanghua et al. 2001.

PREHISTORIC WALLED SETTLEMENTS 111

The roughly square-shaped walled settlement is oriented to 316°, with the north and east walls relatively well preserved (Figure 2). The highest point is an approximately 5 m high section at the northern tip of the east wall known as Baodunzi, 寶墩子. The site covers an area of 60 ha. Excavation of building foundations of the Neolithic period within the settlement reveals that houses were built on ground level with mud walls on a wooden framework. Wooden posts were erected with their bases set into foundation trenches. An excavated section of the settlement wall shows that it was constructed with the “pile-and-ram” method, with preserved dimensions of 4 m in height, 8 m in width across the upper portion, and 30 m in width at the base. The walls were built along the

Figure 1. Distribution of early walled settlement sites (indicated by triangles) on the Chengdu Plain.

Sanxingdui

Yufucheng

Gucheng

Baodun

DujiangyanGuanghan

Mangcheng

Shuanghe

ZizhuChengdu

Longmens

han M

t.

Qionglaishan Mt.

Mianyang

50 km

112 WANG YI

edges of terraces that rose 3 m above the ground level. The “pile-and-ram” construction technique for the walls involved piling layers of earth and compacting them with wooden planks and rods (Figure 3). Pottery from the Baodun site includes two general wares: coarse fabric and fine clay. Most of the coarse pottery is gray, followed by brown and by vessels with brown exteriors and black interiors. The majority of the fine clay pottery features a relatively high-fired, grayish-white ware and slightly lower-fired, gray-yellow ware with a black slip. There are also small quantities of black-slip pottery and brown pottery. Among the coarse pottery, most have cord-markings, with a smaller number having stamped (or appliqué) and “bowstring” incised patterns, and a

Figure 2. Plan of the Baodun walled settlement site in Xinjin County.

Figure 3. Section profile of the west wall of test trench T1 through the north wall of the Baodun settlement, near Zhenwuguan.

Yulinpan Mahuangdun

Tianjiaolin

Gudunzi

Jianglin

Zhenwuguan

Sijia

ngya

n br

anch

cana

l

0 200 m

T1

1A

4A

1B

1B

QS1

QS2 QS5

5A

QS6QS7

QS8

G1 G2

QN1

QN2

QN3

2

1B

1B

34B

5B

1A

0 2 m


small minority with composite decorations. Stamped and appliqué décor dominate the fine clay pottery, followed by combed, ribbed, and a few composite patterns. Very few vessels are plain surfaced. The major vessel forms include cord-marked and scalloped-rim guan 罐 pots, tall-neck and trumpet-mouth guan, broad-rim zun 尊 beakers, ring-foot zun, everted rim guan, hu 壺 jars, broad-rim and tall-neck guan, constricted-neck guan, ring feet and handles of otherwise unknown vessel forms, and spindle whorls (Figure 4). Small polished stone tools, such as axes, adzes, and chisels, are also common. In 1996, the Chengdu Municipal Cultural Relics and Archaeology Team led an excavation of loci III and IV at Baodun, yielding rich finds. They also conducted an excavation of the southeastern section of the walls at Mahuangdun 螞蝗墩. This provides information concerning

Figure 4. Typical pottery vessel forms from the Baodun site: 1, 2, 4, 7. cord-marked and scalloped-rim guan [IIIT2029(7):129, IIICT2030(7):47, IIIT1829H:15:2,IIIT2030(7):137]; 3. broad-rim pan [IIIT2129(7):65]; 5. ring-foot zun with plate-shaped mouth [IIIT2129(7):76]; 6, 12. tall-neck and trumpet-mouth guan [TIQS(7):43,IIIH2:1]; 8. dou with a shallow plate [IIIT2029H5:59]; 9. guan with plate-shaped mouth [T2(6):92]; 10.broad-rim and flat-bottom zun [IIIT1929(7):46]; 11. flared-mouth and ring-foot zun

[III1929(7):128]; 13. hu [IIIT1929(7):2]. Vessels not to same scale.

1

2

3

4

5

6

7

12

8

9

10

11 13

114 WANG YI

the extent and orientation of the walls. Additional building foundations and “ash pits” were also discovered. The structures had been built on ground level, with mud walls supported by wooden posts. The ash pits are generally shallow with their cross sections displaying a variety of shapes. Moreover, five rectangular graves were excavated, three of which were small enough to suggest that they possibly contained child burials and the two larger ones possibly belonged to adults. The excavators estimate that the site dates back as far as 4500 BP and proposed naming the cultural remains of this type the Baoduncun 寶墩村 Culture. Pottery seriation for the Baodun site allows the site’s occupation to be divided into two major phases and three subphases. The Early Phase comprises Subphases 1 and 2, and the Late Phase corresponds to Subphase 3. The Baodun walls were likely constructed at the end of Subphase 1 or the beginning of Subphase 2 and were used through the end of the settlement (Zhong Ri Lianhe Kaogu Diaochadui 1998; Chengdu Shi Wenwu Kaogu Yanjiusuo et al. 2000).

The Yufucun site in Wenjiang

The walled settlement at Yufucun 魚鳧村 (30° 45' N, 103° 49' E, elevation 556–560 m above sea level) was discovered in 1996 when archaeologists from the Chengdu Municipal Cultural Relics and Ar-chaeology Team were investigating an exposed section in an irrigation channel. In this section they discovered not only brown coarse pottery sherds with stamped décor, but also the piled-and-rammed earth layers of a construction similar to the walls at Baodun. This section is now known to be in the eastern portion of the south wall of the earthen en-closure of the Yufucun site. Following this discovery, in the same year, the Chengdu Municipal Cultural Relics and Archaeology Team and the Department of History of Sichuan University carried out a joint survey and test excavations at Yufucun (Chengdu Shi Wenwu Kaogu Gongzuodui et al. 1998). The Yufucun site lies in the area of Yufu 魚鳧, Zhishu 直屬, and Baoen 報恩 villages in Wanchun 万春 Township, approximately 5.5 km to the north of Liucheng 柳城 Township in Wenjiang 溫江 County, and 7 km to the northeast of the Minjiang 岷江 River. Coring at Yufu-cun shows that the area of the walled settlement covers 40 ha, with the walls themselves following a hexagonal plan (Figure 5). In their present condition, the walls extend to 2 m above the current ground level and are 15 to 20 m wide. The walls date to the Neolithic period, and the Neolithic cultural remains are buried directly under Han 漢 and Song 宋 dynastic period deposits. Under the walls are earlier cultural levels.The majority of pottery unearthed at Yufucun is coarse brown (gray-brown


or red-brown) ware, with the next most abundant ware being fine clay. The fine clay wares clearly increase in abundance from the earlier to later Neolithic occupation at the site. Characteristic vessel forms include tall-neck and trumpet-mouth guan,cord-marked and scalloped-rim guan, ring-foot zun, and vessels with a small, flat base covered by a false ring-foot that has a whorl-like cord-marked pattern on the external side of the base (Figure 6). These false ring-foot vessels are unique to this site. Stone tools from Yufucun are generally small in size. The 1996 excavations at Yufucun are particularly remarkable for the discovery of remains that would later be recognized as belonging to Phase 3 in the three-phase pottery chronology for the site (see below). They also revealed details concerning the careful construction of the walls of the settlement. An excavated section through the wall shows

Figure 5. Plan of the Yufucun walled settlement site in Wenjiang County.

Baoensi

Fengjiadunzi

Luojia-yuanzi

GuangmingChicken

Plant

Xiaojiaxiang

Leijiayuanzi

Wangjiayuanzi

Shuaijiayuanzi

Wen

- Pi

Roa

d

0 100 m

116 WANG YI

that the sloping internal side of the wall was constructed above Neolithic cultural deposits, while the central section of the wall was constructed directly upon culturally-sterile, yellow soil. The existing walls average no less than 3.7 m high, and they measure 19 m thick at the top and 29 m thick at the base. They were built along the edge of a terrace ris-ing about 1.3 m above the ground. Each of the thickest layers in the wall had been built by piling fill in a sloped shape. A sloping layer of cobblestones was discovered on the surfaces of the first and the seventh layers on the south side of the main wall. The cobblestones are of equal size and are neatly arranged, revealing a carefully planned construction. This phenomenon, in which layers of earth are sandwiched between layers of cobblestones, is frequently seen on the exterior side (south side) of the excavated wall section. The interior side (north side) of the wall was different: the base of the first layer is comprised of a 10 cm thick level surface paved with cobbles that must have been related to the wall foundation. The construction of the Yufucun wall featured a piling method that did not use tools for compaction, as there are no clear traces for the use of pounding implements. The interior side of the main wall section featured a protective slope, and the portion of the wall through which the section was dug shows that the wall was built in a one-time construction event. From late 1999 to early 2000, excavations were conducted primarily in Yufucun’s eastern and southern sections. Excavated features include 90 Neolithic ash pits and 12 structures. These buildings were all con-structed from ground level. Their walls were either erected in founda-tion trenches or on elevated earthen floors supported by a framework of wooden posts driven into the ground. Four graves, apparently part of a cemetery, were also excavated. The burial pits were rectangular in shape and contained poorly preserved human remains, but no burial objects or furnishings. Additionally, five “ash trenches” were located. The excavation of the east wall at Yufucun shows that the construction method is equivalent to that of the excavated part of the wall section in the southeastern part of the site. Jumbled and comparatively small cobblestones were found in the layers of the north wall. These cobble-stones were scattered and not deliberately arranged in a horizontal or slanting manner like those excavated in 1996, yet various layers in the wall were sloped, conforming to the trend found in other prehistoric walled settlements in the Chengdu Plain. Excavated artifacts from Yufucun include pottery, stone implements, and bamboo pieces, with pottery being the most abundant. The most common stone implements found are small polished adzes, axes, and chisels. Additionally, there are stone spears that may have been used for ritual purposes, a few pieces of turquoise, chipped stone pieces, and


stone spheroids. The bamboo pieces are 20 by 2 cm, but their usage awaits further analysis. Distinctive types of pottery include cord-marked and scalloped-rim guan, tall-neck and trumpet-mouth guan (some hav-ing a scalloped rim), ring-foot zun, hu, broad-rim zun, ring-foot pan 盤 plates, waist-ridged vessels, narrow-rim guan, constricted-mouth guan, curved-rim guan, small-flat-bottom guan, vessel covers, and high-stem dou 豆 pedestal serving-stands (Figure 6). There are few vessels with small,

Figure 6. Typical pottery vessel forms from the Yufucun site. (1–13 belong to the Early Phase, 14–26 to the Late Phase). 1, 2. tall-neck and trumpet-mouth guan [H29:1, H10(2):55]; 3, 4. flared-mouth and ring-foot zun [H10(1):40, H15(2):34]; 5. broad-shoulder guan [H71(2):74]; 6, 10. cord-marked and scalloped-rim guan [H15(2):62, H15(2):111]; 7. folded-rim guan [H48(1):45]; 8. ring-foot zun with plate-shaped mouth [T10(5):7]; 9, 20, 25, 26. dou [H15(2):47, H73(3):156, H204:9, TG4G18:9]; 11. broad-rim and flat-bottom zun [H48(2):92]; 12, 13. deep-belly guan [H15(2):86, H15(2):85]; 14. bo [T9(5):78];15. bei 杯 cup [H73(3):125];16.constricted-mouth guan [H73(3):122]; 17, 21, 23. constricted-neck and broad-shoulder guan [H73(3):151, H73(3):150, H73(4):172];18.constricted-mouth weng [H73(4):180];19.curved-rim guan [H15(2):86];

22. vessel lid [H73(4):190]; 24. hu (H73(3):127). Vessels not to same scale.

1

2

3

4

5

6

7

128 9

10

11

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Early Phase

Late Phase

118 WANG YI

flat bottoms covered by false ring feet. Black-slip fine clay wares, though not common, were exquisitely made and decorated, with vessel forms including high-stem dou, vessels with openwork, tall ring feet, ring-foot pan, waist-ridged vessels, and vessel covers. Grayish-white fine clay ves-sels with platform-shaped, tall ring feet decorated with fine openwork resemble those of the Shijiahe 石家河 Culture from the middle reaches of the Yangzi River. The majority of pottery uncovered at Yufucun, however, is constituted of gray-brown coarse ware (Chengdu Shi Wenwu Kaogu Yanjiusuo 2001). The preliminary report for the 1996 excavation recognizes three phases at the Yufucun site, but Phase 2 apparently follows Phase 1 closely, whereas a rather distinct gap occurs between Phase 2 and Phase 3. The Yufucun walls were likely constructed in Phase 2 and abandoned by the end of this same phase (Chengdu Shi Wenwu Kaogu Gongzuodui et al. 1998). Given the close relationship between the first two phases, here we will regard them as two subphases of a single phase—the Early Phase—while our Late Phase for the site corresponds to the report’s Phase 3.

The Mangcheng site in Dujiangyan

In November 1996 and March 1997, the Chengdu Municipal Cultural Relics and Archaeology Team conducted a survey and test excava-tions at the Mangcheng 芒城 site (30° 52' N, 105° 35' E, elevation 675 m above sea level) (Mangcheng 1999), located in Qingcheng 青城 Township, approximately 12 km south of Guankou 灌口 Township in Dujiangyan 都江堰 City. About 2.4 km west of the site is Yaowangshan 藥王山 Mountain, part of the Qingchengshan 青城山 mountain range, and 1.4 km east is the Bojiang 泊江 River, flowing from the north to the south. The walled area at Mangcheng forms a relatively regular rectangle oriented at 10° (Figure 7). The walls delimit an inner and an outer enclosure, with the inner enclosure being better preserved. From north to south, the inner walls are about 300 m long, and they are 240 m wide from east to west, covering an area of 7 ha. The inner walls measure 5 to 20 m thick, and they are preserved to heights of 1 to 2.2 m above the surface. The gap between the inner and outer walls of the settlement is 20 m. It is estimated that the outer walls have a width of 300 m from the east to the west, and a length of 350 m from the north to the south, encompassing about 10 ha. The outer walls are 7 to 15 m thick and 1 to 2.5 m high. Excavated ash pits display a variety of shapes—rectangular, oval, and irregular. What remains of the building foundations are trenches for wall foundations that also have concentrations of postholes in them, suggesting that the mud walls were constructed on a wood frame.


Artifacts unearthed at Mangcheng include pottery and stone imple-ments. The pottery consists of two general wares, fine clay and coarse, with the former being the most abundant. Among the fine clay wares, grayish-yellow pieces are most frequent, while gray or brown is most common among the coarse pottery. Vessels were primarily shaped by hand and finished on a slow potter’s wheel. Coarse vessels were frequent-ly decorated with cord-markings, while fine clay vessels feature incised

Figure 7. Plan of the Mangcheng walled settlement site in Dujiangyan.

Mangchengsi

Liujiayuanzi

0 50 m

120 WANG YI

Figure 8. Typical pottery vessel forms from the Mangcheng site. 1, 2: ring-foot zun with plate-shaped mouth [H4:9P207,QJ2:3]; 3. tall-neck and trumpet-mouth guan [G4:265]; 4, 5. broad-rim and flat-bottom zun [T4(13):142, H5:1]; 6, 8, 11. cord-marked and scalloped-rim guan [G4:217, T6(4):22, H1:41]; 7, 12. hu [H8:11, T7(4):41]; 9. broad-rim pen [H11:7]; 10. flared-mouth and ring-foot zun [H11:15]; 13. dou (H4:1). Vessels

not to same scale.

designs. Vessel forms found include cord-marked and scalloped-rim guan, ring-foot zun, tall-neck and trumpet-mouth guan, hu, everted-rim guan, waist-ridged vessels, vessel covers, and ring feet (Figure 8). Recovered stone implements consist of polished axes, adzes, and chisels. In 1998, an excavation was undertaken on the western part of the northern area of the Mangcheng walled settlement (Zhong Ri Lianhe Kaogu Diaochadui 2001a). It was discovered that the settlement’s twin walls and two moats had been built at about the same time, but the inner wall was later repaired and reconstructed. According to the stra-tigraphy of the inner and outer wall ditches, we can assign the features of the inner and outer walls as well as the inner and outer ditches to two phases. Test excavations revealed that the western, northwestern, and northern parts of Mangcheng are mainly areas of cultural deposits, while the central and south-central parts were the residential areas; the situation in the eastern part of the settlement remains unclear. Building No. 5 (F5) at Mangcheng lies at the southeastern corner of the site (Figure 9). It is rectangular, consisting of a front and a rear cham-ber, which together constitute a residential unit. The building extends 8.8 m north to south, and 5.6 m east to west, covering a total area of 50 sq m. The front chamber, located at the northern end of the building, measures 5.6 by 5.2 m. The doorway to the outside, in the middle of the

1

2

3

4

5

6

7 12

8

9

10

11

13


Figure 9. Building foundation F5 at the Mangcheng site. After Zhong Ri Lianhe Kaogu Diaochadui 2001a: 88, fig. 29.

western wall, is 0.83 m wide. The smaller rear chamber measures 3.6 by 5.6 m and is linked to the outer chamber by a doorway measuring 0.8 m wide located towards the eastern end of the partition wall. The front chamber had been filled with a layer of earth about 5 to 20 cm thick that was absent from the rear chamber. The front chamber also features a hearth measuring 0.5 to 0.6 m in diameter. F5 was constructed with mud walls supported by bamboo posts ranging from 3 to 5 cm in diameter. The posts were densely arrayed in wall foundation trenches 16–23 cm wide. The roof is conjectured to have been sloped on both sides. This is the first time that such a construction technique has been documented at one of the prehistoric walled settlements in the Chengdu Plain. In 1999, during a large-scale excavation, a number of wall trenches and postholes were discovered in the western and southwestern parts of Mangcheng. These featured an orderly arrangement and possibly related to buildings. In addition, a test trench was dug outside the eastern wall, providing substantial evidence that there was an outer wall and an inner moat in the eastern part of the site (Zhong Ri Lianhe Kaogu Diaochadui 2001b). In 2000, the Chengdu Municipal Institute of Cultural Relics and Archaeology and Waseda 早稲田 University, in their third collaborative

hearth

settlementwall Song D

ynasty trench

A A´

B´

B A´

A

B

B´

0 1 m

N

122 WANG YI

effort, carried out excavations at Mangcheng. A 10-meter wide break in the inner and outer walls was found at the southwestern corner of the settlement. Moreover, it was determined that the outer walls were not connected in the northeastern corner, allowing water from the southern and northern inner moats to flow out from there into a swamp outside the eastern side of the settlement. It is likely that the inner walls were repaired and constructed for a second time. There is also evidence for flooding of the entire site at least once following the completion of the walls.4 The Mangcheng site was in use for a relatively short amount of time. The preliminary report for the 1996–1997 excavations recognized only one phase (Mangcheng 1999), but the materials from the 1998 exca-vation have made it possible to divide the occupation into two phases (Zhong Ri Lianhe Kaogu Diaochadui 2001a), and the 1999 excavations have further revealed that the walls were constructed in the second phase (Zhong Ri Lianhe Kaogu Diaochadui 2001b).

The Gucheng site in Pixian County

In 1996, the Chengdu Municipal Cultural Relics and Archaeology Team and the Pixian 郫縣 Museum carried out survey and test excavations at the Gucheng 古城 site (30° 54' N, 105° 55' E, elevation 565 m above sea level) (Gucheng 1999). The site, located at Gucheng and Zilu 梓路 villages in Sandaoyan 三道堰 Township, is approximately 8 km north of Pitong 郫筒 Township in Pixian County. North of the site, 3.2 km away, is the Qingbaijiang 青白江 River, and 2.5 km to the south is the Baitiaohe 柏條河 River.The walled enclosure at Gucheng is rectangular and oriented 120° (ESE), measuring 620 m by 490 m and enclosing an area of 30 ha (Figure 10). The walls measure 8 to 40 m wide and 0.8 to 5 m high, and a 12 m wide gap was identified in the northern end of the east wall. The excavations revealed that Neolithic cultural deposits were covered by Han period deposits, although in one excavated locale, a layer of grayish-yellow sandy soil was detected above the Neolithic deposits and under the Han stratum. This layer, though, featured a po-rous fill that contained no artifacts. The Gucheng enclosure walls were regular in shape. Although no excavation was done through the walls themselves, sherds collected from exposed sections of the walls were identical to the Neolithic sherds of the test pits. The exact relationship between the walls and the settlement, though, needs further analysis. Coarse wares with gray interiors and brown exteriors as well as brown wares predominate in the pottery assemblage. In the Early Phase of the

4 A report of the 2000 excavation is forthcoming.


site, coarse fabric pottery and fine clay wares each account for half of the total assemblage, while in the late phase, brownish, coarse fabric wares become predominant. Typical vessel forms include cord-marked and scalloped-rim guan, tall-neck and trumpet-mouth guan, and ring-foot, curved-rim guan (Figure 11). Stone tools are generally small in size (Song Zhimin 2000). From late 1997 to January of the following year, survey was carried out at the Gucheng site, along with the excavation of a section through the city wall that has thrown light on the construction method and the date of the wall (Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan 2001a). The wall was constructed in two phases. The first construction episode occurred in the Early Phase of a two-phase peri-odization of the site, while the second construction epsiode, on top of the earlier wall section, probably took place early in the Late Phase. During this field season, a long, narrow, vertical pit tomb was also excavated

Figure 10. Plan of the Gucheng walled settlement site in Pixian County.

Wangjiayuanzi

Hejiayuanzi

Zhujiayuanzi

Dujiayuanzi

Lijiayuanzi

Zhangjiayuanzi

Tiejushan

Diaojiayuanzi

0 200 m

124 WANG YI

Figure 11. Typical pottery vessel forms from the Gucheng site. (1–8 belong to the Early Phase, 9–26 to the Late Phase). 1. constricted-mouth guan [T3(12):11]; 2, 3, 10, 11. cord-marked and scalloped-rim guan [T3(11):78, H14:1, T5(13):92, H3:1]; 4. hu [H9:69]; 5, 22. flared-mouth and ring-foot zun [H9:17, 97T1614(6):9]; 6, 21. ring-foot zun with plate-shaped mouth [H10:1, T5(13):102]; 7. dou [H22:24]; 8, 17. tall-neck and trumpet-mouth guan [H14:3, 97T9(6):9]; 9. constricted-mouth guan [T5(13):94]; 12. narrow-rim pen [T3(9):43]; 13, 14, 18, 20, 23. narrow-rim guan [T2(9):49, T2(9):49, T4(11):87, T5(12):31, T5(12):13]; 15. curved-rim guan [T5(13):129]; 16. broad-rim pen [T5(13):135]; 19, 24. broad-rim and flat-bottom zun [T4(13):142, H22:35]; 25. folded-

rim and deep-belly guan [H22:34]; 26. bei [T4(13):150]. Vessels not to same scale.

1

2

3

4

5

6

7

12

8

9

10

11

13

14

15

16

17

18

19

20

21

22

23

24

25

26

Early Phase

Late Phase


Figure 12. Plan of building foundation F5 at the Gucheng site in Pixian County. Also see Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan 2001a: 58–60.

featuring a prone, extended occupant. No burial objects were found. Additionally, four building foundations, three small and one large, were unearthed. The small-sized foundations lie east of the center of the site. Rectangular in plan, they are all ground-level structures with mud walls built over wooden frameworks. Building No. 6 (F6) was bet-ter preserved. It measured 8.3 by 5.6 m, covering an area of 46 sq m. Six post holes were arranged in a cruciform shape. A 1.2-meter wide opening, probably a doorway, lied south of the western wall, facing the center of the site, and was paved with cobblestones 10 cm in diameter. A hearth was also discovered. It is a rectangular, shallow pit filled with cobblestones of various sizes. Building No. 5 (F5), the largest among the buildings found, is located at the center of the site: this large foundation was one of the major discoveries from the 1997–1998 season. F5 has a rectangular plan, and its northwest to southeast axis parallels that of the city (Figure 12). The structure measures 50 by 11 m, occupying an area of 550 sq m. Only the foundation has survived, and its western portion had sustained considerable damage during Han times. The layout of the building is still discernible, and its construction sequence was as follows: a large foundation pit was dug, and surfaces of cobblestones 4 to 6 cm in size were laid around the edges of the pit (each of these cobblestone surfaces covered an area 0.9 to 1 m wide and was 0.1 to 0.28 m deep). After wooden posts were erected on the cobblestones, the pit was then filled with soil and a large quantity of fired, reddish earth that protected the construction against moisture. Postholes found among the cobblestones were neatly spaced 0.7 to 1.2 m apart, and they measured 20 to 30 cm in diameter. It is presumed that the walls of the structure were made of

Han pit

Platform5

Platform4

H20

Platform3

H23

Han pit

Songpit

Han pitPlatform

2

Platform 1

H21

Handitch

N

0 5 m

126 WANG YI

wooden posts, the interstices between which were filled with bamboo matting. The matting was then covered on both sides with a mud-straw mixture. No traces of any postholes from the interior of the structure survive, but a pile of cobblestones about 65 cm in diameter was uncov-ered at the center of the eastern part of the building, and this possibly had been used to reinforce vertical supports, as a structure as large as this one would have needed internal supports for the roof beams. No internal partition walls were found, either. There are, however, five or-derly rectangular heaps of cobblestones lined up from east to west. Base pits had been dug around these cobblestones into which concentrations of round bamboo posts, now carbonized, were inserted. These posts might have formed a protective wall, against which cobblestones were piled up on the interior to form platforms on which daub might have been spread. These five platforms, numbered 1 to 5 from east to west, are spaced about 3 m apart. Platforms No. 4 and No. 5 were damaged by Han deposits, and their cobblestones had been taken away, but traces of their base pits were still visible. One to two layers of cobblestones remained from platforms Nos. 1, 2, and 3. The cobblestones range from 10 to 30 cm in length, with the average being 20 cm long. Platforms Nos. 2 and 3 are larger than the others, measuring about 5 m long from east to west and 2.7 m wide from north to south. No. 1 is 3.4 m long from east to west, and 3 m wide from north to south. No. 4 is 3 m long from east to west, and 2.5 m wide from north to south. No. 5 is 2.75 m long from east to west, and 2.35 m wide from north to south. In addition to the rarity of such a large structure in Neolithic China, F5 is also unique in terms of its architectural style. It is likely to have been a site for large-scale rituals. It was built during the Early Phase of the Gucheng site and was abandoned at the end of the Late Phase. From December 1998 to January 1999, the Chengdu Municipal Institute of Cultural Relics and Archaeology conducted a third full-scale excavation at Gucheng, working mostly in loci Nos. I and IV (Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan 2001b). In addi-tion, further excavations were carried out around the large Building F5 in locus No. III. 5 foundations, 11 ash pits (mostly irregular in shape), and 1 ash trench were discovered. After analyzing an opening at the northern part of the east wall, it can be postulated that this might have been the location of a city gate.A moat was discovered outside the walls, but its extent has not been determined. Building No. 13 (F13), found situated in the northwestern part of the Gucheng site, features a different floor construction than the other buildings so far discussed. A series of wooden stakes were driven into the ground to support a wooden floor elevated from the ground sur-face. Posthole remains found buried among cobblestone supports were


set in a neat order, nine by nine. The structure has a rectangular plan, measuring 5.3 by 4 m and occupies an area of 21.2 sq m, with the main walls oriented to 38° (NE). Building No. 10 (F10), a small building with mud walls with an internal wooden framework, lies 70 m south of the large Building F5. It has a roughly rectangular plan, measuring 5.2 by 4.3 m and encloses an area of 22 sq m. Its doorway, about 1 m wide, is located in the central section of its western wall, facing F5 at the site’s center. Outside the foundation of F10 is a surface covered with small cobblestones that possibly had been used for outdoor activities. Excavations show that the Gucheng site evidently spanned two phases. Pottery seriation allows us to further divide the Early Phase into two subphases.

The Shuanghe site in Chongzhou

In November of 1996, while the preliminary excavation at the Mangcheng site was underway, we were informed that a place called “Xia Mangcheng” 下芒城 (“Lower Mangcheng”), in Shuanghechang 雙河場 (the old name for Shangyuan 上元 Township) of Chongzhou 崇州 City, also had a double-ring walled enclosure resembling that of Mangcheng. In December of that year, we surveyed the site and con-firmed the news. In order to differentiate it from the other site named Mangcheng, in Dujiangyan, the newly discovered site was named Shuanghe 雙河. In the winter of 1997, the Chengdu Municipal Cultural Relics and Archaeology Team conducted fieldwork at Shuanghe (30° 46' N, 103° 36' E, 586–595 m above sea level) (Chengdu Shi Wenwu Kaogu Gong-zuodui 2002). The site is 16 km north of Chongyang 崇陽 Township in Chongzhou City; it is 45 km to the northwest of Chengdu, 5 km northeast of the Xihe River, and 10 km southwest of the Minjiang River. The Shuanghe site’s walled settlement has a rectangular plan. Remnants of walls are found on the north, east, and south sides, with the eastern wall being preserved the best. The western wall was either never built or perhaps washed away when the local river changed course during ancient times. The walls form an inner and outer ring with a ditch in between (Figure 13). The outer northern wall is 325 m in length, while the inner northern wall is 270 m long. The outer eastern wall has a length of 500 m, and the inner eastern wall is 420 m long. The outer southern wall measures 120 m, and the inner southern wall is 110 m. The ditch ranges from 12 to 20 m in width. The walls have a height of 2 to 3 m, and a width of 18 to 30 m at the top. In total, the settlement covers an area of 11 ha. Among the remains there were ash pits, building founda-tions, and cobblestone features, with ash pits being the most plentiful.

128 WANG YI

One of the building foundations (F1) features a peculiar construction pattern, as no traces of wall foundation trenches were found, and the excavations revealed only 14 postholes and one cobblestone surface. The 14 postholes are arranged in a cruciform pattern and are either round or oval in plan, and partially filled with compacted, yellow earth; cobblestones were then either laid at the bottom of the hole or positioned at the mid-portion of the postholes, apparently as a load-bearing device

Figure 13. Plan of the Shuanghe walled settlement in Chongzhou City.

Linjiayuanzi

Weijia-yuanzi

MangchengElementary

School

ShangyuanElectric

Plant

Pangxie RiverE

lectri

c Pla

nt D

itch

0 100 m


when the structure was erected. Among the walled settlements excavated on the Chengdu Plain, only one other example of such a construction method has been found, at the Gudunzi 鼓墩子 locus of the Baodun site. Coring at Shuanghe shows that the F1 foundation covered an area of over 70 sq m: the layer comprising the foundation was about 25 cm thick, and its fill was densely packed and contained a mix of cobble-stones, charcoal pieces, and nodules of fired, red earth. The surface of this layer was level, and below this layer was sterile fill. To the west of F1 lies a surface of closely-fitted cobblestones laid directly above the sterile soil. The surface is lower than the layer of the postholes and descends gently away from the building, and it probably served as a walkway or for water dispersal. Most of the artifacts unearthed at Shuanghe are pottery vessels. There are also a few stone implements. The pottery includes fine clay and coarse fabric wares. The majority of the fine clay ware is gray-yellow and relatively soft in texture, while gray-brown pottery is the most abundant among the coarse fabric pottery. The vessels were all formed by hand and then finished on a slow potter’s wheel. The majority of the fine clay vessels are plain surfaced, while the coarse fabric pottery was often decorated with cord-marks (in unidirectional or interlocking patterns). Common decor includes appliqué, fine lines, combed, open-work, stamped, or incised patterns, as well as crescent designs and ring patterns. Characteristic pottery vessel forms include flat-bottom and ring-foot vessels, with the former outnumbering the latter. Among the vessel types are cord-marked and scalloped-rim guan, broad-rim zun, tall-neck and trumpet-mouth guan, folded-rim guan, ring-foot zun and pan. In addition, there are a small number of stems from dou and cylin-drical vessels. The lithic inventory consists of flaked tools, microliths, and polished-stone tools. Flaked stone tools include axes and knives, microliths include cores and flakes, and polished tools include axes, adzes, and chisels. A stone axe with three perforations was exquisitely made and stylistically unique, showing no traces of wear. Excavated remains from the Shuanghe site allow periodization into two phases, Phases 1 and 2.

The Zizhu site in Chongzhou

During our preliminary excavation at Shuanghe, we also discovered the nearby site of Zizhu 紫竹, also in Chongzhou.5 In March, 2000, the Chengdu Municipal Institute of Cultural Relics and Archaeology and the Institute of Archaeology of the Chinese Academy of Social Sciences

5 An excavation report of the Zizhu site is forthcoming.

130 WANG YI

carried out survey and test excavations at the site. Test squares were excavated in the southeast and southwest of the site. In addition, the western and northern walls of its inner enclosure were analyzed, and it was determined that the inner enclosure walls were erected by the piling of earth into a sloping shape. Neolithic remains located include nine ash pits, a red fired clay pile, and one posthole. The ash pits are irregular in shape. Gray fine clay ware is the most abundant pottery found at the site. Artifacts include tall-neck and trumpet-mouth guan, cord-marked and scalloped-rim guan, flared-mouth guan, double-handle zun with folded rims and straight bellies, ring-foot zun, hu, and spindle whorls. Pottery decoration includes a range of cord-marked, combed, appliqué, stamped, and incised motifs. Lithics include axes, adzes, and chisels. Additionally, whetstones and a small number of bone awls were found. The cultural contents of this site resemble what is found at Baodun and nearby Shuanghe.

Major characteristics of the Baoduncun Culture

The cultural assemblages of all six of the sites described above can be identified as coming from the same archaeological culture—aNeolithic culture that was previously unknown. There has been some debate over how to name this culture (which we now call the Baoduncun Culture), and this debate stems from the relationship between this culture and previous finds at the Sanxingdui site. According to Chen De’an 陳德安 (Sichuan Sheng Wenwu Kaogu Yanjiusuo 1999: 16, note 15), one of the lead excavators for Sanxingdui, the Sanxingdui site spans a time period equivalent to beginning in the Longshan 龍山 Culture period and extending into the late Shang and early Western Zhou periods of the Central Plains, with the major components of the Sanxingdui site dating to the historical chronological periods of the Xia, Shang, and early Western Zhou dynasties. The term Sanxingdui Culture should thus refer to remains of this period. The Neolithic deposits at Sanxingdui, dating earlier than this main component of the site, have been labeled the Lower Sanxingdui Culture, or the Sanxingdui Phase I Culture (Sichuan Sheng Wenwu Kaogu Yanjiusuo 1999: 16, note 15), so some researchers have wanted to apply these Sanxingdui-based cultural names to these other Neolithic sites as well. Other researchers have emphasized the importance of the Biandui-shan site. In the words of one report, “[The Bianduishan] site most likely represents a new, distinctive type of indigenous culture found in the hilly areas north of Chengdu in the Sichuan Basin,” and “its determined date coincides with the early period of the Longshan Culture in the Central


Plains, or even earlier” (Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Sichuan Gongzuodui 1990). He Zhiguo (1993) has thus proposed the name the “Bianduishan Culture” for cultural assemblages from the time period of the Bianduishan archaeological site. Sun Hua (1993: 23) suggested that Phase I of the Sanxingdui Culture together with other remains at related sites of the same time period be separated from the Sanxingdui Culture and subsumed under the Bianduishan Culture, as it is represented by the stratigraphically less-complicated Bianduishan site. This earlier nomenclatural debate has been rendered obsolete by fieldwork done since the mid-1990s. The six prehistoric walled settle-ments excavated by the Chengdu Municipal Institute of Cultural Relics and Archaeology display a great deal of consistency in their cultural assemblages. They can therefore be defined as all belonging to a new archaeological culture that should be named after the Baodun site, as this is not only the largest site, but also one of the best preserved among all the sites excavated. The geographical scope of this Baoduncun Culture is limited to the Sichuan Basin, and it is centered on the city of Chengdu and distributed around the Chengdu Plain. The Baoduncun Culture settlement pattern includes not only small villages, but also large-scale settlements enclosed by thick, high walls of piled earth (Wang Yi and Sun Hua 1999). These walls, described above, are the most notable feature of the Baoduncun Culture. Within the Baoduncun Culture walled settlements, various architec-tural features have been documented. Buildings were all constructed at ground level, and they can be characterized into two types: Type I features foundation trenches for the walls, and Type II features the construction technique described for building F13 at Gucheng, with a series of wooden stakes driven into the ground to support a wooden floor elevated from the ground surface. In Type I, a framework of wooden or bamboo poles are set into the foundation trenches, and the building walls are formed by mud daub applied to the framework. Generally, the regular outlines of the wall trenches and a few postholes in them are preserved, and thus, the orientation and interior divisions of the buildings can be discerned. The wall foundations, as well, can be divided into two types. In the first, such as those found in F5 at the Mangcheng site, the foundation trench is simply filled with earth. In the second type, the bottom of the trench is paved with cobblestones, such as in F5 at the Gucheng site. In addition to F13 at Gucheng, several other examples of Type II buildings have been found, but they are less common than the buildings with wall trenches. In these buildings with elevated floors, only regular rows of postholes or pedestal stones have been unearthed.

132 WANG YI

Even more common than above-ground structures, pits are the most ubiquitous archaeological features at Baoduncun Culture sites. They are represented by a number of regularly shaped ash pits found in the course of our excavations. A significant portion of these is clearly fashioned with considerable care, with smooth, vertical walls, rather than being makeshift pits for waste disposal. Large quantities of artifacts, including pottery and stone implements, have been recovered from these pits. Although to date no direct evidence for workshops has been found at any of the Baoduncun Culture sites, the distinctive shape of a few ash pits and some special items found in them suggest that they were related to the firing of pottery or the manufacturing of stone implements. Baoduncun Culture burials were found at the Baodun, Yufucun, and Gucheng sites. The burials all feature rectangular, vertical pits that are shallow, with long, narrow sidewalls and flat bottoms. No coffins or other burial containers were found in the graves, and the orientations of the occupants are inconsistent. When several tombs are found clus-tered together, however, the orientations are nearly identical. Tombs of adults and children differed only in the size of the burial pits. No burial goods were unearthed except for a small number of potsherds in the fill. All tomb occupants were placed on their backs in an extended position. The principal production tools used by the Baodun people were stone implements. These are typically thoroughly polished and small in size. Large stone implements are rarely found. Adzes, axes, and chisels are typical, but there are also a few knives and spades. Adzes were meticu-lously polished and are simple in form. They generally have a straight blade and a curved top, with the blade being wider than the top. They are trapezoidal in cross section. Axes were mostly shaped into long rectangular forms, with a few being trapezoidal. Axe tops are narrower than the blades. Both the blades and the tops are curved. The axes range in size from 6 to 14 cm in length. Chisels were elegantly made of fine stone, in some cases of jade-like material. Some are long, flat pieces, and others are pointed-pentagonal forms. Some are long and narrow with concave blades, while others have two cutting edges (one pentagonal and the other straight-edged). Knives and spades, which are few in number, have perforations drilled from both sides. Knives are rectangular, with cutting edges on both ends. A few small chert microliths and whetstones of red sandstone were also discovered. In addition, a small quantity of pottery tools such as spindle whorls and net sinkers were unearthed. The spindle whorls were variously shaped and undecorated, having no particular distinctiveness. Aside from tools, some weapons were also found. All were made of stone, including arrowheads, spears, and yue 鉞 axes. There are no signs


to indicate that they had ever been used. They may have been associ-ated with early ritual customs. The main category of utilitarian objects preserved in the Baoduncun Culture is pottery. The Baoduncun vessel assemblage is characterized by flat-bottomed and ring-footed objects. Most have a broad rim or a wide, flared mouth. The mouths of these vessels were often decorated with cord markings or scalloped rims. All sites had fine clay and coarse fabric wares. The latter were probably primarily used for cooking and include such vessel forms as cord-marked and scalloped-rim guan and flared-mouth guan. Fine clay pottery probably served mostly as containers; the most prominent forms include tall-neck and trumpet-mouth guan, broad-rim guan, and flat-bottom zun. The coarse fabric pottery tends to be brown or gray, while the fine clay ware varies in color from grayish-white to grayish-yellow, brown, and black slip. The grayish-white ware was generally fired at a higher temperature. Those fired at a lower temperature show rough, powdery surfaces. Grayish-yellow fine clay ware was fired at a lower temperature and was often coated with a black slip that tended to easily come off during the excavation process. Baoduncun pottery was manufactured by coiling clay strips by hand and then finishing the vessel on a slow potter’s wheel. Coiling and rotat-ing marks are sometimes visible on a vessel’s interior wall. For ring-foot vessels, the ring foot and the body were formed separately and then joined. Traces of reinforcement are discernible on both the exterior and interior sides of the join. The coarse fabric ware vessels with thick bottoms were also constructed in two stages. There are clear traces that the assembly was from the base to the body. The exterior is commonly decorated with cord-markings, and the profile of these vessels often features a false ring-foot. Some of the rims of the coarse fabric ware vessels also show traces of a two-stage construction. The coarse fabric ware ring-foot zun often are burnished on the interior of the vessel rim. Some of the fine clay ware vessels have clay disk appliqués on the interior walls, and the collars of some of the fine clay hu vessels show traces of two-stage construction. The surface decoration of the Baoduncun Culture coarse fabric pottery predominantly consists of cord-markings applied mostly on the rims, lips, shoulders, bellies, and bases of vessels. Those decorating the rims, lips, and bases are quite distinct in appearance, including verti-cal, oblique, interlocking, and webbed patterns. Other common decor includes stamped and appliqué designs. Additionally, there are a few pieces decorated with incised, bowstring, crescent, and circular designs, often on the belly and on ring-foot vessels. On the fine clay wares, incised, stamped, and clay-strip designs are prominent. The incised designs consist of wave motifs, parallel lines, and geometric patterns formed by

134 WANG YI

parallel lines. Stamped designs are primarily obliquely stamped dotted patterns and long strip designs. There are also a small number of fine lines, ridges, and bowstrings applied to the mouth, shoulder, and belly of vessels. The tall-neck, trumpet-mouth guan and hu with rims impressed in scalloped shapes are distinctive. A black slip was primarily applied to the grayish-yellow pottery and the grayish-white pottery that were fired at a low temperature. Frequently found utilitarian wares include cord-marked and scalloped-rim guan, tall-neck and trumpet-mouth guan, broad-rim zun, ring-foot zun, and hu. Apart from these, there is a small number of deep-belly guan, rolled-rim guan, cylindrical vessels, folded-belly bo 砵 bowls, narrow-rim guan, curved-rim guan, small-flat-bottom guan, constricted-mouth guan, and constricted-mouth weng 瓮 urns. In the archaeological remains of the Baoduncun Culture, there are very few ornamental items or pieces that could be called art works. This was perhaps due to burial conditions and the environment of the Chengdu Plain. At the Yufucun site, perforated pieces of turquoise were discovered. Some unique and exquisitely decorated pottery vessels suggest a strong sense of ornamentation, such as we see with the black slip, the highly burnished surfaces, and the elegant ridged patterns of various vessels, as well as the black painted designs on pottery sherds unearthed at the Yufucun site. Baoduncun Culture remains of possible religious significance include objects and architecture. Meticulously made objects lacking any trace of wear, such as the stone yue, spears, and arrowheads could be considered ancient religious or ritual objects. The large structure discovered at the center of the Gucheng site in Pixian most likely served as a place for ritual activities.

Periodization and absolute chronology of the Baoduncun Culture

The large amount of pottery sherds excavated at four of the six sites discussed above—Baodun, Yufucun, Mangcheng, and Gucheng—have made it possible to establish developmental sequences for each site based on seriation.Correlating these sequences across sites has further enabled us to produce a periodization scheme for the Baoduncun Culture.6 The periodization within each site is as follows:

Baodun: Early Phase (Subphases 1 and 2) and Late Phase (Subphase 3).

6 Although the Shuanghe site is understood to consist of two phases, they are yet to be correlated with the periodizations in the other sites.


Yufucun: Early Phase (Subphases 1 and 2, which correspond to the excavation reports’ Phase 1 and 2) and Late Phase (Subphase 3, which corresponds to the reports’ Phase 3).

Mangcheng: Early Phase and Late Phase.

Gucheng: Early Phase (Subphases 1 and 2, which correspond to the excavation reports’ Early Phase) and Late Phase (Sub- phase 3, which corresponds to the reports’ Late Phase).

Cross-correlation of similar pottery vessel types across sites allows us to then construct an overall four phase (with further subphases) periodi-zation scheme for the Baoduncun Culture, as follows:

Phase I: Represented by the Early Phase of the Baodun site. As such, Phase I can be further divided into two subphases corresponding to those of the early phase of the Baodun site.

Phase II: Represented by the remains at the Mangcheng site (both phases of the Mangcheng site belong to this phase) and the Late Phase of the Baodun site.

Phase III: Represented by the Early Phase of both the Gu-cheng site and the Yufucun site. Phase III can be further divided into two subphases: Early and Late, with the former represented by Subphase 1 of the Gucheng site and the Early Phase of the Yufucun site, and the latter by Subphase 2 of the Gucheng site.

Phase IV: Represented by the Late Phase of both the Yufu-cun site and the Gucheng site.7

An absolute chronology for the Baoduncun Culture can tentatively be established through a very limited number of radiocarbon determi-nations now available from several sites. These are listed in Table 1. Determinations from samples from the two features at the Baodun site suggest a date between ca. 2600–2300 cal. BC. The calibrated Gucheng site dates presented in Table 1 are more problematic, with BK98009 in the same approximate range as the Baodun dates, but with BK98010, from the same feature (F5 fill) dating several centuries later, to ca. 2200–1875 cal. BC. Because of the close connections apparent in pottery typology be-tween Phase IV of the Baoduncun Culture (corresponding to Phase 3 of the Yufucun site as given in the excavation reports) and the slightly

7 Illustrations and descriptions of the vessel forms found in Phases I–IV can be found in Chengdu Shi Wenwu Kaogu Yanjiusuo et al. 2000: 108–113 figs. 67–72 and Jiang Zhanghua et al. 2001.

136 WANG YI

Tab

le 1

: Rad

ioca

rbon

Dat

es fo

r th

e B

aodi

ncun

Cul

ture

Site

Lab

NoPr

ove

nien

ceSa

mpl

e M

ater

ial

14C

Dat

e (B

P)

T½

=55

68

Cali

brat

ed A

ge

(wit

h r

elat

ive

area

)

1σ (6

8.3%

)

2σ

(95.

4%)

Bao

dun

GrA

5726

1995

Zhe

nwu-

guan

Cha

rcoa

l fra

gmen

ts

with

fine

min

eral

pa

rtic

les

3965

±60

2572

bc

(0.3

8) 2

512

bc25

05 b

c (0

.41)

243

5 bc

2421

bc

(0.0

8) 2

404

bc23

79 b

c (0

.13)

234

9 bc

2631

bc

(0.9

9) 2

286

bc

Bao

dun

GrA

5717

1996

Gud

engz

i bu

rial

site

Cha

rcoa

l fra

gmen

ts

(woo

dy)

3950

±50

2565

bc

(0.2

2) 2

532

bc25

28 b

c (0

.01)

252

5 bc

2496

bc

(0.4

5) 2

432

bc24

24 b

c (0

.13)

240

2 bc

2381

bc

(0.1

9) 2

348

bc

2575

bc

(1.0

0) 2

296

bc

Guc

heng

BK

9800

919

97 C

PGF5

fill

Bam

boo

char

coal

3905

±85

2546

bc

(0.0

1) 2

541

bc24

88 b

c (0

.91)

228

1 bc

2249

bc

(0.0

6) 2

231

bc22

18 b

c (0

.18)

221

3 bc

2620

bc

(1.0

0) 2

137

bc

Guc

heng

BK

9801

019

97 C

PGF5

fill

Bam

boo

char

coal

3650

±70

2135

bc

(0.3

2) 2

069

bc20

64 b

c (0

.68)

193

8 bc

2273

bc

(0.0

1) 2

257

bc22

07 b

c (0

.97)

187

5 bc

1843

bc

(0.0

1) 1

817

bc17

98 b

c (0

.01)

177

9 bc

Not

e: c

alib

ratio

ns b

y C

AL

IB v

. 5.0

1 us

ing

IntC

al04

(Rei

mer

et a

l. 20

04)


later Sanxingdui Culture, it is relatively easy to deduce the dating of the Baoduncun Phase IV assemblage. Some vessels from the Yufucun Phase 3 assemblage, such as the small-flat-bottom guan, are also found in Phases I and II at Sanxingdui, while Yufucun-type vessel lids can only be found in Sanxingdui Phase II. Many vessel forms from Yufucun Phase 3, such as cord-marked and scalloped-rim guan, tall-neck and trumpet-mouth guan, and ring-foot forms, are part of the typical Baoduncun Culture assemblage, while other forms are unique to Yufucun and not found in other Baoduncun Culture sites, such as the curved-rim guan and constricted-mouth guan and weng; these unique Yufucun vessels are not found in the Sanxingdui Culture pottery assemblage, either. We must also note that the characteristics of the Yufucun Phase 3 pottery system, with its majority of wares being coarse fabric and its tendency toward brown ware, as well as its styles of décor and placement of orna-mentation, are generally close to the pottery system of the Sanxingdui Culture. Based on this overall comparison, we can deduce that Phase IV of the Baoduncun Culture should date between Phases I and II of the Sanxingdui Culture: this would be equivalent to the period from the end of the Longshan Culture to the beginning of the Xia Dynasty, or approximately 4000–3700 BP. Phase II was the pinnacle of the Baoduncun Culture. This was also the peak period of cultural developments in the Sichuan Basin during the Longshan Culture period. A number of sites from this time period are found throughout the Sichuan Basin: in the north of the basin there is the Bianduishan site in Mianyang; in the northeast, the Leiguzhai 擂鼓寨 site in Tongjiang 通江 County and the Yueliangyan 月亮岩 site in Bazhong 巴中 County; in the east, the Weijialiangzi 魏家梁子 site in Wushan 巫山 County near the Three Gorges of the Yangzi River; and in the southwest, the Shizishan 獅子山 site in Hanyuan 漢源 County. These sites have varying degrees of cultural affinity with the Baoduncun cultural tradition. For instance, Bianduishan has tall-neck and trumpet-mouth guan, cord-marked and scalloped-rim guan, and broad-rim zun. Leiguzhai and Yueliangyan have well-developed incised designs and saw-tooth rims, as well as a pottery manufacturing system involving the assembly of separate bases and body parts, like those of the Baoduncun Culture. Weijialiangzi’s various characteristics, such as incised wave patterns, vessels with plate-like mouths, bo, cylindrical vessels, and a particular patterning in the cord-marking of vessel bases, are similar to those of the Baoduncun Culture. Shizishan’s pottery is most frequently decorated with cord-marking or appliqué motifs, and the vessel types include tall-neck guan, and large-mouth and scalloped-rim guan; all these features can be found in the Baoduncun Culture also. These shared cultural characteristics are particularly outstanding in Phase II of the

138 WANG YI

Baoduncun Culture, suggesting wide-ranging communication between the Baoduncun Culture and other cultures in the Sichuan Basin. While each of these cultures had its own distinctive features, there was also mutual influence that produced shared cultural characteristics. These shared features clearly show that these cultures were contemporaneous, yet each continuously developed from its own primitive culture founda-tion. These Longshan period sites that were scattered across the Sichuan Basin mark a prosperous stage in the development of archaeological cultures in this region, and this period was also the first climax in the developmental history of ancient cultures in the upper Yangzi River valley. The date for this peak period in the Baoduncun Culture—Phase II—was about 4700–4300 BP. Phase I artifacts for the Baoduncun Culture, so far, have only been found at the Baodun site itself. The pottery system from Phase I, however, reflects evidence for a rather strong similarity to that of the upper stratum of the Dengjiaping site in Guangyuan, in the northern Sichuan Basin. In both places, coarse fabric pottery vessels were found in abundance. Cord-markings and appliqúe designs were widely em-ployed. Scalloped-rims and cord-marking were also popular treatments for vessel mouths. Additionally, stone implements tended to be small, consisting of axes, adzes, and chisels (Wang Renxiang and Ye Maolin 1993). Radiocarbon determinations suggest that the Dengjiaping site dates to around 5500–4700 BP. Only one of the four samples gave a later determination, of about 4200 BP. So, the date of 4700 BP could perhaps be taken as an indication for the lower limit of Phase I of the Baoduncun Culture. Typology and seriation, as well as radiocarbon determinations, in-dicate that Phase III of the Baoduncun Culture should date to around 4300–4000 BP. In short, from the above discussions, we see that the entire chrono-logical range of the Baoduncun Culture should extend from 5000 BP to 3700 BP.

Relationships with neighboring archaeological sites

We now can see that the Baoduncun Culture is dated to the same time period as the Longshan cultures in other regions of China. The distri-bution of Baoduncun Culture sites shows that its cultural center was in the area of Chengdu City, with other sites limited to the Chengdu Plain (Wang Yi and Sun Hua 1999). Contemporaneous and related cultural remains are found at other sites surrounding the Baoduncun Culture. These include, to the north, Zhongzipu, Zhangjiapo, Dengjiaping and


Lujiafen 魯家坋 in Guangyuan, and Bianduishan in Mianyang; to the northeast, Leiguzhai in Tongjiang and Yueliangyan in Bazhong; to the east, Shaopengzui 哨棚嘴 in Zhongxian 忠縣 County and Weijialiangzi in Wushan; and to the southwest, Shizishan in Hanyuan.

The northern-most Sichuan Basin

Wang Renxiang and Ye Maolin (1993) of the Institute of Archaeology, Chinese Academy of Social Sciences, provide a summary of the Zhong-zipu, Zhangjiapo, and Dengjiaping sites in Guangyuan, on the upper Jialingjiang 嘉陵江 River at the northern tip of the Sichuan Basin. Rela-tive dating and radiocarbon determinations allowed them to establish the following sequence for these Neolithic sites: 1. microlithic remains from Zhongzipu, dating to 7000–6000 BP; 2. remains from Zhangjiapo and the late phase at Zhongzipu (6000–5500 BP); and 3. Dengjiaping site remains (5500–5000 BP). Within this sequence, the later remains from the Dengjiaping site date to within the earliest stages of the Baoduncun Culture period. Dengjiaping also shares some cultural content with the Baoduncun Culture, such as an abundance of coarse fabric gray-brown pottery and gray fine clay ware, with only a small amount of black-slip pottery. Cord-marked or appliqué motifs, in addition to carved and incised designs, were common patterns. Cord-markings often adorned rims and lips of pottery vessels. Lithics, too, show many shared features. Small-sized stone implements, a distinctive feature of the Baoduncun Culture, are found at Dengjiaping, as are stone shovels, knives, and ar-rowheads like those of the Baoduncun Culture, such as were found at Baodun, Yufucun and Mangcheng, as well as the microliths from the Baoduncun Culture site of Shuanghe. These characteristics suggest that the Dengjiaping site and the Baoduncun Culture have consider-able similarities. Even so, there are remarkable differences in terms of the complete pottery assemblage, and so, following general classificatory standards, these two sets of cultural remains should not be regarded as belong-ing to the same archeological culture—they in fact belong to different cultural categories. This is further supported by the bulk of the Dengjia-ping cultural remains far predating the period in which the Baoduncun Culture developed. Therefore, the relationship between Dengjiaping and the Baoduncun Culture can tentatively be characterized as a tem-poral relation of earlier to later. A portion of the representative cultural remains of the Dengjiaping site were later absorbed into the Baoduncun Culture in the Chengdu Plain, and these became an integral part of the Baoduncun Culture. Thus, the cultural remains as represented by the Dengjiaping site constituted one source for the Baoduncun Culture.

140 WANG YI

The Zhangjiapo site, preceding and somewhat related to the Deng-jiaping site, has a similar relationship to the Baoduncun Culture. As regards Lujiafeng, its cultural profile is still unclear due to the scarcity of material from the site. But one thing is certain: the Lujiafeng site, together with other Neolithic remains in the northern Sichuan Basin, belongs to a distinctive type of primitive, indigenous culture that origi-nated in the Sichuan Basin.

Bianduishan and connections in the northern Chengdu Plain

Located at the northern edge of the Chengdu Plain, the Bianduishan site is situated in a transitional zone between the plains and highland regions. The central cultural features of Bianduishan, as described by Wang Renxiang and Ye Maolin (1993: 260), are:

The stone implements tend to be small. . . . The types of these stone implements include a large number of rectangular and trapezoidal axes . . . and a relatively large number of adzes and chisels of varied forms. Other items include shovels, knives, spears, arrowheads, pellets, whetstones and grinding stones. Flaked stone tools include choppers and scrapers, etc. The pottery can be divided into two categories: gray-brown coarse fabric ware and gray fine clay ware, with the former being greater in number. Additionally, there is a small amount of black-slip pottery and red-clay pottery. Cord-marking and appliqué motifs are the predominant decorations. Otherwise, there are incised designs and bowstring patterns, among which the awl-marked geometric designs are distinctive. Vessel rims decorated with cord-markings or tooth-shaped scalloped patterns particularly stand out. Vessel forms consist of flat-bottom types such as guan, gang 缸 urns, hu, pen 盆 basins, pan, and wan 碗 bowls, as well as ring-foot types like dou.

Many cultural items of Bianduishan, particularly many of its lithic and pottery forms, are, to various degrees, related to the Baoduncun Culture. Radiocarbon dates also show the Bianduishan site to have existed about 5000–4500 BP, corresponding approximately to Phases I and II of the Baoduncun Culture. Henceforth, Bianduishan can be considered as an exemplary site of the Baoduncun Culture distributed in this northern region (Wang Yi and Sun Hua 1999). Yet quite a number of cultural items at Bianduishan are distinctive to the site itself, such as the toothed, scalloped patterns on vessel mouths, which are extremely rare among Baoduncun remains. The deep-belly guan decorated with appliqué designs on the outer lip found at Bianduishan is absent from Baoduncun cultural remains; in contrast, it is found in the areas near the Yangzi Gorges in the eastern Sichuan Basin, such as at the sites of


Shaopengzui in Zhongxian County, and Weijialiangzi and Suolong 鎖龍 in Wushan County. Also, the commonly seen ring-foot zun of the Baoduncun Culture sites is not found at Bianduishan. This phenom-enon at Bianduishan may be linked to its geographic location. The Bianduishan site is located at the northern entrance to the Chengdu Plain, a position that dominates the pass to the western Sichuan Basin and stands in proximity to the Fujiang 涪江 River—this vantage would have made it easy to build contacts with the heartland of the Baoduncun Culture. Its north and northeast sides are connected to the highland areas of the northern Sichuan Basin. Its eastern side is connected to the Yangzi Gorge region, while its western side could develop connections with the western Sichuan Plateau. These multidirectional geographic connections available from the Bianduishan site allowed it to be an area of cultural confluence, yet judging by the majority of its cultural elements, Bianduishan can still be deemed as part of the Baoduncun Culture.

The northeastern Sichuan Basin: Leiguzhai and Yueliangyan

The Leiguzhai and Yueliangyan sites show the Baoduncun Culture’s connections in the northeastern Sichuan Basin. The basic cultural features of Leiguzhai are as follows: the majority of fine clay wares are either gray or orange-yellow in which incised designs and cord-marking are the most frequently seen decorations, and there are also appliqué and wave motifs on the necks of some vessels. Pottery was hand-formed and then finished on a slow potter’s wheel. For some vessels, their base and body parts were made separately, and then assembled from bottom to top. Flat-bottom pieces are the predominant pottery form, while there is also a small portion of ring-foot vessels. Saw-toothed and wave impressions are a popular vessel lip treatment. Such pieces have some similarities to those of the Baoduncun Culture. The early phase of the Leiguzhai remains has an associated radiocarbon date of 4995±159 BP, which is slightly earlier than the major period of the Baoduncun Culture (Lei Yu and Chen De’an 1991; Sichuan Sheng Wenwu Kaogu Yanjiusuo and Tongjiang Xian Wenwu Guanlisuo 1998). The cultural scope of Leiguzhai is complex, and its cultural features resemble those of the Bianduishan site. At the same time, Leiguzhai carries some cultural characteristics of the eastern Sichuan Basin and could be viewed as a hub of a cultural corridor. The Leiguzhai site appears to represent the slow formation and spread throughout the northeastern Sichuan Basin of a cultural type built upon a foundation of cultural features extensively absorbed from neighbors. Due to its simultaneous embodiment of the cultural elements of both the eastern and western parts of the Sichuan

142 WANG YI

Basin, Leiguzhai must be considered a relatively complex cultural site. The cultural category of Leiguzhai awaits further research and analysis. Preliminarily, Leiguzhai could be regarded as the type site for a distinct archaeological culture of the Longshan period in the Sichuan Basin that differs from the archaeological cultures in the basin’s eastern and western regions. The Yueliangyan site is close to the Leiguzhai site both temporally and in cultural content, and they belong to the same culture. The physi-cal environment in which Yuliangyan is situated is also similar to that of Leiguzhai: both lie in a mountainous region and follow the cultural patterns of mountainous regions (Lei Yu and Chen De’an 1991). Yue-liangyan has some relations to the Baoduncun Culture, but the extent of their contact remains to be explored.

The eastern Sichuan Basin

Another archeological culture related to the Baoduncun Culture spread across the Yangzi Three Gorges in the eastern part of the Sichuan Basin, represented by the remains at Shaopengzui and Laoguanmiao 老關廟 (Wang Yi and Sun Hua 1999). Within this culture, pottery of Phase I at the Shaopengzui site in Zhongxian County appears to be more primitive and simplistic in style than pottery from the Baoduncun Culture. In this region, the early phase of the Zhongba 中壩 site in Zhongxian has the closest cultural affinity to the Baoduncun Culture, but its trumpet-mouth and long-neck hu dispay rather distinct differences in color and form from similar Baoduncun vessels. The late subphase of the early phase of Zhongba has some cultural elements similar to the Baoduncun Culture, while during the same period, there are still remarkable differences in pottery vessels between the Baoduncun Culture in the Chengdu Plain and the various sites in the eastern Sichuan Basin; this suggests that the Baoduncun Culture and the remains in the eastern Sichuan Basin should not be taken as parts of the same archaeological culture. Also, the region from Chongqing 重慶 following along the Yangzi River to Zhongxian does not belong to the basic Baoduncun cultural sphere. Additionally, differences are even more pronounced between pottery vessels unearthed in the lower stratum of Laoguanmiao in Fengjie 奉節 County and those of the Baoduncun Culture. The cultural remains of Weijialiangzi in Wushan appear to be even more divergent than the various Neolithic sites west of the Three Gorges. Only a portion of the site’s cultural elements is congruous with Phase I of Shaopengzui or with the lower stratum of Laoguanmiao. Clearly, the Weijialiangzi site cannot be included in the distribution area of the Baoduncun Culture.


Shizishan and the southwestern Sichuan Basin

The Shizishan site is located on the southern bank of the Daduhe 大渡河 River in the southwestern part of the Sichuan Basin. It was discovered in 1972. The Sichuan Team of the Institute of Archaeology, Chinese Academy of Social Sciences surveyed the site in 1988, and in 1990, archaeologists from Sichuan University undertook excavations at the site. The majority of the pottery was found to be coarse fabric vessels. Decorations include cord-markings, incised, and checker motifs, in addi-tion to a large number of appliqué designs. Painted pottery sherds with black designs painted on a red ground were also found. Vessel forms include tall-neck guan, large-mouth and scalloped-rim guan, pen, bo, and wan. The predominant lithics are axes, adzes, and chisels. Microliths were also found in great quantity (Ma Jixian 1992). Cultural features of the Shizishan site are to a great extent similar to the Baoduncun Culture, but the painted pottery shows an intimate cultural relationship with contemporaneous sites on the western Sichuan Plateau; the channel for this cultural transmission must have been along the eastern portion of the western Sichuan Plateau in the area adjoining the Sichuan Basin. The Shizishan site represents a local Neolithic culture, and it has been radiocarbon dated to 4500–4000 BP (Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shiyanshi 1992). The Shizishan site and the Bao-duncun Culture are two closely related archaeological entities that are both found distributed in the western Sichuan Basin and its adjacent southwestern mountain regions during roughly the same period.

The Significance of the Discovery of the Baoduncun Culture

The walled settlements of the Baoduncun Culture are the earliest walled settlements now known in southwest China. They are also the largest and most densely distributed prehistoric city cluster in this region, and their discovery represents a remarkable breakthrough in the archaeol-ogy of Sichuan. The discovery of the Baoduncun Culture has played a key role in establishing the cultural developmental sequence from the Neolithic to the Bronze Age in Sichuan and filled in a major lacuna in research on the Ba and Shu cultures. As a result, Ba-Shu studies have entered a major, new stage. The discovery of the Baoduncun Culture also demonstrates that the Chengdu Plain was a center in the origins of civilization in the upper Yangzi River valley and that this region was one of the places of origin for Chinese civilization as a whole. These finds have helped to support the theory of a singular Chinese civilization arising from many sources. They have also made great

144 WANG YI

contributions to research on the origin of civilization in the Yangzi River valley, improved our understanding of the origins of civilization within China proper, and contributed to the formation of a theoretical framework regarding Chinese civilization and the processes of early state formation. The prehistoric walled settlements on the Chengdu Plain are a symbol of 5000 years of ancient civilization in the Yangzi River valley, and they represent a portion of the cultural heritage of Chinese civi-lization as well. The various Baoduncun Culture sites have a singular cultural character, and those with typical content are marked by the presence of walls, such as we find in the central settlement sites of the Baoduncun Culture. The Baodun site is the largest prehistoric walled settlement discovered so far in southwest China, and it highlights the characteristics of the Baoduncun Culture and the nature of the social and historical processes in the upper Yangzi River region. The site oc-cupies a key position in studying the evolution of ancient societies, the origins of Chinese civilization, and cultural contact. The settlements of the Baoduncun Culture, centered on the Chengdu Plain, were the direct precursors to the Sanxingdui civilization. The many walled sites and abundant cultural items unearthed from the Baoduncun Culture play a critical role in achieving a comprehensive understanding of the development of archaeological cultures in the Sichuan Basin from the Neolithic to the Bronze Age and the roots and origins of the Sanxingdui civilization. Since the prehistoric Baoduncun Culture walled settlements on the Chengdu Plain date to the late Neolithic period, they are positioned at that crucial juncture when prehistoric societies were undergoing drastic transformations. The immense city walls and structures with various functions that we find are indicative of the emergence of relatively large-scale ruling groups and interest groups at the time. Such Baodun Culture finds as the large, 550 sq m structure (F5) surrounded by a number of smaller buildings that all had their doors set facing toward it at the center of the walled Gucheng site, are of great importance to research on social structure and religious beliefs on the Chengdu Plain during this time period.

References Cited

Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊 (2002). “Sichuan Chongzhoushi Shuanghe shiqian chengzhi shijue jianbao 四川崇州市雙河史前城址試掘簡報 (A brief report on the preliminary excavation of the Shuanghe prehistoric walled settlement in Chongzhou City, Sichuan).” Kaogu 考古 2002.11: 3–19.


Chengdu Shi Wenwu Kaogu Gongzuodui, Sichuan Lianhe Daxue Kaogu Jiaoyanshi 四川聯合大學考古教研室, and Xinjin Xian Wen-guansuo 新津縣文管所 (1997). “Sichuan Xinjin Xian Baodun yizhi diaocha yu shijue 四川新津縣寶墩遺址調查與試掘 (Investigation and preliminary excavation of the Baodun site in Xinjin County, Sichuan).” Kaogu 1997.1: 40–52.

Chengdu Shi Wenwu Kaogu Gongzuodui, Sichuan Lianhe Daxue Lishixi Kaogu Jiaoyanshi 四川聯合大學歷史系考古教研室, and Wenjiang Xian Wenguansuo 溫江縣文管所 (1998). “Sichuan Sheng Wenjiang Xian Yufucun yizhi diaocha yu shijue 四川省溫江縣魚鳧村遺址調查與試掘 (Investigation and preliminary excavation of the Yufucun site in Wenjiang County, Sichuan Province).” Wenwu 1998.12: 38–56.

Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所 (2001). “Wenjiang Xian Yufucun yizhi 1999 niandu fajue 溫江縣魚鳧村遺址 1999 年度發掘 (The 1999 excavation of the Yufucun site in Wenjiang County).” In Chengdu Shi Wenwu Kaogu Yanjiusuo (editor), Chengdu kaogu faxian 1999 成都考古發現1999 (Archeological discoveries in Chengdu 1999), pp. 40–53. Beijing: Kexue Chubanshe.

Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan 郫縣博物館 (2001a). “Sichuan Sheng Pixian Gucheng yizhi 1997 nian fajue jianbao 四川省郫縣古城遺址1997 年發掘簡報 (A brief report on the 1997 excavation of the Gucheng site in Pixian County, Sichuan Province).” Wenwu 2001.3: 52–68.

Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan (2001b). “Sichuan Sheng Pixian Gucheng yizhi 1998–1999 niandu fajue shou-huo 四川省郫縣古城遺址 1998–1999 年度發掘收獲 (Results of the 1998–1999 excavation of the Gucheng site in Pixian County, Sichuan Province).” In Chengdu Shi Wenwu Kaogu Yanjiusuo (editor), Cheng-du kaogu faxian 1999 成都考古發現 1999 (Archeological discoveries in Chengdu 1999), pp. 29–39. Beijing: Kexue Chubanshe.

Chengdu Shi Wenwu Kaogu Yanjiusuo, Sichuan Daxue Lishixi Kao-gu Jiaoyanshi, and Zaodaotian Daxue Changjiang Liuyu Wenhua Yanjiusuo 早稻田大學長江流域文化研究所 (2000). Baodun yizhi: Xinjin Baodun yizhi fajue he yanjiu 寶墩遺址—–新津寶墩遺址發掘和研究 (The Baodun site: Excavation and research of the Baodun site in Xinjin). Tokyo: ARP.

Gucheng 1999: Chengdu Shi Wenwu Kaogu Gongzuodui and Pixian Bowuguan (1999). “Sichuan Sheng Pixian Gucheng yizhi diaocha yu shijue 四川省郫縣古城遺址調查與試掘 (Investigation and pre-liminary excavation of the Gucheng site in Pixian County, Sichuan Province).” Wenwu 文物 1999.1: 32–42.

He Zhiguo 何志國 (1993). “Mianyang Bianduishan wenhua chutan 綿

146 WANG YI

陽邊堆山文化初探 (Preliminary investigation of the Bianduishan Culture at Mianyang).” Sichuan wenwu 四川文物 1993.6: 10–15.

Jiang Zhanghua 江章華, Wang Yi 王毅, and Zhang Qing 張擎 (2001). “Chengdu Pingyuan zaoqi chengzhi ji qi kaoguxue wenhua chulun 成都平原早期城址及其考古學文化初論 (Early walled settlement sites on the Chengdu Plain and a preliminary discussion of their archaeological culture).” In Su Bai 宿白 (ed.), Su Bingqi yu dangdai Zhongguo kaoguxue 蘇秉琦與當代中國考古學, pp. 699–721. Beijing: Kexue Chubanshe.

Lei Yu 雷雨 and Chen De’an 陳德安 (1991). “Bazhong Yueliangyan he Tongjiang Leiguzhai yizhi diaocha jianbao 巴中月亮岩和通江擂鼓寨遺址調查簡報 (A brief report of the investigations of the Yueliangyan site in Bazhong and the Leiguzhai site in Tongjiang).” Sichuan wenwu 1991.6: 52–55.

Ma Jixian 馬繼賢 (1992). “Hanyuan Xian Shizishan xinshiqi shidai yizhi 漢源縣獅子山新石器時代遺址 (The Neolithic site at Shizi-shan in Hanyuan County).” Zongguo Kaogu Xuehui 中國考古學會 (editor), Zhongguo kaoguxue nianjian 1991 中國考古學年鑒 1991 (Almanac of Chinese Archaeology 1991), pp. 270–271. Beijing: Wenwu Chubanshe.

Ma Jixian (1993). “Guanghan Yueliangwan yizhi fajue zhuiji 廣漢月亮灣遺址發掘追記 (Recollections from an excavation at the Yue-liangwan site in Guanghan).” In Nanfang minzu kaogu 南方民族考古 (Ethnology and Archaeology of the South) 5: 310–323. Chengdu: Sichuan Kexue Jishu Chubanshe.

Mangcheng 1999: Chengdu Shi Wenwu Kaogu Gongzuodui and Dujiangyan Shi Wenwuju 都江堰市文物局 (1999). “Sichuan Du-jiangyan Shi Mangcheng yizhi diaocha yu shijue 四川都江堰市芒城遺址調查與試掘 (Investigation and preliminary excavation of the Mangcheng site at Dujiangyan City in Sichuan).” Kaogu 1999.7: 14–27.

Reimer, Paula J., et al. (2004). “IntCal04 terrestrial radiocarbon age calibration, 0–26 cal kyr BP.” Radiocarbon 46(3): 1029–1058.

Sichuan Sheng Wenwu Guanli Weiyuanhui 四川省文物管理委員會, Sichuan Sheng Bowuguan 四川省博物館 , and Guanghan Xian Wenhuaguan 廣漢縣文化館 (1987). “Guanghan Sanxingdui yizhi 廣漢三星堆遺址 (The Sanxingdui site in Guanghan).” Kaogu xuebao 考古學報 1987.2: 227–254.

Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 (editor) (1999). Sanxingdui jisikeng 三星堆祭祀坑 (Sacrificial pits at Sanxingdui). Beijing: Wenwu Chubanshe.

Sichuan Sheng Wenwu Kaogu Yanjiusuo and Tongjiang Xian Wenwu Guanlisuo 通江縣文物管理所 (1998). “Tongjiang Xian Leiguzhai


yizhi shijue baogao 通江縣擂鼓寨遺址試掘報告 (Report of the preliminary excavation of the Leiguzhai site in Tongjiang County).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo (editor), Sichuan kaogu baogaoji 四川考古報告集 (Archaeological reports of Sichuan), pp. 41–58. Beijing: Wenwu Chubanshe.

Song Zhimin 宋治民 (2000). “Shilun Sichuan Wenjiang Yufucun yizhi, Xinjin Baodun yizhi he Pixian Gucheng yizhi 試論四川溫江魚鳧村遺址, 新津寶墩遺址和郫縣古城遺址 (Preliminary discussion of the Yufucun site of Wenjiang, the Baodun site of Xinjin, and the Gucheng site of Pixian in Sichuan).” Sichuan wenwu 2000.2: 9–18.

Sun Hua 孫華 (1993).“Shilun Guanghan Sanxingdui yizhi de fenqi 試論廣漢三星堆遺址的分期 (Preliminary discussion of the periodiza-tion of the remains of the Sanxingdui site at Guanghan).” Nanfang minzu kaogu 5: 10–24.

Wang Renxiang 王仁湘 and Ye Maolin 葉茂林 (1993). “Sichuan pendi beiyuan xinshiqi shidai kaogu xin shouhuo 四川盆地北緣新石器時代考古新收獲 (New archeological results with Neolithic Age sites in the northern borders of the Sichuan Basin).” In Li Shaoming 李紹明, Lin Xiang 林向, and Zhao Dianzeng 趙殿增 (eds.), Sanxingdui yu Ba Shu wenhua 三星堆與巴蜀文化 (Sanxingdui and the Ba-Shu Culture), pp. 257–265. Chengdu: Ba Shu Shushe.

Wang Yi 王毅, Jiang Zhanghua 江章華, Jiang Cheng 蔣成 and Lu Ding 盧丁 (1997). “Zhongguo Changjiang wenming qiyuan yanjiu de xin chengguo: Chengdu pingyuan shiqian chengzhi qun faxian ji 中國長江文明起源研究的新成果—–成都平原史前城址群發現記 (New results from research into the origins of cultures in China’s Yangzi River region: an account of the discovery of a group of prehistoric walled settlements in the Chengdu Plain).” Chengdu wenwu 成都文物 1997.2: 23–29.

Wang Yi and Sun Hua (1999). “Baoduncun wenhua de chubu renshi 寶墩村文化的初步認識 (A preliminary understanding of the Bao-duncun Culture). Kaogu 1999.8: 60–73.

Xi’nan Bowuyuan Choubeichu 西南博物院籌備處 (1954). “Baocheng tielu xiuzhu gongcheng zhong faxian de wenwu jianjie 寶成鐵路修築工程中發現的文物簡介 (Introduction to the cultural artifacts discovered in the construction of the Bao–Cheng Railway).” Wenwu cankao ziliao 文物參考資料 1954.3: 10–34.

Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shiyanshi 中國社會科學院考古研究所試驗室 (1992). “Fangshexing tansu ceding niandai baogao (shijiu) 放射性碳素測定年代報告 (十九) (Report no. 19 of radiocarbon dates).” Kaogu 1992.7: 655–662.

Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Sichuan Gongzuodui 中國社會科學院考古研究所四川工作隊 (1990). “Sichuan Mian-

148 WANG YI

yang Shi Bianduishan xinshiqi shidai yizhi diaocha jianbao 四川綿陽市邊堆山新石器時代遺址調查簡報 (A brief report on the investigation of the Neolithic remains at Bianduishan in Mianyang City, Sichuan).” Kaogu 1990.4: 307–313.

Zhong Ri Lianhe Kaogu Diaochadui 中日聯合考古調查隊 (1998). “Sichuan Xinjin Xian Baodun yizhi 1996 niandu fajue jianbao 四川新津縣寶墩遺址 1996 年度發掘簡報 (A brief report on the 1996 excavation of the Baodun site in Xinjin County, Sichuan).” Kaogu 1998.1: 29–50.

Zhong Ri Lianhe Kaogu Diaochadui (2001a). “Dujiangyan Shi Mang-cheng yizhi 1998 niandu fajue gongzuo jianbao 都江堰市芒城遺址 1998 年度發掘工作簡報 (A brief report on the 1998 excavation of the Mangcheng site in Dujiangyan City).” In Chengdu Shi Wenwu Kaogu Yanjiusuo (editor), Chengdu kaogu faxian 1999 成都考古發現 1999 (Archeological discoveries in Chengdu 1999), pp. 54–98. Beijing: Kexue Chubanshe.

Zhong Ri Lianhe Kaogu Diaochadui (2001b). “Dujiangyan Shi Mang-cheng yizhi 1999 niandu fajue gongzuo jianbao 都江堰市芒城遺址 1999 年度發掘工作簡報 (A brief report on the 1999 excavation of the Mangcheng site in Dujiangyan City).” In Chengdu Shi Wenwu Kaogu Yanjiusuo (editor), Chengdu kaogu faxian 1999 成都考古發現 1999 (Archeological discoveries in Chengdu 1999), pp. 99–126. Beijing: Kexue Chubanshe.

Address:

18 Shi’erqiao Road, Shudu BoulevardChengdu, Sichuan Province [email protected]


DEFINING THE ARCHAEOLOGICAL CULTURES AT THE SANXINGDUI SITE

BY

JAY XU 許杰(The Art Institute of Chicago)

Abstract The Sanxingdui site is a Neolithic and early Bronze Age site in the Chengdu Plain in Sichuan Province, southwest China. Over the years, as archaeological fieldwork at the site has progressed, the understanding of the site and its culture has changed gradually. This article discusses the stratigraphy, pottery typology, periodization, and absolute dating of Sanxingdui, aiming to evaluate major periodization schemes and different definitions of the Sanxingdui Culture, which the site typifies. It is argued that the San-xingdui Culture encompasses Phase II through Phase IV of the four-phase periodization scheme for the Sanxingdui site and that the earliest remains at the Shi’erqiao site in nearby Chengdu, contemporary with the Phase IV remains at Sanxingdui, should be included in the spatial and temporal dimensions of the Sanxingdui Culture.

For archaeologists in China as elsewhere, “archaeological culture” commonly means a distribution of material culture traits that can be observed consistently over a restricted area and within a given period.1 It is delimited in both space and time. Routinely, the archaeological culture is defined mainly by pottery types because of the ubiquitous presence of ceramics in human habitations from the Neolithic period on. Although artifacts of other materials like stone and features such as graves and building foundations often figure in the definition of a given culture, they are clearly of secondary importance in actual practice. An archaeological culture is routinely shared by multiple sites, but a site may have multiple cultural affiliations, as it may have remains be-longing to more than one culture, whether diachronically through time or synchronically across space. In practice, particularly in the early stages of fieldwork at a site, it is often difficult to see the cultural distinctions clearly. The site of Sanxingdui 三星堆 provides a case in point. In this paper, I will review the process by which the pottery-based archaeologi-cal cultures at Sanxingdui have been recognized and defined.

1 I would like to thank Professor Robert Bagley, Professor Lothar von Falkenhausen, and Dr. David Cohen for their valuable comments and editorial aid on this article.

150 jay xu

irrigation ditch

Yazi River

SanxingduiMuseum

mod

ern

road

EastWall

SouthWall

WestWall

YueliangwanTerrace

CangbaobaoTerrace

Rensheng Cemetery

Mamu River

Sanxingdui Pits1and 2 N

0 400 m

Detailed attention will be given to major schemes of periodization of the site and to the definition of the Sanxingdui Culture, with which the site is most prominently associated. A revised definition of the Sanxing-dui Culture’s temporal dimensions and of the spatial distribution of Sanxingdui-related remains will be proposed at the end of the paper.

Summary of fieldwork

The Sanxingdui site is located about 10 km west of the city of Guanghan 廣漢, which is about 40 km northeast of Chengdu 成都, the capital of Sichuan province (Sichuan 1999: 9, 15). Spread along the southern bank of the Yazi River 鴨子河 and both sides of the Mamu River 馬牧河, the site is presently known to cover an area of 10 to 17 sq. km as determined by the distribution of artifacts.2 The center of the site is a

2 The site is estimated to cover an area of about 10 sq. km in Sichuan 1992: 308, about 12 sq. km in Sichuan 1999: 9, about 15 sq. km in Chen De’an et al. 1998: 1, and about 17 sq. km in Zhao 1996: 232. The figures do not necessarily represent the size of the Sanxingdui site at any particular moment in its history since they are obtained,

Figure 1. Map of the Sanxingdui site. After Bagley 2001: 24, fig. 1.

DEFININg THE CULTURES AT SANXINDUI 151

walled enclosure that measures about 3.5 sq. km (350 ha) (Chen De’an et al. 1998: 1) (Figure 1). The site is named after Sanxingdui because of the major excavations in that area in 1980–1981, which established the basic framework of periodization and cultural definition of the site, and because of the discovery there in 1986 of K1 and K2, two “sacrifi-cial pits” filled with immense material wealth. These two pits revealed the site to be a center of an early Bronze Age civilization previously unknown.3 Field archaeology at the Sanxingdui site began in 1934 with the first excavation at Yueliangwan 月亮灣 led by David Graham, then cura-tor of the West China Union University Museum of Archaeology at Chengdu (Graham 1934).4 In the layout of the Sanxingdui city known today, Yueliangwan is situated near the north end of a central axis run-ning roughly north and south across the walled area, with Sanxingdui located towards its south end. Graham organized the excavation to seek stratigraphic and artifactual data following the chance discovery by local farmers of a pit containing stone and jade implements. Nearby the original pit, the excavation exposed a cultural stratum with hundreds of potsherds and numerous fragments of stone and jade. The similari-ties in artifact type made it clear that the cultural stratum and the pit belonged to the same time and culture (Graham 1934: 118–119).5 The excavation thus correlated the stone and jade types from the pit with pottery types. Perhaps most importantly, the excavation established the antiquity of human settlement in the area and an archaeological culture particular to it, called the “Hanchow [= Hanzhou 漢州] Culture” by Graham.6 Judging the pottery to predate the Han period by a large span of time, and finding similarities between the lithic finds and those from elsewhere known at the time to date from the Zhou period, Graham believed that the Hanchow culture should date no later than the begin-ning of the Zhou dynasty, or about 1100 bc, but cautioned that further evidence might make it necessary to date it to a much earlier period,

presumably, by estimating the distribution of all cultural remains that broadly fall into the entire time frame of the site. 3 For a general survey of the Sanxingdui site, see Xu 2001a. Chronological accounts of all the known finds and fieldwork at the Sanxingdui site can be found in Chen Xiandan 2001 and Xiao Xianjin et al. 2001. 4 According to the description in Graham 1934, it may be estimated that roughly 56 sq. m were excavated (not including several test pits). 5 Graham did not report explicitly on the relationship of the pit and the stratum, but evidently regarded them as belonging to the same level. 6 In the 1940s, Zheng Dekun 鄭德坤 referred to the cultural remains from Yue-liangwan by the names of the “Guanghan Culture” (Zheng Dekun 1946: 31–42) and later the “Taipingchang 太平場 Culture” (Cheng 1949). However, no clear definition was offered by Graham or Zheng for these appellations, probably due to the paucity of material.

152 jay xu

perhaps as early as the Neolithic (Graham 1934: 129). In the light of the present understanding of the Sanxingdui site, Graham’s deliberations are remarkably prescient. After the 1934 excavation, fieldwork at the Sanxingdui site came to a halt until the 1950s, when field surveys and test excavations were resumed and carried out intermittently at Yueliangwan. By 1961, several field surveys had led archaeologists to pronounce that the Yueliangwan area was the center of a site spreading over an area about 3 km long between the Yazi and Mamu Rivers, the cultural homogeneity among various loci being attested by similar potsherds. The cultural deposits on the side near the Mamu River were found to be as deep as 2 m. The archaeologists named the site the “Zhongxing 中興 site” after the local commune (Xi’nan 1954, Wang Jiayou and Jiang Dianchao 1958, and Sichuan 1961). In 1963, excavations were conducted in the Yueliangwan area under the directorship of Feng Hanji 馮漢驥. Covering a total of 150 sq. m, they afforded fresh understandings about the site (Ma Jixian 1992).7 The archaeologists identified two cultural strata, which evidently represented two developmental stages of the ancient culture. Artifacts excavated comprised mostly of about 30,000 potsherds. In addition, a small number of features such as graves, parts of building foundations, and possible traces of a bronze foundry were exposed, which began to reveal the complexity of material cultural traits in the area. It must be recognized that the early prominence of Yueliangwan in the exploration of the Sanxingdui site was due to both chance and choice; chance being that it was the locale where the first major discovery took place, and choice being that the later archaeologists, guided by the finds in the 1930s, continued to concentrate their efforts in that area. This choice was reasonable, and it was vindicated by the field results. At the same time, by preventing the archaeologists from employing their limited resources and manpower in other areas, it shaped the archaeological record in a particular way. We must be acutely aware of such factors when evaluating the archaeology of the Sanxingdui site or indeed any other. The Sanxingdui area came to notice in 1956, when a field survey revealed dense deposits of potsherds in that area, prompting the archae-ologists to call it the “Sanxingdui site” for the first time, as distinct from the cultural remains at Yueliangwan (Wang Jiayou and Jiang Dianchao 1958: 30–31). No fieldwork was performed here in the following years, however. In 1975, when a brickyard was set up at Sanxingdui and began

7 The excavations were first briefly described in Song Zhimin 1983:72–73, and 1991: 209–213.


to dig clay on the terraces in the area, a large number of potsherds were exposed, and local archaeologists determined that the terraces contained rich cultural deposits (Xiao Xianjin et al. 2001: 26). After the Cultural Revolution (1966–1976), the archaeologists started to work seriously in the Sanxingdui area due to the urgency to obtain archaeological information about the Sanxingdui terraces before they were completely dug away by the brickyard. In 1980–1981, excavations were carried out at a locale later labeled as Sanxingdui Locus III, led by Wang Youpeng 王有鵬, Chen De’an 陳德安, and Chen Xiandan 陳顯丹. The team first conducted a test excavation over an area of 125 sq. m (five 5 × 5 m test squares) in May 1980, and then from November 1980 through May 1981 excavated an area of 1,100 sq. m (forty four 5 × 5 m test squares). The total area excavated in 1980–81 exceeded 1,600 sq. m, and the excavations yielded a great number of potsherds and an array of features.8 Then, in 1982, Chen De’an and Chen Xiandan led the team in two excavations nearby, at what was to be labeled Sanxingdui Locus I, exposing an area of 100 sq. m (Sichuan 1987a) (Figure 2).9 The 1980–1981 excavations at Sanxingdui Locus III were the most instrumental in defining an archaeological culture at the Sanxingdui site. The archaeologists recognized a series of cultural strata and, based on more than 100,000 potsherds, a core group of pottery types associated with each of the strata. Based on stratigraphy and pottery typology and seriation, they established a periodization documenting three phases of cultural development. To that, the 1982 excavations added one more phase. Typological comparison and radiocarbon dating enabled the archaeologists to estimate the duration of the culture as extending from the late Neolithic period, or about 2500 bc, to about 1000 bc, i.e., the time corresponding to the end of the Shang period or the beginning of the Western Zhou period in the Central Plains. Furthermore, they formally proposed to name the archaeological culture the “Sanxingdui Culture” (Sichuan 1987a: 249). Chronological control and cultural definition of the site were thus finally established. Additionally, features such as building foundations, a pottery kiln, and graves made it possible to describe the content of the culture in a more substantial way than before. The pottery types and features also connected Sanxingdui with loci previously excavated or surveyed and with sites in other parts of the Chengdu Plain or even beyond the Sichuan Basin, thus enabling the

8 The size of 1,600 sq. m also includes more than ten test squares over an area of about 400 sq. m, which yielded no cultural remains. 9 Chen Xiandan (1989: 213, 215) mentions that the area excavated in 1980–81 was later labeled as Sanxingdui Locus III, and the area excavated in 1982 was later labeled as Sanxingdui Locus I; the area covered by the 1982 excavations is mentioned to be 150 sq. m, however.

154 jay xu

Figure 2. Excavation loci at Sanxingdui in 1986. Adapted from Sichuan 1999: 13, fig. 4.

archaeologists to begin mapping the size of the site, the spatial distribu-tion of the Sanxingdui Culture, and its external connections. From March to May 1984, the same team excavated at Xiquankan 西泉坎, about 600 m north of Sanxingdui, covering an area of 175 sq. m (seven 5 × 5 m test squares). The excavation yielded a large amount of potsherds whose types conform to the core group recognized in the

Sacrificial Pits exc.excavation sq.modern structure

major roadminor road

steep ridge

SanxingduiWorkstation

1986 I

1986 II

Sacrificial Pit 1

Sacrificial Pit 2

1986 III

0 60 m


1980–81 excavations. From October 1984 to January 1985, another excavation was carried out at Sanxingdui Locus III over an area of 125 sq. m (Chen Xiandan 1989: 215–216). From March through June 1986, the team launched the largest excavations at the Sanxingdui site since 1980–81, digging at Sanxingdui Locus I and Locus III, as well as a new locale later labeled as Sanxingdui Locus II (Figure 2). At Locus III, the excavation exposed the thickest cultural deposits (about 2.5 m) so far known at the site, with the most numerous strata. The total area excavated in this season was 1,325 sq. m, yielding the largest quantity of features and artifacts up to date, including many small pottery ani-mals (Chen Xiandan 1988: 10 and 1989: 216–217). The excavations confirmed and enriched the results of the earlier excavations in the San-xingdui area. The temporal span was once again determined to range from the Late Neolithic to the beginning of the Western Zhou period, with a duration of about 2,000 years. 1986 proved to be the most rewarding year yet for the Sanxingdui archaeologists. The excavations reached a dramatic climax with the discovery by local brickyard workers of the now world-famous “sacri-ficial pits” K1 and K2, which preserved in staggering abundance the material culture of an elite, and proved without doubt that the site was home to a civilization startlingly different from any other in Bronze Age China.10 As a result of the 1980s excavations, the known size of the site was substantially enlarged, now including the Yueliangwan area in the north and the Sanxingdui area in the south. Subsequently, in surveys from 1984 through the 1990s, a number of previously unnoticed mounds and terraces at Sanxingdui and other locales were found to be remains of man-made walls, and several excavations were conducted under the directorship of Chen De’an to map them out and to understand their plan, extent, method of construction, and date (Taipingyang 1999: 199, Xiao Xianjin et al. 2001: 31). Now measured to cover an area of 3.5 sq. km, the walled enclosure is comparable in size to the walled area of the largest known city of its time, Zhengzhou 鄭州, in the Central

10 The two pits and their contents have been the subject of a number of studies. See Bagley 1988, 1990, 1999: 212–219; Barnard 1990; Ge and Linduff 1990; Goepper 1995: 115–129, 248–278; Rawson 1996: 15–18, 60–84; Zhao 1996; Chen 1999; Liu and Capon 2000; So 2001; Xu 2001a, 2001b, 2001c, 2003; Falkenhausen 2002, 2003, and this volume; Louis 2003; and Thote 2003. The studies published in Chinese are too numerous to list here. For studies by scholars in Japan see Asahi 1998 and Nishie Kiyotaka 2002. Besides K1, K2, and the Yueliangwan pit excavated by Graham in 1934, several other pits were excavated between the 1960s and 1980s at the Sanxingdui site, including a relatively well-published pit excavated in 1987 at Cangbaobao 倉包包, about 400 m east of the 1934 Yueliangwan pit (Sichuan 1998). However, none of those pits was found to have contained pottery or have a clear stratigraphic correlation. They will therefore be omitted from discussion in this article.

156 jay xu

Plains (c. 1500–1300 bc) (Henan 2001: 178).11 Surrounding the walled enclosure at Sanxingdui, related cultural remains have been found over an area of 10 to 17 sq. km, marking the size of the whole site as presently known. Given the large size of the site, the total area covered by the excava-tions is very small: somewhat more than 5,000 sq. m had been excavated by 1994.12 Publication of these excavations and other field surveys is less than adequate. Only the 1980–81 excavations and the 1986 excavations of K1 and K2 have been published in formal reports (Sichuan 1987a, 1999).13 The 1934 excavation was published in a detailed preliminary report (Graham 1934).14 The report of the 1963 excavations had to wait thirty years for publication, the delay probably caused by the Cultural Revolution and consequent dispersal of excavation records. When published, the report appeared in the form of a recollection of personal notes taken by one of the archaeologists (Ma Jixian 1992), and is therefore far from exhaustive.15 Particularly lamentable is the absence of a report of the 1986 excavations at Sanxingdui Loci I, II, and III, the richest so far. Only a very brief account of these loci has been published, included in a summary of fieldwork between 1980 and 1986 (Chen Xiandan 1989). Besides these reports, there are a few brief reports of field surveys, and useful bits of information have occasionally been disclosed in short articles or news releases. None of the fieldwork undertaken since 1994 has been reported. This inadequacy, coupled with the limited scope of the excavations, makes tentative any periodization scheme and definition of the archaeological cultures of the site.

Stratigraphy, pottery typology, periodization, and absolute chronology

Archaeologists in China, as elsewhere, commonly rely on two basic tools to obtain relative dating of sites of the Neolithic and Bronze Age

11 The Zhengzhou excavation report (Henan 2001: 178) does not explicitly give the size of the walled city; instead, it gives the measurements of each wall of this roughly rectangular enclosure: the east wall measures about 1.7 km, the south wall about 1.7 km, the west wall about 1.87 km, and the north wall about 1.69 km. 12 Sichuan (1999: 9) mentions that a total of about 5,000 sq. m was excavated between 1980 and 1994; according to the statistics given in Taipingyang 1999: 23, the 5,000 sq. m seems to include 1,000 sq. m of wall sections. The figure should be significantly enlarged to account for excavations since 1994, but judging from numerous visits to the site, the total area excavated is still quite limited. 13 Sichuan 1999 was preceded by two preliminary excavation reports: Sichuan 1987b, and 1989. 14 Another published preliminary report describes the Cangbaobao pit excavated in 1987: Sichuan 1998. 15 Song Zhimin 1983 and 1991 also have very brief discussions of the material from the 1963 excavations.


periods: stratigraphy and pottery typology and seriation. Absolute dates are provided by radiocarbon dating calibrated by dendrochronology. For the Sanxingdui site in question, cross-dating of remains from San-xingdui with other sites in the region and across China has also helped build the chronological framework. The application of stratigraphy in Chinese archaeology is in some ways unique, as it is based on the conception of “cultural strata.” In the 1930s, Chinese archaeologists recognized that the layering of cultural deposits is routinely associated with different colors and textures of soil, and that artifacts contained in those different layers, most commonly potsherds, show distinctive typological traits. This recognition enabled Chinese archaeologists to construct “cultural strata” that are temporally meaningful. Since its first successful application more than seventy years ago, this method has occupied a primary position in relative dating in Chinese archaeology (Luan Fengshi et al. 2002: 21–39). At the Sanxingdui site, the number of strata (comprising “cultural strata” and strata free of cultural content) revealed at each locus can be summarized as follows (the top stratum of modern farming soil and cultural strata of the Han 漢 dynasty and later that exist in some loci are not included in the list):

(1) 1934 excavation at Yueliangwan: one cultural stratum (Graham 1934: 118–119).

(2) 1963 excavations at Yueliangwan: two cultural strata (Ma Jixian 1992: 312).

(3) 1980–81 excavations at Sanxingdui Locus III: seven strata, comprising five cultural strata and two strata largely free of cultural content interposed between cultural strata (Si-chuan 1987a: 228–232).

(4) 1982 excavations at Sanxingdui Locus I: two cultural strata (Sichuan 1987a: 249).16

(5) 1984 excavation at Xiquankan: three cultural strata (Chen Xiandan 1989: 217 table).17

(6) 1984 excavation at Sanxingdui Locus III: five strata (Chen Xiandan 1989: 217 table).

(7) 1986 excavation at Sanxingdui Locus I: five strata. 1986 excavation at Sanxingdui Locus II: five strata. 1986 excavation at Sanxingdui Locus III: 13 strata (Chen

Xiandan 1989: 217 table).

16 Chen Xiandan (1989: 215), however, reports four cultural strata. Those strata are indicated in Table 1. 17 Chen Xiandan (1989: 217 table) does not explicitly identify the three strata as cul-tural strata, but they must have been such because in the table each stratum represents one phase of his periodization of the Sanxingdui site (see Table 1).

158 jay xu

In the cases of (6) and (7), it is not clear whether the strata are all cultural strata or include one or more strata free of cultural content. When Graham excavated at Yueliangwan in 1934, he recognized only one stratum, but the 1963 excavations at the same locale identified two cultural strata distinguished by differently colored soil. When the report of this season of excavations was published in 1992, its author had the benefit of a four-phase periodization scheme constructed with the results of the 1980–81 and 1982 excavations at the Sanxingdui loci. The two strata at Yueliangwan were accordingly identified as belong-ing to two of these phases, with the lower one corresponding to Phase I and the upper one to Phases III and IV (Ma Jixian 1992: 322–323, Li Mingbin 1999a). Of the five cultural strata discerned in the 1980–81 excavations, the 8th stratum (= the fifth cultural stratum) and the 6th stratum (= the fourth cultural stratum)—the two deepest layers—contained pottery vessels of similar traits although they were separated by a layer of soil free of any cultural content (the 7th stratum). Between the 6th (fourth cultural) and the 4th (third cultural) strata, which were also separated by a layer largely free of artifacts (the 5th stratum), a considerable difference in pottery types was discerned. Pottery traits were again similar between the 4th (third cultural) and the 3rd (second cultural) strata. The archaeologists therefore collapsed those into three phases:

Phase I: the earliest phase, the 8th to the 6th strata (the fifth and fourth cultural strata)

Phase II: the 4th and 3rd strata (the third and second cultural strata)

Phase III: the 2nd stratum (the first cultural stratum)

Of the two cultural strata identified in the 1982 excavations, the lower one was equivalent to the 2nd (first cultural) stratum in the 1980–81 excavations, whereas the upper one showed different traits (Sichuan 1987a: 249). This upper stratum became recognized as a separate Phase IV in 1987 (Sichuan 1987b: 1, Zhao Dianzeng 1987: 19). A four-phase periodization scheme was thus established for the Sanxingdui site:

Phase I: the 8th and 6th strata of 1980–81 Sanxingdui Locus III

Phase II: the 4th and 3rd strata of 1980–81 Sanxingdui Locus III

Phase III: the 2nd stratum of 1980–81 Sanxingdui Locus III, and the lower stratum of 1982 Sanxingdui Locus I

Phase IV: the upper stratum of 1982 Sanxingdui Locus I


This scheme will be referred to as the Report Scheme in the follow-ing discussion. Based on the pottery information so far published and discussed (Sichuan 1987a, Chen Xiandan 1989, Ma Jixian 1992, Sun Hua 1992, Li Boqian 1997, Sichuan 1999: 424–427, Wang Yi and Zhang Qing 1999, Jiang Zhanghua et al. 2002: 7–11), the pottery types typical of each of the Report Scheme’s four phases may be summarized as follows. Phase I is characterized by only a few identifiable types, including ring-foot dou 圈足豆 (bowl or dish), and pots with a sharply everted mouth and rim (Figure 3a). Vessel fabric is mostly grayish and fine in texture, but a smaller number are brownish and coarse in texture. The vessels often bear cord-marked or incised patterns. The dou continued in Phase II with changes in traits, but the phase is distinguished by new vessel forms rather than continuity, with the advent of several new types that would persist through Phases III and IV, including xiaopingdi guan 小平底罐 (jar with tapered profile and small, flat bottom), tripodal he 三足盉 (pitcher), and gaobing dou 高柄豆 (high-stemmed bowl or dish). Also new are ladles with handles in the shape of a bird with a hooked beak (Figure 3b). The brownish coarse ware became the majority. The dominant pattern—cord-marking—became coarser. Phase III shows close affinity with Phase II in fabric texture, vessel types, and decorative patterns, though with gradual changes in specific morphological traits and the addition of a few new types such as gu 觚 (beaker), zun-shaped vessel 尊形器 with broad shoulder and small, flat bottom, and tripodal pot with a broad collar along the neck (Figure 3c). More vessels are plain in surface in this phase. Most vessels of Phase IV are grayish-brown and coarse in texture, but the fraction of fine gray ware increased significantly. Continuity shows in several types such as xiaopingdi guan, tripodal he, and gaobing dou, but morphological changes in those types are somewhat more distinct between Phases IV and III than between Phases III and II, and more vessels are now plain in surface. The major new types of Phase IV were cups and bowls with pointed or nearly pointed bottom (Figure 3d). The excavations from 1980 through 1982 thus, for the first time, established a comprehensive understanding of the developmental stages at the Sanxingdui site. Because of its seminal importance, archaeologists through the years have offered criticism and revisions. A direct critique was first published by Huang Jiaxiang (1990), who pointed out various inconsistencies in the factual information provided by the report of the 1980–81 excavations, as well as possible mistakes in interpretation. The critique, remarkable for its rigor, does not challenge the basic stratig-raphy and periodization of the report, but injects a welcome caution about many details.

Figure 3b. Pottery types from Phase II at the Sanxingdui site. 8–9. xiao-pingdi guan; 10. ring-foot dou; 11. flat-bottom pan 盤 (basin); 12. ring-foot pan; 13. tripodal he; 14. gaobing dou; 15. ladle with a handle in the shape of a bird with a hooked beak. Adapted from Sichuan 1999: fig. 226.

Figure 3a. Pottery types from Phase I at the Sanxingdui site. 1. vessel with broad flat rim; 2–4. vessel with everted rim; 5. vessel with saw- toothed rim. 6–7. ring-foot dou. Adapted from Sichuan 1999: fig. 225.

1

2

3

4

5

6

7

8

9

10

11

12

13 14

15

Figure 3c. Pottery types from Phase III at the Sanxingdui site. 16–17. xiaopingdi guan; 18. high-shoulder weng 瓮 (urn); 19. zun-shaped vessel with broad shoulder and small flat bottom; 20. gu; 21. tripodal he; 22, 24. gaobing dou; 23. ring-foot dou; 25. tripodal pot with a broad collar along the neck; 26. ladle with a handle in the shape of a bird with a

hooked beak. Adapted from Sichuan 1999: fig. 227.

Figure 3d. Pottery types from Phase IV at the San-xingdui site. 27. small xiaopingdi guan; 28–30. cup with nearly pointed bottom; 31. tripodal he; 32–33. gaobing dou. Adapted from Sichuan 1999: fig. 228.

16

17

18

19

20

21

22 24

23 25

26

27

28

29 33

32

30 31

162 jay xu

a Adapted from Chen Xiandan 1989: 217 table. While this table usefully correlates the strata excavated from 1980 to 1986 at Sanxingdui, one must exercise caution in using the correlation because it represents Chen Xiandan’s own problematic periodization scheme as discussed in the present paper.

In a later study, Chen Xiandan (1989: 217–223), a co-author of the report on the 1980–81 excavations, incorporated further results from excavations through 1986 and proposed a four-phase scheme that dif-fers from the Report Scheme in two respects (see Table 1): first, the 4th stratum (= the third cultural stratum) at Sanxingdui Locus III excavated in 1980–81 is now moved from Phase II to his first phase; second, none of the cultural strata at Locus I excavated in 1982 is considered as con-temporary with Phase III, instead all constituting his fourth phase. The first difference is critical, for the 4th stratum at Locus III is known for the advent of new vessel types such as xiaopingdi guan, and this particular type continued through Phase IV. By moving the 4th stratum to his first phase, Chen Xiandan in effect created cultural continuity through the four phases that would otherwise not exist. As argued by Sun Hua (1992: 11–13), however, this move cannot be justified, as pottery fabric and type demonstrate a clear distinction between the Report Scheme’s Phases I and II. On the other hand, Chen Xiandan’s reassignment also

Table 1: Stratigraphy at the Sanxingdui Site from 1980–1986a

Loci and StrataXiquan-

kan

Sanxingdui Periodization 1984 1980–81 1982 1984 1986

III I III I II III

Han to modern times

(1) (1)

(1) (2)

(1) (1) (2) (3)

(1) (2) (3)

(1) (2) (3)

4th Phase

(3) (4) (5) (6)

(2)

(3) (4) (5)

(4) (5) (6) (7) (8)

(4) (5) (6) (7) (8)

(4) (5) (6) (7) (8)

3rd Phase (2) (2) (6) (9) (10)

2nd Phase (3) (3) (11) (12)

1st Phase

(4) (4) (5) (6) (7) (8)

(13) (14) (15) (16)


severs the close affinity between the 4th and the 3rd strata at Locus III, which together constitute the Report Scheme’s Phase II. The second difference, regarding the periodization of the strata uncovered in the 1982 excavations at Locus I, directly contradicts the information provided in Sichuan (1987a: 249). As the material from the excavations is inadequately published, it is difficult to evaluate Chen Xiandan’s judgment, but those pottery types that have been published, by Chen Xiandan himself, seem to demonstrate clear differences (in the cups and bowls with pointed or nearly pointed bottom) as well as close similarities (in the xiaopingdi guan, and tripodal he) (Figure 4).18 In

Figure 4. Pottery types identified in the 1982 excavations at Sanxingdui Locus I. Adapted from Chen Xiandan 1989: 215, fig. 2. 1. xiaopingdi guan, 82GSG2(3):18; 2. xiaopingdi guan, 82GST1(3):4; 3. tall-neck xiaopingdi guan, T2(3):6; 4. tall-neck xiaopingdi guan, T3(3):1; 5. cup with pointed bot-tom, T2(3):1; 6. cup with pointed bottom, T2(3):2; 7. ring-foot pan with rounded bottom, G2(4):3; 8. tripodal he, T2(3):9; 9. dou with perforated high ring foot, G2(3):16; 10. fragment of bowl from gaobing dou, G2(4):3; 11. fragment of bowl from gaobing dou, G2(3):1; 12. vessel lid, G2(3):4; 13. fragment of bowl from gaobing dou, G2(4):13; 14. bottle, G1(3):2; 15. jiandi

zhan, T2(3):13; 16. bowl, T1(3):11.

18 Sun Hua (1992: 12) chooses to emphasize the difference marked by the pointed-bottom vessels, criticizing Chen Xiandan for failing to recognize this distinction.

1 2

3 4

5 6

7

10 8

119

14

15

1613

12

164 jay xu

all probability, the 1982 strata were partly contemporaneous with the Report Scheme’s Phase III. Probably due to these contradictions, Chen Xiandan’s revised periodization scheme has not been widely accepted by other scholars. Sun Hua (1992) has produced the most influential revision of the periodization so far. He endorses the basic validity of the Report Scheme, but proposes a three-phase scheme with more finely divided subphases:

Sun’s First Phase = Report Scheme Phase I

Sun’s Second Phase = Report Scheme Phases II and III

Sun’s Third Phase = K1, and Report Scheme Phase IV19

Sun Hua further divided the three phases into six subphases illustrat-ing the gradual changes in pottery traits:

First Phase: Subphase 1

Second Phase: Subphases 2–4

Third Phase: Subphases 5–6

Sun Hua’s study is solidly anchored in pottery typology, and his subphases more closely mirror the changes of the pottery vessels at the Sanxingdui site than does the Report Scheme, while still being basically compatible with it (Figure 5). His study has since exerted much influence on other efforts at periodization.20 Sun Hua later modified his periodization by moving his Subphase 5, to which he thinks both K1 and K2 belong, to the Second Phase (Sun Hua 2000: 97, 144). His updated periodization, hereafter referred to as the Sun Scheme, now stands as:

First Phase: Subphase 1 = Report Scheme Phase I

Second Phase: Subphases 2–5 = Report Scheme Phases II, III, and K1 and K221

Third Phase: Subphase 6 = Report Scheme Phase IV (minus K2)

19 In 1986, when K1 and K2 were excavated, K1 was found to lie beneath a stratum containing potsherds typical of the Report Scheme’s late Phase III, and K2 beneath a deposit that can be dated by pottery typology to early Phase IV (Sichuan 1999: 427). 20 Recently published periodization schemes such as Li Boqian 1997, Wang Yi and Zhang Qing 1999, and Jiang Zhanghua et al. 2002: 7–11 are all in general agreement with Sun’s periodization. 21 Sun Hua (2000: 67) alternatively suggests that K1 and K2 may be placed between the Second and the Third Phase. Judging from his discussions in other papers in Sun Hua 2000, it is clear that he is not as strongly committed to this alternative as to the Sun Scheme.


Figure 5a. Pottery types in the first periodization scheme of Sun Hua, First Phase (Subphase 1). 1. dou with high ring foot, Yueliangwan (3); 2. vessel with broad flat rim, Xiquankan 84BbT1(4); 3. vessel with saw-toothed rim, 80III; 4. ring-foot dou, 86IIIH19. After Sun Hua 1992: fig. 9. (Note: all the vessels were excavated at Sanxingdui loci

unless otherwise noted.)

Figure 5b. Pottery types in the first periodization scheme of Sun Hua, Second Phase (Subphase 2). 5. tripodal he, 80III(3); 6. tripodal pot with a broad collar along the neck, 86III(12); 7. xiaopingdi guan, 86III(11); 8. fragment of bowl from gaobing dou, 86III(12); 9. fragment of bowl from gaobing dou; 10. ring-foot pan, 80III(3); 11. ladle with a handle in the shape of a bird with a hooked beak, 80III(3); 12. ring-foot pan, 80III(3); 13. ring-foot dou, 80III(3); 14. guan with everted rim, 86III(11); 15. wide-mouth vase, 80III(3); 16. gaobing dou, 80III(3); 17. flat-bottom pan, 80III(3); 18. bottle, 86III(12). After Sun Hua

1992: figs. 10–11. (Note: all the vessels were excavated at Sanxingdui loci.)

1

2 3

4

5

6 7

8

9

10 1112

13

14

15 16

17 18

166 jay xu

Figure 5c. Pottery types in the first periodization scheme of Sun Hua, Second Phase (Subphase 3). 19. narrow-mouth vase, 86III(9); 20. wide-mouth vase, 86III(9); 21. bottle, 86III(9); 22. gu, 86III(9); 23. zun-shaped vessel with broad shoulder and small flat bottom, 86III(9); 24. gaobing dou, 86III(9); 25. ring-foot pan, 80III(2); 26. vessel lid, 86III(9); 27. pot with a broad collar along the neck, 86III(9); 28. xiaopingdi guan, 86III(9); 29. gaobing dou, 86III(9); 30. tripodal he, 80III(2); 31. ladle with a handle in the shape of a bird with a hooked beak, 80III(2). 32. ring-foot pan, 80III(2); 33. gaobing dou, 80III(2). After Sun Hua

1992: figs. 10–11. (Note: all the vessels were excavated at Sanxingdui loci.)

19

20 21 22

23

2425 26

27 2829

33

323130

For the absolute dating of the Sanxingdui site, the report of the 1980–81 excavations (including remarks about the 1982 excavations) relied on the following evidence (Sichuan 1987a: 249). A calibrated radiocarbon date was obtained from a sample of charcoal from Phase I remains, which gives the uncalibrated radiocarbon date of 4075±100 BP (with 5730 half-life) and the calibrated date of 4500±100 BP.22 The fine spiral pattern decorating some of the pottery vessels of Phase III are

22 Prior to the publication of an updated collection of radiocarbon dates in Zhong-guo 1991b, radiocarbon dates were calibrated in China using the calibration curve constructed by Paul E. Damon and his collaborators and published in 1972 (see Zhong-guo 1982: 439 for reference and explanation). In this method, the calibrated dates are expressed in the same manner as the uncalibrated dates. Sichuan 1987a does not specify the lab no. of the sample that gives the dates, but it is easy to see in Table 2 that only the sample of lab no. ZK-0973 has the radiocarbon date of 4075±100. Recalibrated dates using Calib rev.5.0.1 and IntCal04.14c are listed in Table 2.


similar to the leiwen 雷紋 (rectilinear spirals) on Shang bronzes. Besides, the upper stratum of the 1982 excavations yielded pottery vessels with pointed bottom, the likes of which had appeared in abundance at the Shuiguanyin 水觀音 site in Xindu 新都 County, a site that can be dated to the end of the Shang period or the beginning of the Western Zhou period through associations in bronze.23

23 For dating of the Shuiguanyin site, see Sichuan 1959: 410 and Jiang Zhanghua 1998a: 6.

Figure 5d. Pottery types in the first periodization scheme of Sun Hua, Second Phase (Subphase 4), top; Third Phase (Subphase 5), middle; Third Phase (Subphase 6), bottom. Subphase 4: 34. pot with a broad collar along the neck, 86III(8); 35. fragment of bowl from gaobing dou, 86III(8); 36. xiaopingdi guan; 37. tripodal he, 86III(8); 38. ladle with a handle in the shape of a bird with a hooked beak, 86III(8); 39. fragment of stem from gaobing dou, 86III(8); 40. zun-shaped vessel with broad shoulder and small flat bottom, 86III(8); 41. wide-mouth vase, 86III(8); 42. bottle, 86III(8); 43. gu, 86III(8); 44. vessel lid, 86III(8). Subphase 5: 45. jiandi zhan, K1; 46. pedestal, K1. Subphase 6: 47, 48. cup with pointed bottom, 82IT2(3); 49. tall-neck xiaopingdi guan, 82IT2(3); 50–51. jiandi zhan, 82IT2(3); 52. tripodal he, 82IT2(3). After Sun Hua 1992: figs. 10–12. (Note: all

the vessels were excavated at Sanxingdui loci.)

37

3435

36

38 39

4041 42 43

44

4546

47 48 49

50

51

52

168 jay xu

By now, a series of 28 radiocarbon dates is available, suggesting a range of dates for the Sanxingdui site (Table 2). Excluding two outliers at each end, the range of absolute dates appears to be from the early third millennium to about 1000 cal bc. Based on the radiocarbon dates known to him as well as artifact associations, Chen De’an, the lead archaeologist at the Sanxingdui site, has estimated that Phase I lasted from 2800 to 2000 bc. His estimation of the other three phases roughly corresponds to the first half of the second millennium bc for Phase II, and the second half of the second millennium bc for Phases III and IV, with ca. 1200 bc as a convenient, though very approximate, demarcation between Phases III and IV (Sichuan 1999: 424).24 To a large degree, this dividing line between Phases III and IV depends on stylistic evidence supplied by the bronze vessels from K1 and K2; this had been addressed by Bagley (1988, 1990), Chen De’an (Sichuan 1999: 428–431), and other scholars.25

Defining the Sanxingdui Culture

The Report Scheme treats all the phases of the Sanxingdui site as belong-ing to the same culture, namely the Sanxingdui Culture (Sichuan 1987a: 250). In contrast, Sun Hua’s three-phase scheme originally published in 1992 treats each phase as a different culture, which he names as the “Bianduishan 邊堆山 Culture,”26 the “Sanxingdui Culture,” and the “Shi’erqiao 十二橋 Culture” (Sun Hua 1992: 23).27 To estimate the merits of these different cultural definitions, it is necessary to briefly review the differences in pottery type between the phases. In the Report Scheme, a major typological difference exists

24 For somewhat different absolute dates proposed by other scholars, see Sun Hua 1992: 19–21, Li Boqian 1997:273–274, Wang Yi and Zhang Qing 1999: 17–20, Sichuan 1999: 424–427, and Jiang Zhanghua et al. 2002: 10–11. 25 For general discussion of the dating of bronzes of this time, see Bagley 1987: 19–32, and 1999: 146–155. 26 Wang Renxiang and Ye Maolin (1993: 264) first observed similarities in artifact types between Sanxingdui Phase I and the Bianduishan site in Mianyang 綿陽 City and other similar Neolithic sites in northern Sichuan. The Bianduishan site is now identified as having close cultural affinity with the recently identified Baodun 寶墩 Culture. For Bianduishan and related sites in northern Sichuan, see Zhongguo 1990, Wang Renxiang and Ye Maolin 1993, and Ma Xinxin 1993. For their relationship with the Baodun Culture, see Wang Yi and Sun Hua 1999: 63. For a general survey of the Baodun Culture, including radiocarbon dates, see the article by Wang Yi in this volume. 27 For Shi’erqiao and related sites in Chengdu and the Shi’erqiao Culture, see Si-chuan 1987c, Sun Hua 1996, Jiang Zhanghua 1998b, Wang Yi et al. 1999: 4–5, and Jiang Zhanghua et al. 2002: 11–16. Like Sun Hua’s cultural identification, Wang Yi and Zhang Qing 1999, and Jiang Zhanghua et al. 2002 similarly identify the Report Scheme’s Phase IV as belonging to the Shi’erqiao Culture.


between Phases I and II: the latter is distinguished from the former by a whole new set of vessel types, and continuities between the two are limited. In Phase III, the vessel types of Phase II continue with little change, though new types also appear. In Phase IV, on the one hand, several vessel types persist from Phase II through this phase, including but not limited to xiaopingdi guan, tripodal he, and gaobing dou, though their morphological changes are more noticeable. On the other hand, vessels with pointed or nearly pointed bottom become a major presence. The pointed-bottom types have proved most contentious. They did not appear in the 1980–81 excavations at Locus III, but showed up in the 1982 excavations at Locus I (Figure 4: 5, 6, 15), and later in other excavations, including K1 (Figure 6). Clearly these types developed late in the occupation of the Sanxingdui site, but in which phase? In the Report Scheme it is claimed that one of the pointed-bottom types, jiandi zhan 尖底盞 (bowl with pointed bottom), first emerged toward the end of Phase III (Sichuan 1999: 424), as vessels of this type were included in K1.28 In Sun Hua’s first periodization, the jiandi zhan did not emerge until the Third Phase, but in his later Sun Scheme, he dates its appearance to Subphase 5, which belongs to his Second Phase (Sun Hua 1992: 19, 2000: 144). The difference may seem a small technicality, but when different cultural identities are proposed for the phases, as in Sun Hua’s schemes, the difference takes on a weighty significance, as will be discussed below. Since the qualitative differences in pottery types between Phases I and II appear indeed large, it seems warranted to recognize the two phases as representing two different archaeological cultures in the pottery-based definition customary in Chinese archaeology.29 Lending critical support to this distinction is the rise, at Sanxingdui, of the city with its magnifi-cent walled enclosure during Phase II, which signifies a basic change in the nature of the settlement.30 In this light, continuities between the

28 In addition to K1, Chen Xiandan (1999: 167) points out that jiandi zhan “first ap-peared during the third phase of the Sanxingdui site, in the ninth stratum of section B. . . .” According to Table 1, the only excavations that included such a stratum were at Sanxingdui Locus III in 1986. The “third phase” in Chen Xiandan’s periodization is the same as Phase III of the Report Scheme. 29 Recent fieldwork at the Yufu 魚鳧 site in Wenjiang 溫江 County in the Chengdu Plain seems to supply a missing link between Phases I and II (see Chengdu 1998, Li Mingbin 2001a). 30 As described in Chen De’an 1998: 62, the pounded earth that composes the walls contains only potsherds of Phase I, and in all the three surviving walls on the east, west and south, the wall fill lies atop a stratum belonging to Phase I while being broken into by strata of Phase II or later. The moat outside the walls contains deposits mainly of Phases III and IV, helping to suggest that the walls were used through the end of the habitation. In the Chengdu Plain, although walled settlements had appeared earlier, in the late Neolithic, at Baodun Culture settlements, the Sanxingdui city is much larger in size, undoubtedly the product of a very different society.

170 jay xu

two phases must be considered as secondary to the changes, though they indicate that the two cultures were related. The pottery typology recently revealed at Baodun and similar sites has confirmed Sun Hua’s distinction, and Phase I is now commonly accepted as belonging to the Baodun Culture.31 Since Phases II and III are particularly similar, it is also not with-out good reason that Sun Hua first combined them, minus K1 of late Phase III, into one phase, his Second Phase. The reason for excluding K1 from the Second Phase was obviously the presence of jiandi zhan, which heralded the popularity of pointed-bottom types in Phase IV. As expressed in his critique of Chen Xiandan’s periodization scheme, Sun Hua attached great importance to such vessels, and he subsequently grouped K1 and Phase IV into one phase, his Third Phase. However, does the difference between the Second and Third Phase warrant de-

Figure 6. Jiandi zhan from K1. After Sichuan 1999: 146, fig. 76.

31 Wang Yi and Sun Hua (1999: 65) first proposed to change the name of Bianduishan Culture to the Baoduncun 寶墩村 Culture, an alternative name to the Baodun Culture. For a latest summary of Phase I of Sanxingdui in the context of the Baodun Culture, see Li Mingbin 2001b.

0 6 cm 0 6 cm

0 3 cm0 3 cm

0 3 cm 0 3 cm

0 3 cm


32 Sun’s periodization and cultural scheme are part of his effort to establish an overall sequence of archaeological cultures in the Sichuan Basin. For a comprehensive presen-tation of his research, see Sun Hua 2000. 33 Chen Xiandan, however, maintains the position that there was only one culture at the Sanxingdui site throughout its duration. This position is obviously based on his particular scheme of periodization that maintains continuity in pottery type throughout the four phases, as discussed above. He proposes that the Baodun Culture should be re-identified as the “Baodun phase of the Sanxingdui Culture” (see Chen Xiandan and Liu Jiasheng 2002). Like his periodization, Chen Xiandan’s position in archaeological cultural identification is hard to justify. 34 A note of explanation with regard to the practice of naming archaeological cultures is in order here. In field archaeology in China as elsewhere, archaeological cultures have usually been named after the first site at which they are discovered. Some archaeologists have recently adopted a different approach by naming cultures after the sites recognized as the most representative or the richest in cultural remains or occupying a central position in the cultural distribution. This approach seems to be customary in Sichuan archaeology. The naming of Sanxingdui Culture is a case in point; otherwise, the name should be the “Hanchow Culture” as Graham called it

fining a new archaeological culture? Or, would it be more appropriate, given the considerable continuity seen in other vessel types, to consider it as reflecting a gap between two developmental stages within the same archaeological culture, with perhaps some evidence missing in between? Phrased in another way, how much significance should be given to jiandi zhan in defining an archaeological culture? For Sun Hua, the jiandi zhan in K1 signaled the advent of a new ar-chaeological culture because the vessel and related types would not only become a distinctive characteristic of the Third Phase at Sanxingdui, but would also appear widely distributed at a cluster of sites in Chengdu, with Shi’erqiao as their type site. Moreover, the Third Phase represents for him a stage of decline at the Sanxingdui site (Sun Hua 1992: 17–18, 23). This is why he assigns the Third Phase to the Shi’erqiao Culture, separating it from the Sanxingdui Culture of his Second Phase. Sun Hua’s three-culture theory, like his three-phase scheme, is a pioneering contribution that has remained deeply influential.32 Many archaeologists working on ancient Sichuan have come to accept his three-culture identification, including Chen De’an, who had been largely responsible for the Report Scheme (Zhao Dianzeng and Chen De’an 2001: 461–465).33 However, while the identification of Sun Hua’s First Phase as belonging to the Baodun Culture seems justified, his distinc-tion between the Sanxingdui Culture and the Shi’erqiao Culture is problematic in several regards. First of all, the identification of the Third Phase (as originally pro-posed by Sun Hua in his first periodization scheme) as belonging to the Shi’erqiao Culture causes the confusion and incongruity of label-ing the major archaeological finds of the period, i.e., “sacrificial pits” K1 and K2, with a name taken from another far more modest site.34

172 jay xu

Sun Hua’s cultural identification was based on the evidence of pottery alone, without regard to material culture of the elite. Two dozen jiandi zhan in K1 outweighed nearly one metric ton of bronze plus many other valuables from the two pits.35 Judging from the material wealth from K1 and K2, as well as the construction of city walls, it is clear that labeling the archaeological culture of Sun’s Third Phase at the Sanxingdui site after Shi’erqiao is hardly appropriate.36 Sun Hua himself was probably troubled by the problem that K1 and K2 impose. In his updated study, namely the Sun Scheme published in 2000, he modified his periodization by moving K1 and K2 to his Second Phase, allowing the two pits to be identified as belonging to the Sanxingdui Culture (Sun Hua 2000: 142–144). Yet the modification seems to create a new problem for his distinction between the Sanxing-dui Culture and the Shi’erqiao Culture. Now that the jiandi zhan form originates in the Sanxingdui Culture, this vessel type loses its previous status as a signifier of a new archaeological culture, and without this, there is little remaining to justify the separation of a Shi’erqiao Culture from the Sanxingdui Culture. There is yet another problem with identifying the Third Phase as belonging to the Shi’erqiao Culture, whether or not pits K1 and K2 are included in that phase. Actually, Sun Hua’s idea can be challenged from several angles. First, there seems to be no conclusive evidence that the Sanxingdui site was in decline during the Third Phase, as Sun Hua proposes. In Table 1 we can see that, at Sanxingdui, cultural deposits

or the “Guanghan Culture” as Zheng Dekun called it. Another example is the label of the Baodun Culture, which replaced the Bianduishan Culture named after the site where the archaeological culture in question was first observed in 1952 (Xi’nan 1954: 11). The naming of the Shi’erqiao Culture by Sun Hua (1992: 23) represents the same approach. The Shi’erqiao site was discovered in 1985, excavated from 1985 to 1987, and published in a preliminary excavation report at the end of 1987 (Sichuan 1987c), when the four-phase Report Scheme at Sanxingdui had already been established. The Shi’erqiao Culture was obviously not named after the first site, but proposed by Sun because the site was deemed as the most representative. Methodological issues related to either approach of naming an archaeological culture lie beyond the scope of the present article. Instead, my concern here is twofold: whether the Report Scheme’s Phase IV is so different from Phases III and II as to warrant its definition apart as a separate culture; and the importance and representativeness of the Shi’erqiao site in relation to the Sanxingdui site. 35 The estimate of one metric ton is given in Guangming ribao 光明日報 (Guangming Daily), December 10, 1986. 36 The Shi’erqiao site is best known for its architectural remains contemporary with Phase IV at Sanxingdui or earlier (Sichuan 1987c). Wooden beams and planks, bamboo slats from walls, and straw from roofs were preserved well enough to shed much light on building techniques. Some structures may have belonged to the elite class (Sichuan 1987c: 9). No architectural remains are known to have survived in a similar state of preservation at the Sanxingdui site, but this evidence alone does not challenge Sanxing-dui’s position as a more important site.


more or less corresponding to the Report Scheme’s Phase IV or the Sun Scheme’s Third Phase have been excavated at Sanxingdui Locus I in 1982 and 1986, at Locus II in 1986, and at Locus III in 1984 and 1986—clearly at more loci than remains of Phases I and II.37 Moreover, as mentioned earlier, one of the two strata excavated in the 1963 excava-tions spanned both Phases III and IV as reported in Ma Jixian 1992.38 As there has been no quantitative analysis undertaken or published for any period, and since fieldwork has only touched a very small portion of the site, it seems at least premature to state that Phase IV (or the Third Phase) was a time of decline. Indeed, the Sanxingdui city was abandoned at the end of this phase, but that need not have been the result of a drawn-out process of decline. Second, there seems to be no conclusive evidence that vessels with pointed bottom were more extensively distributed at Chengdu than at Sanxingdui, an assertion that is a primary basis for Sun Hua’s cultural identification of the Third Phase. Of the strata at the Shi’erqiao site, the 13th (the lowest) through the 10th represent the earliest phase variously called the Shi’erqiao Culture or an early stage of it (Sichuan 1987c; Sun Hua 1996:123–129,141–142; Jiang Zhanghua 1998b: 146–156; Wang Yi et al. 1999: 4–5; Jiang Zhanghua et al. 2002: 12).39 The 13th stratum contained vessels with pointed bottom, xiaopingdi guan, tripodal he, gaobing dou, and ladles with handles in the shape of a bird with a hooked beak. Most of these types continued in the 12th stratum, while the 11th and 10th strata seem to contain only pointed-bottom types (Sichuan 1987c: 4,14). This artifact list instantly makes clear that some of these strata must be contemporary with the Sanxingdui site. In Sun Hua’s first periodization scheme, the 13th through the 11th strata cor-respond to the Third Phase at Sanxingdui (Sun Hua 1996: 125 fig. 2, and 129), but archaeologists of the Chengdu Municipal Institute of

37 The 8th stratum in the 1986 excavations at Sanxingdui, which Chen Xiandan dates to his fourth phase, is classified in Sun Hua’s periodization as belonging to the Second Phase (Sun Hua 2000: 8), but the removal of that stratum does not affect the statement made here. Moreover, Chen Xiandan (1989: 217 table) states that the fourth phase may be further divided into two subphases. Sun Hua (2000: 8), however, states that the Third Phase remains at the Sanxingdui site have so far only been excavated at a few small areas, including the Xiquankan locus. Jiang Zhanghua and Li Mingbin (2002: 110) similarly state that the cultural remains of the period corresponding to Sun Hua’s Third Phase are thin, and therefore the city was clearly in decline, or perhaps abandoned already. It is unclear what information these statements are based on. 38 A recent excavation led by Chen Zujun 陳祖軍 at Yueliangwan in late 2000 and early 2001 yielded a large quantity of pottery vessels, including some with pointed or nearly pointed bottom typical of Phase IV. I had the opportunity of visiting the excava-tion site on February 14, 2001. The excavation report is yet to be published. 39 Sun Hua (1996: 142) states, in the concluding remarks of the paper, that the 10th stratum should be excluded from the temporal span of the Shi’erqiao Culture, for it shows distinctively different cultural traits.

174 jay xu

Archaeology treat the 13th and 12th strata as one phase and the 11th and 10th strata as another in light of pronounced differences in pot-tery types between them (Jiang Zhanghua 1998b: 154, Wang Yi et al. 1999: 4, Jiang Zhanghua et al. 2002: 12). The demarcation drawn by the Chengdu archaeologists appears to be more appropriate on pres-ent evidence. Among the cluster of sites in Chengdu, few can be dated to this early phase of the Shi’erqiao site, however, except for remains beneath the terraced mound at Yangzishan 羊子山 (Jiang Zhanghua et al. 2002: 12) and a stratum at the Fuqin Xiaoqu 撫琴小區 site.40 Otherwise, no other sites were specified as containing remains of this phase. Moreover, the Shi’erqiao site—the largest settlement in Chengdu known at the time of Sun Hua’s writing—covers a maximum area of 3 hectares (Jiang Zhanghua 1998b: 146),41 much smaller than even the Neolithic Baodun Culture sites. It is hard to perceive that jiandi zhan and other pointed-bottom types had a more extensive distribution at Chengdu than at Sanxingdui. Other sites in Chengdu indeed yielded similar vessels, but they date later than the Third Phase. The continued popularity of vessel types with pointed bottom in those sites should have no bearing on the present subject. On the other hand, there seems to be no scarcity of pointed-bottom types at Sanxingdui. In the report of the 1980–81 excavations at Locus III, it is stated that the upper stratum of the 1982 excavations at Locus I, which corresponds to Phase IV (or Sun Hua’s Third Phase), yielded a large number of such vessels; the number becomes more impressive when we realize that the total space excavated was only 100 sq. m (Sichuan 1987a: 249).42 It would be interesting to compare the distribu-tion ratios per square meter of these vessels at comparable loci in the two sites, but no quantitative analysis of this kind has been done, and appropriate data has not been published to carry out such analyses. In addition to these challenges to Sun Hua’s identification, it is important to point out again that the pottery assemblage in the 13th and 12th strata at Shi’erqiao exhibits a remarkable continuity with

40 For the Yangzishan terraced mound, see Sichuan 1957, Lin Xiang 1988, Sun Hua 1993, and Lin Mingbin 2003. For pottery excavated at Fuqin Xiaoqu, see Wang Yi 1991: 298–302. Sun Hua (1996: 132) dates the stratum at Fuqin Xiaoqu to an interim time between the 11th and 10th strata, thus following the Third Phase. The earliest remains at another site, Fangchijie 方池街, are dated by the Chengdu archaeologists generally to the period encompassing the 13th through the 10th strata (Wang Yi et al. 1999: 5). 41 In Sun Hua 1996: 123, the size is given as 1.5 hectares only. 42 Chen Xiandan (1989: 215), however, states that the 1982 excavation covered 150 sq. m. Pointed-bottom vessels were also found in the 1963 test excavation at Yueliang-wan (Ma Jixian 1992: 322), but they are not published. The 2000–01 excavation at Yueliangwan yielded pointed-bottom vessels as well (see note 38).


Sanxingdui in such vessel types as xiaopingdi guan, tripodal he, gaobing dou, and bird-head ladle. These types form the core of the pottery as-semblage at the Sanxingdui site during Phases II and III (or Sun Hua’s Second Phase—the phase of the Sanxingdui Culture). As mentioned before, the same vessels were present at Sanxingdui in Phase IV (or Sun Hua’s Third Phase). In the context of this continuity, the pointed-bottom types should perhaps not yet be taken as a decisive indicator of a new archaeological culture. Lastly, as Sun Hua himself pointed out, the Shi’erqiao Culture represents a transitional phase (Sun Hua 2000: 104–105), which clearly suggests that this culture lacks a distinctive character identifiable as unique to it. For all these reasons, I propose to drop the appellation of the Shi’er-qiao Culture from the definition of the Sanxingdui site, and propose to define the temporal dimension of the Sanxingdui Culture as starting from the Report Scheme’s Phase II and ending with Shi’erqiao’s 12th stratum. The Sanxingdui Culture so delimited is based on a core pot-tery assemblage that included xiaopingdi guan, tripodal he, gaobing dou, and bird-head ladle, but also takes into consideration elite material culture such as represented in K1 and K2, and the city walls. Moreover, this re-definition expresses a hierarchical relationship in the settlement pattern of the Chengdu Plain, with the Sanxingdui site as the primary center and the Shi’erqiao site as a satellite of Sanxingdui.43 Since the 1950s, but mostly in the past decade, a number of modest-scale early Bronze Age sites have been excavated or surveyed in the Chengdu Plain, and they apparently co-existed partly with Sanxingdui. Such sites include Guilinxiang 桂林鄉 (Chengdu 1997) and Shuiguanyin (Sichuan 1959) in Xindu County, Qingjiangcun 清江村 in Pixian 郫縣 County (Chengdu 2001 and 2003a), Yangzishan (Sichuan 1957) and Hetaocun 核桃村 (Chengdu 2003b) in Chengdu City, and as far as Shaxi 沙溪 at the southwest edge of the Sichuan Basin (Sichuan 1990 and Li Mingbin 1999b).44 The pottery types from those sites indicate that they belong to the Sanxingdui Culture. The scales of the sites and their close proximity (save the Shaxi site) to the Sanxingdui site suggest the possibility that they were closely related with Sanxingdui as satellite settlements. In early 2001, another site was discovered at Jinsha 金沙 in the

43 Li Boqian 1997 also includes Shi’erqiao and related sites in his definition of the Sanxingdui Culture, but the temporal dimension of his definition is broadly extended to include remains in Chengdu as late as the early Spring and Autumn period. His scheme thus does not take into consideration the crucial change in the pottery inven-tory between the 12th and the 11th strata at the Shi’erqiao site as well as the marked changes in bronzes during the long span of time. 44 For general surveys of early Bronze Age sites in the Chengdu Plain, see Sun Hua 1996, 2000: 102–105, 146–148, and Jiang Zhanghua et al. 2002: 11–16.

176 jay xu

western suburb of Chengdu. Though smaller and less rich than San-xingdui, the Jinsha site was far larger and richer than those mentioned above. Hundreds of objects of gold, bronze, jade, stone and other ma-terials were unearthed at the time of the initial discovery. Subsequent fieldwork has yielded a large number of potsherds, and revealed that the site occupied an area of four square kilometers (400 ha), consisting of building foundations of modest to large sizes, cemeteries, ash pits, pottery kilns, and areas that may have been lithic and bone workshops or sacrificial grounds. The time span of the site is estimated from the late Shang period to the early Western Zhou period (Chengdu 2002, Zhu Zhangyi et al. in this volume). Evidently, the Jinsha site was closely related with the Sanxingdui site, for they share pottery types as well as lithic and bronze types. For the period that is contemporaneous with the Sanxingdui site, the Jinsha site should perhaps be regarded as a satellite of Sanxingdui also. Jinsha and the other sites mentioned above thus suggest the spatial distribution of the Sanxingdui Culture in the Chengdu Plain.

Epilogue

Over the past twenty five years, field archaeology in the Chengdu Plain has yielded some of the most important discoveries in China, the Sanxingdui site being a most spectacular example. Furthermore, at least six walled settlements of the Neolithic Baodun Culture have been identified and excavated since 1995 (Wang Yi in this volume), and for the early Bronze Age, the Shi’erqiao and the Jinsha sites are among the most significant discoveries. These discoveries completely re-write the ancient history of Sichuan, which had traditionally been viewed as a cultural backwater. This group of related sites provides rich resources for analyzing and describing ancient material cultures in the Chengdu Plain. They also suggest a great potential for studying the settlement pattern and the development of social and political structure on a regional scale. The Sanxingdui site, so far known to cover the longest time span and the largest size, occupies a central position in these studies. The present paper has attempted to clarify some issues in defining the archaeo-logical cultures at the Sanxingdui site. Results from future fieldwork will undoubtedly continue to revise our understanding of the site, its position in the network of settlements on the Chengdu Plain, and the archaeological culture—the Sanxingdui Culture—that it typifies.

DEFININg THE CULTURES AT SANXINDUI 177T

able

2: R

adio

carb

on D

ates

from

the

Sanx

ingd

ui S

ite

Sam

ple M

ater

ial a

nd P

rove

nien

ceD

ate P

ublis

hed

in C

hina

Cal

ibra

ted D

ate (

BC

) L

ab N

o.Ra

dio

carb

on

Dat

e (B

P)Ca

libr

ated

T½

= 5

730

T½

= 5

568

Dat

e (B

C)1σ

(68.

2%)

2σ (9

5.4%

)

ZK

-259

1C

harc

oal:

woo

d,

91G

SHT

G2(

4B):1

4900

±13

047

60±

130

3700

–337

036

55 (6

4.56

%) 3

485

3474

(35.

44%

) 337

1

3895

(0.4

9%) 3

881

3800

(93.

39%

) 326

5 32

41 (6

.12%

) 310

4

ZK

-233

0C

harc

oal:

woo

d,

T71

4EV

III

4675

±13

545

40±

135

3497

–304

034

97 (9

.55%

) 345

433

77 (8

1.14

%) 3

081

3069

(9.3

1%) 3

025

3630

(3.4

7%) 3

580

3533

(96.

53%

) 290

9

BK

9208

5C

lay-

char

coal

mix

ture

, 91

GSD

T71

19(1

4)45

75±

180

4446

±18

033

61 (1

00%

) 290

8

3632

(2.9

8%) 3

558

3538

(89.

41%

) 283

328

18 (7

.13%

) 266

126

49 (0

.48%

) 263

5

BK

9208

6C

lay-

char

coal

mix

ture

, 91

GSD

T71

19(1

5)44

15±

8542

90±

85

3082

(4.1

0%) 3

067

3027

(77.

43%

) 286

228

07 (1

5.15

%) 2

757

2718

(3.3

2%) 2

706

3319

(1.8

5%) 3

273

3266

(1.7

2%) 3

236

3170

(0.2

5%) 3

163

3113

(96.

18%

) 262

1

BK

9208

4W

ood,

91G

SDF7

011(

16)

4260

±11

041

40±

110

2876

(94.

04%

) 261

826

08 (2

.98%

) 259

925

94 (2

.98%

) 258

4

3011

(1.0

4%) 2

977

2973

(0.6

1%) 2

948

2944

(98.

17%

) 245

724

17 (0

.18%

) 240

9

BK

8604

6C

harc

oal:

woo

d,86

GSI

IIT

1416

(14)

:182

4210

±80

4090

±80

2873

–250

2

2860

(20.

94%

) 280

827

55 (1

3.20

%) 2

720

2704

(57.

63%

) 256

725

21 (8

.23%

) 249

8

2876

(100

%) 2

476

ZK

-232

8C

harc

oal:

woo

d,

T70

9CIV

4180

±10

540

60±

105

2871

–247

028

58 (1

4.87

%) 2

810

2750

(8.0

5%) 2

723

2700

(77.

08%

) 247

4

2890

(99.

53%

) 233

923

21 (0

.15%

) 231

923

15 (0

.32%

) 230

9

178 jay xu

Tab

le 2

: Rad

ioca

rbon

Dat

es fr

om th

e Sa

nxin

gdui

Site

(con

t.)

Sam

ple M

ater

ial a

nd P

rove

nien

ceD

ate P

ublis

hed

in C

hina

Cal

ibra

ted D

ate (

BC

) L

ab N

o.Ra

dio

carb

on

Dat

e (B

P)Ca

libr

ated

T½

= 5

730

T½

= 5

568

Dat

e (B

C)1σ

(68.

2%)

2σ (9

5.4%

)

ZK

-210

4C

harc

oal:

woo

d-ba

mbo

o

mix

ture

, 86

GSI

IIT

1416

(14)

:119

4170

±85

4050

±85

2864

–247

5

2849

(11.

72%

) 281

327

41 (2

.86%

) 273

026

93 (1

.72%

) 268

826

79 (8

3.70

%) 2

472

2879

(96.

57%

) 243

324

23 (1

.35%

) 240

223

80 (2

.08%

) 234

8

ZK

-259

2C

harc

oal:

woo

d,

91G

SHT

G1H

140

80±

115

3970

±11

528

52–2

340

2832

(2.1

3%) 2

820

2657

(0.3

6%) 2

655

2632

(97.

51%

) 228

9

2872

(6.5

1%) 2

800

2793

(0.3

4%) 2

785

2780

(91.

89%

) 219

621

69 (1

.26%

) 214

7

ZK

-259

4C

harc

oal:

woo

d,

91G

SHT

G1(

5D)

4070

±12

039

60±

120

2850

–231

0

2829

(0.9

1%) 2

823

2626

(95.

79%

) 228

322

48 (2

.94%

) 223

322

17 (0

.36%

) 221

5

2871

(5.9

6%) 2

801

2791

(0.0

9%) 2

790

2780

(93.

95%

) 214

0

ZK

-259

3C

harc

oal:

woo

d,

91G

SHT

G1(

5A):1

4100

±60

3990

±60

2854

–246

3

2619

(3.4

7%) 2

607

2599

(2.0

3%) 2

593

2586

(90.

52%

) 245

724

18 (2

.61%

) 240

823

73 (1

.37%

) 236

9

2836

(2.1

0%) 2

815

2671

(97.

90%

) 229

7

ZK

-097

3C

harc

oal:

woo

d,

80G

ZA

aT1(

3)H

4075

±10

039

60±

100

2590

–234

026

18 (1

.78%

) 260

925

97 (0

.64%

) 259

525

82 (9

7.58

%) 2

292

2862

(4.2

3%) 2

807

2758

(2.2

9%) 2

718

2706

(92.

31%

) 219

621

71 (1

.17%

) 214

6

ZK

-232

9C

harc

oal:

woo

d,

T71

1EIV

-VII

3935

±24

038

20±

240

2590

–194

025

78 (1

00%

) 192

328

97 (9

9.83

%) 1

663

1650

(0.1

7%) 1

642


able

2: R

adio

carb

on D

ates

from

the

Sanx

ingd

ui S

ite (c

ont.)

Sam

ple M

ater

ial a

nd P

rove

nien

ceD

ate P

ublis

hed

in C

hina

Cal

ibra

ted D

ate (

BC

) L

ab N

o.Ra

dio

carb

on

Dat

e (B

P)Ca

libr

ated

T½

= 5

730

T½

= 5

568

Dat

e (B

C)1σ

(68.

2%)

2σ (9

5.4%

)

BK

8604

5C

harc

oal:

woo

d,

86G

SIII

T15

16(9

):105

3990

±80

3880

±80

2471

–220

924

69 (8

8.80

%) 2

279

2250

(7.9

4%) 2

230

2220

(3.2

6%) 2

211

2573

(100

%) 2

135

BK

8205

8C

harc

oal:

woo

d, T

1(1)

3765

±80

3660

±80

2183

–193

521

87 (1

.32%

) 218

421

41 (9

8.68

%) 1

929

2286

(2.9

8%) 2

246

2244

(93.

86%

) 187

318

44 (1

.81%

) 181

518

00 (1

.35%

) 177

8

BK

8604

7C

harc

oal:

bam

boo-

woo

d

mix

ture

, 86

GSI

IIT

1415

(8B

):69

3700

±10

036

00±

100

2133

–178

7

2132

(14.

18%

) 208

420

56 (7

2.04

%) 1

873

1844

(7.9

2%) 1

815

1800

(5.8

6%) 1

778

2275

(0.9

0%) 2

255

2227

(0.0

4%) 2

225

2208

(99.

06%

)168

9

ZK

-210

2C

harc

oal:

woo

d,

86G

SIII

T15

16(8

B):6

536

15±

8035

10±

8019

48–1

743

1942

(100

%) 1

740

2109

(0.2

0%) 2

105

2035

(99.

80%

) 162

7

ZK

-232

6C

harc

oal:

woo

d,

89G

SHLT

705(

11)

3525

±17

034

30±

170

1960

–152

019

40 (1

00%

) 152

522

04 (9

9.61

%) 1

377

1337

(0.3

9%) 1

321

ZK

-210

1C

harc

oal:

woo

d-ba

mbo

o

mix

ture

, 86

SIII

T14

15(9

)36

00±

7535

00±

7519

32–1

741

1923

(98.

16%

) 173

917

04 (1

.84%

) 169

920

24 (1

00%

) 163

6

ZK

-210

5C

harc

oal:

woo

d,

86G

SIII

T14

14(9

)H36

(3):1

2336

00±

7535

00±

7519

32–1

741

1923

(98.

16%

) 173

917

04 (1

.84%

) 169

920

24 (1

00%

) 163

6

ZK

-210

3C

harc

oal:

woo

d-ba

mbo

o

mix

ture

, 86

SIII

T14

15(1

1)35

55±

8034

50±

8018

86–1

681

1884

(100

%) 1

681

1955

(95.

68%

) 160

215

91 (4

.32%

) 153

2

180 jay xu

Tab

le 2

: Rad

ioca

rbon

Dat

es fr

om th

e Sa

nxin

gdui

Site

(con

t.)

Sam

ple M

ater

ial a

nd P

rove

nien

ceD

ate P

ublis

hed

in C

hina

Cal

ibra

ted D

ate (

BC

) L

ab N

o.Ra

dio

carb

on

Dat

e (B

P)Ca

libr

ated

T½

= 5

730

T½

= 5

568

Dat

e (B

C)1σ

(68.

2%)

2σ (9

5.4%

)

ZK

-136

5C

harc

oal:

woo

d, T

(3)

3485

±10

533

90±

105

1876

–152

9

1871

(8.1

3%) 1

845

1812

(2.9

4%) 1

803

1776

(70.

80%

) 160

215

90 (1

8.13

%) 1

533

1941

(100

%) 1

450

ZK

-269

4C

lay-

char

coal

mix

ture

,91

GSD

T71

24(6

C)

3382

±11

732

86±

117

1735

–144

017

26 (0

.55%

) 172

416

91 (9

9.45

%) 1

435

1882

(97.

64%

) 136

813

59 (2

.36%

) 131

5

ZK

-232

7C

harc

oal:

woo

d,

89G

SHLT

705(

10)

3165

±28

530

80±

285

1680

–929

1656

(0.1

1%) 1

656

1635

(97.

16%

) 971

960

(2.7

3%) 9

35

2114

(0.2

1%) 2

100

2037

(97.

95%

) 742

689

(0.4

2%) 6

6364

7 (1

.42%

) 549

ZK

-269

3C

lay-

char

coal

mix

ture

, 91

GSD

T70

20(6

A)

3284

±87

3191

±87

1528

–140

816

07 (1

1.73

%) 1

572

1559

(3.1

5%) 1

549

1539

(85.

12%

) 138

616

81 (1

00%

) 126

5

ZK

-249

6C

harc

oal:

woo

d, T

114(

9)32

15±

115

3120

±11

515

20–1

260

1519

(95.

24%

) 125

612

36 (4

.76%

) 121

516

60 (0

.24%

) 165

416

38 (9

9.76

%) 1

049

ZK

-113

8C

harc

oal:

woo

d, T

1(1)

2875

±70

2790

±70

1023

–847

1012

(79.

01%

) 887

884

(20.

99%

) 842

1127

(100

%) 8

07

ZK

-269

5C

lay-

char

coal

mix

ture

, 91

GSD

T71

24H

924

11±

134

2343

±13

475

9–25

1

747

(13.

88%

) 688

665

(4.6

1%) 6

4458

9 (1

.79%

) 580

556

(57.

64%

) 350

305

(22.

08%

) 209

794

(98.

51%

) 149

140

(1.4

9%) 1

12


able

2: R

adio

carb

on D

ates

from

the

Sanx

ingd

ui S

ite (n

otes

)

a. R

ecal

ibra

tion

by C

alib

rev

. 5.0

.1 u

sing

Int

Cal

04.1

4 (S

tuiv

er e

t al.

2005

).

b. T

he s

eque

nce

of d

ates

from

the

earl

iest

to th

e la

test

pre

sent

ed in

this

tabl

e is

bas

ed o

n th

e da

tes

calib

rate

d w

ith o

ne s

igm

a er

ror

term

by

Cal

ib r

ev. 5

.0.1

usi

ng I

ntC

al04

.14c

(Stu

iver

et a

l. 20

05).

For

basi

c pr

inci

ples

and

pro

blem

s of

the

radi

ocar

bon

met

hod,

se

e B

owm

an 1

990.

c. A

bbre

viat

ions

of t

he n

ames

of t

he r

adio

carb

on la

bora

tori

es:

BK

: Rad

ioca

rbon

Lab

orat

ory

of th

e D

epar

tmen

t of A

rcha

eolo

gy, B

eijin

g U

nive

rsity

. Z

K: R

adio

carb

on L

abor

ator

y of

the

Inst

itute

of A

rcha

eolo

gy, C

hine

se A

cade

my

of S

ocia

l Sci

ence

s.

d. T

he d

ates

pub

lishe

d in

Chi

na a

re fr

om Z

hong

guo

1991

b: 2

24-2

26 e

xcep

t for

the

sam

ples

list

ed b

elow

. The

cal

ibra

ted

date

s rep

orte

d in

Chi

na re

pres

ent t

he la

rges

t pos

sible

rang

es o

f dat

es a

rriv

ed b

y ca

libra

tion

with

one

sigm

a er

ror t

erm

by

the

com

pute

r sof

twar

e us

ed

in th

e R

adio

carb

on L

abor

ator

y of

the

Inst

itute

of A

rcha

eolo

gy, C

hine

se A

cade

my

of S

ocia

l Sci

ence

s.

ZK

-259

1, Z

K-2

592,

ZK

-259

4, Z

K-2

593:

pub

lishe

d in

Zho

nggu

o 19

92: 6

60.

B

K92

085,

BK

9208

6, B

K92

084:

pub

lishe

d on

ly w

ith u

ncal

ibra

ted

date

s (T

½=

5730

) in

Bei

jing

1996

: 93.

Z

K-2

694,

ZK

-269

3, Z

K-2

695:

pub

lishe

d in

Zho

nggu

o 19

93: 6

48-6

49.

e. T

he f

ull I

Ds

of t

he s

ampl

es o

ften

have

to

be r

e-co

nstr

ucte

d ba

sed

on t

he i

nfor

mat

ion

publ

ishe

d in

the

ir o

rigi

nal r

adio

carb

on

repo

rts

and

Zho

nggu

o 19

91b.

In

a fe

w c

ases

, as

liste

d be

low,

it is

impo

ssib

le to

asc

erta

in th

e fu

ll ID

s:

Z

K-2

330:

Cha

rcoa

l: w

ood,

T71

4EV

III (

Zho

nggu

o 19

91b)

. In

the

orig

inal

repo

rt, Z

hong

guo

1991

a: 6

61, t

he sa

mpl

e is

iden

tified

as

89G

SHL

EV

III.

ZK

-232

8: C

harc

oal:

woo

d, T

709C

IV (Z

hong

guo

1991

b). I

n th

e or

igin

al re

port

, Zho

nggu

o 19

91a:

661

, the

sam

ple

is id

entifi

ed a

s 89

GSH

LC

IV.

Z

K-2

329:

Cha

rcoa

l: w

ood,

T71

1EIV

-VII

(Zho

nggu

o 19

91b)

. In

the

orig

inal

repo

rt, Z

hong

guo

1991

a: 6

61, t

he sa

mpl

e is

iden

tified

as

89G

SHL

EIV

-VII

.

ZK

-249

6: C

harc

oal:

woo

d, T

114(

9) (Z

hong

guo

1991

b). I

n th

e or

igin

al r

epor

t, Z

hong

guo

1991

a: 6

61, t

he s

ampl

e is

iden

tified

as

88T

119(

9).

182 jay xu

References cited

Asahi 1998: Asahi Shimbunsha 朝日新聞社 (editor) (1998). Sanseitai— Chūgoku 5000 nen no nazo, kyōi no kamen ōkoku 三星堆—中国 5000 年の謎・驚異の仮面王国 (Sanxingdui—China’s 5000-year old mystery, kingdom of startling masks). Tōkyō: Asahi Shimbunsha.

Bagley, Robert (1987). Shang Ritual Bronzes in the Arthur M. Sackler Collec-tions. Washington, DC: Arthur M. Sackler Foundation.

Bagley, Robert (1988).“Sacrificial pits of the Shang period at Sanxingdui in Guanghan County, Sichuan Province.” Arts Asiatiques 43 (1988): 78–86.

Bagley, Robert (1990). “A Shang city in Sichuan Province.” Orientations (November 1990): 52–67. [Chinese translation published in Li Shao-ming et al. 1993: 69–75].

Bagley, Robert (1999). “Shang archaeology.” In Michael Loewe and Edward L. Shaughnessy (eds.), The Cambridge History of Ancient China, pp. 124–231. Cambridge: Cambridge University Press.


Barnard, Noel (1990). “Some preliminary thoughts on the significance of the Kuang-han pit-burial bronzes and other artifacts.” Beiträge zur allgemeinen und vergleichenden Archäologie 9–10 (1990): 249–279. [Chinese translation published in Nanfang minzu kaogu 南方民族考古 5 (1992): 25–66.]

Beijing 1996: Beijing Daxue Kaoguxi Tanshisi Shiyanshi 北京大學考古系碳十四實驗室 (1996). “Tanshisi niandai ceding baogao (yiling) 碳十四年代測定報告 (一○) (Report of radiocarbon dates, No. 10)”. Wenwu 文物 1996.6: 91–95.

Bowman, Sheridan (1990). Radiocarbon Dating. London: British Museum Publications.

Chen De’an 陳德安 (1998). “Sanxingdui yizhi de faxian yu yanjiu 三星堆遺址的發現與研究 (Discovery and study of the Sanxingdui site).” Zhonghua wenhua luntan 中華文化論壇 1998.2: 57–63. Chengdu.

Chen De’an, Wei Xuefeng 魏學峰, and Li Weigang 李偉綱 (1998). San-xingdui: Chang jiang shangyou wenming zhongxin tansuo 三星堆: 長江上游文明中心探索 (Sanxingdui: Exploring the center of civilization in the upper reaches of the Changjiang River). Chengdu: Sichuan Ren-min Chubanshe.

Chen Xiandan 陳顯丹 (1988). “Lun Guanghan Sanxingdui yizhi de xingzhi 論廣漢三星堆遺址的性質 (On the nature of the Sanxingdui site of Guanghan).” Sichuan wenwu 四川文物 1988.4: 9–11, 8.

Chen Xiandan (1989). “Guanghan Sanxingdui yizhi fajue gaikuang,


chubu fenqi—jianlun ‘zao Shu wenhua’ de tezheng jiqi fazhan 廣漢三星堆遺址發掘概況, 初步分期—兼論 ‘早蜀文化’ 的特徵及其發展 (Brief account of excavations and preliminary periodization of the Sanxingdui site in Guanghan—concurrently on the characteristics of the ‘early Shu Culture’ and its development).” Nanfang minzu kaogu 2 (1989): 213–229.

Chen, Xiandan (1999). “The sacrificial pits at Sanxingdui: their nature and date.” In Roderick Whitfield and Wang Tao (translators and eds.), Exploring China’s Past: New Discoveries and Studies in Archaeology and Art, pp. 165–171. London: Saffron Books.

Chen Xiandan (2001). “Guanghan Sanxingdui dashiji 廣漢三星堆大事記 (1929–2000.2) (Chronicle of Guanghan Sanxingdui [1929– 2000.2]).” Zhonghua wenhua luntan 2001.1: 41–45.

Chen Xiandan and Liu Jiasheng 劉家勝 (2002). “Lun Sanxingdui wenhua yu Baodun wenhua zhi guanxi 論三星堆文化與寶墩文化之關係 (On the relationship between the Sanxingdui Culture and the Baodun Culture).” Sichuan wenwu 2002.4: 3–6.

Cheng, Te-k’un [= Zheng Dekun] (1949). “The T’ai-p’ing-ch’ang Cul-ture, Szechwan.” Hsieh-ta Journal of Chinese Studies 1 (1949): 67–81. Chengdu. Reprinted in Cheng, Te-k’un (1982), Studies in Chinese Archaeology, pp. 55–65. Hong Kong: The Chinese University Press.

Chengdu 1997: Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊 and Xindu Xian Wenwu Guanlisuo 新都縣文物管理所 (1997). “Sichuan Xindu Xian Guilinxiang Shang dai yizhi fajue jianbao 四川新都縣桂林鄉商代遺址發掘簡報 (Preliminary report of excavation of a Shang period site at Guilinxiang in Xindu County, Sichuan).” Wenwu 1997.3: 24–34.

Chengdu 1998: Chengdu Shi Wenwu Kaogu Gongzuodui, Sichuan Lianhe Daxue Lishixi Kaogu Jiaoyanshi 四川聯合大學歷史系考古教研室, and Xinjin Xian Wenguansuo 新津縣文管所 (1998).“Si-chuan Sheng Wenjiang Xian Yufucun yizhi diaocha yu shijue 四川省溫江縣魚鳧村遺址調查與試掘 (Survey and test excavation at the Yu-fucun site in Wenjiang County, Sichuan.” Wenwu 1998.12: 38–56.

Chengdu 2001: Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所, and Pixian Bowuguan 郫縣博物館 (2001). “Sichuan Sheng Pixian Qingjiangcun yizhi diaocha fajue shouhuo 四川省郫縣清江村遺址調查發掘收獲 (Results of survey and excavation of the Qingjiangcun site at Pixian County, Sichuan). In Chengdu Shi Wenwu Kaogu Yanjiusuo (editor), Chengdu kaogu faxian 1999 成都考古發現 1999 (Archaeological discoveries in Chengdu 1999), pp. 146–163. Beijing: Kexue Chubanshe.

Chengdu 2002: Chengdu Shi Wenwu Kaogu Yanjiusuo and Beijing Daxue Kaogu Wenboyuan 北京大學考古文博院 (2002). Jinsha

184 jay xu

taozhen—Chengdu Shi Jinshacun yizhi chutu wenwu 金沙淘珍—–成都市金沙村遺址出土文物 (Panning for treasure at Jinsha—Artifacts unearthed at Jinshacun, Chengdu). Beijing: Wenwu Chubanshe.

Chengdu 2003a: Chengdu Shi Wenwu Kaogu Yanjiusuo and Pixian Bowuguan (2003). “Sichuan Pixian Qingjiangcun yizhi fajue jianbao 四川郫縣清江村遺址發掘簡報 (Preliminary report of excavation of the Qingjiangcun site at Pixian County, Sichuan).” Wenwu 2003.1: 56–64.

Chengdu 2003b: Chengdu Shi Wenwu Kaogu Gongzuodui (2003). “Chengdu Shi Hetaocun Shang dai yizhi fajue jianbao 成都市核桃村商代遺址發掘簡報 (Preliminary report of excavation of a Shang period site at Hetaocun in Chengdu).” Wenwu 2003.4: 21–25.

Falkenhausen, Lothar von (2002). “Some reflections on Sanxingdui.” In I-tien Hsing (editor), Regional Culture, Religion, and Arts before the Seventh Century: Papers from the Third International Conference on Sinology, History Section, pp. 59–97. Taipei: Institute of History and Philology, Academia Sinica.

Falkenhausen, Lothar von (2003). “La culture de Sanxingdui: sources et contacts.” In Paris–Musées (editor), Chine, l’énigme de l’homme de bronze: Archéologie du Sichuan (XII–III siècle av. j.-c.), pp. 47–59. Paris: Paris–Musées.

Ge, Yan and Katheryn M. Linduff (1990). “Sanxingdui: a new Bronze Age site in Southwest China.” Antiquity 64 (1990): 505–513. [Chinese translation published in Sichuan wenwu 1991.6: 71–77.]

Goepper, Roger (editor) (1995). Das alte China: Menschen und Götter im Reich der Mitte. Essen: Kulturstiftung Ruhr.

Graham, David C. (1934). “A preliminary report of the Hanchow ex-cavation.” Journal of the West China Border Research Society 6: 114–131. Chengdu.

Henan 2001: Henan Sheng Wenwu Kaogu Yanjiusuo 河南省文物考古研究所 (2001). Zhengzhou Shang cheng—1953–1985 nian kaogu fajue baogao 鄭州商城—1953–1985 年考古發掘報告 (Report of the 1953–1985 excavations of the Shang city at Zhengzhou). Beijing: Wenwu Chubanshe.

Huang Jiaxiang 黃家祥 (1990). “‘Guanghan Sanxingdui yizhi’ de chubu fenxi ‘廣漢三星堆遺址’ 的初步分析 (Preliminary analysis of ‘the Sanxingdui site at Guanghan’).” Kaogu 考古 1990.11: 1030–1036.

Jiang Zhanghua 江章華 (1998a). “Shilun Sanxingdui wenhua Shi’erqiao wenhua yu zhoulin wenhua de guanxi 試論三星堆文化十二橋文化與周鄰文化的關係 (Preliminary discussion of the relationships be-tween the Sanxingdui Culture, the Shi’erqiao Culture, and cultures of neighboring regions).” Chengdu wenwu 成都文物 1998.1: 4–11.

Jiang Zhanghua (1998b). “Chengdu Shi’erqiao yizhi de wenhua xingzhi ji fenqi yanjiu 成都十二橋遺址的文化性質及分期研究 (A study of


cultural identity and periodization of the Shi’erqiao site of Chengdu).” In Sichuan Daxue Kaogu Zhuanye 四川大學考古專業 (editor), Sichuan Daxue Kaogu Zhuanye chuang jian sanshiwu zhounian jinian wenji 四川大學考古專業創建三十五周年紀念文集 (Collection of papers in commemoration of the thirty fifth anniversary of the founding of the archaeology program in the Sichuan University), pp. 146–164. Chengdu: Sichuan Daxue Chubanshe.

Jiang Zhanghua and Li Mingbin 李明斌 (2002). Guguo xunzong: Sanxing-dui wenhua de xingqi jiqi yingxiang 古國尋蹤: 三星堆文化的興起及其影響 (In search of the ancient kingdom: the rise of the Sanxingdui Culture and its influence). Chengdu: Ba Shu Shushe.

Jiang Zhanghua, Wang Yi 王毅, and Zhang Qing 張擎 (2002). “Cheng-du Pingyuan Xian Qin wenhua chulun 成都平原先秦文化初論 (Preliminary discussion of pre-Qin cultures in the Chengdu Plain).” Kaogu xuebao 考古學報 2002.1: 1–22.

Li Boqian 李伯謙 (1997). “Dui Sanxingdui wenhua ruogan wenti de renshi 對三星堆文化若干問題的認識 (Views regarding a few questions on the Sanxingdui Culture).” Kaoguxue yanjiu 考古學研究 3 (1997): 84–94.

Li Mingbin 李明斌 (1999a). “Guanghan Yueliangwan yicun shixi 廣漢月亮灣遺存試析 (Tentative analysis of the remains at Guanghan Yueliangwan).” Huaxia kaogu 華夏考古 1999.1: 26–35, 59.

Li Mingbin (1999b). “Sichuan Ya’an Shaxi yizhi taoqi ji xiangguan wenti de chubu yanjiu 四川雅安沙溪遺址陶器及相關問題的初步研究 (Preliminary study of pottery and related problems of the Shaxi site in Ya’an, Sichuan).” Kaogu 1999.2: 75–81.

Li Mingbin (2001a). “Shilun Yufucun yizhi disanqi yicun 試論魚鳧村遺址第三期遺存 (Preliminary discussion of Phase 3 remains at the Yufucun site).” Kaogu yu wenwu 考古與文物 2001.1: 40–41, 48.

Li Mingbin (2001b). “Guanyu Sanxingdui yizhi diyiqi wenhua yicun de jige wenti 關於三星堆遺址第一期文化遺存的幾個問題 (Some questions concerning Phase I cultural remains in the Sanxingdui site).” Chengdu wenwu 2001.2: 56–61.

Li Mingbin (2003). “Yangzishan tutai zai kao 羊子山土台再考 (A fur-ther study of the terraced mound at Yangzishan).” Gudai wenming 古代文明 2 (2003): 241–251.

Li Shaoming 李紹明, Lin Xiang 林向, and Xu Nanzhou 徐南洲 (eds.) (1991). Ba Shu lishi, minzu, kaogu, wenhua 巴蜀歷史, 民族, 考古, 文化 (History, ethnic identity, archaeology, and culture of Ba and Shu). Chengdu: Ba Shu Shushe.

Li Shaoming, Lin Xiang, and Zhao Dianzeng 趙殿增 (eds.) (1993). Sanxingdui yu Ba Shu wenhua 三星堆與巴蜀文化 (Sanxingdui and the Ba–Shu Culture). Chengdu: Ba Shu Shushe.

Lin Xiang 林向 (1988). “Yangzishan jianzhu yizhi xin kao 羊子山建築

186 jay xu

遺址新考 (A new study of the architectural remains at Yangzishan).” Sichuan wenwu 1988.5: 3–8.

Liu, Yang and Edmund Capon (2000). Masks of Mystery: Ancient Chinese Bronzes from Sanxingdui. Sydney: Art Gallery of New South Wales.

Louis, François (2003). “L’or au Sichuan à l’âge du bronze.” In Paris– Musées (editor), Chine, l’énigme de l’homme de bronze: Archéologie du Sichuan (XII–III siècle av. j.-c.), pp. 157–165. Paris: Paris–Musées.

Luan Fengshi 欒豐實, Fang Hui 方輝, and Jin Guiyun 靳桂雲 (2002). Kaoguxue lilun fangfa jishu 考古學理論方法技術 (Theories, methods, and techniques of Archaeology). Beijing: Wenwu Chubanshe.

Ma Jixian 馬繼賢 (1992). “Guanghan Yueliangwan yizhi fajue zhuiji 廣漢月亮灣遺址發掘追記 (Recollections of the excavations at Guang-han Yueliangwan).” Nanfang minzu kaogu 5: 310–323.

Ma Xinxin 馬辛辛 (1993). “Daba shanmai yu chuanbei shiqian wenhua de tantao 大巴山脈與川北史前文化的探討 (Discussion concerning prehistoric cultures in the Daba Mountains and northern Sichuan).” Sichuan wenwu 1993.5: 8–14.

Nishie Kiyotaka 西江清高 (editor) (2002). Fusang yu Ruomu—Riben xuezhe dui Sanxingdui wenming de xin renshi 扶桑與若木—–日本學者對三星堆文明的新認識 (New understandings by Japanese scholars on the Sanxingdui civilization). Chengdu: Ba Shu Shushe.

Paris 2003: Paris–Musées (editor) (2003). Chine, l’énigme de l’homme de bronze: Archéologie du Sichuan (XII–III siècle av. j.-c.). Paris: Paris–Musées.

Rawson, Jessica (editor) (1996). Mysteries of Ancient China. London: British Museum.

Sichuan 1957: Sichuan Sheng Wenwu Guanli Weiyuanhui 四川省文物管理委員會 (1957). “Chengdu Yangzishan tutai yizhi qingli baogao 成都羊子山土台遺址清理報告 (Salvage report of remains of the Yangzishan terrace in Chengdu).” Kaogu xuebao 1957.4: 17–31.

Sichuan 1959: Sichuan Sheng Bowuguan 四川省博物館 (1959). “Si-chuan Xinfan Shuiguanyin yizhi fajue jianbao 四川新繁水觀音遺址發掘簡報 (Preliminary report of excavation of the Shuiguanyin site at Xinfan, Sichuan).” Kaogu 1959.8: 404–410.

Sichuan 1961: Sichuan Daxue Lishixi Kaoguxue Jiaoyanzu 四川大學歷史系考古學教研組 (1961). “Guanghan Zhongxing gongshe gu yizhi diaocha jianbao 廣漢中興公社古遺址調查簡報 (Preliminary report of field survey of the ancient site at the Zhongxing Commune of Guanghan).” Wenwu 1961.11: 22–27.

Sichuan 1987a: Sichuan Sheng Wenwu Guanli Weiyuanhui, Sichuan Sheng Bowuguan, and Guanghan Xian Wenhuaguan 廣漢縣文化館 (1987). “Guanghan Sanxingdui yizhi 廣漢三星堆遺址 (The San-xingdui site at Guanghan).” Kaogu xuebao 1987.2: 227–254.

Sichuan 1987b: Sichuan Sheng Wenwu Guanli Weiyuanhui, Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 , and


Sichuan Sheng Guanghan Xian Wenhuaju 四川省廣漢縣文化局 (1987). “Guanghan Sanxingdui yizhi yihao jisikeng fajue jianbao 廣漢三星堆遺址一號祭祀坑發掘簡報 (Preliminary report of excavation of No. 1 sacrificial pit at Sanxingdui).” Wenwu 1987.10: 1–15.

Sichuan 1987c: Sichuan Sheng Wenwu Guanli Weiyuanhui, Sichuan Sheng Wenwu Kaogu Yanjiusuo, and Chengdu Shi Bowuguan 成都市博物館 (1987). “Chengdu Shi’erqiao Shang dai jianzhu yizhi diyiqi fajue jianbao 成都十二橋商代建築遺址第一期發掘簡報 (Prelimi-nary report of the first excavation of the Shang period architectural remains at Chengdu Shi’erqiao).” Wenwu 1987.12: 1–23, 37.

Sichuan 1989: Sichuan Sheng Wenwu Guanli Weiyuanhui, Sichuan Sheng Wenwu Kaogu Yanjiusuo, and Guanghan Shi Wenhuaju Wenguansuo 廣漢市文化局文管所 (1989). “Guanghan Sanxingdui yizhi erhao jisikeng fajue jianbao 廣漢三星堆遺址二號祭祀坑發掘簡報 (Preliminary report of excavation of Sacrificial Pit No. 2 at Sanxingdui).” Wenwu 1989.5: 1–20.

Sichuan 1990: Sichuan Sheng Wenwu Guanli Weiyuanhui, Sichuan Sheng Wenwu Kaogu Yanjiusuo, and Sichuan Sheng Ya’an Diqu Wenwu Guanlisuo 四川省雅安地區文物管理所 (1990). “Ya’an Shaxi yizhi fajue ji diaocha baogao 雅安沙溪遺址發掘及調查報告 (Report of excavation and survey of the Shaxi site at Ya’an).” Nanfang minzu kaogu 3: 293–339.

Sichuan 1992: Sichuan Sheng Wenwu Kaogu Yanjiusuo Sanxingdui Gongzuozhan 四川省文物考古研究所三星堆工作站 , Sichuan Sheng Guanghan Shi Wenguanhui 四川省廣漢市文管會, and Shi-fang XianWenguansuo 什邡縣文管所 (1992). “Sichuan Guanghan, Shifang Shang Zhou yizhi diaocha baogao 四川廣漢, 什邡商周遺址調查報告 (Report of field surveys of Shang and Zhou sites in Guang-han and Shifang of Sichuan).” Nanfang minzu kaogu 5: 295–309.

Sichuan 1998: Sichuan Sheng Wenwu Kaogu Yanjiusuo Sanxingdui Gongzuozhan, and Guanghan Shi Wenwu Guanlisuo 廣漢市文物管理所 (1998). “Sanxingdui yizhi Zhenwu Cangbaobao jisikeng diao-cha jianbao 三星堆遺址真武倉包包祭祀坑調查簡報 (Preliminary report of investigation of a sacrificial pit at Zhenwu Cangbaobao in the Sanxingdui site).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo (editor), Sichuan kaogu baogaoji 四川考古報告集 (Collection of excava-tion reports of Sichuan), pp. 78–90. Beijing: Wenwu Chubanshe.

Sichuan 1999: Sichuan Sheng Wenwu Kaogu Yanjiusuo (editor) (1999). Sanxingdui jisikeng 三星堆祭祀坑 (Excavation of the sacrificial pits at Sanxingdui). Beijing: Wenwu Chubanshe.

So, Jenny F. (2001). “Jade and stone at Sanxingdui.” In Robert Bagley (editor), Ancient Sichuan: Treasures from a Lost Civilization, pp. 153–175. Seattle: Seattle Art Museum in association with Princeton University Press.

188 jay xu

Song Zhimin 宋治民 (1983). “Guanyu Shu wenhua de jige wenti 關於蜀文化的幾個問題 (A few questions concerning the Shu Culture).” Kaogu yu wenwu 1983.2: 71–80.

Song Zhimin (1991). “Cong Sanxingdui de xin faxian kan zaoqi Shu wenhua 從三星堆的新發現看早期蜀文化 (The early Shu Culture in the light of new discoveries at Sanxingdui).” In Li Shaoming et al. (eds.), Ba Shu lishi, minzu, kaogu, wenhua 巴蜀歷史·民族·考古·文化 (History, ethnic identity, archaeology, and culture of Ba and Shu), pp. 207–223. Chengdu: Ba Shu Shushe.

Stuiver, Minze, Reimer, Paula J., and Reimer, R. W. (2005). CALIB 5.0. [WWW program and documentation]. <http://radiocarbon. pa.qub.ac.uk/calib/>.

Sun Hua (1992). “Shilun Guanghan Sanxingdui yizhi de fenqi 試論廣漢三星堆遺址的分期 (Preliminary discussion of periodization of the Sanxingdui site of Guanghan).” Nanfang minzu kaogu 5: 10–24. Reprinted in Sun Hua 2000: 49–67.

Sun Hua (1993). “Yangzishan tutai kao 羊子山土台考 (Study of the terraced mound at Yangzhishan).” Sichuan wenwu 1993.1: 3–7.

Sun Hua (1996). “Chengdu Shi’erqiao yizhiqun fenqi chulun 成都十二橋遺址群分期初論 (Preliminary discussion of periodization of the cluster of Shi’erqiao sites in Chengdu).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo (editor), Sichuan kaogu lunwenji 四川考古論文集 (Collection of research papers on Sichuan archaeology), pp. 123–144. Beijing: Wenwu Chubanshe. Reprinted in Sun Hua 2000: 68–88.

Sun Hua (2000). Sichuan pendi de qingtong shidai 四川盆地的青銅時代 (Bronze Age of the Sichuan Basin). Beijing: Kexue Chubanshe.

Taipingyang 1999: Taipingyang Wenhua Jijinhui 太平洋文化基金會 (1999). Sanxingdui chuanqi, Huaxia gu wenming de tansuo 三星堆傳奇, 華夏古文明的探索 (Legend of Sanxingdui: exploring the ancient civilization of Huaxia). Taipei: Taipingyang Wenhua Jijinhui.

Thote, Alain (2003). “Le jade et la pierre à Sanxingdui.” In Paris–Musées (editor), Chine, l’énigme de l’homme de bronze: Archéologie du Sichuan (XII–III siècle av. j.-c.), pp. 113–132. Paris: Paris–Musées.

Wang Jiayou 王家祐 and Jiang Dianchao 江甸潮 (1958). “Sichuan Xinfan Guanghan gu yizhi diaocha ji 四川新繁廣漢古遺址調查記 (Notes of field surveys in Xinfan and Guanghan of Sichuan).” Kaogu tongxun 考古通訊 1958.8: 27–31.

Wang Renxiang 王仁湘 and Ye Maolin 葉茂林 (1993). “Sichuan pendi beiyuan xinshiqi shidai kaogu xinshouhuo 四川盆地北緣新石器時代考古新收獲 (New results in archaeology of the Neolithic period in the northern edges of the Sichuan Basin).” In Li Shaoming et al. (eds.), Sanxingdui yu Ba Shu wenhua (Sanxingdui and the Ba–Shu Culture), pp. 257–265. Chengdu: Ba Shu Shushe.

Wang Yi 王毅 (1991). “Chengdu Shi Ba Shu wenhua yizhi de xin


faxian 成都市巴蜀文化遺址的新發現 (New discoveries of Ba–Shu culture sites in Chengdu).” In Li Shaoming et al. (eds.), Sanxingdui yu Ba Shu wenhua (Sanxingdui and the Ba–Shu Culture), pp. 295–309. Chengdu: Ba Shu Shushe.

Wang Yi, Jiang Cheng 蔣成, and Jiang Zhanghua (1999). “Chengdu diqu jinnian kaogu zongshu 成都地區近年考古綜述 (A general sur-vey of archaeology in the Chengdu region in recent years).” Sichuan wenwu 1999.3: 3–12.

Wang Yi and Sun Hua (1999). “Baoduncun wenhua de chubu renshi 寶墩村文化的初步認識 (Preliminary understanding of the Baoduncun Culture).” Kaogu 1999.8: 60–73.

Wang Yi and Zhang Qing 張擎 (1999). “Sanxingdui wenhua yanjiu 三星堆文化研究 (A study of the Sanxingdui Culture).” Sichuan wenwu 1999.3: 13–22.

Xi’nan 1954: Xi’nan Bowuyuan Choubeichu 西南博物院籌備處 (1954). “Baocheng tielu xiuzhu gongcheng zhong faxian de wenwu jianjie 寶成鐵路修築工程中發現的文物簡介 (Brief introduction of cultural remains discovered during the construction of the Baoji–Chengdu railroad).” Wenwu cankao ziliao 文物參考資料1954.3: 10–34.

Xiao Xianjin 蕭先進, Ao Tianzhao 敖天照, Liu Jiasheng 劉家勝, and Bao Yuzhi 包育智 (2001). Sanxingdui faxian fajue shimo 三星堆發現發掘始末 (The story of the discovery and fieldwork at Sanxingdui). Chengdu: Sichuan Renmin Chubanshe.

Xu, Jay (2001a). “Sichuan before the Warring States Period.” In Robert Bagley (editor), Ancient Sichuan: Treasures from a Lost Civilization, pp. 21–37. Seattle: Seattle Art Museum in association with Princeton University Press.

Xu, Jay (2001b). “Bronze at Sanxingdui.” In Robert Bagley (editor), Ancient Sichuan: Treasures from a Lost Civilization, pp. 59–151. Seattle: Seattle Art Museum in association with Princeton University Press.

Xu, Jay (2001c). “Reconstructing Sanxingdui imagery: some specula-tions.” Orientations (May 2001): 32–44.

Xu, Jay (2003). “Le site de Sanxingdui.” “Métallurgie du bronze à San-xingdui,” “Les arbres en bronze de Sanxingdui.” In Paris–Musées (editor), Chine, l’énigme de l’homme de bronze: Archéologie du Sichuan (XII–III siècle av. j.-c.), pp. 35–46, 61–112, 135–155. Paris: Paris–Musées.

Zhao Dianzeng 趙殿增 (1987). “Ba Shu wenhua jige wenti de tantao 巴蜀文化幾個問題的探討 (Inquiries into a few problems of the Ba Shu culture).” Wenwu 1987.10: 18–21.

Zhao, Dianzeng (1996). “The sacrificial pits at Sanxingdui.” In Jessica Rawson (editor), Mysteries of Ancient China, pp. 232–239. London: British Museum.

Zhao Dianzeng and Chen De’an 陳德安 (2001). “Yige chongman huoli de xueke shengzhangdian—Su Bingqi xiansheng zhidao xia

190 jay xu

de Sanxingdui kaogu 一個充滿活力的學科生長點—–蘇秉琦先生指導下的三星堆考古 (A research area full of vitality—Sanxingdui archaeology under the guidance of Mr. Su Bingqi).” In Su Bai 宿白 (editor), Su Bingqi yu dangdai Zhongguo kaoguxue 蘇秉琦與當代中國考古學 (Su Bingqi and Chinese archaeology at present), pp. 451–469. Beijing: Kexue Chubanshe.

Zheng Dekun 鄭德坤 (= Cheng Te-k’un) (1946). Sichuan gudai wenhua shi 四川古代文化史 (Cultural history of ancient Sichuan). Chengdu: West China Union University Museum Monograph Series No. 1.

Zhongguo 1982: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (editor) (1982). Kaogu gongzuo shouce 考古工作手冊 (Manual of field archaeology). Beijing: Wenwu Chubanshe.

Zhongguo 1990: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Sichuan Gongzuodui 中國社會科學院考古研究所四川工作隊 (1990). “Sichuan Mianyang Bianduishan xinshiqi shidai yizhi diao-cha baogao 四川綿陽邊堆山新石器時代遺址調查報告 (Report of field survey of the Neolithic period Bianduishan site at Mianyang, Sichuan).” Kaogu 1990.4: 307–313.

Zhongguo 1991a: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shiyanshi 中國社會科學院考古研究所實驗室 (1991). “Fangshe-xing tansu ceding niandai baogao (shiba) 放射性碳素測定年代報告 (十八) (Report of radiocarbon dates, No. 18).” Kaogu 1991.7: 657–663.

Zhongguo 1991b: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo (edi-tor) (1991). Zhongguo kaoguxue zhong tanshisi niandai shujuji, 1965–1991 中國考古學中碳十四年代數據集, 1965–1991 (Radiocarbon dates in Chinese archaeology: 1965–1991). Beijing: Wenwu Chubanshe.

Zhongguo 1992: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shi-yanshi (1992). “Fangshexing tansu ceding niandai baogao (shijiu) 放射性碳素測定年代報告 (十九) (Report of radiocarbon dates, No. 19).” Kaogu 1992.7: 655–662.

Zhongguo 1993: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shi-yanshi (1993). “Fangshexing tansu ceding niandai baogao (erling) 放射性碳素測定年代報告 (二○) (Report of radiocarbon dates, No. 20).” Kaogu 1993.7: 645–649.

Address

Department of Asian ArtThe Art Institute of Chicago111 South Michigan AvenueChicago, IL 60603-6110USAEmail: [email protected]


THE EXTERNAL CONNECTIONS OF SANXINGDUI

BY

LOTHAR von FALKENHAUSEN(UCLA)

Abstract An idiosyncratic Bronze Age culture, thought to be roughly contemporaneous with the Late Shang (ca. 1300–1050 BC) in the Yellow River Basin, has recently been dis-covered in the Chengdu Plain 成都平原. The large walled settlement at the type site of Sanxingdui, Guanghan (Sichuan) 四川廣漢三星堆 and its highly developed bronze and jade manufacturing traditions indicate the presence of state-level civilization. This article attempts to clarify some of the relationships to the Sanxingdui culture to earlier, contemporaneous, and later archaeological cultures in the surrounding areas. Objects from the so-called “sacrificial pits” at Sanxingdui and at the slightly later site of Jinsha, Chengdu (Sichuan) 四川成都金沙 are compared to archaeological finds from other parts of China, revealing significant connections to Neolithic and Early Bronze Age cultures along the Middle and Lower Yangzi as well as in the Central Plain, and showing that, for all its unusual features, Bronze Age Sichuan was by no means isolated in the cultural interaction sphere of mainland East Asia. Moreover, it can now be shown that Sanxing-dui traits survived in later archaeological contexts, mainly in Southwest China.

Complementing another article (Falkenhausen 2002), in which I at-tempted to explain the two “sacrificial pits” from Sanxingdui in their local cultural context under the methodological pretense that their as-semblages were internally coherent (Falkenhausen 2002),1 the present study compares the objects from these two pits, to which one can now add the slightly later finds from Jinsha, to archaeological discoveries in other parts of China. As the Sanxingdui finds are now comprehensively published (Sichuan 1999) and lodged in the international consciousness due to having been prominently exhibited in Europe (Goepper [ed.] 1995, Rawson [ed.] 1996, Thote [ed.] 2003), Japan (Sanseitai 1998), Taiwan (Sanxingdui 1999), Australia (Liu [ed.] 2000), and North America

1 An early version of Falkenhausen 2002 followed by a preliminary version of the present article has been published in Chinese as the introduction to Luo Tai (ed.) 2003. An abbreviated version of the present article is appearing in the form of two separate contributions to Thote (ed.) 2003. I am extremely grateful to Jay Xu and Alain Thote for helpful comments and suggestions.

192 Lothar von Falkenhausen

(Yang [ed.] 1999, Bagley [ed.] 2001), and with some of the Jinsha finds accessible through exhibition catalogues (Chengdu 2002; Thote [ed.] 2003), the time seems ripe for a comprehensive look at their wider relationships. Where did the splendid Bronze Age culture of Sanxingdui come from? And where did it go after the end of its florescence in the Chengdu Plain? One way of approaching the first question is to subdivide the Sanxingdui and Jinsha assemblages into different stylistic groups and then trace each of these groups to its area of origin (e.g. Bagley 1988, 1990, 1992; Ge and Linduff 1990). Yan Ge and Katheryn M. Linduff (1990), for instance, who are interested primarily in elucidating the extent of Sanxingdui’s cultural linkages with the Central Plains 中原, propose a binary distinction between those objects that have evident parallels at major sites of the Shang culture (Group A), and those that do not (Group B). They believe Group B to have originated locally in the Sichuan Basin. But as soon as one takes a closer look, one is bound to notice that, in actuality, neither Group A nor Group B is uniform. Items and groups of items under each of the two headings can be connected to a number of different places and periods, testifying to relationships of different nature and time-depth. As we discuss these in the first part of this article, we shall find that the Sanxingdui culture is unlikely to have “come from” any one place. At the same time, we shall also see that Sanxingdui was by no means a unique, isolated, or inexplicable phenomenon. To answer the second question, one may proceed similarly and trace individual Sanxingdui characteristics in later archaeological contexts in the surrounding areas. Here, too, our findings in the second part of this article suggest a complex picture. Evidently, the Sanxingdui culture did not “go” to any one place in its entirety. But even though the particular combination of traits that characterizes Sanxingdui in the Chengdu Plain during the late second millennium BC was not perpetuated either locally or anywhere else during later centuries, the survival of piecemeal Sanxingdui features attests to its lasting impact on the development of material culture in a number of different contexts in southwestern China and beyond.

Part I: Earlier and Contemporaneous Parallels

Shijiahe Iconographic Parallels. The enigmatic human and animal imag-ery of the Sanxingdui bronze statuary (fig. 1) seems at first sight to be without parallel anywhere else. A possible clue for its derivation may, however, be found in the jade carvings of the Late Neolithic Shijiahe 石

THE external connections of SANXINDUI 193

家河 culture in Hubei (ca. 2500 to 1900 BC [cf. Zhongguo 1991: 183–84; Hubei 1999a: 348]). The best-documented instances come from the site cluster of Xiaojiawuji, Tianmen (Hubei) 湖北天門蕭家屋脊, which has yielded a num-ber of representations of human heads in jade (Hubei 1999a).2 Although they measure but a few centimeters in size, they display some remarkable similarities to the Sanxingdui bronze heads.3 If, for instance, we compare the physiognomic details of the Shijiahe jade head in fig. 2.1 to the Sanxingdui bronze head in fig. 2.2, we see the same drawn-out mouths with downward-bent corners, almond-shaped eyes (with eyeballs at Shijiahe, not always at Sanxingdui),4 noses with nostrils that end in spirals, and stylized ears with holes for earrings. The variously shaped upper portions of the Shijiahe jade heads, as well, show remarkable parallels to the San-xingdui bronze statuary. The specimen in fig. 2.1, for instance, wears an ornate diadem which, though simpler in execu-tion, is structurally similar to those seen on the large standing figure in fig. 1 and on other Sanxingdui bronze heads; both the conical top of the Shijiahe specimen in fig. 2.3 and the flat top of the one in fig. 2.5 have parallels at Sanxingdui (figs. 2.4 and 2.6). Most convincing is the case of

Figure 1. Standing human figure from Pit no. 2 at Sanxingdui, Guanghan (Sichuan). Second half of second millennium BC. From Sichuan 1999: 162 (insert), fig. 82.

2 Jades of some of the same types represented at Xiaojiawuji have also been found at other Shijiahe sites, e.g. at Zaolingang, Jingzhou (Hubei) 湖北荊州棗林崗 (Hubei 1999b: 15–42). For general remarks see Hayashi 1998. 3 Several of these jades are now on permanent loan to the Shanghai Museum. 4 On the closed eyes of Sanxingdui statuary, see Wu Hung 1997; Thorp 1999: 219; Falkenhausen 2002: 82–83. It is likely that eyeballs were sec-ondarily painted or mechanically affixed on the bronze statues now appearing to lack them.

Figure 2. Juxtaposition of jade figures from Xiaojiawuji and bronze statuary from San-xingdui. 2.1: Jade figurine from Xiaojiawuji, Tianmen (Hubei). Shijiahe culture, late third millennium BC. From Hubei 1999a: 316, fig. 251. 2. (flat top, large diadem). 2.2: Bronze mask from Pit no. 2 at Sanxingdui. From Sichuan 1999: 189, fig. 103. (flat top). 2.3: Jade figurine from Xiaojiawuji. From Hubei 1999a: 316, fig. 251.4. (conical top, seen in profile). 2.4: Bronze head from Pit no. 2 at Sanxingdui. From Sichuan 1999: 178, fig. 96. (round top). 2.5: Jade figurine from Xiaojiawuji. From Hubei 1999a: 316, fig. 251.1. (flat top). 2.6: Bronze head from Pit no. 2 at Sanxingdui, Guanghan (Si-chuan). From Sichuan 1999: 175, fig. 91. (flat top). 2.7: Jade figurine from Xiaojiawuji. From Hubei 1999a: 316, fig. 251.3. (has braid around head). 2.8: Bronze head from Pit no. 2 at Sanxingdui. From Sichuan 1999: 172, fig. 88. (with braid around forehead).


Figure 2 (cont). 2.9: Jade unspecified animal mask from Xiaojiawuji. From Hubei 1999a: 324, fig. 256.1. 2.10: Bronze unspecified animal mask from Pit no. 2 at Sanxingdui. From Sichuan 1999: 198, fig. 111.3. 2.11: Jade bird images from Xiaojiawuji. From Hubei 1999a: 327, fig. 258.4 and 329, fig. 260.1. 2.12: Bronze bird-shaped clapper-bell from Sanxingdui. From Sichuan 1999: 299, fig. 163.2. 2.13: Jade cicada from Xiaojiawuji. From Hubei 1999a: 319, fig. 253.4 2.14: Cicada-derived (?) bell from Sanxingdui. From Sichuan 1999: 299, fig. 164.2.


the jade head in fig. 2.7, which features a hair braid (?) slung around the head;5 it has an exact counterpart in Pit no. 2 at Sanxingdui (fig. 2.8). Heads dressed in a similar fashion also occur elsewhere in Neolithic human representations (fig. 3),6 suggesting that they carried a spe-cific iconographic meaning. The same is likely to have been true of the various other types of jade heads.7 Further parallels between Shijiahe jades and Sanxingdui bronzes can be found in animal imagery. In their gen-eral outlines and internal structure, for instance, the unspecified animal masks (which may be among the iconographic forerunners of the taotie 饕餮 face ubiq-uitous on Shang bronzes) among the Xiaojiawuji jades (fig. 2.9) are quite comparable to the bronze masks found at Sanxingdui (fig. 2.10)—more so, indeed, than are contemporaneous instances of the taotie from the centers of the Shang culture sphere (fig. 4). Bird imagery is also

present at both sites (figs. 2.11 and 2.12), and perhaps not accidentally, the overall shape of one of the jingle-bells from Pit no. 2 at Sanxingdui (fig. 2.13) reminds of that of the jade cicadas found at Xiaojiawuji (fig.

5 Admittedly, the attachment of the braided head covering is not clear on either the Xiaojiawuji or the Sanxingdui specimen; my suspicion that a hair braid is represented stems from the fact that the flat-top bronze heads at Sanxingdui all have braided pony-tails (see fig. 1.6). One should admit the possibility that the braid is actually some kind of crown or diadem. (Conversation with Prof. Hayashi Minao, November 16, 2002). 6 Instances include a human-shaped plaque in the Palace Museum, Beijing (Deng Shuping 1985, color plate 24), a plaque in the C. T. Loo collection, Paris (Salmony 1938, pl. 32.6), and a rectangular mask in The Art Institute, Chicago (ex Sonnenschein collection) (Salmony 1938, pl. 9.4). Even though, in the case of such unprovenienced pieces, the locus of manufacture remains a matter of conjecture, at least the first two of these three items may be cautiously assigned to the Shijiahe jadeworking tradition as characterized by Hayashi (1998). An instance of such a headdress on a (reworked?) figurine of Han or later date in the Royal Ontario Museum (Dohrenwend 1971: 92) must remain enigmatic for the moment. 7 An additional specific parallel to Shijiahe (and perhaps more generally Longshanoid) figural imagery on jades may be observed on a fragmentary sceptre found in the vicin-ity of the Sanxingdui site (Sanseitai 1998, no. 153). It shows an abstacted version of two facing heads with long flowing hair, resembling representations on known Longshanoid specimens (Deng Shuping 1998).

Figure 3. Human-shaped plaque. Shijiahe culture (?), late third millennium BC. Palace Museum, Beijing. After Deng Shuping 1985, color plate 24.


2.14).8 Aside from the motifs, there is also a more fundamental parallel of style. The elegant curvilinear contour lines and filler motifs on the Shijiahe jades are remarkably similar in ductus, execution, and, seem-ingly, artistic intention to those observable on objects from Sanxingdui, whether in bronze (e.g. on the base of the standing figure from Pit no. 2 [fig. 1] and on the small figures of the “spirit altars” [see fig. 25.4]), jade (e.g. on the engraved prism-shaped blade in fig. 5.1), or gold (e.g. on the scepter sheath from Pit no. 1 [fig. 5.2], and on various gold objects from Jinsha [Chengdu 2002, nos. 2–4]). Geographically, of all the various Longshanoid cultures of the third millennium BC, Shijiahe is the closest to Sichuan (see map). Still, the two culture areas are separated by more than 500 km of mountainous terrain that poses forbidding difficulties to premodern traffic. Moreover, there is a chronological hiatus of more than half a millennium between the end

8 It was no doubt with a mimetic intention that Sanxingdui bronze casters gave some of the jingle-bells that were to be suspended from the “spirit trees” the shape of sound-generating animals, as shown in figs. 2.12 and 2.14 (though bird and insect imagery is by no means limited to such contexts).

Figure 4. Unspecified animal masks from Tomb no. 5 at Yinxu (“Fu Hao’s tomb”), Anyang (Henan). Late Shang (early 13th c. BC).

From Zhongguo 1980: 21, fig. 10.1 and 22, fig. 11.2.


of Shijiahe and the time when the objects from the two sacrificial pits at Sanxingdui are thought to have been interred. The recent large-scale rescue excavations in the Three Gorges Dam Area 三峽庫區 have begun to bridge this double gap by revealing the existence, in the valley of the Yangzi 揚子江 (長江) river downstream from Chongqing 重慶, of Late Neolithic phases that are contemporaneous and apparently related to Shijiahe (Guojia Wenwuju 2001; Changjiang 2002; Chongqing 2001; Sun Hua 2000). These remains in turn are linked to the roughly contem-poraneous Baodun 寶墩 culture of the Chengdu Plain, which features walled settlements (Chengdu 1997, 2000; Zhongri 1998; Xu 2001c: 23) that may be comparable to those of the Shijiahe culture (Zhang Xuqiu 1992: 238–315; Zhao Hui 1999; Nakamura 1997). There is thus little doubt that cultural links between the Chengdu Plain and the Middle Yangzi existed during Late Neolithic times. Baodun, moreover, is in a continuous sequence with the Bronze Age civilization of Sanxingdui in the same geographical area. Since by Sanxingdui times the Shijiahe

Panlong-cheng

Xiaojia-wuji

Anyang

Fen R.

Nu R

iver

Lancang (Mekong) R

iver

Jinsha River

Yalong River

Dadu River

Fu Riv er Min R

iver

Jialin g R.

Chengdu

Sanxing-dui

Jinsha

Shizhai-shan

Wushan

Moutuo

North China Plain

Yellow Sea

EastChinaSea

Yangzi River

Xia

ng R

.Yua

n R.

Gan

R.

Han Shui

VIETNAM

INDIA

Hong

Mekong

Dayang-zhou

Wei River

Hong Shui

Xi River

Ruo

Shu

i

Xi’an

Taiyuan

Chongqing

Kunming

Changsha

Zhengzhou

Guangzhou

Wuhan

Shanghai

Hangzhou

Nanjing

Beijing

Ye

llow

River (Huang He)

South China Sea

ErligangErlitou

Laoniupo

Sidun

Rujia-zhuang

BURMA

0 500 1000 Miles

0 500 1000 1500 Kilometres


culture had long disappeared, it seems likely that iconographic elements of Shijiahe derivation reached Sichuan during Neolithic times. Here they were preserved and adapted to dif-ferent media. Whether they retained any of their original meanings is as unclear as are the meanings them-selves. In any case, it appears that the iconography of the Sanxingdui bronzes constitutes a survival, albeit probably in considerably changed form, of a far more ancient tradition of representation that had become obsolete in its likely area of origin. As we shall see, such a situation is typical for the ancient cultures of the Sichuan Basin. Ritual Jades: Longshanoid Connec-tions. Ritual lithics such as cong 琮 tubes (fig. 6.1), bi 璧 disks (fig. 6.2), and zhang 璋 scepters (fig. 6.3) were first discovered in the Sanxingdui area in the 1930s, and locally made specimens in a variety of materials— jade, stone, and bronze—have been excavated at both Sanxingdui and Jinsha. Cheng Te-k’un 鄭德坤 (1949) was the first to offer wide-ranging comparisons of these objects with finds from elsewhere in East Asia. With the incomparably richer archaeological data reported to date, we can now be far more precise than

Cheng in drawing such parallels. It is now known, for instance, that cong, bi, and zhang, all of which Cheng blithely assumed to date to the Zhou dynasty, go back to Neolithic times, more specifically to the Longshan 龍山 period (ca. 3000–1900 BC). In Sichuan, as well, the so far earliest archaeological manifestation of polished-stone working in the Chengdu Plain is at Renshou 任壽, a Baodun period cemetery on the western side just outside the Bronze Age enclosure at Sanxingdui,9 which yielded a

9 Unpublished; seen at Sanxingdui in March 1999 thanks to Prof. Chen Xiandan.

Figure 5. Drawn figural décor on objects from Sanxingdui. 5.1: Prism-shaped blade from Pit no. 2. From Sichuan 1999: 361, fig. 197.1. 5.2: Décor of gold sheath from Pit no. 1. From Sichuan 1999: 61, fig. 34.1.


number of objects made of chlorite, a locally occurring mineral that, when polished, looks like jade.10 That such originals did indeed reach Sichuan is now attested by a cong tube from Jinsha (Chengdu 2002: no. 20; discussed by Xie Hui 2003) (fig. 7.1). It is clearly recognizable as a product of the Liangzhu 良渚 culture, the Longshanoid culture of the Lower Yangzi basin and adjacent areas (present-day southern Jiangsu, Shanghai, and northern Zhejiang). With ten tiers of carefully carved mask décor on the square middle portion, it is very close to specimens from the cemetery of Sidun, Wujin (Jiangsu) 江蘇武進寺墩 (Nanjing1984) (fig.7.2), which exemplify

10 A number of chlorite objects in the shape of ritual jades are on display in the Mu-seum of Sichuan University. Acquired on the art market before World War II, they are said to come from Sanxingdui. Almost all of them have had carved ornaments added to make them more saleable, and some may be all-out forgeries. The exact date of the genuine pieces is currently impossible to determine.

Figure 6. Ritual lithics found in the Sanxingdui area in the early 1930s. 6.1: Jade cong tube. After Cheng 1949 (1982 edition), pl. 13.d. 6.2: Stone bi disk. After Cheng 1949 (1982 edition), pl. 13.b. 6.3: Stone zhang

scepter. After Cheng 1949 (1982 edition), pl. 14.e.


the late phase (ca. 2200–1900 BC) of the Liangzhu culture. Posteriorly engraved near its upper edge is a symbolic marker that resembles a human figure with its arms outstretched and wearing an extravagant headdress; while somewhat gauche in its execution, it may also be linked to similar motifs seen on other Liangzhu jades (Teng Shu-p’ing 2000). It stands to reason that such an object could not have reached Sichuan directly from its distant area of origin: it must have passed through the hands of various intermediaries along the Yangzi river. Perhaps, indeed, it followed a somewhat circuitous path of transmission; for curiously, although cong and bi are among the commonest ritual jade types in the Longshanoid cultures along the East Coast of China and in the Yellow River Basin, no specimens appear to be present at the Shijiahe sites that are geographically intermediary between those areas and Sichuan.11

11 Two alleged cong fragments have been reported from Zaolingang (Hubei 1999b: pls. 10.6 and 11.2), but their identification is somewhat questionable. At Xiaojiawuji,

Figure 7. Liangzhu-type cong from Jinsha and Sidun. 7.1: Cong from Jinsha, Chengdu (Sichuan). Probably Liangzhu culture, late third millennium BC. From Chengdu 2002: 84. 7.2: Cong from tomb no. 3 at Sidun, Wujin (Jiangsu). Late phase of Liangzhu culture. From Nanjing 1984: 119, fig. 9.4. 7.3: Cong from Jinsha. Possibly

made in Sichuan, date uncertain. From Chengdu 2002: 87.


Such objects may thus have reached the Chengdu Plain by a different route, and/or in a different time, from the Shijiahe figural motifs dis-cussed above. Since the interment at Jinsha of the Liangzhu cong in fig. 7.1 post-dates its manufacture by at least a millennium, it is impossible to know now whether it had been in Sichuan since Neolithic times. We do know that in other parts of China, Neolithic jades were sometimes transmitted over many centuries (Hayashi 1991: 515–576; Rawson in Rawson [ed.] 1996:22–26 et passim). An active local engagement with cong of Liangzhu type seems to be indicated, at any rate, by a second cong from Jinsha (Chengdu 2002: no. 21) (fig. 7.3), which features a much simplified version of the mask décor and differs in shape, proportions, and material from specimens found in the Liangzhu core area; Alain Thote (in Thote [ed.] 2003: 115) believes it to be an imitation made in Sichuan.12 While we cannot be certain about the exact date of this particular specimen—the archaeological context merely imposes a vague terminus ante quem of ca. 1000 BC— the possibility that it dates back to the time of the Baodun culture may at least be preliminarily entertained, as the polished-stone objects from Renshou attest to a local interest in the material during that time. The finds from Sanxingdui and Jinsha abundantly show that such interest continued during the Bronze Age. In areas east of Sichuan, the widespread distribution, during Late Neolithic times, of ritual lithics and certain types of pottery vessels that are in part ancestral to the principal shapes of Shang bronzes, attests to élite ritual practices (and, presumably, religious ideas) that were shared over large parts of continental East Asia, transcending the differences among the many regional archaeological cultures and phases. K. C. Chang (1986: 234–94) has famously argued that this “Chinese inter-action sphere” archaeologically manifests significant cultural con-vergence during the period just preceding the rise of the historically known royal dynasties in the Yellow River basin. It seems unlikely that the Chengdu Plain was fully integrated into this network of élite-level exchange, as none of its characteristic ritual-jade and élite-pottery types have been found at sites of the Baodun culture so far. But even if we put the Jinsha cong aside for the moment, the chlorite objects from Renshou, along with the labor-intensive construction of walled enclosures and the

the only item vaguely relatable to the standard Longshanoid ritual lithic ensemble also occurring at Sanxingdui is a “handle-shaped object” (Hubei 1999 a, v. 2: pl. 176.5) that resembles a zhang with bifurcated end. Some of the less complex and distinctive lithic types at Sanxingdui—those classified in the Final Report as “spades,” “axes,” “long adzes,” and “drills,” do have counterparts among the nonjade stone artifacts from Xiaojiawuji and Zaolingang. 12 As long as the lithic material has not been scientifically sourced, one should perhaps remain open to the possibility that this cong was made in some other region intermediary between the Lower Yangzi region and Sichuan.


iconographic parallels to Shijiahe mentioned above, suggest that the area was at least indirectly affected by ongoing trends in areas further east. Such connections were undoubtedly an important factor in the rise of the local Bronze Age culture of Sanxingdui.

Ritual Jades: Erlitou Connections. In Sichuan as well as in the Yellow River Basin, the jade-producing industries of the Longshanoid cultures were the point of departure of strong regional traditions of élite lithic manu-facture during the Bronze Age. Even though both assemblages and typological features of the jade objects from Sanxingdui and Jinsha are highly idiosyncratic (cf. Bagley 1988: 79–80), their detailed analysis suggests a continued connection with areas to the East and North after the distinctive Longshanoid jade-producing traditions in the Middle and Lower Yangzi Basin (Shijiahe and Liangzhu) and along the East Coast of China had come to an end around 1900 BC. Zhang scepters are by far the most numerous among the ritual jade blades of Longshanoid derivation at both Sanxingdui and (apparently) Jinsha. The vast majority of objects thus classified by the Sanxingdui Report feature bifurcated ends and ornate, serrated hilts (fig. 8.1). Some of them are enhanced by additional openwork decoration and incised motifs relating to those seen in Sanxingdui bronze art. The most unusual instance, from Pit no. 1 at Sanxingdui, features a small openwork bird placed between the two branches of the bifurcated end and the outlines of another zhang with bifurcated end incised on the surface (fig. 8.2). Besides, the Sanxingdui Report also classifies as zhang a variety of objects of a shape that is completely different from zhang with bifurcated ends. Among them are some objects I have provisionally called “prism-shaped blades” (figs. 5.1 and 8.3), remarkable for their size and for their incised ornamentation. Although ultimately they go back to Neolithic origins (cf. various con-tributions to Deng Cong [ed.] 1994: 1–117) (fig. 9.1), the zhang scepters from Sanxingdui and Jinsha are most closely similar in shape to those from Periods III and IV of the Erlitou 二里頭 culture, North China’s first full-fledged Bronze Age culture (ca. 1900–1500 BC) (fig. 8.4) (for specimens from the Erlitou type site, see Zhongguo 1999: pls. 118.5, 168.2). The same is true of the very numerous jade “dagger-axes” (ge 戈) from Sanxingdui, which lack immediate Neolithic antecedents (for two Period III specimens from Erlitou, see Zhongguo 1999, pl. 118.3, 4) (figs. 8.5 and 8.6). Not used as weapons, they apparently served as symbolic objects comparable to zhang with bifurcated ends, and Robert W. Bagley (1988: 79) has noted a number of “hybrids” that combine features of zhang and ge (fig. 8.7; also fig. 8.2); such objects seem to be without parallel at Erlitou. Of another type of ritual jades seen in Period IV at Erlitou, yue 鉞 axes with serrated sides (fig. 8.8), no specimens were


Figure 8. Juxtaposition of jade blades from Sanxingdui and Erlitou. 8.1: Zhang sceptre with bifurcated end from Pit no. 1 at Sanxingdui. From Sichuan 1999: 64, fig. 36.2. 8.2: Ornate blade with incised depiction zhang sceptre from Pit no. 1 at Sanxingdui. From Sichuan 1999: 81, fig. 41.1. 8.3: Prism-shaped blade from Pit no. 1 at Sanxingdui. From Sichuan 1999: 64, fig. 35. 8.4: Zhang sceptre with bifurcated end from Erlitou, Yanshi (Henan). Erlitou Culture (ca. 1900–1500 BC). After Zhongguo 1999: 342, fig. 238.3 (improved). 8.5: Jade dagger-axe (ge) from Pit no. 2 at Sanxingdui. From Sichuan 1999: 379, fig. 206.5. 8.6: Jade dagger-axe (ge) from Erlitou. From Zhongguo 1999: 250, fig. 162.5. 8.7: Hybrid blade from Pit no. 1 at Sanxingdui combining features

of zhang and ge. From Sichuan 1999: 75, fig. 39.1.


Figure 8 (cont.). 8.8: Axe (yue) with serrated sides from Erlitou. After Zhongguo 1999: 342, fig. 238.1 (improved). 8.9: Axe (yue)

with serrated sides from Jinsha. From Chengdu 2002: 117.

Figure 9. Longshanoid ritual jades from North China. 9.1: Zhang sceptre with bifur-cated end from Simatai, Haiyang (Shandong). Coastal (Haidai) Longshan culture, late third millennium BC. After Wang Hongming 1985: 1062, fig. 6.3. 9.2: Cong tube from Tomb no. 3168 at Taosi, Xiangfen (Shanxi). Taosi (Shanxi Longshan) culture, ca. 2000 BC. From Gao Wei 1998: 194, fig. 20.1.5. 9.3: Bi disk from Shi’ang, Shenmu (Shaanxi). Keshengzhuang (Shaanxi Longshan) culture, late third millennium BC. From

Yang Yachang 1998: 212, fig. 22.3.17.


found at Sanxingdui, but one has been reported from Jinsha (Chengdu 2002, no. 33) (fig. 8.9). Since zhang with bifurcated ends occur over a wide area ranging from northern Shaanxi all the way to Hong Kong and Vietnam (Deng Cong [ed.] 1994: 116–218 and 451–454; Jay Xu in Bagley [ed.] 2001: 32–33; So 2001), one hesitates to assume that they were introduced directly from the Erlitou culture area in central Henan, but such a possibility does exist. Since the speci-mens from Sanxingdui are stylistically distinctive, e.g. in the execution of their hilts, they were probably manufactured locally, perhaps some time after their initial diffusion. That diffusion very probably occurred in the time contem-porary with the Erlitou culture in the Central Plains; for the huge “prism-shaped blade” from Pit no. 1 (fig. 8.3), with incised fine-line ornament on both its slanted narrow ends, has an exact counterpart at Erlitou (Zhongguo 1999, pl. 168.1; Fong 1980, no. 3). This object, at least, must thus have been obtained as an import from there, or both must have been derived from a common contemporaneous source.

The Neolithic antecedents of the zhang scepters seen at Erlitou as well as at Sanxingdui and Jinsha come from the Coastal (Haidai 海岱) Longshan culture of Shandong (fig. 9.1); they do not seem to have been current in the Liangzhu culture that produced the above-discussed cong from Jinsha. The single cong found in the sacrificial pits at Sanxingdui (fig. 10)13 also represents a type of cong that is represented in the Coastal Longshan culture and other Longshanoid phases of north China (fig. 9.2) but not in the Liangzhu culture area further to the south.14 Like

Figure 10. Cong from Pit no. 1 at Sanxingdui. From Sichuan

1999: 81, fig. 42.1.

13 Goepper’s statement (in Goepper [ed.] 1995: 249) that the Sanxingdui pits did not contain cong is incorrect. 14 The same is true of the cong collected before 1949 now in the Sichuan University Museum. A stray find from Yueliangwan within the area of the Sanxingdui walled site is in Sanseitai 1998, no. 152. It is the simple, Coastal Longshan cong type that was ancestral


most of the zhang, the cong in fig. 10 is presumably of local manufacture; whether such specimens also exist at Jin-sha side by side with the two Liangzhu-derived ones already mentioned is so far unknown. Both stone and jade specimens of cong have been unearthed at the Sanxingdui site (see, e.g., Cheng 1947); at least some of these may well antedate the time of the sacrificial pits. Their paucity in the Sanxingdui sac-rificial pits and at Jinsha—compared to the large number and considerable typological variety of zhang—bespeaks the idiosyncratic character of ritual-jade use in Bronze Age Sichuan. Flat bi disks—the third principal type of ritual jades in élite contexts in both

Liangzhu and the Bronze Age civilization of the Yellow River Basin (fig. 9.3)—are likewise virtually absent from the sacrificial pits at Sanxingdui and from the Jinsha deposits.15 Instead, a number of collared rings (fig. 11) have been found at both places. Even though Jessica Rawson (in Rawson [ed.] 1996: 84) considers these as representative of an “exotic, non-Shang type of jade,” such items have in fact been found, sometimes in considerable numbers, at Bronze Age sites in North China (e.g. in tomb no. 5 at Yinxu, Anyang [Henan] 河南安楊殷墟 [“Fu Hao’s tomb” 婦好墓], see Zhongguo 1980: pls. 87.1–2, 88.1–2, 90.3, 91.1–6, 93.3). They may well have been derived from, and be functionally related to, bi. Flat bi have been found elsewhere within the area of the Sanxingdui walled settlement (Cheng Te-k’un 1949; Sanseitai 1998, nos. 161–163; Bagley [ed.] 2001: 159, figs. 7–8) (fig. 6.2), e.g. at Zhenwu Cangbaobao 真武蒼包包, a little to the north of the two sacrificial pits, in a hoard of objects that, while dating to the Bronze Age, seem on the whole to be significantly earlier in date than those from the two “sacrificial pits” at Sanxingdui (Sichuan 1998). It is clear from the discovery of unfinished specimens (made of stone as well as jade) that bi, as well as cong, were

to cong seen throughout China in Shang and Western Zhou times. Conversely, the tiered, Liangzhu-type cong shape was revived in early Imperial times and has enjoyed great popularity as a shape of archaic reference in various media ever since. This shift, and the circumstances that occasioned it, deserve separate investigation. 15 At Sanxingdui, only one debatable instance of a flat bi was found in pit no. 2 (Si-chuan 1999: 412, fig. 219.1), but even it shows intimations of a ring around the center, which may have been fashioned in a secondary attempt at transforming a flat bi into a ritually appropriate collared bi.

Figure 11. “Collared disk” from Pit no. 2 at Sanxingdui. From Sichuan

1999: 370, fig. 202.4.


manufactured at the Sanxingdui site, though perhaps no longer in the time of the sacrificial pits. Regardless of the Chengdu Plain’s early connections with Long-shanoid cultures in areas further east, the major impetus for local ritual-lithic production—and, perhaps, usage—during Sanxingdui times thus seems to have occurred during the Bronze Age, quite pos-sibly beginning as early as Erlitou times. That there was a connection with the Central Plains can hardly be controversial, though it cannot be known at present whether this connection was a direct one or through some intermediary cultural entity.

Erlitou Connections: Ceramics. While Longshanoid ritual ceramics, as noted, were not introduced to Sichuan during Neolithic times, their Erlitou typological descendants are prominently present at the Sanxingdui settlement site (Sichuan 1987; Sun Hua 2000: 153–156; on the Erlitou materials see Zhongguo 1995, 1999). Sanxingdui pottery types with close Erlitou affinities include closed-spouted pitchers with pouch-shaped

Figure 12. Comparison of pottery types from Sanxingdui (left) and Erlitou (right). After Sun Hua 2000: 154, fig. 6.5.


16 He vessels of Erligang type do appear at Panlongcheng (in Huangpi, Hubei) (see Hubei 2001: 222, fig. 159.4), thought to be a Shang outpost in non-Shang territory (Bagley 1992). 17 In fact, only Pit no. 1 yielded any pottery at all (see Falkenhausen 2002 for further discussion).

Figure 13. Pottery vessels of pointed-bottom phiale shape (jiandizhan), from Pit no. 1 at Sanxingdui. From

Sichuan 1999: 146, fig. 76.1 and 3.

feet (he 盉) and high-stemmed bowls (dou 豆) (fig. 12). The excavators persuasively state that “these elements of resemblance must be the result of influence and contact between different groups” (Sichuan 1987: 250). Objects of these shapes have also been found at second-millennium BC sites in the Three Gorges Dam Area in Chongqing and western Hubei (Sun Hua 2000:153–156), though not, curiously, at sites of local Bronze Age cultures further downstream, such as Jingnansi (in Jiangling, Hubei) and Zaoshi (in Shimen, Hunan).16 Their arrival in Sichuan, perhaps by way of the Upper Han 漢 river valley in southern Shaanxi (Zhao Congcang 1996), may have coincided with that of the Erlitou lithics. Sun Hua (2000: 230–231) traces their local typological sequence into the time contemporaneous with the Shang dynasty—a time when the shapes represented in Sichuan had long fallen into disuse in their areas of origin. No Erlitou-derived pottery vessels are comprised among the very limited ceramic sample from the Sanxingdui “sacrificial pits.”17 The vast majority of potsherds found there belonged to pointed-bottom cups (jiandizhan 尖底盞), the commonest locally-made vessel type in Sichuan during the Bronze Age (fig. 13). Erlitou-related pottery types seem to be absent from later stages of the Sichuan Bronze Age culture sequence. It is unclear whether they were in use at the same occasions as the large-scale bronze statuary from Sanxingdui.


Erlitou Connections: Metallurgy. The casting technique of the various figural bronzes from Sanxingdui and Jinsha is identical in principle to that used elsewhere in Mainland East Asia throughout the Bronze Age. It involved the use of piece molds assembled around a core; the seams at the interstices between two molds are often visible on the objects. The technical achievement of the Sanxingdui casters consisted in making bronze objects of unprecedentedly large size. They did this by casting them in pieces that were assembled secondarily (Barnard 1990: 255–57), and Jay Xu (2001a: 60–65, 2001b) has offered a detailed study of the way in which this was done. Often, three-dimensional attachments were fitted (or possibly soldered) into preformed spaces. In spite of evidence for considerable technical versatility, Barnard’s (1990: 264) cautiously-expressed suggestion that the Sanxingdui casters may have used the lost-wax casting technique does not stand up in the light of scrutiny. Pace Barnard, moreover, the Sanxingdui bronzes display no technical particularities that would warrant a post-Shang date.18 To the contrary, recent finds suggest a surprisingly early origin of the Sichuan bronze-casting industry, perhaps as early as Erlitou times. Two oval-shaped bronze plaques with openwork decoration possibly once containing turquoise inlay (fig. 14.1) were excavated at Zhenwu Cangbaobao (Sichuan 1998; for another, fragmentary specimen, see Ao and Wang 1980); they probably predate the bronzes from the Sanxingdui “sac-rificial pits.” These objects closely resemble turquoise-inlaid bronze plaques excavated at the Erlitou type site (fig. 14.3) (Zhongguo 1984, 1986, 1992; Yang Meili 2002), and Jay Xu (2001a: 59–60) considers them as imports. But their shape, style, and technical execution are subtly distinctive, leading Yang Meili (2002: 31, 32) to conclude that they must be of local manufacture and possibly post-Erlitou in date. If indeed they were made in Sichuan, it would follow that the regional bronze industry that eventually produced the stunning statuary found in the Sanxingdui pits must have been established at least as early as the middle of the second millennium BC. This would make it earlier in date than currently known bronze-producing cultures in the middle and lower reaches of the Yangzi river basin. While surprising, this is not unthinkable. If this scenario can be substantiated, one might be justified in char-acterizing the Sanxingdui bronze-manufacturing industry as stylistically and technologically conservative vis-à-vis contemporaneous bronze-manufacturing traditions in other parts of China that were derived from the same roots.

18 None of the three criteria that Barnard (1990: 264) marshals in support of an Eastern Zhou date—stage-by-stage construction, large-area core-extension spacers, and extensive post-cast tooling—are at all convincing.


Metallurgy: Metropolitan Shang Connections. Even if, in the light of future research, bronze-casting in the Chengdu Plain does indeed turn out to go back to pre-Shang epochs, there is ample evidence for contact with bronze-casting traditions in areas further east during the time contemporaneous with the Shang. In the locally manufactured bronzes from the Sanxingdui sacrificial pits, connections to metropolitan Shang metalworking are intimated by similarities in ornament. Robert Thorp (1999: 209) observed that “conventional Shang motifs” are cited, for instance, on the plinth of the large standing statue from Pit no. 2 (fig. 1); executed in thread-relief, they were produced by a technique well-

Figure 14. Bronze plaques from Sichuan and Erlitou. 14.1: Bronze plaque (originally inlaid) from Zhenwu Cangbaobao, Guanghan (Sichuan). From Sichuan 1998: 81, fig. 3.2. 14.2: Tiger-shaped plaque. Stray find, allegedly excavated near Sanxingdui. From Sanseitai 1998, no. 149. 14.3: Turquoise-

inlaid bronze plaque from Erlitou. From Zhongguo 1981.38, fig. 5.1.


known from Erligang 二里岡 (Early/Middle Shang) period bronzes, though already pioneered on their Erlitou predecessors (e.g. Zhongguo 1999: pl. 169.2). The bold contour and filler lines of the elephant faces (thus identified by Jessica Rawson [in Rawson (ed.) 1996: 62; see also Xu 2001a: 75, n. 3]) on the basis of the standing statue, rendered in sunken-line, also have stylistic parallels on Erligang bronzes (fig. 15.2). The flat masks from Pit no. 2 (fig. 2.10), as well, are generally taotie-like, but Rawson (in Rawson [ed.] 1996: 69–70) is right in stressing differ-ences vis-à-vis the metropolitan Shang. Both Shang and Sanxingdui masks may ultimately be derived from Neolithic (e.g. Shijiahe) proto-types. Robert Thorp (1999: 214) has perspicaciously pointed out that “compared with the many permutations of the so-called taotie known from Shang tradition, this [sc. the face motif in Sanxingdui statuary] is a relatively stable imagery.” In keeping with their essentially conserva-tive character, the Sichuan bronze workshops (at least up to the time encapsulated by the Sanxingdui “sacrificial pits”) did not follow the stylistic tendencies of the Shang workshops at Anyang, which led at the latter to the separation of the taotie motifs from a distinctively decorated background surface (fig. 4).

Metallurgy: Connections to Shaanxi Workshops. An unknown number of flat, animal-shaped panels of bronze inlaid with turquoise were found in undocumented contexts at or near Sanxingdui (fig. 14.2) (Sanseitai 1998, no. 149; Bagley [ed.] 2001, no. 40). Technically, they remind of the turquoise-incrusted plaques from Erlitou and Zhenwu Cangbao-bao, and they may well, like the latter, predate the bronzes found in the “sacrificial pits” at Sanxingdui. The only place where objects of a vaguely similar character have been discovered to-date is tomb no. 41 at Laoniupo, Xi’an (Shaanxi), dating to the Late Shang period. The two fragmentary tiger-shaped plaques found there (Liu Shi’e [ed.] 2001: 297, 298 fig. 258.3,4; pl. 155.1,2) probably represent an idiosyncratic bronze-casting workshop operating in the lower Wei river basin during that period. They are smaller, thinner, and less well-executed than those from Sichuan, and it is unclear whether the angular depressions on the body of the more complete specimen could have accommodated inlay. As in the case of the small bronze figures from the Middle Western Zhou period tombs at Rujiazhuang discussed below, one wonders whether the inspiration for these objects might in fact have come from Sichuan. The nature of early cultural contacts across the Qinling mountains deserves further investigation.19

19 Yang Meili (2002: 29 and 51 figs. 7, 9) introduces a previously unpublished turquoise-inlaid plaque of Erlitou style collected in the 1980s in the Qincheng District of Tianshui (Gansu), as well as another, similar object of alleged Gansu provenance,


Shang Type Bronzes Imported from the Middle Yangzi Region. That Sanxingdui was, again in the words of Robert W. Bagley (1990: 52), “in close touch with other Shang centers” is also evidenced by the presence, in both of the Sanxingdui “sacrificial pits,” of bronze vessels of common Shang shapes (fig. 15.1); such vessels were made in North China since Erligang times (ca. 1500–1300 BC) (fig. 15.2). At Jinsha, the only possibly related object is a bronze fitting in the shape of a bovine head (Chengdu 2002, no. 9), perhaps originally an attachment to a bronze vessel of Shang type. The shape and ornamentation style of the Sanxingdui bronze vessels are closest to products from the Middle Yangzi region and the Han river valley (fig. 15.3), where local workshops were initially set up in Erligang times and flourished throughout the second half of the first millennium BC. That these “southern bronzes” form a manufacturing tradition different from that of the Shang and Western Zhou metro-politan centers was first demonstrated in a trailblazing article by the late Virginia C. Kane (1974/75), which has since been followed up by considerable additional scholarship (Hayashi 1980; Asahara 1984, 1985; Bagley 1987, 1992, 1999).20 While this manufacturing tradition owes its principal stylistic and technical inspiration to the metropolitan Shang bronze foundries during the Erligang period, local stylistic developments thereafter followed their own trajectory with but occasional reference to the Late Shang bronze art of Anyang.

now in a private collection. She points out a stylistic resemblance of their decoration to the headdress of the raised-arm figurines from the Western Zhou period tombs at Rujia-zhuang (see below). She elaborates (p. 41, n. 17): “From a geographical perspective, the Tianshui region may have been, in antiquity, a conduit of cultural contact between the Central Plains and northern Sichuan. Historians point to two main avenues between the Sichuan Basin and the Central Plains, the eastern route through the Three Gorges of the Yangzi River and the northern route through the Jialing River valley [...]. Thus, communication from southern Gansu to northern Sichuan proceeded first downstream along the Bailong River and then joined the main road along the Jialing River. Archaeo-logical surveys along the Bailong River have revealed numerous prehistoric sites of the Dadiwan, Majiayao, Qijia, and Siwa cultures; Qijia and Siwa sites, in particular, are distributed in areas close to northern Sichuan. Chronologically, these two archaeological cultures span the time between Erlitou and Western Zhou. [...] As to the cultural and ethnic affiliation of the [Rujiazhuang] materials, as well, the analysis by Lu Liancheng and Hu Zhisheng [1988] mainly points to close connections with Gansu and Qinghai, especially to the Qijia and Siwa cultures of southern Gansu.” This discussion indirectly raises the possibility of a common non-Erlitou derivation, perhaps in an as-yet un-explored area further to the west, of the turquoise-inlaid bronze plaques from Erlitou and Zhenwu Cangbaobao. Further explorations are urgently needed to ascertain the date of possible Erlitou parallels in the as-yet ill-documented local metallurgical traditions of the areas to the west of the Shang (on this issue, see also by Fitzgerald-Huber 1995 and Mei 2003). 20 The statement in Ge and Linduff (1990: 513) that “Although others have suggested a multi-centered development for early China, this material from Sichuan is the first excavated evidence from the Bronze Age to confirm that hypothesis” is untenable.


Figure 15. Bronze amphora-shaped vessels from the Shang period. 15.1: Zun vessel from Pit no. 2 at Sanxing-dui. From Sichuan 1999: 252, fig. 139. 15.2: Zun vessel from Zheng-zhou (Henan). Early/Middle Shang, ca. 1500–1300 BC. From Henan 2001, vol. 2: 817, fig. 549.1. 15.3: Zun ves-sel from Sucun, Chenggu (Shaanxi). Local culture contemporary with Shang. From Tang Jinyu et al. 1980, pl. 2.4. 15.4: Zun vessel from Lijia-tan, Wushan (Chongqing). From Yu Weichao 2000: 48. 15.5: Frag-mentary pou vesselfrom Pit no. 1 at Sanxingdui. From Sichuan 1999: 42,

fig. 25.1.


In spite of some debate over their place of manufacture,21 the style and typology of these vessels unambiguously bespeak a Middle Yangzi provenience.22 These vessels are among the latest objects among the Sanxingdui finds that are amenable to cross-dating on the basis of their external parallels, although their date of ca. 1400–1250 BC can only provide a terminus post quem for the “sacrificial pits.” Barnard (1990: 261–62), who believes that the local-style bronze figures date only to the time contemporary with the Springs and Autumns period, has insisted that the vessels were interred only after long history of usage; but the stratigraphic evidence at the excavation site would suggest that their use-life at Sanxingdui cannot have exceeded a century or so at the most. I strongly agree with Bagley’s assessment that the presence of these vessels indicates “far-reaching connections between the bronze-using cultures of the Chengdu Plain and those of regions to the east down the Yangzi River” (Bagley 1988: 82; similarly ibid. p. 84 and 1990: 65). But how did they reach this peripheral region? Recently, a zun 尊 vessel stylistically comparable to those from Sanxingdui has been found at Lijiatan, Wushan (Chongqing) 重慶巫山李家灘 (fig. 15.4) (Zhongguo 1997, no. 43; Yu Weichao 2000: 48), intimating the existence of a trade route through the Yangzi Gorges, in spite of the well-known difficulties of navigation. Another possible avenue of contact is the Han river valley (Zhao Congcang 1996; Wang Shouzhi 1988; Tang Jinyu et al. 1980; Li and Zhang 1996; Lu and Hu 1983, Li Boqian 1983; Bagley 1988: 84 and 86 n. 17, 1990: 66), where numerous bronzes and ceramics resembling those from Sanxingdui have been found, including close parallels to the pou [瓿] from Pit no. 1 (fig. 15.5). It would be a gross mistake to assume that the presence of such ritual vessels indicates the adoption of Shang ritual by the Sanxingdui élite. This seems, in fact, exceedingly unlikely given that the two pits yield only a small selection of the many types of vessels used in Shang ancestral sacrifices (Rawson in Rawson [ed.] 1996: 70 and passim; Goepper in

21 Goepper (in Goepper [ed.] 1995: 266) considers the issue undecidable; Rawson (in Rawson [ed.] 1996: 70–74) follows Chen and Chen (1987: 27) in arguing for local Sichuan manufacture (but this idea must be considered in connection with Rawson’s own earlier hypothesis that the Western Zhou-style container vessels found at Zhuwajie were made in Sichuan. I discuss the weaknesses of that argument in Falkenhausen 2003); Barnard (1990: 255 et passim), Bagley (1988: 80, 82 and 1990: 65), and Thorp (1999: 225)—correctly, in my opinion—contend that they were brought in from the Middle Yangzi region. 22 The technological analyses so far published (Zeng Zhongmao 1989; Jin Zhengyao et al. 1995), while not designed to address the provenience issue, at least are not in contradiction to such a conclusion.


23 This point is well made by Rawson (in Rawson [ed.] 1996: 70), although the final report shows that the typological array was wider than indicated in the preliminary reports alone (see Falkenhausen 2002, Table 1). I also overstated such preferences in Falkenhausen 2001, written before the final report on Sanxingdui became available. 24 Goepper (in Goepper [ed.] 1995: 261–62; similarly on p. 264) uncritically assumes this to have been the case in Sanxingdui as well, and he questionably asserts that the ritual system of Sanxingdui “must have been principally oriented toward the offering of wine or water.” Lin Xiang (1995: 113–117) propounds a similar point of view.

Goepper [ed.] 1995: 262); zun liquid-containers, in particular, seem to have enjoyed considerable popularity, while some of the most important kinds of vessels, such as ding 鼎 meat-offering tripods, are entirely lack-ing.23 The single-minded preference of the Sanxingdui consumers for amphora-shaped vessels—shared by later bronze users in the Sichuan basin and indeed in other areas that imported bronzes from the Shang and Zhou metropolitan regions (Li Xueqin 1991: 78–80; Sun Hua 2000: 258–261 and 232–234; Gao Dalun 2001; Falkenhausen 2001, 2003)—is aberrant by Shang standards: of the 22 bronze vessels found in the two pits, all but two (a pan 盤 basin and the already-mentioned pou) belong to this category. In the Shang culture sphere, such vessels would have served for the presentation of liquids (water or alcohol) (Hayashi 1964: 244–247);24 that their function was different in Sichuan is suggested by the fact that some of the specimens from Sanxingdui were filled with cowrie shells. They seem to have served to hoard or sacrificially pres-ent precious items, similar to amphora-shaped vessels in the Middle Yangzi region, some of which were found filled with jades (Rawson in Rawson [ed.] 1996: 70 and 72 n. 3; Thorp 1999: 225). Perhaps, these bronze vessels arrived in Sichuan filled with such treasure, and, like Greek amphoras in their time, they were valued in the first place for their contents. However, they must have also been considered valuable in and of themselves, and they were apparently displayed in ritual set-tings—although perhaps not in the same way as in their places of origin. For as Jay Xu has observed (Xu 2001a: 68, 140–149), holes were often drilled in their ring-feet, suggesting that they were affixed to some kind of support, possibly sculptural. They may thus have been a constituent feature of the “fabulous world” (to use Rawson’s expression [in Rawson (ed.) 1996: 18]) that was composed from the life-size bronze figures and heads, the “spirit-trees,” and the various other kinds of extravagant objects found in the “sacrificial pits” at Sanxingdui.

Shang Period External Connections: Jades. Wu Hung (1997: 66) discusses a jade carving of an eyeless human head that was part of a collection of exotic jades found in the “Tomb of Fu Hao” at Anyang, datable to slightly after 1300 BC (Zhongguo 1980, pl. 131.3) (fig. 16). Since it dif-fers stylistically from other jades made in the Shang core area, and its


Figure 16. Jade carving of a human head from “Fu Hao’s tomb.” From Zhongguo

1980, pl. 131.3.

eyelessness evokes an association with some of the bronze heads from Sanxingdui, Wu reasons that it may have been made in Sichuan. In fact, it bears closer resemblance to the stone statues from Jinsha (Chengdu 2002, nos. 51–54) than to the bronze statuary from the sacrificial pits at Sanxingdui. This object is a rare instance of a possible Sichuan export to the Yellow River Basin during this time, suggesting that the Chengdu Plain was not a completely passive recipient of outside culture transfer. In the local cultures of the Middle Yangzi Basin contemporaneous with the Shang, an item inviting comparison with Sanxingdui is a winged anthropomorphic figure with a bird-shaped headdress that has a loose-link chain attached to it—all painstakingly carved from a single piece of reddish jade (fig. 17). This unusual piece was excavated from the large tomb at Dayangzhou, Xin’gan (Jiangxi) 江西新干大洋洲, which probably dates to around 1200 BC (Jiangxi 1997: 159) (fig. 17). Aside from a certain stylistic resemblance (e.g. to the facial traits on the bound figures from Jinsha), the motif itself may derive from the Chengdu Plain, the only other place in China where depictions of winged animals

Figure 17. Red-jade figure of winged kneeling human from Dayangzhou, Xin’gan (Jiangxi). From Jiangxi

1997: 158, fig. 80.1.


(figs. 18.1, 2) and human-avian hybrids (fig. 18.3) are documented in this period; here they are apparently linked to the iconography of the enigmatic “spirit-trees” (Xu 2001b: 40–42). The occurrence of such motifs in the art of the later Chinese Bronze Age has been taken as an indication of contacts to areas in Central and Western Eurasia, where the depiction of angels, griffins, and other winged beasts has a long and distinguished history (Li Ling 2001); but whether such connections can also be elucidated in the time contemporaneous with the Shang still awaits further elucidation. (One would feel more comfortable suggesting this if there were any other indications of transasiatic contacts in early Sichuan, but this does not seem to be the case.)

Discussion. By now we have come to realize that the question, “Where does the Sanxingdui culture come from?” is wrongly phrased. Sanxing-dui was not a sudden flash of civilization irrupting into an uncharted wil-derness. Recent excavations in the Chengdu Plain have shown ever more clearly the existence of a continuous regional sequence of archaeological cultures, from Baodun down to the Qin conquest in 316 BC (Sun Hua 2000), within which the startling discoveries at Sanxingdui and Jinsha have their clear and logical place. But it is also clear that these regional cultures did not develop in isolation. The finds from Sanxingdui and Jinsha provide indications of significant cultural contact during each of the three main time periods in the early development of dynastic civili-zation as defined in North China: the Longshan period (Late Neolithic or Chalcolithic), the Early Bronze Age represented in central Henan by the Erlitou culture (i.e. the timespan Chinese historians assign to the Xia dynasty), and the florescent Bronze Age of the Shang civilization. In each period, the inhabitants of the Chengdu Plain interacted with both the Yellow River Basin and the Middle Yangzi region (including the Han River Basin). Rather than assuming that the unique bronze-producing culture of Sanxingdui “came from” one of these areas during any one of the aforementioned periods, it is probably accurate to say that it followed its own local trajectory of development within a context of interregional communication. In this process, the prehistoric cultures of Sichuan creatively absorbed, recombined, and transformed various elements of different origins. It seems to be particularly characteristic of Bronze Age Sichuan that cultural elements of outside derivation were often perpetuated here long beyond the time when they had been current in their areas of origin. In time, thus, Sichuan developed into a sort of repository for objects and cultural practices of early origin (Falkenhausen 2001, 2003). We wonder: What form did such cultural interaction take from phase to phase; how often or regularly did it occur; and by what mechanisms


Figure 18. Winged animals and avian-human hybrids at Sanxingdui. 18.1: Winged dragon (?) on socle of large “spirit tree” from Pit no. 2. From Sichuan 1999: 218 (insert), fig. 120. 18.2: Finial of uncertain function, topped with winged goat (?), from Pit no. 1. From Sichuan 1999: 34, fig. 20. 18.3: Small “spirit tree” topped by avian-human

hybrids from Pit no. 2. From Sichuan 1999: 226, fig. 125.


were local idiosyncrasies maintained? To what extent was it reciprocal? What kinds of evidence are we missing? Were the inhabitants of the Chengdu Plain acquainted with the intellectual background of the ritual objects they obtained from neighboring areas? And were the inhabitants of the latter aware, at any level at all, of the “fabulous world” that the Sanxingdui bronze casters created from the transmitted visual vocabu-lary? These questions are currently unanswerable, but they should be addressed in future archaeological fieldwork. Perhaps the greatest surprise emerging from the preceding consider-ations is the relative prominence of Erlitou elements in the cultural mix of the Sichuan Bronze Age. In the future, this phenomenon should be considered within a comprehensive assessment of the spread of Erlitou-type élite ritual inventory. Interpreting it in terms of a “Xia” political impact on the Sichuan region would probably be a distortion, but one might well try to contextualize it within an East Asia-wide analysis of the genesis of early royal dynasties. To understand such processes bet-ter, one hopes for archaeological evidence that might indicate whether Erlitou contact with the Chengdu Plain was direct or indirect. The notion of a stimulus from Shijiahe observed in Sanxingdui ico-nography also deserves further exploration in a wider context. Above all, one should consider it in connection with the transmission of southern élite culture elements from Shijiahe to the Erlitou area (Asahara 1984: 25). In a somewhat limited sense, one might conceive of Shijiahe as a fountainhead of bifurcating trajectories, one leading to central Henan, the other to the Chengdu Plain—two areas that remained in contact even after the demise of their common source of inspiration. If tenable, such a scenario might necessitate some rethinking of the privileged posi-tion conventionally granted to the Yellow River basin in accounting for the rise of the earliest Chinese states.

Part II: Later Parallels

When the Sanxingdui pits were first discovered, it seemed that their magnificent treasures were not only virtually without parallel elsewhere, but also isolated in the local cultural environment of Sichuan. This impression has since been modified considerably as the local cultural sequences of the Chengdu Plain have become better known. To begin with, the Jinsha finds now provide some continuity from Sanxingdui into a slightly later period. Even so, it still appears that the material culture of the Chengdu Plain and adjacent areas underwent consider-able change fairly soon after the time of Sanxingdui and Jinsha. The casting of monumental figural bronzes seems to have come to an end,


and the products of the local bronze foundries from the succeeding epoch, while still ornamented with considerable skill, betray a far lesser degree of technical sophistication than those from Sanxingdui. Closer observation nevertheless hints at some continuity with Sanxingdui and Jinsha in later regional cultures, both within the Chengdu plain and further afield.

Continuities in Imagery and Style. The only known bronze images com-parable to Sanxingdui statuary are two small figures with large hands wielding round objects (fig. 19) found in the tombs of a ruler of Yu A and his principal wife at Rujiazhuang, Baoji (Shaanxi) 陜西寶雞茹家莊, dating to Middle Western Zhou (ca. 950–850 BC) (Lu and Hu 1988: 315–16 and 375; Bagley 1988: 85–86 n. 10; Rawson in Rawson [ed.] 1996: 61–62). With their disproportionately enlarged hands, which apparently were meant to hold some object or objects of circular cross- section,25 they resemble the much larger standing bronze figure from Pit no. 2 at Sanxingdui (fig. 1). The two figures from Rujiazhuang differ

25 Perhaps this was something comparable to the “odd, rope-like object whose com-plete shape is unfortunately not known,” observed by Jay Xu (2001b: 38) on the smaller figures on the “spirit altars” from Pit no. 2 at Sanxingdui.

Figure 19. Figures from Tombs no. 1B and 2 at Rujiazhuang, Baoji (Shaanxi). Middle Western Zhou period (ca. 975–850 BC). From

Lu and Hu 1988, vol. 1: 315, fig. 221.2 and 375, fig. 257.


in dress and facial expression, seemingly constituting a male-female pair, the sex in each case corresponding to that of the re-spective tomb occupant. Perhaps significantly, the “male” figure, at 17.9 cm, is 50% higher than the “female” (11.6 cm), and the “female” figure is notable for its flamboyant headdress. Both figures were apparently intended to be mounted on wooden sticks, and the excavators cautiously suggest that they may relate to shamanistic sacrifices at the tombs. Since they occur in association with full sets of

Zhou ritual vessels, Jessica Rawson rightly cautions (in Rawson [ed.] 1996: 70) that their cultural meaning may have been quite different from their Sanxingdui pendants. But it is also clear that the Baoji area, which is located at the northern end of a major transit route across the Qinling 秦嶺 mountains, was in contact with the Chengdu Plain and the Han river valley, perhaps through the intermediary of the Han river valley (Lu and Hu 1983; Zhao Congcang 1996). At Rujiazhuang, such connections are also attested by the occurrence of ceramic speci-mens of the typically Sichuan pointed-bottom bowl (fig. 20), albeit of a different and probably later type from that seen in Pit no. 1 at Sanx-ingdui; Rujiazhuang has also yielded bronze specimens of such vessels, so far unparalleled in Sichuan. If the two bronze figures are imports from Sichuan, or were made under Sichuan stimulus, they would be the only extant Western Zhou period manifestation of the Sanxingdui sculptural tradition; they are certainly not sufficient evidence for dating Sanxingdui to Western Zhou. Even though both Sanxingdui and Jinsha have yielded a wealth of weapon-shaped objects of jade (cf. fig. 8.5) and bronze (fig. 21), none of these objects appear to have been intended for use as weapons. By contrast, various successive first-millennium BC bronze assemblages from Sichuan—Zhuwajie in Pengzhou 彭州竹瓦街 (10th century BC?) (Wang Jiayou 1961; Sichuan 1981; Feng Hanji 1980; Li Xueqin 1996; Falkenhausen 2001, 2003), Moutuo in Mao Xian 茂縣牟托 (mid- to late 5th century BC) (Mao Xian 1994; Falkenhausen 1996; Shi Jingsong 1996), and those of the Qingyanggong 青羊宮 (or “Ba-Shu” 巴蜀) cul-ture (4th–3rd centuries BC) (Pirazzoli-t’Serstevens 2001: 39–48; SunHua 2000)—have yielded considerable quantities of weapons ornamented

Figure 20. Pointed ceramic bowl from tomb no. 1B at Rujiazhuang. From Lu and Hu 1988, vol. 1: 359, fig. 244.8.


in a highly distinctive regional style. They often render motifs known from Shang and Zhou bronzes, which are, however, interpreted quite differently, just as in the above-mentioned Sanxingdui local adaptations of Shang bronze art (as seen, e.g. on the pedestal of the standing figure in fig. 1). A fairly orthodox example of Shang-style taotie on a much later object of Sichuan manufacture is on the dagger-axe in fig. 22.1. Residual allusions to Sanxingdui bronze art may be observed in several axes (fu 斧 or yue) found at Zhuwajie (fig. 22.2), which have no counterparts in north China, although similar axes with simpler decora-tion are present elsewhere in Sichuan (Yue Runlie 1983: fig. 1.1, 2) and in the upper Han river valley (Shaanxi 1979, no. 108; Tang Jinyu et al. 1980: 216, 213 fig. 3.5, pl. 4.5; Li Xixing [ed.] 1994, no. 246; Zhao Congcang 1996: 10, 16 fig. 7.5; Li Boqian 1983: 68). A stone object of the same shape has turned up at Jinsha (Chengdu 2002, no. 58 ) (fig. 22.3); its lack of a blade may bespeak a habit, also observable at Sanx-ingdui, of fashioning objects of purely symbolic significance in imitation of weapons. The décor of the item in fig. 22.2 is ambiguous—it could be a bird with outstretched wings seen from above, or a deformed, doubled-up version of the Shang taotie face motif; the sunken-line diamonds on the central axis remind of those seen on bovine faces in Western Zhou bronzes (though it is unusual that there are two such diamonds); yet the framed, rounded-rectangular “eyes” placed on the inward-bent “horns”(?) distantly recall Sanxingdui forerunners. Li Xueqin (1996: 121) points to Sanxingdui masks (e.g. the one depicted in Zhongguo 1994, no. 36) or bells (ibid., no. 59) as sources of inspiration for the decor of these axes. An even closer resemblance to Sanxingdui anthropomorphic masks is manifest on the two-piece halberd (ji 戟) in fig. 22.4. The decora-tion appearing on each of the component blades features an amalgam

Figure 21. Bronze dagger-axe with serrated edge from Pit no. 2 at Sanxingdui. From Sichuan 1999: 291, fig. 160.3.


Figure 22. Weapons from Zhuwajie and related finds. 22.1. Dagger-axe (ge) from Zhuwajie, Pengzhou (Sichuan). Early Western Zhou (ca. 1050–975 BC) or later. From Sichuan 1981: 499, fig. 6.3. 22.2: Axe (yue or fu) from Zhuwajie. From Sichuan 1981: 499, fig. 6.1. 22.3: Stone Axe-shaped object from Jinsha. From Chengdu 2002, no. 58. 22.4: Halberd (qi) from Zhuwajie. From Feng Hanji 1980: 40, fig. 5.1. 22.5: Spearhead (mao) from Zhuwajie. From Feng Hanji 1980: 40, fig. 5.2. 22.6: Dagger-axe from Jiaotongxiang, Chengdu (Sichuan). Early phase of Qingyang-gong culture (mid-first millennium BC). From Zhongguo 1994: 7, fig. 4.


of human and avian traits. A long curved wing extends down the midrib. Next to it, near the base of the blade, a face is shown in profile, its curved nose resembling a bird’s beak and its comma-shaped beard almost like an-other, smaller wing. The presence of pointed ears and a mouth underneath the “beak,” however, leaves no doubt that the face is anthropomorphic. The motif harks back to the human-avian hybrids on the Sanxingdui “spirit trees” (fig. 18.3). With due caution, one may wonder whether it might be an icono-graphic precursor to the yuren 羽人 (feather-men) immortals in Han popu-lar religious iconography. Even though the weapon type itself is derived from the Yellow River Basin, the decoration of this halberd constitutes a rare indica-tor of some continuity between Sanxing-dui and later bronze-casting in Sichuan.The framed, rounded-rectangular eyes and wavy mouth of the face evoke specific stylistic associations with the large human masks from Sanxingdui (e.g. fig. 2.2). Other noteworthy features include the attractively asymmetrical deployment of the wings on the weapon surface and a quite daring dichotomy in the execution of the face, which is shown in relief on the lateral (dagger-axe) blade, while on the vertical (spear-head) blade it emerges in three dimen-sions and forms part of the object’s contour. The difference between such

objects (see also figs. 22.5 and 22.6) and bronzes from Shang and Zhou workshops could hardly be greater; the stylistic idiom is closer to that of later Southeast Asian bronze weapons (Zhou Wei 1993: 55–112). While the vast majority of Bronze Age weapons found in Sichuan can be traced to earlier prototypes from the Shang and Zhou culture sphere, the jade short-sword (or dagger; jian 劍) blade found in Pit no. 1 at Sanxingdui (fig. 23.1) is earlier in date than any bronze sword

Figure 23. Jade sword from San-xingdui and related object. 23.1: Jade sword or dagger blade (?)

from Pit no. 1 at Sanxingdui. From Sichuan 1999: 97, fig. 50. 23.2: Bronze sword from Moutuo. From

Mao Xian 1994: 36, fig. 55.4.


found elsewhere in East Asia. The earliest known bronze specimens in China, some of them in elaborate sheaths, appear in Western Zhou tombs from the northern and western fringes of the Zhou realm (in-cluding Rujiazhuang) (Yang Hong 1985: 116; Lu and Hu 1988), and in Sichuan, several early specimens have been found (see Jiang Zhanghua 1992). They continue with little change into the mid-first millennium BC; several ornate specimens have been found, for instance, at Moutuo (fig. 23.2), in an ensemble remarkable for its extraordinary diversity. Since the jade sword from Sanxingdui is concave on the back (Jay Xu, personal communication, 2001), it may not have been made in imitation of a metal prototype but of an object made of bamboo or bone. One cannot exclude the possibility that short swords of this type originate in Sichuan, but such a far-reaching argument should not be based on a single, isolated, and still poorly understood specimen.26

Ritual Continuities. Although vessels are not particularly numerous in first-millennium BC bronze assemblages from Sichuan, both imported and locally made specimens have been found. The predilection for amphora-shaped vessels, so noticeable at Sanxingdui, continues in those later assemblages. This may bespeak some degree of continuity in ritual priorities (see Falkenhausen 2001, 2003). Even more interestingly, some of the vessel shapes current in Sichuan during the time of Sanxingdui are taken up as well in the much later and geographically more remote bronze-manufacturing industry of the “Dian 滇 culture” (more appro-priately called Shizhaishan 石寨山 culture) in Yunnan and adjacent areas, which began sometime before the middle of the first millennium BC and appears to have been snuffed out by the Han conquest in 109 BC (Yunnan 1959; Pirazzoli-t’Serstevens 1974; Rawson [ed.] 1983). Its casting techniques are essentially identical to those of the Shang and Zhou, and its ultimate origins must probably be sought in the Yellow and Yangzi River basins; that Sichuan played an intermediary rôle in the transmission is suggested by similarities of Dian weapons (fig. 24) to Late Bronze Age weapons from the Chengdu Plain (Tong Enzheng 1977, 1979). Noel Barnard (1990: 259 n. 12) has observed a specific technical parallel in the “bending or hammering of casting fins” that is seen both at Sanxingdui and in certain Dian bronzes. But instead of following Barnard in considering this an indicator of a late date for Sanxingdui, I would rather explain it as the consequence of an early link between the two areas. In other words, Sanxingdui should be considered as one potential fountainhead for the transmission of ultimately Shang-derived

26 When I considered the absence of swords from Zhuwajie as a possible indicator of their relatively early age (Falkenhausen 2001: 187), I was still unaware of the Sanxingdui jade specimen.


bronze-casting techniques into the Dian culture area. Among the highly ornate vessels found at cemeteries from the late phase of the Dian culture, we find a kind of zun with trumpet-shaped neck and narrow central por-tion (fig. 25.1). Its derivation from objects such as the zun from Sanxingdui (fig. 15.1) can hardly be doubted— a remarkable instance of ar-chaic survival, considering that no zun were made in North China after the aboli-tion of wine vessels from ritual assemblages in ca. 850 BC (cf. Rawson 1990, vol. 1: 100–116 et passim; Luo Tai 1997), and no specimens are known from Sichuan after the time of Zhuwajie (afterwards, lei 罍 , hu 壺 , and fang 鈁 became the preferred amphora-shape

vessel types).27 Another probably relevant Dian parallel to Sanxingdui is the custom of filling bronze vessels with cowries (Goepper in Goepper [ed.] 1995: 264; Rawson in Rawson [ed.] 1996: 72 n. 3). Looking at the well-known Dian cowrie containers (fig. 25.2; fig. 27.2) one notices that their profile resembles that of (upturned) zun, and it seems quite plausible that they are ultimately zun derivates—as may be, incidentally, their close relatives the Southeast Asian bronze drums (fig. 25.3). The presence of a great deal of sculptural ornament on the Dian cowrie containers reminds of the Sanxingdui “Spirit Altars” (fig. 25.4) whose affinity to zun vessels has been convincingly demonstrated by Sun Hua (2000: 261 and this volume). Enhancing the impression of ritual continuity, Dian cemeteries have also yielded specimens of the same kinds of ritual jades seen at Sanxingdui,

Figure 24. Dagger-axes from Shizhaishan, Jin-ning (Yunnan). Dian culture (second half of first millennium BC). From Yunnan 1959: 30/31,

fig. 7.1, 2, 3.

27 Hu of Han type have also been found at Dian cemeteries (e.g. at Shizhaishan, see Yunnan 1959: 68, fig. 20.6). While doubtless functionally identical to the zun, they are not typologically related to them.


Figure 25. The Dian transformation of early amphora-shaped vessels, and related objects. 25.1. Zun vessel from Shizhaishan. From Yunnan 1959: 68, fig. 20.2. 25.2. Cowrie container from Shizhaishan. From Lutz (ed.) 1986: 81. 25.3: Bronze drum from Wanjiaba, Chuxiong (Yunnan). Early phase of Dian culture. From Li and Huang 1991: 387, fig. 4. 25.4: “Spirit altar” from Pit no. 2 at Sanxingdui. From Sichuan 1999: 233, fig. 129. 25.5: Plaque ornament from tomb no. 1 at Moutuo. After Mao

Xian 1994: 30, fig. 49.4.


derived from the Bronze Age of the Central Plains and ultimately from the Longshanoid cultures (fig. 26). Since known specimens are incom-pletely and poorly published, it is difficult to judge at present whether these were heirlooms from very early times or local imitations. In spite of the chronological and stylistic gap vis-à-vis Sanxingdui and Jinsha and of the great physical distance from the Chengdu Plain, Dian’s specific parallels vis-à-vis Sanxingdui cannot be ignored. That there was a connection seems likely also in light of other evidence to be discussed momentarily. But if so, it cannot have been a direct one; a missing link (or chain of links) between the two in all probability still awaits discovery.

Continuities in Habits of Representation. Rawson (in Rawson [ed.] 1996: 18 et passim) has pointed out that multifigure representational art of the sort seen at Sanxingdui is alien to the Shang-Zhou tradition, even though it is a mainstay of the art of most other early civilizations (Egypt, Mesopota-mia, Mesoamerica, Peru etc.). The geographically closest occurrence to Sanxingdui of such habits of representation are the sculptural ensembles placed on Dian cowrie-containers (figs. 25.2 and 27.2) as well as on other products of Dian metallurgical craftsmanship (see Yunnan 1959; Rawson [ed.] 1983); as in the Sanxingdui “spirit shrines” and “spirit-altars” (fig.

Figure 26. Collared bi excavated at Shizhaishan.From Lutz (ed.) 1986: 105 (one of two).


25.4) the preferred subjects are ritual scenes.Once again, the parallel between Dian and Sanxingdui is first and foremost a conceptual one rather than one of style. As a possible connecting link, one might adduce a stylistically isolated “plaque ornament” from tomb no. 1 at Moutuo (fig. 25.5), which seems to perpetuate the openwork technique of San-xingdui while at the same time intimating the more three-dimensional rendering of figures characteristic of Dian (Falkenhausen 1996: 44). Where this intriguing object was made is still a mystery. An additional Sanxingdui-Dian commonality at the level of habits of representation is the use of gold-foil wrapping to enhance a bronze figure or its face (figs. 27.1 and 27.2) (Goepper in Goepper [ed.] 1995: 258). The technique seems to be the same in both places, and it evokes mental association to a much older tradition of anthropomorphic statues wholly or partially wrapped in gold in the Mediterranean world (Miho 1997, no. 5; Negbi 1976, pls. 18.1307, 33.1463). It would be pure folly at the present state of research to start speculating about a possible connection; all we can say is that in China, gold does not seem to have been used to similar effect anywhere else during pre-Imperial times.

Figure 27. Gold-foil wrapping at Sanxingdui and Dian. 27.1: Bronze head with gold-wrapped face from Pit no. 2 at Sanxingdui. From Sichuan 1999: 183, fig. 98. 27.2: Cowrie container with gold-wrapped rider from Shizhaishan.

From Lutz (ed.) 1986: 57.


Ideological Continuity. Perhaps the most evocative among all the extraor-dinary bronzes from the two “sacrificial pits” at Sanxingdui are the so-called “spirit trees” (fig. 18.1, 18.3). Nothing is known about their original meaning. It has been proposed that they are somehow linked to the money-shaking trees (yaoqianshu 搖錢樹) (Erickson 1994) that were a local specialty of the Chengdu Plain and adjacent areas during the second and third centuries AD (Barnard 1990: 262–63; Goepper in Goepper [ed.] 1995: 271; Rawson in Rawson [ed.] 1996: 18). The idea that the popular religious imagination (sometimes misleadingly called Taoist) of abundance and immortality reflected in these much later objects might have their origin in a specific local mythological and iconographic tradition and go back to a very early time is an attractive one. But so far, archaeological activity has brought to light only one single potential piece of evidence to bridge the time gap of one mil-lennium and a half: a bronze bird (fig. 29) from tomb no. 1 at Moutuo (Falkenhausen 1996: 43). It superficially resembles specimens from both Sanxingdui (fig. 18.1; fig. 2.12) and Jinsha (Chengdu 2002: no. 8), and it may once have been perched on a “spirit tree,” though the mode of attachment is now unclear. It may be a heirloom from Sanxingdui times, but it is also possible that it dates to a later time. In any event, such a single item does not suffice as a proof of the continuity between Sanxingdui “spirit trees” and Eastern Han money-shaking trees. Until more supporting material is discovered, that relationship must remain a matter of speculation. In the ethnographic present, Sanxingdui elements are alleged to have survived in the sacrificial customs of the Yi nationality now inhabiting the hilly regions of southwestern Sichuan (Goepper [ed.] 1995: 249–50, quoting Qian Yuzhi 1992). Barnard (1990: 267 and pl. 91.11) depicts a modern wooden mask with protruding eyeballs (fig. 29), which ap-parently comes from non-Han populations near Chengdu. Rawson (in Rawson [ed.] 1996: 63, n. 2) rightly considers it “unwise to draw any firm conclusion from such tenuous evidence,” yet the possibility of very longterm local transmission of religious ideas and iconography should not be entirely discounted. It would indeed be disingenuous to posit links over similarly long timespans at the early end of the chronologi-cal spectrum while disallowing them at the late end. Feng Shi (1994) has interpreted the enigmatic pottery inscription from the Neolithic Dawenkou culture in Shandong (fig. 30) as an early form of the present-day Yi script, whose origins had heretofore been unexplained. Feng hypothesizes that literate members of the aristocracy of coastal polities moved to Sichuan sometime during late prehistoric times. If true, this theory might provide some context for Sanxingdui’s long-distance cul-tural connections.


Discussion. One might liken the Chengdu Plain in the time of Sanxingdui and Jinsha to a cultural laboratory where some of the same elements that had defined the greatness of the cultures of Erlitou, Erligang, Anyang, and Western Zhou in the Yellow River Basin, as well as elements from other regional cultures in the surrounding areas, were combined in a different mix, with very distinctive results (cf. Thorp 1999: 214). The extent to which this distinctive culture radiated outward is so far unclear, as is the extent of its continuing impact on Sichuan itself. It seems sig-nificant, however, that all the early comparisons offered in PartI—those through which essential features of the Sanxingdui phenomena were in all probability derived—are with the Yellow and the Middle and Lower Yangzi River Basins, while cultural elements to which Sanxing-dui may conceivably have been ancestral, discussed in Part II, for the

Figure 28. Bronze bird from Moutuo. From Mao Xian 1994: 24, fig. 36.1

Figure 29. Wooden mask collected at Chengdu. From Barnard 1990, pl. 91.11.


most part come from southwestern China. By contrast, Sanxingdui elements cannot be found in areas east anytime after Middle Western Zhou.28 At the risk of oversimplification, one may thus construe a trajectory in which cultural stimuli were transferred to the Sanxingdui culture area from the Yellow and Yangzi river basins, transformed locally, and then transmitted on to areas further afield. Here they may have survived in marginal zones even after the indigenous high culture in the Chengdu Plain—represented archaeologically by Baodun and Sanxingdui at the early end and Qingyanggong at the late end—had been absorbed into the civilization of Imperial China. Particularly important is the appar-ent survival of Sanxingdui elements in the Dian culture; the nature of the relationship and the intervening stages of development in between the two, remain a subject for sustained further research. Someone might now ask: Was Sanxingdui Chinese? Ultimately, the answer depends on how one defines one’s terms. To colleagues in China, this is a non-issue. The Chengdu Plain is geographically and culturally part of China, and the local cultural sequences in this part of the country are demonstrably continuous with, and at least in part ancestral to, cultural developments that followed the area’s incorpora-tion into the Qin and Han empires (cf. Su Bingqi 1994). Although the old paradigm of uniform cultural development following the unilinear dynastic sequence mentioned in the traditional histories has now been replaced, in China, by a consensus on the regional character of cultural

Figure 30. Inscribed (?) potsherd from Dinggong, Zouping (Shandong). Dawenkou culture (5th–4th millennia BC). From Feng Shi 1994: 38, fig. 1.

28 The connections between Sanxingdui and Eastern Zhou period Chu 楚 posited by Barnard (1990: 268–69 et passim) are all highly unconvincing.


developments during the Neolithic and the early Bronze Age, these separate developments are held to form part and parcel of a greater entity—Chinese civilization. Jessica Rawson problematizes this when she writes in connection with Sanxingdui (in Rawson [ed.] 1996: 74): “The possessions and thus the roles of the élite in different areas of the landmass of China were quite distinct. Such distinctions imply not just different practices but also different beliefs. And different beliefs point in turn to very different societies.” Strongly insistent on their non-Shang identity, Rawson disbelieves (ibid., p. 75) “that the peoples in Sichuan at that time [sc. in the Shang period] shared a language and a set of beliefs with the peoples of Henan. It would indeed be dangerous to assume that they did.” Were Shang material culture in central Henan to be taken as a fixed standard of “Chineseness” in that period, it would indeed be difficult to accommodate Sanxingdui. But even though the inhabitants of the Chengdu Plain probably differed in language, customs, appearance, and socio-political affiliation from their contemporaries in the Shang and Zhou heartland, the archaeological connections we have traced show that, since Late Neolithic times, Sichuan was linked to an “inter-action sphere” (Chang 1986: 234–94) that also comprised other early archaeological cultures regarded as ancestral, ultimately, to the cultural mainstream of Imperial China. In such a sense, there are legitimate archaeological reasons for a definition of “Chinese” in the Bronze Age that would include Sanxingdui.

References Cited

Ao Tianzhao 敖天照 and Wang Youpeng 王有鵬 (1980). “Sichuan Guanghan chutu Shang dai yuqi 四川廣漢出土商代玉器 (Shang jades excavated at Guanghan, Sichuan).” Wenwu 1980.9: 76.

Asahara Tatsurō 淺原達郎 (1984). “Ka bunka tansakuno michi 夏文化探索の道 (The Xia culture: Avenues of investigation).” Koshi shunjū 古史春秋 1: 18–28.

Asahara Tatsurō (1985). “Shoku hei tangen—Nirikō inpakuto to Shū, Shoku, So 蜀兵探源: 二里岡インパクトと周, 蜀, 楚 (Classification and derivation of the weapons of the Shu Culture: The Erligang impact on Zhou, Shu, and Chu).” Koshi shunjū 2: 23–52.

Bagley, Robert W. (1987). Shang Ritual Bronzes in the Arthur M. Sackler Col-lections. Cambridge, Mass.: Harvard University Press.

Bagley, Robert W. (1988). “Sacrificial Pits of the Shang Period at San-xingdui in Guanghan County, Sichuan Province.” Arts Asiatiques 43: 78–86.


Bagley, Robert W. (1990). “A Shang City in Sichuan Province.” Orienta-tions 21.11: 52–67.

Bagley, Robert W. (1992). “Changjiang Bronzes and Shang Archae-ology.” In Proceedings, International Colloquium on Chinese Art History, 1991: Antiquities, Part 1, pp. 209–254. Taipei: National Palace Museum.

Bagley, Robert W. (1999). “Shang Archaeology.” In The Cambridge History of Ancient China, Michael Loewe and Edward L. Shaughnessy (eds.), pp. 124–231. Cambridge: Cambridge University Press.

Bagley, Robert W. (ed.) (2001). Ancient Sichuan: Treasures from a Lost Civilization. Seattle: Seattle Art Museum and Princeton University Press.

Barnard, Noel (1990). “Some Preliminary Thoughts on the Significance of the Kuang-han Pit-burial Bronzes and Other Artifacts.” Beiträge zur Allgemeinen und Vergleichenden Archäologie 9–10: 249–279.

Chang, Kwang-chih (1986). The Archaeology of Ancient China, 4th edition. New Haven: Yale University Press.

Changjiang Shuili Weiyuanhui 長江水利委員會 (2002). Yichang Lujiahe: Chang jiang Sanxia kaogu fajue baogao 宜昌路家河: 長江三峽考古發掘報告 (The site of Lujiahe, Yichang: Excavation Report in the Three Gorges of the Yangzi River). Beijing: Kexue Chubanshe.

Chen Xiandan 陳顯丹 and Chen De’an 陳德安 (1987). “Shixi San-xingdui yizhi Shang dai yihaokeng de xingzhi ji youguan wenti 識析三星堆遺址商代一號坑的性制及有關問題 (Exploratory analysis of the form of Shang period Pit no. 1 at the site of Sanxingdui, and related issues).” Sichuan wenwu 1987.4: 27–29.

Cheng Te-k’un (1949). “The T’ai-p’ing-ch’ang Culture, Szechwan.” Hsieh-ta Journal of Chinese Studies 1: 57–81. Also in Cheng, Studies in Chinese Archaeology, pp.55–65. Hong Kong: Chinese University Press, 1982.

Chengdu 1997. Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊, Sichuan Lianhe Daxue Kaogu Jiaoyanshi 四川聯合大學考古教研室, and Xinjin Xian Wenguansuo 新津縣文管所 (1997). “Sichuan Xinjin Xian Baodun yizhi diaocha yu shijue 四川新津縣寶墩遺址調查與識掘 (Survey and trial excavations at the site of Baodun at Xinjin, Sichuan).” Kaogu 1997.1: 40–52.

Chengdu 2000. Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所, Sichuan Daxue Lishixi Kaogu Jiaoyanshi 四川大學考古教研室, and Zaodaotian [Waseda] Daxue Changjiang Liuyu Wenhua Yanjiusuo 早稻田大學長江流域文化研究所 (2000). Bao-dun Yizhi: Xinjin Baodun yizhi fajue he yanjiu 寶墩遺址—–新津寶墩遺址發掘和研究 (The Site of Baodun: Excavation and Research on the site of Baodun, Xinjin). Chengdu: Chengdu Shi Wenwu Kaogu Yanjiusuo.


Chengdu 2002. Chengdu Shi Wenwu Kaogu Yanjiusuo and Beijing Daxue Kaogu Wenboyuan 北京大學考古文博院 (2002). Jinsha tao-zhen: Chengdu shi Jinshacun yizhi chutu wenwu 金沙淘珍—–成都市金沙村遺址出土文物 (Panning for treasure at Jinsha: Cultural relics excavated at the Jinshacun site in the city of Chengdu). Beijing: Wenwu Chubanshe.

Chongqing 2001. Chongqing Shi Wenwuju 重慶市文物局 and Chong-qing Shi Yiminju 重慶市移民局 (ed.) (2001). Chongqing kuqu kaogu baogaoji, 1997-juan 重慶庫區考古報告集, 1997 卷, Changjiang San-xia Gongcheng Wenwu Baohu Xiangmu baogao 長江三峽工成文物保護項目報告, Series A, vol. 1. Beijing: Kexue Chubanshe.

Deng Cong 鄧聰 (= Tang Chung) (ed.) (1994). Nanzhongguo ji linjin diqu gu wenhua yanjiu: Qingzhu Zheng Dekun jiaoshou congshi xueshu huodong liushizhounian lunwenji 南中國及臨近地區文化研究—–慶祝鄭德坤教授從事學術活動六十周年 (Ancient Cultures of South China and Neighboring Regions: Essays in Honor of Professor Cheng Te-K’un on the Occasion of the Sixtieth Anniversary of His Academic Career). Xianggang Zhongwen Daxue Zhongguo Wenhua Yanjiu-suo, Zhongguo Kaogu Yishu Yanjiu Zhongxin zhuankan 香港中文大學中國文化研究所中國考古藝術研究中心專刊, vol. 8. Hong Kong: Zhongwen Daxue Chubanshe.

Deng Cong (1998). Dongya yuqi 東亞玉器/East Asian Jade: Symbol of Excellence. 3 vols. Hong Kong: Centre for Chinese Archaeology and Art, Chinese University of Hong Kong.

Deng Shuping 鄧叔蘋 (= Teng Shu-p’ing) (1985). Zhonghua wuqian-nian wenwu jikan, Yuqi bian 1 中華五千年文物集刊玉器編 (一 ) (A collection of cultural relics from China’s 5000-year long history: Jades, vol. 1).Taibei: Zhongguo Wuqiannian Wenwu Jikan Bianjiwei-yuanhui.

Deng Shuping (1998). “Zailun shenzumianwen yuqi 再論神祖面文玉器 (Further remarks on jades with divine-face decoration).” In Deng Cong (ed.) 1998: 45–60.

Dohrenwend, Doris (1971). Chinese Jades in the Royal Ontario Museum. Toronto: Royal Ontario Museum.

Erickson, Susan N. (1994). “Money-trees of the Eastern Han Dynasty.” Bulletin of the Museum of Far Eastern Antiquities 66: 5–155.

Falkenhausen, Lothar von (=Luo Tai) (1996). “The Moutuo Bronzes: New Perspectives on the Late Bronze Age in Sichuan.” Arts Asiatiques 51: 29–59.

Falkenhausen, Lothar von (2001). “The Chengdu Plain in the Early First Millennium B.C.: Zhuwajie.” In Bagley (ed.) 2001, pp. 177–201.

Falkenhausen, Lothar von (2002). “Some Reflections on Sanxingdui.” In Papers from the Third International Conference on Sinology, History Section:


Regional Culture, Religion, and Arts Before the Seventh Century, pp. 59–97. Taipei: Institute of History and Philology, Academia Sinica.

Falkenhausen, Lothar von (2003). “Les bronzes de Zhuwajie: une énigme archéologique.” In Thote (ed.) 2003:195–203.

Feng Hanji 馮漢驥 (1980). “Sichuan Peng Xian chutu de tongqi 四川彭縣出土的銅器 (The bronzes excavated at Peng Xian, Sichuan).” Wenwu 1980.12: 38–47.

Feng Shi 馮時 (1994).“Shandong Dinggong Longshan shidai wenzi jie-du 山東丁公龍山時代文字解讀 (Decipherment of Longshan period writing from Dinggong, Shandong).” Kaogu 1994.1: 37–54.

Fitzgerald-Huber, Louisa G. (1995). “Qijia and Erlitou: The Question of Contact with Distant Cultures.” Early China 20: 17–67.

Fong, Wen (ed.) (1980). The Great Bronze Age of China. New York: Metro-politan Museum of Art.

Gao Dalun (2001). “Bronze Ritual Artefacts of the Shu Culture: A Pre-liminary Survey.” Orientations 32.5: 45–51.

Gao Wei 高偉 (1998).“Taosi wenhua yuqi ji xiangguan wenti 陶寺文化玉器及相關問題 (Jades of the Taosi culture and related problems).” In Deng Cong (ed.) 1998: 192–200.

Ge Yan and Katheryn M. Linduff (1990). “Sanxingdui: A new Bronze Age site in southwest China.” Antiquity 64: 505–513.

Goepper, Roger (ed.) (1995). Das alte China: Menschen und Götter im Reich der Mitte. Essen: Kulturstiftung Ruhr.

Guojia Wenwuju Sanxia Kaogudui 國家文物局三峽考古隊 (2001). Chaotianzui yu Zhongbaodao 朝天嘴與中堡島 (The sites of Chaotian-zui and Zhongbaodao). Beijing: Wenwu Chubanshe.

Hayashi Minao 林巳奈夫 (1964). “In-Shū seidō iki-no meishō to yōto 殷周青銅彝器の名稱と用途 (Nomenclature and usage of Shang and Zhou ritual bronze vessels).” Toho gakuho 34: 199–298.

Hayashi Minao (1980). “In Seishū jidai-no chihōkei seidōki 殷、西周時代の地方型青銅器 (Regional bronzes of Shang and Western Zhou times).” Kōkogaku memowāru 考古学メモワール 1980: 17–58.

Hayashi Minao (1991). Chūgoku kogyoku no kenkyū 中國古玉の研究 (Studies on Chinese ancient jades). Tōkyō: Yoshikawa kōbunkan.

Hayashi Minao (1998). “Guanyu Shijiahe wenhua de yuqi 關於石家河文化的玉器 (On the jades from the Shijiahe culture).” In Deng Cong (ed.) 1998, vol. 1: 287–297.

Henan 2001. Henan Sheng Wenwu Kaogu Yanjiusuo 河南省文物考古研究所 (2001). Zhengzhou Shang cheng 鄭州商城 (The Shang walled settlement at Zhengzhou), 3 vols. Beijing: Wenwu Chubanshe.

Hubei 1999a. Hubei Sheng Jingzhou Bowuguan 湖北省荊州博勿館, Hubei Sheng Wenwu Kaogu Yanjiusuo Shijiahe Kaogudui 湖北省文物考古研究所石家河考古隊, and Beijing Daxue Kaoguxue Xi


北京大學考古學系 (1999a). Xiaojiawuji (The site of Xiaojiawuji).Tianmen Shijiahe kaogu fajue baogao 天門石家河考古發掘報告, no. 1, 2 vols. Beijing: Wenwu Chubanshe.

Hubei 1999b. Hubei Sheng Jingzhou Bowuguan (1999b). Zaolingang yu Duijintai: Jing jiang dati Jingzhou Mashan duan kaogu fajue baogao 棗林崗與堆金臺—荊江大堤荊州馬山段考古發掘報告 (Zaolingang and Duijintai: Report on archaeological excavations along the Jiangling Mashan section of the Great Jingjiang Dyke). Beijing: Kexue Chu-banshe.

Hubei 2001. Hubei Sheng Wenwu Kaogu Yanjiusuo 湖北省文物考古研究所 (2001). Panlongcheng: Yijiuliusannian–yijiujiusinian kaogu fajue bao-gao 盤龍城: 一九六三年–一九九四年考古發掘報告 (Panlongcheng: Report on archaeological excavations from 1963 to 1994), 2 vols. Beijing: Wenwu Chubanshe.

Jiang Zhanghua 江章華 (1992).“Ba Shu liuyexingjian yuanyuan shitan 巴蜀柳葉形劍淵源試探 (Exploratory discussion of the derivation of Ba Shu willow-leaf shaped sword blades).” Sichuan wenwu: Sanxingdui gu Shu wenhua yanjiu zhuanji 四川文物: 三星堆古蜀研究專輯 (1992), pp. 81–84. Chengdu.

Jiangxi 1997. Jiangxi Sheng Bowuguan 江西省博物館, Jiangxi Sheng Wenwu Kaogu Yanjiusuo 江西省文物考古研究所, and Xin’gan Xian Bowuguan 新干縣博物館 (1997). Xin’gan Shang dai damu 新干商代大墓 (The large Shang period tomb at Xin’gan). Beijing: Wen-wu Chubanshe.

Jin Zhengyao 金正耀, Mabuchi Hisao 馬淵久夫, Tom Chase, Chen De’an 陳德安, Miwa Karoku 三輪嘉六, Hirao Yoshimitsu 平尾良光, and Zhao Dianzeng 趙殿增 (1995).“Guanghan Sanxingdui jisikeng qingtongqi de huaxue zucheng he qiantongweisu bizhi yanjiu 廣漢三星堆祭祀坑青銅器的鉛銅位素比值研究 (Study of the lead isotope ratios on the bronzes from the sacrificial pits at Sanxingdui, Guang-han).” Wenwu 1995.2: 80–85. Also in Sichuan 1999: 490–499.

Kane, Virginia C. (1974–75). “The Independent Bronze Industries in the South of China Contemporary with the Shang and Western Chou Dynasties.” Archives of Asian Art 28: 77–107.

Li Boqian 李伯謙 (1983). “Chenggu tongqiqun yu zaoqi Shu wenhua 城固銅器群與早期蜀文化 (The bronze hoard at Chenggu and early Shu culture).” Kaogu yu wenwu 1983.2: 66–71. Republished in Xu Zhongshu 徐中舒 (ed.), Ba Shu kaogu lunwenji 巴蜀考古論文集, Beijing: Wenwu, 1987: 33–39 and in Li Boqian, Zhongguo qingtong wenhua jiegou tixi yanjiu 中國青銅文化結構體繫研究, Beijing: Kexue Chubanshe, 1998: 260–267.

Li Ling 李零 (2001).“Lun Zhongguo de youyi shenshou 論中國的有翼神獸 (Winged sacred beasts in China).” Zhongguo xueshu 5: 62–134.


Li Xixing 李西興 (ed.) 1994. Shaanxi qingtongqi 陜西青銅器 (The Shaanxi bronzes). Xi’an: Shaanxi Renmin Meishu Chubanshe.

Li Xueqin 李學勤 (1991). Bijiao kaogu suibi 比較考古隨筆 (Comparative archaeological notes). Hong Kong: Zhonghua.

Li Xueqin (1996). “Peng Xian Zhuwajie qingtongqi de zaikaocha 彭縣竹瓦街青銅器的再考查 (Renewed investigation into the bronzes from Zhuwajie, Peng Xian).” In Sichuan Sheng Wenwu Kaogu Yan-jiusuo 四川省文物考古研究所 (ed.), Sichuan kaogu lunwenji 四川考古論文集, pp. 118–122. Beijing: Wenwu Chubanshe.

Li Ye 李燁 and Zhang Liwen 張歷文 (1996). “Yang Xian chutu Yin Shang tongqi jianbao 洋縣出土殷商銅器簡報 (Shang bronzes ex-cavated in Yang Xian).” Wenbo 1996.6: 73–75.

Lin Xiang 林向 (1995). Ba Shu wenhua xinlun 巴蜀文化新論 (New dis-cussion of the Ba Shu culture). Ba Shu xuelin congshu. Chengdu: Chengdu Chubanshe.

Liu Shi’e 劉士莪 (ed.) (2001). Laoniupo 老牛坡 (The site of Laoniupo). Xibei Daxue kaogu zhuanye tianye fajue baogao 西北大學考古專業田野發掘報告. Xi’an: Shaanxi Renmin Chubanshe, 2001.

Liu Yang (ed.) (2000). Masks of Mystery: Ancient Chinese Bronzes from San-xingdui. Sydney: Art Gallery of New South Wales.

Lu Liancheng 盧連成 and Hu Zhisheng 胡智生 (1983). “Baoji Rujia-zhuang, Zhuyuangou mudi chutu bingqi de chubu yanjiu—jianlun Shushi bingqi de yuanyuan he fazhan 寶雞茹家莊, 竹園溝墓地出土兵器的初步研究—–兼論蜀式兵器的淵源和發展 (Initial study of the weapons excavated at the cemeteries of Rujiazhuang and Zhu-yuangou in Baoji, and some remarks on the derivation and develop-ment of Shu-type weapons).” Kaogu yu wenwu 1983.5: 50–65.

Lu Liancheng and Hu Zhisheng (1988). Baoji Yu guo mudi 寶雞A國墓地 (The cemetery of the Yu polity at Baoji), 2 vols. Beijing: Wenwu Chubanshe.

Luo Tai 羅泰 (= Lothar von Falkenhausen) (1997). “Youguan Xi Zhou wanqi lizhi gaige ji Zhuangbai qingtongqi niandai de xin jiashuo: Cong shixi mingwen shuoqi 有關西周晚期禮制改革暨莊白青銅器年代的新假說—–從世繫銘文說起 (A new hypothesis concerning the Late Western Zhou ritual reform and the date of the bronzes from Zhuangbai: Reflections on the genealogical terminology in the bronze inscriptions),” Li Ling (trans.). In Zhongguo kaoguxue yu lishi-xue zhi zhenghe yanjiu 中國考古學與歷史學之整合研究 (Integrated Studies of Chinese Archaeology and Historiography), Zang Zhen-hua (Tsang Cheng-hwa) 臧振華 (ed.), vol. 2, pp. 651–676. Taipei: Academia Sinica, Institute of History and Philology.

Luo Tai (ed.) (2003). Qiyi de tumu: Xifang xuezhe kan Sanxingdui wenhua 奇異的凸目—西方學者看三星堆文化 (Strange protruding eyes:


Western scholars look at the Sanxingdui culture). Chengdu: Ba Shu Shushe.

Lutz, Albert (ed.) (1986). Dian: Ein versunkenes Königreich in China. Zürich: Museum Rietberg.

Mao Xian 1994. Mao Xian Qiangzu Bowuguan 茂縣羌族博物館 and Aba Zangzu Qiangzu Zizhizhou Wenwu Guanlisuo 阿壩藏族羌族自治州文物管理所 (1994). “Sichuan Mao Xian Moutuo yihao shi-guanmu ji peizangkeng qingli jianbao 四川茂縣牟託一號石棺墓及陪葬坑情理簡報 (Preliminary report on the clearing of stone-cist tomb no. 1 and its accessory pits at Moutuo, Mao Xian, Sichuan).” Wenwu 1994.3: 4–40.

Mei, Jianjun (2003). “Qijia and Seima-Turbino: the Question of Early Contacts between North China and Eurasia.” Paper presented at the conference on New Perspectives in Eurasian Archaeology, Stockholm, November 21–23, 2003.

Miho Museum: South Wing. Shigaraki: Miho Museum, 1997. Nakamura Shin’ichi 中村慎一 (1997). “Sekikaga iseki wo meguru sho-

mondai 石家河遺址をめぐる諸問題 (Some issues concerning the Shijiahe sites).” Chūgoku kōkogakkai kaihō 7 (1997): 41–55.

Nanjing Bowuyuan 南京博物院 (1984). “1982 nian Jiangsu Chang-zhou Wujin Sidun yizhi de fajue 1982 年江蘇常州武進寺墩遺址的發掘 (Excavation of the Sidun site at Wujin, Changzhou, Jiangsu, in 1982).” Kaogu 1984.2: 109–129.

Negbi, Ora (1976). Canaanite Gods in Metal: An Archaeological Study of An-cient Syro-Palestinian Figurines. Tel Aviv: Tel Aviv University, Institute of Archaeology.

Pirazzoli-t’Serstevens, Michèle (1974). La civilisation du royaume de Dian à l’époque Han. Publications de l’École Française d’Extrême-Orient, vol. 94. Paris: École Française d’Extrême-Orient.

Pirazzoli-t’Serstevens, Michèle (2001). “Sichuan in the Warring States and Han Periods.” In Bagley (ed.) 2001: 39–57.

Qian Yuzhi 錢玉趾 (1992). “Sanxingdui qingtong lirenxiang kao 三星堆青銅立人像考 (On the bronze statue of a standing person from Sanxingdui).” In Sichuan wenwu: Sanxingdui gu Shu wenhua yanjiu zhuanji (1992), pp. 50–55.

Rawson, Jessica (ed.) (1983). The Chinese Bronzes of Yunnan. London: Sidgwick and Jackson.

Rawson, Jessica (1990). Western Zhou Ritual Bronzes from the Arthur M. Sackler Collections. Ancient Chinese Ritual Bronzes from the Arthur M. Sackler Collections, vol. 2, 2 parts. Cambridge, Mass.: Harvard University Press.

Rawson, Jessica (1995). Chinese Jade from the Neolithic to the Qing. London: British Museum Press.


Rawson, Jessica (ed.) (1996). Mysteries of Ancient China: New Discoveries from the Early Dynasties. London: British Museum, and New York: George Braziller.

Salmony, Alfred (1938). Carved Jade of Ancient China. Berkeley: Gillick Press (reprint London: Han Shan Tang, 1982.)

Sanseitai: Chūgoku 5000 nen no nazo—kyōi no kamen ōkoku 三星堆。中國 5000 年の謎―狂威の假面王國 (Sanxingdui: A 5000-year mystery of China—The kingdom of the awe-inspiring masks). Tōkyō: Setagaya Art Museum et al., 1998.

Sanxingdui 1999. Sanxingdui chuanqi: Hua Xia gu wenming de tansuo 三星堆傳奇: 華夏古文明的探索. Taipei: Guoli Gugong Bowuyuan, 1999.

Shaanxi 1979. Shaanxi Sheng Kaogu Yanjiusuo 陜西省考古研究所, Shaanxi Sheng Wenwu Guanliweiyuanhui 陜西省文物管理委員會, and Shaanxi Sheng Bowuguan 陜西省博物館 (1979). Shaanxi chutu Shang Zhou qingtongqi 陜西出土商周青銅器 (Shang and Zhou bronzes unearthed in Shaanxi), vol. 1. Beijing: Wenwu Chubanshe.

Shi Jingsong 施勁松 (1996). “Guanyu Sichuan Moutuo yihao shiguan-mu ji qiwukeng de liangge wenti 關於四川牟托一號石棺墓及棄物坑的兩個問題 (Two problems concerning the stone-cist tomb no. 1 and its paraphernalia pits at Moutuo, Sichuan).” Kaogu 1996.5: 461–466.

Sichuan 1981. Sichuan Sheng Bowuguan 四川省博物館 and Peng Xian Wenhuaguan 彭縣文化館 (1981). “Sichuan Peng Xian Xi Zhou jiao-cang tongqi 四川彭縣西周窖藏銅器 (Bronzes from a Western Zhou hoard at Peng Xian, Sichuan).” Kaogu 1981.6: 496–499, 555.

Sichuan 1987. Sichuan Wenwu Guanliweiyuanhui 四川文物管理委員會, Sichuan Sheng Bowuguan, and Guanghan Xian Wenhuaguan 廣漢縣文化館 (1987). “Guanghan Sanxingdui yizhi 廣漢三星堆遺址 (The Sanxingdui site in Guanghan).” Kaogu xuebao 1987.2: 227–54.

Sichuan 1998. Sichuan Sheng Wenwu Kaogu Yanjiusuo Sanxingdui Gongzuozhan 四穿省文物考古研究所三星堆工作站 and Guang-han Shi Wenwu Guanlisuo 廣漢市文物管理所 (1998). “Sanxingdui yizhi Zhenwu Cangbaobao jisikeng diaocha jianbao 三星堆遺址真武蒼包包祭祀坑調查簡報 (Preliminary investigation report on the sacrificial pit at Zhenwu Cangbaobao at the Sanxingdui site).” In Sichuan kaogu baogaoji, Sichuan Sheng Wenwu Kaogu Yanjiusuo (ed.), pp. 78–90. Beijing: Wenwu Chubanshe.

Sichuan 1999. Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文勿考古研究所 (1999). Sanxingdui jisikeng 三星堆祭祀坑 (The sacrificial pits at Sanxingdui). Beijing: Wenwu Chubanshe.

So, Jenny F. (2001). “Jade and Stone at Sanxingdui.” In Bagley (ed.) 2001: 153–175.


Su Bingqi 蘇秉琦 (1994). Huaren, long de chuanren, Zhongguoren: Kaogu xun-genji 華人, 龍的傳人, 中國人—考古尋根記 (The Hua People; the Descendants of the Dragon; Chinese: Notes from the archaeological search for our roots). Shenyang: Liaoning Daxue Chubanshe.

Sun Hua 孫華 (2000). Sichuan pendi de qingtongshidai (The Bronze Age of the Sichuan Basin). Beijing: Kexue Chubanshe.

Tang Jinyu 唐金鈺, Wang Shouzhi 王守智, and Guo Changjiang 郭長江 (1980). “Shaanxi sheng Chenggu xian chutu Yin Shang tongqi zhengli jianbao 陜西省城固縣出土殷商銅器整理簡報 (Preliminary report on the sorting out of the Shang bronzes unearthed at Chenggu county, Shaanxi province).” Kaogu 1980.3: 211–218.

Teng Shu-p’ing (= Deng Shuping) (2000). “The Original Significance of Bi Discs: Insights Based on Liangzhu Jade Bi with Incised Sym-bolic Motifs,” Lothar von Falkenhausen (trsl.) Journal of East Asian Archaeology 2.1–2: 165–194.

Thorp, Robert (1999). “Sacrificial Pits at Sanxingdui, Guanghan, Si-chuan Province.” In Yang Xiaoneng (ed.) 1999: 206–227.

Thote, Alain (ed.) (2003). Chine, l’énigme de l’homme de bronze. Archéologie du Sichuan (XIIe–IIIe siècle av. J.-C.). Paris: Paris Musées/éditions Findakly.

Tong Enzheng 童恩正 (1977).“Woguo xinan diqu qingtongjian de yan-jiu” 我國西南地區青銅劍的研究 (Studies on bronze swords in our country’s southwestern regions).” Kaogu xuebao 1977.2: 35–55. Also in: Tong Enzheng, Zhongguo xinan minzu kaogu lunwenji 中國西南民族考古論文集, Beijing: Wenwu Chubanshe, 1990: 96–117.

Tong Enzheng (1979). “Woguo xinan diqu qingtongge de yanjiu 我國西南地區青銅戈的研究 (Studies on bronze dagger-axes in our country’s southwestern regions).” Kaogu xuebao 1979.4: 441–60. Also inTong Enzheng, Zhongguo xinan minzu kaogu lunwenji, Beijing: Wenwu Chubanshe, 1990: 118–134.

Wang Hongming 王宏明 (1985). “Shandong sheng Haiyang xian shi-qian yizhi diaocha 山東省海陽縣史前遺址調查 (Survey of prehis-toric sites in Haiyang county, Shandong province).” Kaogu 1985.12: 1057–1067.

Wang Jiayou 王家祐 (1961). “Ji Sichuan Peng Xian Zhuwajie chutu de tongqi 記四川彭縣竹瓦街出土的銅器 (Notes on bronzes unearthed at Zhuwajie, Peng Xian, Sichuan).” Wenwu 1961.11: 28–31.

Wang Shouzhi 王守智 (1988). “Shaanxi Chenggu chutu de Shang dai qingtongqi 陜西城固出土的商代青銅器 (Shang period bronzes un-earthed at Chenggu, Shaanxi).” Wenbo 1988.6: 3–9.

Wu Hung (1997). “All About the Eyes: Two Groups of Sculptures from the Sanxingdui Culture.” Orientations 28.8: 58–66.

Xie Hui 謝輝 (2003). “Dui Jinsha yizhi chutu bufen yuqi de jidian renshi


對金沙遺址出土部分玉器的幾點認識 (Insights concerning some of the jades excavated at Jinsha).” Sichuan wenwu 2003.3: 60–63.

Xu, Jay (2001a). “Bronze at Sanxingdui.” In Bagley (ed.) 2001: 59–151. Xu, Jay (2001b). “Reconstructing Sanxingdui Imagery: Some Specula-

tions.” Orientations 32.5: 32–44. Xu, Jay (2001c).“Sichuan Before the Warring States Period.” In Bagley

(ed.) 2001: 21–37. Yang Hong 楊泓 (1985). “Jian he dao 劍和刀 (Swords and knives).” In

Yang Hong, Zhongguo gudai bingqi luncong (zengdingben) 中國古代兵器論叢, pp. 115–130. Beijing: Wenwu Chubanshe.

Yang Meili 楊美莉 (2002). “Erlitou wenhua de qianlüsongshi tongpai 二里頭文化的嵌綠松石銅牌 (Erlitou Bronze plaques inlaid with tur-quoise)”, followed by Japanese translation. Miho Museum kenkyū kiyō 3 (2002): 27–77.

Yang Xiaoneng (ed.) (1999). The Golden Age of Chinese Archaeology: Celebrated Discoveries from the People’s Republic of China. Washington: National Gal-lery of Art and Kansan City: Nelson-Atkins Museum of Art.

Yang Yachang 楊亞長 (1998). “Shaanxi shiqian yuqi de faxian yu chu-bu yanjiu 陜西史前玉器的發現與初步研究 (Discovery and initial study of prehistoric jades in Shaanxi).” In Deng Cong (ed.) 1998, vol. 1: 208–215.

Yu Weichao 俞偉超 (ed.) (2000). Chang jiang Sanxia wenwu cunzhen 長江三峽文物存珍 (Precious cultural relics from the Three Gorges of the Yangzi River). Chongqing: Chongqing Chubanshe.

Yue Runlie 岳潤烈 (1983). “Sichuan Hanyuan chutu Shang Zhouqing-tongqi 四川漢源出土商周青銅器 (Shang and Zhou bronzes un-earthed at Hanyuan, Sichuan).” Wenwu 1983.11: 91.

Yunnan Sheng Bowuguan 雲南省博物館 (1959). Yunnan Jinning Shizhai-shan gumuqun fajue baogao 雲南晉寧石寨山古墓群發掘報告 (Excava-tion report on the ancient cemetery at Shizhaishan, Jinning, Yunnan). 2 vols. Beijing: Wenwu Chubanshe.

Zeng Zhongmao 曾中懋 (1989). “Guanghan Sanxingdui yi, erhao jisikeng chutu tongqi chengfen de fenxi 廣漢三星堆一, 二號祭祀坑出土銅器成分的分析 (Component analysis of bronzes excavated from Sacrificial Pits nos. 1 and 2 at Sanxingdui, Guanghan).” Sichuan wenwu: Sanxingdui yizhi yanjiu zhuanji (1989), pp. 76–80.

Zhang Xuqiu 張緒球 (1992). Chang jiang zhongyou xinshiqishidai wenhua gailun 長江中游新石器時代文化概論 (Overall discussion of Neolithic cultures along the middle reaches of the Yangzi river). Wuhan: Hubei Kexue Jishu Chubanshe.

Zhao Congcang 趙叢蒼 (1996). “Chenggu Yang Xian tongqiqun zonghe yanjiu 城固洋縣銅器群綜合研究 (Comprehensive study of bronze assemblages from Chenggu and Yang Xian).” Wenbo 1996.4: 2–26.


Zhao Hui 趙輝 (1999). “Shijiahe yizhiqun 石家河遺址群 (The site cluster of Shijiahe).” In Su Bai 宿白 (ed.), Zhonghua Renmin Gongheguo zhongda kaogu faxian (1949–1999) 中華人民共和國重大考古發現 (Great archaeological discoveries in the People’s Republic of China), pp. 108–109. Beijing: Wenwu Chubanshe.

Zhongguo 1980. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (1980). Yinxu Fu Hao mu 殷墟婦好墓 (The tomb of Fu Hao at Yinxu). Zhongguo tianye kaogu baogaoji, Kao-guxue zhuankan 中國田野報告集考古學專刊, Series IV, no. 23. Beijing: Wenwu Chubanshe.

Zhongguo 1984. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo Erlitou Gongzuodui 中國社會科學院考古研究所二里頭工作隊 (1984). “1981 nian Henan Yanshi Erlitou muzang fajue jianbao 一九八一年河南偃師二里頭墓葬發掘簡報 (Preliminary report on tombs exca-vated in 1981 at Erlitou, Yanshi, Henan).” Kaogu 1984.1: 37–40.

Zhongguo 1986. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo Er-litou Gongzuodui (1986). “1984 nian Henan Yanshi Erlitou yizhi faxian de jizuo muzang 一九八四年河南偃師二里頭發現的幾座墓葬 (Some tombs excavated in 1984 at Erlitou, Yanshi, Henan).” Kaogu 1986.4: 318–23.

Zhongguo 1991. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo (1991). Zhongguo kaoguxuezhong tan shisi niandai shujuji, 1965–1991 中國考古學中碳十四年代數據集 1965–1991 (Collected radiocarbon dates from Chinese archaeology, 1965–1991). Kaoguxue zhuankan, Series II, no. 28. Beijing: Wenwu Chubanshe.

Zhongguo 1992. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo Erlitou Gongzuodui (1992). “1987 nian Erlitou yizhi muzang fajue jianbao 一九八七年二里頭遺址墓葬發掘簡報 (Tombs at the Erlitou site excavated in 1987).” Kaogu 1992.4: 294–303.

Zhongguo 1994. Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅器全集編輯委員會 (1994). Zhongguo qingtongqi quanji 中國青銅器全集 vol. 13 (Ba Shu) (A complete survey of Chinese bronzes, Sichuan/Chongqing volume). Beijing: Wenwu Chubanshe.

Zhongguo 1995. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (1995). Erlitou taoqi jicui 二里頭集翠 (High-lights of Erlitou ceramics). Kaoguxue zhuankan, Series II, no. 30. Beijing: Zhongguo Shehuikexueyuan Chubanshe.

Zhongguo 1997. Zhongguo Wenwu Jinghua Bianji Weiyuanhui 中國文物精華編輯委員會 (1997). Zhongguo wenwu jinghua 1997 中國文物精華 1997 (Highlights of Chinese cultural relics, 1997). Beijing: Wenwu Chubanshe.

Zhongguo 1999. Zhongguo Shehuikexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (1999). Yanshi Erlitou: 1959 nian–1978 nian kaogu


fajue baogao 偃師二里頭 —–1959 年–1978 年考古發掘報告 (The site of Erlitou, Yanshi: Report on excavations from 1959 to 1978). Zhongguo tianye kaogu baogaoji, Kaoguxue zhuankan, Series IV, no. 59. Beijing: Zhongguo Dabaikequanshu Chubanshe.

Zhongri Lianhe Kaogu Diaochadui 中日連合考古隊 (1998). “Sichuan Xinjin Xian Baodun yizhi 1996 nian fajue jianbao 四穿新津寶墩遺址 1996 年發掘簡報 (Preliminary report on excavations in 1996 at the site of Baodun, Xinjin, Sichuan).” Kaogu 1998.1: 29–50.

Zhou Wei 周維 (1993). Yazhou gu bingqi tushuo 亞洲古兵器圖說 (An illus-trated guide to the ancient arms of Asia). Shanghai: Shanghai Guji Chubanshe (first published in 1957).


THE JINSHA SITE: AN INTRODUCTION

BY

ZHU ZHANGYI 朱章義, ZHANG QING 張擎, and WANG FANG 王方

(Chengdu Municipal Institute of Cultural Relics and Archaeology)

Translated by Mark Pittner

Abstract This paper introduces the recently discovered site at Jinsha in the western suburbs of Chengdu, capital of Sichuan Province. In 2001, hundreds of objects of gold, bronze, jade, stone, and other materials were unearthed at Jinsha, and subsequent fieldwork has revealed that the site was a major settlement of the late Shang and Western Zhou periods, occupying an area of 3 sq km and following in date the famous Sanxingdui site. This paper situates the Jinsha site in its geographic and natural environment, discusses previous excavations in the area, describes the major localities and associated finds at the site, and analyzes the chronology and cultural classification of the site as well as the archaeological significance of its discovery. This is the first survey of the Jinsha site in English.

In February 2001, archaeological discoveries in the Chengdu 成都Plain once again made front page news around the world. While constructing the Shufeng Huayuancheng 蜀風花園城 Road at Jinsha 金沙 village in the Qingyang 青羊 district of the western suburbs of Chengdu, the China Real Estate Development Group-Chengdu Company 中房集團成都房地產開發總公司 discovered a large number of bronze, jade, stone, and ivory artifacts in the process of digging a drainage ditch.1 After hearing about this, the Chengdu Municipal Institute of Cultural Relics and Archaeology immediately sent a team to the site, and in keeping with cultural relics protection legislation, ordered construction to cease immediately and then took steps to secure the site. The following day, excavations began, gradually expanding from Locus Meiyuan 梅苑,

1 According to the ordinances of Chengdu relating to cultural heritage preservation in effect in 2001, road construction projects under the auspices of the city government were not required to first conduct a preliminary archaeological investigation.

248 ZHU ZHANgYI, ZHANg QINg, and WANg FANg

where the artifacts had been originally unearthed, to Locus Lanyuan 蘭苑 and Locus Tiyu Gongyuan 體育公園. Excavation work has con-tinued into 2005. In the course of research, and in light of data previously gathered in the same area, it has become clear that the Jinsha site covers an area of over three sq km—the area enclosed by Shuhan 蜀漢 Road to the north, Qingyang 青羊 Road to the east, Sanhuan 三環 Ring Road to the west, and Qingjiang 青江 Road to the south (see Figures 2 and 3 below). We preliminarily feel that it was the location of a large-scale Bronze Age site contemporaneous with the Shang 商 and Western Zhou 西周 periods, and that it may very well have been the capital city of an ancient state that was centered on the Chengdu Plain, perhaps comparable to, but somewhat later in date than, the site of Sanxingdui 三星堆 in Guanghan 廣漢, less than 40 km to the east. Over 1300 gold, bronze, jade, stone, earthenware, and ivory objects were excavated during the first season at Jinsha. Quite a few of these artifacts seem extremely similar to the pieces unearthed at Pits Nos. 1 and 2 at Sanxingdui. In order to provide timely information to the scholarly community, an exhibition was held at Peking University in 2002, and a catalogue was published (Chengdu 2002), providing the first opportunity for preliminary study. In this brief overview of the Jinsha finds,2 we will begin with a description of the geographic environs of the site.

Geographic location and surroundings

The Jinsha excavations reveal an early stage in the history of ancient Sichuan. Abutting the eastern edge of the Qinghai-Tibet (Qingzang 青藏 ) Plateau and forming part of the Sichuan Basin, the Chengdu Plain, where the site is located, is the only large plain in Sichuan. It consists of the alluvial fan deposited by the Minjiang 岷江 River and the Tuojiang 沱江 River, as well as their tributaries (the Jianjiang 湔江, Shitingjiang 石亭江, and Mianyuanhe 綿遠河), as they exited the mountain passes leading out from their sources in the mountainous regions of the northwestern Sichuan Basin. The Chengdu Plain is protected by the steep eastern escarpment of the Qingzang Plateau to the northwest. The foothills of the Longquan 龍泉 Mountains run northeast to southwest, forming a barrier in the southeast and separating the Chengdu Plain from the hilly region of central Sichuan. The and

2 The present essay is an adapted translation of the authors’ introduction to Chengdu 2002: 3–15. The draft translation by Mark Pittner has been extensively revised by the Editors.

THE JINSHA SITE 249

the Chaping 茶坪 Mountains in the northwest come together with the southeastern foothills of the Longquan Mountains to enclose the plain on the northeast and southwest. The Chengdu Plain is thus a physio-graphically independent unit in the western part of the Sichuan Basin (Figure 1). The Chengdu Plain is fan-shaped, 170 km long from north to south and 60 to 70 km wide from east to west, covering an area of approxi-mately 9,500 sq km. The entire plain is 400 to 750 m above sea level with the altitude decreasing from northwest to southeast. In the north-west, the Dujiang Weir 都江堰 is approximately 750 m above sea level, while the Chengdu area to the southeast drops to 500 m, and further to the south, within a distance of 50 km from Chengdu, the plain drops precipitously to 200 m at a grade of 3% to 4%. These rather striking differences in elevation have proven very useful to irrigation. In addition, the warm and moist subtropical climate makes the area highly suitable to agricultural production. Consequently, since early times the Chengdu Plain has been a place where humans have settled, leaving behind many ancient legends and rich deposits of cultural remains (Sichuan 1996). Through several thousand years of development, and particularly due to the implementation of the famous Dujiang Weir water control project

Figure 1. Map showing Chengdu and locations on the Chengdu Plain.

Dujiang Weir

Pixian

Chengdu

Guanghan

Deyang

Mianyang

Jianyang

Jinmahe R.

Min

jiang

R.

Jiang’anhe R.

Bait i aohe R.

TuojiangR.

Fu jiang R.

Long

quan

shan

Mts.

Fuhe

R.

0 40 km

N


in the third century BC, the Chengdu Plain was transformed into one of the most prosperous areas of China. As a consequence, Chengdu, at the center of the plain, became the administrative, economic, and cultural center of Southwest China. The urban area of Chengdu is in the southeastern area of the Chengdu Plain. With the Sanhuan Ring Road marking its perimeter, the modern city is about 14 km long from north to south, and 17 km wide from east to west, with an area of about 200 sq km (Figure 2). The topography of the city is generally flat, without large variations in elevation except for an area 20 km to the southeast of the city center where urbanization has reached the foothills of the Longquan Moun-tains, and an area in the north of the city where it extends into the Fenghuang 鳳凰 Mountains. The rivers around Chengdu generally flow from northwest to southeast. No later than the Warring States period, there were two relatively large waterways navigable by boat within the

Figure 2. Map locating the Jinsha site in Chengdu City, and the location of the Shi’erqiao site. After Chengdu 2004: 4.

North TrainStation

First Ring RoadSeco

nd Ring Road

Third Ring Road

Ren

min

Rd.

S.

Shudu Dadao

Yangshijie Rd. W. Ext.

Modihe River

Qingshuihe RiverNanhe River

Fuhe River

(Chengwen) Rd.

Shi’er-qiao

JinshaQingyang

Dadao

THE JINSHA SITE 251

city of Chengdu. The Shiji 史記 records that during the time when this area had been conquered by Qin 秦, “Governor of Shu 蜀 [Li 李] Bing 冰channeled through the levees and prevented harm from the Moshui 沫水 River, and created two canals through the middle of Chengdu. These canals could both be navigated by boats, and when there was excess [water], the water was used for irrigation. The common people thus benefited” (Shiji: “Hequshu” 河渠書, Zhonghua ed. 29: 1407). These two canals crossing through Chengdu were known as Liangjiang 兩江 during the Qin and Han periods, and Yang Xiong 揚雄, in his “Shudu fu 蜀都賦 (‘Rhapsody on the Shu capital’),” described them as follows: “The Liangjiang cross through its fore, seven bridges bound its flow” (Quan Shanggu Sandai Qin Han Sanguo Liuchao wen 全上古三代秦漢三國六朝文 51: 26: 402). The fourth century Ad Huayangguo zhi 華陽國志 states: “After the destruction of the Zhou 周, King Xiaowen 孝文of Qin made Li Bing Governor of Shu . . . Bing regulated the rivers by building dykes, he dug the Pi 郫 Canal and the Jian 檢 Canal, and created branches flowing into different parts of the commandery, thereby making them accessible by boat and watercraft” (Huayangguo zhi “Shu zhi” 蜀志 3: 6, pp. 132–133). Today, there are three major canals that cross the city, all of which are tributaries of the Zouma 走馬 River. From northeast to southwest, they are the Shahe 沙河, the Fuhe 府河 (one of its other names is the Pijiang 郫江),and the Nanhe 南河 (its other names include Qingshuihe 清水河 and Jianjiang 檢江). The Nanhe is the main channel of the Zouma River; it enters Chengdu in the city’s southwest, and there it is joined by the Modihe 磨底河 coming in from the northwest. The Nanhe flows along Wainanrenmin 外南人民 Road, Chengbian 城邊 Road, and Binjiang 濱江 Road; then, near Qinglianjie 青蓮街 Road, it joins together with the Fuhe River. Gradually turning south, the newly joined canal flows out of the city (Sichuan Sheng Wenshiguan 1987: 118–164). The Jinsha site sits on the banks of the Modihe River toward the northern part of the western district of Chengdu. The Jinsha site is nearly 5 km west of the center of Chengdu. The topography within the site is level, with the difference in elevation not exceeding 5 m. Within and around the site there are numerous water-ways. About 1.5 km to the south is the Qingshuihe River, to the north is the original channel of the Pijiang River, and the Modihe River passes through the center of the site from west to east, dividing the site into a northern and a southern half. The northern half now belongs to Huangzhongcun 黃忠村 village of the Jinniu 金牛 District, and the southern half is in Jinsha village of the Qingyang District. Not far to the southeast of the Jinsha site, in the southwestern sec-tion of present-day Chengdu, many Shang and Zhou period sites have


been found along the Pijiang’s original course and along the banks of the Nanhe River. These sites are spread out over an area ca. 10 km from west to east. Sites that have been archaeologically excavated include Fuqinxiaoqu 撫琴小區 (Wang Yi 1991), Shi’erqiao 十二橋 (Sichuan Sheng Wenwu Guanli Weiyuanhui 1987), Fangchijie 方池街 (Xu Pengzhang 1984), Junpingjie 君平街, Zhihuijie 指揮街 (Sichuan Daxue Bowuguan and Chengdu Shi Bowuguan 1987), Yandaojie 鹽道街, Minshan Fandian 岷山飯店, and Minjiangxiaoqu 岷江小區.3 As the Shi’erqiao site has the largest excavated area, some scholars have referred to this entire group of nearby sites as the “Shi’erqiao Site Cluster” (Sun Hua 1996: 123–144). In this cluster, the site closest to Jinsha is Fuqinxiaoqu, about 3 km to the southeast, and the most distant site is Minjiangxiaoqu, which is about 9 km away. The site featuring an earthen platform at Yangzishan 羊子山, discovered in 1954, is about 8 km to the northeast (Sichuan Sheng Wenwu Guanli Weiyuanhui 1957), and the noted Sanxingdui site is about 38 km from the Jinsha site.

Discovery and excavation

Startling discoveries attracted attention to Jinsha in 2001, but in fact, this site had already been located in 1995. Prior to 2001, our Institute (the Chengdu Municipal Institute of Cultural Relics and Archaeology) had undertaken excavations at three sites in the Jinsha area. Since all three lay north of Jinsha, in Huangzhongcun village, they were originally called the Huangzhongcun site (Figure 3). Between December 1995 and April 1996, our Institute conducted field survey and excavated an area of approximately 700 sq m at Locus Huangzhongxiaoqu 黃忠小區, on the eastern side of Huangzhong-cun, bordered on the east by Tonghe 同和 Road. In that excavation, a large quantity of pottery was unearthed, with the representative types including pointed-bottom bei 杯 cups, small, pointed-bottom zhan 盞 cups, ring-foot guan 罐 jars, tall-neck guan, and vessel lids. These dated between the late Shang and the early Western Zhou chronological periods, and the site is a typical settlement site of this period (Chengdu 2001: 164–181). Between June 1999 and April 2000, our Institute started fieldwork once again at Locus Sanhe Huayuan 三合花園 in Huangzhongcun, excavating an area of about 2026 sq m. Sanhe Huayuan lies in the northeastern part of Huangzhongcun; it is bordered by Tonghe Road in the east, and abuts the Yangshijie 羊市街 Road West Extension in

3 The as yet unpublished materials from Junpingjie, Yandaojie, Minshan Fandian, and Minjiangxiaoqu are held at the Chengdu Municipal Institute of Cultural Relics and Archaeology.

THE JINSHA SITE 253

the north and Huangzhongxiaoqu in the south. The cultural deposits in this region are rather thick, and many artifacts and features were discov-ered. In total, the remains of 17 buildings, 17 kilns, over 300 refuse pits, and 13 graves were discovered. The five buildings that appeared below Layer 5A were all large-size row buildings over 20 m long (Figure 4). The largest one (F6) had at least five rooms and measured nearly 8 m wide and over 54.8 m in length, covering a total area of over 430 sq m. These five structures are placed in a regular arrangement and possibly form a complex of buildings. All 17 of the kilns found at Locus Sanhe Huayuan are small-sized, bulbous-shaped (mantou 饅頭-shaped) kilns. The graves are of different types: some are primary burials, with the bodies supine and extended,

Figure 3. Excavation loci within the Jinsha site. After Chengdu 2004: 5.

SanheHuayuan

ThirdRing

Road

Qin

gyan

gD

adao

Rd.

Shufenglu Rd.

Modihe River

Lanyuan

Meiyuan

Huang-zhongcun

N

Yangshijie Rd. West Extension

YanshaTingyuan &Tiyugongyuan

Jinshayuan

MingyangfangchanJingangwan

JiangwangFudi

BoyaTingyun

Yinxinglu R

d.

Tong

heR

d.


while others contain secondary burials. Aside from one burial that in-cluded one bronze sword, all of the graves were without burial objects. The refuse pits are predominately round-shaped. The major types of pottery unearthed include pointed-bottom zhan, pointed-bottom bei, ring-foot guan, and medium-high-stem dou 豆pedestal serving-stands. The excavation also confirmed that an ancient waterway crossing from the northwest to the southeast wiped out the northwestern part of the site, probably sometime around the Warring States to the early Han period (Chengdu 2001). The excavation at Sanhe Huayuan was the largest excavation at Jinsha prior to 2001. Between July and September 2000, our Institute started excavating an area of about 500 sq m at Locus Jindu Huayuan 金都花園, once again in Huangzhongcun. Jindu Huayuan lies in the northern part of Huangzhongcun, abutting the Yangxixian Road in the north, and about 500 m west of Sanhe Huayuan. The cultural deposits are relatively thin. We discovered a few kilns, refuse pits, and graves, and the excavated pottery dates approximately from the late Shang to the early Western Zhou chronological periods. After these excavations, it was recognized that Huangzhongcun was a site of the Shi’erqiao Culture, the local culture of the Chengdu Plain during the late Shang and early Western Zhou historical periods, and that the site covered an area of approximately one sq km. At the time, however, it was not understood that this site was one locus within a large-

Figure 4. Aerial view of exposed wall foundations of the large row buildings excavated at Locus Sanhe Huayuan. Photo courtesy of the Chengdu Municipal Institute of Cultural

Relics and Archaeology.

THE JINSHA SITE 255

scale urban center, nor that it extended south of the Modihe River to Jinsha. This was only realized after numerous major cultural treasures were discovered at Jinsha, which has since given its name to the Jinsha site as a whole. After the discoveries at Jinsha, in order to coordinate the excavations with the ongoing construction project, a comprehensive survey and excavation plan for the area was launched on February 9, 2001, con-centrating on the localities of Meiyuan, Lanyuan, and Tiyu Gongyuan, which lie within the Shufeng Huayuancheng construction site of the China Real Estate Development Group-Chengdu Company, and Jin-shayuan 金沙園, which is being developed by the Chengdu Zhixinshiye Company 成都置信實業公司. These sites are all to the south of the Modihe River in Jinsha. The survey encompassed an area of one sq km, and a total of more than 800 test squares were excavated within the boundaries of the four localities. As of mid-2002, the total area excavated was approximately 16,000 sq m. These excavations, along with coring, have given us a tentative understanding of the distribution of remains and the nature of the Jinsha site.

Jinsha site archaeological loci

Locus Meiyuan. Locus Meiyuan is bordered to the north by the south-ern bank of the Modihe River and on the east by Qingyang Dadao and extends over at least 22 ha. Survey determined that the greatest concentration of finds was in an 8 ha area in the north-central part of Meiyuan. Moreover, a large number of precious artifacts were un-earthed in the northeastern part of Meiyuan in an area of about 0.8 ha. More than 1300 gold, bronze, jade, stone, ivory, and bone objects were unearthed by construction activities here before the start of systematic excavations. Between February and June 2001, our Institute excavated a total of 145 test squares covering an area of 3625 sq m in the north-eastern part of Meiyuan (Figure 5). At the end of June, when we had already excavated to Level 8, we ran into a layer of hardened earth that proved difficult to excavate. Since the rainy season was about to com-mence, work was suspended. Due to the preciousness of the artifacts unearthed in this area, the excavations were particularly detailed. All the test squares were excavated in synchronization so as to acquire a full understanding of the formation of the site, the patterns of artifact distribution, and the relationships between the features found and their stratigraphic context. Stylistic cross-dating of the artifacts allows us to date Level 5 to the Spring and Autumn period, and Levels 6 and 7 to the Western Zhou period. Judging from the section profile exposed by the mechanical equipment of the real estate construction company,


Figure 6. Cache of elephant tusks excavated in pit K1 at Locus Meiyuan. Photo courtesy of the Chengdu Municipal Institute of Cultural Relics and Archaeology.

this locus has thick cultural deposits down to a depth of approximately 4.5 m and consisting of at least 16 cultural strata. Over 700 important artifacts of gold, bronze, jade, and oracle bones were unearthed in the Western Zhou period strata. Additionally, a large number of ivory and pottery pieces were found. The three most important features were: pit K1, a pit full of elephant tusks, which also contained a number of jade and bronze pieces (Figure 6), feature L3, an area containing unfinished blanks for such objects as stone bi 璧disks and zhang 璋 (axe-shaped

Figure 5. Oblique view of the excavation grid at Locus Meiyuan. Photo courtesy of the Chengdu Municipal Institute of Cultural Relics and Archaeology.

THE JINSHA SITE 257

Figure 7. A section of feature L3 at Locus Meiyuan showing part of a scatter of unfinished or broken lithic artifacts and possible manufacturing preforms, including bi disks and zhang. Photo courtesy of the Chengdu Municipal Institute of Cultural

Relics and Archaeology.

Figure 8. Feature L2 at Locus Meiyuan, showing a portion of a dense scatter of modi-fied boar tusks and deer antler. Photo courtesy of the Chengdu Municipal Institute of

Cultural Relics and Archaeology.


objects, often with a forked blade), scattered over about 300 sq m, piling up layer upon layer (Figure 7), and finally, feature L2, an area covering about 300 sq m featuring an accumulation of sliced wild boar tusks, deer antlers, and elephant tusks, as well as pottery and decorative stones (Figure 8).

Locus Lanyuan. Lanyuan lies in the south-central section of Jinsha, 30 m west of Meiyuan. It forms a rectangle of 16 ha, within which finds were concentrated in a 2 ha area in the center. From July 2001 to January 2002, our Institute carried out a large-scale excavation here, opening 512 test squares and excavating an area of 1.28 ha (Figure 9). A large number of building remains were found, along with red, fired-earth surfaces, clusters of cache pits, more than 400 refuse pits, over 80 tombs, and one pottery kiln. Approximately 10,000 pottery vessels and sherds, and a small number of jades, bronze, and gold objects were also unearthed. These artifacts date from around the late Shang period. Building remains were generally distributed in the north-central portion of Lanyuan, whereas tombs were generally distributed in the western and southern areas.

Locus Tiyu Gongyuan. Tiyu Gongyuan reaches to the Modihe River in the north and borders the northern area of Lanyuan and the western area of Meiyuan, covering an area of at least 9 ha, including a dense concentration of finds of about 3.6 ha. From October to November 2001, our Institute excavated an area here of about 162 sq m. Building remains, fired-earth surfaces, and 15 graves were found. The 15 burials were concentrated in an area of 81 sq m. They have been preliminar-ily confirmed to be secondary burials. Among them, three were found to have burial objects, including a few jade and stone pieces as well as pottery; this allows their date to be fixed around the early Western Zhou period. Since these burials cut into cultural strata and building remains, this seems to have originally been a residential area that became a burial site after it was abandoned.

Our work so far suggests that the Jinsha site has a discernible plan. The eastern part of the site, particularly the northeastern corner of Locus Meiyuan, was perhaps an area of religious and ceremonial activities or workshops. The south central area, particularly Locus Lanyuan, seems to have been a residential area with a small burial ground. The center of the site, at Locus Tiyu Gongyuan, appears to have been a residential area that was later given over to burials. The area of Huangzhongcun in the northeast was most likely an elite residential area, because it includes large palace structures.

THE JINSHA SITE 259

Figure 9. Top: Aerial oblique view of the open excavation grid at Locus Lanyuan, 2002. Bottom: Locus Lanyuan excavations in progress. Photos courtesy of the Chengdu

Municipal Institute of Cultural Relics and Archaeology.


Archaeological remains

Features

Buildings: Buildings are found predominantly at Lanyuan and Sanhe Huayuan. All feature foundation trenches for the walls, in addition to the walls themselves, had been constructed of mud over a bamboo frame. There are two styles of building construction. The first features foundation trenches filled with clusters of small post holes. Buildings of this type are relatively small. The second type also has foundation trenches of clustered small postholes, but in addition, there are large postholes spaced regularly throughout, approximately 1 m apart. This construction style is found in buildings relatively large in area. All build-ings are generally oriented northwest to southeast.

Pottery Kilns: Kilns have been found at Lanyuan, Sanhe Huayuan, and Jindu Huayuan. All are small, bulbous-shaped kilns with an area of about 6 sq m. The various parts of the kiln are all preserved: the work surface,kiln door, fire chamber, and kiln chamber, with the kiln chambers mostly sloping up toward the back. These are the earliest kilns found to date in Sichuan Province.

Burial Sites: The most prominent cemeteries were found at Lanyuan and Tiyu Gongyuan. The grave pits are all oriented northwest to south-east. Heads are positioned either in the northwest or southeast, with the southeast orientation being predominant. There are both primary burials and secondary burials. In all of the primary burials, the bodies were supine and outstretched with their limbs straight. In about half of the burials there were no burial objects, and in those cases in which there were objects, they were few in number and generally restricted to pottery. There were only five graves with relatively plentiful amounts of burial objects, including not only pottery but also a number of bronze and jade items.

Pits: Pits were found spread across the whole site, but the densest concentrations were found at Lanyuan. All of the pits were rounded in plan. There were a small number of perfectly circular pits, which appear to have been arranged in a regular pattern. One of these produced a large number of peculiar pottery objects of no obvious practical function. The functions of the various kinds of pits still need clarification.

Elephant Tusk Pit: This feature (pit K1) lies in the northeastern corner of Lanyuan. The pit commenced below Layer 7, but it had been dam-aged by construction equipment before excavation, leaving a portion

THE JINSHA SITE 261

triangular in plan measuring 1.6 m in length and 0.6 m in width. The pit contained two layers of earthen fill. The first layer consisted of brown fill about 0.6 m thick, and the second of sandy fill, within which a large number of elephant tusks, the longest 150 cm long, were placed in a regular fashion (Figure 6). These tusks have been preliminarily deter-mined to be of the Asian elephant. The vertical section through the pit reveals that the tusks were piled up in eight layers. The pit also contained a large number of bronze and jade items. Since the conservation work on the ivory is still at an experimental stage, most of the tusks have been kept in situ for the time being.

A Surface Covered with Unfinished Ritual Lithics: This feature (L3) lies in the southern part of the northeast sector of Meiyuan. Within an area of ap-proximately 300 sq m was found a thick accumulation of stone bi-disks and zhang (axe-shaped objects often with a forked blade) (Figure 7). The deposit was especially thick at the northwest end, from where it slopes down and thins toward the southeast. The layers of stone objects are embedded in a yellow fill. No remains of building structures, activity areas, or stone-working tools have been found, though the excavation has not yet been completed, and the exact nature of the feature still awaits further research.

Concentration of Boars’ Tusks, Deer Antlers, and Semi-precious Stones: This feature (L2) lies in the northern part of the northeastern sector of Meiyuan. The area of this feature, exposed by construction equipment before excavation, measures over 300 sq m. The accumulation seems to be much more disordered than the lithic deposit previously described, but the boar tusks are generally found above the deer antlers (Figure 8). Preliminary analysis reveals that all of the boar tusks are lower canines, suggesting that they were selected for a specific purpose.

Artifacts

As of early 2002, more than 2000 important artifacts had been un-earthed at Jinsha, including approximately 40 gold objects, more than 700 bronze items, over 900 jade items, nearly 300 stone items, and more than 40 ivory and bone artifacts, in addition to a large number of elephant tusks and approximately 10,000 pottery vessels and sherds. Among the gold objects, the most important decorative motifs include human faces, fish pierced by arrows, four birds grouped around a “sun” design, and bands of fish with bird-like heads. These motifs are found on horn-shaped objects, box-shaped objects, racket-shaped objects, fish-shaped objects, and a large number of fragments. The imagery of birds, fish pierced by arrows, and a face-like motif on a gold foil band from


Jinsha (object 2001CQJC:688) is almost identical to that on a gold sheath unearthed from Pit No. 1 at Sanxingdui (Figure 10). This design symbolically conveys authority. Likewise, an openwork, round gold sheet cut with imag-ery of four birds harmoniously grouped concentrically around a “sun” motif has great emotive qualities (Figure 11). The bronze pieces unearthed at the Jinsha site are small in size. The most important pieces include a standing human figure, ox-heads, serrated ge 戈 dagger-axes, bi-shaped pieces, hoe-shaped pieces with square holes, eye-shaped pieces, clapper-bells, and cowrie shell-shaped ornaments. Most noteworthy is a human figure standing about 20 cm tall upon a base with both arms raised before its chest, resembling the positioning

Figure 10. Top: Roll-out drawing showing the bird and fish pierced by arrow motifs on a gold foil band (2001CQJC:688) from Jinsha; diameter 19.6–19.9 cm, width 2.68–2.8 cm, thickness 0.02 cm. Also see Chengdu 2002: 23–26. Unexcavated provenience (Note: This and other Jinsha objects with the “2001CQJC” catalogue code were uncovered during the construction activities that led to the discovery of the site. As such, their stratigraphic context is unclear). Bottom: A gold sheath (K1:1) with similar motifs from

Pit No. 1 at Sanxingdui; after Sanxingdui 1999: 61.

Figure 11. Line drawing of a gold foil disk with a cut-out design of four birds grouped concentrically around a “sun” motif (2001CQJC:477); outer diameter 12.5 cm, inner diameter 5.29 cm, thick-ness 0.02 cm. Also see Chengdu 2002:

29–31. Unexcavated provenience.

THE JINSHA SITE 263

of the arms of the large standing figure from Pit No. 2 at Sanxingdui (see Sanxingdui 1999: 162) (Figure 12). Most of the bronzes probably did not originally stand as individual objects but were instead probably components of larger, composite bronze objects.A small number of fragments of bronze vessels were also found, including ring-footed zun 尊 and other objects of unusual shapes. It seems likely that large-sized bronzes will be unearthed in the future. Among all the objects unearthed at Jinsha, the more than 900 jades are typologically the most varied. The most important types are the cong 琮 tubes (Figure 13), bi disks, axe-shaped objects, ge-shaped objects, objects shaped like zao 鑿 chisels, heads of divine figures, rings, and cowrie shell-shaped objects. Many of the axe-shaped objects feature incised decorations stained by cinnabar.

Figure 12. Drawing of a bronze standing human figurine (2001CQJC:17); overall height 19.6 cm, weight 641g. Also see

Chengdu 2002: 43–47. Unexcavated provenience.


Among the stone objects, the most important are sculptures of kneel-ing humans (Figure14), crouching tigers, coiled snakes, and turtles, as well as yue 鉞 battle-axes, axe-shaped objects, bi disks, adze-shaped objects, and chisel-shaped objects. The eyes, ears, and mouth portions of the human, tiger, and snake sculptures have traces of cinnabar. The most important pottery types include small, flat-bottom guan, lobed he 盉 pitchers, high-stem dou and dou-shaped vessels, tall-neck guan, ring-foot guan, bottles, pointed-bottom zhan, pointed-bottom bei cups,

Figure 13. Jade cong tube (2001CQJC:1); height 16.57 cm, weight 3.92 kg. Unexcavated provenience. Also see Chengdu 2002: 86–88. Photo courtesy of the Chengdu Municipal

Institute of Cultural Relics and Archaeology.

THE JINSHA SITE 265

and stands with constricted waists (Figure 15). The quantity of pottery is quite substantial, including some rather large pieces. The vessels also show a large amount of variation in their manufacturing techniques, and there are a number of typological differences from other, smaller sites in the general region during this time period. This would indicate that the Jinsha site was different from an ordinary settlement. Such forms as the small, flat-bottom guan, lobed he, high-stem dou, dou-shaped vessels, and narrow-neck bottles, are clearly derived from vessel forms of the Sanxingdui Culture. Pointed-bottom zhan, short-stem dou, and gui 簋-shaped vessels, on the other hand, are ancestral to the pottery of the Ba-Shu 巴蜀 Culture of the Warring States period.

Figure 14. Stone kneeling figure (2001CQJC:716); height 21.72 cm, weight 2.12 kg. Carved, drilled, and polished serpentinite originally painted with pigments. Also see

Chengdu 2002: 174–178. Unexcavated provenience.


The gold, bronze, jade, and stone artifacts from Jinsha show many similarities to those unearthed from Pits Nos. 1 and 2 at Sanxingdui. The human faces and fish pierced by arrows on the gold objects, the standing human figure, collared bi, and square-holed hoe-shaped ob-jects among the bronzes, and axes, bi-shaped objects, ge-shaped objects, and chisel-shaped objects among the jades, all have their counterparts at the Sanxingdui site. There is little doubt that the origins of Jinsha are in the Sanxingdui Culture. At the same time, the artifacts from the Jinsha site also exhibit some particular qualities of their own, such as: 1. A comparatively large quantity and variety of gold objects, of which many types are not seen at Sanxingdui; 2. A voluminous and typologically comprehensive assemblage of jades, including specimens of almost all the important jade types from Bronze Age China, with both large-sized and miniature-sized specimens occurring together; 3.A large quantity of stone sculpture carved in the round, such as kneeling human figurines, crouching tigers, and coiled snakes that are rarely seen elsewhere in China; 4. Distinctive pottery types such as pointed-bottom zhan, pointed-bottom bei cups, tall-neck guan, and ring-foot guan that are

Figure 15. Periodization of Jinsha pottery types. 1. guan with small flat bottom (H202:14); 2. guan with small flat bottom (H22:16); 3. hu (H437:1); 4.bottle (H244:2); 5. dou-shaped vessel (H147:1); 6. lobed he (H22:5); 7. pointed-bottom zhan (H305:19); 8. pointed-bottom zhan (H305:11); 9. pointed-bottom zhan (H305:20); 10. pointed-bottom zhan (H305:9); 11. pointed-bottom bei (H305:2); 12. pointed-bottom bei (H305:12); 13. pointed-bottom zhan (T6413(7):12); 14. pointed-bottom bei (L28:39); 15. guan with flat bottom (T7609(5):2); 16. guan with flat bottom (T7609(5):16); 17. pointed-bottom guan (T6413(7):17); 18. guan with flat bottom (T6413(7):2). Vessels 2, 3, 5, and 6 are drawn to Scale B; all others to Scale A. Vessels 1–12 excavated at Locus Lanyuan; vessels 13–18

excavated at the northeast sector of Locus Meiyuan (the “sacrificial zone”).

1 2 3 4 5 6

7 8 9 10 11 12

13 14 15 16 17 18

I

II

III

Scale

A: 10 cmB: 20 cm

THE JINSHA SITE 267

characteristic of the Shi’erqiao Culture. Indeed, the typological variety of the Jinsha ceramic assemblage exceeds by far that of the cluster of Shi’erqiao sites. Jinsha therefore should be considered the new type site of the Shi’erqiao Culture.

Dating and cultural classification

The discovery of the Jinsha site has given rise to questions about its date, its relationship with the Sanxingdui site, and whether or not it was a capital of an ancient Shu 蜀 state. Since none of the finds have yet been radiocarbon dated, and since there are no textual sources to follow, we can only base our investigation of its date and cultural classification on Jinsha’s typological position in the sequence of pre-Qin archaeological cultures on the Chengdu Plain. After many years of fieldwork on the Chengdu Plain, the basic se-quence of the area’s pre-Qin cultural history has now been established (Sun Hua 2000: 91–115; Jiang Zhanghua et al. 2002). This sequence is as follows: the Baodun 寶墩 Culture → the Sanxingdui Culture → the Shi’erqiao Culture → the culture represented by the Shangwangjiaguai 上汪家拐 site. The late Neolithic Baodun Culture is represented by a cluster of six walled settlements (see Wang Yi, this volume). This culture already used walls of wood (or bamboo) framing covered with mud to con-struct rectangular buildings. A small number of “post and fence” type structures have also been found. The six prehistoric walled settlements all had large and for the most part rectangular-shaped enclosures, and some also had large, wide moats. Graves were rectangular, vertical pits. Burial objects differed from site to site. At the Shijiefang 十街坊 site near Chengdu’s South Train Station, 7 of the 19 graves excavated yielded burial objects of bone; at the Renshou 仁壽 cemetery at the Sanxingdui site, the predominant materials of burial objects were jade and stone. Graves found sporadically at other sites have been devoid of any funerary items. The stone objects found were mostly tools for daily use; most were made solely by grinding, and they tended to be small in size. Trapezoidal axes, trapezoidal adzes, and double-ended bladed chisels were the predominant types. Pottery was constructed by coiling and then finished on a slow wheel. The main varieties are gray coarse pottery, gray-white fine-clay pottery, and yellowish-gray pottery. Pat-terned decoration is dominated by cord-marking, incised lines, impressed patterns, and appliqué. The predominant vessel types are cord-mark patterned scallop-rim guan, everted-mouth ring-foot zun, ring-foot zun with plate-shaped mouth, tall-neck guan with trumpet-shaped mouth,


wide-rim flat-bottom zun, and hu 壺 . This culture is approximately contemporaneous with the Longshan Period of Central China and the early Xia historical chronological period, and thus may be dated to ca. 2700 bc to 1700 bc. The Baodun Culture is the earliest archaeological culture on the Chengdu Plain known to date. The early Bronze Age Sanxingdui Culture is exemplified by the San-xingdui site in Guanghan County. The predominant form of architec-ture in this period still consisted of rectangular structures constructed with walls of wood framing covered in mud. Everyday tools were small items made of ground stone, including trapezoidal axes, adzes, long strip-shaped chisels, and semi-lunate knives. The stone cores left over from making bi disks were re-bored to make stone spindle whorls. Bor-ing with a solid drill or a tubular cylindrical drill were the two methods used to make holes. Utilitarian ceramic implements include spindle whorls and net weights. Coarse pottery accounts for the majority, of which the most common variety was a brown ware. Decorative patterns were not particularly pronounced, consisting mostly of cord-marking, bowstrings, impressed patterns, appliqué, incised-line, and geometric patterns. Typical vessel forms included guan with small flat-bottoms, high-stem dou and dou-shaped vessels, li 鬲-shaped vessels, lobed he, ring-foot pan 盤, ladles with bird-head handles, hu, bottles, gu 觚, and lids. The site dates to the Xia and Shang periods, with absolute dates ranging 1700 bc to 1150 bc. Continual discoveries of jade and stone ritual objects at Sanxingdui since the 1930s, followed by the discovery, in 1986, of two artifact caches yielding a large number of fantastic objects, and the discovery of extensive city walls, all suggest that in this period there was already an urban-based civilization in the Chengdu Plain, of which the Sanxingdui site was the political, economic, and cultural center. The Sanxingdui Culture certainly inherited certain elements of the Baodun Culture, but its development must have been greatly influenced by contemporaneous cultures, including both the Erlitou Culture in central China and the Bronze Age cultures of the central Yangzi River region. Following the Sanxingdui Culture and dating from approximately the late Shang to the early Spring and Autumn periods (ca. 1200 bc to 600 bc), the Shi’erqiao Culture was first defined at the Shi’erqiao site. Shi’erqiao building techniques include the “post and fence” type seen at the Shi’erqiao site, as well as the mud-over-wood-frame construc-tion found at Huangzhongcun. Some buildings were relatively large in size, perhaps being palaces. Utilitarian stone tools still include small, ground stone axes, adzes, and chisels. In addition, a large number of plate-shaped stone objects of unclear use have been preserved. There is a large number of polished bone objects, including combs, needles,

THE JINSHA SITE 269

awls, and arrowheads. Bronze objects include long, thin, strip-like chisels, square-awl-shaped bronze bars, arrowheads with raised spines and reversed points, as well as primitive, willow-leaf-shaped swords. The prevalence of oracle bones in the Shi’erqiao period is notable, clearly showing the influence of the Shang and Zhou cultures from Central China. The oracle bones are for the most part turtle plastrons, some of which have both round bored hollows and chiseled square hollows, while others feature only the round bored hollows. Ceramic implements include a large number of spindle whorls but only a smaller number of net weights. The ceramic vessel assemblage is dominated by coarse brown and gray pottery. Decoration is seldom seen, with cord marks being the most common, followed by some instances of diamond-shaped patterns, bird designs, bowstrings, appliqué, and a circlet pattern. Rep-resentative vessel forms include pointed-bottom bei, pointed-bottom guan, pointed bottom zhan, pen 盆 basins, ring-foot guan, tall-neck guan, trumpet-mouth guan, weng 瓮 urns, fu 釜 cauldrons, and lids with either double-loop shaped knobs, or three-petaled knobs. Chronology and typology show that the Shi’erqiao Culture is en-tirely derived from the Sanxingdui Culture. The first chronological phase of Shi’erqiao adopts such Sanxingdui elements as he, ladles with bird-head handles, hu, bottles, guan with small flat bottoms, and high-stem cup-shaped stands. But due to the appearance in this period of a large number of pointed-bottom bei, pointed-bottom guan, pointed-bottom zhan, we are able to differentiate the Shi'erqiao Culture from the Sanxingdui Culture.4 Additional new forms in this period include trumpet-mouth guan, pen, constricted-mouth weng, guan, and weng with scalloped lips. The pointed-bottom bei was peculiar to Phase 1 of the Shi’erqiao Culture. During Phase 2 of the Shi’erqiao Culture, all of the ceramic vessel forms inherited from the Sanxingdui Culture had died out, and from then on the differences between the two cultures began to increase. Due to the extent of these differences found in Phase 2, some scholars such as Sun Hua (2000) differentiate this phase as a separate archaeological culture, the Xinyicun 新沂村 Culture. Following Shi’erqiao, the site of Shangwangjiaguai represents the archaeological culture of the Chengdu Plain during the Warring States period (Chengdu 1993). The number of contemporaneous settlements found so far is relatively small. The most important, apart from Shangwangjiaguai, is the Qingyanggong 青羊宮site, after which some scholars have named this period the Qingyanggong Culture (Sichuan Sheng Bowuguan 1959b). As there are literary records that

4 Editors’ Note: The paper by Jay Xu in this same volume presents an alternative view on the relationship between the Sanxingdui and Shi’erqiao Cultures.


indicate that the Sichuan Basin was occupied by the Ba 巴 and Shu kingdoms during this period, most archaeologists tend to use the term the “Ba-Shu Culture” to describe this period in Sichuan. There are a large number of tombs from this period, the largest being the wooden-chamber tomb at Jiuliandun 九聯墩 in Majia 馬家 Township of Xindu 新都 County, and the boat-coffin tombs at Shangyejie 商業街 Road in Chengdu. Tomb typology is quite diverse, including narrow and rectangular vertical-pit tombs, boat-coffin chamber tombs, wooden chamber tombs, and rectangular earthen chambers lined with wooden planks. Typical pottery vessels include rope-patterned round-bottomed fu cauldrons, shallow-basin low-stem dou, shallow-basin medium-stem dou, bowl-shaped dou, weng, guan, and ding 鼎 tripods. Among the bronze vessels there are ding, dou, hu, pan, yi 匜, zeng 甑, mou 鍪, lei 罍, fou 缶, and yan 甗 steamers. Weaponry includes dagger-axes, spears, swords, and battle-axes. For hand tools there are axes, jin 斤 narrow axes, adzes, and saws. Aside from these there are a few other categories including chime bells, Ba-Shu seals, and belt hooks. Shangwangjiaguai dates from approximately the Warring States period to the beginning of the Han period, or ca. 500 bc to 100 bc. There is a chronological gap between Shi’erqiao and Shangwangjiaguai, but it is clear that they are in cultural succession. Among these four cultural and chronological stages, the Jinsha site most closely resembles the third—the Shi’erqiao Culture. The predomi-nant pottery at Jinsha includes pointed-bottom bei, pointed-bottom zhan, ring-foot guan, tall-neck guan, weng, high-stem dou cup-shaped stands, and guan with trumpet-shaped mouths. These are all typical of the Shi’erqiao Culture, and so these Jinsha objects should also date from the late Shang period to the early Spring and Autumn period. Specific parallels between Jinsha and Shi’erqiao include the following: First, pointed-bottom bei, ring-foot guan, and tall-neck guan were un-earthed from Layer 7 in the northeastern sector of Meiyuan that are the same as or similar to those of the latter part of Phase 1 of the Shi’erqiao site and thus date to the Early Western Zhou period. In Layer 6 of this area a relatively small number of artifacts have been unearthed, which on the whole are close in style to those of Layer 7 and should be similar in date. In Level 5, trumpet-mouth guan and straight-mouth pointed-bottom zhan have been unearthed that correspond to the typical artifacts of the latter part of Phase 2 at Shi’erqiao, dating from approximately the early Spring and Autumn period. Accordingly, cultural deposits in this area continued over a relatively long period. The gold, bronze, and jade pieces unearthed from Meiyuan are datable to the late Shang and early Western Zhou periods due to two factors. First, it is not until the early Western Zhou strata that gold, bronze, and jade objects appear,

THE JINSHA SITE 271

and second, these objects are the same as or similar to those from Pits Nos. 1 and 2 at Sanxingdui, to which they should be close in date. Second, the date of the cultural assemblage at Lanyuan is a bit earlier than that of Meiyuan. The predominant pottery types, which include guan with small flat-bottoms, pointed-bottom zhan with constricted mouths, pointed-bottom bei, high-stem dou, high-stem cup-shaped stands, and he, resemble those from the early part of Phase 1 of Shi’erqiao as well as the beginning of the late part of Phase 1. Based on all the finds at the Jinsha site, we preliminarily understand that the site dates from the late Shang to the Spring and Autumn period, but most of the excavated remains date between the late Shang and early Western Zhou periods, and they belong to the early Shi’erqiao Culture. Extending over approximately 3 sq km, Jinsha was thus a large-scale Shang-Zhou period site of the Shi’erqiao Culture. The different parts of the site each have their own functions. These include a palace area, a ceremonial-ritual area, a residential area, and a burial ground. The discovery of ritual objects and building remains sets Jinsha apart from ordinary settlements of its time. It is therefore reasonable to suggest that Jinsha was the central site on the Chengdu Plain and that it was the capital of an ancient state.

The significance of the discovery of Jinsha

Jinsha is the most important archaeological discovery in Sichuan since Sanxingdui. Its geographical location, size, rich and varied remains, as well as the quality of the objects unearthed all suggest that this site was a major political, economic, and cultural center of the Chengdu Plain after the decline of Sanxingdui. Therefore, the discovery of the Jinsha site is important in a number of ways. First, Jinsha was occupied continuously for a relatively long time, from the late Shang all the way to the Spring and Autumn periods, but the most important remains date from around the Western Zhou period. Even though similar pottery had been unearthed already at Shuiguanyin 水觀音 in Xinfan 新繁 County (Sichuan Sheng Bowuguan 1959a), Shi’erqiao (Sichuan Sheng Wenwu Guanli Weiyuanhui 1987), and other sites, none of these sites can compare with Jinsha in size, richness, and length of occupation. Beyond enriching our knowledge of the Shi’erqiao Culture, Jinsha also connects the Sanxingdui Culture on the one end with Shangwangjiaguai on the other. The Jinsha site thus provides essential material for establishing an archaeological sequence in the Chengdu Plain for pre-Qin times, and also for exploring cultural developments following the Sanxingdui Culture.


The Jinsha site was merely 38 km from the Sanxingdui site. Both are located on the Chengdu Plain, and both were large-scale settlements (though unlike Sanxingdui, no walls have been found at Jinsha). Jinsha directly followed Sanxingdui in time. It has been generally accepted that the Sanxingdui site was the central urban site of the Chengdu Plain, or even the entire Sichuan Basin, during the late Xia to early Shang period (or slightly later). Where did this center shift following the decline of Sanxingdui? The discovery of Jinsha answers or at least provides valuable clues for this question, which before had been impossible to answer. The discovery of the Jinsha site greatly enriches our knowledge of the Bronze Age material culture of the Chengdu Plain and provides a large amount of material evidence for exploring the origins and history of the early Shu state. Connecting this evidence with the semi-mythical accounts of the ancient history of Sichuan, it can be postulated that the ancient state of Shu—which is documented in the legends and early literary sources about Sichuan—may have had its capital at Jinsha during the period from the late Shang to the Western Zhou period (provided, of course, that this state was in fact already in existence on the Chengdu Plain in the time period before the Eastern Zhou). Early Shu cultural beliefs and social customs may be reflected in the artifacts unearthed a Jinsha. Such items as the gold ornament with four birds encircling the sun, for example, reflect the worship of a sun god. Other decorated gold objects—similar to the golden sheath from Pit No. 1 at Sanxingdui—also reflect the practice of bird worship, and the bronze standing figure and stone kneeling figures demonstrate with their dif-ferent hair styles and other features that there were class distinctions among the people of the time. The discovery of Jinsha also allows us to trace back the origins of the city of Chengdu, the historical capital of ancient Shu. According to literary sources, ever since King Kaiming 開明 of Shu moved there, Chengdu was continuously the principal city of the entire Sichuan Basin. The discovery of the Jinsha site makes us aware that the history of Chengdu as a capital city began earlier than the Kaiming kingdom of the Warring States period, and that it had already become a principal city by the late Shang period. The discovery also shows that the city of Chengdu began in the area of what is now the northwest section of modern Chengdu, and that it then gradually shifted and expanded from the northwest towards the southeast, until the waning years of the Tang 唐 Dynasty, when it reached the area where the two canals dug by Li Bing converge. The urban growth of Chengdu can now be traced from Jinsha to the city of King Kaiming, to the Dacheng 大城 and Shaocheng 少城 cities of the Qin and Han periods, to Luocheng 羅城

THE JINSHA SITE 273

of the terminal period of the Tang Dynasty, and finally, to the modern extent of the city of Chengdu. Research at Jinsha has also helped to deepen our understanding of the two pits at Sanxingdui. As is well-known, after the discovery of these two pits, varying ideas about their nature and dating were proposed. Were they contemporaneous or not? Should they be dated as early as the Shang or the Western Zhou periods, or as late as the Eastern Zhou period? Were they sacrificial pits, cremation pits, or caches of treasures from conquered states? Of course, the finds from Jinsha cannot solve all of these questions, but through comparisons between the sites, one can at least ascertain that the date of the Sanxingdui pits cannot be later than the Western Zhou period. We have also gained many insights into early jades in the region. The 900 jades so far unearthed at Jinsha are richly varied in type and form, including all the common types from the Shang and Zhou periods but also many specimens of types never before unearthed. The Jinsha jades provide abundant evidence for research into the origin and develop-ment of the art of jade manufacture in the Sichuan Basin, the relation-ship between different jade-producing areas in China, the relationship between the different Shang and Zhou period Bronze Age cultures of China in general, as well as the use of jades in ritual activities of the time.The discovery of finished pieces, partially finished pieces, raw jade material, and a large number of pieces preserving the markings made during production should help us to better understand the processes used in working jade at the time, as well as the tools used to do so. Lastly, although the Jinsha site was discovered merely a few years ago, the excavation has mapped out the basic layout of the site and the functions of different localities. Further research and excavation at the site will enable us to understand more clearly the arrangement and structure of capital cities of the Shang-Zhou period, and it will facilitate comparisons with other contemporaneous capital cities in the Central Plain.

Guest Editor’s note:

The ongoing excavations at the Jinsha site frequently produce new information. The most recent preliminary report now estimates that the Jinsha site covers more than four sq km (Chengdu 2004: 58). In a personal conversation on March 8, 2005, Wang Fang, one of the authors of this paper, reported that nearly 200 tombs have been excavated at the Yansha Tingyuan 燕沙庭院Cemetery expansion to Locus Tiyu Gongyuan (Figure 16). Our knowledge of the Jinsha site will continue to be updated on a regular basis.


References Cited

Chengdu 1993: Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊 and Sichuan Daxue Lishixi 四川大學歷史系 (1993). “Chengdushi Shangwangjiaguaijie yizhi fajue baogao 成都市上汪家拐街遺址發掘報告 (Report on the excavation of the site at the Shangwangjiaguai Street in Chengdu).” Nanfang minzu kaogu 南方民族考古 5: 325–358.

Chengdu 2001: Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所 (2001). “Chengdu Shi Huangzhongcun yizhi 1999 niandu fajue de zhuyao shouhuo 成都市黃忠村遺址 1999 年度發掘的主要收穫 (The notable results in the 1999 excavation of the Huangzhongcun site in Chengdu).” In Chengdu Shi Wenwu Kao-gu Yanjiusuo (editor), 1999 Chengdu kaogu faxian 1999 成都考古發現 (Chengdu archaeological discoveries for 1999), pp.164–181. Beijing: Kexue Chubanshe.

Chengdu 2002: Chengdu Shi Wenwu Kaogu Yanjiusuo and Beijing Daxue Kaogu Wenboyuan 北京大學考古文博院 (2002). Jinsha tao-zhen: Chengdu shi Jinshacun yizhi chutu wenwu 金沙淘珍—–成都市金沙村遺址出土文物 (Panning for treasure at Jinsha: Cultural relics excavated at the Jinshacun site in the city of Chengdu). Beijing: Wenwu Chubanshe.

Chengdu 2004: Chengdu Shi Wenwu Kaogu Yanjiusuo (2004). “Chengdu Jinsha yizhi I qu ‘Meiyuan’ didian fajue yiqi jianbao 成

Figure 16. Aerial oblique view of the 2002–2003 excavations at the Yansha Tingyuan 燕沙庭院 cemetery expansion of Locus Tiyu Gongyuan. Photo courtesy of the Chengdu

Municipal Institute of Cultural Relics and Archaeology.

THE JINSHA SITE 275

都金沙遺址 I 區 ‘梅苑’ 地點發掘一期簡報 (Preliminary report on the first phase excavation of Locus Meiyuan at Sector I in the Jinsha site, Chengdu).” Wenwu 文物 2004.4: 4–65.

Huayangguo zhi jiao bu tu zhu 華陽國志校補圖注, by Chang Qu 常璩; Ren Naiqiang 任乃強 edition. Shanghai: Shanghai Guji Chubanshe, 1987.

Jiang Zhanghua 江章華, Wang Yi 王毅, and Zhang Qing 張擎 (2002). “Chengdu Pingyuan xian Qin wenhua chulun 成都平原先秦文化初論 (A preliminary discussion of the pre-Qin cultures of the Chengdu Plain).” Kaogu xuebao 考古學報 2002.1: 1–22.

Quan Shanggu Sandai Qin Han Sanguo Liuchao wen 全上古三代秦漢三國六朝文, ed. by Yan Kejun 嚴可均 . Shanghai: Zhonghua Shuju edition, 1958.

Sanxingdui 1999: Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 (editor) (1999). Sanxingdui jisikeng 三星堆祭祀坑 (Ex-cavation of the sacrificial pits at Sanxingdui). Beijing: Wenwu Chu-banshe.

Shiji 史記, by Sima Qian 司馬遷. Beijing: Zhonghua Shuju edition, 1959.

Sichuan Daxue Bowuguan 四川大學博物館 and Chengdu Shi Bowu-guan 成都市博物館 (1987). “Chengdu Zhihuijie Zhoudai yizhi fajue baogao 成都指揮街周代遺址發掘報告 (Report on the excavation at the Zhou-period Zhihuijie site in Chengdu).” Nanfang minzu kaogu 1987.1: 171–210.

Sichuan Sheng Bowuguan 四川省博物館 (1959a). “Sichuan Xinfan Xian Shuiguanyin yizhi shijue jianbao 四川新繁縣水觀音遺址試掘簡報 (A brief report on the test excavations at the Shuiguanyin site in Xinfan County, Sichuan).” Kaogu 考古 1959.8: 404–410.

Sichuan Sheng Bowuguan (1959b). “Chengdu Qingyanggong yizhi shijue jianbao 成都青羊宮遺址試掘簡報 (A brief report on the test excavations at the Qingyanggong site in Chengdu).” Kaogu 1959.8: 411–414.

Sichuan Sheng Difangzhi Bianzuan Weiyuanhui 四川省地方志編纂委員會 (1996). Sichuan Sheng Dilizhi 四川省·地理志 (Sichuan Prov-ince—Geography). Chengdu: Chengdu Ditu Chubanshe.

Sichuan Sheng Wenshiguan 四川省文史館 (1987). Chengdu cheng fang guji kao 成都城坊古跡考 (Investigation of the ancient remains of walls and streets in Chengdu). Chengdu: Sichuan Renmin Chubanshe.

Sichuan Sheng Wenwu Guanli Weiyuanhui 四川省文物管理委員會 (1957). “Chengdu Yangzishan tutai qingli jianbao 成都羊子山土台清理簡報 (A brief report on the excavation of the earthen mound at Yangzishan in Chengdu).” Kaogu xuebao 1957.4: 17–31.

Sichuan Sheng Wenwu Guanli Weiyuanhui (1987). “Chengdu Shi’er-


qiao Shangdai jianzhu yizhi diyiqi fajue jianbao 成都十二橋商代建築遺址第一期發掘簡報 (Brief report on the first excavation season of the Shang Dynasty building structures at the Shi’erqiao site in Chengdu).” Wenwu 1987.12: 1–23, 37.

Sun Hua 孫華 (1996). “Chengdu Shi’erqiao yizhiqun fenqi chulun 成都十二橋遺址群分期初論 (A preliminary discussion of the periodiza-tion of the Shi’erqiao site cluster in Chengdu).” In Sichuan Sheng Wenwu Kaogu Yanjiuso 四川省文物考古研究所 (eds.), Sichuan kaogu lunwenji 四川考古論文集 (Collection of essays on the archaeology of Sichuan), pp.123–144. Beijing: Wenwu Chubanshe.

Sun Hua (2000). “Chengdu Pingyuan de xian Qin wenhua 成都平原的先秦文化 (Pre-Qin culture of the Chengdu Plain).” In Sun Hua, Sichuan Pendi de qingtong shidai 四川盆地的青銅時代 (The Bronze Age in the Sichuan Basin), pp. 91–115. Beijing: Kexue Chubanshe.

Wang Yi 王毅 (1991). “Chengdu Shi Ba Shu wenhua yizhi de xin faxian 成都市巴蜀文化遺址的新發現 (New discoveries of the Ba Shu culture in Chengdu). In Li Shaoming 李紹明, Lin Xiang 林向, and Xu Nanzhou 徐南洲 (eds.), Ba Shu lishi, minzu, kaogu, wenhua 巴蜀歷史·民族·考古·文化 (Ba-Shu history, people, archaeology, culture), pp. 295–309. Chengdu: Ba Shu Shushe.

Xu Pengzhang 徐鵬章 (1984).“Woshi Fangchijie faxian gu wenhua yi-zhi 我市方池街發現古文化遺址 (An ancient cultural site discovered on Chengdu’s Fangchijie Street). Chengdu wenwu 成都文物 1984.2: 90–91, and 89.

Address:

18 Shi’erqiao Road, Shudu BoulevardChengdu, Sichuan Province [email protected] (Zhu Zhangyi)[email protected] (Zhang Qing)[email protected] (Wang Fang)


THE ZHUWAJIE BRONZES

BY

SUN HUA 孫華(Peking University)

Translated by Adam Smith, with the assistance of Mark Pittner

Abstract For the first half of the first millennium Bc, the most notable find of elite artifacts in the Chengdu Plain is a pair of hoards of bronzes found at Zhuwajie in Pengxian (now Pengzhou City) in 1959 and 1980. The present paper is a comprehensive study exam-ining and discussing the circumstances and the date of the burials, the styles and the dates of the bronzes, and their casting technology and likely place of manufacture. It also situates the Zhuwajie bronzes in the context of the bronze cultures in the Chengdu Plain as well as the general historical background of the Western Zhou period.

Since the successive discoveries, in 1959 and 1980, of two assemblages of bronzes at Zhuwajie 竹瓦街 in Pengzhou 彭州(formerly Pengxian 彭縣 County), Sichuan 四川 Province (Wang Jiayou 1961; Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981), the bronze objects themselves and the site where they were discovered have both attracted scholarly attention. Researchers have discussed such questions as the date of the bronzes, the date of their deposition, the location of their manufacture, the reasons for their appearance in the Chengdu 成都Plain, and the historical circumstances of their deposition, but they have yet to reach a consensus, and all of these issues remain unresolved (Xu Zhongshu 1962; Feng Hanji 1980; Li Xueqin 1996; Shi Jingsong 1998). The discovery of the two sacrificial pits at Sanxingdui 三星堆, Guanghan 廣漢 Municipality, Sichuan, however, has spurred great progress in our understanding of the archaeology of the Sichuan Basin. With the gradual clarification of the Bronze Age cultural sequence in the Chengdu Plain and the progressive elucidation of the relationship between the Sichuan Basin and surrounding Bronze Age cultures, we are now in a position to enter into a more comprehensive and more profound discussion of the unresolved questions concerning the Zhu-

278 SUN HUA

wajie bronzes—and to attempt to provide more reasonable answers to some of them.1 The approach adopted here is as follows. First, we shall examine systematically and in detail the circumstances of the interment of the bronzes, their stylistic traits, and the casting technology employed in their manufacture. Then we shall compare them with the earlier bronzes from the pits at Sanxingdui and with the later bronzes from the large tomb at Jiuliandun 九聯墩 (also known as Shaiba 晒壩) in Majia 馬家 Township, Xindu 新都 County, Sichuan, and then with the Western Zhou bronzes from the Guanzhong 關中 region of Shaanxi 陝西 Prov-ince. Finally, the results of these comparisons will be placed against the

1 A Chinese version of this paper is also available, see Sun Hua 2002.

Figure 1. Location of the Zhuwajie site in the Chengdu Plain of Sichuan Province.

DujiangyanGuanghan

Chengdu

Longmens

han M

t.

Qionglaishan Mt.

Zhuwajie

50 km

THE ZHUWAJIE BRONZES 279

broader cultural background of the cultural affiliation of the Zhuwajie site, the developments on the Chengdu Plain in the Bronze Age, and the relationships between the Bronze Age cultures of the Sichuan Basin and their neighbors, with a view of providing a reliable reconstruction of the historical context. At the outset, a few words should be said regarding the periodiza-tion of Western Zhou bronzes that is employed throughout this paper. Different approaches for handling the periodization of Western Zhou bronzes from the Central Plains have been adopted, with scholars vari-ously dividing them into two, three, five or six periods (Guo Baojun 1981; Hayashi Minao 1984; Lu Liancheng and Hu Zhisheng 1988, v. 1: 470–529; Li Feng 1988; Cao Wei 1994; Zhu Fenghan 1995, ch. 11). If one considers the stylistic and typological development of the bronzes in their own right, a division into two major periods and six sub-periods is most fitting. “Early Western Zhou” in this article refers to the reigns of kings Wu Wang 武王 (ca. 1049–1053 Bc) through Xiao Wang 孝王 (ca. 872–866 Bc), while “Late Western Zhou” refers to the reigns of kings Yi Wang 夷王 (ca. 865–858 Bc) through You Wang 幽王 (781–771 Bc). The Early Western Zhou period will be further sub-divided into four subperiods, Subperiod I comprising the reigns of Wu Wang and Cheng Wang 成王(ca. 1041–1006 Bc), Subperiod II those of Kang Wang 康王 (ca. 1005–978) and Zhao Wang 昭王 (ca. 977–957 Bc), Subperiod III those of Mu Wang 穆王(ca. 956–918 Bc) and Gong Wang 共王 (ca. 917–900), and Subperiod IV those of Yih Wang 懿王(ca. 899–873 Bc) and Xiao Wang. The Late Western Zhou period can be subdivided into two subperiods, comprising the reigns of Yi Wang 夷王 and Li Wang 厲王 (ca. 858–828 Bc), and the reigns of Xuan Wang 宣王 (827–782) and You Wang, respectively.

The archaeological context

Zhuwajie lies south-west of Mengyang 濛陽 Township in Pengxian, 25 km north of Chengdu. The Chengdu-Dujiangyan 都江堰 railway line runs past the site from east to west. The Mengyanghe 濛陽河 runs three km to the north and the Qingbaijiang 青白江 three km to the south of the site. The two rivers flow westwards, converging east of Wanfu 萬福 Township in the Guanghan Municipality, and at Jintang 金堂 they join the Tuojiang 沱江, one of the four great rivers of the Sichuan Basin. The site is thus located within the Tuojiang river drainage, as is Sanxingdui eight km to the north on the banks of the Yazihe 鴨子河, whereas other localities further south near Chengdu, such as the site of Jinsha 金沙, fall within the Minjiang 岷江 drainage (Figure 1) (Sichuan Sheng Cehuiju and Zhongguo Kexueyuan Chengdu Dili Yanjiusuo 1981: 48).

280 SUN HUA

The 1959 finds at Zhuwajie were located south of the line of the Chengdu-Dujiangyan railway. The 8 bronze vessels and 13 bronze weapons and tools were all placed within a large earthenware vat, which was itself buried in a pit over two meters deep.2 The shape of the pit is unknown, but the fact that the pit, once the earthenware vat and bronzes were placed inside, was filled with fine yellow sand before the pit was backfilled, suggests that the interment was accomplished without pressure of time and with considerable care. By the time archaeologists reached the scene, sherds remained of only about half of the earthen-ware vat that had contained the bronze objects. The vat, a vessel with a large mouth and small base, was said to have been made of gray-brown coarse clay with cord-marking applied to the outer surface.3 The original arrangement of the bronzes within the vat was possibly already unclear and the report makes no mention of this, stating only that “a fine, white clay was smeared onto the bronze ge 戈 dagger-axes, and clearly was deliberately applied at the time of deposition.” It is difficult to imagine why weapons should have been smeared with fine clay. However, we do know that when wooden coffins or burial chambers are buried in a damp clay environment, the clay around the decayed timbers frequently takes on a fine consistency and white color. For this reason one suspects that these bronze ge were originally encased in wooden sheaths akin to scabbards, or perhaps that the ge had been deposited inside one of the bronze vessels together with other wooden objects, and that the decaying wood, subject to the action of the clay and rain water that penetrated the pottery vat, gradually took on a fine, white appearance.4 No attempt was made by Wang Jiayou 王家祐, the author of the report, to clarify for what purpose the bronzes had been buried. Feng Hanji (1980) later

2 In describing the circumstances of the find, the excavation report states, “From an inspection of the trench profile below the level of the road bed, it appeared that the original topsoil had already been removed, and that the remains of the bottom of the large vat that had contained the bronzes were about 2m below the surface of the road bed.” Since the vat was described as being “perhaps about a meter tall,” the mouth of the vessel was likely about a meter below the existing surface at the time of discovery (Wang Jiayou 1961: 28–31). 3 This follows the description of Feng Hanji (1980: 38–47). According to Wang Jiayou’s (1961: 28–31) description, the vat “was of gray-black, undecorated, sand-tempered coarse pottery.” 4 During the Eastern Zhou period in the southern state of Chu 楚, a “sticky white mud” (baigao ni 白膏泥) or “sticky black mud” (qinggao ni 青膏泥) is commonly found outside wooden burial chambers. This is generally thought to be a material delib-erately used to fill and seal tombs. However, in my experience, this phenomenon is associated with tombs subject to the permeation of water below the water table, and is absent from tombs that are kept comparatively dry. This white or black mud, rather than being a deliberate feature of tomb construction, may thus in fact be due to the differential preservation of the wooden burial chamber under different environmental conditions.


suggested that this was a “cache” (jiaocang 窖藏), and, reckoning on the basis of textual accounts that Zhuwajie was a zone of activity of the early Shu 蜀 people, further proposed that the bronzes were buried during the struggles surrounding the political transition in the eighth or seventh century Bc, when Wang Di 望帝 of the Duyu lineage 杜宇氏 is said to have abdicated in favor of Cong Di 叢帝 of the Kaiming lineage 開明氏. The finds of 1980 came to light during the extraction of clay for brick making. The finds were located on the northern side of the Chengdu-Dujiangyan railway line, only 25 m or so northeast of the site of the 1959 discoveries. Four bronze vessels and 15 bronze weapons were found placed within a 1.2 m high earthenware vat, with the vessels packed inside the vat in order of size, largest first and smallest last, and with the weapons placed inside the vessels. The earthenware vat was buried in a trench 3–4 m in width and of a depth not exceeding 2.5 m, and the trench was filled with gray earth. The length of the trench and the form that it took are not clear, but under normal circumstances, a more economical means of burying a vat of round cross-section would be to dig a pit rather than a trench. The interment of the vat of bronzes in a trench thus suggests the possibility that those responsible for the caches were exploiting a pre-existing trench or moat. The gray color of the trench fill suggests that it would have contained ash and other organic detritus of human activity, implying that the locality was not in the wilds beyond the bounds of human occupation, but rather in close proximity to a site of habitation. The excavators reckoned the date of the objects in the assemblage to be “no earlier than Late Shang 商 or Early Western Zhou,” and the nature of their deposition to be “probably a cache.” They further proposed that the date of deposition, and that of the 1959 finds, probably fell around the “end of the Western Zhou or beginning of the Spring and Autumn period,” coinciding with the displacement of the Du Yu lineage as the rulers of Shu by the Kaiming lineage and the subsequent “retreat of the Duyu lineage into seclusion in the Western Mountains,” a conclusion having much in common with that of Feng Hanji mentioned above (Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981). The foregoing is a summary account of the discovery of the two batches of bronzes at Zhuwajie. We now turn to the interpretation of their mode of deposition. Since neither of the two bronze caches at Zhuwajie was discovered by archaeologists, and the original reports give only a cursory description of how they were buried, this is a difficult task. However, understanding the way in which these deposits came into being is a vital precondition if we are to make sense of them.

282 SUN HUA

The two caches were separated by only about 25 m. The depths of the pit and trench in which they were buried were approximately the same, two m or so. In addition, the manner in which the bronzes were deposited was identical in both cases, with each batch being packed into a large-mouth, small-bottom earthenware vat, and with a similar order to the packing. These similarities in mode of deposition, together with the occurrence of similar types, forms, ornaments, and number of the bronzes led the authors of the original reports and later researchers to conclude that the two caches were contemporary and identical in nature, and that both were deliberately interred at the same time in response to some particular emergency—a plausible preliminary conclusion. Note, however, that the two assemblages of Zhuwajie bronzes are made up of both large vessels and small bladed objects. The explanation as caches deposited in an emergency poses no difficulties with respect to the bronze vessels, which were bulky and could not easily be carried off. However, why were the associated weapons not taken away by their owners? The answer is, very likely, that all the weapons in the Zhuwajie caches are of types requiring for their use the fitting of a long wooden shaft. Note that, by contrast, not a single sword was present, even though short-swords (or daggers) with willow-leafshaped blades and a flattened tang (liuyexing bianjing tong jian 柳葉形扁莖銅劍; hereafter “willow-leaf swords”) were in continuous use in the Bronze Age cultures of the Sichuan Basin at least since the later phase of the Sanxingdui Culture and had become the standard weapon type carried by members of the elite by the Western Zhou period. In the cemeteries of the Yu A lineage in Baoji 寶雞 City, Shaanxi Province, for instance, a large majority of tombs yielding bronze weapons included willow-leaf swords (Lu Liancheng and Hu Zhisheng 1988: 463–469).5 In the later Warring States Ba 巴-Shu tombs, swords of this type were even more commonplace and became a highly distinctive regional trait. On the basis of the above, it has been proposed that the willow-leaf sword originated in the Sichuan Basin and was transmitted from there to other regions dominated by the Zhou kingdom during the Early Western Zhou period, but was subsequently little employed in areas other than Sichuan (Jiang Zhanghua 1996). These short-swords are noticeably more portable than weapons requiring a shaft. Their absence from the Zhuwajie bronze assemblages would be most logically

5 The Yu tombs near Baoji are located on the southwestern periphery of the Guan-zhong region, and the design of the tombs, the system of coffins and burial chambers, and the use of bronze ritual vessels all show the deep influence of Bronze Age cultures of the Central Plains. However, each tomb also contains objects—ceramics, undecorated bronze vessels, weapons, and bronze ornaments—that are typical of the Sichuan Basin. This, together with the presence of the duck-headed bronze scepters in the high-ranking tombs, suggests that this may have been the cemetery of a group associated with the bronze cultures of the Sichuan Basin.


explained if one were to imagine that the possessor of the bronzes had deposited a number of unwieldy vessels and surplus weapons, taking their short-swords with them. If so, it would indeed seem plausible that the impulse behind the interment was a sudden event of some kind, and that they were interred with the intention of recovering them later for use. I feel thus vindicated in referring to the two Zhuwajie bronze assemblages as “Cache 1” and “Cache 2” (referring to the earlier and later discoveries respectively). The proximity between the two caches suggests that the bronzes may have belonged to a single elite residence, and interment in two separate locations may have been for added security (were one cache to be dis-covered, the other might still remain intact). It is also possible that the two caches represent bronzes from two distinct but adjacent elite resi-dences.6 After their departure from their estate or town, those responsible for the interment either removed themselves to distant regions or met with misfortune and were subsequently unable to return. The bronzes consequently remained buried right up until their recent rediscovery. The date and background of the interment will be discussed below. The Zhuwajie bronze caches are located between two important ancient sites: Sanxingdui, less than 10 km to the north (Fan Yi 1998), and the site cluster comprising Jinsha (Li Xuemei 2001) and Shi’erqiao 十二橋 in Chengdu, 30 km due south (Sichuan Sheng Wenwu Guan-liweiyuanhui et al. 1987). We know that Sanxingdui was a political center for the Sichuan Basin during the early Bronze Age (the period contemporaneous with the Xia 夏 and Shang dynasties in the Central Plains), and that Chengdu was the capital of the Shu kingdom during the Warring States (Liu Lin 1978). Although the area near the Zhuwajie caches is nowhere claimed to have been a capital in textual and legend-ary accounts, both its geographical location and the archaeological data nevertheless suggest that it was a major settlement during the pre-Qin 秦 period. In 1992, the Sichuan Team of the Institute of Archaeology, CASS, surveyed the site, carrying out excavations at the locality of Qing-longcun 青龍村, approximately 1000 m from the place of discovery of the bronze caches (Zhongguo Shehuikexueyuan Kaogu Yanjiusuo and Chengdu Shi Wenwu Kaogu Gongzuodui, forthcoming). For the time being, we are still unable to assert with confidence that cultural deposits are distributed evenly between the Qinglongcun excavations and the site of the caches, but their proximity and the similar chronological

6 “Elite residence” is simply a catch-all term used until we are able to provide a more specific account of the social status of the individuals concerned. It is possible that these “elite” were the rulers of the ancient state centered on the Chengdu Plain. The size and elaboration of the bronze vessels from Zhuwajie compel us to entertain this possibility.

284 SUN HUA

characteristics of the material recovered suggest that both loci were indeed once part of a single site. If we are correct in so thinking, then the Zhuwajie site would have been of considerable magnitude, a major town and certainly not a minor village.7 At the times of their discovery in 1959 and 1980, archaeologists still knew little of the cultural sequence of the Chengdu Plain prior to the Warring States period and were thus unable to determine the cultural affiliations of the Zhuwajie site and the bronze caches. However, as exca-vation work got under way at Sanxingdui and Shi’erqiao, and following the discovery of the two sacrificial pits at Sanxingdui, an understanding of these issues began to emerge. We are now aware that, between the Terminal Neolithic Baoduncun 寶墩村 and Yufucun 魚鳧村 Cultures and the conquest of Ba and Shu by Qin in 316 Bc, the Chengdu Plain experienced four cultural phases, namely the Sanxingdui Culture, the Shi’erqiao Culture, the Xinyicun 新一村 Culture and the Qingyang-gong 青羊宮 Culture (the latter also known as the “Ba-Shu Culture”) (Sun Hua 2001), and only a few points of controversy remain regarding their developmental progress and chronology (see Jay Xu, this volume). The large pottery vats that housed the Zhuwajie bronze caches were badly damaged, and the original report provided only a rubbing of the surface patterning of a single sherd from Cache 2. Jiang Zhanghua (1998a: 9) has already made an important study of the relationship between this sherd and the known Bronze Age cultures of the Chengdu Plain. He points out that

“The cleanly executed and rectilinear repeating-diamond pattern visible on the rubbing from the pottery vat, and the widely-spaced distribution of the pattern in particular, are identical to the style of the early part of Period II of the Shi’erqiao Culture and clearly distinct from that of Shi’erqiao Period I (Figure 2). The repeating diamonds of Shi’erqiao Period I generally have curved sides and are densely packed so as to interconnect or even overlap. On this basis we may estimate the date of interment of the bronzes as equivalent to the early part of Period II of the Shi’erqiao Culture.”

Elsewhere, Jiang refers to Periods I and II of the Shi’erqiao Culture as the “Lower” and “Upper Shi’erqiao Culture,” which are synonymous, respectively, with the “Shi’erqiao Culture” and the “Xinyicun Culture” as used here (Jiang Zhanghua 1998b). As Jiang correctly points out, the patterning on the vat from the Zhuwajie cache resembles that seen on pottery vats from the earlier stage of the Xinyicun Culture, which suggests a date for the interment of the caches.

7 Wang Renxiang 王仁湘 (personal communication), who has directed the survey of the Zhuwajie site, states that it is of exceptional size.


Further leads are provided by the pottery from Locus Qinglongcun at the Zhuwajie site. There, the vast majority (97%) of pottery was sand-tempered and mostly (over 80%) gray or yellow in color. The vast majority (99%) was plain-surfaced, but a small amount was embel-lished with repeating-diamond, large cross-hatching, cord-impression, or circular patterns. Pottery objects had flat, ringfoot, or pointed bases. The main pottery forms were round-shoulder jar (guan 罐) with short straight neck, large-mouth vat (gang 缸) with constricted neck, “artillery shell”-shaped pointed-bottom cup (bei 杯), and pointed-bottom bowl (bo 缽) with inward-turning mouth and rim (Zhongguo Shehuikexueyuan Kaogu Yanjiusuo and Chengdu Shi Wenwu Kaogu Gongzuodui, forth-coming). By way of comparison, we know that the vast majority of the pottery pieces of the Shi’erqiao Culture were plain, with only a very small number having cord-marking or whorl patterns. Most common are the pointed-bottom types, among which are pointed-bottom bowls with flaring or vertical rim, “artillery shell” pointed-bottom cups, and straight-necked pointed-bottom jars. Additionally, there are straight-neck hu 壺 flasks, round-shoulder jars, gui 簋 basins, and covers with pinched knobs. In addition to these, the earlier material includes several pottery types that persist from Period I of the Sanxingdui Culture: small flat bottom pen 盆, tall-stem dou 豆 and dou-style objects, and hollow-leg he 盉. Xinyicun pottery, for its part, is tempered with fine sand, and is gen-erally of a gray-brown color. There is an increase in the occurrence of cord-marking and a modest amount of overlapping impressed lozenges. The main pottery types are the small-mouth cord-impressed fu 釜, the pointed–bottom zhan 盞 with inward-turning mouth, the pointed- bottom jar (guan) with a swelling body, the tall cord-impressed jar (guan), the round-shoulder vat (weng 瓮) with inward-turning mouth, and the flat-bottom pen with flaring mouth. Of these, the tall cord-impressed jars are the most prevalent and distinctive.

Figure 2: Potsherds from Xinyicun (left) and Zhuwajie Cache 2 (right). After Jiang Zhanghua 1998a: 8, fig. 4. 1.

286 SUN HUA

Comparing the ceramics from Zhuwajie with those already known to us from the Bronze Age cultures of the Chengdu Plain, we find that the composition is distinct from that of both typical Shi’erqiao and typical Xinyicun, but appears intermediate between the two, showing greatest similarity to the Late Shi’erqiao typified by the remains from Fuqinxiaoqu 撫琴小區 in Chengdu (Figure 3) (Unpublished; for pre-liminary information see Wang Yi 1991). The Zhuwajie bronze caches

Figure 3: Zhuwajie site pottery types in comparison with other Bronze Age pottery in the Chengdu Plain. Zhuwajie pottery appears stylistically intermediate between earlier Shi’erqiao Culture pottery (top) and later Xinyicun Culture pottery (bottom), and similar

to Late Shi’erqiao pottery at Fuqinxiaoqu.

Late Shi’erqiaoCulture

Xinyicun site / Xinyicun Culture

Shi’erqiao site

Zhuwajie site

Fuqinxiaoqu site


should thus be placed within the temporal span from Late Shi’erqiao to Early Xinyicun. From the foregoing we may derive the following conclusions: 1) the two Zhuwajie bronze assemblages are indeed “caches,” that is to say, deliberate interments for storage purposes; 2) the Zhuwajie site was a major town of the Chengdu Plain region when the bronzes were buried; 3) the Zhuwajie site belongs to the later part of the Shi’erqiao Culture and possibly continued to be occupied in the early period of the Xinyicun Culture. The two bronze caches should not be divorced from the cultural and chronological background of the site.

The bronze assemblages

With this in mind, we may now embark on a preliminary analysis of the Zhuwajie bronzes. Based on their function, they may be divided into vessels, weapons, and tools; however, as it is sometimes difficult to discriminate accurately between weapons and tools, I shall consider them jointly under the rubric of “bladed implements.”8 Cache 1 yielded 8 bronze vessels and 13 bladed implements. The vessels include five lei 罍, one zun 尊, and two zhi 觶, and all can be categorized as wine vessels, as the report indicates. The five lei form an orderly group, composed of one large, two medium-sized, and two small specimens, with lei belonging to the same category being more or less identical in terms of dimensions, form, and decoration. One may imagine the lei arranged with the largest centrally placed, flanked symmetrically by the medium-sized and small specimens. Feng Hanji (1980: 41) argues that these five lei may be considered to be ‘lei in series’ (lie lei 列罍). Feng uses the term lie lei (“lei in series”) as an analogy to the terms lie ding 列鼎 (“ding in series)” or lie gui 列簋 (“gui in series”), which, in Zhou cultural contexts, refer to sets of vessels of match-ing form and ornament that are either identical in size or of graduated sizes. The five lei from Cache 1 differ from the usual Zhou vessel series in that they feature three different sizes, forms, and decorative designs. Nevertheless, one might still be tempted to say that such sets of five lei, when arranged in the above-mentioned manner, constituted a variant form of “vessels in series.” In fact, the Zhou culture has a rich variety of vessels in series. Besides sets of ding tripods of graduated sizes and sets of gui tureens of identical size, hu containers also occur in matching

8 For detailed illustrations of the Zhuwajie bronzes, see Wang Jiayou 1961; Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981; Feng Hanji 1980; Liu Ying 1983a, 1983b; Sichuan Sheng Bowuguan 1992; Zhongguo Qingtongqi Quanji Bianji Weiyuan-hui 1994.

288 SUN HUA

Figure 4: Bronze lei vessels from Zhuwajie Cache 1.

pairs. The pair-wise grouping of the Zhuwajie lei (excluding the largest) is somewhat reminiscent of the pairing of hu under the Zhou system, and it may be related to the common function and employment of lei and hu, which are both lidded bronze wine containers (Figure 4). Cache 2 yielded four vessels and fifteen bladed implements. The ves-sels include only lei, one large, two medium-sized, and one small. Not only are the numbers and categories of vessels fewer than for Cache 1, but the lei also cannot be grouped symmetrically. The first thing one should ask is whether this latter group of lei also belonged to the same set. The single instance of a five-part set from Cache 1 is not enough to decide the matter, but fortunately Feng Hanji (1980: 41) provides us with a further example:

This is not the first occasion that lei in series have been unearthed in western Sichuan. A set, similarly comprising one large and four small, was brought to light during the War of Resistance Against Japan, and the antiquarians of Chengdu are able to recall it still. Sadly it was dispersed and lost during that time, and the vessels’ whereabouts today are unknown.

It is said that one lei from this set is now in the collection of the Shanghai Museum. To these two instances one may add a much later example of bronze container-vessels occurring in a set of five from the large Warring States tomb at Jiuliandun. We thus have reasonable grounds to declare that the four bronze lei from Cache 2 are an incom-plete set, and that the set originally comprised five members, like the one from Cache 1. For some reason, one of the two small lei was not packed into the large pottery vat (Figure 5). Perhaps the missing lei had been lost during its period of use or during transportation, or perhaps


Figure 5: Bronze lei vessels from Zhuwajie Cache 2.

more likely there was simply no room inside the pottery vat for this last vessel.9 Although only a single specimen of a bronze zun was found in the Zhuwajie caches, this vessel type plays an exceedingly important role in the earlier Sanxingdui Culture. Of the 23 vessels excavated from the two sacrificial pits at Sanxingdui, 14 are zun. Furthermore, the bronze object known as a “spirit altar” (shentan 神壇, K②:296) from Pit 2 at Sanxingdui, when reconstructed appropriately, is akin to a zun in its overall shape: from the bottom upward, it comprises four tiers, the first being a pair of animals, the second four standing figures, the third the bulging central portion of the zun, and the fourth the trumpet-shaped mouth of the zun. The piece viewed as a whole represents a massive sacrificial zun of square cross-section born aloft by weird beasts and dancing spirit-mediums (wu 巫) (Sichuan Sheng Wenwu Kaogu Yan-jiusuo 2000: fig. 129). From Pit 2 at Sanxingdui we also have a small (just

9 The lei from Cache 2 are all larger than those found in Cache 1, and the medium-sized ones also have protruding flanges. In a vat of identical size, they would thus have taken up more space. The excavation report provides some indication of the order in which the pieces were placed into the vat in Cache 2. “At the time of excavation, the large lei ornamented with rams’ heads was placed on its side at the bottom of the vat, the two lei with animal-mask ornament and elephant heads were placed above it, and the small lei with animal-mask ornament at the very top” (Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981: 496). In other words, the vessels were deposited in order of size. Perhaps the vat was full after four lei had been placed within it.

290 SUN HUA

over 10 cm tall) kneeling human figure with a zun on his head (K②:48). The single figure kneels atop a circular altar platform, balancing on his head a zun of the angled-shoulder type bearing an animal-mask motif (Sichuan Sheng Wenwu Kaogu Yanjiusuo 2000: p. 170, fig. 86). These show clearly that the central vessel type among the bronzes of the Sanxingdui Culture was the zun rather than the lei (Figure 6).

Figure 6: A possible reconstruction of the original appearance of the “spirit altar” from Pit 2 at Sanxingdui.


During the period of the Shi’erqiao Culture, the zun began to slip from its preeminent position among the ritual bronzes. The several bronze assemblages from Chenggu 城固 and Yangxian 洋縣 counties in southern Shaanxi, dating to a period transitional between Sanxingdui and Shi’erqiao, include a newly complex array of bronze vessels. Vessels with a Central Plains pedigree increase sharply in number, and although zun remain numerous, they make up only about 20% of the total number of vessels. Lei in the strict sense (we here restrict the definition to vessels with rounded raised shoulders equipped with loops on the shoulder [er 耳] and a lower loop [bi 鼻]) have begun to make an appearance, but they are not numerous, and there is as yet no sign of the lei being of a particular importance among the bronze vessels (Zhao Congcang 1996). By the Qingyanggong (or Ba-Shu) Culture of the Warring States period, the bronze zun is already quite extinct. Although the number of lei discovered in contexts dating to this period is not large, it can nevertheless be seen from a number of phenomena that the vessel type remained of considerable importance. The large “boat-coffin” tombs (chuanguanmu 船棺墓) found at Shangyejie 商業街 in Chengdu (Yan Jingsong and Sun Hua 2001), dating to the Early Warring States period, had been looted at an early date. There were no bronze vessels extant, but the large majority of the pottery vessels that remained were finely made pottery lei. The bronze vessels found in the “waist pit” (yaokeng 腰坑) of the large Middle Warring States tomb at Jiuliandun comprised sets of five and sets of two. The following seven vessel categories were represented by five items each: ding 鼎, fu, mou 鍪, hui F, hu 壺 (of which there were two sets), lei 罍 and pan 盤. From the main chamber of the same tomb was recovered a bronze seal, the design on which was divided into two registers, the uppermost depicting a centrally-placed suit of armor flanked by two erect bells, and the lower a centrally-placed lei flanked by two standing figures. It would appear that at this time, even though lei were now less numerous than hu, the lei was still thought of as the major bronze vessel for sacrificial and ritual use in the Bronze Age culture of the Sichuan Basin (Sichuan Sheng Bowuguan and Xindu Xian Wenwu Guanlisuo 1981: 1–25). The grouping of vessels from the Zhuwajie caches undoubtedly points to a date later than that of the zun-dominated assemblages of Late Shang from Chenggu and Yangxian (that is to say, later than the Early Shi’erqiao period), but earlier than those of the Qingyanggong Culture of the Warring States period when the lei was already declining in its status as a ritual object. The use of five bronze lei appears to be conceptually central to the Zhuwajie vessel group. This is different from the situation in Shang

292 SUN HUA

period Sichuan where the zun was the core of the bronze vessel as-semblage. It differs also from the eclecticism of bronze vessels that is characteristic of Sichuan from the Late Spring and Autumn period on-ward. During this later period, bronze lei introduced from central China continued to be used. At the same time, a great number of Chu-style bronze vessel forms, such as ding, fou 缶, zhan, and dui 敦, were adopted from the Middle Yangzi region, while native Sichuanese innovations like the round-bottomed hu, yan 甗, and mou were current as cooking vessels. This substitution of one wine vessel type for another to form the core of the ritual bronze assemblage suggests that the way in which the users of the Zhuwajie bronzes conceived of them had more in common with the thinking of their Shang period predecessors than with the mindset evident later on in the Qingyanggong Culture. Ge 戈 dagger-axes are the most numerous among the bladed imple-ments. Cache 1 contained eight ge, two yue 鉞 axes, one spearhead (mao 矛), and one ji 戟 halberd, which is a composite of spear and dagger-axe. Cache 2 contained ten ge, two ji and three yue, but no spearheads. An abundance of ge and a relative paucity of spearheads is generally characteristic of the Bronze Age cultures of the Sichuan Basin. Only ge occurred among bronze weapons of ritual function from the pits at Sanxingdui—no spearheads were present (Sichuan Sheng Wenwu Kaogu Yanjiusuo 2000). A collection of bronzes recovered from the site of Pilusi 毗盧寺 in Guanghan comprised four ge and two spearheads (Sichuan Sheng Wenwu Kaogu Yanjiusuo Sanxingdui Gongzuozhan, Sichuan Guanghan Shi Wenguansuo et al. 1992). The bronzes from the Chenggu and Yangxian region in southern Shaanxi included 111 ge and 27 spearheads (Zhao Congcang 1996, chart 1). Bordering on the Zhou cultural sphere, the Yu tombs at Baoji yielded 48 ge, a number which, even if one discounts 76 additional ming qi 明器 ge, vastly exceeds the eight spearheads found with them (Lu Liancheng and Hu Zhisheng 1988: 463–467). In Warring States period tombs of the Qingyanggong Culture, as well, the number of ge still tends greatly to exceed the number of spearheads. Take, for example, the bronzes from the waist-pit of the Jiuliandun tomb. The bronze articles in the pit were mostly in groups of five or multiples thereof, and included 30 ge of four different types, and yue and swords each in a set of ten, but only one set of five spear-heads (Sichuan Sheng Bowuguan and Xindu Sian Wenwu Guanlisuo 1981). We can thus trace the prominence of ge in the Chengdu Plain back to the early part of the Late Shang period, when curved-blade specimens turned up as ritual implements in the Sanxingdui caches. The abun-dance of bronze ge and their mingqi at the tombs of the Yu cemeteries would suggest that the inhabitants of Sichuan and culturally related areas


in the Early Western Zhou period set particular store in the number of ge in a tomb, the number perhaps symbolizing how many warriors wielding such weapons were under the control of the tomb occupant. This observation may be relevant to the understanding of the weapons assemblages from Zhuwajie.

Vessel typology and ornamentation

Lei

Discounting lei of square cross-section, Late Shang and Western Zhou round lei of Central Plains origin come in three basic shapes: flat- bottomed without a ring foot, with a straight-tapering ring foot, and with a ringfoot of complex profile that comprises a tapering and a vertical portion (hereafter referred to as “the double-ringfoot type”). The first of these is a rather primitive form current during the Middle Yinxu 殷虛 period (a sub-period of the Late Shang period). The straight-tapering ringfoot is the standard form for bronze lei for the Late Yinxu and West-ern Zhou periods. The double-ringfoot form, with its complex profile, is of sumptuous execution and associated with rather elevated social status. So far, of the bronze lei of Terminal Shang to Western Zhou date discovered in the distribution area of the Central Plains bronze cultures, almost all are of the standard form with a straight-tapering ringfoot, the exception being a single instance of the sumptuous, complex profile form from Hejiacun Xihao 賀家村西壕, in Qishan 岐山, Shaanxi Province. This perhaps reflects the fact that no Western Zhou elite tombs of the very highest rank have yet been found intact in that area. The nine Zhuwajie lei, as well, include no examples of the early type lacking a ringfoot. The other two principal forms are, however, both represented; they will here be referred to as Type I and Type II.

Type IOf comparatively large size, elaborate decoration, and double-ringfoot profile, this is the most numerous type of lei at Zhuwajie, comprising all examples except two from Cache 1. The form of the cover can serve to further divide Type I lei into four subtypes.10

10 This procedure is admittedly somewhat arbitrary, as subtypes could be defined by a number of alternative criteria. For example, one could take the construction of the body as one’s basis and divide them into two categories: the lei with rounded shoulders and bodies, and those lei with straighter bodies. Or, taking the loop handles er as the basis, one could differentiate between lei with handles that stick up and those that are comparatively flat.

294 SUN HUA

Subtype I.1. Characterized by a cover with a ring-foot style knob, only the largest example from Cache 1 belongs to this subtype. The domed cover rises in the manner of an inverted bowl. The body of the ves-sel is elongated, is nearly flat above the shoulder, and has a sudden, almost angular turn between shoulder and belly. The high ringfoot is smaller in diameter than the mouth of the vessel. The two loops at the shoulder are largish and with an upward turn, each embellished with a forward-curling hook, while the semicircular loop low on the belly is adorned with a bovine head. The surface of the object is smooth and plain, but the cover, neck, belly, and ringfoot each bear four vertical flanges dividing the object into four equal segments. The flanges on the cover and belly are F-shaped in profile. The surface of the cover has four raised whorl motifs, and six more appear on the shoulder. With a height of 68 cm, this lei is imposing but lacks a sense of stability and has a somewhat rigid profile. Nevertheless, the simply executed relief animal heads and well-spaced whorls set against the unadorned surface, as well as the crisply-handled flanges, combine to give a well-balanced match of ornamentation to form.

Subtype I.2. Having a beast-head knob on the cover, the largest and small-est lei from Cache 2 are of this subtype. The larger of the two, which stands 79 cm tall, is both the largest bronze from Zhuwajie and the largest lei discovered to date.11 Unlike the elongated form of the large lei from Cache 1, this subtype has a rounder and more handsomely solid appearance. The domed cover has a slight curvature, and the upper part of the massive knob on top is composed of four beast masks. The top of the knob also has a coiled dragon design with the head orientated inwards. The surface of the cover bears “nose-shaped” elaborations (actually, compressed versions of the standard animal-mask motif) on the front and back sides, while to either side in profile is a bovine creature with its head bent inwards and its tail bent outwards. The bovine kneels on its front legs as though rising from the ground. There is a consider-able “lift” to the round shoulders of this full-bodied vessel form, and the large loop handles extend outwards almost horizontally, unlike the vertically rising examples more commonly seen on lei. A bovine head in relief tops a pair of loop handles, and the body behind divides in two, continuing in relief on the shoulder at either side. Further bovine

11 Generally speaking, lei of square cross-section are taller than round lei, since square lei were more prestigious than round ones. Yet the largest square lei discovered so far, in the collection of the Nezu Institute of Fine Arts (Nezu Bijutsukan 根津美術館) in Tōkyō 東京, measures 66.5 cm in height, shorter not only than the large lei from Zhuwajie Cache 2 but also than the large lei in Cache 1 (Hayashi 1984: 291 #24). If one adopts the system of prestige marking used in the Central Plains, the items from the Zhuwajie cache would rank very highly indeed.


heads in relief are placed front and back on the shoulder, and together with those at the side all but cover the whole of the shoulder zone. If we suppose that this lei, over half a meter in height, were to be placed on the ground, and its users to stand or seat themselves before or around it, it would be the richly ornamented shoulder, cover, and knob that would meet their gaze, whereas the unadorned neck and belly of the vessel would lie largely outside the field of view. This emphasis on the decoration around the shoulder is characteristic of bronzes of late Shang and early Western Zhou date, and the rendering of animals with heads modeled in the round and bodies done in shallow relief was also cur-rent during the period. The closest comparative example in the Central Plains tradition is the bronze lei with bovine ornament from Shanwanzi 山灣子 in Kalaqin Left Banner 喀拉沁左旗 (commonly abbreviated as Kazuo 喀左), Liaoning 遼寧 Province (Kazuo Xian Wenhuaguan et al. 1977: 23–27, figs. 2–3). The size of the Shanwanzi piece does not reach half that of the Zhuwajie example, and the form and manufacture are inferior, but the treatment of the ring handles is very similar to the bovine decoration on the Zhuwajie piece (Figure 7). Some scholars have wondered whether these two bronzes might be somehow related, given that decoration of this kind has not been seen in China’s central region

Figure 7: Comparison of lei from Zhuwajie and Kazuo, Liaoning. Left: the largest lei from Zhuwajie Cache 2, height 79 cm. Right: a lei from Kazuo Shanwanzi. Not to same scale. Left: from Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981: fig. 2. Right: from Kazuo Xian Wenhuaguan

et al. 1977: fig. 2.

296 SUN HUA

but nevertheless appears in both the northeast and southwest (Hirokawa 1997). In fact, as early Western Zhou tombs and caches of the highest rank have been only seldom encountered in the central regions, it is to be expected that bronze lei, which only occur in elite bronze assemblages anyway, should also be met with rather rarely, and it is therefore not strange that examples of this bovine decoration have not been found. The lei from Shanwanzi was excavated from a cache that contained a total of 22 bronzes, including examples from the early state of Yan 燕, such as the Bo Ju yan 伯矩甗. Apart from the Shu Yin fangding 叔尹方鼎, which is slightly later, being transitional between Subperiods II and III of the Early Western Zhou, the other bronzes in the Liaoning cache all belong to Subperiods I and II. From the point of view of typological seriation, the large lei with bovine decoration from Cache 2 at Zhuwajie could only be chronologically prior to the Shanwanzi piece and could not be later. This much will likely be obvious to anyone versed in the patterns of formal evolution of Shang and Zhou bronze lei. The lid of the smallest lei from Cache 2 is generally similar to that of the largest lei in form and decoration, but the body matches that of Subtype I.4 with dragon finial to be discussed below. Lothar von Falkenhausen (2001: 183) suspects that the lid does not match the body, and this is quite likely.

Subtype I.3. Featuring a mushroom-shaped knob on the lid, this subtype is represented by the pair of medium-sized lei from Zhuwajie Cache 2. The bodies of these two lei are basically the same in form and decora-tion. Both are of elongated form and overlaid with emphatic and lavish decoration, giving an impression of great splendor. However, there are clear distinctions between the two examples in the shape, decoration, and flanges of the lid. To begin with the features in common, the necks of these two lei are the tallest and narrowest among all of those from Zhuwajie. In each case, the shoulder is rather narrow and smoothly rounded. The tallish ring-foot is of larger diameter than the mouth of the vessel, giving an impression of stability. The large loops on the shoulder are adorned with elephant-like beast heads. The perforation through each loop is almost entirely closed off. The two loops rise to a considerable height, and the horizontal horns and vertical flange on each of the beast heads reaches almost as high as the bottom of the lid flange. Placed on the front and back at the shoulder are further three dimensional elephant-like beast heads, with pairs of horns and ears spreading out to either side, forming a sharp contrast with the exaggeratedly prominent flange running down the center of each head (the lower section of the flange is interpretable as the elephant’s trunk). The animal-head knob on the reverse side of


the belly is rather small, and the corresponding flange runs down only as far as the top of the animal-head knob, whereas the remaining belly flanges run right down to the bottom, coming almost into contact with those of the ringfoot. The flanges are highly varying in shape, from free at the top to tight and orderly below. The flanges on the shoulder join with the trunks of each elephant head to form an asymmetric unit. The flanges on the belly consist of pairs of connected hooks turned in opposite directions, as do the flanges on the legs. Except for the neck, the body of each lei is adorned with animal designs against a ground of dense spirals, with separate registers for shoulder, belly, and foot. Each of the four shoulder panels between pairs of flanges contains the coiled body of a dragon (kuilong 夔龍) in shallow relief, complemented by kuilong heads turned away from the coils. The decorative scheme on the belly is centered on the front and back flanges, with pairs of large spiral-bodied kuilong facing one another on either side of a central “nose-like” orna-ment.The arrangement on the ringfoot resembles that of the belly, with pairs of reclining bovines facing one another across the central axes of the front and back flanges. We turn now to the features that distinguish these two lei from one another. The first lei has a kuilong pattern on the lid, the surface of which gathers in a natural manner around the knob. Four exaggerated avian flanges produce a four-fold partition of the lid surface, each of which holds a coiled-bodied kuilong on a background of dense spirals. The maximal diameter of the body is somewhat larger than that of the second lei. The second lei has a lid that is almost flat, and it is decorated with beast-mask designs. The four flanges on the lid perhaps represent birds, but they are highly abstracted. The decoration on the lid is organized around the axes of the front and back flanges, with a curled-horned beast mask against dense-spiral ground at the front and another at the back. Each beast mask has bulging eyeballs and clearly delineated eye-sockets, while the beast’s two horns curl outwards and forwards, the tips protruding from the surface of the lid. This is typical for elite bronzes during Subperiods I and II of the Early Western Zhou. Comparable examples include the Yu Bo A伯 double-handled gui on a square plinth from tomb 1 at Zhifangtou 紙坊頭, Baoji City (Shaanxi); the Bo Ge zun 伯格尊 and Bo Ge hu with carrying handle from tomb 7 at Zhuyuangou 竹園溝 Baoji (Lu Liancheng and Hu Zhisheng 1988: 27, fig. 19; 102, fig. 79; 105, fig. 83); the He zun 何尊 unearthed at Jiacunzhen 賈村鎮, Baoji (Tang Lan 1976: 60–63, fig. 1); and the Bo Ju li 伯矩鬲 from tomb 251 at Liulihe 琉璃河, Fangshan 房山 County, Beijing (Beijing Shi Wenwu Yanjiusuo 1995: 156–159 fig. 94). The dates of all of these bronzes (and of other historically handed-down pieces with same type of beast-mask ornament) fall within the time range of Subperiod I to early

298 SUN HUA

Subperiod II of the Early Western Zhou. Elite wine vessels of earlier or later date do not display this highly date-specific decoration.

Subtype I.4. This subtype has a finial on the lid in the form of a rising dragon. It is represented by the two middle-sized lei from Cache 1. The body of the small lei from Cache 2, despite its differing lid, may also be assigned to this subtype. The two middle-sized lei are largely the same in form and in the ar-rangement and content of their decoration. Each has a rising-dragon lid, dragon-loop handles, a bovine-head lower knob, pairs of coiled-bodied kuilong facing one another across beast heads placed centrally front and back on the shoulder, four large beast-mask designs on the belly centered at 45 degrees away from the loop handles, and two pairs of kuilong with bent bodies on the ringfoot. The major decoration is set against a ground of dense spirals. Despite the similarities, there are clear differences between the two items. One vessel features a rather flat lid, and the rising dragon has bottle-shaped horns and legs that extend forwards. The lid of the second example slopes down towards the rim, and its dragon has fan-shaped horns and legs that extend sideways. The first lei has a longer neck, a narrower band dividing the registers on the shoulder and belly, and a ringfoot in which the upper tapering portion has a slightly concave profile. The second lei has a shorter neck, a broader band separating the shoulder and belly registers, and a ringfoot on which the tapering portion is slightly convex in profile. The loops on the shoulder of the first lei are equipped with ring handles, are without flanges, and are fashioned so that the beast above and the pendant lobe below are entirely independent. The loops on the second lei lack rings but have “dorsal fin” flanges, and the pendant lobe is a natural extension of the loop rather as though it were the tail of the kuilong. The ornamentation on the first lei is distinguished by a bodiless ram’s head in relief at the shoulder, and by the rather larger size of the beast-mask designs on the belly, where they occupy all of the available space. The second lei had instead a bovine head on the shoulder (now missing), with the two flanks of the animal opening out to either side in shallow relief; the beast masks on the belly are smaller, but the nose of each mask is elaborated with inverted triangles and inverted kuilong to either side (Figure 8.1). The small lei from Cache 2 is narrow at the shoulder, somewhat elon-gated in overall form, and has a ringfoot intermediate in form between those of the two lei just discussed. The form of the shoulder loops and the decoration on the belly match those of the lei with a fan-horned dragon on the lid. The decoration on the shoulder matches that on the lei with bottle-horned dragon.

Figure 8: Decorative designs on the body of the dragon-lidded lei vessels from Zhuwajie (top) and Kazuo (bottom). Not to same scale.

300 SUN HUA

The form and decoration of these lei, and of the lids in particular, are rather unusual. The closest comparative example was discovered at Beidongcun 北洞村 in Kazuo, Liaoning, and is a particularly good match for the fan-horned dragon example from Zhuwajie (Figure 8.2) (Liaoning Sheng Bowuguan 1974). In each of the three examples from these two locations, the lid is in the form of a coiled dragon raising its head and fore parts up from the surface with the support of two spindly forelegs. The dragon serves both as ornament and as a handle for the lid. If we put the lid aside and consider only the body of the vessel, then comparisons may be drawn with the small lei from Zhuwajie Cache 2 (as mentioned above), and with the above-mentioned example from a tomb at Hejiacun Xihao, Qishan (Shaanxi Sheng Kaogu Yanjiusuo et al. 1979: fig. 69). Besides the lei with a dragon lid, the cache at Kazuo contained five further bronze objects. Apart from a bowl-shaped object with spout, the date of which is hard to determine, the others all date to the last years of the Shang and the early years of the Western Zhou period, and the date of the dragon-lidded lei is likely to be within the same range.12 The objects found together with the lei from Hejiacun Xihao are all typical for the early phase of the Early Western Zhou, and at the latest do not post-date the reign of Kang Wang. The Hejiacun lei should also date to around the reign of the Kang Wang.13 Comparing the Zhuwajie dragon-lidded lei (including the small-sized lei from Cache 2 with the beast-knobbed lids) with the examples from Beidongcun and Hejiacun Xihao, the Zhuwajie specimens are stylistically later than the Beidongcun dragon-lidded lei and rather close to the example from Hejiacun. Thus it is most likely that the lei with coiled-dragon lids from Zhuwajie date to late within the Early Western Zhou Subperiod II, i.e., to the reign of Kang Wang.

Type II Lei vessels of this type are of smaller size, with a smooth and unadorned surface, and with straight-tapering ringfeet. Only the two small lei from Cache1 belong to this type. They are characterized by a ringfoot-shaped knob on a domed lid (but note that the lid is less domed than on Subtype I.1), a high raised shoulder, large loop handles, and a tallish ringfoot. These two lei are largely similar but with significant differences. The first has a rather soft turn from the shoulder to the belly, four bulging

12 Li Xueqin (1990: 46–53) believes that the assemblage from Kazuo dates to the beginning of the Western Zhou, no later than the reign of Kang Wang. Zhu Fenghan (1995: 795) believes that two pieces date to the “late Shang period” and three pieces to “the early part of Early Western Zhou.” 13 Li Feng (1988: 416) places the Hejiacun Xihao lei in Early Western Zhou Subperiod II, corresponding to the Cheng Wang and Kang Wang reigns.


whorl motifs on the lid and again on the shoulder, a three-dimensionally modeled bovine head front and back on the shoulder, and a pair of fine raised lines separating the shoulder from the belly, the lower of which is overlaid by the loop handles. The second lei has a rather more abrupt turn in the profile between shoulder and belly, the lid and shoulder each have six bulging whorls rather than four, the shoulder and belly are separated by a single raised line elaborated with key-shaped motifs, and the two loop handles are placed so as to bridge over this raised line. Several specimens of lei of this type and decoration are known from contexts within the Zhou domain and date to the early part of the early Western Zhou period. The most important of these must surely be the Yan Hou Ke lei 燕侯克罍 from tomb 1193 in the Yan cemetery at Liulihe, near Beijing (Zhongguo Shehuikexueyuan Kaogu Yanjiusuo et al. 1990: 24, fig. 3.1, plate 2.2). Apart from the added roundness of the shoulder and the slightly smaller loop handles, all other features are exceedingly similar to the second of the two lei just described. As the Yan Hou Ke lei is inscribed with the Zhou king’s edict to the figure known by the title Taibao 太保, and the order to “enfeoff Lord Ke at Yan,” the vessel may be dated to the reign of Cheng Wang. A similar lei was found in Tomb 1 at Zhifangtou, Baoji, a tomb which is also estimated to date to the reign of Cheng Wang (Lu Liancheng and Hu Zhisheng 1988: 36, fig. 28.1). Therefore, the two small lei from Zhuwajie Cache 1 should be dated no earlier than the reign of Cheng Wang and no later than the reign of Zhao Wang (Figure 9).

Figure 9: Similar decorative patterns on lei vessels with six whorls on shoulder from Zhuwajie Cache 1 (left) and the Yan Hou Ke lei from Tomb 1193 at Beijing

Liulihe (right). Not to same scale.

302 SUN HUA

Zun 尊

Only a single vessel from Zhuwajie is classifiable as a zun. It has a flaring mouth with a rim diameter larger than that of the ringfoot. The neck is rather tall and straight, and the belly swells outward slightly. The high ringfoot is in two segments, of which the lower segement has vertical sides and is rather tall. The neck and foot are without decoration save for a pair of raised lines at the interface with the belly, which serve to mark the boundary of the decorative scheme on the belly. The front and back of the belly each have a large beast mask against a ground of dense spirals. The beast masks extend no further than a depiction of the head, with a broad brow, large curling horns, small ears, eyebrows above the socketless eyeballs, and teeth just visible within the mouth. The beast mask is flanked by simplified kuilong with heads turned back. This was the most popular zun type during the Early Western Zhou, and the use of large beast masks on the belly was also the fashion of the period, as can be seen from a number of comparative specimens. Of the three zun of this shape excavated from the Early Western Zhou cemetery of the Ge 戈 polity at Gaojiabu 高家堡, Jingyang 涇陽 County, Shaanxi Province, two had large beast masks on their bellies (the third had spiral-bodied kuilong and large beast masks on its ringfoot) (Shaanxi Sheng Kaogu Yanjiusuo 1995: 21, 48, 83, figs. 16, 40, and 68). Of the four zun of this shape from the Yan cemetery at Liulihe, three have large beast masks on the belly (Beijing Shi Wenwu Yanjiusuo 1995:174–182). The horns and eyes of the masks, and the subsidiary animal motifs to either side, vary widely on different specimens. Some examples clos-est to the Zhuwajie zun in size, form, and decoration include the Hei zuo Mu Xin zun 潶作母辛尊 from the Baicaopo 白草坡 cemetery in Lingtai 靈臺 County, Gansu 甘肅 Province (Gansu Sheng Bowuguan 1972: 3 fig. 4.19), and the Zu Ding zun 祖丁尊 from Tomb 20 at Yun-tang 云塘 in Fufeng 扶風, Shaanxi Province (Shaanxi Sheng Kaogu Yanjiusuo et al. 1979: fig. 69), both of which date to the early phase of the early Western Zhou. However, as Li Xueqin (1996: 118–122) has observed, a composition involving a beast mask flanked by kuilong with turned heads appears again on the Li gui 利簋 and on the historically handed-down De gui 德簋. The Li gui may be dated on the basis of its inscription to the reign of Wu Wang or Cheng Wang,14 and the De gui is generally reckoned as being a Cheng Wang period piece (Guo Moruo

14 The Li gui inscription records a gift to a functionary, Li, after the “campaign against the Shang by Zhou Wu Wang.” The title “Wu Wang” used in the inscription is likely a posthumous appellation, meaning that the vessel should be dated to the reign of Cheng Wang.


1959: 1–2). For this reason, Chen Gongrou and others who have made studies of the Shang and Zhou beast mask assign this type of beast mask to the period of Wu Wang and Cheng Wang (Chen Gongrou and Zhang Changshou 1990: 137–168). On this account, it would be quite reasonable to attribute the Zhuwajie zun with beast mask ornament to the reign of Cheng Wang or thereabouts. Since the form and decoration of this zun show no sign of any local traits proper to the Sichuan Basin, Feng Hanji (1980) and others believe that this item was transported to Sichuan sometime after being manufactured elsewhere. The likelihood of this is of course very high. However, neither the date of manufacture of the zun nor the date of its transmission to Sichuan are likely to have been as early as Feng suggests.

Zhi 觶

Two vessels from Cache 1 belong to the category of zhi with oval cross-section. The two may be classified separately on the basis of the form of the ringfoot.

Type IThe first is the Tan Fu Gui zhi 覃父癸觶 with segmented ringed foot. Its elongated form comprises a tall neck, a gently rolling belly, and a ringfoot, the lower part of which flares outwards. The surface is smooth and unadorned, save for a band of interlocking spirals on the upper shoulder. An inscription of three graphs on the base reads Tan Fu gui 覃父癸 (“Father Gui of the Tan [lineage]). The form and ornamentation of the zhi are typical for the early phase of the Early Western Zhou. The zhi from Tomb 61.106 at Zhangjiapo 張家坡, Chang’an 長安 County, Shaanxi (Zhao Yongfu 1984: Plate 2.3), and the Bao Ding Fu zhi 保丁父觶 from Tomb 4 at Gaojiabu are both of this form (Shaanxi Sheng Kaogu Yanjiusuo 1995: 100, fig. 83.2).15 These zhi also have a band of interlocking spirals on the shoulder (this decoration sometimes also oc-curs on the ringfoot) in the manner of the Zhuwajie example, and are likely to be of similar date. Xu Zhongshu (1962) proposed that the Tan of Tan Fu Gui was the name of an eastern clan, possibly the anteced-ents of the later Tan guo 譚國 polity. Whether this view is correct or not need not trouble us here, but we can be sure that Tan is a lineage name and that the Tan Fu Gui zhi was made by a member of that lineage

15 Lu Liancheng and Hu Zhisheng (1988: 470–529, ap. 1) point out that bronze zhi of this form “were distinctive and commonly-occurring drinking vessels current from the Late Shang through the first two reigns of the Western Zhou.”

304 SUN HUA

for a deceased father whose temple name was Gui. This name recurs on the unprovenanced Tan Fu Gui ding 覃父癸鼎 and on the Tan Fu Gui jue 覃父癸爵 excavated from Tomb 13 at Zhuyuangou, Baoji (Luo Zhenyu 1936: 2.29; Lu Liancheng and Hu Zhisheng 1988: 66, fig. 54). The form of the Tan Fu Gui ding is unknown, but features of the jue, such as a spout that is deep to the fore and shallow to the rear, and a pair of rim posts that have been moved back to a point between the handle and spout, are indicative of a date early in the Early Western Zhou period. The original report dated Tomb 13 at Zhuyuangou to “early in the reign of Kang Wang” and declared that the tomb occupant “was active mainly during the later years of Cheng Wang and the early years of Kang Wang” (Lu Liancheng and Hu Zhisheng 1988: 263). The Tan Fu Gui of the jue and of the zhi from Zhuwajie are likely the same individual, and the Zhuwajie piece should be dated to the reigns of Cheng Wang and Kang Wang, and not, as some have suggested, to the Late Shang period.

Type IIThe second zhi from Cache 1, the Mu Zheng Fu Yi zhi 牧正父乙觶, has a segmented ringfoot (Figure 10). It is squatter in appearance than the Tan Fu Gui zhi, with a shorter neck and a dramatically sagging belly. The wall of the ringfoot is nearly straight and has a protruding lip at the rim. There is a band of animal motifs against a dense-spiral ground at the interface between the neck and shoulder, and another such band is found around the ringfoot. The decorative band at the neck is centered on protruding tiger heads to the front and back, flanked by four paired kuilong with their heads turned back, the two kuilong on the inner side having two feet and those on the outer having only a single foot. The band on the ringfoot is centered on the front and back flanges, with spiral-and-eye patterns on either side. The decoration on this zhi is highly specific to around the reign of Zhao Wang of the Western Zhou. In the earliest years of the Western Zhou, during the reigns of Wu Wang and Cheng Wang, this type of slim-bodied, abstract kuilong with turned-back head was seldom seen, but it then became current around the reign of Zhao Wang. This is demonstrated by the following. The chronological sequence of the three large-scale, well-preserved Yu Bo A伯 tombs in the cemetery at Zhuyuangou, Baoji, Shaanxi, starts with Tomb 13 as the earliest, followed by Tomb 7 and finally by Tomb 4, and the kuilong with head turned back only appear on the bronzes of Tomb 4, dating to late in the reign of Zhao Wang. They are absent by the time of Tombs 1 and 2 at Rujiazhuang 茹家莊, which date to late in the reign of Mu Wang (Lu Liancheng and Hu Zhisheng 1988). The situation at the cemetery of the the Jin 晉 polity at Tianma-Qucun 天


Figure 10: Comparison between the Mu Zheng Fu Yi zhi (top) and a zhi from Tombs 6214 and 6231 at Tianma-Qucun.

馬曲村, Quwo 曲沃 County, Shanxi 山西 Province, is the same: a sim-plified kuilong with its head turned back appears regularly on bronzes from Tombs 6069, 6210, 6214, and 6231 (Beijing Daxue Kaoguxuexi Shang Zhou Zu and Shanxi Sheng Kaogu Yanjiusuo 2000), but at the cemetery of the Marquises of Jin at Beizhao 北趙, which belongs to the same site and which started to be utilized during the reign of Mu Wang, this motif is already absent on the bronzes from the earliest pair of tombs (Beijing Daxue Kaoguxuexi and Shanxi Sheng Kaogu Yanjiusuo 1994: 4–34). Thus we can say not only that the Zhuwajie Mu Zheng Fu Yi zhi cannot be dated as early as the late Shang period, but that it also post-dated the Zhuwajie Tan Fu Gui zhi and must be assigned to the reign of Zhao Wang or thereabouts. The bronze vessel form customarily termed zhi by archaeologists and bronze scholars belongs to the general class of small vessels for alcohol consumption. Small wine vessels were an indispensable element of the ritual bronze assemblages in the greater Central Plains region prior to the Late Western Zhou. The Bronze Age cultures of the Sichuan region, on the other hand, whether of Late Shang date or of the Spring and Autumn and Warring States periods, essentially lacked such vessels. The two zhi from Zhuwajie are unlikely to have been of any great importance

306 SUN HUA

among the bronze vessels and ritual activities of the region, and their appearance here is likely no more than a reflection of the intimate rela-tionship between the bronze-using cultures of the Chengdu Plain and that of the early Zhou culture centered on the Guanzhong region.

Typology and ornamentation of the bladed implements

Ge 戈 (dagger-axes)

As mentioned, ge are the most numerous among the bronze weapons associated with the Zhuwajie caches and with the bronze cultures of the Sichuan Basin in general. The caches at Zhuwajie yielded eighteen specimens, all of which have a straight tang (na 內), and a blade lack-ing the downward elongation (hu 胡) seen in many Late Bronze Age ge, although a faint, initial intimation of such an elongation is visible in a few cases. A typology of these ge was provided in the original preliminary reports as well as in some later research articles; these categorizations will here serve as a foundation for further analysis. The ge would have been hafted to a wooden shaft by means of the protruding crosspiece (lan 闌) and adjacent perforations (chuan 穿). The shape of the crosspiece, as well as the position and number of perforations, should be among the primary variables by which the ge are classified, in addition to the criteria put forward in the previous typologies. Based on this standard, we can first divide the bronze ge from Zhuwajie into two types (Figure 11).

Type IGe of this type have a long, narrow blade from which the lan crosspiece protrudes, tooth-like, upward and downward; when viewed in profile, however, the crosspiece is indistinct. It was principally by means of these protrusions on the upper and lower rim of the blade that the ge would have been hafted to a long shaft with cords. Specimens of this type from Zhuwajie have a perforation in the bottom half of the rear part of the blade. Based on the shape of that part of the blade, they can be subdivided into two subtypes.

Subtype I.1. The top and bottom halves of the rear part of the blade are symmetrical. The two ge with protruding crosspieces unearthed from Cache 1 both belong to this type.

Subtype I.2. The bottom portion of the rear of the blade extends slightly downward, widening that part of the blade—an anticipation of the much more prominent hu portions of later specimens. The three ge with protruding crosspieces unearthed from Zhuwajie Cache 2 all belong to this subtype. Their blades bear designs of birds with hooked beaks


and long, awl-shaped bodies. At the Yu cemeteries in Baoji, ge with the same highly distinctive traits were found in Tomb 18 at Zhuyuangou (Li Xueqin 1996: 121), which dates to the transition from the Early Western Zhou I to II periods (i.e. between the reigns of Cheng Wang and Kang Wang); the Zhuwajie specimen should be dated likewise (Lu Liancheng and Hu Zhisheng 1988: 223, fig. 162.2–3).

Type IIThe lan crosspieces of ge of this type do not protrude upward and down-ward. Instead, these specimens have short, wide blades that are nearly triangular in outline. The crosspiece is usually quite prominent when viewed in profile. Fixing of the blade to the handle would have depended mainly on the hafting cords running through the perforations. Based on the characteristics of the perforations in the back part of the blade, this category can be subdivided into two subtypes, which in turn fall into variants and subvariants.

Subtype II.1. This type is characterized by rectangular perforations at the top and bottom of the crosspiece. Based on whether or not there is top-bottom symmetry with respect to the midrib of the blade, this shape can be further subdivided into two variants.

Variant II.1.A. Ge of this variant are characterized by symmetrical top and bottom halves. A ge with an owl-shaped animal-mask design found at the bottom of Cache 2 belongs to this variant. This ge is similar to a ge

Figure 11: The author’s classification of ge from the Zhuwajie site. Labels follow the hierarchical categories (Type).(Subtype).(Variant).(Subvariant).

I.1

I.2

II.1.A

II.1.B

II.2.A.a

II.2.A.b

II.2.B.a

II.2.B.b

II.2.C

308 SUN HUA

from Tomb 4 at Zhuyuangou (M4: 105), with the single exception that the animal head on the Zhuyuangou ge is in profile, rather than forward-facing (Lu Liancheng and Hu Zhisheng 1988: 162, fig. 125.4).

Variant II.1.B. This variant is characterized by its slightly downward hooking blade. A ge from Cache 1 with two sets of parallel lines on the back of the tang belongs to this variant. This type of ge does not occur at the Yu cemeteries.

Subtype II.2. This type is characterized by two rectangular perforations on the blade near the side rim and one round perforation on the midrib at the back end of the blade. Another round or drop-shaped perforation appears on the tang. Ge of this shape are the most numerous and exhibit the greatest degree of variation. Based on the width of the blade, this shape can be subcategorized as follows.

Variant II.2.A The body of this variant of ge is relatively long. There are eight specimens, which can be further subdivided into two subvari-ants based on the presence or absence of symmetry with respect to the midrib.

Subvariant II.2.A.a. This subvariant is characterized by top-bottom symmetry. Two such ge were found in Cache 2.

Subvariant II.2.A.b. This subvariant is characterized by blades that hook downward slightly. Two examples were found in Cache 1 and four in Cache 2. Similarly shaped ge of both subvariants were unearthed at the Yu cemeteries.

VariantII.B.2. Ge of this variant are relatively short and broad, almost triangular in shape. There are only two examples. Based on the same criteria as those used for Variant II.2.A, this type can also be subdivided as follows.

Subvariant II.2.B.a. This subvariant is characterized by top-bottom symmetry and by the small angle formed by its upper and lower edges, both of which are relatively straight. Ge of this subvariant do not bear any surface decoration, but they are similar in shape to a ge decorated with an animal mask from Hejiacun, Qishan County, Shaanxi, which date to the later part of the Late Shang period (Shaanxi Sheng Bowuguan and Shaanxi Sheng Wenwu Guanliweiyuanhui 1976: 31–38, fig. 8.4). Sty-listically, these are the simplest among the ge unearthed at Zhuwajie.

Subvariant II.2.B.b. This subvariant is characterized by a slightly downward-hooking blade, the edges of which are concave. The bronze ge from Cache 2 that are decorated with tiger-head patterns on the back end of the blade are of this subvariant.


Variant II.2.C. This variant, characterized by an extremely broad blade, is a typical example of triangular-blade ge. The one undecorated ge with triangular blade unearthed from Cache 2 belongs to this vari-ant. The arc-shaped edges of the blade come together at a wide angle. The rim is slightly arched toward the point of the blade, and the tang is short and broad. This ge is similar in shape to two ge (M3: 3 and M8: 16) unearthed at Zhuyuangou (Lu Liancheng and Hu Zhisheng 1988: 432, fig. 276.7–8) and can be dated later than a ge from Tomb 1 (M1: 11) at Hejiacun (Shaanxi Sheng Bowuguan et al. 1976: 35, fig. 8.1), which dates to the later part of the Late Shang. On the other hand, they are quite far removed from a similarly shaped ge found in the large Warring States tomb at Jiuliandun (Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981: 15, fig. 38.1).

Mao (spearheads)

Only one spearhead was found at Zhuwajie. Its body is long and broad, with a protruding midrib, while the tubular stem is short and narrow. On each of the two sides of the stem is a small ring. Both sides of the spear are decorated with the same pattern: a lizard-like animal on the stem biting a beetle-like insect on the blade. This type of spearhead is extremely rare; in fact, it has not been observed before in either the central region of China or in the Sichuan region. Li Jianmin (1996: 78–87) has compared it with the bronze spearheads from the large tomb at Dayangzhou 大洋洲, Xin’gan 新干, Jiangxi 江西 Province, and based upon similarities with a bronze spearhead from Shuiguanyin 水觀音, Xinfan 新繁, Sichuan, alluded to the possibility that the bronze spears in the Sichuan region during the Shang-Zhou transitional period may have had their origin in the Middle Yangzi Basin. The types of bronze weaponry found in the Zhuwajie caches, espe-cially the ge, disappeared from central China by the end of the Western Zhou period. However, in the Sichuan area, such weapons remained common up until the Warring States period. The special characteristics of these weapons have attracted the attention of researchers such as Tong Enzheng (1979a), who, in discussing the bronze ge of the southwest region, compared those from Zhuwajie to those of the central region of China. Tong believed that such ge originated in the Central Plains region and developed into distinctive local types upon transmission to the Sichuan area. Lu Liancheng and Hu Zhisheng (1988: 435–443) have compared the bronze ge from Zhuwajie with those unearthed from Sichuan and Shaanxi, arguing that the Chenggu region in southern Shaanxi is the point of origin for the “Sichuan style” bronze ge. They further argue that the bronzes from the Zhuyuangou tomb and the

310 SUN HUA

Zhuwajie caches date to the same period and may represent parallel courses of development. Based on the shapes of bronze ge found in Sichuan, it is reasonable to conclude that the long-bladed ge with protruding crosspieces originated in the central region of China, while the triangular-bladed ge with no protrusions originated locally in Sichuan, where over the long course of their development, these weapons came to have their own new and unique style. As bronze ge have a complicated shape, stylistic changes over the course of their development are obvious. If we take the datable ge from Sichuan and the neighboring region of southern Shaanxi and seriate them according to their assemblages and groupings, it is not dif-ficult to see that those from Zhuwajie are very different from those found at Shuiguanyin, which date to the early part of the Late Shang period (Sichuan Sheng Bowuguan 1959: 404–410). They are also different from the Shang/Zhou transitional specimens from Chenggu (Tang Jinyu et al. 1980: 211–218) and those found at the Early Warring States period

Figure 12: A comparison of ge from different time periods from Zhuwajie and related sites. Top: Late Shang period ge excavated in 1976 at Chenggu Sucun 蘇村; middle rows: Early Western Zhou period ge from Baoji Zhuyuangou (above) and Pengxian Zhuwajie

(below); bottom: Warring States period ge from Xindu Jiuliandun.

Chenggu Sucun

Baoji Zhuyuangou

Pengxian Zhuwajie

Xindu Jiuliandun


Tomb M1 at Moutuo 牟托 in Maoxian 茂縣 County, Sichuan (Maoxian Qiangzu Bowuguan 1994: 4–52), as well as from those found in the large Warring States period tomb at Jiuliandun (Sichuan Bowuguan and Xindu Xian Wenwu Guanlisuo 1981). On the other hand, the bronzes from Zhuwajie are extremely similar to those from the Early Western Zhou Yu tombs at Zhuyuangou (Figure12). The chronological position of the bronze ge from the caches at Zhuwajie, which are grouped together with those of the Zhuyuangou assemblages, is thus as follows:

Chenggu, Shuiguanyin → Zhuwajie, Zhuyuangou → Jiuliandun

Since the Zhuyuangou tombs that yielded bronze ge similar to those from Zhuwajie date from the first and second parts of the Early West-ern Zhou period, the dates of the ge from Zhuwajie should be similar. Although they are developmentally related to both earlier and later bronze weapons in the Sichuan area, there is a wide chronological gap between them.

Casting technology

The bronze weapons from the Zhuwajie caches are relatively un-sophisticated and have a strong local flavor, which leads one unprob-lematically to the conclusion that these weapons were locally cast by people in the Sichuan Basin. As to the bronze vessels, scholars have unanimously concluded that the lone zun and the two inscribed zhi from Cache 1 must have been brought from the Central Plains region. Thus the casting technology of these three vessels reflects that of the Central Plains tradition, which need not be discussed here. This leaves the nine lei, which are relatively complicated in their mold structure. Some scholars have argued that they may be replicas cast locally in Sichuan after prototypes from the Central Plains. Among the nine ves-sels, there are some differences in casting techniques that necessitate further explanation. The large bronze lei from Cache 1 (59J1: 1) has a plain surface. Four casting seams—a leftover from the piecemold casting process—are clearly discernible, coinciding with the vessel’s four vertical flanges. These four seams are evident on the flanges, on the two loop handles on the shoulder, and on the “nose-shaped” relief elements on the lower part of the belly. The seams indicate that the lei body was cast with four body molds, one bottom mold and one core. We can also see that the loop handles on the shoulder and the “nose-like” relief elements on the body were all produced as part of a single casting. In order to ensure the precision of the componential construction, a large number of bronze spacers were placed in four rows between the core and the

312 SUN HUA

outer molds. Aside from the shoulder region, where two spacers were used (positioned at each side of the round whorl patterns), each of the three rows on the lower vessel body features three spacers. The construction of the two medium-sized bronze lei from Cache 1 (59J1: 2, 3) is more complicated. The covers are in the form of a coiled dragon (though between the two vessels there are some differences in the shape of the dragon’s horns: on one piece they take the form of forked deer antlers, while on the other they each take the form of an individual mushroom). Feng Hanji (1980) has argued that these lei were cast using a single pour, but the complex shape of the raised dragon heads would have made it very difficult to remove the mold pieces from the model. So, not only should the casting mold for the covers have been split into at least two pieces, but the situation would most probably have required the use of special joining techniques such as inserting pre-cast elements into the core-mold assemblage. In fact, we notice an obvious casting seam running from the center of one coiled dragon’s head down along the flange on the spine of the dragon. As on the vessel body, the dividing line between the lids’s mold pieces followed the flanges, and the lid was cast with two principal outer molds that are, in this case, asymmetrical. On the vessel body, we can see traces of the casting seam on the shoulder loop handles, on the relieved elements of the body, and on correspond-ing positions in between. As Feng Hanji (1980) observed, “The mold for the body of the lei was split into four pieces, plus a single piece for the bottom.”The two looped handles on the shoulder and the relief elements on the belly are closely joined to the main body of the vessel, making it likely that they were cast together with the vessel body in a single pour. On one of the vessels, there is a small strut protruding from the place where one of the three-dimensional animal-head decorations on the shoulder fell off. On the other side, where the head remained attached, there is an obvious crack at the back. The main part of the vessel was thus first cast with protruding struts, to which the molds for the animal heads were later attached; these would be joined to the vessel body as they were cast. This technique is called “casting on.” The two smaller bronze lei (59J1: 4, 5) have plain surfaces, but it seems that the quality of the bronze used was not high, since after the vessel corroded its color and luster became mottled. Of the two vessels, the surface of the lei with six whorl patterns suffered the most wear. Perhaps because it had been used for a longer period of time, or because it had been carefully polished, the traces of the casting process have become difficult to discern. However, on the lei with four whorl patterns, along the central axis of the loop handles on the shoulder and the bovine-head relief ornamentation located to either side, we can discern a vertical casting seam, making it certain that the outside molds for the body of


this vessel consisted of four pieces. The two loop handles and “nose-shaped” elaborations are tightly joined. Underneath these, we can also discern traces of two vertical casting seams on the vessel body, so that we know that the loop handle’s “nose-shaped” elaborations and the vessel body were cast together in a single pour. The casting technique for the large bronze lei from Cache 2 (80J1: 1) is very similar to that of the large lei from Cache 1. On the loop handles along the shoulder, on the animal heads located to either side, as well as on the loop handle centrally placed on the lower back side of the belly, we can clearly discern the casting seams. These indicate that the vessel body was cast using a four-piece mold together with an additional mold for the bottom and a central core. The two bovine heads on the loop handles on the shoulder are con-nected to the bovine body on the vessel shoulder without any trace of separation. The inside of the loop handles shows two vertical casting seams that result from the use of a separate clay core for casting the handles. The inside of the tabs that hang off the loop handles still bears the traces of bronze that seeped into the interstices between the piece molds of the vessel walls and the mold of the handle. Such a phenomenon indicates that the handles were cast together with the main body in a single pour. Due to the fact that the two horns on the bovine head at the top of the shoulder’s loop handles are rather long and twisted, which would have made it difficult to remove the mold after casting, the casters adopted the technique of inserting a stand-alone mold at the top of the bovine head. Such stand-alone mold pieces were often used for casting elements of irregular, rounded shape. The plane of division was located between the protrusion of the ox’s horns and the ox’s shoulder, which is why one can see the clear traces of a casting seam at these places. On the walls of the vessel, the spacers were placed in six rows, one at the shoulder and five down the belly. Additionally, it seems that bronze core-spacers were placed on either side of the bovine head ornaments at the shoulder. These spacers served to maintain uniform thickness of the walls during the casting process. The cover of this lei has casting seams located between the tail ends of the two reclining-oxen motifs, indicating that the mold for the cover was divided into two symmetrical pieces, unlike the asymmetric molds used for the coiled-dragon covers of the medium-sized lei from Cache 1. The pair of medium-sized lei from Cache 2 (80J1: 2, 3) are quite complex in their casting technique. A casting seam is clearly visible at the bottom of the mushroom-shaped knob on the lid, perhaps indicating that the knob was cast-on afterwards. No mold seams are visible on the lid itself, but the two pairs of curved animal horns that protrude from the front and back of the lid could not have been cast with ordinary

314 SUN HUA

piecemolds; they must have been made using separate mold inserts. Mold seams are evident on the vessel body along the axis of the vertical flanges. These indicate that the body mold was primarily composed of four pieces. The flanges are very tall, and on the inside wall of the belly’s loop handle, one can observe traces of the loop’s clay core, indicating that the flanges and “nose-shaped” element were cast together with the vessel body in a single pour. At the place where the two ears of the vividly rendered elephant head on the shoulder join with the vessel, there are obvious crevices. Such gaps are also visible where the relief elephant-head ornaments, positioned between the two looped handles, meet the shoulder of the vessel. Above and below the loop handles are small, round perforations. The crevices between the loop handles and the shoulder had all been filled in with molten bronze, which served as a solder and which also covered some of the surface decoration on the shoulder of the vessel. From this, we can see that these appendages were first cast separately from the main body of the vessel and then soldered together, the join being secured by the presence of interlocking struts and mortises. The casting thus proceeded as follows. First, the body of the lei, the three loop handles, and the relief elephant-head ornaments were separately cast; then, any pieces of the clay core were removed from the two ends of the loop handles and ornaments, and from the places on the vessel where these elements were to be attached. The loop handles and ornaments were then placed atop struts protruding from the vessel body, and then a simple mold for the area around each con-nection was made. The gaps were filled in by pouring molten bronze into the interstices between the vessel and the loops and ornaments, thereby tightly attaching the components to the body. As for the single small lei from Cache 2, its shape and ornamentation were exactly the same as the two medium-sized lei from Cache 1 with coiled dragons on the cover. Its casting method and technology were also roughly the same. Among the casting techniques observable on these nine lei, two may serve to help date these bronzes. One is the method of joining append-ages to the main body of the vessel; the other is the use of stand-alone mold inserts to cast complicated, animal-head ornaments. To under-stand their position in the history of Chinese bronze-casting, one must realize that a major technological shift occurred at the transition between the Middle and Late Yinxu periods. Before that time, componential casting in multiple pours was common. Vessel attachments were either cast-on or (less commonly) pre-cast, and in either case they were kept in place on the vessel by means of mortise-and-tenon fits. The technique of joining two separately cast components with a third pour of metal is only rarely observed. Starting in the Late Yinxu period, casting in a single pour became common: the molds for appendages and attachments


were now integrated into the core-mold assemblage of the vessel body. When this was impossible, pre-casting and casting-on of components were still used, and a host of “self-locking” devices were invented to ensure the join of the separately cast components to the vessel body (Su et al. 1995: 110–145; Zhou Jianxun 1995: 254–270). The first great period of florescence of Chinese bronze-casting ended approximately with the reign of Mu Wang. Only a very few vessels cast after that time have been subject to an analysis of their casting technol-ogy. From my observations of the vessels from Cache 1 at Zhuangbai 庄白, Fufeng, Shaanxi (e.g. the Sannian Xing hu 三年E壺 and the Shisannian Xing hu 十三年E壺), it appears that the single-pour casting technique became even more widespread during this period (Beijing Daxue Kaogu Wenbo Xueyuan 2002). As for the use of stand-alone insert molds to cast animal-head ornaments (especially the curved horns on animal heads), the known examples all belong to the early part of the Early Western Zhou period. This can be seen as a technique devised to overcome the difficulties associated with casting vessels with complicated shapes during the time when the tendency to cast vessels in single pours was becoming stronger. For the most part, the vessels from Zhuwajie had their components cast together with the vessel body in a single pour. Some vessels were made by pre-casting the components and joining them to the vessel body by means of a self-locking join cast—a technique that dates to the earlier part of the Early Western Zhou—but we do not see the type of casting-on that results in rivet-like joins between the compo-nents common after the middle of Early Western Zhou. This would indicate that the Zhuwajie bronze vessels must date to before the middle of Early Western Zhou. More precisely, the preceding analysis suggests that the datable technical features of the Zhuwajie bronzes are all characteristic of Subperiods II or III of Early Western Zhou as defined at the beginning of this article. The examples closest to the Zhuwajie bronzes are the vessels from the Yu cemeteries of Zhifangtou and Zhuyuangou in Baoji. There is no indication that the Zhuwajie bronzes contain any elements characteristic of Late Western Zhou bronzes.

Chronology

Previous studies of the Zhuwajie bronzes show disagreement over several core issues, including their date of manufacture; their date of interment; their place of manufacture (i.e., which of them were cast in the Central Plains and which were of local manufacture); and the reasons why they were interred.

316 SUN HUA

As long as the chronological discrepancy between the casting of the objects and their interment is not great, the interval between the two can be effectively discounted. In the case of the Zhuwajie bronzes, however, their stylistic complexity suggests the possibility of a chronological gap of some magnitude. It is even possible that their actual date of manu-facture differs from the date implied by their stylistic features. Scholars have taken different positions on these issues. Some scholars hold that most or all of the bronzes recovered from the Zhuwajie cache were produced in the Early Western Zhou period and were also buried during this period, even though the specific dating of the bronzes is still disputed in some cases. Xu Zhongshu (1962) argued that the bronze zun and two of the bronze zhi from Zhuwajie Cache 1 are Late Shang period objects, while the rest were copies made in Sichuan “whose latest date should be the Early Western Zhou,” and that the burial date of the cache “was not far removed from the founding of the Western Zhou.” This argument has been very influential and continues to be endorsed by most specialists in the study of Sichuan bronzes (Du Naisong 1985: 62–65; Lin Chun 1991: 164–173; Shi Jingsong 1998). Hayashi Minao (1984), however, does not accept that the bronze zun with the beast mask design and the two bronze zhi with cast inscrip-tions can be pushed back to the Late Shang. Instead, in his In Shū jidai seidōki no kenkyū (Hayashi 1984), he argues that these three objects, as well as the other nine bronze lei, should all date to the late phase of Early Western Zhou. Lu Liancheng and Hu Zhisheng (1988: 436, 442) believe that the Zhuwajie bronzes are basically similar in date to those recovered from the Zhuyuangou cemetery in Baoji, and they propose that “the cache possibly dates to the Early Western Zhou.” Li Xueqin (1996) likewise denies that the Zhuwajie bronzes could possibly be later in date, believing that:

“altogether the bronze assemblage from the caches at Zhuwajie is homo-geneous, lacking items of obviously late date. Based on comparison with bronzes of the Central Plains, in particular the examples from the places nearest to Sichuan such as the Chenggu and Baoji in Shaanxi, the Zhuwa-jie assemblage should date to the beginning of the Western Zhou. This is a fundamentally different situation from that concerning the bronzes from Shaiba [i.e. Jiuliandun] in Xindu, even though that assemblage also con-tained elements of preceding cultural stages” (Li Xueqin 1996: 122).

The implication is that the objects were deposited relatively soon after their time of manufacture. Somewhat more cautiously, Lothar von Falkenhausen (2001: 188) states, “It is possible . . . that the vessels, weapons, and tools were all made at about the same time, perhaps around 1000 Bc, and deposited not long thereafter.”


Li Boqian’s (1983) view is similar to this, except that he admits the pos-sibility of a slightly wider chronological range. In his words, “the shape and inscriptions of the zun and zhi and the surface treatment of the lei vessels are achieved according to the practices of the Late Shang or of the Zhou during its formative stage.The ge with long blades and straight tangs are similar to those unearthed at Shuiguanyin and may date to as early as the Shang period, while the molding of the lei and the dragon motif with backward-curling tail are typical forms and decorative motifs of the Early Western Zhou. Compared with the specimens unearthed at Chenggu and Shuiguanyin, the triangular-bladed straight-tanged ge are longer and thinner. This form is somewhat later in date, and many of them may be from the Early Western Zhou. Some indeed may date to as late as the Middle Western Zhou” (Li Boqian 1983: 66–70). “Middle Western Zhou” in Li Boqian’s definition refers to the reigns from Mu Wang to Li Wang. Since the burial of the cache cannot predate the date of manufacture of any of the objects within it, this would then suggest that the burial took place during the Middle Western Zhou, that is, no earlier than the reign of Mu Wang. A third viewpoint was first put forward by Feng Hanji (1980), who proposed that the zun and the two zhi were possibly “objects belonging to late Shang Yin peoples,” while the other bronze lei and weapons were “of local Sichuan manufacture.” Feng admitted that “a judgment based solely on surface decoration would conclude that the casting of these objects was likely no later than the beginning of the Western Zhou, but because of the particular developmental histories of both politics and metallurgy in the Shu region, they may in fact be as late as the termi-nal Western Zhou or indeed the Spring and Autumn Period.” In their report of Cache 2, Fan Guijie 范桂杰 and Hu Changyu 胡昌玉 also suggest that the upper limit for the date of the cache should match that of Zhuwajie Cache 1, and that “none of the bronzes should be earlier than the terminal Shang or the beginning of Western Zhou,” and fur-ther, that “the time of burial of both may date to the end of Western Zhou or to the Spring and Autumn Period” (Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981). This position has also been greatly influential. Based on his understanding of the bronzes from Cache 1, Tong Enzheng (1979b: 66–67) argued that the five lei, one mao, and one ji were “local products manufactured by Shu casters during the Spring and Autumn Period.” I support the first of these three positions, for stylistically, as discussed above, the bronzes most strongly resemble items produced in the Central Plains during the first two subperiods of the Early Western Zhou. Never-theless, we must address the possibility that these might be stylistically faithful later copies produced locally in Sichuan.

318 SUN HUA

There is currently a tendency in studies of bronzes from southern China to suggest the existence of a stylistic lag with respect to the core regions. Frequently, southern bronzes that are morphologically or stylistically comparable to examples from central China are assigned to a later date than their central Chinese parallels, a stylistic lag being assumed to exist between the two areas concerned.16 I agree that Late Western Zhou bronzes from the southern area of China often tenaciously preserved the styles of earlier bronzes from the Central Plains, which had been transmitted during Early Western Zhou times. In the begin-ning, the southern bronzes must have been copied directly from pieces imported from the central area. For some reason, later innovations in techniques and style were no longer transmitted from the Central Plains to the south, where local bronze manufacturers could only continue to perpetuate the preexisting techniques and artistic traditions. For this reason, southern bronzes made during the Late Western Zhou and Spring and Autumn Period continue to display many Early Western Zhou anachronisms. One might ask, therefore, whether is it possible that, despite their stylistic similarity to Early Western Zhou bronzes from the Central Plains, the date of manufacture of the Zhuwajie bronzes was in fact later. Might the similarity be due to the preservation, in Sichuan, of old Early Western Zhou styles and production methods down to the Late Western Zhou and Early Eastern Zhou periods? With deference to Feng Hanji, I find this unlikely. Although it is certainly true that early bronzes were frequently imitated by later casters, it is impossible in such cases to avoid certain deviations from the models in morphology, surface decora-tion, vessel proportions, and details of casting technique, and in many cases new local or contemporary elements will be incorporated. This is certainly true in the case of the typologically atavistic south Chinese bronzes from Late Western Zhou and Spring and Autumn times; even when imitators were intentionally trying to cheat others, such as in the example of deliberate fakes produced from the Ming 明 and Qing 清 onwards, it is hard to avoid such anachronisms. It deserves notice that, among the many bronzes from the Zhuwajie caches, there is no example of a piece demonstrating any characteristics of Late Western Zhou or even later date. We can therefore be certain that the caches cannot postdate the Early Western Zhou. Moreover, unlike what seems to have happened in other areas of South China, contact between Sichuan and the Central Plains was by no means interrupted after the middle of the Western Zhou. Archaeological evidence shows that dur-

16 This phenomenon is particularly apparent in the discussions surrounding the dating of bronzes from the tumulus tombs (tudunmu 土墩墓) in the southeast, the Sanxingdui pits in Guanghan, Sichuan, and from the Dayangzhou tomb in Xin’gan County, Jiangxi.


ing the later stages of the Early Western Zhou, and perhaps even as late as the transition to the Late Western Zhou, Sichuan still preserved close connections with the Shaanxi Central Plains region and the Zhou court; the Yu cemeteries at Baoji, which show the strongest evidence of mutal interaction between the Sichuan bronze culture and the Zhou cultural sphere, were in use until late in the reign of Mu Wang. Such ties are corroborated by textual sources. The Bamboo Annals (as quoted in Taiping yulan 太平御覽, “Huangwang” 皇王; see Guben Zhushu jinian jizheng, p. 58) state, for instance: “In the second year of Yi Wang, the people of Shu and Lü came to make tribute of fine jade [qiongyu]. The bin ceremony to the god of the Yellow River was performed using the jiegui [jade scepter]. 夷王二年，蜀人、呂人來獻瓊玉，賓于河，用介珪).” The state of Shu mentioned here is commonly believed to have been situated within the area of modern Sichuan during this period. All of the chronological indicators gathered in the preceding analysis point to a date in Subperiod I or II of the Early Western Zhou period for the Zhuwajie bronzes. They exhibit no stylistic elements from the central region that postdate the reign of Zhao Wang. If one were to push the date of the Zhuwajie bronzes much later, it would be very difficult to explain the persistence of early elements in light of the continued connections between Sichuan and central Shaanxi. Having thus excluded the possibility of a later local manufacture, we can now consider the possible chronological gap between the Early Western Zhou manufacture of the bronzes and their deposit in the two caches. This issue is more difficult to resolve. In order to address it, we must rely on three lines of evidence: the date of the pottery vat in which the bronzes were buried, the date of the site with which the cache was associated, and a consideration of whether the Zhuwajie lei played any role as a stylistic stimulus for the later Warring States period bronze lei in Sichuan. The pottery vats in which the two caches of bronzes at Zhuwajie were buried have disappeared except for some fragments of the vat from Cache 2. As mentioned, Jiang Zhanghua (1998b) has demon-strated that its surface decoration is characteristic of Xinyicun Culture pottery. The Xinyicun Culture has recently been recognized to be intermediate between the Shi’erqiao and Qingyanggong Cultures in the Sichuan area. Due to a lack of relevant data, I have hitherto been uncertain about the chronological extent of this culture, and for a time I set its start date at the beginning of the Western Zhou (Sun Hua 2000: 91–115). However, new data from the site of Jinsha in Chengdu include fragments of a loop-handled bronze hu vessel dating roughly to Subperiod III of the Early Western Zhou that was recovered from the Late Shi’erqiao Culture layers at the site, implying that the Shi’erqiao Culture must have persisted until the later stages of the Early Western

320 SUN HUA

Zhou, and that the beginning of the Xinyicun Culture should be set near the boundary between the Early and Late Western Zhou.17 As mentioned, the settlement to which the Zhuwajie caches belonged is a more or less typical Shi’erqiao period site. There is little pottery with Sanxingdui cultural features of the kind commonly seen at sites of the earlier stage of the Shi’erqiao Culture—flat-bottom basins, pouch-foot he, long-stem dou, ladles with bird-head handles—suggesting that the site dates to the late phase of the Shi’erqiao Culture. The Early Xinyicun Culture is chronologically immediately subsequent to the Late Shi’erqiao. As the site of Zhuwajie belongs to the Late Shi’erqiao Culture, and the earthenware vat in which the cache bronzes were buried exhibits characteristics of Early Xinyicun pottery, the time of burial of this bronze cache should be just about the time when the site was abandoned and right at the terminus of the Shi’erqiao Culture. Typical Xinyicun pottery seems to be otherwise lacking at the Zhuwajie site. If we suppose that the beginning of the Xinyicun Culture dates to between Subperiods III and IV of the Early Western Zhou, or to Subperiod IV, and remembering that, for the most part, the bronzes were manufactured during Subperiods I and II, then the length of time between the casting and interment of these vessels was, in fact, quite short (only the zun with the beast-mask motif was somewhat earlier in date, possibly from Subperiod I). The style of the bronze vessels from both caches corresponds to that of counterparts from the Central Plains dating to the end of the Shang or the early phase of the Early Western Zhou. Regardless of whether these bronze vessels were cast in Sichuan or in some other region, they are certainly associated with the historical context of the close relations between the Sichuan region and the Central Plains, and were influenced by the latter’s techniques and artistic traditions. Close ties between the Sichuan Basin and the Guanzhong region of Shaanxi emerged during the Late Shang and Early Western Zhou periods, but were essentially severed after the Late Western Zhou. From then on, since the region no longer received supplementary information on further developments in the Central Plains, the Sichuan Basin continued to preserve Early Western Zhou technical and aritistic norms. Only after the Middle Spring and Autumn period did this situation begin to change, when, following the rise of the Chu state in the Middle Yangzi Basin, Chu bronzes entered the Chengdu Plain, transmitted along the Yangzi and Minjiang Rivers. Under Chu influence, Sichuan bronze manufacture

17 My thanks are due to the director of the excavations, Zhu Zhangyi 朱章義, for access to the material, which is in the charge of the Chengdu Municipal Institute of Cultural Relics and Archaeology.


Figure 13: A comparison of lei of the Western Zhou period from central China with Warring States period lei from the Sichuan region. 1. lei from Tomb M1043 from the Liulihe site in Fangshan, Beijing (early Western Zhou period); 2. lei from the Qijiacun 齊家村 site in Fufeng, Shaanxi (middle Western Zhou); 3. lei from the Sandongqiao site in Chengdu, Sichuan (early Warring States); 4. lei from the Jiuliandun site in Xindu, Sichuan (middle Warring States). 1. From Liulihe 1984: 142, fig. 7; 2. from Qingtongqi

1997: 171; 3. from Chengdu Shi Wenwu Guanlichu 1989: fig. 4.

1 2

3 4

322 SUN HUA

regained vitality and entered a period during which new and old styles were combined (Sun Hua 2000: 41–44). In this connection, it is relevant to observe that the form and decoration of the bronze lei from Zhuwajie all exhibit the styles of the Early Western Zhou Subperiods II and III, while the most common lei types seen in Sichuan during the Eastern Zhou period imitated models of the Shaanxi area dating to the Early Western Zhou Subperiod IV (or the Middle Western Zhou in the chronological sequence terminology followed by other scholars). This can be clearly seen when comparing both the bronze lei from Tomb 1 at Sandongqiao 三洞橋 in Chengdu (Chengdu Shi Wenwu Guanlichu 1989: fig. 4.4) and the lei from the large tomb at Jiuliandun 九聯墩 to Early Western Zhou IV specimens from Shaanxi, such as the square lei (fanglei 方罍) from Puducun 普渡村 in Chang’an County (Shaanxi Sheng Wenguanhui 1957: 75–85) and the round lei unearthed at Quanducun 勸讀村 in Fengxiang 鳳翔 County (Figure 13) (Shaanxi Sheng Kaogu Yanjiusuo 1979: 189). Objects like the lei deposited in the Zhuwajie caches thus do not seem to have served as models for casting lei vessels in Sichuan during later times. From this we can infer that the deposition date of these two assemblages was probably not far removed from the date of manufacture of the objects themselves, and that they were buried prior to the onset of the Late Western Zhou period.

Place of manufacture

The question of where the Zhuwajie bronzes were made was first raised by Wang Jiayou (1961) in his report of the discovery of Cache 1. Believing that the vessels were indistinguishable from Shang products, he implied that they, together with the bronze ge, mao, and ji weapons, had all been “imported from outside the region,” and that only the bronze yue 鉞, because of its distinctive shape and large size, “was possibly cast locally” (Wang Jiayou 1961). By contrast, Feng Hanji (1980) argued that among the Zhuwajie bronzes, only the zun and the two inscribed zhi came from the Central Plains, and that the rest of the vessels were cast locally in Sichuan. At present, there is consensus about the northern provenience of the zun and the two zhi, and about the local Sichuan manufacture of the bladed implements. Only the place of manufacture of the nine bronze lei still remains the subject of some debate. Jessica Rawson (1990: 30) argues that the Zhuwajie lei were locally cast in Sichuan, but feels that the decoration of the vessels, with their realistically rendered animal motifs and exaggerated flanges, displays the influence of bronze styles from the Middle Yangzi region; she further believes that these Sichuan


products in turn influenced Early Western Zhou bronze manufacture in the Central Plains. By contrast, Lothar von Falkenhausen (2001) feels that the most reasonable hypothesis is that the vessels were produced at some foundry either in the Guanzhong area or in Henan, and that even if the vessels were produced in Sichuan, they must certainly have been reproductions modeled on vessels that came from central China and could not have been the source of inspiration for simlar vessels cast in the royal domain. I agree with this point of view and will now offer some additional analysis in support of it, based on the relationship be-tween the Zhuwajie lei with rising-dragon lids and similar vessels from other regions. In the view of many scholars advocating a Sichuan place of manu-facture for the Zhuwajie lei, the most “local” feature is the three- dimensionally executed rising-dragon motif on the lids of the two medium-sized lei from Cache 1 (Subtype I.4). Their bodies, like that of the small lei from Cache 2 featuring a mismatched cover, are com-pletely covered by coiled-dragon decor. As mentioned above, however, a very similar lei with rising-dragon cover was excavated from Cache 2 at Beidongcun, Kazuo, in far-away Liaoning, and there must be some type of relationship between them. Du Naisong (1985) has stated that “since the size of the [Zhuwajie vessels], their sculpturally executed coiled dragons on the covers, and the coiled-body kuiwen decoration on the vessel bodies are all similar in style to those on the lei from Kazuo, Liaoning, it is not difficult to see that in ancient times the Sichuan re-gion must have had some relationship with cultures to the north.” Tong Enzheng (1986) cited the similarity between the lei from Zhuwajie and Beidongcun in support of his notion of a “crescent-shaped zone of cul-tural transmission between northeast and southwest China” spanning the Inner Mongolian Plain, the loess plain, and the Qinghai-Tibetan Plateau. Hirokawa Mamoru (1997: 24), however, cautions that “it would be somewhat forced to claim a connection between places as distant as Liaoning and Sichuan based on this one vessel type.” Indeed, their similarity is probably not the result of a cultural transmission between the two regions, but rather of the expansion of Zhou bronze styles from the central area to peripheral areas. We know that, except for a few vessels that can possibly be dated to the Late Yinxu period, all of the bronzes found in various caches at Kazuo are typical examples of Early Western Zhou bronzes from the Central Plains tradition. In those vessels that feature inscriptions, the state and lineage affiliations documented are all connected with the state of Yan, which had its capital at Liulihe in Fangshan, Beijing, during Early Western Zhou times. Their interment may have followed a decline in Yan’s political and military position in the Daling 大凌 River Basin that

324 SUN HUA

precipitated a hurried evacuation.18 Given their stylistic similarity, the two Zhuwajie lei and the specimen from Beidongcun must have been made at the same workshop sometime during the early part of the Early Western Zhou. The location of their common place of manufacture is intimated by the discovery of the above-mentioned lei from Hejiacun Xihao, Qishan, in the area of the Western Zhou capital (Figure 14) (Shaanxi Sheng Kaogu Yanjiusuo el al. 1979: fig. 149). Its cover is missing, but the size, shape, surface decoration, and casting technique of the vessel body are very similar to the lei from Zhuwajie and Beidongcun. It was found in an Early Western Zhou tomb that also yielded other bronze vessels, such as the Shi Su fangding 史速方鼎, the Shi Su jiao 史速角, the Shi Yi gui 史頤簋, and the Yin Cheng ding 尹丞鼎. The inscrip-tion on the Shi Yi gui records that “the king made an announcement to Bi Gong 王誥畢公.” Commentators have connected this with the “Biming 畢命” chapter of the Shangshu 尚書 , the preface of which records that “Kang Wang commanded Recorder Bi Gong to estab-lish separate residence in the suburb of Cheng Zhou, and recorded the ‘Command to Bi Gong’” 康王命作冊畢公分居里成周郊作畢命 (Shangshu “Biming;” Shisanjing zhushu 19.132, p. 245), concluding that the event mentioned in the Shi Yi gui inscription must have taken place on the ninth day of the sixth month of the twelfth year of Kang Wang’s reign. On the basis of a stylistic analysis of the Shi Shu and Shi Yi vessels, it has been further proposed that Shi Shu preceded Shi Yi in the genealogy of this family (Shaanxi Sheng Kaogu Yanjiusuo et al. 1979). If so, the lei from Hejiacun Xihao should date to around the time of Kang Wang, and so should the similar lei from Zhuwajie and Beidongcun. In fact, the Zhuwajie specimens just discussed are not regionally distinctive at all, but instead reflect the casting techniques and artistic style of the elite bronze culture associated with the Zhou royal house in central China. They must be the products of a Zhou royal foundry, or at least the work of royal Zhou bronze artisans. The Zhuwajie bronze lei with the rising-dragon covers should also be the work of a Zhou royal foundry, and if this is true, we can be confident in concluding that the other lei from Zhuwajie, as well, must have been produced within the royal Zhou bronze-making tradition.

18 Before the establishment of Yan in the reign of Cheng Wang, this area had been occupied by the Weiyingzi 魏營子 Culture. It was brought under Zhou control for strategic reasons, but the occupying forces from Yan were soon pushed back by the aboriginal populations. Several caches of ritual bronzes have been found in Kazuo, containing objects too unwieldly to be carried off when their owners were driven out.


Figure 14: Comparison and map showing the influence of Western Zhou bronze lei from the central region (1) in peripheral regions (2, 3). 1. Lei 1 from the Pengxian Zhuwajie Western Zhou cache (Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 1981: 497, fig. 2); 2. lei from the Hejiacun Xihao Zhou tomb in Qishan County, Shaanxi (Shaanxi Sheng Kaogu Yanjiusuo et al. 1979: fig. 149); 3. lei from the Kazuo Beidongcun Cache,

Liaoning (Kazuo Xian Wenhuaguan et al. 1974: 368, fig. 5).

1 2

3

326 SUN HUA

Historical background

Bronze ritual vessels were among the most precious objects during the Bronze Age, and there are many possible reasons why they might have come to be buried. When aristocratic families came off badly in court struggles and were forced to flee their state, they might have buried their bronzes among those objects that they were unable to take with them. When power relations shifted as a result of dynastic change or similar political transitions, if the transition were violent, those on the losing side might bury their own bronzes. When one state or city was captured by another, as long as there was sufficient time to dispose of those valu-able objects that could not be taken away, the local elite might bury its bronzes. The deposition of bronzes into the many caches discovered at the Zhouyuan 周原 site cluster in Fufeng and Qishan Counties, Shaanxi, for instance, was motivated either by the uprisings during the reign of Li Wang, or by the destruction of the Zhou capital by the Quanrong 犬戎 during the reign of You Wang (the latter possibility is more likely). Simi-larly, albeit during a much later time, the many Southern Song period (1127–1279) caches of gold, silver, bronze, and porcelain objects found in Sichuan were deposited during the long period of fighting between the Southern Song and the Mongolian armies at the end of the Southern Song, when many residents were killed or left homeless. In order to understand the reason for the burial of the bronze hoards at Zhuwajie in Pengxian, we must not only investigate the caches themselves, but also examine the entire cultural background of the sur-rounding site and the Chengdu Plain. We should not limit ourselves to the written records that contain the traditions of the dynastic history of the Shu state, but should consider all possibilities as to why valuable objects might have been buried in such caches. As we have seen, Feng Hanji (1980) thought that the bronzes were deposited at a rather late date. Based upon late textual accounts of Shu history that preserve the tradition of the displacement of the Duyu lin-eage by the Kaiming lineage, and that record that the Kaiming dynasty ruled the Shu state for twelve generations, he hypothesized an association between the caches and the Duyu lineage. He supposed that the Duyu buried the objects when they ceded rule to the Kaiming and fled to the Western Mountains. Feng Hanji should be commended for his bold-ness in attempting to link the bronze caches at Zhuwajie with historical records that purport to describe the period during which he thought the bronzes were buried. However, based on our above discussion, we can now state confidently that the date of deposition of the bronze caches at Zhuwajie cannot have been as late as proposed by Feng. Therefore, it does not seem possible that the deposition of the caches can be asso-


ciated with the traditional account of the supplanting of the Duyu by the Kaiming; it must be due to a different reason. Archaeological data from the Chengdu Plain already hint at a number of important social changes that might provide an historical context for the Zhuwajie bronze caches, and that should receive our careful attention. One such change is the abandonment of the large scale, walled site of Sanxingdui, which occurred around the time of transition from the Early to the Late Shang period, or perhaps slightly later (approximately 1200 Bc). At this time, the residents of Sanxingdui migrated to other areas, with some of the population moving northward to the Hanzhong Basin in southern Shaanxi. The many Shang-style bronzes recovered from Chenggu and Yangxian may be the remnants of this group. As Sanxingdui declined, the Jinsha site to the southwest rose to prominence. Thereafter, at approximately the time of the Zhou conquest of Shang in the middle of the eleventh century Bc, the Zhu-wajie site in Pengxian also became one of the important sites of the Chengdu Plain, along with Jinsha. Zhuwajie, in turn, was abandoned during the later part of the Early Western Zhou (about 900 Bc), when valuable vessels, including those used in ritual, were buried. This last event coincides with the abandonment of the Yu cemetery in Baoji on the northern periphery of the distribution area of the Sichuan bronze cultures. Coinciding with these changes in the archaeological record, the traditional historical accounts of ancient Sichuan reflect ongoing social changes in their legends concerning succession in political power. Ac-cording to these legends, the state of Shu,which was centered around the Chengdu Plain, may have gone through at least four shifts in dynastic political authority, being ruled by a total of five different dynasties: Cancong 蠶叢, Baiguan 柏灌,Yufu 魚鳧, Pubei 蒲卑 (also called Duyu), and Kaiming.19 According to the tradition, each one of these dynasties, all of which centered their rule in the Chengdu Plain, lasted several hundred years. According to the legends, however, except for the final Kaiming dynasty, which is said to have ruled for eleven generations over

19 Yang Xiong’s 楊雄 Shuwang benji 蜀王本紀, quoted in the commentary by Liu Kui 劉逵 to Zuo Si’s 左思 (太沖) “Shudu fu 蜀都賦,” states that, “The predecessors of the kings of Shu were called Cancong, Bohuo [Guan], Yufu, Puze [Bei], and Kaiming. At this time, the people [of Shu] were doltish, with drum-stick knot hair and strange language. They did not know writing. They had no rites or music. From Kaiming back to Cancong it was thirty-four thousand years.” 蜀王之先名蠶叢，柏濩（灌），魚鳧，蒲澤（卑），開明，是時人萌椎髻左言，不曉文字，未有禮樂。從開明上到蠶叢積三萬四千歲。(Wenxuan 文選 “Shudu fu,” You Mao 尤袤 edition 4.14b, Zhonghua Shuju 1977 reprint of 1809 edition by Hu Kejia 胡克家). “Bohuo 柏濩” and “Puze 蒲澤” in the Shuwang benji are likely mistranscriptions of the Boguan 柏灌 and Pubei 蒲卑 found in the Huayang guozhi 華陽國志.

328 SUN HUA

approximately 350 years, the total number of generations is unclear, as is the total number of years for which each dynasty ruled.20 The artifacts from the sacrificial pits at Sanxingdui preserve some clues that suggest a linkage between the Sanxingdui culture and the Baiguan and Yufu dynasties mentioned in the legends of ancient history. Related evidence comes from the Late Shang and Early Western Zhou cemeteries of the lords of Yu in Baoji at the western edge of the Guanzhong plain, which have a strong Sichuan flavor. The discovery in some of these tombs of duck-headed scepters and certain inscribed lineage names suggests that these may have been the remains of a lineage of local lords loyal to the Yufu dynasty. After the reigns of Mu Wang and Gong Wang, remains of the state of Yu suddenly disappear from their previous location in the northern foothills of the Qinling 秦嶺 mountains, and at about the same time, relations between the Sichuan Basin and the Guanzhong area were severed. There is no evidence for Zhou cultural influence on the Sichuan Basin after the end of the Early Western Zhou period. It is very possible that, at that time, the state then dominating the Sichuan Basin underwent an extraordinary internal or external transformation. Perhaps the royal house that ruled over the Chengdu Plain was over-thrown, and the dynasty (probably that of the Yufu lineage) that had maintained relations with the Western Zhou dynasty, either as an allied state or as a subordinate branch of the royal lineage, was replaced by a new dynasty that was more detached from the Zhou court. Even if we disregard the legends and restrict our analysis to archae-ological materials, we may say that the interment of the Zhuwajie hoards should reflect the historical situation of the decline of the Shi’erqiao Culture, as also represented at the adjacent settlement site, at the Shi’erqiao site in Chengdu, and at the Yu cemeteries in Baoji. If

20 The Huayang guozhi by Chang Qu 常璩 (Ad 4th c.) states: “Kaiming was enthroned,

and known as Congdi. . . . In the ninth generation, Kaiming Shang established lineage temples. . . . The Kaiming king, having dreamt that the city walls moved, founded Chengdu and relocated there . . . In the autumn of the fifth year of Zhou Shen Wang, the Qin daifu Zhang Yi and Sima Cuo, the Governor (du wei) Mo, and others campaigned against Shu via the Shiniu Route . . . The Kaiming lineage fell, having ruled Shu for twelve generations.” 開明位，號曰叢帝。…九世有開明尚，始立宗廟。…開明王自夢郭移，乃徙治成都，…周慎王五年秋，秦大夫張儀、司馬錯、都尉墨等從石牛道伐蜀，…開明氏遂亡，凡王蜀十二世。(Huayang guozhi “Shu zhi 蜀志”; Huayangguozhi jiaobu tu zhu 華陽國志校補圖注, Shanghai: Shanghai Guji Chubanshe, 3.1.113–114. Luo Bi 羅泌 in his Lushi 路史 draws on the Shuji 蜀記 (i.e. Shuwang benji) in stating: “According to the current Shuji and the Wangdi yuanji, from Zhou Xiang Wang to King Bieling of Shu was eleven generations, or three hundred and fifty years.” 以今蜀記望帝遠記周襄王至鱉靈王蜀，十一代三百五十年。(Lushi “Yulun 餘論”; Sibu beiyao edi-tion, 1.4b). There are many difficulties with the punctuation and interpretation of this passage, but however one understands it, the period until Qin’s conquest of Shu was over three hundred years. If one takes one generation to be twenty-five to thirty years, this would be approximately equal to eleven or twelve generations.


the accounts in later textual sources are reliable, this would reflect the ascent of the Pubei dynasty, who succeeded the Yufu dynasty in ruling over the Sichuan Basin at that time. The Zhuwajie bronze caches might consequently be the remains of Yufu aristocrats from the time when their dynasty was superseded by the Pubei dynasty.

References Cited:

Beijing Daxue Kaogu Wenbo Xueyuan 北京大學考古文博學院 (2002). Jijin zhu guoshi: Zhouyuan chutu Xi Zhou qingtongqi jingcui 吉金鑄國史: 周原出土西周青銅器精粹 (National history cast in bronze: Fine Western Zhou bronzes unearthed at Zhouyuan). Beijing: Kexue Chubanshe.

Beijing Daxue Kaoguxuexi 北京大學考古學系 and Shanxi Sheng Kaogu Yanjiusuo 山西省考古研究所 (1994).“Tianma-Qucun yizhi Beizhao Jin Hou mudi erci fajue 天馬—–曲村遺址北趙晉候墓地二次發掘 (Second excavation of the Jin Hou cemetery at Tianma-Qucun).” Wenwu 1994.1: 4–34.

Beijing Daxue Kaoguxuexi Shang Zhou zu 北京大學考古學系商周祖 and Shanxi Sheng Kaogu Yanjiusuo 山西省考古研究所 (eds.) (2000). Tianma-Qucun (1980–1989) 天馬—–曲村(1980–1989) (Tianma-Qucun [excavation report for] 1980–1989). Beijing: Kexue Chubanshe.

Beijing Shi Wenwu Yanjiusuo 北京市文物研究所 (1995). Liulihe Xi Zhou Yanguo mudi (1973–1977) 琉璃河西周燕國墓地 (The Western Zhou cemetery of the state of Yan at Liulihe [1973–1977]). Beijing: Wenwu Chubanshe.

Cao Wei 曹瑋 (1994). “Xi Zhou tongqi de fenqi 西周銅器的分期 (Periodization of Western Zhou bronzes).” In Beijing Daxue Kaoguxi 北京大學考古系 (eds.), Kaoguxue yanjiu 考古學研究 (Research in archaeology), pp. 144–165. Beijing: Beijing Daxue Chubanshe.

Chen Gongrou 陳公柔 and Zhang Changshou 張長壽 (1990). “Yin Zhou qingtong rongqi shang shoumian wen de duandai yanjiu 殷周青銅容器上獸面紋的斷代研究 (Research into the periodization of the beast mask ornament on Yin and Zhou bronze vessels).” Kaogu xuebao 1990.2: 137–168.

Chengdu Shi Wenwu Guanlichu 成都市文物管理處 (1989). “Chengdu Sandongqiao Qingyangxiaoqu Zhanguo mu 成都三洞橋青羊小區戰國墓 (Warring States burials from Qingyangxiaoqu at Sandong-qiao, Chengdu). Wenwu 1989.5: 30–35.

Du Naisong 杜迺松 (1985). “Lun Ba Shu qingtongqi 論巴蜀青銅器 (A discussion of Ba-Shu bronzes).” Jiang Han kaogu 江漢考古1985.3: 62–65.

330 SUN HUA

Falkenhausen, Lothar von (2001). “The Chengdu Plain in the early first millennium Bc: Zhuwajie.” In Robert Bagley (editor), Ancient Sichuan: Treasures from a Lost Civilization, pp.177–201. Princeton: Princeton University Press.

Fan Yi 樊一 (1998). Sanxingdui xunmeng 三星堆尋夢 (Seeking dreams of Sanxingdui). Chengdu: Sichuan Minzu Chubanshe.

Feng Hanji 馮漢驥 (1980). “Sichuan Pengxian chutu de tongqi 四川彭縣出土的銅器 (Bronzes excavated in Peng County, Sichuan).” Wenwu 1980.12: 38–47.

Gansu Sheng Bowuguan Wenwuzu 甘肅省博物館文物組 (1972). “Lingtai Baicaopo Xi Zhou mu” 靈臺白草坡西周墓 (The Western Zhou tombs at Baicaopo, Lingtai). Wenwu 1972.12: 2–8.

Guben zhushu jinian jizheng 古本竹書紀年輯証 (Collated edition of the Old Bamboo Annals), edited by Fang Shiming 方詩銘. Shanghai: Shanghai Guji Chubanshe, 1981.

Guo Baojun 郭寶鈞 (1981). Shang Zhou tongqiqun zonghe yanjiu 商周銅器群綜合研究 (Comprehensive research on Shang and Zhou dynasty bronze assemblages). Beijing: Wenwu Chubanshe.

Guo Moruo 郭沫若 (1959). “You Zhou chu si De qi de kaoshi tandao Yindai yijing jinxing wenzi jianhua 由周初四德器的考釋談到殷代已經進行文字簡化 (A discussion of simplification in the writing system in the Yin Dynasty based on the analysis of the four early Zhou De vessels).” Wenwu 1959.7: 1–2.

Hayashi Minao 林巳奈夫 (1984). In Shū jidai seidōki no kenkyū 殷周時代青銅器の研究 (Research on the bronzes of the Yin Shang and Zhou Dynasties). Part 1 of In Shu seidōki sōran 殷周青銅器綜覧 (Conspectus of Yin and Zhou bronzes), 2 vols. Tōkyō: Yoshikawa Kōbunkan.

Hirokawa Mamoru 広川守 (1997). “Liaoning Dalinghe liuyu de Yin Zhou Qingtongqi 遼寧大淩河流域的殷周青銅器 (Shang and Zhou bronzes from the Daling river basin, Liaoning). In Liaoning Sheng Wenwu Kaogu Yanjiusuo 遼寧省文物考古研究所 and Chūgoku kōkogaku kenkyūkai 中国考古学研究会 (eds.), Dongbeiya kaoguxue yanjiu: Zhong Ri hezuo yanjiu baogaoshu 東北亞考古學研究: 中日合作研究報告書 (Archaeological studies in Northeast Asia: Report on cooperative Sino-Japanese research), pp. 217–236. Beijing: Wenwu Chubanshe.

Huayangguo zhijiaobu tuzhu 華陽國志校補圖注, edited by Ren Naiqiang 任乃強. Shanghai: Shanghai Guji Chubanshe, 1987.

Jiang Zhanghua 江章華 (1996).“Ba Shu liuyexing jian yanjiu 巴蜀柳葉形劍研究(Research into the Ba-Shu “willow leaf-shaped” sword).” Kaogu 1996.9: 74–80.

Jiang Zhanghua (1998a). “Shilun Sanxingdui wenhua, Shi’erqiao wen-hua yu zhoulin wenhua de guanxi 試論三星堆文化, 十二橋文化與


周鄰文化的關係 (An examination of the relationship between the Sanxingdui and Shi’erqiao Cultures and their neighbors).” Chengdu wenwu 成都文物 1998.1: 4–11.

Jiang Zhanghua (1998b). “Chengdu Shi’erqiao wenhua de wenhua xingzhi ji fenqi yanjiu 成都十二橋文化的文化性質及分期研究(Research on the cultural characteristics and periodization of the Shi’erqiao Culture in Chengdu).” In Sichuan Daxue Kaogu Zhuanye 四川大學考古專業 (editor), Sichuan Daxue kaogu zhuanye chuang jian sanshiwuzhounian jinian wenji 四川大學考古專業創建三十五周年紀念文集 (Collected essays commemorating thirty-five years of the Archaeology Program at Sichuan University), pp. 146–164. Cheng-du: Sichuan Daxue Chubanshe.

Kazuo Xian Wenhuaguan 喀左縣文化館, Chaoyang Diqu Bowuguan 朝陽地區博物館, Beidong Wenwu Fajue Xiaozu 北洞文物發掘小組, and Liaoning Sheng Bowuguan 遼寧省博物館 (1974). “Liaoning Kazuo Xian Beidongcun chutu Yin Zhou qingtongqi 遼寧喀左縣北洞村出土的殷周青銅器 (Yin and Zhou bronze vessels unearthed in Beidongcun, Kazuo County, Liaoning).” Kaogu 1974.6: 356–372.

Kazuo Xian Wenhuaguan, Chaoyang Diqu Bowuguan, and Liaoning Sheng Bowuguan (1977). “Liaoning Sheng Kazuo Xian Shanwanzi chutu Yin Zhou qingtongqi 遼寧省喀左縣山灣子出土殷周青銅器 (Yin and Zhou bronzes unearthed at Shanwanzi in Kazuo County, Liaoning Province). Wenwu 1977.12: 23–27.

Li Boqian 李伯謙 (1983). “Chenggu tongqiqun yu zaoqi Shu wenhua 城固銅器群與早期蜀文化 (The Chenggu bronzes and the early Shu Culture).” Kaogu yu wenwu 考古與文物 1983.2: 66–70.

Li Feng 李豐 (1988). “Huanghe liuyu Xi Zhou muzang chutu qingtong liqi de fenqi yu niandai 黃河流域西周墓葬出土青銅禮器的分期與年代 (Dating and periodization of ritual bronzes from Western Zhou tombs in the Yellow River Basin).” Kaogu xuebao 考古學報 1988.4: 383–419.

Li Jianmin 李健民 (1996). “Lun Sichuan chutu de qingtong mao 論四川出土的青銅矛 (Bronze spears unearthed in Sichuan).” Kaogu 1996.2: 78–87.

Li Xuemei 李雪梅 (2001). “You yige Sanxingdui: Jinsha yizhi 又一個三星堆: 金沙遺址 (Another Sanxingdui: the Jinsha site).” Zhongguo guojia dili 中國國家地理 2001.6.

Li Xueqin 李學勤 (1990). “Beijing, Liaoning chutu qingtongqi yu Zhou chu de Yan 北京、遼寧出土青銅器與周初的燕 (Bronzes from Bei-jing and Liaoning, and the early Zhou state of Yan).” In Li Xueqin, Xin chu qingtongqi yanjiu 新出青銅器研究 (Studies on newly unearthed bronzes), pp. 46–53. Beijing: Wenwu Chubanshe.

Li Xueqin (1996). “Pengxian Zhuwajie qingtongqi de zai kaocha 彭

332 SUN HUA

縣竹瓦街青銅器的再考察 (A reexamination of the bronzes from Zhuwajie, Pengxian County).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 (editor), Sichuan kaogu lunwenji 四川考古論文集 (Essays on the Archaeology of Sichuan), pp. 118–122. Beijing: Wenwu Chubanshe.

Liaoning Sheng Bowuguan 遼寧省博物館 and Chaoyang Diqu Bowu-guan 朝陽地區博物館 (1974). “Liaoning Kazuo Xian Beidongcun chutu de Yin Zhou qingtongqi 遼寧喀左縣北洞村出土的殷周青銅器 (Yin and Zhou bronzes from Kazuo County, Liaoning).” Kaogu 1974.6: 364–372.

Lin Chun 林春 (1991). “Ba Shu de qingtongqi yu lishi 巴蜀的青銅器與歷史 (Ba Shu bronzes and history).” In Li Zhaoming 李紹明, Lin Xiang 林向, and Xu Nanzhou 徐南洲 (editors) (1991). Ba Shu lishi, minzu, kaogu, wenhua 巴蜀歷史, 民族, 考古, 文化 (History, ethnic identity, archaeology, and culture of Ba and Shu), pp. 164–173. Chengdu: Ba Shu Shushe.

Liu Lin 劉琳 (1978). “Chengdu chengchi bianqianshi kaoshu 成都城池變遷史考述 (Historical changes of location of the Chengdu city).” Sichuan Daxue xuebao 四川大學學報 1978.2: 78–84.

Liu Ying 劉瑛 (1983a). “Ba Shu tongqi wenshi tulu 巴蜀銅器紋飾圖錄 (An illustrated record of Ba Shu bronze ornament).” Wenwu ziliao congkan 文物資料叢刊 7: 1–12.

Liu Ying (1983b). “Ba Shu bingqi ji qi wenshi fuhao 巴蜀兵器紋飾符號 (Ba Shu weapons and their decorations and symbolic marks).” Wenwu ziliao congkan 7: 13–23.

Liulihe 1984 = Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 and Beijing Shi Wenwu Gongzuodui 北京市文物工作隊 (1984). “1981–1983 nian Liulihe Xi Zhou Yan guo mu di fajue jianbao 1981–1983 年琉璃河西周公燕國墓地發掘簡報 (Preliminary report for the excavation of the Yan state Western Zhou cemetery at Liulihe, 1981–1983).” Kaogu 1984.5: 405–416, 404.

Lu Liancheng 盧連成 and Hu Zhisheng 胡智生 (1988). Baoji Yuguo mudi 寶雞A國墓地 (The Yu Cemetery at Baoji). Beijing: Wenwu Chubanshe.

Luo Zhenyu 羅振玉 (editor) (1936). Sandai jijin wencun 三代吉金文存 (A corpus of bronze inscriptions from the Three Dynasties). Shanghai: privately published. New edition, Beijing: Zhonghua shuju, 1983.

Lushi 路史, by Luo Mi 羅泌. Sibu beiyao 四部備要 edition. Shanghai: Zhonghua Shuju, 1933.

Maoxian Qiangzu Bowuguan 茂縣羌族博物館 (1994). “Sichuan Mao-xian Moutuo yihao shiguanmu ji qi peizangkeng qingli jianbao 四川茂縣牟托一號石棺墓及其陪葬坑清理簡報 (Summary report on the excavation of stone tomb M1 and associated pits at Moutuo, Maoxian County, Sichuan).” Wenwu 1994.3: 4–52.


Qingtongqi 1997 = Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅器全集編輯委員會 (editor) (1997). Zhongguo qingtongqi quanji 5 (Xi Zhou 1) 中國青銅器全集 5 (西周 1) (Complete collec-tion of Chinese bronzes, vol. 5 (Western Zhou 1). Beijing: Wenwu Chubanshe.

Rawson, Jessica (1990). Western Zhou Ritual Bronzes from the Arthur M. Sackler Collections. Ancient Chinese Ritual Bronzes from the Arthur M. Sackler Collections, vol. 2, 2 parts. Cambridge, MA: Harvard Uni-versity Press.

Shaanxi Sheng Bowuguan 陝西省博物館 and Shaanxi Sheng Wen-wu Guanliweiyuanhui 陝西省文物管理委員會 (1976). “Shaanxi Qishan Hejiacun Xi Zhou muzang 陝西岐山賀家村西周墓葬 (Western Zhou burials at Hejiacun, Qishan, Shaanxi). Kaogu 1976.1: 31–38.

Shaanxi Sheng Kaogu Yanjiusuo 陝西省考古研究所 (1995). Gaojiabu Geguo mu 高家堡戈國墓 (The Ge state cemetery, Gaojiabu). Xi’an: San Qin Chubanshe.

Shaanxi Sheng Kaogu Yanjiusuo, Shaanxi Sheng Wenwu Guanliwei-yuanhui, and Shaanxi Sheng Bowuguan (1979). Shaanxi chutu Shang Zhouqingtongqi (1) 陝西出土商周青銅器 (1) (Shang and Zhou Bronzes from Shaanxi, Vol. 1). Beijing: Wenwu Chubanshe.

Shaanxi Sheng Kaogu Yanjiusuo, Shaanxi Sheng Wenwu Guanliwei-yuanhui, and Shaanxi Sheng Bowuguan (1980). Shaanxi chutu Shang Zhouqingtongqi (3) 陝西出土商周青銅器 (3) (Shang and Zhou Bronzes from Shaanxi, Vol. 3). Beijing: Wenwu Chubanshe.

Shaanxi Sheng Wenguanhui 陝西省文管會 (1957). “Chang’an Pudu-cun Xi Zhou mu de fajue 長安普渡村西周墓的發掘 (Excavation of Western Zhou tombs at Puducun, Chang’an).” Kaogu xuebao 1957.1: 75–85.

Shi Jingsong 施勁松 (1998). “Lun Sichuan chutu de Shang dai Xi Zhou qingtong liqi 論四川出土的商代西周青銅禮器 (Shang and Western Zhou ritual bronzes excavated in Sichuan).” In Sichuan Daxue Kaogu Zhuanye (editor), Sichuan Daxue kaogu zhuanye chuang jian sanshiwuzhounian jinian wenji 四川大學考古專業創建三十五周年紀念文集 (Collected essays commemorating thirty-five years of the Archaeology Program at Sichuan University), pp. 179–189. Cheng-du: Sichuan Daxue Chubanshe.

Shisanjing zhushu 十三經注疏 (The Thirteen Classics with commentaries and subcommentaries), edited by Ruan Yuan 阮元. New edition, Beijing: Zhonghua Shuju, 1981.

Sichuan Sheng Bowuguan 四川省博物館 (1959). “Xinfan Xian Shui-guanyin yizhi shijue jianbao 新繁縣水觀音遺址試掘簡報 (Summary report of test excavations at Shuiguanyin, Xinfan County).” Kaogu 1959.8: 404–410.

334 SUN HUA

Sichuan Sheng Bowuguan (1992). Ba Shu qingtongqi 巴蜀青銅器 (Ba-Shu bronzes). Chengdu: Chengdu Chubanshe and Macao: Ziyunzhai Chuban Youxiangongsi.

Sichuan Sheng Bowuguan and Pengxian Wenhuaguan 彭縣文化館 (1981). “Sichuan Pengxian Xi Zhou jiaocang tongqi 四川彭縣西周窖藏銅器 (Bronzes from a Western Zhou cache in Pengxian County, Sichuan).” Kaogu 1981.6: 496–499.

Sichuan Sheng Bowuguan and Xindu Xian Wenwu Guanlisuo 新都縣文物管理所 (1981). “Sichuan Xindu Zhanguo muguo mu 四川新都戰國木槨墓 (A Warring Sates tomb with wooden burial chamber from Xindu, Sichuan).” Wenwu 1981.6: 1–25.

Sichuan Sheng Cehuiju 四川省測繪局 and Zhongguo Kexueyuan Chengdu Dili Yanjiusuo 中國科學院成都地理研究所 (1981). Sichuan sheng dituji 四川省地圖集 (Atlas of Sichuan Province). Chengdu: Sichuan Cehuiju.

Sichuan Sheng Wenwu Guanliweiyuanhui 四川省文物管理委員會 (1987). “Chengdu Shi’erqiao Shang dai jianzhu yizhi diyiqi fajue jianbao 成都十二橋商代建築遺址第一期發掘簡報 (Summary report of the first season of excavations of Shang period structures at Shi’erqiao, Chengdu).” Wenwu 1987.12: 1–23, 37.

Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所 (2000). Sanxingdui jisikeng 三星堆祭祀坑 (The sacrificial pits at San-xingdui). Chengdu: Wenwu Chubanshe.

Sichuan Sheng Wenwu Kaogu Yanjiusuo, Sichuan Guanghan Shi Wen-guansuo 四川廣漢市文管所, Shifang Xian Wenguansuo 什邡縣文管所 (1992). “Sichuan Guanghan, Shifang Shang Zhou yizhi diaocha 四川廣漢、什邡商周遺址調查 (Survey of Shang and Zhou sites in Guanghan and Shifang Counties, Sichuan).”In Nanfang minzu kaogu 南方民族考古 5: 295–309.

Su Rongyu 蘇榮譽, Hua Jueming 華覺明, Li Kemin 李克敏, and Lu Benshan 盧本珊 (1995). Zhongguo shanggu jinshu jishu 中國上古金屬技術 (Ancient Chinese metallurgy). Ji’nan: Shandong Kexue Jishu Chubanshe.

Sun Hua 孫華 (2000). Sichuan pendi de qingtong shidai 四川盆地的青銅時代 (The Bronze Age of the Sichuan Basin), pp. 2–46. Beijing: Kexue Chubanshe.

Sun Hua (2002). “Pengxian Zhuwajie tongqi zai fenxi—maizang xing-zhi, niandai, yuanyin jiqi wenhua beijing 彭縣竹瓦街銅器再分析—–埋葬性質，年代，原因及其文化背景 (Re-examination of the Bronzes from Pengxian Zhuwajie—the Nature, Dates, Causes of the Burials and their Cultural Context). In Gao Chongwen 高崇文 and Yasuda Yoshinori 安田喜憲 (eds,), Chang jiang liuyu qingtong wenhua yanjiu 長江流域青銅文化研究 (Studies of the bronze cultures of the Changjiang River), pp. 126–168. Beijing: Kexue Chubanshe.


Tang Jinyu 唐金裕, Wang Shouzhi 王壽芝, and Guo Changjiang 郭長江 (1980). “Shaanxi Sheng Chenggu Xian chutu Yin Shang tongqi zhengli baogao 陝西省城固縣出土殷商銅器整理報告 (Report on the typological ordering of Yin Shang bronzes excavated in Chenggu County, Shaanxi).” Kaogu 1980.3: 211–218.

Tang Lan 唐蘭 (1976). “He zun mingwen jieshi 何尊銘文解釋 (Com-mentary on the He zun inscription).” Wenwu 1976.1: 60–63.

Tong Enzheng 童恩正 (1979a). “Woguo xi’nan diqu qingtong ge de yanjiu 我國西南地區青銅戈的研究 (Bronze ge dagger-axes from the southwest of our country).” Kaogu xuebao 1979.4: 441–457.

Tong Enzheng (1979b). Gudai de Ba Shu 古代的巴蜀 (Ba and Shu in antiquity). Chengdu: Sichuan Renmin Chubanshe.

Tong Enzheng (1986). “Shilun woguo cong dongbei zhi xi’nan de biandi banyuexing wenhua chuanbodai 試論我國從東北至西南的邊地半月形文化傳播帶 (The crescent-shaped zone of cultural transmission reaching from the northeastern to the southwestern borderlands of China). In Wenwu Chubanshe Bianjibu 文物出版社編輯部 (editor), Wenwu yu kaogu lunji 文物與考古論集 (Essays on material culture and archaeology), pp. 17–43. Beijing: Wenwu Chubanshe.

Wang Jiayou 王家佑 (1961). “Ji Sichuan Pengxian Zhuwajie chutu de tongqi 記四川彭縣竹瓦街出土的銅器 (A note on the bronzes from Zhuwajie, Pengxian County, Sichuan).” Wenwu 1961.11: 28–31.

Wang Yi 王毅 (1991). “Chengdu Shi Ba Shu wenhua yizhi de xin faxian 成都市巴蜀文化遺址的新發現 (New discoveries of the Ba Shu culture in Chengdu). In Li Shaoming 李紹明, Lin Xiang 林向, and Xu Nanzhou 徐南洲 (eds.), Ba Shu lishi, minzu, kaogu, wenhua 巴蜀歷史·民族·考古·文化 (Ba-Shu history, people, archaeology, culture), pp. 295–309. Chengdu: Ba Shu Shushe.

Xu Zhongshu 徐中舒 (1962). “Sichuan Pengxian Mengyangzhen chutu de Yindai er zhi 四川彭縣濛陽鎮出土的殷代二觶 (Two Yin period zhi excavated in Mengyangzhen, Pengxian, Sichuan).” Wenwu 1962.6: 15–18, 23.

Yan Jingsong 顏勁松 and Sun Hua (2001). “Chengdu naoshi dixia da faxian—ju xing chuan guanmu 成都鬧市地下大發現—–巨型船棺墓 (A great discovery beneath the city center of Chengdu: huge boat-coffin burials).” Zhongguo guojia dili 2001.5: 8–17.

Zhao Congcang 趙叢蒼 (1996). “Chenggu Yangxian tongqiqun zonghe yanjiu 城固洋縣銅器群綜合研究 (A comprehensive study of the bronze assemblages from Chenggu and Yangxian counties).” Wenbo 文博 1996.4: 3–26.

Zhao Yongfu 趙永福 (1984). “1961–1962 nian Chang’an Fengxi fajue jianbao 1961–1962 年長安灃西發掘簡報 (Preliminary report on the 1961–1962 excavations at Fengxi, Chang’an).” Kaogu 1984.9: 784– 789.

336 SUN HUA

Zhongguo Qingtongqi Quanji Bianji Weiyuanhui 中國青銅全集編輯委員會 (1994). Zhongguo qingtongqi quanji, Ba Shu 中國青銅器全集· 巴蜀 (Corpus of Chinese bronzes: Ba-Shu). Beijing: Wenwu Chu-banshe.

Zhongguo Shehuikexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 and Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊 (forthcoming). “Sichuan Qinglongjie yizhi fajue jianbao 四川青龍街遺址發掘簡報 (Summary report of excavations at the Qinglongjie site, Sichuan).” Publication n.a.

Zhongguo Shehuikexueyuan Kaogu Yanjiusuo, Beijing Shi Wenwu Yanjiusuo, and Liulihe Kaogu Dui 琉璃河考古隊 (1990). “Beijing Liulihe 1193 hao da mu fajue jianbao 北京琉璃河 1193 號大墓發掘簡報 (Summary report of the excavation of large Tomb 1193 at Liulihe, Beijing).” Kaogu 1990.1: 20–31.

Zhou Jianxun 周建勛 (1995). “Shang Zhou qingtongqi zhuzao gongyi de ruogan tantao 商周青銅器鑄造工藝的若干探討 (A number of questions concerning Shang and Zhou bronze casting techniques).” In Beijing Shi Wenwu Yanjiusuo 1995: 254–270.

Zhu Fenghan 朱鳳翰 (1995). Gudai Zhongguo qingtongqi 古代中國青銅器(Ancient Chinese bronzes). Tianjin: Nankai Daxue Chubanshe.

Address:

Department of ArchaeologyPeking UniversityBeijing 100870People’s Republic of [email protected]


LACQUER CRAFTSMANSHIP IN THE QIN AND CHU KINGDOMS:

TWO CONTRASTING TRADITIONS (LATE 4TH TO LATE 3RD CENTURY bc

BY

ALAIN THOTE(Ecole Pratique des Hautes Etudes 4ème Section

(Sciences historiques et philologiques) Paris, France)

AbstractThis article deals with the lacquer traditions of the late Warring States period as seen in the Qin and Chu kingdoms and in relation to recent discoveries from Sichuan. Qin lacquer working techniques and style are different from those of Chu, which are evi-denced by hundreds of objects. By contrast, due to poor conditions of preservation, Qin lacquer craftsmanship can only be known through very few objects, some of which show a strong influence from nomadic art. During the late fourth and early third century bc, Sichuan appears to have been a key region in the development of the lacquer art of Qin. Several Qin lacquers have marks either stamped on the core before the application of lacquer or incised with a needle on the lacquer surface. They confirm that the lacquer workshops operated under the control of the state administration. Until the destruction of the Chu capital in 278 bc, the Chu lacquer tradition in the Jiangling area had a more diverse range of shapes than Qin. However, the techniques used to make the cores were less sophisticated than in Qin, except in the case of luxury objects. Only a small number of Chu lacquer pieces have marks. At the same time, the comparisons show that both traditions exerted a mutual influence on each other to some extent until the early third century bc. Thereafter, Qin seems to have dominated the production of lacquer, even in the area that was in earlier times the core of the Chu kingdom, in present-day Hubei and Hunan. Most of the lacquers found in Changsha tombs of the third century bc that have been traditionally considered as Chu products were in fact produced in workshops working in a Qin cultural environment. Keywords: Lacquer, craftsmanship, Sichuan, Hubei, Changsha, Qin, Chu, Warring States period.

Introduction

The imperial Sichuan 四川 factories were extremely well known within and beyond the Chinese empire during the later part of the Western Han

338 ALAIN THOTE

西漢 dynasty.1 Many lacquer pieces produced at the official workshops (gongguan 工官) of the Shu 蜀 and Guanghan 廣漢 commanderies have been discovered in a few sites spread all over China, from Guizhou 貴州 to Jiangsu 江蘇, and even in Korea, testifying to a wide distribution of their products. In fact, the origin of the lacquer pieces excavated on those sites could be easily identified because the long inscriptions they bear record their provenance and provide precise dates of manufacture. These factories were created around the middle of the second century bc based on earlier Han local workshops (Yu and Li 1975; Pirazzoli-t’Serstevens 1990: 525–527).2 But the latter had their own roots in the lacquer traditions of the kingdom of Qin 秦. Not only were the Han workshops organized on the model of earlier Qin workshops, but they also made use of lacquer techniques that had been invented in Sichuan at least two centuries earlier. Even the style of their own products betrays earlier Qin artistic traditions. In fact, the 316 bc conquest of Shu in central Sichuan had allowed the kingdom of Qin to take im-mediate advantage of the natural resources of Sichuan (Sage 1992: 83–156; Pirazzoli-t’Serstevens 2001: 39–40). Soon after the conquest, Qin created lacquer workshops or, more probably, took control of local lacquer production. The states of Ba 巴 and Shu, as well as the westernmost regions of Sichuan, already had lacquer traditions when Qin invaded the Chengdu Plain. This is documented by a few objects found in tombs from early sites located near Chengdu and in the upper Minjiang 岷江 valley.3

1 This paper was originally presented at the symposium entitled “Treasures from a Lost Civilization: Ancient Chinese Art from Sichuan,” organized by the Seattle Art Museum, Seattle, 3–4 August, 2001. I wish to express my deep gratitude to Mimi Gardner Gates, Illsley Ball Nordstrom Director, Seattle Art Museum, and Jay Xu, Foster Foundation Curator of Chinese Art, Seattle Art Museum, for their invitation to participate in the symposium, as well as Michèle Pirazzoli-t’Serstevens, Philip K. Hu and Peter J. Lu for their valuable comments, and to Lothar von Falkenhausen and Robert E. Murowchick for their editorial suggestions. 2 To the famous site of Lelang 樂浪 in Korea, where a large number of lacquers from the Sichuan official factories were found in the early twentieth century (Umehara Sueji 1943), can be added a few other examples, such as Tombs 13, 15 and 17 at Qingzhen 清鎮 in Guizhou 貴州 (lacquer earcups from the Shu and Guanghan commanderies) (Guizhou 1959: 99–100 and pl. IV); Yaoziling 鷂子岭Tombs 2 and 3 in Yongzhou Shi 永州市, Hunan (one basin, three earcups, and a few containers of the zun 樽 and zhi 觶 types) (Hunan 2001 : 52–56 and fig. 18 p. 56, fig. 19–20 p. 57, fig. 21 p. 58, fig. 22 p. 59, pl. VI.2, pl. VI.4, pl. VII, pl. VIII.1–3); and Hanjiang Yaozhuang 邗江姚莊 Tomb 102 in Jiangsu (a ke 榼 wine container, and probably one basin and some earcups) (Jiangsu 2000: 61 and p. 61 fig. 20.1). Recently a fragment from a lacquerware of exceptional quality was discovered in a tomb of the Western Han to Eastern Han period at Yuejincun 躍進村 in Qingbaijiangqu 青白江區, Chengdu 成都 (Chengdu 1999: 35, and p. 36 fig. 40). Comparison with the lacquers from Lelang suggests that this piece was also probably produced by the Sichuan official factories. 3 These are for example Shifang Chengguan 什邡城關 Tomb 33, dated to the second

LACQUER IN THE QIN AND CHU KINGDOMS 339

Figure 1. Lacquer eared cup from Tomb 1 at Xindu Majiaxiang (Sichuan), ca mid-fourth century bc. After Sichuan 1981, p. 5 fig. 9.

These sites are dated to the period before the conquest. In particular, an eared cup from Xindu Majiaxiang 新都馬家鄉, though inspired by a model from the Zhou 周 cultural sphere, has a very distinctive shape. Its two non-symmetrical ears would be extremely rare anywhere else in China during the Warring States period [Fig. 1].4 Moreover, a Ba-Shu emblem has been painted on the inner part of the cup. Although this emblem is related to the owner of the cup, and therefore was added after the cup had been finished, it could only have been painted in Sichuan. Also, the boat-coffin from the same tomb, a typical burial of the Ba-Shu cultures, was painted with red and black lacquer with some traces of gold. A cist burial at another site, at Mao Xian Moutuo 茂縣牟托, contained a pottery jar closed by a lacquered wooden lid in the shape of two buffalo heads put together back to back.5 The wall of the jar was ornamented with a red lacquer decoration showing a human head [Fig. 2]. Both pieces—the jar and its lid—have a very specific style that is totally unknown in either a Zhou context or in a Ba-Shu context. They can only have been made locally or brought from a nearby site.

half of Period II of the site (ca. 350–300 bc) (Beijing 1998a); the large tomb at Xindu Majiaxiang (ca. 350–316 bc) (Sichuan 1981: 5, and p. 5 fig. 9); Chengdu Xijiao 成都西郊 tomb (4th century bc) (Sichuan 1983: 597); and Dayi Wulong 大邑五龍 Tomb 4 (4th century bc) (Sichuan 1985: 38). Other Ba-Shu sites as well are notable for having traces of laquerware. Unfortunately, due to poor conditions of preservation, these are merely mentioned in archaeological reports without any further information. 4 As far as I know, only one tomb has provided similar examples. Tomb 1 at Jingzhou Jicheng 荊州紀城 contained a set of four eared-cups of this shape. The lacquer decora-tion on these cups suggests that they were made in a Chu 楚 workshop (see Hubei 1999: 10–11 and p. 11 fig. 15.5). This tomb can be dated to about 350 bc. 5 See Mao Xian 1994: 6, fig. 7.1–2 p. 9, fig. 10 p. 10.

340 ALAIN THOTE

It seems that the lacquer craftsmanship in the Sichuan Basin may have been stimulated by imported objects. A recent discovery in Chengdu Shangyejie 成都商業街 has revealed the existence of lacquers of very high quality in Sichuan as early as in the fifth century bc. Judging from their style, most of the lacquers from Shangyejie are strongly con-nected to Chu 楚 lacquerware. In particular, some of the pieces can be compared with those from the Zeng Hou Yi 曾候乙 tomb at Suizhou Leigudun 隨州擂鼓墩, dated to the late fifth century bc (ca. 433 bc). They were probably imported from Chu. Many others betray a specific taste that could be local (Chengdu 2002; compare with Beijing 1989).

The ancient tradition of Qin lacquer craftsmanship

Before invading Sichuan, Qin already possessed a tradition of lacquer craftsmanship. However, due to poor conditions of preservation, almost no examples of early Qin lacquerware survive. In the archaeological reports of Qin sites, traces of lacquered coffins, boxes, and chariots are mentioned, but none of these have been illustrated, making it difficult to have a clear idea of their craftsmanship and to identify the artistic tradi-tion to which they belonged.6 In spite of this unfavorable archaeological

Figure 2. Pottery jar with wooden and bronze lid from Maoxian Moutuo Tomb 1 (Sichuan), ca. late fifth or early fourth century bc. After Mao Xian 1994, p. 9 fig. 7.2.

6 For example, traces of lacquerware were found in the cemetery of Longxian Dianzi 隴縣店子 (Shaanxi 1998: 109). Moreover, fourteen from a total of 21 lacquer boxes discovered at Xianyang Huangjiagou 咸陽黃家溝 (Qin Du Xianyang Kaogudui 1982:


situation, several insights, in my opinion, seem to indicate that Qin had an indigenous lacquer tradition with distinct characteristics. First, it is highly probable that Qin had inherited artistic traditions that had developed in the Wei 渭 River valley under the patronage of the Zhou court. What has already been shown for the bronze industry should apply equally to other artistic traditions. For example, large trays made in wood and lacquer, as well as lacquer containers (three dou 豆 for food offerings and a lei 罍 vessel for wine), have been discovered at Baoji Zhuyuangou 寶雞竹園溝 Tombs 4, 7, and 13, in an Early to Middle Western Zhou cemetery. The dou and lei vessels were inlaid with mother-of-pearl, a common mode of decoration on Western Zhou lacquers. Shields from the same site (Tombs 13 and 20) were also coated with lacquer.7 Apart from the burial furniture, the coffins in most of the tombs had also been lacquered.8 Qin probably inherited the Western Zhou palace workshops and maintained their earlier artistic traditions. In spite of the poor conditions of preservation, one remarkable example from around the middle of the sixth century bc has survived [Fig. 3]. Originally, this wood carving was probably an ornament at-

Figure 3. Part of a lacquered wooden ornament from the tomb of Qin Jing Gong (r. 576–537 bc), at Fengxiang Nanzhihui (Shaanxi), ca. mid sixth century bc. Dimen-sions unknown. Drawing from a photograph given by Prof. Han Wei, Director fo the

Archaeological Institute of Shaanxi.

8, 9, 12–14) are said to come from Middle to Late Warring States tombs (the remaining seven boxes are dated between the Late Warring States period and the Qin dynasty). 7 Lu and Hu 1988: 61, 94, 156 (trays); p. 181 (one lei vessel); 109 (two dou footed cups); 233 (one dou footed cup); 76, 194 (shields). Tomb 1 at Rujiazhuang 茹家莊 contained also two dou footed cups (ibid: 281). 8 Lu and Hu 1988: 94; 141; 189; 200; 220 (respectively from Tombs 7, 4, 20, 19, 18 at Zhuyuangou); 274; 361 (Tombs 1 and 2 at Rujiazhuang). Also, the coffins from Tomb 13 at Zhuyuangou (p. 45) could possibly have been lacquered.

342 ALAIN THOTE

tached to a larger object of unknown function (sculptures did not exist as autonomous works of art in that early period). This carving, lacquered in red and black, comes from the tomb of Qin Jing Gong 秦景公 who reigned from 576 to 537 bc. Although the tomb, which is the largest excavated to date in pre-imperial China, had been plundered several times since remote antiquity, it contained a few objects of excellent craftsmanship. The lacquer ornament in question represents a running boar, with its front legs bent so as to connect with its hind legs, a pos-ture that was intended to suggest speedy movement. The surface of the wood is lacquered in black and ornamented with small motifs painted in red. The fur is suggested by small dots covering the whole surface, while the muscles of the upper part of their legs are emphasized by two large scrolls. The application of lacquer with red or brown dots against a black surface was used in ancient China by lacquer artists before they could mix natural pigments of other colors with raw lacquer, in the fifth century bc (Thote 1990). The posture of the animal and the two large spirals used to represent its legs can be associated immediately with the animal style of nomadic people. Also, the representation of a boar with such naturalism testifies to borrowings from the art of peoples living to the west and north of Qin.9 This piece cannot have been imported from another area of the Zhou cultural sphere. Neither does the lacquer technical process seem to have been used by nomadic people, even by those who lived close to the Chinese peripheral regions. It is therefore most probable that the ornament in question was made in Qin, in spite of the strong influence from nomadic art. Apparently, the craftsmen from Qin, whatever their specialty, were open to artistic traditions from their nomadic or semi-nomadic neighbors. Other pieces from the same burial, such as the gold ornaments, are also of nomadic origin.10 The nomadic art influence was important enough to leave a deep impression on Qin art, which lasted until the end of the Warring States period.

Sichuan, a key region in the development of the lacquer art of Qin

A few Warring State period sites in Sichuan have provided lacquerwares produced by workshops under Qin administration.11 One of these sites

9 In animal style artworks, the impression of speed given by the posture of the rear and front legs bent towards each other has been described most accurately by Schiltz (1994). On the appropriation of nomadic elements by Chinese craftsmen, see also Jacobson (1988). 10 I wish to express my deep gratitude to Professor Han Wei 韓偉, Director of the Shaanxi Provincial Institute of Archaeology, for giving photographs of these valuable objects to me in 1988. Such taste influenced by nomadic art in Qin could be exempli-fied by several other examples. 11 Among these sites, the cemetery of Chengdu Yangzishan 羊子山 shows strong


in particular seems to belong to Qin immigrants or to people living in a Qin cultural environment (Song Zhimin 1984).12 This is the cemetery of Qingchuan Haojiaping 青川郝家坪 (Sichuan 1982, Mase 1984) located at the junction of Sichuan, Gansu 甘肅, and Shaanxi 陜西 provinces.13 The two rivers that flow close to the site, the Bailongjiang 白龍江 and the Jialingjiang 嘉陵江, not only connected Sichuan with the valleys of southern Gansu, but also served as trade routes with the area of Baoji in Shaanxi through the Qinling 秦嶺 mountain range. The cemetery comprised about a hundred tombs, of which 72 were excavated in 1979–1980. Since the tombs do not overlap, one can ascer-tain from their furniture that the cemetery remained in use continuously during a rather brief period that can be estimated to have lasted slightly less than a century. Forty-five tombs had an outer coffin (guo 槨) and an inner coffin (guan 棺) [Fig. 4]. The structure of the tombs and their protection with thin white clay are reminiscent of the burial practices followed in the Chu kingdom (Thote 2000). The slightly later Qin tombs from Yunmeng Shuihudi 雲夢睡虎地 in Hubei 湖北 have confirmed that the Chu burial model was adopted by people living under the rule of Qin during the third century bc (Hubei 1981). Earlier on, by the fourth century bc, Chu burial practices had been partly adopted in the Chengdu plain in Sichuan, as exemplified by the large tomb from Xindu Majiaxiang, dating to about the third quarter of the fourth century bc (Thote 2001: 214–215).14 Some of the features of the tombs at Haojiaping are typical of third-century bc Qin tombs. For instance, a piece of silver birch bark had been placed on the lid of the outer coffin of Tomb 1, in exactly the same way as on the outer coffin of Tomb 7 at Yunmeng Shuihudi, dating to 256 bc [Fig. 5]. This Qin custom was completely unknown in the Chu kingdom, as well as in the local traditions of Ba and Shu. Some pieces of furniture are also typical of Qin, such as a hu vessel with garlic-shaped mouth, several types of pottery jars, and bronze cooking pots (fu 釜). Another specificity of Qin tombs is the absence or near-absence of bronze vessels and weapons, a striking difference from Chu

affinities with Qin culture. Tomb 172 contained seven lacquers, two round boxes, two zun, and five round or square containers. Some of their wooden cores were strength-ened by bronze fittings inlaid with silver. However, the tomb–the dating of which is still problematic—is probably of the late third century bc (Qin dynasty). See Sichuan 1956: 1–20, and Plates 7 and 8. 12 Mase Kazuyoshi (1984) is of the opinion that the owners of the tombs at Qingchuan Haojiaping were people from Chu forced to migrate to this area after 278 bc. But as we shall see, the cultural environment of the people buried in the cemetery seems to be mainly of Qin character. 13 As to other Warring States sites from Sichuan, Qin lacquers have also been dis-covered, though in far smaller numbers, at Longquanyiqu Beigandao 龍泉驛區北干道 Tomb 21, in Chengdu Municipality (Chengdu 2000: fig. 17.6 p. 29). 14 Less precise dating is proposed by Song Zhimin 1994: 245.

344 ALAIN THOTE

Figure 4. Plan of Tomb 1 at Qingchuan Haojiaping (Sichuan). After Sichuan 1982, p. 2 figs. 3–6.


Figure 5. Plan of Tomb 7 at Yunmeng Shuihudi (Hubei). After Hubei 1981, p. 5 figs. 4–5; p. 6 figs. 6–7; and p. 9 fig. 13.

and Ba-Shu tombs of the fourth and third centuries bc. This lack of bronze, in addition to a general use of cheap substitutes (mingqi 明器), seems to be a dominant feature of Qin burials until the beginning of the empire, with the exception of rulers’ tombs.15

15 Throughout the Eastern Zhou period there was an increased use of mingqi vessels in Qin tombs, whether in bronze or in ceramic. For details, see Falkenhausen 1999a: 489–494.

346 ALAIN THOTE

Figure 6. Lacquerwares from Qingchuan Haojiaping (Sichuan): eared cups from Tomb 26 (1), and Tomb 37 (2, 3); basin yu, Tomb 1 (4); spatula, Tomb 1 (5); rounded boxes turned on a lathe, from Tomb 2 (6), and Tomb 13 (6); zhi goblets, bent wood and lac-quer, from Tomb 26 (8), and Tomb 2 (9); hu vessel, wood and lacquer, from Tomb 1 (10), third century bc. After Sichuan 1982, p. 16 figs. 38, 39 (1, 2); p. 15 fig. 34 (3) and fig. 33 (4); p. 16 fig. 46 (5); p. 10 fig. 18 (6); p. 21 fig. 60 (7); p. 16, figs. 45 and 42 (8, 9);

p. 15 fig. 32 (10).


Figure 7. Lacquerwares from Qingchuan Haojiaping (Sichuan): lian boxes, bent wood with lacquer and dry lacquer, third century bc. Tomb 41 (1, 2), Tomb 23 (3), Tomb 26 (4); combs, wood and lacquer, Tomb 13 (5). After Sichuan 1982, p. 15 figs. 37, 35, 36

(1, 2, 3); p. 16 fig. 41 (4); p. 10 fig. 16 (5).

At Haojiaping, 177 lacquer objects were found, amounting to slightly less than half of the total burial furniture at the site [Figs. 6, 7]. Most of them have a wooden core. Hemp cloth was pasted on the surface of some of the cores before the lacquer was applied, perhaps to protect the object against deformation under changes in temperature or humidity. The manufacturing techniques show considerable variety. The wood was cut from a block and subsequently carved with chisels, or turned on a lathe, or cut into a thin board that was then bent under certain condi-tions of temperature and humidity. The use of the lathe seems to have appeared during the fourth century bc. It creates a wooden core with rather thick walls, contrasting with the thin cylindrical containers that result from the technique of bent wood, which was probably invented after the lathe came into use. In particular, it seems that the regions situ-ated west of the Zhou cultural sphere had a special taste for cylindrical shapes, since several cylinder-shaped bronze vessels, probably imitating objects made of wood or bark, have been discovered in those regions. Among the earliest and most interesting examples is a toilet-box from Tomb 20 at Baoji Zhuyuangou (Shaanxi) that was made of both wood

348 ALAIN THOTE

and bronze [Fig. 8] (Lu and Hu 1988: 192 and p. 193, fig. 143.5).16 In the fourth and third centuries bc, many cylindrical containers had a core made of wood, bamboo, or fabric. Making use of fabric, the so-called dry lacquer technique was invented during the fourth century bc. It consists of a fabric core shaped with a kind of mold. On the fabric, which is generally hemp, layers of raw lacquer were applied one after another. Once one layer had become dry, it was smoothed, and then several more layers were successively applied by the same process until the core had become firm and was ready to receive painted decoration. The objects made in this way are light, thin, and resistant (Garner 1979: 34). Since it required a lot of time and incurred considerable cost, the technique was mainly used for luxury objects. At Qingchuan Haojia-ping, the colors of the lacquers are red, black, and brown. Some of the

Figure 8. Toilet-box from Tomb 20 at Baoji Zhuyuangou (Shaanxi), wood and bronze, Western Zhou Dynasty (ca. 1050–771 bc). After Lu and Hu 1988, p. 193 fig. 143.5.

16 Two Western Zhou cylindrical bronze you 卣 of two different sizes come also from Zhuyuangou Tomb 13 (Lu and Hu 1988: 63–65, and pp. 62–63 figs. 49–50). Compa-rable examples, also of two different sizes, have been found at Lingtai Baicaopo 靈台白草坡 Tomb 1 and Tomb 2, in eastern Gansu (Gansu 1977: 110, and p. 107 fig. 7.5). Both tombs are dated to early Western Zhou. Apparently the manufacture of such cylindrical containers was not limited to the western part of the Zhou cultural sphere. A few early Western Zhou you vessels of this shape came from other areas of China, for example from Tengzhou Qianzhangda 滕州前掌大 Tomb 119 in Shandong (Shandong 2000: 18 and p. 18 fig. 5.2).


wares are inlaid, while others have decoration incised with a needle. The burial furniture from Hao-jiaping comprised 86 eared cups, which were very common during the Warring States period. Aside from these, many pieces are more typical of Qin culture: long oval-shaped boxes (12), flat wine con-tainers or bian hu 扁壺 (5), several bowls (yu 盂), cylindrical boxes or lian 奩 (55), and zhi 卮 goblets. The lian boxes have flat or rounded lids, and their diameters range between 13 and 20 cm; that is to say, they are much smaller than the Western Han period lian from Changsha Mawangdui 長沙馬王堆 , which measure between 26 and 35 cm in diameter. At least thirteen lacquer objects, about seven percent of the total, have workshop marks that were either stamped on the core before the application of lacquer or incised with a needle on the lacquered sur-face [Fig. 9]. The stamped marks were applied on the wooden surfac-es of the objects with seals heated on fire. The two characters Cheng ting 成亭 (guardpost of Cheng[du]) appear on at least three items with such marks, testifying to the existence of workshops in Chengdu. According to Shen and Huang (1987: 187), the characters belong to the Qin type. Most of the marks are composed of one or two Chinese characters, whereas some of them are mere signs without any relationship to Chinese written characters. It must be noted that not a single mark resembles the emblems of Ba and Shu. Curiously, with the exception of the Cheng ting mark, it seems that the

Figure 9. Marks incised in the lacquer surface or branded with a seal on the core surface, Tombs 1, 2, 26, 31, 36, 37, and 40 at Qingchuan Haojiaping (Sich-uan). After Sichuan 1982, p. 9 figs. 15.

350 ALAIN THOTE

same mark never appears on any two wares, and that any one object may have several different marks, up to five in one case. Moreover, the wares on which marks have been applied belong to several different categories. Clearly, the marks testify to the control that was exercised on the craftsmen in order to regulate either the quality of their work or the supply to markets. The administration was responsible for the ap-plication of seals on the wooden surfaces of the objects, while the marks incised by the craftsmen held them accountable for the quality of their work. In the first case, the control was exerted upstream (at an early stage in the making of the objects), and in the second case downstream (as the objects were completed and could be sold), in the flow of production. The legal documents from Tomb 11 at Yunmeng Shuihudi, which date to 217 bc at the latest, have revealed that the quality of the raw lacquer and manufactured wares, as well as the quantity of production, had to meet precise standards.17 Since the use of marks stamped or incised on lacquerwares existed already in the first half of the third century bc, as shown by the lacquer assemblage from Tomb 7 at Yunmeng Shuihudi, one may assume that such laws had already been in practice since at least the first half of the third century bc. In effect, Tomb 7 at Yunmeng Shuihudi, which dates to 256 bc, contained fourteen lacquer vessels, of which ten had a brand—about 71% of the total (Hubei 1981: 27–37; 104–138, table 2). Tomb 11 at the same site, dating to 217 bc, had 37 lacquer objects, of which 33 showed incised and stamped marks, amounting to 89% of the total. By comparison, the lacquers from the Haojiaping cemetery are less numerous, and they betray a less systematic administrative practice. The marks from Qingchuan could be interpreted as reflect-ing a first step in the organization of Qin administration in Sichuan. In other words, the laws concerning the work of the craftsmen in the Qin kingdom and thereafter in the empire became stricter during the second half of the third century bc. The low rate of the occurrence of marks at Haojiaping could also indicate that at an earlier stage, only lacquer products made under certain conditions could be or had to be branded. Other sites from Sichuan testify to the development in the Chengdu basin of lacquer workshops operating under the control of the Qin administration during the third century bc. For instance, in a cem-etery situated about 150 km to the southwest of Chengdu, at Yingjing Zengjiagou 熒經曾家溝, many comparable lacquerwares were found in

17 Hulsewé 1978, in particular p. 46 n. 8; p. 57 (A52, A53); p. 97 (B22), p. 110 (C11); Yamada 1990; Lewis 1999: 610. The making of Qin weapons was also subject to bureau-cratic control (Sumiya 1982).


a few tombs (Tombs 11 to 16 and Tomb 21) [Fig. 10].18 Their similarity with the lacquers from Qingchuan Haojiaping can be interpreted in terms of cultural uniformity, as a result of the colonization of Sichuan

18 Sichuan 1989, and Sichuan 1984: 1072–1077. Tomb 1 at Yingjing Guchengping 滎經古城坪 seems to be of the Qin dynasty. See Yingjing 1981: 70–74, and Sichuan 1984: 1079 fig. 10.3 (inscription on the base of a round box from Tomb 1 at Guchengping). The Warring States boat-coffins from Yingjing Tongxincun 滎經同心村 also contained round boxes and eared cups, which were all made with the dry lacquer technique (Beijing 1998b).

Figure 10. Lacquers from Tomb 21 at Yingjing Zengjiagou (Sichuan), lacquered wood, second half of the third century bc. After Sichuan 1989, p. 25 fig. 9.

352 ALAIN THOTE

by Qin immigrants. If we now compare these tombs with Tombs 7 and 11 at Yunmeng Shuihudi in Hubei, datable to the third quarter of the second century bc and located a good thousand kilometers away from the Sichuan sites, this cultural uniformity is even more conspicuous (Hubei 1981). This also suggests that the burial furniture and especially the lacquers from Haojiaping should be dated to the first half of the third century.19

The lacquer tradition in Chu in the late fourth and early third centuries

Located near present-day Jiangling 江陵, the Chu capital city of Ying 郢 was seized and destroyed by the armies of Qin in 278 bc. From that time on, the elite of Chu lost most of the Hubei area, and the city never recovered until the Qin dynasty. Economic wealth was passed on to other areas such as Changsha, which greatly benefited from this event. Meanwhile, the lacquer workshops from Jiangling probably ceased operation. Indeed, most of the tombs in the area of Jiangling that can be dated to between 278 bc and the late Warring States period are rather impoverished. By contrast, prior to 278 bc, the numerous wealthy tombs in the Jiangling area have provided us with very large quantities of lacquerware of the highest quality. These tombs, listed in chronological order, are Tomb 1 at Tianxingguan 天星觀 (Hubei 1982), Tombs 1 and 2 at Wangshan 望山 (Hubei 1996), Tomb 1 at Baoshan 包山 (Hubei 1991), Tomb 1 at Shazhong 沙塚 (Hubei 1996), and finally Tomb 1 at Mashan 馬山 (Hubei 1985). To these wealthy tombs can be added several cemeteries from the same period, whose tombs were less well furnished, in particular those at Yutaishan 雨台山 and Jiudian 九店 (Hubei 1984; Hubei 1995). The excellent state of preservation of the lacquers from Chu is due to several factors, including the methods used by the Chu elite to protect their deceased against decay, which continually improved from the fifth century to the Han dynasty. The burial furniture benefited also greatly from this improvement (Guo 1995; Thote 2000). The lacquers found in Chu and Qin tombs of comparable size show great differences. Firstly, the lacquer sets are much richer and more di-verse in Chu. Chu burial furniture not only comprises all kinds of dishes, such as eared cups, food boxes, and containers, but also chests, musical

19 The dating proposed by the excavators in the brief report on the Haojiaping cemetery is Middle Warring States for the earliest tombs and Late Warring States for the latest (Sichuan 1982). According to Michèle Pirazzoli-t’Serstevens (2001: 41), the whole cemetery can be more precisely dated to the first half of the third century bc.


instruments, tomb guardians (zhen-mushou 鎮墓獸 ), pieces of furniture for the house (armrests, or backrests, pillows, tables) and weapons (shields, bows, etc.).20 Obviously, some of these pieces were specific to Chu, such as the dou footed cups for offering food [Fig. 11], the multi-tiered boxes for eating and drinking, and the round food contain-ers with rounded or flat lids [Fig. 12]. Some of the latter had bronze feet and fittings; sometimes, entire vessels were imitated in bronze, and their painted decoration was rendered in gold and silver inlay, testifying to increasing elite taste for lacquerware from the fourth century bc on [Fig. 13]. The dominant artistic trends now come from the lacquer decoration, and no longer from the bronze decoration. The techniques used in Chu to prepare the cores of the lacquer con-tainers are generally simpler than in Qin. The great majority of cores were carved from a single piece of wood. Even the complex shapes of vessels or containers with thin walls were carved with chisels, instead of being made with more sophisticated techniques. These time-consuming techniques account for the improvement of carv-ing tools during the later half of the fourth century bc, probably made possible by the greater availability of iron. In general, the carvings of the Chu craftsmen were exceptionally sharp and energetically drawn. The round boxes and containers were turned with a lathe, just as it was the case in Sichuan at the same time. However, very few objects were made with the bent-wood technique. To achieve cylindrical shapes, one easy solution would be to use bamboo, as done in a double cup from

Fig. 11 Lacquered wood dou container for presenting food, Tomb 33 at Jiang-ling Jiudian (Hubei), ca late fourth or early third century bc. After Hubei 1995, p. 277 fig. 181.

20 On Warring States lacquer craftsmanship, see Ma Wenkuan 1981 and Thote 1990; on Chu lacquerware, see Chen Zhenyu 1991, Hong Kong 1994, and Hou Dejun 1995.

354 ALAIN THOTE

Figure 12. Rounded lacquer containers, wood carved and turned on a lathe, and lacquered, bronze fittings, Tomb 438 at Jiangling Jiudian and Tomb 1 at Jingmen Baoshan (Hubei), fourth to third century bc. After Hubei 1995, p. 270 fig. 176, and

Hubei 1991, p. 34 fig. 20.1.


Tomb 2 at Baoshan [Fig. 14] (Hubei 1991: 137–141).21 Almost no lian made in the bent-wood technique were found in the Jiangling area, with the exception of one piece at Jiudian (Tomb 712) [Fig. 15] and another one from Tomb 1 at Mashan [Fig. 16.3] (Hubei 1995: 288 fig. 189, and Hubei 1985 fig. 64.3). A few dry lacquers coming from the Jiangling area can be mentioned, but it appears that this technique was used only for luxury pieces. An example are the two lian boxes from Tomb 2 at Baoshan, dated to 316 bc, which is the terminus ante quem for this technique (Hubei 1991: 144–146).22 The more famous of the two pieces is decorated with a narrative scene representing an embassy from one state being greeted by an officer of another. In this painting, there are six colors in addition to black.23 The decoration on the upper part of the cover is clearly of Chu style, being close to inlaid bronze vessels of the same provenance [Fig. 17]. Another dry lacquer piece was found at Jiangling Mashan Tomb 1. It is a toilet-basin ornamented with a decoration using remarkable optical effects [Fig. 16.1]. The design at the

21 The tomb also contained a thin cylindrical box, almost 29 cm high, made of bamboo (Hubei 1991: 147, and fig. 90). 22 Apparently, the scabbard from Tomb 1 at Jiangling Wangshan that is described as made with the dry lacquer technique (Hubei 1996:106), actually has a wooden core (Tōkyō 1998: 82). 23 The technique is described in Hubei 1991: 144–146. See also Wu Hung 1999: 705, and Thote 1999: 207.

Figure 13. Container for food, bronze inlaid with gold and silver, and lacquered in red (inner part of the vessel), Tomb 2 at Jingmen Baoshan (Hubei), ca. mid-fourth

century bc (date of burial: 316 bc). After Hubei 1991, p. 190 fig. 120.

356 ALAIN THOTE

center of the basin is composed of geometric and zoomorphic elements (dragons and birds), and it is based on the opposition of black and red colors. If one focuses on the black elements, then a certain composition appears; if one concentrates on the red elements, another compositional scheme becomes evident. The interplay between both compositions compels the viewer constantly to adjust focal lengths (Hubei 1985: 79–80, and p. 78 fig. 64.3; Thote 1996: 159 and p. 161, Thote 2003: 81–82). In the Chu kingdom, the production of luxury lacquerwares devel-oped during the fifth and fourth centuries bc. The items issued from the few workshops that specialized in this kind of production are conspicu-

Figure 14. Double cup, lacquered wood and bamboo, Tomb 2 at Jingmen Baoshan, ca. mid-fourth century bc (date of burial: 316 bc). After Hubei 1991, p. 140 fig. 85 (C).


24 One mark is made on a wooden ji 几 stand, which is not lacquered. Also, the marks on the three eared cups from Jiangling Jiudian seem to relate to the owner, not to the craftsman. One inscription, da guan 大官 (high officer), was inscribed on a cup in red with a brush. 25 This is true except for one of the inscriptions, the reading of which has initiated some controversy. It was found in Tomb 1 at Jingmen Guodian 荊門郭店. While some read it as “the master of the [crown prince] of the Eastern palace” (Dong gong zhi shi 東宮之師) (Li Xueqin 2000: 110, Luo Yunhuan 2000), other specialists think that it can be read as the “cup of the Eastern palace” (Dong gong zhi bei 東宮之杯) (Wenwu 1997.7 p. 41, Li Ling 2000: 47). The second interpretation seems much more appropriate.

ous for the inventiveness in the decoration and variety of techniques. However, before 278 bc, the pieces from these workshops were few and none of them has a workshop mark, whether incised or stamped. Some Chu pieces of lesser quality than those luxury items do have marks. They amount to twelve examples from Jiudian24 and six from Yutaishan (Hubei 1984: 93, and pl. 52). Those wares were not excavated from wealthy tombs. All of them have round shapes, being boxes, zun 樽 containers, and zhi 觶 goblets. All the marks, except one, were incised on the lacquer surface. Contrary to most of the Qin marks, they tend to have no relationship to Chinese writing.25 The one exception is different from the others in that it has been made on a wooden core with a seal heated on a fire be-fore the lacquer was applied against the surface. It is placed on a zun with metal fittings and its mark consists of two characters, Xian ting 咸亭, indicating that the vessel was manufactured in Xianyang 咸陽, the capital of Qin, and controlled by the officers in charge [Fig. 18] (Hubei 1995: 269, and fig. 177). This is, therefore, actually a Qin product.

Figure 15. Lian from Tomb 712 at Jiangling Jiudian (Hubei), lacquered bent wood, ca. late fourth or early third century bc. After Hubei

1995, p. 288 fig. 189.3.

358 ALAIN THOTE

Figure 16. 1. toilet basin, dry lacquer; 2. rounded box for food, wood turned with the lathe (outer surface) and carved (inner part); 3. lian box containing earthware cakes (mingqi), bent wood. From Tomb 1 at Jiangling Mashan (Hubei), ca. late fourth or early

third century bc. After Hubei 1985, p. 78 figs. 64.5, 64.3, and p. 79 fig. 65.5.

Relationship between the lacquers from Qin and Chu

The Qin and Chu lacquer-making traditions show many contrasting characteristics (Li Zhaohe 1980, Yuan Wenqing 1987, Chen Zhenyu 1991: 52–53).26 The Qin pieces are more limited typologically; more-over, within any one category of objects, the range of shapes is much narrower among the Qin lacquers than those of the Chu. This accords with the “Statutes for Artisans” (gonglü 工律) discovered in Tomb 11 at Yunmeng Shuihudi (217 bc), which gives a set of regulations stating that objects of the same type must have the same dimensions. There are further laws concerning the borrowing of tools by craftsmen and the application of marks. As a principle, the artisans were considered to be responsible for their own work. Consequently the quality of their work was controlled by officers, and they could be charged fines,

26 However, Hou Dejun (1995: 246–247) is convinced that the lacquer craftsmanship in Ba-Shu developed under strong Chu influence, in spite of their strong differences.


Figure 17. Lian box from Jingmen Baoshan Tomb 2. Dry lacquer technique, ca. mid fourth century bc (date of burial: 316 bc). After Hubei 1991, p. 144–145 figs. 89 (A)

and (B).

Figure 18. Zun from Tomb 483 at Jiang-ling Jiudian, lacquer on wood and bronze fittings. Late third century bc. After

Hubei 1995, p. 271 fig. 177.

360 ALAIN THOTE

which undoubtedly would have been rather heavy.27 The division of labor within the Qin workshops is visible, furthermore, in the different names given to the specialized craftsmen and to the officials who were responsible for the work. As already stated, the making of lacquers in Qin implied a wider use of elaborate techniques than in Chu, in particular the dry lacquer process and the bent-wood mode of fabrication. More surprisingly, these refined techniques were commonly used for objects that were put into small- or medium-sized tombs of Qin. By contrast, the lacquers made in Chu involved less sophisticated techniques, although Chu society was, as a general rule, much wealthier than Qin society. Another difference between both traditions can be observed in the decoration of the lacquers. Qin decoration often shows representations of animals right at the center of the composition in a panel. In other cases, the representation of an animal occupies the whole panel, as in the case of combs [Fig. 7.5], lian [Figs. 7.2–7.3], and eared cups [Figs. 6.1–6.3]. By contrast, the compositions on Chu lacquers are mainly geometric, although the decorative units are made of stylized dragons and birds. Those zoomorphic motifs are so distorted that they cannot be identified at first glance [Figs. 13, 16.1, and 17]. The main lines of the compositions are diagonals, made of straight lines mixed with spirals, and the areas defined by these diagonals are filled with contrasting colors. The visual effects created in this way remind one of optical games, as previously noted (Thote 1996: 159, 161). Besides the stylistic differences between Qin and Chu lacquers, products from the Qin workshops bear, as already noted, many more makers’marks than the Chu wares. Moreover, the increasing number of marks during the third century bc indicates that a system of control must have grown to such an extent as to have become institutionalized by the Qin rulers. During the same time, Chu marks remained very small in number, and their use seems always to have been marginal. This is also confirmed by the fact that no brand made with a heated seal on the core (laoyin 烙印) has to date been identified on Chu lacquers. The same discrepancy could be observed on ceramic jars [Fig. 19]. If we compare Qin and Chu jars, the former were often marked with a seal before being fired, while the latter do not show any marks (Thote 1990: 85). It is therefore very probable that the few marks found on Chu lacquers had no relationship with local administration, but had been incised either by craftsmen, shopkeepers, or even owners without being intended to be checked by officers from the administration. Instead,

27 Beijing 1978, in particular pp. 69–73 (gonglü); 112–126 (xiaolü 效律); and pp. 127–148 (Qin lü zachao 秦律雜抄), Hülsewé 1985, Yamada 1990.


Figure 19. Qin earthenwares with the mark ting (guard-post), from Fengxiang Xian Gaozhuang Tomb 1 (Shaanxi).

After Yongcheng 1980, p. 12 figs. 3, 4.

the use of clay seals in Chu that is evidenced from the fourth century bc on was probably limited to the goods to be placed in storage rooms [Fig. 20] .28 Besides the clay seals, the earliest tallies ever found in China, dat-ing to 323 bc, come also from Chu. Documents of this kind, which would be issued every year at the Chu capital, allowed the bearer to be exempted from tolls while he was trading along certain land and water routes (Falkenhausen 1999b). In this case, the person to whom such documents were issued was a high officer from Chu named Ejun Qi 鄂君啟, who probably delegated to his own people the transportation and selling of goods. As to marks made by craftsmen, whether they specialized in lacquer products or in pottery, it seems that their use in Chu was not confined to objects of high quality, nor were they intended to facilitate the control

28 The earliest examples in the Chu kingdom are from Jiangling Tianxingguan Tomb 1, ca. 350 bc (Hubei 1982: 107 fig. 31.4) and Jingmen Baoshan Tomb 2, 316 bc (Hubei 1991: 196–197). It must also be mentioned that Chu was the only state to cast gold plates stamped with seals.

362 ALAIN THOTE

of the craftsmen’s work. It was mainly the content of the containers to be sold or preserved untouched, and the circulation of goods, that were under control. It is also possible that before 278 bc, under the pressure of political events, Chu rulers had neither the time nor the opportunity to organize a system of control comparable to the system elaborated in Qin. Later on, the urgency in responding to attacks by Qin armies

Figure 20. Chu earthenwares with clay seals, from Tomb 1 at Jiangling Tianxing-guan (Hubei), ca. mid-fourth century bc (1, 2), and from Jingmen Baoshan Tomb 2 (date of burial: 316 bc) (3, 4). After Hubei 1982, p. 107 fig. 31.1, 2, 4; and Hubei

1991, p. 196 fig. 126 and 197 fig. 127.


and its consequences did not allow Chu to engage in administrative reforms. In spite of their contrasting properties, the material cultures of Chu and Qin had some common relationship. The Yangzi 揚子江 and Han 漢水 rivers were convenient trade routes, first between Sichuan and west central Hubei 湖北, and later between southern Shaanxi and north central Hubei. The influence exerted by Chu on Ba and Shu cultures in the fourth century bc left traces in Sichuan that could in turn have been influential on Qin after the conquest of Sichuan. That Qin lacquerware could circulate in the Chu area is illustrated by Tomb 1 at Jiangling Mashan. At least two lacquer objects found there, a lian and a round box, were likely manufactured by Qin workshops during the early third century bc [Figs. 16.2–16.3]. However, such examples were still rare before the conquest of central Hubei by Qin. After 278 bc, a new cultural context developed in the middle Yangzi region. One suspects that after the destruction of the former capital of the Chu kingdom near Jiangling, Chu lacquer workshops ceased their activ-ity, or, if they could maintain production, were reorganized according to the new standards imposed by the Qin rulers and their administration on the inhabitants of the area. The court of Chu escaped from the Qin armies and settled in a new capital in eastern Henan 河南, while many inhabitants of Ying had to flee to northern Hunan 湖南. In fact, the economic prosperity of Changsha began from that time on, possibly right after the arrival of wealthy Chu immigrants. Between 1952 and 1994, Chinese archaeologists excavated more than two thousand Chu tombs in Changsha and its surburbs. Many of the discoveries testify to the prosperity of Changsha in the third century bc.29 A recently published report (Hunan 2000), synthesizes all the work Chinese archaeologists have conducted in this area.30 From this synthesis, one can see that excavated lacquer objects illustrate the new economic situation in the material culture of the Chu people between 278 and 223 bc. First, 416 pieces come from 130 tombs, which makes an average of about three lacquer items per tomb. The figure of 130 tombs seems low, by comparison with earlier Chu cemeteries near Jiangling, since it represents only six percent of the tombs discovered in Changsha.31 Second, the 174 eared cups from Changsha show a major difference

29 For a reassessment of the dating of the Changsha tombs, see Wagner 1987. 30 It seems likely that some of the tombs said to be of Chu people are slightly later, dating either to the Qin dynasty, or even to the early Han. 31 These are approximate figures, since the amount of lacquer objects found in a tomb depends on the conditions in which they were preserved. In fact, as Chu burial practices were considerably improved during the third century bc, only a small part of the pieces should have decayed over time.

364 ALAIN THOTE

with those from Jiangling. The inner part of the Changsha cups is cov-ered with decoration, while the inner part of the Jiangling cups is in general plain red or, in rare cases, black [Fig. 21]. Moreover, the inner decoration of the cups from Changsha and Sichuan are very close [compare with Figs. 7.1–7.3]. Third, another new tendency can be observed from the lacquer con-tainers found at Changsha. Some types had a low frequency at Jiangling while they are quite well represented at Changsha.32 For example, 23 cylindrical zun with thin walls, bronze fittings, and mounted on three bronze feet were found at Changsha, with only one unique example at Jiangling Jiudian. New categories in the lacquer assemblages at Changsha include rounded boxes (5), lian boxes (14), and pan 盤 basins for water (6). Besides these new shapes, there were only two dou footed cups, the category that was best represented earlier in Chu tombs to-gether with the eared cups. Actually, all these shapes are precisely those representative of Qin lacquer craftsmanship, in Sichuan as well as at Yunmeng Shuihudi. Through these new categories of lacquer objects

Figure 21. Two lacquer eared cups from Changsha Tomb 185 (Hunan). After Hunan 2000, p. 348 fig 278.

32 Since the conditions of preservation are equally good in Jiangling and in Changsha, one may presume that the proportions of different lacquer types we see today are more or less representative of those at burial.


Figure 22. 1a. Lian lacquer boxes from Tomb 1195 at Changsha, Hunan (earlier labeled Tomb 3 at Changsha Wulipai), third century bc. After Hunan 2000, p. 366 fig. 294. 1b

and 1c. After Hunan 2000, vol. 2, plate 117.2.

in Chu tombs at Changsha, it is clear that new commercial exchanges and trade routes had been established. Furthermore, one may suggest that lacquer workshops from Sichuan operating under the control of the Qin administration greatly benefited from the conquest of central Hubei by Qin in 278 bc. The techniques involved, which allowed the manufacturing of light and thin wares, and the refined decoration of these wares, were probably considered very attractive in comparison with earlier Chu lacquers. Some astonishing similarities between lacquer pieces from sites sepa-rated by such distances as those of Qingchuan Haojiaping in Sichuan, Changsha in Hunan, and Yunmeng Shuihudi in Hubei, exemplify the mass production at Qin lacquer workshops and the wide distribution of their products. For example, a lian from Tomb 1195 at Changsha can be compared with another lian from Tomb 7 at Yunmeng Shuihudi, dating to 256 bc [Figs. 22.1–22.2]. In the center of the composition, a very rare theme is represented: a horse whose body is adapted into a circle. The style of the painting has nothing in common with the Chu style of decoration. In reality, this motif reflects a strong influence of the animal style [Fig. 23], an influence we already noted in the lacquer art of Qin by the middle of the sixth century bc.

Figure 22. 2a. Lian from Tomb 7 at Yunmeng Shuihudi, Hubei (256 bc). After Hong Kong 1994, no. 34. 2b. Painted reconstruction of this lian. After Yunmeng Shuihudi

chutu Qin Han qiqi tulu (Wuhan: Hubei Meishu Chubanshe, 1986), p. 93.

Figure 23. Animal style tattoo, kurgan 2 at Pazyryk, in the Altai mountains, ca. fourth century bc. After Rudenko 1953, p. 312 fig. 181.


The exchanges through new trade routes from Sichuan to Changsha that have been observed by the middle of the third century bc seem to have grown continuously during the Qin and early Han periods. Several lacquers from Tomb 1 at Mawangdui 馬王堆, Changsha 長沙 (ca. 168 bc), had marks indicating that they were made in Chengdu (Yu and Li 1975). During the early Han period, the trade routes initiated by Qin from Sichuan into the Middle Yangzi Basin to Jiangling and Changsha were still very active, bringing the lacquer products from Chengdu to what two centuries earlier had been the heart of the Chu lacquer artistic tradition.

References cited

Beijing 1978: Shuihudi Qin mu zhujian zhenglixiaozu 睡虎地秦墓竹簡整理小組 (1978). Shuihudi Qin mu zhujian 睡虎地秦墓竹簡 (The bamboo slips from a Qin tomb at Shuihudi [Hubei province]). Beijing: Wenwu Chubanshe.

Beijing 1989: Hubei Sheng Bowuguan 湖北省博物館 (1989). Zeng Hou Yi mu 曾候乙墓 (The tomb of the Marquis Yi of Zeng). 2 vols. Beijing: Wenwu Chubanshe.

Beijing 1998a: Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所, Shifang Shi Wenwu Baohu Guanlisuo 什邡市文物保護管理所 (1998). “Shifang Shi Chengguan Zhanguo Qin Han muzang fajue baogao 什邡市城關戰國秦漢墓葬發掘報告 (Excavation report on tombs at Chengguan, Shifang Municipality [Sichuan province], of the Warring States, Qin and Han periods).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所, ed., Sichuan kaogu baogaoji 四川考古報告集. Beijing: Wenwu Chubanshe: 112–185.

Beijing 1998b: Sichuan Sheng Wenwu Kaogu Yanjiusuo and Yingjing Yandao Gucheng Yizhi Bowuguan 滎經嚴道古城遺址博物館 (1998). “Yingjingxian Tongxingcun Ba Shu chuanguanzang fajue baogao 滎經縣同心村巴蜀船棺葬發掘報告 (Excavation report on the boat-coffin tombs in the ancient city of Yandao, Yingjing [Sichuan province]).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo, ed., Sichuan kaogu baogaoji, pp. 212–280. Beijing: Wenwu Chubanshe.

Chen Zhenyu 陳振裕 (1991). “Shilun Chu mu chutu qiqi de chandi wenti 試論楚墓出土漆器的產地問題 (On the sites of production of lacquer wares discovered in Chu tombs).” In Chu Wenhua Yanjiuhui 楚文化研究會, ed., Chu Wenhua yanjiu lunji, di er ji 楚文化研究論集, 第二集 (The second volume of collected papers on Chu culture), pp. 44–56. [Wuhan]: Hubei Renmin Chubanshe.

Chengdu 1999: Chengdu Shi Wenwu Kaogu Gongzuodui 成都市文物考古工作隊 and Qingbaijiang Qu Wenwu Guanlisuo 青白江區文

368 ALAIN THOTE

物管理所 (1999). “Chengdu Shi Qingbaijiang Qu Yuejincun Han mu fajue jianbao 成都市青白江區躍進村漢墓發掘簡報 (Excavation report on Han tombs at Yuejincun, Qingbaijiang section, Chengdu municipality [Sichuan province]).” Wenwu 1999.8: 19–37.

Chengdu 2000: Chengdu Shi Wenwu Kaogu Yanjiusuo 成都市文物考古研究所 and Longquanyiqu Wenwu Guanlisuo龍泉驛區文物管理所 (2000). “Chengdu Longquanyi qu Beigandao muguo muqun fajue jianbao 成都龍泉驛區北干道木槨墓群發掘簡報 (Brief exca-vation report on a group of tombs with outer coffins at Beigandao, Longquanyi sector, Chengdu municipality [Sichuan province]).” Wenwu 2000.8: 21–32.

Chengdu 2002: Chengdu Shi Wenwu Kaogu Yanjiusuo (2002). “Chengdu shi Shangyejie chuanguan, dumuguanzang fajue baogao 成都市商業街船棺, 獨木棺葬發掘報告 (Excavation report on the boat-coffins and coffins hewn from logs at Chengdu Shangyejie).” In Chengdu kaogu faxian 成都考古發現 (2000), p. 78–136. Beijing: Kexue Chubanshe.

Falkenhausen, Lothar von (1999a). “The Waning of the Bronze Age: Material Culture and Social Developments, 770–481 bc.” In Mi-chael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China. From the Origins of Civilization to 221 BC, pp. 450–544 (Chapter 7). Cambridge: Cambridge University Press.

Falkenhausen, Lothar von (1999b). “Bronze Ejun Qi jie tally.” In Yang Xiaoneng, editor, The Golden Age of Archaeology: Celebrated Discoveries from the Peoples’ Republic of China, pp. 340–344 (entry 117). Washington: The National Gallery of Art.

Gansu 1977: Gansu Sheng Bowuguan Wenwudui 甘肅省博物館文物隊 (1977). “Gansu Lingtai Baicaopo Xi Zhou mu 甘肅靈台白草坡西周墓 (Western Zhou tombs at Baicaopo, Lingtai district, Gansu province).” Kaogu xuebao 1977.2: 99–130.

Garner, Sir Harry (1979). Chinese Lacquer. London: Faber and Faber.Guizhou 1959: Guizhou Sheng Bowuguan 貴州省博物館 (1959).

“Guizhou Qingzhen Pingba Han mu fajue baogao 貴州清鎮平壩漢墓發掘報告 (Excavation report on Han tombs in Qingzhen and Pingba districts, Guizhou province).” Kaogu xuebao 1959.1: 85–103.

Guo Dewei 郭德維 (1995). Chu xi muzang yanjiu 楚系墓葬研究 (Research on Chu tombs). Wuhan: Hubei Jiaoyu Chubanshe.

Hong Kong 1994: Lacquerware from the Warring States to the Han Periods ex-cavated in Hubei Province. Hong Kong: Hubei Provincial Museum/The Art Gallery, The Chinese University of Hong Kong.

Hou Dejun 後德俊 (1995). Chu guo de kuangye, xiuqi he boli zhizao 楚國的礦冶, 髹漆和玻璃制造 (Mining, lacquer craftsmanship and glass making in the state of Chu). Wuhan: Hubei Jiaoyu Chubanshe.

Hubei 1981: “Yunmeng Shuihudi Qin mu” Bianxiezu “雲夢睡虎地


秦墓” 編寫組 (1981). Yunmeng Shuihudi Qin mu 雲夢睡虎地秦墓(The Qin tombs at Shuihudi, Yunmeng [Hubei province]). Beijing: Wenwu Chubanshe.

Hubei 1982: Hubei Sheng Jingzhou Diqu Bowuguan 湖北省荊州地區博物館 (1982). “Jiangling Tianxingguan yi hao Chu mu 江陵天星觀一號楚墓 (The Chu Tomb No. 1 at Tianxingguan, Jiangling district [Hubei province].” Kaogu xuebao 1982.1: 71–116.

Hubei 1984: Hubei Sheng Jingzhou Diqu Bowuguan (1984). Jiangling Yutaishan Chu mu 江陵雨台山楚墓 (The Chu tombs at Yutaishan, Jiangling district [Hubei province]).” Beijing: Wenwu Chubanshe.

Hubei 1985: Hubei Sheng Jingzhou Diqu Bowuguan (1985). Jiangling Mashan yi hao Chu mu 江陵馬山一號楚墓 (The Chu Tomb No. 1 at Mashan, Jiangling district [Hubei province]).” Beijing: Wenwu Chubanshe.

Hubei 1991: Hubei Sheng Jing Sha Tielu Kaogudui 湖北省荊沙鐵路考古隊 (1991). Baoshan Chu mu 包山楚墓 (The Chu tombs at Baoshan [Jingmen district, Hubei province]).” Beijing: Wenwu Chubanshe.

Hubei 1995: Hubei Sheng Wenwu Kaogu Yanjiusuo 湖北省文物考古研究所 (1995). Jiangling Jiudian Dong Zhou mu 江陵九店東周墓 (The Eastern Zhou tombs at Jiudian, Jiangling district [Hubei province]).” Beijing: Wenwu Chubanshe.

Hubei 1996: Hubei Sheng Wenwu Kaogu Yanjiusuo (1996). Jiangling Wangshan Shazhong Chu mu 江陵望山沙塚楚墓 (The Chu tombs at Wangshan and Shazhong, Jiangling district [Hubei province]).” Beijing: Wenwu Chubanshe.

Hubei 1997: Hubei Sheng Jingmen Shi Bowuguan 湖北省荊門市博物館 (1997). “Jingmen Guodian yi hao Chu mu 荊門郭店一號楚墓 (The Chu Tomb No. 1 at Guodian, Jingmen district).” Wenwu 1997.7: 35–48.

Hubei 1999: Hubei Sheng Wenwu Kaogu Yanjiusuo (1999). “Hubei Jingzhou Jicheng yi, er hao Chu mu fajue jianbao 湖北荊州紀城一, 二號楚墓發掘簡報 (Brief report on the Chu Tombs No. 1 and No. 2 at Jicheng, Jingzhou municipality, Hubei province).” Wenwu 1999.4: 4–17.

Hülsewé, A.F.P. (1985). Remnants of Ch’in Law. Leiden: Brill.Hunan 2000: Hunan Sheng Bowuguan 湖南省博物館 et al. (2000).

Changsha Chu mu 長沙楚墓 (The Chu tombs in Changsha). Beijing: Wenwu Chubanshe.

Hunan 2001: Hunan Sheng Wenwu Kaogu Yanjiusuo 湖南省文物考古研究所 and Yongzhou Shi Zhishan Qu Wenwu Guanlisuo 永州市芝山區文物管理所 (2001). “Hunan Yongzhou Shi Yaoziling er hao Xi Han mu 湖南永州市鷂子岭二號西漢墓 (Tomb No. 2 at Yaoziling, Municipality of Yongzhou, Hunan province).” Kaogu 2001.4: 45–62.

370 ALAIN THOTE

Jacobson, Esther (1988). “Beyond the Frontier. A Reconsideration of Cultural Interchange Between China and the Early Nomads.” Early China 13: 201–240.

Jiangsu 2000: Yangzhou Bowuguan 揚州博物館 (2000). “Jiangsu Hanjiang Xian Yaozhuang 102 hao Han mu 江蘇邗江縣姚莊 102 號漢墓 (The Han Tomb No. 102 at Yaozhuang, Hanjiang district, Jiangsu province). Kaogu 2000.4: 50–65.

Lewis, Mark Edward (1999). “Warring States Political History.” In Michael Loewe and Edward L. Shaughnessy, eds., The Cambridge History of Ancient China. From the Origins of Civilization to 221 BC, pp. 587–650. Cambridge: Cambridge University Press.

Li Ling 李零 (2000). “Guodian Chu jian yanjiu zhong de liang ge wenti—Meiguo Damusi Xueyuan Guodian Chu jian ‘Laozi’ guoji xueshu taolunhui ganxiang 郭店楚簡研究中的兩個問題﹣美國達慕思學院郭店楚簡 “老子” 國際學術討論會感想 (On two problems concerning the bamboo slips of the Chu kingdom from Guodian—Reflections on the international conference at Dartmouth College, United States, on the ‘Laozi’ manuscript from the Chu kingdom discovered in Guodian).” In Wuhan Daxue Zhongguo Wenhua Yan-jiuyuan 武漢大學中國文化研究院, ed., Guodian Chu jian guoji xueshu yantaohui lunwenji 郭店楚簡國際學術討論會論文集 (Essays from the international conference on the Chu bamboo slips from Guodian), pp. 47–52. Wuhan: Hubei Renmin Chubanshe.

Li Zhaohe 李昭和 (1980). “ ‘Ba Shu’ yu ‘Chu’ qiqi chutan ‘巴蜀’ 與 ‘楚’ 漆器初探 (Discussion on lacquers from Ba-Shu and from Chu).” In Zhongguo Kaoguxuehui di er ci nianhui lunwenji 中國考古學會第二次年會論文集 (Essays from the second annual conference of the Society for Chinese Archaeology), pp. 93–99. Beijing: Wenwu Chubanshe.

Li Xueqin (2000). “The Confucian Texts from Guodian Tomb Number One: Their Date and Significance.” In Sarah Allan and Crispin Wil-liams, eds., The Guodian Laozi. Proceedings of the International Conference, Dartmouth College, May 1998, pp. 107–111. Berkeley: The Society for the Study of Early China and The Institute of East Asian Studies, University of California.

Lu Liancheng 盧連成 and Hu Zhisheng 胡智生 (1988). Baoji Yu guo mudi 寶雞A國墓地 (The cemetery of the princes of Yu at Baoji [Shaanxi province]). Beijing: Wenwu Chubanshe.

Luo Yunhuan 羅運環 (2000). “Lun Guodian yi hao Chu mu suo chu qi erbei wen ji muzhu he zhujian de niandai 論郭店一號楚墓所出漆耳杯文及墓主和竹簡的年代 (On the owner and date of the Chu Tomb No. 1 at Guodian [Jingmen, Hubei province]) after the inscription on a lacquer ear cup.” Kaogu 2000.1: 68–71.

Ma Wenkuan馬文寬 (1981). “Lüetan Zhanguo shiqi de qiqi 略談戰國


時期的漆器 (A discussion on lacquers of the Warring States period).” Zhongguo Lishi Bowuguan guankan 中國歷史博物館館刊 3 (1981): 109–114.

Mao Xian 1994: Mao Xian Qiangzu Bowuguan 茂縣羌族博物館, Aba Zangzu Qiangzu Zizhizhou Wenwu Guanlisuo 阿壩藏族羌族自治州文物管理所 (1994). “Sichuan Mao Xian Moutuo yihao shiguanmu ji peizangkeng qingli jianbao 四川茂縣牟托一號石棺墓及陪葬坑清理簡報 (Brief report on cist Tomb No. 1 at Moutuo, Mao Xian, and associated funerary pits).” Wenwu 1994.3: 4–40.

Mase Kazuyoshi 間瀬收勞 (1984). “Shin teikoku keiseikatei no ikko kōsatsu—Shisenshō Seisen sengoku bo no kentō ni yoru 秦帝国形成過程の一個考察—四川省青川縣戰國墓の檢討による (An investigation on the formation process of the Qin empire—Research on the Warring States tombs at Qingchuan, Sichuan province).” Shirin 史林, 67.1 (1984).

Pirazzoli-t’Serstevens, Michèle (1990). “Ateliers, patronage et collections princières en Chine à l’époque Han.” Comptes rendus de l’Académie des In-scriptions et Belles-Lettres des séances de l’année 1990, avril-juin: 521–535.

Pirazzoli-t’Serstevens, Michèle (2001). “Introduction Part 2: Sichuan in the Warring States and Han Periods.” In Robert Bagley, ed., Ancient Sichuan. Treasures from a Lost Civilization, pp. 39–40. Seattle/Princeton: Seattle Art Museum/Princeton University Press.

Prüch, Margarete (1997). Die Lacke der Westlichen Han-Zeit (206 v.–6 n. Chr.). Bestand und Analyse. Francfort am Main: Peter Lang–Europäischer Verlag des Wissenschaften.

Qin Du Xianyang Kaogudui 秦都咸陽考古隊 (1982). “Xianyang Shi Huangjiagou Zhanguo mu fajue jianbao 咸陽市黃家溝戰國墓發掘簡報 (Brief report on the Warring States tombs at Huangjiagou, Xianyang municipality [Shaanxi province]). Kaogu yu wenwu 1982.6: 6–15.

Rudenko, S.I. (1953). Kul’tura naselenija Gornogo Altaja v skifskoe vremja. Moskva-Leningrad.

Sage, Steven F. (1992). Ancient Sichuan and the Unification of China. Albany: State University of New York Press.

Schiltz, Véronique (1994). Les Scythes et les nomades des steppes—VIIIe siècle avant J.-C.–Ier s. après J.-C. Paris: Gallimard.

Shaanxi 1998: Shaanxi Sheng Kaogu Yanjiusuo 陜西省考古研究所, ed. (1998). Long Xian Dianzi Qin mu 隴縣店子秦墓 (The Qin tombs at Di-anzi, Long Xian [Shaanxi province]). Xi’an: San Qin Chubanshe.

Shandong 2000: Zhongguo Shehui Kexueyuan Kaogu Yanjiusuo Shandong Gongzuodui 中國社會科學院考古研究所山東工作隊 (2000). “Shandong Tengzhou Shi Qianzhangda Shang Zhou mudi 1998 nian fajue jianbao 山東藤州市前掌大商周墓地 1998 發掘

372 ALAIN THOTE

簡報 (Brief report on the excavation of Shang and Zhou tombs at Qianzhangda, Tengzhou municipality, Shandong province).” Kaogu 2000.7: 13–28.

Sichuan 1956: Sichuan Sheng Wenwu Guanli weiyuanhui 四川省文物管理委員會 (1956), “Chengdu Yangzishan di 172 hao mu fajue baogao 成都羊子山第 172 號墓發掘報告 (Excavation report on Tomb No. 172 at Yangzishan, Chengdu municipality).” Kaogu xuebao 1956.4: 1–20.

Sichuan 1981: Sichuan Sheng Bowuguan 四川省博物館 and Xindu Xian Wenwu Guanlisuo 新都縣文物管理所 (1981), “Sichuan Xindu Zhanguo muguo mu 四川新都戰國木槨墓 (The Warring States tomb with wooden outer coffin at Xindu, Sichuan province).” Wenwu 1981.6: 1–16.

Sichuan 1982: Sichuan Sheng Bowuguan and Qingchuan Xian Wen-huaguan 青川縣文化館 (1982), “Qingchuan Xian chutu Qin geng xiu Tian Lü mudu 青川縣出土秦更修田律木牘 (Wooden strips on the modifications of the Statute on Agriculture found in Qingchuan Xian).” Wenwu 1982.1: 1–21.

Sichuan 1983: Sichuan Sheng Bowuguan (1983). “Chengdu Xijiao Zhanguo mu 成都西郊戰國墓 (a Warring States tomb in the western suburbs of Chengdu).” Kaogu 1983.7: 597–600.

Sichuan 1984: Sichuan Sheng Wenguanhui 四川省文管會, Ya’an Diqu Wenguanhui 雅安地區文化館 and Yingjing Xian Wenhuaguan 滎經縣文化館 (1984). “Sichuan Yingjing Zengjiagou Zhanguo muqun di yi, er ci fajue 四川滎經曾家溝戰國墓群第一、二次發掘 (The first and second excavation campaigns of the Warring States tombs at Zengjiagou, Yingjing Xian, Sichuan province).” Kaogu 1984.12: 1072–1084, 1091.

Sichuan 1985: Sichuan Sheng Wenguanhui and Dayi Xian Wenhua-guan 大邑縣文化館 (1985). “Sichuan Dayi Wulong Zhanguo Ba-Shu muzang 四川大邑五龍戰國巴蜀墓葬 (The Ba-Shu tombs of the Warring States period at Wulong, Dayi Xian, Sichuan province).” Wenwu 1985.5: 29–40.

Sichuan 1989: Sichuan Sheng Wenwu Guanli Weiyuanhui and Yingjing Xian Wenhuaguan 滎經縣文化館 (1989). “Sichuan Yingjing Zengjia-gou 21 hao mu qingli jianbao 四川滎經曾家溝 21 號墓清理簡報 (Brief report on the excavation of Tomb No. 21 at Zengjiagou, Yingjing Xian, Sichuan province).” Wenwu 1989.5: 21–30.

Song Zhimin 宋治民 (1984). “Lüelun Sichuan de Qin ren mu 略論四川的秦人墓 (On the tombs of Qin people in Sichuan).” Kaogu yu wenwu 1984.2: 83–90, 82.

Sumiya Sadatoshi 角谷定俊 (1982). “Shin ni okeru seidō kōgyō no ikkōsatsu: kōkan o chūshin ni 秦における青銅工業の一考察—–工


官を中心に (An investigation on the bronze craftsmanship of Qin, centrered on the officers in charge of the workshops).” Sundai shigaku 駿台史学 55 (1982): 52–85.

Shen Zhongchang 沈仲常 and Huang Jiaxiang 黃家祥 (1987). “Cong chutu de Zhanguo qiqi wenzi kan "Chengdu" de deming 從出土的戰國漆器文字看 “成都”的得名 (The appearance of the name “Chengdu” based upon the inscriptions found on excavated lacquers of the Warring States period).” In Xu Zhongshu 徐中舒, ed., Ba Shu kaogu lunwenji 巴蜀考古論文集 (Collected papers on the archaeology of Ba-Shu), pp. 186–190. Beijing: Wenwu Chubanshe.

Thote, Alain (1990). “Innovations techniques et diversification des commandes, l’artisanat du laque en Chine aux Ve–IVe siècles avant J.-C.” Arts Asiatiques 45: 76–89.

Thote, Alain (1996). “I Zhou orientali.” In M. Pirazzoli-t’Serstevens, ed., Storia Universale dell’Arte : La Cina, pp. 95–165. Turin: UTET.

Thote, Alain (1999). “De quelques conventions picturales : le char et ses représentations aux Ve–IVe siècles avant notre ère.” Études chinoises 18: 179–220.

Thote, Alain (2000). “Burial practices in the Chu kingdom in the light of recent discoveries: Continuities and discontinuities.” In Roderick Whitfield and Wang Tao, eds., Exploring China’s Past. New Discoveries and Studies in Archaeology and Art, pp. 189–204. Londres: Saffron.

Thote, Alain (2001). “The Archaeology of Eastern Sichuan at the End of the Bronze Age (fifth–third centuries bc).” In Robert Bagley, ed., Ancient Sichuan. Treasures from a Lost Civilization, pp. 203–225 and 230–251. Seattle/Princeton: Seattle Art Museum/Princeton University Press.

Thote, Alain (2003). “Du message à l’image: le décor des bronzes Shang et Zhou (XVème–IIIème s. avant notre ère),” Arts Asiatiques 58: 73–85.

Tōkyō 1997: Tōkyō Kokuritsu Hakubutsukan 東京国立博物館. Dai-sōgen no kiba minzoku—Chūgoku hokuhō seidōki 大草原の騎馬民族—–中国北方の青銅器 (Horse riding peoples of the steppes—Bronzes from northern China). Tōkyō, 1997.

Tōkyō 1998: Tōkyō Kokuritsu Hakubutsukan (1998). Urushi de kakerareta shinpi no sekai—Chūgoku kodai shikki ten 漆で描かれた神秘の世界—– 中国古代漆器展 [A Mysterious world painted in lacquer—Lacquer-ware from the Tombs of Hubei, China]. Tōkyō: Tōkyō Kokuritsu Hakubutsukan.

Umehara Sueji 梅原末治 (1943). Shina Kandai kinenmei shikki zusetsu 支那漢代紀年銘漆器圖説 (Explanations and illustrations of the Chinese lacquers with dated inscriptions of the Han dynasty). Kyōto: Kuwana Bunshōdō.

374 ALAIN THOTE

Wagner, Donald B. (1987). “The Dating of the Chu Graves of Chang-sha: the Earliest Iron Artifacts in China?” Acta Orientalia [Copen-hagen] 48: 111–156.

Wu Hung (1999). “The Art and Architecture of the Warring States Period.” In Michael Loewe and Edward L. Shaughnessy, eds. The Cambridge History of Ancient China: From the Origins of Civilization to 221 BC, pp. 651–744. Cambridge: Cambridge University Press, 1999.

Yamada Katsuyoshi (1990). “Offices and Officials of Works, Markets and Lands in the Ch’in Dynasty.” Acta Asiatica. Bulletin of the Institute of Eastern Culture 58: 1–23.

Yingjing 1981: Yingjing gumu fajue xiaozu 滎經古墓發掘小組 (1981). “Sichuan Yingjing Guchengping Qin Han muzang 四川滎經古城坪秦漢墓葬 (The Qin and Han tombs at Guchengping, Yingjing Xian, Sichuan province).” Wenwu ziliao congkan 文物資料叢刊 4 (1981): 70–74.

Yongcheng 1980: Yongcheng Kaogu Gongzuodui 雍城考古工作隊 (1980). “Fengxiang Xian Gaozhuang Zhanguo Qin mu fajue jian-bao 鳳翔縣高莊戰國秦墓發掘簡報 (Brief report on Warring States and Qin tombs at Gaozhuang, Fengxiang Xian [Shaanxi province]). Wenwu 1980.9: 10–24.

Yu Weichao 俞偉朝 and Li Jiahao 李家浩 (1975). “Mawangdui yi hao Han mu chutu qiqi zhidi zhu wenti—cong Chengdu Shi Fu zuofang dao Shu Jun gongguan zuofang de lishi bianhua 馬王堆一號漢墓出土漆器製地諸問題—–從成都市作坊到蜀郡工官作坊的歷史變化 (The main problems concerning the site of production of the lacquers from Mawangdui Tomb No. 1—the historical development from the workshops of the Chengdu prefecture to the official workshops administred by the Shu Commandery).” Kaogu 1975.6: 344–348.

Yuan Wenqing 院文清 (1987). “Chu yu Qin Han qiqi de ji ge wenti 楚與秦漢漆器的幾個問題 (A few questions concerning the lacquers of Chu and of the Qin and Han periods).” Jiang Han kaogu 江漢考古 1987.1: 64–69.

Personal address

66–68 rue du Faubourg-Poissonnière75010 Paris, FranceE-mail: <[email protected]>


ORDINARY MYSTERIES: INTERPRETING THE ARCHAEOLOGICAL RECORD OF HAN SICHUAN

BY

MICHAEL NYLAN(University of California, Berkeley)

AbstractIn this article I present three ideas designed to challenge us to look harder at the riches that the Han tomb tiles and figurines from Sichuan represent. After noting briefly the ways in which the tomb represents a “well-protected home,” I will argue first that the scenes that now appear most straightforward and easily interpreted may once have communicated far more complex ideas; second, that multiple metaphorical meanings attached to certain images can be established by reference to other sites as far away from Sichuan as Shandong, on the northeastern seacoast; and third, that what we now—quite anachronistically—call the “art of the Sichuan tombs” nonetheless exhibits several noteworthy features that are entirely distinctive to the area.

It is the mysterious and the unusual that tend to spark our interest—not the conventional and the familiar. At first glance, the Sichuan 四川 sculptures and tomb tiles that date to the Han 漢 Dynasty (the same time period as the Roman empire) look familiar, even humdrum. However lively their style, they appear to depict mundane scenes that require little or no explanation. They show, for example, carriage processions, scenes of local industry, the classical masters lecturing to their disciples, and young boys shooting birds with slingshots—all scenes that can be accommodated within our own experiential world. Such pictorial narra-tives do not seem to require our close attention—not when we compare them with the startling Sanxingdui 三星堆 figures or the magnificent ritual drums from Yunnan 雲南. To show what is distinctive and noteworthy about the Sichuan ma-terials, let me begin with what was true across the whole span of the present geographic area known as China in Han times—roughly the last two centuries bc and the first two centuries ad. In Han times, the tomb (whether situated in cliffs or below ground) represented a home for the dead, on an analogy with regular homes for the living, and the place where both the living and the dead gathered together during ritual

376 michael nylan

occasions. (By the canonical prescriptions, the living members of the extended family visited the tombs repeatedly after the first encoffining of the dead to conduct memorial services.) The chief function of tomb architecture and the arrangement of motifs, in consequence, was to establish for all onlookers, living or dead, human or god, the absolute security of the tomb as a “well-protected home.”1 What must strike even the most casual viewer about Sichuan tombs is how strongly they convey this theme both in their decoration and their construction (with husband and wife and often all other members of the family buried together).

1 Cf. Susan Erickson’s essay in this issue; also Christian de Pee (forthcoming).

Figure 1. Rubbing of a Sichuan armory. Stone carving. Height 300 cm, width 275 cm. First or second century ad. Excavated from a tomb at Chengdu Zengjiabao. Chengdu Municipal Museum. After Gao Wen et al. 2000: 7, figs. 43–45 (cf. the stone carving

from Chengdu Tuqiao 土橋, illustrated in Gong Tingwan et al. 1998: fig. 217).

THE ARCHAEOLOgICAL RECORD OF HAN SICHUAN 377

In tomb after tomb, we see images that enhance the protection af-forded the afterlife home. Frequently, pictorial images in the tombs portray guards or armories stocked with weapons, as in Figure 1. Images associated with water, like the lotus, can be read as images to ward off destructive fires, and the tombs depict high walls and watchtowers from which to see potential enemies approach. Exorcists are portrayed, often in the shape of a big bear, and apotropaic mask imagery appears on tomb doors and tower-gates. The tomb builders moreover made good provision for the long-term financial needs of the family after death. Hence, the inclusion among the grave goods of those money-trees that are distinctive to Sichuan. These trees are literally composed of bronze cash and the bases of the trees show them growing more money for the family to harvest (Figure 2) (Erickson 1994). The long-term economic health of the family can also be insured by portraying money-making ventures in the tombs, and plenty of examples in the tombs do just that: one of the most famous pictorial bricks depicts large-scale salt-mining ventures in Sichuan, while another, less famous, depicts the making of wines. The families who could afford this sort of well-protected tomb—all of them headed by men trained in the classics and enjoying wealth from land or trade—intended, no doubt, to live on in some sense after death. By tradition, the most reliable means by which to live on after death, of course, was sacrifices made by descendants to the soul of the

Figure 2. Ceramic base of a money tree. Height 50 cm. Second century ad. Excavated from tomb 176 at Pengshan. After Susan Erickson 1994: fig. 8a (cf.

Sichuan Pengshan 1991: fig. 36).

378 michael nylan

deceased. The belief in this way to achieve a kind of immortality long predates Confucius in Chinese history, being traceable at least to the late Shang 商 period (ca. 1200 bc). Obviously, as a precondition for continuity over the generations, one had to produce descendants, and a number of visual units in the tomb refer to that necessity, including, I would guess, the three fish joined at the head (Figure 3). For three is what we call the “completion number” (as in our own expression, third time’s the charm); and here the three are arranged like the legs of a bronze tripod (an image that symbolized absolute stability). Moreover, the word for fish (yu 魚)—like that for leaves—is a pun of the word for abundance (yu 餘 and 祐) may explain why a sculpted child—a scion of the family perhaps?—fishes for still more fish in one decorated tomb.2 Other allusions to the auspicious production of new generations over time are absolutely unambiguous: there is everything from the gourds (noted for their many seeds, and also for being the gestation site within which yin 陰 and yang qi 陽氣 mysteriously interact to produce life) to the coupled deities Fuxi 伏羲 and Nuwa 女媧, the primeval couple from whom all life comes, and the image of red-budded lotus flower that bears fruit. And after one has produced the children, the love and respect that prevail within the family circle constitute the best assurance that sacrifices will be offered in perpetuity to the dead. And Sichuan scenes—in marked contrast to those from other regions within the Han empire—explicitly allude to that, too (Figures 4 and 5). (Before mov-ing on, I would just point out the wonderful mutuality of the conjugal embrace in Figure 5—and hazard the comment that the presence of such scenes in Sichuan should cause us to question our modern Euro-American stereotypes of female oppression in traditional China.) But even when healthy children born to the family have grown to

Figure 3. Drawing of a three-headed fish motif. See Susan Erickson’s paper in the present volume.

2 A boy fishing at the rafters has been found in cliff tomb no. 1 at Leshan Mahao (Gao Wen et al. 2000: 4, fig. 5–6).


maturity in a nurturing family, it sometimes happens that sons and grandsons are less than perfectly filial, and then they perform the sacrifices in a perfunctory manner, or not at all. For that reason, some men sought also to immortalize themselves through their writings, their deeds, and the monuments erected to their memory. A third option, available only to those who commanded vast resources, was to seek

Figure 4. Ceramic figure of a woman nursing a baby. Height 22 cm. First or second century ad. Excavated in 1987 from a tomb at Deyang Huangxu. Deyang Municipal

Museum (Dezi 87–0072). After Bagley 2001: 328 (lower).

380 michael nylan

Figure 5. Rubbing of a household scene that includes a kissing couple. Panel of a stone coffin. Height, 60 cm, width 62 cm. First or second century ad. Excavated in 1969 from the suburbs of Yingjing 滎經. Yingjing County Museum. After Gao Wen et al.

2000: vol. 7, fig. 114.

Figure 6. Brick with Xiwangmu, attendants, and a worshipper. Height 40.3 cm, width 45.5 cm. Second century ad. Excavated in 1954 from tomb 1 at Xinfan 新繁 Qingbaixiang 清白鄉. Sichuan Provincial Museum.

After Wenwu cankao ziliao 1956.6: 36.


personal immortality, in the hopes that one might live as a material be-ing in an afterlife realm, either on earth or in the heavens. That search is depicted in other examples from Sichuan tombs, for example, where Xiwangmu 西王母 reigns (Figure 6) or men play the liubo 六博 game that “extends time” (Figure 7). Any and all such attempts to secure life beyond the grave represented the same search for personal welfare. They are not specifically Confucian or Daoist, despite what the conventional accounts have told us. As is clear from representative texts of the time, for the Han elites, expressions of filial devotion and experiments with alchemical elixirs went hand-in-hand, with rare exceptions. Only a few highly sophisticated thinkers of the period—Yang Xiong 揚雄 (53 bc–ad 18) and Wang Chong 王充 (ad 27–97) to name two—insisted that man’s days were numbered and one’s good name was the only form of immortality ever attainable (see Nylan 1999: 17–56; 2001a: chapter 1). Most members of the literate elite hoped to preserve the family unit over generations via multiple ways. By definition, the preservation of the well-protected home required the householder to maximize physical light and inner enlightenment,

Figure 7. Rubbing of a liubo game (ceramic brick). Height 39 cm, width 32 cm. Excavated from Pengshan. After Gong Tingwan et al. 1998: fig. 75.

382 michael nylan

and this, too, is a prominent theme in Sichuan tombs. The notion is that the light of personal wisdom, of day, and of yang qi can together mitigate the worst aspects of the long dark nights associated with death itself, with the cutting off of descendants, and with the consequent loss of vivid memories of the deceased. Two Han depictions, one from outside Sichuan and one from inside, make this link between light and protection clear: a house-hold guard bears the sun or moon in his belly region, in the same place where the gods do (Figures 8 and 9). Attaining light and enlightenment, in related ways, virtually guarantees that pleasures will attend both the living and the dead, in this world and in the next.3 People are always debating whether the pleasures depicted in the tomb commemorate specific events in the life of the tomb occupants or whether they are pleasures the dead may anticipate in the afterworld. The debate makes little sense, for the com-mon religion of the time held that a continuum exists between the realms of the living and the dead for all those who are cared for adequately by their descendants (cf. Harper 1995; Poo 1998). At first glance, the idea that family safety provides one of the best bases for secure and serene pleasure-taking may not seem strange at all, though our own culture is more inclined to pit the individual against the family, to assume that the family recedes into the background as the indi-

3 See chapter 2 of Nylan 2001; Nylan 2001c, for theories of pleasure-taking in rela-tion to aesthetic and political theories.

Figure 8. Stone carving of a door guard, tentatively identified in Han Wei and Wang Weilin 2000 as the god Goumang 勾莽. Height 116 cm, width 33.5 cm, thickness 6 cm. First century ad. Shanxi Shenmu Da-baodang. After Han Wei and Wang Weilin 2000: 69, fig. 64; 71, fig. 66.


vidual matures, and to equate pleasure-taking with danger. It is worth reminding ourselves, therefore, how very strong the ties between safety and pleasure were in the early period whose artifacts we examine today. Literary tropes and philosophical essays of the classical period insistently ask precisely the same question (in a manner, I might add, which has numerous parallels to the classical Greek and Roman think-ers): How may thinking humans convert the consuming passions (e.g., those for wine, women, song, the hunt, for palace-building, and for fine rhetoric)—passions that by definition required vast expenditures of the scarce and unreplenishable resources of wealth and bodily energy—into pleasures that will ultimately sustain, rather than corrupt, the main societal units of family and state (Nylan 2001c)? With the depletion or exhaustion of the physical resources so manifestly apparent in death, the problems involved in converting the consuming passions into the sustaining pleasures became all the more pressing, one presumes. Like the mouse nibbling in secret away at the flesh of the gourd symbolic of fertility (Figure 10), ill-considered and excessive activities—no matter

Figure 9. Rubbing of a ceramic brick. Excavated from the suburbs of Chengdu. After Gong Tingwan et al. 1998: fig. 345 (cf. Dayi 1998: fig. 12).

384 michael nylan

Figure 10. Sculpted mouse nibbling away at a gourd. Height, 12 cm, width 40 cm. First or second century ad. Excavated from Leshan Hutouwan 樂山虎頭灣 cliff

tomb. After Gao Wen 2000: 10, fig. 12.

how seemingly inconsequential—could lead to the destruction of the family line supporting the individual. And since Han tombs were by no means “fine and private places,” but instead were one of the regular gathering places for members of the elite, their modes of decoration, by careful design, reminded onlookers of the basic beliefs to which they should trust. Those beliefs were: first, that each ephemeral expenditure of resources should ideally lead to a corresponding increase in social-con-nectedness and conviviality; and second, that pleasure-taking need only be condemned when it provoked social isolation or the commodification of others. We see in representative tomb tiles from Sichuan, then, family and friends convening to enjoy themselves at a feast, with wine, dancing, and music. One of my favorite Han dynasty literary anecdotes tells of a man who had himself depicted in his tomb entertaining four paragons of virtue who lived long before his time (Hou Han shu 64.2124). Such associations linking goodness, moral rectitude, safety, and pleasure are not often made, I think, in today’s political life—though they were a central trope in that of the Han. Once we understand the basic significance of the Han tomb as a public place designed to display the strength and continuity of the fam-ily unit to wider communities, both living and dead, including friends, clients, and allies, then we can begin to ask what is distinctive about Sichuan—what deserves, even demands, a second and a third look. But


Figure 11. Tianmen, line drawing of gilt-bronze coffin plaque from Wushan . Diam-eters between 23 and 27.5 cm. After Robert Bagley 2001: 54, fig. 15a and d (cf. the tree labeled a “heavenly pillar,” i.e., the axis mundi, whose character [桂] may also,

by an alternative reading, mean a cassia or cinnabar pillar).

in order to assess that, we must acknowledge something that is generally true for all early artifacts: that the pictorial images we hope to read are much more complex and rich than we usually care to imagine. Let me note here what I see as an obvious problem in many modern pieces of research: the impulse to identify one symbol as bearing only one meaning and to assume that once we have tied an image to a particular historical event, we have somehow explained its function in the tomb. To the contrary, it is clear that the early inhabitants of the Chengdu 成都 Plain were enormously sophisticated in their tastes, and in their sophistication they demanded emblems that had lodged in them more than one meaning. In fact, the relative power of the figure (xiang 象) in early China was thought to be in direct proportion to the number of meanings that were packed into it. A mere one-to-one correspondence, where X means Y, struck them as a fairly paltry and uninteresting pat-tern— something any country bumpkin could come up with. Signs were made to do double-, triple-, and quadruple-duty, whenever possible. Three examples from the recent exhibition Ancient Sichuan (Bagley 2001) can be used to illustrate this

(1) Que 闕 as the gateway of Heaven (Figure 11). More than a sign of the entrance to Heaven, Tianmen 天門 (that is what the in-scription says on this line drawing based upon a gilt-bronze or-nament for a wooden coffin), the tower-gate has multiple mean-ings: It is an indubitable sign of status and wealth; it represents a protected entrance into a place or a state of mind; it stands in for the axis mundi; it signifies farsightedness and providence,

386 michael nylan

Figure 12. Figure of a winged immortal at an open door (situated at the south side of the coffin). Stone coffin, from the tomb of Wang Hui 王暉 at Lushan 蘆

山. ad 211. After Gao Wen and Gao Chenggang 1996: figs. 3–5.

since one can climb to the top of its watchtower and “see” far into the distance, maybe even the future.

(2) The door ajar (Figure 12) bespeaks the journey through long passageways—this parallel imagery stands for death and for birth, the coming out and the going in of life; even the journey to another state of being.

(3) The image of money itself, which seems utterly pedestrian (and decidedly secular) when the workers on the ceramic base of a money tree from Pengshan 彭山 County (Figure 2) harvest money off the trees and then carry it away on shoulder poles. Still, money represents much more than the mere medium of payment for goods. In many texts from the classical period, it symbolizes communication between different realms since coins circulated throughout and well beyond the Han polity. It rep-resents also the transmutation of objects since value is so easily transferred from objects sold to cash, which can then be used to buy different objects.


In the Sichuan tombs, of course, images do double-duty for a good reason: given the limited space of the tomb (and its even more limited decorative space), as many auspicious signs as possible needed to be packed into the space to facilitate the tomb occupants’ long life over generations to come. Only if the householder has done his best to pro-vide such signs can he be said to have made adequate provision for the continued well-being and pleasure of all those in the home. But partly, I suspect, the early Chinese just liked complexity, juxtaposition, and overlap. They located in them a source of vitality, insisting that the same aspects of life can be good or bad, depending on the situation, and the angle from which they are viewed. Just as aspects of life can be good or bad, depending on their context, no image is inherently good or bad in China. One has to examine the entire repertoire of images in a tomb to intuit its full import, examining how any single image fits into the entire visual program implied by the complete range of grave goods, including the wall paintings and carvings. To illustrate this point, I would have us consider the “demon-quelling postures” (e.g., Figure 13) positioned usually at the tomb entranceway or on nearby que towers. These explicitly sexual postures, said to repel

Figure 13. Panel from a stone que showing a demon-quelling posture (principal tower stone of the Yang 楊 family que at Zitong 梓潼, front face). After Xu Wenbin et al. 1992: 104, pl. 80 (cf. a similar demon-quelling posture on the Yang family que at Mianyang 棉陽, north face, excavated in 1974 at Xinshengxiang 新勝鄉, Zhuwapu 竹瓦鋪, in Pixian 陴縣 County, illustrated in Xu Wenbin et al. 1992: 40, fig. 11, and in Gao Wen et al. 2000: 99, fig. 126). The latter is now in the Sichuan

Provincial Museum.

388 michael nylan

demons (legs asplay, for example), are the same postures that must be avoided at all costs when receiving important guests, or so the Han ritual texts tell us. Casual displays of intimacy were deemed offensive, for it was normally only servants or slaves who witnessed intimate acts without some loss of face to the viewers and the intimate couple. At the same time, it needs to be said, sexual postures do not repel demons because sex was regarded as inherently bad. The power of generation is thought—quite simply—to be so stupendous that it poses a serious threat even to the types of creatures who are normally impervious to threats—creatures like grave-robbers and demons. Having broached the topics of money and sex, I would remind the reader that the production of money-trees and of explicit depictions of sexual activity appear, so far at least, to be unique to the Chengdu Plain area.4 Other motifs that are nearly unique to the Chengdu Plain area prior to the importation of Buddhism into the rest of China in-clude the flower, as in the flower-becrowned pottery figurines (see e.g. Bagley 2001: no. 115) (according to Pollan 2001, the flower is hardly a universal motif in art). But as soon as I have mentioned money, sex, and flower-crowns as three motifs confined to Han Sichuan, I would have us direct our attention to the most curious lacuna from Han Sichuan motifs, when contrasted with tomb art from other regions under Han rule. Most peculiar, I would say, is the near-absence of pictorial reference to famous historical events and exemplary figures. Typically, outside Sichuan, the most famous shrines and tombs in the Han period devote a larger proportion of their engraved and painted areas to the depiction of famous historical events and persons. Presumably, in doing so, the artists in service to their local patrons bolstered the family claims that the tomb occupants had by right now joined the grand sweep of his-tory. It is odd, then, that to date, so far as I know, only three historical scenes can be identified with any certainty in the art of the Chengdu Plain region. The first supposedly portrays an assassination attempt made on the future First Emperor of Qin (Qin Shihuang 秦始皇) by the assassin-retainer Jing Ke 荊軻 (Figure 14); the second, the legendary meeting between the two greatest thinkers of early China, Confucius (Kongzi 孔子) and Laozi 老子5; and the third, the chaste wife of Qiu Hu 秋胡 being accosted by a virtual stranger on the roadway while away from home, mulberry picking (Figure 15) (the stranger, of course, turns

4 A naked woman harrowing the ground under a lianli 連理 tree (symbol of family harmony and continuity) has been found in a post-Han tomb at Gansu Jiuquan 甘肅酒泉, an area long in communication with Sichuan via one of the major silk routes. While possibly metaphorical, this scene may also represent a peasant genre scene. 5 See Gao Wen et al. 2000: 164, fig. 200. It is no coincidence that Sichuan is the place known best for the merging of the teachings associated with Laozi, Zhuangzi 莊子, and Confucius.


demons (legs asplay, for example), are the same postures that must be avoided at all costs when receiving important guests, or so the Han ritual texts tell us. Casual displays of intimacy were deemed offensive, for it was normally only servants or slaves who witnessed intimate acts without some loss of face to the viewers and the intimate couple. At the same time, it needs to be said, sexual postures do not repel demons because sex was regarded as inherently bad. The power of generation is thought—quite simply—to be so stupendous that it poses a serious threat even to the types of creatures who are normally impervious to threats—creatures like grave-robbers and demons. Having broached the topics of money and sex, I would remind the reader that the production of money-trees and of explicit depictions of sexual activity appear, so far at least, to be unique to the Chengdu Plain area.4 Other motifs that are nearly unique to the Chengdu Plain area prior to the importation of Buddhism into the rest of China in-clude the flower, as in the flower-becrowned pottery figurines (see e.g. Bagley 2001: no. 115) (according to Pollan 2001, the flower is hardly a universal motif in art). But as soon as I have mentioned money, sex, and flower-crowns as three motifs confined to Han Sichuan, I would have us direct our attention to the most curious lacuna from Han Sichuan motifs, when contrasted with tomb art from other regions under Han rule. Most peculiar, I would say, is the near-absence of pictorial reference to famous historical events and exemplary figures. Typically, outside Sichuan, the most famous shrines and tombs in the Han period devote a larger proportion of their engraved and painted areas to the depiction of famous historical events and persons. Presumably, in doing so, the artists in service to their local patrons bolstered the family claims that the tomb occupants had by right now joined the grand sweep of his-tory. It is odd, then, that to date, so far as I know, only three historical scenes can be identified with any certainty in the art of the Chengdu Plain region. The first supposedly portrays an assassination attempt made on the future First Emperor of Qin (Qin Shihuang 秦始皇) by the assassin-retainer Jing Ke 荊軻 (Figure 14); the second, the legendary meeting between the two greatest thinkers of early China, Confucius (Kongzi 孔子) and Laozi 老子5; and the third, the chaste wife of Qiu Hu 秋胡 being accosted by a virtual stranger on the roadway while away from home, mulberry picking (Figure 15) (the stranger, of course, turns

out to be her long-lost husband, Qiu Hu, who had failed to recognize the bride he abandoned many years before).6 Given the rather dogged preoccupation with depicting exemplary historical figures and events in the rest of Han realm, what are we to make of the fact that we can identify only three historical scenes in Sichuan art of the period?7 I would suggest that the scenes so far identified as “historical” may not carry their expected cultural meaning in Sichuan.8 Let me take the

Figure 14. Rubbing of a stone relief, said to depict the attempted assassination of the future First Emperor of Qin. Height 276 cm, width 63 cm. Second century ad. Located at the entrance to a cliff tomb at Leshan Mahao. After Gong Tingwan et al.

1998: fig. 242.

Figure 15. Rubbing of a stone carving depicting the attempted assignation scene with Qiu Hu. Height 68 cm, width 205 cm. First or second century ad. Excavated from a cliff tomb at Xinjin 新津. (The original has been broken.) After Gao Wen

et al. 2000: 166 fig. 202.

6 See Gao Wen et al. 2000: 166, fig. 202. 7 There is a possible fourth historical scene, that of Dong Yong 董永 attending his father. See Gao Wen et al. 2000: 56, fig. 68. A fifth scene said to be historical (Gao Wen et al. 2000: fig. 84, explicated p. 27), which supposedly shows Gaozu 高祖 cutting a snake in half, may have a different interpretation. Some attempts to link Sichuan pictures with historical or pseudo-historical events are fairly ludicrous, in fact, as in Gao Wen et al. 2000: 71, fig. 89, where unknown stuff in a bowl is boldly identified as blood used in swearing a blood-oath. On the tendency to over-interpret texts, see Snodgrass 1998. 8 I began thinking about the wide range of uses to which verbal images and allusions were once put when considering the Lunyu 論語 citations found in the Liji 禮記. The Liji sometimes uses the proof text in expected ways, but sometimes it appears to cite a Lunyu passage only because of the coincidence of a single character or phrase, which the Liji then directs to quite another use. I posit here a symmetrical range of uses for visual and verbal images.

390 michael nylan

scene of the assassination attempt of the First Emperor of Qin as an example. Outside Sichuan, for example in Shandong, we see this scene depicted at numerous sites, and there is no doubt outside Sichuan of the scene’s general import, for it is invariably juxtaposed with portray-als of other famous assassination attempts in history. In all likelihood, therefore, in memorial structures outside Sichuan, this scene celebrates the utility to the continued security of the family and state of a fierce ethical commitment to one’s patron, even unto death. By contrast, in the comparable Sichuan depiction, the same figures appear alone, without related associations. In Sichuan, in other words, the failed assassination attempt lacks a visual context, and its possible historical or ethical import is therefore muddied. Evil, in fact, appears to triumph, if we take this depiction at face value. (After all, the assassination attempt made on the evil First Emperor’s life fails abysmally and the noble assassin dies most horribly.) But the stock composition of a man pursued by another dagger-throwing man can be easily adapted for a completely new use. Put at the entrance, where normally we would find frightening images of wild beasts or fierce guards, it serves as an explicit and unambigu-ous warning to grave-robbers. That seems to explain why we see the composition (hitherto misread as a historical scene, the assassination attempt) appearing at the entrance to tombs, as with the cliff tomb at Leshan Mahao 樂山麻浩 . Similar arguments could be constructed about the sexually-charged encounter between the mulberry-picking wife of Qiu Hu and her long-lost husband—especially as that scene does not appear amid related scenes of virtuous women, as it does elsewhere in Chinese literary and visual images, but rather at the beginning of a long procession of unidentified but different men. Perfectly in line with Sichuan’s unusual disinterest in historical events and in exemplary figures, there are three larger themes—as opposed to specific pictorial forms—that may be unique to Sichuan at the early period under review. These three larger themes are: first, a strong awareness of time’s passing (an awareness that supplants interest in any eternal lessons that can be transmitted by historical events and exemplary figures); second, a willingness to insert human figures con-vincingly within a natural landscape; and third, a delight in the comic that in many cases depends on the juxtaposition of the incongruous, especially the sacred and mundane. Let me begin with the Sichuan preoccupation with time’s passing, shown with equal force in the poetic form for which Han Sichuan is justly famous (the long prose-poem, or fu 賦) and in two influential ways of thinking associated with the Chengdu Plain, the so-called “Mystery Learning” of Yan Junping 嚴君平 (fl. ca. 40 bc) and Yang Xiong and that of the religious Daoists from ca. ad 150 on. Their ways of thought


I have analyzed elsewhere (Nylan 2001b). Simply put, the famous poets, thinkers, and religious figures from the area all insisted in their writings that Sichuan was a “hallowed land” under Heaven’s special protec-tion. As evidence of Sichuan’s hallowed status, they remarked upon its breathtaking natural scenery and the incredible fecundity of its land (Sichuan is still one of the “breadbaskets” of China). And yet there was trouble in paradise. Given the relative geographic isolation of the Sichuan Plain—and the linguistic barriers that existed between Sichuan and the court—the poets, thinkers, and religious figures from the Chengdu Plain region felt a strong sense of alienation from both the political center and from mainstream ways. That sense of alienation was compounded in the first and second centuries ad by the powerful trend toward ever-bigger landed estates run by a few magnate families. While they had long viewed themselves as outsiders in the current world of power relations, the local visionaries increas-ingly came to believe that the old opportunities were fast disappearing to men of the middle-socioeconomic standing. As a result, the leading lights of Sichuan displayed over and over again a nostalgic longing to recreate the presumably simpler, agriculturally-based communities they attributed to a distant, but recoverable past. To their minds, an earlier, better world had been supplanted by a world of danger, corruption, and dehumanization. Historians of religion are still debating the precise beliefs of the Han residents of Sichuan in this early period, but this much is clear: shared memories of a sharp break with a better past and the pain and discomfort occasioned by such memories propelled residents of the Chengdu Plain to envision themselves situated differently within time, and because of their different sense of time, they sought new and better modes of reintegration with community and cosmos—as we can see from both their writings and their visual productions. In support of this point, I should like to look carefully now at more sets of images drawn from tomb murals that date either from the very end of Eastern Han or the first years after the downfall of the Han in ad 220. Figure 16, a mural showing the master and mistress of a house seated on a dais, represents one of the very best examples of metropolitan art from the capital of Luoyang 洛陽. Seated on the dais, the tomb occupants, husband and wife, as master and mistress of the house, wear ceremonial dress, as if they were about to receive guests. Neatly flanked by two personal servants each, they epitomize unshake-able wealth and power, in part because the metropolitan painter has formally positioned the figures in flat iconic poses, in part because the tomb throughout depicts so many perquisites of high rank and landed wealth—the banquets, the canopies, the receptions for clients, and, in the entrance-way, horse-drawn carriages in procession—as if by these

392 michael nylan

Figure 16. Detail of the painted mural of the tomb occupants poised to receive guests. Height 146 cm, length 250 cm. Late Eastern Han or Cao-Wei 曹魏 period (early third century ad). Excavated in 1991 from Henan Luoyang Zhucun 朱村.

After Huang Minglan 1966: 190 pl. 2.


luxury objects the viewer could establish the moral worth of the tomb occupants commensurate with their high status. Judging from these and similar examples of tomb art from the capital at Luoyang and from else-where in the Han realm, from the east coast, the most advanced artists outside Sichuan were asked to display their patrons formally as paragons of rectitude and probity, whose motionless attitudes conveyed not only a sense of their unwavering virtue, but also an idea of their eternal place in history. In consequence, these artists emphasized the fine delineation of facial expressions, which they thought served to (a) reaffirm the just character of the current hierarchy; and (b) propagate certain didactic lessons.9 Where there was a perceived need to suggest interaction be-tween figures in the scene, it was shown primarily through the figures’ facial expressions and secondarily by a twist of limbs or postures. It was surely not that artists outside Sichuan were unable to convey movement in space convincingly. They often had the so-called lower orders scurrying from place to place, industriously working, for instance, in kitchens or butchering the animals to be used in sacrifices. Artists outside Sichuan, in other words, definitely knew how to create a sense of three-dimensionality by using devices such as overlapping figures in space, but outside Sichuan, it seems, bustle and change were associated with the bottom of the sociopolitical scale. The ruling elite were evidently content to have themselves portrayed as the latest worthy heirs to a divinely-appointed hierarchy which might experience cyclical change over dynasties, but no substantial changes over time. Long accustomed to the pleasures of this life, the tomb occupants confidently looked to see those exactly replicated in the next. By contrast, when we turn to Sichuan, scenes whose main purpose seems to be to proclaim to the viewer a sense of the subject’s high status are comparatively few. In their place, we find in Sichuan—and only in Sichuan, so far as we now know—numerous scenes of human intimacy (a subject to which I will return). Just as importantly, the Sichuan scenes, rather than picturing eternal and unchanging time, typically seek to give the viewer a sense of momentary time, which necessitates, in turn, a quite different relation to the larger community and cosmos. It seems that, in Shu 蜀, the ideal was to exult in a free and easy integration with one’s soulmate or with nature itself in order to immerse oneself in the pleasurable insights that can come only from a close personal connection with someone or something else. One tomb tile, dating to the second century ad (Figure 17) and immediately recognizable as a

9 Note the importance in Han of physiognomy, as attested in numerous anecdotes. For further information on a recent Mawangdui 馬王堆 manuscript that describes judging horses by their physical characteristics, see Roel Sterckx 2002.

394 michael nylan

Figure 17. Brick with scenes of hunting and harvesting. Height 40 cm, width 48 cm. Second century ad. Excavated in 1975 from tomb 2 at Chengdu, Zengjiabao

曾家包. After Bagley 2001: no. 106.

Sichuan piece because of its style and subject matter, shows a scene of hunting fowl and harvesting rice. In the upper register of the scene, the single bowman on the left strains to shoot geese flying high in the sky; in the lower register, men in groups work with scythes to cut down grasses (probably wet-rice plants). Note that the surprisingly convincing repre-sentation of rhythmic action in three-dimensional space is built by taut diagonals (in this case, in the bowman’s body and his bow) juxtaposed against unmovable objects (in this case, the ground plane and trees). To me, this scene as a whole is inexpressibly beautiful precisely because it captures a single moment in time with such stunning clarity. In another moment, all will have changed forever. The scythe will have cut down the rice stalks; the birds will either be shot or they will have flown free. During the Han, it is first and only in the art of Sichuan that we have human figures placed naturally within a landscape that reminds us of the inevitable processes of change and transformation. Turning to a second famous tomb tile from Sichuan (Figure 6), we find a portrayal of the highest female deity recognized in Sichuan during


Han times: the Queen Mother of the West, Xiwangmu 西王母, whose faithful worship promises to confer immortality. I include this image to counter the possible objection that the most memorable Sichuan scenes depict movement only in relation to local industry or the bucolic. Even when showing one of the very highest deities in the local pantheon, the Chengdu Plain artist resisted the static and the eternal. The Queen Mother may face us, sitting solemnly on her dais, but around her, her attendants, such as the frog and the hare, fairly leap and prance. The bowing subordinate figure in the left hand corner cannot help but look around in a somewhat dazed fashion at the scene, wondering whether the two supplicants who have presented themselves for admission to paradise on the right—they are presumably the tomb occupants—will pass muster during the audience. The last unusual feature of Sichuan art I should like to mention is the comic: Whereas the tomb art of other regions consistently tries to impress you with its grandeur and solemnity, the tomb figurines and tiles from Sichuan as often as not have been artfully crafted to induce a broad smile, if not an outright belly laugh. Hence, we find the pottery drummers from the area (see Bagley 2001: nos. 110–111), whose pants are sometimes in danger of slipping down, and the Sichuan howling dogs. It is worth asking ourselves the larger question: Why would the residents of this particular region have wanted to introduce the comic into the tomb? Our own modern-day proprieties tell us that such scenes are quite ill-suited to decorate the lying-in place of the dead, and the texts inscribed on memorial steles above ground are uniformly lugubri-ous. I cannot come up with a satisfactory answer to this question, but I would urge us to consider the possibility that whenever we run into something that seems so bizarre, we have hit upon a good entry point into the alien mentality of the past. Needless to say, the insertion of the comic element in the tomb bears testimony to a highly sophisticated way of thinking whose underlying logic we can only hope will be further revealed by future archaeological finds. The interpretation of the Han tomb figurines and tiles we know from Sichuan archaeology is not nearly so straightforward and unproblem-atic as it may appear at first glance. I will end with a final speculation based on an observation: He Zhiguo (1994) has shown how very early foreign influence comes to Sichuan, in the form of little Buddha images perched on the money tree in precisely the same position we more often find the Queen Mother of the West. Barbieri-Low (2001) shows Roman putti comporting themselves on Sichuan lacquers made in the late Han or Three Kingdoms period.10 It seems quite plausible, then, that some

10 The objects discussed by Barbieri-Low 2001 come from the third-century ad tomb of Zhu Ran 朱然 at Anhui Ma’anshan 安徽馬鞍山.

396 michael nylan

of the distinctive features of Sichuan art, including those scenes of love and intimacy not usually found elsewhere within China’s geographic borders, were influenced by foreign art, specifically that from the Medi-terranean. Pompeii’s above-ground villas (terminus ante quem ad 79) compare well with the below-ground tombs of Sichuan of a century later in both the range of topics and intensity of expression. Pairing images, where the first two are from Pompeii (Figures 18 and 19) and the second from Sichuan (figs. 5 and 20), is suggestive enough. And while the true import of the Sichuan and the Pompeii scenes continue to elude specialists, it seems inconceivable that such complexly conceived constructions are meant only to provoke voyeurism in the viewer. The man from Pompeii may well represent the god Priapus, but why weigh his enormous penis? Are we meant to think that the ability to spawn progeny is literally worth its weight in gold? The Sichuan couple set off equally strong reactions in all possible viewers—we ourselves, the monkey delightedly swinging overhead, and the startled male passerby. The scene is sometimes misidentified as a man and woman engaged in

Figure 18. Wall painting of a satyr with a maenad. From Pompeii, casa di Cecilio Giocondo (V 1, 26). Napoli. Museo Nazionale (110590).

After Varone 1994: fig. 7.


Figure 19. Wall painting of Priapus phallostates. From Pompeii, casa dei Vettii (VI 15,1). After Varone 1994: fig. 22 (no dimensions given).

398 michael nylan

Figure 20. Brick depicting an erotic scene. Height 29 cm, length 50 cm. Second century ad. Unearthed in 1979 in Xindu 新都 Xinlongxiang 新龍鄉. Xindu

County Bureau of Cultural Relics. After Bagley 2001: no. 107.

a solemn and highly choreographed Daoist ritual called “uniting the qi,” but all we really know is that the woman’s silk skeins have been set aside for the time being. Still, the exuberance of such pieces makes for playful elaborations on the categorical theories of sharing pleasures and reaction-response proposed by the most famous theoreticians of the Han age, including Liu An 劉安 (179–122 bc), Dong Zhongshu 董仲舒 (179–104 bc) and Yang Xiong.11 Viewers cannot but be struck (and most of us are delighted) by the sense of incongruity engendered by another seemingly straightforward Sichuan narrative scene. It is not only, as in Figure 20, that multiple perspectives are employed in such a way as to bring home a spectrum of human emotions. The Sichuan tombs fairly demand of us that we accept into our lives the unknown and the unknowable along with the familiar.

References cited


Barbieri-Low, Anthony (2001). “Roman Themes in a Group of Eastern Han Lacquer Vessels.” Orientations (May, 2001): 52–58.

Chang Renxia 常任俠 (editor) (1988). Zhongguo meishu quanji: huihua bian

11 For ganying 感應 theory, see Lloyd and Sivin 2003.


中國美術全集繪畫編 18: huaxiangshi huaxianggzhuan 畫像石畫像磚 (Anthology of Chinese art: paintings 18: pictorial stones and pictorial tiles). Shanghai: Shanghai Renmin Meishu Chubanshe.

Dayi 1988 = Dayi Xian Wenhuaju 大邑縣文化局 (1998). “Dayi xian Dongchang xiang Sanguo huaxiang zhuan mu 大邑縣董場三國畫像磚墓 (A Three Kingdoms period tomb with pictorial tiles at Dongchang Township, Dayi County).” In Sichuan Sheng Wenwu Kaogu Yanjiusuo 四川省文物考古研究所, Sichuan kaogu baogaoji 四川考古報告集 (Collected reports of excavations in Sichuan), pp. 382–397. Beijing: Wenwu Chubanshe.

de Pee, Christian (forthcoming). “The practice of the text, the practice of the tomb, and the ritual practice of weddings in premodern China.” Journal of Asian Studies.

Erickson, Susan N (1994). “Money trees of the Eastern Han Dynasty.” Bulletin of the Museum of Far Eastern Antiquities 66 (1994): 5–115.

Gao Wen 高文 et al. 2000 = Zhongguo Huaxiang Shi Quanji Wei-yuanhui 中國畫像史全集委員會 (eds.) (2000). Zhongguo huaxiangshi quanji 中國畫像史全集 7, Sichuan Han huaxiangshi 四川漢畫像史 (Comprehensive collection of the history of pictorial stones in China, vol. 7, History of Han pictorial stones in Sichuan). Zhengzhou: Henan Meishu Chubanshe.

Gao Wen and Gao Chenggang 高成剛 (1996). Zhongguo huaxiang shiguan yishu 中國畫像石棺藝術 (The art of carved stone coffins in China). Taiyuan: Shanxi Renmin Chubanshe.

Gong Tingwan 龔廷萬 et al. (eds.) (1998). Ba Shu Han dai huaxiang ji 巴蜀漢代畫像集 (Collected pictorial images of the Han period in Sichuan). Beijing: Wenwu Chubanshe.

Han Wei 韓偉 and Wang Weilin 王煒林 (2000). Shanxi Shenmu Dabaodang Han caihui huaxiang shi 陜西神木大保當彩繪畫像石 (Color-painted pictorial stones of the Han period at Shanxi Shenmu Dabaodang). Chongqing: Chongqing Chubanshe.

Harper, Donald (1995). “Warring States, Ch’in and Han periods: Chinese religion: the state of the field.” Journal of Asian Studies 54.1: 152–160.

He Zhiguo 何志國 (1994). “Yaoqianshu foxiang chuyi 搖錢樹銅佛像芻義 (Preliminary discussions of Buddha images on bronze money trees).” Gugong wenwu yuekan 故宮文物月刊 1994.9: 118–125. Tai-bei.

Hou Han shu 後漢書 (History of Latter Han), by Fan Ye 范曄. Beijing: Zhonghua Shuju edition, 1965.

Huang Minglan 黃明蘭 (editor) (1966). Luoyang Han mu bihua 洛陽漢墓壁畫 (Wall paintings in Han tombs of Luoyang). Beijing: Wenwu Chubanshe.

400 michael nylan

Lloyd, Geoffrey and Sivin, Nathan (2003). The Way and the Word. New Haven: Yale University Press.

Nylan, Michael (1999). “A problematic model: the Han “orthodox synthesis,” then and now.” In Kai-wing Chow, On-cho Ng, and John B. Henderson (eds.), Imagining Boundaries: Changing Confucian Doctrines, Texts, and Hermeneutics, pp. 17–56. Albany: SUNY Press.

Nylan, Michael (2001a). The Five “Confucian” Classics. New Haven: Yale University Press.

Nylan, Michael (2001b). “Legacies of the Chengdu Plain.” In Robert Bagley (editor), Ancient Sichuan: Treasures from a Lost Civilization, pp. 309–325. Seattle: Seattle Art Museum in association with Princeton University.

Nylan, Michael (2001c). “On the politics of pleasure.” Asia Major 14: 73–124.

Pollan, Michael (2001). The Botany of Desire: A Plant’s Eye View of the World. New York: Random House.

Poo, Mu-chou (1998). In Search of Personal Welfare. Albany: SUNY Press.

Sichuan Pengshan 1991 = Nanjing Bowuyuan 南京博物院 (editor) (1991). Sichuan Pengshan Han dai yai mu 四川彭山漢代崖墓. Beijing: Wenwu chubanshe.

Sterckx, Roel (2002). The Animal and the Daemon in Early China. Albany: SUNY Press.

Varone, Antonio (1994). Erotica Pompeiana: Inscrizioni d’amore sui muri di Pompei. Roma: L’Erma di Bretschneider.

Wenwu cankao ziliao 文物參考資料 1956.6, p. 36.Xu Wenbin 徐文彬 et al. (1992). Sichuan Handai shique 四川漢代石闕

(Han Dynasty stone que of Sichuan). Beijing: Wenwu Chubanshe.

Address:

History DepartmentUC-Berkeley3212 Dwinelle HallBerkeley, CA [email protected]


EASTERN HAN DYNASTY CLIFF TOMBS OF SANTAI XIAN, SICHUAN PROVINCE

BY

SUSAN N. ERICKSON(University of Michigan – Dearborn)

AbstractRock-cut tombs were a popular choice for burial during the Eastern Han dynasty in Sichuan province. A number of sites in Santai Xian in the Qijiang area have yielded this type of tomb, featuring multiple chambers excavated deep into the cliffs. Many of these tombs have pseudo-architectural elements and wall decorations, with applied pig-ment still intact in some cases. This article explores the kinds of architectural structures that were copied in these tombs and the meaning of the décor that is carved in relief or painted on interior walls.

Basic characteristics

The cliff tombs of Santai Xian 三台縣 exemplify the preferred type of tomb in Sichuan province during the Eastern Han dynasty (AD 25–220). Rock-cut tombs in other areas of Sichuan, including those at Leshan 樂山 and Pengshan 彭山, are better known since they were published by Western scholars in the early twentieth century (Torrance 1910 and 1930–1931; Ségalen 1915 and 1917; Ségalen et al. 1923; Bishop 1916; Bedford 1937; and Edwards 1954; for a very early account, see Baber 1882). Luo Erhu has grouped the cliff tombs in Sichuan into three geographic regions: the Chengdu 成都 Plain, the lower Min River 岷江 valley where Leshan is located, and the middle Fu River 涪江 valley (Luo 1988b:163). This paper focuses on the rock-cut tombs of northern Santai Xian (Nanmingzhen 南溟 (明) 鎮) in the vicinity of the Fu River, as well as those of the southern part of the county (Anjuzhen 安居鎮) along the Qi River (郪江) and Jin River (錦江) (Figure 1: map). These tombs exhibit characteristics shared by tombs of all areas of Sichuan, but they also have some special features; they have yielded finely pre-served examples of wall paintings and painted relief sculpture, as well as distinctive architectural forms.

402 SUSAN N. ERICKSON

Figure 1a: Map of related sites in China (after Santai 2002, figs. 1 and 2)

Cho

ngqi

ng

Eastern HAN CLIFF TOMBS, SANTAI XIAN, SICHUAN 403

Figure 1b: Detail of Santai Xian (after Santai 2002, figs. 1 and 2)


The tombs of Santai Xian were cut from the sandstone cliffs along the rivers. As at other sites in Sichuan, these tombs are organized as groups laid out in horizontal layers along the upper reaches of the hills. Unfortunately, most of the tombs were robbed of their funerary goods long ago. But on the other hand, the entrances to some of the tombs came to be filled with earth, helping to obscure the location of the cave. Thus many of these tombs have yet to be re-discovered, and the interiors of those recently opened are preserved in fine condition, even though they are devoid of their original contents. The date of the Santai tombs ranges from mid to late Eastern Han, and a few date to the post-Han period. One tomb in northern Santai at Yuanbaoshan 元寶山 is dated by inscription to AD 139 (Jing Zhuyou 1997:69; also see dating analysis by Sun Hua et al. 1998: 373–375). The tombs vary in size. Some are shallow caves, like tomb No. 1 at Jinzhongshan II 金鐘山, which is several meters deep and now is used by a farmer as a storage place for his sweet potato crop. Others are quite vast, like tomb No. 1 at Fentaizui 墳台嘴, which is over 27 meters deep (Figure 2) (Santai 2002:19). This tomb is one of the most complex tombs yet excavated in Santai Xian. It has a long tomb passage, a doorway, an entry chamber, two middle chambers each with side chambers to the right and left, and a back chamber with side chambers to the right and left. The maximum height is 2.55 meters. The structure was carved with a continuous decline from back to front, and three steps are placed between each chamber. This slope is characteristic of the Santai tombs, and its function is to allow water to drain from the interior. This feature may help to explain the fine condition of the graves. Indeed, water within the tombs was a problem, and in some caves, cracks in the walls and ceilings have calcium deposits resulting from water dripping through fissures in the sandstone. Rock-cut tombs did not originate in Sichuan province, but developed in eastern China during the Western Han period (206 BC–AD 8). An early example is found at the tomb site of Liu Sheng 劉貹and his consort at Mancheng 滿城, Hebei province (Figure 3). Liu Sheng, the half-brother of Emperor Wu 武帝, was the king of Zhongshan 中山. His large tomb, comprised of several chambers, was cut on a horizontal axis into the side of a limestone mountain. Although Liu’s tomb resembles a tunnel, it was given the appearance of a house by the creation of rooms within the cave (Wu Hung 1995:130–133). The excavators believe that a wooden structure once stood within the main chamber. The wood has since decayed, but roof tiles and metal fittings remain. The burial chamber in the back was also made more house-like with the addition of stone doors, walls and a pitched roof. The concept of replicating elements of


Figure 2: Line drawing of the plan, elevation and ceiling of tomb one at Fentaizui (after Santai 2002, figs. 7–8)

a home for the living in the resting place of the dead was carried over from tombs like that of Liu Sheng to many areas of China during the Eastern Han period. When the type was transplanted in Sichuan, local preferences created a distinctive style.

Figu

re 3

: Rec

onst

ruct

ion

of th

e ro

ck-c

ut to

mb

of L

iu S

heng

at M

anch

eng,

Heb

ei p

rovi

nce

(a

fter

Zho

nggu

o Sh

ehui

Kex

ueyu

an K

aogu

Yan

jiusu

o, M

anch

eng

Han

mu

(Bei

jing:

Wen

wu

Chu

bans

he, 1

978,

fig.

6)

Tom

b of

Liu

She

ng

1. to

mb

pass

age

fille

d w

ith r

ubbl

e

2. e

ntra

nce

hall

3.

sto

rage

are

a

4. s

tabl

e

5. c

entr

al c

ham

ber

6.

bur

ial c

ham

ber

7.

bat

hroo

m

8. e

ncir

clin

g co

rrid

or

①②

③

④

⑧

⑦

⑤

⑥


Architectural elements

The rock-cut tombs in Sichuan are not just deep caves: as previously discussed, they also feature architectural elements that make them look like houses. While Liu Sheng’s tomb had rooms constructed within the cave, the cliff tombs in Sichuan have architectural members carved into the cave walls and ceilings.1 In addition, some of the entrances to these tombs have relief carvings that simulate the eaves of a roof, complete with rows of rounded tiles and prominent end tiles. At tomb No. 2 at Jinzhongshan I, the end tiles are clearly visible, and the roof is supported by what appear to be engaged columns topped with a kind of capital, called dougong 斗栱 (Figure 4). The dou, or block, and the gong, or bow-shaped bracket, operate according the principle of the cantilever in real architecture. In this case they are simply carved in high relief and provide no support. The walls of the tombs are frequently carved in relief to resemble a timber-framed construction (Figure 5). Upright posts form the outer edge of the frame that creates the wall, and a bressumer or cross rail divides the frame horizontally. Vertical studs divide the horizontal levels into a series of squares or rectangles. If one were to actually construct this kind of wall, the spaces between the wooden framework would be filled with thinner panels of wood or woven bamboo matting that could be covered with plaster. In some tombs, the pseudo-wood frame is enhanced with applied pigments of either red or blue-black. In a few cases, the structure has been suggested with paint only, rather than by relief carving. This style of wood-frame architecture is also represented in contemporary Eastern Han relief carvings from elsewhere in Sichuan. The best example comes from a tomb outside of Santai Xian, near Chengdu at Zengjiabao 曾家包 (Figure 6). A relief carving on the wall of the back chamber of this tomb depicts an estate, and the building to the right has the same features as those simulated in the interior of the tombs of Santai. This technique of building walls with an exposed-wood frame can be seen in traditional architecture still standing in the villages at the foot of the hills where the tombs are located (Figure 7). Inside tomb No. 1 at Fentaizui, additional architectural elements and furnishings have been borrowed from the homes of the living. For instance, the left sideroom of the first middle chamber has fully carved

1 Elements of wooden architecture also can be found in earlier, Western Han dynasty underground wood chamber tombs. Although the basic structure of the tomb is a box, it included small doors and windows that would associate it with a house of the living. Examples can be seen in the early Western Han tomb (no. 168) at Fenghuangshan 風凰山, Jingzhou 荊州, Hubei province; or the late Western Han tomb (no. 101) at Yaozhuang 姚庄 in Hanjiang 邗江, Jiangsu province. See Ji’nan 1975; Yangzhou 1988: figures 2 and 4; and Yangzhou 1980: figure 1.


column shafts with dougong to either side of the central passageway (Fig-ure 8). To the sides of the columns, half-walls are created by the panels, which seem to be constructed of angled slats that look like louvers. This same architectural structure is seen in small models of houses that were found in tombs in Sichuan dating to the Eastern Han period, as in the example from Lushan 蘆山 shown in Figure 9 (also see Zhong Jian 1987). A similar arrangement is also seen in an underground brick tomb dating to the Wang Mang 王莽 interregnum (AD 8–23) in the vicinity of Luoyang in Henan province. This tomb, at Xincun 辛村 in Yanshi Xian 偃師縣, utilizes louvered half-walls to delineate rooms within the tomb just as in the side chamber in Santai (Figure 10). One of the most important aspects of the cliff tombs of Santai in

Figure 4: Entrance of tomb no. 2 at Jin-zhongshan I (Line drawing after Sun Hua, Zheng Dingli, and He Zhiguo 1998:355,

fig. 9; and photo by Yang Cunguan)


comparison to other areas of Sichuan is the variety of dougong that have been discovered. Most of these dougong from the tomb interiors have brackets that parallel the wall line. They are carved in high relief with color sometimes added to the surface to help them stand out from the rest of the wall (Figure 11). In other cases, the columns are nearly in the round, and they have an additional bracket called a huagong 華栱 that is placed at a right angle to the other bracket. One of the most complex examples was found in the back of tomb No. 1 at Hujiawan

Figure 5: Wall with pseudo timber-framed construction, tomb no. 2 at Jinzhongshan I (Line drawing after Sun Hua, Zheng Dingli, and

He Zhiguo 1998:355, fig. 9; and photo by Yang Cunguan)

Figure 6: Rubbing of a relief carving on the wall of the back chamber of tomb no. 1 at Zengjiabao, 296 × 285 cm (after Yu

Weichao 2000a: 40)

Figure 7: Extant wood-frame architecture in a village in Santai Xian (photo by S. Erickson)

Figure 8: Elevation of the left sideroom of the first middle chamber of tomb no. 1 at Fentaizui (after Santai 2002, figs. 7 and 17)

Figure 9: Model of a house from tomb at Lushan, H. 58 cm (after Lim 1987, colorplate 6)

Figure 10: Line drawing of a tomb in Yanshi Xian, Henan province (after Luoyang 1992, figs. 4–6)

Figure 11: Painted dougong, tomb no. 1 at Tiantaishan (photo by S. Erickson)


胡家灣 (Figure 12). The date of this tomb is uncertain, but it is either from the late Eastern Han or from the immediate post-Han period (Yang Cunguan and Zhong Zhi 2000). The brackets project in three directions and are arranged in two tiers. Black pigment has been added to outline the edges of the units comprising the bracket, and patterns of zigzags and crossing diagonal lines adorn the surfaces of the capitals.

Figure 12: Painted dougong, tomb no. 1 at Hujiawan (photos by Yang Cunguan)


The column shaft is octagonal and is placed on a stepped base. In a second example from tomb no. 1 at Bailinpo 柏林坡, the structure of the dougong is less complicated, but the column shaft is cut with sixteen facets. The application of red pigment on the brackets and column shaft creates a bold design (Figure 13). Some of the tombs are furnished. For example, just beyond the en-trance of many Santai tombs, there is often a stove carved from the living rock (Figure 14). Portable ceramic stoves are a popular item in tombs of the Qin and Han periods generally, but a built-in stove is a charac-teristic of the rock-cut tombs of Sichuan. The addition of this necessary household equipment further substantiates the equation of tomb and

Figure 13: Painted column shaft and dougong, tomb no. 1 at Bailinpo (photo by Yang Cunguan)


home.2 Another piece of furniture found in a few of the tombs of Santai Xian is a spectacular couch or bed carved from the living rock (Figure 15). In other cliff tombs, doors are carved in relief on interior walls, suggesting that there are more rooms in the house. On the back wall

2 Wu Hung, following Luo Erhu, suggests that “a kind of ‘stove’ found in many cliff tombs was used, either symbolically or practically, to make elixirs––a suggestion that again connects these tombs with Taoist practices.” See Wu Hung 2000: 84.

Figure 14: Stone stove in tomb no. 1 at Tiantaishan (photo by S. Erickson)

Figure 15: Stone bed in tomb no. 3 at Liujiayan 劉家堰 (after Santai 2002, fig. 31)


of tomb No. 1 at Jinzhongshan II, a figure peers out from behind the doors (Figure 16). The entrance of this tomb also has a much smaller set of doors placed to the right side (figure 17). Here, the peering figure is missing, but the doors are set ajar, suggesting that passage is possible. The motif of the partially open doors is found in many other Eastern Han tombs in Sichuan as part of wall decoration and as an image carved on stone coffins. It is an enduring image in Chinese mortuary archi-tecture, but textual explanation for the motif during the Han dynasty is lacking.3 At Jinzhongshan II, the figure appears to be anticipating the approach of someone, but in several examples from other areas in Sichuan, the figure behind the door is clearly not human. On the end of the sarcophagus from Lushan that is dated to AD 221 by inscrip-tion, the figure has a small wing, a scaled leg and two ears at the top of its head (Figure 18) (for a stone relief with this motif from Rongjing Xian, see Lim 1987: plate 38). Large ears and wings are attributes of immortals.4 Not every one of these figures standing in the doorway has these telltale characteristics, but there seems to be the implication that someone will pass physically or spiritually to another realm. The destination is made somewhat more clear by the incorporation of this motif on the base of a lamp unearthed in a cliff tomb at Xindexiang 新德鄉 in Santai Xian and on the base of a money tree (yaoqianshu 搖錢樹) from nearby Mianyang 綿陽 (Figure 19) (Jing Zhuyou 1993: fig. 13; Mianyang 2000: colorplate 3.4). The upper parts of both bases feature images of the Queen Mother of the West (Xiwangmu 西王母), an im-portant Han dynasty deity associated with immortality and paradise. She is seated on her tiger and dragon throne, and below her throne there is an image of open doors with guards standing to the sides. The doors, set slightly ajar, may be the entrance to a passageway that ascends to the paradise of the Queen Mother of the West. If so, the doorway may be linked to the Gate of Heaven (tianmen 天門) that is labeled on some tomb artifacts from Sichuan, including a stone coffin from Jianyang 簡陽 and several gilded bronze coffin ornaments discovered in Wushan Xian 巫山縣 (Figure 20). The final and perhaps most striking architectural element in the Santai tombs is the ceiling treatment. In general, the ceiling of the main corridor is slightly arched as is the case at other sites in Sichuan. In

3 For a Six Dynasties tomb, see Gansu 1985: plate 59.2; and for a late eleventh cen-tury Song dynasty tomb in Baisha, Yu Xian, Henan, see Su Bai 1957, plates 28–29. For a list of Song tombs that include this motif on the wall or on the sarcophagus in the tomb, see Kuhn 1996: 50, note 30. Kuhn notes that Chinese archaeologists have linked the motif to “the notion of filial piety as practiced by daughters-in-law” in the Song dynasty context. Also see Laing 1978: 13–14 and 17–18, and Rawson 1996b: 35–36, n. 44. For another interpretation, see Goldin 2001. 4 A que pillar in Qu Xian has the same kind of figure sculpted in relief on the surface of the pillar. See Chongqing 1992: 149, figure 232.

Figure 16: Carved relief on the back wall of tomb no. 1 at Jinzhongshan II (photos by S. Erickson and Yang Cunguan)


Figure 17: Carved relief at the entrance of tomb no. 1at Jinzhongshan II (photo by S. Erickson)

Figure 18: Rubbing of a relief on the end of the stone sarcophagus of Wang Hui, Lushan (after Yu Weichao 2000a, p. 72)


Figure 19: Ceramic base for a lamp from a cliff-tomb at Xinde, Santai, and a ceramic money tree base from Mianyang (Photo by S. Erickson and after Mian-yang 2000, colorplate 3.4)


Figure 20: Rubbing of the carving on a stone coffin from Jianyang and line drawing of a gilded bronze coffin ornament (D. 25.4 cm) from Wushan Xian

(after Neijiang 1991, fig. 16; and Chongqing 1998, fig. 4.1)


Santai, however, it is often embellished further with pseudo-architectural elements. In tomb no. 4 at Jinzhongshan II, a central ridge beam con-nected to side rafters has been carved in relief to look like the underside of a pitched roof (Figure 21). In other cases the illusion is even more complex, with shorter rafters abutting a framed panel in the center of the ceiling, and this center panel is sometimes elaborated by the instal-lation of corbelled layers to form a well (Figure 22).5 The uppermost surface of the caisson is flat and is sometimes decorated with reliefs or paintings. Eastern Han texts mention a similar structure, called zaojing 藻井 or “aquatic plant well.” In the Xijing fu 西京賦 (“Western Metropolis Rhapsody”), Zhang Heng 張衡 (AD 78–139) evokes this kind of structure in the Weiyang gong 未央宮 (“Everlasting Hall”) in Chang’an 長安, the capital of the Western Han dynasty (Wen xuan “Xijing fu” 文選西京賦 38 (2.5a); translation: Knechtges 1982:187–188, ll. 100–103):

They ran crosswise long beams of the masculine arc, Tied purlins and rafters to link them together, Rooted inverted lotus stalks on the figured ceiling (zaojing), Which bloomed with red flowers joined one to another.

Figure 21: Pseudo ridge beam and rafters, tomb no. 4 at Jinzhongshan II (photo by Yang Cunguan)

5 A house model (H. 36 cm) from a tomb in Hunan illustrates this type of roof from an exterior view. See Hunan 1981: fig. 17.1.


Although not as spectacular as the palace architecture described in Zhang Heng’s poem, the main corridor and many of the side chambers of these tombs of Santai have ceiling wells. The intriguing image of the inverted lotus in the zaojing in the Han palace will be discussed below, but first, a more common pair of motifs will be considered: images of the sun and moon that are either carved in relief or painted in the center of the well. A beautifully preserved example comes from the first chamber of tomb no. 1 at Yuanbaoshan in Santai Xian (Figure 23) ( Jing

Figure 22: Zaojing in tomb no. 1 at Dongzipai (after Santai 2002, figs. 43 and 39)


Zhuyou 1997).6 Images of the sun and moon are not unusual in Han dynasty tombs throughout Sichuan, but a more typical representation is seen in a pair of pictorial bricks from Peng Xian 彭縣 that were part of the wall decoration of the tomb (Figure 24) (Sichuan 1956). These representations of the sun and moon deities include symbols on the surfaces of the orbs. At the center of the image of the sun, a bird with outstretched wings appears, and at the center of the moon, a toad and a tree are depicted. Despite the fact that representations of the sun and moon are com-mon on wall tiles and also on sarcophagi, the incorporation of these images in the ceilings of tombs in Santai more closely relates them to a number of tombs dated prior to the Eastern Han period from Henan province, primarily in the environs of Luoyang. The ceilings of several underground hollow brick tombs dating to the late Western Han period in this area also feature painted images of the sun and full moon, and they too are identified by their characteristic symbols. In the tomb of Bu

Figure 23: Painted relief images of the sun and moon in the zaojing of the first chamber of tomb no. 1 at Yuanbaoshan (photo by S. Erickson)

6 A similar arrangement of the sun and the moon is utilized in the design on the surface of a stone coffin from Pi Xian (Sichuan 1979). On one end of the coffin, the deities of sun and moon are depicted, and a bird with outstretched wings is between them. The tomb from Yuanbaoshan in Santai has painted images of birds with outstretched wings placed underneath the images of the sun and moon.


Qianqiu 卜千秋 (1st c. BC), the sun is painted red with a black bird in the center, and the moon has an image of a toad and a cassia tree (Fig-ure 25) (Huang Minglan and Guo Yinqiang 1996: 61–76; Sun Zuoyun 1977). They are part of a long composition that includes dragons and immortals amid swirling clouds, perhaps making a journey through the heavens. The journey to a paradise in the heavens was a popular theme in Han dynasty art and literature, and the inclusion of the sun and the moon in the tomb must be related to the hoped-for destination of the soul of the deceased. The ceiling of the tomb of Bu Qianqiu is constructed of a row of tiles, and the images of the sun and the moon are painted on separate tiles creating the look of an individual panel.7

Figure 25: Line drawing of the ceiling of the tomb of Bu Qianqiu, Luoyang, Henan province (after Sun Zuoyun 1979:48, 54–55)

Figure 24: Rubbing of ceramic bricks with images of sun and moon deities from Taipingxiang, Peng Xian, 28.5 × 47 cm each (after Chang Renxia 1988, p. 172)

7 A nearly identical treatment is seen at the Western Han period tomb no. 61 at Shaogou 燒溝 (Huang Minglan and Guo Yinqiang 1996: 87–100). The central chamber of this tomb is divided by a post supporting a lintel and pediment. On the back side of the pediment (that is, facing the back wall), the motif of a pair of slightly parted doors is incorporated just like in many tombs of Sichuan. Also see Henan 1964 and Chaves 1968. For a related Western Han tomb at Qianjingtou 淺井頭, see Huang and Guo 1996: 77–86; Luoyang 1993.


A tomb at Jinguyuan 金谷園 in the environs of Luoyang, dating to the Wang Mang interregnum (AD 8–23), includes paintings on the ceiling of the back chamber that even more closely resemble the Santai zaojing images (Figure 26) (Huang and Guo 1996: 105–120; Luoyang 1985). The paintings are separated into framed panels, and the images of the sun and the moon are at the center of layered well constructions. At Jinguyuan, the utilization of painted architectural members on the walls of the front room and on the high relief posts and beams in the back chamber of the tomb suggest another link to the tombs at Santai discussed above. However, the cliff tombs of Santai also have certain motifs not found in the Luoyang tombs, and these images may be viewed as a regional preference. The center of some panels is decorated with an object referred to as gua 瓜 in the excavation reports. Gua is translated as “melon”, but it can also designate vegetables like squash, gourds, cucumbers, and pumpkins. The motif hangs from the center of the panel much like the lotus described by Zhang Heng in the zaojing of the Everlasting Hall palace in Chang’an quoted above. In tomb No. 22 at Fentaizui, the object located in the recessed ceiling panel is lobed with a circular depression in the bottom, and it still retains some of its original, reddish-orange color that was applied to its surface (Figure 27). Models of houses made to be placed in the tomb sometimes feature this decor too. In an example from Tujing 涂井 in Zhong Xian 忠縣, several gua hang from the bracket of the central pillar of the lower story and also from brackets of the side pillars on the second story of the house (Figure 28).8 This bright red fruit could be just a decorative object, but it most likely meant much more. In the Classic of Mountains and Seas, a similar image is described (Shanhai jing “Xishan jing” 山海經 “西山經”).9 According to the text, a cinnabar tree (dan mu 丹木) on the summit of Mt. Yanzi 崦嵫山 produces a fruit that is said to be like a gua. Its skin is red with black veins, and it was believed to provide protection against jaundice and to ward off fire. In the wood-framed buildings of contemporary Han dynasty settlements that is imitated by the pseudo-architecture of the

8 The house model (#99) is from tomb no. 5 and is one of several examples in this tomb. Another has a gua placed on top of the column under the dougong. The authors of the article date the tomb to the post-Han period, but other sources date it to the late Eastern Han (Yu Weichao 2000b: 70). Another house model (H. 62 cm) with this motif was found in Suining Xian 遂寧縣, and it is dated to the Eastern Han period (Lim 1987: 139, plate 46 and colorplate 7). 9 Shanhai jing jiaozhu, 1.65; see Mathieu (1983, vol. 1: 130): “Trois cent soixante li au sud-ouest il y a le mont Yanzi; à son sommet il y a quantité d’arbres rouge-cinabre, leurs feuilles ressemblent à celles du mûrier de Chine, leurs fruits sont gros comme des courges; ceux-ci ont une queue rouge et des veines noires, qui en mange guérit de la jaunisse; ils peuvent protéger de feu.”


Figure 26: Ceiling of the back chamber of a tomb at Jinguyuan (after Huang Minglan and Guo Yinqiang 1996, pp. 107, 111 and 113; and Luoyang 1985, fig. 6)

Figure 26 (cont.):


Figure 27: Carved gua in ceiling, tomb no. 22 at Fentaizui (photo by Yang Cunguan)

Figure 28: Ceramic model of house, from Tujing, Zhong Xian, H. 45 cm (after Bagley 2001:263, fig. 8)


tomb interior, fire would have been a great concern, and an auspicious gua might have been desirable. Other possible associations relate to the numerous seeds within pumpkins and melons, which could suggest a desire for progeny. My initial investigation of Eastern Han tombs in Sichuan reveals that use of this motif is rather localized. Only rock-cut tombs from the neighboring areas of Mianyang and Zhongjiang 中江 also have this unusual element hanging from the ceiling (He Zhiguo 1988: 224, figs. 7.1–2; Huang Chonghua 1989; and Wang Qipeng and Wang Kongzhi 1989). Outside of Sichuan, a few examples can be cited. A similar gua was found in the center of a zaojing in tomb No. 1 in Donghai Xian 東海縣, Jiangsu province (Figure 29) (Nanjing 1957).10

Figure 29: Line drawing of zaojing with gua and dragon, tomb no. 1 in Donghai Xian, Jiangsu province (after Nanjing 1957:37, 39 and 42)

10 Other tombs that might be related are from Shandong and Jiangsu, but the dates of the tombs are post-Han (Shandong 1975: figure 1; Nanjing 1981b). In the Eastern Jin tomb in the environs of Nanjing, a stone object was found (D. 11.2 cm; H. 7.7). The


In this case it is surrounded by the undulating body of a dragon, sug-gesting a watery environment in the ceiling well. If this fruit was believed to serve the function of warding off fire, it may be related to another element of décor described as having been part of the zaojing of palace buildings—a lotus or a similar flowering water plant. And indeed, in a tomb in the neighboring county of Mian-yang, a gua and a lotus pod both hang from the frame of the ceiling panel (Figure 30) (He Zhiguo 1988). In several rock-cut tombs of Santai, high relief images that have connections to water decorate the beams forming the well in a comparable fashion. At tomb No. 1 at Fentaizui, for instance, a gua and two turtles were carved in relief on the frame surrounding the well (Figure 31). Although an “inverted lotus” is not among the motifs used at the center of zaojing in the Santai tombs, there are many tombs of the late Eastern Han period outside of Sichuan that do include a lotus within the ceiling well. Marylin Rhie notes the use of this image in ceiling wells at sites that range from Leitai 雷台 in Wuwei 武威, Gansu province, in the western part of China, to Dahuting 打虎亭 in Mi Xian 密縣, Henan province, in central China, and to Yi’nan 沂南 in Shandong province in

Figure 30: Line drawing of the plan, elevation, ceiling and details of sculpted lotus and gua on the frame of the ceiling panel, Baishabao 白沙包 tomb no. 2, Mianyang

(after He Zhiguo 1988, figs. 4 and 7.1–2)

authors of the excavation report do not know its function, but its appearance is like the gua found in Sichuan. It has eight lobes and a stem-like knob on the bottom. A lotus design surrounds the “stem”. The top is without decor.


eastern China (Rhie 1999: 64f).11 In Sichuan, however, the lotus motif is more commonly located in other parts of the tomb, and models of houses included in tombs frequently have lotus blossoms decorating the

11 For Leitai in Wuwei, Gansu province, see Gansu 1974: plate 2.3; and for Dahu-ting see Henan 1993: figures 159–160, and plates 6–9. Tomb no. 2 at Dahuting has painted images of lotus flowers in the center of ceiling panels. These panels alternate with painted “illusionistic” zaojing and decorative windows. At Yinan (Zeng Zhaoyu, Jiang Baogeng, and Li Zhongyi 1956: plates 18.1–2), zaojing with orbs alternate with panels that have high relief lotuses. The center of the lotus has a lobed appearance. Also see Nanjing 1981a: figure 5 for a zaojing with carved flower-like design. Marylin Rhie (1999) also discusses the relationship of these well ceilings of the Han dynasty to similarly constructed ceilings in the Buddhist context during the post-Han era. On page 66, she notes that Uehara Kazu (1994) linked the lotus in square ceiling wells with early Han palace architecture. Uehara lists the Han texts where these wells are described as zaojing. In this connection, see also Hayashi Minao’s classic article on the lotus in pre-Buddhist China (Hayashi 1987).

Figure 31: Frame of zaojing with sculpted decor, tomb no. 1 at Fentaizui (line draw-ing after Santai 2002, detail of fig. 7; and photo by Yang Cunguan)


exterior of the building (Figure 28). In a cliff tomb at Pengshan, lotus buds and blooms are included among auspicious images at the entry of the tomb and on the front of a stone coffin inside the cave (Figure 32) (Tang Changshou 1993:97). Indeed, this plant was an important agricultural product then as it is today, and it is often depicted among the plants in agricultural scenes on tomb tiles (Figure 6, see left fore-ground).12 Its association with water is clear, but another gilded bronze coffin ornament unearthed in Wushan Xian suggests a connection to the heavens (Figure 33) (Chongqing 1998). On the face of this disk, a lotus bud (or perhaps a gua) prominently hangs between que 闕 towers and over a figure.13 As noted above, several other objects like this have been unearthed in Wushan Xian, and in these other examples, the place between the que towers is designated by inscription as tianmen or Gate of Heaven (Figure 20). The connection between Heaven and zaojing is made clearer in another rhapsody composed during the Eastern Han period by Wang Yanshou 王延壽 (fl. AD 163). In his “Rhapsody on the Hall of Numi-nous Brilliance in Lu” (Lu Lingguangdian fu 魯靈光殿賦),” Wang calls the ceiling well with layered structures a tian chuang 天窗, or heaven or sky window. Wang wrote (Wenxuan “Lu Lingguangdian fu”; Wenxuan 170 [11.18a]); translated by Knechtges 1987: 271, ll. 102–109):14

12 For a model of a pond from Suining (51 × 31.5 cm), see Rawson 1996a: 214, plate 116; Liu Zhiyuan 1958: plate 4.1. 13 Another gilded bronze ornament (Chongqing 1998: 78; fig. 1.2) has a depiction of a bloom and a hanging bud to the side of one of the towers. A tiger appears off to the side of the pillar on the opposite side of the ornament. 14 A similar ceiling is described in He Yan’s 何晏 “Rhapsody on the Hall of Great Blessings (Jingfudian fu 景福殿賦)” (Wenxuan “Jingfudian fu”; Wenxuan 175 [11.27a]); Knechtges 1987: 289, ll. 150–155.

Figure 32: Relief images of lotuses on the upper edge of stone sarcophagus and on the lintel of the tomb entrance, tomb no. 2, Shuangjiangzhen, Pengshan

(after Yu Weichao 2000a, pp. 12 and 16)


And then, Suspended purlins tied to the sloping roof, Sky windows (tian chuang) with figured filigree: In a round pool (yuan yuan 園淵 ) on the square well (fang jing 方井 ), Invertly planted are lotus (he qu 菏蕖 ) Bursting with florescence, erupting in bloom, Their blossoms spread and open, Their green pods and purple fruits, Bulging and bloated like dangling pearls.

The ceiling well as an architectural structure adds splendor to the Hall of Numinous Brilliance just as it would to a tomb interior. In both the tomb and the palace it can also be viewed as a compluvium allowing light to enter the space and providing a glimpse of heaven beyond the skylight.15

15 Rolf Stein (1990: 147–155) has studied this type of ceiling in the Han and also the post-Han Buddhist period, and he notes that “the symbolic presence of water in the middle of the roof was felt to ward off fires.” He also relates the tianchuang or heavenly window to the household compluvium (zhong liu 中霤) that was a source of light. Stein suggests that the form may be making reference to the five household gods and their place of worship. In the Eastern Han Comprehensive Discussions in the White Tiger Hall (Bai-hutong 白虎通), the five deities (wu si 五祀) of the home are spirits associated with the

Figure 33: Line drawing of a gilded bronze coffin ornament unearthed at Beixincheng in Wushan Xian, D. 27 cm (after

Chongqing 1998, fig. 1.1)


Wall Decoration

In addition to the intricate architectural elements, the cliff tombs of Santai Xian are decorated with low-relief images carved from the flat areas within the pseudo-wooden frame of the wall. The iconographic range is not very extensive, and the approach to composition is consistent by and large. Usually, a single figure or creature is placed within the panel, although on occasion several figures may be found. This kind of composition is quite unlike those seen in many of the tiles that come from the Chengdu area, or in the tombs near Leshan where figures are engaged in an activity, and the figures are often placed in a setting (Figure 34). In some of those cases, the lines of the architecture or the shore of a pond are depicted with diagonals to suggest recession from front to back. This kind of composition is rarely seen in the rock-cut tombs of Santai. The difference in approach may be explained, in part, by the fact that the images from the Chengdu area are sculpted or molded on panels that are executed outside of the tomb and then installed, whereas the images from Santai are carved from the living rock. Another explanation for this preference may be influence from another area of China, espe-

outer door, the inner door, the well, the hearth, and the impluvium (Baihutong “Wusi”; Baihutong shuzheng, 2.77; Tjan Tjoe Som 1952: vol. 2: 376f). Also see Wang Chong’s 王充 Lun heng “Siyi” 祀義; Lunheng jiaoshi 25.1049; trans. Forke 1962: vol. 1: 510; and Bodde 1975: 51.

Figure 34: Rubbing of a ceramic tile with scenes of hunting and harvesting, tomb no. 2 at Zengjiabao, Jinniuxiang, Chengdu, 40 × 48 cm (after Chang

Renxia 1988:158)


cially Henan province, as suggested by Michèle Pirazzoli-t’Serstevens.16 The best comparison is with a tomb dated to the late Western Han or Wang Mang period at Tanghe 唐河 in Nanyang 南陽 city (Figure 35) (Nanyang 1982). The walls of this stone-brick tomb are divided into a series of panels, many of which contain only single figures. Another difference between the Santai reliefs and those of other areas of Sichuan is the amount of detail in the carving. Typically, the Santai examples are rendered very simply as silhouettes, but some have patches of pigments still remaining on the surface of the relief carvings. Most likely, these reliefs relied on color for much of their effect rather than on intricately carved details, such as were utilized in pictorial bricks from other regions. Considering iconography, the Santai tombs feature subjects popular

16 Michèle Pirazzoli-t’Serstevens (2001: 52–53) notes that, in terms of style, the “Sichuan stone carvings have close parallels in the art of Nanyang in southern Henan, and it is very likely that artists or designs from Nanyang played a role in the formative years of the Sichuan workshops.” She also calls attention to the “airy compositions with much open space” seen in the Nanyang carvings.

Figure 35: Line drawing of the elevation of side walls of a stone brick tomb in Tanghe, Henan province (after Nanyang 1982, figs. 5–6.)


Figure 36: Images of flute player (tomb no. 1, Jinzhongshan II); dancer (tomb no. 1, Guloushan); figures playing a game of liubo (tomb no. 1, Jinzhongshan II); and figure grinding substance with mortar and pestle

(tomb no. 1, Zijingwan) (photos by Yang Cunguan)


throughout China, especially those that are associated with ceremonies for the spirit of the deceased, and those that are linked to the goddess associated with immortality, the Queen Mother of the West. A solo flute player from Jinzhongshan II or a dancer from Guloushan 鼓摟山 may stand in for the orches-tras and troupes of dancers who participated in ceremonies for the spirit of the deceased (Figure 36). In another example, a pair of figures playing a game of liubo 六博 from the same tomb at Jinzhongshan II most likely rep-resents immortals who often play in the company of the Queen Mother of the West in more detailed reliefs from other loca-tions in Sichuan. In tomb one at Zijingwan 紫荊灣, a figure with mortar and pestle is probably

grinding materials to concoct the elixir of immortality for Xiwangmu. These images are not part of a grand, overarching composition, but they seem to have been selected from an index of subjects to articulate concerns about death and also about the afterlife that awaited the de-ceased. Only one example of a scene that is based on a historical event can be cited, and it is extremely simplified in comparison to other versions of the theme. It is from tomb no. 1 at Jinzhongshan II, and it depicts the attempted assassination on the future First Emperor of Qin (Figure 37) (Shi ji 史記 “Cike liezhuan” 刺客列傳 86, 2527f). The scene is well-known from the shrines in Shandong and also from representations on tomb walls and sarcophagi from a number of locations in Sichuan including Jiang’an 江安 (Figure 38).17 In the panel from Santai, only the dagger in the column and the figure of the Qin king pointing at

Figure 37: Relief image of the attempted assassination of the future First Emperor of Qin, tomb no. 1 at Jinzhongshan II (photo

by Yang Cunguan)

17 For a discussion of the theme see Nylan 2001: 323f. Also see Gong Tingwan, Gong Yu and Dai Jialing 1998: figures 241–244 showing stone carving from Hechuan (272 × 45 cm); wall carving from Leshan, Mahao (376 × 63); carving on stone sarcophagus no. 1, Jiang’an (221 × 66); and carving on que tower from Qu Xian (123 × 35).


the post indicate that this event is being depicted. Rather than the box containing the map of the Dukang 督亢 region in Yan 燕 or the head of Fan Wuqi 樊於期, which were used as decoys in the plot, the designer of this scene has included a large turtle that defies explanation. My preliminary analysis suggests that a number of the images seen in the tombs of Santai basically function as auspicious or apotropaic motifs. One of the most interesting emblems in this category is related to the popularity of the Queen Mother of the West. Unlike other tombs in Sichuan, or in other areas of China for that matter, representations of the goddess are not found on the cave walls or on sarcophagi. Instead, in the cliff tombs of Santai, and also of Pengshan and Mianyang, a very simple depiction of part of the headdress of the Queen Mother of the West, called a sheng 勝, is placed over a doorway or on the edge of the lid of a sarcophagus (Figure 39).18 It is also part of the design carved on the end of a stone coffin from Lushan discussed earlier, in which an immortal stands in a doorway, and the sheng appears on the architrave (Figure 18). This image can be identified by examining pictures of the Queen Mother of the West wearing her sheng. For instance, a tile from Xindu 新都 and the long side of a stone coffin from Yingjing Xian 縈經縣 illustrate the basic form of the sheng worn by the goddess (Figure 40) (Gong Tingwan, Gong Yu and Dai Jialing 1998: 368–69; and Yu Weichao 2000a: 88, plates 111–114). At the Wu Liang 武梁 shrine in Shandong, Xiwangmu is represented several times, but on one of the ceiling panels, only her sheng is depicted (Figure 41). The ceiling is considered to be a catalogue of good omens. The damaged inscrip-tion beside the image reads: “ . . . Sheng: it appears when a ruler . . .”

Figure 38: Rubbing of a relief image of the attempted assassination of the future First Emperor of Qin; from Jiang’an; 221 × 66 cm (after Gong Tingwan, Gong

Yu and Dai Jialing 1998,fig. 243)

18 Wu Hung (2000: 79) notes the presence of the sheng pattern above the entrances of tombs in Pengshan and Leshan, and he links it to Wudoumi 五斗米 practices. Tang Changshou 1993: 62, fig. 18 illustrates various forms of the sheng pattern.


Figure 39: Sheng above doorway, tomb no. 2, Yuanbaoshan (photo by S. Erickson)

Figure 40: Images of Xiwangmu wearing her sheng. Rubbings of a tile from Xindu (49 × 29 cm) and of the side of a stone coffin from Yingjing Xian (79 × 232 cm) (after Gong Tingwan, Gong Yu and Dai Jialing 1998, figs. 368–69; and Chang

Renxia 1988:84)


Wu Hung (1989:79) notes that a related version of the inscription is recorded in the Song shu 宋書 (History of the Liu-Song dynasty), and it reads: “The gold sheng: it appears when the country is without rebels, and when all barbarian peoples submit and pay visits.”19 Although no contemporary texts address the use of this image in a tomb, it would seem that in the vicinity of Santai, the sheng symbolized Xiwangmu and all the good fortune, including immortality, associated with her. Another distinctive image found in tombs in Santai and the neigh-boring areas is that of a bear, most often squatting on its haunches. In

19 Wu Hung (1989:79) also cites a passage purportedly from Ying Shao’s 應劭 Fengsu tong 風俗通 (Records of customs), which is not in present versions of the text: “[In the seventh day of the seventh month,] every family used to cut colorful silk or gold foil into figures. People put these figures on screens or wore them as headdresses. Nowadays, [instead of making figures,] people often cut [silk or gold foil] into the flower sheng, which resembles the ‘golden sheng’ as shown in the ruitu 瑞圖.” (For bibliographic details, see Wu Hung: 355, n. 12).

Figure 41: Good omen image of Xiwangmu’s sheng, reconstruction, Wu Liang shrine in Shandong (after Feng Yunpeng 1821, vol. 10)


tomb no. 1 at Yuanbaoshan, a squatting bear is represented in a frontal pose and acts as a kind of atlantid supporting horizontal architectural members on walls in the first chamber of the tomb (Figure 42). In cliff tomb no. 3 at Shuangjiangzhen 雙江鎮 at Pengshan, the entrance to the first chamber has a similarly posed bear carved above the lintel (Figure 43).20 It was placed at the center of the doorway, and in this case, the phallus of the bear is prominent and erect. The bear, which squats on its haunches rather than stands on all fours, has a human quality. Images like these frequently are interpreted as fangshi 方士 or fangxiangshi 方相氏, demon expellers who wore bearskins and led sea-sonal exorcisms and also those rites associated with making the tomb free of unfavorable infiuences (Bodde 1975: 75–85). Representations of bears are also found on tomb objects like a lamp stand from Santai and the trunks of money trees (yaoqianshu) found just outside of Sichuan in northern Guizhou and in southern Shaanxi (Figure 44) (Chen Moxi, Mou Yinghang, and Chen Heng’an 1959: 97, figure 20-right; Wang

20 A recently excavated cliff tomb in the environs of Mianyang has a squatting bear with erect phallus placed at the center of the back wall of the tomb. The excavation has yet to be published.

Figure 42: Painted relief image of a bear, tomb no. 1 at Yuanbaoshan (photo by

S. Erickson)

Figure 43: High relief image of a bear above the entrance to the first chamber of tomb no. 3 at Shuangjiangzhen, Pengshan (after Yu Weichao 2000a. p.

17, plate 20)

Figure 44: Ceramic lamp stand from Santai and line drawings of sections of the trunks of money trees from Guizhou and Shaanxi (photo by S. Erickson; line drawings after Chen Moxi, Mou Yinghang and Chen Heng’an 1959, fig. 20-right; and after

Luo Erhu 1998, fig. 5.5)


Shouzhi 1987. fig. 1–2; Luo Erhu 1998: figs. 5.5, 7 and 9).21 In general, the squatting bear with a frontal view of its face (and sometimes with exposed genitalia) can be seen as a powerful apotropaic image, meant to protect the tomb and ward off anything harmful.22 The use of images in doorways is discussed by the Eastern Han writer, Wang Chong 王充, in his Lunheng 論衡 (Lunheng “Ding gui” 訂鬼 [Lunheng jiaoshi 22:939] and “Luanlong” 亂龍 [Lunheng jiaoshi 16:699]; Forke 1962: vol. 1:243 and vol. 2: 352–353; this translation by Bodde 1975: 128–129).23 He states that “prefectural officials of today have peach trees cut down to be made into human forms set up at the sides of doors, and tigers painted to be placed on gates and gate screens.” And he explains that “the carving and painting of such likenesses is done in the hope of repelling baneful influences.”24 Also within the tomb, there are depictions of mythological creatures, including those that symbolize the constellations of heaven: the red bird

21 In an earlier article (Erickson 1994:25), I stated (following the excavation report) that these creatures are monkeys, but a first-hand observation I made in May 2001 reveals that they are bears. Luo Erhu (1998: figures 3–4) also describes them as bears, and in addition, he identifies the central figure at the top of this money tree as an image of the Buddha. It had previously been identified as the image of a woman. 22 This depiction of the bear may help us understand the function of a seated figure, often interpreted as the Buddha, used in many of the same locations. In the neighboring county of Mianyang, a money tree was discovered with small figures placed at intervals on the trunk in the same fashion as the trees with bears (He Zhiguo 1991: figs. 19 and 20, pls. 1.1 and 1.2). Other fragments of yaoqianshu trunks with images of the Buddha have been found in Qingzhen, Guizhou (Chen Moxi, Mou Yinghang and Chen Heng’an 1959: fig. 10.5); in Zhong Xian, Chongqing (Zhao Dianzeng and Yuan Shuguang 1991: figs. 1 and 2); and Yun Xian, Hubei (Hubei 1996: figure 6.3). A money tree base from tomb no. 166 at Pengshan also has a Buddha image on its surface (Nanjing 1991: fig. 44). In addition, at cliff tomb no. 9 at Mahao, Leshan, a seated figure was carved above one of the doorways of the main vestibule of the tomb complex (Nanjing 1993: 159, colorplate 1; H. 37 cm). A second seated image of a Buddha (H. 28 cm) was found in virtually the same location at tomb no. 1 in nearby Shiziwan (Nanjing 1993: colorplate 2). It is not insignificant that these are among the earliest representations of the Buddha in China, but because of their location in the tomb or on the trunk of a money tree, they seem to have little to do with Buddhist worship. They more likely functioned as potent talismans, similar to the squatting bear and the sheng of the Queen Mother of the West. 23 In the medical text Wushi’er bingfan 五十二病方, excavated from tomb no. 3 at Mawangdui (second century BC), peachwood figures are used to cure an affliction in one section. Donald Harper (1998: 302) has translated the passage: “Perform the Pace of Yu thrice. Take a branch from the east side of the tao (peach). In the middle separately fashion figurines [3 missing graphs], and fasten them above the doorway, one on each side.” 24 One other motif that fits in this category is a “human” face seen in frontal view filling the field of a wall panel. Examples are found in tomb no. 1 at Zijingwan and at tomb one at Songlinzui 松林嘴. In a recently excavated tomb in the environs of Mianyang city, a similar face but with extended tongue was found on the back wall of the tomb. Its function of warding away dangerous elements is clear, and it is related to ceramic figures like zhenmuyong 鎮墓俑 tomb guardians (for an additional specimen found near Chengdu, see Bagley 2001: 307, cat. no. 118; H. 89 cm).


of the south (zhuque 朱雀), the black tortoise of the north, the azure dragon of the east, and the white tiger of the west (Sun Hua, Zheng Dingli, and He Zhiguo 1998: 362, figs. 17–19). Another fantastic motif is that of three fish that are joined at the head. It is found on a stone sarcophagus in tomb no. 2 at Jinzhongshan II, as well as in other areas of Sichuan (Figure 45). For instance, at Hejiang 合江, the image occurs on the side of a stone coffin along with other auspicious or fantastic creatures including a toad, a rabbit, a nine-tailed fox, a three-legged bird, and another bird with outstretched wings (Figure 46) (Wang Tingfu and Li Yihong 1995: fig. 4; on p. 66, the authors also list coffins in Luzhou M4; M9 and M11). These creatures are often seen in conjunction with representations of the Queen Mother of the West, and so like them, the fish with joined heads may also have this specific association. But more generally speaking, it is a motif of good omen. At the Wu Liang shrine in Shandong, a related image is depicted among a group of omens (Figure 47). It appears on the ceiling, and it is labeled: “The fish joined at the eye” (bimu yu 比目魚) (Wu Hung 1989: 75 and 242). The rest of the inscription reads: “They appear when the virtue of a ruler is widespread and his brightness extends to everywhere” (trans. by Wu Hung 1989: 243). The rendering of the fish with their bodies parallel to one another at the Wu Liang shrine varies from the Sichuan images. Nevertheless, an undated stone recently unearthed in Weishan Xian 微山縣 in southern Shandong has the same version as the sites in Sichuan, and perhaps more importantly, this representation is also found on the face of a que tower at Dengfeng 登封, Henan (Figure 48) (Weishan 2000: fig. 4; for the Henan que see Gu Sen 1997: 776 and Xin Lixiang 2000: 317–321, fig. 163). This tower, known as the Taishi que 太室闕, is dated AD 118, and “the fish joined at the eye” is one of many motifs segregated in a framework of panels comparable to the compositions on walls of Santai tombs. Traveling artisans with motif indexes in hand may help to explain the similarities in these images found in Henan and in Sichuan. Que towers like the Taishi que were used to signal the entrance to a cemetery and to protect the gravesite. Stone que were also built in Sich-uan, and some are still extant (Chongqing 1992). Like the interiors of the Santai tombs, these stone towers imitate wood constructions (Figure 49). Relief images of small que towers are frequently found within Sichuan tombs. In the cliff tombs of Santai Xian, a single tower or a pair are at the entrances to the tombs or the rooms within a tomb. At tomb no. 1 at Yuanbaoshan, color still remains on the surface of a relief carving of a tower that stands alone on the face of a juncture wall between two chambers of the tomb (Figure 50) (Jing Zhuyou 1997). This motif is also found on the surface of sarcophagi. A pair of que towers flanked by


Figure 45: Relief image of three fish on a sarcophagus, tomb no. 2 at Jinzhongshan II (photo by Yang Cunguan)


Figure 46: Rubbing of the side of a stone coffin from Hejiang Xian (after Wang Tingfu and Li Yihong 1995, fig. 4)

Figure 47: Good omen image of “fish joined at the eye” (bimu yu), reconstruction, Wu Liang shrine, Shandong

(after Feng Yunpeng 1821, vol. 10)

Figure 48: Rubbing of bimu yu motif and line drawing of the Taishi que tower at Dengfeng, Henan province (after Gu Sen 1997:776; and

Xin Lixiang 2000, fig. 163)

Figure 49: Que in Mianyang Xian (after Ségalen, Gilbert de Voisins, and Lartigue 1923, atlas, vol. 1, pl.37)

Figure 50: Painted relief image of a que tower on the side wall of tomb no. 1 at Yuanbaoshan (photo by S. Erickson)


plumed birds is depicted on the main face of a stone coffin from tomb no. 1, at Dongzipai 洞子排 (Figure 51) (Yang Cunguan and Zhong Zhi 2000:1). This combination of images is very common in Han funerary art.25 Other images connoting protection include doorguards, weapons hanging on racks, and men loading crossbows. In tomb no. 1 at Fen-taizui, a wide range of figures and objects is carved on the walls of the entry (Figure 52: line drawing of right and left sides). Standing figures are positioned to either side of the door. One is depicted in a frontal pose, but unfortunately, it has suffered extensive damage. The body of

25 For a Western Han example from tomb no. 1 at Huchang 胡場, Yangzhou, see Powers 1983: figures 8–9. One of the wood panels in the box-like tomb chamber features this design.

Figure 51: Stone sarcophagus with que and birds, tomb no. 1 at Dongzipai (after Santai 2002, figs. 35 and 44)


the other figure is seen in profile view, but his head is turned to stare into the central corridor (Figure 53). This figure leans forward dramatically and has a staff across his shoulder. Further ahead on the left side, pseudo-columns with dougong support a roof. The columns also provide surfaces for mount-ing several weapons. Between the columns, a man with a crossbow is depicted, and to the side, three fish are carved (Fig-ure 54). They take a Y-shaped formation, and their heads touch a central circular boss, thus almost becoming joined at the heads as in the image of good omen mentioned above. On the opposite side, the wall is divided into panels because

of the characteristic wooden frame (Figure 55). In one panel, a man on a rock or stool is sitting with his legs boldly splayed outward, and in another, a standing figure is represented. The second image is difficult to see, but perhaps it is a sleeve dancer. The figure with a crossbow on the opposite wall is frequently found in tombs in Sichuan as well as most other parts of China. The display of weapons is also common, and it is easy to understand its function at the entrance of the tomb. On the other hand, the postures of two other figures are quite strange.

Figure 52: Line drawing of the right and left walls of the entry to tomb no. 1 at Fentaizui (after Santai 2002, fig. 7)

Figure 53: Figure in the leaning pose, tomb no. 1 at Fentaizui (photo by S. Erickson)


Figure 54: Three fish and figure with crossbow, tomb no. 1 at Fentaizui (photo by Yang Cunguan)


Figure 55: Seated figure and dancing figure, tomb no. 1 at Fentaizui (photos by Yang Cunguan)

A text excavated from a third century BC tomb at Shuihudi 睡虎地 in Yunmeng 雲夢, Hubei may help to explain the unusual figure leaning forward and the other sitting with legs spread open. The text, which has been translated by Donald Harper, involves formulas for spellbinding demons. The prologue reads (Harper 1985: 480f):

Spellbinding to inflict odium on demons.

The Wanghang 罔行 who injure people treat the people unpropi-tiously.

Let the way for how to spellbind them be declared, to enable the people to not encounter the baleful and calamitous.

What demons detest are namely reclining in a crouch (qu wo 屈臥), sitting like a winnowing basket (ji zuo 箕坐 ), interlinked motion (lian xing 連行 ), and the leaning stand (qi li 奇立 ).

The last pose, “the leaning stand,” seems to have been taken by the figure in the entry with a staff over his back.26 It is possible that this

26 In tomb no. 1 at Jinzhongshan, another figure in the pose of the “leaning stand” can be cited (Sun Hua, Zheng Dingli, and He Zhiguo 1998: figure 6).


leaning figure is simply showing deference to someone of high rank passing before him, but in no other example of bowing guards is the body thrown into such a dramatic angle. Given its location in the tomb, this prophylactic stance seems especially appropriate.27 Another pose mentioned in the Shuihudi text, “to sit like a winnowing basket,” has been explained by Harper (1985:484) as “imitat[ing] the shape of the Chinese winnowing basket by sitting with the legs spread apart and fully extended.” The man with splayed legs most likely is seated like a win-nowing basket even though his legs are not completely extended. The pose termed lian xing or “linked motion,” could possibly be suggested by the form of the three fish joined at the head. Although the last pose, qu wo or “reclining in a crouch,” is not represented here, a figure on the Mianyang que tower is depicted in this protective pose (Figure 56). In the second to the top register of the que tower, a figure placed be-tween two gong brackets reclines on his side but curls his knees forward, or crouches, to help support himself. He is not sleeping as one might expect, but rather he appears very alert. In tomb no. 1 at Fengtaizui, the ensemble of figures in demon-quelling poses plus the man with a crossbow and other weapons at his disposal presents a formidable force to protect the deceased.

27 Harper (1985: 488) also states: “The prophylactic nature of the leaning stand in the prologue may be related to the limping and hopping movements used in ancient shamanic dance steps. The Pace of Yu was the classic shaman’s limp: it was performed by having one foot trail behind while the other foot stepped forward. This lurching shuffle was a widely used magical device for coercing the spirits and overcoming spectral perils.”

Figure 56: Figure in the “reclining in a crouch” pose, que tower in Mianyang (after Chongqing 1992:74, fig. 11)


Tomb no. 1 at Bailinpo, Qijiang area

The last site to be considered is tomb no. 1 at Bailinpo, a tomb that has yet to be fully excavated (Zhong Zhi 2000: 4). Officials of the Santai Culture Bureau are waiting for special funding to preserve the paint-ings on the walls in the course of full-scale excavation. This tomb has architectural elements that are typical of the Santai area. For example, the column and bracket painted red in the back chamber of this tomb were noted earlier in this paper (Figure 13). Although the ceiling in the central corridor of the middle room has collapsed partially, other areas of the tomb, like the back room and the side chamber of the middle room, have zaojing with painted images of the sun and moon still intact. As for the wall decoration, an overarching interpretation of the iconography cannot be offered at this point since the lower parts of some walls are still concealed by deposits of earth. Nevertheless, this tomb promises to change our view of tomb decoration in Santai. Two areas of the tomb are especially interesting in terms of wall decoration. One of them is in the rear chamber. The right side wall

and the surface of the wall that separates this chamber from the middle chamber have painted relief carvings. The walls are divided into a grid pattern because of the presence of the pseudo-wood frame that is also painted. On the dividing wall, a unique depiction of a rampant feline or canine fills each of two panels (Figure 57). The body of the animal is painted in a red that is still vibrant. On the side wall of the back chamber, a column with dougong divides the walls vertically, and a crossrail divides the wall horizontally. To the right of the column there are three panels, and to the left there are two panels arranged in two registers. The lower panels are filled with a variety of motifs, but the best preserved example is a depiction of a dragon with details picked out in black paint

Figure 57: Painted relief of an image of a rampant animal, dividing wall of back chamber, tomb one at Bailinpo (photo by

Yang Cunguan)


(Figure 58).28 In the upper register of the side wall, human figures are represented. The most interesting are three panels that have pairs of figures in each. The excavators believe that these three panels should be considered as a sequence (Figure 59). In the first one, the figures seem to be conversing, and in the second panel, one figure reaches out for the other, who is rising. And finally, in the third panel, the couple is copulating. Although the pair is not depicted as graphically as the figures in ceramic tiles found in a brick tomb at Xindu (Figure 60), the theme is basically the same (Bagley 2001: 293).29 Representations of amorous couples are much more common in Sichuan than in other parts of China, but their precise meaning has remained elusive. A possible interpretation for the stages of “amour” depicted in these three panels comes from texts devoted to the “sexual arts” excavated from tomb no. 3 at Mawangdui 馬王堆 in Hunan province that was sealed in 168 BC. Donald Harper has translated and

28 Other motifs include a tiger, a rabbit, and a strange scene of a figure (or perhaps a monkey) hanging at the top of the panel and an animal standing at the bottom of the panel. 29 A second brick from the same tomb at Xindu with a similar scene is reproduced in Bagley 2001:293. A brick with a couple in the same posture but underneath a canopy was unearthed in 1955 in Deyang 德陽 (Fan Xiaoping 1998: figure 1).

Figure 58: Painted relief image of a dragon, lower register of panels on the side wall of back chamber, tomb no. 1 at Bailinpo (photo by Yang Cunguan)


Figure 59: Painted relief of images of human figures, upper register of panels on the side wall of the back chamber, tomb no. 1 at Bailinpo (photos by Yang Cunguan)


annotated the two texts assigned the titles, “Conjoining yin and yang (He yinyang 合陰陽)” and “Discourse on the culminant way in Under Heaven (Tian xia zhi daotan 天下至道談)” by the editors of the materials (Harper 1987). Harper (1987:539) states that “along with breath cultivation, callisthenic exercises and dietetics, sexual intercourse was one of the methods of ‘nurturing life (yang sheng 養生)’.” He then explains (1987:581) the need for such texts: “Intercourse––intercourse, that is, for the sake of physical cultivation rather than as a consequence of passion––had to be carefully controlled so that the couple at no point committed a tactical error that would lead to failure. The primary function of the sex manu-als was to transform the erotic subtleties and complicated permutations of intercourse into a technique that, like other cultivation techniques, guided the practitioner toward success.” In these texts, the first section involves foreplay—stroking various parts of the woman’s body in a specified order. It is only about midway into the text that penetration takes place. A similar “sequence of procedures” may be suggested in visual form in these scenes painted in the rear chamber. Other images of amorous but not yet copulating couples are found on the stone coffin from Yingjing and also above the doorway in a cliff tomb at Mahao 麻浩, Leshan (Figures 40 and 61) (Yu Weichao 2000a: 7, plate 9).30 It is likely that all of these images are auspicious, and they represent one of

30 Also see an amorous couple from cliff tomb no. 550 at Pengshan (Yu Weichao 2000a: 18, plate 21 [H. 49]).

Figure 60: Ceramic tile with erotic scene from Xinlongxiang in Xindu Xian, 29 × 50 cm (after Bagley 2001:293)


several ways of “nurturing life (yang sheng)” that was associated with the hope of extending life or avoiding death altogether. The middle chamber of the tomb also has painted reliefs on the right side wall and on the surfaces of the two walls that separate this chamber from both the rear and front chambers. Just as in the rear chamber, the wall has panels created by the wood frame element. The side wall has elegant images of dancing cranes and also of an acrobat or dancer carved in relief. The upper tier of this side wall has images of standing figures, but in this case they are painted directly on the wall. These figures are painted with great attention to detail, and they have Chinese characters written beside them. The excavators have interpreted the characters as names of individuals.31 Both figures stand on the backs of animals, and one holds a staff (Figure 62). Zhong Zhi (private communication) believes that these are beings who have become immortals, who now ride their animal vehicles. At the bottom of the narrow wall separating the rear and middle chambers, a tiger with open mouth and glaring eyes is rendered in relief with color added. The tiger symbolizes the west, but in this case, it seems to be primarily warding away the inauspicious. Its tail arches upward to touch the lower edge of a much more complicated scene located in the upper section of the wall. It is the scene of a banquet with people

31 Other wall paintings in tombs of the late Eastern Han dynasty have inscriptions beside painted motifs. In tomb no. 1 at Wangdu, Hebei (Hebei 2000: 29–56), titles are written beside figures. For a late Eastern Han tomb in Henan see Zhengzhou 1996.

Figure 61: Relief image of an amorous couple, cliff-tomb at Mahao, Leshan, H. 29 cm (after Yu Weichao 2000a:7, plate 9)


seated on a white mat and with red lacquered trays placed in front of them (Figure 63). Above them, a red canopy protects them and helps to define the space they occupy. Space is also created by the placement of the figures, which suggests that the center person is in back of the other two people. This kind of scene is popular in other areas of Sich-uan. For instance, a sculpted tile from Peng Xian has a depiction of a similar group of figures (Figure 64). The edge of the mat under them and part of the canopy over them define the space the figures occupy. Banqueting scenes can be related to ceremonies to call back the soul of the deceased or they can represent one of the pleasurable activities in life that hopefully will be perpetuated in the afterlife. The painted reliefs and wall paintings in tomb no. 1 at Bailinpo are important for several reasons. The reliefs in this tomb, as well as in others in the county, illustrate the fact that the original appearance of the im-age is altered significantly when pigments are still preserved. With the addition of color and black lines, the reliefs, and also the architectural members for that matter, look completely different from most of those at other locations where only the stone surface is intact. The paintings in this tomb reveal that, in some cases, the panels within the pseudo-

Figure 62: Painting of a figure standing on the back of an animal, upper tier of the side wall, middle chamber, tomb no. 1, Bailinpo (photo by Yang Cunguan)


Figure 63: Painting of a tiger and a banquet, short wall separating the rear and middle chambers, middle chamber, tomb no. 1, Bailinpo (photo by

Yang Cunguan)

Figure 64: Rubbing of a ceramic brick with banquet scene, Peng Xian, 48 × 28 cm (after Gong Tingwan, Gong Yu and

Dai Jialing 1998, fig. 69)


wood frame of the wall were filled with paintings that only are preserved under the best conditions. It also is clear that several styles of creating an image are in operation in the tomb. Fairly simple single images are represented within some panels, but in the same tomb, much more complicated groups of figures set into a defined space are also depicted. Final analysis of the iconography of the relief sculptures and the wall paintings must await the complete excavation of all areas of the tomb and the publication of detailed photographs. Nevertheless, a tantalizing view of the original appearance of a tomb interior created during the Eastern Han period in Santai is possible through the fortuitous pres-ervation of this tomb, complementing the evidence from a number of recently documented tombs in the county mentioned above.

Acknowledgements:

I am indebted to scholars from Sichuan who facilitated my study of the cliff tombs of Santai during the spring of 2001, especially to Zhong Zhi 鍾治 and Yang Cunguan 楊存貫 of the Santai Xian Culture Bureau. Zhong Zhi accompanied me to the sites and shared with me his article (Santai 2002). He also gave me a copy of a manuscript in progress, “Si-chuan Santai Yuanbaoshan yamu qingli jianhua 四川三台元寶山崖墓清理簡報.” Yang Cunguan generously allowed me to produce reprints of his photographs of the tombs for this article. I am grateful to Chen De’an 陳德安 (Sichuan Archaeological Institute), and Sun Hua 孫華 (Peking University) for helping me make contacts in Santai. Wang Yi 王毅 (Chengdu Archaeological Institute) made arrangements for me to see the cliff tombs of Pengshan that provided a comparison to those of Santai. Su Rongyu 蘇榮譽 (Institute for the History of the Natural Sciences, Chinese Academy of Sciences) accompanied me to all the sites and helped document the tombs. I would like to thank Ellen Johnston Laing and Michèle Pirazzoli-t’Serstevens for commenting on this essay when it was in preparation. I also would like to thank Chad Weinberg, GIS specialist in the University of Michigan Map Library, for designing the maps for this paper.

References cited:

Baber, E. Colborne (1882). “Travels and Researches in the Interior of China.” Royal Geographical Society Supplementary Papers (London: John Murray) 1, part 1, 129–139.

Bagley, Robert (ed.) (2001). Ancient Sichuan: Treasures from a Lost Civili-


zation. Seattle and Princeton: Seattle Art Museum and Princeton University Press.

Bedford, O. H. (1937). “Han Dynasty Cave Tombs in West China.” The China Journal 26, no. 4 (April 1937): 175–176.

Bishop, Carl W. (1916). “The Expedition to the Far East.” The Museum Journal (University of Pennsylvania) 7, no. 2 (June 1916): 97–124.

Bodde, Derk (1975). Festivals in Classical China: New Year and Other Annual Observances during the Han Dynasty 206 BC –AD 220. Princeton: Princ-eton University Press.

Chang Renxia 常任俠 (editor) (1988). Zhongguo meishu quanji: Huihua bian, 18: Huaxiangshi huaxiangzhuan 中國美術全集: 繪畫編, 18: 畫像石畫像磚 (Corpus of Chinese Fine Arts: Paintings, Vol. 18: Picto-rial stone reliefs and pictorial bricks). Shanghai: Shanghai Renmin Meishu Chubanshe.

Chaves, Jonathan (1968). “A Painted Han Tomb at Loyang.” Artibus Asiae 30, no. 1: 5–27.

Chen Moxi 陳默溪, Mou Yinghang 牟應杭, and Chen Heng’an 陳恆安 (1959). “Guizhou Qingzhen Pingba Han mu fajue baogao 貴州清鎮平壩漢墓發掘報告 (Report on the excavation of a Han tomb at Pingba, Qingzhen, Guizhou [province]).” Kaogu xuebao 1959.1: 85–103.

Chengdu 1981: Chengdu Shi Wenwu Guanlichu 成都市文物管理處 (1981). “Sichuan Chengdu Zengjiabao Dong Han huaxiang zhuan shi mu 四川成都曾家包東漢畫像磚石墓 (An Eastern Han stone tomb with pictorial bricks at Zengjiabao, Chengdu, Sichuan [prov-ince]).” Wenwu 1981.10: 25–32.

Chongqing 1992: Chongqing Shi Wenhuaju 重慶市文化局and Chongqing Shi Bowuguan 重慶市博物館 (1992). Sichuan Handai shi que 四川漢代石闕 (Han dynasty stone towers in Sichuan). Beijing: Wenwu Chubanshe.

Chongqing 1998: Chongqing Wushan Xian Wenwu Guanlisuo 重慶巫山縣東漢管理所 (1998). “Chongqing Wushan Xian Dong Han liujin tongpai shi de faxian yu yanjiu 重慶巫山縣東漢鎏金銅牌飾的發現與研究 (The discovery of and research into gilded bronze ornaments of the Eastern Han period at Wushan Xian, Chongqing).” Kaogu 1998.12: 77–86.

Edwards, Richard (1954). “The Cave Reliefs at Ma Hao,” parts one and two. Artibus Asiae 17, no. 1: 5–28; and no. 2: 103–29.

Erickson, Susan N. (1994). “Money Trees of the Eastern Han Dynasty.” Bulletin of the Museum of Far Eastern Antiquities 66: 7–115.

Fan Xiaoping 范小平 (1998). “Sichuan Han dai xing ticai huaxiang yanjiu 四川漢代性題材畫像研究 (A study of Han dynasty reliefs with erotic subjects found in Sichuan).” Dongnan wenhua 1998.4: 78–85.


Feng Yunpeng 馮雲鵬(1821). Jin shi suo 金石索. n.p.: Shuang Tong Shu Wu.

Forke, Alfred (trans.) (1962). Lun-heng, vols. 1–2. New York: Paragon Book Gallery.

Gansu 1974: Gansu Sheng Bowuguan 甘肅省博物館 (1974). “Wuwei Leitai Han mu 武威雷台漢墓 (A Han tomb at Leitai, Wuwei [Gansu province]).” Kaogu xuebao 1974.2: 87–109.

Gansu 1985: Gansu Sheng Wenwudui 甘肅省文物隊, Gansu Sheng Bowuguan 甘肅省博物館 and Jiayuguan Shi Wenwu Guanlisuo 嘉峪關市文物管理所 (1985). Jiayuguan bihua mu fajue baogao 嘉峪關壁畫墓發掘報告 (Report on the excavation of painted tombs at Jiayuguan). Beijing: Wenwu Chubanshe.

Goldin, Paul R. (2001). “The motif of the woman in the doorway and related imagery in traditional Chinese funerary art.” Journal of the American Oriental Society 121 (4): 539–548.

Gong Tingwan 龔廷万, Gong Yu 龔玉, and Dai Jialing 戴嘉陵 (edi-tors) (1998). Ba Shu Handai huaxiang ji 巴蜀漢代畫像集 (A selection of pictorial tiles of the Han dynasty from Sichuan). Beijing: Wenwu Chubanshe.

Gu Sen 顧森 (editor) (1997). Zhongguo Han huatu dian 中國漢畫圖典 (Illustrated index of Han dynasty pictures). Hangzhou: Zhejiang Dianying Chubanshe.

Harper, Donald (1985). “A Chinese Demonography of the Third Cen-tury B.C.” Harvard Journal of Asiatic Studies 45: 459–498.

Harper, Donald (1987). “The Sexual Arts of Ancient China as Described in a Manuscript of the Second Century B.C.” Harvard Journal of Asiatic Studies 47: 539–593.

Harper, Donald (1998). Early Chinese Medical Literature: The Mawangdui Medical Manuscripts. London: Kegan Paul International.

Hayashi, Minao 林已奈夫 (1987). “XX (The symbolism of the lotus in ancient China).” Tōhō gakuhō 東方學報 59: 1–62.

He Zhiguo 何志國 (1988). “Sichuan Mianyang Hebian Dong Han yamu 四川綿陽河邊東漢崖墓 (The Eastern Han cliff tombs at Hebian, Mianyang, Sichuan [province]).” Kaogu 1988.3: 219–226.

He Zhiguo (1991). “Sichuan Mianyang Hejiashan 1 hao Dong Han yamu qingli jianbao 四川綿陽何家山 1 號東漢崖墓清理簡報 (Brief report on the clearing of Eastern Han cliff tomb no. 1 at Hejiashan, Mianyang, Sichuan [province]).” Wenwu 1991.3: 1–8.

Hebei 2000: Hebei Sheng Wenwu Yanjiusuo 河北省文物研究所 (2000). Hebei gudai muzang bihua 河北古代墓葬壁畫 (Wall paintings from ancient tombs in Hebei [province]). Beijing: Wenwu Chubanshe.

Henan 1964: Henan Sheng Wenhuaju Wenwu Gongzuodui 河南省文化局文物工作隊 (1964). “Luoyang Xi Han bihua mu fajue baogao


洛陽西漢壁畫墓發掘報告 (Report on the excavation of Western Han tombs with painted murals in Luoyang [Henan province]).” Kaogu xuebao 1964.2: 107–125.

Henan 1993: Henan Sheng Wenwu Yanjiusuo 河南省文物研究所 (1993). Mi Xian Dahuting Han mu 密縣打虎亭漢墓 (The Han tomb at Dahu-ting, Mi Xian [Henan province]). Beijing: Wenwu Chubanshe.

Huang Chonghua 黃崇華 (1989). “Zhongjiang Han huaxiangshi yi cun chukao 中江漢畫像石遺存初考 (A preliminary study of the remains of pictorial stone reliefs in the Han tombs at Zhongjiang [Sichuan province]).” Sichuan wenwu 1989.4: 18–19.

Huang Minglan 黃明蘭 and Guo Yinqiang 郭引強 (editors) (1996). Luoyang Han mu bihua 洛陽漢墓壁畫 (Painted murals from Han tombs at Luoyang). Beijing: Wenwu Chubanshe.

Hubei 1996: Hubei Sheng Wenwu Kaogu Yanjiusuo 湖北省文物考古研究所 (1996). “Nanshui beidiao gongcheng Danjiangkou shuiku Yun Xian Yanmoqu kaogu diaocha 南水北調工程丹江口水庫鄖縣淹沒區考古調查 (Archaeological investigation of the sinking zone at Yun Xian in Danjiangkou Reservoir conducted during the engineering project to move water from south to north).” Jianghan kaogu 1996.2: 20–33.

Hunan 1981: Hunan Sheng Bowuguan 湖南省博物館 (1981). “Hunan Changde Dong Han mu 湖南常德東漢墓 (An Eastern Han tomb at Changde, Hunan [province]).” Kaoguxue jikan 1 (1981): 158–176.

Ji’nan 1975: Ji’nancheng Fenghuangshan 168 hao Hanmu fajue Zhenglizu 紀南城風凰山一六八號漢墓發掘整理組 (1975). “Hubei Jiangling Fenghuangshan 168 hao Han mu fajue jianbao 湖北江陵風凰山一六八號漢墓發掘簡報 (Brief report on the excavation of Han tomb no. 168 at Fenghuangshan, Jiangling, Hubei [province]).” Wenwu 1975.9: 1–8, and 22.

Jing Zhuyou 景竹友 (1993). “Santai Xindexiang Dong Han yamu qingli jianbao 三台新德鄉東漢崖墓清理簡報 (Brief report on the clearing of Eastern Han cliff tombs at Xindexiang, Santai [Sichuan province]).” Sichuan wenwu 1993.5 68–69.

Jing Zhuyou (1997). “Santai Yongmingxiang yamu diaocha jianbao 三台永明鄉崖墓調查簡報 (Brief report on investigations into the cliff tombs of Yongmingxiang, Santai [Sichuan province]).” Sichuan wenwu 1997.1: 63–71.

Knechtges, David R. (1982). Wen xuan or Selections of Refined Literature, volume one: Rhapsodies on Metropolises and Capitals. Princeton: Princeton University Press.

Knechtges, David R. (1987). Wen xuan or Selections of Refined Literature, volume two: Rhapsodies on Sacrifices, Hunting, Travel, Sightseeing, Palaces and Halls, Rivers and Seas. Princeton: Princeton University Press.


Kuhn, Dieter (1996). A Place for the Dead: An Archaeological Documentary on Graves and Tombs of the Song Dynasty (960–1279). Heidelberg: Ed. Forum.

Laing, Ellen (1978). “Patterns and Problems in Later Chinese Tomb Decoration.” Journal of Oriental Studies 16, 1/2: 3–20.

Lim, Lucy (ed.) (1987). Stories of China’s Past: Han Dynasty Pictorial Tomb Reliefs and Archaeological Objects from Sichuan Province, People’s Republic of China. San Francisco: The Chinese Culture Foundation of San Francisco.

Liu Zhiyuan 劉志遠 (1958). “Chengdu Tianhuishan yamu qingli ji 成都天迴山崖墓清理記 (Notes on the clearing of a cliff tomb at Tianhuishan, Chengdu [Sichuan province]).” Kaogu xuebao 1958.1: 87–103.

Luo Erhu 羅二虎 (1988a). “Santai Xian Qijiang yamu chutan 三台縣郪江崖墓初探 (Preliminary explorations of the cliff tombs at Qijiang, Santai Xian).” Sichuan wenwu 1988.4: 67–72.

Luo Erhu (1988b). “Sichuan yamu de chubu yanjiu 四川崖墓的初步研究 (Preliminary research into the cliff tombs of Sichuan).” Kaogu xuebao 1988.2:133–167.

Luo Erhu (1998). “Shaanxi Chenggu chutu de qianshu foxiang jiqi yu Sichuan diqu de guanxi 陜西城固出土的錢樹佛像及其與四川地區的關系 (The Buddhist images and money trees excavated at Chenggu, Shaanxi [province] and their relationship with Sichuan).” Wenwu 1998.12: 63–70.

Luoyang 1985: Luoyang Bowuguan 洛陽博物館 (1985). “Luoyang Jinguyuan Xinmang shiqi bihua mu 洛陽金谷園新莽時期壁畫墓 (Painted tombs of the Xinmang period at Jinguyuan, Luoyang [Henan province])” Wenwu ziliao congkan 9 (1985): 163–173.

Luoyang 1992: Luoyang Shi Di’er Wenwu Gongzuodui 洛陽市第二文物工作隊 (1992). “Luoyang Yanshi Xian Xin Mang bihua mu qingli jianbao 洛陽偃師縣新莽壁畫墓清理簡報 (Brief report on the clearing of Xinmang period painted tombs in Yanshi Xian, Luoyang [Henan province]).” Wenwu 1992.12: 1–8.

Luoyang 1993: Luoyang Shi Di’er Wenwu Gongzuodui (1993). “Luoy-ang Qianjingtou Xi Han bihua mu fajue jianbao 洛陽淺井頭西漢壁畫墓發掘簡報 (Brief report on the excavation of a Western Han tomb with painted walls at Qianjingtou, Luoyang [Henan province]).” Wenwu 1993.5: 1–16.

Mathieu, Rémi (1983). Étude sur la mythologie et l’ethnologie de la Chine ancienne. Paris: Collège de France, Mémoires de l’Institut des Hautes Études Chinoises.

Mianyang 2000: Mianyang Bowuguan 綿陽博物館 (2000). “Mianyang Bowuguan guancang wenwu 綿陽博物館館藏文物 (Cultural relics in the collections of the Mianyang Museum).” Sichuan wenwu 2000.6: four inserted colorplates.


Nanjing 1957: Nanjing Bowuyuan 南京博物院 (1957). “Changli shuiku Han muqun fajue jianbao 昌梨水庫漢墓群發掘簡報 (Brief report on the excavation of a group of Han tombs at the Changli Reservoir).” Wenwu cankao ziliao 1957.12: 29–40.

Nanjing 1981a: Nanjing Bowuyuan (1981). “Xuzhou Qingshanquan Baiji Dong Han huaxiangshi mu 徐州青山泉白集東漢畫象石墓 (Eastern Han tombs with pictorial reliefs at Baiji, Qingshanquan, Xuzhou).” Kaogu 1981.2: 137–150.

Nanjing 1981b: Nanjingshi Bowuguan 南京市博物館 (1981). “Nanjing beijiao Guojiashan Dong Jin muzang fajue jianbao 南京北郊郭家山東晉墓葬發掘簡報 (Brief report on the excavation of Eastern Jin tombs at Guojiashan, in the northern outskirts of Nanjing).” Wenwu 1981.12: 1–7.

Nanjing 1991: Nanjing Bowuyuan (1991). Sichuan Pengshan Handai yamu 四川彭山漢代崖墓 (Cliff tombs of the Han period at Pengshan, Si-chuan). Beijing: Wenwu Chubanshe.

Nanjing 1993: Nanjing Bowuyuan (1993). Fojiao chuchuan nanfang zhi lu wenwu tulu 佛教初傳南方之路：文物圖錄 (Illustrated record of relics concerning the early transmission of Buddhism via the southern route). Beijing: Wenwu Chubanshe.

Nanyang 1982: Nanyang Han huaxiangshi Bianweihui (1982). “Tanghe Xian dianchang Han huaxiangshi mu 唐河縣電廠漢畫像石墓 (A Han tomb with pictorial reliefs near the Tanghe Xian electric plant).” Zhongyuan wenwu 1982.1: 5–11.

Neijiang 1991: Neijiangshi Wenguansuo 內江市文管所 and Jianyang Xian Wenhuaguan 簡陽縣文化館 (1991). “Sichuan Jianyang Xian Guitoushan Dong Han mu 四川簡陽縣鬼頭山東漢崖墓 (An East-ern Han cliff tomb at Guitoushan, Jianyang Xian, Sichuan).” Wenwu 1991.3 20–25.

Nylan, Michael (2001). “The Legacies of the Chengdu Plain.” In Robert Bagley (ed.) 2001:309–325.

Pirazzoli-t’Serstevens, Michèle (2001). “Sichuan in the Warring States and Han Periods.” In Robert Bagley (ed.) 2001:39–57.

Powers, Martin (1983). “A Late Western Han Tomb near Yangzhou and Related Problems.” Oriental Art 29: 275–290.

Rawson, Jessica (ed.) (1996a). Mysteries of Ancient China: New Discoveries from the Early Dynasties. London: British Museum Press.

Rawson, Jessica (1996b). “Changes in the Representation of Life and the Afterlife as Illustrated by the Contents of Tombs of the T’ang and Sung Periods.” In Maxwell K. Hearn and Judith G. Smith (eds.), Arts of the Sung and Yüan, pp. 23–43. New York: Metropolitan Museum of Art.

Rawson, Jessica (2001). “Tombs and Tomb Furnishings of the Eastern Han Period (AD 25–220).” In Robert Bagley (ed.) 2001:253–267.


Rhie, Marylin Martin (1999). Early Buddhist Art of China and Central Asia, vol. 1. Leiden: Brill.

Santai (2002). Santai Xian Wenhua Tiyuju 三台縣文化體育局 and Santai Xian Wenwu Guanlisuo 三台縣文物管理所. “Sichuan Santai Xian Qijiang yamu qun dianxing yamu qingli jianbao 四川三台縣郪江崖墓群典型崖墓清理簡報 (Brief report in 2000 concerning the excavation of a group of cliff tombs at Qijiang, Santai Xian, Sichuan).” Wenwu 2002.1:16–41.

Ségalen,Victor (1915). “Premier exposé des résultats archéologiques obtenus dans la Chine occidentale par la mission Gilbert de Voisins, Jean Lartigue et Victor Ségalen,” part 1: La sculpture sur pierre dans la Chine antique; and part 2: Les tombes de falaises du Sseu-tch’ouan. Journal Asiatique, 11.5: 467–486; and 11.6: 282–306.

Ségalen, Victor (1917). “Recent Discoveries in Ancient Chinese Sculp-ture.” Journal of the North China Branch of the Royal Asiatic Society 48: 145–162.

Ségalen,Victor, Auguste Gilbert de Voisins, and Jean Lartigue (1923). Mission archéologique en Chine, atlas, vol. 1: La sculpture et les monuments funèraires. Paris: Paul Geuthner.

Shandong 1975: Shandong Sheng Bowuguan 山東省博物館 and Cangshan Xian Wenhuaguan 蒼山縣文化館 (1975). “Shandong Cangshan yuanjiayuan nian huaxiangshi mu 山東蒼山元嘉元年畫象石墓 (A tomb with pictorial reliefs dated AD 424 at Cangshan, Shandong).” Kaogu 1975.2: 124–134.

Sichuan 1956: Sichuan Sheng Wenwu Guanli Weiyuanhui 四川省文物管理委員會 (1956). “Sichuan Xinfan Qingbaixiang Dong Han huaxiangzhuan mu qingli jianbao 四川新繁清白鄉東漢畫像磚墓清理簡報 (Brief report on the excavation of an Eastern Han tomb with pictorial bricks at Qingbaixiang, Xinfan, Sichuan [province]).” Wenwu cankao ziliao 1956.6: 36–44.

Sichuan 1979: Sichuan Sheng Bowuguan 四川省博物館 and Pi Xian Wenhuaguan 郫縣文化館 (1979). “Sichuan Pi Xian Dong Han zhuanmu de shiguan huaxiang 四川郫縣東漢磚墓的石棺畫像 (Pictorial scenes on the stone coffin from an Eastern Han brick tomb in Pi Xian, Sichuan).” Kaogu 1979.6: 495–503.

Sichuan 1985: Sichuan Sheng Wenwu Guanli Weiyuanhui 四川省文物管理委員會 (1985). “Sichuan Zhong Xian Tujing Shu Han yamu 四川忠縣涂井蜀漢崖墓 (Cliff tombs of the Kingdom of Shu Han at Tujing in Zhong Xian, Sichuan).” Wenwu 1985.7: 49–95.

Stein, Rolf A. (1990). The World in Miniature: Container Gardens and Dwellings in Far Eastern Religious Thought. Stanford: Stanford University Press.

Su Bai 宿白 (1957). Baisha Song mu 白沙宋墓 (The Song tombs at Bai-sha). Beijing: Wenwu Chubanshe.


Sun Hua 孫華 (1991). “Santai Qijiang yamu suojian Han dai jianzhu xingxiang shulüe 三台郪江崖墓所見漢代建築形象述略 (Han dy-nasty architectural figures seen in the cliff tombs at Qijiang, Santai [Sichuan province]).” Sichuan wenwu 1991.5: 10–16.

Sun Hua, Zheng Dingli 鄭定理, and He Zhiguo 何志國 (1998). “Santai Xian Qijiang yamu 三台縣郪江崖墓 (Cliff tombs along the Qijiang in Santai Xian).” In Sichuan kaogu baogaoji, pp. 350–381. Beijing: Wenwu Chubanshe.

Sun Zuoyun 孫作云 (1977). “Luoyang Xi Han Bu Qianqiu bihua kaoshi 洛陽西漢卜千秋墓壁畫考釋 (An Analysis of the Western Han Murals in the Luoyang Tomb of Bu Qianqiu).” Wenwu 1977.6: 17–22.

Sun Zuoyun (1979). “An Analysis of the Western Han Murals in the Luo-yang Tomb of Bu Qianqiu.” Trans. of Sun Zuoyun 1977 by Suzanne Cahill. Chinese Studies in Archeology 1, no. 2 (Fall 1979): 44–78.

Tang, Changshou 唐長壽 (1993). Leshan yamu he Pengshan yamu 樂山崖墓和彭山崖墓 (The cliff tombs at Leshan and at Pengshan). Chengdu: Dianzi Keji Daxue Chubanshe.

Tjan Tjoe Som (trans.) (1952). Po Hu T’ung: The Comprehensive Discussions in the White Tiger Hall. Leiden: E. J. Brill.

Torrance, T. (1910). “Burial Customs in Sz-chuan.” Journal of the North China Branch of the Royal Asiatic Society 41: 57–75.

Torrance, T. (1930–31). “Notes on the Cave Tombs and Ancient Burial Mounds of Western Szechuan.” Journal of the West China Border Research Society 4: 88–96.

Uehara Kazu 上原和 (1994). “Kōguri kaiga no Nihon-e oyoboshita eikyō 高句麗絵画の日本へ及ばした影響 (The influence of Koguryo paintings on Japan. Relations among China, Korea, and Japan in ancient times, viewed from the expression of lotus designs).” Bukkyō geijutsu 仏教芸術 215: 74–103.

Wang Qipeng 王啟鵬 and Wang Kongzhi 王孔智 (1989). “Zhongjiang Xian Yuguixiang Dong Han yamu diaocha jianbao 中江縣玉桂鄉東漢崖墓調查簡報 (Brief report on investigations of an Eastern Han cliff tomb at Yuguixiang, Zhongjiang Xian).” Sichuan wenwu 1989.5: 64–66.

Wang Shouzhi 王壽芝 (1987). “Chenggu chutu de Handai taodu 城固出土的漢代桃都 (Han dynasty peach tree unearthed at Chenggu).” Wenbo 1987.6: 91–92.

Wang Tingfu 王庭福 and Li Yihong 李一洪 (1995). “Hejiang Zhangjia-gou erhao yamu huaxiangshiguan fajue jianbao 合江張家溝二號崖墓畫像石棺發掘簡報 (Brief report on the excavation of a stone coffin with reliefs in tomb no. 2 at Zhagjiagou, Hejiang).” Sichuan wenwu 1995.5: 65–66.


Weishan 2000: Weishan Xian Wenwu Guanlisuo 微山縣文物管理所 (2000). “Shandong Weishan Xian chutu de Han huaxiangshi 山東微山縣出土的漢畫像石 (Han pictorial reliefs excavated from Weishan Xian, Shandong).” Wenwu 2000.10.68–73.

Wu Hung (1989). The Wu Liang Shrine: The Ideology of Early Chinese Pictorial Art. Stanford: Stanford University Press.

Wu Hung (1995). Monumentality in Early Chinese Art and Architecture. Stan-ford: Stanford University Press.

Wu Hung (2000). “Mapping Early Taoist Art: The Visual Culture of Wudoumi Dao.” In Stephen Little (ed.), Taoism and the Arts of China, pp. 77–93. Chicago: The Art Institute of Chicago.

Xin Lixiang 新立祥 (2000). Han dai huaxiangshi zonghe yanjiu 漢代畫像石綜合研究 (Comprehensive studies into pictorial reliefs of the Han dynasty). Beijing: Wenwu Chubanshe.

Yang Cunguan 楊存貫 and Zhong Zhi 鍾治 (2000). “Santai diaocha Han Jin yamuqun wenwu yicun 三台調查漢晉崖墓群文物遺存 (Cultural remains of a group of cliff tombs of the Han and Jin periods in Santai).” Zhongguo wenwu bao 中國文物報 (August 23, 2000): 1.

Yangzhou 1980: Yangzhou Bowuguan 揚州博物館 and Hanjiang Xian Wenhuaguan 邗江縣文化館 (1980). “Yangzhou Hanjiang Xian Huchang Han mu 揚州邗江縣胡場漢墓 (Han tombs at Huchang, Hanjiang Xian, Yangzhou).” Wenwu 1980.3: 1–10.

Yangzhou 1988: Yangzhou Bowuguan (1988). “Jiangsu Hanjiang Yao-zhuang 101 hao Xi Han mu 江蘇邗江姚庄 101 號西漢墓 (Western Han Tomb No. 101 at Yaozhuang, Hanjiang, Jiangsu).” Wenwu 1988.2: 19–43.

Yu Weichao 俞偉超 (editor) (2000a). Zhongguo huaxiang shi quanji: juan 7: Sichuan Han huaxiang shi 中國畫像史全集: 卷 7: 四川漢畫像史 (The

history of Chinese pictorial painting, Vol. 7: The history of Han picto-rial painting in Sichuan). Zhengzhou: Henan Meishu Chubanshe.

Yu Weichao (chief editor) (2000b). Chang jiang Sanxia wenwu cunzhen 長江三峽文物存真 (Cultural remains from the Three Gorges region of the Changjiang). Chongqing: Chongqing Chubanshe.

Zeng Zhaoyu 曾昭燏, Jiang Baogeng 蔣寶庚, and Li Zhongyi 黎忠義 (1956). Yi’nan gu huaxiangshi mu fajue baogao 沂南古畫像石墓發掘報告 (Report on the excavation of ancient tombs with pictorial reliefs at Yi’nan). Beijing: Wenhuabu Wenwu Guanliju 文化部文物管理局.

Zhao Dianzeng 趙殿曾 and Yuan Shuguang 袁曙光 (1991). “Sichuan Zhongxian San Guo tong foxiang ji yanjiu 四川忠縣三國銅佛像及研究 (A study of bronze Buddha images of the Three Kingdoms period from Zhong Xian, Sichuan).” Dongnan wenhua 東南文化 1991.5.55–61.

Zhengzhou 1996: Zhengzhou Shi Wenwu Kaogu Yanjiusuo 鄭州市文


物考古研究所 and Xingyang Shi Wenwu Baohu Guanlisuo 滎陽市文物保護管理所 (1996). “Henan Xingyang Changcun Han dai bihua mu diaocha 河南滎陽萇村漢代壁畫墓調查 (Investigations of painted tombs of the Han dynasty at Changcun, Xingyang, Henan).” Wenwu 1996.3: 18–27.

Zhong Jian 鐘堅 (1987). “Sichuan Lushan chutu Handai shike loufang 四川蘆山出土漢代石刻樓房 (Han dynasty stone house model un-earthed at Lushan, Sichuan).” Wenwu 1987.10: 95.

Zhong Zhi 鍾治 (2000). “Han dai fanhua yi xue zhong 漢代繁華一穴中 (The prosperity of the Han dynasty reflected in a grave).” Zhongguo wenwu bao (Nov. 29, 2000), yuemo jianshang (月末鑒賞), p. 4.

Zhong Zhi (manuscript). “Sichuan Santai Yuanbaoshan yamu qingli jianbao 四川三台元寶山崖墓清理簡報 (Brief report about a cliff tomb at Yuanbaoshan, Santai, Sichuan).”

Editions of Classical texts:

Baihutong shuzheng 白虎通疏證, by Chen Li 陳立. Beijing: Zhonghua, 1994.

Lunheng Jiaoshi 論衡校釋 , by Huang Hui 黃暉 . Beijing: Zhonghua, 1990.

Shanhaijing jiaozhu 山海經校注, by Yuan Ke 袁珂. Shanghai: Shanghai Guji, 1980.

Shiji 史記, by Sima Qian 司馬遷. Beijing: Zhonghua Shuju, 1959 (new ed. 1985).

Wen xuan 文選, reprint of 1810 ed., with index. Beijing: Zhonghua Shuju, 1977.

Address:

Humanities DepartmentUniversity of Michigan-Dearborn4901 Evergreen Road, Dearborn, MI 48128–1491E-mail: <[email protected]>


Antiquities of Northern Tibet: Pre-Buddhist Archaeological Discoveries on the High Plateau, John Vincent Bellezza. Adroit Publishers, Delhi, 2001, 436 pages, 276 photographs, 24 maps.

Antiquities of Upper Tibet: Pre-Buddhist Archaeological Sites on the High Plateau, John Vincent Bellezza. Adroit Publishers, Delhi, 2002, 304 pages, 219 photographs, 16 maps.

Ethnogenesis, Migration, Diffusion or All of the Above? The Pre-Buddhist Archaeology of Northern Tibet

Despite the able efforts of Chinese archaeologists who have conducted archaeological research on the Tibetan plateau since the 1970s, much, if not most, of Tibetan prehistory remains a mystery. Their efforts have provided us with important, but still fragmentary, outlines of the past. We remain uncertain about many critical facets, such as the routes and timing of the arrival of the first inhabitants of the plateau, the develop-ment of social and political complexity, and the nature of interaction of the plateau’s inhabitants with their neighbors. To be sure, the plateau is vast and hard to traverse, and combined with the real stressors of seasonal cold and hypoxia, these conditions have made it difficult for lowlander archaeologists to work the plateau in a systematic fashion. Thus it is within this context of difficulty, hardship, and ignorance that we should examine these two books published by John Vincent Bellezza on one of the most remote parts of the Tibetan plateau—the Chang Tang (a.k.a. Xizang gaoyuan beibu 西藏高原北部 or the “northern Tibetan plateau”). There is no question that what is contained in these books, as well as other research he has conducted on the plateau over the past fifteen years, has substantially improved our understanding of the later prehistory of the plateau, and for this they are welcome indeed. How-ever, while they have provided us with new and valuable ways to think of this remote place, we must also acknowledge their limitations as well so as to move beyond them and find avenues to place his findings more securely into the Tibetan past. In many ways, Bellezza is a reminder of a more recent Tibetan past, when scholar-explorers like Jacques Bacot (1877–1965) traveled through eastern Tibet in the early 20th c. and described what he called “mega-liths” (Macdonald 1953). George Roerich (1902–1960), who in the 1920s

472 book review

crossed northern Tibet and parts of the Chang Tang, also described a number of megalithic sites, some of which Bellezza visited some 50 years later (Roerich 1930). He also reminds me of the famous Italian Tibetologist, Giuseppe Tucci (1895–1984), who undertook a number of scholarly expeditions across Tibet that were focused primarily upon art, archaeology, and aspects of material culture, and who published much of his research in the four volumes that comprise the Indo-Tibetica series (Tucci 1932–45). Like these earlier travelers, Bellezza has compiled an impressive body of new and important knowledge about the Tibetan past. The volumes under review, as well as a third, Divine Dyads (Bellezza 1997), represent the culmination of almost 20 years of research in north-ern Tibet. His work has been directed at the definition of “pre-Buddhist” culture, a nebulous concept that potentially spans from deep antiquity until the appearance of Buddhism on the plateau after 620 CE. Bellezza narrows it, however, and focuses his attention on a construct labeled “Zhang zhung-Mon”, which he uses to “ . . . designate the pre-Buddhist population of Upper Tibet west of 88 degrees east longitude from the First Millennium BCE to the beginning of the Imperial Era in the early 7th century as reflected in the common tradition of monuments and rock art.” (Bellezza 2002: 10). The extent of his explorations is large, and including the area cov-ered in Divine Dyads, it stretches across the Chang Tang in a band from roughly 30° N latitude to 33° N, and from Lake Namtso (a.k.a. Namco or Namucuo 納木錯) in the east to Rutok (a.k.a. Rutog or Ritu 日土) at the western border with India. This band includes most of the major lakes of the plateau and includes the most densely inhabited part of the Chang Tang. He also examined extreme southwestern Tibet including the headwaters of the Sutlej (a.k.a. Sutlei or Xiangquan He 象泉河, a tributary to the Indus) where the reputed capital of the Zhang zhung polity—Kyunglung (a.k.a. Yinghui liuyu 鷹徽流域)—is located. Multiple ecologies are represented here, ranging from the very high, relatively flat plains to the north to the rugged, narrow, and very dry valleys to the southwest. His research style is informed by an interesting and sometimes very effective blend of textual sources, oral tradition, ethnography, and field reconnaissance. However, it is important to note that he is not trained as an archaeologist, and further, because of the nature of his permis-sion to work on the plateau, he was unable to collect artifacts from the sites he discovered, and could not excavate. Moreover, his survey strat-egy was not systematic. That is, he relied extensively upon informant interviews to locate sites. Consequently, his reporting on the spatial distributions of the sites discovered is likely to be affected by unknown

book review 473

biases. The lack of even grab samples of ceramics and other cultural materials from these sites is perhaps the greatest obstacle to their long-term potential. Without these materials, as I shall discuss below, the sites float in time, thus making it difficult to assign them confidently to a meaningful chronological period. It is important, however, to keep this apparent shortcoming in perspective. Because there have been almost no systematic stratigraphic excavations in the entire Chang Tang, there are no robust ceramic (or lithic) chronologies for the region in any case. While samples of materials would have been of very great value could he have collected them, they would have still “floated in time” until those excavations were realized. Over the course of his expeditions, he has recorded almost 500 sites, and has placed them into the following categories: hilltop forts and other edifices on summits, caves and their local settings, free-standing religious edifices, sedentary villages, graves with superficial superstructures of all types, cubic-shaped graves built on summits, isolated pillars, stelae erected within rectangular enclosures, monolithic arrays with structures, other monuments, and pictographs and petroglyphs (Bellezza 2001: 28–39). Although he has few site maps, he illustrates many of these site types with photographs, and describes the sites within their local settings. When appropriate, he also connects sites to their textual and mythi-cal contexts. The photographs of these sites are especially important, because it is this visual component that Bellezza turns to frequently to place these sites in time. Lacking chronology via material culture, Bellezza was forced to rely heavily upon two sources to place the sites he discovered into his con-struct: the textual sources of Bon, and local oral tradition. Secondary lines of evidence, somewhat circular in nature, included evidence of sites with “non-Buddhist characteristics,” site locations in “pre-Bud-dhist geographical settings,” and comparisons of discovered materials, usually architecture or mortuary monuments with regions beyond the Chang Tang (Bellezza 2001: 16–26). Bon is an important religious tradition on the plateau today, and according to Kvaerne (1997: 9–10), it has three meanings to western scholars: it is the pre-Buddhist religion of Tibet, a religion similar to Bud-dhism that appeared on the plateau after 1000 CE, and finally, a body of what might be best seen as “folk” or popular beliefs about the nature of spirits, demons, and ritual. Many scholars, among them Bellezza, believe that Bon was the state religion of the Zhang zhung polity. Bon authors have generated an enormous historical and ecclesiastical literature over the centuries, and it is this to which Bellezza turns first to define an archaeological locality as having a “Zhang zhung-Mon” affiliation. The problem, of course, is that the Bon literature is rarely specific about

474 book review

a locality, and further, as Bellezza (2001: 18) notes, “ . . . the historicity and chronology of Zhang zhung, as derived from Bon literary refer-ences, have yet to be satisfactorily established . . . ”. Thus while some assignments may be accurate, the problem is in knowing when this is the case. The use of the term “Mon” is also problematic, which Bellezza (2001: 40–41) also acknowledges. Pommaret (1999: 52) notes that the term is used variously to describe groups that have little or nothing in common. It has wide currency in Tibet, and has been used to describe groups from Ladahk to Bhutan. It tends to be an exonym, that is, a term used by some group to denote the ethnicity of some ‘other.’ The term is of relevance to this discussion because Bellezza (2002: 9) observes that in the oral tradition of those who currently live in the Chang Tang, the creators of the archaeological sites within this region are routinely referred to as the Mon-pa. Although oral tradition is now more widely accepted as a powerful source of insight into the past in many parts of the world, like the Bon literature, it has not been subjected to a thorough critique of its historicity. We are thus left with a good deal of uncertainty about the temporal assignments of these sites, but some may, in fact, be accurate. In my own research in defining “Zhang-zhung ethnicity” at Dindun in far western Tibet (Aldenderfer 2003a; in press; Aldenderfer and Moyes 2004), for example, I have discovered a standing stone-structure complex which dates to ca. 550–100 BCE that is very similar to a structure Bellezza (2001: 411) describes from the Chang Tang. Since domestic architecture is one powerful way to recognize ethnicity in the archaeological record of many peoples, at least some of Bellezza’s speculations may prove to be valid. The debate about “Zhang zhung-Mon” or “Zhang zhung” ethnic-ity provides an interesting window into how Tibetan prehistory insofar as we know it is understood and conceptualized. If sites float in time, it is only sensible, as Bellezza does, to seek sources of architectural tradition, for instance, in nearby places where such traditions have been well-dated. What becomes problematic, however, is that when a potential analog or source is located, the appearance of the traits on the plateau are “explained” either by migration of peoples or the diffu-sion of traits into an undefined cultural milieu. Models of the Tibetan past are rife with this kind of thinking. Peoples move onto the plateau from neighboring regions, or ideas or traits appear on the plateau and in some instances, simply cross the plateau and leave no trace behind. The indigenous inhabitants of the plateau in most of these explanations of change are either passive recipients of culture from other groups, or are simply replaced by others. More nuanced models of exchange, ac-

book review 475

culturation, and frontier interactions have yet to make their appearance on the plateau. In other words, there is no indigenous ethnogenesis in the Tibetan past. Two examples illustrate my contention: explanations for the Tibetan Neolithic, and Bellezza’s definition of the origins of the Zhang zhung-Mon cultural entity. Zhang Yinong and I (Aldenderfer and Zhang 2004: 34–40) have summarized the most recent models of the Tibetan Neolithic by Su et al. (2000), Ren (2000), and van Driem (2001, 2002). The Su et al. model is based upon the distribution of Y haplotypes in east Asia, and they argue that around 6000 BP, proto-Tibeto-Burman speakers move from Qinghai 清海 onto the central plateau. At this time, a “Central Asian” set of haplotypes is added to these migrants. Ren’s model is based primarily upon paleoenvironmental data, and he sees that around 6000 BP, a “Neolithic package” moves from the east up the major river valleys onto the central plateau. The carriers of this package are apparently lowlanders moving into the high elevations of the plateau. Finally, van Driem’s model is most comprehensive of the explanations for the ap-pearance of the Neolithic on the plateau, and combines archaeological, DNA, and linguistic data. He postulates the existence by 11,500 BC of an ancestral homeland of Tibeto-Burmese speakers in the middle and upper Yangtze river of Sichuan 四川. This population served as the source of the Early Neolithic groups in both the eastern (Peiligang 裴里岡) and western (Dadiwan 大地灣) Yellow River basin. Both areas were occupied as early as 6,500 B.C., and van Driem labels them as “Northern Tibeto-Burmese” speakers. He suggests that the clearest archaeological correlate of this migration is what he terms the “abrupt” replacement of local microlithic-using cultures by the Dadiwan cultivators (van Driem 2001: 417). These migrants brought with them cord-marked pottery and polished stone tools, which van Driem takes as cultural and ethnic markers. The postulated linguistic divergence that created Northwestern Tibeto-Burmese (Bodic) languages is directly correlated by van Driem with the appearance of the Majiayao 馬家窯Neolithic in Gansu 甘肅 and extreme eastern Qinghai, which he dates as early as 3900 BC These Bodic-speaking migrants spread across the plateau after this date. None of these models considers whether indigenous peoples were present on the plateau during these migrations. Current thinking about the earliest occupation of the plateau places it around 30–35,000 years ago (Aldenderfer 2003b), and Brantingham et al. (2003) have developed a model of how this occupation may have taken place. Materials confi-dently, if not yet precisely, dated to the Paleolithic are found in the Chang Tang (Brantingham, Olsen, and Schaller 2001) as well as across the central plateau (Aldenderfer and Zhang 2004:15–21). Assuming these

476 book review

peoples survived the Last Glacial Maximum, an assertion supported by the recent discovery of Chusang on the central plateau near Lhasa which is dated to ca. 20,000 years ago (Zhang and Li 2002), what were they doing at ca. 6000 BP? How did they interact with these presumed migrants? These indigenous peoples could have been the source of Su et al.’s Central Asian genetic contribution, implying a process of assimila-tion or intermarriage with the migrants. Unfortunately, the models are silent about these peoples. Considering the the Zhang zhung-Mon construct, on the one hand, Bellezza believes that it faithfully represents at least one indigenous population of the Chang Tang and northern Tibet, but is open to the possibility that there may be other ethnic groups present. In some instances, especially sites with rock art of animals and hunting scenes, he is certain that the distribution and diversity of these animal motifs “ . . . argues in favor of a long process of indigenous development not limited to a foreign vector of transmission from the First Millennium BC” (Bellezza 2002: 13–138). This seems to be an argument for a kind of ethnogenesis, or in situ development from an as-yet unidentified source. On the other hand, he believes that a strong case can be made for a foreign origin for many of the core traits of the construct, specifically certain forms of mortuary architecture, isolated pillars, stelae erected within a quadrangular (or rectangular) superstructure, monolithic ar-rays, and rock art. Bellezza believes that stylistic features of these site types can be seen in various Central Asian locations, more specifically the steppelands of eastern Mongolia, southern Siberia, and beyond. He places the sites into a time frame that is no earlier than 1000 BCE, a reasonable assertion, but then argues that while the evidence is still problematic, “These types of close physical correspondences can only be attributable to cultural affinities between the steppes and the high plateau of Tibet. However, the exact nature and period in which this cultural diffusion took place is still obscure” (Bellezza 2002: 106). Dif-fusion of traits into an accepting local population is now invoked as an explanation for their presence. While Bellezza’s work shows sensitivity to the possibility that indigenous peoples were present on the plateau, they still seem to be relatively passive acceptors of traits from elsewhere. This explanation also suggests that his Zhang zhung-Mon construct owes more of its expression to peoples beyond the plateau rather than the indigenous peoples to whom these traits diffused. What we are left with, then, is considerable uncertainty about the applicability of his construct in time, space, and perhaps more importantly, process. We have much to learn about the Tibetan past, and efforts like those of Bellezza are important contributions toward this end. Although this progress ultimately depends on a far greater scale of excavation and

book review 477

systematic research than seems likely in the foreseeable future, it will also benefit from more creative model building. This will entail more nuanced thinking about how to define ethnicity and identity on the plateau, and to avoid using either diffusion or migration as the explanations of first choice. This effort will benefit from the introduction of new technologies, specifically the analysis of ancient DNA (Kaestle and Horsburgh 2002). While it would be very naïve to think that DNA tracks “ethnicity”, it does provide information on population variability. When these data are combined with better control of time and the variation of material culture through it, dealing with archaeological questions of ethnicity becomes far simpler. If we can successfully combine these studies with new historical research on the ethnic composition of the plateau in the period immediately prior to the founding of the Tibetan empire (Vitali 2003) we will then be in an excellent position to truly understand the ethnogenesis of the Tibetan people.

References cited

Aldenderfer, Mark (2003a). “Domestic rdo-ring? A new class of standing stone from the Tibetan plateau.” Tibet Journal 28 (1&2): 3–20.

Aldenderfer, Mark (2003b). “Moving up in the world.” American Scientist 91: 542–54.

Aldenderfer, Mark (in press). “Defining Zhang zhung ethnicity: An archaeological perspective from far western Tibet.” In Amy Heller and Giacomella Orofino (eds.), Western Tibet and Western Himalayas: Essays on History, Literature, Archaeology, and Art, Proceedings of the Tenth IATS, pp. to be determined. Leiden: Brill.

Aldenderfer, Mark and Holley Moyes (2004). “Excavations at Dindun, a pre-Buddhist village site in far western Tibet.” In Huo Wei and Li Yongxian (eds.), Proceedings of the First International Conference on Tibetan Archaeology and Art, pp. 47–69. Chengdu, China: Center for Tibetan Studies, Sichuan University.

Aldenderfer, Mark and Zhang Yinong (2004). “The prehistory of the Tibetan Plateau to the Seventh C. A.D.: Perspectives from China and the West since 1950.” Journal of World Prehistory 18 (1): 1–55.

Bellezza, John (1997) Divine Dyads: Ancient Civilization in Tibet. Dharamsa-la, India: Library of Tibetan Works and Archives.

Brantingham, P. Jeffrey, John W. Olsen, and George B. Schaller (2001). “Lithic assemblages from the Chang Tang region, northern Tibet.” Antiquity 75 (288): 319–327.

Brantingham, P. Jeffrey, Ma Haizhou, John W. Olsen, and Gao Xing, David Madsen, and David E. Rhode (2003). “Speculation on the

478 book review

timing and nature of Late Pleistocene hunter-gatherer colonization of the Tibetan Plateau.” Chinese Science Bulletin 48(14): 1510–1516.

Kaestle, Frederika and K. Ann Horsburgh (2002). “Ancient DNA in anthropology: Methods, applications, and ethics.” Yearbook of Physical Anthropology 45: 92–130.

Kvaerne, Per (1997). The Bon Religion of Tibet. Boston: Shambhala.Macdonald, A. W. (1953). “Une note sur les mégaliths Tibétains.” Journal

Asiatique 241: 63–76.Pommaret, Françoise (1999). “The Mon-pa revisited: In search of

Mon.” In Toni Huber (ed.), Sacred Spaces and Powerful Places in Tibetan Culture, pp. 52–76. Dharamsala, India: Library of Tibetan Works and Archives.

Ren, G. (2000). “Decline of the mid-to-late Holocene forests in China: Climate change or human impact?” Journal of Quaternary Science 15: 273–281.

Roerich, George (1930). The Animal Style among the Nomad Tribes of Northern Tibet. Prague: Seminarium Kondakovianum.

Su, Bing, Junhua Xiao, Peter Underhill, Ranjan Deka, Weiling Zhang, Joshua Akey, Wei Huang, Di Shen, Daru Lu, Jingchun Luo, Jiayou Chu, Jiazhen Tan, Peidong Shen, Ron Davis, Luca Cavalli-Sforza, Ranajit Chakraborty, Momiao Xiong, Ruofu Du, Peter Oefner, Zhu Chen, Li Jin (1999). “Y-chromosome data evidence for a northward migration of modern humans into eastern Asia during the last Ice Age.” American Journal of Human Genetics 65: 1718–1724.

Tucci, Giuseppe (1932–45). Indo-Tibetica. Rome: Reale Accademia d’Italia.

van Driem, George (2001). Languages of the Himalayas: An Ethnolinguistic Handbook of the Greater Himalayan Region. 2 vols, Leiden: Brill.

van Driem, George (2002). “Tibeto-Burman phylogeny and prehistory: Languages, material culture, and genes.” In Peter Bellwood and Colin Renfrew (eds.), Examining the Farming/Language Dispersal Hypothesis, pp.233–249. Cambridge: McDonald Institute for Archaeological Research, University of Cambridge.

Vitali, Roberto (2003). “Tribes which populated the Tibetan plateau, as treated in the texts collectively called the Khungs chen po bzhi.” Lungta 16: 37–63.

Zhang, David and Shenghua Li. (2002). “Optical dating of Tibetan human hand- and footprints: An implication for the palaeoenviron-ment of the last glaciation of the Tibetan Plateau.” Geophysical Research Letters 29: 16-1–16-3.

Reviewed by Mark Aldenderfer (Department of Anthropology, Uni-versity of Arizona)


Greiff, Susanne and Yin, Shenping, Das Grab des Bin Wang. Wandmalereien der Östlichen Han-Zeit in China (The Tomb of Bin Wang. Murals of the Eastern Han in China). Mainz: Römisch-Germanisches Zentralmu-seum, 2002. 99 pages, 42 color plates + 15 bw plates.

In October 2000, three Eastern Han brick tombs were discovered on the grounds of a brickyard at Baizi 百子 in Xunyi 旬邑 County, Shaanxi Province. One of them, the tomb of Bin Wang 邠王 (Lord Bin), was found to be richly decorated with wall paintings covering some 50 m2 of the inner surfaces. Das Grab des Bin Wang (The Tomb of Bin Wang) presents the results of a co-operative endeavor between the Roman Germanic Central Museum (Römisch-Germanisches Zentralmuseum) in Mainz, Germany, and the Archaeological Institute of Shaanxi Prov-ince in Xi’an, China, in rescuing the murals by removing them from the tomb. This task had to be accomplished within only four weeks. The tomb was destroyed in May 2001 by the brickyard owner, immediately after the removal. The murals are now conserved in Xi’an. The photo-graphs of the vibrantly colored murals that are published in this book, however, were taken in situ. They are the most valuable documentation of these 1,800 year-old murals, reminding us at the same time of the modern dilemma of how economic development often takes place at the expense of cultural heritage. The book comprises three parts: color plates and diagrams, the Ger-man text (32 pp.), and its Chinese translation (35 pp.). The text begins with a short but impressive introduction describing the circumstances and condition of the find, and the difficulties encountered in removing the murals (p. 2). An overview covering the history, philosophy, mythol-ogy, religion, sepulchral culture, and tomb types of the Han Dynasty (p. 3–10) is followed by the description of the tomb structure, grave goods, and the murals of this tomb (p. 11–13). The main body of the text (p. 15–30) contains detailed discussions of the motifs depicted in the murals, as well as their historical background. The murals are designed as follows: the entrance corridor is painted with four guardian figures, while two warriors face outward (one car-ries the inscription lishi 力士), and two gate guardians (identified by the inscriptions tingzhang 亭長 and menzhe 門者) face the chambers (p. 15). On the walls surrounding the antechamber are scenes with animal herds and landscapes. The animal herds are combined with two large figures, interpreted as “exorcists” by the authors (p. 15). The landscape

480 book review

scenes include ploughs and harvest in the front part of the antechamber and groups of people under trees in the rear. The antechamber is the only room with a cloister vault, on which a celestial sphere is depicted. The directional animals are painted above the entrance corridor (south) and above the entrances of both side chambers (east and west). The murals above the entrance of the rear chamber are badly damaged, and therefore hardly recognizable. Above the dragon (west) and the tiger (east) are the moon and the sun and stars (p. 16). Scrolls of clouds rise from the base of the vault and climb along its four ridges towards the apex of the vault. The murals in the eastern side chamber illustrate the preparation of banquets: two butchers in front of a beef rack, and maids carrying away food (p. 25). The murals of the western side chamber are badly damaged; only one standing female figure remains. The authors relate this scene, together with the figures under the trees of the antechamber, to the “private sphere” (p. 23). The scenes in the rear chamber show an audience scene: Bin Wang and his wife (p. 15, 18) are receiving General Guo 郭氏將軍, several officials (one entitled xiao shi 小史 meaning “junior scribe,” and two subordinates of General Guo), and their wives (p. 22), as identified by inscriptions. The wives of Bin Wang, General Guo, and the scribe are depicted with children and maids (p. 19, 23). The males, except for Bin Wang and General Guo, are standing in three quarter profile, the ladies sitting frontally. On the back wall a huge, painted yellow gate with a single closed door panel is interpreted by the authors of the report as tianmen 天門 “Heaven’s Gate” (p. 20–21). Two carriages that are reserved, according to the inscriptions, for Bin Wang’s personal use, are driving from the closed door toward the male figures (p. 19). A special section of the book is dedicated to the composition of the ground layers of the paintings and the preliminary results of the pigment analysis with a Raman microscope (p. 27–29). The authors observe a dominance of red and brown colors against a beige and grey colored ground, while no clear blue or green colors are adopted. Even the “Blue Dragon” of the East is colored pink violet instead of blue (p. 15). This violet color, which is also used for the blossoms on the trees and for the robes, is identified as one of the earliest synthetic pigments in China, the so-called “Han purple,” a barium-copper-silicate (p. 29). The color plates, together with the technical analyses of the ground layers and the pigments, form the most impressive and valuable parts of the book. The chemical analyses, though short, provide an insight into the techniques of local painters, the availability of materials used for the pigments, and the color spectrum of their works. Like only a few predecessors, this book provides a clear overview and details of

book review 481

the major motifs of the murals of a single tomb. It allows readers to reconstruct the paintings and, most of all, to experience visually the polychrome world that this local community had once appreciated. On the other hand, the sections on history and art history are over-loaded with general surveys of social, economic, political, and cultural topics concerning the entirety of the Han Dynasty from 206 BC to AD 220. Some of these elaborations turn out to be irrelevant and confus-ing. For example, the authors draw attention to the important fact that celestial deities as well as other motifs symbolizing the world beyond death are lacking, and interpret the tomb correctly as being intended to represent the deceased person’s dwelling in this world (p. 20). But then the authors go on to discuss in great length those absent deities, and speculate whom or what the soul of the dead is going to meet beyond “Heaven’s Gate” without questioning whether the depiction of the door can indeed be identified as Heaven’s Gate. Likewise, the authors examine the Confucian system during the whole Han dynasty, even though this tomb —unlike some others from the period—contains no motifs what-soever reflecting Confucian ways of thinking and moral standards such as righteous assassins, loyal ministers, filial sons, or exemplary women. In fact, I would argue that the lack of those most popular Eastern Han pictorial themes, the absence of depictions of the other-world, and the presence of the unusual banquet scene (see below) in the rear chamber may be the key to comprehending the concept of values in Bin Wang’s own world. The scene in the rear chamber is explained by the authors as either an audience (p. 18) or a banquet (p. 22). As plain as it may seem, it deserves a closer look. Never before have so many women been seen depicted in a tomb of this period. Their noble sitting posture is like that of the sovereign Queen Mother of the West. Women have an unusual presence in the murals of this tomb. Most of them are not servants; nor are they exemplary women from the Confucian ideology. Instead, they are wives of local dignitaries. This indicates a local community totally different from that of Confucian scholars in the East that constitutes the frame of reference of the authors of this book (p. 23). The content of the murals of the rear part of the antechamber and the rear cham-ber can be perceived as very formal through the presence of the local dignitaries. This, in turn, emphasizes the high status of Bin Wang in his own community. In spite of what is maintained by the authors, nothing private— such as hunting or music playing for leisure— can be observed. The titles of the visitors all reveal a connection to military staff, thus indicating that Bin Wang as well might have been a local warlord. This raises the next question of why the authors did not take ad-vantage of the numerous inscriptions, which unfortunately are only

482 book review

partially mentioned, to identify the motifs and to explore the histori-cal background of the tomb occupant and of this area. Such a great number of inscriptions is unusual and should not be taken for granted. Other interesting or unusual motifs, such as the warriors (lishi) at the entrance and the so-called “exorcists” in the antechamber, may need to be reappraised. In sum, we are missing a closer iconographic and art historic research of the many local particularities which could give us a close-up picture of this tomb and this area during the later Eastern Han. The German text does not mention why the tomb was attributed to Bin Wang, although this is explained vaguely in the Chinese translation. The Chinese translation is probably based on an earlier German ver-sion, because some parts of the Chinese text deviate from the printed German text. Despite these flaws, the book is an important source of material for further research on this tomb and for pictorial arts in the Xi’an region during Eastern Han times, and it sets a standard for the documenta-tion and publication of murals. We are looking forward to a detailed excavation report of the tomb with which to complement this work.

Reviewed by Shing Müller (Ludwig-Maximillian-Universität, München)


Radiocarbon Chronology of the Stone Age of Northeast Asia, Yaroslav V. Kuzmin (ed.). Vladivostok: Pacific Institute of Geography, 1998, 127 pp.

This little booklet by an international team of eight researchers led by Kuzmin (V. Alkin, A. Ono, H. Sato, T. Sasaki, S. Matsumoto, K. Orimo, and S. Ito) is a compilation of radiocarbon determinations on Palaeolithic and Neolithic assemblages in Northeast Asia. A short (5 page) Introduction is followed by Chapter 1, consisting of lists of radiocarbon determinations from four regions of Northeast Asia (pp. 10–80), and Chapter 2, entitled “Radiocarbon chronology of the Stone Age of southern Russian Far East and adjacent areas in Northeast Asia” (pp. 81–105). The booklet ends with a short (one page) “Conclusion,” “References,” and very brief “Summaries” in English, Chinese, Korean and Japanese. To a reader without Russian language skills (like this reviewer), the fully accessible part of this useful and interesting publica-tion is Chapter 1, where the tables of radiocarbon determinations are presented in English (for Korea and Japan), and in both Russian and English (for Russian Far East and China). A bilingual text that appears with the tables explains the chronological framework for each of the four regions according to which the dates are presented. Chapter 2 consists of two sections, whose titles are shown in English as “Radio-carbon chronology of the Paleolithic and Neolithic of southern Russian Far East” and “Correlation of chronological boundaries of the Stone Age of Russian Far East, Korea, Japan, and Northeast China.” For the content of the discussions, however, the reader without Russian language skills should turn to many articles Kuzmin and his colleagues have pub-lished in recent years, including those in Journal of World Prehistory 12(1) [1998]:1–53; Journal of East Asian Archaeology 3 (3–4) [2002]: 227–254; Radiocarbon 44(2) [2002]: 503–530; and The Review of Archaeology 24(2) [2003] (entire issue). The area covered under “Southern Russian Far East” is the Primorye, Sakhalin Island, the southern part of the Kurile Islands, and the Amur River basin east of about 128°E. The table, listing 168 dates (17 Palaeo-lithic and 151 Neolithic) from 54 sites from this region, seems quite comprehensive and up-to-date. “Northeast China” here refers to Hei-longiang 黑龍江, Jilin 吉林, and Liaoning 遼寧 provinces and Inner Mongolia, east of about 110°E, not including the North China Plain proper. The 144 dates from 56 sites are judiciously chosen, excluding those with dubious associations, yet the list is quite comprehensive.

484 book review

Particularly impressive are the 54 (8 Palaeolithic and 46 Neolithic) dates from Inner Mongolia. The Northeast China list provides far more detailed information regarding cultural association of the samples than those for other regions. For example, the 4605±125 date (ZK-1354) for Niuheliang 牛河梁 is identified as a determination made on charcoal that was obtained from Tomb #8 of Stone Mound 1 in excavation area II and was associated with the late period of Hongshan 紅山 culture. This is particularly useful, because precise information about sample provenience is often missing from Chinese publications, such as Zhong-guo kaoguxue zhong tanshisi niandai shujuji 1965–1981 中國考古學中碳十四年代數据集 1965–1981 (Radiocarbon Dating in Chinese Archaeol-ogy 1965–1981), edited by Shehui Kexueyuan Kaogu Yanjiusuo 中國社會科學院考古研究所 (Institute of Archaeology, Chinese Academy of Social Sciences; Beijing: Wenwu Chubanshe 文物出版社 (Cultural Relics Publishing House, 1983). For Korea, only the dates from the Republic of Korea are listed, and almost all are taken from Sarah Nelson’s The Archaeology of Korea (Cambridge, 1993). Comparison with more recent lists (e.g., Kidong Bae (배기동 ) 裴基同 in Radiocarbon vol. 44, 2002; C. Norton in Journal of Human Evolution vol. 38, 2000) suggests, however, that not a great deal has changed in the last few years as far as the Palaeolithic period is concerned. Spelling and other minor errors of proper nouns could be confusing to those not familiar with the literature. For example, the Amsadong (암사동) 岩寺洞 and Turubong 두루봉 sites are misspelled as “Amsadon” and “Tuburong,” respectively, and Pokee Sohn’s (손보기 ) 孫寶基 works are cited by his personal name “Pokee.” By far the largest number of determinations is from Japan, even though the Satsunan 薩南 and Ryūkyū 琉球 Islands at the southern end of the Archipelago are excluded from consideration. The long list is divided into two parts: 208 dates from 65 Palaeolithic sites and hundreds more from 147 Jomon 縄文 sites. The compiled list is most welcome, since it is very difficult to keep track of these dates, because radiocarbon determinations, along with the results of other “scientific” analyses, are usually published as appendices to the numerous excavation reports that appear every year. The reports are published, often in very small quantities, by various prefectural and municipal organizations re-sponsible for local cultural resources management, and quickly become unavailable. As useful as the lists are, there are a few minor errors that have crept into the Tables. For example, the lab code numbers for the two Palaeolithic dates from the famous Fukui Cave 福井洞穴 of Kyushu 九州 (site No. 62, on Table 4, p. 57) are GaK-951 and GaK-952. Fur-thermore, the GaK-951 is for the specimen from Cultural Layer VII, not IV, according to the original investigators of the site (Y. Kamaki 鎌

book review 485

木義昌 and C. Serizawa 芹沢長介, in Cave Sites of Japan 日本の洞穴遺跡, 1978, Heibonsha, p. 265). For the reasons given above, it is not possible, unfortunately, for this reviewer to determine how prevalent such errors are by systematically comparing the information given in the Tables against each of the original reports. The long Jomon list is based entirely on the paper C. Keally and Y. Muto 武藤康弘 prepared for the 10-volume Studies of Jomon Culture 縄文文化の研究 (Yuzankaku, 1982). The internal organization of the date lists, by geographical and chronological subdivision, is also exactly as it was in the Japanese publication. For those unfamiliar with Japanese geography, this may prove difficult to use. In order to find the date for the Stone Circle site of Oyu 大湯, for example, one has to know that it is situated in “Northern Tohoku Region 東北北部” before locating the appropriate entry. The list, in any event, was quite complete at the time, but is now sadly out-of-date. In recent years, Japanese archaeologists and physicists have been busily working to produce high-precision chronol-ogy of Jomon Period. In fact, S. Tsuji 辻誠一郎 and T. Nakamura 中村俊夫, two scientists who are active in this field, wrote an article with precisely that title in a special issue of Quaternary Research 第四紀研究, devoted to the question of calibration of radiocarbon dates (vol. 40, No. 6, 2001). It is hoped that Japanese researchers would embark once again on the daunting task of compiling the radiocarbon determinations that have accumulated since Keally and Muto did over 20 years ago. Until that is done, this booklet is a very useful source of reference. Kuzmin and his associates are to be congratulated for their initiative and accomplishments.

Reviewed by Fumiko Ikawa-Smith (McGill University, Montreal, Canada)

Art and Archaeology of the Sichuan Basin

Documents

Transcript of Art and Archaeology of the Sichuan Basin