Ancient East Asia
Special Report
On the Chronology of the Three Dynasties
by LIU Li
The current discussions on EAAN (an email list for early East Asian archaeology and history) on the Three Dynasties Chronology in China were triggered by two reports published in The New York Times (In China, Ancient History Kindles Modern Doubts, November 10, 2000, by Erik Eckholm and Far Eastern Economic Review (China: Nationalism Digging into the Future, July 20, 2000, by Bruce Gilley These reports represent suspicions held by many Western scholars regarding a controversial project launched by the Chinese government a few years ago. The EAAN discussions have been primarily focused on some important issues concerning the study of Chinese civilization and the influence of modern ideology, especially nationalism, in archaeology. Although no scientist in the world could be value free from his or her social environment, wrong judgement about the nature of a particular civilization may be made simply because of insufficient knowledge about it.
It is commonly held by Chinese scholars that the Erlitou culture is equivalent to the Xia dynasty, and that the Xia and Shang were state-level societies which constituted large centralized political systems throughout their reigns. In contrast, many western scholars have been sceptical about the reality of the Xia, and believe that a pristine state did not develop until the late Shang. The reason that the interpretations of the political organization in early Chinese civilization were so contradictory may relate to the data used in these studies. Most of the discussions by Western scholars are based on old archaeological information and on ancient texts including oracle-bone inscriptions. Oracle-bone inscriptions provide invaluable data on the late Shang, but little information on earlier periods. Most texts were written hundreds, if not thousands, of years after the Erlitou and early Shang periods. These interpretations, therefore, were inevitably modified by oral transmission and the court historians who made the records.
It is clear that up-dated archaeological information with a refined chronological scale is crucial to the study of processes of early state formation in China. The discussion below is derived from recent research on this subject conducted by Xingcan Chen and myself, which will be published in the near future. I post some relevant discussions and the abstracts of these two papers here in order to promote more dialogues among scholars on these issues.
Question: Was Erlitou a state-society?
The earliest unbanization in Bronze Age China emerged during the Erlitou culture, centered at the type-site Erlitou in the Yi-Luo River basin, western Henan. Some 38 calibrated radiocarbon dates derived from Erlitou sites in Henan (Institute of Archaeology 1991) indicate that this culture may have flourished during a period between 1900 and 1550 BC (Figure 1). Erlitou (400 ha in area) is the largest among all its contemporary sites in China, and sites containing the Erlitou material assemblages have been found over a very broad region mainly including Henan, southern Shanxi, Eastern Shaanxi, and Hubei (Figures 2, 12). Because of its spatial and temporal correlation with the Xia dynasty, which is traditional believed to be the first dynasty in Chinese history, most Chinese archaeologists and historians now agree that the Erlitou culture represents the material remains of the Xia dynasty, and that the Erlitou site may have been a capital city of the late Xia (e.g., Chang 1986: 307-16; Du et al. 1999; Zhao 1987; Zou 1990).1 This argument is adopted in the present paper.
Several marked changes in settlement pattern and material culture took place when the Longshan culture in the late Neolithic period developed into the Erlitou culture, as follows:
(1) Settlement hierarchy in western and central Henan changed from two- or three-tiered systems in the late Longshan to a four-tier system in the Erlitou culture (Figure 3) and rank-size curves changed from convex to strong primate (Figure 4). All these measurements point to the development of centralized political and economic control.
(2) The site number decreased dramatically, dropping from some 700 Longshan sites to less than 200 Erlitou sites in an area including southern Shanxi and Henan; in the same period, settlement nucleation took place, as the largest site size increased from 75 ha to 400 ha. These changes may suggest nucleation of population accompanying with urbanism, rather than simply population decline.
(3) The political structure changed from the coexistence of multiple competing polities to one in which a single large center dominated smaller centers and villages over a very broad region.
(4) Ceramic styles changed from diversity (six variants of the Longshan culture) to relatively uniformity (two variants of the Erlitou culture: Erlitou and Dongxiafeng). This may suggest an increase in craft specialisation and standardisation of production relating to the development of political centralization (cf. Longacre 1999; Rice 1981; Rice 1996).
(5) In addition to jade objects, which were the traditional means for representing high social ranking in the Neolithic, bronze objects, mainly weapons and ritual vessels, became status symbols for the first time. Bronze production, which was largely based in capital cities, may have become a state-controlled industry, ensuring a state monopoly of bronze products for military and ritual use (Liu 1996, in press).
(6) Long-distance exchange of precious goods developed to a new level - cowrie shells (Monetaian moneta or Monetaria annulus) with possible origins from the Indian Ocean region (Peng and Zhu 1999) were added to the inventory of grave goods in elite burials. Some artifacts and decorative motifs with characteristics of Central Asian cultures also appeared at Erlitou (Fitzgerald-Huber 1995).
These changes unquestionably signify the emergence of a state-level social organization characterized by a centralized political and economic control in its core area, as well as expanded cultural contacts and influence over a broad region.
1 There are divided opinions among scholars on the nature of the Erlitou culture. The majority of archaeologists in China agree that the Erlitou culture represents a state-level society, and that the Erlitou site represents a capital city of the Xia or Shang dynasties, although they disagree about to which capital city named in textual record the Erlitou site corresponds.
Definition of a state
Much of the disagreement on this Erlitou-state issue may be affected by employment of different definitions of state/civilization. I list three definitions here. The first one by V. Gordon Childe published in the 1950s was influential for decades, but has been questioned by many archaeologists. This checking list is simply not applicable to all civilizations as a universal law. I prefer the two definitions given by Wright and Marcus and Feinman.
Gordon Childe: a list of primary and secondary characteristics. The primary characteristics include: 1) cities - dense, nucleated demographic concentrations; 2) full-time labor specialization; 3) state organization, based on territorial residence rather than kin connections; 4) class stratification - the presence of a privileged ruling stratum; and 5) the concentration of surplus. The secondary characteristics are: 1) monumental public works; 2) long-distance exchange; 3) writing; 4) arithmetic, geometry, and astronomy; and 5) highly developed, standardized artwork.
Henry Wright (Recent research on the origin of the state. Ann. Rev. Anthropol. 1977:383): 'a state can be recognized as a cultural development with a centralized decision-making process which is both externally specialized with regard to the local processes which it regulates, and internally specialized in that the central process is divisible into separate activities which can be performed in different places at different times. States, like chiefdoms, usually exist in networks. Among simple states these networks seem to be regulated by competition and alliance, as was briefly noted for chiefdoms. A difference is that developing state networks are periodically centralized into a single political unit incorporating most previously existing polities.
Joyce Marcus and Gary Feinman (in Archaic States 1998: 4) 'archaic states were societies with (minimally) two class-endogamous strata (a professional ruling class and a commoner class) and a government that was both highly centralized and internally specialized.'
It is true that Chinese archaeologists have not found a writing system at Erlitou. But the question is how decisive writing is to define a state? It is the function of writing that matters. Writing would be important if it is related to power acquisition in religious or economic domains. If we could learn the sociopolitical system of Erlitou based on other archaeological data, the presence/absence of writing would not be a crucial factor. Furthermore, the Erlitou people may have indeed had writing, but written on perishable materials, which have not been preserved in the archaeological record.
The Erlitou culture (at least a part of it) represents the first social entity that meets the two criteria given by Marcus and Feinman. Erlitou was characterized by a centralized and internally specialized government, indicated by a great concentration of palatial foundations and various craft production workshops in an urban center (Erlitou), and rapid cultural expansion over a large region. The second criterion, marked social stratification, is indicated by burial differentiation.
Some people may argue that some of these characteristics can be observed in the Longshan culture. But they did not occur altogether. Erlitou manifested a qualitative social change from the Longshan culture, while the Erligang represented a quantitative change from the Erlitou.
If we use different criteria to examine the political organization of the Erlitou polity, it is likely that we should agree to disagree on this issue.
Question: Was Erlitou related to the Xia?
The glory of this first Chinese city came to an end in the Erlitou Phase IV, when a walled enclosure (80 ha.; known as the Yanshi small city, xiaocheng) was constructed 6 km. northeast of Erlitou. This change is generally believed to relate to the political transition from the Xia to the Shang dynasty, and the enclosure at Yanshi may have been the earliest Shang capital built after the conquest (Du et. al 1999; (Gao et al. 1998); (Zhao 1989), or a special garrison established by the Shang for preventing a rebirth of the Erlitou state.
According to the traditional chronology, it has been believed that the Xia dynasty existed in a period between 2100 and 1600 BC, which would have included a part of the late Longshan and most of the Erlitou culture (Phases I-III). Based on archaeological information, however, the Longshan culture may be characterized as a chiefdom-level social organization, with many small competing polities coexisting in the Yellow River valley. It was not until Phase II of the Erlitou culture (c. 1800 BC) that a state-level society emerged (Liu 1996). The rise of the early state seems to have been a relatively rapid event, with many changes in social organization taking place during a short period around the Erlitou Phase II. This interpretation of Erlitou state formation does not match the traditionally believed image of Chinese early history, in which the Xia was a great dynastic power from the outset. The Xia dynasty, if it existed, may have started as a chiefdom society during its early period, and then developed into a territorial state only during the late part of its duration.
Chang, Kwang-chih 1986 Archaeology of Ancient China. Yale University Press, New Haven.
Du, Jinpeng, Xuerong Wang, and Liangren Zhang 1999 Shilun Yanshi Shangcheng xiaocheng de jige wenti (On several issues related to the small city inside the Yanshi Shang city). Kaogu 2:35-40.
Fitzgerald-Huber, Louisa 1995 Qijia and Erlitou: The question of contacts with distant cultures. Early China 20:17-68.
Gao, Wei, Xizhang Yang, Wei Wang, and Jinpeng Du 1998 Yanshi Shangcheng yu Xia Shang wenhua fenjie (The Yanshi Shang city and the demarcation between the Xia and Shang cultures). Kaogu 10:66-79. Institute of Archaeology, CASS
1991 Zhongguo kaoguxue zhong tan shisi niandai shujuji. Wenwu Press, Beijing.
Liu, Li 1996 Settlement patterns, chiefdom variability, and the development of early states in north China. Journal of Anthropological Archaeology 15:237-288.
Liu, Li (in press) The development and decline of social complexity in China: Some
social and environmental factors. The Indo-Pacific Prehistory: The Melaka Papers. Bulletin of the Indo-Pacific Prehistory Association 19 3.
Longacre, W. A. 1999 Standardization and specialization: What's the link. In Pottery and people: A dynamic interaction, edited by J. Skibo and G. Feinman, pp. 44-58. The University of Utah Press, Salt Lake City.
Marcus, Joyce & Gary Feinman 1998 Introduction. In Archaic states, Gary Feinman and Joyce Marcus, eds. pp.1-13. Santa Fe: School of American Research Press.
Peng, Ke, and Yanshi Zhu 1999 Zhongguo gudai suoyong haibei laiyuan xintan (New inquiry in the sources of cowries in ancient China). Kaoguxue jikan 12:119-147.
Zhao, Zhiquan 1987 Lun Erlitou yizhi wei Xiadai wanqi duyi (On the Erlitou site as a capital of the late-Xia). Huaxia Kaogu 2:196-204, 217.
Zhao, Zhiquan 1989 Erlitou wenhua yu Erligang wenhua (The Erlitou culture and Erligang culture). In Qingzhu Subingqi kaogu wushiwu nian lunwenji, edited by The Editorial Group for Qingzhu Subingqi kaogu wushiwu nian lunwenji, pp. 273-279. Wenwu Press., Beijing.
Zou, Heng 1990 Xia wenhua yantao de huigu yu zhanwang (The study of the Xia culture in retrospect and prospect). Zhongyuan Wenwu 2:1-12.

Archaeology and Nationalism in China
Li Liu and Xingcan Cheng
(from the article contributed to Encyclopedia of the History of Archaeology, ed. Tim Murray. Santa Barbara: ABC-CLIO, 2001)
The development of Chinese archaeology has been intertwined with the ever-changing political environment during the twentieth century. Archaeologists have worked extremely hard to overcome all kinds of economic, social, and political difficulties during turbulent eras, and have made extraordinary contributions to the field. Our understanding of ancient China has been markedly improved because of these archaeological achievements. In many cases archaeology has been driven by the changing concept of nationalism and used as an instrument to support, rather than to evaluate, particular theoretical themes or political agenda. In other situations, it has provided independent data for creating new paradigms, which changed traditional perspectives of Chinese national history. State promoted nationalism has indeed played an important role in shaping the discipline.
Many individual archaeologists, however, have spontaneously exercised nationalist ideology in their research. For them, the building of national history implies dignity and pride as human beings. China is certainly not the only nation in which archaeology is relevant and meaningful primarily in the context of the connection between modern cultural and national identities and ancient indigenous traditions (e.g., Kohl & Fawcett 1995). Therefore, in spite of growing influences from Western ideology and technology during the past seventy years, which in many cases are positive, the general objective for the mainstream of Chinese archaeology seems not to have changed - the discipline has been committed to the reconstruction of national history. This mission will probably continue to be carried out for many years to come. It is possible, however, that more various research approaches will emerge in future. While some archaeologists are continuously pursuing regional historical issues, others may become engaged in theory building and cross-cultural comparative studies, which will develop the discipline with a more international outlook.
Kohl, Philip and Clare Fawcett (eds.) 1995 Nationalism, politics, and the practice of archaeology. Cambridge: Cambridge University Press.  

State Formation in Early China
LIU Li ( La Trobe University, Melbourne, Australia)
CHEN Xingcan (Institute of Archaeology, Chinese Academy of Social Sciences, Beijing, PRC)
This paper investigates the political organization of the earliest state-level societies in northern China, the Erlitou and Erligang cultures, from the perspectives of settlement pattern and political economy. While a number of small political entities coexisted after the Longshan period, only the Erlitou developed into the first centralized territorial state in the Central Plains. Its expansion to the surrounding regions was motivated by the procurement of key resources, especially metal and salt. The political-economic organization remained largely unchanged in the Erligang culture (the early Shang), as the state further expanded its territory to all directions where rich natural resources were available. The core-peripheral interactions may be characterized as the tributary relationships.
On the one hand, the regional centers in the peripheries were established to obtain key resources, and to transport them as tribute to the capitals. On the other hand, the capitals monopolized the production of bronze ritual vessels, which were used as ritual paraphernalia and status symbols. These bronze ritual objects, in turn, may have been redistributed as gifts by the royal court to regional elites to establish their local sociopolitical hierarchies. This centralized system, however, appears to have been broken down during the middle Shang period when multiple large urban centers emerged simultaneously in different regions, suggesting the collapse of the territorial state. This process may have led to the new development of regional bronze cultures beyond the core area during the late Shang period.
Cities and Towns: the Control of Natural Resources in Early States, China
Paper contributed to Population and Preindustrial Cities: A Cross-Cultural Perspective.
Ed. Glenn Storey. University of Alabama Press, 2001
LIU Li ( La Trobe University, Melbourne, Australia)
CHEN Xingcan (Institute of Archaeology, Chinese Academy of Social Sciences, Beijing, PRC)
This paper examines the developmental processes and functions of five of the earliest cities and towns dating to the Erlitou and Erligang periods: Erlitou and Yanshi Shangcheng in the core area, and Dongxiafeng, Yuanqu Shangcheng, and Panlongcheng in the peripheries. Regional settlement patterns, demographic variations, locations of copper and salt resources in the peripheries, the internal structure of the cities, as well as artefacts from some of these urban sites are analysed. It is argued that the development and decline of some early "urban centers" were closely related to changing strategies of early state rulers in procurement of vital natural resources. The effort made by elites to control and transport these resources may have generated major affects toward shaping unique patterns of urban expansion in early Chinese civilisation.

Cities and Towns: the Control of Natural Resources in the Xia and Shang periods
Cheng: Xia Shang shiqi dui ziran ziyuan de kongzhi wenti
Li Liu and Xingcan Chen, Dongnan wenhua (Southeastern Culture) 2000. 3:45-60.
This is a preliminary research on the procurement of vital resources by the early states (Erlitou and Erligang) published in Chinese. This article presents our initial hypotheses relating to the processes of state formation in China from a perspective of political economy. In order to test these hypotheses we conducted investigations at several sites in southern Shanxi in Sept. 1999, including Dongxiafeng, Yuanqu Shang town, an ancient copper mine, and the Hedong salt lake. As a result, we have gained more confidence on our approaches. We have also changed some of the opinions presented in this paper, particularly on the two types of pottery jars found in Dongxiafeng. We originally hypothesized that they were used for transporting salt, but based on our recent research this conclusion is problematic. The results of this fieldwork and further research will be published in Population and Preindustrial Cities: A Cross-Cultural Perspective. Ed. Glenn Storey. University of Alabama Press. 2001.
The Chinese text of this paper is available in pdf form (which requires Acrobat Reader 4.0) at:
The file is 5.4 megabytes in size.
LIU Li, Dr
La Trobe University, Melbourne, Australia)
20 November 2000
Please note that my surname in Chinese is Liu. In some of the publications listed above my names have been reversed according to American usage.