Introduction
For many decades Chinese archaeologists have been interested in the ancient peoples who lived along and beyond the Great Wall in a region called the beifang, or Northern Corridor.  They were described by ancient Chinese historians as living in ways different from the agricultural lifeways of dynastic China of the second and first millennia BCE.  These written reports and excavated cemeteries across the region suggest that agro-pastoralism and/or nomadic lifeways emerged in the beifang, but and why that happened has not been studied using systematic archeological field methods.  That change in lifeway has been marked by archaeologists by noting the difference in material culture during the period.  The earlier materials were characterized by specialized painted pottery, by use of carved jades for ritual implements and by utilitarian and decorative items made of metal.  The later period tombs have yielded greatly increased amounts of portable metal artifacts, especially weapons and personal ornaments.
Until very recently little contextual information was collected beyond that of the immediate find-spot.  The change in lifeway and thereby in material culture cannot be understood with the currently available data recovered almost exclusively from burial remains.  Different kinds of information, for example on regional social and political organization, on ecology, climate, trade, and other types of interaction are needed.  Our goal is to put into regional context the relevant archaeological materials already accumulated from burials and caches and, by using regional survey methods followed by test excavations, to systematically investigate the emergence of pastoralism in our targeted area near Chifeng, Inner Mongolia.  This sort of analysis has not been attempted before in the Northern Corridor.

The Project
The Chifeng Collaborative Archaeological Research Project (CICARP) is a cooperative project among the University of Pittsburgh (USA), The Hebrew University (Israel), Jilin University (PRC) and the Inner Mongolia Institute of Archaeological Research (PRC).  It was initiated with a seminar and  planning season held in the summer of 1998 followed by two full field seasons in 1999 and 2000.  Initial base funding was provided by a Chiang Ching-kuo Foundation grant in 1997, and has just been extended by them to support the years 2001-2003.  So far we have surveyed an area of about 550 square km and have located over 836 sites, most not known previously.  The region projected for survey will require an estimate of three more field seasons.
Our main objective is to address the process of socio-political change in northeast China.  The issues we address are pertinent to a better understanding of regional history--China and her northern neighbors--but are also relevant to the general understanding of socio-political processes worldwide.  Such issues include the identification of different trajectories of social complexity and different types of complex societies, sources of political power and legitimization, interregional interaction and its effect on local developments as well as the causes and effects of economic change.

Theoretical and Methodological Implications
Our research, therefore, focuses on development and change in complex societies that is regional in scope (Drennan 1992; Earle 1987).  The theoretical underpinnings of the project, the questions it addresses and the methods used, all include explicit regional components.  One implication of this approach is that much of the data collected during the fieldwork is meaningful only within its regional context.
Intensive full-coverage survey of a substantial area is the method used to address our research goals.  Such a survey, complemented by test excavations and designed to record data suitable for GIS analysis, will enable us to address synchronic and diachronic processes at the regional level.  Such a method is well suited to the archaeological data already available from the Chifeng region.  Ceramic sequences, arranged on the basis of extensive stratigraphic excavations carried out by the Chinese archaeologists working in this area, are the basic tool for dating sites discovered by our survey, and the regional data we are generating can best supplement previous research in the area which tended to focused on artifacts and individual sites.  Cooperative work has broadened the scope of our understanding by bringing together different theoretical perspectives and methodological rigor.

The Research Area
Chifeng is located in the southeastern part of the Inner Mongolia Autonomous Region, bordering with Liaoning Province.  Its climate is cold and dry in the winter, dry and very windy in the spring.  Summer is relatively short and hot, with temperatures dropping rapidly during the autumn.  The average temperatures in the area are between ?11 and -15o C during January and 20 to 23o C in July (Kong et al. 1991).  Most of the precipitation occurs during the summer season with 70% of the average falling between June and August (Wang 1990; Kong et al. 1991).
Chifeng is ideally located between the Yellow River basin and the steppe to the north and northwest which the Chinese tradition associates with a pastoral nomadic way of life.  This location thus allows for the study of interaction among peoples occupying different ecological zones.
The data collected by previous research in the area is limited as to the type of questions that can be addressed and the analyses that can be performed.  As the main goals of archaeological research in China have been to illustrate historical events and to discover sites and artifacts known from written sources, this results in a unilinear model of social evolution (An 1989; Chen 1989) and the common view that Yellow River basin was the source of all social, political, economic and cultural change (Du 1991; Sun 1987; Sun 1989; Tong 1986; Zheng 1988).  Although other surveys have been conducted in the area (Xu 1986; Zhongguo 1998a) they were seen mainly as a way of finding sites where excavations could take place and therefore analysis of patterns of distribution based on the data collected in these surveys is seriously crippled.
The reconstruction of changing political centralization and socioeconomic differentiation is as our starting pint.  Our objectives center on gathering evidence to evaluate it, to extend its chronological scope, and to assess the forces that produced the sequence of change.  Our strategy for achieving these objectives focuses heavily on expanding the size of the area subjected to systematic archaeological survey, the area surveyed to date is still small, compared to the scale of the social, political, and economic phenomena of the region.  These issues can be discussed most easily in the framework of the sequence of chronological phases.

Xinglongwa and Zhaobaogou ; (c. 6000-4500 BC)
Early Neolithic cultures in the Chifeng area--the Xinglongwa and Zhaobaogou cultures--were identified during the 1980�s (Xu 1989).  To date, very few habitation and burial sites of these periods have been excavated (Aohan 1991; Neimenggu 1994; Zhongguo 1985; Zhongguo 1987; Zhongguo 1998b; Zhu 1997).  These excavations yielded the earliest evidence in the area known so far for permanent habitation in villages, ceramic production, and the domestication and cultivation of plants and animals.  The small number of sites so far identified from these cultures and the limited area of excavations permits only tentative hypotheses about the way of life in this important period of transition to agriculture.
Results from work to date indicate that systematic survey locates larger quantities of these scarce remains than survey methodologies that have been applied previously in the region.  So far we have recovered Xinglongwa ceramics in 16 different collection units in 14 spatially discrete small sites and Zhaobaogou ceramics in 28 collection units in 22 spatially discreet small sites.  The largest of these sites covers less than 3 ha.  These are presumably the remains of small economically self-sufficient egalitarian villages.  Further survey will be required to determine whether this occupation is relatively evenly spread throughout the region or tends to concentrate in certain sectors based on resource distribution or other factors.  Further site excavation and comparative study of features and artifact assemblages at the household scale will be required to delineate possible patterns of social or economic differentiation, productive specialization, etc.  Faunal and botanical remains reported to date demonstrate the presence of domesticated species, but quantitative study will be required to evaluate their relative importance as well as that of wild plants and animals and thus the completeness of reliance on agriculture and herding.

Hongshan (c. 4500-3000 BC) and Xiaoheyan (c. 3000-2200 BC)
The Hongshan culture has attracted much attention.  However, in spite of--and maybe even because of--the zeal with which these ritualistic expressions are being studied, the more mundane aspects of the Hongshan culture have been largely neglected.  Only meager information on domestic sites of the Hongshan has been published and almost nothing at all is known about internal site structure and about settlement patterns.
The ritual structures, the elaborate jade sculptures, and the concentration of offerings in the burials of a few individuals that have been reported for Hongshan sites are highly suggestive of the kinds of societies to which the term "chiefdom" or some similar label is often applied (e.g. Drennan 1995; Earle 1987, 1997).  In the survey area to date, Hongshan ceramics were encountered in 203 different collection units, forming some 129 spatially discrete sites, suggesting quite substantial population growth for this period.  Altogether 1527 Hongshan sherds were recovered, compared to only 309 for Xinlongwa and Zhaobaogou combined, reinforcing the impression of considerable increase in population.  (These latter two together account for about the same amount of time as Hongshan does.)  Some sites continue to be quite tiny, as before, but one was as large as 11 ha, and there is a tendency for larger sites to group together, quite possibly representing communities within which inter-settlement interaction had taken on greater importance.  The social, political, ritual, and/or economic aspects of such interaction remain to be explored, but it may represent the initial webs of regional relationships upon which complex societies are based.  Completion of survey over a larger area will be required to delineate these patterns comprehensively, to see how they emerge from those of the previous period and give way, in turn, to later ones. Identifiable Xiaoheyan ceramics are relatively rare, 174 sherds found in 31 collections in 28 discreet sites, suggesting that these ceramics transitional between Hongshan and Lower Xiajiadian may not represent an entirely distinct period in their own right.  Further stratigraphic chronological work will be required to clear up this question.

Lower Xiajiadian (c. 2200-1600 BC)
Population growth appears to be stronger as the sequence continues into Lower Xiajiadian times, since these ceramics were recovered from 608 collection units in nearly 299 discreet sites in the survey area to date.  Altogether, Lower Xiajiadian sherds numbered 7289 (compared to 1527 for Hongshan, which represents more than twice as much time as Lower Xiajiadian does).  The largest single Lower Xiajiadian site covers 23 ha, and several groupings of large sites, probably represent sociologically meaningful communities better than individual sites do.  These site groupings tend to be very densely occupied areas on the low bluffs overlooking the prime agricultural lands of the valley floors with a very few sites in the bottomland.  They are at least loosely associated with hilltop fortified sites well back in the upland zone between river valleys.  These hilltop-fortified sites tend to be much smaller in area and to have extremely low densities of surface ceramics despite ideal surface collecting conditions.  Their cultural deposits tend to be shallow, but the architectural remains (often quite visible on the surface) include the stone foundations of circular structures, large terraces, and massive walls and gateways, representing a major investment in construction.  Such features are altogether lacking from earlier sites.
Expanding the survey area is particularly important for studying Lower Xiajiadian political organization.  Previous surveys in the Chifeng area indicate the presence of numerous site clusters similar to those recorded in the survey area to date (Xu 1986; Zhongguo 1998a; Zhu 1994), possibly representing a series of different Lower Xiajiadian polities.  Surveying a larger area will make it possible to identify their relative geographic positions, to delineate the patterns of association between densely occupied residential zones and more sparsely occupied hilltop sites with monumental defensive works, and to investigate the distribution of the much smaller settlements that make up the majority of Lower Xiajiadian sites.  The distribution of these smaller settlements in the areas between such polities, for example, in the form of buffer zones whose presence or absence might be related to varying degrees of competition between polities.  Comparing these polities can also help us understand their political and economic makeup.

Typologies of Lower Xiajiadian sites have been suggested previously (Shelach 1999; Zhongguo 1998a), but they have not entirely resolved the question of functional differences between small and large sites, between very densely occupied and very sparsely occupied sites, between sites with and without defensive works.  Rigorous comparison of sites in terms of such variables as location, accessible resources, artifact assemblage composition, and types of architectural features would help us to understand possible differences in site functions and the nature of relationships between sites.  To add to the data recorded on regional survey, we began to use a Total Station to produce detailed maps of a selected sample of Lower Xiajiadian sites and to make intensive artifact collections in association with individual architectural features.  This additional data will help us to reconstruct patterns of intra-site organization.

Upper Xiajiadian (c. 1000-600 BC)
The Upper Xiajiadian represents a transformation no less dramatic that the Lower Xiajiadian.  The three-tiered central place settlement hierarchy proposed by Shelach (1999) all but disappears, and Upper Xiajiadian sites display no evidence of massive defensive works.  On the other hand, a local bronze industry flourished during this time and burial practices became very elaborate.  In comparison to the richest Lower Xiajiadian graves which contained no more than twenty ceramic vessels, a few bone and stone artifacts, bones of sacrificed animals (usually pigs and dogs), and an occasional small bronze artifact (Aohan 1976; Guo 1995b; Liu and Xu 1989;Zhongguo 1996), the richest Upper Xiajiadian grave contained over a thousand artifacts including more than ninety large bronze artifacts along with imported Chinese bronze vessels (Liaoning 1973; Xiang and Li 1995).  In the 1999/2000 survey area, the number of collection units with Upper Xiajiadian ceramics drops to 499 in some 245 spatially discrete sites.  The total number of sherds is 6470, a decrease from the 7288 Lower Xiajiadian sherds recovered, even allowing for the fact that Lower Xiajiadian represents about 50% more time than the Upper Xiajiadian does.  A tendency for settlements to group together continues, but the groupings are often smaller and more spread out than before.  Surprisingly,  the total area of Upper Xiajiadian sites of 7.57 square km, including large clusters often found in the valleys between mountains, exceeds that of the Lower Xiajiadian total of 7.03.  What these clusters indicate is not yet understood, but these areas are obvious candidates for test excavation.
This evidence of population increase, of overall loosening of settlement groupings, and less clear settlement hierarchy seems to contradict the much more developed socioeconomic hierarchy suggested by the burial evidence�at least in terms of what one might expect on the basis of general schemes advanced to account for the development of complex societies in global comparative perspective.  Rich Upper Xiajiadian graves may be evidence of a society in which political power is personally associated with the leaders or upper social strata, rather than with the communal functions such as leaders performed (Shelach 1999).
Surveying a larger area will make it possible to test the association of Upper Xiajiadian sites with locations most suitable to a mixed agro-pastoral economy.  The distribution of Upper Xiajiadian graves, very frequently visible on the surface, shows a considerable variability in different sectors of the study area, being much more common in the region surveyed by Shelach (1999) than in the 1999/2000 survey area.  Data on the sizes of the individual graves and their number can be collected in regional survey and help reconstruct patterns of stratification and variation between possible political units.
 
 

Warring States�Han (c. 500 B.C ? A.D 200);  Period from 200 AD through the Liao Dynasty (907-1122)
This period represents the first real intrusion of the Chinese states into the area.  It has been recently suggested that this close interaction between the Chinese and non-dynastic polities helped shape the history of China during this and later periods (Di Cosmo 1999). The increasing geographical scale of polities and of meaningful political interactions make our regional approach less suitable for the study of socio-political process during this period. Nevertheless, since settlement pattern studies of this period have never been done, our work can contribute to the understanding of the local effects of the large-scale processes.
Warring States�Han material was found in 306 collection units in about 189 spatially discreet sites in the 1999/2000 survey area.  Altogether, there were 2869 sherds, compared to 6470 sherds for the preceding Upper Xiajiadian (which is about 100 years shorter).  The largest single site was nearly 12 ha.  The overall population level changed and the population was dispersed into a larger number of smaller sites.  There is less apparent tendency for these sites to form groupings than in earlier periods.  Sherds from the periods following through the Liao numbered 5097 from 325 sites in 529 collections were also collected.  These data will be compared to those from current population censuses as well as settlement locations in the region.  Land-use patterns will also be of interest as we continue to be able to document earlier patterns.

Future Field Seasons:
In order to further understand and explain the changes in social patterns in this area, we will continue the full-coverage survey as well as begin test excavation and refined mapping of selected sites.  All records are made in both Chinese and English, and the data is stored on laptop computers and kept in Hohhot (Institute of Archaeology, (Zhang Wenping); the Department of Archaeology, Jilin University (Teng Mingyu); The Hebrew University (Shelach); and the University of Pittsburgh (Drennan and Linduff).  Computer technology used to store and manipulate all the data as well as to create these maps was introduced by the US team and was quickly absorbed.
The new methods of study were eagerly tested by the Chinese members of the team while their expertise in regional topography, ceramic analysis and previous excavation results was provided has been fundamental to the progress of the project to date.  Preliminary results of the survey have being drafted and will be published in both Chinese and English next year.