Alice Yao1 & Jiang Zhilong2 Surface collection, exposed sections and the use of irrigation wells and channels enabled the authors to map the settlement pattern of the elusive Dian kingdom before it became a subsidiary of the Han empire. The pattern showed that the Dian were already hierarchical, with settlements of different sizes and a political centre in which ritual bronzes featured. The empire redrew the landscape, with settlement migrating away from the wetlands into the hills where it could oversee the routes of communication into Southeast Asia. Keywords: south China, Yunnan province, Yizhou, Dian, Bronze Age, first millennium BC, landscape survey, settlement hierarchy Introduction During the first millennium BC, regions south of the Yangzi River simultaneously witnessed the onset of sociopolitical complexity. Foremost among these ‘indigenous’ political formations include the Yelang, Dian, Dongson and Lingnan (Higham 1996; Allard 1998; Lee 2001; Yao 2010), which all shared a sophisticated bronze tradition characterised by the production of large ritual vessels, ornate socketed weapons and agricultural implements. Whether these societies emerged from internally driven processes over the long term or sudden, punctuated events (e.g. migrations and conquests) has been much debated. Recent research in the core of Dongson polity of northern Vietnam suggests important political transformations emerging in the fourth century BC, long before the social impact of Han 1 2 Department of Anthropology, University of Toronto Mississauga, Room 396 Terrence Donnelly Health Science Complex, 3359 Mississauga Road North, Mississauga, ON L5L 1C6, Canada (alice.yao@utoronto.ca) Yunnan Institute of Cultural Relics and Archaeology, 15-1, Chunmingli, Chunyuanxiaoqu, Kunming, Yunnan 650118, China C Antiquity Publications Ltd.  ANTIQUITY 86 (2012): 353–367 http://antiquity.ac.uk/ant/086/ant0860353.htm 353 Research Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China imperial conquest in 111 BC and as described in Chinese texts (Kim 2010 et al.). These findings are enhanced by discoveries directly north of the Dongson, where Bronze Age polities appeared between the eighth and fifth centuries BC in eastern Yunnan. It is certain that social inequality and political consolidation predate imperial conquest, and investigations now seek to identify the internal dynamics and timing behind these complex formations. That factional competition, warfare or prestige goods trade (e.g. metal ores and exotic ornaments) stimulated increasing social difference is beyond doubt and highlight the importance of this region. An important corollary is just how complex were these Bronze Age polities, raising the proverbial problem of scale (Wright 1977; Feinman 1998). Opinions differ, with terms such as chiefly society, paramount chiefdom and protostate being variably applied (Tong 1991; Kim et al. 2010). Bronze Age cemeteries in eastern Yunnan imply entrenched social hierarchies with the division of individuals into social ranks and the separation of elite and commoner burial spaces (Lee 2001; Yao 2005). This display of social hierarchy in the mortuary context renders an impression of political and economic control, yet leaves its exact territorial extent largely unanswered. To adequately address sociopolitical scale requires a corresponding understanding of spatial scale. However, in the absence of detailed settlement patterns, archaeological investigation cannot effectively tackle this issue. This paper presents results from the South Dian Survey, the first full coverage survey in the heartland of ‘Dian’ polity located in central Yunnan province, PRC (Figure 1). Using a combined programme of surface survey, coring and section cleaning, investigations recovered the remnants of a Bronze Age settlement system associated with the ‘Dian’ polity and important shifts coincident with Han regional incorporation in 109 BC. These findings reveal the concentration of sites and locations on rich lacustrine wetlands and a settlement hierarchy dominated by the large central mound of Hebosuo that was likely to have marked the political centre of a Bronze Age polity in the Dian basin. Historical and geographical background Ancient Chinese texts refer to the Dian kingdom as one of the most powerful entities in south-western China. Han records state that the Dian king could command 30 000 men, and the discovery in 1955 at the cemetery of Shizhaishan of a gold seal bearing the king’s title verified the historical existence of the polity. However, before 2008 archaeological investigations had found only two Bronze Age settlements in the Dian basin, while subsequent finds at cemetery sites only contributed to the enigma (Watson 1993; Zhang 2001). If there was indeed a formidable polity, where was its centre? Is it, as archaeologists have long suspected, buried beneath the Han dependency of Yizhou (Figure 1) (Allard 2005)? More important, the spectacular nature of Dian bronze manufacture suggests political control over craft and agricultural production (Murowchick 2002). In the absence of occupational sites, however, we can assess neither its developmental sequence nor its organisational complexity. Chinese archaeologists divide the regional chronology into two broad periods: the Bronze Age or ‘Dian culture’ period (eighth–sixth centuries BC to 109 BC) and post-conquest period spanning the Han to Jin dynasties (109 BC to AD 420). Despite the use of iron C  Antiquity Publications Ltd. 354 Research Alice Yao & Jiang Zhilong Figure 1. The Dian basin and the south-east survey region (in red). Bronze Age cemetery sites, which have been excavated and documented, are shown as triangles. The only Bronze Age settlements previously reported in the lake basin were Wangjiadun and Longmen near Kunming City, and a shell midden on top of Shizhaishan (map generated from ArcScene 10.0 using SRTM90 World Map). implements from the Central Plains by the third century BC, Chinese chronologies do not take into account an Iron Age period since bronze continues to be the main metallurgical tradition. The survey area and method The Dian basin (Figure 1) is highly important to the investigation of Bronze Age societies in southern China and Southeast Asia. Historical sources and archaeological discoveries at Shizhaishan indicate that this area was the probable centre of Dian activity. Other discoveries indicate that the seat of the subsequent Han imperial government was nearby. South-western China is characterised by rugged topography consisting of dissected mountain chains with C  355 Antiquity Publications Ltd. Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China dispersed flat and arable land limited to the lakeside plains. In addition to providing the most agriculturally productive land, the Dian basin occupies a pivotal position along trade and communication routes connecting China, Southeast Asia and India that emerged in the fourth–third centuries BC (Yang 2005). Given the economic and interregional importance of the Dian basin, the detection of cultural activity in this region can contribute significantly to the study of social complexity emerging along China’s southern borders. Our survey focused on the south-eastern part of the Dian basin because this area contains the widest expanse of flat and arable land as well as being a candidate for the ‘heartland’ of the Dian polity. In addition to the discovery of burials on Shizhaishan, excavations in the 1950s exposed cultural layers along the edge of the hill, revealing a prehistoric occupation site (Yunnan Provincial Museum 1963). It could be assumed that the agricultural productivity of this region would have attracted prehistoric and Han inhabitants alike. The survey region encompasses three distinctive ecological zones (foothills, alluvium and lakeshore) totalling 64km2 . The plain consists primarily of lacustrine wetlands and alluvium irrigated by the lake and the main south-east to north-west draining river channel. Surrounding hills provide a natural physical barrier secluding this area from adjacent parts of the basin. Due to the density of modern greenhouse structures in the lowland areas, surface visibility was poor. Fortunately, the extensive distribution of shallow irrigation wells associated with greenhouses significantly improved site visibility by allowing direct observation of sub-surface deposits. The survey methods are summarised in Figure 2. The survey aimed at full coverage, combining surface and sub-surface observations with coring. Following surface methods successfully applied elsewhere in China, crew members were spaced at 50–75m and collected diagnostic sherds of the Bronze Age and Han periods (Figure 3; Underhill et al. 1998; Linduff et al. 2004; Liu et al. 2004). When sherds and shells in well debris indicated a site, coring was undertaken to define the extent and character of cultural layers, and samples were taken from these layers for fine wet screening and flotation. Three sherds served as a minimum requirement for site designation. Most of the occupation sites consisted of mounds formed from accumulated layers of gastropod shells. These gastropods are predominantly represented by the species Margarya melanoides, which were until recently an abundant and a heavily exploited lacustrine resource that has now largely disappeared in the Lake Dian area (Ellen Strong pers. comm. 2007). On the floodplain, the recovery of wet preserved wood posts from these shell matrices, which often extended more than 3m below the surface, suggest that architectural structures were constructed into the shell layers. Some Bronze Age occupation sites were also discovered on top of numerous limestone and flysh (cretaceous sandstone) hillocks, where exposed stratified shell layers indicated prehistoric occupation on the outcrops. The use of coring enhanced the assessment of individual site size, with important implications for understanding the scale of Bronze Age settlement organisation. Results: the Bronze Age landscape Whereas only two settlements had been previously documented in the basin, the survey discovered 44 discrete sites: 25 of which were occupied during the Bronze Age, 9 spanned C  Antiquity Publications Ltd. 356 Research Alice Yao & Jiang Zhilong Figure 2. Survey method: (right) in addition to surface collection, observations were made from coring (a), exposed sections (b) and irrigation wells (c) which revealed shell midden deposits, Bronze Age ceramics and wooden posts (d), allowing the survey team to deduce with greater accuracy site locations and boundaries. The map (left) shows the resulting sites plotted on Quickbird imagery (2008) using ArcGIS 10. Table 1. Sites documented in south Dian basin. Bronze Age site Bronze–Han period site Han site Bronze Age cemetery Han tomb 25 9 5 2 3 the Bronze Age to Han/Jin period, and 5 yielded strictly Han ceramics and tiles (Figure 4; Table 1). In addition to the two known Bronze Age cemetery sites, three Han period mounded tombs were discovered along the spurs in the south-eastern valley exiting the plain. The 34 sites in the study area occupied during the Bronze Age divide into three discrete size classes. The village of Hebosuo proved to have been constructed on top of a prehistoric shell mound and was the largest Bronze Age settlement, measuring 31ha (Figure 4a, n◦ 1). The proximity of Hebosuo to the necropolis of Shizhaishan (n◦ 2) endorses the political prominence of this site. Sites in the next size category range from 4–10ha and include the shell mound sites located on top of Shizhaishan and at Gucheng (n◦ 4); the remaining sites were estimated to be less than 2ha in size (Figure 5a). The survey documented 16 Bronze Age sites within 6km of Hebosuo, distributed in an arc-like pattern along the lakeshore. Based on the distribution of surface finds and a survey of approximately 156 irrigation wells in this area, most of the 16 sites, with the exception C  357 Antiquity Publications Ltd. Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China Figure 3. Dian ceramic assemblage. Ceramics diagnostic of Bronze Age period occupations (upper) include (a) oxidised rough ware bowls and (b) limestone and sand tempered vessels. Rough ware bowls include an incised and plain ware form, which occurred in higher frequency than other ware types. Limestone and sandy wares include jars with everted rims and collared necks. Both types of ceramic wares occur infrequently in Bronze Age cemeteries. Ceramics diagnostic of the Han period (lower) include stamped limestone-tempered body sherds and grey fabric impressed roof tiles with a silty fabric. of Shizhaishan, were 1ha or smaller in size. These sites possibly comprised the satellite communities of Hebosuo. The more extensive occupation on top of Shizhaishan possibly indicates a symbolic locale given its association with the necropolis and position above the surrounding plain. A carbonised seed taken from a stratified cultural layer on Shizhaishan was dated to 779–488 cal BC (BA091158) (Table 2). Besides the complex of sites around Hebosuo, a secondary locus of cultural activity is to the north-east at Gucheng (Figure 4a, n◦ 4). While considerably smaller than Hebosuo at C  Antiquity Publications Ltd. 358 Research Alice Yao & Jiang Zhilong Figure 4. Results of survey project. Location of Bronze Age (a) and Han period settlements (b). The sites are primarily concentrated near Hebosuo (1), Shizhaishan (2), Jincheng (3), Gucheng (4) and Haibaoshan (5). Cemetery sites are outlined in red. C  359 Antiquity Publications Ltd. Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China Figure 5. Site size distributions in the south Dian basin: Bronze Age period sites (a) divided into three primary classes with Hebosuo being the largest at 31ha; Han period sites (b) divided into three primary classes with Jincheng being the largest at 63ha. 4ha, the mound at Gucheng is the largest site in the north-east extent of the survey area. Dated deposits from stratified levels suggest occupations as early as the eighth century BC (Table 2). Like Hebosuo, Gucheng is located in close proximity to a Bronze Age cemetery that was similarly positioned on top of an adjacent hillock, 500m north of the mounded site. If Gucheng represents the primary occupation site, the smaller settlements in its vicinity may represent the “supra-local community” in the north-eastern corner of the survey region (Drennan & Dai 2010: 458). As was the case at Hebosuo, these sites also contained stratified shell deposits containing a considerable amount of cultural material. These sites were also smaller than 1ha and most were located along the terraced slope and at the base of the Bronze Age cemetery. Other associations between occupation density and site size suggest that the Bronze Age settlement system was differentiated and organised according to socioeconomic factors. The kernel density map generated from counts of Bronze Age C  Antiquity Publications Ltd. 360 Alice Yao & Jiang Zhilong Site & context Gucheng layer 2 Gucheng layer 3 Shizhaishan layer 6 Yangfutou burial M22 Yangfutou burial M19 Yangfutou burial M297 Yangfutou burial M147 Yangfutou M96 Tianzimiao burial Tianzimiao burial Sample type Lab sample no. Calibrated 2σ (95.4%) Seed Seed Seed Charcoal Charcoal Charcoal Charcoal Charcoal Charcoal Charcoal BA091155 BA091156 BA091158 ZK3110 ZK3111 BK200209 BK200210 BK200208 PV328 PV329 810–530 BC 900–760 BC 779–488 BC 756–394 BC 787–415 BC 361 BC–AD 22 366 BC–AD 5 AD 53–382 390 BC–AD 1 40 BC–AD 350 Figure 6. Density map of Bronze Age ceramics in the south Dian basin. Sherd distributions are displayed as smoothed contours based on kernel density estimation (KDE) set to 100m radius in ArcGIS 10. While surface collections could not be made according to defined units due to extensive greenhouse and urban cover, investigators did record the full spatial extent of sherd scatter and made controlled collections at 249 locations. The greatest concentration of Bronze Age sherds is located in the Hebosuo-Shizhaishan settlement area. Within this areal extent, the highest density (green peak) corresponds spatially with the site of Hebosuo. Drop-offs in sherd frequencies indicate possible separation between sites within this supra-local community. ceramics show patterned artefact concentrations with the highest density peak occurring at Hebosuo and a secondary peak at Gucheng (Figure 6). Botanical remains from a range of Bronze Age sites substantiate the presence of a wetland habitat long thought to have characterised this prehistoric landscape. Sediment C  361 Antiquity Publications Ltd. Research Table 2. Calibrated radiocarbon dates from prehistoric settlements and cemeteries in the Dian region. Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China cores from sites in the Hebosuo-Shizhaishan and Gucheng areas contained an abundance of wet preserved uncharred materials dominated by annual plants related to wet shallow microenvironments. The two most prevalent aquatic plants found in both locations are Potamageton and Najas. Other aquatic plants include Characeae, Cyperaceae, Myriophyllum, Rumax and Juncus. The prevalence of these plants suggests that the depiction of pile dwellings in Dian iconography and the discovery of wooden posts from survey are consistent with habitation constructions suited to marshy environments. Furthermore, the extensive nature and depth of Bronze Age shell mounds suggest that inhabitants possibly constructed these features to raise occupation levels above marshlands. Cultigens from Shizhaishan and Gucheng cultural levels reveal that in addition to rice (Oryza sativa), millet (Seteria italica) and wheat (Triticum sp.) were also cultivated. A longstanding question surrounding the Dian polity concerns the nature of craft production. Based on the opulent display of bronze implements and jade and agate ornaments found in elite burials, the detection of craft production in the HebosuoShizhaishan core would provide some resolution on the scale of Dian economic control over ‘prestige goods’. Flotation samples from three sites in the Hebosuo-Shizhaishan area produced slag and copper debris, indicating metallurgical activity being undertaken in the centre of Dian settlement system. In addition to the presence of metalworking in this core area, worked stone materials resembling turquoise were found in cultural levels at two separate sites. More recently, the excavation of a modern sewage trench uncovered agate debris and a broken piece of reworked Central Plains jade disk at Hebosuo (City Museum of Kunming, pers. comm. 2010). These findings—however small given sample size constraints—show that these objects were not imports but probably manufactured in the Dian region. Changes in the Han period Following Han incorporation, the Dian basin witnessed the establishment of the imperial capital of Yizhou in present-day Jincheng city (Figure 4b, n◦ 3). Survey observations indicate that this was an extensive settlement occupying 63ha. A large community of this order was certainly unprecedented in this region. Archaeological support for the textual identification of Jincheng city with Yizhou includes the discovery of a dense cluster of Han fabric impressed tile along the north-east corner and Han tomb bricks in the south-western edge of the city. In addition to the imperial centre of Jincheng, the survey documented 13 other sites with Han material divided into two size classes. Nine sites were smaller than 2ha while four sites ranged in size from 4–10ha (Figure 5b). The occupations at Shizhaishan and Hebosuo continued to be the most extensive, ranging in size from 8–10ha. By the Han period, occupation as a whole also shifted towards the eastern foothills and the Han governmental centre founded at Jincheng in 109 BC. Sites located on outcrops to the west of Jincheng produced Han bricks and the stamped sherds typical of Han ceramic assemblages. These sites are smaller than 2ha in size. Their position overlooks the two river valleys that provide entry into the Dian region from the Yuxi basin, a rich lacustrine basin providing the main communication route connecting highland Yunnan with the Dongson polity in northern Vietnam. Perhaps the population was concentrated at the sizable site of C  Antiquity Publications Ltd. 362 Yizhou during the Han period, leading to a less extensive occupation of the lowlands. The shift indicates a reorganisation of the landscape, perhaps guided either by imperial concerns over security and defence or an interest in expanding the availability of cultivable land on the basin floor through draining the wetlands. Discussion The Bronze Age activity in Dian proves to have been hierarchical and, contrary to expectations, did not coincide with the imperial settlement system. Beyond the detection of a settlement hierarchy for the two periods, survey results also indicate locational trends that contrast the Bronze Age and Han period settlement systems. While many Bronze Age sites continued to be occupied into the Han period, the frequency of Han sherds was very low. The number of settlements in the Hebosuo-Shizhaishan satellite area decreased to three primary locations with sites appearing to have also declined in size. The Gucheng area also experienced a similar decline in site frequencies but the main mound continued to be occupied. But questions concerning the scale of the Dian polity and related issue of its genesis still remain. How big was this Bronze Age political formation? Although most scholars assume that the polity was limited to the lake basin, integration across this expansive area (450km2 ) would have necessitated differentiated structures of administration. In the north-eastern extent of the basin, the cemetery of Yangfutou (Figure 1) equals Shizhaishan in its opulence and is similarly associated with a complex of mounded settlements (YPICRA et al. 2005: 862). Equally important is control over the western extent of the basin as major copper and tin ores can be directly accessed through a single river valley pass that links the basin with the Yangzi River watershed. The limited availability of arable lands along the western shore, due to the craggy landscape and rugged topography, presumably could not have supported large communities without staple provisions. Whether the ‘Dian’ polity emerged from cyclical or punctuated trajectories of development requires much finer chronological resolution (e.g. ceramic sequences and radiometric dates). If, as indicated by findings in Vietnam, Bronze Age complexity emerged from internal competition and warfare beginning in the fourth century BC, then an understanding of temporal trends in the Dian basin could similarly detect fluctuations in the settlement order. If the revised dating scheme for the Dian basin is considered, the earliest evidence of social stratification is located in the north part of basin where elaborate graves with bronze drums begin to be interred at the cemetery of Yangfutou by the fifth century BC and possibly even earlier (Table 2), preceding burial activity at Shizhaishan by more than a century (YPICRA et al. 2005: 715; Chiou-Peng 2011). Thus, when Yangfutou became a consecrated elite cemetery site, Shizhaishan was an occupation site. The differential timing behind the erection of these elite cemeteries may reflect political fluctuations associated with escalating regional factionalism in the Dian basin. Radiometric dates from Gucheng and Shizhaishan indicate Bronze Age cultural activity in the south-eastern extent of the basin as early as the eighth century BC (Table 2). Given the depth of cultural layers at Gucheng and Hebosuo, the lengthy occupation sequence at these sites could reveal diachronic trends in settlement organisation documenting cycles of fluctuation (Wright 1984; Anderson 1994). C  363 Antiquity Publications Ltd. Research Alice Yao & Jiang Zhilong Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China In particular, understanding the timing and growth of the settlement cluster around Hebosuo and Yangfutou could potentially clarify the developmental sequence that is necessary for evaluating models of political formation. Shifts in the settlement at the dawn of the Han conquest naturally raise questions about the nature of imperial incorporation and the ‘waning’ of the Bronze Age. Simply put, what happened to the local peoples? The Shiji notes that the Dian population was peacefully incorporated into the Han state while the decline in mortuary activity at Bronze Age cemeteries across the basin suggests that indigenous traditions came under pressure (Watson 1993; YPICRA et al. 2005: 834). The loss of symbolic funerary spaces in the landscape would signal shifts in the political order. Later documentary sources record successive uprisings across numerous prefectures in Yizhou from 105 BC–AD 176 with some skirmishes culminating in the murder of the imperial governor (Allard 2005: 246). In this light, the concentration of Han settlements near the provincial centre and towards more defendable locations in the foothills may reflect a level of uncertainty. In any event, these textual accounts suggest that Han imperialism was far from a succinct episode of conquest and incorporation but rather a more protracted and uneven process. Another possible explanation proposed for the post-conquest settlement shift is the introduction of new measures to increase agricultural production. In AD 19, the Han governor created 121km2 of arable land in the prefecture by introducing terracing and irrigation (Sun & Xiong 1983: 249). Palynological and geomorphological data from the Dian and neighbouring Erhai basins, however, revealed limited anthropogenic disturbance coincident with Han conquest. The disappearance of valley floor and hillside taxa along with increased erosion became marked in both regions around 1500 BP, well after Han incorporation (Sun et al. 1986; Dearing 2008). While cores from the southern extent of the basin would help corroborate the timing and extent of these anthropogenic changes, current observations do not support a sudden reorganisation of the population in the basin coincident with agricultural intensification. Without an understanding of agricultural productivity during the Bronze Age, imperial cultivation strategies should not be assumed to be more intensive in scope and interpreted as the basis for population relocation. To pit imperial ambitions against indigenous interests focuses on the event of conquest, projecting a view of cultural irreconcilability. Survey data provide compelling evidence of cultural continuity in the lake basin, which, when examined with ‘indigenous’ tombs in the post-conquest period, could suggest lengthy periods of intercultural engagement. The stamped pottery sherds recovered during survey showed an undeniable resemblance to the surface treatment typical of Han ceramic assemblages (Figure 3). However, unlike their grey ware Han counterparts, the fabric of these stamped pottery sherds consists of a fine sandy ware with crushed limestone temper that is identical to their Bronze Age precursors. The concurrent production and use of both grey ware and local variants in the Dian basin is more reflective of a transitional rather than displacement process. Paralleling these trends is the incorporation of Han materials into local burial assemblages. At both Shizhaishan and Yangfutou, graves from the conquest period include Han C  Antiquity Publications Ltd. 364 bronze serving ware, iron implements, Chinese coins and even ceramic house models typical of Han burial customs, while the Bronze Age tradition of pit graves with an easterly orientation are retained (YPICRA et al. 2005: 834). For Han material culture to be manifest in the tomb repertoire of conquered peoples for two centuries following conquest suggests that Bronze Age peoples continued to reside in the basin or ostensibly engaged with Han both economically and socially so as to become familiar with imperial customs. Conclusion Without an understanding of the organisation of people and resources over space and time, debates surrounding the processes leading to political formations and the nature of complexity cannot move forward. Here regional survey work provided the foundational support for centralised political leadership that has until now been fixed to one ‘Dian’ site and one object—the king’s seal. The discovery of a complex of occupation sites concentrated around the formidable settlement of Hebosuo—in addition to the presence of metal- and stoneworking there—point to the economic and political influence exercised by this centre over the region. If habitation in this wetland area was achieved by raising sites above the plain through the construction of shell mounds and levées, then Hebosuo would have drawn on considerable corporate labour from its satellite communities. Perhaps even more striking is how Shizhaishan now figures in this landscape as a symbolic anchor in the ‘Dian’ heartland. By the same token, an understanding of scale at the regional level has implications for conceptualising interregional levels of interaction, thus providing a solid basis for reintegrating external factors and taking core-periphery relationships into consideration. If these regions were part of an expansive trade network, the configuration of these polities in this network depended on size, scale, economic base, geographic position and access to key raw materials (Stein 1999; Galaty & Parkinson 2010). Recent metallurgical analysis indicates that copper ores from Yunnan were used in bronze casting by powerful states in central China as early as the late second millennium BC and even circulated to the Dongson polity (Jin et al. 1999; Chiou-Peng 2011). Determining which regional polities controlled access to specific ores and transport nodes in this dissected landscape can elucidate the structure of interregional interaction. In this light, Bronze Age societies in this part of the world can provide a promising cross-regional test case to further stimulate ways to model core-periphery and peer-polity relations. Acknowledgements The South Dian survey is an international project involving the collaboration of the Yunnan Provincial Institute of Cultural Relics Archaeology, Jinning Cultural Relics Office, and the Museum of Anthropology, University of Michigan. Special thanks go to Henry Wright, without whose guidance and feedback, this preliminary season of fieldwork would have been impossible. The Oriental Archaeology Research Center (OARC) at Shandong University and the Radiocarbon Lab at Beijing University assisted with the botanical and radiometric analyses. Dr Chen Xuexiang (OARC) graciously undertook the paleobotanical analyses. We are grateful to the funding received from the National Science Foundation (BCS 0835153), Henry Luce Foundation, and American Council C  365 Antiquity Publications Ltd. Research Alice Yao & Jiang Zhilong Rediscovering the settlement system of the ‘Dian’ kingdom, in Bronze Age southern China of Learned Societies. In addition, the lead author extends her thanks to the Institute for the Study of the Ancient World, NYU, for providing time off to direct the field project. References LINDUFF, K., R. DRENNAN & G. SHELACH. 2004. Early complex societies in NE China: the Chifeng International Collaborative Archaeological Research Project. Journal of Field Archaeology 29: 45–73. LIU, LI, CHEN XINGCAN, LEE YUN KUEN, H. WRIGHT & A. ROSEN. 2004. Settlement patterns and development of social complexity in the Yiluo region, north China. Journal of Field Archaeology 29: 75–100. MUROWCHICK, R. 2002. The political and ritual significance of bronze production and use in ancient China. Journal of East Asian Archaeology 3(1–2): 133–92. STEIN, G. 1999. Rethinking world-systems: diasporas, colonies, and interaction in Uruk Mesopotamia. Tucson (AZ): University of Arizona Press. SUN, TAICHU & XIONG XI. 1983. Chronology, in J. Rawson (ed.) The Chinese bronzes of Yunnan: 243–52. London: Sidgwick and Jackson and Cultural Relics Publishing House. SUN, XIANGJUN, WU YUSHU, QIAO YULOU & D. WALKER. 1986. Late Pleistocene and Holocene vegetation history at Kunming, Yunnan province, southwest China. Journal of Biogeography 13: 441–76. TONG, ENZHENG. 1991. Chiefdoms in southwest China: the Dian culture as an example. High Bronze Age of southeast China and south China. Hong Kong: Institute of Chinese Studies, Chinese University of Hong Kong. UNDERHILL, A., G.M. FEINMAN, L. NICHOLAS, G. BENNETT, FENGSHU CAI, HAIGENG YU, FENGSHI LUAN & HUI FANG. 1998. Systematic regional survey in southeast Shandong province. Journal of Field Archaeology 25: 453–74. WATSON, B. 1993. Sima Qian: records of the Grand Historian of China (translation). New York: Columbia University Press; Hong Kong: University of Hong Kong. WRIGHT, H.T. 1977. Recent research on the origin of the state. Annual Review of Anthropology 6: 379–97. – 1984. Pre-state political formations, in T.K. Earle (ed.) The evolution of complex societies: the Harry Hojier Lecture for 1982: 41–77. Malibu (CA): Undena Press. YANG, BIN. 2005. Horses, silver and cowries: Yunnan in global perspective. Journal of World History 15(3): 281–322. ALLARD, F. 1998. Stirrings at the periphery: history, archaeology, and the study of the Dian. International Journal of Historical Archaeology 2(4): 321–41. – 2005. Frontiers and boundaries: the Han empire from its southern periphery, in M. Stark (ed.) Archaeology of Asia: 233–54. Malden (MA): Blackwell. ANDERSON, D. 1994. The Savannah River chiefdoms: political change in the late prehistoric Southeast. Tuscaloosa (AL): University of Alabama Press. CHIOU-PENG, TZEHUEY. 2011. The technical history of early Asian kettle drums, in E. Bunker & D. Latchford (ed.) Khmer and related bronzes: 17–26. Chicago (IL): Art Media Resources. DEARING, J. 2008. Landscape change and resiliance theory: a palaeoenvironmental assessment from Yunnan, SW China. The Holocene 18: 117–27. DRENNAN, R. & DAI XIANGMING. 2010. Chiefdoms and states in the Yuncheng basin and the Chifeng region: a comparative analysis of settlement systems in north China. Journal of Anthropological Archaeology 29: 455–68. FEINMAN, G. 1998. Scale and social organization: perspectives on the archaic state, in G. Feinman & J. Marcus (ed.) Archaic states: 95–134. Sante Fe (NM): SAR Press. GALATY, M. & W. PARKINSON (ed.). 2010. Archaic state interaction: the eastern Mediterranean in the Bronze Age. Santa Fe (NM): SAR Press. HIGHAM, C. 1996. The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press. JIN ZHENGYAO, MABUCHI HISAO, TOM CHASE, CHEN DE’AN, MIWA KAROKU, HIRAO YOSHIMITSU & ZHAO DIANZENG. 1999. Guanghan Sanxingdui jisigeng qingtongqi de huaxue zucheng he qian tongweisu bizhi yanjiu [Research on the lead isotope rations and chemical composition of bronzes from the Sanxingdui ritual pits in Guanghan], in Sichuan Sheng Wenwu Kaogu Yanjisuo (ed.) Sanxingdui jisikeng [The ritual pits of Sanxingdui]: 490–99. Beijing: Wenwu chubanshe. KIM, NAM C., LAI VAN TOI & TRINH HOANG HIEP. 2010. Co Loa: an investigation of Vietnam’s ancient capital. Antiquity 84: 1011–1027. LEE, YUN KUEN. 2001. Status, symbol, and meaning in the Dian culture. Journal of East Asian Archaeology 3(1–2): 103–131. C  Antiquity Publications Ltd. 366 YAO, A. 2005. Scratching beneath iconographic and textual clues: a reconsideration of the social hierarchy in the Dian culture of southwestern China. Journal of Anthropological Archaeology 24: 378–405. – 2010. Recent developments in the archaeology of southwestern China. Journal of Archaeological Research 18: 203–239. Yunnan Institute of Cultural Relics and Archaeology, City Museum of Kunming & Guandu District Museum (YPICRA et al.). 2005. Kunming Yangfutou mudi [Kunming Yangfutou cemetery]. Beijing: Science Press. Yunnan Provincial Museum. 1963. Fourth season of excavations of the ancient cemetery at Shizhaishan, Jinning, Yunnan. Kaogu 9: 480–85. ZHANG, ZENGQI. 2001. Dian wenhua [Dian culture]. Beijing: Wenwu chuban she. Received: 7 April 2011; Accepted: 22 May 2011; Revised: 15 August 2011 C  367 Antiquity Publications Ltd. Research Alice Yao & Jiang Zhilong