Cutting a Gordian Knot: the Bronze Age of Southeast Asia: origins, timing and impact.

Higham, Charles ; Higham, Thomas ; Kijngam, Amphan 等

[ILLUSTRATION OMITTED]

Introduction

There are two irreconcilable models for the timing and origin of the Bronze Age in Southeast Asia. One is based on the site of Ban Chiang, the other on Ban Non Wat, both located in the Khorat Plateau of north-east Thailand. This paper cuts the Gordian Knot by redating the former site, and then reviews the cultural implications.

Ban Chiang

The first model is founded on the 1974-75 excavations at the site of Ban Chiang (Figures 1 & 2). As members of the research team, we (CH and AK) opened two areas, one of which began with Neolithic occupation and graves before the transition into the Bronze Age, while the 1975 season encountered a Bronze Age cemetery in the basal layers (Table 1, Figure 2). White and Hamilton's model (2009) is anchored in a set of radiocarbon determinations from ceramic vessels placed as mortuary offerings (White 1997, 2008). The seven dates were selected from 20 determinations, which formed the basis of a pretreatment experiment on the direct dating of pottery (Glusker & White 1997). White and Hamilton (2009) state that these place the transition to the Bronze Age at about 2000 BC. They further identify, as supporting evidence, the chaff-based determinations from Non Nok Tha potsherds (Higham 1996: 191) and a bronze bar from Ban Mai Chaimongkol, a site with no radiometric dates.

How did this complex technology reach the site of Ban Chiang as early as 2000 BC? It cannot have involved Chinese states because their bronze technology was established too late to have any bearing on origins at Ban Chiang. So they turn to the Seima-Turbino metallurgical 'transcultural phenomenon' (Chernykh 1992:218). Sites of this grouping centre on the Urals, with related sites reaching west into Finland and east to the Altai (Gorodtsov 1916). While dating remains fugitive (Mei 2003), one radiocarbon date from the site of Satyga suggests that the bronzes were being cast in the vicinity of 2000 BC (Hanks et al. 2007). The Seima cemetery has yielded socketed spearheads and decorated axes, knives decorated with cast animals on the hilts and rings. From Turbino come tanged chisels, knives decorated with cast-on sheep and horses on the pommel, and a range of spearheads often with a ring on the hilt to facilitate hafting. Some aspects of this repertoire are similar to the socketed copper-base tools cast in Southeast Asia, but many others are missing.

[FIGURE 1 OMITTED]

White and Hamilton then follow Mei (2003) in identifying Seima-Turbino influence in the metallurgical tradition of the Qijia culture in Gansu. This culture has for long been seen as seminal in the transmission of a metallurgical tradition into China (Higham 1996). It is at this point in their argument that White and Hamilton propose that bronze founders moved south to Thailand, a distance of at least 2500km along the eastern foothills of the Himalayas, to the site of Ban Chiang. This great march evidently had no impact on any known Neolithic community in China, nor at any other site in Southeast Asia. Again, Haimenkou, Yinsuodao and Yeshishan, the earliest Bronze Age sites on the line of this march in Yunnan, are 1000 years later than the proposed date of 2000 BC for Ban Chiang (Chiou-Peng 2009).

[FIGURE 2 OMITTED]

Ban Non Wat

The second model is based on Ban Non Wat, located 280km south of Ban Chiang (Figure 3). The cultural sequence here began with late hunter-gatherers, and then proceeded through two Neolithic, six Bronze Age and three Iron Age phases (Higham & Kijngam 2009). The chronological framework for this sequence is based on 76 radiocarbon determinations modelled using a Bayesian approach with OxCal. 4.0 (Bronk Ramsey 2009; Higham & Higham 2009). Neolithic settlement began in the mid seventeeth century BC and lasted in the vicinity of 150 years, while the Neolithic 1 burials date from about 1460 cal. BC, and lasted for about 50 years. This was followed by Neolithic 2 graves, which are dated to 1260-1055 cal BC. The transition to the Early Bronze Age settlement took place between 1050 and 995 cal. BC. The date of c. 1000 BC for the beginning of the Bronze Age makes it feasible to identify the origins of metallurgical knowledge through the exchange of goods, ideas and the necessary skills with long-established bronze-casting societies in China.

[FIGURE 3 OMITTED]

Resolving the dilemma

It is important to resolve this dilemma. The model based on Ban Chiang supposes that one or more individuals journeyed 2500km from Gansu, possibly further still from the Altai, to reach Ban Chiang by 2000 BC, at which point there was a standstill of40 generations before the knowledge spread further to incorporate Ban Non Wat and other southern Chinese or Southeast Asian sites. This model also indicates that the knowledge of bronze-casting had little if any impact on the social life of the inhabitants of Ban Chiang over the ensuing 1500 years of the Bronze Age, for the graves at Ban Chiang 1975 contained very modest mortuary offerings. The alternative model allows for a progressive expansion of the knowledge of metallurgy over the course of the second millennium BC, in a context where the exchange of bronzes and jades moved south and cowrie shells and turtle carapaces found their way north to the states of the Yangtze Valley and Central Plains. This accompanied a veritable outburst of social display with the initial Bronze Age at Ban Non Wat.

White and Hamilton's model is raised entirely on the foundation of six radiocarbon determinations derived from rice chaff temper in mortuary vessels and one from rice phytoliths. The Gordian Knot of the above dilemma can be quickly severed if these determinations prove unreliable and are replaced by a new series of determinations based on a proven source material.

Although the AMS dating of organic material in pottery once seemed a straightforward proposition, experience has shown it to be particularly challenging. There are several possible sources of carbon within the fabric of a pot, and occasionally as surface residues. Often the sherd is of low carbon, which means that a larger than usual amount of material is required for a direct date. The clay matrix might contribute old carbon, while the smoke and soot generated during firing can be absorbed into the temper. If firing uses old wood, results may be too early (de Atley 1980; Bonsall et al. 2002).

Hedges et al. (1992) have concluded that this technique is often unreliable due to the incorporation of carbon from the clay. Manning et al. (2011) have demonstrated in Mali that direct pot fabric radiocarbon dates are 300-400 years older than OSL dates on the same sherd and at odds with dates of other more reliable organic remains. Kuzmin et al. (2001) have shown that a stepped combustion approach to dating pottery sometimes produces quite different ages with different temperatures, because temper-derived carbon tends to be preferentially removed with a lower temperature combustion than the carbon bonded within the clay fraction. Samples from the same vessel were taken from the inner and outer parts of the sherd, and combusted with different temperatures. The determination for the inner sherd combusted at 400[degrees] C was 9020 [+ or -] 65 BP, while that combusted at 800[degrees] C was 11375 [+ or -] 75 BE For this reason, Kuzmin (2002) has suggested that the most reliable carbon fraction is obtained using the lower temperature combustion. The Oxford Ban Chiang samples were all combusted at 900[degrees] C. Berstan et al. (2008: 702) conclude that: 'Direct radiocarbon dating of pottery is relatively uncommon due to the presence of carbon sources with differing ages, for example geological carbon remaining in the clay after firing, added organic temper, carbon from fuel in the kiln and exogenous contaminants absorbed from the burial environment. "

The fallibility of the technique is elegantly illustrated in White's own results obtained from Oxford by Glusker, designed to 'test and refine procedures for dating carbon rich pottery" (Glusker & White 1997: 259; Hedges et al. 1997; Figure 4). This involved subjecting parts of the same vessel to different chemical pretreatments. In treatment A, fragments of organic matter were teased out of the potsherd and treated with acetone, 0.2 N HCI and 0.5 N NaOH successively to extract lipids, carbonates and humic acids. With treatment B the above procedure was applied to crushed potsherds, and followed by either a mixture of4 M HF in 6 M HC1 or concentrated HF followed by multiple washings. This procedure 'aims to concentrate the carbon content of the residue and also makes soluble a considerable quantity of clay-bound humic material' (Glusker & White 1997: 259). The offset between the two forms of pretreatment is often significant. A pot from Middle Period (MP) VIII burial 12 processed under treatment A produced a result of 1970 [+ or -] 60 BP while that for treatment B was 2980 [+ or -] 50 BP. Again, a vessel from burial 19/24 of MP VII has provided a treatment A determination of 2190 [+ or -] 70 BP and 2545 [+ or -] 65 for treatment B.

[FIGURE 4 OMITTED]

Such contradictions have encouraged White to conclude that Southeast Asian archaeologists must tolerate "some degree of chronological fuzziness' (2008:101). Nevertheless, she has persisted with this dating technique, identifying treatment B, that which invariably furnishes the earliest date, as the preferable method. Bur this again produces difficulties, as when the same ceramic vessel from burial 47 of Early Period (EP) III, the earliest Bronze Age phase at the site, has yielded a result of 4810 [+ or -] 90 BP on the basis of pretreatment B, and 2910 [+ or -] 90 BP under treatment A. Results such as these imply that while the pretreatment applied may be making soluble a quantity of humic complexes from the sample, it is also likely to be concentrating clay-bound carbon, which may very well be of an older age. It is likely that there is carbon in the temper of variable age because of the mixed nature of the carbon within the sherds. This ranges from young (humic acids) to potentially quite old (carbon sourced from within clays), as well as variably aged material from the manufacturing of the pots themselves. The oldest determination accepted by White comes from a sample of phytoliths found within a ceramic vessel (White 2008). She terms this "the highest quality radiocarbon determination with the least possibility of contamination' (White 2008: 97). The methods used to isolate the phytoliths are based on those outlined by Mulholland and Prior (1993). In the most recent and detailed attempt to develop methods to date phytoliths, Santos et al. (2010) have concluded that this material is not, at present, a reliable one for independent age determination and much further work is required to demonstrate this (Santos et al. 2010).

[FIGURE 5 OMITTED]

We conclude that other means are necessary if we are to obtain a reliable chronology for this site.

A new series of radiocarbon determinations from Ban Chiang

White has concentrated on the dating of ceramic temper, yet bone is an ideal candidate for dating because collagen, which is almost always the target for radiocarbon dating, is chemically characterisable. In Oxford, gelatinisation followed by an ultrafiltration protocol is applied to extract purified collagen that is reliable for AMS dating (Bronk Ramsey et al. 2004; Higham et al. 2006). One advantage of dating collagen is that, by using a suite of analytical parameters and considering the yields obtained from the bone, it is possible to assess the structural integrity of the collagen, and its preservation state. Sadly, in Thailand bone from some sites has proved difficult to date owing to poor presentation of collagen.

Prior to attempting a new dating programme, we decided therefore to pre-screen suitable bone for AMS dating from Ban Chiang by measuring the %nitrogen in whole bone from animal skeletons placed with the dead as mortuary offerings. Samples of bone containing <0.76%N and CN atomic ratios above 4-5 are usually not dated (Brock et al. 2010). We found that the bones we analysed averaged 0.8% collagen (modern bone contains 4-4.5%N), which indicates poor preservation. Nevertheless, we were able to isolate several bones that were above 1% and these were taken for full collagen pretreatment and dating. The bones were extracted using the ORAU ultrafiltration protocol and AMS dated. Some of the bones were < 1%wt. collagen, which is the lower limit for reliability in Oxford, bur the yields themselves were all > 5mg, and the other analytical parameters measured were acceptable (e.g. the CN atomic ratio), so the results are likely to be accurate. The new determinations are set out in Table 2 and Figure 5.

The earliest stratigraphically comes from bovine bone found in the lower half of a ceramic vessel from the southern section of square D4, layer 32 (Figure 6). The vessel lay just above the subsoil, below the earliest Bronze Age graves. The remaining burial determinations come from graves of EP III-V, EP III representing the earliest phase of the Bronze Age. Four lie well within the first millennium BC, while OxA-22378 from EP V appears a possible outlier and too early. The disparity between the AMS dates obtained by White and this new series is best seen in the calibrated determinations for burial 47, the grave of a man from EP III. Under White's pretreatments A and B, the results are 3770-3370 and 1124-995 cal BC. A pig bone from the same burial is 976-902 cal BC.

In addition to dating mortuary offerings, we have also obtained a determination on human bone from burial 76. This grave contained the flexed skeleton of a young man interred with a bronze socketed spear, described by White as the earliest bronze item found with a burial at Ban Chiang. The result is Ox-24047 2868 [+ or -]26 BP, cal 1128-972 BC (90.7%), 960-936 BC (4.7%).

The relationship between the new series of Ban Chiang dates and those from Ban Non Wat is shown in Figure 7. With the exception of OXA-22378, the results overlap closely with the chronology developed for the Bronze Age phases Ban Non Wat. No results are significantly earlier than 1200 BC. Even OXA-22378 overlaps with the boundary range identified for the start of the Bronze Age at Ban Non Wat (Figure 7).

The new radiocarbon determinations that place the Bronze Age cemetery of Ban Chiang in the early first millennium BC provide a near perfect harmony with an emerging pattern, and are an essential prerequisite to appreciating the course of later Southeast Asian prehistory. The problem posed by the two sites nominated by White and Hamilton to support their early chronology for Ban Chiang are easily set aside. The charcoal-based dates for Non Nok Tha have no scientific credibility and the series of determinations from chaff-tempered sherds suffer from the same reservations as do those from Ban Chiang. A grave from Ban Mai Chaimongkol containing a bronze bar comes from a site with no radiometric dates, and Eyre's (2006) finding that rim forms of the ceramic vessels from the grave parallel those from Neolithic Ban Kao fails the test of careful scrutiny.

It is concluded that White and Hamilton's long and exhaustive argument in favour of a Seima-Turbino long match from the Urals to Ban Chiang is raised on the foundation of seven unreliable radiocarbon determinations. It has often been stressed that the progress of research will provide the evidence necessary to assess properly the course of Southeast Asian prehistory. On this basis we present a new synthesis freed from the anchor of erroneous chronologies.

The implications of a new chronological framework for Ban Chiang

New excavations employing precise and tested dating methods together with archaeogenetic and linguistic evidence are now presenting an emerging pattern of remarkable cultural dynamism in Southeast Asia. Three aspects to hunter-gatherer settlement are now archaeologically visible. One is the occupation of upland rockshelters by Hoabinhian foragers. The second is a vibrant sedentary coastal settlement documented at such sites as Khok Phanom Di. A third aspect comes in the form of large settlements containing inhumation cemeteries and much evidence for stone working but again, no remains of cultivated plants or domestic animals (Xie et al. 2003; Xie & Peng 2006).

[FIGURE 6 OMITTED]

[FIGURE 7 OMITTED]

As Fuller et al. (2009) have shown, there is a long history of rice exploitation in the Yangtze Valley before it became a major part of the subsistence base during the third millennium BC. Zhang and Hung (2010) have followed the expansion of Neolithic rice farmers south into Lingnan and Southeast Asia, showing that rice cultivation was established in southern China by about 2500 BC. Five centuries later, the southern movement of rice farmers was reaching Southeast Asia. The Neolithic rice farmers at Ban Non Wat, dated from the seventeeth century BC, reveal considerable mortuary wealth. Results from other Neolithic sites, such as An Son in southern Vietnam, confirms initial occupation during the early second millennium BC, ending in about 1000 BC (Masanari & Nguyen 2002; Bellwood pers. comm.; Oxenham pers. comm.).

It was during the course of this Neolithic settlement of Southeast Asia that the Xia, Shang and Yangtze valley civilisations developed. Their bronze founders, under an original stimulus from the Qijia and Longshan cultures, experimented with new techniques such as piece-mould casting to produce festive and ritual items destined for the court elite. White and Hamilton (2009) have difficulty in identifying how such a sophisticated and complex casting technology could have influenced the simpler Southeast Asian industry. Yet the Shang founders also cast tools, such as socketed axes, which did not find their way into royal graves (Ciarla 2007: 18; Higham 2008). These court societies extended their exchange orbits into Neolithic Southeast Asia. In northern Vietnam, jade yazhang blades and ge axe halberds have been found in Neolithic Phung Nguyen culture burials. Again, as Ciarla has shown, Shang and early Zhou bronzes reached southern Chinese sites (Ciarla 2007; Pigott & Ciarla 2007). In return, the Shang leaders valued southern cowries and turtle shell, the latter to assist in divination rituals. It is suggested that it was within this proven context of contact and exchange that the knowledge of bronze founding reached southern China and further south.

The sequel in Lingnan and Southeast Asia is seen in sites like Yuanlongpo (Figure 1). There, 110 bronze artefacts came from 350 graves, dated both in terms of imports from the north and a series of eight [sup.14]C determinations to the eleventh--eighth centuries BC (Ciarla 2007). This is a crucial site, for there are two exotic ritual bronzes stylistically belonging to the Western Zhou dynasty (1050-771 BC), together with bronze tools and weapons similar to those of Southeast Asia and the bivalve casting moulds.

We stress the virtually identical casting moulds and socketed axes seen in Lingnan, Vietnam and Thailand. At Non Pa Wai in central Thailand, we find both types of artefact interred with the dead. Again, a bronze founder from Ban Non Wat was interred with two clay moulds for casting socketed axes, and another 25 for casting bangles. At Ban Non Wat, where the area opened exceeds by far that from Ban Chiang (Figure 3), we also encountered compelling evidence that the adoption of copper-base metallurgy coincided with a remarkable expression of social display. Graves far larger than were needed to accommodate a coffin were filled with up to 83 ceramic vessels often finely decorated with complex painted patterns. More often than not, copper-base axes were found not only with men and women, but also with infants. There were also copper-base anklets, bells, chisels and awls. Exotic bangles of marble and marine shell covered the arms of some people from shoulder to wrist. Some women wore over 23 000 shell beads strung as necklaces or belts, and up to 22 shell earrings. This presence of social aggrandizers at last fills a void in our understanding of the Southeast Asian Bronze Age, for it is clear that at this site, at least one social group projected their high status through ostentatious mortuary rituals that almost certainly included the provision of lavish feasts.

Conclusion

As Movius stated with reference to the Upper Palaeolithic in Europe: 'Without ... such a [chronological] framework the over-all picture becomes confused and, in certain instances, almost meaningless. Time alone is the lens that can throw it into focus' (Movius 1960:355).

This is nowhere better illustrated than in the elucidation of Southeast Asian prehistory since the 1960s. Seminal changes have come from both the techniques of radiocarbon dating and the programs used in the exegesis of a set of determinations from a given site. Unless identified to a short-lived species, charcoal determinations must now be seen as providing only a terminus post quem, for it is impossible to quantify the distorting effect of inbuilt age. Whereas the AMS dating of rice-tempered pottery once seemed an elixir, it is now known to be so open to inaccuracies that in our view its use should be set aside and any result, again, seen at best as a terminus post quem. White has recently presented a rather pessimistic view of such issues when she wrote that: 'The degree of resolution possible for many problems in regional prehistory is coarse for now. Often, "within the correct half millennium" and when lucky, the correct third of a millennium, are adequate for many issues" (White 2008: 101).

This contrasts with the improving precision and reliability of AMS dating on reliable materials, integrated with careful sample selection and the use of Bayesian statistics in the modelling of radiocarbon dates with respect to inferred archaeological sequences. In such cases, as noted by Bayliss et al. (2007): 'Timetables of the sort presented in these papers are now not only achievable on a routine basis, but are a necessity if we are to address fundamental questions about our pasts, including the experience of the flow of life, the social marking of time.'

We argue that the chronological fuzziness that has characterised White's interpretation of the site of Ban Chiang for so long has begun to disperse. We have presented a set of determinations obtained from a key Early Bronze Age skeleton and the animal bones that we excavated at Ban Chiang in 1975, which reliable and precise techniques now enable us to date. Our new determinations allow us to place a missing piece of the jigsaw with direct relevance to the transmission of metallurgical knowledge. Southeast Asia was the last region in the mainland of Eurasia to receive and adopt bronze technology (Roberts et al. 2009). At Ban Non Wat, we have identified a sophisticated Neolithic community whose potters, by the seventeeth century BC, commanded the pyrotechnological expertise needed to adapt to the casting of copper-base artefacts (Higham & Kijngam 2009). As Linduff and Mei (2009) have stressed, it is necessary to identify, as we have done at Ban Non Wat, communities with established technical skill and social demands before metallurgy can be established. By the late eleventh century, the occupants of this site came to share the knowledge of how to produce copper-base tools, weapons and ornaments. The vital importance of chronology is illustrated in the finding that the Bronze Age at Ban Chiang is a millennium later than claimed by White and Hamilton (2009). Their model was constructed on the foundation of seven impaired radiocarbon determinations from techniques that we argue, should no longer be employed. When this foundation is removed, their edifice collapses.

Acknowledgements

We wish to thank the late Dr C.F. Gorman and Dr Pisit Charoenwongsa for inviting us to participate in the excavations and subsequent research on the site of Ban Chiang. The Ford Foundation supported our participation with a training grant. Staff of the Oxford Radiocarbon Accelerator Unit are thanked for their careful analytical work on the bone samples from Ban Chiang. We thank Leslie O'Neill for drawing the burials in Figure 6.

References

BAYLISS, A.C., C. BRONK RAMSEY, J. VAN DER PLICHT & A. WHITTLE. 2007. Bradshaw and Bayes: towards a timetable for the Neolithic. Cambridge Archaeological Journal 17(1) (Supplement): 1-28.

BERSTAN, R., A.W. STOTT, S. MINNIT, C. BRONK RAMSEY, R.E.M. HEDGES & R.P. EVERSHED. 2008. Direct dating of pottery from its organic residues: new precision using compound-specific carbon isotopes. Antiquity 82: 702-13.

BONSALL, C., G. COOK, J. MANSON & D. ANDERSON. 2002. Direct dating of Neolithic pottery: progress and prospects. Documenta Praehistorica 29: 47-59.

BROCK, F., T.EG. HIGHAM & C. BRONK RAMSEY. 2010. Pre-screening techniques for identification of samples suitable for radiocarbon dating of poorly preserved bones. Journal of Archaeological Science 37(4): 855-65.

BRONK RAMSEY, C. 2009. Bayesian analysis of radiocarbon dates. Radiocarban 51: 337-60.

BRONK RAMSEY, C., T.F.G. HIGHAM, A. BOWLES & R. HEDGES. 2004. Improvements to the pretreatment of bone at Oxford. Radiocarbon 46(1): 155-63.

CHERNYKH, E.N. 1992. Ancient metallurgy in the USSR. Cambridge: Cambridge University Press.

CHIOU-PENG, T. 2009. Incipient metallurgy in Yunnan: new data for old debates, in J. Mei & Th. Rehren (ed.) Metalllurgy and civilization : Eurasia and beyond: 79-84. London: Archetype.

CIARLA, R. 2007. Rethinking Yuanglongpo. The case for technological links between Lingnan (PRC) and central Thailand during the Bronze Age. East and West 57: 1-23.

DE ATLEY, S.P. 1980. Radiocarbon dating of ceramic materials: progress and prospects. Radiocarbon 22(3): 987-93.

EYRE, C. 2006. Prehistoric and proto-historic communities in the eastern Upper Chao Phraya river valley, Thailand: analysis of site chronology, settlement patterns, and land use. Unpublished PhD dissertation, University of Pennsylvania.

FULLER, D., L. QIN, Y. ZHENG, Z. ZHAO, X. CHEN, L.A. HOSOYA & G.P. SUN. 2009. The domestication process and domestication rate in rice: spikelet bases from the Lower Yangtze. Science 323: 1607-1610.

GLUSKER, D. & J. WHITE. 1997. Comment. Archaeometry 39: 259-60.

GORODTSOV, V.A. 1916. Kultury bonzovoi epokhi v Srednei Rossii, in Otchety Istoricheskogo muzeya za 1916 g. Moscow.

HANKS, B.K., A.V. EPIMAKHOV & A.C. RENFREW. 2007. Towards a refined chronology for the Bronze Age of the southern Urals, Russia. Antiquity 81: 353-67.

HEDGES R.E.M., C. TIEMEI & R.A. HOUSLEY. 1992. Results and methods in the radiocarbon dating of pottery. Radiocarbon 34(3): 906-915.

HEDGES R.E.M., P.B. PETTITT, C. BRONK RAMSEY & G.J. VAN KLINKEN. 1997. Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 23. Archaeometry 39: 247-62.

HIGHAM, C.F.W. 1996. The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press.

--2008. Recasting Thailand: new discoveries at Ban Non Wat. Current World Archaeology 31: 38-41.

HIGHAM, C.F.W. & T.F.G. HIGHAM. 2009. A new chronological framework for prehistoric Southeast Asia, based on a Bayesian model from Ban Non Wat. Antiquity 82: 125-44.

HIGHAM, C.F.W. & A. KIJNGAM. 2009. The origins of the civilization of Angkor, Volume 3. The excavation of Ban Non Wat: introduction. Bangkok: Fine Arts Department.

HIGHAM, T.F.G., R.M. JACOBI & C. BRONK RAMSEY. 2006. AMS radiocarbon dating of ancient bone using ultrafiltration. Radiocarbon 48(2): 179-95.

KUZMIN, Y.V. 2002. The earliest centers of pottery origin in the Russian Far East and Siberia: review of chronology for the oldest Neolithic cultures. Documenta Praehistorica 29: 37-46.

KUZMIN, Y.V., S. HALL, M.S. TITE, R. BAILEY, J.M. O'MALLEY & V.E. MEDVEDEV. 2001. Radiocarbon and thermoluminescence dating of the pottery from the early Neolithic site of Gasya (Russian Far East): initial results. Quaternary Science Reviews 20: 945-48.

LINDUFF, K.M & J. MEI. 2009. Metallurgy in eastern Eurasia: retrospect and prospects. Journal of World Prehistory 22: 265-81.

MANNING, K., R. PELLING, T.F.G. HIGHAM, J.-L. SCHWENNIGER & D.Q. FULLER. 2011. 4500-year-old Domesticated Pearl Millet (Pennisetum glaucum) from the Tilemsi Valley, Mali: new insights into an alternative cereal domestication pathway in Africa: Journal of Archaeological Science 38: 312-322.

MASANARI, N. & K.D. NGUYEN. 2002. Excavation of An Son: a Neolithic mound site in the middle reach of the Varo Co Dong river, southern Vietnam. Bulletin of the Indo-Pacific Prehistory Association 22: 101-109.

MEI, J. 2003. Cultural interaction between China and Central Asia during the Bronze Age. Proceedings of the British Academy 121: 1-39.

MOVIUS, H.L. 1960. Radiocarbon dates and Upper Palaeolithic archaeology in Central and Western Europe. Current Anthropology 1: 355-91.

MULHOLLAND, S.C. & C. PRIOR. 1993. AMS radiocarbon dating of phytoliths, in D.M. Pearsall & D.R. Piperno (ed.) Current research in phytolith analysis: applications in archaeology and paleoecology (MASCA Research Papers in Science and Archaeology 10): 21-23. Philadelphia (PA): The University Museum of Archaeology and Anthropology, University of Pennsylvania.

PIETRUSEWSKY, M. & M.T. DOUGLAS. 2002. Ban Chiang, a prehistoric village site in northeast Thailand h the human skeletal remains. Philadelphia (PA): The Museum of Archaeology and Anthropology, University of Pennsylvania.

PIGOTT, V.C. & R. CIARLA. 2007. On the origins of metallurgy in prehistoric Southeast Asia: a view from Thailand, in S. La Niece, D. Hook & E Craddock (ed.) Metals and mines. Studies in Archaeometallurgy: 76-88. London: Archetype in association with the British Museum.

ROBERTS, B.W., C.E THORNTON & V.C. PIGOTT. 2009. Development of metallurgy in Eurasia. Antiquity 83: 1012-1022.

SANTOS, G.M, A. ALEXANDRE, H.H.G. COE, P.E. REVERSON, J.R. SOUTHON & C.N. DE CARVALHO. 2010. The Phytolith [sup.14]C puzzle: a tale of background determinations and accuracy tests. Radiocarbon 52(1): 113-28.

WHITE, J.C. 1997. A brief note on new dates for the Ban Chiang cultural tradition. Bulletin of the Indo-Pacific Prehistory Association 16: 103-106.

-- 2008. Dating early bronze ar Ban Chiang, Thailand, in J.-P. Pautreau, A.-S. Coupey, V. Zeitoun & E. Rambault (ed.) From Homo erectus to the living traditions: 91-110. Bougon: European Association of Southeast Asian Archaeologists.

WHITE J.C. & E. HAMILTON. 2009. The transmission of early Bronze Age technology to Thailand: new perspectives. Journal of World Prehistory 22: 357-97.

XIE G.M., Q. LIN & C.L. PENG. 2003. Gexinqiao Neolithic site in Baise City, Guangxi. Archaeology 12: 3-6 (with English abstract).

XIE G.M. & C.L. PENG. 2006. An important discovery from the excavation of Baida site in Baise, Guangxi. China Cultural Relics News 1409 (in Chinese).

ZHANG, C. & H.-C. HUNG. 2010. The emergence of agriculture in southern China. Antiquity 84:11-25.

Charles Higham, (1) Thomas Higham (2) & Amphan Kijngam (3)

(1) Department of Anthropology, University of Otago, P.O. Box 56, Dunedin, New Zealand (Email: charles.higham@otago.ac.nz)

(2) Oxford Radiocarbon Accelerator Unit, Research Laboratory of Archaeology and the History of Art, Dyson Perrins Building, University of Oxford, OX1 3QY, UK (Email: thomas.higham@rlaha.ox.ac.uk)

(3) Archaeology Division, Fine Arts Department, Sri Ayutthaya Road, Bangkok, Thailand (Email: ampankij@yahoo.co.th)

Received: 28 June 2010; Accepted: 3 September 2010; Revised: 7 September 2010

Table 1. The cultural sequence of Ban Chiang (after Pietrusewky
& Douglas 2002: 5).

Burial phase Date according to White Major period *

Late Period X 300 BC AD 200 Iron Age
Late Period IX
Middle Period VIII 900-300 BC
Middle Period VII
Middle Period VI Bronze Age
Early Period V 2100-900 BC
Early Period IV
Early Period III
Early Period II Neolithic
Early Period I

* As far as can be ascertained by CH.

Table 2. Radiocarbon determinations from Ban Chiang. All
determinations are of ultrafiltered gelatin. Stable isotope ratios
are expressed in % relative to vPDB. Mass spectrometric precision is
[+ or -] 0.2% for C and [+ or -] 0.3% for N. Wt. used is the amount
of bone pretreated and the yield represents the weight of gelatin or
ultrafiltered gelatin in milligrams. % yield is the wt.% collagen
(the amount of collagen extracted as a percentage of the starting
weight). %C is the carbon present in the combusted gelatin. For
ultrafiltered gelatin this averages 41.02%. CN is the atomic ratio of
carbon to nitrogen. At ORAU this is acceptable if it ranges between
2.9-3.5.

 Conventional Wt. used
OxA Species radiocarbon age BP Context (mg)

22378 Sus scrofa 2965 [+ or -] 29 Burial 29 EPV 1250
22380 Sus scrofa 2516 [+ or -] 26 Burial 56 EPV 1210
22383 Sus scrofa 2819 [+ or -] 26 Burial 54EPIV 1450
22381 Sus scrofa 2786 [+ or -] 26 Burial 47EPIII 1070
24047 Homo sapiens 2868 [+ or -] 26 Burial 76 EPIII 620
22379 Bos sp. 2515 [+ or -] 26 D4 SEQ32 ?EPII 1400

 Yield
OxA (mg) %Yld %C [delta] [sup.13]C (%)

22378 16 1.3 44.7 -20.9
22380 8.87 0.7 44.5 -19.4
22383 23.9 1.6 45.2 -19.6
22381 5.5 0.5 44.8 -19.1
24047 6.07 1.0 43.2 -18.3
22379 5.7 0.4 44.2 -14.1

OxA [delta] [sup.15]N (%) CN

22378 8.8 3.4
22380 8.7 3.2
22383 7.9 3.1
22381 7.8 3.1
24047 10.6 3.2
22379 4.2 3.1