Linking the early historic cultures of Southeast Asia to their prehistoric antecedents has, despite some decades of research, proved difficult for a number of reasons. Until the 1960s, this did not seem to be an issue for scholars of the region. For George Coedès,Footnote 1 the region was simply occupied by diverse and backward ethnic and linguistic communities, cultivators using stone tools, or at best, people using bronze with poorly developed social organisations until impacted by the expanding Han peoples during the third century CE: ‘There is very little direct information about the societies of South East Asia before they entered into contact with India and China.’Footnote 2
Field archaeology over the past four decades has subsequently revealed the remains of many complex societies with developed hierarchies and networks of trade spanning thousands of kilometres. Sites of late prehistoric communities are regularly marked by substantial cemeteries in which the dead were buried with a wide variety of grave goods including pottery, stone, bronze, iron tools, weapons and ornaments made from shell, stone and glass.
Sites from the historic period, on the other hand, are recognised by the remains of brick or stone religious buildings; some sites may have substantial surrounding walls covering several hectares that are ringed by moats. Until recently, settlement areas have barely been investigated. In early historic sites, burials are considered rare or not yet linked with certainty to the culture of the monuments.Footnote 3 These problems come for the most part from the different methodologies used by field archaeologists and art historians, the nature of the materials studied (pottery, metal finds, burials, stratified settlement refuse) and the reliance of radiometric dating methods on one hand, and standing monuments, stone and bronze sculpture, stuccos, paintings, inscriptions and stylistic comparisons on the other. This is to emphasise the differences rather than the methods and approaches shared by field archaeologists and art historians.
Chronology is also a major problem; in parts of mainland Southeast Asia such as Vietnam and especially northeast Thailand, a fairly robust radiometric chronology is available based on around one thousand C-14 determinations. These determinations document the arrival of polished stone-using field-based rice agriculturalists in the late third millennium BCE, as well as copper and bronze tools and weapons that were used by increasingly complex, stratified communities before the end of the second millennium. By about 500 BCE, bronze was replaced by iron for most tools and weapons in the western and central parts of the region.Footnote 4
Dating early historic sites
The dating of early historic sites has, for the most part, depended on stylistic analysis of sculpture and ornamentation on buildings, palaeographic comparisons with South Asian inscriptions, and the assumption that developments in Southeast Asia must have occurred later, usually by several hundred years, than in India. However, Hermann Kulke long ago argued that complex state-like polities developed at more or less the same time on both sides of the Bay of Bengal: ‘If we look at the societies on both sides of the Bay of Bengal in the middle of the first millennium A.D., we observe … a certain convergence of the social evolution in both regions.’Footnote 5
Stone and metal cannot be directly dated and thermoluminescence dating of fired bricks has shown to be somewhat unsuccessful in dating early Southeast Asia buildings.Footnote 6 But looking back a little, we should recognise that H.G. Quaritch Wales was the first person to recognise the need for subsurface investigation of early historic monuments with his work at the Dvaravati site of P'ong Tuk in western Thailand and in Peninsular Malaya.Footnote 7
In the 1960s Jean Boisselier had some sondages made at the Dvaravati town of U-Thong and reported a long sequence of occupational deposits that he felt could be dated from early in the first millennium CE and perhaps to a much earlier Neolithic settlement.Footnote 8
Dating organic material within and below the foundations of structures has only occasionally been attempted. In the late 1960s Bennett Bronsen and George Dales excavated at the minor Dvaravati site of Chansen in western Thailand with the express aim of establishing a chronological sequence. Until then, brick monuments and art objects had only been dated by comparison with Indian sculpture.Footnote 9
The second systematic attempt to investigate the Dvaravati levels at U-Thong from the field archaeology point of view was the Thai–British Archaeological Expedition of 1966–70 led by William Watson and Helmut Loofs who excavated at a location referred to as Tha Muang.Footnote 10 Radiocarbon dates from Bronson's work at Chansen and Watson's and Loofs’ excavations at U-Thong were later revised by Andrew Barram, who made it clear that neither of these dated sequences support the traditional dating for Dvaravati culture between 600 and 1000 CE. Instead, they strongly support the notion that communities heavily influenced by Indian ceramic and other traditions of material culture had occupied these sites from early in the first millennium CE.Footnote 11 Thus, the dating of what we might call an ‘Early Dvaravati’ culture was established in the region.Footnote 12 Such a revision would bring Thailand more into line with the picture now being revealed by excavations at early ‘Indianised’ sites in Cambodia, Vietnam and Java and Burma.
In West Java, Pierre-Yves Manguin and Agustijanto Indradjaya undertook a project specifically aimed at bridging the gap between prehistory and the historic period, demonstrating that the first phase of construction of Buddhist temples at Batujaya could not have been earlier than the fifth century CE, but more likely during the seventh century.Footnote 13
Likewise in central Vietnam, excavations at and around Tra Kieu between 1993 and 2003 show that occupation of this large Cham urban site started no later than the first to second century with distinctive Indic-style ceramics and brick structures following soon after.Footnote 14 In the 1920s J.Y. Claeys of the École Française d'Extrême-Orient (EFEO) had previously excavated at Tra Kieu hoping to verify whether or not it was the location of a Cham capital attacked by the Chinese in the fifth century CE — with ambiguous results.Footnote 15 The slightly earlier and nearby site of Go Cam is the polity known from Chinese sources as Lin Yi, a transitional phase between the late prehistoric Iron Age Sa Huynh culture and the emergence of Champa.Footnote 16 The emergence of the historic Champa kingdoms is generally dated to the mid-fifth century CE as marked by the Sanskrit inscriptions of Bhadravarman.Footnote 17 Field archaeology at Tra Kieu extends the life of this urban site back several hundred years. Taking into account the new dating from Tra Kieu, we cannot rule out that the earlier, often disputed, date for the third century Vo Canh inscription may be correct.Footnote 18 The continuing project at Tra Kieu shows that the perimeter brick walls of the ancient Cham city of Simapura at Tra Kieu were built in the early centuries CE.Footnote 19
Excavations at Angkor Borei, in southeastern Cambodia, by a joint University of Hawai'i–Cambodian team shows that a typical Pre-Angkorian Khmer culture began occupying the ancient settlement no later than the fourth or fifth centuries CE. The excavation overlies a cemetery with inhumation burials dated between c.200 BCE and 200 CE.Footnote 20
Just across the present-day border into Vietnam, renewed work by Vietnamese and a joint French–Viet expedition has expanded our understanding of Funan culture as set out in the classic reports by Louis Malleret.Footnote 21 Excavations at the Oc Eo/ Ba Thê complex have revealed a Phase I settlement dating to the mid-first to mid-third centuries CE, with many India-related materials such as kendi and roofing tiles with brick buildings. These materials began appearing in the Phase II of the third century, marking a transition to the historical period in this area.Footnote 22 These floodplain settlements at Oc Eo seemed to have flourished in subsequent years only to be abandoned by the mid-seventh century CE, soon after the traditional date for the establishment of Dvaravati culture in Thailand.Footnote 23
In 2007–08 a joint Vietnamese–American team dated the construction of the ramparts at Co Loa in Northern Vietnam to the fourth century BCE through a series of radiocarbon dates associated with pottery inside and below the construction.Footnote 24
To summarise I argue that Hermann Kulke's 1990 ‘convergence hypothesis’ has some heuristic value. It shows that the almost simultaneous emergence of complex societies on both sides of the Bay of Bengal was based on long established and local processes of social and technical development, as exemplified by the results obtained at Ban Don Ta Phet and Khao Sam Kaeo. Societies in Thailand incorporated many elements of Indic civilisation between the third and sixth centuries CE, as now shown by recent excavations at Nakhon Pathom and U-Thong by San Thaiyanonda, giving rise to the regionally discrete configurations of the later first millennium CE.Footnote 25
