Socio-Economic Dynamics in Peninsular Thailand: Food-Producing Strategies in Stone Age communities

Based on field studies in Songkhla, lower peninsular Thailand

Masters thesis. Department of Archaeology and Ancient History. Uppsala University. 2010. Henrik Lekenvall. Supervisor: Paul J.J. Sinclair

Masters thesis at Uppsala University May 2011. ABSTRACT Lekenvall, H. Socio-Economic Dynamics in Peninsular Thailand: Food-Producing Strategies in Stone Age communities. Department of Archaeology and Ancient History. Uppsala University. The conventional view of Stone Age communities in Southeast Asia is that they consisted of foraging hunter-gatherers, who became colonized by Austroasiatic speaking agriculturalists from southern China in the second millennium BC. For this reason, it has been argued that rice agriculture was the earliest type of food-production in the region. Consequently, agriculture and later urbanism and state formation have become associated with rice domesticating communities, neglecting alternative interpretations, such as indigenously developed cultivation strategies. Recent environmental data from various parts of Southeast Asia and elsewhere, however, suggest that forager groups practiced forest clearance in the late Pleistocene, and that there was an increase in landscape management practices during the Holocene. During this time, communities also started to cultivate plants. The archaeological survey of Songkhla Province in lower peninsular Thailand presented here was aimed at locating Stone Age cave and rock shelter sites. Up till now only a few sites from this period have been found in peninsular Thailand. I base my analysis on the notion that environment and present day land-use reflect the diversity of past cultivation strategies. The aim is therefore to enhance our archaeological understanding of the area and connect the survey results with environmental data and current land-use strategies. One of the most important tasks for the study is to address archaeological site visibility in landscapes associated with land forms, soils and water. In doing so, the sites may be used to explain the changes in land-use which took place around 4000 BP. The archaeological record attests the presence of Late Stone Age groups from 4000 BP. By reviewing environmental research and archaeological data, I establish a link between past and present land-use and sites found during the archeological survey. The function of cave and rock shelter sites changing from temporary residential sites to their role as burial localities after 4000 BP supports this notion. On this basis, I suggest investigations of cave and rock shelter sites benefits our understanding of Holocene land-use and subsistence strategies, as they indicate a connection between local economic systems and site use. However, environmental evidence suggests a much earlier adaptation to a food-producing economy, which intensified in later Holocene periods. The increase in disturbed taxa and land-use seems to be associated with the supposed contemporary immigration of agricultural groups. Archaeological and environmental data from neighboring regions suggests intensified land-use connected with foodproducing strategies of horticulture and arboriculture during this period. I propose similar interpretations for peninsular Thailand, and support this by local environmental data. It is also important to note that the environmental data does not support the idea of a cultural and chronological gap between hunter-gatherers and expansionary agriculturalists. There are also indications of overlapping in the archaeological assemblages commonly associated with hunter-gatherer communities (the so-called Hoabinhian material culture) and those attributed to agriculturalists (Neolithic material culture). This further questions the necessity of continuing the use of these concepts. Local level explanations are also crucial to exploring specific socio-environmental interactions and the resulting cultural implications. This does not preclude regional patterns but local level explanations stresses the possibility the possibility of several interrelated factors, which affected the agricultural development in the region. Similarly, sub-cultures and predominant-cultures are always affecting and influencing one another, resulting in an inter-marriage of cultures. On this basis, I suggest a continuation of internal (indigenously) developed food-producing strategies in interaction with external (immigrating) cultural and economical influences attributed to Austroasiatic speaking agriculturalists, rather than relying on a simplistic monocausal explanation, namely the introduction of domesticated rice and rapid replacement of hunter-gatherers by expanding agriculturalists. In this framework, both indigenous activities and external stimuli, had an effect on and contributed to the development of agriculture, and in the long run, also urbanism and state formation in the region.

Keywords Peninsular Thailand; archaeology; caves; rock shelters; land-use; cultivation; agriculture; foodproduction; Late Stone Age; hunter-gatherers; Pleistocene; Holocene; environmental data. Henrik Lekenvall Layout, maps, photographs and illustrations by Henrik Lekenvall, unless otherwise indicated.

TABLE OF CONTENTS
ILLUSTRATIONS .................................................................................................................................. 8 ACKNOWLEDGEMENTS .................................................................................................................. 10 INTRODUCTION ................................................................................................................................. 11 1. RESEARCH BACKGROUND ......................................................................................................... 15 1.1 Southeast Asia ............................................................................................................................. 15 1.2 Previous research and general trends .......................................................................................... 16 The Thai-Malay Peninsula ............................................................................................................ 18 The issues of definitions ................................................................................................................ 20 1.3 The archaeological and environmental evidence ........................................................................ 22 Prime indicator (1): stone implements, pottery and rock paintings ............................................... 23 Prime indicator (2): pollen and phytolith ...................................................................................... 26 1.4 Defining the problem................................................................................................................... 28 2. THE RESEARCH AREA ............................................................................................................. 37 2.1 Environmental history and archaeology ...................................................................................... 37 Seal level changes ......................................................................................................................... 38 GIS ................................................................................................................................................ 39 2.2 Environmental setting.................................................................................................................. 40 Main landforms ............................................................................................................................. 40 Land development ......................................................................................................................... 42 Sediments ...................................................................................................................................... 43 Soils ............................................................................................................................................... 46 Land-use ........................................................................................................................................ 49 3. THE ARCHAEOLOGICAL SURVEY ........................................................................................ 54 5

3.1 Survey strategies.......................................................................................................................... 54 Oral testimony associated with Satingphra ................................................................................... 56 Khao Daeng Rockshelter ............................................................................................................... 57 Khao Rakian Mountain ................................................................................................................. 58 Khao Nui Mountain ....................................................................................................................... 60 Khao Khua Cave ........................................................................................................................... 60 Khao Joompa Hill.......................................................................................................................... 62 Khao Tok Nan Hill ........................................................................................................................ 63 Khao Chang Lon Mountain ........................................................................................................... 65 The museum collections ................................................................................................................ 67 3.2 Is the material culture evidence of food-production? .................................................................. 69 3.3 Sites in GIS.................................................................................................................................. 72 3.4 Site location and environment ..................................................................................................... 72 4. DISCUSSION ............................................................................................................................... 81 4.1 Summary and a local model ........................................................................................................ 81 Relating the data ............................................................................................................................ 81 An alternative perspective ............................................................................................................. 87 REFERENCES CITED ......................................................................................................................... 91

ILLUSTRATIONS
Figures: 1.1. Map of Southeast Asia showing the research area, p. 16. 1.2. Lang Rongrien Rockshelter after Anderson 1990: 2, p. 19. 1.3. Material culture of contemporary Semang hunter-gatherers, p. 33. 2.1. Land development: geological land forms, p. 41. 2.2. Land development: sediments, p. 42. 2.3. Land development: spit formations, p.45. 2.4. Land development: river meanders, p. 45. 2.5. Land development: soil types, p. 47. 2.6. Current land-use system over lower peninsular Thailand. The research area emphasized p. 50. 2.7. Rubber tree plantation close to Khao Rakian Cave. 51. 2.8. Current land-use and topography in Rattaphum district after Panapitukkul et al. 2005: 150, p. 51. 3.1. Late prehistoric and historic sites on the Satingphra Peninsula and Stone Age sites in the inland as indicated by potsherds, stone tools and oral testimony, p. 55. 3.2. Map of cities and road system in lower peninsular Thailand, p. 55. 3.3. The wall of the rockshelter showing disturbance of ground surface, p. 57. 3.4. Khao Daeng Rockshelter and present day altar and house, p. 57. 3.5. View of the landscape from the highest point of Khao Rakian Mountain, p. 59. 3.6. The entrance of Khao Khua Cave, p. 61. 3.7. Potsherds found inside Khao Khua Cave, p. 61. 3.8. Khao Joompa hill from the ground level, p. 62. 3.9. Khao Tok Nan from ground surface, p. 63. 3.10. View towards Khao Joompa and Khao Khua, p. 63. 3.11. The entrance of Khao Tok Nan Hill where a polished stone axe is reported to have been found, p. 64. 3.12. Khlong Paum Canal south of Khao Tok Nan Hill, p.64. 3.13. Polished stone axe found in the vicinity, p. 65. 3.14. Khao Chang Lom situated at the foot of the mountain, p.65. 3.15. The wall inside the cave showing removal of soil, p. 66. 3.16. Hammerstone found in one of the chambers, p. 66. 3.17. The entrance of Khao Chang Lom from the inside, p. 67. 3.18. The entrance of Tham Kra Duk from the inside, p. 67. 3.19. Polished stone axe with cutting edge. Songkhla National Museum, p. 68. 3.20. Polished stone axe. Songkhla National Museum, p. 68. 3.21. Polished shouldered stone adze. Songkhla National Museum, p. 68. 3.22. Stone tool collection. Wat Machimawat National Museum, p. 68. 3.23. Stone knife? Wat Machimawat National Museum, p. 68. 3.24. Late Stone Age pottery. Thaksin Folklore Museum, p. 68. 3.25. Tripod. Thaksin Folklore Museum, p. 69. 3.26. Hoabinhian stone tools. Kedah State Museum, p. 69. 3.27. Polished stone tools. Kedah State Museum, p. 69. 3.28. The surveyed archaeological sites seen in relation with the different environmental features and distance between them, p. 73. 3.29. Current land use and topography of the research area after Panapitukkul et al, 2005: 150, with the archaeological sites found during survey, p. 74.

Tables: 1.1. Evolutionary scheme of plant-food production adapted from Harris 1996: 4, p. 30. 4.1. Scheme over socio-economic dynamics during contemporary time periods in the research area by Lekenvall, p. 88.

ACKNOWLEDGEMENTS
My research in Thailand has been carried out with financial support from Sida (Minor Field

Studies) and the Rydbergs fond and African and Comparative Archaeology, Department of Archaeology and Ancient History, Uppsala University. I am grateful for the support, without which the archaeological fieldwork could not have been carried out. In Thailand, many people contributed to the study. Firstly, I want to acknowledge the 13th Fine Arts Department, Songkhla. I should like to thank Supot Prommanot and Pornthip Pantukowit for accepting and allowing me to carry out fieldwork in the area. I am especially grateful to Siriporn Limwijitwong for assistance and cooperation in the field. I also wish to thank the staff at the office. You all made me feel welcome and I will never forget your hospitality. My gratitude also goes to Songkhla National Museum, Thaksin Folklore Museum, Wat Machimiwat National Museum, Nakhon Si Thammarat National Museum, Satun National Museum and Kedah State Museum. I should like to thank the personnel for answering questions regarding collections and allowing me to take photographs. In addition, the local residents must not be forgotten. Your words became an important part of this thesis. Restaurant owners, farmers, fishermen, Monks, you all contributed to this final work. Your warmth and hospitality charmed me. Last but not least, my deepest gratitude goes to my colleagues at the Institute of Archaeology, Uppsala University. I should like to thank Anna Karlstrm, who helped me with the initial work for this thesis, suggesting literature and people to contact in Thailand, and Karl-Johan Lindholm, who supervised my bachelors thesis and was one of the lecturers in a course on Landscape Archaeology and GIS at Uppsala University. He has contributed to my development as a researcher regarding writing academic texts and has also been helpful with GIS. Thanks to Anneli Ekblom for her comments and Marjaana Kohtamki for helping me with the English and for being such a good friend. I am indebted to my supervisor, Paul Sinclair, for proposing the research area for my Masters thesis and his constant support throughout my studies at Uppsala. I am grateful for the comments which he has provided on this thesis and all other help.

10

INTRODUCTION
The prevailing questions in Southeast Asian archaeology have traditionally focused on the origins of rice agriculture in the Late Stone Age and the organizational changes during the Bronze Age, which are believed to have been the foundation for state development along the coast and inland river valleys. Northeast and central Thailand is also characterized by a rich Hoabinhian and Neolithic past. The aforementioned cultural periods have been used to explain two distinctive food procurement systems: hunting and gathering and agriculture. The conventional view is that Austroasiatic-speaking (AA) rice agriculturalists (i.e. Neolithic communities), from southern China colonized the continent in the second millennium BC. For this reason, it has been argued that communities prior to this intrusion must have been hunter-gatherers (i.e. Hoabinhian foragers). An indisputable result of this perception is that alternative food-producing strategies and ensuing cultural trajectories have usually been ignored. However, in the last decade or so, an alternative view has emerged. An increasing amount of environmental data suggests that communities from the late Pleistocene and onwards developed different strategies to cope with environmental change. Forest-burning by human agencies in Thailand is evident from the pollen and phytolith record. This has also been considered evidence for the existence of agro-ecosystems and food-production prior to the colonization by AA agriculturalists. In the Thai-Malay Peninsula has archaeological research predominantly focused on the development of urbanism in Satingphra and Kedah during the first centuries AD. Regional prehistory has mostly remained unexplored. The peripheral location of the area, which lies in the southern parts of Mainland Southeast Asia, is probably the reason for its exclusion from research. Current knowledge on Stone Age communities is based on archaeological information extracted from a few, often very disturbed, caves and rock shelters. Lang Rongrien Rockshelter, in Krabi Province, is one of few well-excavated sites in the region, which has revealed a continuous occupation from the late Pleistocene until early historic times. Polished stone implements, rock paintings and pottery have also been recovered at many locations on the peninsula but the provenience of the remains has rarely been determined. Very often, they derive from excavations carried out in the first half of the last century. At that time, not much was known about site stratigraphy, and radiocarbon dating had not yet been invented. Based on similar Late Stone Age material culture found in the Thai-Malay Peninsula and in the rest of Southeast Asia, scholars have argued for a similar chronological socio-economic
11

sequence in these areas. As a result of this concept, all local groups are believed to have been hunter-gatherers prior to the agriculturalist influence. The existence of indigenous foodproduction has not been acknowledged. Yet, different food-procurement strategies prevailed in the area prior to the introduction of rice agriculture. On the basis of environmental data and comparative studies from similar ecological settings elsewhere, a subsistence regime of arboriculture and horticulture and other crops can be interpreted from burned taxa in the pollen and phytolith record. This advocates a local/regional trajectory of agro-ecosystem development and food-production, which predates the adoption of rice agriculture. Moreover, this thesis is also partially based on an archaeological survey carried out in Songkhla Province in south-eastern peninsular Thailand. Situated in the humid tropics, the local land has experienced substantial land development throughout time. Fluctuating sea levels, chemical weathering and transportation of sediments by wind and water have formed the contemporary landscape, composition of soils and the extent of land. The past and current environment is used to facilitate an understanding of the landscape in which Stone Age communities interacted. The possibility of conducting food-production for past communities in this kind of environment is also discussed in relation with current local land-use. An archaeological survey carried out by the author revealed eight caves and rock shelters with direct (surface finds) or indirect (oral testimonial) evidence of Late Stone Age material culture. Since the majority of these sites is associated with Late Stone Age pottery and skeletal remains, it is argued that the material culture can be linked with burial practices, which signals some sort of societal change in the Late Stone Age. It is possible that this change is related with the arrival of early farming communities. Based on this hypothesis, it could be argued that incoming agriculturalists had a great impact on the socio-economic organization of the region. Nevertheless, in my opinion, it also shows that the complexity of farmer/huntergatherer interaction requires an approach alternative to merely assigning cultural associations to material culture of immigrant farmers or resident hunter and gatherers, as different sources of data points towards an intertwined process of change in lower peninsular Thailand. By taking on a landscape approach to the understanding of these processes, rather than just focusing on material culture, I aim to succeed where the cultural historical orientation has failed, namely to argue in favor of food-production and socio-economic diversification during contemporary time periods. The main argument in this thesis is that even if rice agriculture contributed to the emergence of urbanism and state formation in the region, other food-producing strategies and crops cultivation, were already employed prior to its arrival. This implies that both internal
12

and external factors contributed to the subsequent rise of urbanism and state formation in the region. The indigenous food-producing communities had also developed certain cultivation strategies, which were beneficial to later rice cultivation communities. Notwithstanding, hunting and gathering did continue and has done so until today. Therefore, these communities must also have had an impact on the development of the region, considering that contact and interaction between forager and farmer groups surely occurred.

13

14

1. RESEARCH BACKGROUND

1.1 Southeast Asia

Mainland Southeast Asia usually refers to Burma (Myanmar), Laos, Vietnam, Cambodia, Thailand and (peninsular) Malaysia although some scholars, e.g., Higham (1996) incorporates southern China into the same geographical classification. To the north, hill ranges divide Southeast Asia from China, spreading south and subdividing the region into geographically delimited areas, many of which constitute contemporary national boundaries. Complex societies rose in regions which witnessed frequent contact and trade between people of contrasting cultural backgrounds from different environments. Great river valleys flowing into the wet-dry boundaries provided ideal environments for such encounters (Yasuda, 2004: 15). People have settled along the Red, Mekong, and Chao Phraya Rivers for thousands of years and also today (e.g. see Higham, 1989, 1996, 2005). Southeast Asia is one of the most fertile and bio-diverse areas of the world (Baker & Phongpaichit, 2005: 1-2; Higham, 1989: 3; Tougard, 2001: 337). The tropical and subtropical temperatures and moisture regimes create a particularly abundant environment, which is directly affected by the southwest monsoons and the northwest monsoons which cause the dry season (Yasuda: 2004: 12). The proximity to the equator means that high annual humidity, precipitation and temperatures prevail all year round. Some areas also maintain micro climatic variations. Individual climate characteristics on macro and micro scale are determined by high or low altitude and inland or littoral location (Strahler & Strahler, 2005; Horton et al. 2005: 1200). Generally, the lowlands in northern Southeast Asia have a wet equatorial climate with stable abundant precipitation, whereas the southern parts have a monsoon and trade-wind influenced coastal climate, with abundant rainfall and a stronger seasonal pattern (Strahler & Strahler, 2005: 258-61). In the north vegetation, generally consists of deciduous thick forest, while tropical rainforest dominates further south with dense mangroves along the coast (Baker & Phongpaichit, 2005). Today, much of the lowlands are cultivated and the highlands are to a larger extent still forested (Higham, 1989: 5-14). The research area comprises Surathani, Nakhon Si Thammarat, Trang, Phattalung, Satun and Songkhla provinces (fig 1.1.) but the archaeological survey was carried out in Songkhla and Phattalung provinces. The southern provinces of peninsular Thailand, namely Pattani,
15

Yala and Narathiwat, are excluded from the study as political instability has hindered archaeologists from carrying out fieldwork in the region during the past few decades.

Figure 1.1. Map of Southeast Asia showing the research area.

1.2 Previous research and general trends

Prehistory in Southeast Asia is relatively novel area of research (Higham, 1996: 3) for which there are several reasons. For instance, the wide spread of Buddhism in region has played part in the lack of prehistoric studies as the inception of Buddhism in historic times used to be the focal point of research. Thus, archaeological research primarily focused on the origins of Buddhism, art, architecture and artifacts (Karlstrm, 2009: 20). From the 1960s and onwards, however, prehistoric archaeology has advanced significantly (Bayard, 1993). Ban Chiang, a prehistoric site on the Khorat Plateau, northeast Thailand, which was discovered in the 1960s
16

(Higham: 2002a: 24), marked the beginning of prehistoric archaeology in the region. Since then, research on the Khorat Plateau and Chao Phraya basin in central Thailand has been favored (e.g. see Higham, 1989, 1996, 2002a). Subsequently, more sites from the same time period, such as Non Nok Tha, Ban Na Di and Chansen, have been found, dating back to c. 5000 2500 BP (Higham, 1989, Higham, 1996: 2). The excavations resulted in an increased emphasis on social organization (e.g. Bayard, 1992; Bronson, 1985, 1992; Glover, 1991; Glover et al. 1992; Higham & Thosarat, 1998; White, 1982, 1995; White & Pigott, 1996) and the organizational changes which occurred from approximately 500 BC to 500 AD. The changes which took place during this period are believed to have led to the formation of the earliest states of the region along the coast and inland river valleys (Stark & Allen, 1998; Stark, 2006: 408). These changes have usually been attributed irrigated rice agriculture and trade (for example see Allen, 1997; Bentley, 1986; Glover, 1990; Manguin, 2000; Stargardt, 1973, 1976, 1977, 1983, 1992). Nonetheless, the Stone Age in the region remains relatively unknown. What is known about this long cultural sequence is based on material culture remains from cave and rock shelter sites. Cave archaeology has shown a rich Pleistocene/Palaeolithic and late Pleistocene to mid Holocene/Hoabinhian human past in the region. Major sites include Sakai Cave, Spirit Cave, Banyan Valley Cave, Lang Rongrien Rockshelter and the Ban Kao Cave sites. It has been proposed that pre-agrarian communities in the late Pleistocene to the mid-late Holocene occupied both interior and coastal tracts (Higham, 2002b: 228-30). Late and postPleistocene hunter-gatherer settlement patterns have been reviewed by Shoocongdej (2000), who emphasizes the role of foraging strategies, mobility and settlement organization in the analyses of late Pleistocene and early Holocene cave sites in western Thailand. Shoocongdej (2000) has proposed a model of a dual mobility organizational system in seasonal tropical environments. She argues for residential mobility in the wet season and logistical mobility as an organizational response to the dry season. Stone Age archaeology has also often been interpreted in relation to linguistics and the modern distribution of ethnic groups in Southeast Asia. The general theory is that AAspeaking farming communities migrated from their native homeland in southern China (Bellwood, 1997: 28687, Diamond & Bellwood, 2005; Bellwood, 2006: 112). According to Higham (2002b: 231), this migration of peoples led to cultural and socio-economic assimilation of hunter-gatherer groups by AA farming communities. Similarly, the early (AA-speaking) farmers were later overtaken by other intrusive groups, such as the Thais (Austro-Thai-

17

languages), the Chams (Austronesian), the Burmese (Sino-Tibetan) and Indo-Aryan languages (ibid, also see Bellwood, 2005). Some of the fundamental issues discussed in contemporary archaeological research are the chronology of the Neolithic and Bronze Age and the origins of rice agriculture (Bellwood, 1997; Bayard 1972, 1979; Higham, 1989, 1996; Solheim 1968). However, it is now generally accepted, that rice derived from Yangtze Valley, southern China, in the second millennium BC (Higham, 2005, C. Higham & T. Higham, 2009; Liu et al. 2007). It has even been proposed that communities in the Yangtze River Valley domesticated rice and produced pottery as early as 20, 000 BP (Yasuda, 2002). Recent excavations at Ban Non Wat, a Late Stone Age site with long cultural continuity from c. 1750 BC 500 AD, have added substantial knowledge to the issue. The earliest introduction of rice cultivation has previously been argued to have taken place c. 2300 BC among inland and coastal groups of hunter-gatherers in central and northeast Thailand (Higham, 1996: 6; Higham, 2005a: 250-51). A more recent set of radiocarbon dates from Ban Non Wat suggest a later date for the introduction of rice cultivation as the samples have been dated to c. 1650 BC (C. Higham & T. Higham, 2009: 134). Khok Phanom Di (4000-3500 BP) in central Thailand is a key site in this period. Although hunting and gathering and fishing was practiced at the site, rice was present both in cultivated and wild form. One of the key questions has been whether hunter-gatherers at the site actively engaged in rice cultivation. It has been proposed that foragers traded cultivated rice from inland agricultural groups (Higham & Bannanurag, 1990, 1991; Higham & Thosarat, 1993, 1994; Higham, 1996, Tayles, 1999, also see Halcrow et al. 2008). Yet, it has also been suggested that forager groups at least periodically cultivated rice (Higham, 2005: 254-55). The aforementioned site, together with Ban Chiang, Non Kao Noi, Ban Non Wat, Ban Lom Khao in the Mekong catchment, and Ban Kao, Non Pai Wai and Ban Tha Kae in the Chao Phraya River Area (see Higham, 2002b: 228) are among the only open air Stone Age sites found in the region. Therefore, more field research on open air sites is urgently needed.

The Thai-Malay Peninsula

Research on the prehistory of peninsular Thailand is synonymous with cave research, which begun at the end of the 1980s and still lags behind in other parts of Southeast Asia (Anderson, 2005: 137). More information on the archaeology of the late Pleistocene and subsequent periods has only recently become available (e.g. see Anderson, 1986, 1988, 1990, 1997, 2005;
18

Matsamura & Pookajorn, 2005; Mudar & Anderson, 2007; Pookajorn, 1991, 1994, 1996). Excavations at the Long Rongrien Rockshelter site in Krabi (fig 1.2.), southwestern peninsular Thailand (Anderson, 1988, 1990, 1997, 2005), and at one of the sites in Perak in northern peninsular Malaysia (Zuraina, 1994, 2005, Matsamura & Zuraina, 1999) have revealed that the earliest sequences of human occupation in the region date back to the late Pleistocene (< 40, 000 BP). A continuation of occupation in the region is evident from contexts dated to c. 26, 000 BP at the sites of Moh Khiew I and Sakai Cave in Krabi and Trang Provinces, respectively (Albrecht et al. 1993; Pookajorn, 1991, 1994, 1996). During the period between c. 22, 000 14, 000 BP, there is a gap in the archaeological record. This has been attributed to rising sea levels because it has been assumed that some people resided in caves/rock shelters in coastal tracts which today lie under water (Auetrakulvit, 2005: 50-1; Srisuchat, 2005: 33, also see Higham, 2002b).

Figure 1.2. Lang Rongrien Rockshelter after Anderson 1990: 2.

The archaeological data of the Thai-Malay Peninsula from 12, 000 5000 years BP primarily derives from cave and rock shelter sites and is attributed to the makers of the Hoabinhian pebble tool industry (Adi, 2005: 46). In peninsular Thailand, these sites are Lang Rongrien and Khiew II in Krabi Province and Thung Nong Nien and La Sawang in Satun Province (Auetrakulvit, 2005: 50-62). In peninsular Malaysia, the sites of Gua Sagu, Gua Kechil and
19

Gua Peraling in Pahang, Gua Taat in Terengganu and Gua Cha, in Ulu Kelantan are associated with the same tradition (Adi, 2005: 47). Many of the sites have a continuous, long-term archaeological record, spanning from the late Pleistocene to the Holocene, although some breaks occur in the assemblages. The only Stone Age open sites found in the peninsula are the now destroyed shell middens in Malaysia, which have been associated with Hoabinhian cultural remains (Adi, 1983: 53-4). The eldest open settlement found and excavated in the upper parts of peninsular Thailand is Khao Sam Kaeo, which has been dated to the midlate of the first millennium BC (Murillo-Barroso et al. 2010). Several ethnographic and ethnoarchaeological studies have been conducted so as to understand past hunter-gatherer ways of life in this part of the world. Hunter-gatherer populations have persisted in parts of Southeast Asia until the present. In north Thailand, the Phi Tong Luang (Pookajorn, 1985: 208-9), the Semang (or the Sakai/Negrito) in the forested areas of Trang, Satun and Phattalung provinces in peninsular Thailand (Albrecht & Moser, 1996: 1-2; also see Pookajorn, 1991) and the Orang Asli (or the Semai/Senoi) living in peninsular Malaysia (Zuraina, 2003: 102, also see Bellwood, 1993) are known to have practiced hunting and gathering until recently. DNA analyses have also been conducted on human skeletal remains found at the sites of Moh Khiew Cave and Sakai Cave, ranging in date from the late (terminal) Pleistocene to early-mid Holocene. The results of the analyses have shown genetic similarities between past and present hunter-gatherers (Oota et al. 2001).

The issues of definitions

Chia has argued that the most pressing issues to be solved when it comes to research on the prehistory of the Thai-Malay Peninsula include (i) establishment of regional culture sequences, (ii) clear definitions of archaeological periods and cultures and (iii) the classification of archaeological artifacts (2005: 80-1). The present author agrees with Chias formulation of problems needed solving. The lack of a fundamental understanding of the archaeological record has created a situation of cross-usage of traditional terms and concepts of the Neolithic and Bronze Age, which is not grounded in sound empirical data. The current methodology also neglects the possibility of cultural variation as local assemblages may deviate from the established chronological cultural phases. Polished stone tools and pottery are assumed to be part of Neolithic material culture though very few undisturbed caves and rock shelters attributed to the Neolithic have been found and verified by dates, and when it comes to open air sites, no Neolithic sites have been found. The phase (or the cultural-sequence) of the so20

called Neolithic is only present in the material record, which is dated by correlation with material culture from central and northeast Thailand and southern China. Thus the Neolithic is not a well defined term and has within it an embedded circularity of reasoning. The confusion among scholars in aspects of chronology, terms and conceptions of the major phases in human history in the region is therefore profound. This confusion extends to the discussion on the origins of agriculture in the region, which is the main topic of this thesis. Based on the above, it can be argued that archaeological research in this region has been biased. However, in my opinion, it is rather explained by the lack of archaeological data available. In this situation of paucity of data, archaeologists have tried to explain the prehistory of peninsular Thailand by means of comparative archaeology and clearly distinguishable chronological ages. The material culture recorded in the area is similar to artifacts found in core areas in Southeast Asia, and as a result scholars have argued for similar cultural trajectories. The problem is that the use of the term Neolithic hinders discussion on alternative processes of development in the same time period. At the same time, it is usually used liberally causing confusion about its actual meaning. For central and northeast Thailand C. Higham and T. Higham define the Neolithic as: a context by mortuary or occupation remains with domestic animals or plants but no evidence for metallurgy. (2009: 127-8). In southern Thailand, flaked implements associated with Hoabinhian hunter-gatherers and edge-polished stone tools and ceramic fragments have been found in the same cultural layer in a mid Holocene context. The subsequent late Holocene is characterized by pottery-using and presumably agricultural groups (White et al. 2004: 127-8). Overlapping in the archaeological record is also evident from excavations in cave sites in Surathani Province (Chiew Lam Project, 1986). This shows that clearly defined cultural sequences based on the archaeological evidence is difficult to attest. Criticism against this concept also exists elsewhere. Denham (2004: 611) has questioned the use of Neolithic packages for New Guinea and island Melanesia. In Africa south of the equator, Smith (2005) argues that the concept of the Neolithic is problematic because it makes assumptions and reflects how the archaeological record is interpreted. Considering the implications the definition Neolithic generates, the expressions: Late Stone Age or the climatic time unit midlate Holocene are used in this thesis. This terminology is preferred as it is not loaded with the same amount of preconceptions as the use of the term Neolithic The term Hoabinhian is also problematic but it is well established in Southeast Asian archaeology. The Hoabinhian chronological phase (first proposed at the Congress of the Far Eastern Archaeologists in 1932) refers to the end of the Pleistocene and Holocene eras, span21

ning the time period of 12000 - 5000 years ago. Adi (2005) has briefly presented the shortcomings of the usage of the term Hoabinhian. Despite his criticism, he maintains that there is no other suitable term to replace it. He uses Hoabinhian to refer to: ... a specific industry, characterized by a technique of knapping pebbles bifacially to produce morphologically standardized implements (Adi, 2005: 46-7, also see Shoocongdej, 2000: 14, on the same topic). Here, the expressions late Pleistocene and Holocene communities are used in connection or in place of term Hoabinhian. Using this particular terminology has the advantage that it does not preclude other possibilities of alternative cultural trajectories or elements which are usually associated with the term Hoabinhian. Climatic time units are important because they do not necessarily follow accepted truths of human behaviour during a given period in time. For example, the term Hoabinhian is usually associated with hunter-gatherer societies. This is important because in this thesis it will be argued that early food-producing communities probably emerged in the late Pleistocene and early Holocene and further developed their foodproducing strategies in the later Holocene periods. Archaeological remains of domesticated animals dating back to the firsts millennium BC have been found in both mainland and insular Southeast Asia. The remains have been associated with local agriculturalists. However, there is no archaeological evidence for animal husbandry in peninsular Thailand. Therefore, animal husbandry is not discussed in this thesis. Despite the apparent lack of evidence, it is likely that Late Stone Age communities kept animals. Notwithstanding, I believe food production or the cultivation of plants began long before domesticated animals.

1.3 The archaeological and environmental evidence

Archaeological remains attributed to the Stone Age have been recovered from around 50 sites in peninsular Thailand (Srisuchat, 1987, 1997 cited in Anderson, 2005: 144) and about 100 sites in peninsular Malaysia (Adi, 2005: 47). The majority of these are surface finds in cave and rock shelter contexts recovered during archaeological surveying. A common characteristic of almost all sites is that they have been heavily disturbed by local farmers, who use the soil found at these sites as a fertilizer (ibid, also Lekenvall, pers. obs.). The archaeological evidence solely consists of stone implements, pottery and rock paintings. Environmental data comes from pollen and phytolith samples which point towards anthropogenic related changes in land-use during the late Pleistocene, which further intensified in the later Holocene periods
22

(Kealhofer, 2002, 2003; White et al. 2004). Because land-use is related to past and present day subsistence regimes, proxy data is fundamental for the understanding of emerging foodproducing communities.

Prime indicator (1): stone implements, pottery and rock paintings

The Stone Age finds were mainly recovered during previous archaeological survey and excavations carried out in the beginning of the 20th century. At that time not much was known about site stratigraphy and no absolute dating techniques existed (Chia, 2005: 81). Exceptions to this are the sites Lang Rongrien and Moh Khiew. The stratigraphic contexts of the sites were predominantly characterized by Late Pleistocene mid Holocene/Hoabinhian remains, although Lang Rongrien Rockshelter in Krabi Province shows a long historical continuity from c. 43000 BP 3500 BP (Anderson, 1997: 607). More recent Stone Age contexts have been excavated in Surathani and Songkhla Provinces. In the first half of the 1980s archaeological fieldwork was carried out as part of a rescue project in connection with the construction of a dam in the area of Phum Duang River in Surathani. Around 60 sites were discovered and three caves, namely Tham Pak Om, Tham Bueng Bab and Khao Kichan were excavated. Shell beads, potsherds (of various types) stone tools (made of volcanic basalt, quartzite and quarts); flaked tools, chopping tools, bark beaters and polished stone axes were found in the caves. These included birds, shells, crabs, fish, turtle and plant seeds. The cave assemblages differ to some extent from one another in terms of the choice of stone technologies and raw materials. Tham Bueng Bab was the only cave which showed evidence of burial practices. Something striking is that Hoabinhian and Late Stone Age stone implements and the remains of a tripod are reported to have been found in the same cultural layer. In total, 38 samples from shells, animal bones and charcoal were radiocarbon dated of which 28 could be used. The chronological sequence was set between 6500 4200 BP and defined as Hoabinhian/Neolithic (Chiew Lam Project, 1986). Khao Rakian Cave in Songkhla was excavated in 1986 by the 13th Fine Arts Department, Songkhla. The excavation revealed vast amounts of potsherds. These included rim, neck shoulder and body fragments with sandy temper with strong brown, and yellowish-red and dark-gray color. They had been decorated with plain or cord-marked design and were between 0.4 1 cm thick. Some sherds had the characteristic style of Late Stone Age pottery found elsewhere. The thinner more fragile pottery sherds found were thought to have been fired in
23

lower firing temperatures. The remains of a pottery tripod were also recovered, along with human skeletal remains indicating burial activities. On the basis of the potsherds and tripod the site was considered Neolithic. However, when the cave was excavated it had already been disturbed by local people (Archaeological survey from Khao Rakian, 1986). In 1987 1988 the Fine Arts Department carried out surveys in the western littoral areas of Phangnga, Krabi, Trang and Satun provinces. In the mountains surrounded by mangrove forests, survey of caves and rock shelters revealed rock paintings and archaeological surface finds. In Phangnga and Krabi nine rock painting sites were recorded with motifs of human, animal and abstract designs in red, yellow, brown and black color, suggesting the usage of ochre and charcoal. It was noted that that the usage of the sites in the area seems to have been different depending on the geographical setting. Sites with rock paintings were situated on islands and only one site revealed surface finds in the form of flakes, polished stone axes and potsherds. Conversely, no rock paintings were recovered from inland sites, but an extensive distribution of surface finds including hammer stones, grinding stones, potsherds, animal bones and shells was recorded. This lead to the belief that the coastal/island rock painting sites were used temporary and/or as ceremonial sites whereas the inland areas were habited more frequently. The dates of the sites have not determined but they are considered prehistoric (Chaimongkol, 2005: 90-4). In addition, prehistoric rock paintings have been found in Yala Province (Limwijitwong, 2005: 97), Kanchaburi Province and peninsular Malaysia (Adi, 2005). Moreover, several sites in southern Thailand with similar archaeological remains exist but the specific characteristics of the remains the context of their recovery are often blurred. In the Satingphra catchment at the western edges of the Songkhla Lakes, Stargardt (1983: 4, 2003: 104) ((presumably referring to Phattalung Province), has reported secondary burials with skulls and long bones in rock shelters and caves. Bones have been buried together with polished stone adzes and skulls. The skulls did show traces of painting with haematite. From caves and rock shelters (Khao Khanab Nam, Na Ching and Tham Phi Huato) in Trang Province, White et al (2004) have reported the presence of pottery ground stone axes and adzes, which may date between 5000 - 6000 BP. However, neither Stargardt (1983, 2003) nor White et al. (2004) provide details of the provenance of these finds. On that basis, it is possible that this information was obtained from local people in the same way as the author gained access to similar knowledge (see chapter 3 in this thesis). The archaeological reports from the Chiew Lam Project and Khao Rakian Cave have not, as far as I know, been previously published. Parts of the reports were translated by Siriporn
24

Limwijitwong from the 13th Fine Arts Department during a joint archaeological field project in 2009. Combined with the older data from peninsular Malaysia and from Lang Rongrien Rockshelter, these new sources of data further point to a rich late Pleistocene to midHolocene/Hoabinhian and mid-late Holocene/Late Stone Age cultural past in peninsular Thailand. The remains of pottery and polished stone tools are particularly important since these items are usually associated with early farming communities. Finding mixed assemblages consisting of objects usually associated with either mid-Holocene/Hoabinhian or mid-late Holocene/Late Stone Age together in the same cultural layer at one of the sites in Surathani is interesting. The contemporariness of the assemblages implies that human groups with a diverse subsistence economy and material culture lived side by side during an extended period of time. At the same time, the radiocarbon dates spread between 6500 4200 BP, which corresponds rather well with an earlier Hoabinhian cultural phase, prior to the arrival of the supposed immigrating agriculturalists in second millennium BC. The rock paintings further address the need of absolute dating methods. There are several examples of rock paintings which clearly were made by the contemporary Semang (e.g. see Adi, 2005: 107). Even if many of them are considered prehistoric, their actual dates remain unknown. Most rock paintings in the world were created by hunter-gatherer communities or early farming communities. Judged by the artifact scatter in association with the rock paintings in this area, they could have been made by early farmers. Recently developed direct radiocarbon dating by accelerator mass spectrometry can solve this problem by sampling the pigment, instead of relying on the archaeological remains in the vicinity for relative dating (see Valladas, 2003, on this subject). There also seems to be an important association between the initiation of burial practices and Late Stone Age material culture. The archaeological evidence from Tham Bueng Bab Cave, Khao Rakian Cave and the cave or rock shelter site in Phattalung Province reported by Stargardt, suggests that communities started to bury their deceased in caves/rock shelters in the Late Stone Age. This has also been proposed by Anderson on the basis of burials associated with pottery in the cultural layer from mid-late Holocene in Lang Rongrien Rockshelter (Anderson, 1990, 1997, 2005). Anderson has taken this even further and suggests that the beginning of burials in Lang Rongrien Rockshelter may be related to a shift from a huntergatherer economy to one based on farming (1997). In order to explore this further, proxy data is needed, because environmental data provide the means to study the wider, ecological context for long-term changes in peoples economy and land-use (Kealhofer, 2003: 74).
25

Prime indicator (2): pollen and phytoliths

In terms of the environmental record, pollen and phytolith samples from southern peninsular Thailand suggest that the late Pleistocene environment was dominated by grasses or woodland savannah which transformed during the early Holocene by an expansion of lowland forests. Late Pleistocene woodland savannah is argued to have supported a wide range of herbivores. The expansion of forest vegetation reduced grass or appropriate ground-level foods. During the early Holocene, an increase in plant foods typical of lowland forests resulted in high diversity and low concentrations of species (Hutterer, 1983, cited in Kealhofer 2002: 184-5). These profound changes in environment and animal life must have had a substantial effect on the late Pleistocene and early Holocene populations. The climatic changes affected the distribution of food resources. Kealhofer (2002, 2003) argues for the development of alternative food-producing strategies among hunter-gatherers as a result of this environmental stress. Forest-burning by human agencies is evident around Thailand and south Thailand in the late Pleistocene (White et al. 2004: 124). Based on pollen and phytolith retrieved from sediment contexts in southern Thailand, Kealhofer (2002, 2003) and White et al (2004: 124), stress that people began to manage the landscape in the late Pleistocene and early Holocene through extensive burning, as evidenced from the environmental record. Pollen and phytolith samples extracted from Lake Thalee Song Hong in Krabi Province show a rapid forest expansion in the region from c. 8000 6500 BP and also a continuation of burning. The phytolith sequence from the mid seventh millennium BC indicates intensifying or diversifying land-use, which together with the presence of economic species in the pollen samples have been considered signs of intentional cultivation and unintentional disturbance factors (Kealhofer, 2003: 87). However, the most substantial evidence of burning comes from the early Holocene and after 6500 BP, particularly around 3500 BP, during which grass burns appear to have been intensified. Furthermore, burnt wood becomes prominent in the archaeological record c. 2700 BP (Kealhofer, 2003: 80-1). The latter peak seems to correspond in time to the findings of Horton et al. (2005) and Stargardt (1983), who have found evidence for burning in the Satingphra catchment in late prehistoric times prior to urban Satingphra. Economic plants, such as palms, banana, and rice, appear by 5000 BP in the phytolith record, and possibly much earlier in the pollen record. This suggests that Holocene groups in tropical rainforests may have practised arboriculture, horticulture, hunting and gathering in an economy which Kealhofer describes in the following way: could not be considered
26

pure hunting and gathering . . . nor conventionally agricultural. (Kealhofer, 2003: 87). According to Kealhofer, the arboreal diversity between 8000 4000 BP supports an interpretation of forest management (Kealhofer, 2003: 86). She argues that hunter-gatherers dealt with the loss of savannah woodland (and animals and species) in the late Pleistocene by attempting to maintain forest gaps in order to concentrate food resources (Kealhofer, 2002: 186). Indeed disturbance taxa and indicators of slash-and-burn agriculture increase dramatically after 4000 BP (Kealhofer, 2003: 88). Kealhofer (2002: 186) argues that the development of agro-ecosystems is easy to identify as they are large-scale regional transformations of vegetation and soils. Changes in soils and vegetation result in identifiable patterns and signatures in agro-ecosystems, which include deforestation, domesticated plants, weeds and regrowth taxa in the vegetation sequence. The environmental data clearly shows that humans profoundly changed the landscape during the late Pleistocene and continued to do so in the Holocene. From the late Pleistocene and onwards, the intensified land-use may be related to the shift in stone technology from smaller flakes to heavy Hoabinhian woodworking sort of implements (see Solheim: 2006: 132). Does this mean that communities engaged in food-production as early as the late Pleistocene or beginning of the Holocene? Considering the amount of evidence visible in the phytolith and pollen record, this interpretation is feasible. However, as noted by Kealhofer (2002) communities probably employed a diverse economy, and were involved in practices of both food-gathering and production. The increase in land-use in the mid-late Holocene, witnessed from disturbance taxa and indicators of slash-and-burn agriculture, is also of interest. The date 4000 BP corresponds rather well to the assumed earliest date for burials at the cave and rock shelter sites and Andersons hypothesis of burials being associated with early farming, bearing in mind that the expansion of agriculturalists into Southeast Asia is believed to have taken place in the second millennia BC. Judging from the chronology of these phenomena, they appear to be related. Notwithstanding, the presented environmental data from Kealhofer points to an earlier development of food-production. In my opinion, this data reflects an earlier phase of an indigenously developed set of food-production strategies, which was supplemented by other food-producing strategies and foreign domesticated plant seeds at a later stage. This approach is different from the conventional view of agricultural development in the region and will be discussed in the next section.

27

1.4 Defining the problem

All current available archaeological data indicates that rice cultivation and domestication originated in the Yangtze River Valley in southern China. Phytoliths of semi-domesticated rice have been found at the Xianrendong and Diaotonghuan cave sites, and the dated samples predate 15, 000 BP (Yasuda, 2002). The appearance of settled villages during the Late Stone Age has been attributed to the introduction of this important plant seed. On this basis, the conventional view of agriculturalists expanding southwards in the second millennium BC is plausible, and similarities in material culture, linguistics and genetics also support such a hypothesis. That rice agriculture eventually reached northeast and central Thailand from China is also evident from cultural layers in Ban Non Wat. A great number of radiocarbon dates from this site implies that rice agriculture began around 1650 BC (C. Higham & T. Higham, 2009: 134). For the Thai-Malay Peninsula, archaeological evidence of` rice agriculture is far later than that found in Ban Non Wat. In southern Thailand, rice agriculture is argued to have been introduced approximately 2000 years ago in the Satingphra catchment (Stargardt, 1973, 1983), and similar dates has been proposed on the basis of archaeological evidence from Kedah in peninsular Malaysia (Allen, 1997, 2000). There is more recent data which might support an early presence of domesticated rice in peninsular Thailand. The petrographic analysis of pottery found in a Late Stone Age context from the site of Khao Lamu Cave in Satun Province revealed remains of rice husks in the pottery. Using a polarizing microscope on three samples showed that the earthenware consisted of rice husks of various shapes, which indicates that the rice from these samples had undergone domestication or semi-domestication (Asawamas, 2008: 39, 62, 72). However, no radiocarbon dates for the samples are yet available. That wild rice, semi-domesticated, or domesticated rice have been found does not tell us much without radiocarbon dates since the Late Stone Age material culture persisted well beyond the second millennium BC. Nevertheless, the mere presence of rice at these sites supports Kealhofers (2003) study, which, based on the presence of wild staples in phytolith and pollen samples, suggests that rice might have been domesticated in the second millennia BC. On the other hand, an early date of 4000 BP is questionable because it precedes the radiocarbon dates obtained from the rice farming context in Ban Non Wat. Besides these dates, there is limited evidence for rice cultivation in the Thai-Malay Peninsula in the Stone Age (Chia, 2005: 81). The intensified cultivation of plant species visible in the environmental record does not appear to have included rice or only to a small degree (Kealhofer, 2003: 88).
28

A similar developmental sequence for the Late Stone Age has been assumed for entire Southeast Asia, which usually is seen in relation to settlement aggregation and rice agriculture in Neolithic China. In this context, the hypothesis of expanding agriculturalists implicitly stresses that indigenous communities were hunters and gatherers prior to the influence of an agricultural based economy practiced by more complex societies. The sparse evidence of rice agriculture in some areas has typically been attributed to extended periods of huntergatherer occupations or simply a lack of sufficient archaeological research in the region. In my mind, such a view is too narrow as it disregards the existence of alternative foodproducing strategies, which may not be related to rice. Consequently, alternative models of food-production have been set aside. Kealhofer writes: The emphasis on the origins of rice domestication in Southeast Asia has been at the expense of a wider investigation of the development of food production strategies. (2003: 72). Kealhofer (2002: 179-80) has also criticized Higham (c.f. 1996) for not separating initial domestication from agro-ecosystems. Kealhofer argues that Highams explanation of rice agriculture as a gradual expansion of rice farmers implies that domestication and agriculture are one and the same process. This critique is well-founded because Higham does not acknowledge the possibility of communities having been involved in food-production prior to the introduction of rice. However, Higham has mainly focused on rice agriculture and its subsequent importance for urbanism and state development, and is therefore working within another framework and with another set of aims. Considering the often intersected usage of the terms cultivation, domestication and agriculture, it is important to clarify the meaning of the concepts used here. Domestication usually refers to the biological process in plant and animal genotypes in which the reproduction of the physical character becomes dependent on humans. The process of domestication can also often be unintentional, resulting from reoccurring interactions between humans and wild species. Cultivation refers to the intentional preparation of fields by humans; sowing, harvesting, storing seeds and other plants. Cultivation requires advanced technologies as well as changes in subsistence and socio-economic perspectives. Agriculture is the result of domestication and cultivation. It involves profound changes in the human use of the environment, such as extensive clearance of forest. This is usually followed by transitions in the socio-political structures and organization of societies. Agriculture is usually associated with the emergence of hierarchical structures in human societies and control over the agricultural surplus. Agriculture is followed by advances in technology, such as plows, field systems and irrigation. Combined,

29

these factors lead to a larger population and a more complex socio- political organization and settlement aggregation (Scarre, 2005: 183). It is important to note here that cultivation precedes agriculture and need not be related to domesticated crops. This is made clearer with Harris (1996: 4) classification of the concepts within an evolutionary scheme (table 1.1), which ranges from an initial stage of gathering and collecting to small-scale cultivation and large-scale wild plant production. Domesticated plant-crops are introduced in the final stage in connection with the introduction of agriculture. As noted by Harris (ibid) the scheme perhaps oversimplifies the very complex and diverse nature of plant exploitation but it also clarifies the terms and circumstances related to the different food-production phases/strategies.
Table 1.1. Evolutionary scheme of plant-food production adapted from Harris 1996: 4.

However, Spriggs (1996) understanding of the different food-production strategies used in the Pacific is the most suitable framework for the purpose of this thesis. Firstly, he makes a distinction between simple and complex foraging. Spriggs stresses that complex foragers were perhaps involved in burning of vegetation so as to open up forests and encourage the growth of certain plants. This strategy may also have included tending or nurturing of plants by weeding around them. Wild food-production is an extension of complex foraging. It involves replacement planting of wild tubers or vines, transplanting useful species such as nut trees, selective weeding or clearing and perhaps even simple forms of irrigation and drainage. Furthermore, Spriggs has criticized Harris for placing domestication as the threshold between cultivation and agriculture (1996: 525), simply because many plants used in food-production in the Pacific even today not are fully domesticated (Yen 1985: 323, cited in Spriggs, ibid).
30

This is important because it implies that domestication is not as relevant for food-production as one may think. It could also mean that the morphological evidence of domestication in the plants might be absent or weak. In the same manner as Spriggs (ibid), I treat the term cultivation as not significantly different from wild plant-food production. Therefore, the beginning of wild food-production is the key issue here as opposed to domestication. Everything thus far indicates that the start of full-scale agriculture in Southeast Asia occurred in the second millennium BC and had a fundamental impact on the region. Agriculture is argued to have given rise to settlement aggregation and more complex societies. However, while Spriggs stresses that the initial wild food-production strategies did not contribute to the overall development of agriculture in the region, I am inclined to argue the opposite. In my opinion, a total replacement of a culture (and a food-related economic system) can never be fully carried out. Cultures are constantly changing and when subjected to external stimuli, they evolve into new entities with customs from both cultures. In other words, an intermarriage of cultures takes place. Considering that the indigenous community engaged in food-production early on, it is not odd that they might have adopted domesticates, such as rice, at a later date. The overemphasis on the role of domestication in food-production is also illustrated by the method of ratooning. Hill (2009) has acknowledged that this simple method of cultivation might have been a possible step prior to the adoption of rice agriculture. The method consists of repeated rounds of tillage, seed-time, planting, growth and harvest. The main point of the strategy is that rice (oryza sativa) is a perennial grass. In other words, the plant will continue to form shoots after the initial harvest as long as it is supplemented with enough sun and water. Therefore, no replanting is necessary (Hill, 2009: 1). The advantage of this method is that it must have demanded less human labor input and less advanced irrigation and drainage technology. For this reason, it would therefore be a small-scale plant food-production industry. However, finding evidence for this kind of food-production would be difficult since the rice does not develop morphologically distinguishable traits as grains do. The method ratooning is similar to the food-production strategies suggested by Spriggs (1996) for the Pacific. Despite the fact that communities probably did engage in activities which are conventionally not associated with a forager economy, a connection between the Thai-Malay Peninsula and the rest of Mainland Southeast Asia is evident. In addition to linguistics, similarities in material culture between the Thai-Malay Peninsula and the rest of Southeast Asia indicates a cultural bound. The most obvious similarities are Late Stone Age pottery and tripods found all over Southeast Asia, the Thai-Malay Peninsula included. Whether this link is the result of
31

agriculturalist immigration, trade/exchange or indigenous developments, remains to be solved. Compositional studies of tripods found in peninsular Thailand have revealed that they were made locally (Srisuchat: 2003: 249) but this only shows that the technology was present in the area. The actual identity of the makers and users of tripods, whether foragers or colonizing groups, remains unknown. Moreover, pottery similar to that found in Ban Kao has also been excavated from Lang Rongrien Rockshelter (Bellwood, 1993: 47), which further indicates a close cultural relationship between the areas. Pottery found in different areas of Southeast Asia should undergo more comparative analyses. The Late Stone Age pottery containing tempered rice husks recovered by Asawamas (2008) in Satun Province could, for instance, perhaps be more accurately dated by extracting charred organic remains from the ware (e.g. see Klln, 2004: 85 on this topic). As argued here the discussion should not revolve around domesticated rice as a staple food source and prime instigator (along with trade) for the subsequent development of urban societies, despite its assumed importance. Rather, I propose that this plant seed should be seen as an important factor and stimulus for later urbanism and state development. Prior to the supposed immigration of agriculturalists as proposed by most scholars (e.g. Bellwood, 1995, 1997; Higham, 1996; Higham & Lu, 1998; Higham, 2005a; C. Higham & T. Higham, 2009), food-production was already practiced. From this viewpoint, my argument is that we should consider both internal and external factors as responsible for the development of agriculture and, indirectly, for later developing urbanism and state formation. In my opinion, there is clear evidence for an expansion of rice agriculturalists in the second millennium BC. The problem is that the possible existence of earlier food-production strategies which are not related to rice has been overlooked. The negligence of earlier strategies is partly due to ethnography/ethnoarchaeology and research on human prehistoric DNA. Because hunter-gatherer populations have persisted in parts of Southeast Asia until the present, the relationship between past and contemporary hunter-gatherers has received a lot of attention from scholars (c.f. Albrecht & Moser, 1996; Pookajorn, 1985, 1991). This has even extended to research on the DNA of late Pleistocene and early-mid Holocene human skeletal remains, which have been found from the Moh Khiew Cave and Sakai Cave. Oota et al (2001) stress that Mitochondrial DNA phylogenetic relationships and the ethnographi-

cal/ethnoarchaeological research combined show a genetic and cultural link between these prehistoric humans and the present Semang. This ancestry has also been proposed by Bellwood: the Semang are an indigenous population of Peninsular Malaysia with a local ancestry dating from at least 10, 000 years B. P (1993: 54). Considering that the Semang are
32

contemporary hunter-gatherers, a similar economy has therefore also been assumed for Stone Age communities. The problem with this approach is that it only indicates a genetic link, but not necessarily a cultural one. For instance, a lithic culture and pottery are absent among contemporary Semang (Albrecht & Moser, 1996: 4-8). Instead, their material culture consists of implements and containers of wood (fig 1.3.). Communities are always subjected to a variety of interacting factors resulting in societal change, and it cannot be assumed that contemporary hunter-gathers and past ones share an identical socio-economic structure.

Figure 1.3. Material culture of contemporary Semang hunter-gatherers. Satun Museum.

As a result, there has been unwillingness among scholars to consider forager communities food-producers prior to the AA influence but this is of course not the only motive. The late Pleistocene to mid Holocene/Hoabinhian period is not associated with cultivation of food because there is meager evidence for such practices in the archaeological record during this period, other than from a few uncertain excavated sites. The remains of plants together with new technological elements in the Hoabinhian layers at Spirit Cave, led Gorman to formulate the hypothesis of an early agricultural development in the region (e.g. see Gorman, 1970, 1972; also see Bellwood, 1993 on this topic). Hoabinhian contexts from the three cave sites at Ban Kao have revealed possible food-plants associated with habitation debris. Five types of seed or fruit were identified but they showed no morphological evidence of domestication (Pyramarn, 1989).

33

I agree with the skeptics (e.g. Bellwood, 1993), in that neither Gormans nor Pyramarns results can be considered evidence for early agriculture. However, considering that cultivation, domestication and agriculture are not a single process, evidence for food-production is probably not discernable in the archaeological documentation. The reason for this is that plants in a food-production system did not undergo a morphological transition from wild to domesticated, which means that it is not possible distinguish between plants gathered in the wild and those obtained from a food-production system in the archaeological record. Hence, archaeological research has focused on the final product, namely domestication and agriculture, and overlooked the initial stages of food-production. Hills (2009) argument concerning ratooning is useful since it shows that the cultivation of rice does not depend on domestication. Moreover, it can also explain how rice finally becomes domesticated through its reoccurring exploitation by local communities. Scarres (2005: 183) proposal that domestication is often unintentional and the result of interactions between humans and wild species would support this theory. Other significant work carried out by Denham (2004) has resulted in an alternative perspective on agricultural development in New Guinea. Firstly, Denham criticizes the common prevailing hypothesis of expanding Austronesian agriculturalists and Neolithic packages replacing pre-existing hunter-gatherer communities in the mid- to late Holocene, but also the theories of agricultural development in the region solely being of indigenous origin (Denham, 2004: 610-1). Contesting these theories, Denham (ibid) advocates inter-regional interaction in regards to the agricultural transition in the region, something which he argues has been previously neglected in favor of the Austronesian migration hypothesis. Further interesting information on early innovations of food-production in the Pacific has been reported by Athens et al (1996), Denham et al (2003), Denham et al (2004), Fairbairn (2005), Fairbairn & Swadling (2005), Haberle (2003) and Matthews & Gosden (1997), to name a few. The situation in the Pacific provides also a plausible scenario of socio-economic adaptation and transition for peninsular Thailand and will be further discussed in the final chapter. Bellwood is one of the most prominent supporters of the farmer immigration hypothesis, which he primarily bases on linguistic evidence. In Bellwoods view: agriculture and the distinctive associated artifact forms were simply adopted wholesale by preexisting Hoabinhian foragers, without considerable pressure from immigrant farmers. (1993: 48). The main objective of this thesis is to question this concept of a wholesale adoption of agriculture by indigenous hunter-gatherer communities. Instead, on basis of the environmental data and comparative studies from tropical settings elsewhere, I argue here that food-production started
34

much earlier. Communities in the late Pleistocene early Holocene were already involved in food-production prior to the supposed expansion of AA language-speakers, and this initial production was not necessarily related to rice. The main argument is that an indigenously developed food-production system was supplemented by more advanced agricultural methods, and perhaps domesticated rice in the second millennium BC. For this reason, it is argued, that both internal and external factors are responsible for the agricultural development, and thus also for subsequent urbanism and state formation in the region. The emphasis of chapter 2 lies on the environmental setting of the research area. It aims to evaluate the feasibility of Stone Age hunter-gatherers practicing agriculture and wild foodproduction/cultivation in the area. Chapter 3 presents the research results from the archaeological fieldwork carried out in Songkhla and Phattalung Provinces. Eight caves and rock shelters with Stone Age surface finds were found. Archaeological and oral testimonies show that caves and shelters were perhaps employed as burial localities in the Late Stone Age. In the final chapter, the research results are discussed and intertwined with environmental and archaeological data from tropical contexts in Island Southeast Asia, the Pacific, Africa and Mesoamerica. In so doing, we might be able to understand past food-production system in peninsular Thailand.

35

36

2. THE RESEARCH AREA

2.1 Environmental history and archaeology

Archaeologists primarily deal with two sets of phenomena - past human behavior and its material consequences (Bird & OConnell, 2006: 144). But how can we claim to understand past human societies; social organization, choice of settlement, subsistence patterns, material culture etc., if we do not consider the landscape in which humans interacted? A landscape perspective faciltates cross-disciplinary understanding over environmental and human-related sciences. Archaeologists should whenever possible take advantage of the data obtained through palaeoenvironmental research (Anderson, G. D. et al. 2007: 13) because knowledge of climatological, vegetational, geological, settings and changes can help us to understand environmental contexts in which past humans interacted. The relationship between cultural deposits and associated soils and landforms has been noted for some time (Mandel & Arthur Bettis III: 2001: 173). Sediment and soil development is critical for contextualizing pedologic data in archaeological interpretations because soils are the products of both human and natural processes, which evolved from the event of origin into the present observed phenomenon (Frink, 2003: 5). Thus, sediment and soil development in archaeology is useful in reconstructing the depositional and landscape history. It can also assist in locating cultural deposits, interpreting artefact associations and contexts, defining site stratigraphy and establishing cultural chronologies (Mandel & Arthur Bettis III, 2001: 175). An understanding of the environmental history of a given research area is fundamental as environmental factors were one of the main drivers behind human settlement patterns in prehistoric times. Environmental conditions establish the limits for human actions in a broad sense, although the range of possibilities for human activities is broad (Stargardt, 1976: 212). A landscape perspective provides an alternative approach than the cultural historical one. The objective of the following chapter is to elucidate the past and present environmental setting in the research area with special emphasis on land development; sediments, soils and watersystems. In other words, the aim is to place prehistoric settlements within their natural environmental setting. In doing so, it is also possible to investigate whether Stone Age communities could have engaged in food-production.

37

Seal level changes

Fluctuating sea levels greatly influence human settlement and adaptive patterns. Sea level rise and sunk land drove human and fauna to new places. Similarly, sea level drops pushed humans and fauna to exploit new coastal lands (Simajuntak, 2005: 372). Sea level history is important to archaeology because: sea level history is a key reference for other climate records, an accurate history of these changes is essential for establishing an order of events and causal relationships in climate history. (Cutler et al. 2003: 254). In addition, sea levels do not only affect coastal environments but also trigger other climatic processes with a variety of ecological responses (Stenseth, et al. 2002). Environmental data attests that at the beginning of the Holocene around 11, 000 BP sea levels were 50 meters below present levels. At this time, land bridges still connected the ThaiMalay Peninsula with Sumatra, Java and Borneo (Voris, 2000: 1155, 1164). The estimation of 22 meters below presented levels at 9700-9250 cal. yr BP by Horton et al. (2005: 1206-8) means, if Voriss (2000: 1162) estimation is accepted as valid (cf. Dunn, 1970: 1044), that the land bridge to Island Southeast Asia was lost at this time. This shows that the extent of land was much greater in the late Pleistocene and early Holocene than it is today. Communities could, without any considerable effort, move between the areas which today constitute the Thai-Malay Peninsula and Island Southeast Asia. From this aspect, the separation between Mainland and Island Southeast Asian archaeology in this time is not essential. It would rather advocate that similar subsistence strategies would have prevailed in these tracts in the late Pleistocene and early Holocene. The climatic and environmental similarities between the Thai-Malay Peninsula and Island Southeast Asia also speak in favor of this. In the mid Holocene between 4850-4450 cal. yr. BP, sea levels had a maximum height of 5 meters over present levels (Horton et al. 2005: 1206-8). Nevertheless, some scholars have argued for more than one Holocene highstand. Tjia (1996: 32), for instance, has argued for several highstands in the mid-late Holocene. However, Tjiass results have rendered critique for not having taken into consideration all environmental data available (see Horton et al. 2005: 1205). According to Bellwood (2007: 22), the sea level between 6000-3000 years ago might have been 6 meters above present but this may reflect tectonic or isostatic movement rather than absolute sea level rise. This theory may be supported by dated fossil coral from Phuket on the west coast. Here, sea levels have been reported to be 1 meter above present levels around 6000 BP (Scoffin & Tissier, 1998: 273).
38

In total, the environmental data generally shows high sea levels in the early Holocene and a decrease of sea levels over time. It means that in the early Holocene the extent of land around coastal tracts and the Satingphra Peninsula was greater than today. In the later Holocene periods the extent of land decreased and meant that much land became unavailable for human communities. After 6000 BP, however, land once again increased in size and became available for human exploitation. The ecological response was that a freshwater swamp developed. The ecological setting transformed between 2720 2350 cal. yr BP into an early stage of peat swamp before burning resulted in the destruction of the swamp forest, an intervention which Horton et al (2005: 1205, also see Stargardt, 1983) suggest might have been caused by human communities. This corresponds to Kealhofers (2003: 88) data, which points at an increase of grass burns around the same time. Fluctuating sea levels must also have been responsible for the destruction of archaeological sites in the coastal tracts (Bellwood, 1993).

GIS
In archaeology, the spatial dimension of human behavior over time is studied because variables of geography exert a strong influence on human behavior in past human societies. GIS (geographical information systems) is a computer based tool for compiling and analyzing spatial data and can facilitate mapping for descriptive analyzing and identifying spatial patterns in the landscape. This is done in order to create new data sets of information about patterns in relation to archaeological sites, which may provide information on aspects of environmental strategies and site and social organization, among other things, The initial data was first created for a landscape archaeology course at Uppsala University by the author and Anja Mortensen in spring 2009 but it was later conceptually extended and adapted to the research area by the author. The GIS data contains raster information about demography of Thailand, rivers and streams in a resolution of 300 m. The lack of digitalized data available for GIS utilization in the region forced the author to digitalize layers of map material, e.g. a GIS model was created with the use of ArcView Spatial Adjustment to combine multiple environmental data layers so as to illustrate the environmental setting of the area (this chapter) and for more simple site analysis (chapter 4). Nevertheless the precision of the produced GIS data on a high level of detail cannot be assumed as data was compiled using adjustment tools; one should thus be aware of the need to include a possible margin of error in the analysis. It also presupposes that the original maps
39

(geology, soils and land-use) are accurate which cannot be taken for granted as most of the published material is several decades old. However, all efforts have been made to minimize these possible errors by comparing this data to contemporary data. Google Earth and Google Maps have been used to locate and compare the distribution of landforms e.g., mountains and hills and courses of rivers and streams.

2.2 Environmental setting

The environmental history of lower peninsular Thailand is poorly researched and documented compared to the north. The conventional view of the landscape history of the region is mainly based on macro-scale environmental research related to Mainland and Island Southeast Asia in the Pleistocene with emphasis on sea level fluctuations (e.g. Hanebuth et al. 2000; Hanebuth & Stattegger, 2004; Heaney, 1991; Voris, 2000). Research on the Holocene of lower peninsular Thailand has only been undertaken by a few scholars (e.g. Horton et al. 2005; Stargardt, 1976, 1983). Nevertheless, published studies by Thai scholars and departments have added knowledge on current ecological conditions (Forestry Statistics of Thailand, 2007; Jintanugool & Round, 1989) and landforms (Joint Project Kingdom of Thailand and United Nations Special Fund, 1972; Pendleton, 1949; Vijarnsorn & Jongpakdee, 1979). The USDA system (Soil Survey Staff, 1999) was used to classify and explain the soils. The considerable amount of material published in Thai has not been included here.

Main landforms
The research area comprises a western mountain spine of granitoid and residual mountains (fig 2.1.) ((Dheeradilok et al. 1992: 745). The lower elevated boundaries have been made up of loams of shales, conglomerates, limestone and intrusions of gneiss. High alluvial terraces run parallel with the mountain range, consist of sandstone and quartzites and are intersected with lower alluvial terraces of semi-recent and old alluvium (Pendleton, 1949; Virgo & Holmes, 1977: 209). The estuarine coastal environment of Songkhla is a semi-enclosed water system which forms a lagoon-system (Jintanugool & Round, 1989). Three lakes Thale Sap Songkhla, Thale Luang and Thale Noi - form a shallow water-system of fresh to brackish water. These are connected to the sea through a narrow channel from Thale Sap Songkhla at the

40

town Songkhla, while the Satingphra Peninsula is a strip of land separating the lakes from the sea.

Figure 2.1. Geological land forms

The coastal environment is formed by organic materials transported from sea and inland rivers (Darnsawasdi & Chitpong, 2002: 44). Lower perennial river wetlands consist of meandered rivers and riverine marshes. The Satingphra Peninsula constitute of transported organic sediments with alluvial sediments of marine origin. The open sea coast and sandy beaches are associated with intertidal mudflats (ibid, 43). While some fossilized, mudflats are also under continuing formation today (Stargardt, 1976, also Lekenvall, pers. obs.). From the western margins of the Songkhla Lakes and eastwards the topography is more or less flat except for the cluster of heights at Khao Daeng (Red Mountain) and Khao Ko Yai (Big Mountain Island) ((Stargardt, 1976). In the Satingphra Peninsula late prehistoric and early historic urban settlements have been found and excavated by Stargardt (1976, 1977, 1983).

41

Land development
Land development is related to climate. During the Last Glacial Maximum precipitation is argued to have been lower than today (Wang et al. 1999: 278). Depending on deglaciation more water in the oceans meant more precipitation. The increase of water also led to changes in moisture, temperature and vegetation (Heany, 1991: 59). The area experienced land changes in the Holocene because of the increasing levels in oceans. These factors combined changed the landscape. Pollen analysis from the sediment sequence of Thale Noi lake basin in the coastal zone indicates that the coastline was covered by mangroves 8500 cal. yr BP. They were replaced by a freshwater swamp between 7880-7680 cal. yr BP. During the time period of 2720 2350 cal. yr BP, the freshwater swamp transformed into a peat swamp before burning resulted in the destruction of the swamp forest (Horton et al. 2005: 1202-05, also see Stargardt, 1983).

Figure 2.2. Sediments.

42

The isostatic processes during the stadials and interstadials affected the distribution of land and ecology. That much of the coastal land was under current sea level in the early-mid Holocene is important because it shows the extent of land available for past communities. In contrast to the many Hoabinhian cave sites in the inland (e.g. see Bellwood, 1993), coastal shell middens at Guar Kepah (Pinang) and Seberang Perak in peninsular Malaysia attributed to Hoabinhian communities have been reported by Adi (1983: 53-4). This indicates that communities used the littoral areas for subsistence related activities, e.g. mollusk-gathering (for examples of this see Bentley et al. 2007), but it does not necessarily mean that inland and coastal communities were the same people. Burning of the forest vegetation started in the inland of the peninsula and it is suggested that food-production took place here first. It was not until land became available for exploitation that food-production emerged on the Satingphra Peninsula. Since current environmental data suggests that this occurred in the first millennium BC, one can speculate whether fullscale agriculture was introduced in the region at this time, almost a millennium later than in the rest of Southeast Asia, the island region included. I argue that it shows that inland areas were subjected to both earlier food-production and later rice agriculture. It was not until people learned how to master the environment and make use of sub-surface fresh water with a hydraulic system of irrigation and water tanks that agriculture began. The coastal location of the Satingphra Peninsula and its close association with rivers and lakes provided good opportunities for both irrigated rice agriculture and trade (Stargardt, 1973, 1976, 1983).

Sediments
Transgression and regression of the sea level are an important part of land development but other processes also contribute to the formation and alteration of landscapes. Progradation is the incoming sediment and growth of river delta and it has had major impact on coastlines in Southeast Asia (Stark, 2006: 411). In tropical climates the movement of water is one of the prime factors behind the distribution of sediments and creation of distinct landforms and the transformation of parent material into soil differentiation. Bedrock and regolith are the sources for sediments which are transported downhill and become accumulated as colluviums or alluvium (fig 2.2.) ((Sanchez & Buol, 1975: 598; Strahler & Strahler, 2005: 428). Limestone and granitic are abundant in the area and are important as parent materials in soil formation (Jenny, 1994: 54-5). Because of high temperatures, humidity and precipitation, chemical
43

weathering is intense. Chemical weathering is: the combined processes that cause rock to disintegrate physically and decompose chemically because of exposure near the Earths surface. (Strahler & Strahler, 2005: 419). Because of these reasons, sedimentation processes contribute to land formation and therefore also determine the extent and type of land available for past human communities. This has been best described by Stargardt (1976) in relation to the coastal area of the Songkhla Lakes. Fluvial processes have transported silts from the inland rivers to the coast, which has resulted in sediments becoming deposited over time and, ultimately, in alluvial segmentation at the coast. In addition, the abrasive actions of storms in the South China Sea has led to particles of sand and mud becoming deposited along the land strips of land. These dynamic forces can be seen through previous coastlines still visible in the landscape today (Stargardt: 1976: 212-20, also see Stargardt, 1983). Moreover, tidal currents are active in the area (Stargardt, 1976) and ought to have accumulated silt and clay, which contributes to the coastal condtions and extent of land (see Strahler & Strahler, 2005: 542-55, on this subject). Stargardt (1976) argues that Khao Daeng (Red Mountain) and Khao Ko Yai (Big Mountain Island) were once off-shore islands in a shallow sea. The land development is continuing to this day. Beach drift, which is the littoral drift of material, can be seen along the coast. It is plausible that the spit formations (fig 2.3.) contributed to the formation of the lagoon-system through longshore drift of material. This suggests that fluvial processes interact with coastal processes which led to the formation of the lagoon-system and the Satingphra Peninsula. These observations are important since they suggest that the extent of land in the past has always been changing and so did the ecology. On this basis, Stone Age communities surely acted in a different environment than that encountered today. This also account for the location of present watersheds. Rivers probably played an important role in the choice of settlement among earlier Holocene societies. The porous character of alluvial material, however, means that rivers erode easily, change direction and create new watersheds. Meandering is common in these environmental contexts (fig 2.4.). Considering the possibility that rivers and streams might have changed direction, potential archaeological sites can be located at places in the landscape which are not always visible.

44

Because environmental stress has altered the Holocene landscape, it is not surprising that so few archaeological sites from the Stone Age have been found. It has also been noted that in tropical contexts, hydrolysis and oxidation over time have resulted in the decay of materials in depths of 100 meters (Strahler & Strahler, 2005: 419). It means that many sites have probably been exposed to environmental decay and are today buried far below the ground surface, making them almost impossible to find. For this reason, the purpose of the arFigure 2.3. Spit formations.

chaeological survey (chapter 3) was also

to focus on higher elevated areas in the interior because these areas have presumably suffered less of this environmental stress and alteration of the landscape. This decision was taken because it was believed that the possibility of finding archaeological remains was higher in higher elevated mountainous areas than on the Satingphra Peninsula. Environmental data from the inland of Krabi Province (Kealhofer, 2003) also indicates that cultivation strategies started in the interior of the peninsula. It is important to note here the coast and the interior lie close to one another in distance. The distance between the east and the west coast roughly measures between 100 km and 150 km. The distance today between the coast (e.g. the Satingphra Peninsula) and the research area the interior of Songkhla Province only measures approximately 50 km. The area might thus be seen as a single entity and the separation be-

Figure 2.4. River meanders.

45

tween the coast and inland is not as important as one may think. Instead, it suggests that there were many opportunities for communities to engage in different subsistence related strategies. The next section focuses on the soils of lower peninsular Thailand with the objective of determining the possibility of human engagement in food-production strategies in the area.

Soils
In soil classification either the FAO or USDA system is used. The USDA-system is used here because the original soil map on which the GIS-illustration is based is the USDA version from 1979. The soil information from the FAO of peninsular Thailand is very general and does not take local and regional variants into consideration. However, in the last decades, some of the sub-soils and soil names in the USDA-system have been conceptually changed. In such instances, the soil with the most similar features has been used to replace it. Note that the soil names or the system used are not the important issue here. Instead, their specific characteristics are of importance if one is to investigate the possibilities for practicing agriculture and wild food-production/cultivation in the region. Sediment transformation of rock into soil is the result of the interaction between climate, relief and parent material but it is also affected by living organisms such as vegetation, fauna and humans. These are all part of the land-forming processes (Dudal, 2005: 95). In comparison to deposits of material which result from sedimentary processes, soil formation is a secondary alteration related to weathering at a relatively stable land surface. Because chemical weathering is particularly intense in tropical climates these regions account for approximately one-third of the worlds soils (Hartemink, 2004: 373). Soils can also provide a record of human activities (Mandel & Arthur Bettis III, 2001: 174-5). For this reason, studying soils is useful since they can reveal information on land-use and subsistence strategies related to agriculture. The definition of soils has been explained: natural bodies, made up of mineral and organic materials that contain living matter and can support vegetation out of doors, and have in places been changed by human activity. (Keys to Soil Taxonomy, 1996: 1). Soils consist of sand, silt and clay and it is the portion of material and the texture class which determine the soil type (Strahler & Strahler, 2005: 609) and also agricultural productivity. Ultisols are the dominant soil in the research area (fig 2.5.) and in the interior as a whole. The regional variants of ultisols in warm and humid places with high annual precipitation are the udult and aquult. The aquult is a humus-poor utisol with an udic moisture regime in wet places in which ground water is close to the surface. They were formed in alluvium and ma46

rine deposits in the Pleistocene (USDA, 1999: 726). The udult is the more freely drained and humus-poor ultisol and originates from developments in sediments and on surfaces from the Pliocene to the late Pleistocene. Most of these soils have or had forest vegetation (ibid, 747, 761). Ultisols in tropical contexts are usually subjected to slash-and-burn agriculture since, without fertilizer, they can only sustain crops for a short period of time before nutrition is lost (Strahler & Strahler, 2005: 621). That ultisols have supported forest vegetation and is suitable for shifting agriculture are an important association because proxy data from Krabi Province indicates forest clearance in the late Pleistocene and subsequent slash-and-burn agriculture in the Holocene (Kealhofer, 2003). A similar trajectory may therefore be assumed for Songkhla Province.

Figure 2.5. Soil types.

The coastal environment predominantly consists of inceptisols which have been formed by

47

loamy and clayey parent materials. They usually occur both in young and old surfaces or deposits (USDA, 1999: 489). Most of the Satingphra Peninsula is dominated by the aquept, an inceptisol occurring in wet areas such as floodplains and delta lands with natural poor drainage capabilities (Strahler & Strahler, 2005: 628). The Aquept usually formed in late Pleistocene or younger deposits on level plains or on flood plains (USDA, 1999: 494-95). This soil is also suitable for rice agriculture (Foss et al. 1983: 374) because wetland areas are rich in organic matter (Strahler & Strahler, 2005: 628). The inceptisol south of Thale Sap Songkhla is of a loamier regime than aquepts on the Satingphra Peninsula and is therefore more freely drained than the latter (Steila & Pong, 1989: 96). These wetland areas were probably the primary zones for rice cultivation in the Satingphra catchment prior to development of irrigated rice agriculture associated with canals and tanks in the first centuries AD. The outer coastline consists of spodosols. The Humod is a sandy and freely drained spodosol with accumulation of organic carbon. It was formed in the Pleistocene and Holocene. In tropical regions most humods have supported rain forest (USDA, 1999: 709), which has contributed to the acidic humus of the soil. It means that opportunities for cultivation are limited, except for acid-tolerant species and unless fertilizer is applied (Strahler & Strahler, 2005: 626). Along the narrow channel running between Thale Sap Songkhla and Thale Sap the soil type is entisols. The aquent is the entisols which occurs in wet areas such as tidal marshes, margins of lakes and flood plains (USDA, 1999: 393). In moist climates they are among the highest in agricultural output (Strahler & Strahler, 2005: 627). This means that if rice cultivation was carried out without fertilizer it would have been in the wetlands with high organic matter. The soils most appropriate for rice agriculture are the entisols and aquepts. Based on the properties of the soils, the inland area is less productive in an agricultural sense because the soils are less favourable for the growth of plants. It means that fertilizer is required in order to acquire a continuous agricultural output from the same unit of land; otherwise shifting cultivation is likely to have been the food-producing strategy. Slash-and-burn agriculture is also supported by environmental data (Kealhofer, 2002, 2003; White et al. 2004). In the coastal area the conditions for farming are better due to the wet and organic soils. Rice farming is also believed to have begun on the Satingphra Peninsula in the first centuries AD when sub-surface freshwater could be exploited for the first time. Prior to the technological advance of the hydraulic system, Stargardt (1976, 1983) argues that the Satingphra Peninsula did not uphold permanent settlements owing to the dearth of surface freshwater. The adaptation to the environmental extremes of the Satingphra Peninsula, however, is be-

48

lieved to have started c. 2500 2200 BP with small-scale cultivation along river banks and lake shores with experimentation of grain selection (Stargardt, 1983: 31). It is conceivable that the inland reflects an earlier phase of food-production which is associated with shifting cultivation and different crops than rice. Stargardt (ibid) stresses that communities in the coastal tracts experimented with grains. However, this notion has not been supported by archaeological or environmental data. It is therefore arguable whether agriculture based on domesticated grain seeds did develop without external influence. Considering that the agriculturalist expansion is believed to have occurred more than a millennium before this assumed experimentation of grains, as suggested by Stargardt, it remains questionable whether a similar process would have occurred twice in the same region. This suggests that early cultivators focused on other crops than domesticated rice. Because the initial food-production was probably not related to rice, it is important to investigate which other plants might have been available to Stone Age communities. Observing present day land-use is one way of investigating how the land might have been used in prehistoric times. This will be done below.

Land-use
According to Forestry Statistics of Thailand (2007), paddy land is the most common practice in respect to land-use, followed by field crops, vegetable plantations and fruit tree plantations. The map information of land-use in Thailand is based on a survey conducted by Omakupt in 1972 and has been applied to GIS by the author (fig 2.6). It shows that the upper reaches of the inland are generally characterized by rubber tree (fig 2.7) and orchard plantations. In the Satingphra catchment land is mainly attributed to rice paddy fields and other crops in the western margins of the Songkhla Lakes. Along the eastern coastline coconut and orchard are cultivated. This general map, however, does not show the full extent of land-use in the research area. In a more recent study, Panapitukkul et al (2005) have accounted for the land-use in Rattaphum district (fig 2.8). In the coastal plain of the Songkhla Lake Basin, paddy fields and vegetables is evident. Rubber trees are grown and rice agriculture is also carried out on the lower terraces of the inland. The mountainous areas are mostly used for growing rubber and fruit trees.

49

Figure 2.6. Current land-use system in the research area.

Land-use and the type of crop cultivated are dependent on soil feature and landform. The lower elevated areas in the Satingphra catchment uphold more nutritious soils which are more favorable for cultivating aquatic crops than acid soils. The coastal area and the Songkhla Lakes have a high abundance of water and are seasonally flooded. Today, a drainage system controls the amount of flooding but water from the flooded rivers is still used for growing crops as flooding recedes (Darnsawasdi & Chitpong, 2002: 43-4). Water is also stored in water tanks (Lekenvall, pers. obs.) and is probably used for irrigation during the dry season. It has been argued that this kind of hydraulic system supporting rice agriculture has been used in the Satingphra Peninsula since at least the 4th century AD (Stargardt, 1976, 1983). The inland soils are generally more acid and the inland environment is more mountainous. For this reason, fruit and rubber trees are mostly cultivated in these parts. In the lower terraces of the inland, however, paddy fields are present in more acid soils. Acid soils prohibit long
50

term agricultural productivity unless fertilizer is applied. Since the inland soils could not have supported sustainable crop yields it is reasonable to suggest that past foodproducers practiced slash-and-burn cultivation. This picture is also supported by pollen and phytolith samples from Krabi Province (Kealhofer, 2003), not far from the research area. Whether this is connected to rice is not
Figure 2.7. Rubber tree plantations close to Khao possible to say but rice is certainly a staple Rakian Cave.

food even today in shifting agriculture in (insular) Malaysia (see Bruun et al. 2006; Neergard et al. 2008) and Laos (Klln, 2004: 71), for example. Ratooning is an alternative foodproduction strategy in this region (Hall, 2009). Therefore, even if the soils are acid, rice cultivation need not be impossible, although crop yields would be small per land unit exploited. This point is important because it suggests that there might have been good reasons to grow other more acid tolerant crops than rice in order to maximize agricultural surplus. Arboriculture and horticulture are also quite reasonable strategies for this area, based on the tree cultivation carried out today.

Figure 2.8. Current land-use and topography in Rattaphum district after Panapitukkul et al 2005: 150. 51

Moreover, certain crops must have been favored because they offered more protein or carbohydrates than others. For instance, the staple diet in New Guinea today comprises taro, jams, sugar cane, edible grasses and large-leaved vegetables, which are grown in the high lands of the island. Irrigation in the highlands associated with these crops is evident from the late Pleistocene early Holocene and is still used today. These are also indigenous foods on the island, which suggests that food-production probably did evolve without external influence (Diamond, 1997: 311-12). These foods, however, do not provide enough protein to the diet. Therefore, subsistence strategies probably relied on hunting, fishing and mollusk gathering to a large extent. Trade/exchange of food-products between communities might have also contributed to the local diet. An interesting aspect of the example from New Guinea is that a large variety of subsistence economies is practiced in diverse environments on the island today. The highlands support farmers, littoral areas fishers and people are practicing shifting cultivation based on bananas and jams, which is complemented by hunting and gathering, in the drier areas away from the coast and rivers. The marshy grounds support mobile hunter-gatherers, who depended on the sago palm, which is argued to give three times more energy per hour of work than horticultural products (Diamond, 1997: 313). In my opinion, this suggests that there is no uniform transition from one economic system to the other during a given period of time. Instead, diverse economies and communities coexist and interact with one another, perhaps through trade/exchange. I believe that the situation in New Guinea could provide a plausible subsistence scenario for other places, peninsular Thailand included. Considering that fruit tree cultivation is a common feature in contemporary land-use, it is conceivable that this is a remnant from early food-producing communities. However, the examples from New Guinea also show that many different crops might have been utilized for subsistence related purposes. Still there are reasons to assume that we are dealing with economic and cultural variation during contemporary periods of time. This will be discussed more thoroughly in chapter 4. The archaeological survey is discussed in the next chapter.

52

53

3. THE ARCHAEOLOGICAL SURVEY

3.1 Survey strategies

The archaeological survey was conducted together with Siriporn Limwijitwong from the 13th Fine Arts Department (FAD), Songkhla. The intended survey strategy was supposed to be systematic walking in transects. This method, however, turned out difficult to conduct; due to the dense forest vegetation covered the research area. Instead, the local community was addressed and archaeological sites were mostly identified through formal interviews and informal talks (see Karlstrm, 2009: 95-6; Kllen, 2004: 77, on a similar survey strategy in Laos). An initial cursory survey in the Satingphra catchment was carried out with the objective of visiting previous known archaeological sites and to gain knowledge of any Stone Age activity on the peninsula. Owing to external circumstances, such as time and financial constraints, the initial survey only constituted a quick field inspection. Previous surveys and excavations carried out by the FAD had suggested that the likelihood of finding prehistoric archaeological remains would be high in the Limestone Mountains in the interior. The Songkhla and Phattalung Provinces was deteromed as suitable survey areas through cartographic studies. Since some people are more competent in answering certain types of questions than others (see discussion in Bernard: 2006: 200-1), it became obvious that elder villagers would know more about artifacts found in the area than younger people. One frequently used method to determine archaeological potential was to initiate contact with elder local people living in the area of interest and ask whether or not stone axes/adzes (Khaun hin kat in Thai, Khauon fa, south Thailand) had been found near the village. The reason for asking about stone axes is that they are perhaps the easiest recognizable attribute of the Stone Age. Interviews were combined with display of photographs of stone implements to the villagers. The collection of cultural objects, prior to the opening of the first museums, was conducted by Buddhist monks. Because most Thais are Buddhist, local people often make merit by presenting offerings to temples. Temples in the past had a curatorial role of artifacts found by the local community and as a consequence many museum collections derive from temples (Chaorenpot et al. 2008: 45, 51, 63). For this reason, monks often possess knowledge on archaeological remains and the location of potential cave and rock shelter sites. Museums and temples
54

were visited with the aim of documenting objects attributed to Stone Age material culture and acquiring knowledge on where such items have been found.

3.1. To the left: late prehistoric and historic sites on the Satingphra Peninsula and Stone Age sites in the inland as indicated by potsherds, stone tools and oral testimony. 3.2. To the right: Map of cities and road system in lower peninsular Thailand.

To test the assumption that Stone Age communities might have favored Limestone Mountains (or that the chances of finding remains in Limestone Mountains are higher owing to the delimitation of caves/rock shelters), two Stone Rock Mountains, namely Khao Want and Ton Pliu (or Plew) in Rattaphum district were cursory surveyed. Contacts with the Ministry of Natural Resources and Management revealed that no Stone Age artifacts had been found in this area. According to a forest manager, who had traversed the mountain area on several occasions, no caves or rock shelters have ever been recorded. Local informants confirmed the same. This quick field inspection cannot exclude the possibility of Stone Age activity but it indicated that it would be difficult to find archaeological remains. The position of caves and rock shelters were recorded by a handled Global Positioning System (GPS). However, this turned out impossible at many places because it was occasionally difficult to get signal. In these cases, coordinates were taken wherever this was possible. One should thus be aware of the margin of errors and that the coordinates can differ up to 200 meters. GIS have been used for illustrative and simple analyzing purposes so as to supply information on site location in context of environmental attributes.

55

Oral testimony associated with Satingphra

Oral testimonies from the local community in the Satingphra Peninsula indicate the presence of Stone Age communities. Contacts were first initiated with villagers living in the vicinity of already known archaeological sites previously excavated by Stargardt, who has reported vast amounts of protohistoric remains and historic urban settlements on the peninsula (1973, 1976, 1977, 1983). An interview conducted by the author with a fisherman showed that there is potential to find Stone Age material culture on the Peninsula. He explained that a teacher close to his house owned a polished stone axe but he did not believe that the teacher would agree to let us see it. The second interview took place at Khokt Ong, an archaeological site located in the city Satingphra, which was excavated by Stargardt in the 1970s. It came to our attention that a man living close to the site had gathered artifacts in the area. His collection consisted of dozens of stone and glass beads, spindle whorles and an earring. These items were similar to the finds recorded by Stargardt and her team made some 30 years ago. According to him, stone axes had also been found but he was reluctant to show us where. Nevertheless, he introduced us to a man who, according to the interviewee, was known to be possession of stone axes. The man was not at home but his wife confirmed that her husband had such in his possession. A polished stone axe did eventually turn up with the assistance of staff working in the office at the FAD. It was in the possession of a woman living on the Satingphra Peninsula. The woman had received it from her grandparents who had lived in a village in the area their whole lives. Thus, she assumed that it had been found in the area. Based on this information there are strong indicators that the Satingphra Peninsula was inhabited during an earlier phase of the Holocene. Nonetheless, this sparse evidence in the form of a single object alone cannot alone prove such claims. In addition, without a known provenance for the items there are reasons to question whether they were found in the Satingphra Peninsula. Considering the significance of old items and polished stone axes/adzes, in particular among Buddhists, such items might have been inherited by families for generations, originating from other areas. Because no Stone Age sites are known from the Satingphra Peninsula this coastal strip of land needs to be further investigated. It is recommended that further archaeological surveying is carried out in the Satingphra Peninsula.

56

Khao Daeng Rockshelter

Khao Daeng is a mountain which lies west of the Satingphra Peninsula and the Songkhla Lakes in Phattalung Province. It is one of several mountains close to the Great Lakes but one of few Limestone Mountains. The mountain is surrounded by houses and temple lies at the foot of it. The area is largely cultivated and major cultigens include rice and rubber tree plantations. The Khao Daeng Rockshelter is situated in the middle of the mountain and it is easily accessed. Quite recently, it seemed, the area closest to the rock shelter had been dug out (fig 3.3) and the material from the pit had been used to build an altar around a tree (fig 3.4). The altar also showed evidence of recently burnt incense after Buddhist stupa practices.

3.3. The wall of the rock shelter showing disturbance of ground surface.

3.4. Khao Daeng Rockshelter and present day altar and house.

57

The rock shelter has previously been surveyed by the FAD but no surface finds were recorded at that time. Stargardt mentions that Khao Daeng, as an inland mountain, is likely to have been inhabited in prehistoric times before settlement moved to the Satingphra Peninsula in later times (1983: 19). For this reason and because of its physical characteristics (limestone) and proximity to the coast, it seemed as a suitable place for archaeological survey. The prospection of the site was carried out by the author and Limwijitwong in May 2009. The objective of this inspection was to look for surface finds. Contrary to the previous survey, potsherds of prehistoric and historic origin were found. This could be due to the recent removal of material of an estimated depth of 30-40 cm. Potsherds were also found in association with the altar and near the wall of the rock shelter. The majority of the potsherds was historic but one or two of the sherds were most likely of Late Stone Age origin. These were identified through the red/brown/yellow color and cord mark design on the sherds, which were similar to Late Stone Age pottery found elsewhere. It is recommended that the rock shelter is excavated in the future despite the fact that it has been disturbed. In sum, the area seems archaeologically rewarding and should be further investigated, both around the cluster of heights of Khao Daeng and further in Phattalung Province. Little archaeological research has been conducted in Phattalung and it should perhaps become a priority when it comes to future archaeological fieldwork in southern Thailand. (Note that Khao Daeng is not illustrated or analyzed in GIS because of its distance from the cluster of sites in Songkhla Province).

Khao Rakian Mountain

Tham Khao Rakian is a mountain in Rattaphum district (fig 3.5) in eastern Songkhla Province. The area is surrounded by tree plantations and sparsely positioned houses. One stream flows about 200 meters from the cave. Khao Rakian Cave was previously excavated in 1986 by the FAD (see chapter 2). The excavations revealed that people resided in the area in the Late Stone Age. The team had found pottery and skeletal remains from the site in 1986 (Archaeological survey from Khao Rakian, 1986). Our research objective was two-folded. The first aim was to find out more about the archaeology in the area. For this reason, the local community was involved. Secondly, objectives was to survey the mountain for archaeological sites, as this area has not been surveyed previously. The survey was carried out by the author and Limwijitwong in May 2009. The in58

tended systematic survey strategy was abandoned because of topographical constraints. The survey relied on information from local residents instead. With the assistance of two local guides, the author and Limwijitwong recorded two caves and three rock shelters but no surface finds were recorded. Of these five, one of the rock shelters seemed more
3.5. View of the landscape from the highest point of Khao Rakian Mountain.

promising than the others. It had an estimated size 20x30 meter platform and nearby

there was a small cave. It also resembles Khao Daeng Rockshelter visually and in regard to elevation. Contacts with the local community revealed that artifacts had been found in the area surrounding the mountain. An old man living in the area told us that he had found a polished stone axe and pots containing bones 30-40 years ago in the cave. According to the chief of the village, human skeletal remains including 11 craniums were found inside the cave when local farmers dug there for fertilizer a few decades ago. The excrement from cave dwelling bats is a very effective fertilizer due to its high levels of phosphor. Local farmers therefore use the cave soil as fertilizer. The human remains were kept by the local temple but their present location is unknown. Note that these finds were made before the 1986 excavation and that the cave had already been disturbed at the time of the excavation. Moreover, in the area surrounding the mountain, plenty of animal bones have been recovered during the plantation of rubber trees by the local farmers. Most of these bones were believed to have come from an elephant. The recorded evidence from the 1986 excavation and the newly obtained oral testimony show that the area is archaeologically rewarding. It is therefore recommended that future archaeological fieldwork is carried out in the region. The rock shelter is a good place for excavation because of the above stated reasons. It also seemed undisturbed. Considering the archaeological potential of the area, it is also recommended that more surveying is carried out. These surveys should be oriented to areas nearby rivers and streams, which might result in the recovery of open archaeological sites.

59

Khao Nui Mountain

Kao Nui is a mountain in eastern Rattaphum district. According to the FAD, the main cave in the mountain, Sekysoune Cave, has been surveyed previously but the date of the survey is unknown. At that time, Late Stone Age potsherds were found. Today the area is under severe disturbance as a new tourist center is under construction in the area. Guided tours will be carried out inside the big and deep cave in the near future. The purpose of the survey was to find additional artifacts and sites. The survey was carried out by the author and Limwijitwong in May 2009. The survey technique relied on informant interviews so as to acquire information on potential caves/rock shelters and artifacts found in the area. A quick inspection of the Sekysoune Cave showed no evidence for archaeological remains. However, potsherds were found in a rock shelter close to the entrance. These sherds were identified as Late Stone Age pottery. Another cave a few hundred meters away from the other sites was inspected but nothing of interest was found. All of the sites had been heavily disturbed. According to the villagers, pottery vessels and human skeletal remains had been found in the Sekysoune Cave in the past but their current location remains unknown. The local guide working in the area stated that he had found a polished stone axe while planting rubber trees in the past. He explained that it had the same length as his finger (compare with the one found in Khao Joompa, later this chapter), but he did not know where it was located today. Similar to the previous survey areas, this region has archaeological potential, but the problem is that the sites have been heavily disturbed. Even so, the area should not be neglected and archaeological surveys in the area might prove rewarding. The field in which the local farmer/guide found the stone tool is suggested as a potential site for future excavations. Test pits may determine the potential for further excavations.

Khao Khua Cave

Khao Khua (fig 3.6.) is located in the northwestern part of the mountain of the same name. Khlong Takian is the nearest stream, which runs approximately 300 meters away from the cave. The mountain has undergone industrial mining for the past 20 years, and, as a result, a result a big part of the mountain has been destroyed. The cave was found during archaeological surveying by the FAD in 2007 but no artifacts were found at that time. Local residents, however, reported that human skeletal remains had been found in the cave in the past (Archaeological Survey for Mining, 2007). The purpose of the survey was to inspect the cave once
60

again and establish contacts with the local community in order to investigate the archaeological potential of the area. The survey was conducted in June 2009 and it was carried out by the author and Limwijitwong. Similarly to the other caves investigated during the survey, the cave had been disturbed as the cave floor had been dug out to an estimated depth of 10 meters. Nevertheless, it came to our attention that soil had been filtered, which was marked by piles of soil. In these piles potsherds were found (fig 3.7.). While most of the pottery were historic, one sherd was considered prehistoric. We
3.6. The entrance of Khao Khua Cave.

learnt during our fieldwork that local residents

had found skeletal remains in the cave. Pottery has also been found in the agricultural fields of the area but these were believed to be historic. We were told that no additional caves/rock shelters existed in the mountain, but a villager who used to hunt in the area claimed that there were additional caves.

3.7. Potsherds found inside Khao Khua Cave.

There are good reasons to further investigate this area. Because of the industrial mining, additional surveying of the mountain is recommended. The archaeological evidence from Khao Khua and from nearby Khao Tok Nan and Khao Joompa make the area archaeologically interesting and worthy of further investigation. The use of several survey strategies during fu61

ture field research is recommended. These should include increased contact with local residents as well as random and systematic survey strategies. By combining these methods, the possibility of finding archaeological sites might prove fruitful.

Khao Joompa Hill

Khao Joompa (fig 3.8) lies between Khao Tok Nan to the southwest and Khao Khua in the northeast in Rattaphum district, Songkhla Province. The hill has not been previously surveyed. Because of its topographic similarities and vicinity to Khao Tok Nan hill, it was determined appropriate for archaeological investigation. The research objective was to find caves/rock shelters and archaeological remains. The survey was carried out by the author, Limwijitwong and a local guide in May and June 2009. The survey method relied on local knowledge obtained by interviewing residents. The survey resulted in the recovery of a cave and a rock shelter. The rock shelter was not visited because of time constraints but the cave was investigated. The path to this cavern-like structure was hard, and access to its interior demanded climbing equipment. Surprisingly, the cave had been disturbed and soil appeared to have been removed from the bedrock-floor. No surface finds were recovered. Because of safety reasons, only the first chamber was surveyed, and the continuing tunnels were left unexplored. This cavern seems unlikely to contain archaeological remains.

3.8. Khao Joompa Hill from the ground level.

62

While the cave did not reveal any archaeological evidence, contacts with local residents did suggest the presence of sites in the area. A polished stone axe was in the ownership of an old man living in the area. He used it (together with several other items) in a Buddhist ceremony called kheun ban mai. According to the man, his father had found it in the area. He believed that it had been found in the agricultural fields during farming activities. Owing to the proximity of Khao Khua and Khao Tok Nan, the area in total requires more archaeological attention. The area should be subjected to archaeological surveying, which ought to focus on the agricultural fields close to rivers and streams.

Khao Tok Nan Hill

Khao Tok Nan (fig 3.9.) lies as the third of three limestone outcrops in a northheastern direction (fig 3.10.). Khlong Paum Canal flows near the hill. The landscape is characterized by rice cultivation and ponds used for irrigation. The hill area has not been previously surveyed and the objective was therefore to find out whether archaeological remains and caves/rock shelters are present in the area. The fieldwork was carried out by the author together with Limwijitwong in May 2009. Two survey strategies were used. First, the local community was addressed so as to obtain information on archaeological items and caves/rock shelters in the area. Secondly, the hill was systematically surveyed by walking in transects (2-3 meters between each survey).

Figure 3.9. To the left: Khao Tok Nan from ground surface. Figure 3.10. To the right: View towards Khao Joompa and Khao Khua.

63

Interviews with the local community revealed features which had been found in the area in the past. According to a monk in the local temple, a rock shelter had once existed in the southern part of the hill close to the canal. Stone artifacts had also been found around the rock shelter but according to the monk, the rock shelter had collapsed since then. However, it is possible that the rock shelter in question was in fact located during the survey, although no surface finds were recovered. The monk also told us that two polished stone axes had been found during the construction of the Khlong Paum Canal a few decades ago (fig 3.11).

Figure 3.12. To the left: The entrance of Khao Tok Nan Hill where a polished stone axe is reported to have been found. Figure 3.11. To the right: Khlong Paum Canal south of Khao Tok Nan Hill.

According to a local resident, local narratives state that the hill had two caves. Two caves were found during surveying but it is likely that they constitute different entrances to the same one (fig 3.12.). This was also later claimed by a local man living beside the hill. It was possible to enter into the first chamber at one of the entrances but the steepness of the cave prohibited further exploration without safety equipment. We were unable to enter the cave from the second entrance as we did not have proper safety equipment. No surface finds were recorded in the caves. Despite the fact that no archaeological remains were found, a local resident informed us that he had found a polished stone axe in the entrance of the cave (fig 3.13) when he was a child. At that time, he had given it to his brother and its current location is unknown. The collected evidence from this area indicates high archaeological potential. The cave should be further investigated and the rock shelter is perhaps suitable for excavation. It is recommended that the areas around the hill, agricultural fields and canal, are surveyed.

64

Figure 3.13. Polished stone axe found in the vicinity.

Khao Chang Lon Mountain

Khao Chang Lon Mountain is located in the northwestern part of Rattaphum district. It is a large Limestone Mountain situated in a similar environmental context to the other limestone formations. Khao Chang Lon has not been previously surveyed and the objective was therefore to determine the archaeological potential of the area. The survey was carried out by Lekenvall and Limwijitwong in May and June 2009. The survey method relied on local informants and local guides.

Figure 3.14. Khao Chang Lom situated at the foot of the mountain.

The local community confirmed the presence of four caves. Two of them were small-sized and very disturbed. No surface finds were recorded at either one of the caves. One of the two remaining caves, Khao Chang Lom, was detectable from ground level (fig 3.14). It is a big
65

cave with one main chamber and many small chambers. The cave had been heavily disturbed by farmers who have removed soil from the cave, which is used as fertilizer in agriculture. A soil depth of estimated 10 meters of the cave floor had been removed (fig 3.15). Because of this, possible cultural layers have been lost. In one of the smaller chambers a hammer stone was found (fig 3.16) but nothing else of archaeological interest was recovered. The fourth cave, Tham Kra Duk, is situated at higher altitude than the previous caves. Like Khao Chang Lom, it had been disturbed and only piles of stones on the bedrock remained. No finds were recovered here.

Figure 3.15. To the left: The wall inside the cave showing removal of soil. Figure 3.16. To the right: Hammerstone found in one of the chambers.

According to the local community, bone remains have previously been found in Khao Chang Lom (fig 3.17.) and Tham Kra Duk (figure 3.18.). It is not known whether the bones from the cave are from animals or humans but human bones and a cranium from Tham Kra Duk indicate that humans might have been buried in the cave in the past. The bones were later given to the local temple but their current location remains unknown. Vast amounts of shells have also been reported from Khao Chang Lom. However the specific species of mussel/snail found from the cave has not been identified. Therefore, it cannot be determined whether the shells were of fresh water or salt water origin. Nevertheless, it is possible that people brought them to the cave for subsistence purposes.

66

Figure 3.17. To the left: The entrance of Khao Chang Lom from the inside. Figure 3.18. To the right: The entrance of Tham Kra Duk from the inside.

The oral testimony obtained from local residents suggests that humans were buried in Tham Kra Duk Cave. Therefore, Andersons (1997) hypothesis of burials being associated with early farming communities could be relevant. Moreover, it would also appear that the area has a long, continuous record of human occupation from prehistoric times to the historic period since local villagers have reported that Buddhist statues and relics had been placed inside and outside the cave in the past. They were later moved to the local temple. The area presents interesting possibilities for future archaeological fieldwork. Although the caves themselves will not supply us with more information, archaeological surveys would probably be of interest. I propose that surveys should be carried out in the agricultural fields close to the rivers and streams in the area. Elevated areas, such as mounds, should also be investigated because settlements frequently occur in such places.

The museum collections

All national museums and provincial Fine Arts Departments in Thailand are under supervision of the National Fine Arts Department in Bangkok. As mentioned earlier, many of the museum collections in Thailand derive from local temples, which before the establishment of museums nationwide played the role of the keepers of the cultural heritage. The temples received artifacts from people who made merit by offering cultural objects to Buddhist monks (Chaorenpot, 2008: 45, 51, 63). Several museums were visited with the objective of making an inventory of their collections.

67

Figure 3.19. Polished stone axe with cutting edge. Songkhla National Museum. Figure 3.20. Polishedstone axe. Songkhla National Museum. Figure 3.21. Polished shouldered stone adze. Songkhla National Museum.
.

In Songkhla National Museum the Stone Age collection consists of polished stone tools with cutting edges (fig 3.19), polished stone axes (fig 3.20) and polished stone adzes (fig 3.21). Wat Matchimiwat is a temple museum in Songkhla city and also displays stone artifacts associated with the Late Stone Age (fig 3.22). A knife made out of stone (fig 3.23) is particularly interesting because it might have been used in cultivating activities, for example as a sickle. In Satun National Museum the prehistoric artifacts displayed are Late Stone Age pottery.

Figure 3.22. Stone tool collection. Wat Machimawat National Museum. Figure 3.23. Stone knife (?). Wat Machimawat National Museum. Figure 3.24. Late Stone Age pottery. Thaksin Folklore Museum.

Another museum in Songkhla Province is that of Thaksin Folklore. The prehistoric collections of the museum include Late Stone Age axes and pottery (fig 3.24 and fig 3.25). The exhibition of the Nakhon Si Thammarat National Museum also includes stone implements similar to those displayed in Songkhla. In addition the museum is in possession of a bark beater, an item which was used for the extraction of fibers from tree bark aimed at textile production. The linear motif of the item also suggests that it might have been used for imprinting motifs on pottery. A stone bangle is also on display. Kedah State Museum in Alor Setar, northern peninsular Malaysia, has the largest collection of Stone Age objects of the visited museums.

68

This is exemplified by the displays of Hoabinhian hand axes and hammer stones (fig 3.26) and Late Stone Age tools (fig 3.27).

Figure 3.25. Tripod. Thaksin Folklore Museum. Figure 3.26. Hoabinhian stone tools. Kedah State Museum. Figure 3.27. Polished stone tools. Kedah State Museum.

Because of the background of the history of museums in Thailand the provenance of the archaeological objects is usually unknown. This is a problem since it is not possible to access the information which is usually made available through archaeological excavations. According to museum personnel, the archaeological objects in Songkhla and Wat Machimiwat museum derive from the local community. In Thaksin Folklore Museum the items have been donated from other provinces in southern Thailand. The provenance of the pottery in Satun National Museum is not known but the same type of pottery has been discovered elsewhere in the province (see Awasamas, 2008). It also resembles some of the pottery found during the archaeological survey. The find context of the objects in Nakhon Si Thammarat and Kedah State Museum is specified. In Nakhon Si Thammarat the artifacts derive from the three provinces, namely Nakhon Si Thammarat, Surathani and Krabi. They were found during archaeological excavations and surveying in the 1980s. In Kedah, the origin of the objects is even better documented. The collection derives from archaeological excavations carried out in the first half of the 20th century.

3.2 Is the material culture evidence of food-production?

Even though the provenance of the items is usually unknown, they clearly demonstrate the presence of Stone Age communities in the area. Whether they are associated with agriculture is up for interpretation. Pottery is commonly associated with the material culture of early farming communities. The combined evidence from the surveys and archaeological fieldwork carried out in the past shows that Late Stone Age pottery frequently occurs in the area. The
69

pottery may therefore be the product of immigrating agriculturalists in the second millennium BC. Gorman (1970), however, stresses that Hoabinhian communities made pottery. He based this claim on archaeological evidence from the Spirit Cave in northwestern Thailand. In addition, in Laang Spean Cave in Cambodia, pottery has been attributed to late Hoabinhian groups (Stark, 2003: 212-13). Whereas Solheim II (2003: 1) supports Gormans hypothesis of a Hoabinhian pottery production, others - for example Bellwood (2007, also see Matsumura et al. 2008) - consider the Hoabinhian a pre-ceramic phase. Presumably, the uncertainty among scholars regarding the dates of pottery is a reflection of disturbed excavation contexts and contamination of radiocarbon samples. Uncertainty also resides in the fact that there are two different schools of thought. One acknowledges the possibility of early agriculture amongst Late Stone Age communities, and the other emphasizes a traditional hunter-gatherer economy amongst Late Stone Age communities. If pottery was produced amongst Late Stone Age communities, agriculture is argued to have been part and parcel of ceramic production, and, consequently, if ceramic production is absent, communities were hunter-gatherers. The link between farming and beginning of ceramic production needs to be further tested. A landscape and environmental perspective provides an independent source of data through which we can approach the question whether Stone Age communities practiced foodproduction and also when full-scale agriculture began. As discussed in chapter 2, there is no evidence for a rapid replacement of a hunter-gatherer way of life by an agricultural one in the pollen record. Although the intrusion of farming communities in the second millennium BC is plausible, based on the comparison of pottery in the research area and in other parts of Southeast Asia, food-production most likely already existed prior to this influence. Archaeological data which might support Gorman and Solheims claims of an early pottery production comes from China and Japan. Pottery has been documented among late Pleistocene communities in China (e.g. see Yasuda, 2002) and Japan (e.g. see Habu, 1996). These communities were sedentary or semi-sedentary communities who were involved in some form of food-production. In the Yangtze River Area, semi-domesticated plant seeds from several cave sites dated to c. 20, 000 BP were accompanied by pottery not long after (Yasuda, 2002). This indicates that the initial food-producers were the makers of early pottery, rather than the residents of fullfledged agricultural villages. However, it is important note that the absence of pottery need not mean that communities were not involved in food-production. Nevertheless, in many parts of Southeast Asia, the development of pottery seems to be related to the beginning of foodproduction. In order to ascertain the situation for southern Thailand, more archaeological data is required.
70

Polished stone tools are usually associated with farming communities. The polishing of stone implements started in some places in the late Pleistocene - early Holocene. In China the earliest completely polished tools have been dated to around 10,000 BP (Zhao et al. 2004). In Southeast Asia, on the other hand, these items appear several thousand years later. Polishing increased the mechanical strength of the axe, and, for this reason, they were important tools for forest clearance. An increase in forest clearances associated with slash-and-burn agriculture is evident in the environmental record from circa 4000 BP (Kealhofer, 2003). The appearance of these items thus seems to correspond rather well with the recorded increase in land-use and might therefore also be related to early farming activities. Micro-wear analysis of stone tools from Samrong Sen, a Late Stone Age site in Cambodia, has shown that they were employed as multi-purpose tools in woodworking activities. For instance, adzes were probably used to cut down trees and chop wood or bamboo. Chisels, on the other hand, were employed for more delicate carpentry, such as grooving, smoothing or carving in wood. It is also stressed that the inhabitants of Samrong Sen produced these tools since the construction of pile houses and boats was essential for settlement in the flooding zone (Heng, 2008: 89-90). These implements were probably also used for other activities, such as agriculture, fishing, hunting, food-processing and weaving (Zhao et al. 2004: 131). Moreover, in Neolithic villages in China, reaping knives or sickles made out of stone have been found (Chang, 1986: 93). The stone knife located in Wat Machimawat National Museum and displayed in fig. 3.23 (see page 68) is interesting and perhaps represents farming activities in the Late Stone Age. If so, it may be the eldest archaeological evidence for agriculture in southern Thailand. The transition from a flake tool industry to Hoabinhian woodworking implements in the late Pleistocene implies that this shift is associated with the increase of forest vegetation around the same time. That early Holocene communities transformed their stone tool industry probably had the purpose of cutting down trees. There could be several explanations for this phenomenon. It is possible that forager groups wished to concentrate the remaining herbivores and other animals in forest gaps, with the aim of making them easy targets to hunt. The change in technology could also be related to food-production. Spriggss (1996: 525) hypothesis of the different food-production strategies in the Pacific appears to fit the situation in lower peninsular Thailand. The hypothesis is also supported by the environmental record. The burning of forests may have been carried out with the purpose of clearing forests in order to encourage the growth of certain plants or so as to replant or cultivate them. Considering that communities practice several different specialized subsistence economies and live side by side
71

even today in New Guinea (e.g. Diamond, 1997; Fairbairn, 2005), it does not seem so implausible that a similar sequence could be found in peninsular Thailand.

3.3 Sites in GIS

The results of the surveys and information gathered from local residents, the material in the museums and our knowledge of sites in peninsular Thailand and Malaysia, together clearly indicate Stone Age activity in the area. This section presents a GIS-based landscape analysis, the aim of which is to investigate aspects of prehistoric site settlement locations with the backdrop of the finds gathered during the archaeological survey by the author and the 13th FAD in spring 2009. Economic and environmental studies tend to predominate in archaeological applications of GIS. Studies on ritual landscapes and perception of landscapes are more recent developments. Most of the GIS applications which focus on cognition use GIS to create viewsheds and cost surfaces in order to reconstruct how people perceived the landscapes in which they operated (e.g. see Wheatley, 1995). The lack of trustworthy height data prohibited any further analysis of slope and elevation. Instead, the purpose of using GIS was to identify key environmental variables in relation to the sites. GIS data on watersheds, soils and topography is used in order to facilitate an understanding of site locations. The next part discusses how the locations of the sites in relation with the landscape may be associated to food-production through time.

3.4 Site location and environment

The sites are located on limestone mountains/hills in an environment associated with lower elevated terraces of semi-recent and old alluvium. The landscape is characterized by a valleylike depression and is surrounded by higher elevated terraces (fig 3.28). The majority of the sites is situated in limestone outcrops overlooking the lower ground level plains and inland swamps, land which today is mainly used for rubber plantations and paddy fields. Khao Nui Cave site, on the other hand, lies on a foot slope of the western mountain range but is situated close to the lowlands. Parra rubber and fruit trees are cultivated in its surroundings (fig 3.29).

72

Figure 3.28. The surveyed archaeological sites seen in relation with the different environmental features.

Countless streams and rivers intersect the landscape and there tends to be a concentration of watercourses around the sites. It shows that accessibility to water did not constitute a problem for Stone Age communities. Rivers were probably also used for mollusc gathering and fishing as well as for transportation to adjacent coastal areas. For example, Songkhla River which flows north of the sites may have been one of the rivers used for such endeavours. Water is also important to the cultivation of different crops. The soils are suitable for shifting cultivation, which probably varied in intensity and methods and was practised from the late Pleistocene onwards. The sites are also located near lithic sources and the resource base of material for stone implements is therefore abundant in the research area. Metamorphic rocks in the form of hornfels, quartzite and sedimentary rocks, such as shale, sandstone and chert, all occur frequently. Igneous rock types, such as basalt, is also present. A key question is the nature and function of these sites. The type of usage is often linked to diverse time periods in human history. From an evolutionary perspective, they have considered the habitats of the first humans, and for this reason, their inhabitants in more recent
73

times have been seen as primitive people, i.e. hunter-gatherers. That they were used by hunter-gatherers is uncontested but foragers have also occupied open-air settlements. The use of caves and rock shelters has typically attracted functionalist explanations. Anderson (1997: 619) argues from a functionalist position that cave sites were used because they offered protection from the elements. On this basis, hunter-gatherers probably used them as temporary campsites. The choices behind occupying caves can have other explanations. Based on evidence from Lang Kamnan, a post- and late Pleistocene cave site in western Thailand, Shoocongdej (2000) suggests that two mobile organizational systems are present in seasonal tropical environments. She argues that the archaeological evidence indicates residential mobility in the wet season and logistical mobility as an organizational response to the dry season. Semang hunter-gatherers today also use caves as shelters together with shelters constructed on the ground, depending on the season and supply of subsistence. In this manner their settlement pattern resembles Shoocongdejs model. Reynolds (1992: 96) also contributes to the discussion regarding usage. He suggests that cave sites, such as Banyan Valley Cave and Spirit Cave in north Thailand, served as specialized sites for the exploitation of upland and forest products. These were then perhaps exchanged or transported down to larger villages in the plains.

Figure 3.29. Current land-use and topography of the research area after Panapitukkul et al 2005: 150, with the sites found during survey.

74

Caves and rock shelters might also have been places of social and religious significance (Anderson, 2005: 137). In the later Holocene periods, caves and rock shelters have been ascribed a role as burial sites (Anderson, 1997: 614). The upper cultural layers (2500 4000 years ago) of Lang Rongrien Rockshelter are interpreted as a burial site with pottery-associated inhumations and cremations (Anderson, 1990). The majority of sites in our research area also is associated with burials and in most cases also with pottery. This indicates that people in the Late Stone Age buried their dead in caves and rock shelters. Anderson (1997) discusses that while inaccessible locations seem to have been favored by late Pleistocene and early Holocene group, the use of cliff-foot hills appears to have begun in later Holocene periods. He considers it a shift in the economic system towards farming and suggests that shelters at ground surface became the prime location for settlements. However, landscapes does not mean only the availability of resources for acquiring food but could also be cognitive, namely a sacred landscape (Knapp & Ashmore, 1999: 10-11, cited in Macamo, 2005: 13), or places with ceremonial and spiritual value (ibid). It is interesting to speculate whether the burial sites functioned as some sort of monuments in the landscape. Archaeological and anthropological studies have shown that caves and rock shelters may play a religious role and some caves have been perceived as vital aspects of a sacred landscape (Healy, 2007). In Neolithic England, for example, funeral long burrows have been attributed to geographical and political groupings consisting of families and/or clans in a ranked society (Wheatley, 1995). It has been argued that the construction and presence of burials were used to signal claims to land during the Swedish Iron Age (Lwenborg, 2010: 20). Interpreted in such a fashion, the sites in peninsular Thailand could suggest the emergence of ranked communities and sedentary agricultural villages. However more archaeological research is needed in Thailand in order to test such hypotheses against empirical data. Anderson (1997) suggests that the beginning of cave/rock shelter inhumations is related to a shift in the economic system. However, there is no secure archaeological evidence for a change from a hunter-gatherer mode of subsistence to an agricultural one in the period between 4000 2500 BP, and a fast transition from one subsistence regime to another is not supported in the environmental data. Because of this, Kealhofer suggests that a more complex pattern of local and regional interactions influenced the transition rather than a simple immigration (Kealhofer, 2003: 89). This might be associated with the increase in deforestation and the subsequent evidence of slash-and-burn agriculture in the same time period. Nevertheless, immigration and/or increase in contacts with agriculturalists is evident in the material culture. The existence of contact networks does not imply that the indigenous hunter-gatherer soci75

ety was replaced by agriculturalists. Instead, internal food-production was complemented by an external food-production. It is possible that indigenous food-production was supplemented by more refined agricultural methods from the outside, which were more complex in terms of domesticated plant seeds (e.g. rice), agricultural methods (e.g. irrigation), material culture (e.g. pottery, polished stone tools) and settlements (e.g. villages). Despite these advantages, internal processes also contributed to the development of agriculture and succeeding urbanism in Satingphra and Kedah. As argued here, food-production probably has an even longer history in the region, which developed prior to the external influence by agriculturalists. One dominant theory when it comes to explain the development of food-production is that it was forced upon human. Flora and fauna changed due to climate changes during the late Pleistocene and early Holocene periods which affected the distribution and type of food resources. This must have led to socioeconomic diversification amongst forager communities and they had to develop other strategies to cope with the environmental situation. Food-production was one of these new foodrelated strategies. That wild foods became more important in the late Pleistocene - mid Holocene/Hoabinhian is also evidenced from three sites in Kanchaburi Province, namely the sites of Khao Talu, Ment Caves and Heap Cave. The excavations show that communities made plenty of use of plant and root foods as well as fruit (Pyramarn, 1989). Ecological studies have also shown that the tropics are resource rich environments suitable for pre-agrarian subsistence economies. On this account, it is very credible that cultivation was carried out prior to the arrival of AA farming communities. Kealhofer (2003: 86-8) proposes that groups throughout the Holocene may have practised arboriculture, horticulture and hunting and gathering in a mixed economy. The evidence from 4000 BP onwards shows that disturbance taxa and indicators of slashand-burn agriculture increase dramatically (Kealhofer, 2003: 88). The latter peak of disturbance taxa coincides rather well in time with the start of burials in caves. It is currently not possible to evaluate whether this change is also related to the emergence of rice agriculture as there is a lack of pure rice samples from archaeological radiocarbon dated contexts. Environmental data indicates that wild rice grew in the area in the earlier Holocene but the transition from gathering rice in the wild to domestication and agriculture would have been a big step at that time. It is unlikely that communities domesticated rice in peninsular Thailand. Instead, the most plausible explanation is that domesticated rice came from the outside, considering that the first secure evidence for domesticated rice from Ban Non Wat has been radiocarbon dated to 1650 BC (C. Higham & T. Higham, 2009).
76

Considering that some communities cultivated food, it is also reasonable that they were sedentary or semi-sedentary. The presence of burials also indicates that communities frequently resided in the local area. However, there is a lack of information on settled villages in the archaeological record from the Thai-Malay Peninsula until the emergence of urbanism on the Satingphra Peninsula and Kedah during the first centuries AD. Harris (1996) argues that the most far-reaching consequence of agriculture was that it enabled more food to be obtained while it supported more people per unit area of exploited land. Harriss argument is perhaps an adequate explanation for full-fledged agricultural societies with settlement aggregation but it does not explain the effects which food-production might have had on late Pleistocene and earlier Holocene communities. It also leads to the question: how much food and which type of food is needed to provide for a sedentary village population? The key to becoming sedentary depends on a regular and stable supply of subsistence. Agriculture is not a requirement for sedentism. A sedentary lifestyle can be supported as along as mammals, fishes, mollusks, plants and so forth are abundant and cover the basic demand of energy. Examples of such societies are found in Japan, for instance. Late Pleistocene huntergatherers of the Jomon culture had semi-permanent settlements. It is believed that Jomon settlements were occupied for part of the year as seasonal base camps to which people returned on an annual basis (Habu, 1996). However, such communities differ from agricultural ones in that they are unable to store resource surplus. It has been argued that storing surplus also led to the emergence of complex hierarchical societies and settlement aggregation. Since foodproduction is a step closer to agriculture than wild food-procurement, it is possible that it could give rise to more advanced settlements than the Jomon settlements, for example. However, village settlements from the Stone Age have not been found in the Thai-Malay Peninsula. It is probable that initial food-production did not give rise to settlement aggregation in the same way as agriculture did. The reason for this might be that there was not enough surplus to support large sedentary populations. This implies that settlements may have been semipermanent constructions which were moved when the emaciated soils demanded so, in a food-production system which resembles slash-and-burn agriculture. Moreover, if the people responsible for the cave burials were early farmers, there are strong reasons to assume that open settlements are located in the vicinity of the sites. The choice of location for these agrarian or semi-agrarian settlements should logically be the result of combined settlement-subsistence strategies which are determined by different environmental variables (e.g. see Binford, 1980). It can be argued that groups settled down in areas in which the environment offered a maximum output of subsistence. It is also expected
77

that food-producing and farming communities, unlike hunter-gatherer societies, depended on favorable soil conditions in order to practice cultivation. However, since both food-producing and farming communities usually continued to practice hunting/fishing and gathering as part of a mixed economy, several environmental variables were probably considered before settling into an area. In general, Late Stone Age settlements are likely to be found in higher elevated places protected from water and wind. These places are usually mounds and hillsides. Proximity to arable land and fresh water are also basic conditions for agricultural practices. Rivers and streams also contribute to the diet (fishing and mollusk gathering), stone tool industry (raw material) and trade/exchange (transportation). Ban Chiang (early period 2100-900 BC) on the Khorat Plateau is the most famous and well-excavated Late Stone Age site in Thailand. The site is located on a large mound in the lowlands at the junction of three streams (Pietrusewsky & Douglas, 2002: 5). Other Late Stone Age sites in Thailand are Ban Non Wat in Nakhon Ratchsima Province in the northeast (Higham & Thosarat, 2006: 98) and Ban Kao in Kanchanburi Province in the south-central parts (Ooi, 2004: 208). They are all mounded villages situated in a similar environmental context as the Ban Chiang site. One site south of Ban Kao has shown evidence for postholes from a raised-floor house (Bellwood, 1993: 46), which could provide a basic understanding of early house structures. Similar housing strategies have also been reported in southeastern China. At the Late Stone Age site of Hemudu (5000 BC onwards), agricultural villagers constructed wooden houses built on stilts (Buckley-Ebrey, 1996: 17). Houses on raised poles are constructed in parts of Southeast Asia also today (c.f. Klln, 2004: 223). Hence, this seems to be a general trait of how houses were and still are constructed in tropical environments. In doing so, it is possible to avoid moist and heavy monsoonal rains. On this basis, Late Stone Age settlements are probably found on mounds which lie in proximity to arable land and near or at the junction of rivers and streams. The caves and rock shelter sites may have acted as burial localities within the limits of the settlement area. Settlements should be looked for in areas which fulfill these environmental conditions, as suggested by the location of sites in northeast and central Thailand. It is not known whether early foodproduction in the late Pleistocene to mid Holocene gave rise to similar settlement constructions since such sites have not been found in Southeast Asia thus far. It could be speculated, however, that early food-producer settlements were of a semi-permanent nature. Therefore, their settlements probably did not leave any long standing evidence of habitation. This would also explain why settlements from this time period have not been found. Environmental decay
78

may also be responsible for this invisibility in the archaeological record. Even if the lack of Stone Age open settlements is problematic, the food-related endeavors of these communities are visible in the environmental record. In the final chapter I will further argue for an early origin of food-production, which was complemented by agriculture at a later date. Regarding the difficulties in interpreting and separating hunter-gatherers, early food-producers and agriculturalists in the archaeological and environmental record, I propose that these different food-related strategies have co-existed through time in a close interaction, even until this day. In doing so, I also suggest that several factors rather than only one (immigrating agriculturalists) contributed to the development of urbanism and later state formation in the region.

79

80

4. DISCUSSION

4.1 Summary and a local model

Previous archaeological research carried out in peninsular Thailand has shown that communities have inhabited the region from the Pleistocene period onwards. Nevertheless, the evidence for these communities is only based on a handful cave and rock shelter sites in the northern and southern parts of the Thai-Malay Peninsula. The archaeological fieldwork conducted by the author did reveal the presence of pottery and stone artefacts in caves, rock shelters and museum collections. This shows that groups inhabited lower peninsular Thailand in the Stone Age. Indeed, this has been known amongst many Thai archaeologists working in the area for quite some time but the results of the excavations in (e.g.) Khao Rakian Cave (see chapter 1) have not reached the international domain. The main topic of this thesis, however, is food-production, which is argued to have been practiced by indigenous communities prior to the supposed arrival of farming communities in the second millennium BC. Archaeological, ethnographic and environmental data combined, the results imply the presence of cultural variability from the late Pleistocene until today. Food-production probably first appeared in the late Pleistocene early Holocene as a human response to environmental change. At a later stage, agriculture and domesticated plant seeds were introduced into the area. Meanwhile, some groups persisted with a hunting and gathering way of life. Owing to this disclosure, I propose that a variety of factors contributed to the adoption of agriculture, which later led to urbanism and state formation in the region.

Relating the data

For the time being, the archaeological data from the region alone is insufficient to prove the existence of food-producing communities in peninsular Thailand. Proxy data is therefore crucial for the interpretation of food-production in the environment along with comparative perspectives from similar environmental settings from elsewhere. Traditionally, the emergence of agriculture has been attributed to climatic changes. In the late Pleistocene early Holocene large herbivores died out because of climatic changes (or because they were hunted out) and
81

some hunter-gatherers had to abandon wild food-procurement in favour of food-production as the main subsistence strategy. The disappearance of large mammals and the introduction of new plant species probably encouraged the spread of food-production. Nevertheless, hunting and gathering remained a substantial part of the diet for a considerable period of time, and some groups never abandoned this sort of food-procurement. Environmental data, as discussed by Kealhofer (2002, 2003) and White et al (2004), suggests the presence of foodproducing activities in peninsular Thailand from the Late Pleistocene onwards. These practices appeared to have increased in the Holocene. Data from other tropical settings might assist the interpretation of the archaeological record in peninsular Thailand. In Mesoamerica a variety of crops is argued to have been cultivated several thousand years ago (Rosenwig, 2006: 331). Horticultural subsistence strategies evident in the environmental data suggest that forest clearances took place from c. 9000 BP (Piperno & Pearsall, 1998: 78). Neff et al (2006: 281-91) propose that forest clearance began in the same region by 6000 BP the latest. In addition, archaeological evidence does indicate extensive land-use by horticultural groups, who modified the forest ecosystem by incorporating domesticates or semi-domesticates into their subsistence. In Island Southeast Asia and the Pacific, which are considerably closer to our area, arboriculture and horticulture are evident from the late Pleistocene period onwards. Palynological and archaeological evidence from the Kuk site in New Guinea suggests that food-production was under way 10,000 years ago (Latinis, 2000: 49). In Borneo, research on stable carbon isotopes from several Late Stone Age sites implies that some communities were devoted to open-forest horticulture (Krigbaum, 2003). The development of agro-ecosystems in tropical regions appears to depend on the high percentage of rainforest taxa. However, not all hunter-gatherer societies in tropical regions became food-producing communities. Although rainforest settings in West and Central Africa have a high percentage of arboreal rainforest taxa (Mercader, 2002: 122-23), evidence for food-production has not been found in the environmental record. It can be argued that a foraging lifestyle resulted in a higher yield of resources than a food-producing one. Excluding the ecological niches in western and central Africa, the lack of early food-production in Africa and South America could be due to the large sizes of the continents, which enable a variety of subsistence strategies. Assuming that foragers had to change subsistence strategies due to climatic changes, the transition from foraging to food-production must have had a much more devastating impact on populations living in closed environments, such as small islands, peninsulas and isthmuses, than those found in Africa and South America. Similarly, communities in
82

Africa and South America were probably not affected by population pressures as badly as those living on islands, for example. In sum, the availability of resources meant that there were no pressing reasons to begin food-production in tropical Africa early on. Notwithstanding, land-use in Late Stone Age Africa differs between arid grassland and wetter wood lands. Improvements on fishing and fowling technologies and hunting implements are argued to have led groups in grassland areas to become more mobile, while foragers in wetter parts appear to have intensified the procurement of marine and plant foods, which could have led to a more sedentary lifestyle in woodlands and inland littoral areas (Kusimba, 1998). This also implies that semi-sedentary lifestyles need not be directly related to the emergence of food-production. In Island Southeast Asia and the Pacific, on the other hand, food-production seemed to have emerged because it was forced upon humans. Climate change and human involvement led to the extinction of large mammals, which disturbed local ecologies. Considering that these are islands, the effects on the food economy must have been severe, and the choice of subsistence strategies limited. A similar sequence of events might have also taken place in the Thai-Malay Peninsula based on the similarities in flora, fauna and climate between this area and those of Island Southeast Asia. Pollen and phytolith data also support this theory. Moreover, the Thai-Malay Peninsula also constitutes a rather small landmass. The average distance from the west to the east coast of the peninsula, measures no more than approximately 150 km. This distance could have been shorter in the late Pleistocene and earlier Holocene periods because the sea levels were higher then than they are today. For instance, the east coast and the Satingphra Peninsula are today also accessible by foot. This is a one- or two-day journey from our archaeological sites (Lekenvall. pers. obs.). The extinction of large terrestrial animals in this environmental context must have forced communities to explore alternative subsistence strategies in peninsular Thailand. Food-production was reasonably one of them. Alternative food strategies would have included specialized ways of wild food procurement and production; hunting and gathering of small mammals, mollusk gathering and fishing in littoral tracts and wild food-production/cultivation. These strategies may have been practiced separately by different and economic oriented communities during the same time period. However, while food-producing communities continued to make use of hunting, gathering and fishing, which remained a substantial part of the diet, the other communities were probably not at all involved in food-production. Instead, they become devoted to specialized hunting and gathering, which are dominated either by marine or terrestrial resources but contact and interaction between these communities surely persisted. Stargardt (1983: 20) speculates
83

whether the urban settlers of Satingphra in the first centuries AD exchanged the surplus crop plants in return for aromatics, medicinal plants and poisons extracted from the forests. This kind of exchange probably occurred earlier and also involved other products. As part of a food-production system which resembles slash-and-burn agriculture, burning of vegetation might have been carried out with the purpose of clearing forests in order to encourage the growth of certain plants. Spriggs (1996) argues that this was a food-production strategy in Melanesia, and it is also a plausible strategy for peninsular Thailand. The emergence of food-production in Melanesia has been dated to 20, 000 3500 BP. Subsequently, agriculture, domestic animals, large villages and pottery were introduced to the area in connection with the Lapita culture (Spriggs, 1996: 530, also see Spriggs et al. 2002). The transition to agriculture during the second millennium BC corresponds rather well with the supposed AA expansion in the region. Available environmental documentation from both areas indicates that a similar phase of food-production was carried out prior to the introduction of agriculture, even if the transition to agriculture in Island Southeast Asia mainly is associated with Austronesian agriculturalists in the first centuries BC. Following the Hoabinhian period, Bellwood (1997: 255) proposes a rapid Neolithic expansion and colonization of the peninsula after 4000 BP. The almost contemporary early dates for pottery, stone implements and tripods speak in favour of this theory. There is also a relationship between the beginning of burials in caves/rock shelters and the appearance of Late Stone Age material culture. This association has been previously made for sites in peninsular Thailand (Anderson, 1997; Chew Lam Project, 1986), Malaysia (Bellwood, 1997: 165, 268; Bellwood, 2001: 304), Borneo (Barker et al 2007: 245; Krigbaum, 2003: 295-6) and now also in southern peninsular Thailand. The earliest burials have been dated to the second millennium BC, which coincides with the alleged agriculturalist expansion into Southeast Asia during this time. Therefore, it seems logical that some scholars (e.g. Anderson 1997) have associated the early burials with the arrival of agriculturalists. These changes also coincide with an increase in disturbed taxa and slash-and-burn agriculture around 4000 BP, as suggested by environmental data (Kealhofer, 2003: 88). I do not argue against the view of regional farming influence in the region in the first millenniums BC. In fact, this relationship is perhaps even strengthened by the results from the recent archaeological fieldwork carried out by the author. What might be the largest problem with this interpretation, however, is that it overlooks the possibility of indigenous change and adaption to food-production, which occurred prior to this external stimulus. Moreover, the environmental data from peninsular Thailand does not support the idea of a cultural and chro84

nological gap between hunter-gatherer and expansionary agricultural groups (Kealhofer, 2003: 89). This is also reflected in the excavations in Surathani Province from which Hoabinhian and Neolithic artefacts have been found in the same cultural layer. There also seems to be a significant overlap between the assemblages over time (Chiew Lam Project, 1986). A distinct overlap through time of Hoabinhian and Neolithic archaeological remains has earlier also been pointed out by Reynolds (1992: 85) for Banyan Valley Cave and by Gorman (ibid) at Spirit Cave in north Thailand. Contact and interaction between people is an integrated part of the history of the region. Bellwoods (1993: 48) concept of a whole sale adoption of agriculture by pre-existing Hoabinhian foragers needs to be tested by alternative theoretical perspectives. I suggest here that food-production was most likely already under way prior to this agriculturalist intervention. Secondly, neither the environmental nor the archaeological record suggests a fast transition from a hunter-gatherer way of life to an agricultural one. So, what does this mean? Keahofers (ibid) proposal of an ongoing pattern of local and regional interactions facilitating change rather than a simple immigration is a reasonable assessment of the situation dealt with in this study. The Khok Phanom Di site (c. 2100 1500 BC) in the Bangkok Plain has revealed direct evidence for contact and interaction between agriculturalists and huntergatherers. Sedentary, ceramic-using communities appeared to have practiced hunting and gathering and fishing in the midst of the transition to rice agriculture in the interior (Bentley et al. 2007: 301, Higham, 1996). The earliest archaeological evidence (of which the author is aware) for rice agriculture in the Thai-Malay Peninsula is rice husks derived from Late Stone Age pottery, which resemble domesticated or semi-domesticated rice (Asawamas, 2008). However, this data remains contested as no radiocarbon dates are available for the samples. Rice appears in the phytolith record rice around 5000 BP (Kealhofer, 2003: 87) but the earliest instances presumably derive from wild stands. Consequently, the possibility of early domestication and cultivation of rice during the mid-late Holocene cannot be completely excluded. It would seem likely that domesticated rice first appeared in the region in the second millennium BC. A substantial amount radiocarbon dates from stratified pre-agrarian to agrarian contexts at the Ban Non Wat site in northeast Thailand strongly indicates that rice agriculture began c. 1650 BC (C. Higham & T. Higham, 2009). Nonetheless, rice probably played a role in past human subsistence regimes but as one of many sources of foods gathered in the wild. The method of ratooning would perhaps explain how the initial rice cultivation was carried out (Hill, 2009). However, it is important to note that this cultivation strategy does not result
85

in morphological changes associated with domesticated since rice did not actually become domesticated through this method (Spriggs, 1996). For me it stands clear that archaeologists have placed too much emphasis on rice agriculture at the expense of other aspects of food-production. Blench has focused on this issue, writing: One unconscious bias of the development and spread of agriculture is an emphasis on cereals and tubers. Since these are the basis of agriculture in the developed world, when students of prehistory construct narratives in the tropics they tend to focus on these classes of cultigen and to downplay both trees and herbs. (Blench, 2002: 2). Food-production strategies of arboriculture and horticulture, which included different root foods, were probably practiced prior to the agriculturalist expansion into the region. It would also seem inadvisable to focus on the cultivation of a single species in the tropics since it puts one at high risk of crop failure (Chang, 1970: 183, cited in Kealhofer, 2002: 189). It is likely that hunting and gathering continued to constitute an important part of the diet. Despite new data and alternative hypotheses the specifics of this food-production industry and its cultural implications remains unsolved. Local level explanations are necessary to further explore this transitional period in the Thai-Malay Peninsula. Rosenwig stresses the importance of this: local level explanations are required to explore specific sequences of human/plant interaction and their resulting cultural implications. This does not preclude the possibility of common processes; it simply recognizes that there is a wide variety of interrelated factors that may not always have occurred for the same reasons. (Rosenswig, 2006: 331). The stone axes recovered from the cultivated domains around our sites suggest a connection between lithics, land-use and perhaps forest clearance with the aim of cultivating lands. Regardless, no harvesting implements have been found in the area. In East Asia and Southeast Asia, sickles are usually small flat knives out of stone with a curved blade but harvesting knives of shell and pottery are also common (Needham & Bray, 1984: 323-24). One item which might have been used for such a purpose is the stone knife in Wat Machimiwat Museum (figure 3.27). Although the provenance of this knife is unknown, it presumably originates from the Late Stone Age because it has been polished, although stone tools were also polished prior to this time and also after. The locations of sites and burials and their close association with the lower terraces, which today are used for agriculture, may suggest that sedentary farming settlements from the Stone Age could also be found in this area. Future archaeological surveys are needed in order to explore this possibility.

86

An alternative perspective
Rather than using simple distinctions between hunter-gatherers and agriculturalists, the complexity of the situation requires an alternative model for explaining major changes in the prevailing economic systems in the region. The socio-economic structures of past and contemporary hunter-gatherers, the development of food-production and later agriculture could be explained by the following scheme (table 4.1.). The first stage involves hunting and gathering (H-G). Harris (1996) also calls this stage wild plant-food procurement. It is the basic manner in which humans obtain food. The next two next stages are (continued) hunting and gathering (H-G) and early food-production (EFP). Spriggs (1996: 525) argues that EFP (although he does not use the term food-production) is a logical extension from foraging to wild food production/cultivation of crops. While some started with food-production, others seem to have continued to forage. For this reason, the scheme is divided into two directions of the foodrelated economic development (H-G and EFP). Because middle-large herbivores became extinct (due to climate change or anthropogenic related reasons), the remaining hunter-gatherers were probably forced to specialize on certain types of subsistence. The present day Semang hunt small animals with blowpipes and toxin, which probably would have not been an appropriate strategy for hunting large herbivores in the past. Both these strategies (H-G and EFP) suggest an adaptation to changing ecological conditions in relation to flora and fauna which occurred in the late Pleistocene and early Holocene. The shell middens attributed to the Hoabinhians along the Malaysian coast perhaps represent a specialized method of wild foodprocurement among a community, or one of many food-related strategies employed by another group. Besides, the dates for the earliest appearance of shell middens remains uncertain as they have not been radiocarbon dated. Despite the fact that the subsistence strategies have been separated in the model, these communities also had a lot in common; food-producing communities continued with hunting and gathering, which probably remained a substantial part of their economy. In addition, contact and interactions between foragers and farmers are also very probable. Therefore, the long-term relationship between hunting and gathering and food-production is emphasized in the scheme. In the next stage, three factors contributed to agricultural development in peninsular Thailand. The hunting and gathering economy persisted but it decreased in importance regarding its influence on agriculture and later developing urbanism. During this stage, the relationship between hunting and gathering and agriculture is vague and its importance is slightly undefined. The indigenous food-producing economy, on the other hand, continued to thrive and
87

its cultivation strategies were probably refined during this stage. This stage is followed by the introduction of external agricultural practices or stimuli, which could have been the result of an immigration or contact between different groups. These practices might have included advanced cultivation methods and domesticated plant seeds. This way was also the pathway to subsequent urbanism and state formation. Overlapping in the archaeological record is also evident from Hoabinhian and Late Stone Age material culture. It further suggests an ongoing pattern of regional and local change which did not occur over night.
Table 4.1. Scheme over socio-economic dynamics during contemporary time periods in the research area by Lekenvall.

This rather simple scheme shows how the different subsistence strategies interacted and affected one another during the later phases of the Stone Age. Indigenous wild foodprocurement and wild food-production/cultivation have co-existed from the time of the earliest forms of local food-production. Pollen records and other environmental data have shown that this human intervention in the landscape started in the late Pleistocene, continued in the early Holocene and increased in intensity during later times. A total replacement of huntergatherers by food-producing communities never occurred in the late Pleistocene to the mid Holocene. Similarly, the incoming agriculturalists did not completely replace earlier foodproducers and hunter-gatherers in the mid-late Holocene. Tanurdijos contribution to the debate regarding the effect of predominant cultures on marginal areas is applicable to the proposed model of intrusive agriculturalist communities: globalization produces a global (global-local) culture in which the global88

predominant culture has been articulated by local subcultures according to their own values and symbols in order to maintain their identity. (Tanurdijo, 2006: 84). This concept is useful because it suggests that communities maintained some of their own cultural and perhaps economic characteristics despite having been exposed to external stimuli. In addition, it also shows that the so-called replacement of a marginal culture by a dominant counterpart is too simplistic. Rather, it should be acknowledged that we are dealing with an intertwined intermarriage of cultures, which all contributed to cultural and food-related transitions in the region. The archaeological and environmental record from lower peninsular Thailand suggest that local communities engaged in foraging and crop cultivation, probably within wide ranging and dynamic networks of contact and interactions. I would argue that mixed economies were already present in this area prior to the introduction of rice agriculture by expanding agriculturalists in the second millennium BC. Similarities in the local material culture suggest that indigenous Hoabinhian populations were assimilated into agriculturalist groups but the extent of the alleged assimilation cannot be assessed from the archaeological and environmental record presented in this study. Finally, and, as a conclusion of this study, I propose that both internal and external factors are responsible for the local/regional transition to agriculture. Therefore, both indigenous transitions/developments, as well as influences from outside the area, must have been important for the subsequent rise of urbanism and state formation in the region. On this basis it should not be stated that these socio-economic transitions occurred in total isolation, rather it points at a dynamic interplay between people throughout the history of the region. These local/regional interactions continued to increase as time went by. The aforementioned socioeconomic transitions, together with an increase of inter-regional exchange of exotic trade goods (circa 2000 years ago), which has been reported by Stargardt (e.g. 1973, 1983) and Allen (e.g. 1990, 1997, 2000) for the Satingphra Peninsula and Kedah, ultimately gave rise to urbanism and state formation on the peninsula.

89

90

REFERENCES CITED
Adi Haji Taha., 1983. Recent archaeological discoveries in Peninsular Malaysia (1976-1982). JMBRAS 56 (1), pp.47-63. 2005. Overview of Archaeology Research in Malaysia. In: J. Manavid. & P. Pisnupong. eds. 2005. Proceeding of the seminar Thailand-Malaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 43-48. Albrecht, G. Moser, J. & Pookajorn, S., 1993. Sakai Cave, Trang Province Southern Thailand: Preliminary report on the field work 1993. Tbingen: Institut fr Urgeschichte der Universitt Tbingen. Albrecht, G. & Moser, J. 1996., Recent Mani settlements in the Satun Province, southern Thailand. Report on the 1995 & 1996 field campaigns. Tbingen: Institut fr Urgeschichte der Universitt Tbingen. Allen, J. S., 1990. Agriculture, Hydraulics, and Urbanism at Satingpra (Review Article). Asian Perspectives, 28 (2), pp. 163-77. 1997. Inland Ankor, coastal Kedah: landscapes, subsistence systems and state development in early Southeat Asia. Bulletin of the Info-Pacific Prehistory Association, 16, pp. 79-87 - 2000. In Support of Trade: Coastal Site Location and Environmental Transformation in Early Historical-Period Malaysia and Thailand. Indo-Pacific Prehistory Association Bulletin, 20, pp. 62-78. Anderson, D. D., 1986. Recent Archaeological Findings from Southwestern Thailand. Muang Boran , 12 (4), pp. 97-98. 1988. Excavations of a Pleistocene Rock Shelter in Krabi and the Prehistory of Southern Thailand. In: P. Charoenwongsa. & B. Bronson, eds. 1988. Prehistoric Studies: The Stone and Metal Ages in Thailand, (Papers in Thai Antiquity 1). Bangkok: Thai Antiquity Working Group, pp. 4360. - 1990. Lang Rongrien rockshelter: a Pleistoceneearly Holocene archaeological site from Krabi, Southwestern Thailand. Philadelphia: University of Pennsylvania Press. - 1997. Cave Archaeology in Southeast Asia. Geoarchaeology: an international journal, 12 (6), pp. 607-638. - 2005. Late Pleistocene and Early to Middle Holocene Use of Caves in Peninsular Thailand. Asian Perspectives, 44 (1), pp. 137-153. Anderson, G. D.et al., 2007. Climate and culture change: exploring Holocene transitions. 2007. In: D. G. Anderson, A. K, Maasch. & H. D. Sandweiss, eds. 2007. Climate Change & Cultural Dynamics: A Global Perspective on Mid-Holocene Transitions. USA: Academic Press, pp. 1-23. Auetrakulvit, P., 2005. The subsistence of Prehistoric Man during Late Pleistocene-Early Holocene in Peninsular Thailand. In: J. Manavid. & P. Pisnupong, eds. 2005. Proceeding of the seminar ThailandMalaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 50-62. Archaeological survey for mining of Pirapon Mining, Tanbon Khuga Tai., 2007. Unpublished report. The 13th Fine Arts Department, Songkhla. (In Thai). Archaeological survey and excavation from Khao Rakian., 1986. Unpublished report. The 13th Fine Arts Department, Songkhla. (In Thai). Archaeological Study: Chiew Lam Project: By the Bureau of Archaeology. 1986., Fine Arts Department. (In Thai). Asawamas, D., 2008. Petrographic analysis of pottery from Kok Thong, Ranot Amphoe in Songkhla Province and Khao Lamu, Amphoe Khaunkalong in Satun Province. Fine Arts Department. (In Thai). Athens, S. J., Ward, J. V. & Murakami, G. M., 1996. Development of an agroforest on a Micronesian high island: prehistoric Kosraean agriculture. Antiquity, 70, pp. 834-46. Baker, C. & Phongpaichit, P., 2005. A history of Thailand. Cambridge: Cambridge University Press. Barker, G. et al., 2007. The human revolution in lowland tropical Southeast Asia: the antiquity and behavior of anatomically modern humans at Niah Cave (Sarawak, Borneo). Journal of Human Evolution, 52, pp. 243-261. Bentley, R. A. et al., 2007. Lapita migrants in the Pacifics oldest cemetery: Isotopic analysis at Teouma, Vanuatu, American Antiquity, 72, pp. 645-56. 91

Bayard. D. 1972. Early Thai bronze: analysis and new dates. Science: 176 (4042), pp. 1411-12. - 1979. The chronology of prehistoric metallurgy in Northeast Thailand: Silabhumi or Samrddhabhumi? In: R. B. Smith. & W. Watson, eds. 1979. Early South East Asia. New York & London: Oxford University Press for School of Oriental and African Studies, pp. 15-32. - 1992. Models, Scenarios, Variables and Suppositions: Approaches to the Rise of Social Complexity in Mainland Southeast Asia, 700 B.C. 500 A.D. In: I, Glover. P, Suchitta. & J, Villiers, eds. Early Metallurgy, Trade and Urban Centers in Thailand and Southeast Asia. Bangkok: White Lotus, pp. 13-38. - 1993. (Review). Higham. C., 1989. The Archaeology of Mainland Southeast Asia: From 10, 000 B.C. to the Fall of Angkor. Journal of Archaeological Science, 20, pp. 363-64. Bellwood, P., 1993. Cultural and Biological Differentiation in Peninsular Malaysia: The Last 10,000 years. Asian Perspectives, 32 (1), pp. 37-60. - 1997. Prehistory of the Indo-Malaysian Archipelago. 2nd ed. Honolulu: University of Hawaii Press. - 2001. Southeast Asia Neolithic and Early Bronze. In: P. N. Peregrine & M. Ember, eds. Encyclopedia of Prehistory: East Asia and Oceania. Volume 3. United States: Human Relations Area Files, Inc, pp. 287-306. - 2005. The First Farmers: the Origins of Agricultural Societies. United Kingdom: Blackwell Publishing Ltd. - 2006. Austronesian Prehistory in Southeast Asia: Homeland, Expansion and Transformation. In: P. Bellwood. J. J, Fox. & T. T. Darrel, eds. 2006. The Austronesians: historical and comparative perspectives. 2nd ed. The Australian National University: Anu E Press, pp. 103-14. - 2007. Prehistory of the Indo-Malaysia Archipelago. 2nd ed. Canberra: ANU E-Press. Bellwood. P. & Diamond. J., 2005. On Explicit Replacement Models in Island Southeast Asia: A reply to Stephen Oppenheimer, World Archaeology, 37 (4), pp. 503-06. Bentley Carter, G., 1986. Indigenous States of Southeast Asia. Annual Review of Anthropology, 15, pp. 275-305. Bentley, R. A. et al., 2007. Lapita migrants in the Pacifics oldest cemetery: Isotopic analysis at Teouma, Vanuatu, American Antiquity, 72, pp. 645-56. Bernard, R. H., 2006. Research Methods in Anthropology: Qualitative and Quantitative Approaches. 4th ed. UK: AltaMira Press. Binford, L. R., 1980. Willow Smoke and Dogs Tails: Hunter-Gatherer Settlement Systems and Archaeological Site Formation, American Antiquity. 45 (1), pp. 4-20. Bird, W. D. & OConnell, F. J., 2006. Behavioral Ecology and Archaeology. Journal of archaeological research, 14, pp. 143-88. Blench, R., 17th congress of the Indo-Pacific Prehistory Association. Taipei, Taiwan 9 to 15 September 2002. Fruits and arboriculture in the Indo-Pacific region. United Kingdom: Cambridge. Bronson, B., 1985. Notes on the Iron Age in Thailand. Journal of the Siam Society, 73(1), pp: 205-25. - 1992. Patterns in the early Southeast Asian Metal trade. In: I, Glover. P. Suchitta. & J. Villiers, eds. 1979. Early Metallurgy, Trade, and Urban Centers in Thailand and Southeast Asia. Bangkok: White Lotus, pp. 63-114. Bruun, T.B. Mertz, O. & Elberling, B., 2006. Linking yields of upland rice in shifting cultivation to fallow length and soil properties. Agriculture, Ecosystems and Environment, 113, pp. 139-49. Buckley-Ebrey, P., 1996. The Cambridge Illustrated History of China. Cambridge: Cambridge University Press. Chang, K-C., 1986. The Archaeology of Ancient China. 4th ed. New Haven and London: Yale University Press. Chia, S., 2005. The Prehistory of the Northern Malay Peninsula during the Holocene era, with special reference to Southern Thailand. In: J. Manawid. & P. Pisnupong, eds. 2005. Proceeding of the seminar Thailand-Malaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 79-85. Chaimongkol, S., 2005. Prehistorical Cave Arts Studied By the Southern Thailand Archaeological Research Project. In: J. Manawid. & P. Pisnupong, eds. 2005. Proceeding of the seminar ThailandMalaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 90-94. 92

Charoenpot, S. Incherdchai, J. & Kojaranont, R., 2008. 44 National Museums of Thailand. Group of Promotion and Public Relations, Fine Arts Department. Bangkok: Rungship Publisher (1977) Company Limited. Cutler, K.B. et al., 2003. Rapid Sea-level fall and deep-ocean temperature change since the last interglacial period. Earth and Planetary Science Letters, 206, pp. 253-71. Darnsawasdi, R. & Chitpong, P., 2002. Wetland Systems in Southern Thailand. The Essential Resources For Sustainable Regional Development. In: M. M. Holland. M. L. Warren. & J. A. Stanturf, eds. 2002. Proceedings of a conference on sustainability of wetlands and water resources: how well can riverine wetlands continue to support society in the 21st century. Oxford, Mississippi May 23 25 2002. Gen. Tech. Rep. SRS-50. Asheville, NC: U.S. Department of Agriculture, Forest Service, Southern Research Station, pp. 43-46. Denham, T. et al., 2003. Origins of agriculture at Kuk Swamp in the Highlands of New Guinea. Science, 301, pp. 189-93. Denham, T., 2004. The Roots of Agriculture and Arboriculture in New Guinea: Looking beyond Austronesian Expansion, Neolithic Packages and indigenous Origins. World Archaeology, 36 (4), pp. 610620. Denham, T., 2005. Envisaging early agriculture in the Highlands of the New Guinea: landscapes, plants and practices. World Archaeology, 37 (2), pp. 290306. Dheeradilok, P., 1992. An introduction to Geology in Thailand. In: National Conference on Geologic Resources of Thailand: Potential for Future Development. 17-24 November, 1992. Department of Mineral Resources. Bangkok, Thailand, pp. 737-52. Diamond, J., 1997. Guns, Germs and Steel: A short history of everybody for the last 13, 000 years. London: Vintage. Dudal, R., 2005. Soils of Southeast Asia. In: A. Gupta, ed. Physical Geography of Southeast Asia. Oxford University Press, pp. 94-104. Duff, R., 1970. Stone Adzes of Southeast Asia. An Illustrated Typology by Roger Duff. Canterbury Museum, Bulletin No. 3. Fairbairn, A., 2005. An Archaeobotanical Perspective on Holocene Plant-Use Practices Lowland Northern New Guinea. World Archaeology, 37 (4), pp. 487-502. Fairbairn, A. & Swadling, P., 2005. Re-dating mid-Holocene betelnut (Areca catechu L) and other plant use at Dongan, Papua New Guinea. Radiocarbon, 47 (3), pp. 377-382. Forestry Statistics of Thailand. 2007. Forest Land Assessment by Province in 2004 2006. [Online] Available at: http://www.forest.go.th/stat/stat50/stat2550.htm [Accessed 9 October 2009] Foss, F. E., Moornamn, F. R. & Rieger, S., 1983. Inceptisols. In: L. P. Wilding, N. E. Smeck. & G. F. Hall. eds. Pedogenesis and soil taxonomy, II. The Soil Orders. Amsterdam, Elsevier, pp. 355-81. Frink, S. D., 2003. Temporal Values in a Universe of Turbations: Application of the OCR Carbon Dating Procedure in Archaeological Site Formational Analyses and Pedogenic Evaluations. In: Soils and Archaeology: Papers of the 1st International Conference on Soils and Archaeology. Szzhalombatta, Hungary 30 May 3 June 2001. BAR International Series 1163, 2003. Oxford: Archaeopress, Publishers of British Archaeological Reports. Glover, I., 1990. Early Trade Between India and Southeast Asia: A Link in the Development of a World Trading System. Hull University: Centre for South-East Asian Studies. - 1991. The Bronze-Iron Age in western Thailand. Paper presented at the conference: The High Bronze Age of Southeast Asia and South China. Hua Hin, Thailand 1991. Glover, I, Suchitta, P. & Villiers, J., 1992. Early metallurgy, trade and urban centres in Thailand and southeast Asia. Bangkok: White Lotus. Gullaprawit, C., 2002. Thailand. In: Zanko, M, ed. 2002. The handbook of human resource management policies and practices in Asia-Pacific economies. Massachusetts: Edward Elgar Publishing Limited, pp. 575-628. Gorman, C., 1970. The Hoabinhian and after: subsistence patterns in Southeast Asia during the late Pleistocene and Early Recent period. World Archaeology, 2 (3), pp. 300-20. - 1972. Excavations at Spirit Cave, North Thailand. Asian Perspectives, 13, pp. 79-107. 93

Gupta, A., 2005. Landforms of Southeast Asia. In: A, Gupta, ed. The physical geography of Southeast Asia. Gupta. 2005. New York: Oxford University Press, pp. 38-64. Haberle, S. G., 2003. The emergence of an agricultural landscape in the highlands of New Guinea. Archaeology in Oceania, 38, pp. 149-58. Habu, J., 1996. Jomon Sedentism and Intersite Variability: Collectors of the Early Jomon Moroiso Phase in Japan. Arctic Anthropology, 33 (2), pp. 38-49. Halcrow, S. Tayles, N. & Livingstone, V., 2008. Infant Death in Late Prehistoric Southeast Asia. Asian perspectives, 47 (2), pp. 371-403. Hanebuth, T. J. J. Stattegger, K. & Grootes. P. M., 2000. Rapid Flooding of the Sunda Shelf: A LateGlacial Sea-Level Record. Science, 12 (288), pp. 1033-35. Hanebuth, T. J. J. & Stattegger. K., 2004. Depositional sequences on a late Pleistocene-Holocene tropical siliciclastic shelf (Sunda Shelf, southeast Asia). Journal of Asian Earth Sciences, 23, pp. 113-26. Hartemink, E. A., 2004. Soils of the tropics. Geoderma, 123, pp. 37375. Harris. R. D., 1996. The Origins and Spread of Agriculture and Pastoralism in Eurasia. London: UCL Press. Healy, F. P., 2007. The anthropology of Mesoamerican Caves. Reviews in Anthropology, 36, pp. 245-78. Heaney, R, L., 1991. A Synopsis of Climatic and Vegetational Change in Southeast Asia. Climatic Change, 19, pp. 53-61. Horton, R.P. et al., 2005. Holocene sea levels and palaeoenvironments, Malay-Thai Peninsula, Southeast Asia. The Holocene, 15 (8), pp. 1199-1213. Higham, C. F. W., 1989. The Archaeology of Mainland Southeast Asia. Cambridge: Cambridge University Press. - 1996. The Bronze Age of Southeast Asia. Cambridge: Cambridge University Press. - 1998. Noen U-Loke and the implications for the origins of early states. Paper presented at the 16th Congress of the Indo-Pacific Prehistory Association. Melaka, Malaysia 1998. - 2002a. Early Cultures of Mainland Southeast Asia. Bangkok: River Books. - 2002b. Languages and Farming Dispersals: Austroasiatic Languages and Rice Cultivation. In P. Bellwood & A. C. Renfrew, eds. 2002. Examining the Farming/Language Dispersal hypothesis. Cambridge: The McDonald Institute, pp. 223-32 - 2005a Complex Societies of East and Southeast Asia. In C. Scarre, ed. 2005. The Human Past. World Prehistory and the Development of Human Societies. London: Thames & Hudson Ltd, pp. 552-93. - 2005b. Corpus of complete pottery vessels. In: C. Higham, C. & P. Thosarat, eds. 2005. The Origins of the Civilization of Angkor. Volume 1. The Excavation of Ban Lum Khao. Bangkok: The Fine Arts Department of Thailand, pp. 239-300. Higham, C. F. W. & Kijngam, A., 1979. Ban Chiang and Northeast Thailand: The Palaeoenvironment and Economy. Journal of Archaeological Science, 6, pp. 211-33. Higham, C. F. W. & Bannanurag, R., 1990. The Excavation of Khok Phanom Di: Volume I, The Excavation, Chronology and Human Burials. London: The Society of Antiquaries of London and Thames and Hudson. Higham, C. F. W. & Bannanurag, R., 1991. The Excavation of Khok Phanom Di: Volume II: The Biological Remains Part 1. London: The Society of Antiquaries of London and Thames and Hudson. Higham, C.F.W. & Thosorat, R., 1993. Khok Phanom Di: Volume III: The Material Culture Part 1. London: The Society of Antiquaries of London. Higham, C.F.W. & Thosorat, R., 1994. Khok Phanom Di: Prehistoric Adaptation to the World's Richest Habitat. Fort Worth : Harcourt Brace Jovanovich. Higham, C. F.W. & Thosarat, R., 1998a. Prehistoric Thailand: From Early Settlement to Sukhothai. Bangkok: River Books. Higham, C. F. W. & Thosarat, R., 2006. Ban Non Wat: the first three seasons. In: E. A: Bacus. I. Glover & V. C. Pigott, eds. 2006. Uncovering Southeast Asias Past. Singapore: NUS press, pp. 98-104. Higham, C. F. W. & Higham. T. F. G., 2009. A New chronological framework for prehistoric Southeast Asia, based on a Bayesian model from Ban Non Wat. Antiquity, 83, pp. 125-44. Hill. R. D., 2009. The cultivation of perennial rice, an early phase in Southeast Asia agriculture? Journal of Historical Geography, xxx, pp. 19. 94

Hutterer. K., 1983. The Natural and Cultural History of Southeast Asian Agriculture. Anthropos, 78, pp. 169-212. Jintanugool, J. & Round, P. D., 1989. Thailand. In: D. A. Scott, ed. 1989. A Directory of Asian Wetlands. Switzerland: IUCN, Gland, pp. 667-736. Jenny, H., 1994. Factors of Soil Formation: A System of Quantitative Pedology. United States of America: Dover Publications, Inc. Joint Project Kingdom of Thailand and United Nations Special Fund. Executed For the UN by the Food and Agriculture Organization, 1972. Main Landforms of Thailand as Related to Soil Formation. Scale: 2, 500, 000. Bangkok: Printed by the Soil Survey Division. [Electronic Print] Available at: http://library.wur.nl/isric/index2.html?url=http://library.wur.nl/WebQuery/isric/19030 [Accessed 15 October 2009]. Karlstrm, A., 2009. Preserving impermanence: the creation of heritage in Vientiane, Laos. Ph. D. Uppsala: Uppsala University. Kealhofer, L., 2002. Changing Perceptions of Risk: The Development of Agro-Ecosystems in Southeast Asia. American Anthropologist, New Series, 104 (1), pp. 178-94. 2003. Looking into the Gap: Land Use and the Tropical Forests of Southern Thailand. Asian Perspectives, 42 (1), pp. 72-95. Keys to Soil Taxonomy: Soil Survey Staff., 1996. 7th ed. United States Department of Agriculture Natural Resources Conservation Service. Kirch, V. P., 1989. Second millennium B. C. arboriculture in Melanesia: Archaeological evidence from the Mussau Islands. Economic Botany, 43 (2), pp. 225-40. Kusimba, B. S., 1998. Hunter-Gatherer Land Use Patterns in Later Stone Age East Africa. Journal of Anthropological Archaeology, 18, pp. 165200. Krigbaum, J., 2003. Neolithic subsistence patterns in northern Borneo reconstructed with stable carbon isotopes of enamel. Journal of Anthropological Archaeology, 22, pp. 292304. Klln, A., 2004. And Through Flows the River: archaeology and the pasts of Lao Pako. Ph. D. Uppsala: Uppsala University. Latinis, K. D., 2000. The development of subsistence system models for Island Southeast Asia and Near Oceania: the nature and role of arboriculture and arboreal-based economies. World Archaeology, 32 (1), pp. 41- 67. Liu, L. Ah-Lee, G. Jiang, L. & Zhang, J., 2007. Evidence for the early beginning (c .9000 cal BP) of rice domestication in China: a response. The Holocene, 17, pp. 1059-68. Limwijitwong, S., 2005. Historic Rock Arts in the South of Thailand. In: J. Manawid. & P. Pisnupong, eds. 2005. Proceeding of the seminar Thailand-Malaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 95-101. Lwenborg, D., 2010. Excavating the Digital Landscape. GIS analyses of social relations in central Sweden in the 1st millennium AD. Ph. D. Uppsala: Uppsala University. Macamo, S., 2005. Privileged Places in South Central Mozambique: The Archaeology of Manykeni, Niamara Songo and Degue-Mufa. Ph. D. Uppsala: Uppsala University. Mandel, R. D. & Arthur Bettis III, A., 2001. Use and Analysis of Soils by Archaeologists and Geoscientists: A North American Perspective. In: P, Goldberg. T. V, Holliday. & C, Reid Ferring, eds. Earth Sciences and Archaeology. New York: Kluwer Academic/Plenum Publishers, pp. 173-95. Manguin, P. Y., 2000. The archaeology of the early maritime polities of Southeast Asia. In: P, Bellwood & I, Glover, eds. 2000. Southeast Asia: Origins to Civilisation. Oxford: Curzon Press. Ch.11. Matthews, P. J. & Gosden, C., 1997. Plant remains from waterlogged sites in the Arawe Islands, West New Britain Province, Papua New Guinea: implications of the history of plant use and domestication. Economic Botany, 51, pp. 121-33. Matsamura, H. & Zuraina, M., 1999. Metric Analyses of an Early Holocene Human Skeleton from Gua Gunung Runtuh, Malaysia. American journal of physical anthropology, 109, pp. 32740. Matsamura, H. & Pookajorn, S., 2005. A morphometric analysis of the Late Pleistocene Human Skeleton from the Moh Khiew Cave in Thailand. Homo Journal of Comparative Human Biology, 56 (2), pp. 93-118.

95

Matsamura, H. et al., 2008. Terminal Pleistocene human skeleton from Hang Cho Cave, northern Vietnam: implications for the biological affinities of Hoabinhian people. Anthropological Science. 116 (3), pp. 201-17. Mudar, K. & Anderson, D., 2007. New Evidence for Southeast Asian Pleistocene Foraging Economies: Faunal Remains from the early Levels of Lang Rongrien Rockshelter, Krabi, Thailand. Asian Perspectives, 46 (2), pp. 298-334. Murillo-Barroso, et al., 2010. Khao Sam Kaeo an archaeometallurgical crossroads for trans-asiatic technological traditions. Journal of Archaeological Science xxx, pp. 1-12 (in press). Mercader. J., 2002. Forest People: The Role of African Rainforests in Human Evolution and Dispersal. Evolutionary Anthropology, 11, pp. 11724. Needham, J. & Bray, F., 1984. Science and Civilization in China: Volume 6, Biology and Biological Technology. Part 2, Agriculture: Agriculture Pt. 2. Cambridge: Cambridge University Press. Neff, H. et al., 2006. Early Maya Adaptive Patterns. Mid-Late Holocene Paleoenvironmental Evidence From Pacific Guatemala. Latin American Antiquity, 17 (3), pp. 287-315. Neergard, A. Magid, J. & Mertz, O., 2008. Soil erosion from shifting cultivation and other smallholder land use in Sarawak, Malaysia. Agriculture , Ecosystems & Environment, 125 (1-4), pp. 182-90. Ooi, K. G. ed., 2004. Southeast Asia: A Historical Encyclopedia from Angkor Wat to East Timor. Volume 1. United States: ABC-CLIO, Inc. Omakupt, M. 1972. Atlas of Thailand. Major Land Uses. Scale: 2,500,000. Royal Thai Survey Department. [Electronic Print] Available at: http://eusoils.jrc.ec.europa.eu/esdb_archive/EuDASM/Asia/maps/TH2000_3LU.htm [Accessed 17 October 2009]. Oota, H. et al., 2001. Genetic Study of the Paleolithic and Neolithic Southeast Asians. Human Biology, 73 (2), pp. 225-31. Panapitukkul, N. et al., 2005. Hydrogeomorphological Controls on Groundwater Quality in the Ratthapum Catchment (Songkhla Lake Basin), Thailand: Water, Air, and Soil Pollution. Focus, 5, pp. 14963. Paz, V., 2005. Rock shelters, Caves, and Archaeobotany in Island Southeast Asia. Asian Perspectives, 44 (1), pp. 108-18. Pendleton, R. L., 1949. Provinsional Map of the Soils and Surface Rocks of the Kingdom of Thailand. Scale: 2, 500, 000. Bangkok: Cadastral Survey Office. [Electronic Print] Available at: http://eusoils.jrc.ec.europa.eu/esdb_archive/EuDASM/Asia/maps/TH2003_SO.htm [Accessed 17 October 2009]. Piperno, D. R. & Pearsall, D. M., 1998. The Origins of Agriculture in the Lowland Neotropics. San Diego, CA: Academic Press. Pookajorn, S., 1991. Preliminary Report of Excavations at Moh Khiew Cave, Krabi Province, Sakai Cave, Tran Province and Ethnoarchaeological Research of Hunter-Gatherer Group, so-called Sakai or Semang at Trang Province. Bangkok: Silpakorn University Press. 1985. Ethnoarchaeology with the Phi Tong Luang (Mlabrai): Forest Hunters of Northern Thailand. World Archaeology, 17, pp. 206-21. 1994. Final Report of Excavations at Moh Khiew Cave, Krabi Province; Sakai Cave Trang Province and Ethnoarcheological Research of Hunter-Gatherer Group, so called Sakai or Semang at Trang Province. Bangkok: Silpakorn University Press. 1996. Human activities and environmental changes during the Late Pleistocene to Middle Holocene in Southern Thailand and Southeast Asia. In: L. Straus. et al, eds. 1996. Humans at the End of the Ice Age: The Archaeology of the Pleistocene-Holocene Transition. New York: Plenum Press. Ch.10. Pyramarn, K., 1989. New evidence on plant exploatation and environment during the Hoabinhian (Late Stone Age) from Ban Kao Caves, Thailand. In : R. D, Harris. & C, Hillman, C, eds. 1989. Foraging and Farming: The Evolution of Plant Exploitation. London: Unwin Hyman Ltd, pp. 282-91. Pietrusewsky, M. & Douglas, M. T., 2002. Ban Chiang, A Prehistoric Village Site in Northeast Thailand I: The Human Skeletal Remains. Philadelphia: University of Pennsylvania. Reynolds, T. G. D., 1992. Excavations at Banyan Valley Cave, Northern Thailand: A Report on the 1972 Season. Asian Perspectives, 31 (1), pp. 77-97. 96

Rosenswig, M. R., 2006. Sedentism and food production in early complex societies of the Soconusco, Mexico. World Archaeology, 38 (2), pp. 330-50. Rugmail, W. et al., 2008. A Late Pleistocene palynoflora from the coastal area of Songkhla Lake, southern Thailand. Science Asia, 34, pp. 13745. Sanchez, P. A. & Buol, S. W., 1975. Soils of the Tropics and the World Food Crisis. Science, 188, pp. 598-603. Scarre, C., 2005. The Human Past: World Prehistory & the Development of Human Societies London: Thames & Hudson. Scoffin, T. P. & Le Tissier , M. D. A., 1998. Late Holocene sea level and reef-flat progradation. Phuket South Thailand. Coral Reeds, 17, pp. 273-76. Shoocongdej, R., 2000. Forager Mobility Organization in Seasonal Tropical Environments of Western Thailand. World Archaeology, 32 (1), pp. 14-40. Smith, A. B., 2005. The concepts of Neolithic and Neolithisation for Africa? Before Farming, 2, pp. 1-6. Soil Survey Staff., 1999. Soil Taxonomy. A basic system of soil classification for making and interpreting soil surveys. 2nd ed. Agricultural Handbook 436. Natural Resources Conservation Service, USDA. Washington DC, USA. Solheim II. W. G., 1968. Early bronze in northeastern Thailand. Carr. Anthropol, 9, pp. 5962. - 1976. Reflection on the New Data of Southeast Asian Prehistory: Austronesian Origin and Consequence. Asian Perspectives, 18 (2), pp. 146-60. - 2003. Southeast Asian Pottery and its Spread. In: J, Miksic, ed. Earthenware ceramics in Southeast Asia. Singapore: Singapore University Press, pp. 1-21. - 2006. Archaeology and culture in Southeast Asia: unraveling the Nusantao. Philippines: The University of Philippines Press. Spriggs, M., 1996. Early agriculture and what went before in Island Melanesia: continuity or intrusion? In: R. D, Harris. 1996, ed. The Origins and Spread of Agriculture and Pastoralism in Eurasia. London: UCL Press, pp. 525-37. Spriggs, M. OConnor, S. & Veth, P., 2002. Vestiges of Early Pre-Agricultural Economy in the Landscape of East Timor. Recent Research. In: A, Karlstrm, & A, Klln, A, eds. 2002. Fishbones and Glittering Emblems. Stockholm: Museum of Far Eastern Antiquities, pp. 49-58. Simajuntak, T., 2005. Indonesia Southeast Asia: Climates, settlements, and cultures in Late Pleistocene. C. R. Palevol, 5, pp. 37179. Srisuchat, A., 1987. Prehistoric caves and some important prehistoric sites in southern Thailand. In: country report of Thailand for SPAFA Seminar in Prehistory of Southeast Asia (T-W11). Bangkok, Surat Thani, Phangnga, Phuket, and Krabi, Thailand, January 12-25 1987, pp. 103-17. 2003. Earthenware from Archaeological Sites in Southern Thailand: The First Century BC to the Twelfth Century AD. In: J, Miksic, ed. 2003. Earthenware in Southeast Asia: Proceedings of the Singapore Symposium on Premodern Southeastern Earthenwares. Singapore University Press, pp. 249-60. 2005. Overview of Southern Thai-Malay Peninsula Archaeology. In: J, Manavid. & P, Pisnupong, eds. Proceeding of the seminar Thailand-Malaysia: Malay Peninsula Archaeology Programme. Bangkok: Samaphan Publishing CO., LTD, pp. 32-42. Stargardt, J., 1973. Southern thai waterways: archaeological evidence on agriculture, shipping and trade in the srivijayan period. Man, New Series, 8 (1), pp. 5-29. 1976. Man's impact on the ancient environment of the satingpra, south Thailand. I. The natural environment and natural change. Journal of Biogeography, 3 (3), pp. 211-28. 1977. Man's impact on the ancient environment of the satingpra, south Thailand. II. Ancient agriculture. Journal of Biogeography, 4 (1), pp. 35-50. - 1983. Satingpra I: The Environmental and Economic Archaeology of South Thailand. Oxford: British Archaeological Reports (BAR). 1992: 'Water for Court or Countryside? Archaeological Evidence from Central Burma and South Thailand Reconsidered. ' In: J, Riggs, ed. 1992. London: School of Oriental and African Studies, pp. 59-72.

97

2003. Mapping the Mind. Some Cultural Cargos of the Sea-Trade in Southeast Asia, 5th 13th century. In: A, Karlstrm. & A, Klln, eds. 2002. Fishbones and Glittering Emblems. Stockholm: Museum of Far Eastern Antiquities, pp. 103-17. Stark, M. T., 2003. The chronology, technology and contests of earthenware ceramics in Cambodia. In: J, Miksic, ed. Earthenware ceramics in Southeast Asia. Singapore: Singapore University Press. pp. 208-29. 2006. Early Mainland Southeast Asian Landscapes in the First Millennium A.D. Annual Review of Anthropology, 35, pp. 407-32. Stark, M. T & Allen. J. S., 1998. The Transition to History in Southeast Asia: An Introduction. International Journal of Historical Archaeology 2 (3), pp. 163-74. Stenseth. C.N. et al., 2002. Ecological Effects of Climate Fluctuations. Science, 297 (5585), pp. 1292 96. Strahler. A, H. & Strahler, A., 2005. 3rd ed. Physical Geography: Science and Systems of the Human Environment. United States of America: Wiley. Steila, D. & Pong, E. T., 1989. 2nd ed. The Geography of Soils: Formation, Distribution and Management. United States: Rowman & Littlefield. Tanudirjo, A. D., 2006. The Dispersal of Austronesian-speaking people and the Ethnogenesis of Indonesian People. In: T, Simanjuntak. H. E.I, Pojoh. & M, Hisayam, eds. 2006. Austronesian Diaspora and the Ethnogenesis of People in Indonesian Archipelago: Proceedings of the International Symposium. Jakarta: LIPI Press, pp. 83-89. Tayles, N., 1999. The excavation of Khok Phanom Di: A prehistoric site in Central Thailand. Vol. 5: The People. London: The Society of Antiquaries of London, Research Report L. Thomas, R. F., 2002. An evaluation of Central-Place Foraging among Mollusk Gatherers in Western Kiribati, Micronesia: Linking Behavioral Ecology with Ethnoarchaeology. World Archaeology , 34 (1), pp. 182-208. Tjia. H. D., 1996. Sea-level changes in the tectonically stable Malay-Thai Peninsula. Quaternary International, 31, pp. 95-101. Tjia, H. D. & Liew, K. K., 1996. Changes in tectonic stress field in northern Sunda Shelf basins. London: Geological Society, Special Publications, 106, pp. 291-306. Tougard, C., 2001. Biogeography and migration routes of large mammal faunas in South.-East Asia during the Late Middle Pleistocene: focus on the fossil and extant faunas from Thailand. Palaeogeogr Palaeoclimat Palaeoecol, 168, pp. 337-58. Tougard, C. & Montuire, S., 2006. Pleistocene paleoenvironnemental reconstructions and mammalian evolution in South-East Asia: focus on fossil faunas from Thailand. Quaternary Science Review, 25, pp. 126-41. Treerayapiwat, C., 2005. Pattern of the Habitation and Burial Activity in the Ban Rai Rockshelter, Northwest Thailand. Asian Perspectives, 44 (1), pp. 231-45. United States Department of Agriculture (USDA) Natural Resources Conservation Center. Soil Taxonomy : A Basic System of Soil Classification for Making and Interpreting Soil Surveys., 1999. 2nd ed. Valladas, H., 2003. Direct radiocarbon dating of prehistoric cave paintings by accelerator mass spectrometry. Measurement Science and Technolog. 14, pp. 1487-92. van Heekeren, H.R. & Knuth, C. E., 1967. Archaeological Excavations in Thailand 1: SaiYok, Stone-Age Settlements in the Kanchanaburi Province. Copenhagen: Munksgaard. Voris, K. H., 2000. Maps of Pleistocene sea levels in Southeast Asia: Shorelines, river systems and time durations. Journal of Biogeography, 27, pp. 1153-67. Vijarnsorn, P. & Jongpakdee, C., 1979. General Soil Map of Thailand. Scale: 1, 000. 000. [electronic print] Available at: http://eusoils.jrc.ec.europa.eu/esdb_archive/EuDASM/Asia/maps/TH2005_3SO.htm [Accessed 18 October 2009]. Virgo, J. K. & Holmes, A. D., 1977. Soil and Landform Features of Mountainous Terrain in South Thailand. Geoderma, 18 (3), pp. 207-25. Waelbroeck, C. et al., 2002. Sea-level and deep water temperature changes derived from benthonic foraminifera isotopic records. Quaternary Science Reviews, 21, pp. 295-305. Wang, L. et al., 1999. East Asian monsoon climate during the Late Pleistocene: high-resolution sediment records from the South China Sea. Marine Geology, 156 (1), pp. 24584. 98

Wheatley, D., 1995. Cumulative viewshed analysis: a GIS-based method for investigating intervisibility, and its archaeological application. In: G, Lock. & Z, Stancic, eds. 1995. Archaeology and Geographical Information Systems. London: Francis Ltd, pp. 171-86. White. E. R., 1997. Principles and Practise of Soil Development: The soil as a Natural Resource. 3rd ed. Cornwall: Blackwell Science Ltd. White, J, C., 1982. The Discovery of a Lost Bronze Age, Ban Chiang. Philadelphia: University of Pennsylvania Press. 1995. Incorporating Heterarchy into Theory on Socio-Political Development: The Case from Southeast Asia. In: R. Ehrenreich, C. Crumley. & J. Levy, eds. 1995. Heterarchy and the Analysis of Complex Societies. Washington D.C: American Anthropological Association, pp. 101-23. White, J, & Pigott, V., 1996. From community craft to regional specialization: Intensification of copper production in pre-state Thailand. In: B. Wailes, ed. 1996. Craft Specialization and Social Evolution: In Memory of V. Gordon Childe. Philadelphia: University of Pennsylvania Museum and Archaeology and Anthropology, pp. 151-75. White, J. C. et al., 2004. Vegetation changes from the late Pleistocene through the Holocene from three areas of archaeological significance in Thailand. Quaternary International, 113, pp. 111-32. Yasuda, Y. ed., 2002. The Origins of Pottery and Agriculture. New Delhi: Roli Books. Yasuda, Y. & Vasant. S. eds., 2004. Monsoon and Civilization. New Delhi: Roli Books. Zhao, C. et al., 2004. Early polished stone tools in south China evidence of the transition from Palaeolithic to Neolithic. Documenta Praehistorica XXXI. Neolithic studies, 11, pp. 131-37. Zuraina M., 1994. The excavation of Perak man, an epi-Palaeolithic burial at Gua Gunung Runtuh. In: M, Zuraina, ed. 1994. The excavation of Gua Gunung Runtuh and the discovery of the Perak man in Malaysia. Kuala Lumpur: Department of Museums and Antiquity, Malaysia, pp. 2347. 2003. Archaeology in Malaysia = Arkeologi di Malaysia. Centre for Archaeological Research Malaysia, Universiti Sains Malaysia, Penang 2005. Perak Man and Other Prehistoric Skeletons of Malaysia. Penang, Malaysia: University of Science Malaysia.

99