Citation
Lithic technology and mobility in late Pleistocene southeast Asia

Material Information

Title:
Lithic technology and mobility in late Pleistocene southeast Asia a whole assemblage approach to assessing technological behavior in modern and archaic human populations
Creator:
Cicero, Daniel Ryan ( author )
Place of Publication:
Denver, CO
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
1 electronic file (105 pages) : ;

Thesis/Dissertation Information

Degree:
Master's ( Master of arts)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Anthropology, CU Denver
Degree Disciplines:
Anthropology
Committee Chair:
Stone, Tammy
Committee Members:
Riel-Salvatore, Julien
Musiba, Charles

Subjects

Subjects / Keywords:
From 10 thousand to 2 million years ago ( fast )
Stone implements ( lcsh )
Paleoecology -- Pleistocene ( lcsh )
Paleoecology ( fast )
Pleistocene Geologic Epoch ( fast )
Stone implements ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
Assessing human behavior during the Upper Pleistocene of Southeast Asia from a technological standpoint has long been viewed as problematic due to the indistinctiveness of the stone tool artifacts. Often expediently produced from poor quality stone and lacking standardization in tool form, these 'low-input' industries have been taken as evidence that the lithic economies of Pleistocene hunter-gatherers here were likely deemphasized in favor of locally abundant organic materials. While the unique prehistoric ecological constraints appear to be the primary limiting factor in the production of complex lithic industries, the extent to which lithic economies of Late Pleistocene foragers in Southeast Asia were truly liberated from the type of embedded procurement strategies and broader landscape-use patterns inferred from contemporaneous hunter-gatherer stone tool traditions has not been explored at the regional scale. Using a new methodology to reconstruct technological organization from a whole assemblage perspective, the reconstruction of artifact discard patterns and mobility patterns indicates that lithic technologies of Southeast Asia were an integral component of the technological system and were structured in ways fundamentally similar to hunter-gatherers of the Late Pleistocene foragers of Western Eurasia. The findings from this thesis suggest that stone tools of Southeast Asia can be modeled within a larger anthropological framework emphasizing the relationship between mobility strategies and human technological behavior. Furthermore, the preliminary results indicate that non-human assemblages, which have been difficult to distinguish from modern assemblages typologically, deviate from the expectations of the model and could be behaviorally significant and not simply a broader function of the environment.
Bibliography:
Includes bibliographic resource.
System Details:
System requirements: Adobe Reader.
General Note:
n3p
Statement of Responsibility:
by Daniel Ryan Cicero.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
on10198 ( NOTIS )
1019880730 ( OCLC )
on1019880730
Classification:
LD1193.L43 2017m C54 ( lcc )

Downloads

This item has the following downloads:


Full Text
LITHIC TECHNOLOGY AND MOBILITY IN LATE PLEISTOCENE SOUTHEAST
ASIA: A WHOLE ASSEMBLAGE APPROACH TO ASSESSING TECHNOLOGICAL BEHAVIOR IN MODERN AND ARCHAIC HUMAN POPULATIONS
by
DANIEL RYAN CICERO B.A., University of Colorado, Denver, 2013
A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements
for the degree of Master of Arts Anthropology Program
2017


©2017
DANIEL RYAN CICERO ALL RIGHTS RESERVED


This thesis for the Master Arts Anthropology degree by Daniel Ryan Cicero Has been approved for the Anthropology Department by
Tammy Stone, Chair Julien Riel-Salvatore Charles Musiba
Date: May 13th, 2017
111


Cicero, Daniel Ryan (M.A. Anthropology)
Lithic Technology and Mobility in Late Pleistocene Southeast Asia: A Whole Assemblage Approach to Assessing Technological Behavior in Modem and Archaic Human Populations Thesis directed by Associate Professor Tammy Stone
ABSTRACT
Assessing human behavior during the Upper Pleistocene of Southeast Asia from a technological standpoint has long been viewed as problematic due to the indistinctiveness of the stone tool artifacts. Often expediently produced from poor quality stone and lacking standardization in tool form, these ‘low-input’ industries have been taken as evidence that the lithic economies of Pleistocene hunter-gatherers here were likely deemphasized in favor of locally abundant organic materials. While the unique prehistoric ecological constraints appear to be the primary limiting factor in the production of complex lithic industries, the extent to which lithic economies of Late Pleistocene foragers in Southeast Asia were truly liberated from the type of embedded procurement strategies and broader landscape-use patterns inferred from contemporaneous hunter-gatherer stone tool traditions has not been explored at the regional scale. Using a new methodology to reconstruct technological organization from a whole assemblage perspective, the reconstruction of artifact discard patterns and mobility patterns indicates that lithic technologies of Southeast Asia were an integral component of the technological system and were structured in ways fundamentally similar to hunter-gatherers of the Late Pleistocene foragers of Western Eurasia. The findings from this thesis suggest that stone tools of Southeast Asia can be modeled within a larger anthropological framework emphasizing the relationship between mobility strategies and human technological behavior. Furthermore, the preliminary results indicate that non-human assemblages, which have been
IV


difficult to distinguish from modern assemblages typologically, deviate from the expectations of the model and could be behaviorally significant and not simply a broader function of the environment.
The form and content of this abstract are approved. I recommend its publication.
Approved: Tammy Stone


TABLE OF CONTENTS
I. INTRODUCTION.................................................................11
Specific Aims................................................................15
II. THEORETICAL BACKGROUND......................................................19
Interpretive Framework.......................................................20
Human behavioral ecology............................................21
Optimal foraging strategies.........................................24
Middle range theory.................................................25
The Forager-Collector Continuum..............................................27
Mobility patterns...................................................28
Curation concept....................................................29
III. BACKGROUND..................................................................32
Pleistocene Technological Behavior...........................................33
Technological origins...............................................36
The Movius Line..............................................................37
Non-lithic industries...............................................39
Archaic Homo in Pleistocene Southeast Asia...................................40
Indonesian hominins.................................................41
Interpretive Approaches......................................................44
Pacitanian industries...............................................45
Sangiran industries.................................................47
Flake and core dichotomy............................................49
vi


IV. CONTEXT
52
H. sapiens of Pleistocene Southeast Asia.....................................53
Modem human expansion..............................................54
Late Pleistocene Archaic Hominins............................................56
Paleoenvironmental Conditions................................................58
Ecological challenges..............................................59
Site Background..............................................................63
V. METHODOLOGY..................................................................65
Whole Assemblage Behavioral Indicator........................................66
Evaluating retouch.................................................66
Lithic technological organization..................................68
Collating data.....................................................69
Analytical considerations..........................................69
Statistical interpretation.........................................72
VI. RESULTS......................................................................75
General findings...................................................77
Modem human assemblage patterning..................................78
Specialized point technology.......................................80
Chronological changes..............................................82
Non-modem assemblage patterning....................................84
VII. CONCLUSIONS.................................................................86
Southeast Asian Technological Organization...................................87
Hominin assemblages................................................90
vii


SUMMARY OF FIGURES
FIGURE 1: DISTRIBUTION OF SITES ACROSS SOUTHEAST ASIA....12
FIGURE 2: ASSEMBLAGE PATTERNING ACROSS ALL SITES.........77
FIGURE 3: ASSEMBLAGE PATTERNING AT LIANG BUA.............79
FIGURE 4: ASSEMBLAGE PATTERNING IN SPECIALIZED POINT TECHNOLOGY 81
FIGURE 5: ASSEMBLAGE PATTERNING WITHOUT OUTLIERS.........82
FIGURE 6: ASSEMBLAGE PATTERNING IN SE ASIA...............87
FIGURE 7: ASSEMBLAGE VARIATION IN EAST TIMOR SITES.......90
FIGURE 8: ASSEMBLAGE VARIATION IN NON-MODERN SPECIES.....91
viii


SUMMARY OF TABLES
TABLE 1: SUMMARY OF ANALYTICAL CONSIDERATIONS.............70
TABLE 2: SUMMARY OF SITES AND ASSEMBLAGES.................73
TABLE 3: SI MM ARY OF RESULTS.............................76
1X




CHAPTERI
INTRODUCTION
This thesis explores lithic technological organization and human adaptation in prehistoric foragers of Southeast Asia (SE Asia). The chronological period of the research spans the geological period of the Upper Pleistocene, or Late Pleistocene, which broadly coincides with the Middle to Upper Paleolithic transition, as well as the African Middle to Late Stone Age. For the purpose of this study, which draws parallels to Pleistocene Eurasia, the language of Paleolithic studies is occasionally used as a point of contrast; other times it is simply convenient for broader reference to the cultural histories of the SE Asian Pleistocene and Holocene periods due to the relatively late occurrence of ‘Neolithic’ manifestations, which largely occurs without a lithic component.
The prehistoric sites selected for this study extend across much of the SE Asian subregion, covering territories on Peninsular Malaysia (Cambodia, Thailand), the Malay-Archipelago (Indonesia, Sulawesi), Philippine-Archipelago, and eastern New Guinea. The assemblages considered include the sites of Lang Rongrien, Liang Lemdubu, Lang Kamnan, Laang Spean, Moh Khiew, Ulu Leang, Song Keplek, Leang Sarru, Pia Hudale, Ille Cave, Kria Cave, Toe Cave, Uai Bobo 1, Uai Bobo 2, Lie Siri, Bui Ceri Uato, Song Terns and Liang Bua (Figure 1). In covering many parts of the region and drawing from multiple occupational phases, the assemblage data presented here are thought to reflect behavioral differences in technology at the organizational level, while also indicating that underlying this variability are fundamentally similar strategies to raw material use and mobility in SE Asian Paleolithic foragers on the whole. Furthermore, the assemblage data taken from behaviorally modern foragers present a backdrop to address differences in technological behavior at the species level


Laang Spean
Cambodia
Moh Khiew
Thailand
Ulu Leang
Sulawesi
' i r ri*
* r l
Lang Kamnan
Thailand

Lang Rongrien
Thailand
llle Cave
Palawan,
Philippines
r y •
Kria Cave
Eastern New Guinea
Leang Sarru
Taiaud Islands, Indonesia • V •
Toe Cave
Eastern New Guinea


Song Keplek
|ava,
Indonesia
East Timor
Liang Lemdubu
ftru Islands, Indonesia
UaiBobol&
East Timor

Song Terus
lava.
Liang Bua
Flores. Indonesia
Bui Cen Uato
East Timor

Pia Hudale
Rote Island, Indonesia
Figure 1: Distribution of sites across Southeast Asia
at the sites of Liang Bua and Song Terus, or in the very least identify them as anomalous before turning to investigate other formative processes that are more taphonomic in nature than behavioral.
Technological simplicity and continuity over time is not altogether unique in human evolutionary contexts and certainly not those of Central and East Asia, but the regional scale of this technological stasis and the temporal reach into Early and Middle Holocene is a uniquely SE Asian phenomenon. Hence, SE Asia is a historical context in need of an alternative approach to stone tool analyses due to their importance in addressing behavior and other chronologically diagnostic features. Although the lithic industries of Pleistocene SE Asia are often described as unexceptional and unchanging, the stone tools here provide a line of
17


evidence to major ongoing debates in paleoanthropology and the nature of human adaptation more broadly. In the absence of a well-defined technological sequence, the characteristically ^characteristic stone tools discovered here are at all times relevant and uniformly important data to the major evolutionary questions surrounding the nature of this variability.
Overview
In general, the stone tool artifacts from these assemblages are representative of the broader technological system of Pleistocene SE Asia. However, little is known of the prehistoric behavioral strategies underlying this broader technological system. In particular, two major issues are regularly noted by scholars: the first deals broadly with the fact that SE Asian lithic artifacts do not appear to easily lend themselves for distinguishing behaviorally modern assemblages from those accumulated by archaic Homo species; the second concerns the extent to which this technological system was actually embedded within wider land-use patterns of Late Pleistocene forager communities. The research presented in this thesis tests this premise by examining the relationship between economic behavior, lithic technological organization, and mobility strategies in Paleolithic SE Asian foragers. The results from the behaviorally modern sample are then turned back toward addressing the question of authorship in important late-surviving hominin localities. The non-modern assemblages from Liang Bua and Song Terns, included in this study, present an opportunity to test whether the stone tool industries of archaic humans conform to the expectations of a model now informed through the study of assemblage-scale patterning in Paleolithic foragers of SE Asia.
The Paleolithic period of SE Asia generally lacks typologically distinct forms, and it would appear degrees of technological complexity is generally not compatible with the behavioral developments of Homo species in this area of the world. These Late Pleistocene
n


foragers retained the basic technological system of the former indigenous hominins of the region, the components of which persisted well into the Holocene. Attempts at distinguishing between the technological behavior of anatomically modern humans (AMH) and former archaic inhabitants such as H. erectus often fall on securely-dating sequences, which has often been highly problematic. Moreover, the possibility that Late Pleistocene foragers interacted or nearly overlapped with archaic hominin contemporaries in H. floresiensis and ‘evolved’ H. erectus reintroduce the problem of stone tools and authorship irrespective of the context.
Recently, Paleolithic researchers have turned to assemblage-scale variability to address important behavioral shifts in technological organization and land-use strategies of the Middle to Upper Paleolithic transition. Late Pleistocene foragers in Western Eurasia recurrently display a strong negative relationship between the density of lithics they contain and the frequency of retouched pieces therein (Riel-Salvatore and Barton, 2004). The ends of the spectrum defined by this relationship can also shed light about the dominant mobility strategies (e.g., residential, logistical) employed by the foragers who accumulated these assemblages (Riel-Salvatore et al., 2008). This trend seems to hold at a continental level (Barton et al., 2011), and appears to be a fundamental dimension of lithic technological organization from the Mousterian on (Kuhn, 2004; Clark, 2008; Riel-Salvatore et al., 2008). Given that this trend characterizes both Neanderthals and modern humans after ca. 130 ka, it could be reasonably assumed that modern humans radiating into SE Asia might also employ similar strategies. Moreover, the methodology’s alignment within the temporal parameters of the Middle to Late Pleistocene transition have fundamentally the same questions to answer by using this model. The ‘whole assemblage behavioral analysis’ method used in this thesis offers the potential to determine whether the hominins responsible for accumulating the Liang Bua lithic
14


assemblages were characteristic of AMH Paleolithic foragers or unique in their own right, and thus yielding new measures of behaviorally, and perhaps taxonomically sensitive information from stone tools.
Specific Aims
Late Pleistocene hunter-gatherers of SE Asia are broadly characterized as technological opportunists; incorporating locally available lithic raw material on an encounter basis (Rabbet, 2012). In contrast, the adaptive strategies of foragers of the European Late Pleistocene are believed to have mobilized in accordance with deeply embedded procurement strategies, which were most sharply influenced by select fine-grained materials (Riel-Salvatore et al., 2008). The behavioral model applied in this thesis is founded in ethnographic and ethnoarchaeological evidence, which have shed light on how hunter-gatherers structured their stone tool use behaviors in relation to myriad factors including resource predictability, raw material availability, efficiency, transport, and stylistic preference, to name a few (Binford 1979; Nelson 1991; Shott 1986; Torrance 1983; Wiessner 1982; Wobst 1983).
At present, only a handful of studies have applied the principles of this research program directly to the study of SE Asia lithic assemblages (e.g., Anderson, 1990; Simanjuntak, 2001), while other scholars have put forward new interpretive methods as an alternative to typological approaches (Borel et al. 2016; Marwick, 2007). The ecologically based perspectives on human behavior have been difficult to operationalize in the study of prehistoric SE Asia due to lingering questions surrounding the Pleistocene environment, particularly in terms of the sub-standard quality of lithic raw material (i.e. Movius Line), the onset and composition of the Late Pleistocene tropics, and the nature of tropical adaptation in human species more generally. While the ecology of SE Asia and the notion of tropical


adaptation more generally have always figured into explanations of low-input industries (e.g. Kelly, 1982), the nearly total lack of any standardized elements in stone tool forms throughout Late Pleistocene period in particular have led some to question whether stone tool technology had been liberated from the adaptive strategy entirely (Reynolds, 1992). This wider cultural phenomenon across SE Asia is the basis for the first set of questions explored with this model and are listed below:
1. ) Do prehistoric foragers of Southeast Asia structure their lithic technology in relation to wider landscape-use and mobility patterns as seen in Pleistocene Western Eurasian hunter-gatherers?
2. ) Is a regional context with poor quality raw material and dominated by expedient industries with minimally retouched tools incompatible with behavioral models which emphasize the relationship between mobility and curationl
The second set of questions explored in this model addresses the uniquely persistent technological continuities, and the potential diversity of the Upper Pleistocene hominin landscape. The evolutionary context of Pleistocene SE Asia has hosted a number of peculiar hominin forms, at least one of which inhabited the region as recently as the Middle to Late Pleistocene boundary. Although the revised chronology at Liang Bua, Indonesia indicates that the youngest occupational period of H. floresiensis predate the arrival of H. sapiens in SE Asia, the older dates have bolstered the general acceptance of H. floresiensis as a distinct, archaic relative of H. sapiens. The evidence for multiple archaic hominin populations surviving up to
16


this period is controversial, but the notion of other late hominin extinctions is broadly recognized as a likely scenario here. For the time being, defining the spatiotemporal range of SE Asian hominins at the close of the Middle Pleistocene is limited to Liang Bua, whose species holotype has been re-dated to 100 - 80 ka, while the associated stone tools span into the Late Pleistocene (190 - 50 ka), immediately prior to the first evidence of H. sapiens in SE Asia (Sutikna and Tocheri et al., 2016).
Although only a handful of assemblages can be securely-dated to the Middle Pleistocene here, the Paleolithic sequence is defined as one of the longest continuous sequences in human history due to the prolonged technological stasis. In fact, was only recently that stone tools were definitively-dated to a period unquestionably linked with archaic hominins at the site of Mata Menge on Flores, Indonesia (Moorwood, 1998). Prior to these assemblages dating to the Middle Pleistocene, there were some of the opinion that H. erectus in SE Asia did not use stone tools at all. Incidentally, the findings from Mata Menge would come to be described as the stone tool using ancestors of H. floresiensis, not H. erectus. The stone tools and fossil remains at Mata Menge are believed to be those of the founding population of Homo on Flores, Indonesia, which exhibit strong anatomical ties with more recent, Late Pleistocene populations of H. floresiensis discovered at the site of Liang Bua, which is also located in the Soa Basin (van den Bergh et al., 2016). Prior to the recent fossil discoveries at Mata Menge, published just this last year no less, former excavations had yielded only one assemblage with unambiguous stone flakes (n=14) confidentially associated with Pleistocene-era hominin distributions of SE Asia.
The theme of SE Asian technological continuity is largely defined by a few well-dated fossil and archaeological contexts of H. floresiensis and their ancestors on Flores in the early
17


Middle Pleistocene, and the widely-distributed industries associated with modem human foragers of the Late Pleistocene. The actual ‘continuity’ and whether hominin populations sustained this technological behavior throughout the Middle Pleistocene lacks robust archaeological evidence. Nevertheless, the arrival of modem humans on the SE Asian archipelago appears to have only minimally impacted the lithic reduction and land-use strategies pioneered by archaic hominins of the Early to Middle Pleistocene, which raises the question of whether ecological constraints are responsible for assemblage patterning due to some combination of poor material quality and the unique distribution of non-lithic resources in Pleistocene SE Asia. The second series of questions stated below considers whether assemblage patterning can be used as a strictly defined behavioral correlate of AMH and thus tap into taxonomically sensitive behavior, or if assemblage-scale patterning in this region here follows typological analyses in blurring technological boundaries, indicating how strongly the broader environmental pressure facing all Homo species inhabiting the Pleistocene SE Asia impacted the use of lithic technologies here.
3. ) Does the assemblage scale patterning in Southeast Asian foragers hold across all Homo including non-modern hominins, which might indicate that regional patterning is not taxonomically significant but rather environmentally consistent?
4. ) Is the model proposed in this thesis a compatible alternative to typology in Southeast Asia in terms of identifying behavioral sensitive information which could be used to distinguish archaic hominins from modern humans?
18


CHAPTER II
THEORETICAL BACKGROUND
The study of hunter-gatherers, both past and present, draws heavily on the perspectives of middle-range theory (MRT) and human behavioral ecology (HBE). In particular, these perspectives have been attractive to scholars of prehistory because they provide a genuinely testable model for a wide range of hypotheses that can be applied in the archaeological record. Over the last several decades, the subject of forager mobility and technological organization have become tightly integrated within this explanatory framework, emphasizing the role of lithic technology and its relationship to forager mobility, resource exploitation, and economizing behavior (Binford, 1980; Bamforth, 1991, Bamforth and Bleed, 1997; Kelly, 1983; Shott, 1986; Torrence, 1989; Winterhalder and Smith, 2000).
The theory reviewed in this chapter reflect the fundamentals of the contemporary processual approach (also variously referred to as Processual-plus) used in this thesis. Notably, the relevance of evolutionary principles in the study of HBE and the more narrowly applied study of optimal foraging strategies (OFS), and the middle-range approaches which links general theories to the cultural and natural forces influencing their formation.
Another component of the processual-plus approach touched on in this chapter is the construction of research and design that actively contributes to anthropological theory through developing methodical tools and testable hypotheses. In approaching the archaeological record as fundamentally testable, the relationship between data and the explanatory construct can develop into a powerful concept in its own right. Of particular relevance to the subject matter in this thesis are the conceptual measures of mobility and curation, which together provide the
19


interpretive framework for identifying assemblage patterning and linking them with the specific behaviors believed to have formed them.
Interpretive Framework
In viewing aspects of Pleistocene human mobility strategies and technological behavior as part of larger cross-cultural phenomenon, the contemporary processual approach adheres to a theoretical foundation in close company with the methods and theory of the natural sciences. The explanatory power of HBE and MRT is strengthened through the differing roles they play in interpreting the archaeological record. The middle-range emphasis on observable data and how it came to rest in a particular state, provides a line to mid-level empirical content (i.e. regularities in artifact ‘behavior’). These data can be tested with reference to the ‘high-theory’ evolutionary perspectives of HBE, which together brings to bear a positivist approach to interpreting cultures of the past.
The first arm of this interpretive framework, HBE, concerns the theory of actual, or perceived rules governing all behavior, which projects universalities observed in biological survival strategies onto the prehistoric study of humans. In this sense, HBE defines the behavioral parameters for which an individual, or group of individuals are expected to operate within, given the conditions for a various prehistoric context. The organization of these ideas in the study of archaeological cultures is the primary subject matter of optimal foraging strategies.
If the evolutionary themes of HBE are to bear out the adaptive strategies of non-living people, the interpretation must flow from a scientific understanding of the archaeological data. Consequently, the second arm of this interpretive framework, MRT, is the critical theory building component which links behavioral trends with a line of empirically tested
90


observations concerning the formation of the archaeological record. The fundamental goal of MRT is to link the interpretation of prehistoric data to the actual processes which are believed to have shaped their archaeological formations, thus providing some level of scientific grounding in the interpretation of archaeological data.
Human behavioral ecology.
As was alluded to above, HBE, or behavioral archaeology, is a broader theoretical approach drawing on evolutionary principles— namely a biological organism’s predilection to behave optimality in response to ecological pressure —which outlines how humans should interact within their surroundings when aspects of a given environment are known. Both modern and prehistoric human behavior is conceptualized as fundamentally adaptive, not unlike that of any other organism in the natural world. However, our place in the natural world is uniquely different, and the multitude of behaviors which drive adaptive strategies in human populations are intertwined with, and often orchestrated through various dimensions of sociocultural organization. Thus, as it pertains to the study of prehistoric people, HBE is concerned with the study of human behavior in relation to ecological and socio-cultural conditions (Bird and O’Connell, 2006).
Technology, subsistence, and settlement are vital to survival strategies and, as such, are among the most relevant matters in the study of human culture and ecology. These facets of social organization are viewed as the primary buffering mechanisms by which prehistoric humans modulated new or worsening constraining pressures within their environs (Binford, 1980; Winterhalder and Smith, 1992). Underlying the levers of cultural systems is an ecologically driven adaptive strategy believed to be influenced most strongly by fitness-maximizing behaviors (Bird and O’Connell, 2006). Since the forces of nature do not place
21


pressure squarely on the biology of humans, but rather is mitigated through behaviors embedded within complex social systems, the structural reconfigurations in human sociocultural systems are viewed as adaptive traits, which can be studied with explicit reference to the process of natural selection (Kelly, 1995).
In evolutionary terms, the locus of change still occurs at the individual level. However, the mechanism of inheritance is not strictly essential to our analyses and understanding of human behavior through this lens. From the perspective of HBE, the evolutionary circumstances which endowed the human species with highly sophisticated cultural faculties is only tangentially related to the research questions they pursue, and is not presumed to “seriously constrain adaptive responses to ecological variation” (Nettle, 2013: 1055).
The basic notion here is known as the phenotypic gambit and, as it relates to its broader application in HBE, is the idea that favorable traits which maximize fitness will be differentially selected regardless of the pathway of inheritance (Smith and Winterhalder, 1992). Thus, practitioners of HBE start with the premise that behavioral variability is not routed through changes in gene frequencies, but rather is representative of the human capacity to adapt in a manner suitable with the proximate environmental and social conditions (Codding and Bird, 2015).
Certainly, the role of genes in producing behavioral variability presumably much more involved than is accounted for with the logic provided by the phenotypic gambit. And, indeed, advocates of the strong sociobiological thesis envision a closer “link between genetics and behavioral variation,” (Kelly, 1995: 51). However, while these behavioral directives are entangled in both the genetic and social domain of human behavior, the capacity to respond to changing environmental conditions is the actual trait that evolved under natural selection, not
22


the particulars of the behavior itself. Thus, the capacity for culture evolved under selective pressure which favored the various building blocks of behavioral flexibility, such as the ability to exchange new information socially and utilize the memory of that knowledge across variable and changing environments (Nettle, 2013) In this way, human behavior is viewed as phenotypic extensions which emerge from a basic genomic configuration, whereby variability in human behavior occurs within the ‘norms of reaction’. Accordingly, the efficacy of cultural behavior is rooted in an individual’s evolutionarily tailored capacity to tune one’s own behavior in accordance with their proximate environment and future environment.
The type of adaptive latitude described above is referred to as behavioral plasticity and, in essence, refers to the elasticity and ephemerality of human adaptive responses, which are made possible and coordinated through various channels of complex social behavior. As it relates to prehistoric humans, aspects of artifact patterning in the archaeological record is thought to represent the distillation of these behaviors, and their retention over time indicates its success via differential selection. In this sense, these behaviors are cultivated under selective pressures not unlike those of other organisms, which live and reproduce in full exposure to the forces of nature. And while behavioral traits in humans amass and find inventory in the culture system, which buffers against the sort of raw environmental pressure that results in biological change, the differential selection of behaviors, although culturally-mediated, are still treated as behavioral variants that flourish and wane depending on their optimality under a given set of ecological conditions. In the study of human strategies, the individual fitness of an organism is replaced by behavioral variants, which refer to the mean fitness of a specific behavior, and serve as the basic analytical unit in the study of optimality in human foraging (Kelly, 1995).
23


Optimal foraging strategies.
Despite the position held HBE, which rejects the notion that human interaction and ecology necessarily require a fundamentally different interpretive framework than any other organism (Nettle, 2013), the evolutionary approach to human culture is far more nuanced. In a sense, behavioral plasticity freed-up ‘adaptation’ from the hardened evolutionary channels within which organisms become masterfully adapted to their present habitat and, generally speaking, increasingly vulnerable to change in the future habitat. In non-human animals and plant species, optimality is reflected in a fairly straightforward relationship between morphological and structural commitments (i.e. anatomy) and subsistence-related tasks; it is also quite manageable in comparison to humans in the sense that fitness is a relatively easy measure when working with non-human animal populations because the relationship between behavioral phenotypes and genotype is more straightforward.
From an ecological standpoint, the decisions and activities underwriting human behavior are similar to the biological tendencies to maximize one’s individual fitness. However, human populations resolve ecological challenges by channeling reproductive and subsistence-related tasks through a network of social and cultural behaviors carried out at the group level. The body of work advancing the study of OFS is primarily concerned with outlining specific sets of contingent relationships between components of the social unit and the frequency and distribution of environmental variables (Bird and O’Connell 2006; Winterhalder and Smith, 2000). Consequently, inferring behavioral optimality in human populations involves several more variables, many of which are materialistic. Behavioral variants are measured in relation to these specific variables, or currencies.
24


In calling on the general theories of behavior provided by HBE, OFS concerns itself primarily with the indirect measure of energy expenditures, the various cost-benefit tradeoffs, and the application of a testable hypothesis. The simplicity of behavioral models allows hypothesis to be put forward and easily tested in the archaeological record. The use of OFS simplify behavior and focus on the relationship between specific variables. For instances, a heterogeneous distribution of resources and the time and energy used to extract those resources as they deplete versus the time and energy saved by moving camp. In this way, OFS describes a series of probabilistic relationships which articulate with the various aspects of technology, society and the environment.
Middle range theory.
Not unlike many other traditions of archaeological inquiry concerned with the lifeways of prehistoric societies, MRT approaches are chiefly concerned with deriving meaningful interpretations in the archaeological record, but lean far more heavily on the role of empirical insights and addressing its interpretive significance in relation to the formative processes which drive archaeological patterning. In viewing the archaeological record as heavily shaped by generalizable processes, the depositional context and artifact content observed in the present can be used to substantiate claims about the past (Binford 1977, 1982). From these data, the archaeological record can be used to reconstruct a theory on how it formed. Moreover, the reconstructive function of observations made in the present, limited as prehistoric context are, is bolstered by contemporary ethnographic studies and experimental testing. In essence, though, the ‘middle’ in MRT refers to any link between a particular theory and specific type of empirical content with limited scope (Kosso, 1991; Schiffer, 1988). Nonetheless, the popularity of MRT research in archaeology has been fostered by the its close alignment with


the methodological and theoretical modeling of the natural sciences, which others have strongly contested (e.g. Hodder, 1986).
In viewing human behavior as fundamentally adaptive, MRT practitioners attempt to explain the relationship between artifacts, ecology, social organization, and depositional environment through linkages to a higher level of foraging theory. As was mentioned above, the ‘higher theory’ of human behavior is not generally the focal point of the MRT research, nor is it exclusively paired with HBE, but rather serves as a precept that allows variability in the archaeological record to be treated as a series of probable outcomes which guides the interpretation of evidence. With this format, the flow of information is not hierarchical; theories can be used to inform observations, but the observations themselves can form the basis of a theory to be tested against the evidence (Kosso, 1991). As such, an all-encompassing definition for MRT is difficult to find agreement on because of the extensive interdisciplinary input used to formulate archaeological expectations. As its used here, the body of theory guiding this model is informed most strongly by neo-evolutionary perspectives on behavior (HBE, OFT), actualistic and ethnographic studies (MRT), and the quantitative methods in which archaeological patterning is explained.
Actualistic studies are essentially an empirical check on both the theoretical validity of neodarwinian perspectives, and the full breadth off formative processes believed to shape archaeological patterning. Ethnographic analog, taphonomic experimentation, and historical reference among other forms of inquiry, derive data to formulate expectations to be tested against the archaeological record (Flannery, 1982). Contemporary studies of traditional societies still practicing hunter-gathering present a wealth of information likely relevant to prehistoric modes of life; but also presents the opportunity to identify cross-cultural trends and
96


normative behavior to be tested in the archaeological record. Prior to the use of MRT in archaeology, ethnographic data was more or less borrowed to fit with the archaeological evidence. Experimental studies allow artifacts to be reverse engineered in a scientifically controlled setting, permitting an extremely close and deliberate study on the processes which drive the formation of an artifact, or assemblage of artifacts. These experimental observations offer a unique contribution that has direct bearing on the recovery, analysis, and inference of archaeological material (Schiffer, 2013).
In summary of the processual-plus interpretive framework, this theoretical orientation is heavily influenced by evolutionary principles, with a notable emphasis the relationship between natural selection and cultural variability. Physical artifacts are viewed as an adaptive phenotypical variable extension of biologically-rooted behavior, and as such is subject to the same selective forces shaping behavior on the whole.
In simplistic terms, the tinkering and dispensing with behaviors that are not optimal in performance is what makes human behavior predictable and therefore testable. The embeddedness of these behaviors in the cultural systems (e.g. technology, subsistence, and settlement), and their material extensions provide the data to measure optimality across prehistoric human survival strategies. The interrelatedness of these cultural systems is such that a change in one system can have a cascading effect and lead to a change in another system. In this way, researchers can potentially isolate a single component of the system, say, technological organization, and measure a change in another variable, like settlement strategy.
The Forager-Collector Continuum
Mobility in the archaeological record is a conceptual measure landscape-use patterns in pre-agricultural communities. The concept of mobility is grounded in the relationship
27


between human adaptation and the natural pressures and particulars of habitats. Mobility in this sense, is managed around strategies of resource exploitation (e.g. subsistence strategies, raw material procurement) and is measured along a continuous spectrum which roughly correspond with the demands of a particular environment. In the study of Pleistocene adaptive strategies, forager mobility is often the cornerstone of the interpretive framework with the conceptual distinction between logistical and residential settlement patterns in hunter-gatherers being the dominant theme in hunter-gatherer research. In some ways, mobility is viewed as the critical adaptive system powering Pleistocene human survival strategies, around which technology, subsistence, and other socio-cultural aspects of human behavior are organized.
Mobility patterns.
Binford (1980) characterized logistical mobility as a small group of individuals whose site-occupation was of longer duration and whose movements consisted of carrying-out task-oriented forays within a larger forager radius. Comparatively, residential mobility refers to a smaller group of individuals whose hunting expeditions were restricted to a smaller foraging radius, but also involved frequent camp relocation as a means to stabilize subsistence economy as resources diminished. Essentially, Binford (1980) dichotomized the human settlement configuration in terms of how subsistence resources were pursued, extracted, and consumed by hunter-gatherers. In the simplest terms, residential hunter-gatherers move camp to resources, while logistical hunter-gatherers move resources back to the camp.
The basis for interpreting assemblage patterning as a measure of human organization is founded in the conceptual distinction between logistical and residential mobility and the notion of curation (Binford, 1979; 1980). What was important about these observations is that Binford noticed that different settlement strategies resulted in different assemblage formation,


the consequence of which were distinct material signatures that could be extrapolated into the archaeological record. To evaluate mobility, stone tool artifact composition and density within assemblages are commonly used quantitative measure. Because of the central role of stone material in the wider web of forager activities and survival, it has been possible to articulate the various states of artifacts from their recovered context with the broader mobility strategies from which they were embedded (Andrefsky, 1994; 2001; Barton, 1998; Hiscock, 2002; Barton and Riel-Salvatore, 2014; Riel-Salvatore and Barton, 2004; Riel-Salvatore et al. 2008).
Curation concept.
An important distinction in the interpretation of lithic variability are contrasting views concerning the underlying mechanisms driving patterning in artifact morphology. From one perspective, stone tool traditions are organized through their morphological similarities in artifact form, and are conceptualized as the consequence of selective processes imposed by culturally descendant groups. From this viewpoint, variability in artifact form is largely constrained by social descent, and as such, were often treated as cultural markers (sometimes referred to as “fossil-types”), which when plotted along spatial and temporal boundaries were felt to retrace the succession of cultural lineages.
The interpretation of stone tools accumulations as a measure of occupational intensity assumes that the technological behavior of prehistoric foragers was sharply shaped through deliberate consideration of the current and future utility of chipped stone tools they produced, and their proximity the next cache or quarry. The broader topic of economizing behavior and lithic technology is known as the curation concept, which emphasize the relationship between the dynamic life- history of lithic artifact and their final morphology, wherein myriad factors contribute to the condition in which the artifacts are discovered in, most of which are the
29


unintended consequence of tool use behaviors. In particular, the identification of a host of potential factors have been outlined, ranging from the intensity and breadth of tool activities performed, the length of use, to the quality of the raw material used. (Bamforth, 1986; Binford, 1979; Odell, 1996; Shott, 1996). Importantly, these use-life patterns in stone tools are believed to articulate within a wider technological system, whereby important economic decisionmaking behaviors such as lithic provisioning, the distribution of lithic raw material, and the quality and quantity of lithic material available (Andrefsky, 1994; Bleed, 1986) regularly factor into the character of lithic assemblages encountered in the archaeological record.
The curated-expedient continuum of lithic organization frames technological behavior within corresponding mobility strategies of hunter-gatherers. One the one hand, curated industries are characterized by heavily modified lithics, core reduction is usually extensive, and retouched tools can be larger in size and ranging in shape- allowing flexible modification of stone tools as access to raw material diminished. In Pleistocene hunter-gatherers, this type of lithic organization appears to have been favored by highly mobile foragers and so consequently would be most often represented in ephemeral archaeological contexts. Comparatively, high densities of artifacts and low frequencies of retouched tools characterize expedient assemblages. Retouched tools are often function-specific and are manufactured for immediate use. Raw material is either locally abundant or raw material procurement is embedded within the use of the site; this type of technological organization is indicative of less mobile groups. These lithic qualities would be most often represented in archaeological contexts with repeated-and sustained occupation such as a residential camp. However, this type of relationship becomes difficult to demonstrate when working with disparate stratigraphic contexts, especially when it is unclear how rapidly assemblages formed. The model applied


in this thesis approaches curation at the assemblage-level, minimizing detailed accounts of the tools themselves and instead, analyzing the collective frequency of retouched artifacts in relation to the density of all other artifacts within a given assemblage (e.g. Riel-Salvatore and Barton, 2004).


CHAPTER III
BACKGROUND
The various cobble and flake-based industries of ‘Paleolithic’ SE Asia (ca. 1.1 ma - ca. 11 ka) have been widely recognized as a technological system unconnected to the lines of development that define the classic stone tool traditions of the western Old World (Corvinus, 2004; Movius, 1948). In fact, the peculiarity of these stone tools has been so extensively noted that virtually all recent literature on the subject of Paleolithic in SE Asia comment on this distinction, which usually calls attention to the major interpretive challenges faced by all archaeologists working in this regional context. Namely, the many problems in addressing the indistinctiveness of these lithic artifacts and often unconstrained geological context from which they derive. Prominent scholars quite accustom to chronicling these issues, characterize SE Asia as a regional context where “... meaningful typological categories have remained elusive” (Brumm and Moore, 2012: 386), and “...definitive ‘cultural’ allocations are geographically mercurial and temporally ambiguous” (Rabbet, 2013:102). The unstandardized and expedient character of the stone artifacts blur technological divisions and contribute to the problem of addressing the very unique distribution of hominin taxa during the Pleistocene, and the controversy and debate concerning stone tool use among these archaic human species (e.g. Bartstra, 1982; Semah, 1992).
The lithic artifacts of the Late Pleistocene in SE Asia are amorphous and undiagnostic, and in many cases, are indistinguishable from earlier stone tool industries manufactured by archaic Homo species during the Middle Pleistocene. The ‘least effort’ or ‘low-input’ strategies associated with these lithic industries have often been explained in relation to a broader function of the Asian Pleistocene environment— irrespective of the cognitive ability and
42


technical competence of the stone tool producers. Thus, the similar nature of their expedient manufacture has been taken as an indication of a common underlying strategy, one which apparently transcends both cultural and species boundaries.
Although the vast majority of lithic artifacts from SE Asia are fundamentally similar in their informal character, their membership within this broader classification is offset by subtle internal variability. The multitude of regional differences are often subsumed into the coarser division of ‘chopper-chopping tool’ and ‘flake-based’ techno-complexes in an effort to differentiate between the handiwork of modern human and archaic populations. However, this dichotomy has been somewhat problematic in application because it is predicated on the typological compatibility of SE Asia’s heavy core tools with ‘Mode-1’ artifacts from African Pleistocene sequences (2.5 - 1.7 Ma), which significantly predate the earliest securely-dated stone tools in the region (1.1 Ma). Furthermore, it would appear that the extension of these industries by modern humans in the Late Pleistocene preclude the use of stone tools as a proxy for behavioral modernity in the archaeological record.
Pleistocene Technological Behavior
Pleistocene lithic artifacts are the most abundant and durable source of information available in the study of our evolutionary history. The preference for lithic raw material throughout human history underscores the importance of this technological medium to the evolutionary success of Pleistocene hominins. Accordingly, the nature of technological variability is an important archaeological measure in the study of our evolutionary history. In the broadest sense, patterning in technological behavior are equated with the various modalities underwriting the cultural evolution of our species. Modalities, in this sense, refer to the broad range and interplay of cultural, neural, and physiological mechanisms operating on the


evolution of technological behavior. In the archaeological record, their fossil and material cultural extensions are treated as correlates of behavior and phylogeny, which divide up several phases in our evolutionary history.
In general, the trend toward complex tool forms and systematic reduction strategies are thought to reflect major biological and cultural developments in human evolution (Ambrose, 2001). However, while it is fairly evident that the structure of technological change trends broadly with evolutionary history of the genus Homo, narrowly defining the specific evolutionary context behind the succession of each technological phase has not been as forthcoming. In particular, what remains unclear is whether technological diversity of the Lower Paleolithic should be explained in terms of biological differences or a combination of environmental and functional factors (Foley et al., 2003; de la Torre and Mora, 2009).
The recognition of variation within major categories of technological change (e.g. Oldowan vs. Developed Oldowan) had long been viewed as intermediate sequences of a continuous technological phase. However, this view has recently been critiqued upon reexaminations of Oldowan assemblages, which have demonstrated consistencies in knapping competency and technical skill from beginning to late phase Oldowan (Delagnes and Roche, 2005). Alternatively, recent approaches have explored the possibility that intersite variability of the Oldowan tradition could reflect local responses to environmental factors (e.g. the flaking properties of different materials), indicating a strong adaptive element in the lithic technology used by early hominins (Semaw et al. 2009; Stout et al. 2010).
The earliest appearance of stone-tool industries with robust documentation is represented by the Oldowan techno-complex; it is important to note here that stone tools found in association with hominin fossils have been influential in distinguishing early Homo from
34


contemporary species (e.g. Leakey et al., 1964), while species distinction within the genus Homo and australopithecines continues to be problematic due to the morphological variation (e.g. Cela-Conde and Ayala 2003; Wood and Collard 1999). Furthermore, the paucity of stone tools which derive from hominin-bearing localities is particularly problematic prior to the emergence of Homo due to the technological timespan of the Oldowan, which overlaps with a number of australopithecine species.
The contextual problems and uncertainties in the origin and development of lithic technology in the Early Pleistocene is pertinent to the study of SE Asian Paleolithic industries on two fronts. Firstly, the widespread use of stone tools as cultural, phylogenetic, and chronological distribution markers in hominin populations noticeably fails east of the so-called Movius Line, which is a technological and geographic demarcation apparently affecting human populations in SE Asia as recently as those Pleistocene and Holocene foragers included in this study. Moreover, the expedient qualities of SE Asian industries have become deeply embedded in a historical connotation with archaic hominin populations thought to have arrived in late Early, or Middle Pleistocene Sundaland. Secondly, the associated stone tool behaviors of H. floresiensis at Liang Bua redefine the standard inventory of correlates used in defining the emergence of Homo in the fossil record. The assemblages at Liang Bua are exceptional in this regard, because they have some bearing on the evolution of stone tool manufacture. Establishing the cognitive and manipulatory requirements necessary in stone-tool manufacture are difficult to estimate in extinct hominin taxa, yet it has been precisely these estimates that have converged to define the genus Homo. At 380cm, the endocranial volume of H. floresiensis is well below the accepted range for genus Homo, and registers on the low end of estimates for small-brained australopiths (Brown, 2004). Moreover, the retention of primitive wrist


morphology inferred from metacarpal elements of H. floresiensis indicated some level of morphological restriction in stone tool manufacture (Tocheri 2007). The puzzling combination of an absolutely small brain (with comparison to body size), primitive stature, and relatively complex stone tool industry raise interesting questions concerning the nature and timing of stone tool behaviors.
Technological origins.
Currently, the earliest evidence of Oldowan technology dates to 2.6—2.5 Ma at Gona, Ethiopia, and was found to be contemporaneous with the nearby hominin-bearing site of Bouri, which had previously produced cut-markings on modified bone in the same geological deposits as Australopithecus gar hi (Asfaw et al., 1999; de Heinzelin et al., 1999; Semaw et al., 2003). However, the known spatial and temporal distribution of Homo, as well that of earlier and contemporaneous hominins, is regularly reconfigured or expanded in light of new fossil evidence and increasingly precise dating techniques. For example, a new discovery of a partial mandible from Ethiopia was recently dated to 2.8 Ma, and may represent an intermediate form of Homo (Villmoare et al., 2015).
Reintroducing Homo in this region well before the lithic accumulations at Gona lithic might calm speculation that these stone tools were made by australopithecines. More recently, however, the discovery of in situ lithic artifacts in West Turkana, Kenya, were dated to the Pliocene (3.3. Ma), which are .5 Ma older than the earliest evidence for the genus Homo, and fall within the spatiotemporal range of Kenyanthropusplatyops (Harmond et al. 2015). While these stone-tools were found to bear distinctive patterns in flake detachment—a fundamental criterion used to assess technical skill associated with the Oldowan techno-complex (e.g. unidirectional flakes, debitage)—the knapping-skill was considered underdeveloped by


comparison, which the authors believe warrants the techno-complex distinction of Lomekwi, and hence the “pre-Oldowan” designation (Harmond et al., 2015). However, it is worth mentioning here that K. platyops are still considered by some to be an archaic predecessor to the genus Homo (Cela-Conde and Ayala, 2003).
Given this taxonomic uncertainty, the relationship between non -Homo hominins (i.e. A. garhi, K. platyops) and stone tools is, for the moment, contingent on the spatial and temporal distribution of known stone-tool manufactures in Homo. In this sense, stone-tools still have retained a high level of classificatory power, and as such, heavily influence the taxonomic definition of Homo. However, despite the fact that stone tools continue to factor heavily into designating Homo, establishing an exclusive criterion (e.g. cranial capacity; morphology) strongly suggestive of technological origins within Homo has also yet to materialize.
Although debate continues, the overarching patterning from simple to complex is still the cornerstone of Paleolithic research. Yet contemporary perspectives on the nature of this variability are largely informed by material cultures of Africa, Central and Southern India, and Western Eurasia. In contrast, the Asian Paleolithic is represented by an extraordinarily long sequence of technological stasis, lacking the type of directional change in lithic technologies observed in western portions of the Old World. Here, simple core and flake tool technologies persist into the Late Pleistocene at a near continental level (excluding West Asia) and as recently as the Terminal Pleistocene in SE Asia, indicating environmental constraints may have been a powerful limiting factor in the development of complex lithic industries.
The Movius Line
There has long been speculation that easily accessible and overabundant organic material such as bamboo (Pope, 1984) and wood (Hutterer, 1977) were preferred over course


grain lithic material by early Homo in SE Asia, or alternately, that H. erectus embarked out into the Asian continent before Acheulean technology was developed (Reynolds, 2007). However, these theories, particularly the former, are largely founded in the noticeable absence of specific data (i.e. standardized stone tools) rather than actual physical evidence.
There have been a number of explanations put forward in an attempt to explain the persistence of simple core/flake industries of SE Asia (ca. 1 Ma— Terminal Pleistocene), and more broadly, the similar commitment to low-input technologies which characterize Central and East Asia during the Paleolithic (1.7 Ma— Late Pleistocene). Currently, no single explanation is presented to account for the technological phenomenon. Rather, the lack of bifacial elements in the Asian Paleolithic is believed to be influenced by a combination of methodological biases, environmental factors, and perhaps to a greater or lesser extent, the biogeographical movement of archaic hominins. The later hypothesis would involve an early hominin migration out of Africa before the rise of bifacial industries e.g. Acheulean (ca. 1.9— 1.7 Ma). This idea was originally proposed by Foley and Lahr (1997), at the time was supported by the earliest Indonesian fossil evidence dated around 1.7 Ma (Swisher, 1994). However, the earliest conservatively accepted appearance of H. erectus in ISE Asia much closer to the beginning of the Middle Pleistocene.
An important issue deepening the typological divide between the Eastern and Western Old World are methodological biases operating on both sides of the Movius line. Although bifacial technology is not altogether absent in Asia, and reports of them have grown recently (e.g. Corvinus, 2004; Mokhtar and Saidin, 2006; Pawlik, 2004), the bifacial tool discoveries in SE Asia are often categorically downgraded to “proto” and “crude” versions, and are generally believed to differ from the classic Acheulean in terms of symmetry and degree of elaboration


(Lycett and Bae, 2010). However, the question of whether this reflects interpretive biases toward the recognition of complex forms has been raised (Rabbet, 2012: pg. 98).
In a recent morphological analysis of Asian ‘hand-axe’ varieties and the classic Acheulean, some degree of metrical overlap was found in each of the key dimensions of bifacial technology (e.g. weight, refinement, thickness), indicating that some Asian chopping tools fall within the range of variation of Acheulean manufacture (Petraglia and Shipton, 2008). And although bifaces are argued to dominant lithic assemblages in Africa and Western Eurasia by comparison (Norton et al., 2006), a large number of bifaces east of the Movius line are reported from un-dated surface contexts (Moore and Brumm, 2012). Comparatively, assemblages that are reliably dated in SE Asia are limited, but scholars working regionally have argued that bifacial technology is more common to surface finds than is commonly imagined (Brumm and Moore, 2012; Simanjuntak, 2010). The implication here is that the industries east of the Movius Line (i.e. pebble and flake) has not shared the same typological enclosure afforded by Acheulean and its affinities west of the Movius line. In this sense, the robust sample of Acheulean industries east of the Movius line is inflated, whereas bifaces of the Asian Paleolithic are potentially greatly underrepresented, perhaps creating a greater technological gap than is warranted for the stone-tool industries of the Lower Paleolithic.
Non-lithic industries.
The alternative use of organic tools has long been suggested, but until recently this perspective operated as a theory independent of any direct line of evidence due to the remote chance of organic preservation. However, two recent studies of surface modification concerning cut-markings on faunal bone and a single shell tool with engraving dating to the Middle Pleistocene in Central Java, have been interpreted as evidence for organic tool-use in


H. erectus of Middle Pleistocene ISE Asia (Choi and Driwantoro, 2007; Joordens et al., 2015). The former study hinges on potential cut-markings on faunal assemblages dated to the Middle Pleistocene (Choi and Driwantoro, 2007), which were compared against experimentally produced cut marks, leading the authors to conclude that the modification was “most plausibly” inflicted by thick clamshell (Choi and Driwantoro, 2007). The later study is restricted to just one geometric engraving on shell. The authors found that the marking inflicted on the shell were “unambiguously consistent” in their morphological patterning, which is to say they could not have occurred naturally through taphonomic processes.
Archaic Homo in Pleistocene Southeast Asia
As is often the case, drawing the boundaries of Pleistocene hominin distributions calls on a combination of fossil and archaeological evidence. The hominin biogeography of the Pleistocene here is informed through fossil evidence and, to a much greater extent, the stone tool artifacts. The Paleolithic character of these stone tools and their perceived antiquity is deeply set into the archaeological perspectives of the region. The assemblages of Song Terns and Liang Bua in Indonesia represent two of the more thoroughly studied sequences with occupational levels dating to the late Middle Pleistocene, both of which occur in regions with well-documented with fossil hominins. While H. floresiensis was discovered with associated stone tools at Liang Bua, the particular Homo species responsible for the Song Terns accumulations on Java Island will remain less clear until a better representation of the late-surviving hominins emerges. Although these assemblages have been dated prior to the arrival of modern humans, the late period of the Middle Pleistocene of SE Asia is somewhat unmoored chronologically in a number of respects and difficult to assess. Due to the archaic morphological qualities of H. floresiensis in particular, as well as the character of stone tools
40


more regionally, approaches to stone tools often draw from perspectives on Early Pleistocene technological behavior. Stone tools are widely abundant and on occasion have been found in, or around ancient Pleistocene geological deposits and continually factor into the spatial distribution and chronological boundaries of primitive Homo across greater SE Asia, despite rarely being reported in any direct or peripheral association with the physical remains of archaic hominins. In SE Asia, this subject is especially complicated by a very murky understanding of archaic Homo adaptation, their technological capacity, and their distributional extent throughout the Pleistocene.
Indonesian hominins.
The chronological and evolutionary perspectives on early Homo in SE Asia have largely been confined to fossil-bearing contexts of Pleistocene Indonesia (southeastern-most Sundaland). Although far from straightforward, indications are that the arrival of early Javanese H. erectus occurred sometime around the Early to Middle Pleistocene boundary [1.2 ma - 0.8 ma; (Larick et al., 2001; Hyodo et al., 2011)], or perhaps more firmly in the Early Pleistocene as some scholars’ have attested [<1.5 ma (Swisher et al. 1994; Semah et al. 2000)]. However, little is known about the geographic distribution of these colonizers, nor is it clear how the Pleistocene environment impacted settlement of early Javanese H. erectus of Pleistocene ISE Asia.
During arid glacial intervals of the Quaternary period, the island of Java would have episodically formed land bridges between peninsular Malaysia and the Greater Sunda Islands (i.e. Borneo, Sumatra) to the east (Bush and Fairbanks, 2003). The Lower and Middle Pleistocene in particular, are thought to correspond with the expansion of savanna and open forest habitat. These suggested “savanna-like” paleo-climatic conditions would have been
41


suitable habitats for human and animal dispersals (Bettis III et al., 2009), and perhaps offered significant refugia for hominins (Louys and Turner, 2012). While recent ecological approaches have begun to develop a picture of the environmental constraints and landmass configuration during the initial dispersal of H. erectus (e.g. Dennell, 2009; Semah et al., 2016), the precise timing of their arrival and subsequent evolutionary trajectory remains less clear. For instance, the periodic passage of migratory H. erectus into the southernmost extent of Sundaland would have temporarily stranded populations in the Greater Sunda Islands during interglacial intervals, exposing one or more H. erectus populations to highly variable island conditions (Semah et al., 2016), but it is still an open question whether the Javanese H. erectus represent a long sequestered lineage, or alternately, intermittent migrations of variant H. erectus which either stabilized, or replaced previous H. erectus forms during periods of low sea stand (Falgueres et al. 2015).
Unlike neighboring Java Island, the island of Flores remained isolated during the lowest sea-stand periods of the Pleistocene and was unreachable by land (Morwood et al., 1998). Importantly, securely dated lithic assemblages define the oldest human occupation here. Stone tools recovered alongside datable material from Mata Menge (1.1 ma - 0.8 ma) attest to the presence of a hominin species by the Middle Pleistocene (Morwood et al., 1998; Brumm et al., 2010). In addition to the stone artifacts, the site of Mata Menge recently produced hominin fossils dated to ca.8—0.65 (van der Bergh et al., 2016). With whom do the island inhabiting hominins share lineages with, and how did they come to colonize a remote volcanic island in Indonesia are among the most important and timely paleoanthropological topics debated today. The two scenarios which have traction thus far involve differing migratory routes which trace back to different source populations.
49.


The most geographically-logical model involves the transit of H. erectus populations from east to west, crossing a deep underwater channel separating Java from Flores. However, the island-hopping event is important, because it would require that//, erectus were not only capable of successfully navigating difficult sea crossings, but also capable of marshalling the resources and co-operation at least as many times necessary to become a biologically viable population on Flores Island. The alternative hypothesis is that Flores was populated from the north via Sulawesi (Morwood and Oosterzee van, 2007; Morwood and Jungers, 2009; van der Berg et al., 2008). This model proposes that the source population became dislocated from Sulawesi through force of nature, and were transported to Flores clutching natural rafts held together by debris and vegetation (Smith, 2001; van der Bergh, 2008).
However likely, it should be noted that the model requires that early hominins were displaced in multiple evolutionarily significant ‘accidental rafting’ events. The latter described above, presumably followed a much earlier separation from the upper margins of SE Asia, in which hominins were swept downward via oceanic currents running north to south and onto Sulawesi, which may have been connected to Flores (Dennell, 2014). In either scenario, paleoanthropologists are asked to entertain surprisingly complex scenarios with considerable evolutionary ramifications. The former involves a greatly expanded behavioral range for H. erectus stalking the coast lines only to give rise to a descendent hominin species with significant evolutionary reversals (e.g. diminished brain size). The later hypothesis involves either an earlier, unknown hominin, or perhaps H. habilis, migrating from Africa far earlier than the fossil record currently indicates.
Unfortunately, mainland SE Asia is conspicuously void of fossil hominins which has made it difficult to assess the differing models of hominin dispersal (Marwick, 2009).
44


Alternatively, stone tools have helped extend regional chronologies back into the Middle Pleistocene in lieu of fossil evidence, which is slowly reestablishing the notion of archaic hominin occupation in adjacent areas. In Sulawesi, for instance, a lithic assemblage has been securely dated to the late Middle Pleistocene (van der Berg et al. 2016b), indicating, at the very least, the possibility that multiple hominin dispersals routes are conceivable, including those which skew the current debate toward evolutionary linkages between H. erectus and H. floresiensis.
Interpretive Approaches
Traditionally, a number of lithic industries of ISE Asia have been recognized as of Lower—Middle Pleistocene origin in Java (i.e. Sangiran, Sambungmacan, Pacitanian, Ngandong), Sulawesi (Cabengian), and the Philippines (Cabalwanian). And while these industries are perpetually reconsidered in light of new data or alternative theories, these artifacts have received persistent scrutiny since their discoveries (e.g. Bartstra and Basoeki, 1989; Bartstra et al., 1994; Jacob et al., 1978; Pawlik and Ronquillo, 2003), discovered in close proximity to the fossil-bearing localities have not only played a significant role in expanding on the behavioral antiquity of SE Asian fossil hominins, but also the defining criteria used in lieu of fossil evidence while developing regional chronologies. The disputed artifacts of early H. erectus in ISE Asia and the antiquity of lithic technological behavior more broadly across the entire region, has its origins here in Java, Indonesia. The geological formation (Kabuh, and the directly overlain ‘Grenzback layer’) from which the majority of classic H. erectus fossils were discovered have also been known to yield small concentrations of imported artifacts, variously identified as ‘Sangiran Flakes,’ or Ngebung artifacts, which have been described as morphological similar to the Sangiran industries (Simanjuntak and
44


Semah, 1996). The Ngebung artifacts consist of a small number of flakes and a cleaver which has been described as an ‘Acheluean-like’ assemblage (Semah et al., 1992). More recent excavations from the ‘Grenzback deposit’ yielded a small number of flakes, which were given age determination of more than 1 Ma (Widianto, 2006). However, these artifacts occur in unconstrained geological deposits, which have long been noted for their severe erosional uplifting from younger deposits (Bartstra 1985; Bartstra and Baseoki, 1989; von Koenigswald, 1973). Accordingly, many scholars are cautious in accepting Middle Pleistocene ages for artifacts excavated from Sangiran localities, and hence skeptical of their affiliation with the H. erectus fossil materials. Nevertheless, a review of the artifacts that have come to define early archaic technological behavior in SE Asian hominin species is described below.
Pacitanian industries.
The Gunung Sewa region of Java is widely known for the abundance of stone tools discovered along the terraces and riverbeds of the Baksoka River. The Baksoka River stone tools are broadly characterized by crude hand-axes, hand-adzes, and other chopper-chopping tools made from silicified limestone, silicified tuff, and petrified wood (Soejono, 2001). Establishing the antiquity of the Baksoka River artifacts, popularly known as the “Pacitanian” industry (also formerly known as the Pajitan), has been problematic.
Historically, the Pacitanian chopping tools have been attributed to the Middle Pleistocene on account of their ancient appearance and the geological context from which they derive (e.g. Van Heekeren, 1972). The original Pacitanian distinction proposed by von Koeningswald, (1936) was conceived before radiometric dating techniques were available. At the time, Pleistocene chronologies in Europe were devised from the systematic identification of glacial and interglacial events within the horizons of river-terraces (Moore and Brumm,


2007). Studies of the greater Asian Paleolithic followed in this way, with the geomorphological chronologies patterned after the European Pleistocene terrace sequence in an attempt to develop a compatible prehistoric framework (Pope, 1997). However, while the age of stone tools defined by riverine context during the early 20th century continue to figure into interpretation of Paleolithic SE Asia, the various opinions and explanations flowing from them have been left unanswered in favor of research conducted in cave and rock-shelters (Anderson, 1997).
The Pacitanian assemblages primarily occur in secondary context, and the few artifacts discovered in situ have not been recovered alongside associable fauna and human fossils, or chronometrically datable materials (Bartstra, 1984; Simanjuntak, 2004). Up to now, scholars have not reached consensus on the Pacitanian and continue to debate the authenticity of differing age-estimations concerning the stone tools in question. For instance, in a geomorphological study of the surrounding Baksoka River landscape, Bartstra (1983) estimated that the artifacts eroding from fluvial terraces were likely terminal Pleistocene or Holocene in origin. Moreover, Bartstra (1982) contends that the co-occurrence of ‘Paleolithic types’ at river localities as well as in ‘Neolithic’ surface assemblages located away from rivers confounds the historical tendency to always associate Pacitanian industries with Lower Paleolithic technologies. Interestingly, Bartstra (1982) further comments that the various sites that make up the Pacitanian tradition might be better explained as “different seasonal or occupational activities,” which is a concept elaborated in a recently proposed model of stone tool use by Moore and Brumm (2007). In any event, the mixture of Pacitanian stone tools with elements of Neolithic industries (e.g. adzes) underscores the uncertainty that comes with
40


determining the age of ‘old-looking’ Pacitanian stone tools (Brumm and Moore, 2012; Simanjuntak, 2004).
For the most part, opinions are still divided on the chronological placement of the Pacitanian industry and hinge on typological interpretations. On the one hand, it is problematic to link the Pacitanian industry with primitive Homo by extension of Early Pleistocene typologies because, as Corvinus (2004) has rightly pointed out, the heavy, cobble core-tools that typify the Pacitanian are “part and parcel with the chopper/chopping tool complex,” which by most accounts is wholly different than the Acheulean concept used in the Western Old World (2004:145).
Sangiran industries.
The ‘Sangiran Dome’ is an important geological feature in Central Java known for its complete Upper Pliocene through Middle Pleistocene stratigraphic sequence (Bouteax and Moigne, 2010). The Pucangan (Early Pleistocene) and Kabuh (Middle Pleistocene) sequences are particularly important, with the former having yielded H. erectus fossils sometimes described as archaic, and the later yielding the largest concentration of classic Javanese H. erectus. The first Sangiran flake stone tools were discovered atop Ngebung Hill in gravels overlying the fossil-bearing layers containing hominin and faunal remains, and hence were subsequently interpreted as Middle Pleistocene in age (Corvinus, 2004). The Ngebung artifacts have twice been dated to the Middle Pleistocene [250 ± 70 ka (Suzuki et al., 1985) and ca. .9 ma (Saleki, 1997)], but have since been reinterpreted as Upper Pleistocene in age on account of the ambiguous geological context (Choi and Driwantoro, 2007).
In more recent excavations at Ngebung, 20 chalcedony artifacts from the Kabuh Beds (Middle Pleistocene) were recovered, which are claimed to be comparable to the Sangiran
47


flakes and interpreted to be the handiwork of H. erectus (Simanjuntak and Semah, 1996). However, the authenticity of the artifacts has been questioned on grounds that they may not represent true artifacts (Corvinus, 2004).
Roughly 60 km to the east of the Sangiran site, crude chalcedony flakes were collected from high terraces above the Solo River near the Ngandong H. erectus fossils (2004:143— 144). However, any association of H. erectus with the Ngandong stone tools has been mostly unconvincing due to the indirect relationship between the artifacts and the geological context (Choi and Driwantoro, 2007). Nevertheless, others maintain that a direct association between late H. erectus and the Sangiran and Ngandong artifacts is possible, while also citing both the typological similarities in the flake industries and the corresponding geological contexts from which they were discovered (Corvinus, 2004).
Remarkably, there are still no stone tool artifacts that have been securely removed from early Javanese//, erectus fossil contexts (Dennell, 2016). In fact, only recently have ISE Asia lithic assemblages actually returned dates confidently placed in the Early—Middle Pleistocene (Brumm et al., 2010; Moorwood et al., 1998). Currently, the best evidence for early Middle Pleistocene stone tool use in archaic hominins occur in dated assemblages within Flores, Indonesia (Mata Menge), which have produced ages consistent with the early Middle Pleistocene occupation of H. erectus in Java [.8-1 Ma (Moorwood et al., 1998; van den Bergh et al., 2016)]. Recently, supporting evidence from the nearby Mata Menge assemblages, located on Flores, Indonesia, have recently been found in association with fossil evidence within a constrained layer dating to between .65 - .8 Ma.
4R


Flake and core dichotomy.
An oft-cited typological distinction in SE Asia Paleolithic technology is the simple dichotomy differentiating ‘flake-based’ industries from so-called expedient ‘core-tool’ industries. Within the literature this dichotomy is also referred to as the ‘chopper/chopping -tool industries’ and the ‘pebble-and-flake technocomplex’ (e.g. Anderson, 1990; Bellwood,
1997) and has generated various interpretations concerning the nature of the patterning (e.g. Moore and Brumm, 2007). The conventional view on the meaning of this patterning has been driven by the association of core-tools with putative archaic Homo populations believed to have occupied the region. The bulk of core-tools are derived from open-air river-terrace contexts and therefore lack the chronometric dating necessarily to confirm their antiquity. By comparison, pebble and flake-based industries are often discovered in cave and rock-shelters (Moore and Brumm, 2007), which have provided a temporal distribution of these technologies primarily confined to the Lower Pleistocene. Although the typological scope of this distinction condenses lithic variability within SE Asia by their relative differences, a definitive cultural division between these industries are tempered by the high incidence of internal variation and significant overlap (Rabbet, 2012).
Currently, the most compelling evidence toward identifying the first unambiguous lithic “tradition” of SE Asia is represented by the rounded-base point technologies of Java (Bellwood, 2007), and the flaked-based Tolean industries of Sulawesi (Pasqua and Bulbeck,
1998) , but the appearance and distribution of these technologies is poorly understood. Nevertheless, there has been no shortage of cultural descriptors used in reference to many regional variants of Paleolithic technology, especially with regard to the pebble and core-tool industries and are the source of much ‘outsider’ confusion (Reynolds, 2007).
49


The core-tool industries, and to a lesser extent flake-based industries of insular and island SE Asia have been routinely interpreted as the technological equivalent of the African Oldowan “Mode-1” (Clark, 1977). In ISE Asia in particular, H. erectus is fairly well represented by fossil evidence and have assisted the long-held assumptions that the character of stone tools was indicative of hominin manufacture. Historically, the simple nature of coretool technologies and the typological affinities they share with classic Lower Paleolithic industries (i.e. Oldowan) has been the primary basis for using large, heavy core-tools as population markers for the distribution of archaic hominins (Moore and Brumm, 2007). There is a compendium of core-tool industries attributed.to H. erectus on typological grounds. Namely, those decidedly characteristic of the ‘Pacitanian’ from Java, or the ‘Cabalwanian’ of the Philippines, but also the undesignated bounty of core-tools reported from across greater SE Asia.
Similarly, finding a chronological place for the ‘flake-based industries has suffered from the same ambiguity between fossil evidence and stone tools. In particular, the ‘Sangiran’ and ‘Ngandong,’ both small-flaked based industries named for their purported affiliation with H. erectus bearing sites in Java, would be inversely related to the proposed dichotomy (Moore and Brumm, 2007). Adding to the typological confusion, “flaked-based” industries are occasionally lumped together with the presumably older ‘core-tooT tradition (Bellwood, 1997), presumably due to the inferred shared technological simplicity of both industries. Further compounding the issue, the arrival of H. sapiens apparently had little impact on the stone-tool kits of the relict hominin inhabitants.
The lack of chronological resolution renders typological approaches inadequate in assessing hominin behavior, and for that matter identifying the arrival of H. sapiens. The


technological simplicity and continuity exhibited in SE Asia present interpretive challenges in contrast to the rather straightforward relationship between behavioral modernity and technological complexity seen in western parts of the Old World. However, analyses aimed toward the interpretation of this variability are far from being abandoned. In contrast to the fossil evidence, stone tools are widely abundant and on occasion have been found in, or around ancient Pleistocene geological deposits. Unsurprisingly, stone tool artifacts continue to factor into the spatial distribution and chronological boundaries of primitive Homo across greater SE Asia, despite rarely being reported in any direct or peripheral association with the physical remains of archaic hominins.
In lieu of direct fossil evidence, it would be particularly informative if a broad classificatory schema of stone tools could shed light on the various artifact forms of SE Asia which, in turn, could help suss out the archaeological manifestations of archaic and modern populations across the wider region Whether these typological differences can be reconciled with the biogeographical history of archaic hominins is ultimately still subject to debate, but the dichotomy itself has earned its stay as a useful point of contrast in a region in want of comparative attributes. Recently, an alternative explanation proposed by Moore and Brumm (2007) has suggested that the technological differences between flake and core dominated assemblages is spatially sensitive, and in fact represent different phases of transport within a single reduction strategy. Even if it turns out that this patterning is different ends of a single technological spectrum, it still does not address how to tease apart behavioral differences between species.


CHAPTER IV
CONTEXT
The possible coexistence of archaic species and modem humans in ISE Asia is controversial and many of the evolutionary parameters are, at present, ill-defined. Nevertheless, the simple stone tool artifacts found here undoubtedly straddle this important evolutionary context within which mounting biological, ecological, or cultural pressures beset the dwindling archaic Homo populations at Flores, and quite possibly multiple localities on Java. However, mapping out the technological landscape during this dynamic biogeographical period faces inherent difficulty in distinguishing AMH from archaic hominin species on typological grounds alone. At roughly the same period hominin occupation of Indonesia is ending on Flores, late-surviving Neanderthal populations and H. sapiens cohabitated parts of Western Eurasia. And reconstructing the lifeways of contemporaneous hominins in Paleolithic Europe has drawn heavily on the study of transitional industries and the emergence of modern foraging behavior, although these archeological representations are not observable in every region (Bar-Yosef, 2002). To that point, there is no other region more underrepresented in their artifact histories than the various Homo species of Pleistocene SE Asia. Attempts at defining the cultural, spatial, and biogeographical boundaries in the Pleistocene era have been slow to develop here, in large part because evidence of technological behavior in H. sapiens, H. erectus, or H. floresiensis is difficult to prove on a case by case basis.
By comparison, it has become fairly evident that the sequence of technological change of the African and Eurasian Pleistocene broadly trends with the evolutionary history of the genus Homo (e.g. see Ambrose, 2001; Clark, 1977; Foley, 1987; Foley and Lahr, 2003). For instance, the origin of Acheulean bifacial industries and the wave of H. erectus dispersals into


Eurasia during the middle Early Pleistocene (Bar-Yosef and Belfer-Cohen, 2001), or the emergent complexity and diversity of Paleolithic tool forms in the Late Pleistocene and the underlying behaviors which drive technological diversification among these industries. The later have often figured into the adaptive package of incipient H. sapiens and their subsequent geographic dispersal (Mellars, 2007).
However, recent studies have indicated that the forward leap in Late Pleistocene technological and behavioral innovation which popularize the notion of the ‘Upper Paleolithic Revolution’ does not uniformly emerge in the archaeological record. Scholars have cautioned against transposing behavioral correlates of one region onto another (Henschilwood and Marean, 2003). In fact, an increasingly popular opinion on Paleolithic variability is that these newly acquired modern behaviors of H. sapiens do not go entirely unmatched by late-surviving hominin populations. For example, scholars have recently reconsidered the possible involvement of Neanderthals in the early stages of the Aurignacian industry, which has been a key temporal diagnostic exclusive to H. sapiens in Europe (Conard et al., 2004). Nevertheless, the diversification and refinement of stone tool artifacts is still generally a very resourceful point of contrast in assessing behavioral variability across and within human species. Stone tools offer deep temporal resolution, gross measureable attributes, and ubiquity over time. But intriguingly, the study of lithic artifacts of the Late Pleistocene in SE Asia have not yet yielded any kind of technological pathway to behavioral modernity.
H. sapiens of Pleistocene Southeast Asia
The Upper Pleistocene period (MIS 4/MIS 3). is most notably associated with the rapid geographic expansion of ‘behaviorally modern’ H sapiens out of Africa which roughly coincides with the replacement of indigenous Neanderthal populations in Eurasia. The


centerpiece of this evolutionary period—the Middle to Upper Paleolithic transition— has been the benchmark for describing the timing and cultural processes behind modem human colonization. While the dawn of this age (MIS 5) records the earliest evidence of an anatomically modern population in the Levant, archaeological and molecular data have preliminarily indicated that this initial expansion beyond the African continent was likely a failed dispersal (Macualay, 2005; Mellars, 2006; Shea, 2008). Certainly, there is good evidence that behavi orally modern//, sapiens were equipped with widening range of complex tool forms and higher capacity for symbolic representations, ritual paraphernalia, elaborate body ornamentation, among other classic archaeological markers. However, the popular synthesis of cultural traits considered germane to distinguishing AMH from those of non-modern predecessors, as well as contemporaneous populations of non-human hominins, are far from compatible across all Late Pleistocene cultures.
Modern human expansion.
The premise of the popular single-dispersal model is largely predicated on the demographic expansion of newly acquired adaptive behaviors, which quickly take hold in Africa as closely-related populations are swiftly replaced, or integrated (Mellars, 2006). Despite never bearing anywhere near the full spectrum of modern behavioral markers, the Late Pleistocene of SE Asia holds some of the earliest fossil evidence for modern humans outside of Africa ca. 42—46 ka [Laos, ca 46—63 ka (Demeter et al., 2012; Demeter et al., 2015); Sulawesi 46—30 ka (Barker et al., 2007)]. More surprisingly, the initial colonization of Australia is now confidently estimated at ca. 55—50 ka (Hiscock, 2008), with genetic evidence suggesting an even earlier migration at ca. 62—75 ka (Rasmussen et al., 2011). Such an early occupation of Sahul is somewhat unexpected, particularly because it is among the earliest


evidence for modern humans outside of Africa [ca. 60—40 ka (Mellars, 2006)], but also because it predates the earliest well-dated fossil evidence for H. sapiens in SE Asia. The conventional explanation has contended that a costal corridor facilitated a rapid migration along the Indian Ocean, essentially bypassing interior mainland (Dennell 2003; 2008). However, the archaeological evidence from Australia, not unlike that of SE Asia, occurs without as much as trace of the classic ‘innovative package’ that signals the advent of modern human behavior in Africa (Boivin et al., 2013).
Despite the conflicting sequence, the earliest dates from both Sunda and Sahul are considered chronological significant in configuring a timeline for a single-dispersal model. In contrast, the possibility that the first dispersal of modern humans was, in fact, a successful dispersal whereby a rouge band of early H. sapiens trekked much deeper into the Asia continent, is still considered by some to be a viable, and understudied hypothesis. Dated archaeological material from ISE Asia continues to establish initial human occupation more firmly into the Late Pleistocene (O’Connor, 2007), but there have been a few indications that have hinted such an early colonization is possible.
In South China at the upper margins of the SE Asia mainland, U-series dating of a skull and post-cranial elements were found to be no younger than 68 ka, but more likely ca. 111— 139 ka (Shen et al., 2002), which if confirmed would indicate that an earlier population of H. sapiens was at the doorstep of the SE Asia sub-region at the close of the Middle Pleistocene. Moreover, a small number of teeth (which have since been lost) were thought to exhibit dental elements of modern H. sapiens, were found in association with the Punung faunal assemblage dated to 128 ± 15 and 118 ± 3 ka in Java, Indonesia (Storm, 2005; Westaway et al., 2007). However, the evidence for such an early chronology is at present very limited and questionable,


and also out of step with the robust archaeological visibility that follows in the Late Pleistocene of both mainland and ISE Asia (Rabbet, 2012).
The uncertainty which surrounds the initial colonization brings attention to the fact that human presence in SE Asia around the transition from the Middle Pleistocene to the Late Pleistocene is not fully discounted by scholars working in the field. Although it is merely speculated on at this point rather than offered as an alternative perspective, the major implication within SE Asia is that it significantly widens an already confounding period of time of which there is considerable uncertainty regarding the behavioral differences and potential interaction between archaic hominin species and modern humans.
The subtle footprint of behavioral modernity in SE Asia presents difficulties in ruling out the scenario, however unlikely. Problematically, by whichever selective pressure primitive Homo were ultimately vanquished, modern humans were able to manage without major technological innovation. The technological continuity from the Middle to Upper Pleistocene obfuscate a crucial line of evidence commonly used to reconstruct the movement of modern H. sapiens throughout the Old World. However, the evidence placing modern humans in Middle Pleistocene ISE Asia is razor thin and highly contentious, but remains an entirely possible evolutionary scenario.
Late Pleistocene Archaic Hominins
The fairly routine consideration of a late H. erectus extinction is mostly perpetuated by regional perspective on simple lithic industries in Java. To some extent, the good evidence for H. erectus persisting throughout the Middle Pleistocene is predicated on a historical understanding of typological variation that has fallen out of favor in recent years. While there is sizable and convincing artifact evidence of H. erectus ’ handiwork outside of SE Asia during


the Early and Middle Pleistocene, the many ‘old stone tool’ claims originating within the region are fraught with controversy and dispute (as discussed in the previous Chapter). At present, the fossil evidence alone is not currently sufficient enough to address the circumstances surrounding a possible late extinction of H. erectus. Nonetheless, the debate surrounding the taxonomic relationship between archaic populations e.g., H. erectus, H. ergaster, H. hablis, and the remains of the small-bodied hominin H. floresiensis (Aiello, 2010; Argue et al., 2007; Baab, 2016; Brown and Maeda, 2009; Dennell et al., 2014; Moorwood et al., 2004) hint at the possibility of a remarkably dynamic and variable biogeographic hominin landscape throughout the Pleistocene SE Asia (Marwick, 2008).
The clearest evidence of a relict hominin species surviving into the Late Pleistocene comes from H. floresiensis on Flores, Indonesia (Brumm et al., 2010; Brumm et al. 2016; Moorwood, 2004; van der Bergh et al., 2016), which also happens to hold the only unambiguous evidence for stone tool use and manufacture in non-human hominins of this period (Moore and Brumm, 2007). Other scholars have presented radiometrically-dated evidence suggesting that a late variety of Javanese erectus persisted into the Late Pleistocene (Swisher et al., 1996; Yokoyama et al., 2008), while more recently Larick and Ciochon (2015) have hinted more at the possibility of another small-bodied hominin occupying the Philippines in the Late Pleistocene. However, the reconstruction of hominin biogeography throughout Pleistocene SE Asia continues to be treated cautiously; crucial information is still needed in order to establish the evolutionary histories of the early period and late period Javanese H. erectus, with the later formerly described as archaic H. sapiens (Hawk et al., 2000; Stringer, 1987) and the antecessors of H. floresiensis on nearby Flores Island still unclear (van der Berg


et al., 2016a). The chronological prime for each of these SE Asian fossil hominin populations remains an important, but mostly unanswered question in the field of paleoanthropology.
Paleoenvironmental Conditions
The end of the Middle Pleistocene also coincides with the beginning of the last glacial period, which is marked by a global cooling trend and glaciation, which characterizes much of the Upper Pleistocene period. While generally marked by an overall temperature reduction, the climactic conditions during the Late Pleistocene were highly variable; regular climatic amelioration brought on rapid changes in sea levels, as well as local and regional-level compositional changes to vegetation. The instability in climate would have introduced localized and regional-wide ecological stress to modern humans and any relic archaic hominin populations; but also, resulted in the formation of contiguous land bridges with the exposure of the Sunda shelf. At the height of the Last Glacial Maximum (LGM) when sea-levels were at their lowest, the Sunda subcontinent would have included the southern islands of Borneo, Sulawesi, Sumatra, and to the north, the Philippine island of Palawan. Overall, the climatic fluctuations would have periodically linked chains of islands to mainland SE Asia, increased inter-island visibility, and migratory opportunities would have vacillated with the peaks of glacial and interstadial periods.
The tropical environment has long been viewed as a formidable ecological barrier to both archaics and modern humans, but these limitations are largely informed by contemporary tropical foragers and varied ethno-historical accounts from tropical SE Asia, the Amazon, and the Congo River Basin (e.g. Bailey et al., 1991). The reach and intensity of tropical rainforest coverage at various intervals of the Late Pleistocene are still regularly debated, but the


prevailing opinion of many in the field is that the modern extent of the tropical rainforest in insular and ISE Asia is not analogous to environmental conditions during the Late Pleistocene.
The characterization of Pleistocene SE Asia as consisting of a belt of tropical rainforest running east to west across the mainland has recently evolved into a more complex picture of mosaic habitats, with large tracts of savanna and open woodland bordered by peripheral concentrations of tropical rainforest (van der Berg, 2001; Bird et al., 2005). The archaeological evidence from the peninsular Thailand, Malaysia, and ISE Asia largely conform to this ecological description, with a good proportion of the mid-Late Pleistocene faunal assemblages showing a reliance on auxiliary resources extracted from diverse habitats, including tropical rainforests (Conrad 2015; Roberts and Petraglia, 2015).
Ecological challenges.
From a purely ecological standpoint, the exploitative potential of the modern tropical environment suggests that closed canopy forests were uninviting and unproductive environments, which would have, or have historically required access to alternative resource patches, or seasonal reliance on agricultural economies in order to sustain an adaptive strategy centering on tropical subsistence (Bailey et al., 1991). Given how persuasive the distribution of resources in the coordination of movements and subsistence strategies in H. sapiens, it has been long argued that modern humans would have found tremendous difficulty in striking a balance between the distribution of resources and tropical-subsistence strategies without the contemporary reliance on agricultural communities (Bailey et al., 1989; Bailey et al., 1991; Headland and Bailey, 1991; Kelly, 1995). Ecological stability is obviously paramount to the survival of human populations, and it is the relatively uniformity in the range of adaptive responses employed by modern humans— with relation to specific environmental


constraints— which provide the basis for ecological, or behavioral modelling used in this thesis.
However, as discussed above, modeling hunter-gatherer activities in prehistoric H. sapiens of SE Asia is difficult due to the environmental variability, and the little knowledge available to simply approximate adaptive strategies in this region. As was reviewed above, typological succession is not the cornerstone of assessing variability in hominin behavior in SE Asia What’s more, whole assemblage appraisals—often require details of the smaller but important elements that make up the bulk of an assemblage— have been historically disregarded in search of more diagnostic artifacts (Reynolds, 1992). However, a number of comprehensive reports have been published over the years that have tilted subsequent studies in the direction of whole assemblage analysis, although no attempt has been made to synthesize these data sets under a single interpretative framework.
Traditionally, the lack of regional specialization throughout the SE Asian Paleolithic has been explained in terms of limitations imposed by the natural environment. In a manner of speaking, this is an ecological explanation for a material culture phenomenon. But importantly, the notion of a raw-material boundary has not advanced the study of lithics much further toward the explanatory framework of behavioral ecology. The archaeological reality that is the “Movius Line,” and the arguments that closely follow the relationship between the geographic demarcations and technological division are more of convenient explanation than a theoretically guided conclusion. Clearly, the lack of fine-grained raw material and the ample availability of organic material impacted the intensification of stone tool technology. But it is also apparent from the numerous bone and shell technologies that the understanding of


fracturing mechanics necessary for elaborative artifact forms is not contingent on high quality lithic material.
The technological intensification of non-lithic materials demonstrated the propensity for Pleistocene H. sapiens to produce standardized forms, suggesting that the general lack of fine-grained raw material was not an overriding restriction in the production of complex technology. By comparison, the adaptive strategies of contemporaneous European Late Pleistocene groups are strongly directed toward locating a limited number of select finegrained raw materials (Riel-Salvatore et al. 2008; Rabbet 2012). In this sense, access to raw material quality and quantity were important factors guiding hunter-gatherer landscape use and this type of landscape interaction has been shown to possess traceable archaeological patterning (Andrefsky 1994; Parry and Kelly 1987; Torrance 1983). Comparatively, Pleistocene hunter-gatherers of SE Asia are often characterized as technological opportunists; in contrast to their Western Eurasia counterparts, it appears that prehistoric inhabitants of SE Asia were guided to lithic raw material by the specific activities they carried out on the landscape (Rabbet 2012). In western Eurasia, Pleistocene hunter-gatherer landscape-use was likely dictated by narrow and unpredictable resources and fewer subsistence options while foragers in the tropics would have likely needed to tailor their adaptive system in response to the forces of resource diversity (Rabbet, 2007) Binford (1980; 1982) suggested that foragers subsisting in regions of high primary mass, such as the tropics, are much more inclined to practice high residential mobility by moving frequently between centers of foraging radii. Kelly (1983) demonstrated the relationship between environment and forager mobility strategies on a global scale. However, it has been noted that tropically adapted hunter-gatherers, which rely on agricultural communities to offset key deficiencies in carbohydrates
01


and fats, may not reflect an accurate representation of mobility and economic strategies practiced by prehistoric hunter-gatherer populations in SE Asia.
Currently, very little research has been carried out in an attempt to reconcile what such a behavioral strategy would result in archaeologically. While the research presented in this thesis cannot resolve these questions relating to tropical or otherwise adaptive strategies unique to the SE Asian Paleolithic and how they might manifest archaeologically, the data compiled for this project can directly test the possibility that, in fact, lithic technology was integrated into the mobility strategy despite the deemphasized aspects of fine-grained lithic material and formal tool production.
The exploitative potential of the modern tropical environment suggests that closed canopy forests are unproductive environments for foraging, and would have likely been a formidable barrier to archaic and modern humans. In part, this has been supported by the thorough review of historical and contemporary ethnographies on foragers subsisting in interior rainforests indicating they require access to alternative resource patches, or seasonal reliance on agricultural economies in order to sustain an adaptive strategy centering on tropical subsistence items (Bailey et al., 1989; Bailey and Headland, 1991). However, the reach and intensity of tropical rainforest coverage at various intervals of the Late Pleistocene are still regularly debated; the prevailing opinion of many in the field is that the modem extent of the tropical rainforest in insular and island Southeast Asia is not analogous to environmental conditions existing during the Late Pleistocene.
However, the characterization of Pleistocene Southeast Asia as consisting of a belt of tropical rainforest running east to west across the mainland has recently evolved into a more complex picture of mosaic habitats, with large tracts of savanna and open woodland bordered
eo.


by peripheral concentrations of tropical rainforest (van der Berg 2001; Bird et al. 2005). The archaeological evidence from the peninsular Thailand, Malaysia, and island Southeast Asia largely conform to this ecological description, with a good proportion of the mid-Late Pleistocene faunal assemblages showing a reliance on auxiliary resources extracted from diverse habitats, including tropical rainforests (Conrad 2015; Roberts and Petraglia 2015). Despite the complexities of the Southeast Asian paleoecology, recent and earlier approaches have turned to modeling hunter-gatherer mobility around “peaks of abundance.”
Evidence suggests that a savanna-like corridor facilitated rapid colonization of the Southeast Asia peninsula. While open savanna-like land access may have opened up accessibility to more segregate and predictable resources, it was likely a constricted and competitive habitat for exploitation. Dependence on auxiliary resources appears to be the norm, not the exception. Comparatively, resource distribution in the modern tropics is somewhat easy to place on a behavioral ecology map. Resources are unevenly distributed far and wide, and tropical primary biomass is rich in diversity and abundance but mostly inaccessible for animals, and terrestrial secondary biomass is consequently unpredictable, diverse, and solitary.
Site Background
Currently, the origin and definability of the SE Asian Late Pleistocene is not easily described in relation to stone tool industries. The title ‘Paleolithic’ is referenced often within the literature, but is used more tacitly as a general category for ‘old stone tools’ rather than a temporally significant sequence. For the purposes of making a chronological distinction, the ‘Upper’ refers vaguely to a period upon which modern humans begin occupying greater SE Asia. However, from a technological perspective, it is often as challenging to pinpoint a


transition to Holocene stone tools as it is discriminating between the stone tools of the Middle and Late Pleistocene. At best, the typology of the SE Asian Upper Paleolithic is nebulously defined by the manufacture of informal tools, that is, tools which show little evidence of elaboration, function, or standardization in form. With the possible exception of a slight increase in flake-based assemblages among modem humans occupying rock-shelters in the Terminal Pleistocene, very little is diagnostic about the stone artifacts themselves. And while countless regional and cultural variations have been recognized, ultimately their uniqueness is diluted by the high degree of morphological overlap shared by various other industries, making it difficult to gauge any long-term distribution of artifacts Consequently, among the assemblages included in this study none are clearly linked with a robustly defined industry. Instead, the lithic assemblages are treated as a cohesive sample of assemblages which exhibit shared characteristics which permit their inclusion in this analysis. Chiefly, the informal quality and either their origin from in archaeological context which indicate a hunter-gatherer mode of survival.
f, 4


CHAPTER V
METHODOLOGY
The methodology used in this thesis was originally developed by Barton (1998) in order to link assemblage scale variability with patterns in technological behavior of the Late Pleistocene on the southern Iberian Peninsula (Villaverde et al., 1998). The focus on whole assemblage variability provides a simple measure of changes in techno-economic behaviors and shifting land-use strategies over long periods of time (Riel-Salvatore and Barton, 2004). To date, this approach has continually demonstrated a simple and repeatable method for distinguishing between various patterns of forager mobility and lithic organization at the local and regional scale, and recently at the continental level, with a wider group of assemblages incorporated from the whole of west Eurasia (Barton et al., 2011).
Previous applications of this model have been used to demonstrate continuity between Middle and Upper Paleolithic land-use strategies in southeastern Italy (Riel-Salvatore 2007; Riel-Salvatore and Barton 2004; 2007; Riel-Salvatore et al. 2008), Gibraltar and eastern Spain (Barton 1998; Villaverde et al. 1998), which was later found to hold at the inter-regional level of analysis (Barton et al., 2011). Thus, this model is particularly well-tailored to evaluate adaptive behavior for fundamental differences in supposedly behaviorally transitional periods in Eurasia (i.e. Upper Paleolithic), where traditional typological divisions are not necessarily considered a clean break from archaic to behavioral modernity.
In light of these findings, this model should have appeal in the study of the SE Asian Paleolithic— a spatiotemporal context where typological ambiguity has hindered efforts to develop culturally definable sequences, and where many questions of adaptation and


technology are entangled with issues of environment. The methodology described briefly is oriented toward addressing the questions outlined in the specific aims and are reviewed below:
Whole Assemblage Behavioral Indicator
The basic requirements for this model (WABI) are count data for the whole lithic assemblage including retouched pieces, cores, and debitage, in addition to the total volume for each corresponding unit, layer, or horizon from which the artifacts were retrieved (Riel-Salvatore and Barton, 2004). Here, the formal properties of retouched tools are deemphasized, and instead these data establish a relative frequency that is compared to artifact volumetric density. Following Riel-Salvatore and Barton (2004: 259), “artifact volumetric density is defined as the total number of pieces of chipped stone per cubic meter of excavated sediment” and “...the relative frequency of retouched pieces is simple the count of retouched ‘tool s’... divided by the total number of pieces.” Due to the complex depositional environment of SE Asia rock-shelters and the temporal scope of the project, calculating actual sedimentation rates was essentially required in order to make direct comparisons between assemblages. Time is not often reported as precise data, so the preferred chronological sequence provided by the authors often involved a combination of radiometric dates and relative dating techniques. Nevertheless, the chronological data were broken down in relation to discrete assemblage data, which allowed differential artifact accumulation rates to be estimated.
Evaluating retouch.
The model presented above (WABI) is part of a larger family of quantitatively driven techniques which hinge on the identification and interpretation of retouching intensity. In response to the classic forager studies by Binford (1979,1980), retouch indices were developed to demonstrate that assemblage patterning were strongly influenced technological


organization, mobility patterns, and land-use strategies (Andrefsky, 1991; Bamforth, 1986; Bleed, 1986; Hiscock, 1994; Kuhn, 1991). Despite there being a general consensus on the analytical value of retouch in monitoring these dimension of human behavior, concerns on how best to measure retouch intensity with efficiency and reliability have been raised (Andrefsky 2005; Clarkson, 2002). Studies examining the most effective metrics in analyzing reduction (i.e. retouch) intensity have been put forward (e.g. Clarkson 2005; Hiscock and Clarkson 2005a, Hiscock and Clarkson 2005b).
Notwithstanding these concerns, it has been noted that these diagnostics are ill-suited measurements for ‘Hoabinhian’ assemblages of SE Asia because of the unstandardized tool form and low proportions of retouched tools (Marwick, 2007), which unofficially qualifies as a Litmus test for gauging methodological compatibility with almost any Pleistocene SE Asia assemblage. Other issues stem from the straightforward identification of retouched flakes and core tools, which lacking continuity in standard measures such as size and shape, are difficult to distinguish from generally unintended waste products such as flake shatter and exhausted cores. Consequently, the nature of chipped stone tools in this regional context presents methodological challenges in interpreting retouching intensity that have been largely unstudied.
Despite the concerns raised above, the WABI methodology varies from conventional approaches to lithic reduction strategies in significant ways. The simple requirement of retouched pieces irrespective of tool shape or form is particularly well-suited for the ambiguous tool morphologies common to SE Asia. The most problematic issues inherent in this approach is related to representation; specifically, whether minimal proportions of retouched tools will
67


be substantial enough in number to show meaningful ‘technological separation’ through interassemblage patterning, or if the nature of the variability appears to be relatively stochastic.
Lithic technological organization.
The curated-expedient continuum of lithic organization has generally been framed as a corresponding technological trend which roughly aligns with a particular mobility strategy of hunter-gatherer groups. One the one hand, curated industries are characterized by heavily modified lithics, core reduction is usually extensive, and retouched tools can be larger in size and range in shape- allowing flexible modification of stone tools as access to raw material diminished. In Pleistocene hunter-gatherers, this type of lithic organization appears to have been favored by highly mobile foragers and so consequently would be most often represented in ephemeral archaeological contexts. Comparatively, high densities of artifacts and low frequencies of retouched tools characterize expedient assemblages. Retouched tools are often function-specific and are manufactured for immediate use. Raw material is either locally abundant or raw material procurement is embedded within the use of the site; this type of technological organization is indicative of less mobile groups. These lithic qualities would be most often represented in archaeological contexts with repeated-and sustained occupation such as a residential camp. However, this type of relationship becomes difficult to demonstrate when working with disparate stratigraphic contexts, especially when it is unknown how rapidly assemblages formed. The methodology used in this analysis builds on simple approaches to assessing curation that minimize detailed accounts of the tools themselves and instead, analyze the collective frequency of retouched artifacts in relation to the density of all other artifacts within a given assemblage.


Collating data.
The data compiled for this study come from 17 sites across greater SE Asia region (n=64 assemblages), with the majority of assemblages coming from ISE Asia, and a comparably small but still modest sample coming from the Malay-Thai peninsula within the mainland (Table 2). The simple informal character of stone tools permits the inclusion of diverse assemblages originating from a wide range of spatiotemporal contexts into this model. Although it would be preferable to more narrowly focus on the assemblages of Pleistocene foragers of SE Asia in keeping with the previous applications of this model, there are only a small number of well-dated Pleistocene sites with lithic artifacts, and far fewer had published the required data necessary for this model. In order to expand our database and generate a more robust sample size for analysis, assemblages from the Early and Mid-Holocene have been provisionally incorporated into the general forager model under the assumption that Holocene technological organization differed only marginally from the former Pleistocene occupants of SE Asia, though their mobility strategies tend to favor longer sustained use of base camps.
Analytical considerations.
In deciding to test the relationship between mobility and assemblage patterning while assuming continuity in technological organization throughout all of Pleistocene SE Asia, much of the archaeological content and occupational histories of these assemblages were deemphasized in accordance with the generalized parameters of the WABI methodology. Many of these aspects were considered too cumbersome to incorporate, and ultimately, they are highly important datasets to return to following the results of this model. However, the regional and temporal latitude of this project dictates that such an undertaking is likely beyond the scope of this thesis. Nevertheless, there were a number of selected variables which were


Table 1: Summary of analytical considerations
FILTER n Analytical Consideration
N/A (ALL) 63
LIANG BUA 3 Correlation between AVD and
ALL AMH* 59 retouch frequencies:
AMH UNCALIBRATED 59 • Modern human sample
AMH <25% RETOUCH 56 versus H. floresiensis
AMH + LIANG BUA 59
AMH + SONG TERUS 57 Predictive strength of the model:
AMH EXCLUDING POINT TECH. 52 • Impact of outliers (highly curated industries above
AMH* PLEISTOCENE 14 25% retouch)
AMH PLEISTOCENE 11 • Specialized point technology
AMH EARLY HOLOCENE 13
AMH MIDDLE HOLOCENE 20 Chronological Trends:
AMH LATE HOLOCENE 12 • Pleistocene versus Holocene
AMH TERMINAL PLEISTOCENE 16 • Holocene phases
AMH PLEISTOCENE + LIANG BUA 14 • Terminal Pleistocene
AMH PLEISTOCENE + SONG TERUS 12 (transitional)
relatively easy to code and “keep tabs on” without fundamentally altering the data or straying too far from original questions driving this research. These variables amounted to series of analytical considerations which were applied as data filters to test their effects on the output. In some instances, it merely warranted a closer look at outliers, but others were thought to be pertinent to interpreting the results. The full table of analytical considerations is presented above (Table 1) and the reasoning behind their incorporation is briefly discussed here.
The relatively small percentage of Holocene assemblages with overtly specialized elements, namely unifacial points, are atypical in SEA lithic assemblages. Unifacial points are not found in late Pleistocene assemblages, and their presence in the Holocene assemblage is minimal. The second, more concerning issue with including Holocene assemblages is that the chronological boundary between strictly forager and predominately agricultural adaptive strategies is not often starkly defined by depositional changes in material culture, and recent
70


review of the archaeological evidence for emerging “Neolithic” farmers of ISE Asia at ca 4,500 years ago were found to be limited and inconclusive (O’Connor, 2015). These Holocene assemblages have generally been interpreted under similar ambiguity regarding site function, settlement configuration, and the role of lithic economy during the transition to agricultural communities. This makes it difficult to address the spatiotemporal expansion of agriculture into SEA as prehistoric societies continue to utilize raw stone material, or at least manufacture stone tools using local methods comparable to early Holocene and Late Pleistocene hunter-gatherers. Given that our main concern is the relationship between mobility, assemblage patterning, and the relative position of modern versus archaic groups rather than subsistence, the mid-late Holocene assemblages should be emerging “Neolithic” farmers of ISE Asia at ca 4,500 years ago were found to be limited and inconclusive (O’Connor, 2015). These Holocene assemblages have generally been interpreted under similar ambiguity regarding site function, settlement configuration, and the role of lithic economy during the transition to agricultural communities. This makes it difficult to address the spatiotemporal expansion of agriculture into SEA as prehistoric societies continue to utilize raw stone material, or at least manufacture stone tools using local methods comparable to early Holocene and Late Pleistocene hunter-gatherers. Given that our main concern is the relationship between mobility, assemblage patterning, and the relative position of modern versus archaic groups rather than subsistence, the mid-late Holocene assemblages should be included to provide greater depth to the analysis that can be viewed, at least tentatively, as a technological system closely aligned with earlier periods, therefore representing another important source of data in illustrating the usefulness of the whole assemblage approach in understanding subtle variation or alternatively, continuity in the lithic technology for the region. Lastly, the assemblages of Liang Bua and Song Keplek
71


are evaluated against the broader regional data-set, both which are the strongest evidence for stone tool use in non-modem humans near the end of the Pleistocene. This is done to explicitly compare the assemblages of archaic to modem groups and evaluate the possibility of using lithic assemblages to differentiate hominin species in SE Asia.
Statistical interpretation.
The relationship between AVD and retouch frequency across these assemblages are tested using a linear regression analysis, which was coded in the statistical program Rstats. As mentioned above, the diversity of assemblages included in this analysis required that a series of computations were carried out, as to allow the influence of specific sets of assemblages to be independently tested. The linear regression analysis is interpreted through the strength of the R square, and R value, and the closer to -1 and 1 the value the stronger the relationship. These outputs are described in relation to a number of graphs plotting the distribution of the assemblages. Thus, the statistical analyses ran in this study are aimed at explaining the variability through a linear regression analysis, specifically by assessing the correlation between AVD and retouch frequency as described by the R value, and testing the predictability of these assemblages to conform along the negative linear regression by evaluating R square.
17


Table 2: Summary of sites and assemblages
SITE Level Period Beginning Ending Location
BUI CERI UATO 1 Early Holocene 8500 7500 East Timor
BUI CERI UATO II Early Holocene 7500 6500 East Timor
BUI CERI UATO III Mid Holocene 6500 5500 East Timor
BUI CERI UATO IV Mid Holocene 5500 4500 East Timor
BUI CERI UATO V Mid Holocene 4500 3500 East Timor
BUI CERI UATO VI Late Holocene 3500 2500 East Timor
BUI CERI UATO VII Late Holocene 2500 1500 East Timor
BUI CERI UATO VIII Late Holocene 1500 750 East Timor
BUI CERI UATO X-IX Late Holocene 750 0 East Timor
LIESIRI III Early Holocene 7600 7200 East Timor
LIESIRI IV Early Holocene 7300 7000 East Timor
LIESIRI V Early Holocene 7000 6700 East Timor
LIESIRI V(C) Mid Holocene 6000 5500 East Timor
LIESIRI VI Mid Holocene 4300 3700 East Timor
UAI BOBO 1 1 (a.b.) Early Holocene 9000 7500 East Timor
UAI BOBO 1 II (a.b.) Mid Holocene 6000 5000 East Timor
UAI BOBO 1 Ill (a.b.) Mid Holocene 4200 3500 East Timor
UAI BOBO 1 Ill (c.) Late Holocene 3200 2300 East Timor
UAI BOBO 1 IV (a.b.) Late Holocene 2500 2000 East Timor
UAI BOBO 1 V Late Holocene 2000 1600 East Timor
UAI BOBO 1 VI Late Holocene 1600 1200 East Timor
UAI BOBO 1 VII Late Holocene 1000 700 East Timor
UAI BOBO 1 VIII Late Holocene 800 500 East Timor
UAI BOBO 2 IV Early Holocene 8200 7400 East Timor
UAI BOBO 2 V Early Holocene 7000 6300 East Timor
UAI BOBO 2 VI Mid Holocene 6500 5900 East Timor
UAI BOBO 2 VII Mid Holocene 6200 5700 East Timor
UAI BOBO 2 VIII Mid Holocene 5200 5000 East Timor
UAI BOBO 2 IX Mid Holocene 4300 3800 East Timor
UAI BOBO 2 X Mid Holocene 3800 3200 East Timor
UAI BOBO 2 XII Late Holocene 1700 1400 East Timor
KRIA CAVE II Mid Holocene 5106 4367 New Guinea
KRIA CAVE III Mid Holocene 5567 5106 New Guinea
KRIA CAVE IV Mid Holocene 5936 5567 New Guinea
KRIA CAVE V Early Holocene 7136 5936 New Guinea
TOE CAVE 1 Pleistocene 26000 6000 New Guinea
TOE CAVE II Mid Holocene 6000 3000 New Guinea
ULU LEANG 1 Mid Holocene 4650 3470 Sulawesi
77


SITE Level Period Beginning Ending Location
ULU LEANG II Early Holocene 7500 5500 Sulawesi
ULU LEANG III Early Holocene 9450 7500 Sulawesi
LIANG LEMDUBU zone II Pleistocene 13910 9255 Aru Islands, IND
LIANG LEMDUBU zone III Pleistocene 17785 13910 Aru Islands, IND
LEANG SARRU 1 (ia) Holocene 10000 8000 Talaud Islands
LEANG SARRU l(ib) Pleistocene 21000 18000 Talaud Islands
LEANG SARRU III Pleistocene 35000 32000 Talaud Islands
PIA HUDALE 1,11,111 Pleistocene 11500 10000 Roti Island, IND
ILLE CAVE 334-807 Pleistocene 11000 9400 Philippines
SONG KEPLEK lb,II,III Mid Holocene 6000 4500 Java, IND
MOH KHIEW 1 Pleistocene 25800 25000 Thailand
MOH KHIEW III Pleistocene 11170 8330 Thailand
MOH KHIEW IV Early Holocene 7160 5520 Thailand
MOH KHIEW VI Mid Holocene 5520 4240 Thailand
LANG KAMNAN 1 Pleistocene 27000 16000 Thailand
LANG KAMNAN II Early Holocene 7500 2500 Thailand
LANG RONGRIEN X Pleistocene 43000 37265 Thailand
LANG RONGRIEN IX Pleistocene 37000 34380 Thailand
LANG RONGRIEN VIII Pleistocene 32180 27110 Thailand
LANG RONGRIEN V-VI Pleistocene 9655 7575 Thailand
LAANG SPEAN Hoabin. Early Holocene 11000 5000 Cambodia
SONGTERUS Terus Mid Holocene 341000 254000 Java, IND
LIANG BUA III Pleistocene 100000 95000 Flores, IND
LIANG BUA IV Pleistocene 74000 61000 Flores, IND
LIANG BUA V Pleistocene 55000 50000 Flores, IND
74


CHAPTER VI
RESULTS
The early indications from these statistical analyses give the impression that overall, assemblages across SE Asia conform to the expectations of the model outlined in the previous chapter. The series of graphs depicted below correspond with the results of the multiple data filters (sub-sets of data) which are presented below in Table 2 and are also described in more detail in the previous chapter. Due to the contrasting relationship between certain data filters, there are fewer graphs than there are statistical outputs. This arrangement allows smaller data-samples to be more easily viewed as a cross-section of the whole aggregate of assemblages. In other words, the graphs discussed below correspond with various data filters (pre-sorted assemblages), which have implications across multiple sub-sets of assemblage data depending on the analysis. Additionally, the assorted sub-data are presented against the backdrop of the entire data-set as to allow easy contrast to the wider sample.
The results presented here generally correspond with the four-major aims discussed in Chapter 1, and have helped influence the presentation of statistical representations relevant to those questions. However, while the results presented here are directed toward the specific aims of the proj ect, the results are further broken down in an effort to account for biases created while the data was collated. Thus, the arrangement of Figures and Tables in this chapter reflect an evaluation of the methodology and compatibility of specific data sets concurrent with the results from each new series.
To begin, the overall conformity of SE Asia lithic assemblages with the expected negative relationship between artifact density (AVD) and retouched tool frequencies is reported, which provides a benchmark for comparing assemblages across human species in


Table 3: Summary of Results
FILTER n r ? p value
N/A (ALL) 63 -0.603237 0.3638952 1.67E-07
LIANG BUA 3 0.2126667 0.0452271 0.8636
ALL AMH* 59 -0.757739 0.5741688 3.72E-12
AMH UNCALIBRATED 59 -0.68856 0.4741154 1.67E-09
AMH < 25 % RETOUCH 56 -0.659008 0.4342919 3.35E-08
AMH + LIANG BUA 59 -0.457436 0.2092473 0.0002701
AMH + SONG TERUS 57 -0.6846975 0.4688106 4.27E-09
AMH EXCLUDING POINT TECH. 52 -0.687994 0.4733353 3.38E-08
AMH* PLEISTOCENE 14 -0.856735 0.7339951 9.10E-05
AMH PLEISTOCENE 11 -0.811668 0.6588045 0.002416
AMH EARLY HOLOCENE 13 -0.76045 0.5782835 0.002547
AMH MIDDLE HOLOCENE 20 -0.371746 0.1381949 0.1065
AMH LATE HOLOCENE 12 -0.544328 0.2962932 0.06729
AMH TERMINAL PLEISTOCENE 16 -0.712973 0.5083307 0.001933
AMH PLEISTOCENE + LIANG BUA 14 -0.391338 0.1531456 0.1665
AMH PLEISTOCENE + SONG TERUS 12 -0.806685 0.6507407 0.001522
*AMH assemblages including those with >25% retouch freguencies (Lang Rongrien).
Pleistocene SE Asia. However, before proceeding with these analyses, specific sub-sets of assemblages are identified as potential outliers, or otherwise at odds with the overall assemblage patterning. Among those identified are assemblages with specialized point technology (SPT), ostensibly high frequencies of retouch, and the potentially semi-sedentary Holocene assemblages of SE Asia. After scrutinizing these sub-sets of data, the most representative sample of Pleistocene era forager strategies, unsurprisingly, comes from exclusively Pleistocene assemblages. However, in spite of the various effects of the many data configurations listed in Table 3, we think the greater sample is illustrative of a common technological strategy employed by AMH prehistoric populations of SE Asia. Therefore, more than one data series is emphasized to contrast assemblage pattering across certain non-modern assemblages.
76


100
o
o O
10
o
o
T3
D
o
+-*
<1)
cc.
° <9 c
So oo o
o o ooo -
* ° o og
o O o ooo
A non-modern O modern
0.1-------------------------------------------------------------------------------
0.1 1 10 100 1000 10000 Artifact Volumetric Density (m3)
Figure 2: assemblage patterning across all sites
General findings.
The results of the analysis demonstrated a moderately high correlation between AVD and retouch frequency across all assemblages (r = - .61). In contrast, when only those lithic assemblages created by modem humans were incorporated the data set showed a stronger, fairly high correlation between AVD and retouch frequency (r = - .75). Directly above in Figure 2 is a representation of the general variability along a negative regression. Among the modern assemblages, the data-set conformed well with the expectations of the model (r2 = .57), although with some noticeable outliers in the upper left comer and bottom left center of Figure 1. By comparison, the Liang Bua assemblages, which are part of a slightly larger group of assemblages considered non-modern in this analysis, produced a weak positive linear relationship which deviates significantly from the predicted negative relationship outlined in this model [Figure 3 ( 77


of Lang Rongrien, Thailand, particularly the three assemblages with retouch frequencies > 25% (Figure 5), as well as the assemblages with SPT [ Figure 4 (n =6)] from ISE Asia. These SPT assemblages from the sites of Uai Bobo 1, East Timor, and Ulu Leang, Sulawesi, were also noted for potentially distorting the Holocene series. Adherence to this model is most apparent in the Pleistocene and Early Holocene phases, which exhibits the strongest predictive statistical relationship and steadily declines across the Middle and Late Holocene data sets.
Modern human assemblage patterning.
The results from this study indicate that the directionality of the linear regression is correlates well with a negative relationship between AVD and retouch frequency across all AMH data sets except for the Middle Holocene [Table 3 (r = -.37)], excluding those series which incorporated the assemblages from Liang Bua (see Table 2 for multiple r values associated with Liang Bua). However, as alluded to above, there are a number of assemblages within the modern human sample which stand out as potential outliers driving the strength of the overall assemblage patterning.
The first consideration were the assemblages in the upper left comer [Lang Rongrien assemblages with retouch (Figure 3). These data were removed to test their influence on the overall statistical output. Indeed, the predictive strength in modem assemblages was noticeably impacted, with the predictability of variability in the negative regression dropping sharply after removing these assemblages from the analysis [(r2 = .43), which is even lower than uncalibrated output (r2 =.47)]. These results indicate that overall, the moderately high
7R


IQ
O
o
3
o
-*-<
0)
QC
o o ° o ° ° o °° ° o
, o 0 o o o o <
o o o o
0.1 1 10 100 1000 10000
Artifact Volumetric Density (m3)
O modern â–² Liang Bua
Figure 3: Assemblage patterning at Liang Bua
predictive strength between retouch frequency and AVD in modern assemblages (r2 = .57) is not equally represented across the data set.
The subsequent analyses are represented by a new baseline which excludes the outlier assemblages (retouch > 25 %) from the site of Lang Rongrien. This core data sample is used in a number of this is subsequent data series in order to compare modern human assemblage patterning across temporal, species, and technological boundaries without the possibility of outliers driving the negative relationship (Table 3).
Notwithstanding the impact of anomalous retouch frequencies on the statistical strength in predicting retouch frequency and AVD, the remaining modem assemblages exhibit a moderate to low predictive strength [r2 = .43 (n = 56)], which broadly defines the core patterning unique to SE Asia modem assemblages. Furthermore, the core patterning predicted by the model (described with r2) is shown to hold across nearly all of the remaining data filters, with the notable exceptions being the inclusion of the Liang Bua assemblages, as well as the
7Q


Mid—Late Holocene assemblages. For example, when the Liang Bua assemblages were refigured into the new baseline sample [Figure 3 (n = 59)] the predictive strength of the linear regression was again steeply dropped (r2= .20). Directly above in Figure 3, the assemblages from Liang Bua can clearly be identified as extreme outliers in relation to the larger data sample.
Specialized point technology.
While the Pleistocene assemblages created by H. floresiensis were predicted to vary from those of AMH, other filters exposed the impact of aggregated data within the AMH sample which were unanticipated. For example, the sub-grouping of assemblages with overt specialization in point technology carried a slightly higher retouch frequency in relation to AVD, although still adhering strongly with the expected negative linear relationship. As such, these data were also considered for potentially biasing some of the smaller data sets, particularly the Middle and Late Holocene series. Figure 4 (above) identifies the slightly elevated retouch frequencies within this sub-set, drawing attention to the possible impact of this data concentration on other assemblage assemblages. In particular, the Holocene sample could be affected more acutely because all assemblages with specialized tools came from this period. Overall, after removing the assemblages from the larger grouping, the predictive strength of the pattern returned with a moderately high outcome (r2= .58), indicating that SPT assemblages may hover above the core pattern, slightly decentered from the expected negative linear trend (see trend line on Figure 4) seen among the larger collection of modem human assemblages seen across SE Asia. Additionally, when the assemblages with specialized elements were isolated and tested independently, these demonstrated the strongest predictive


100
A
Uai Bobo 1
10
o
o
TO
<1)
UB1 9 \
UB1 ^ UB1
13
O
modern A non-modern • point tech.
0.1
0.1 1 10 100 1000 10000 Artifact Volmetric Density (m3)
Figure 4: Assemblage patterning in specialized point technology
strength among the sample [r2 = .92 (n = 6)], although, this series is the smallest of any aggregate sample tested.
These findings are not altogether unexpected since culturally imposed forms are virtually non-existent across SE Asia, or at least they are not evident to many archaeologists, and in a region nearly devoid of retouched tools, occurrences of these specialized forms may be expected to slightly impact the core patterning across SE Asia, which is the case here when the assemblages with unifacial and Maros points from Ulu Leang, and Uai-Bobo are included (,r2 = .43) versus when they are absent (r2 =.57). The apparent impact on the negative patterning in the wider sample is likely magnified in the smaller data sets from the Middle and Late Holocene series.
R1


A non-modern O modern + Outliers
0.1 1 10 100 1000 10000 Artifact Volumetric Density (m3)
Figure 5: Assemblage patterning without Outliers
Chronological changes.
The assemblages were further tested in data series which were bound by chronological periods. Of particular importance were the modem human assemblages bracketed by the Late Pleistocene phase, and how they compared to the modern human assemblages which occurred during the Holocene phases. The Pleistocene assemblages which formed at the terminal Pleistocene and prior to the Last Glacial Maximum (LGM) were found to be the most congruent sample with respect to fitting the expectations of the model (r2 = .73). Notably, the Pleistocene assemblage patterning also held even when controlling for the effect of outliers, specifically the assemblages of Lang Rongrien [(/'2= 65) Figure 5], By comparison, the Holocene assemblages did not conform well to the expected negative relationship between AVD and retouch frequencies. It should be reviewed, though, that this model is designed for true hunter-gatherer populations, and, as there is some uncertainty whether the mid-Holocene
L ffl * A O ° ° o °° ° <9 ~ O S°o00o° ooo° * ° ° °>o V A 0 o O 000
° o o°
R?


and Late Holocene assemblage localities were truly adapted around hunter-gathering. The Middle and Late Holocene assemblages were to be treated with caution until more is known about lithic organization and settlement strategies in “Neolithic” Southeast communities. Nevertheless, it was our contention that some measure of logistical mobility which characterizes Pleistocene forager lifeways would have been evident in assemblage composition under the pretense that settlement strategies during the Middle—Late Holocene of SE Asia. The results discussed above diminish our confidence in the compatibility of the majority of the Holocene assemblages.
Conversely, the weak relationship exhibited across Holocene assemblages indirectly strengthen the findings and suitability of this methodology in Pleistocene archaeological contexts. It is worth noting that the model used here (WABI) is an interpretive framework designed to draw inferences on landscape-use and adaptive strategies in Pleistocene era hunter-gatherers. In retrospect, the expectation that SE Asia Holocene survival strategies were fundamentally similar to Pleistocene foragers has little to do with physical evidence, but rather the lack thereof and hence the inability to differentiate these populations from former inhabitants. While the low predictive strength found across the majority of Holocene sites certainly calls into question their inclusion within the wider sample, the potential biasing effects of the SPT may need to be reconsidered in light of the fact that all six SPT assemblages derive from Holocene contexts. Despite the possibility that SPT assemblages are pulling the overall Holocene r2 value down because of their clustering in the upper right of the core patterning (Figure 3), any further consideration of the Middle and Late Holocene sample would be difficult to manage and is beyond the scope of this research. For the time being, the Middle and Late Holocene samples are not fully discounted as outliers and are kept in the larger AMH


SE Asia data sample, but these assemblages are excluded from the subsequent series, which most convincingly captures assemblage patterning which formed under Pleistocene adaptive configurations, or otherwise strictly hunting and gathering mobility strategies.
In consideration of the possible incompatibility of Middle and Late Holocene adaptive systems with this model, the Early Holocene assemblages were combined with terminal Pleistocene assemblages (>12 ka), and constitute a transitory phase which is much more apt for this model, and incidentally demonstrated a much stronger predictive relationship between AVD and retouch frequency [(r2 = .50)]. Notably, when non-modern assemblages are reconfigured into this the Pleistocene data series, the sites of Liang Bua and Song Terns have differing effects on the core patterning and are discussed briefly below.
Non-modern assemblage patterning.
The site localities with assemblages created by relict hominins include Liang Bua (n=3) and Song Terns (n=l), the alledgedly non-modern assemblages on the curated/expedient continuum, are contrasted against the core patterning exhibited in Pleistocene and tranistionary Holocene series. After the Liang Bua assemblages are incorporated into the Pleistocene series, it is clear that these assemblages fall outside of the models expected neagitve linear regression (r2 = . 153). In contrast, the site of Song Terns, when calibrated, fits more neatly with the overall relationship predicted by the model (r2 =. 65) . While only one assemblage, the assemblage from Song Terns is one of only a few assemblages that have been recently dated the Middle Pleistocene (ca. 265 ka - 170 ka) and is believed to have been left by a relic “evolved” H. erectus population. Although the assemblage from Song Terns is dislocated from the core patterning in the uncalibrated sample, its minimal impact on the overall predictive strength in the calibrated patterning (r2 = .47) positions it more closely aligned with the modern
R4


assemblage series range than are any of the assemblages from Liang Bua, even after they are corrected for duration.


CHAPTER VII
CONCLUSIONS
The preliminary results from these analyses suggest that the lithic assemblages across SE Asia can be compatible with middle range approaches to prehistoric behavior, which often emphasize the linkages which exist between the technological interface, the distribution of resources, and aspects of landscape-use, subsistence strategies, and economy. Although hunter-gatherer behavior in prehistoric H. sapiens of SE Asia is complicated by the little knowledge available to approximate adaptive strategies in this region, early indications from this research support the notion that lithic organization was structured around landscape-use and economic decisions in Pleistocene foragers of SE Asia (Figure 6).
Ideally, this model works best when it is able to explore these differing mobility strategies, or land-scape use in relation to specific ecological variables such as the distribution or material resources, or in relation the subsistence-based activities. However, the findings from this thesis are not quite polished enough to comprehensively define the cultural structure of AMH, or speculate on the shaping forces behind regional and ecologically specific organizational qualities of SE Asia survival strategies. Nevertheless, these results hit on higher methodological and theoretical levels of archaeological inquiry, while also setting up a framework to re-analyze and incorporate new data into a working hypothesis of hunter-gatherer adaptation in SE Asia.
In review, there is a widely-held consensus among researchers working toward a regional perspective of lithic technology in SE Asia that stone tools may not be a behavioral proxy for differentiating between modern and archaics. In particular, the conventional typological divisions (i.e. core/flake dichotomy) have proven to be inadequate in addressing


Figure 6: Assemblage patterning in SE Asia
aspects of technological development and hominin behavior (e.g. Moore and Brumm, 2007; Marwick, 2008; O’Connor, 2007; Reynolds, 2007). The informal character of lithic industries is argued to transcend species boundaries; thus, the remarkable regional continuity in stone tool traditions makes them an unreliable behavioral proxy for differentiating between hominin species. The findings from this research suggest that it is at least possible to get at modern human organizational strategies, and these data sets may be used to define differences in technological behavior when better defined occupation sequences of hominins of SE Asia are available.
Southeast Asian Technological Organization
Making sense of the why and where the assemblages of Pleistocene SE Asia would cluster along the curated-expedient continuum is somewhat paradoxical. On the one hand, we know that the expedient manufacture of stone tools is often tied into logistically oriented mobility strategies, which centers around larger foraging radii, involves longer occupations,
R7


and where resources flow back to camp. On the other hand, the distribution of resources in lower latitudes, or environments with more evenly distributed resources like those of the Pleistocene SE Asia, are more commonly responded to with residentially oriented mobility strategies. In consideration of these expectations, the expedient character of stone tools in SE Asia are somewhat in conflict with the ecological framework of tropical and pre-humid climates, which would otherwise predict prehistoric foragers were highly mobile and practicing residentially settlement strategies, which in theory would lead to stronger assemblage patterning indicative of highly curated industries.
Of course, even as the paleoenvironmental conditions remain vague, it is clear that in order to characterize forager organization in this region the conventional parameters along which mobility is defined in this model will need realignment to account for the unique demands encountered by Pleistocene SE Asian foragers. Firstly, in a cross-cultural study of hunter-gatherer mobility strategies Kelly (1982) noticed several incongruences with a number of forager mobility strategies across tropical environments, but noted there where logistical mobility strategies dominated in tropical conditions the forager communities were reliant on aquatic resources. Studies on Pleistocene era foragers of SE Asia have increasingly considered the role of littoral resources in shaping the survival strategies and technological behavior here. Additionally, however, there are tropical-specific ecological demands which result in resource stress such as the lack of carbohydrates and low density of secondary biomass, which impose important caloric deficits on forager populations subsisting on predominately tropical available foods. Moreover, the notion of environment constraints on lithic technology continues to garner popularity as a working hypothesis— the fact that the arrival of modern humans did little to improve upon the preexisting lithic technologies hints at the high probability that non-


lithic resources played an important role in prehistoric technology here. The archaeological signatures of SE Asia are also convincingly expedient, often put down to the lack of availability to fine-grained lithic raw material resources. At present, conceptualizing the ecological pressures encountered by Pleistocene foragers in SE Asia is a topic still facing numerous uncertainties., among the most debated topics in Pleistocene SE Asia is the issue of prehistoric climatic conditions— of particular concern is the timing of tropical conditions, and the compositional changes to forest and the configuration of coastlines throughout the Upper and Late Pleistocene. Moreover, scholars have expressed concern over the impact of tropical ecology constraints on hunter-gatherer organization, and questioned the extent true tropical adaptation is possible (e.g. Bailey et al., 1989).
With these caveats in mind, and what is currently known and observable about these industries (low-input, expediently produced) in contrast to what remains largely unknown about Pleistocene environments across SE Asia, it seems more than reasonable to assume assemblage patterning would cluster near the logistical axis of forager strategies along the expedient/curated continuum (Figure 6). Nevertheless, when the density and rate by which artifacts accumulated are factored in, it appears that there is, in fact, much more variability in technological organization than previously assumed as well as what was predicted before our analyses (Figure 6). Although episodic, retouching stone tools in Pleistocene SE Asia forager strategies may have had a similar role in economizing lithic resources in relation to broader mobility across Pleistocene SE Asia
There is still much to be learned with respect to foraging adaptation and economic strategies in Pleistocene SE Asia. For example, what type of mobility existed in Pleistocene and whether these hunter-gatherers were strictly foraging within the interior forest or,
R9


O modern ♦ Bui Ceri Uato AUai Bobo 2
1 10 100 1000 10000
Artifact Volmetric Density (m3)
Figure 7: Assemblage variation in East Timor sites
alternatively, broad spectrum foragers exploiting auxiliary habitats because of the deficiencies or challenges of resource extractions in tropical habitat has yet to have been addressed. However, scholars have recently noted important accumulation differences in stone tools, occupational intensity, in one case between of Bui Ceri Uato and Uai Bobo 2, and questioned whether the differences in occupational intensity might be embedded in a larger subsistence/settlement strategy (Veth et al., 2005). While developing a robustly defined human adaptation strategy in SE Asia is out of reach, the later topic is well within the purview of this model (Figure 7).
Hominin assemblages.
Figure 8 displays the predicted negative relationship between lithic density and retouch frequency, and the pattern documented at Liang Bua. As was reported in Results, Liang Bua assemblages do not conform to the expected relationship, showing instead a statistically insignificant, weak positive relationship. Thus, these results would seem to indicate that the
QO


ss
o
o
o o ° o ° ° o °° ° <9 n O <£0 00 0 O O o ooo * ° » °° °>oV>
, o 0 o o o o â– 4
o o o o
0.1 1 10 100 1000 10000
Artifact Volumetric Density (m3)
O modern â–² Liang Bua
Figure 8: Assemblage variation in non-modern species
toolmakers who occupied Liang Bua did not organize their mobility and lithic technology similarly to contemporary Homo sapiens and Neanderthal populations. This pattern could be interpreted as a taxonomically significant behavioral signal indicating a ‘non-modern’ hominin population, perhaps one descended from an isolated Lower Pleistocene archaic population. However, it was also possible that it more simply reflects how tropical foragers organize this technology. This is because all prior applications of the method used here have employed datasets drawn from temperate contexts. The datasets from prehistoric hunter-gatherer populations from various sites in SE Asia were analyzed to test this possibility. If the ‘tropical forager’ interpretation is correct, the expectation was that these other assemblages attributed to modem humans would not have shown the negative relationship predicted by the whole assemblage behavioral indicator.
The pattern that emerged is that SE Asian tropical foragers appear to have tailored their
Q1


lithic technological organization to their mobility strategies in ways very similar to Late Pleistocene foragers from Western Eurasia. This suggests that the methodology used to reconstruct Paleolithic hunter-gatherer mobility in temperate conditions is also applicable to hunter-gatherers in tropical environments. Further, it provides a new tool to potentially distinguish the assemblages from modern populations from those of archaic populations. Overall, these data indicate that the lithic assemblages from Liang Bua are not organized like those of contemporary foragers, and that this difference is not the result of ecological conditions. One potential interpretation is that this reflects a fundamental behavioral difference of the Liang Bua hominins relative to contemporary human populations. However, recent redating of the Liang Bua sequences requires a reanalysis of the stone tool assemblages; specifically whether the stone tool assemblages are constrained by the same geo-chronological layers. Similarly, the assemblage from Song Terns requires additional sequences to test for organizational changes over time; and the possible fluvial deposit of this assemblage is methodologically problematic.
However, major preliminary conclusion of this study is that a growing number of lithic assemblages from tropical hunter gatherers appear to capture the same adaptive signal as that of Late Pleistocene Homo sapiens and Neanderthal lithic assemblages from Western Eurasia. The manifestations of these discard patterns in the archaeological record reflect mobility strategies that are apparently not manifested at Liang Bua, and they therefore suggest a fundamental difference in lithic technological organization and mobility. This has behavioral implications that reach beyond techno-typological classifications and may be significant in the context of the debate over the taxonomy of the stone tool manufacturers at Liang Bua and Song Terns once better sequences can be defined.
Q?


Although the SE Asian site selection is fairly limited due to the requirement of a majority of Holocene content, the reported trend is very strong and the developing pattern should encourage researchers to pull together a more comprehensive analysis of lithic assemblages to corroborate the predictions of mobility patterns in tropical ecologies. A robust data set pertaining to lithic densities and frequency of modification in SE Asia can provide a new understanding of tropical mobility, as well as potentially shed light on whether the behavior of the Liang Bua hominins was comparable to those of coeval foragers.


REFERENCES:
Ambrose, Stanley H.
2001 Paleolithic Technology and Human Evolution. Science 291(5509): 1748—1753. Anderson, Douglas D.
1990 Lang Rongrien Rockshelter: A Pleistocene, Early Holocene Archaeological Site from Krabi, Southwestern Thailand. University Museum Monograph 71. The University Museum, University of Pennsylvania, Philidelphia.
1997 Cave archaeology in Southeast Asia. Geoarchaeology: An International Journal 12: 607—638.
Andrefsky, William, Jr.
1994 Raw Material Availability and the Organization of Technology. Society for American Archaeology 59(1): 21—34.
2006 The Analysis of Stone Tool Procurement, Production, and Maintenance. Journal of Archaeological Science 17: 65—103.
Anthony, David W.
1990 Migration in Archaeology: The Baby and the Bathwater. American Anthropologist 92(4): 896—914.
Bahuchet S., D. Mckey, and Igor de Garine
Wild yams revisited: Is independence from agriculture possible for rain forest hunter-gatherers? Human Ecology 19(2): 213—243.
Bailey, Robert C. and Thomas N. Headland
1991 The tropical rain forest: Is it a productive environment for human foragers? Human Ecology 19(2): 261—285.
Bailey, Robert C., Genevieve Head, Bruce Owen, Robert Rechtman, and Elzbieta Zechenter 1989 Hunting and Gathering in Tropical Rain Forest: Is It Possible? American Anthropologist 91(1): 59—82.
Bamforth, Douglas B.
1986 Technological Efficiency and Tool Curation. American Antiquity 51: 38—50.
Bar-Yosef, Ofer
2002 The Upper Paleolithic Revolution. Annual Review of Anthropology 31: 363—393. Bar-Yosef, Ofer and Philip Van Peer
2009 The Chaine Operatoire in Middle Paleolithic Archaeology. Current Anthropology 5(1):103—131.
94


Barton, Huw, Graeme Barker, David Gilbertson, Chris Hunt, Lisa Kealhoper, Helen Lewis, Victor Paz, Philip J. Piper, Ryan J. Rabett, Tim Reynolds, and Katherine Szabo.
2013 Late Pleistocene Foragers, c. 35,000—11,500 Years Ago. In Rainforest foraging and farming in Island Southeast Asia, Vol. 1, edited by Graeme Barker, pp. 173—216.. McDonald Institute for Archaeological Research, Cambridge.
Bartstra, Gert-Jan
1982 Homo Erectus Erectus: The Search for His Arifacts. Current Anthropology 23(3): 318—320.
Bartstra, G. J. and Basoeki
1989 Recent Work on the Pleistocene and the Paleolithic of Java. Current Anthropology 30(2): 241—241.
Bartstra, G.J., D.A. Hooijer, B. Kallupa, and M. Anwar Akib
1994 Notes on fossil vertebrates and stone tools from Sulawesi, Indonesia, and the stratigraphy of the Northern Walanae Depression. Palaeohistoria 33(34): 1—18.
Bay-Petersen, Jan
1983 Excavations at Bagumbayan, Masbate, Central Philippines: An Economic Analysis. Asian Perspectives 25(2): 67—98.
Bettis III, E. Arthur, Adrianne K. Milius, Scott J. Carpenter, Roy Larick, Yahdi Zaim, Yan Rizal, Rusell L. Ciochon, Stephanie A. Tassier-Surine, Daniel Murray, Suminto, Sutinko Bronto
2009 Way out of Africa: Early Pleistocene paleoenvironments inhabited by Homo erectus in Sangiran, Java. Journal of Human Evolution 56(1): 11—24/
Binford, Lewis R.
1979 Organization and Formation Processes: Looking at Curated Technologies. Journal of Anthropological Research 35(3): 255—273.
1980 Willow Smoke and Dog's Tails: Hunter-Gatherer Settlement System and Archaeological Site Formation. American Antiquity 45: 4—20.
Bleed, Peter
2001 Trees or Chains, Links or Branches: Conceptual Alternatives for Consideration of Stone Tool Production and Other Sequential Activities. Journal of Archaeological Method and Theory 8(1): 101—127.
Brantingham, Jeffery P.
2003 A Neutral Model of Stone Raw Material Procurement. American Antiquity 68(3): 487—509.


Brumm, Adam, F. Aziz, G.D. van derBergh, M.J. Mowood, D.R. Hobbs, R. Fullagar
2006 Early stone technology at Flores and its implications for Homo floresiensis. Journal of Human Evolution 57: 503—526.
Brumm, Adam, Mark W. Moore, Gert van den Bergh, Iwan Kumiawan, and Michael J. Morwood
2010 Stone technology at the Middle Pleistocene site of Mata Menge, Flores, Indonesia. Journal of Archaeological Science 37: 451—473.
Brumm, Adam, and Mark W. Moore
2012 Biface distributions and the Movius Line: A Southeast Asian perspective. Australian Archaeology 32—46.
Brosius, Peter
1991 Foraging in Tropical Rain Forests: The Case of the Penan of Sarawak, East Malaysia (Borneo). Human Ecology 19(2): 123—150.
Bulbeck, David F.
2003 Hunter-Gatherer Occupation of the Malay Peninsula from the Ice Age to the Iron Age. In Under the Canopy. The Archaeology of Tropical Rain Forests, edited by Julio Mercader, pp. 119—160. Rutgers University Press, New Brunswick.
Choi, Kildo and Dubel Driwantoro
2007 Shell tool use by early members of Homo erectus in Sangiran central Java: cut mark evidence. Journal of Archaeological Science 34(1): 48—58.
Colinvaux, Paul A. and Mark B. Bush
1991 The Rain-Forest Ecosystem as a Resource for Hunting and Gathering. American Anthropologist 93(1): 153—160.
Corvinus, Gudrun
2004 Homo erectus in East and Southeast Asia, and the question of the age of the species and its association with artifacts, with special attention to handaxe-like tools. Quaternary International 117(1): 141—151.
Cowan, Frank L.
1999 Making Sense of Flake Scatters: Lithic Technological Strategies and Mobility. American Antiquity 64(4): 593—607.
Detroit, Florent, Eusebio Dizon, Christophe Falgueres, Sebastien Hameau, Wilfredo Ronquillo, and Francois Semah
2004 Upper Pleistocene Homo sapiens from the Tabon cave (Palwan, The Philippines): description and dating of new discoveries. Comptes Rendus Palevol 3(8): 705—712.
96


Dibble, Harold J.
1995 Middle Paleolithic Scraper Reduction: Background, Clarification, and Review of the Evidence to Date. Journal of Archaeological Method and Theory 2(4): 299—368.
Dizon, Eusebio Z.
2010 The lower Paleolithic record in the Philippines. Quaternary International 1040(6182): 444—450.
Dwyer, Peter D. and Monica Minnegal
1991 Hunting in Lowland, Tropical Rain Forest: Towards a Model of Non-Agricultural Subsistence. Human Ecology 19(2): 187—212.
Falgueres C., F. Semah, H. Saleki, S. Hameau, H. Tu, G. Feraud, H. Simanjuntak, H. Widianto
2015 Geochronology of early human settlements in Java: What is at stake? Quaternary International xxx: 1—7.
Foley, Robert A.
1987 Hominid species and stone tool assemblages: how are they related? Antiquity 61(233):
380—392.
1997 Mode 3 technologies and the evolution of modern humans.Cambridge Archaeological Journal 7(1): 3—36.
Foley, Robert and Marta M. Lahr,
2003 On Stony Ground: Lithic Technology, Human Evolution, and the Emergence of Culturq. Evolutionary Anthropology 12(3): 109—122.
Forestier, Hubert
1999 L'assemblage industriel de Song Keplek, Java Est [Un nouveau regard sur l'outillage lithique del'homme modeme au debut de l'Holocene en Indonesie. Bulletin de I'Ecole franqaise d'Extreme-Orient 86: 129—159.
2007 Ribuan Gunung, Ribuan Alat Batu Prasejarah Song Keplek Gunung Sew a, Jawa Timur. Kepustakaan Populer Gramedia, Jakarata.
Forestier, Hubert and Elise Patole-Edoumba
2000 Les industries lithiques du Paleolithique tardif et du debut de l'Holocene en Insulinde. Aseanie 6: 13—56.
Forestier, H., Simanjuntak, T., Detroit, and V. Zeitoun
2010 Finite et diversite prehistorique entre Java et Sumatra. Archipel 80: 19—80.
97


Forestier, Hubert, Heng Sophady, Simon Puaud, Vicenzo Celiberti, Stephane Frere, Valery
Zeitoun, Cecile Mourer-Chauvire, Roland Mourre, Heng Than, Laurence Billault
2015 The Hoabinhian from Laang Spean Cave in its stratigraphic, chronological, typo-technological and environmental context (Cambodia, Battambang province/ Journal of Archaeological Science: Reports 3: 194—206.
Fox, Robert B.
1970 The Tabon Caves: Archaeological Explorations and Excavations on Palawan Island, Phillipines. Monograph of the National Museum No. 1. National Museum,
Manila.
1987 The Philippine Paleolithic. In Early Paleolithic in South and East Asia, edited by Fumiko Ikawa Smith, pp. 59—86. Mouton Publishers, Paris.
Glover, Ian C
1977 Late Stone Age in Eastern Indonesia. World Archaeology 9(1): 42—61.
1986 Archaeology in East Timor 1966-67. Terra Australis 11. Australian National University, Canberra.
Grove, Matt
2009 Hunter-gatherer movement patterns: Causes and constraints. Journal of Anthropological Archaeology 28: 222—233.
Headland, Thomas N.
1987 The wild yam question: How well could independent hunter-gatherers live in a tropical rainforest environment? Human Ecology 15(4): 463—491.
Headland, Thomas N. and Robert C. Bailey
1991 Introduction: Have Hunter-Gatherers Ever Lived in Tropical Rain Forest Independently of Agriculture? Human Ecology 19(2): 115—122.
Hiscock, Peter
2005 Artefacts on Aru: Evaluating the Technological Sequence. In The Archaeology of the Am Islands, Eastern Indonesia, terra australias 22, edited by S. O'Connor, M. Spriggs, and P. Veth, pp. 205—234. Australian National University Press, Canberra.
Hutterer, K.L.
1977 Reinterpreting the Southeast Asian Paleolithic. In Sunda and Sahul: Prehistoric Studies.in Southeast Asia, Melanesia and Australia, edited by J. Allen, J. Golson, and R. Jones, pp. 31—71. Academic Press, New York.
Jacob, T., R.P. Soejono, L.G. Freeman, and F.H. Brown
1978 Stone tools from Mid-Pleistocene sediments in Java. Science 202: 885—887
9R


Joordens, Josephine C.A., Francesco d'Errico, Frank P. Wesselingh, Stephen Muhro, John de
Vos, and Jakob Wallinga
2015 Homo erectus at Trinil on Java used shells for tool production and engraving. Nature 518(7538): 228.
Kelly, Robert
1983 Hunter-Gatherer Mobility Strategies. Journal of Anthropological Research 39(3): 277—306.
Lycett, Stephen J., and Christopher J. Bae
2010 The Movius Line controversy: the state of the debate. World Archaeology 42(4): 521—544.
Marwick, Ben
2009 Biogeography of Middle Pleistocene hominins in mainland Southeast Asia: A
review of current literature. Quaternary International 202: 51—58.
Mijares, Armand S. B.
2002 The Minori Cave Expedient Lithic Technology. Quezon City, University of the Philippines Press.
2004 Lithic analysis of recently excavated Tabon Cave flakes. Proceedings of the Society of Philippine Archaeologists 2: 15—19.
2005 The Archaeology of Penablanca Cave Sites, Northern Luzon, Philippines. Journal of Austronesian Studies 1(2): 65—93.
2008 The Penablanca Flake Tools: An Unchanging Technology? Hukay 12: 13—34.
Mijares, Armand S., Florent Detroit, Philip Piper, Rainer Grim, Peter Bellwood, Maxime
Aubert, Guillaume Champion, Nida Cuevas, Alexandra De Leon, and Eusebio Dizon
2010 New Evidence for a 67,000-year-old human presence at Callao Cave, Luzon, Philippines. Journal of Human Evolution 59: 123—132.
Moore, Mark W., and Adam Brumm
2007 Stone artifacts and hominins in island Southeast Asia: New insights from Flores, eastern Indonesia. Journal of Human Evolution 52: 85—102.
2009 Homo floresiensis and the African Oldowan. In Interdisciplinary Approaches to the Oldowan, edited by E. Hovers and D.R. Braun, pp. 61—69. Springer, New York.
Moore, Mark W., and T. Jatmiko Sutikna, M. J. Morwood, and A. Brumm
2009 Continuities in stone flaking technique at Liang Bua, Flores. Journal of Human Evolution 57: 503—526.
99


Morgan, Christopher
2009 Climate change, uncertainty and prehistoric hunter-gatherer mobility. Journal of Anthropological Archaeology 28: 382—396.
Morwood, M.J., P.B. O'Sullivan, F. Aziz, and A. Raza
1998 Fisson-track ages of stone tools and fossils on the east Indonesian island of Flores. Nature 392: 173—176.
Mudar, Karen and Douglas Anderson
2007 New Evidence for Southeast Asian Pleistocene Foraging Economies: Faunal Evidence From the Early Levels of Lang Rongrien Rockshelter, Krabi, Thailand. Asian Perspectives 46(2): 298—334.
Nelson, Margret C.
1991 The Study of Technological Organization. Archaeological Method and Theory 3: 57—100.
O'Connor, Sue
2002 Pleistocene Timor: Further Corrections, a Reply to Bednarik Australian Archaeology 54: 46—51.
2006 New Evidence from East Timor contributes to our understanding of earliest modem human colonization east of the Sunda Shelf. Antiquity 81: 523—535.
O'Connor, Sue, and Peter Veth
2005 Early Holocene shell fish hooks from Lene Hara Cave, East Timor establish complex fishing technology was n use in Island South East Asia five thousand years before Austronesian settlement. Antiquity 79: 249—256.
O'Connor, Sue, Ono, Rinataro, and Clarkson, Chris
2011 Pelagic Fishing at 42,000 Before the Present and the Maritime skills of Modem
Humans. Science 334(6059): 1117—1121.
O'Connor, Sue, and Matthew Spriggs, and Peter Veth
2002a Excavation at Lene Hara Cave establishes occupation in East Timor at least 30,000—35,000 years ago. Antiquity 76: 45—50.
2002b Direct Dating of shell beads from Lene Hara Cave, East Timor. Australian Archaeology 55: 18—21.
O'Connor, Sue, Ken Aplin, Katherine Szabo, Juliette Pasveer, Peter Veth, and Matthew Spriggs
2007 Liang Lemdubu: A Pleistocene Cave Site in the Am Islands. In The Archaeology of the Aru Islands, Eastern Indonesia, Terra Australis 22, edited by S. O'Connor, M. Spriggs, and P. Veth, pp. 171—204. Australian National University Press, Canberra.
100


Full Text

PAGE 1

LITHIC TECHNOLOGY AND MOBILITY IN LATE PLEISTOCENE SOUTHEAST ASIA: A WHOLE ASSEMBLAGE APPROACH TO ASSESSING TECHNOLOGICAL BEHAVIOR IN MODERN AND ARCHAIC HUMAN POPULATIONS b y DANIEL RYAN CICERO B.A., University of Colorado, Denver , 2013 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfillment of the requirements for the degree of Master of Arts Anthropology Program 2017

PAGE 2

ii ! 2017 DANIEL RYAN CICERO ALL RIGHTS RESERVED

PAGE 3

iii This thesis for the Master Arts Anthropology degree by Daniel Ryan Cicero Has been approved for the Anthropology Department by Tammy Stone, Chair Julien Riel Salvatore Charles Musiba Date: May 13th, 2017

PAGE 4

iv Ci cer o, Da niel Ryan (M.A. Anthropology) Lithic Technology and Mobility in Late Pleistocene Southeast Asia: A Whole Assemblage Approach to Assessing Technological Behavior in Modern and Archaic Human Populations Thesis directed by Associate Professor Tammy St one ABSTRACT Assessing huma n behavior during the Upper Pleistocene of Southeast Asia from a technological standpoint has long been viewed as problematic due to the indistinctiveness of the stone tool artifacts. Often expe diently produced from poor quality s tone and lacking standardization in tool form, these Ôlow input' industries have been taken as evidence that the lithic economies of Pleistocene hunter gatherers here were likely deemphasized in favor of locally abundant organic materials. While the unique prehistor ic ecological constraints appear to be the primary limiting factor in the production of complex lithic industries, the extent to which lithic economies of Late Pleistocene foragers in Southeast Asia were truly liberated from the type of embedded procurement strategies and broader landscape use patterns inferred from contemporaneous hunter gatherer stone tool tradition s has not been explored at the regional scale . Using a new methodology to reconstruct technological organization from a whole assemb lage perspective, the reconstruction of artifact discard patterns and mobility patterns indicates that lithic technologies of Southeast Asia were an integral component of the technological system and were structured in ways fundamentally similar to hunter gatherers of the Late Pleistocene foragers of Western Eurasia. The findings from this thesis suggest t hat stone tools of Southeast Asia can be modeled within a larger anthropological framework emphasizing the relationship between mobility strategies an d hu man technological behavior. Furthermore, the preliminary results indicate that non h uman assemblages, which have been

PAGE 5

v difficult to distinguish from modern assemblages typologically, deviate from the expectations of the model and could be behaviorally signi ficant and not simply a broader function of the environment. The form and content of this abstract are approved. I recommend its publication. Approved: Tammy Stone

PAGE 6

vi TABLE OF CONTENTS I. INTRODUCTION ................................ ................................ ................................ ............... 11 ! Specific Aims ................................ ................................ ................................ ................... 15 ! II. THEORETICAL BACK GROUND ................................ ................................ ................... 19 ! Interpretive Framework ................................ ................................ ................................ ... 20 ! Human behavioral ecology. ................................ ................................ ............ 21 ! Optimal foraging strategies. ................................ ................................ ............ 24 ! Middle range theory. ................................ ................................ ....................... 25 ! The Forager Collector Continuum ................................ ................................ .................. 27 ! Mobility patterns. ................................ ................................ ............................ 28 ! Curation concept. ................................ ................................ ............................ 29 ! III. BACKGROUND ................................ ................................ ................................ .............. 32 ! Pleistocene Technological Behavior ................................ ................................ ................ 33 ! Technological origins. ................................ ................................ ..................... 36 ! The Movius Line ................................ ................................ ................................ .............. 37 ! Non lithic industries. ................................ ................................ ....................... 39 ! Archaic Homo in Pleistocene Southeast Asia ................................ ................................ .. 40 ! Indonesian hominins. ................................ ................................ ...................... 41 ! Interpretive Approaches ................................ ................................ ................................ ... 44 ! Pacitanian industries. ................................ ................................ ...................... 45 ! Sangiran industries. ................................ ................................ ......................... 47 ! Flake and core dichotomy. ................................ ................................ .............. 49 !

PAGE 7

vii IV. CONTEXT ................................ ................................ ................................ ........................ 52 ! H. sapiens of Pleistocene Southeast Asia ................................ ................................ ........ 53 ! Modern human expansion. ................................ ................................ .............. 54 ! Late Pleistocene Archaic Hominins ................................ ................................ ................. 56 ! Paleoenvironmental Conditions ................................ ................................ ....................... 58 ! Ecological challenges. ................................ ................................ ..................... 59 ! Site Background ................................ ................................ ................................ ............... 63 ! V. METHODOLOGY ................................ ................................ ................................ ............. 65 ! Whole Assemblage Behavioral Indicator ................................ ................................ ........ 66 ! Evaluating retouch. ................................ ................................ ......................... 66 ! Lithic te chnological organization. ................................ ................................ .. 68 ! Collating data. ................................ ................................ ................................ . 69 ! Analytical considerations. ................................ ................................ ............... 69 ! Statistical interpretation. ................................ ................................ ................. 72 ! VI. RESULTS ................................ ................................ ................................ ......................... 75 ! General findings. ................................ ................................ ............................. 77 ! Modern human assemblage patterning. ................................ .......................... 78 ! Specialized point technology. ................................ ................................ ......... 80 ! Chronological changes. ................................ ................................ ................... 82 ! Non modern assemblage patterning. ................................ .............................. 84 ! VII. CONCLUSIONS ................................ ................................ ................................ ............. 86 ! Southeast Asian Technological Organization ................................ ................................ .. 87 ! Hominin assemblages. ................................ ................................ .................... 90 !

PAGE 8

viii SUMMA RY OF FIGURES FIGURE 1: DISTRIBUTI ON OF SITES ACROSS S OUTHEAST ASIA ............................ 12 ! FIGURE 2: ASSEMBLAGE PATTERNING ACROSS AL L SITES ................................ ... 77 ! FIGURE 3: ASSEMBLAGE PATTERNING AT LIANG B UA ................................ ........... 79 ! FIGURE 4: ASSEMBLAGE PATTERNING IN SPECIA LIZED POINT TECHNOLO GY 81 ! FIGURE 5: ASSEMBLAGE PATTERNING WITHOUT O UTLIERS ................................ 82 ! FIGURE 6: ASSEMBLAGE PATTERNING IN SE ASI A ................................ ................... 87 ! FIGURE 7: ASSEMBLAGE VARIATION IN EAST TI MOR SITES ................................ . 90 ! FIGURE 8: ASSEMBLAGE VARIATION IN NON MODERN SPECIES ......................... 91 !

PAGE 9

ix S UMMARY OF TABLES TABLE 1: SUMMARY OF ANALYTICAL CONSIDERA TIONS ................................ ..... 70 ! TABLE 2: SUMMARY OF SITES AND ASSEMBLAGE S ................................ ................ 73 ! TABLE 3: SUMMARY OF RESULTS ................................ ................................ ................. 76 !

PAGE 11

11 CHAPTER I INTRODUCTION This thesis e xplores lithic technological organization and human adaptation in prehistoric foragers of S outheast Asia (SE Asia) . The chronological period of the research spans the geological per iod of the Upper Pleistocene , or Late Pleistocene , which bro adly coincides with the Middle to Upper Paleolithic transition, as well as the African Middle to Late Stone Age . For the purpose of this study , which draws parall els to Pleistocene Eurasia, the language of Paleolithic studies is occasionally u sed as a point of contrast ; other times it is simply convenient for broader reference to the cultural histories of the SE Asian Pleistocene and Holocene period s due to the relatively late occurrence of Ô Neolithic ' manifestations , which largely occurs without a lithic component. The prehistoric sites selected for this study extend across much of the SE Asian subregion, covering territories on Peninsular Malaysia (Cambodia, Thailand), the Malay Archipelago (Indonesia, Sulawesi), Philippine Archipelago, and eastern New Guin ea . The assem blages considered include the sites of Lang Rongrien, Liang Lemdubu, Lang Kamnan , Laang Spean, Moh Khiew, Ulu Leang, Song Keplek, Leang Sarru, Pia Hudale, Ille Cave, Kria Cave, ToŽ Cave, Uai Bobo 1, Uai Bobo 2, Lie Siri, Bui Ceri Uato , Song Terus and Liang Bua (Figure 1) . In covering many parts of the region and drawing from multiple occupational phases , the assemblage data presented here are thought to reflect behavioral differences in technology at the organizational level , while also indicating that unde rlying this variability are fundamentally similar strategies to raw material use and mobility in SE Asian Paleolithic foragers on the whole. Furthermore, the assemblage data taken from behaviorally modern foragers present a backdrop to address differences in technological behavior at the species level

PAGE 12

12 at the sites of Liang Bua and Song Terus, or in the very least identify them as anomalous before turning to investigate other formative processes that are more taphonomic in nature than behavioral. Technolog ical simplicity and continuity over time is not altogether unique in human evolutionary contexts and certainly not those of Central and East Asia, but the regional scale of this technological stasis and the temporal reach into Early and Middle Holocene is a uniquely SE Asian phenomenon. Hence, SE Asia is a historical context in need of an alternative approach to stone tool analyses due to their importance in addressing behavior and other chronologically diagnostic features. Although the lithic industries of Pleistocene SE Asia are often described as unexceptional and unchanging, the stone tools here provide a line of Figure 1 : Distribution of sites across Southeast Asia

PAGE 13

13 evidence to major ongoing debates in paleoanthropology and the nature of human adaptation more broadly. In the absence of a well defined techno logical sequence, the characteristically un characteristic stone tools discovered here are at all times relevant and uniformly important data to the major evolutionary questions surrounding the nature of this variability. Overview In general, the stone to ol artifacts from these assemblages are representative of the broader technological system of Pleistocene SE Asia . H owever, little is known of the prehistoric behavioral strategies underlying this broader technological system . In particular, two major issu es are regularly noted by s cholars : the first deals broadly with the fact that SE Asian lithic artifac ts do not appear to easily lend themselves for distinguishing behaviorally modern assemblages from those accumulated by archaic Homo species; the second c oncerns the extent to which this technological system was actually embedded within wider land use patterns of Late Plei stocene fo rager communities . The research pr esented in this thesis tests this premise by examining the relationship between economic beha vior, lithic technological organization, and mobility strategie s in Paleolithic SE Asian forage rs. The results from the behaviorally modern sample are then turned bac k toward addressing the question of authorship in i mportant late surviving hominin localit ies. The non modern assemblages from Liang Bua and Song Terus , included in this study, present an opportunity to test whether the stone tool industries of archaic humans conform to the expectations of a model now informed through the study of assemblage sc ale patterning in Paleolithic foragers of SE Asia . The Pale olithic period of SE Asia generally lacks typological ly distinct forms, and it would appear degrees of technological complexity is g enerally not compatible with the behavioral de velopments of Homo species in this area of the world . The se Late Pleistocene

PAGE 14

14 foragers retained the basic technological system of the former indigenous hominins of the region, the components of which persisted well into the Holocene. Attempts at distinguishing between the te chnological behavior of anatomically modern humans (AMH) and former archaic inhabitants such as H. erectus often fall on securely dating sequence s, which has often been highly problematic . Moreover , the possibility that Late Pleistocene foragers interacted or nearly overlapped with archaic hominin contemporaries in H. floresiensis and Ôevolved' H. erectus reintr oduce the problem of stone tools and authorship irresp ective of the context. Recently, Paleolithic researchers have turned to assemblage scale vari ability to address important behavioral shifts in technological organization and land use strategies of the Middle to Upper Paleolithic transition. Late Pleistocene foragers in Western Eurasia recurrently display a strong negative relationship between the density of lithics they contain and the frequency of retouched pieces the rein (Riel Salvatore and Barton, 2004). The ends of the spectrum defined by this relationship can also shed light about the dominant mobility strategies (e.g., residential, logistical ) employed by the foragers who accumulated these assemblages (Riel Salvatore et al. , 2008). This trend seems to hold at a continental level (Barton et al. , 2011), and appears to be a fundamental dimension of lithic technological organization from the Moust erian on (Kuhn , 2004; Clark , 2008; Riel Salvatore et al. , 2008). Given that this trend characterizes both Neanderthals and modern hum ans after ca. 130 ka , it could be reasonably assumed that modern humans radiating into SE Asia might also employ similar st rategies. Moreover, the methodology's alignment within the temporal parameters of the Middle to Late Pleistocene transition have fundamentally the same questions to answer by using this model. The Ôwhole assemblage behavioral analysis' m ethod used in this thesis offers the potential to determine whether the hominins responsible for accumulating the Liang Bua lithic

PAGE 15

15 assemblages were characteristic of AMH Paleolithic foragers or unique in th eir own right, and thus yielding new measures of behaviorally, and pe rhaps taxonomically sensitive information from stone tools. Specific Aims Late Pleistocene hunter gatherers of SE Asia are broadly characterized as technological opportunists; incorporating locally available lithic raw material on an encounter basis (Rabbe t, 2012). In contrast, the adaptive strategies of foragers of the European Late Pleistocene are believed to have mobilized in accordance with deeply embedded procurement strategies, which were most sharply influenced by select fine grained materials (Riel Salvatore et al., 2008 ). The behavioral model applied in this thesis is founded in ethnographic and ethnoarchaeological evidence, which have shed light on how hunter gatherers structured their stone tool use behaviors in relation to myriad factors includin g resource predictability, raw material availability, efficiency, transport, and stylistic preference, to nam e a few (Binford 1979; Nelson 1991; Shott 1986; Torrance 1 983; Wiessn er 1982; Wobst 1983). At present, only a handful of studies have applied the p rinciples of this research program directly to the study of SE Asia lithic assemblages (e.g., Anderson, 1990; Siman juntak, 2001), while other scholars have put forward new interpretive methods as an alternative to typological approaches (Bor el et al. 2016; Marwick, 2007). The ecologically based perspectives on human behavior have been difficult to operationalize in the study of prehistoric SE Asia due to lingering questions surrounding the Pleistocene environm ent, particularly in terms of the sub standard q uality of lithic raw material (i.e. Movius Line), the onset and composition of the Late Pleistocene tropics, and the nature of tropical adaptation in human species more generally . While the ecology of SE Asia and the notion of tropical

PAGE 16

16 adaptation more gene rally have always figured into explanations of low input industries (e.g. Kelly, 1982), the nearly total lack of any standardized elements in stone tool forms throughout Late Pleistocene period in particular have led some to question whether stone tool tec hnology had been liberated from the adaptive strategy entirely (Reynolds, 1992). This wider cultu ral phenomenon across SE Asia is the basis for the first set of questions explored with this model and are listed below : 1.) Do prehistoric foragers of Southe ast Asia structure their lithic technology in relation to wider landscape use and mobility patterns as seen in Pleistocene Western Eurasian hunter gatherers? 2.) Is a regional context with poor quality raw material and dominated by expedient industries wi th minimally retouc hed tools incompatible with behavioral models which emphasize the relationship between mobility and curation ? The second set of questions explore d in this model addresses the uniquely persistent technological continuities , and the poten tial diversity of the Upper Pleistocene hominin landscape . The evolutionary context of Pleis tocene SE Asia has hosted a number of peculiar hominin forms, at least one of which inhabited the region as recently as the Middle to Late Pleistocene boundary. Alt hough the revised chronology at Liang Bua, Indonesia indicates that the youngest occupational period of H. floresiensis predate the arrival of H . sapiens in SE Asia, the older dates have bolstered the general acceptance of H. floresiensis as a distinct, ar chaic relative of H. sapiens . The evidence for multiple archaic hominin populations surviving up to

PAGE 17

17 this period is controversial, but the notion of other late hominin extinctions is broadly recognized as a likely scenario here. For the time being, defining the spatiotemporal range of SE Asian hominins at the close of the Middle Pleistocene is lim ited to Liang Bua, whose species holotype has been re dated to 100 " 80 ka, while the associated stone tools span into the Late Pleistocene (190 " 50 ka), immediate ly prior to the first evidence of H. sapiens in SE Asia (Sut ikna and Tocheri et al., 2016). Although only a handful of assemblages can be securely dated to the Middle Pleistocene here, the Paleolithic sequence is defined as one of the longest continuous se quences in human history due to the prolonged technological stasis. In fact, was only recently that stone tools were definitively dated to a period unquestionably linked with archaic hominins at the site of Mata Menge on Flores, Indonesia (Moorwood, 1998). Prior to these assemblages dating to the Middle Pleistocene, there were some of the opinion that H. erectus in SE Asia did not use stone t ools at all. Incidentally, the findings from Mata Menge would come to be described as the stone tool using ancestors of H. floresiensis , not H. erectus . The stone tools and fossil remains at Mata Menge are believed to be those of the founding population of Homo on Flores, Indonesia, which exhibit strong anatomical ties with more recent, Late Pleistocene populations of H . floresiensis discovered at the site of Liang Bua, which is also located in the Soa Basin (van den Bergh et al., 2016). Prior to the recent fossil discoveries at Mata Menge, published just this last year no less, former excavations had yielded only one ass emblage with unambiguous stone flakes (n=14) confidentially associated with Pleistocene era hominin distributions of SE Asia . T he theme of SE Asian technological continuity is largely defined by a few well dated fossil and archaeological contexts of H. f loresiensis and their ancestors on Flores in the early

PAGE 18

18 Middle Pleistocene, and the widely distributed industries associated with modern human foragers of the Late Pleistocene. The actual Ôcontinuity' and whether hominin populations sustained this technolog ical behavior throughout the Middle Pleistocene lacks robust archaeological evidence. Nevertheless, the arrival of modern humans on the SE Asian archipelago appears to have only minimally impacted the lithic reduction and land use strategies pioneered by a rchaic hominins of the Early to Mid dle Pleistocene, which raises the question of whether ecological constraints are responsible for assemblage patterning due to some combination of poor material quality and the unique distribution of non lithic resources i n Pleistocene SE Asia. The second series of questions stated below considers whether assemblage patterning can be used as a strictly defined behavioral correlate of AMH and thus tap into taxonomically sensitive behavior , or if assemblage scale patterning i n this region here follows typological analyses in blurring technological boundaries, indicating how strongly the broader environmental pressure facing all Homo species inhabiting the Pleistocene SE Asia impacted the use of lithic technologies here . 3.) D oes the assemblage scale patterning in Southeast Asian foragers hold across all Homo including non modern hominins, which might indicate that regional patterning is not taxonomically significant but rather environmentally consistent? 4.) Is the model prop osed in this thesis a compatible alternative to typology in Southeast Asia in terms of identifying behavioral sensitive information whic h could be used to distinguish archaic hominins from modern humans ?

PAGE 19

19 C HAPTER II THEORETICAL BACKGROU ND The study of h unter gatherers, both past and present, draws heavily on the perspectives of middle range theory (MRT) and human behavioral ecology (HBE). In particular, these perspectives have been attractive to scholar s of prehistory because they provide a genuinely tes table model for a wide range of hypotheses that can be applied in the archaeological record . Over the last several decades , the subject of forager mobility and technological organization have become tightly integrated wit hin this explanatory framework, emp hasizing the role of lithic technology and its relationship to forager mobility, resource exploitation, and economizing behavior (Binford, 1980; Bamforth, 1991, Bamforth and Bleed, 1997; Kelly, 1983; Shott, 1986; Torrence, 1989 ; Winterhalder and Smith, 200 0 ). The theory reviewed in this chapter reflect the fundamentals of the contemporary processual approach (also variously referred to as Processual plus) used in this thesis. Notably, the relevance of evolutionary principles in the study of HBE and the more narrowly applied study of optimal foraging strat egies (OFS) , and the middle range approaches which links general theories to the cultural and natural forc es influencing their formation. Another component of the processual plus approach touched on in this chapter is the construction of research and design that actively contributes to anthropological theory through developing methodical tools and testable hypotheses . In approaching the archaeological record as fundamentally testable, the relationship between data and the explanatory construct can develop into a powerful concept in its own right. Of particular relevance to the subject matter in th is thesis are the conceptual measure s of mobility and curation, which together provide the

PAGE 20

20 interpretive framework f or identifying assemblage patterning and linking them with the specific behaviors believed to have form ed them. Interpretive Framework In viewing aspects of Pleistocene human mobility strategies and technological behavior as part of larger cross cultural phenomenon, the contemporary processual approach adheres to a theoretical foundation in close company w ith the methods and theory of the natural sciences. The explanatory power of HBE and MRT is strengthened through the differing roles they play in interpr eting the archaeological record. The middle range emphasis on observable data and how it came to rest in a particular state, provides a line to mid level empirical content (i.e. regularities in artifact Ôbehavior'). These data can be tested with reference to the Ôhigh theory' evolutionary perspectives of HBE, which together brings to bear a positivist approach to int erpreting cultures of the past. The first arm of this interpretive framework, HBE, concerns the theory of actual, or perceived rules governing all behavior, which projects universalities observed in biological survival strategies onto the prehistoric study of humans. In this sense, HBE defines the behavioral parameters for which an individual, or group of individuals are expected to operate withi n, given the conditions for a various prehistoric context. The organization of these ideas in the study of archaeological cultures is the primary subject matter of opti mal foraging strategies. If the evolutionary themes of HBE are to bear out the adaptive strategies of non living people, the interpretation must flow from a scientific understanding of the archaeological data. Consequently, the second arm of this interpretive framework, MRT, is the critical theory building component which links behavioral tre nds with a line of empirically tested

PAGE 21

21 observations concerning the formation of the archaeological record. The fundamental goal of MRT is to link the interpretation of prehistoric data to the actual processes which are believed to have shaped their archaeol ogical formations, thus providing some level of scientific grounding in the interpretation of archaeological data. Human behavioral ecology. As was alluded to above, HBE, or behavioral archaeology, is a broader theoretical approach drawing on evolutionary principles Ñ namely a biological organism's predilection to behave optimality in response to ecological pressure Ñ which outlines how humans should interact within their surroundings when aspects of a given environment are known. Both modern and prehistoric human behavior is conceptualized as fundamentally adaptive, not unlike that of any other organism in the natural world. However, our place in the natural world is uniquely different, and the multitude of behaviors which drive adaptive strategies in human populations are intertwined with, and often orchestrated through various dimensions of socio cultural organization . Thus, as it pertains to the study of prehistoric people, HBE is concerned with the study of human behavior in relation to ecological and soc io cultural conditi ons (Bird and O'Connell, 2006). Technology, subsistence, and settlement are vital to survival strategies and, as such, are among the most relevant matters in the study of human culture and ecology. These facets of social organization ar e viewed as the primary buffering mechanisms by which prehistoric humans modulated new or worsening constraining pressures within their environs ( Binford , 1980 ; Winterhalder and Smith, 1992 ). Underlying the levers of cultural systems is an ecologically dri ven adaptive strategy believed to be influenced most strongly by fitness maximizing behaviors (Bird and O'Connell, 2006) . Since the forces of nature do not place

PAGE 22

22 pressure squarely on the biology of humans, but rather is mitigated through behaviors embedded within complex social systems, the structural reconfigurations in human sociocultural systems are viewed as adaptive traits, which can be studied with explicit reference to the process of natural selection (Kelly , 1995) . In evolutionary terms, the locu s of change still occurs at the individual level. However, the mechanism of inheritance is not strictly essential to our analyses and understanding of human behavior through this lens. From the perspective of HBE, the evolutionary circumstances which endow ed the human species with highly sophisticated cultural faculties is only tangentially related to the research questions they pursue, and is not presumed to "seriously constrain adaptive respo nses to ecological va riation" (Nettle, 2013: 1055). The basic no tion here is known as the phenotypic gambit and, as it relates to its broader application in HBE, is the idea that favorable traits which maximize fitness will be differentially selected regardless of the pathway of inheritance (Smith and Winterhalder, 199 2). Thus, practitioners of HBE start with the premise that behavioral variability is not routed through changes in gene frequencies , b ut rather is representative of the human capacity to adapt in a manner suitable with the proximate environmental and socia l conditions (Codding and Bird, 2015) . Certainly, the role of genes in producing behavioral variability presumably much more involved than is accounted for with the logic provided by the phenotypic gambit. And, indeed, advocates of the strong sociobiologi ca l thesis envision a closer "link between genetics and behavioral variation," (Kelly , 1995: 51). However, while these behavioral directives are entangled in both the genetic and social domain of human behavior, the capacity to respond to changing environm ental condition s is the actual trait that evolved under natural selection, not

PAGE 23

23 the particulars of the behavior itself. Thus, the capacity for culture evolved under selective pressure which favored the various building blocks of behavioral flexibility, such as the ability to exchange new information socially and utilize the memory of that knowledge across variable and changing environments ( Nettle, 2013) In this way, human behavior is viewed as phenotypic extensions which emerge from a basic genomic configur ation , whereby variability in human behavior occurs within the Ônorms of reaction' . Accordingly, the efficacy of cultural behavior is rooted in an individual's evolutionarily tailored capacity to tun e one's own behavior in accordance with their proximate e nvironment and future environment. The type of adaptive latitude described above is referred to as behavioral plasticity and, in essence, refers to the elasticity and ephemerality of human adaptive responses, which are made possible and coordinated through various channels of complex social behavior. As it relates to prehistoric humans, aspects of artifact patterning in the archaeological record is thought to represent the distillation of these behaviors, and their retention over time indicates its success via differential selection. In this sense, these behaviors are cultivated under selective pressures not unlike those of other organisms, which live and reproduce in full exposure to the forces of nature. And while behavioral traits in humans amass and find inventory in the culture system, which buffers against the sort of raw environmental pressure that results in biological change, the differential selection of behaviors, although culturally mediated, are still treated as behavioral variants that flourish and wane depending on their optimality under a given set of ecological conditions. In the study of human strategies, the individual fitness of an organism is replaced by behavioral variants, which refer to the mean fitness of a specific behavior, and serve as the basic analytical unit in the study of optimality in human foraging (Kelly , 1995).

PAGE 24

24 Optimal foraging strategies . Despite the position held HBE, which rejects the notion that human int eraction and ecology necessarily require a fundamentally differen t interpretive framework than a ny other organism (Nettle, 2013), the evolutionary approach to human cu lture is far more nuanced. In a sense, behavioral plasticity freed up Ôadaptation' from the hardened evolutionary channels within which organisms become m asterfully adapted to their present habitat and, generally speaking, increasingly vulnerable to change in the future habitat. In non human animals and plant species, optimality is reflected in a fairly straightforward relationship between morphological and structural commitments (i.e. anatomy) and subsistence related tasks; it is also quite manageable in comparison to humans in the sense that fitness is a relatively easy measure when working with non human anima l populations because the relationship between behavioral phenotypes and genotype is more straightforward. From an ecological standpoint, the decisions and activities underwriting human behavior are similar to the biological tendencies to maximize one's individual fitness. However, human populations r esolve ecological challenges by channeling reproductive and subsistence related tasks through a network of social and cultural behaviors carried out at the group level. The body of work advancing the study of OFS is primarily concerned with outlining speci fic sets of contingent relationships between components of the social unit and the frequency and distribution of environmental variables (Bird and O'Connell 2006; Winterhalder and Smith, 2000). Consequently, inferring behavioral optimality in human populat ions involves several more variables , many of which are materialistic. Behavioral variants are measured in relation to these specific variables, or currenci es.

PAGE 25

25 In calling on the general theories of behavior provided by HBE, OFS concerns itself primarily with the indirect measure of energy expenditures, the various cost benefi t tradeoffs, and the application of a testable hypothesis. T he simplicity of behavioral models allows hypothesis to be put forward and easily tested in the archaeological record. The use of OFS simplify behavior and focus on the relationship between specific variables. For instances, a heterogeneous distribution of resources and the time and energy used to extract those resources as they deplete versus the time and energy saved by mov ing camp. In this way, OFS describes a series of pro babilistic relationships which articulate with the various aspects of technology, society and the environment. Middle range theory. Not unlike many other traditions of archaeological inquiry concerned wi th the lifeways of pre historic societies, MRT approaches are chiefly concerned with deriving meaningful interpretations in the archaeological record, but lean far more heavily on the role of empirical insights and addressing its interpretive significance i n relation to the formative processes which drive archaeological patterning. In viewing the archaeological record as heavily shaped by generalizable processes, the depositional context and artifact con tent observed in the present can be used to substantiat e claims about the past (Binford 1977, 1982). From these data, the archaeological record can be used to reconstruct a theory on how it formed. Moreover, the reconstructive function of observations made in the present, limited as prehistoric context are, is bolstered by contemporary ethnographic studies and experimental testing. In essence, though, the Ômiddle' in MRT refers to any link between a particular theory and specific type of empirical content with limited scope (Kosso, 1991; Schiffer, 1988). Noneth eless , the popularity of MRT research in archaeology has been fostered by the its close alignment with

PAGE 26

26 the methodological and theoretic al modeling of the natural sciences, which others have strongly contested (e.g. Hodder, 1986). In viewing human behavior as fundamentally adaptive, MRT practitioners attempt to explain the relationship between artifacts, ecology, social organization, and depositional environment through linkages to a higher level of foraging theory. As was mentioned above, the Ôhigher theor y' of human behavior is not generally the focal point of the MRT research, nor is it exclusively paired with HBE, but rather serves as a precept that allows variability in the archaeological record to be treated as a series of probable outcomes which guide s the interpretation of evidence. With this format, the flow of information is not hierarchical; theories can be used to inform observations, but the observations themselves can form the basis of a theory to be tested against the evidence (Kosso, 1991). As such, an all encompassing de finition for MRT is difficult to find agreement on because of the extensive interdisciplinary input used to formulate archaeological expectations. As its used here , the bod y of theory guiding this model is informed most strongl y by neo evolutionary perspectives on behavior (HBE, OFT), actualistic and ethnographic studies (MRT), and the quantitative methods in which archaeological patterning is explained . Actualistic studies are essentially an empirical check on both the theoreti cal validity of neodarw inian perspectives, and the full breadth of f f ormative processes believed to shape archaeological patterning. Ethnographic analog , taph onomic experimentation, and historical reference among other forms of inquiry, derive data to form ulate expectations to be tested against the archaeological record (Flannery, 1982). Contemporary studies of traditional societies still practicing hunter gathering present a wealth of information likely relevant to prehistoric modes of life; but also prese nts the opportunity to identify cross cultural trends and

PAGE 27

27 normative behavior to be tested in the archaeological record. Prior to the use of MRT in archaeology , ethnographic data was more or less borrowed to fit with the archaeological evidence. Experiment al studies allow artifacts to be reverse engineered in a scientifically controlled setting, permitting an extremely close and deliberate study on the processes which drive the formation of an artifact, or assemblage of artifacts. These experimental observa tions offer a unique contribution that has direct bearing on the recovery, analysis, and inference of archaeological material (Schiffer, 2013). In summary of the processual plus interpretive framework, this theoretical orientation is heavily influenced b y evolutionary principles , with a notable emphasis the relationship between natural selection and cultural variability. Physical artifacts are viewed as an adaptive phenotypical variable extension of biologically rooted behavior, and as such is subject to the same selective forces shaping behavior on the whole. In simplistic terms , the tinkering and dispensing with behaviors that are not optimal in performance is what makes human behavior predic table and therefore testable . The embeddedness of these behaviors in th e cultural systems ( e.g. technology, subsistence, and settlement ) , and their material extensions provide the data to measure optimality across prehistoric human survival strategies. The interrelatedness of these cultural systems is such that a change in on e system can have a cascading effect and lead to a change in another system. In this way, researchers can potentially isolate a single component of the system, say, technological organization, and measure a change in another variable, like settlement stra tegy. The Forager Collector Continuum Mobility in the archaeological record is a conceptual measure landscape use patter ns in pre agricultural communities . The concept of mobility is grounded in the relationship

PAGE 28

28 between human adaptation and the natural pre ssures and particulars of habitat s . Mobility in this sense , is managed around strategies of resource exploitation (e.g. subsistence strategies, raw material procurement) and is measured along a continuous spectrum which roughly correspond with the demands of a particular environment . In the study of Pleistocene adaptive strategies , forager mobility is often the corn ersto ne of the interpretive framework with the conceptual distinction between log istical and residential settlement patterns in hunter gatherer s being the dominant theme in hunter gatherer research. In some ways, mo bility is viewed as the critical adaptive system powering Pleistocene human survival strategies, around which technology, subsistence, and other socio cultural aspects of human behavior are organized. Mobility patterns. Binford (1980) characterized logistical mobility as a small group of individuals whose site occupation was of longer duration and whose movements consisted of carrying out task oriented forays within a larger forager ra dius. Comparatively, residential mobility refers to a small er group of individuals whose hunting expeditions were restricted to a smaller foraging radius, but also involved frequent camp relocation as a means to stabilize subsistence economy as resources d iminished . Essentially, Binford (1980) dichotomized the human settlement configuration in terms of how subsistence resources were pursued, extracted, and consume d by hunter gatherers. In the simplest terms , residential hunter gatherers move camp to resourc es, while logistical hunter gatherers move resources back to the camp. The basis for interpreting assemblage patterning as a measure of human organization is founded in the conceptual distinction between logistical and reside ntial mobility and the notion of curation (Binford, 1979; 1980). What was important about these observations is that Binford noticed that different settlement strategies resulted in different assemblage formation,

PAGE 29

29 the consequence of which were distinct material signatures that could be extrapolated into the archaeological record. T o evaluate mobility , stone tool artifact composition and density within assemblages are commonly used quantitative measure. Because of the central role of stone material in the wider web of forager activities and survival, it has been possible to articulate the various states of artifacts from their recovered context with the broader mobility strategies from which they were embedded (Andrefsky, 1994; 2001; Barton, 1998; Hiscock, 2002; Barton and Riel Salvatore, 2014; Riel Salvatore and Barton, 2004; Riel Salvatore et al. 2008). Curation concept. An important distinction in the interpretation of lithic variability are contrasting views concerning the underlying mechanisms driving patterning in artifact morphology . From one perspective, stone tool traditions are organized through their morphological similarities in artifact form, and are conceptualized as the consequence of selective processes imposed by culturally descendant groups. From this viewpoint, variabilit y in artifact form is largely constrained by social descent, and as such, were often treated as cultural markers (sometimes referred to as "fossil types"), which when plotted along spatial and temporal boundaries were felt to retrace the succession of cult ural lineages. The interpretation of stone tools accumulations as a measure of occupational intensity assumes that the technological behavior of prehistoric foragers was sharply shaped through deliberate consideration of the current and future utility of chipped stone tools they produced, and their proximity the next cache or quarry. The broader topic of economizing behavior and lithic technology is known as the curation concept , which emphasize t he relationship between the dynamic life history of lithic artifact and their final morphology, wherein myriad factors contribute to the condition in which the artifacts are discovered in, most of which are the

PAGE 30

30 unintended consequence of tool use behaviors. In particular, the identification of a host of potential f actors have been outlined, ranging from the intensity and breadth of tool activities performed, the length of use , to the quality o f the raw material used. (Bamforth, 1986; Binford, 1979; Odell, 1996; Shott, 1996). Importantly, these use life patterns in stone tools are believed to articulate within a wider technological system, whereby important economic decision making behaviors such as lithic provisioning, the distribution of lithic raw material, and the quality and quantity of lithic material available (Andrefsky, 1994; Bleed, 1986) regularly factor into the character of lithic assemblages encountered in th e archaeological record. The curated expedient continuum of lithic organizati on frames technolo gical behavior within corresponding mobility strategi es of hunter gatherers . One the one hand, curated industries are characterized by heavily modified lithics, core reduction is usually extensive, and retouched tools can be larger in size and ranging in shape Ð allowing flexible modification of stone tools a s acces s to raw material diminished. In Pleistocene hunter gatherers, this type of lithic organization appears to have been favored by highly mobile foragers and so consequently would be most often represented in ephemeral archaeological contexts. Comparat ively, high densities of artifacts and low frequencies of retouched tools charac terize expedient assemblages. Retouched tools are often function specific and are manufactured for immediate use. Raw material is either locally abundant or raw material procur ement is embedded within the use of the site; this type of technological organization is ind icative of less mobile groups. These lithic qualities would be most often represented in archaeological contexts with repeated and sustained occupati on such as a re sidential camp. However, this type of relationship becomes difficult to demonstrate when working with disparate stratigraphic c ontex ts, especially when it is unclear how rapidly ass emblages formed. The model applied

PAGE 31

31 in this thesis approaches curation at t he assemblage level, minimizing detailed accounts of the tools themselves and instead, analyzing the collective frequency of retouched artifacts in relation to the density of all other artifacts within a given assemblage (e.g. Riel Salvatore and Barton , 20 04).

PAGE 32

32 CHAPTER III BACKGROUND The various cobble and flake based industries of Ô Paleolithic ' SE Asia (ca. 1.1 ma Ð ca. 11 ka) have been widely recognized as a technological system unconnected to the lines of development that define the clas sic stone tool traditions of the western Old World (Corvinus, 2004; Movius, 1948). In fact, the peculiarity of these stone tools has been so extensively noted that virtually all recen t literature on the subject of Paleolithic in SE Asia comment on this distinction, which usually calls attention to the major interpretive challenges faced by all archaeologists working in this regional context. Namely, the many problems in addressing the indistinctiveness of these lithic artifacts and often unconstrained geological context f rom which they derive. Prominent schol ars quite accustom to chronicling these issues, characterize SE Asia as a regional context where "Émeaningful typological categories have remained elusive" (Brumm and Moore, 2012: 386), and "...definitive Ôcultural' al locations are geographically mercurial and temporally ambiguous" (Rabbet, 2013:102). The unstandardized and expedient character of the stone artifacts blur technological divisions and contribute to the problem of addressing the very unique distribution of hominin taxa during the Pleistocene , and the controversy and debate concerning stone tool use among these archaic human species ( e.g. Bartstra, 1982; SŽmah, 1992). The lithic artifacts of the Late Pleistocene in SE Asia are amorphous and undiagnostic, and in many cases, are indistinguishable from earlier stone tool industries manufactured by archaic Homo species during the Middle Pleistocene. The Ôleast effort' or Ôlow input' strategies associated with these lithic industries have often been explained in re lation to a broader function of the Asian Pleistocene environment Ñ irrespective of the cognitive ability and

PAGE 33

33 technical competence of the stone tool producers. Thus, the similar nature of their expedient manufacture has been taken as an indication of a comm on underlying strategy, one which apparently transcends both cultural and species boundaries. Although the vast majority of lithic artifacts from SE Asia are fundamentally similar in their informal character, their membership within this broader classifica tion is offset by subtle internal variability. The multitude of regional differences are often subsumed into the coarser division of Ôchopper chopping tool' and Ôflake based' techno complexes in an effort to differentiate between the handiwork of modern hu man and archaic populations. However, this dichotomy has been somewhat problematic in application because it is predicated on the typological compatibility of SE Asia's heavy core tools with ÔMode 1' artifacts from African Pleistocene sequences (2.5 " 1.7 Ma), which significantly predate the earliest securely dated stone tools in the region (1.1 Ma). Furthermore, it would appear that the extension of these industries by modern humans in the Late Pleistocene preclude the use of stone tools as a proxy for beh avioral modernity in the archaeological record. Pleistocene Technological Behavior Pleistocene l ithic artifacts are the most abundant and durable source of information available in the study of our evolutionary history. The preference for lithic raw mater ial throughout human history underscores the importance of this technological medium to the evolutionary success of Pleistocene hominins. Accordingly, the nature of technological variability is an important archaeological measure in the study of our evolut ionary history. In the broadest sense, patterning in technological behavior are equated with the various modalities underwriting the cultural evolution of our species. Modalities, in this se nse, refer to the broad range and interplay of cultural , neural, a nd physiological mechanism s operating on the

PAGE 34

34 evolution of technological behavior . In the archaeological record, their fossil and material cultural extensions are treated as correlates of behavior and phylogeny, which divide up several phases in our evoluti onary history. In general, the trend toward complex tool forms and systematic reduction strategies are thought to reflect major biological and cultural developments in human evolution (Ambrose, 2001). However, while it is fairly evident that the structure of technological change trends broadly with evolutionary history of the genus Homo , narrowly defining the specific evolutionary context behind the succession of each technological phase has not been as forthcoming. In particular, what remains unclear is w hether technological diversity of the Lower Paleolithic should be explained in terms of biological differences or a combination of environmental and functional factors (Foley et al., 2003; de la Torre and Mora, 2009). The recognition of variation within m ajor categories of technological change (e.g. Oldowan vs. Developed Oldowan) had long been viewed as intermediate sequences of a continuous technological phase. However, this view has recently been critiqued upon reexaminations of Oldowan assemblages, whic h have demonstrated consistencies in knapping competency and technical skill from beginning to late phase Oldowan (Delagnes and Roche, 2005). Alternatively, recent approaches have explored the possibility that intersite variability of the Oldowan tradition could reflect local responses to environmental factors (e.g. the flaking properties of different materials), indicating a strong adaptive element in the lithic technology used by early hominins (Semaw et al. 2009; Stout et al. 2010). The earliest appeara nce of stone tool industries with robust documentation is represented by the Oldowan techno complex; it is important to note here that stone tools found in association with hominin fossils have been influential in distinguishing early Homo from

PAGE 35

35 contemporar y species (e.g. Leakey et al., 1964), while species distinction within the genus Homo and australopithecines continues to be problematic due to the morphological variation (e.g. Cela Conde and Ayala 2003; Wood and Collard 1999). Furthermore, the paucity o f stone tools which derive from hominin bearing localities is particularly problematic prior to the emergence of Homo due to the technological timespan of the Oldowan, which overlaps with a number of australopithecine species. The contextual problems and uncertainties in the origin and development of lithic technology in the Early Pleistocene is pertinent to the study of SE Asian Paleolithic industries on two fronts. Firstly, the widespread use of stone tools as cultural, phylogenetic, and chronological di stribution markers in hominin populations noticeably fails east of the so called Movius Line, which is a technological and geographic demarcation apparently affecting human populations in SE Asia as recently as those Pleistocene and Holocene foragers inclu ded in this study. Moreover, the expedient qualiti es of SE Asian industries have become deeply embedded in a historical connotation with archaic hominin populations thought to have arrived in late Early, or Middle Pleistocene Sundaland. Secondly, the assoc iated stone tool behaviors of H. floresiensis at Liang Bua redefine the standard inventory of correlates used in defining the emergence of Homo in the fossil record. The assemblages at Liang Bua are exceptional in this regard, because they have some bearin g on the evolution of stone tool manufacture. Establishing the cognitive and manipulatory requirements necessary in stone tool manufacture are difficult to estimate in extinct hominin taxa, yet it has been precisely these estimates that have converged to d efine the genus Homo . At 380cm, the endocranial volume of H. floresiensis is well below the accepted range for genus Homo , and registers on the low end of estimates for small brained australopiths (Brown , 2004). Moreover, the retention of primitive wrist

PAGE 36

36 morphology inferred from metacarpal elements of H. floresiensis indicated some level of morphological restriction in stone tool manufacture (Tocheri 2007). The puzzling combination of an absolutely small brain (with comparison to body size), primitive stat ure, and relatively complex stone tool industry raise interesting questions concerning the nature and timing of ston e tool behaviors . Technological origins. Currently, the earliest evidence of Oldowan technology dates to 2.6 Ñ 2.5 Ma at Gona, Ethiopia, and w as found to be contemporaneous with the nearby hominin bearing site of Bouri, which had previously produced cut markings on modified bone in the same geological deposits as Australopithecus garhi (Asfaw et al., 1999; de Heinzelin et al., 1999; Semaw et al. , 2003). However, the known spatial and temporal distribution of Homo , as well that of earlier and contemporaneous hominins, is regularly reconfigured or expanded in light of new fossil evidence and increasingly precise dating techniques. For example, a ne w discovery of a partial mandible from Ethiopia was recently dated to 2.8 Ma, and may represent an intermediate form of Homo (Villmoare et al., 2015). Reintroducing Homo in this region well before the lithic accumulations at Gona lithic might calm specula tion that these stone tools were made by australopithecines. More recently, however, the discovery of in situ lithic artifacts in West Turkana, Kenya, were dated to the Pliocene (3.3. Ma), which are .5 Ma older than the earliest evidence for the genus Homo , and fall within the spatiotemporal range of Kenyanthropus platyops (Harmond et al. 2015). While these stone tools were found to bear distinctive patterns in flake detachment Ñ a fundamental criterion used to assess technical skill associated with the Oldow an techno complex (e.g. unidirectional flakes, debitage) Ñ the knapping skill was considered underdeveloped by

PAGE 37

37 comparison, which the authors believe warrants the techno complex distinction of Lomekwi, and hence the "pre Oldowan" designation (Harmond et al., 2015). However, it is worth mentioning here that K. platyops are still considered by some to be an archaic predecessor to the genus Homo (Cela Conde and Ayala, 2003). Given this taxonomic uncertainty, the relationship between non Homo hominins (i.e. A. ga rhi, K. platyops ) and stone tools is, for the moment, contingent on the spatial and temporal distribution of known stone tool manufactures in Homo . In this sense, stone tools still have retained a high level of classificatory power, and as such, heavily i nfluence the taxonomic definition of Homo . However, despite the fact that stone tools continue to factor heavily into designating Homo , establishing an exclusive criterion (e.g. cranial capacity; morphology) strongly suggestive of technological origins wi thin Homo has also yet to materialize. Although debate continues, the overarching patterning from simple to complex is still the cornerstone of Paleolithic research. Yet contemporary perspectives on the nature of this variability are largely informed by ma terial cultures of Africa, Central and Southern India, and Western Eurasia. In contrast, the Asian Paleolithic is represented by an extraordinarily long sequence of technological stasis, lacking the type of directional change in lithic technologies observe d in western portions of the Old World. Here, simple core and flake tool technologies persist into the Late Pleistocene at a near continental level (excluding West Asia) and as recently as the Terminal Pleistocene in SE Asia, indicating environmental const raints may have been a powerful limiting factor in the development of co mplex lithic industries. The Movius Line There has long been speculation that easily accessible and overabundant organic material such as bamboo (Pope, 1984) and wood (Hutterer, 1977) were preferred over course

PAGE 38

38 grain lithic material by early Homo in SE Asia , or alternately, that H. erectus embarked out into the Asian continent before Acheulean technology was developed (Reynolds, 2007). However, these theories, particularly the former, are largely founded in the noticeable absence of specific data (i.e. standardized stone tools) rathe r than actual physical evidence . There have been a number of explanations put forward in an attempt to explain the persistence of simple co re/flake industr ies of SE Asia (ca. 1 Ma Ñ Terminal Pleistocene), and more broadly, the similar commitment to low input technologies which characterize Central and East Asia during the Paleolithic (1.7 Ma Ñ Late Pleistocene). Currently, no single explanation is presented to account for the technological phenomenon. Rather, the lack of bifacial elements in the Asian Paleolithic is believed to be influenced by a combination of methodological biases, environmental factors, and perhaps to a greater or lesser extent, the biogeogr aphical movement of archaic hominins. The later hypothesis would involve an early hominin migration out of Africa before the rise of bifacial industries e.g. Acheulean (ca. 1.9 Ñ 1.7 Ma). This idea was originally proposed by Foley and Lahr (1997), at the tim e was supported by the earliest Indonesian fossil evidence dated around 1.7 Ma (Swisher, 1994). However, the earliest conservatively accepted appearance of H. erectus in ISE Asia much closer to the beginning of the Middle Pleisto cene . An important issue d eepening the typological divide between the Eastern and Western Old World are methodological biases operating on both sides of the Movius line. Although bifacial technology is not altogether absent in Asia, and reports of them have grown recently (e.g. Cor vinus, 2004; Mokhtar and Saidin, 2006; Pawlik, 2004), the bifacial tool discoveries in SE Asia are often categorically downgraded to "proto" and "crude" versions, and are generally believed to differ from the classic Acheulean in terms of symmetry and degr ee of elaboration

PAGE 39

39 (Lycett and Bae, 2010). However, the question of whether this reflects interpretive biases toward the recognition of complex forms has been raised (Rabbet, 2012: pg. 98). In a recent morphological analysis of Asian Ôhand axe' varieties a nd the classic Acheulean, some degree of metrical overlap was found in each of the key dimensions of bifacial technology (e.g. weight, refinement, thickness), indicating that some Asian chopping tools fall within the range of variation of Acheulean manufac ture (Petraglia and Shipton, 2008). And although bifaces are argued to dominant lithic assemblages in Africa and Western Eurasia by comparison (Norton et al., 2006), a large number of bifaces east of the Movius line are reported from un dated surface conte xts (Moore and Brumm, 2012). Comparatively, assemblages that are reliably dated in SE Asia are limited, but scholars working regionally have argued that bifacial technology is more common to surface finds than is commonly imagined (Brumm and Moore, 2012; S imanjuntak, 2010). The implication here is that the industries east of the Movius Line (i.e. pebble and flake) has not shared the same typological enclosure afforded by Acheulean and its affinities west of the Movius line. In this sense, the robust sample of Acheulean industries east of the Movius line is inflated, whereas bifaces of the Asian Paleolithic are potentially greatly underrepresented, perhaps creating a greater technological gap than is warranted for the stone tool indust ries of the Lower Paleol ithic. Non lithic industries. The alternative use of organic tools has long been suggested, but until recently this perspective operated as a theory independent of any direct line of evidence due to the remote chance of organic preservation. However, two recent studies of surface modification concerning cut markings on f aunal bone and a singl e shell tool with engraving dating to the Mid dle Pleistocene in Central Java, have been interpreted as evidence for organic tool use in

PAGE 40

40 H. erectus of Middle Pleistoce ne ISE Asia (Choi and Driwantoro, 2007; Joordens et al., 2015). The former study hinges on potential cut markings on faunal assemblages dated to the Middle Pleistocene ( Choi and Driwantoro, 2007 ), which were compared against experimentally produced cut mar ks, leading the authors to conclude that the modification was "most plausibly" inflicted by thick clamshell (Choi and Driwantoro, 2007). The later study is restricted to just one geometric engraving on shell. The authors found that the marking inflicted on the shell were "unambiguously consistent" in their morphological patterning, which is to say they could not have occurred naturally through taphonomic processes. Archaic Homo in Pleistocene Southeast Asia As is often the case, drawing the boundaries of P leistocene hominin distributions calls on a combination of fossil and archaeological evidence. The h ominin biogeogra phy of the Pleistocene here is informed through fossil evidence and, to a much greater extent, the stone tool artifacts. The Paleolithic cha racter of these stone tools and their perceived antiquity is deeply set into the archaeological perspectives of the region. The assemblages of Son g T erus and Liang Bua in Indonesia represent two of the more thoroughly studied sequences with occupational le vels dating to the late Middle Pleistocene , both of which occur in regions with well documented with fossil hominins. While H. floresiensis was discovered with associated stone tools at Liang Bua, the particular Homo species responsible for the Song Terus accumulations on Java Island will remain less clear until a better representation of the late surviving hominins emerges . Although these assemblages have been dated prior to the arrival of modern humans, the late period of the Middle Pleistocene of SE Asia is somewhat unmoored chronologically in a number of respects and difficult to assess. Due to the archaic morphological qualities of H. floresiensis in particular, as well as the character of stone tools

PAGE 41

41 more regionally, approaches to stone tools often dra w from perspectives on Early Pleistocene technological behavior. S tone tools are widely abundant and on occasion have been found in , or around ancient Pleistocene g eological deposits and continually factor into the spatial dis tribution and chronological bo undaries of primitive Homo across greater SE Asia , despite rarely being reported in any direct or peripheral association with the physical remains of archaic hominins. In SE Asia, this subject is especially complicated by a very murky understanding of arch aic Homo adaptation, their technological capacity, and their distributional extent throughout the Pleistocene. Indonesian hominins. The chronological and evolutionary perspectives on early Homo in SE Asia have largely been confined to fossil bearing conte xts of Pleistocene Indonesi a (southeastern most Sundaland). Although far from straightforward, indications are that the arrival of early Javanese H. erectus occurred sometime around the Early to Middle Pleistocene boundary [1.2 ma Ð 0.8 ma; (Larick et al., 2001; Hyodo et al., 2011)], or perhaps more firmly in the Early Pleistocene as some scholars' have attest ed [<1.5 ma (Swisher et al. 1994; SŽmah et al. 2000)]. However, little is known about the geographic distribution of these colonizers, nor is it clear how the Pleistocene environment impacted settlement of early Javanese H. erectus of Pleistocene ISE Asia . During arid glacial intervals of the Quaternary period, the island of Java would have episodically formed land bridges between peni nsular Malaysia a nd the Greater Sunda Islands (i.e. Borneo, Sumatra) to the east (Bush and Fairbanks, 2003). The Lower and Middle Pleistocene in particular, are thought to correspond with the expansion of savanna and open forest habitat. These suggested "savanna like" pale o climatic conditions would have been

PAGE 42

42 suitable habitats for human and animal dispersals (Bettis III et al., 2009), and perhaps offered significant refugia for hominins (Louys and Turner, 2012). While recent ecological approaches have begun to develop a pic ture of the environmental constraints and landmass configuration during the initial dispersal of H. erectus (e.g. Dennell, 2009; SŽmah et al., 2016), the precise timing of their arrival and subsequent evolutionary trajectory remains less clear. For instanc e, t he periodic passage of migratory H. erectus into the southern most extent of Sundaland would have temporarily stranded populations in the Greater Sunda Islands during interglacial intervals, exposing one or more H. erectus populations to highly variable island conditions (SŽmah et al., 2016), but it is still an open question whether the Javanese H. erectus represent a long sequestered lineage, or alternately, intermittent migrations of variant H. erectus which either stabilized, or replaced previous H. e rectus forms during periods of low sea stand ( Falgures et al. 2015) . Unlike neighboring Java Island, the island of Flores remained isolated during the lowest sea stand periods of the Pleistocene and was unreachable by land (Morwood et al., 1998). Import antly, securely dated lithic assemblages define the oldest human occupation here. Stone tools recovered alongside datable material from Mata Menge (1.1 ma Ð 0.8 ma) attest to the presence of a hominin species by the Middle Pleistocene (Morwood et al., 1998 ; Brumm et al., 2010). In addition to the stone artifacts, the site of Mata Menge recently produced hominin fossils dated to ca.8 Ñ 0.65 (van der Bergh et al., 2016). With whom do the island inhabiting hominins share lineages with, and how did they come to c olonize a remote volcanic island in Indonesia are among the most important and timely paleoanthropological topics debated today. The two scenarios which have traction thus far involve differing migratory routes which trace back to different source populati ons.

PAGE 43

43 The most geographically logical model involves the transit of H. erectus populations from east to west, crossing a deep underwater channel separating Java from Flores. However, the island hopping event is important, because it would require that H. e rectus were not only capable of successfully navigating difficult sea crossings, but also capable of marshalling the resources and co operation at least as many times necessary to become a biologically viable population on Flores Island. The alternative hy pothesis is that Flores was populated from the north via Sulawesi (Morwood and Oosterzee van, 2007; Morwood and Jungers, 2009; van der Berg et al., 2008). This model proposes that the source population became dislocated from Sulawesi through force of natur e, and were transported to Flores clutching natural rafts held together by debris and vegetation (Smith, 2001; van der Bergh, 2008). However likely, it should be noted that the model requires that early hominin s were displaced in multiple evolutionarily s ignificant Ôaccidental rafting' events. The latter described above, presumably followed a much earlier separation from the upper margins of SE Asia, in which hominins were swept downward via oceanic currents running north to south and onto Sulawesi, which may have been connected to Flores (Dennell, 2014 ). In either scenario, paleoanthropologists are asked to entertain surprisingly complex scenarios with considerable evolutionary ramifications. The former involves a greatly expanded behavioral range for H. e rectus stalking the coast lines only to give rise to a descendent hominin species with significant evolutionary reversals (e.g. diminished brain size). The later hypothesis involves either an earlier, unknown hominin, or perhaps H. habilis , migrating from Africa far earlier than the fossil record currently indicates. Unfortu nately, mainland SE Asia is conspicuously void of fossil hominins which has made it difficult to assess the differing models of hominin dispersal (Marwick , 2009).

PAGE 44

44 Alternatively, stone t ools have helped exte nd regional chronologies back into the Middle Pleistocene in lieu of fossil evidence, which is slowly reestablishing the notion of archaic homin in occupation in adjacent areas. In Sulawesi, for instance, a lithic assemblage has b een se curely dated to the late Middle Pleistocene (van der Berg et al. 2016b), indicating, at the very least, the possibility that multiple hominin dispersals routes are conceivab le, including those which skew the current debate toward evolutionary linkages betw een H. erectus and H. floresiensis. Interpretive Approaches Traditionally, a num ber of lithic industries of ISE Asia have been recognized as of Lower Ñ Middle Pleistocene origin in Java ( i.e. Sangiran, Sambungmacan , Pacitanian, Ngandong ), Sulawesi (Cabengia n), and the Philippines (Cabalw anian) . And while these industries are perpetually reconsidered in light of new data or alternative th eories, these artifacts have received persistent scrutiny since their discoveries (e.g. Bartstra and Basoeki, 1989; Bartstr a et al., 1994; Jacob et al., 1978; Pawlik and Ronquillo, 2003), discovered in close proximity to the fossil bearing localities have not only played a significant role in expanding on the behavioral antiquity of SE Asian fossil hominins, but also the defi ning criteria used in lieu of fossil evidence while developing regional chronologies . The disputed artifacts of early H. erectus in ISE Asia and the antiquity of lithic technological behavior more broadly across the entire region, has its origins here in J ava, Indonesia. The geological formation (Kabuh, and the directly overlain ÔGrenzback layer') from which the majority of classic H. erectus fossils were discovered have also been known to yield small concentrations of imported artifacts, variously identifi ed as ÔSangiran Flakes,' or Ngebung artifacts, which have been described as morphological similar to the Sangiran industries (Simanjuntak and

PAGE 45

45 SŽmah, 1996 ). The Ngebung artifacts consist of a small number of flakes and a cleaver which has been described as an ÔAcheluean like' assemblage (SŽmah et al., 1992). More recent excavations from the ÔGrenzback deposit' yielded a small number of flakes, which were given age determination of more than 1 Ma (Widianto, 2006). However, these artifacts occur in unconstrain ed geological deposits, which have long been noted for their severe erosional uplifting from younger deposits (Bartstra 1985; Bartstra and Baseoki, 1989; von Koenigswald, 1973). Accordingly, many scholars are cautious in accepting Middle Pleistocene ages f or artifacts excavated from Sangiran localities, and hence skeptical of their affiliation with the H. erectus fossil materials . Nevertheless, a review of the artifacts that have come to define early archaic technological behavior in SE Asian hominin specie s is described below. Pacitanian industries. The Gunung Sewa region of Java is widely known for the abundance of stone tools discovered along the terraces and riverbeds of the Baksoka River. The Baksoka River stone tools are broadly characterized by crude hand axes, hand adzes, and other chopper chopping tools made from silicified limestone, silicified tuff, and petrified wood (Soejono, 2001 ). Establishing the antiquity of the Baksoka R iver artifacts, popularly known as the "Pacitanian" industry (also forme rly known as the Pajitan), has been problematic . Historically, the Pacitanian chopping tools have been attributed to the Middle Pleistocene on accou nt of their ancient appearance and the geological context from which they derive (e.g. Van Heekeren , 1972). The original Pacitanian distinction proposed by von Koeningswald, (1936) was conceived before radiometric dating techniques were available. At the time, Pleistocene chronologies in Europe were devised from the systematic identification of glacial and inte rglacial events within the horizons of river terraces (Moore and Brumm ,

PAGE 46

46 2007). Studies of the greater Asian Paleolithic followed in this way, with the geomorphological chronologies patterned after the European Pleistocene terrace sequence in an attempt to develop a compatible prehistoric framework (Pope, 1997). However , while the age of stone tools defined by riverine context during the early 20 th century continue to figure into interpre tation of Paleolithic SE Asia, the various opinions and explanations f lowing fro m them have been left unanswered in favor of research conducted in cave and rock shelters (Anderson , 1997). The Pacitanian assemblages primarily occur in secondary context, and the few artifacts discovered in situ have not been recovered alongsid e associable fauna and human fossils, or chronometrically datable materials (Bartstra, 1984; Simanjuntak, 2004). Up to now, scholars have not reached consensus on the Pacitanian and continue to debate the authenticity of differing age estimations concerni ng the stone tools in question. For instance, in a geomorphological study of the surrounding Baksoka River landscape, Bartstra (1983) estimated that the artifacts eroding from fluvial terraces were likely terminal Pleistocene or Holocene in origin. Moreove r, Bartstra (1982) contends that the co occurrence of ÔPaleolithic types' at river localities as well as in ÔNeolithic' surface assemblages located away from rivers confounds the historical tendency to always associat e Pacitanian industries with Lower Pale olithic technologies. Interestingly, Bartstra (1982) further comments that the various sites that make up the Pacitanian tradition might be better explained as "different seasonal or occupational activities," which is a concept elaborated in a recently pro posed model of stone tool use by Moore and Brumm (2007) . In any event, the mixture of Pacitanian stone tools with elements of Neolithic industries (e.g. adzes) underscores the uncertainty that comes with

PAGE 47

47 determining the age of Ôold looking' Pacitani an ston e tools (Brumm and Moore, 2012; Simanjuntak , 2004). For the most part, opinions are still divided on the chronological placement of the Pacitanian industry and hinge on typological interpretations. On the one hand, it is problematic to link the Pacitanian industry with primitive Homo by extension of Early Pleistocene typologies because, as Corvinus (2004) has rightly pointed out, the heavy, cobble core tools that typify the Pacitanian are "part and parcel with the chopper/chopping tool complex," which by m ost accounts is wholly different than the Acheulean concept used in the Western Old World (2004:145). Sangiran industries. The ÔSangiran Dome' is an important geological feature in Central Java known for its complete Upper Pliocene through Middle Pleistoc ene stratigraphic sequence (Bouteax and Moigne, 2010). The Pucangan (Early Pleistocene) and Kabuh (Middle Pleistocene) sequences are particularly important, with the former having yielded H. erectus fossils sometimes described as archaic, and the later yie lding the largest concentration of classic Javanese H. erectus . The first Sangiran flake stone tools were discovered atop Ngebung Hill in gravels overlying the fossil bearing layers containing hominin and faunal remains, and hence were subsequently interpr eted as Middle Pleistocene in age (Corvinus, 2004). The Ngebung artifacts have twice been dated to the Middle Pleistocene [250 ± 70 ka (Suzuki et al., 1985) and ca. .9 ma (Saleki, 1997)], but have since been reinterpreted as Upper Pleistocene in age on acc ount of the ambiguous geological context (Choi and Driwantoro, 2007). In more recent excavations at Ngebung, 20 chalcedony artifacts from the Kabuh Beds (Middle Pleistocene) were recovered, which are claimed to be comparable to the Sangiran

PAGE 48

48 flakes and int erpreted to be the handiwork of H. erectus (Simanjuntak and SŽmah, 1996). However, the authenticity of the artifacts has been questioned on grounds that they may not represent true artifacts (Corvinus, 2004). Roughly 60 km to the east of the Sangiran site , crude chalcedony flakes were collected from high terraces above the Solo River near the Ngandong H. erectus fossils (2004:143 Ñ 144). However, any association of H. erectus with the Ngandong stone tools has been mostly unconvincing due to the indirect rela tionship between the artifacts and the geological context (Choi and Driwantoro, 2007). Nevertheless, others maintain that a direct association between late H. erectus and the Sangiran and Ngandong artifacts is possible, while also citing both the typologic al similarities in the flake industries and the corresponding geological contexts from which they were discovered (Corvinus, 2004). Remarkably, there are still no stone tool artifacts that have been securely removed from early Javanese H. erectus fossil c ontexts (Dennell , 2016 ). In fact, only recently have ISE Asia lithic assemblages actually returned dates confidently placed in the Early Ñ Middle Pleistocene (Brumm et al., 2010; Moorwood et al., 1998). Currently, the best evidence for early Middle Pleistoce ne stone tool use in archaic hominins occur in dated assemblages within Flores, Indonesia (Mata Menge ), which have produced ages consistent with the early Middle Pleistocene occupation of H. erectus in Java [.8 " 1 Ma (Moorwood et al., 1998 ; van den Bergh et al., 2016 ) ] . Recently, supporting evidence from the nearby Mata Menge assemblages, located on Flores, Indonesia, have recently been found in association with fossil evidence within a constrained layer dating to between .65 " .8 Ma.

PAGE 49

49 Flake and core dichot omy. An oft cited typologic al distinction in SE Asia Paleolithic technology is the simple dichoto my differentiating Ôflake based' industries from so called expedient Ôcore tool' industries. Within the literature this dichoto my is also referred to as the Ôc h opper/chopping Ð tool industries' and the Ô pebble and flake technocomplex' (e.g. Anderson , 1990; Bellwood , 1997) and has generated various interpretations concerning the nature of the patterning (e.g. Moore and Brumm , 2007). The conventional view on the me aning of this patterning has been driven by the association of core tools with putative archaic Homo populations belie ved to have occupied the region. The bulk of core tools are derived from open air river terrace contexts and therefore lack the chronometr ic dating necessarily to confirm their antiquity. By comparison, pebble and flake based industries are often discovered in cave and rock shelters (Moore and Brumm, 2007), which have provided a temporal distribution of these technologies primarily confined to the Lower Pleistocene. Although the typological scope of this distinction condenses lithic variability within SE Asia by their relative differences, a definitive cultural division between these industries are tempered by the high incidence of internal v ariation and significant overlap (Rabbet , 2012). Currently, the most compelling evidence toward identifying the first unambiguous lith ic "tradition" of SE Asia is represented by the rounded base point technologies of Java (Bellwood , 2007), and the flaked based Tolean industries of Sulawesi (Pasqua and Bulbeck , 1998), but the appearance and distribution of these technologies is poorly understood. Nevertheless, there has been no shortage of cultural descriptors used in reference to many regional variants of Paleolithic technology, especially with regard to the pebble and core tool industries and are the source of much Ôoutsider' confusion (Reynolds , 2007).

PAGE 50

50 The core tool industries, and to a lesser extent flake based industri es of insular and island SE Asia have been routinely interpreted as the technological equivalent of the African Oldowan "Mode 1" ( Clark , 1977). In ISE Asia in particular, H. erectus is fairly well represented by fossil evidence and have assisted the long held assumptions that the charact er of stone tools was indicative of hominin manufacture. Historically, the simple nature of core tool technologies and the typological affinities they share with classic Lower Paleolithic industries (i.e. Oldowan) has been the primary basis for using larg e, heavy core tools as population markers for the distribution of archaic hominins (Moore and Brumm , 2007). There is a compendium of core tool indu stries attributed.to H. erectus on typological grounds. Namely , those d ecidedly characteristic of the ÔPacita nian' from Java, or the ÔCabalwanian' of the Philippines, but also the undesignated bounty of core tools reported from across greater SE Asia. Similarly, finding a chronolog ical place for the Ôflake based industries has suffered from the same ambiguity be tween fossil evidence and s tone tools. In particular, the ÔSangiran' and ÔNgandong,' both small flaked based industries named for their purported affiliation with H. erectus bearing site s in Java, would be inversely related to the proposed dichotomy (Moore and Brumm , 2007). Adding to the typological confusion, "flaked based" industries are occasionally lumped together with the presumably older Ôcore tool' tradition (Bellwood , 1997), presumably due to the inferred shared technological simplicity of both indu stries. Further compounding the issue, the arrival of H. sapiens apparently had little impact on the stone tool kits of the relict hominin inhabitants. The lack of chronological resolution renders typological approaches inadequate in assessing hominin beha vior, and for that matter identifying the arrival of H. sapiens. The

PAGE 51

51 technological simplicity and continuity exhibited in SE Asia present interpretive challenges in contrast to the rather straightforward relationship between behavioral modernity and techno logical complexity seen in western parts of the Old World. However, analyses aimed toward the interpretation of this variability are far from being abandoned. In contrast to the fossil evidence, stone tools are widely abundant and on occasion have been fo und in, or around ancient Pleistocene g eological deposits. Unsurprisingly, stone tool artifacts continue to factor into the spatial distribution and chronological boundaries of primitive Homo across greater SE Asia, despite rarely being reported in any dir ect or peripheral association with the physical remains of archaic hominins . In lieu of direct fossil evidence, it would be particularly informative if a broad classificatory schema of stone tools could shed light on the various artifact forms of SE Asia w hich, in turn, could help suss out the archaeological manifestations of archaic and modern populations across the wider region Whether these typological differences can be reconciled with the biogeographical history of archaic hominins is ultimately still subject to debate, but the dichotomy itself has earned its stay as a useful point of contrast in a region in want of comparative attributes. Recently, an alternative explanation proposed by Moore and Brumm (2007) has suggested that the technological differ ences between flake and core dominated assemblages is spatially sensitive, and in fact represent different phases of transport within a sin gle reduction strategy. Even if it turns out that this patterning is different ends of a single technological spectru m, it still does not address how to tease apart behavioral differences between species.

PAGE 52

52 CHAPTER IV CONTEXT The possible coexistence of archaic species and modern humans in ISE Asia is controversial and many of the evolutionary parameters are, at presen t, ill defined . Nevertheless, the simple stone tool artifacts found here undoubtedly straddle this important evolutionary context within which mounting biological, ecological, or cultural pressures beset the dwindling archaic Homo populations at Flores, an d quite possibly multiple localities on Java. However, m apping out the technological landscape during this dynami c biogeographical period faces inherent difficulty in distinguishing AMH from archaic hominin species on typological grounds alone . At roughly the same period hominin occupation of Indonesia is ending on Flores, late surviving Neanderthal populations and H . sapiens cohabi tated parts of Western Eurasia. And r econstructing the lifeways of contemporaneous hominins in Paleolithic Europe has drawn hea vily on the study of transitional industries and the emergence of modern foraging behavior, although these archeological representations are not observable in every region (Bar Yosef, 2002). To that point, there is no other region more underrepresented in their artifact histories than the various Homo species of Pleistocene SE Asia. Attempts at defining the cultural, spatial, and biogeographical boundaries in the Pleistocene era have been slow to develop here, in large part because evidence of technological behavior in H. sapiens , H. erectus , or H. floresiensis is difficult to prove on a case by case basis. By comparison, it has become fairly evident that the sequence of technological change of the African and Eurasian Pleistocene broadly trends with the ev olutionary history of the genus Homo (e.g. see Ambrose, 2001; Clark, 1977; Foley, 1987; Foley and Lahr, 2003). For instance, the origin of Acheulean bifacial industries and the wave of H. erectus dispersals into

PAGE 53

53 Eurasia during the middle Early Pleistocene (Bar Yosef and Belfer Cohen, 2001), or the emergent complexity and diversity of Paleolithic tool forms in the Late Pleistocene and the underlying behavi ors which drive technological diversification among these industries . The later have often figured into the adaptive package of incipient H. sapiens and their subsequent geographic dispersal ( Mellars, 2007). However, recent studies have indicated that the forward leap in Late Pleistocene technological and behavioral innovation which popularize the notion o f the ÔUpper Paleolithic Revolution' does not uniformly emerge in the archaeological record. Scholars have cautioned against transposing behavioral correlates of one region onto another (Henschilwood and Marean, 2003). In fact, an increasingly popular opini on on Paleolithic variability is that these newly acquired modern behaviors of H. sapiens do not go entirely u nmatched by late surviving hominin populations . For example, scholars have recently reconsidered the possible involvement of Neanderthals in the e arly stages of the Aurignacian industry, which has been a key temporal diagnostic exclusive to H. sapiens in Europe (Conard et al., 2004). Nevertheless, the diversification and refinement of stone tool artifacts is still generally a very resourceful point of contrast in assessing behavioral variability across and within h uman species . Stone tools offer deep temporal resolution, gross measureable attributes, and ubiquity over time. But i ntriguingly, the study of lithic artifacts of the Late Pleistocene in SE Asia have not yet yielded any kind of technological pathway to behavioral modernity. H. sapiens of Pleistocene Southeast Asia The Upper Pleistocene period (MIS 4/MIS 3). is most notably associated with the rapid geographic expansion of Ôbehaviorally mode rn' H sapiens out of Africa which roughly coincides with the replacement of indigenous Neanderthal populations in Eurasia. The

PAGE 54

54 centerpiece of this evolutionary period Ñ the Middle to Upper Paleolithic transition Ñ has been the benchmark for describing the tim ing and cultural processes behind modern human colonization. While the dawn of this age (MIS 5) records the earliest evidence of an anatomically modern population in the Levant, archaeological and molecular data have preliminarily indicated that this initi al expansion beyond the African continent was likely a failed dispersal (Macualay, 2005; Mellars, 2006; Shea, 2008). Certainly, there is good evidence that behaviorally modern H. sapiens were equipped with widening range of complex tool forms and higher ca pacity for symbolic representations, ritual paraphernalia, elaborate body ornamentation, among other classic archaeological markers. However, the popular synthesis of cultural traits considered germane to distinguishing AMH from those of non modern predece ssors, as well as contemporaneous populations of non human hominins, are far from compatible across all Late Pleistocene cultures. Modern human expansion. The premise of the popular single dispersal model is largely predicated on the demographic expansio n of newly acquired adaptive behaviors, which quickly take hold in Africa as closely related populations are swiftly replaced, or integrated (Mellars, 2006). Despite never bearing anywhere near the full spectrum of modern behavioral markers, the Late Pleis tocene of SE Asia holds some of the earliest fossil evidence for modern humans outside of Africa c a. 42 Ñ 46 ka [Laos, ca 46 Ñ 63 ka (Demeter et al. , 2012; Demeter et al. , 2015); Sulawesi 46 Ñ 30 ka (Barker et al. , 2007)]. More surprisingly, the initial coloniza tion of Australia is now confidently estimated at ca. 55 Ñ 50 ka (Hiscock , 2008), with ge netic evidence suggesting an even earlier migration at ca. 62 Ñ 75 ka (Rasmussen et al., 2011). Such an early occupation of Sahul is somewhat unexpected, particularly beca use it is among the earliest

PAGE 55

55 evidence for modern humans outside of Africa [ca. 60 Ñ 40 ka (Mellars, 2006)], but also because it predates the earliest well dated fossil evidence for H. sapiens in SE Asia. The conventional explanation has contended that a cost al corridor facilitated a rapid migration along the Indian Ocean, essentially bypassing interior mainland (Dennell 2003; 2008). However, the archaeological evidence from Australia, not unlike that of SE Asia, occurs without as much as trace of the classic Ôinnovative package' that signals the advent of modern human behavior in Africa (Boivin et al., 2013). Despite the conflicting sequence, the earliest dates from both Sunda and Sahul are considered chronological significant in configuring a timeline for a single dispersal model. In contrast, the possibility that the first dispersal of modern humans was, in fact, a successful dispersal whereby a rouge band of early H. sapiens trekked much deeper into the Asia continent, is still considered by some to be a v iable, and understudied hypothesis. Dated archaeological material from ISE Asia continues to establish initial human occupation more firmly into the Late Pleistocene (O'Connor, 2007), but there have been a few indications that have hinted such an early col onization is possible. In South China at the upper margins of the SE Asia mainland , U series dating of a sk ull and post cranial elements were found to be no younger than 68 ka, but more likely ca. 111 Ñ 139 ka (Shen et al. , 2002), which if confirmed would i ndicate that an earlier population of H. sapiens was at the doorstep of the SE Asia sub region at the close of the Middle Pleistocene. Moreover, a small number of teeth (which have since been lost) were thought to exhibit dental elements of modern H. sapie ns , were found in association with the Punung faunal assemblage dated to 128 ± 15 and 118 ± 3 ka in Java, Indonesia ( Storm, 2005; Westaway et al. , 2007). H owever, the evidence for such an early chronology is at present very limited and questionable,

PAGE 56

56 and al so out of step with the robust archaeological visibility that follows in the Late Pleistocene of both mainland and ISE Asia (Rabbet, 2012) . The uncertainty which surrounds the initial colonization brings attention to the fact that human presence in SE As ia around the transition from the Middle Pleistocene to the Late Pleistocene is not fully discounted by scholars working in the field. Although it is merely speculated on at this point rather than offered as an alternative perspective, the major implicatio n within SE Asia is that it significantly widens an already confounding period of time of which there is considerable uncertainty regarding the behavioral differences and potential interaction between archaic hominin species and modern humans. The subtle footprint of behavioral modernity in SE A sia pre sents difficulties in ruling out the scenario, however unl ikely. Problematically, by whichever selective pressure primitive Homo were ultimately vanquished, modern humans were able to manage without major tec hnological innovation. The technological continuity from the Middle to Upper Pleistocene obfuscate a crucial line of evidence commonly used to reconstruct the movement of modern H. sapiens thro ughout the Old World. However , the evidence placing modern huma ns in Middle Pleistocene ISE Asia is razor thin and highly contentious, but remains an entirely possible evolutionary scenario . Late Pleistocene Archaic Hominins The fairly routine consideration of a late H. erectus extinction is mostly perpetuated by reg ional perspective on simple lithic industries in Java. To some extent, the good evidence for H. erectus persisting throughout the Middle Pleistocene is predicated on a historical understanding of typological variation that has fallen out of favor in recent years . While there is sizable and convincing artifact evidence of H. erectus' handiwork outside of SE Asia during

PAGE 57

57 the Early and Middle Pleistocene, the many Ôold stone tool' claims originating within the region are fraught with controversy and dispute (a s discussed in the previous Chapter). At present , the fossil evidence alone is not currently sufficient enough to address the circumstances surrounding a possible late extinction of H. erectus . Nonetheless, the debate surrounding the taxonomic relationship between archaic populations e.g., H. erectus , H. ergaster, H. hablis, and the remains of the small bodied hominin H. floresiensis ( Aiello, 2010; Argue et al. , 2007 ; Baab, 2016; Brown and Maeda, 2009; Dennell et al., 2014; Moorwood et al., 2004 ) hint at th e possibility of a remarkably dynamic and variable biogeographic hominin landscape throughout the Pleistocene SE Asia (Marwick , 2008). The clearest evidence of a relict hominin species surviving into the Late Pleistocene comes from H. floresiensis on Flor es, Indonesia (Brumm et al., 2010 ; Brumm et al. 2016; Moorwood, 2004; van der Bergh et al., 2016 ), which also happens to hold the only unambiguous evidence for stone tool use and manufacture in non human hominins of this period ( Moore and Brumm, 2007 ). Oth er scholars have presented radiometrically dated evidence suggesting that a late variety of Javanese erectus persisted into the Late Pleistocene ( Swisher et al., 1996; Yokoyama et al., 2008 ), while more recently Larick and Ciochon (2015) have hinted more a t the possibility of another small bodied hominin occupying the Philippines in the Late Pleistocene. However, the reconstruction of hominin biogeography throughout Pleistocene SE Asia continues to be treated cautiously; crucial information is still needed in order to establish the evolutionary histories of the early period and late period Javanese H. erectus, with the later formerly described as archaic H. sapiens (Hawk et al., 2000; Stringer, 1987) and the antecessors of H. floresiensis on nearby Flores Is land still unclear (van der Berg

PAGE 58

58 et al., 2016a). The chronological prime for each of these SE Asian fossil hominin populations remains an important, but m ostly unanswered question in the field of paleoanthropology. Paleoenvironmental Conditions The end of the Middle Pleistocene also coincides with the beginning of the last glacial period, which is marked by a global cooling trend and glaciation, which characterizes much of the Upper Pleistocene period. While generally marked by an overall temperature reduct ion, the climactic conditions during the Late Pleistocene were highly variable; regular climatic amelioration brought on rapid changes in sea levels, as well as local and regional level compositional changes to vegetation. The instability in climate would have introduced localized and regional wide ecological stress to modern humans and any re lic archaic hominin populations; but also, resulted in the formation of contiguous land bridges with the exposure of the Sunda shelf. At the height of the Last Glacial Maximum (LG M) when sea levels were at their lowest, the Sunda subcontinent would have included the southern islands of Borneo, Sulawesi, Sumatra, and to the north, the Philippine island of Palawan. Overall, the climatic fluctuations would have periodicall y linked chains of islands to mainland SE Asia , increased inter island visibility, and migratory opportunities would have vacillated with the peaks of glacial and interstadial periods . The tropical environment has long been viewed as a formidable ecologi cal barrier to both archaics and modern humans, but these limitations are largely informed by contemporary tropical foragers and varied ethno historical accounts from tropical SE Asia, the Amazon, and the Congo River Basin (e.g. Bailey et al., 1991). The r each and intensity of tropical rainforest coverage at various intervals of the Late Pleistocene are still regularly debated, but the

PAGE 59

59 prevailing opinion of many in the field is that the modern extent of the tropical rainforest in insular and ISE Asia is not analogous to environmental conditions during the Late Pleistocene. The characterization of Pleistocene SE Asia as consisting of a belt of tropical rainforest running east to west across the mainland has recently evolved into a more complex picture of mos aic habitats, with large tracts of savanna and open woodland bordered by peripheral concentrations of tropical rainforest (van der Berg, 2001; Bird et al., 2005). The archaeological evidence from the peninsular Thailand, Malaysia, and ISE Asia largely conf orm to this ecological description, with a good proportion of the mid Late Pleistocene faunal assemblages showing a reliance on auxiliary resources extracted from diverse habitats, including tropical rainforests (Conrad 2015; Roberts and Petraglia, 2015). Ecological challenges. From a purely ecological standpoint, the exploitative potential of the modern tropical environment suggests that closed canopy forests were uninviting and unproductive environments, which would have, or have historically required acc ess to alternative resource patches, or seasonal reliance on agricultural economies in order to sustain an adaptive strategy centering on tropical subsistence (Bailey et al., 1991). Given how persuasive the distribution of resources in the coordination of movements and subsistence strategies in H. sapiens , it has been long argued that modern humans would have found tremendous difficulty in striking a balance between the distribution of resources and tropical subsistence strategies without the contemporary r eliance on agricultural communities (Bailey et al., 1989; Bailey et al., 1991; Headland and Bailey, 1991; Kelly, 1995). Ecological stability is obviously paramount to the survival of human populations, and it is the relatively uniformity in the range of ad aptive responses employed by modern humans Ñ with relation to specific environmental

PAGE 60

60 constraints Ñ which provide the basis for ecological, or behavioral modelling used in this thesis. However, as discussed above, modeling hunter gatherer activities in prehis toric H. sapiens of SE Asia is difficult due to the environmental variability, and the little knowledge available to simply approximate adaptive strategies in this region. As was reviewed above, typological succession is not the cornerstone of assessing va riability in hom inin behavior in SE Asia What's more, wh ole assemblage appraisals Ñ often requir e details of the smaller but important elements that make up the bulk of an assemblage Ñ have been historically disregarded in search of more diagnostic artifacts (Reynolds, 1992). However, a number of comprehensive reports have been published over the years that have tilted subsequent studies in the direction of whole assemblage analysis, although no attempt has been made to synthesize these data sets under a singl e interpretative framework. Traditionally, the lack of regional specialization throughout the SE Asian Paleolithic has been explained in terms of limitations imposed by the natural environment. In a manner of speaking, this is an ecological explanation fo r a material cu lture phenomenon. But importantly, the notion of a raw material boundary has not advanced the study of lithics much further toward the explanatory framework of behavioral ecology. The archaeological reality that is the "Movius Line," and the arguments that closely follow the relationship between the geographic demarcations and technological division are more of convenient explanation than a theoretically guided conclusion. Clearly, the lack of fine grained raw material and the ample availabil ity of organic material impacted the intensification of stone tool technology. But it is also apparent from the numerous bone and shell technologies that the understanding of

PAGE 61

61 fracturing mechanics necessary for elaborative artifact forms is not contingent o n high quality lithic material. The technological intensification of non lithic materials demonstrate d the propensity for Pleistocene H. sapiens to produce standardized forms, suggesting that the general lack of fine grained raw material was not an overri ding restriction in the pro duction of complex technology. By comparison, the adaptive strategies of contemporaneous European Late Pleistocene groups are strongly directed toward locating a limited number of select fine grained raw materials (Riel Salvatore et al. 2008; Rabbet 2012). In this sense, access to raw material quality and quantity were important factors guiding hunter gatherer landscape use and this type of landscape interaction has been shown to possess traceable archaeological patterning (Andref sky 1994; Parry and Kelly 1987; Torrance 1983). Comparatively, Pleistocene hu nter gatherers of SE Asia are often characterized as technological opportunists; in contrast to their Western Eurasia counterparts, it appears tha t prehistoric inhabitants of SE A sia were guided to lithic raw material by the specific activities they carried out on the landscape (Rabbet 2012). In western Eurasia, Pleistocene hunter gatherer landscape use was likely dictated by narrow and unpredictable resources and fewer subsistence options while foragers in the tropics would have likely needed to tailor their adaptive system in response to the fo rces of resource diversity (Rabbet, 2007) Binford (1980; 1982) suggested that foragers subsisting in regions of high primary mass, such a s the tropics, are much more inclined to practice high residential mobility by moving frequently between centers of foraging radii. Kelly (1983) demonstrated the relationship between environment and forager mobility strategies on a global scale. However, i t has been noted that tropically adapted hunter gatherers, which rely on agricultural communities to offset key deficiencies in carbohydrates

PAGE 62

62 and fats, may not reflect an accurate representation of mobility and economic strategies practiced by prehistoric hunter gatherer populations in SE Asia. Currently, very little research has been carried out in an attempt to reconcile what such a behavioral strategy would result in archaeologically. While the research presented in this thesis cannot resolve these que stions relating to tropical or otherwise adaptive strateg ies unique to the SE Asian Paleolithic and how they might manifest archaeologically, the data compiled for this project can directly test the possibility that, in fact, lithic technology was integrat ed into the mobility strategy despite the deemphasized aspects of fine grained lithic materi al and formal tool production. T he exploitative potential of the modern tropical environment suggests that closed canopy forests are unproductive environments for foraging, and would have likely been a formidable barrier to archaic and modern humans. In part, this has been supported by the thorough review of historical and contemporary ethnographies on foragers subsisting in interior rainforests indicating they requ ire access to alternative resource patches, or seasonal reliance on agricultural economies in order to sustain an adaptive strategy centering on tropical subsistence items (Bailey et al., 1989; Bailey and Headland, 1991). However, the reach and intensity o f tropical rainforest coverage at various intervals of the Late Pleistocene are still regularly debated; the prevailing opinion of many in the field is that the modern extent of the tropical rainforest in insular and island Southeast Asia is not analogous to environmental conditions existing during the Late Pleistocene. However, the characterization of Pleistocene Southeast Asia as consisting of a belt of tropical rainforest running east to west across the mainland has recently evolved into a more complex picture of mosaic habitats, with large tracts of savanna and open woodland bordered

PAGE 63

63 by peripheral concent rations of tropical rainforest (van der Berg 2001; Bird et al. 2005). The archaeological evidence from the peninsular Thailand, Malaysia, and island So utheast Asia largely conform to this ecological description, with a good proportion of the mid Late Pleistocene faunal assemblages showing a reliance on auxiliary resources extracted from diverse habitats, including tropical rainforests (Conrad 2015; Rober ts and Petraglia 2015). Despite the complexities of the Southeast Asian paleoecology, recent and earlier approaches have turned to modeling hunter gatherer mobility around "peaks of abundance . " Evidence suggests that a savanna like corridor facilitated r apid colonization of the Southeast Asia peninsula. While open savanna like land access may have opened up accessibility to more segregate and predictable resources, it was likely a constricted and competitive habitat for exploitation. Dependence on auxilia ry resources appears to be the norm, not the exception. Comparatively, resource distribution in the modern tropics is somewhat easy to place on a behavioral ecology map. Resources are unevenly distributed far and wide, and tropical primary biomass is ric h in diversity and abundance but mostly inaccessible for animals, and terrestrial secondary biomass is consequently unpredictable, diverse, and solitary . Site Background Currently, the origin and def inability of the SE Asian Late Pleistocene is not easily described in relation to stone tool industries. The title ÔPaleolithic' is referenced often within the literature, but is used more tacitly as a general category for Ôold stone tools' rather than a temporally significant sequence . For the purposes of maki ng a chrono logical distinction, the ÔUpper' refers vaguely to a perio d upon which modern humans begin occupying greater SE A sia . However, from a technological perspective, it is often as challenging to pinpoint a

PAGE 64

64 transition to Holocene stone tools as it is discriminating between the stone tools of the Middle and Late Pleistocene. At best, the typology of the SE Asian Upper Paleolithic is nebulously defined by the manufacture of informal tools, that is, tools which show little evidence of elaboration, functi on, or standardization in form. With the possible exception of a slight increase in flake based assemblages among modern humans occupying rock shelters in the Terminal Pleistocene, very little is diagnostic about the stone artifacts themselves. And while c ountless regional and cultural variations have been recognized, ultimately their uniqueness is diluted by the high degree of morphological overlap shared by various other industries, making it difficult to gauge any long t erm distribution of artifacts Cons equently, among the assemblages included in this study none are clearly linked with a robustly defined industry. Instead, the lithic assemblages are treated as a cohesive sample of assemblages which exhibit shared characteristics which permit their inclusi on in this analysis. Chiefly, the informal quality and either their origin from in archaeological context which indicate a hunter gatherer mode of survival.

PAGE 65

65 CHAPTER V M ETHODOLOGY The methodology used in this thesis was originally developed by Barton ( 1998) in order to link assemblage scale variability with patterns in technological behavior of the Late Pleistocene on the southern Iberian Peninsula (Villaverde et al., 1998). The focus on whole assemblage variability provides a simple measure of changes in techno economic behaviors and shifting land use strategies over long periods of time (Riel Salvatore and Barton, 2004). To date, this approach has continually demonstrated a simple and repeatable method for distinguishing between various patterns of for ager mobility and lithic organization at the local and regional scale, and recently at the continental level, with a wider group of assemblages incorporated from the whole of west Eurasia (Barton et al., 2011). Previous applications of this model have bee n used to demonstrate continuity between Middle and Upper Paleolithic land use strategies in southeastern Italy (Riel Salvatore 2007; Riel Salvatore and Barton 2004; 2007; Riel Salvatore et al. 2008), Gibraltar and eastern Spain (Barton 1998; Villaverde et al. 1998), which was later found to hold at the inter regional level of analysis (Barton et al., 2011). Thus, this model is particularly well tailored to evaluate adaptive behavior for fundamental differences in supposedly behaviorally transitional period s in Eurasia (i.e. Upper Paleolithic), where traditional typological divisions are not necessarily considered a clean break from archaic to behavioral modernity. In light of these findings, this model should have appeal in the study of the SE Asian Paleol ithic Ñ a spatiotemporal context where typological ambiguity has hindered efforts to develop culturally definable sequences, and where many questions of adaptation and

PAGE 66

66 technology are entangled with issues of environment. The methodology described br iefly is oriented toward addressing the questions outlined in the specific aims and are reviewed below: Whole Assemblage Behavioral Indicator The basic requirements for this model (WABI) are count data for the whole lithic assemblage including retouched pieces, c ores, and debitage, in addition to the total volume for each corresponding unit, layer, or horizon from which the artifacts were retrieved (Riel Salvatore and Barton, 2004). Here, the formal properties of retouched tools are deemphasized, and instead these data establish a relative frequency that is compared to artifact volumetric density. Following Riel Salvatore and Barton (2004: 259), "artifact volumetric density is defined as the total number of pieces of chipped stone per cubic meter of excavated sedim ent" and "...the relative frequency of retouched pieces is simple the count of retouched Ôtools'...divided by the total number of pieces." Due to the complex depositional environment of SE Asia rock shelters and the temporal scope of the project, calculati ng actual sedimentation rates was essentially required in order to make direct comparisons between assemblages. Time is not often reported as precise data, so the preferred chronological sequence provided by the authors often involved a combination of radi ometric dates and relative dating techniques. Nevertheless, the chronological data were broken down in relation to discrete assemblage data, which allowed differential artifact accumulation rates to be estimated. Evaluating retouch. The model presented a bove (WABI) is part of a larger family of quantitatively driven techniques which hinge on the identification and interpretation of retouching intensity. In response to the classic forager studies by Binford (1979, 1980), retouch indices were developed to d emonstrate that assemblage patterning were strongly influenced technological

PAGE 67

67 organization, mobility patterns, and land use strategies (Andrefsky, 1991; Bamforth, 1986; Bleed, 1986; Hiscock, 1994; Kuhn, 1991). Despite there being a general consensus on the analytical value of retouch in monitoring these dimension of human behavior, concerns on how best to measure retouch intensity with efficiency and reliability have been raised (Andrefsky 2005; Clarkson, 2002). Studies examining the most effective metrics in analyzing reduction (i.e. retouch) intensity have been put forward (e.g. Clarkson 2005; Hiscock and Clarkson 2005a, Hiscock and Clarkson 2005b). Notwithstanding these concerns, it has been noted that these diagnostics are ill suited measurements for ÔH oabinhian' assemblages of SE Asia because of the unstandardized tool form and low proportions of retouched tools (Marwick, 2007), which unofficially qualifies as a Litmus test for gauging methodological compatibility with almost any Pleistocene SE Asia ass emblage. Other issues stem from the straightforward identification of retouched flakes and core tools, which lacking continuity in standard measures such as size and shape, are difficult to distinguish from generally unintended waste products such as flake shatter and exhausted cores. Consequently, the nature of chipped stone tools in this regional context presents methodological challenges in interpreting retouching intensity that have been largely unstudied. Despite the concerns raised above, the WABI m ethodology varies from conventional approaches to lithic reduction strategies in significant ways. The simple requirement of retouched pieces irrespective of tool shape or form is particularly well suited for the ambiguous tool morphologies common to SE As ia. The most problematic issues inherent in this approach is related to representation; specifically, whether minimal proportions of retouched tools will

PAGE 68

68 be substantial enough in number to show meaningful Ôtechnological separation' through inter assemblag e patterning, or if the nature of the variability appears to be relatively stochastic. Lithic technological organization. The curated expedient continuum of lithic organization has generally been framed as a corresponding technological trend which roughl y aligns with a particular mobility strategy of hunter gatherer groups. One the one hand, curated industries are characterized by heavily modified lithics, core reduction is usually extensive, and retouched tools can be larger in size and range in shape Ð a llowing flexible modification of stone tools as access to raw material diminished. In Pleistocene hunter gatherers, this type of lithic organization appears to have been favored by highly mobile foragers and so consequently would be most often represente d in ephemeral archaeological contexts. Comparatively, high densities of artifacts and low frequencies of retouched tools characterize expedient assemblages. Retouched tools are often function specific and are manufactured for immediate use. Raw material i s either locally abundant or raw material procurement is embedded within the use of the site; this type of technological organization is indicative of less mobile groups. These lithic qualities would be most often represented in archaeological contexts wi th repeated and sustained occupation such as a residential camp. However, this type of relationship becomes difficult to demonstrate when working with disparate stratigraphic contexts, especially when it is unknown how rapidly assemblages formed. The meth odology used in this analysis builds on simple approaches to assessing curation that minimize detailed accounts of the tools themselves and instead, analyze the collective frequency of retouched artifacts in relation to the density of all other artifacts w ithin a given assemblage.

PAGE 69

69 Collating d ata. The data compiled for this study come from 17 sites acr oss greater SE Asia region (n=64 assemblages), with the majority of assemblages coming from ISE Asia, and a comparably small but still modest sample coming from the Malay Thai peninsula within the mainland (Table 2) . The simple informal character of stone tools permits the inclusion of diverse assemblages originating from a wide range of spatiotemporal contexts into this model. Although it would be preferable to more narrowly focus on the assemblages of Pleistocene foragers of SE Asia in keeping with the previous applications of this model, there are only a small number of well dated Pleistocene sites with lithic artifacts, and far fewer had published the requ ired data necessary for this model. In order to expand our database and generate a more robust sample size for analysis, assemblages from the Early and Mid Holocene have been provisionally incorporated into the general forager model under the assumption th at Holocene technological organization differed only marginally from the former Pleistocene occupants of SE Asia, though their mobility strategies tend to favor longer sustained use of base camps. Analytical considerations. In deciding to test the relation ship between mobility and assemblage patterning while assuming continuity in technological organization throughout all of Pleistocene SE Asia, much of the archaeological content and occupational histories of these assemblages were deemphasized in accordan ce with the generalized parameters of the WABI methodology. Many of these aspects were considered too cumbersome to incorporate, and ultimately, they are highly important datasets to return to follow ing the results of this model. However, the regional and temporal latitude of this project dictates that such an undertaking is likely beyond the scope of this thesis. Nevertheless, there were a number of selected variables which were

PAGE 70

70 relatively easy to code and "keep tabs on" without fundamentally altering the data or straying too far from original questions driving this research. These variables amounted to series of analytical considerations which were applied as data filters to test their effects on the output. In some instances, it merely warranted a clos er look at outliers, but others were thought to be pertinent to interpreting the results. The full table of analytical conside rations is presented above (Table 1 ) and the reasoning behind their incorporation is briefly discussed here. The relatively small percentage of Holocene assemblages with overtly specialized elements, namely unifacial points, are atypical in SEA lithic assemblages. Unifacial points are not found in late Pleistocene assemblages, and their presence in the Holocene assemblage is minimal . The second, more concerning issue with including Holocene assemblages is that the chronological boundary between strictly forager and predominately agricultural adaptive strategies is not often starkly defined by depositional changes in material culture , and recent ! !"#"$%&'#($$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$ $ ! "))(*+'",() ! $ %&'#($ " ! #"$%&'()$%!*+",(-./$'(+" ! 01#!2#334 ! 56 ! ! * +//.%$'(+"!7.'8.."!#9:!$"-! /.'+;);.")(., ? ! • ! @+-./"!<;A$"!,$AB%.! C./,;,! !"#$%&'()*(+)*) ! ! D/.-()'(C.!,'/."E'
PAGE 71

71 review of the archaeological evidence for emerging "Neolithic" farmers of ISE Asia at ca 4,500 years ago were found to be limited and inconclusive (O'Connor, 2015). These Holocene assemblages have generally been interpreted under similar ambig uity regarding site function, settlement configuration, and the role of lithic economy during the transition to agricultural communities. This makes it difficult to address the spatiotemporal expansion of agriculture into SEA as prehistoric societies conti nue to utilize raw stone material, or at least manufacture stone tools using local methods comparable to early Holocene and Late Pleistocene hunter gatherers. Given that our main concern is the relationship between mobility, assemblage patterning, and the relative position of modern versus archaic groups rather than subsistence, the mid late Holocene assemblages should be emerg ing "Neolithic" farmers of ISE Asia at ca 4,500 years ago were found to be limited and inconclusive (O'Connor, 2015). These Holocene assemblages have generally been interpreted under similar ambiguity regarding site function, settlement configuration, and the role of lithic economy during the transition to agricultural communities. This makes it difficult to address the spatiotemporal expansion of agriculture into SEA as prehistoric societies continue to utilize raw stone material, or at least manufacture stone tools using local methods comparable to early Holocene and Late Pleistocene hunter gatherers. Given that our main concern is th e relationship between mobility, assemblage patterning, and the relative position of modern versus archaic groups rather than subsistence, the mid late Holocene assemblages should be included to provide greater depth to the analysis that can be viewed, at least tentatively, as a technological system closely aligned with earlier periods , therefore representing another important source of data in illustrating the usefulness of the whole assemblage approach in understanding subtle variation or alternatively, c ontinuity in the lithic technology for the region. Lastly, the assemblages of Liang Bua and Song Keplek

PAGE 72

72 are evaluated against the broader regional data set, both which are the strongest evidence for stone tool use in non modern humans near the end of the P leistocene. This is done to explicitly compare the assemblages of archaic to modern groups and evaluate the possibility of using lithic assemblages to differentiate hominin species in SE Asia. Statistical interpretation. The relationship between AVD and r etouch frequency across these assemblages are tested using a linear regression analysis, which was coded in the statistical program Rstats. As mentioned above, the diversity of assemblages included in this analysis required that a series of computations we re carried out, as to allow the influence of specific sets of assemblages to be independently tested. The linear regression analysis is interpreted through the strength of the R square, and R value, and the closer to 1 and 1 the value the stronger the rel ationship. These outputs are described in relation to a number of graphs plotting the distribution of the assemb lages. Thus, the statistical analyses run in this study are aimed at explaining the variability through a linear regression analysis, specifical ly by assessing the correlation between AVD and retouch frequency as described by the R value, and testing the predictability of these assemblages to conform along the negative linear regression by evaluating R square.

PAGE 73

73 Table 2 : Su mmary of sites and assemblages ! ! ! ! ! ! "#$% ! &'(') ! *'+,-. ! /'0,11,10 ! %1.,10 ! &-234,-1 ! !"#$%&'#$"()* $ # $ &+,-.$/0-01232 $ 4566 $ 7566 $ &+89$):;0, $ !"#$%&'#$"()* $ ## $ &+,-.$/0-01232 $ 7566 $ <566 $ &+89$):;0, $ !"#$%&'#$"()* $ ### $ =:>$/0-01232 $ <566 $ 5566 $ &+89$):;0, $ !"#$%&'#$"()* $ #? $ =:>$/0-01232 $ 5566 $ @566 $ &+89$):;0, $ !"#$%&'#$"()* $ ? $ =:>$/0-01232 $ @566 $ A566 $ &+89$):;0, $ !"#$%&'#$"()* $ ?# $ B+92$/0-01232 $ A566 $ C566 $ &+89$):;0, $ !"#$%&'#$"()* $ ?## $ B+92$/0-01232 $ C566 $ D566 $ &+89$):;0, $ !"#$%&'#$"()* $ ?### $ B+92$/0-01232 $ D566 $ 756 $ &+89$):;0, $ !"#$%&'#$"()* $ E F #E $ B+92$/0-01232 $ 756 $ 6 $ &+89$):;0, $ B#&$G#'# $ ### $ &+,-.$/0-01232 $ 7<66 $ 7C66 $ &+89$):;0, $ B#&$G#'# $ #? $ &+,-.$/0-01232 $ 7A66 $ 7666 $ &+89$):;0, $ B#&$G#'# $ ? $ &+,-.$/0-01232 $ 7666 $ <766 $ &+89$):;0, $ B#&$G#'# $ ? H%I $ =:>$/0-01232 $ <666 $ 5566 $ &+89$):;0, $ B#&$G#'# $ ?# $ =:>$/0-01232 $ @A66 $ A766 $ &+89$):;0, $ "(#$!*!*$D $ #$H+JKJI $ &+,-.$/0-01232 $ L666 $ 7566 $ &+89$):;0, $ "(#$!*!*$D $ ##$H+JKJI $ =:>$/0-01232 $ <666 $ 5666 $ &+89$):;0, $ "(#$!*!*$D $ ###$H+JKJI $ =:>$/0-01232 $ @C66 $ A566 $ &+89$):;0, $ "(#$!*!*$D $ ###$H1JI $ B+92$/0-01232 $ AC66 $ CA66 $ &+89$):;0, $ "(#$!*!*$D $ #?$H+JKJI $ B+92$/0-01232 $ C566 $ C666 $ &+89$):;0, $ "(#$!*!*$D $ ? $ B+92$/0-01232 $ C666 $ D<66 $ &+89$):;0, $ "(#$!*!*$D $ ?# $ B+92$/0-01232 $ D<66 $ DC66 $ &+89$):;0, $ "(#$!*!*$D $ ?## $ B+92$/0-01232 $ D666 $ 766 $ &+89$):;0, $ "(#$!*!*$D $ ?### $ B+92$/0-01232 $ 466 $ 566 $ &+89$):;0, $ "(#$!*!*$C $ #? $ &+,-.$/0-01232 $ 4C66 $ 7@66 $ &+89$):;0, $ "(#$!*!*$C $ ? $ &+,-.$/0-01232 $ 7666 $ $/0-01232 $ <566 $ 5L66 $ &+89$):;0, $ "(#$!*!*$C $ ?## $ =:>$/0-01232 $ $/0-01232 $ 5C66 $ 5666 $ &+89$):;0, $ "(#$!*!*$C $ #E $ =:>$/0-01232 $ @A66 $ A466 $ &+89$):;0, $ "(#$!*!*$C $ E $ =:>$/0-01232 $ A466 $ AC66 $ &+89$):;0, $ "(#$!*!*$C $ E## $ B+92$/0-01232 $ D766 $ D@66 $ &+89$):;0, $ M'#($%(?&$ $ ## $ =:>$/0-01232 $ 5D6< $ @A<7 $ N2O$PQ:32+ $ M'#($%(?&$ $ ### $ =:>$/0-01232 $ 55<7 $ 5D6< $ N2O$PQ:32+ $ M'#($%(?&$ $ #? $ =:>$/0-01232 $ 5LA< $ 55<7 $ N2O$PQ:32+ $ M'#($%(?&$ $ ? $ &+,-.$/0-01232 $ 7DA< $ 5LA< $ N2O$PQ:32+ $ )*&$%(?&$ $ # $ R-2:8901232 $ C<666 $ <666 $ N2O$PQ:32+ $ )*&$%(?&$ $ ## $ =:>$/0-01232 $ <666 $ A666 $ N2O$PQ:32+ $ "B"$B&(NP $ # $ =:>$/0-01232 $ @<56 $ A@76 $ GQ-+O28: $

PAGE 74

74 ! ! ! ! ! ! !"#$ % &'(') % *'+,-. % / '0,11,10 % $1.,10 % &-234,-1 % "#"!#$%&' ! ( ( ! $)*+,!-.+./010 ! 2344 ! 3344 ! 56+)7089 ! "#"!#$%&' ! ((( ! $)*+,!-.+./010 ! :;34 ! 2344 ! 56+)7089 ! #(%&'!#$<=">" ! ? .10! (( ! @+098A./010 ! BC:B4 ! :D33 ! %*6!(8+)1E8F!(&= ! #(%&'!#$<=">" ! ?.10!((( ! @+098A./010 ! B22G3 ! BC:B4 ! %*6!(8+)1E8F!(&= ! #$%&'!5%HH" ! ( I 9) J ! -.+./010 ! B4444 ! G444 ! K)+)6E!(8+)1E8 ! #$%&'!5%HH" ! ( I 9L J ! @+098A./010 ! DB444 ! BG444 ! K)+)6E!(8+)1E8 ! #$%&'!5%HH" ! ((( ! @+098A./010 ! C3444 ! CD444 ! K)+)6E!(8+)1E8 ! @(%!-"=%#$ ! (F((F((( ! @+098A./010 ! BB344 ! B4444 ! H.A9!(8+)1EF!(&= ! (##$!M%N$ ! CC; O G42 ! @+098A./010 ! BB444 ! :;44 ! @P9+9QQ9108 ! 5R&'!S$@#$S ! (LF((F((( ! <9E!-.+./010 ! T444 ! ;344 ! U)V)F!( &= ! "% ! ((( ! ! @+098A./010 ! B44444 ! :3444 ! [+.*08F!(&= ! #(%&'!>"% ! (N ! @+098A./010 ! 2;444 ! TB444 ! [+.*08F!( &= ! #(%&'!>"% ! N ! !!! @+098A./010 ! 33444 ! 34444 ! [+.*08F!( &= !

PAGE 75

75 CHAPTER VI R ESULTS The early indications from these statistical analyses give the impression that overall, assemblages across SE Asia conform to the expectations of the model outlined in the previous chapter. The se ries of graphs depicted below correspond with the results of the multiple data filters (sub sets of data) which are presented below in Table 2 and are also described in more detail in the previous chapter . Due to the contrasting relationship between certai n data filters, there are fewer graphs than there are statistic al outputs. This arrangement allows smaller data samples to be more easily viewed as a cross section of the whole aggregate of assemblages. In other words, the graphs discussed below correspon d with various data fi lters (pre sorted assemblages), which have implications across multiple sub sets of assemblage data depending on the analysis. Additionally, the assorted sub data are presented against the backdrop of the entire data set as to allow e asy contrast to the wider sample. The results presented here generally correspond with the four major aims discussed in Chapter 1 , and have helped influence the presentation of statistical representations relevant to those questions . However, while the results presen ted here are directed toward the specific aims of the project, the results are further broken down in an effort to account for biases created while the data was collated. Thus, the arrangement of Figures and Tables in this chapter reflect an evaluation of the methodology and compatibility of specific data sets concurrent with the results from each new series. To begin, the overall conformity of SE Asia lithic assemblages with the expected negative relationship between artifact density (AVD) and retouched tool frequencies is reported, which provides a benchmark for comparing assemblages across human species in

PAGE 76

76 Pleistocene SE Asia. However, before proceeding with these analyses, specific sub sets of assemblages are identified as potential outliers, or ot herwise at odds with the overall assemblage patterning. Among those identified are assemblages with specialized point technology (SPT), ostensibly high frequencies of retouch, and the potentially semi sedentary Holocene assemblages of SE Asia. After scruti nizing these sub sets of data, the most representative sample of Pleistocene era forager strategies, unsurprisingly, comes from exclusively Pleistocene assemblages. However, in spite of the various effects of the many data configurations listed in Table 3 , we think the greater sample is illustrative of a common technological strategy employed by AMH prehistoric populations of SE Asia. Therefore, more than one data series is emphasized to contrast assemblage pattering across certain non modern assemblages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able 3 : Summary of Results

PAGE 77

77 G eneral findings . The results of the analysis demonstrated a moderately high correlation between AVD and retouch frequency across all assemblages ( r = .61). In contrast, when only those lithic assemblages created by modern humans were incorporated the data set showed a stronger, fairly high correlation between AVD and retouch frequency ( r = .75). Directly above in Figure 2 is a representation of the general variability along a negative regression. Among the modern assemblages, the data set conformed w ell with the expectations of the model ( r 2 = .57), although with some noticeable outliers in the upper left corner and bottom left center of Figure 1. By comparison, the Liang Bua assemblages, which are part of a slightly larger group of assemblages consid ered non modern in this analysis, produced a weak positive linear relationship which deviates significantly from the predicted negative relationship outlined in this model [Figure 3 ( r 2 = .045)]. The noticeable outliers in the AMH data come from the site Figure 1: All Assemblages Figure 2 : assemblage patterning across all sites

PAGE 78

78 o f Lang Rongrien, Thailand, particularly the three assemblages with ret ouch frequencies > 25% (Figure 5 ), as well as the assemblages with SPT [ Figure 4 (n =6)] from ISE Asia. These SPT assemblages from the sites of Uai Bobo 1, East Timor, and Ulu Leang, Su lawesi, were also noted for potentially distortin g the Holocene series . Adherence to this model is most apparent in the Pleistocene and Early Holocene phases, which exhibits the strongest predictive statistical relationship and steadily declines across the Middle and Late Holocene data sets. Modern human assemblage patterning. The results from this study indicate that the directionality of the linear regression is correlates well with a negative relationship between AVD and retouch frequency across all AMH data sets except for the Middle Holocene [Table 3 ( r = .37)], excluding those series which incorporated the assemblages from Liang Bua (see Table 2 for multiple r values associated with Liang Bua). However, as alluded to above, there are a number of asse mblages within the modern human sample which stand out as potential outliers driving the strength of the overall assemblage patterning. The first consideration were the assemblages in the upper left corner [Lang Rongrien assemblages with retouch (Figure 3 ). These data were removed to test their influence on the overall statistical output. Indeed, the predictive strength in modern assemblages was noticeably impacted, with the predictability of variability in the negative regression dropping sharply after re moving these assemblages from the analysis [( r 2 = .43), which is even lower than uncalibrated output ( r 2 =.47)]. These results indicate that overall, the moderately high

PAGE 79

79 predictive strength between retouch frequency and AVD in modern assemblages ( r 2 = .5 7) is not equally represented across the data set. The subsequent analyses are represented by a new baseline which excludes the outlier assemblages (retouch > 25 %) from the site of Lang Rongrien. This core data sample is used in a number of this is subseq uent data series in order to compare modern human assemblage patterning across temporal, species, and technological boundaries without the possibility of outliers driving th e negative relationship (Table 3 ). Notwithstanding the impact of anomalous retouch frequencies on the statistical strength in predicting retouch frequency and AVD, the remaining modern assemblages exhibit a moderate to low predictive strength [ r 2 = .43 (n = 56)], which broadly defines the core patterning unique to SE Asia modern assembl ages. Furthermore, the core patterning predicted by the model (described with r 2 ) is shown to hold across nearly all of the remaining data filters, with the notable exceptions being the inclusion of the Liang Bua assemblages, as well as the Figure 2: Liang Bua Figure 3 : Assemblage patterning at Liang Bua

PAGE 80

80 Mid Ñ Late Holoce ne assemblages. For example, when the Liang Bua assemblages were refigured into the new baseline sample [Figure 3 (n = 59)] the predictive strength of the linear regression was again steeply dropped (r 2 = .20). Directly above in Figure 3, the assemblages fr om Liang Bua can clearly be identified as extreme outliers in relation to the larger data sample. Specialized point technology. While the Pleistocene assemblages created by H. floresiensis were predicted to vary from those of AMH, other filters exposed the impact of aggregated data within the AMH sample which were unanticipated. For example, the sub grouping of assemblages with overt specialization in point technology carried a slightly higher retouch frequency in relation to AVD, although still adhering st rongly with the expected negative linear relationship. As such, these data were also considered for potentially biasing some of the smaller data sets, particularly the Middle and Late Holocene series. Figure 4 (above) identifies the slightly elevated retou ch frequencies within this sub set, drawing attention to the possible impact of this data concentration on other assemblage assemblages. In particular, the Holocene sample could be affected more acutely because all assemblages with specialized tools came f rom this period. Overall, after removing the assemblages from the larger grouping, the predictive strength of the pattern returned with a moderately high outcome ( r 2 = .58), indicating that SPT assemblages may hover above the core pattern, slightly decenter ed from the expected negative linear trend (see trend line on Figure 4) seen among the larger collection of modern human assemblages seen across SE Asia. Additionally, when the assemblages with specialized elements were isolated and tested independently, t hese demonstrated the strongest predictive

PAGE 81

81 strength among the sample [ r 2 = .92 (n = 6)], although, this series is the smallest of any aggregate sample tested. These findings are not altogether unexpected since culturally imposed forms are virtually non existent across SE Asia, or at least they are not evident to many archaeologists, and in a region nearly devoid of retouched tools, occurrences of these specialized forms may be expected to slightly impact the core patterning across SE Asia, which is the c ase here when the assemblages with unifacial and Maros points from Ulu Leang, and Uai Bobo are included ( r 2 = .43) versus when they are absent ( r 2 =.57). The apparent impact on the negative patterning in the wider sample is likely magnified in the smaller dat a sets from the Middle and Late Holocene series. !" !#$%"&'() !*+ !*+ !*+ !',%*-.-%+ /0+ + +/ +// /0+ + +/ +// +/// +//// 1&2-$34&5%2--#6%7 892,:'32%;-#<&29,3%=&(6,2>%?< @ A <-5&9( (-( B <-5&9( C-,(2%2&340 Figure 4 : Assemblage patterning in specialized point technology

PAGE 82

82 Chronological changes. The assemblages were further tested in data series which were bound by chronological periods. Of particular importance were the modern human assemblages bracketed by the Late Ple istocene phase, and how they compared to the modern human assemblages which occurred during the Holocene phases. The Pleistocene assemblages which formed at the terminal Pleistocene and prior to the Last Glacial Maximum (LGM) were found to be the most cong ruent sample with respect to fitting the expectations of the model ( r 2 = .73). Notably, the Pleistocene assemblage patterning also held even when controlling for the effect of outliers, specifically the assemblages of Lang Rongrien [( r 2 =.65) Figure 5]. By comparison, the Holocene assemblages did not conform well to the expected negative relationship between AVD an d retouch frequencies . It should be reviewed, though, that this model is designed for true hunter gatherer populations, an d, as there is some unce rtainty whether the mid Holocene Figure 4 : Modern Assemblages II Figure 5 : Assemblage patterning without Outliers

PAGE 83

83 and Late Holocene assemblage localities were truly adapted around hunter gathering. The Middle and Late Holocene assemblages were to be treated with caution until more is known about lithic organization and settlement strat egies in "Neolithic" Southeast communities. Nevertheless, it was our contention that some measure of logistical mobility which characterizes Pleistocene forager lifeways would have been evident in assemblage composition under the pretense that settlement s trategies during the Middle Ñ Late Holocene of SE Asia. The results discussed above diminish our confidence in the compatibility of the majority of the Holocene assemblages. Conversely, the weak relationship exhibited across Holocene assemblages indirectly strengthen the findings and suitability of this methodology in Pleistocene archaeological contexts. It is worth noting that the model used here (WABI) is an interpretive framework designed to draw inferences on landscape use and adaptive strategies in Plei stocene era hunter gatherers. In retrospect, the expectation that SE Asia Holocene survival strategies were fundamentally similar to Pleistocene foragers has little to do with physical evidence, but rather the lack thereof and hence the inability to differ entiate these populations from former inhabitants. While the low predictive strength found across the majority of Holocene sites certainly calls into question their inclusion within the wider sample, the potential biasing effects of the SPT may need to be reconsidered in light of the fact that all six SPT assemblages derive from Holocene contexts . Despite the possibility that SPT assemblages are pulling the overall Holocene r 2 value down because of their clustering in the upper right of the core patterning (Figure 3), any further consideration of the Middle and Late Holocene sample would be difficult to manage and is beyond the scope of this research. For the time being, the Middle and Late Holocene samples are not fully discounted as outliers and are kept i n the larger AMH

PAGE 84

84 SE Asia data sample, but these assemblages are excluded from the subsequent series, which most convincingly captures assemblage patterning which formed under Pleistocene adaptive configurations, or otherwise strictly hunting and g athering mobility strategies. In consideration of the possible incompatibility of Middle and Late Holocene adaptive systems with this model, the Early Holocene assemblages were combined with terminal Pleistocene assemblages (>12 ka), and constitute a transitory ph ase which is much more apt for this model, and incidentally demonstrated a much stronger predictive relationship between AVD and retouch frequency [( r 2 = .50)]. Notably, when non modern assemblages are reconfigured into this the Pleistocene data series, th e sites of Liang Bua and Song Terus have differing effects on the core patterning and are discussed briefly below. Non modern assemblage patterning. The site localities with assemblages created by relict ho minins include Liang Bua (n=3) and Song Terus (n= 1), the alledgedly non modern assemblages on the cura ted/expedient continuum, are contrasted against the core patterning exhibited in Pleistocene and tranistio nary Holocene series . After the Liang Bua assemblages are incorporated into the Pleistocene serie s, it is clear that these assemblages fall outside of the models expected neagitve linear regression ( r 2 = .153). In contrast, the site of Song Terus, when calibrated, fits more neatly with the overall relat ionship predicted by the model (r 2 =. 65) . While only one assemblage, the assemblage from Song Terus is one of only a few assemblages that have been recently dated the Middle Pleistocene (ca. 265 ka Ð 170 ka) and is believed to have been left by a relic "evolved" H. erectus population. A lthough the ass emblage from Song Terus is dislocated from the core patterning in the uncalibrated sample, its minimal impact on the overall predictive strength in the calibrated patterning (r 2 = .47) positions it more closely aligned with the modern

PAGE 85

85 assemblage series ra nge than are any of the assemblages from Liang Bua, even after they are corrected for duration.

PAGE 86

86 CHAPTER VII C ONCLUSIONS The preliminary results from these analyses suggest that the lithic assemblages acros s SE Asia can be compatible with middle range a pproaches to prehistoric behavior, which often emphasize the linkages which exist between the technological interface , the distribution of resources, and aspects of landscape use , subsistence strategies, and economy. Although hunter gatherer behavior in pr ehistoric H. sapiens of SE Asia is complicated by the litt le knowledge available to approximate adaptive strategies in this region , earl y indications from this research support the notion that lithic organization was structured around landscape use and eco nomic decisions in Pleistocene foragers of SE Asia (Figure 6) . Ideally, this model works best when it is able to explore these differing mobility strategies, or land scape use in relation to specific ecological variables such as the distribution or materi al resources, or in relation the subsistence based activities. However, the findings from this thesis are not quite polish ed enough to comprehensively define the cultural structure of AMH , or speculate on the shaping forces behind regional and ecologically specific organizational qualities of SE Asia survival strategies. Nevertheless, these results hit on higher methodological and theoretical levels of archaeological inquiry, while also setting up a framework to re analyze and incorporate new data into a wo rking hypothesis of hunter gatherer adaptation in SE Asia. In review, there is a widely held consensus among researchers working toward a regional perspective of lithic technology in SE Asia that stone tools may not be a behavioral proxy for differentiatin g between modern and archaics . In particular, the conventional typological divisions (i.e. core/flake dichotomy) have proven to be inadequate in addressing

PAGE 87

87 aspects of technological development and hominin behavior (e.g. Moore and Brumm, 2007; Marwick, 20 08; O'Connor, 2007; Reynolds, 2007). The informal character of lithic industries is argued to transcend species boundaries; thus, the remarkable regional continuity in stone tool traditions makes them an unreliable behavioral proxy for different iating betw een hominin species. The findings from this research suggest that it is at least possible to get at modern human organizational strategies, and thes e data sets may be used to define differences in technological behavior when better defined occupation seque nces of hominins of SE Asia are available . Southeast Asian Technological Organization Making sense of the why and where the assemblages of Pleistocene SE Asia would cluster along the curated expedient continuum is somewhat paradoxical. On the one hand, w e know that the expedient manufacture of stone tools is often tied into logistically oriented mobility st rategies, which centers around larger foraging radii , involves longer occupations, !"# # #! #!! # #! #!! #!!! #!!!! $%&'()*%+,&''-.,/ 01&234)&,5'-6%&12),7%8.2&9,:6 ; <, ! 6'+%18 Figure 6 : Assemblage patterning in SE Asia

PAGE 88

88 and where resources flow back to camp. On the other hand, the distri bution of resources in lower latitudes , or e nvironments with more evenly distributed resources like those of the Pleistocene SE Asia, are more commonly responded to with residentially oriente d mobility strategies . In consideration of these expectations, th e expedient character of stone tools in SE Asia are somewhat in conflict with the ecological framework of tropical and pre humid climate s, which would otherwise predict prehistoric foragers were highly mobile and practicing residentially settlement strateg ies, which in theory would lead to stronger assemblage patterning indicative of highly curated industries. Of course, even as the paleoenvironmental conditions remain vague, it is clear that in order to characterize forager organization in this r egion the conventional parameters along which mobility is defined in this model will need realignment to account for the unique demands encountered by Pleistocene SE Asian foragers . Firstly, in a cross cultural study of hunter gatherer mobility strategies Kelly (19 82) noticed several incongruences with a number of forager mobility strategies across tropical environments , but noted there where logistical mobility strategies dominated in tropical conditions the forager communities were reliant on aquat ic resources. St udies on Pleistocene era foragers of SE A sia have increasingly considered the r ole of littoral resources in shaping the survival strategies and technological behavior here. Additionally, however, there are tropical specific ecological demands which result in resource stress such as the lack of carbohydrates and low density of secondary biomass, which impose important caloric deficits on forager populations subsisting on predominately tropical available foods. Moreover, the notion of environment constraints on lithic technology continues to garner popularity as a working hypothesis Ñ the fact that the arrival of modern humans did little to improve upon the preexisting lithic technologies hints at the high probability that non

PAGE 89

89 lithic resources played an importa nt role in prehistoric technology here . The archaeological signatures of SE Asia are also convincingly expedient, often put down to the lack of availability to fine grained lithic raw material resources. At present, conceptualizing the ecological pressures encountered by Pleistocene foragers in SE Asia is a topic still facing numerous uncertainties. , a mong the most debated topics in Pleistocene SE Asia is the issue of prehistoric climatic conditions Ñ of particular concern is the timing of tropical condition s, and the compositional changes to forest and the configuration of coastlines throughout the Upper and Late Pleistocene. Moreover, scholars have expressed concern over the impact of tropical ecology constraints on hunter gatherer organization, and questio ned the extent true tropical adaptation is possible (e.g. Bailey et al., 1989). With these caveats in mind , and what is currently known and observable about these industries (low input, expediently produced) in contrast to what remains largely unknown abo ut Pleistocene environments across SE Asia, it seems more than reasonable to assume assemblage patterning would cluster near the logistical axis of forager strategies along the exped i ent/curated continuum (Figure 6) . Nevertheless, when the density and rate by which artifacts accumulated are factored in, it appears that there is, in fact, much more variability in technological organization than previously assumed as well as what was predict e d before our analyses (Figure 6 ). Although episodic, retouching ston e tools in Pleistocene SE Asia forager strategies may have had a similar role in economizing lithic resources in relation to broader mobili ty across Pleistocene SE Asia There is still much to be learned with respect to foraging adaptation and economic stra tegies in Pleistocene SE Asia. For example, what type of mobility existed in Pleistocene and whether these hunter gatherers were strictly foraging within the interior forest or,

PAGE 90

90 alternatively, broad spectrum foragers exploiting auxiliary habitats becaus e of the deficiencies or challenges of resource extractions in tropical habitat has yet to have been addressed. However, scholars have recently noted important accumulation differences in stone tools, occupational intensity, in one case between of Bui Ceri Uato and Uai Bobo 2, and questioned whether the differences in occupational intensity might be embedded in a larger subsistence/settlement strategy (Veth et al., 2005). While developing a robustly defined human adaptation strategy in SE Asia is out of rea ch, the later topic is well within the purview of this model (Figure 7). Hominin assemblages . Figure 8 displays the predicted negative relationship between lithic density a nd retouch frequency , and the pattern docume nted at Liang B ua. As was reported in R esults , Liang Bua assemblages do not conform to the expected relationship, showing instead a statistically insignificant, weak positive relationship. Thus, these results would seem to indicate that the !"# # #! #!! # #! #!! #!!! #!!!! $%&'()*%+,&''-.,/ 01&234)&,5'-6%&12),7%8.2&9,:6 ; <, ! 6'+%18 =(2,>%12,?4&' ?42,='@',A Figure 7 : Assemblage variation in East Timor sites

PAGE 91

91 toolmakers who occupied Liang Bua did not organize their mobility and lithic technology similarly to contemporary Homo sapiens and Neanderthal populations. This pattern could be interpreted as a taxonomically significant behavioral signal indicating a Ônon modern' hominin population, perhaps one descended from an isolated Lower Pleistocene archaic population. However, it was also possible that it more simply reflects how tropical foragers organize this technology. This is because all prior applications of the method used here have employed datasets drawn f ro m temperate contexts. The datasets from prehistoric hunter gatherer populations from various sites in SE Asia were analyzed to test this possibility . If the Ôtropical forager' interpretatio n is correct, the expectation was that these other assemblages at tributed to modern humans would not have show n the negative relationship predicted b y the whole a ssemblage behavioral indicator. The pattern that emerged is that SE Asian tropical foragers appear to have tailored their Figure 2: Liang Bua Figure 8 : Assemblage variation in non modern species

PAGE 92

92 lithic technological organization to their mobility strategies in ways very similar to Late Pleistocene foragers from Western Eurasia. This suggests that the methodology used to reconstruct Paleolithic hunter gatherer mobility in temperate conditions is also applicable to hunter gatherers in tropical environments. Further, it provides a new tool to potentially distinguish the assemblages from modern populations from those of archaic populations. Overall, these data indicate that the lithic assemblages from Liang Bua are not organized like thos e of contemporary foragers, and that this difference is not the result of ecological conditions. One potential interpretation is that this reflects a fundamental behavioral difference of the Liang Bua hominins relative to contemporary human populations. Ho wever, recent re dating of the Liang Bua sequences requires a reanalysis of the stone tool assemblages; specifically whether the stone tool assemblages are constrained by the same geo chronological layers. Similarly, the assemblage from Song Terus requires additional sequences to test for organizational changes over time; and the possible fluvial deposit of this assemblage is methodologically problematic. However, major preliminary conclusion of this study is that a growing number of lithic assemblages from tropical hunter gatherers appear to capture the same adaptive signal as that of Late Pleistocene Homo sapiens and Neanderthal lithic assemblages from Western Eurasia. The manifestations of these discard patterns in the archaeological record reflect mobili ty strategies that are apparently not manifested at Liang Bua, and they therefore suggest a fundamental difference in lithic technological organization and mobility. This has behavioral implications that reach beyond techno typological classifications and may be significant in the context of the debate over the taxonomy of the stone tool manufacturers at Liang Bua and Song Terus once better sequences can be defined .

PAGE 93

93 Although the SE Asian site sel ection is fairly limited due to the requirement of a majority of Holocene content , the reported trend is very strong and the developing pattern should encourage researchers to pull together a more comprehensive analysis of lithic assemblages to corroborate the predictions of mobility patterns in tropical ecologies. A robust data set pertaining to lithic densities and frequency of modification in SE Asia can provide a new understanding of tropical mobility, as well as potentially shed light on whether the behavior of the Liang Bua hominins was comparable to those of c oeval foragers.

PAGE 94

94 REFERENCES : Ambrose, Stanley H. 2001 Paleolithic Technology and Human Evolution. Science 291(5509): 1748 Ñ 1753. Anderson, Douglas D. 1990 Lang Rongrien Rockshelter: A Pleistocene, Early Holocene Archaeological Site from Krabi, Southwes tern Thailand . University Museum Monograph 71. The University Museum, University of Pennsylvania, Philidelphia. 1997 Cave archaeology in Southeast Asia. Geoarchaeology: An International Journal 12: 607 Ñ 638. Andrefsky, William, Jr. 1994 Raw Material Ava ilability and the Organization of Technology. Society for American Archaeology 59(1): 21 Ñ 34. 2006 The Analysis of Stone Tool Procurement, Production, and Maintenance. Journal of Archaeological Science 17: 65 Ñ 103. Anthony, David W. 1990 Migration in Ar chaeology: The Baby and the Bathwater . American Anthropologist 92(4): 896 Ñ 914. Bahuchet S., D. Mckey, and Igor de Garine Wild yams revisited: Is independence from agriculture possible for rain forest hunter gatherers? Human Ecology 19(2): 213 Ñ 243. Bail ey, Robert C. and Thomas N. Headland 1991 The tropical rain forest: Is it a productive environment for human foragers? Human Ecology 19(2): 261 Ñ 285. Bailey, Robert C., Genevieve Head, Bruce Owen, Robert Rechtman, and Elzbieta Zechenter 1989 Hunting and Gathering in Tropical Rain Forest: Is It Possible? American Anthropologist 91(1): 59 Ñ 82. Bamforth, Douglas B. 1986 Technological Efficiency and Tool Curation. American Antiquity 51: 38 Ñ 50. Bar Yosef, Ofer 2002 The Upper Paleolithic Revolution. Annual Review of Anthropology 31: 363 Ñ 393. Bar Yosef, Ofer and Philip Van Peer 2009 The Cha” ne OpŽratoire in Middle Paleolithic Archaeology. Current Anthropology 5(1):103 Ñ 131.

PAGE 95

95 Barton, Huw, Graeme Barker, David Gilbertson, Chris Hunt, Lisa Kealhoper, Helen L ewis, Victor Paz, Philip J. Piper, Ryan J. Rabett, Tim Reynolds, and Katherine Szab—. 2013 Late Pleistocene Foragers, c. 35,000 Ñ 11,500 Years Ago. In Rainforest foraging and farming in Island Southeast Asia , Vol. 1, edited by Graeme Barker, pp. 173 Ñ 216.. M cDonald Institute for Archaeological Research, Cambridge. Bartstra, Gert Jan 1982 Homo Erectus Erectus : The Search for His Arifacts. Current Anthropology 23(3): 318 Ñ 320. Bartstra, G.J. and Basoeki 1989 Recent Work on the Pleistocene and the Paleolithic of Java. Current Anthropology 30(2): 241 Ñ 241. Bartstra, G.J., D.A. Hooijer, B. Kallupa, and M. Anwar Akib 1994 Notes on fossil vertebrates and stone tools from Sulawesi, Indonesia, and the stratigraphy of the Northern Walanae Depression. Palaeohistoria 3 3(34): 1 Ñ 18. Bay Petersen, Jan 1983 Excavations at Bagumbayan, Masbate, Central Philippines: An Economic Analysis. Asian Perspective s 25(2): 67 Ñ 98. Bettis III, E. Arthur, Adrianne K. Milius, Scott J. Carpenter, Roy Larick, Yahdi Zaim, Yan Rizal, Rusell L. Ciochon, Stephanie A. Tassier Surine, Daniel Murray, Suminto, Sutinko Bronto 2009 Way out of Africa: Early Pleistocene paleoenvironments inhabited by Homo erectus in Sangiran, Java. Journal of Human Evolution 56(1): 11 Ñ 24/ Binford, Lewis R. 1979 Orga nization and Formation Processes: Looking at Curated Technologies. Journal of Anthropological Research 35(3): 255 Ñ 273. 1980 Willow Smoke and Dog's Tails: Hunter Gatherer Settlement System and Archaeological Site Formation. American Antiquity 45: 4 Ñ 20. Bleed, Peter 2001 Trees or Chains, Links or Branches: Conceptual Alternatives for Consideration of Stone Tool Production and Other Sequential Activities. Journal of Archaeological Method and Theory 8(1): 101 Ñ 127. Brantingham, Jeffery P. 2003 A Neutral Mod el of Stone Raw Material Procurement. American Antiquity 68(3): 487 Ñ 509.

PAGE 96

96 Brumm, Adam, F. Aziz, G.D. van der Bergh, M.J. Mowood, D.R. Hobbs, R. Fullagar 2006 Early stone technology at Flores and its implications for Homo floresiensis . Journal of Human Ev olution 57: 503 Ñ 526. Brumm, Adam, Mark W. Moore, Gert van den Bergh, Iwan Kurniawan, and Michael J. Morwood 2010 Stone technology at the Middle Pleistocene site of Mata Menge, Flores, Indonesia. Journal of Archaeological Science 37: 451 Ñ 473. Brumm, Adam , and Mark W. Moore 2012 Biface distributions and the Movius Line: A Southeast Asian perspective. Australian Archaeology 32 Ñ 46. Brosius, Peter 1991 Foraging in Tropical Rain Forests: The Case of the Penan of Sarawak, East Malaysia (Borneo). Human Ecology 19(2): 123 Ñ 150. Bulbeck, David F. 2003 Hunter Gatherer Occupation of the Malay Peninsula from the Ice Age to the Iron Age. In Under the Canopy : The Archaeology of Tropical Rain Forests , edited by Julio Mercader, pp. 119 Ñ 160. Rutgers University Press, N ew Brunswick. Choi, Kildo and Dubel Driwantoro 2007 Shell tool use by early members of Homo erectus in Sangiran central Java: cut mark evidence. Journal of Archaeological Science 34(1): 48 Ñ 58. Colinvaux, Paul A. and Mark B. Bush 1991 The Rain Forest Eco system as a Resource for Hunting and Gathering. American Anthropologist 93(1): 153 Ñ 160. Corvinus, Gudrun 2004 Homo erectus in East and Southeast Asia, and the question of the age of the species and its association with artifacts, with special attention to handaxe like tools. Quaternary International 117(1): 141 Ñ 151. Cowan, Frank L. 1999 Making Sense of Flake Scatters: Lithic Technological Strategies and Mobility. American Antiquity 64(4): 593 Ñ 607. DŽtroit, Florent, Eusebio Dizon, Christophe Falgures, SŽbastien Hameau, Wilfredo Ronquillo, and Franois SŽmah 2004 Upper Pleistocene Homo sapiens from the Tabon cave (Palwan, The Philippines): description and dating of new discoveries. Comptes Rendus Palevol 3(8): 705 Ñ 712.

PAGE 97

97 Dibble, Harold J. 19 95 Middle Paleolithic Scraper Reduction: Background, Clarification, and Review of the Evidence to Date . Journal of Archaeological Method and Theory 2(4): 299 Ñ 368. Dizon, Eusebio Z. 2010 The lower Paleolithic record in the Philippines. Quaternary Internati onal 1040(6182): 444 Ñ 450. Dwyer, Peter D. and Monica Minnegal 1991 Hunting in Lowland, Tropical Rain Forest: Towards a Model of Non Agricultural Subsistence. Human Ecology 19(2): 187 Ñ 212. Falgures C., F. SŽmah, H. Saleki, S. Hameau, H. Tu, G. FŽraud, H. Simanjuntak, H. Widianto 2015 Geochronology of early human settlements in Java: What is at stake? Quaternary International xxx: 1 Ñ 7. Foley, Robert A. 1987 Hominid species and stone tool assemblages: how are they related? Antiquity 61(233): 380 Ñ 392. 1997 Mode 3 technologies and the evolution of modern humans. Cambridge Archaeological Journ al 7(1): 3 Ñ 36. Foley, Robert and Marta M. Lahr, 2003 On Stony Ground: Lithic Technology, Human Evolution, and the Emergence of Culture. Evolutionary Anthropology 12( 3): 109 Ñ 122. Forestier, Hubert 1999 L'assemblage industriel de Song Keplek, Java Est [Un nouveau regard sur l'outillage lithique del'homme modeme au dŽbut de l'Holocne en IndonŽsie. Bulletin de l'Ecole fran aise d'Extr me Orient 86: 129 Ñ 159. 2007 Rib uan Gunung, Ribuan Alat Batu Prasejarah Song Keplek Gunung Sewa, Jawa Timur . Kepustakaan Populer Gramedia, Jakarata. Forestier, Hubert and Elise Patole Edoumba 2000 Les industries lithiques du PalŽolithique tardif et du dŽbut de l'Holocne en Insulinde. AsŽanie 6: 13 Ñ 56. Forestier, H., Simanjuntak, T., DŽtroit, and V. Zeitoun 2010 UnitŽ et diversitŽ prŽhistorique entre Java et Sumatra . Archipel 80: 19 Ñ 80.

PAGE 98

98 Forestier, Hubert, Heng Sophady, Simon Puaud, Vicenzo Celiberti, StŽphane Frre, ValŽry Zeitoun, CŽcile Mourer ChauvirŽ, Roland Mourre, Heng Than, Laurence Billault 2015 The Hoabinhian from Laang Spean Cave in its stratigraphic, chronological, typo technological and environmental context (Cambodia, Battambang province ). Journal of A rchaeological Scien ce : Reports 3: 194 Ñ 206. Fox, Robert B. 1970 The Tabon Caves: Archaeological Explorations and Excavations on Palawan Island, Phillipines. Monograph of the National Museum No.1. National Museum, Manila. 1987 The Philippine Paleolithic. In Early Paleolith ic in South and East Asia , edited by Fumiko Ikawa Smith, pp. 59 Ñ 86. Mouton Publishers, Paris. Glover, Ian C 1977 Late Stone Age in Eastern Indonesia. World Archaeology 9(1): 42 Ñ 61. 1986 Archaeology in East Timor 1966 67 . Terra Australis 11. Australia n National University, Canberra. Grove, Matt 2009 Hunter gatherer movement patterns: Causes and constraints. Journal of Anthropological Archaeology 28: 222 Ñ 233. Headland, Thomas N. 1987 The wild yam question: How well could independent hunter gatherers live in a tropical rainforest environment? Human Ecology 15(4): 463 Ñ 491. Headland, Thomas N. and Robert C. Bailey 1991 Introduction: Have Hunter Gatherers Ever Lived in Tropical Rain Forest Independently of Agriculture? Human Ecology 19(2): 115 Ñ 122. Hi scock, Peter 2005 Artefacts on Aru: Evaluating the Technological Sequence. In The Archaeology of the Aru Islands, Eastern Indonesia, terra australias 22, edited by S. O'Connor, M. Spriggs, and P. Veth, pp. 205 Ñ 234. Australian National University Press, Can berra. Hutterer, K.L. 1977 Reinterpreting the Southeast Asian Paleolithic. In Sunda and Sahul: Prehistoric Studies.in Southeast Asia, Melanesia and Australia , edited by J. Allen, J. Golson, and R. Jones, pp. 31 Ñ 71. Academic Press, New York. Jacob, T., R .P. Soejono, L.G. Freeman, and F.H. Brown 1978 Stone tools from Mid Pleistocene sediments in Java. Science 202: 885 Ñ 887

PAGE 99

99 Joordens, Josephine C.A., Francesco d'Errico, Frank P. Wesselingh, Stephen Muhro, John de Vos, and Jakob Wallinga 2015 Homo erectus at Trinil on Java used shells for tool production and engraving. Nature 518(7538): 228. Kelly, Robert 1983 Hunter Gatherer Mobility Strategies. Journal of Anthropological Research 39(3): 277 Ñ 306. Lycett, Stephen J., and Christopher J. Bae 2010 The Moviu s Line controversy: the state of the debate. World Archaeology 42(4): 521 Ñ 544. Marwick, Ben 2009 Biogeography of Middle Pleistocene hominins in mainland Southeast Asia: A review of current literature. Quaternary International 202: 51 Ñ 58. Mijares, Arma nd S. B. 2002 The Minori Cave Expedient Lithic Technology . Quezon City, University of the Philippines Press. 2004 Lithic analysis of recently excavated Tabon Cave flakes. Proceedings of the Society of Philippine Archaeologists 2: 15 Ñ 19. 2005 The Archa eology of Pe–ablanca Cave Sites, Northern Luzon, Philippines. Journal of Austronesian Studies 1(2): 65 Ñ 93. 2008 The Pe–ablanca Flake Tools: An Unchanging Technology? Hukay 12: 13 Ñ 34. Mijares, Armand S., Florent DŽtroit, Philip Piper, Rainer GrŸn, Peter Bellwood, Maxime Aubert, Guillaume Champion, Nida Cuevas, Alexandra De Leon, and Eusebio Dizon 2010 New E vidence for a 67,000 year old human presence at Callao Cave, Luzon, Philippines. Journal of Human Evolution 59: 123 Ñ 132. Moore, Mark W., and Adam Brumm 2007 Stone artifacts and hominins in island Southeast Asia: New insights from Flores, eastern Indonesia . Journal of Human Evolution 52: 85 Ñ 102. 2009 Homo floresiensis and the African Oldowan. In Interdisciplinary Approaches to the Oldowan, edited by E. Hovers and D.R. Braun, pp. 61 Ñ 69. Springer, New York. Moore, Mark W., and T. Jatmiko Sutikna, M.J. Morw ood, and A. Brumm 2009 Continuities in stone flaking technique at Liang Bua, Flores. Journal of Human Evolution 57: 503 Ñ 526.

PAGE 100

100 Morgan, Christopher 2009 Climate change, uncertainty and prehistoric hunter gatherer mobility . Journal of Anthropological Archa eology 28: 382 Ñ 396. Morwood, M.J., P.B. O'Sullivan, F. Aziz, and A. Raza 1998 Fisson track ages of stone tools and fossils on the east Indonesian island of Flores. Nature 392: 173 Ñ 176. Mudar, Karen and Douglas Anderson 2007 New Evidence for Southeast Asian Pleistocene Foraging Economies: Faunal Evidence From the Early Levels of Lang Rongrien Rockshelter, Krabi, Thailand. Asian Perspectives 46(2): 298 Ñ 334. Nelson, Mar gret C. 1991 The Study of Technological Organization. Archaeological Method and Theory 3: 57 Ñ 100. O'Connor, Sue 2002 Pleistocene Timor: Further Corrections, a Reply to Bednarik Australian Archaeology 54: 46 Ñ 51. 2006 New Evidence from East Timor contribut es to our understanding of earliest modern human colonization east of the Sunda Shelf. Antiquity 81 : 523 Ñ 535. O'Connor, Sue, and Peter Veth 2005 Early Holocene shell fish hooks from Lene Hara Cave, East Timor establish complex fishing technology was n u se in Island South East Asia five thousand years before Austronesian settlement. Antiquity 79: 249 Ñ 256. O'Connor, Sue, Ono, Rinataro, and Clarkson, Chris 2011 Pelagic Fishing at 42,000 Before the Present and the Maritime skills of Modern Humans. Science 334(6059): 1117 Ñ 1121. O'Connor, Sue, and Matthew Spriggs, and Peter Veth 2002a Excavation at Lene Hara Cave establishes occupation in East Timor at least 30,000 Ñ 35,000 years ago. Antiquity 76: 45 Ñ 50. 2002b Direct Dating of shell beads from Lene Hara Ca ve, East Timor. Australian Archaeology 55: 18 Ñ 21. O'Connor, Sue, Ken Aplin, Katherine Szab— , Juliette Pasveer, Peter Veth, and Matthew Spriggs 2007 Liang Lemdubu: A Pleistocene Cave Site in the Aru Islands. In The Archaeology of the Aru Islands, Eastern Indonesia, Terra Australis 22, edited by S. O'Connor, M. Spriggs, and P. Veth, pp. 171 Ñ 204. Australian National University Press, Canberra.

PAGE 101

101 Odell, George H. 1996 Economizing behavior and the concept of curation. In Stone Tools: Theoretical Insights int o Human Prehistory, edited by George H. Odell, pp. 51 Ñ 80. Springer, New York. Ono, Rintaro, Santoso Soegondho, and Minoru Yoneda 2009 Changing Marine Exploitation During Late Pleistocene in Northern Wallacea: Shell Remains from Leang Sarru Rockshelter i n Talaud Islands. Asian Perspectives 48(2): 318 Ñ 341. Ono, Rintaro, Naoki Nakajima, Hiroe Nishizawa, Shizuo Oda, and Santoso Soegondho 2015 Maritime Migration and Lithic Assemblage on the Talaud Islands in Northern Wallacea during the Late Pleistocene to E arly Holocene. In Emergence and Diversity of Modern Human Behavior in Paleolithic Asia, edited by Yousuke Kaifu, Masami Izuho, Ted Goebel, Hiroyuki Sato, and Akira Ono, pp. 201 Ñ 213. Texas A&M University Press, College Station. Patole Edoumba, Elise 2009 A Typo Technological Definition of Tabonian Industries. Bulletin of the Indo Pacific Prehistory Association 29: 21 Ñ 25. Patole Edoumba, Elise, Alfred F. Pawlik, and Armand S. Mijares 2011 Evolution of prehistoric lithic industries of the Philippines du ring the Pleistocene. Comptes Redus Palevol 11: 213 Ñ 230. Pawlik, Alfred F. and Ronquillio P. Wildredo 2003 The Paleolithic in the Philippines. Lithic Technology 28(2): 79 Ñ 93. Pawlik Alfred F., Philip J. Piper, Marie Grace Pamela G. Faylona, Sabino G. Padilla, Jane Carlos, Armand S. B. Mijares, Benjamin Vallejo. Marian Reyes, Noel Amano, Thomas Ingicco, and Martin Porr 2014 Adaptation and foraging from the Terminal Pleistocene to the Early Holocene: Excavation at Bubog on Ilin Island, Philippines. Journ al of Field Archaeology 39(3): 230 Ñ 247. Pope, G. 1984 The Antiquity and Paleoenvironment of the Asian Hominindae. In The Evolution of the East Asian Environment, edited by R. Whyte, pp. 827 Ñ 847. University of Hong Kong, Hong Kong. Rabett, Ryan J. 2012 a Human Adaptation in the Asian Paleolithic: hominin dispersal and behaviour during the late Quaternary . Cambridge University Press, Cambridge.

PAGE 102

102 2012b Techno Modes, Techno Facies, and Palaeo Cultures: Change and Continuity in the Pleistocene of Southeast , Central and North Asia. In Investigating Archaeological Cultures: Material Culture, Variability, and Transmission , edited by Benjamin W. Roberts and Mark V. Linden. Springer, New York. Rabett, Ryan J., Graeme Barker, Huw Barton, Chris Hunt, Lindsay Llo yd Smith, Victor Paz, Philip J. Piper, Rasmathiri Premathilake, Garry Rushworth, Mark Stephans, and Katherine Szab—. 2013 Landscape Transformation and Human Responses, c. 11,500 Ñ c. 4,500 Years Ago. In Rainforest foraging and farming in Island Southeast A sia , Vol. 1, edited by Graeme Barker, pp. 217 Ñ 254. McDonald Institute for Archaeological Research, Cambridge. Reynolds, T.E.G. 1989 Techno typology in Thailand: a case study of the Tham Khao Khi Chan. Indo Pacific Prehistoric Association 9: 33 Ñ 43. 1993 Problems in the Stone Age of South East Asia. Proceedings of the Prehistoric Society 59: 1 Ñ 15. 2007 Problems of the Stone Age of South east Asia revisited. Proceedings of the Prehistoric Society 73: 39 Ñ 58. Reynolds, Tim, Graeme Barker, Huw Barton, Gatho rne Cranbrook, Lucy Farr, Chris Hunt, Lisa Kealhofer, Victor Paz, Alistair Pike, Philip J. Piper, Ryan J. Rabett, Garry Rushworth, Christopher Stimpson, and Katherine Szab—. 2013 The First Modern Humans at Niah, c. 50,000 Ñ 35,000 Years Ago. In Rainforest f oraging and farming in Island Southeast Asia , Vol. 1, edited by Graeme Barker, pp. 135 Ñ 172. McDonald Institute for Archaeological Research, Cambridge. Riel Salvatore, Julien and C. Michael Barton 2004 Late Pleistocene Technology, Economic Behavior, and L and Use Dynamics in Southern Italy. American Antiquity 69(2): 257 Ñ 274. Riel Salvatore, Julien, Gabriel Popescu, and C. Michael Barton 2008 Standing at the gates of Europe: Human behavior and biogeography in the Southern Carpathians during the Late Pleisto cene. Journal of Anthropological Archaeology 27: 399 Ñ 417. Ronquillo, Wilfredo P. 1985 Archaeological Research in the Philippines, 1951 Ñ 1983. Bulletin of the Indo Pacific Prehistory Association 6: 74 Ñ 88.

PAGE 103

103 SŽmah, Franois, Anne Marie SŽmah, and Truman S imanjuntak 2003 More than a Million Years of Human Occupation in Insular Southeast Asia: The Early Archaeology of Eastern and Central Java. In Under the Canopy : The Archaeology of Tropical Rain Forests , edited by Julio Mercader, pp. 161 Ñ 190. Rutgers Univer sity Press, New Brunswick. SŽmah, Anne Marie, Franois SŽmah, Anne Marie Moigne, Thomas Ingicco, Andre Purnomo, Truman Simanjuntak, Harry Widianto 2016 The paleoenvironmental context of the Paleolithic of Java: A brief review. Quaternary International x xx:1 Ñ 8. Semaw, S., P. Renne, J.W.K. Harris, C.S. Feibel, R.L. Bernor, N. Fesseha, and K. Mowbray 1997 2.5 Million year old stone tools at Gona, Ethopia. Nature 385: 333 Ñ 336. Semaw, Sileshi, Michael J. Rodgers, Jay Quade, Paul R. Renne, Robert F. Butler, Manuel Dominquez Rodrigo, Dietrich Stout, William S. Hart, Travis Pickering, and Scott W. Simpson 2003 2.6 Million year old stone tools and associated bones from OGS 6 and OGS 7, Gona, Afar, Ethiopia. J ournal of Human Evolution 45(2): 169 Ñ 177. Shott, Mi chael J. 1996 An Exegesis of the Curation Concept. Journal of Anthropological Research 52(3): 259 Ñ 280. Shott, Michael J., Andrew P. Bradbury, Phillip Carr, and George H. Odell 2000 Flake Size from Platform Attributes: Predictive and Empirical Approaches. Journal of Archaeological Science 27: 877 Ñ 894. Shott, Michael J. and Paul Sillitoe 2005 Use life and Curation in New Guinea experimental used flakes. Journal of Archaeological Science 32: 653 Ñ 663. Soejono, R. P. 2001 Remarks on the Developments and Pr oblems of the Paleolithic in Indonesia. In Sangiran: Man, Culture, and Environment in Pleistocene Times , edited by Truman Simanjuntak, Bagyo Prasetyo, and Retno Handini, pp. 143 Ñ 153. The National Research Center of Archaeology. Yayasan Obor, Indonesia. Tanudirjo, Daud A. 2006 Long Continuous or Short Occasional Occupation? The Human Use of Leang Sarru Rockshelter in the Talud Islands, Northeastern Indonesia. Bulletin of the Indo Pacific Prehisory Association 25: 15 Ñ 20.

PAGE 104

104 Thiel, Barbara 1987 Excavatio ns at Arku Cave, Northeast Luzon, Philippines. Asian Perspectives 27(2): 229 Ñ 264. 1987 Early Settlement of the Philippines, Eastern Indonesia, and Australia New Guinea: A New Hypothesis. Current Anthropology 28(2): 236 Ñ 241. 1988 Excavations at Musang C ave, Northeast Luzon, Philippines. Asian Perspectives 28(1): 61 Ñ 81. Tuq, Alain, Wil Roebroeks, Laurence Bourguignon, and Jean Philippe Faivre 2013 The fragmented character of Middle Paleolithic stone tool technology . Journal of Human Evolution 65: 641 Ñ 6 55. van den Bergh, G.D., H.J.M. Meijer, Rokhus Due Awe, M.J. Morwood, K. Szab—, L.W. can den Hoek Ostende, T. Suitikna, E.W. Saptomo, P.J. Piper, K.M. Dobney 2009 The Liang Bua faunal remains: a 96 k.yr. sequence from Flores, East Indonesia. J ournal of Hu man Evolution 57: 527 Ñ 537. Van Heteren, Anneke H. and John de Vos 2012 Stone implements from Java and Flores: A history of discoveries. Comptes Rendus Palevol 11(2 Ñ 3): 181 Ñ 189. White, J. Peter 1967 Ethno Archaeology in New Guinea: Two Examples. Mank ind 6(9): 409 Ñ 414. White, J. Peter and David H. Thomas 1972 What mean these stones? Ethnotaxonomic models and archaeological interpretations in the New Guinea Highlands. In Models in Archaeology , edited by David L. Clark, pp. 276 Ñ 308. Routledge, London . Yasuoka, Hirokaza H. 2013 Dense Wild Yam Patches Established by Hunter Gatherer Camps: Beyond the Wild Yam Question, Toward a Historical Ecology of Rainforests. Human Ecology 41(3): 465 Ñ 475. Zeitoun, Valery, Florent DŽtroit, Dominique Grimaud HervŽ , and Harry Widianto 2010 Solo man in question: Convergent views to split H. erectus into two categories. Quaternary International 223 Ñ 224: 281 Ñ 292. Zeitoun, Valery, Hubert Forestier, Michel Rasse, Prasit Auetrakulvit, Jeongmin Kim, and Chaturaporn Tiam tinkrit 2013 The Ban Don Mun artifacts: A chronological reappraisal of human occupations in the Lampang province of Northern Thailand. Journal of Human Evolution 65: 10 Ñ 20.

PAGE 105

105