Desert Peoples: Archaeological Perspectives

Desert Peoples Archaeological Perspectives Edited by Peter Veth (Australian Institute of Aboriginal and Torres Strait I...

Author: Peter Veth | Mike Smith | Peter Hiscock

77 downloads 1457 Views 4MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

Desert Peoples Archaeological Perspectives Edited by

Peter Veth (Australian Institute of Aboriginal and Torres Strait Islander Studies), Mike Smith (National Museum of Australia), and Peter Hiscock (The Australian National University)

Desert Peoples

Desert Peoples Archaeological Perspectives Edited by

Peter Veth (Australian Institute of Aboriginal and Torres Strait Islander Studies), Mike Smith (National Museum of Australia), and Peter Hiscock (The Australian National University)

ß 2005 by Blackwell Publishing Ltd BLACKWELL PUBLISHING 350 Main Street, Malden, MA 02148-5020, USA 108 Cowley Road, Oxford OX4 1JF, UK 550 Swanston Street, Carlton, Victoria 3053, Australia The right of Peter Veth, Mike Smith, and Peter Hiscock to be identiﬁed as the Authors of the Editorial Material in this Work has been asserted in accordance with the UK Copyright, Designs, and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the prior permission of the publisher. First published 2005 by Blackwell Publishing Ltd Library of Congress Cataloging-in-Publication Data Desert peoples: archaeological perspectives / edited by Peter Veth, Mike Smith, Peter Hiscock. p. cm. Includes bibliographical references and index. ISBN 1-4051-0090-7 (hardcover: alk. paper) — ISBN 1-4051-0091-5 (pbk. : alk. paper) 1. Desert people. 2. Archaeology. I. Veth, Peter Marius. II. Smith, M. A. III. Hiscock, Peter, 1957– GN390.D47 2005 306’.09154—dc22 2004014009 A catalogue record for this title is available from the British Library. Set in 10.5/13pt Minion by Kolam Information Services Pvt. Ltd, Pondicherry, India Printed and bound in the United Kingdom by T J International Ltd, Padstow, Cornwall The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards. For further information on Blackwell Publishing, visit our website: www.blackwellpublishing.com

Contents

Notes on Contributors

vii

1 Global Deserts in Perspective

1

Mike Smith, Peter Veth, Peter Hiscock, and Lynley A. Wallis

Part I Frameworks

15

2 Theoretical Shifts in the Anthropology of Desert Hunter-Gatherers

17

Thomas Widlok

3 Pleistocene Settlement of Deserts from an Australian Perspective

34

Peter Hiscock and Lynley A. Wallis

4 Arid Paradises or Dangerous Landscapes: A Review of Explanations for Paleolithic Assemblage Change in Arid Australia and Africa

58

Peter Hiscock and Sue O’Connor

Part II Dynamics

79

5 Evolutionary and Ecological Understandings of the Economics of Desert Societies: Comparing the Great Basin USA and the Australian Deserts

81

Douglas W. Bird and Rebecca Bliege Bird

6 Cycles of Aridity and Human Mobility: Risk Minimization Among Late Pleistocene Foragers of the Western Desert, Australia

100

Peter Veth

________________________________________________________________________________________________________

v

________________________________________________________________________________

Contents

7 Archaic Faces to Headdresses: The Changing Role of Rock Art Across the Arid Zone Jo McDonald

116

8 The Archaeology of the Patagonian Deserts: Hunter-Gatherers in a Cold Desert Luis Alberto Borrero

142

Part III Interactions

159

9 Perspectives on Later Stone Age Hunter-Gatherer Archaeology in Arid Southern Africa Anne I. Thackeray

161

10 Long-Term Transitions in Hunter-Gatherers of Coastal Northwestern Australia Kathryn Przywolnik

177

11 Hunter-Gatherers and Herders of the Kalahari during the Late Holocene Karim Sadr

206

12 Desert Archaeology, Linguistic Stratigraphy, and the Spread of the Western Desert Language Mike Smith

222

13 People of the Coastal Atacama Desert: Living Between Sand Dunes and Waves of the Paciﬁc Ocean Calogero M. Santoro, Bernardo T. Arriaza, Vivien G. Standen, and Pablo A. Marquet

243

14 Desert Solitude: The Evolution of Ideologies Among Pastoralists and Hunter-Gatherers in Arid North Africa Andrew B. Smith

261

15 Hunter-Gatherer Interactions with Sheep and Cattle Pastoralists from the Australian Arid Zone Alistair Paterson

276

16 Conclusion: Major Themes and Future Research Directions Peter Veth

293

General Index

301

Index of Archaelogical Features and Subjects

303

_______________________________________________________________________________________________________

vi

________________________________________________________________________________

Notes on Contributors

Bernardo T. Arriaza is Professor of Physical Anthropology at the University of Nevada, Department of Anthropology and Ethnic Studies, Las Vegas, and Visiting Researcher at Centro de Investigaciones del Hombre en el Desierto, Universidad de Tarapaca´ Arica, Chile. The focus of his research is the bio-cultural adaptation and evolution of prehistoric Andean populations. He has taught bio-anthropology at both undergraduate and postgraduate levels in the Unites States, Chile, Argentina, and Mexico. Douglas W. Bird is an Assistant Research Professor in the Climate Change Institute and Department of Anthropology at the University of Maine. He grew up in the Great Basin and has spent the last few years working with Martu peoples in Australia’s Western Desert. He was educated at the University of Utah and University of California, Davis. His research focuses on foraging ecology and its archaeological implications in small-scale societies. Rebecca Bliege Bird is an Assistant Professor in the Department of Anthropology at the University of Maine. She received her PhD in Anthropology from the University of California, Davis in 1996. After a long-term ethnographic project in the Torres Strait, she began working with the Martu community in 2000. She is especially interested in the ecology of foraging and gender, and the role of women’s economic activities in shaping Martu land management strategies. Luis Alberto Borrero received his PhD from the Universidad de Buenos Aires in 1986. He is a Senior Archaeologist at the Consejo Nacional de Investigaciones Cientiﬁcas y Tecnicas and Professor at the Universidad de Buenos Aires. He has worked widely in Patagonia, with an active research interest in hunter-gatherers.

______________________________________________________________________________________________________

vii

______________________________________________________________________________

No t e s o n C o n t r i b u t o r s

Peter Hiscock is a Reader in the School of Archaeology and Anthropology at The Australian National University. He is a leading researcher into prehistoric stone technology, with a record of developing models about prehistoric (Paleolithic) technology, both in Australia and elsewhere. He has worked extensively in, and published widely on, the Australian deserts, although his current focus is the analysis of European Paleolithic technologies. Jo McDonald is Managing Director of Jo McDonald Cultural Heritage Management Pty Ltd. She completed her PhD at The Australian National University in Canberra. Her thesis used information exchange theory to understand the regional variability in a dual media art system in the Sydney Basin, in southeastern Australia. In addition to managing a cultural heritage consultancy business in Sydney, her current research interests lie in the prehistoric graphic systems of the Western Desert. Pablo A. Marquet is Associate Professor in the Center for Advanced Studies in Ecology and Biodiversity and Departamento de Ecologı´a, Pontiﬁcia Universidad Cato´lica de Chile, Santiago. The focus of his research is in the emergent discipline of macroecology, centered in the search for general principles that underlie the structure and function of ecological systems. Sue O’Connor is a Senior Research Fellow in the Department of Archaeology and Natural History at The Australian National University in Canberra. Her research interests include issues of colonization and technological and subsistence change in northern Australia and Island Southeast Asia. Alistair Paterson is an archaeology lecturer at the University of Western Australia. He researches the archaeology of colonial Australia and culture contact. He has published in a range of leading Australian journals and contributed to the Encyclopedia of Historical Archaeology. Kathryn Przywolnik completed her PhD at the University of Western Australia in 2000. Her thesis explored issues of long-term hunter-gatherer occupation and mobility on the arid Cape Range Peninsula in coastal northwest Western Australia. She has worked on a range of historic and precolonial archaeological excavations and ﬁeld surveys throughout Australia, and her current research interests include environmental and social constructions in archaeology. She is currently employed as an archaeologist with the Department of Environment and Conservation (NSW) in Sydney. Karim Sadr’s PhD is based in the School of Geography, Archaeology, and Environmental Studies at the University of the Witwaterstrand in Johannesburg. His current research interest is the transition to food production in Southern Africa.

_____________________________________________________________________________________________________

viii

______________________________________________________________________________

Notes on Contributors

Calogero M. Santoro is a Full Professor of Archaeology, Executive Director of the Centro de Investigaciones del Hombre en el Desierto, Universidad de Tarapaca´ Arica, Chile and co-editor of the journal Chungara Revista de Antropologı´a Chilena. His research focuses on late prehistoric farming societies and the Inka state, and Holocene hunter-gatherers in the Atacama Desert. Andrew B. Smith is Associate Professor in the Department of Archaeology, University of Cape Town, Rondebosch. His research interests include the origins and development of herding societies in Africa, the prehistory of modern Saharan pastoralists, relations between hunters and herders at the Cape, and the origins of Khoekhoen. Mike Smith is Director of Research and Development at the National Museum of Australia. He is an internationally recognized ﬁeld archaeologist, researcher, lecturer, and museum curator, having pioneered research into late Pleistocene settlement of the Australian desert. He has published widely in leading international journals. Vivien G. Standen is Professor of Bioarchaeology at the Centro de Investigaciones del Hombre en el Desierto, Universidad de Tarapaca´ Arica, Chile and co-editor of the journal Chungara Revista de Antropologı´a Chilena. Her research focuses on both the cultural and environmental factors that shaped the prehistoric populations that inhabited the Atacama Desert, examining both hunter-gatherers and farmers. Anne I. Thackeray is an Honorary Research Associate in the Archaeology Division of the School of Geography, Archaeology, and Environmental Studies at the University of the Witwaterstrand in Johannesburg. Her research interests focus on the stone tools of the southern African Middle and Later Stone Ages, and she writes on wide-ranging African archaeological topics for non-specialist readers. Peter Veth is Head of the Research Section at the Australian Institute of Aboriginal and Torres Strait Islander Studies in Canberra. He has conducted extensive archaeological research throughout the arid zone of Australia and has produced over 200 academic articles, chapters, books, and peer-reviewed reports focusing on desert archaeology, ethno-economics, cultural heritage issues, and native title. Peter has given invited lectures on these topics in the USA, UK, Europe, and Southeast Asia. Lynley A. Wallis is a Visiting Research Fellow in the Department of Archaeology and Natural History at the Australian National University in Canberra, where she completed her PhD in 2001. She is currently researching the timing and nature of Aboriginal occupation in the semi-arid region of northwest Queensland, human– paleoenvironmental interactions during the Late Quaternary, and the application

_______________________________________________________________________________________________________

ix

________________________________________________________________________________

No t e s o n C o n t r i b u t o r s

of phytolith analysis to archaeological and paleoenvironmental research questions in northern Australia Thomas Widlok completed his MSc and PhD in anthropology at the London School of Economics, where he has also taught. Currently he lectures in anthropology at the University of Heidelberg and leads an interdisciplinary research project at the Max Planck Institute for Psycholinguistics in Nijmegen (Netherlands). He has conducted long-term ethnographic ﬁeld research in Namibia and Australia and has published on these regions as well as comparative work on settlement patterns, language change, and ritual dynamics.

________________________________________________________________________________________________________

x

________________________________________________________________________________

---------------------------------------------------------------------------------

1

-----------------------------------------------------------------------------

Global Deserts in Perspective Mike Smith, Peter Veth, Peter Hiscock, and Lynley A. Wallis

Introduction For centuries, deserts have captured the public imagination as places of extremes. These are landscapes that might be perceived as impenetrable barriers to human occupation or instead as the domain entered into by individuals pursuing a revelatory experience. They are of course also the same terrain through which the Tigris and Euphrates Rivers passed and which, when hydraulically ‘‘tamed,’’ became the agricultural powerhouses of the Near East. Desert societies have also been central to the anthropological imagination. The classic ethnographies of hunter-gatherer societies – of the Ju/’hoansi (!Kung), the Paiute and Shoshone, the Arrernte, Pitjantjatjara, and Pintubi – all deal with desert peoples. Surprisingly, very few analyses have adopted a comparative perspective on a global scale (however, see Peterson 1979). This volume aims to bring together studies which, as a corpus, allow us to take a comparative approach to the emergence and diversity of global desert societies. Over the last century, hunter-gatherer studies have moved from a social evolutionary perspective at the close of the nineteenth century (Spencer and Gillen 1899), to structural-functionalist or cultural ecology frameworks in the twentieth century (Gould 1969; Lee 1979; Steward 1938). The Man the Hunter symposium (Lee and DeVore 1968) promulgated a now popular model of hunter-gatherer society, the ‘‘generalized forager’’ model, which was based substantially on these desert hunter-gatherer groups. Under this model, generalized foragers shared ﬁve basic characteristics: egalitarian society; low population density; lack of territoriality; minimal food storage; and ﬂuid band composition with changes in residential mobility used to maintain social ties and reduce intragroup conﬂict – though it remains to be determined whether these are characteristic of foragers in desert environments, rather than hunter-gatherers in general. The elements of this socioeconomic model had been formulated in Steward’s (1938) pioneering study of

____________________________________________________________________________________

1

________________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s

Great Basin Shoshone and Paiute foragers. This model was reinvigorated during the 1970s, and at that time !Kung bushmen came to be seen as the quintessential hunter-gatherers. A review of social and behavioral variability in hunter-gatherers shows that there is a wide spectrum of hunter-gatherer societies if groups living in other types of habitats are included in the analysis (Kelly 1995). Over the last 20 years, hunter-gatherer research has shifted towards either behavioral ecology (Smith and Winterhalder 1992) or historical analyses of these societies (e.g., Schrire 1984; Wilmsen 1989). (For accessible interdisciplinary overviews of hunter-gatherer studies, see Lee and Daly 1999 or Panter-Brick et al. 2001.) In putting this volume together, we felt that it was time to reframe questions about the structure and dynamics of foraging groups, using the desert environment as a frame of reference and comparison. Deserts have a special role in human evolution and adaptation. They appear to be the major terrestrial habitat that channeled early human dispersal, representing barriers at some times, corridors at others (cf. Gamble 1993). Studies of desert societies have also provided some of the most fertile ground for debates about human adaptability and how societies cope with marginal – often precarious – environmental circumstances, and about the effects of these environmental conditions on human land use, mobility, and dispersal (Kelly 1995). How do societies in marginal environments actually deal with risk in either a reactive or strategic sense? Many desert foragers in the ethnographic record appear to have responded by changing their diet-breadth and residential mobility. For others, such as in the Old World deserts of the Northern Hemisphere, the proximity of deserts to the major zones of plant and animal domestication appears to have provided a mutual ecology for change, in both the mode and relations of production. The long-term dynamics of both desert societies and the desert environment are not readily accessible to analysis using standard ethnographic or historical approaches. For this, the longer perspective provided by archaeology is necessary. The emphasis of this volume is therefore squarely on deserts as a major world habitat, on hunter-gatherer peoples in deserts, and on the rapidly growing body of archaeological data on the deep history of these groups.

Deserts: A Modicum of Facts and Figures Deserts are one of the world’s major habitats, forming large bands of drylands along the tropics in both the Northern and Southern Hemispheres (see Mares 1999 and Middleton and Thomas 1997 for overviews of world deserts). A recent map of the extent of world deserts has been produced by the United Nations Environment Program (UNEP) (see Middleton and Thomas 1997) and is reproduced here as Figure 1.1. Deserts cover around 25,500,000 sq km, approximately 20 percent of the land area of the world (see Table 1.1). The boundaries of these drylands are neither static nor abrupt: they have changed throughout the Quaternary in response to shifts in global climate and weather systems – and will no doubt change over the next century as human-induced global warming takes effect. The deﬁning

____________________________________________________________________________________

2

________________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e Table 1.1 Area of world deserts by region and zone. Figures are square kilometers x 1,000,000 (after Middleton and Thomas 1997: table 1.1). Zone

Africa

Asia

Australia

Europe

North America

Arid Hyperarid Total

5.04 6.72 11.76

6.26 2.77 9.13

3.03 0.00 3.03

0.11 0.00 0.11

0.82 0.31 1.13

South America

Total

0.45 0.26 7.02

15.69 9.78 25.47

characteristic of world deserts – aridity – can be measured in a number of ways. The current UNEP deﬁnition is that it represents a moisture deﬁcit under normal climatic conditions where P/PET <0.20; that is, where rainfall is less than 20 percent of potential moisture loss through evaporation. Several factors interact to determine the intensity of aridity: . Atmospheric stability. Most deserts are low latitude deserts, made of two latitudinal bands along the tropics (238N and 238S) produced by patterns of atmospheric circulation. . Continentality. Deserts are often found in the interior of continents where the reach of maritime air masses is less pronounced. This is often accentuated by topography and by cold ocean currents. . Topography. High mountain ranges can form barriers to moist maritime air masses and create rain-shadow deserts, as in the case of central Asia and Patagonia. . Cold ocean currents. Low sea-surface temperatures along the west coasts of continents reduce sea-surface evaporation and contribute to aridity, such as along the Namib and Atacama coasts, by reducing the effectiveness of maritime air masses. This highlights the extreme variability in these habitats, which range from great continental deserts (such as the Sahara, Kalahari, and Australian deserts) to basinand-range or montane deserts (such as North America’s Great Basin or the Puna in northwestern Argentina), coastal deserts (like the Namib or Atacama), or regions where aridity is substantially increased by the rain-shadow effect of nearby mountains (such as in the central Asian deserts, or in the Patagonian deserts). There is also great variability in the intensity of aridity in world deserts. The eastern Sahara in North Africa, the Atacama in northern Chile, and the Namib in southwestern Africa all receive little or no rainfall today and are referred to as hyperarid regions. In these environments life revolves around springs or shallow groundwater seepages, stream ﬂows from the Andes Cordillera (in the case of the Atacama), or moisture from coastal fogs (in the case of the Namib). Outside of the scattered oases or well-watered ravines, absolute desert has few resources for a huntergatherer population. In contrast, the Kalahari Desert receives relatively good rainfall (250–500 mm per annum) but the deep porous Kalahari sands mean that

____________________________________________________________________________________

3

________________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s

NORTH AMERICA Atlantic Ocean Great Basin Deserts (Bird & Bliege Bird, Chapter 5)

Sahara Desert (Hiscock & O'Connor, Chapter 4; A. Smith, Chapter 14)

Pacific Ocean SOUTH AMERICA Atlantic Ocean

Atacama Desert (Santoro et al., Chapter 13)

0

Patagonian Deserts (Borrero, Chapter 8)

5000 km

Hyperarid

____________________________________________________________________________________

4

________________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e

ASIA

EUROPE Pacific Ocean

Western Desert (Bird & Bliege Bird, Chapter 5; Veth, Chapter 6; McDonald, Chapter 7; M. Smith, Chapter 12)

AFRICA

Indian Ocean

NW Coastal Desert (Przywolnik, Chapter 10) AUSTRALIA Namib & Kalahari Deserts (Widlok, Chapter 2; Hiscock & O'Connor, Chapter 4; Thackeray, Chapter 9; Sadr, Chapter 11)

Southern Ocean Australian Deserts (Widlok, Chapter 2; Hiscock & Wallis, Chapter 3; Paterson, Chapter 15)

Arid

Semi-arid

Figure 1.1 Map of world drylands based on UNEP aridity index (after Middleton and Thomas 1997: ﬁgure 6). Hyperarid ¼ areas that have very limited and highly variable rainfall amounts, both interannually and on a monthly basis; Arid ¼ areas that have mean annual precipitation up to about 200 mm in winter rainfall areas and 300 mm in summer rainfall areas; interannual variability in the 50–100 percent range; Semi-arid ¼ areas with highly seasonal rainfall regimes and mean annual values up to ca. 800 mm in summer rainfall areas and ca. 500 mm in winter rainfall areas; high (25–50 percent) interannual variability.

____________________________________________________________________________________

5

________________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s

this is quickly lost: in effect, the Kalahari ‘‘thirstland’’ is an edaphic desert. The Australian deserts also receive comparatively good rainfall and like the Kalahari are well vegetated: the challenge here for people is the pronounced interannual and decadal variability in rainfall in such deserts, which are subject to a ‘‘boom and bust’’ cycle over decades.

Desert People: Some Issues Current UNDP/UNSO statistics indicate that around 313 million people (about 13 percent of the total population) currently live in the world’s arid zones, with 92 million alone residing in hyperarid deserts. These ﬁgures include signiﬁcant urban populations, reliant on resources (and often water) imported from outside the zone, especially in Africa, South America, Arabia, and central Asia. It is likely that prior to the rise of cities and agricultural or pastoral communities, the proportion of world population in deserts would have been signiﬁcantly greater. Deserts are difﬁcult environments for hunter-gatherers not just because scarcity of water and other resources are limiting factors. These are environments where resources are patchy and highly variable in both time and space. Often, small parts of the wider landscape – springs, groundwater discharge zones, run-on areas – are the key to utilization of the wider region. Rainfall events create pulses of biological productivity separated by long dormant periods and these are largely unpredictable in time and space. Desert environments are characteristically subject to high interannual, decadal, and millennial variability in rainfall. In deserts, much of the ecosystem is geared towards a pattern of ‘‘pulse and reserve’’ (a term coined by the desert ecologist Immanual Noy Meir) or ‘‘boom and bust’’ that people also use and exploit for their own needs – social as well as economic. Deserts are also highly patchy environments in which nutrients and/or water are concentrated in patches within a larger, less productive landscape. Most desert hunter-gatherers are water-tethered to some degree, but in the Australian and southern African deserts, where there are signiﬁcant plant and animal resources thinly distributed throughout the desert, it is the distribution of small surface waters, seepages, wells, and springs that provides access to the desert hinterland and a means of stepping through the country; like navigating through islands on the sea (cf. Veth 1993). Hunter-gatherers in these environments have strategies which involve high residential mobility, broad-spectrum foraging, and a high degree of organizational and technological ﬂexibility. Not all contemporary deserts, however, conform to this pattern. In the Atacama, for instance, the desert hinterland is absolute desert with no appreciable biological resources. All productivity is concentrated in the widely separated oases and ravines. Such an extreme environment favors greater investment in territoriality, ‘‘landesque’’ infrastructure such as semi-sedentary villages and storage facilities, and a higher level of management of wild herd animals and wild plant foods. Generalities about the relationship between the economic/social strategies of human groups and the nature of deserts in which they live provide an insight into

____________________________________________________________________________________

6

________________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e

the historical patterns of human existence in arid lands. However, the present is not simply an iteration of the past, and explorations of ancient human colonization and settlement of deserts must deal with at least two complexities. First, past environments in which humans lived were often different, sometimes radically so, to those where people were observed in recent times. The opportunities and constraints of those different ecosystems provided the context for early settlement strategies, and archaeological interpretations of past desert lifeways must therefore be set within a framework of the environmental history of each landscape. The second complexity involves the question of what adaptive strategies equipped people to move into deserts for the ﬁrst time; what kinds of economic tactics had emerged in other landscapes that prepared human groups for survival in the variable and extreme conditions of a desert. A related question is when, and in what conditions, did the economic and social systems visible in historic desert settlement arise; and how far back in time can we recognize those forms of cultural organization? When did people ﬁrst settle these precarious environments? Archaeological evidence indicates that people have a long presence in the African deserts. Early Stone Age or Acheulian sites are reported from the Namib and Kalahari, where they are associated with ancient river courses, and in the eastern Sahara, where they are associated with artesian springs and lake deposits (e.g. at Bir Tarfawi). One of the perennial difﬁculties with the interpretation of such data is in determining whether this reﬂects exploitation of favorable patches within a desert environment or whether these areas were semi-arid savannas at that time. By the Middle Stone Age (ca. 60,000 years ago) there is good evidence for establishment of a resident hunter-gatherer population in the southern Namib Desert. The Australian deserts were widely occupied by 40,000–30,000 bp, though the nature of this occupation is still being worked out – and most researchers agree that early settlement may have been patchy. Any discussion of the initial colonization of deserts now draws on a signiﬁcant body of archaeological and biological theory about the likely pattern of dispersal into new environments and the use of patchy or mosaic environments such as deserts (for overviews, see Clobert et al. 2001; Forman 1995; Rockman and Steele 2003; Shigesada and Kawasaki 1997). In both the archaeological and biological literature, a distinction is conventionally made between the exploratory or pioneer phase of initial colonization and later establishment of a full settlement system. For instance, Beaton (1991) distinguishes between ‘‘transient explorers’’ and ‘‘estate settlers.’’ Where settlement of the desert took place early, this distinction is hard to isolate using current archaeological data. Our best chance of analyzing the process of settling a new desert is in the Americas, given the relatively short prehistory of these regions. A recent review of the early settlement of North America (Haynes 2002) suggests that this two-phase model of desert colonization may be applicable there. During the Clovis period, hunters used lake and spring sites within the desert southwest (including the Mojave Desert), but wider settlement of the desert – and adaptation to desert plant resources such as seeds – did not occur until later during the early Archaic phase. Most deserts have also seen periods of enhanced rainfall, ﬂuvial activity, groundwater discharges, and greater biological activity in

____________________________________________________________________________________

7

________________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s

the past. There is an opportunity here – provided we can get a clear deﬁnition and resolution of these changes in time and space – to look at how human societies have dealt with these periods of expanding or contracting living space and opportunities.

Outline of this Book The chapters in this volume help shape a fresh and more comparative perspective on desert archaeology. When approached by Blackwell Publishing to commission a series of comparative chapters on this topic we quickly found that, except for pockets of research, this was largely a new area. Much of the archaeological research in deserts has simply been framed in other terms – responding either to general questions embedded in national and regional prehistories, or dominated by wider debates (e.g. the great Kalahari debate; megafaunal extinctions; the emergence of complexity in hunter-gatherer society; interpretation of lithic assemblages, etc.). Only in Australia, the world’s driest continent, with more than 3,000,000 sq km of desert or drylands, has the desert become an explicit focus for research into the dispersal and adaptation of humans, and the long-term dynamics of huntergatherer settlement in desert environments. The chapters in this volume have been written to introduce a new generation of students and general readers to broad issues in the archaeology of desert huntergatherers, complementing an earlier book, The Archaeology of Drylands (Barker and Gilbertson 2000), that looked at the archaeology of agricultural societies in drylands. The focus here is on hunter-gatherers. Over half of the case studies presented draw on Australian data, reﬂecting the level of interest in desert research there. About half the chapters look at southern African or American deserts (the Great Basin, the Atacama, and Patagonia). The major gaps in global coverage are the Middle Eastern and central Asian deserts, despite the editors’ attempts to commission material on these areas. The book is organized into three parts, which are developed further as introductory comments before the relevant chapters. In summary, the parts are: Frameworks. This part provides wide-ranging discussion of key temporal, ethnographic, and interpretive frameworks employed in studies of desert huntergatherer groups from around the world. Dynamics. This part provides a range of archaeological perspectives on the longterm dynamics of desert societies. Archaeology now provides a ﬁner-grained historical picture of desert hunter-gatherers, showing long-terms shifts in economy and land use – but as these chapters caution, such changes are not necessarily incremental or directional. Interactions. Clearly, there are a range of factors shaping desert societies that are not environmental. Some are social and/or political, and derive from the position of desert societies on the margins of demographically dominant populations surrounding the deserts, and their interactions with these groups.

____________________________________________________________________________________

8

________________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e

Further themes which emerge from many of these chapters include the nature of early desert hunter-gatherer societies, the technological and organizational responses of such groups to encroaching aridity, the role of art as a mediating factor in desert occupation, the relationship between language spread and huntergatherer settlement patterns, and the timing of occupation of all desert habitats and the climatic backdrop against which people’s estates became established. One of the fundamental issues addressed by this volume is the degree to which landscapes were indeed ‘‘marginal’’ or challenging when ﬁrst explored/occupied by humans (e.g. Chapters 3 and 5). We know from regional paleoenvironmental records in Australia that climate has changed signiﬁcantly since colonization by anatomically modern humans approximately 50,000 years ago and importantly that phases of aridity have followed more lacustral regimes (and vice versa). The freshwater lakes of western New South Wales which once supported abundant ﬁsh and shellﬁsh resources are now salt lakes bordered by shifting sand ﬁelds (Johnston et al. 1998). In many regions early colonists will likely have experienced desert landscapes which were less ‘‘marginal’’ than their contemporary conﬁgurations. Indeed, it has been argued that early colonists of Australia were competent exploiters of the interior portions of the world’s most arid continent and were not effectively tethered to the coast (cf. O’Connor and Veth 2000). Critical to any such review is the global timing for desert occupation and what this implies about the competencies of anatomically and culturally modern humans. Do in fact humans occupy deserts early on, or only where these were semi-arid and less ‘‘marginal’’ landscapes? Is there evidence for occupation of all desert habitats early on? Are these occupations as early as other modern behaviors, such as the ﬁrst sea crossings (from the Wallacean islands to the landmass of Greater Australia) or the earliest dated expression of art? The complexities of addressing this question are raised in Hiscock and O’Connor’s discussion (Chapter 4) of the ambiguities found in many identiﬁcations of modernity, revealing that this is not merely an issue that can be resolved by substantive investigations of the timing of desert occupation but which is additionally entangled in our conception of modernity in humans. Another critical issue is the degree to which deserts are actually homogeneous and represent a uniform bloc of physical attributes. Do they have uniform characteristics both across subregions and through time? How real is the concept of the desert culture bloc and the conservative nature of the societies occupying such a monolithic construct? The chapters from Australia, Africa, and the Americas suggest that this assumption needs to be challenged at a number of levels. When an ecological and biogeographic approach is taken in an examination of desert systems (after Smith 1989; Veth 1993) it becomes apparent that adjacent areas are likely to have presented a variety of optimal situations for huntergatherers to establish different kinds of habitation loci, to target varied prey, to engage in different forms of residential and logistical mobility patterns, and to engage in different rhythms of aggregation and dispersion depending on local and regional climate patterns. For Aboriginal groups to structure their use of the Lake Eyre Basin of Australia, which only receives ﬂoodwaters from the north on average

____________________________________________________________________________________

9

________________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s

once every ten years, is a very different scenario to the persistence of human occupation in the central Australian ranges, where permanent water ‘‘oases’’ have been present since the Tertiary period (cf. Hughes and Hiscock in press). Not only is physical variation across desert systems arguably of relevance to hunter-gatherers; there are also the different climate histories which will have impact at both the regional and local levels and which will serve to make conditions either more marginal and ‘‘risky’’ or increase the productivity and reliability of economic resources – both inviting changes in social and technological organization. The need to accurately contextualize the archaeological record against an environmental framework has, in the past, been hampered by our rudimentary knowledge of the impacts of changing climatic conditions on arid regions across the world (Veth et al. 2000; see also Chapter 3, this volume). This is largely seen as a direct consequence of the general absence of appropriate, uninterrupted data-sets on which to base interpretations, and environmental conditions that are often not conducive to the preservation of organic materials. While dry, arid conditions can be excellent for the preservation of macroscopic organic remains in sheltered archaeological sites, they usually result in extreme degradation of fragile pollen grains in open contexts that otherwise would serve as vegetation markers allowing inferences to be made about climatic change. Reconstruction of climate based on pollen analyses in arid regions is further frustrated by the absence of useful indicator types and the generally non-speciﬁc nature of the grassland pollen types which usually dominate such assemblages. Nevertheless, in recent decades a more sophisticated understanding of environmental change in arid regions has begun to emerge. Forced by necessity, paleoenvironmental researchers interested in the arid zone have focused on sedimentary, stratigraphic, geochemical, mineralogical, and geomorphological investigations of ﬂuvial and lacustrine systems (e.g., Bowler 1998; Magee and Miller 1998; Vogel 1989). In addition, recent developments in luminescence and U-series dating have greatly assisted the study of aeolian sediments, tufa formations, and speleothems in arid landscapes (e.g., Brook et al. 1996; Brook et al. 1999). More novel data-sets relevant to paleoenvironmental research in lowaltitude deserts within Australia, South America, and North America include botanical materials preserved in stick-nest rat middens (e.g., Betancourt et al. 1990; Latorre et al. 2002; Pearson and Dodson 1993). Tufa formations, speleothems, and mud-wasp nests have also been recently recognized for their value in preserving vegetation indicators, such as pollen and phytoliths, in arid regions of the world (e.g., Burney et al. 1994; Roberts et al. 1997; Wallis 2002, 2003). Other data of relevance to reconstruction of desert environments may sometimes be available from offshore marine cores that preserve indicators such as diatoms, foraminifera, pollen, phytoliths, isotopes, and charcoal, as well as from high resolution coral records. Similarly, ice cores in or adjacent to high-altitude deserts offer unique insights into climate change, often at a much higher resolution than is typically afforded by the archaeological record (e.g., Shimida et al. 1991; Thompson et al. 1998). Proxy data from archaeological cave sites (such as faunal remains, macroﬂoral material, phytoliths, and charcoal) may also provide valuable clues as to what the

__________________________________________________________________________________

10

_______________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e

environment was like in the past (e.g., Bowdery 1998; Esterhuysen and Mitchell 1996; McConnell and O’Connor 1997; Robbins et al. 1996; Smith et al. 1996; Wallis 2001). However, the interpretation of such assemblages in terms of paleoenvironmental reconstruction is considerably more difﬁcult, owing to the potential bias in the materials present caused by anthropogenic behaviors. In summary, there is still an enormous amount of work to be carried out in placing archaeological sequences from what are now deserts in to accurately reconstructed past environmental contexts. Overall, this book is intended as an issues-based volume, rather than a series of regional overviews, but each chapter includes extended case studies illustrating key issues. A similar interdisciplinary approach (but on a different topic) was successfully adopted in The Archaeology of Prehistoric Coastlines (Bailey and Parkington 1984) and we believe that a comparable examination of human life in deserts provides an equally productive and provocative framework for future work. There is a strong international demand for knowledge about the dynamics of human settlement in arid regions. The history of humans in arid lands can be used as a tool for developing knowledge about the evolution of desert systems, and understanding environmental changes currently underway there or likely to take place in the near future. We hope this volume satisﬁes at least some of this demand. Finally, the editors wish to pay special thanks to Carrie Robson and Lynley Wallis from the Australian Institute of Aboriginal and Torres Strait Islander Studies for their expertise during ﬁnal preparation and copy-editing of the manuscript; and also to Emily Martin and Jane Huber from Blackwell Publishing. References Bailey, G. and Parkington, J. (eds) 1984: The Archaeology of Prehistoric Coastlines. Cambridge: Cambridge University Press. Barker, G. and Gilbertson, D. (eds) 2000: The Archaeology of Drylands. London: Routledge. Beaton, J. M. 1991: Colonizing continents: Some problems from Australia and the Americas. In T. D. Dillehay and D. J. Meltzer (eds), The First Americas: Search and Research, Boca Raton: CRC, 209–30. Betancourt, J. L., Van Devender, T. R., and Martin, P. S. (eds) 1990: Packrat Middens: The Last 40,000 Years of Biotic Change. Tuscon: University of Arizona Press. Bowdery, D. 1998: Phytolith Analysis Applied to Pleistocene–Holocene Archaeological Sites in the Australian Arid Zone. Oxford: BAR International Monograph Series no. 695, Hadrian Books. Bowler, J. M. 1998: Willandra Lakes revisited: Environmental framework for human occupation. Archaeology in Oceania, 33, 120–55. Brook, G. A., Cowart, J. B., and. Marais, E. 1996: Wet and dry periods in the southern African summer rainfall zone during the last 300 kyr from speleothem, tufa, and sand dune age data. Palaeoecology of Africa, 24, 147–58. Brook, G. A., Marais, E., and Cowart, J. B. 1999: Evidence of wetter and drier conditions in Namibia from tufas and submerged cave speleothems. Cimbebasia, 15, 29–39. Burney, D. A., Brook, G. A., and Cowart, J. B. 1994: A Holocene pollen record for the Kalahari Desert of Botswana from a U-series dated speleothem. The Holocene, 4 (3), 225–32.

__________________________________________________________________________________

11

_______________________________________________________________________________

S mith , V e t h, H is c oc k , W a l l i s Clobert, J., Danchin, E., Dhondt, A. A., and Nichols, J. D. 2001: Dispersal. Oxford: Oxford University Press. Esterhuysen, A. B. and Mitchell, P. J. 1996: Paleoenvironmental and archaeological implications of charcoal assemblages from Holocene sites in western Lesotho, southern Africa. Palaeoecology of Africa, 24, 233–40. Forman, T. T. 1995: Land Mosaics: The Ecology of Landscapes and Regions. Cambridge: Cambridge University Press. Gamble, C. 1993: Timewalkers: The Prehistory of Global Colonization. Phoenix Mill: Alan Sutton Publishing. Gould, R. A. 1969: Subsistence behavior among the Western Desert Aborigines of Australia. Oceania, 39, 253–74. Haynes, G. 2002: The Early Settlement of North America. Cambridge: Cambridge University Press. Hughes, P. and Hiscock, P. in press: The archaeology of the Lake Eyre South area. In Aboriginal Occupation of the Lake Eyre South Region: Ethnography and Archaeology, Adelaide: Lake Eyre Monograph Series no. 6, Royal Geographical Society of South Australia. Johnston, H., Clarke, P., and White, J. P. (eds) 1998: The Willandra Lakes: People and Palaeoenvironments. Sydney: Archaeology in Oceania no. 33 (3), Oceania Publications. Kelly, R. L. 1995: The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, DC: Smithsonian Institution Press. Latorre, C., Betancourt, J. L., Rylander, K. A., and Quade, J. 2002: Vegetation invasions into absolute desert: A 45,000 year rodent midden record from the Calama-Salar de Atacama basins, northern Chile (lat 220 – 240S). Geological Society of America Bulletin, 114 (3), 349–66. Lee, R. B. 1979: The !Kung San: Men, Women, and Work in a Foraging Society. Cambridge: Cambridge University Press. Lee, R. B. and Daly, R. 1999: The Cambridge Encyclopedia of Hunters and Gatherers. Cambridge: Cambridge University Press. Lee, R. B. and DeVore, I. (eds) 1968: Man the Hunter. Chicago: Aldine. McConnell, K. and O’Connor, S. 1997: 40,000 year record of food plants in the southern Kimberley ranges, Western Australia. Australian Archaeology, 45, 20–31. Magee, J. W. and Miller, G. H. 1998: Lake Eyre paleohydrology from 60 ka to the present: Beach ridges and glacial maximum aridity. Palaeogeography, Palaeoclimatology, Palaeoecology, 144, 307–29. Mares, M. A. 1999: Encyclopedia of Deserts. Norman: University of Oklahoma Press. Middleton, N. and Thomas, D. 1997: World Atlas of Desertiﬁcation (2nd edn). London: Arnold. O’Connor, S. and Veth, P. (eds) 2000: East of Wallace’s Line: Studies of Past and Present Maritime Cultures in the Indo-Paciﬁc region. Rotterdam: Modern Quaternary Research in South East Asia no. 16, A. A. Balkema. Panter-Brick, C., Layton, R. H., and Rowley-Conwy, P. 2001: Hunter-Gatherers: An Interdisciplinary Perspective. Cambridge: Biosocial Symposium Series no. 13, Cambridge University Press. Pearson, S. and Dodson, J. R. 1993: Stick-nest rat middens as sources of paleoecological data in Australian deserts. Quaternary Research, 39, 347–54. Peterson, N. 1979: Territorial adaptations among desert hunter-gatherers: The !Kung and Australians compared. In P. Burnham and R. Ellen (eds), Social and Ecological Systems, London: Academic Press, 111–29.

__________________________________________________________________________________

12

_______________________________________________________________________________

Gl ob a l D e s e r t s i n P e r s p e c t i v e Robbins, L. H., Murphy, M. L., Stevens, N. J., Brook, G. A., Ivester, A. H., Haberyan, K. A., Klein, R. G., Milo, R., Stewart, K. M., Matthiesen, D. G., and Winkler, A. J. 1996: Paleoenvironment and archaeology of Drotsky’s Cave: Western Kalahari Desert, Botswana. Journal of Archaeological Science, 23, 7–22. Roberts, R. G., Walsh, G., Murray, A., Olley, J. M., Jones, R., Morwood, M., Tuniz, C., Lawson, E., Macphail, M., Bowdery, D., and Naumann, I. 1997: Luminescence dating of rock art and past environments using mud-wasp nests in northern Australia. Nature, 387, 696–9. Rockman, M. and Steele, J. (eds) 2003: Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation. London: Routledge. Schrire, C. (ed.) 1984: Past and Present in Hunter Gatherer Studies. Orlando, FL: Academic Press. Shigesada, N. and Kawasaki, K. 1997: Biological Invasions: Theory and Practice. Oxford: Oxford University Press. Shimida, I., Schaff, C. B., Thompson, L. G., and Mosley-Thompson, E. 1991: Cultural impacts of severe droughts in the prehistoric Andes: Applications of a 1,500 year ice core precipitation record. World Archaeology, 22 (3), 247–70. Smith, E. and Winterhalder, B. (eds) 1992: Evolutionary Ecology and Human Behavior. New York: Aldine. Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Smith, M. A., Vellen, L., and Pask, J. 1996: Vegetation history from archaeological charcoals in central Australia: The late Quaternary record from Puritjarra rock shelter. Vegetation History and Archaeobotany, 4, 171–7. Spencer, B. and Gillen, F. J. 1899: The Native Tribes of Central Australia. London: Macmillan. Steward, J. 1938: Basin-Plateau Aboriginal Sociopolitical Groups. Smithsonian Institution, Bureau of American Ethnography Bulletin 120. Washington, DC: Government Printing Ofﬁce. Thompson, L. G., Davis, M. E., Mosley-Thompson, E., Sowers, T. A., Henderson, K. A., Zagorodnov, V. S., Lin, P. N., Mikhalenko, V. N., Campen, R. K., Bolzan, J. F., Cole-Dai, J., and Francou, B. 1998: A 25,000 year tropical climate history from Bolivian ice cores. Science, 282, 1858–64. Veth, P. M. 1993: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory, Archaeological Series no. 3, Ann Arbor, MI. Veth, P. M., O’Connor, S., and Wallis, L. A. 2000: Perspectives on ecological approaches in Australian archaeology. Australian Archaeology, 50, 54–66. Vogel, J. C. 1989: Evidence of past climate change in the Namib Desert. Palaeogeography, Palaeoclimatology, Palaeoecology, 70, 355–66. Wallis, L. A. 2001: Environmental history of northwest Australia based on phytolith analysis at Carpenter’s Gap 1. Quaternary International, 83–5, 103–17. Wallis, L. A. 2002: AMS dates and paleoenvironmental data from mud-wasp and bird nests at Carpenter’s Gap 1, northern Australia. Australian Archaeology, 55, 35–9. Wallis, L. A. 2003: Phytoliths and other microfossils in tufa formations as a novel source of palaeoenvironmental data. In D. M. Hart and L. A. Wallis (eds), Proceedings of the Stateof-the-Art in Phytolith and Starch Research in the Australian, Paciﬁc, and South East Asian Regions, The Australian National University: Pandanus Press, 31–42. Wilmsen, E. N. 1989: Land Filled with Flies. Chicago: University of Chicago Press.

__________________________________________________________________________________

13

_______________________________________________________________________________

-------------------------------------------------------------------

Part I

----------------------------------------------------------------

Frameworks

The authors of part one provide a series of fundamental chronological, ethnographic, and interpretive frameworks that reveal patterns of desert hunter-gatherer life. The case studies are diverse, with Thomas Widlok (Chapter 2) employing African and Australian ethnographies; Peter Hiscock and Lynley Wallis (Chapter 3) focusing on the archaeology of Australian foragers; and Peter Hiscock and Sue O’Connor (Chapter 4) comparing the African and Australian archaeological sequences as a way of evaluating explanatory frameworks for tool assemblage changes in deserts and beyond. Despite the different case studies presented in these chapters a series of underlying themes connect them, and these themes are reinforced by their presentation and representation in diverse explorations of desert life. For instance, each of the chapters contains an exploration of the reality of the ‘‘eternal character’’ (Widlok) so often attributed to deserts. The illusory nature of images of unchanging and invariant adaptations to uniformly harsh and static desert environments is clearly explicated in the different arguments of each chapter. While Hiscock and Wallis (Chapter 3) discuss the consequences of long-term climatic shifts for human occupation in the late Pleistocene, Widlok (Chapter 2) neatly explores the complex interaction of social alterations in hunter-gatherer practices in recent times and the nature of anthropological activities. Rejection of normative and ﬁxed pictures of desert life is essential in our search to understand the complexity of human existence in deserts across the globe. In their consideration of cyclical archaeological phenomena, Hiscock and O’Connor (Chapter 4) raise critical questions about the explanation of archaeological changes and the complexities of employing ethnographic models for ancient archaeological materials. These contributions explicitly raise questions about the way in which ethnographic information can be employed in reconstruction of desert lifeways, and challenge archaeologists to develop a critical awareness of the issues involved in employing ethnographic information as an aid to archaeological interpretation. The themes of part

___________________________________________________________________________________

15

_______________________________________________________________________________

Frameworks

one – awareness of the variability in desert life and of the intricacy of employing ethnographic images of that life in archaeological reconstructions – are also displayed in the chapters studying the long-term dynamics of human life in arid lands.

___________________________________________________________________________________

16

_______________________________________________________________________________

-------------------------------------------------------------------------------

2

----------------------------------------------------------------------------

Theoretical Shifts in the Anthropology of Desert Hunter-Gatherers Thomas Widlok

Introduction: Hunter-Gatherer Studies and Anthropological Theory Hunter-gatherer studies have been critical to anthropological thinking in different ways and at various points in the history of the discipline. There have been discernible shifts in the relationship between hunter-gatherer studies, anthropological theory, and public imagery for well over 150 years (Barnard 1999). Huntergatherers played an important role in the imagination and theorizing of many thinkers of the enlightenment period and more speciﬁcally in the development of various kinds of evolutionary theories. These emerging relationships were fueled and intensiﬁed through early ethnographic work on hunter-gatherers, such as that of Boas (1888) and Spencer and Gillen (1899). However, by the time that the practice of extended ﬁeldwork was ﬁrmly established in social and cultural anthropology, interest had moved away from hunter-gatherers to the study of villages or segmentary clan structures. None of the classic ethnographic monographs in the ‘‘golden age’’ of British social anthropology, for instance, deal with a hunter-gatherer group (for example, Radcliffe-Brown’s 1922 study on the Andamans is far less widely read than his other work). This was due to the mutually reinforcing relations between the early twentieth century theoretical focus on social and cultural structures and the preparation of village case studies exemplifying these structures. Scholarly fascination with hunter-gatherers seems to have only returned in full force towards the latter half of the twentieth century. Although this shift was initially against dominant anthropological interests in issues such as complex symbolic systems and lineage structures, it became increasingly informed by new theoretical interests in process, ﬂexibility, and the dynamics of social and cultural forms.

___________________________________________________________________________________

17

_______________________________________________________________________________

T h o m a s Wi d l o k

There are good historically oriented surveys that outline this broad pattern of the shifting role of hunter-gatherer studies within anthropology (e.g., Barnard 1983; Bender and Morris 1988). This chapter will highlight some of the key features that characterize theoretical shifts in the anthropology of desert huntergatherers and the ways in which these are linked to more general developments in hunter-gatherer studies and in anthropology.

Entering the Field – Leaving the Desert When anthropological work based on long-term ethnographic ﬁeld research became the standard research strategy in the early twentieth century, ironically many desert hunter-gatherers were moving in the opposite direction, leaving their home habitat and coming to live as fringe dwellers at the margins of the colonized world. At the same time many anthropologists, and especially those who relied on diffusion to explain the spread and change of culture, continued to talk about desert hunter-gatherers as ‘‘retreating’’ into the desert. The characterization as Rest-und Ru¨ckzugsvo¨lker (remaining and retreating peoples), as they were known in German anthropology (cf. Schott 1956), led to one of the ﬁrst major theoretical debates in the study of desert hunter-gatherers. In the ﬁrst Kalahari Debate, branches of the German culture history school of anthropology were divided on the issue of whether or not the Bushmen of the Kalahari were ‘‘secondary primitives.’’ Were the Bushmen dislocated remnants of a more complex original hunter-gatherer culture in the ecologically more fertile areas of the Cape or, instead, ‘‘original’’ hunter-gatherers (see Szalay 1986)? The original scenario that Bushmen had retreated to previously uninhabited desert regions was seriously ﬂawed mainly because there was no evidence that the communities of living hunter-gatherers of the Kalahari had moved there in recent times. Indeed, there is archaeological evidence to clearly show that the opposite was the case (see Chapter 11, this volume). What remains a critical issue to this day, however, is to what extent their desert dwelling lifestyle is representative for a hunter-gatherer way of life more generally, especially with regard to the hunter-gatherers who have inhabited the moderate and cold zones in Europe in the remote past. The fact that most hunter-gatherers in more moderate climates were culturally and physically destroyed or absorbed before systematic ethnographic ﬁeld research began is one of the major and given limitations for research in this ﬁeld. The question of representativeness is therefore one that can only be resolved in small steps with an investigation into the details of various aspects of hunter-gatherer life and with full recognition of the large spectrum of historical and regional variation that has undoubtedly occurred. Speculations about a ‘‘retreat’’ of hunter-gatherers into the remote deserts notwithstanding, many hunter-gatherers de facto moved out of the desert as ethnographers began to move into the ﬁeld to gain a better understanding of this way of life. For instance, one of the most inﬂuential representations of desert hunter-gatherers, namely Ian Dunlop’s oeuvre of ﬁlms, was made possible only because the ﬁlmmakers took two Aboriginal families back into the desert that they

___________________________________________________________________________________

18

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

had decided to leave. In the 1960s the last remaining desert dwelling Aborigines were leaving their traditional lands in the Western Desert, although some have returned more recently under changed conditions (Tonkinson 1991; see also Berndt and Berndt 1979). This exodus was partly for reasons to do with drought, but also because the continuous trickle of people into settlements had gained momentum as those remaining in the desert found it increasingly difﬁcult to ﬁnd appropriate marriage partners (Gould 1969). Dunlop’s team initially located a Western Desert family of ten individuals who had experienced minimal contact with European settlers. This group left after three days, however, and a newly constituted group from Warburton Mission returned to carry out various subsistence activities (Tindale 1968). Blankets, tins, and rubbish were removed from the scene before ﬁlming started (Dunlop 1966a: 43). The result of the ﬁlming was People of the Australian Western Desert – Dunlop’s series of short thematic ﬁlms – and Desert People, the synthesized account of one day in the desert based on the same material. In a striking parallel in southern Africa, John Marshall achieved a similar role as ﬁlmmaker in the ﬁeld of ‘‘Bushman’’ or San studies (cf. Ruby 1993). His work was based on expeditions by the Marshall family and associated researchers into remote parts of the Kalahari Desert. The Marshalls, too, at this stage were thinking of the Ju/’hoansi as supporting themselves ‘‘with the economy that cradled human evolution.’’ Marshall, like Dunlop, was initially ‘‘cleaning tin cans out of shots to make Ju/’hoansi in Nyae Nyae look real.’’ He later attributed the fact that he was cutting out the presence of what seemed to be ‘‘non-hunter-gatherer’’ features to his ‘‘illusion of ﬁlming in the Pleistocene’’ (Marshall 1993: 23, 32, 56). The Marshalls’ representation of San life, similar to Dunlop’s representation of Australian desert Aborigines, remains very inﬂuential. The ﬁlm and photographic material that they have produced is still used to illustrate a wide range of ethnographic works on southern African and Australian hunter-gatherers. These images extend temporally as well as spatially beyond the setting in which the material was originally collected. In other words, they came to represent not only ethnographic and archaeological work of later decades, but also the San and Aborigines beyond the desert environment. General works on Aboriginal people (e.g., Elkin 1981) and Bushmen (e.g., Barnard 1978) use photographic material from desert settings produced by Dunlop and the Marshalls, respectively. Moreover, their style of representation, showing hunters in their habitat without reference to the world of non-hunter-gatherers, dominated ethnographic works until the 1980s (see the work of Lee 1999a, Silberbauer, and Valiente-Noailles for southern Africa and Meggitt, Gould 1969, and Tindale 1968 for Australia). The visual media, therefore, not only had a particular impact on public imagination and made these hunting and gathering people famous, they also provided a template and a prototypical image of what African or Australian hunter-gatherers are like. Anthropologists working in the tropical north of Australia and outside the Kalahari Desert in southern Africa have occasionally objected to this strong prototypical image, which makes the groups they are dealing with appear either to be ‘‘abnormal’’ or ‘‘hybrid’’ if they do not conform to it. In part, the ﬁercenes of the revisionist critiques of the 1980s and 1990s may

___________________________________________________________________________________

19

_______________________________________________________________________________

T h o m a s Wi d l o k

also be seen as reactions against the way in which stereotyped and harmonious images of remote desert dwellers came to dominate not only ﬁrst year anthropology courses, in general, but also hunter-gatherer studies more speciﬁcally. Another effect of southern African Bushmen and central Australian Aborigines becoming prototypical foragers is that other regions of the world came to be underrepresented in comparative works in hunter-gatherer studies. While today there is again a resurgence of tropical and arctic forager studies, especially with Siberian research since the collapse of the Soviet Union, current desert huntergatherer ethnography is predominantly located in Australia and southern Africa (cf. Burch and Ellanna 1994; Schweitzer et al. 2000; Williams and Hunn 1982). Studies of North American desert hunter-gatherers, which played an important role up to the 1950s in the works of Steward, Eggan, and others (see contributions in Lee and DeVore 1968), have largely disappeared from current debates. Within North America as an ethnographic region, the Northwest Coast overshadows the role that Plains or Californian hunter-gatherers may have had in anthropological theory of the past. This is because the Northwest Coast provides a theoretically productive contrast to hunter-gatherers elsewhere and because, comparatively speaking, very little ethnographic work is currently carried out among desert hunter-gatherers in North America (see Fowler 1999) such that anthropological work in the region is largely reduced to archaeological work. Another issue worth considering is that because North American desert Indians were equipped with horses and ﬁrearms and entangled in intensive trade relations they were no longer considered to be foraging peoples. Kent (2002) has recently argued that, by the standards now applied to southern Africa, most North American foragers would not be considered hunter-gatherers because they include some material culture otherwise associated with farmers. Kent uses this comparison to criticize those who deny a hunter-gatherer identity to occupants of sites that show any pastoral or horticultural products, such as the bones of cattle and sheep or the remains of pottery (see also Chapter 11, this volume). What is true at the level of a continent may also apply within regions. Kalahari Bushmen studies came to stand for southern African hunter-gatherer studies more generally, with little attention given for instance to Namib foragers (who could no longer be studied ethnographically) or to hunter-gatherers with a long record of interaction with agropastoralists such as the Hai//om (see Widlok 1999). In Australia the situation is slightly more complicated, given a growing presence and dominance of Arnhem Land Aborigines in anthropology and in public life (as in art and music performances) over the last decade. There is also the noticeable and growing inﬂuence of desert cultures in areas such as the southern Kimberley region of northwest Australia (see Kolig 1989: 66; Widlok 1992). The ethnographic work on desert peoples collected in the twentieth century continues to be an important and irreplaceable basis for anthropological studies of desert hunter-gatherers. The 1960s and 1970s have not only produced outstanding individual contributions of lasting value, but the density of work during this period also provides the opportunity to pursue comparative questions and to see recurrent patterns in theoretical debates.

___________________________________________________________________________________

20

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

Research Approaches and Comparative Theory Comparison relies on a theoretical framework that allows deﬁning ‘‘cases’’ and the parameters according to which the cases are compared. Along with larger developments in anthropology these frameworks have shifted over time and so has the classiﬁcation of frameworks. One of the most recent classiﬁcations of approaches within hunter-gatherer studies distinguishes ‘‘classic,’’ ‘‘adaptionist,’’ ‘‘revisionist,’’ and ‘‘indigenist’’ perspectives (Lee and Daly 1999a), a distinction that can also be applied to the study of desert foragers. ‘‘Classic’’ approaches in hunter-gatherer studies today, in Lee and Daly’s sense, do not simply continue old topics but have a holistic interest in hunter-gatherers, covering history, ecology, economy, settlement patterns, social and political organization, and religion (see contributions in Lee and Daly 1999b). Characteristically, research highlights the internal working between different cultural domains, that is between social or economic stress and ritual activity, between property rules and egalitarian social organization, and between land use patterns and forms of corporate organization. Work is based on the assumption that these questions can be pursued comparatively, that key terms and categories can be transferred from one context to another. Systematic comparisons involving a well deﬁned sample of cases for testing hypotheses are nevertheless rare. Instead, comparisons are exploratory, and are aimed at clarifying conceptual tools and elucidating the speciﬁcities of one case vis-a`-vis others. These comparisons can involve desert versus non-desert foragers, but also two or more desert forager groups. A typical theoretical result of this approach is the critique of the patrilocal band model (see Hiatt 1996) and the distinction between range (land used by coresidents) and estate (land owned by a descent or totemic group). ‘‘Adaptationists’’ may base their work on the ‘‘classic’’ record. In practice, however, there is a clear focus on areas of research most closely related to hunting and gathering subsistence and most amenable to science-oriented methodology, above all, quantiﬁcation. Recurrent research topics are mobility patterns (dispersal and aggregation), the exchange and distribution of resources, demographic questions of group size and reproduction, the division of labor, and time allocation (cf. Kelly 1995). Typically, research stresses interaction with the natural environment and comparisons involving a larger number of cases are frequently made. Comparisons are explicitly utilized to test existing theories or to develop new theories. The cases involved are often situated in similar habitats, but not necessarily so, and comparisons may also be based on simulation. A typical example is the critique of Sahlins’ notion of ‘‘original afﬂuence’’ on the basis of simulated production, distribution, and consumption dynamics which show that, under certain demographic and ecological conditions, foraging more will sustain less people because of resource depletion (Winterhalder 1993). ‘‘Revisionists’’ were often converted to this approach, after having worked within the classic genre, in that they have reinterpreted their own data as well as that of their adversaries (cf. Solway and Lee 1990; Wilmsen and Denbow 1990). While

___________________________________________________________________________________

21

_______________________________________________________________________________

T h o m a s Wi d l o k

revisionist debates concerning hunter-gatherers in tropical forests were ﬁrmly grounded on ecological arguments, such as the suggested ‘‘impossibility’’ of surviving in forests solely by hunting and gathering, the revisionism concerning desert foragers turned away from ecological arguments towards questions of outside contact leading to political and economic domination. Characteristically, revisionist research foregrounds interaction with dominant non-hunter-gatherer societies (agropastoralists, traders) and its effects. Revisionist arguments of this sort are not necessarily new (see Passarge 1997) and not limited to the Kalahari or to questions of economic dominance (see Swain 1993 for a revisionist debate in Australia on issues of religion). Comparisons are not so much drawn between different huntergatherer cases (since this category is rejected) but between different groups within a region. A useful example is a critical examination of prevalent views about Bushmen, including the seemingly neutral typology into ‘‘tame’’ (farm) and ‘‘wild’’ (veld) Bushmen. These categories have been used by colonists to impose their interests on San who came to be named as cattle thieves and blamed for loss of livestock, the misappropriation of whose land was justiﬁed through the process of ‘‘domestication,’’ and whose ‘‘instinct’’ was exploited for psychological warfare (Gordon 1992: 90). The ‘‘indigenist’’ approach, in the sense described by Lee and Daly (1999a), is not to be confused with the Indigenous perspective, since there are still very few desert foragers who are also anthropologists. Rather, the indigenist approach puts ‘‘the people studied, their goals and aspirations, ﬁrmly into the center of the scholarly equation’’ (Lee and Daly 1999b: 11). In other words, this approach would cover much of what is now known as applied anthropology. This approach has become strong in hunter-gatherer studies for a number of reasons, such as the vulnerability of many of these groups to increasing land pressure and the strategies adopted by these groups to exploit their contacts with outsiders. Characteristically, this research is carried out as consultancy work, highlighting relations between (former) foragers and state or non-governmental agencies and their effects on internal social relations. This may involve a comparative perspective insofar as similar nation-states (such as Australia and Canada) lend themselves to comparison, but also because the descendents of hunter-gatherers now visit one another and meet at international Indigenous rights conferences. An interesting recent example is the joint work by an Indigenous North American Indian, a therapist/ consultant, and an anthropologist on San spirituality (Katz et al. 1997). This project resulted in young Ju/’hoansi generating new knowledge about healing, which then led the authors to emphasize that Indigenous culture is not static, nor is the way in which it was practiced in the past or by the elders more pristine than in the present. The foregoing discussion is but one way of classifying dominant theoretical trends in hunter-gatherer studies and is, at least implicitly, based on the point in time when these schools of thought came into being. It is true that these approaches have emerged from a process of mutual delimitation. That is to say, it only makes sense to talk about a revisionist direction in relation to a classic perspective and vice versa. Moreover, within each of these approaches there is a

___________________________________________________________________________________

22

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

considerable spectrum of views and tendencies. For instance, typologies of huntergatherers, in themselves a feature of the classic approach, differ considerably (cf. Barnard 1983). In most cases, ‘‘desert foragers’’ is not an analytic category; instead, various desert hunter-gatherers are grouped differently within different typologies and cut across major distinctions such as that between immediate versus delayedreturn foragers (see Woodburn 1982). Therefore, there is not only a ‘‘foraging spectrum’’ (Kelly 1995) covering a diversity of hunter-gatherer groups, but there is also a spectrum of forager studies arising from the work of individual scholars which is not always easily subsumed under a single approach. It is reasonable to conclude that for every tendency within anthropology at large an example can be found in the ﬁeld of desert hunter-gatherer studies. Although debates have at times been ﬁerce, and despite the declared intention of major proponents to either debunk an old paradigm or defend appropriate research procedures, I am not persuaded that shifts between the foregoing approaches can be construed as paradigm shifts at a macrolevel. I argue, below, a new paradigm has yet to emerge from the current situation of uncertainty and of dissatisfaction with theoretical approaches of the past. The existing approaches show considerable continuities and share some of their premises with one another. There are clear connections between the classic approach and that of the adaptationists because both were inspired by ideas arising in evolutionary ecology. At the same time, a large number of the classic approaches relied, directly or indirectly, on Marxist theories. They shared many of the same assumptions as the political economy oriented work of the revisionists, although the two schools disagree, for instance, on how to identify classes and underclasses. All three approaches share a materialist perspective with a tendency towards economic determinism, even though proponents may arrive at diverging conclusions depending on whether internal dynamics or external pressures are considered to be crucial. The indigenist perspective most clearly points at something radically new, but in practice it is frequently an extension of the existing research agenda rather than a clear break with ideas from the past. For instance, existing interests in exchange and sharing have been successfully extended to accommodate questions dealing with cash ﬂow and commoditization (see Peterson and Matsuyama 1991). Current issues like deforestation and Indigenous human rights are investigated through the framework of human ecology (Headland and Blood 2002). There are clearly continuities between these approaches. Recent revisionist debates, which have occasionally been announced as challenging the framework of hunter-gatherer studies, have in fact brought together research from archaeology, linguistics, and sociocultural anthropology in a classic fashion. While the category ‘‘hunter-gatherers’’ has been questioned at regular intervals (see, for instance, Burch 1994), the ﬁeld of hunter-gatherer studies remains a vital one. Some of the theoretical models, such as that of the patrilocal band proposed by Service and Steward, have recently been critically revisited (BirdDavid 1994) after having fallen into disuse for several decades, not least because their prime ethnographic cases turned out to be anomalous and aberrant (see Barnard 1983).

___________________________________________________________________________________

23

_______________________________________________________________________________

T h o m a s Wi d l o k

While the polemics of the Kalahari Debate may continue, arguing about how ‘‘deep’’ or ‘‘remote’’ the Kalahari Desert was at different points in time, it is clear that revisionism is much less of a serious challenge than was initially thought. Archaeological evidence from sites at the margin of the desert show that despite prolonged contact there is considerable continuity between Later Stone Age Indigenous hunter-gatherers of the area and the historically known Bushman population (Sadr 2002). Similarly, comparative ethnohistorical evidence from areas with a long record of interaction between hunter-gatherers and their neighbors suggests overwhelming continuity for a long time in the historical record, in terms of material culture and the persistence of certain social institutions (see Marlowe 2002). As for the situation in northern Australia, where historical contact with non-foragers is now considered more seriously than it used to be (e.g., Povinelli 1993; see also Chapter 15, this volume), the point has been made that the question to be raised is not so much why the Aborigines did not engage in agriculture but, rather, ‘‘Why should they?’’ (Triesch 2001). In this regional context researchers ﬁnd it much less problematic to accept that hunting-gathering and horticulture are historically a continuum, and also represent a gradient extending across the Torres Strait to Papua New Guinea (Rumsey 2001). Thus, while the Kalahari Debate became heated on the question as to whether certain groups in the northern or central Kalahari had contact with non-foragers, it lost sight of the fact that contact in itself does not say anything about the relations between the parties involved or about the effects and implications of that contact. By contrast, focusing so much on the question of contact, revisionists at least implicitly subscribe to the notion of pristine and pure hunter-gatherers who must have lived some time somewhere – but not in the Kalahari (see Marlowe 2002). As Barnard and Taylor (2002: 246) have put it: ‘‘It is as if exposure to a non-foraging lifestyle should inevitably lead to cultural corruption and social instability.’’ This point is of particular relevance to desert hunter-gatherers insofar as a lack of contact cannot be considered necessary or sufﬁcient for drawing analogies between contemporary hunter-gatherers and those of the distant past. In other words, depending on the research question under consideration (diet, labor organization, settlement patterns), it may be valid and insightful to compare even ‘‘secondary foragers’’ (who previously had livestock) with those of the past, while taking account of the fact that the pace of change cannot be easily predicted and need not be the same for every arena of life and for every environment. The representativeness of contemporary hunter-gatherers (or some aspects of their lives) for situations in the past is no longer an unquestioned assumption or a self-inﬂicted taboo, but has become a ‘‘normal’’ research question, that is one that has to be empirically investigated and argued about. In view of this situation of diversity, the theoretical landscape of hunter-gatherer studies is appropriately depicted as neither a mosaic of isolated schools nor as a ladder of progress, but as a dynamic research ﬁeld. A new emerging paradigm that may help solve the diversity of problems at hand will necessarily be much broader than the four approaches listed above. Such a paradigm shall be called, for want of better terms, a ‘‘reﬂecting’’ and an ‘‘inﬂecting’’ paradigm.

___________________________________________________________________________________

24

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

A Reﬂecting Mode of Research Despite claims from the revisionists to the contrary, reﬂexivity in desert huntergatherer studies is not a recent phenomenon. Given the prominent place that deserts have in European thinking, there has always been an element of reﬂexivity in desert studies. European naturalists have been attracted by the ‘‘eternal character’’ of the deserts, which seem to preserve the past particularly well but at the same time seem almost paradoxically to be particularly fragile environments. This ambivalence is also found in the popularized anthropology of the deserts. Deserts and the cultural life of desert people have been presented as particularly pristine and harsh and at the same time the fragility of this environment, and the people who are adapted to it, have been highlighted (cf. Marshall-Thomas 1960; Mountford 1950). Given the relative absence of ‘‘third parties,’’ apart from the desert dwellers and the anthropologists themselves, researchers seem to have been much more aware of the inﬂuence they had on the people they were researching. For instance, when Dunlop was ﬁlming the Aboriginal man Minma digging out a bandicoot he was wondering whether the hunter would have shown the same effort in a situation where there was no ﬁlm crew (Dunlop 1966b; see also Marshall 1993). In other words, the mere presence of the ﬁlm crew, however non-intrusive it aimed to be, may have altered the record, leaving a wrong impression about the very basic subsistence activity of digging up a bandicoot. James Woodburn, who was working with the Hadza in Tanzania in the same period, had similar questions about the effects of his presence, especially while he was shooting game for them (see Widlok 2002). Recurrent considerations of this sort are not merely a matter of anthropological ethics or of methodological concern. Rather, they go together with questions concerning the resilience of hunting and gathering in a world of nonhunter-gatherers more generally. It has recently been pointed out that the attention the Hadza have received from anthropologists has probably helped them to continue their hunting and gathering lifestyle for longer than would have otherwise been possible. At the same time, ironically, the increasing popular interest in the Hadza as hunter-gatherers is likely to lead to a demise of hunting and gathering subsistence activities and related social and cultural features, because some Hadza have begun to subsist on regular tourist visits. Similarly, the struggle of southern African hunter-gatherers for recognition by governments and by non-governmental organizations, initiated mainly by anthropologists, certainly has had an effect on policy making in southern Africa and on the policy of donor agencies in the region (cf. Marshall and Ritchie 1986). The integration of human rights and Indigenous peoples’ networks has also produced its speciﬁc dilemmas. As Ingold (2000: 133) has pointed out, huntergatherers who claim special rights to their land because of a relationship of immediacy, sharing, and personalization now have to adopt the hegemonic genealogical model of transferring rights to land by descent in an attempt to make their claims heard with regional, national, and international power holders. Thus, there are feedback processes whereby hunter-gatherer life is changed by the dominant

___________________________________________________________________________________

25

_______________________________________________________________________________

T h o m a s Wi d l o k

environment, not necessarily through direct force or domination, although this has also occurred frequently, but simply because in a nationalized and globalized setting there may be more to be gained by becoming like the ‘‘others.’’ At the same time, there is a strong tendency within anthropology, and within some NGO development circles to advocate that cultivating difference is the more promising prospect for people with a hunter-gatherer background (cf. Gowdy 1998). With opportunities for ﬁeld research in a hunting and gathering context dwindling, social and cultural anthropologists have given increasing attention to questions of stereotyping, myth making, and interethnic relations, not only with regard to interaction between hunter-gatherers and the (colonial/national) state, but also between hunter-gatherers and (agro-) pastoralists. While these questions have always been a topic in hunter-gatherer studies, there seems to be a marked divide now between researchers with an interest in transitions (primarily from foraging to farming) as opposed to those with an interest in interactions (again, primarily between foragers and farmers). Potentially, both questions are of interest to archaeologists and anthropologists alike, but de facto ethnographers tend to think of foraging and farming as interactive systems, while archaeologists tend to ask questions about long-term transformation. Only one strand of research within social and cultural anthropology has so far remained largely unaffected by the reﬂective questions outlined above and by the revisionism of the last two decades, and that is optimal forager theory (see Martin 1983). This approach focuses on questions of demography and the systemic features of foraging as an adaptive system and has a special interest in desert foragers as an extreme form of adaptation. Although this strand of research has a long tradition in hunter-gatherer studies and continues to thrive, it is today a minority position in hunter-gatherer studies. The breadth of (desert) huntergatherer studies may be usefully viewed within the framework of international conferences on hunting and gathering societies (CHAGS), which have become a major bracket that uniﬁes the ﬁeld and that provides a founding genealogy for its constituent discussions and debates. The genealogy is usually traced back to the Man the Hunter conference of 1966 (Lee and DeVore 1968) and includes nine CHAGS conferences and numerous conference volumes which cover all regions of the world (desert, tropical, and arctic foragers) utilizing a wide spectrum of theoretical approaches. With few exceptions, CHAGS volumes include some contributions to optimal forager theory (or other strongly science oriented research) as well as humanities oriented work, but in more recent volumes there is a clear tendency towards the latter and recent CHAGS have always made an effort to include Indigenous voices. An increase in reﬂectivity does not automatically or necessarily mean a decrease in other research activities (cf. Barnard 1992). Indeed, reﬂectivity in huntergatherer studies is not altogether new (see above), and pursuing other research activities does not necessarily mean propagating old theories (see below).

___________________________________________________________________________________

26

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

An Inﬂecting Mode of Research While reﬂecting requires the researcher to alter and adapt research tools referring back to the relationship between researcher and the researched, ‘‘inﬂecting’’ requires the researcher to alter and adapt research tools to a variety of contexts to which they can be applied. To illustrate the point, consider the question of huntergatherer mobility, which is a major issue for ethnography and archaeology. The matter has been succinctly put by Brody (2001: 7): [A] crucial difference between hunter-gatherers and farmers is that one society is mobile . . . whereas the other is highly settled, tending to stay ﬁrmly in one particular area or territory. This difference is established in stereotypes of ‘‘nomadic’’ hunters and ‘‘settled’’ farmers. However, the stereotype has it the wrong way round. It is agricultural societies that tend to be on the move; hunting peoples are far more settled.

Brody provides a good example of reﬂecting about the relationship between hunter-gatherers and farmers (and anthropologists coming from a society with a farming background). He then goes on to invert the dominant stereotype depicting the farmer as ‘‘displaced from his home, roaming the harsh earth looking for land to till’’ (Brody 2001: 76). While this, again, may simply be another stereotype, it does point to a major issue in hunter-gatherer studies. In southern Africa the agropastoralists consider themselves fairly recent arrivals in the area, while most San groups maintain that they have been in the same region all along. This is also reﬂected in the way in which maps are drawn by, or on behalf of, agropastoralist groups in the region, which look very different from maps drawn by, and for, San groups. The former use arrows to indicate migration routes and the history of settlement; the latter show regular, partly overlapping paths that radiate out from waterholes like rays of the sun. In the process of land claim politics, these interlocking routes are represented as blocks of land, similar to the way in which farms and ‘‘native homelands’’ have been mapped on European maps (see Widlok 1999). The same process can be observed in Aboriginal Australia, where the delimitation of territories has led to similar discussions (cf. Sutton 1995). Aboriginal maps with interlocking routes have – under legislative pressure and with the assistance of anthropologists – been changed into maps with ideally nonoverlapping boundaries (Sutton 1995: 36). However, this reﬂexivity about dominant images and practices of mobility and the hunter-gatherer response needs to be complemented with new research that ‘‘inﬂects’’ some of the parameters that have been identiﬁed. Among these is the fact that many hunter-gatherers exhibit particular skills of spatial orientation (see Widlok 1997). Comparative research examining data from southern Africa as well as Australia (Pederson et al. 1998) suggests that in both cases foragers operate with a strong absolute orientation frame. Excellent orientation skills seem to be linked to topographical gossip in which relative orientation terms such as left/right

___________________________________________________________________________________

27

_______________________________________________________________________________

T h o m a s Wi d l o k

and front/back are replaced with absolute (i.e., speaker-independent) terms, especially cardinal terms. The 6¼Akhoe Hai//om San of northern Namibia use directional terms which make use of complex landscape terms based on soil types, characteristic vegetation, and topography; a system that clearly differs from that of agropastoralists in the area (see Widlok 1997). This example underlines the stationary and regional aspect of hunter-gatherer mobility pointed out by Brody. A regional mode of mobility and integration has also been described for Australian desert people (see Myers 1986). Characteristically, this regional mode consists of a network that is larger than a single local group but not as expansive as the migrations of pastoralists and farmers. Based on comparative ethnographic data with hunter-gatherers, Ingold (2000) has recently further synthesized ideas on forager orientation and mobility systems into a more dynamic picture. He argues that the existing maps of hunter-gatherer territories and movements are ﬂawed, but not only because they privilege delineated expanses in favor of interlocking routes and networks. His critique is more radical in that he argues against ‘‘the cartographic illusion . . . that the world is preprepared as a stage upon which living things propel themselves about, from one location to another’’ (Ingold 2000: 242). According to this phenomenology-inspired theoretical approach, forager skills of orientation, and of mobile adaptation more generally, are not a cultural heritage that is transferred like a blueprint out of context, but the skills themselves are only realized in the process, unfolding as they are practiced. Following hunter-gatherers in different contexts and realizing their skill of mobility in different contexts is an example of inﬂecting which has been described in abstract terms, above. There is not one default or ‘‘standard’’ mode of spatial organization and of mobility, nor necessarily clearly distinguishable types, but regular inﬂections do emerge in comparative perspective. Ingold develops this theoretical outline further in a comprehensive theory of life as ‘‘laid down along paths of movement, of action and perception’’ instead of the ‘‘internal property of objects’’ (2000: 242). Whatever the philosophical merits or problems of this approach, hunter-gatherer scholars will note that this perspective can accommodate valuable aspects of what hitherto appeared to be conﬂicting schools in huntergatherer studies. It bridges the classic and adaptionist perspectives insofar as it no longer separates the sociocultural and the natural environment, or history and evolution, but considers all these elements in a more holistic framework. Rather than continue to assume the existence of a cultural core that is subsequently altered through its environment, human beings, culture, and nature are assumed to form a developmental system that is transformed over time and ‘‘inﬂected’’ differently across the world. This approach also creates bridges between revisionist and other perspectives because it is strongly non-essentialist while not denying difference. In other words, it no longer requires the assumption of an ideal-type forager culture that is instilled within forager groups or individuals. Hunter-gatherers do many things differently from non-hunter-gatherers, but this is due to the momentum that practices gain by being practiced, it is not due to a forager template conceived of as a cultural or behavioral program. This also solves a recurrent puzzle in the study of egalitarian hunter-gatherers, namely how social systems that provide

___________________________________________________________________________________

28

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s

considerable personal autonomy can maintain sufﬁcient stability to have lasting social relations at all (Brunton 1989). Or to put it differently: How to maintain the cultural obligation not to impose obligations on one another? In this approach there is no need to assume that rules of sharing and so forth are formulated as abstract rules which are then executed more or less rigorously in a variety of contexts. Rather, the recurrent practice of sharing itself creates the environment, for instance in terms of spatial settlement layouts and kin or friendship ties, that fosters further acts of sharing and an extension of these ties to include other persons (and personalized non-human beings). Finally, it also includes the indigenist perspective in a new way, not simply as an intensiﬁcation of contact between scholars and ‘‘locals,’’ but also by intensifying and facilitating new debates among these groups. By de-essentializing what constitutes a hunter-gatherer way of life it can do away with the genealogical mode that presses hunter-gatherers to identify ultimately through bloodlines or ethnicity and to base their claims on this hereditary identity. It directs attention to hunter-gatherer practices as developmental systems that may involve some persons more than others and that ﬂourishes in some political environments better than in others.

Conclusion: Future Shifts Although desert hunter-gatherers have played an important part as the subjects of anthropological theorizing, it is fair to say that theory in hunter-gatherer studies is still an emerging and open ﬁeld. This means that it is difﬁcult to speculate about future shifts in the theory of desert hunter-gatherer studies. However, if the combination of a reﬂecting and an inﬂecting mode, as outlined above, succeeds, it could also lead to a productive reappraisal of analogy as a research tool, thereby also reconnecting work in archaeology, linguistics, physical anthropology, and sociocultural anthropology. So far, archaeologists have confronted the question of analogy much more forcefully than social and cultural anthropologists and there is a certain degree of asynchrony between the two subdisciplines. In an early phase, ethnographers were keen to make analogies with the archaeological past. The emergence and subsequent disappearance of the phrase ‘‘stone age man’’ in the ethnographic discourse on hunter-gatherers is indicative. Towards the middle of the twentieth century anthropologists made uncritical use of this label (see Sahlins 1988; Spencer and Gillen 1927), and reluctance to accept these analogies initially came from archaeologists. They came to envisage their work to be a self-contained enterprise relying on the recovered materials only, not needing any non-material (and therefore dubious) anthropological analogies, but restricting themselves to middle range research (see Bettinger 1991). Today, the situation in some ways is almost reversed. It is the archaeologists who are rediscovering the importance of ethnographic analogies to make sense of their data (see Eggert 1998), whereas anthropologists now largely reject the use of analogies. A transfer of considerations concerning analogy as a research tool from archaeology back into social and cultural anthropology may be very instructive for future

___________________________________________________________________________________

29

_______________________________________________________________________________

T h o m a s Wi d l o k

theoretical models in the anthropology of desert hunter-gatherers. Archaeologists know that not all aspects of life are equally well captured through archaeological data. This is quite clear in the case of the analysis of settlement patterns, for instance. For those aspects that leave little material traces, ethnographic analogies are therefore most welcome. What is frequently overlooked, however, is that the same holds true for ethnography. Not all aspects of life are equally well captured through standard ethnographic methods, in particular given the tendency to privilege interview material. Complementing an ethnographic perspective with a discussion of historical data (see Sahlins 1985) is one attempt to deal with the problem, but it does not sufﬁce. Ethnographers gather a lot of information on past and present discourses, but the effects of long-term practices, economic transfers, and mobility or marriage patterns are not as easily grasped by standard ethnographic ﬁeld research or by standard historical research. Social and cultural anthropology may have a focus on contemporary issues, but this is largely due to the methods that are being employed and it is not because the past or the future is without importance in the process of generating anthropological knowledge. A productive study of desert hunter-gatherers always needs to be based on the explanatory potential of analogies and other forms of comparison.

References Barnard, A. 1978: Bushmen. London: Museum of Mankind. Barnard, A. 1983: Contemporary hunter-gatherers: Current theoretical issues in ecology and social organization. Annual Review of Anthropology, 12, 193–214. Barnard, A. 1992: The Kalahari Debate: A Bibliographic Essay. Edinburgh: Centre for African Studies, University of Edinburgh. Barnard, A. 1999: Images of hunters and gatherers in European social thought. In R. Lee and R. Daly (eds), The Cambridge Encyclopedia of Hunters and Gatherers, Cambridge: Cambridge University Press, 375–83. Barnard, A. and Taylor, M. 2002: The complexities of association and assimilation: An ethnographic overview. In S. Kent (ed.), Ethnicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 230–46. Bender, B. and Morris, B. 1988: Twenty years of history, evolution, and social change in gatherer-hunter studies. In T. Ingold, D. Riches, and J. Woodburn (eds), Hunters and Gatherers 1: History, Evolution and Social Change, Oxford: Berg, 4–14. Berndt. R. and Berndt, C. (eds) 1979: Aborigines of the West: Their Past and Present. Perth: University of Western Australia Press. Bettinger, R. 1991: Hunter-Gatherers: Archaeological and Evolutionary Theory. New York: Plenum. Bird-David, N. 1994: Sociality and immediacy: Or, past and present conversations on bands. Man, 29, 583–603. Boas, F. 1888: The Central Eskimo, vol. 6. Washington, DC: Bureau of American Ethnology. Brody, H. 2001: The Other Side of Eden: Hunter-Gatherers, Farmers, and the Shaping of the World. London: Faber and Faber.

___________________________________________________________________________________

30

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s Brunton, R. 1989: The cultural instability of egalitarian societies. Man, 24, 673–81. Burch, E. 1994: The future of hunter-gatherer research. In E. Burch and L. Ellanna (eds), Key Issues in Hunter-Gatherer Research, Oxford: Berg, 441–55. Burch, E. and Ellanna, L. (eds) 1994: Key Issues in Hunter-Gatherer Research. Oxford: Berg. Dunlop, I. 1966a: People of the Western Desert. Film leaﬂet. Dunlop, I. 1966b: Technical and production problems of ethnographic ﬁlms: Some aspects of ﬁlming in Australia’s Western Desert. In Round Table on Ethnographic Film in the Paciﬁc Area, Sydney: Australian Institute of Aboriginal þ Torrer Strait Islander Studies. Eggert, M. 1998: Archa¨ologie und Analogie: Bemerkungen zu einer Wissenschaft vom Fremden. Mitteilungen der Anthropologischen Gesellschaft in Wien, 128, 107–24. Elkin, A. 1981: The Australian Aborigines. London: Angus and Robertson. Fowler, C. 1999: The Timbisha Shoshone of Death Valley. In R. Lee and R. Daly (eds), The Cambridge Encyclopedia of Hunters and Gatherers, Cambridge: Cambridge University Press, 66–70. Gordon, R. 1992: The Bushman Myth: The Making of a Namibian Underclass. Boulder, CO: Westview Press. Gould, R. 1969: Subsistence behavior among the Western Desert Aborigines of Australia. Oceania, 39, 253–74. Gowdy, J. 1998: Limited Wants, Unlimited Means: A Reader on Hunter-Gatherer Economics and the Environment. Washington, DC: Island Press. Headland, T. and Blood, D. 2002: What Place for Hunter-Gatherers in Millennium Three? Dallas, TX: SIL. Hiatt, L. 1996: Arguments about Aborigines: Australia and the Evolution of Social Anthropology. Cambridge: Cambridge University Press. Ingold, T. 2000: The Perception of the Environment: Essays in Livelihood, Dwelling, and Skill. London: Routledge. Katz, R., Biesele, M., and St Denis, V. 1997: Healing Makes our Hearts Happy: Spirituality and Cultural Transformation among the Kalahari Ju/’hoansi. Rochester: Inner Traditions. Kelly, R. 1995: The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, DC: Smithsonian Institution Press. Kent, S. 2002: Autonomy or serfdom? Relations between prehistoric neighboring huntergatherers and farmer/pastoralists in southern Africa. In S. Kent (ed.), Ethnicity, HunterGatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 48–92. Kolig. E. 1989: Dreamtime Politics: Religion, World View and Utopian Thought in Australian Aboriginal Society. Berlin: Reimer. Lee, R. and Daly, R. 1999a: Introduction: Foragers and others. In R. Lee and R. Daly (eds), The Cambridge Encyclopedia of Hunters and Gatherers, Cambridge: Cambridge University Press, 1–19. Lee, R. and Daly, R. (eds) 1999b: The Cambridge Encyclopedia of Hunters and Gatherers. Cambridge: Cambridge University Press. Lee, R. and DeVore, I. (eds) 1968: Man the Hunter. New York: Aldine. Marlowe, F. 2002: Why the Hadza are still hunter-gatherers. In S. Kent (ed.), Ethnicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 248–75. Marshall, J. 1993: Filming and learning. In J. Ruby (ed.), The Cinema of John Marshall, Amsterdam: Harwood, 1–133.

___________________________________________________________________________________

31

_______________________________________________________________________________

T h o m a s Wi d l o k Marshall, J. and Ritchie, C. 1986: Where are the Ju/wasi of Nyae Nyae?: Changes in a Bushman Society, 1958–1981. Rondebosch: Communications Centre for African Studies, University of Cape Town no. 9, University of Cape Town. Marshall-Thomas, E. 1960: The Harmless People. London: Secker and Warburg. Martin, J. 1983: Optimal foraging theory: A review of some models and their applications. American Anthropologist, 85, 612–29. Mountford, C. 1950: Brown Men and Red Sand. London: Phoenix House. Myers, F. 1986: Pintupi Country, Pintupi Self: Sentiment, Place, and Politics among Western Desert Aborigines. Washington, DC: Smithsonian Institution Press. Passarge, S. 1997: The Kalahari Ethnographies (1996–98) of Siegfried Passarge, ed. E. Wilmsen (ﬁrst published in German 1907). Ko¨ln: Ko¨ppe. Pederson, E., Danziger, E., Levinson, S., Kita, S., Senft, G., and Wilkins, D. 1998: Semantic typology and spatial conceptualization. Language, 74, 557–89. Peterson, N. and Matsuyama, T. 1991: Cash, Commoditization, and Changing Foragers. Osaka: Senri Ethnological Studies vol. 30, National Museum of Ethnology. Povinelli, E. 1993: Labor’s Lot: The Power, History, and Culture of Aboriginal Action. Chicago: University of Chicago Press. Radcliffe-Brown, A. 1922: The Andaman Islanders. Cambridge: Cambridge University Press. Ruby, J. 1993: The Cinema of John Marshall. Amsterdam: Harwood. Rumsey, A. 2001: Introduction. In A. Rumsey and J. Weiner (eds), Emplaced Myth: Space, Narrative, and Knowledge in Aboriginal Australia and Papua New Guinea, Honolulu: University of Hawai’i Press, 1–18. Sadr, K. 2002: Encapsulated Bushmen in the archaeology of Thamaga. In S. Kent (ed.), Ethnicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 28–47. Sahlins, M. 1985: Islands of History. London: Tavistock. Sahlins, M. 1988: Stone Age Economics. London: Routledge. Schott, R. 1956: Anfa¨nge der Privat-und Planwirtschaft. Wirtschaftsordnung und Nahrungsverteilung bei Wildbeutervo¨lkern. Braunschweig: Limbach. Schweitzer, P., Biesele, M., and Hitchcock, R. (eds) 2000: Hunters and Gatherers in the Modern World: Conﬂict Resistance and Self-Determination. New York: Berghahn. Solway, J. and Lee, R. 1990: Foragers genuine or spurious? Situating the Kalahari San in history. Current Anthropology, 31, 109–46. Spencer, B. and Gillen, F. 1899: The Native Tribes of Central Australia. London: Macmillan. Spencer, B. and Gillen, F. 1927: The Arunta: A Study of a Stone Age People. London: Macmillan. Sutton, P. 1995: Country. Aboriginal Boundaries and Land Ownership in Australia. Canberra: Aboriginal History Monograph vol. 3, The Australian National University. Swain, T. 1993: A Place for Strangers. Towards a History of Australian Aboriginal Being. Cambridge: Cambridge University Press. Szalay, M. 1986: Zur Frage der ‘‘sekunda¨ren Primitivita¨t’’ der ’Buschma¨nner. In R. Vossen and K. Keuthmann (eds), Contemporary Studies on Khoisan, Hamburg: Buske, 257–69. Tindale, N. 1968: Review of Desert People. American Anthropologist, 70, 437–8. Tonkinson, R. 1991: The Martu Aboriginies: Living the Dream of Australia’s Desert. Forth Worth, TX: Holt, Rinehart, and Winston. Triesch, C. 2001: ‘‘Warum sollten sie . . . ?’’ Die Frage nach der Verbreitung des Bodenbaus und die pra¨koloniale Nutzung von Nahrungsressourcen in Australien und Neuguinea. Herbholzheim: Centaurus.

___________________________________________________________________________________

32

_______________________________________________________________________________

T he A n t h r o p o l o g y o f D e s e r t H u n te r -G a t h e r e r s Widlok, T. 1992: Practice, politics, and ideology of the ‘‘traveling business’’ in Aboriginal religion. Oceania, 3 (2), 114–36. Widlok, T. 1997: Orientation in the wild: The shared cognition of Hai//om bush people. Journal of the Royal Anthropological Institute, 3, 317–32. Widlok, T. 1999: Living on Mangetti: ‘‘Bushman’’ Autonomy and Namibian Independence. Oxford: Oxford University Press. Widlok, T. 2002: The long walk IV – Hunter-gatherers and anthropology: An interview with James Woodburn. Nomadic Peoples, 6, 7–22. Williams, N. and Hunn, E. (eds) 1982: Resource Managers: North American and Australian Hunter-Gatherers. Boulder, CO: Westview Press. Wilmsen, E. and Denbow, J. 1990: Paradigmatic histories of San-speaking peoples and current attempts at revision. Current Anthropology, 31, 489–507. Winterhalder, B. 1993: Work, resources, and population in foraging societies. Man, 28, 321–40. Woodburn, J. 1982: Egalitarian societies. Man, 17, 431–51.

___________________________________________________________________________________

33

_______________________________________________________________________________

--------------------------------------------------------------------------------

3

----------------------------------------------------------------------------

Pleistocene Settlement of Deserts from an Australian Perspective Peter Hiscock and Lynley A. Wallis

Introduction: Deserts and Desert Colonization One of the surprising characteristics of deserts is their environmental diversity. Far from the monolithic uniformity of the popular imagination, deserts typically display variation in topography, hydrology, and resources. Within Australian deserts these differences are displayed at a variety of scales. While some areas have coordinated drainage, others have no coherent and articulated drainage systems; some desert areas are principally rocky outcrops, others largely mantled with sand, and still others are a mixture; some have permanent water sources, while many do not; and so on. Within any single desert area the availability of water, stone, and wood suitable for tool making, as well as the nature and abundance of plants and animals, varies across the landscape. All of these features have varied through time as climatic shifts, geological processes such as erosion and sedimentation, and changes in ﬂoral and faunal composition have altered the structure of resources across the landscape. A single landscape may even become or cease being a desert with the passing of time. Variation in the distribution and reliability of resources through time and space is likely to have had a critical inﬂuence on the colonization and settlement of these desert landscapes. Adaptations to deserts by humans are also far from being monolithic. Within Australia images of ethnographic desert dwellers have often dominated discussions of possible lifeways within arid and semi-arid lands. These images of historic Aboriginal desert life often emphasize features like the intensive use of vegetable foods such as seeds, and maintenance of long distance social networks involving reciprocity and rights to territorial access (Gould 1977; Tonkinson 1991: 40). Most importantly, the ethnographic desert dwellers were renowned for their intimate knowledge of the landscapes in which they lived; however, these features are unlikely to have been traits of the colonizers of these desert landscapes! Initial colonization must have been accomplished by foragers who did not know the terrain or the

___________________________________________________________________________________

34

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

distribution patterns of resources and therefore the early settlement of the Australian deserts is likely to have been undertaken by people who had economic and subsistence strategies that were different from those of the historic period. Some researchers have suggested that the ethnographic forms of desert economy and subsistence are relatively recent, and have emerged only since the Last Glacial Maximum (LGM) and in the Holocene (e.g., Veth 1989). This chapter tackles two of the key problems in archaeological investigations of early occupation of the Australian deserts: (1) How different was the settlement system of early settlers of the inland from that found in recent times and (2) how did foragers colonize the interior without the adaptive strategies of historic desert peoples? In answer to these questions we advance the propositions that much of the Australian interior was initially colonized during a period of higher rainfall and more abundant surface water and food resources – a period in which a dedicated desert adaptation of the historic kind was not in place – and that elements of this pattern of early subsistence continued until the LGM, hence supporting a model in which some of the key features found in historic desert life have accumulated only since the terminal Pleistocene. We argue that in Australia Pleistocene foragers did not move into deserts fully equipped with a modern desert adaptation, but rather that climate change created deserts in areas where huntergatherers already lived and those human groups adapted their existing strategies to the new situations or else abandoned the landscape – a ‘‘desert transformation’’ model. A review of deserts around the world suggests that similar colonization models may have broad applicability. Note that the following review incorporates age estimates derived from a range of dating techniques. To reconcile these different estimates we have adopted the approach of expressing all ages as solar (calendar) years rounded to the nearest 500 years, including those derived from radiocarbon dates. To achieve this, age estimates based on uncalibrated radiocarbon dates presented in the cited literature have been calibrated using the procedures developed by the University of Cologne in its CalPal software and by the formula of Miller et al. (1997). Although such calibrations are of uncertain accuracy they facilitate the synthesis of different data sources, such as luminescence and radiocarbon assays. Using this approach our review of the Pleistocene occupation of Australian deserts requires consideration of the question as to when humans ﬁrst arrived in the Australian continent.

Arrival of Humans in Australia Archaeological evidence demonstrates that humans arrived in Australia 60,000–45,000 years ago (Bowler 1998; Bowler and Magee 2000; Bowler and Price 1998; Gillespie 1998; Gillespie and Roberts 2000; Gru¨n et al. 2000; Pearce and Barbetti 1981; Roberts et al. 1990, 1993, 1994). The precise date of initial arrival of humans in Australia has been the subject of vigorous debate, principally between a group of researchers advocating a likely landfall of approximately 45,000 years ago (e.g., O’Connell and Allen 1998, in press) and those preferring an estimate of approximately 55,000 years ago (e.g., Roberts et al. 1990, 1993, 1994).

___________________________________________________________________________________

35

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

We consider these debates of little value given a number of factors. Firstly, they have tended to ignore the uncertainty attached to many age estimates, thereby acting as though the various estimates were more precise than they are. For instance, many of the luminescence, U-series, and ESR dates have uncertainties in the order of 5,000–10,000 years (see O’Connell and Allen, in press, for data summaries), often making debates as to whether material at a speciﬁc site was 50,000 or 45,000 years old merely a speculative adventure. In another context the consequences of these uncertainties have been discussed by Rindos and Webb (1992), and we concur with their conclusion that high resolution models of the initial period of colonization may never be able to be tested using current dating techniques. Although the radiocarbon estimates have higher precision, doubts remain about their accuracy, with even the most recent advances in radiocarbon dating techniques failing to resolve all complications created by contamination or calibration. A second factor we will mention that creates difﬁculties for pinpointing a date for initial human entry into Australia is the lack of systematic sampling of archaeological sites. The probability of ﬁnding archaeological traces of the earliest occupation of Sahul, even within a speciﬁc region, is very small; and the identiﬁcation of one or two old archaeological deposits in any region should be considered evidence of minimum antiquity rather than a deﬁnitive statement of the antiquity of occupation. This factor is exacerbated by the tendency of archaeologists to repeatedly target speciﬁc kinds of site – such as large caves – and the focus of redating investigations during the last decade on those atypical sites. In light of this factor it is noteworthy that the oldest potentially sustainable estimate of human occupation comes not from a cave deposit but from the open lake margin site of Lake Mungo, where ESR, U-series, and OSL techniques have indicated a burial may date to 62,000 6,000 years ago (Thorne et al. 1999). Although a date in excess of 50,000 years appears unlikely for this skeleton, based on the well established age of the sediments in which it was found (Bowler and Magee 2000), the unresolved debate about the antiquity of the Mungo material highlights two aspects of the investigation of human antiquity in the Australian continent: that since the nature of human landscape use may have varied through time, some landscape features may not provide residues of the initial movement of humans into a region, and that even the lowest reasonable estimates of the lowest artifacts at Lake Mungo deny conservative claims of continental colonization of 43,000 bp or less. The third signiﬁcant concern when interpreting age estimates for early Australian occupation is the lack of detailed taphonomic/formational studies for almost every deposit cited in these debates. Formational/disturbance processes, which can be presumed to occur in every archaeological site, create ambiguities in age estimation that are not expressed in the uncertainties provided by dating laboratories. The failure of Australian archaeologists to make taphonomic investigations a high research priority is compounded by the regular dating of non-archaeological components of these old deposits – materials such as sand grains – thereby raising questions of stratigraphic association. Where the formation processes of early sites have been subjected to scrutiny they have often proved problematic and in need of further study (see Hiscock 1990; O’Connell and Allen in press; Richardson 1992).

___________________________________________________________________________________

36

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

Given these complexities in the use of existing chronological evidence we conclude that Australia was most likely colonized by Homo sapiens sapiens at a date of 50,000 10,000 years ago, that is 60,000–40,000 years ago. We consider this estimate to be imprecise but reliable, and to deﬁne the earliest chronological period in which humans were present on the continent and in which occupation of desert landscapes might have occurred. Accepting this estimate of the arrival of humans in Australia, we now turn to the evidence for colonization of desert areas.

Initial Human Penetration of the Australian Deserts In evaluating the timing of initial human movements into the Australian deserts we have identiﬁed the earliest, apparently reliable age estimate for archaeological materials in each region that today is arid or semi-arid. The data-set that we have drawn on is that prepared by O’Connell and Allen (in press), and the following discussion is restricted to those sites from which dates exceeding 35,000 years have been retrieved. Again, the reader is reminded that ages provided here are in solar years, with radiocarbon values calibrated and rounded to the nearest half a millennium. This procedure currently yields seven sites across Australia (see Figure 3.1). In the Kimberly region of northwestern Australia radiocarbon assays at two rock shelters have very similar ages for their lowest occupation levels: Carpenter’s Gap 1 with a date of 45,000 years, or 46,500–43,500 years ago at two standard deviations (Fiﬁeld et al. 2001); and Riwi with a date of 45,500 years, or 47,500–44,000 years (Balme 2000). Further to the southeast, in the MacDonnell Ranges of central Australia the Puritjarra rock shelter has evidence of occupation at 39,000 years ago, or 42,500–36,500 years (Smith et al. 1997). Even further south, on the Nullarbor Plain and edge of the Great Australian Bight, Allen’s Cave has yielded an OSL date of 40,000 years, or 43,000–37,000 years (Roberts et al. 1996). Far to the east of Allen’s Cave, the stratigraphically lowest artifacts in the aforementioned Lake Mungo site complex are estimated, using OSL, to have an antiquity of at least 50,000–46,000 years (Bowler et al. 2003). Further northeast, Cuddie Springs appears to have archaeological material dated using OSL techniques to 35,500 years, or 38,500–32,500 years ago (Roberts et al. 2001). Finally, in the northeast of the continent, in the gulf fall zone of the Barkley Tablelands, the GRE 8 shelter has produced a radiocarbon sample of 41,500 years, or 44,500–37,500 years (O’Connell and Allen in press). Although the radiometric estimate from Cuddie Springs is somewhat younger than the demonstrated antiquity at the other sites, we interpret these data as indicating broadly similar dates for human presence in widely separated regions. Humans are archaeologically visible in each of these locations 50,000–40,000 years ago, a pattern that must be taken to reveal the widespread distribution of huntergatherers across inland Australia at that time. These dates could be seen to imply that occupation of these inland regions took place substantially later than the original landfall, if claims for dates of 55,000 years or more in Arnhem Land

___________________________________________________________________________________

37

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

Figure 3.1 Early Australian sites in landscapes arid or semi-arid today (after O’Connell and Allen in press).

(Roberts et al. 1990, 1993, 1994) are accepted at face value, thereby supporting a modiﬁed version of Bowdler’s (1977) model of coastal colonization; however, as we have already discussed, the imprecision of dates older than 45,000 years makes such a conclusion unsustainable, and we argue that models such as coastal colonization are currently untestable. Imprecise dates also constrain consideration of whether these inland regions were initially occupied contemporaneously or sequentially; the precise rate of human spread across the inland cannot be calculated, although the current evidence of broadly similar antiquities in each region does not favor a spread of people extended over tens of millennia. Nevertheless it is worth noting that there is no obvious directionality in these dates – the northwest and southeast have returned similar ages for occupation; hence, no directional pattern of dispersion can reasonably be read into the pattern described here. Although the dating evidence fails to provide deﬁnitive tests of fast versus slow models of population

___________________________________________________________________________________

38

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

dispersion across the Australian interior, it not only reveals that relatively early (pre-40,000 years) occupation was widespread and not restricted to coasts, but also provides the necessary chronology with which to assess the environmental context of the initial movement of people into the Australian deserts. During the period when humans become archaeologically visible in regions of Australia that are now deserts, the climatic and hydrological regimes created landscapes that were different from those found today. Any consideration of the colonization of Australian deserts must be understood in that environmental context. Three regional paleoenvironmental records are summarized below as exemplars of the context of colonization: the strongly seasonal Kimberley region of northwest Australia; the weakly seasonal Lake Eyre and its catchment; and the relatively aseasonal Willandra Lakes (see Figure 3.2). In the Kimberley the long-term climatic pattern has been reconstructed from northwest shelf marine cores, terrestrial lake sequences, and botanical materials from archaeological sites. The Lombok Ridge marine core indicates that the period from 130,000 until 38,000 years ago was one of high sea level and relatively wet conditions in northern Australia, with rainfall estimated to fall in the range of 500–1000 mm per year (van der Kaars 1991; Wang et al. 1999). Bowler’s (1983a; Bowler et al. 2001) investigations at Lake Gregory, located on the northern fringes 130

SUMMER RAIN

140

0

500 km

Strongly seasonal (S/W >3) Weakly (S/W 1.3-3) seasonal

15

UNIFORM DISTRIBUTION

1 20

2 L Eyre

30

3 35

WINTER RAIN Weekly seasonal (W/S 1.3−3) Strongly seasonal (W/S >3)

135

150

Figure 3.2 The contemporary Australian arid zone (thick line), where potential evaporation exceeds actual evaporation (modiﬁed from Gentilli 1986 and Hesse et al. 2004). Rainfall seasonality is shown as the ratio of summer to winter rainfall. Three broad regions are numbered as (1) Kimberley and the northwest, (2) Lake Eyre and its catchment, and (3) Willandra Lakes.

___________________________________________________________________________________

39

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

of the Great Sandy Desert, have revealed increasing aridity during the last 300,000 years, with several lacustral phases prior to 40,000 years ago. The record from Lake Wood to the east shows similar trends, but differences in the precise timing and extent of high water conditions, indicating interregional and local variations (Bowler et al. 1998). The long-term trend towards increasing aridity is also reﬂected in vegetation change during the late Quaternary. Studies of soil organic matter from Lake Gregory have indicated a trend, over the last 100,000 years, away from exclusively tree and shrub ﬂoral assemblages towards a grass dominated ﬂora as increasing water stress developed (Pack et al. 2003). These data are consistent with the evidence for vegetation change available from the phytolith record at Carpenter’s Gap 1 where, 45,000 years ago, there was a diverse grassland ecology and palms outside their modern geographical range, suggesting the existence of wetter conditions at that time than in any more recent period (Wallis 2001). In the Lake Eyre Basin of central Australia, a region in which water availability reﬂects not only local conditions but also seasonal rainfall patterns in its northerly catchment, broadly similar trends towards desiccation have been well documented. As evidenced by the nature of regional vegetation patterns, the Australian monsoon was more effective before 45,000 years ago than at anytime since (Johnson et al. 1999). This is also seen within Lake Eyre itself, which was for the most part a permanently wet lake until about 60,000 years ago, when it changed to a groundwater dominated system as it progressively dried out (Magee and Miller 1998). Up until about 30,000 years ago enhanced winter westerly circulation might have penetrated more regularly to the Lake Eyre region than under the present climate, bringing with it enhanced winter rainfall and wetter conditions than at any time since. Preserved tree roots and trace fossils of large roots replaced by secondary gypsum also reveal that trees were more common during that period than today (Hesse et al. 2004). These climatic differences in the Lake Eyre Basin probably stem from an interaction of enhanced winter rainfall, greater inﬂow from enhanced storms in northern Australia, and reduced evaporation associated with substantially lower temperatures than at present (Magee and Miller 1998; Miller et al. 1997). In contrast, far from the inﬂuence of monsoonal rainfall in southeastern Australia the archetypal locality for describing arid/semi-arid environmental history and archaeology is the Willandra Lakes, and especially Lake Mungo. Bowler (1998) has demonstrated the alternation of dry and wet phases at Lake Mungo during the late Pleistocene, with a long-term drying trend. Prior to 45,000–42,000 years ago there was a prolonged lacustral phase with high water levels. A further series of water bodies were also present 36,000–22,000 years ago, although low and ﬂuctuating lake levels are features of the period following 42,000 years. Other lake systems in the arid southeast have been shown to display somewhat similar trends, although with distinct hints of local variations in the timing and pattern of high water levels (Harrison 1993). This overall pattern of drying since 50,000–40,000 years ago is also visible in the activation of linear dunes, particularly during the last 25,000 years (Wasson 1984). These brief pictures of surface hydrology and vegetation in widely separated regions of the Australian interior provide an image of the landscapes into which we

___________________________________________________________________________________

40

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

believe the ﬁrst human colonists moved. Prior to about 45,000 years ago conditions were similar to, but probably cooler than, those of today, with a long-term trend towards aridiﬁcation and increased aeolian activity. In many regions it is likely that up until about 45,000 years ago – and perhaps to 30,000 years ago in some locations – the availability of fresh surface water would have been at least as good, if not better, than during the Holocene. Hence at the time humans began exploring these landscapes they were still drier than many parts of the continental margins, but they were noticeably different from their contemporary state. In each of the three regions reviewed above, and many others not summarized here, seasonal ﬂoods and large standing water bodies were common and comparatively predictable prior to 45,000–40,000 years ago in ways that they have not been over the last 35,000 years. We hypothesize that human colonization of these landscapes involved a number of general patterns that reﬂect the environmental conditions and are likely to be unique to this particular period (a point also made by Thorley 1998). The greater relative availability and predictability of resources, including large permanent water bodies, prior to 45,000–40,000 years ago, would have facilitated exploration and exploitation of these unique interior landscapes. It is likely that the pattern and rate of exploration and colonization varied between regions in response to the speciﬁc nature and distribution of landscape features within each region. Both ethnographic and biogeographic perspectives indicate that colonization is most likely to have taken place along corridors of distinct landscape features such as rivers or mountain ranges (Kelly 2003; Veth 1989). Since such features vary between regions we suggest that there is unlikely to have been a uniform pattern of colonization across these diverse interior landscapes. More signiﬁcantly, the movement of people into the interior of Australia prior to 40,000 years ago would have meant that their economic and social systems were developed in response to the existing environmental conditions. Hence, interior colonization may not have required a reﬁned or speciﬁc adaptation to extreme aridity. Thorley (1998, 2001) has previously discussed this proposition for the initial human occupation of the central Australian ranges and plains, where Lake Amadeus and the Finke and Palmer rivers catchment would have provided more regular and more abundant resources at the time of colonization than today. Our review suggests that this notion should be extended to many of the inland regions for which we now have evidence. Without doubt the gradual desertiﬁcation that followed the period of interior colonization (post-35,000 years ago) would have required modiﬁcations to foraging and social strategies. However, these modiﬁcations would have been made by forager groups which by that time had established familiarity with the landscape structure of each region. Readjustment of economic systems to even rapid trends towards aridity would have been facilitated by local group knowledge of the environment; these groups would have been advantaged by modifying their existing systems in situ, building new economic and technological strategies based on knowledge of the local landscape. In this sense we hypothesize that the widespread colonization of the Australian interior prior to the pronounced drying of the late Pleistocene meant that many foraging groups with generalized

___________________________________________________________________________________

41

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

terrestrial economic strategies probably evolved more dedicated desert economic strategies in situ. This ‘‘desert transformation’’ model removes the paradox of explaining how people were able to migrate into Australian deserts in the late Pleistocene; in some important ways the modern deserts of Australia came to inland dwelling people, rather than the reverse. Of course, there is no connection of this model with the idea of coastal colonization (contra Thorley 1998), since the process we hypothesize implies that groups moving into interior lands had effective terrestrial foraging strategies suited to a variety of dry, semi-arid, and arid landscapes. These were not coastally adapted peoples (also as argued by O’Connor and Veth 2000); but our model predicts they were also foragers without many of the speciﬁc desert economic and social traits observable in the late Holocene and historic period. This prediction is consistent with features of the archaeological record in these interior lands. For instance, a number of the strategies taken to be risk reduction responses in the late Holocene desert economies are either absent or little utilized in the archaeological residues of the colonists moving into those regions more than 40,000 years ago. No compelling evidence for trade or long distance transportation of resources has been identiﬁed, perhaps not surprising given the exploratory nature of occupation and the indications in extremely low artifact discard rates that population densities may have been very low and perhaps territories large. Furthermore, although grinding stones are likely to have been a technology employed to exploit plant foods throughout Australian prehistory (Fullagar and Field 1997; Gorecki et al. 1997), the earliest archaeological assemblages contain few grindstones. This pattern probably reveals that the intensive processing of seeds, a key characteristic of risk minimization in historical arid zone economic strategies, emerged in response to speciﬁc conditions in the Holocene and was neither a central nor a universal feature of the initial occupation of the interior. The exploitation of the large lacustrine and riverine resources that existed at the time, plus the low densities of people occupying many inland territories, may be cited as likely reasons that seed processing was not given the same prominence during the pre-40,000 year old colonizing phase as it was in much later times. These earliest inland economies were based on a broad spectrum and ﬂexible foraging strategy. Although not all of the early archaeological sites spread across the Australian interior have good faunal preservation, the key characteristics of hunting and animal foods can be deﬁned as follows. A wide range of small to medium sized game was hunted, including marsupials, reptiles, and, near lacustrine systems, ﬁsh and mussels. The taxa represented varied between sites and regions in response to local conditions. Plant foods were also exploited, but as explained above, the discovery of grindstones at only one locality to date also indicates variation in the way these broad spectrum systems were implemented. A radically different view on the economy of colonizing groups proposes that early foragers targeted large land animals and were responsible for the extinction of all megafauna in Australia, during the ﬁrst millennia of human occupation of the continent (e.g., Miller et al. 1999; Roberts et al. 2001). Many kinds of megafauna (species of animals exceeding 44 kg at adult weight) have become extinct in Australia

___________________________________________________________________________________

42

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

over the last 100,000 years; however, the claim that all megafauna became extinct soon after humans colonized Australia is fraught with sampling problems and fails to consider the discovery of sites where megafaunal species existed until only 30,000–20,000 years ago (Field and Dodson 1999; Wroe et al. 2002). More signiﬁcantly, not a single ‘‘kill’’ site from the initial period of interior colonization has been identiﬁed, and it seems doubtful that there could have been enough humans in interior Australia around 50,000–45,000 years ago to cause the extinction of multiple species (Hughes and Hiscock in press). A more likely explanation is that the environmental transformation of the Australian interior through the process of aridiﬁcation was the trigger for faunal changes of many kinds, including local extinctions. Humans were not immune to those environmental changes and the archaeological sequence records the impact of desertiﬁcation on inland foragers.

Subsequent Use of Desert Areas of Australia One consequence of our ‘‘desert transformation’’ model of pre-hyperarid colonization of the interior by generalist foragers concentrating on riverine and lacustrine resources is that desert adaptations emerged during the subsequent period. In this chapter we are concerned with two aspects of the modiﬁcation of economies: the nature of economic life in the Pleistocene deserts prior to the LGM, and the impact of the LGM on desert dwellers. Archaeological evidence now reveals dynamic and varied subsistence and settlement responses to the onset of desert conditions, and suggests that these responses are not identical to the desert adaptation that emerged in the Holocene. Following a gradual cooling and drying trend initiated at about 45,000 years ago, the last glacial cycle intensiﬁed rapidly with the onset of what is called Oxygen Isotope Stage 2 (OIS2) approximately 30,000 years ago. At that time sea levels dropped rapidly by nearly 50 m over perhaps 1,000 years or less, exposing large areas of the continental shelf and making many inland areas even more continental than they had been (Chappell 1991; Lambeck and Chappell 2001; Lambeck et al. 2002; Yokoyama et al. 2001). From 30,000 to about 20,000 years ago the monsoon was the least effective it had been over the last 100,000 years or so (Johnson et al. 1999); and average air temperatures in low latitude regions, including the continental interior of Australia, appear to have been at least 6–98C lower than at present (Miller et al. 1997). Many inland areas show pronounced reductions in the availability and reliability of surface water, and of related food resources. For instance, by 35,000 years ago Lake Eyre had become dry and remained so until around 10,000 years ago (Magee and Miller 1998). During the same period there were lower and ﬂuctuating lake levels in the Willandra system and the activation of dune building processes (Bowler 1983b, 1986, 1998). In many other lacustrine systems across Australia lake levels show signiﬁcant decreases in this period, although the timing varies locally (Harrison 1993). The radically reduced precipitation/evaporation indices in OIS2 had severe consequences for not only water availability but also for vegetation patterns. Transitions to landscapes with reduced

___________________________________________________________________________________

43

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

amounts of vegetation, ﬂoral assemblages with fewer trees/shrubs, and an increased grassland component have been inferred across many inland regions as extreme desert conditions emerged. Examples of such transformations include Carpenter’s Gap 1 in the northwest (Wallis 2001), Puritjarra rock shelter in central Australia (Bowdery 1998; Smith et al. 1995), and Cuddie Springs and Ulungra Springs in northern and western New South Wales (NSW) (Dodson and Wright 1989; Dodson et al. 1993). Changes in precipitation, temperature, and surface water availability acted in concert to enlarge the portion of Australia that would be regarded as desertic, with the semi-arid zone expanding laterally towards the continental margins, the previously semi-arid zone becoming fully arid, and deserts becoming far more inhospitable than they are today (see Jones and Bowler 1980). At least in the early portion of OIS2 some inland forager groups appeared to have continued and even consolidated their occupation, perhaps reﬁning their generalist economic strategies to suit the landscapes in which they found themselves and in response to the desiccation of those landscapes. Adjustment of groups to different environments implies that localized adaptations were probably emerging at this time. For instance, in some localities, such as the central Murray–Darling lake systems, the number of dated archaeological sites was greater for the period 35,000–25,000 years ago than for earlier periods (Hope 1993). In other regions, such as northwestern Australia, increases in artifact discard rates have been observed within some rock shelters during this period (Morse 1993; O’Connor et al. 1993, 1999). These patterns of increased debris have been interpreted as population growth in some interior localities (e.g., Beaton 1985; Jones 1979; O’Connor et al. 1993). While models of population increase prior to 25,000 years ago are plausible, interpretations of the evidence from some localities indicate the nature and uniformity of occupation clearly varies, and no single trend can be inferred for settlement across the inland. The question of population size raises the unresolved issue of what comprised inland occupation at this time. Although numbers of sites or numbers of artifacts are notoriously unreliable indicators of the amount of human activities or number of people visiting a site, archaeological investigations of Pleistocene rock shelters and lake margins in some regions of interior Australia, such as the northwest and southeast, may indicate relatively high levels of occupation during OIS2 (Balme 1995; O’Connor et al. 1998). In contrast, other landscapes show archaeological sequences with very low artifact discard rates in this period compared to the Holocene (O’Connor et al. 1993; Smith 1989). In some sites artifact discard rates are as low as a few specimens per meter per millennium – rates which are so slow that it is unclear whether this really represents permanent occupation of the territory, or whether occasional but extended excursions by mobile foragers from adjoining landscapes during favorable seasons might be a more accurate interpretation. Irrespective of whether these differences reﬂect the number of people in residence or their mobility, these indicators of occupation intensity may reveal substantial differences in occupation between regions. For instance, while occupation of lake lunettes in several localities shows systematic, intensive, and sometimes dedicated exploitation of lacustrine resources (Balme 1995; Bowler 1998), there is

___________________________________________________________________________________

44

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

minimal evidence of continuing occupation in many of the sandy deserts, except where they are bounded by other kinds of landscapes (see Hughes and Hiscock in press; O’Connor et al. 1998, 1999; Veth 1989). This variation in whether or not we have archaeological evidence for continued occupation probably reﬂects environmental differences between regions. This is an indication that the colonizing economic strategy, perhaps dependent on reliable access to surface water and focusing on the exploitation of riverine and lacustrine resources supplemented with nearby terrestrial resources, continued to form the core theme of many inland foragers in the period after 35,000–30,000 years ago. In a number of landscapes which retained these features, such as desert regions containing uplands, major coordinated drainage, and/or extant lake systems, archaeological evidence for occupation is found from the start of OIS2; in contrast, archaeological investigations in regions without those features have found little or no evidence of occupation. This contrast parallels Veth’s (1989) model that proposed sandy deserts without coordinated drainage and large permanent lakes often acted as barriers to occupation during the Pleistocene and early Holocene. In view of the now apparent widespread colonization of inland Australia, including sandy deserts (O’Connor et al. 1998, 1999), we hypothesize that this pattern of abandonment/avoidance of barriers of sandy deserts arose only during the period of extreme desiccation marking OIS2. If current archaeological evidence can be interpreted as revealing abandonment or greatly reduced use of some desert areas at the start of OIS2, then apparent responses of inland foragers to the intensiﬁcation of arid conditions around 25,000 years ago can be viewed as an exaggeration of the existing land use strategies. The intensity of desiccation and consequent geographical expansion of arid interior Australia peaked towards the end of OIS2, during a relatively short phase spanning approximately 25,000–18,000 years ago, which we refer to as the Last Glacial Maximum (LGM). During this period conditions were both exceptionally cold and dry, and it is likely that evaporation was more effective and windiness enhanced compared to today, at least on a seasonal basis (e.g., Chappell 1991; Hubbard 1995; Magee and Miller 1998). Estimates of rainfall typically suggest levels approximately half that of today across interior Australia (Dodson and Wright 1989; Singh and Geissler 1985). The consequences of these extreme conditions must not be underestimated. Across the continent reduced vegetation cover was probably one trigger of a major phase of dune building initiated at this time (Ash and Wasson 1983; Bowler 1986; Bowler and Wasson 1984; Nanson et al. 1991; Wasson 1984) and aeolian dust storms were intense (Hesse 1994; Hesse and McTainsh 1999). Drying of lakes, and more general decline of surface water, was frequently linked to reductions in the level of water tables and, in some localities, formation of salt crusts on lake surfaces (Magee et al. 1995). These extreme conditions during the LGM caused widespread environmental stress, triggering massive – sometimes irreversible – changes to the landscape and to plant and animal resources. Some of these changes involved the temporary disappearance of food resources, such as in the Willandra Lakes where alterations to temperature, salinity, and water levels caused the loss of ﬁsh and freshwater mussels (Bowler 1998). Other kinds of changes may have involved a reduction in

___________________________________________________________________________________

45

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

the predictability of resources, such as greater periodicity or irregularity in the reﬁlling of water sources from rainfall. During this period it is likely that many portions of inland Australia were substantially drier than today, and archaeological research has yielded many instances of sites in which the absence of cultural material during the LGM indicates local or regional abandonment (Hiscock 1988; O’Connor et al. 1993, 1998, 1999; Veth 1989). Yet in other isolated inland regions occupation persisted, apparently within better resourced refuges. Sometimes these refuges were ranges or desert tablelands with aquifer-recharged water sources which provided plentiful, and more importantly unvarying, sources of water, as well as plants and animals (Hiscock 1984, 1988; Lamb 1996; Smith 1987, 1989). People may have exploited broader territories from these reliable bases, but in at least the case of Lawn Hill and Louie Creek Gorges in northern Queensland a contraction of territory has been demonstrated, with the residents abandoning high risk portions of the landscape from approximately 22,500 until 16,500 years ago (Hiscock 1988). Similar patterns of territorial contraction have been observed on Cape York and in northwestern Australia (Lamb 1996; Marwick 2002). These refuges would have supported only small groups of humans, and perhaps only on a semi-permanent basis, depending on the number of people and resource abundances in and around the refuge. The lake systems of western NSW represent another, perhaps shorter lived kind of refuge; there the appearance of grinding stones in archaeological deposits, either during or immediately after the LGM, is an example of the use of technological means to broaden resource use and obtain reliable access to food (Bowler 1998). Another predicted strategy would have been for groups of foragers to retreat to the semi-arid margins of inland Australia and to undertake long range trips to explore the arid zone during favorable conditions. One implication of decreased exploitation or abandonment of many deserts is that at some time following the termination of the LGM and coincident with the amelioration of climatic conditions foragers recolonized all portions of interior Australia. The timing and process of such recolonization events appear to be complex. In some localities, such as the JSN site in the Strzelecki Desert, there may be near immediate, if ﬂeeting, exploration or exploitation of the desert (see Hughes and Hiscock in press). Another rapid resumption of land use occurred around the Lawn Hill and Louie Creek Gorges, where resource exploitation in once-abandoned landscapes was reinitiated approximately 17,000–16,000 years ago (Hiscock 1988). However, there are regions of the arid zone in which the earliest dated evidence of human reoccupation takes place many millennia after the end of the LGM (Hughes and Hiscock in press; Morse 1988; Veth 1989). These differences may reﬂect the environments being recolonized, especially in the favorable climatic conditions of the early Holocene. However, they also probably reﬂect the historical circumstances of the colonizing groups, namely whether they derived from an LGM refuge or from non-arid lands on the continental periphery, whether the colonizing groups possessed speciﬁc economic or technological capacities that might enhance their ability in the new desert landscapes, and so on. When desert occupation recommenced and/or intensiﬁed in the Holocene the distinctive recent desert economy is clearly visible, most easily recognized by the frequent use of

___________________________________________________________________________________

46

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

grindstones to process reliable and abundant grass seeds, the abundant discard of tula adzes reﬂecting standardized woodworking, and the ramiﬁed long distance social and economic networks (Veth 1989). We can therefore date the emergence of that late Holocene strategy for desert adaptation to the terminal Pleistocene or early Holocene; that is, the period after the LGM. Like Veth (1989), we suspect that the roots of these Holocene adaptive systems may eventually be traced to the experiences of foragers in the refuges of the LGM.

Australian Desert Colonization in a Global Perspective Our synthesis indicates that an arid zone adaptation, of the kind observed historically, was absent from the initial colonization of the Australian deserts, and indeed from inland economies prior to the LGM. We therefore support Veth’s (1989) suggestion that the ethnographic forms of desert economy and subsistence are relatively recent, rather than representing features of inland occupation from the start (contra Gould 1977). Our hypothesis is that the historic form of desert adaptation in Australia emerged subsequent to the LGM, perhaps as a consequence of heavy selection/experimentation during the hyperarid conditions of that period and the later recolonization of desert areas. Because the evidence from Australia, both archaeological and environmental, is exceptional in its extent and detail, we have developed these propositions in reference to the Australian landmass. We now turn to a consideration of whether the pattern of early desert occupation in other parts of the globe displays similarities with the Australian situation. The following review brieﬂy examines available evidence from current desert regions of Africa, North America, and South America. In southern Africa the environmental and archaeological history of the Kalahari Desert is known from only fragmentary records (Deacon and Lancaster 1988). As in Australia, the period leading up to the onset of OIS2 was wetter and more humid than today, while cold, dry, and arid conditions persisted through OIS2, and wetter and more humid conditions returned in the terminal Pleistocene and early Holocene (Blumel et al. 1998; Brook et al. 1996; Lancaster 1989; Robbins et al. 1996; Scott 1989; Shaw and Thomas 1996; Thomas and Shaw 2002; Thomas et al. 2003). Archaeological data from the Tsodilo Hills in the northwest Kalahari (which, it should be noted, is currently considered sub-humid rather than arid) reveals human occupation throughout the last 100,000 years, irrespective of climatic conditions (Robbins et al. 2000). Nevertheless, during periods of higher relative precipitation and more permanent water resources (such as during the terminal Pleistocene) there is evidence for more substantial occupation, with foragers exploiting nearby lacustrine resources. This image of Kalahari foragers with knowledge of the local environment adjusting to the onset of aridity offers some similarity with our interpretation of Australian prehistory. By contrast, while it is often stated that the Namib Desert situated along the southwest coast remained hyperarid throughout the late Quaternary, there is

___________________________________________________________________________________

47

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

evidence to indicate it experienced periods of relatively increased humidity prior to the LGM and in the terminal Pleistocene–early Holocene period (Heine 1992, 1998; Lancaster 2002; Vogel 1989). Current archaeological evidence from within the Namib indicates use of the area during posited periods of relatively greater moisture availability, although given the lack of resolution about local impacts of climatic changes it is unclear whether this represents a permanent or intensive desert adaptation prior to the Holocene (Deacon and Lancaster 1988: 51–2). At the other end of the African continent the also fragmentary archaeological and climatic records from the Sahara–Sahel Desert for the period between 100,000 and 50,000 years ago tentatively reveal a positive relationship between the timing of lacustrine phases and presence of Acheulian and Levallois-Mousterian sites, although the timing of the lacustral phases is not necessarily in concert with those in the south (Petit-Maire 1991). Unfortunately, despite good climatic records of a hyperarid LGM–terminal Pleistocene followed by a substantial wet phase during the early Holocene (Gasse 2002; Haynes et al. 1989; Hoelzmann et al. 2001; Pachur and Hoelzmann 2000; Pachur and Wunnemann 1996; Salzmann 1996), the response of foragers to these environmental changes prior to the introduction of cattle domestication in the Holocene is little known. These three African areas share a pattern of long-term human occupation, perhaps extending back to and beyond the emergence of modern Homo sapiens sapiens, and there is evidence, albeit tentative, for continued occupation in some regions despite dramatic changes in climate, water availability, and food resources. Although current archaeological and paleoenvironmental evidence is limited, it appears that each region displays occupational histories that share both similarities and differences to the one reconstructed for Australia, as well as similarities and differences between each other. This perhaps suggests that no single process of desert colonization is displayed and that settlement patterns reﬂect the environmental and historical features of each landscape. This proposition is clariﬁed by summaries of the situation in the New World. Desert adaptations in the Americas offer a different context in which to test models of desert colonization, primarily because the generally accepted antiquity of humans in the New World dates to only 14,000–12,000 years ago, following the end of the LGM. Our review of the terminal Pleistocene occupation of American deserts examines three landscapes: the Great Basin of North America, the cold deserts of Patagonia and Tierra del Fuego, and the high altitude deserts of the Atacama and Peruvian coast. In the Great Basin available evidence indicates lower temperatures and higher average annual precipitation rates in the terminal Pleistocene period compared to today (Grayson 1993; Thompson 1990; Wells 1976; Woodcock 1986). Under these conditions there would have been greatly expanded water resources available within the Basin, with one estimation suggesting there would have been perhaps ten times more surface water than at present (Grayson 1993: 84–6). When combined with the presence of glaciated mountainous regions (Osborn and Bevis 2001) these lacustrine and riverine features meant that late Pleistocene huntergatherers in the Great Basin operated in a very different environment to the one

___________________________________________________________________________________

48

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s

that exists today. Archaeological evidence indicates people had moved into the Basin by 11,000 years ago and continued to be archaeologically visible and widespread until ca. 7,500 bp (see Chapter 5, this volume). During this period they were utilizing a broad array of resources from a wide range of environmental settings, although there seems little doubt ‘‘the lakes and marshes were key to the adaptations of these people’’ (Grayson 1993: 242). In contrast, high temperatures and low annual precipitation produced substantially more arid conditions during the midHolocene and marked economic change involving a settlement focus on remaining permanent sources of water and the emergence of seed grinding technology (Elston 1986; Grayson 1993; Thomas 1982; Warren and Crabtree 1986). This pattern of a late emergence of the historic desert economic system, during a phase of Holocene aridiﬁcation, parallels the Australian experience. In Patagonia and Tierra del Fuego people probably colonized now arid regions in the terminal Pleistocene as highly mobile foragers concentrating on forest resources, with initial colonization delayed in some deserts until climate change at the Pleistocene/Holocene boundary increased precipitation and water availability (Borrero 1999; Markgraf 1989, 1993; McCulloch et al. 1997; see also Chapter 8, this volume). The Atacama Desert and adjacent Peruvian coastline is today a high altitude desert. Initial occupation of this desert region apparently occurred about 13,000 years ago, when environmental conditions can be characterized as relatively humid, with extensive grass cover, abundant ﬂowing rivers, and high lake levels (Betancourt et al. 2000; Grosjean and Nun˜ez 1994; Latorre et al. 2002, 2003; Nun˜ez et al. 2002; Rech et al. 2002; Sandweiss 2003; see also Chapter 13, this volume). The extensive occupation of the region during a period prior to the transition to current environmental conditions may have facilitated the in situ development of a more speciﬁc desert economic strategy, a mechanism with similarities to the process apparent in Australia.

Conclusion Colonization of deserts is a widespread and varied human experience. Our review of selected desert areas has revealed that the timing and process of colonization has varied within and between continents, but that there are themes that are displayed in the initial movement of people into landscapes that are currently deserts. This exploration of these processes has centered on what we believe to be the most comprehensive environmental and archaeological data-set, the extensive Australian deserts. Prehistoric patterns in Australia are explicable in terms of the ‘‘desert transformation’’ model advanced herein, in which people colonize terrestrial landscapes using ﬂexible and generalized broad spectrum foraging focused on available riverine and lacustrine resources. This model posits that following the adjustment of these foragers to the local resources climatic trends towards aridiﬁcation transformed these landscapes into desert, or more extreme desert, environments. This process of foragers developing modiﬁed economic behaviors in response to

___________________________________________________________________________________

49

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s

changes to their familiar environment provides an explanation for the development of diverse practices suited to different Australian deserts. Similar processes have been observed elsewhere in the world, particularly in the Americas, and represent one mechanism by which Pleistocene foragers colonized deserts. Distinctive desert adaptations of the kind seen in historical records arose much later than the period of initial desert colonization, in many cases only well into the Holocene. The emergence and form of speciﬁc historic desert economic and social systems also varied around the globe in response to speciﬁc contingent environmental and cultural histories. For instance, the occupation of Australian deserts was interrupted, and perhaps fundamentally altered, by the severe, hyperarid conditions of the LGM. In contrast, while it seems likely that the onset of desertic conditions in the Holocene of North and South America was not as extreme as those experienced during the LGM, they were apparently still dramatic enough to precipitate substantial shifts in American forager strategies. While archaeology and paleoenvironmental data from African deserts are not as yet sufﬁciently detailed to allow such interpretations, we have little doubt that future sustained research in these localities will reveal unique histories of desert dwellers and their cultural responses to periods of extreme climatic change during the late Pleistocene and Holocene periods. Exploring the dynamic and changing relationship between climate, landscape, and the human colonization and occupation of dry regions is likely to reﬁne our understanding of the processes by which people moved into and adapted to these landscapes. The central limitation archaeologists currently face in reﬁning models of ancient desert dwellers is the rarity of detailed, paired environmental and archaeological sequences (cf. Veth et al. 2000). As high resolution paleoenvironmental and archaeological information is obtained, the choices that past peoples made about their survival strategies, and the context in which they made them, will become increasingly clear.

References Ash, J. E. and Wasson, R. J. 1983: Mobility of Australian desert duneﬁelds: 182 KA. In J. M. Chappell and J. Grindrod (eds), CLIMANZ: Proceedings of the First CLIMANZ Conference, Howman’s Gap, Victoria, 1981, Canberra: Department of Biogeography and Geomorphology, Research School of Paciﬁc Studies, The Australian National University, 46–7. Balme, J. 1995: 30,000 years of ﬁshery in western New South Wales. Archaeology in Oceania, 30, 1–21. Balme, J. 2000: Excavations revealing 40,000 years of occupation at Mimbi Caves, south central Kimberley, Western Australia. Australian Archaeology, 51, 1–5. Beaton, J. M. 1985: Evidence for a coastal occupation time lag at Princess Charlotte Bay (north Queensland) and implication for coastal colonization and population growth theories for Aboriginal Australia. Archaeology in Oceania, 20, 1–20. Betancourt, J. L., Latorre, C., Rech, J. A., Quade, J., and Rylander, K. A. 2000: A 22,000 year record of monsoonal precipitation from northern Chile’s Atacama Desert. Science, 289, 1542–6.

___________________________________________________________________________________

50

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s Blu¨mel, W. D., Eitel, B., and Lang, A. 1998: Dunes in southeastern Namibia: Evidence for Holocene environmental changes in the southeastern Kalahari based on thermoluminescence data. Palaeogeography, Palaeoclimatology, Palaeoecology, 138, 139–49. Borrero, L. A. 1999: The prehistoric exploration and colonization of Fuego-Patagonia. Journal of World Prehistory, 13 (3), 321–55. Bowdery, D. 1998: Phytolith Analysis applied to Pleistocene–Holocene Archaeological Sites in the Australian Arid Zone. Oxford: BAR International Monograph Series no. 695, Hadrian Books. Bowdler, S. 1977: The coastal colonization of Australia. In J. Allen, J. Golson, and R. Jones (eds), Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia and Australia, London: Academic Press, 205–46. Bowler, J. M. 1983a: 325KA Northern Australia: Hydrologic evidence. In J. M. Chappell and J. Grindrod (eds), CLIMANZ: Proceedings of the First CLIMANZ Conference, Howman’s Gap, Victoria, 1981, Canberra: Department of Biogeography and Geomorphology, Research School of Paciﬁc Studies, The Australian National University, 4–6. Bowler, J.M. 1983b: 325KA Southern Australia; Hydrologic evidence. In J. M. Chappell and J. Grindrod (eds), CLIMANZ: Proceedings of the First CLIMANZ Conference, Howman’s Gap, Victoria, 1981, Canberra: Department of Biogeography and Geomorphology, Research School of Paciﬁc Studies, The Australian National University, 6–9. Bowler, J. M. 1986: Quaternary landform evolution. In D. N. Jeans (ed.), The Natural Environment – A Geography of Australia, vol. 1, Sydney: Sydney University Press, 117–47. Bowler, J. M. 1998: Willandra Lakes revisited: Environmental framework for human occupation. Archaeology in Oceania, 33, 120–55. Bowler, J. M. and Magee, J. W. 2000: Redating Australia’s oldest human remains: A sceptic’s view. Journal of Human Evolution, 38, 719–26. Bowler, J. M. and Price, D. M. 1998: Luminescence dates and stratigraphic analyses at Lake Mungo: Review and new perspectives. Archaeology in Oceania, 33, 156–68. Bowler, J. M. and Wasson, R. J. 1984: Glacial age environments of inland Australia. In J. C. Vogel (ed.), Late Cainozoic Paleoclimates of the Southern Hemisphere, Rotterdam: A. A. Balkema, 183–208. Bowler, J. M., Duller, G. A. T., Perret, N., Prescott, J. R., and Wrywoll, K.-H. 1998: Hydrologic changes in monsoonal climates of the last glacial cycle: Stratigraphy and luminescence dating of Lake Woods, NT, Australia. Palaeoclimates, 3 (1–3), 179–207. Bowler, J. M., Johnston, H., Olley, J. M., Prescott, J. R., Roberts, R. G., Shawcross, W., and Spooner, N. A. 2003: New ages for human occupation and climatic change at Lake Mungo, Australia. Nature, 421, 837–40. Bowler, J. M., Wyrwoll, K.-H., and Lu, Y. 2001: Variations of the northwest Australian summer monsoon over the last 300,000 years: The paleohydrological record of the Gregory (Mulan) Lakes System. Quaternary International, 83–5, 63–80. Brook, G. A., Cowart, J. B., and Marais, E. 1996: Wet and dry periods in the southern African summer rainfall zone during the last 300 kyr from speleothem, tufa, and sand dune age data. Palaeoecology of Africa, 24, 147–58. Chappell, J. 1991: Late Quaternary environmental changes in eastern and central Australia and their climatic interpretation. Quaternary Science Reviews, 10, 377–90. Deacon, J. and Lancaster, N. 1988: Late Quaternary Paleoenvironments of Southern Africa. Oxford: Clarendon Press. Dodson, J. R. and Wright, R. V. S. 1989: Humid to arid to sub-humid vegetation shifts on Pilliga Sandstone, Ulungra Springs, New South Wales. Quaternary Research, 32, 182–92.

___________________________________________________________________________________

51

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s Dodson, J. R., Fullagar, R., Furby, J., Jones, R., and Prosser, I. 1993: Humans and megafauna in a late Pleistocene environment from Cuddie Springs, northwestern New South Wales. Archaeology in Oceania, 28, 94–9. Elston, R. G. 1986: Prehistory of the western area. In W. L. d’Azevedo (ed.), Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 135–48. Field, J. and Dodson, J. 1999: Late Pleistocene megafauna and archaeology from Cuddie Springs, southeastern Australia. Proceedings of the Prehistoric Society, 65, 1–27. Fiﬁeld, L. K., Bird, M. I., Turney, C. S. M., Hausladen, P. A., Santos, G. M., and di Tada, M. L. 2001: Radiocarbon dating of the human occupation of Australia prior to 40 ka bp – successes and pitfalls. Radiocarbon, 43, 1139–45. Fullagar, R. and Field, J. 1997: Pleistocene seed grinding implements from the Australian arid zone. Antiquity, 71, 300–7. Gasse, F. 2002: Diatom-inferred salinity and carbonate oxygen isotopes in Holocene waterbodies of the western Sahara and Sahel (Africa). Quaternary Science Reviews, 21, 737–67. Gentilli, J. 1986: Climate. In D. N. Jeans (ed.), The Natural Environment – A Geography of Australia vol. 1, Sydney: Sydney University Press, 14–48. Gillespie, R. 1998: Alternative timescales: A critical review of Willandra Lakes dating. Archaeology in Oceania, 33, 169–82. Gillespie, R. and Roberts, R. G. 2000: On the reliability of age estimates for human remains at Lake Mungo. Journal of Human Evolution, 38, 727–32. Gorecki, P., Grant, M., O’Connor, S., and Veth, P. 1997: The morphology, function, and antiquity of Australian grinding implements. Archaeology in Oceania, 32 (2), 141–50. Gould, R. A. 1977: Puntutjarpa Rockshelter and the Australian Desert Culture. New York: Anthropological Papers of the American Museum of Natural History no. 54, American Museum of Natural History. Grayson, D. K. 1993: The Desert’s Past: A Natural Prehistory of the Great Basin. Washington, DC: Smithsonian Institution Press. Grosjean, M. and Nun˜ez, L. A. 1994: Late glacial, early, and middle Holocene environments, human occupation and resource use in the Atacama (northern Chile). Geoarchaeology, 9, 271–86. Gru¨n, R., Spooner, N. A., Thorne, A., Mortimer, G., Simpson, J. J., McCulloch, M. T., Taylor, L., and Curnoe, D. 2000: Age of the Lake Mungo 3 skeleton – reply to Bowler & Magee and to Gillespie & Roberts. Journal of Human Evolution, 38, 733–42. Harrison, S. P. 1993: Late Quaternary lake level changes and climates of Australia. Quaternary Science Reviews, 12, 211–31. Haynes Jr, C. V., Eyles, C. H., Pavlish, L. A., Ritchie, J. C., and Rybak, M. 1989: Holocene paleohydrology of the eastern Sahara: Selima Oasis. Quaternary Science Reviews, 8, 109–36. Heine, K. 1992: On the ages of humid Late Quaternary phase in southern African arid areas (Namibia, Botswana). Palaeoecology of Africa, 23, 149–64. Heine, K. 1998: Climate change over the past 135,000 years in the Namib Desert (Namibia) derived from proxy data. Palaeoecology of Africa, 25, 171–98. Hesse, P. P. 1994: The record of continental dust from Australia in Tasman Sea sediments. Quaternary Science Reviews, 13, 257–72. Hesse, P. P. and McTainsh, G. H. 1999: Last Glacial Maximum to early Holocene wind strength in the mid-latitudes of the Southern Hemisphere from aeolian dust in the Tasman Sea. Quaternary Research, 52, 343–9.

___________________________________________________________________________________

52

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s Hesse, P. P., Magee, J. W., and van der Kaars, S. 2004: Late Quaternary climates of the Australian arid zone: A review. Quaternary International, 118–19, 87–102. Hiscock, P. 1984: Preliminary report on the stone artifacts from Colless Creek Cave, northwest Queensland. Queensland Archaeological Research, 1, 120–51. Hiscock, P. 1988: Prehistoric settlement patterns and artifact manufacture at Lawn Hill, northwest Queensland. Unpublished PhD thesis, Brisbane: University of Queensland. Hiscock, P. 1990: Comment: How old are the artifacts at Malakunanja II? Archaeology in Oceania, 25, 122–4. Hoelzmann, P., Keding, B., Berke, H., Kro¨pelin, S., and Kruse, H. J. 2001: Environmental change and archaeology: Lake evolution and human occupation in the eastern Sahara during the Holocene. Palaeogeography, Palaeoclimatology, Palaeoecology, 169, 193–217. Hope, J. H. 1993: Pleistocene archaeological sites in the central Murray Darling Basin. In M. A. Smith, M. Spriggs, and B. Fankhauser (eds), Sahul in Review: Pleistocene Archaeology in Australia, New Guinea, and Island Melanesia, Canberra: The Australian National University, 183–96. Hubbard, N. N. 1995: In search of regional palaeoclimates: Australia 18,000 yr bp. Palaeogeography, Palaeoclimatology, Palaeoecology, 116, 167–88. Hughes, P. and Hiscock, P. in press: The archaeology of the Lake Eyre South area. In Aboriginal Occupation of the Lake Eyre South Region: Ethnography and Archaeology, Adelaide: Lake Eyre Monograph Series vol. 6, Royal Geographical Society of South Australia. Johnson, B. J., Miller, G. H., Fogel, M. L., Magee, J. W., Gagan, M. K., and Chivas, A. R. 1999: 65,000 years of vegetation change in central Australia and the Australian summer monsoon. Science, 284, 1150–2. Jones, R. 1979: The ﬁfth continent: Problems concerning the human colonization of Australia. Annual Review of Anthropology, 8, 445–6. Jones, R. and Bowler, J. M. 1980: Struggle for the savannah: Northern Australia in ecological and prehistoric perspective. In R. Jones (ed.), Northern Australia: Options and Implications, Canberra: The Australian National University, 3–32. Kelly, R. L. 2003: Colonization of new land by hunter-gatherers: Expectations and implications based on ethnographic data. In M. Rockman and J. Steele (eds), Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation, New York: Routledge, 44–58. Lamb, L. 1996: Investigating changing stone technologies, site use, and occupational intensities at Fern Cave, north Queensland. Australian Archaeology, 42, 1–7. Lambeck, K. and Chappell, J. 2001: Sea level change through the last glacial cycle. Science, 292, 679–86. Lambeck, K., Yokoyama, Y., and Purcell, T. 2002: Into and out of the Last Glacial Maximum: Sea level change during Oxygen Isotope Stages 3 and 2. Quaternary Science Reviews, 21, 343–60. Lancaster, N. 1989: Late Quaternary paleoenvironments in the southwestern Kalahari. Palaeogeography, Palaeoclimatology, Palaeoecology, 70, 367–76. Lancaster, N. 2002: How dry was dry? Late Pleistocene paleoclimates in the Namib Desert. Quaternary Science Reviews, 21, 769–82. Latorre, C., Betancourt, J. L., Rylander, K. A., and Quade, J. 2002: Vegetation invasions into absolute desert: A 45,000 year rodent midden record from the Calam-Salar de Atacama basins, northern Chile (lat 228–248S). Geological Society of America Bulletin, 114 (3), 349–66.

___________________________________________________________________________________

53

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s Latorre, C., Betancourt, J. L., Rylander, K. A., Quade, J., and Matthei, O. 2003: A vegetation history from the arid prepuna of northern Chile (22–238S) over the last 13,500 years. Palaeogeography, Palaeoclimatology, Palaeoecology, 194, 223–46. McCulloch, R., Clapperton, C., Rabassa, J., and Currant, A. P. 1997: The natural setting – The glacial and post-glacial environmental history of Fuego-Patagonia. In C. McEwan, L. A. Borerro, and A. Prieto (eds), Patagonia: Natural History, Prehistory, and Ethnography at the Uttermost Ends of the Earth, London: British Museum Press. Magee, J. W. and Miller, G. H. 1998: Lake Eyre paleohydrology from 60 ka to the present: Beach ridges and glacial maximum aridity. Palaeogeography, Palaeoclimatology, Palaeoecology, 144, 307–29. Magee, J. W., Bowler, J. M., Miller, G. H., and Williams, D. L. G. 1995: Stratigraphy, sedimentology, chronology, and paleohydrology of Quaternary lacustrine deposits at Madigan Gulf, Lake Eyre, South Australia. Palaeogeography, Palaeoclimatology, Palaeoecology, 113, 3–42. Markgraf, V. 1989: Paleoclimates in central and South America since 18,000 based on pollen and lake level records. Quaternary Science Reviews, 8, 1–24. Markgraf, V. 1993. Paleoenvironments and paleoclimates in Tierra del Fuego and southernmost Patagonia, South America. Palaeogeography, Palaeoclimatology, Palaeoecology, 102, 53–68. Marwick, B. 2002: Milly’s Cave: Evidence for human occupation of the inland Pilbara during the Last Glacial Maximum. In S. Ulm, C. Westcott, J. Reid, A. Ross, I. Lilley, J. Prangnell, and L. Kirkwood (eds), Barriers, Borders, Boundaries: Proceedings of the 2001 Australian Archaeological Association Annual Conference, St Lucia: University of Queensland, 21–33. Miller, G. H., Magee, J. W., and Jull, A. J. T. 1997: Low latitude cooling in the Southern Hemisphere from amino-acid racemization in emu eggshells. Nature, 385, 241–4. Miller, G. H., Magee, J. W., Johnson, B. J., Fogel, M. L., Spooner, N. A., McCulloch, M. T., and Ayliffe, L. K. 1999: Pleistocene extinction of Genyornis newtoni: Human impact on Australian megafauna. Science, 283, 205–8. Morse, K. 1988: Mandu Mandu Creek rock shelter: Pleistocene human coastal occupation of North West Cape, Western Australia. Archaeology in Oceania, 23 (3), 81–8. Morse, K. 1993: Who can see the sea? Aboriginal occupation of the Cape Range Peninsula. Records of the Western Australian Museum Supplement, 45, 227–42. Nanson, G. C., Price, D. M., Short, S. A., Page, K. J., and Nott, J. F. 1991: Major episodes of climate change in Australia over the last 300,000 years. In R. Gillespie (ed.), Quaternary Dating Workshop 1990, Canberra: Department of Biogeography and Geomorphology, Research School of Paciﬁc Studies, The Australian National University, 45–50. Nun˜ez, L., Grosjean, M., and Cartajena, I. 2002: Human occupations and climate change in the Puna de Atacama, Chile. Science, 298, 821–4. O’Connell, J. F. and Allen, J. 1998: When did humans ﬁrst arrive in Greater Australia and why is it important to know? Evolutionary Anthropology, 6, 132–46. O’Connell, J. F. and Allen, J. in press: Dating the colonization of Sahul (Pleistocene Australia–New Guinea): A review of recent research. Journal of Archaeological Science. O’Connor, S. and Veth, P. 2000: The world’s ﬁrst mariners: Savanna dwellers in an island continent. Modern Quaternary Research in Southeast Asia, 16, 99–137. O’Connor, S., Veth, P., and Barham, A. 1999: Cultural versus natural explanations for lacunae in Aboriginal occupation deposits in northern Australia. Quaternary International, 59, 61–70.

___________________________________________________________________________________

54

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s O’Connor, S., Veth, P., and Campbell, C. 1998: Serpent’s Glen rock shelter: Report of the ﬁrst Pleistocene-aged occupation sequence from the Western Desert. Australian Archaeology, 46, 12–21. O’Connor, S., Veth, P., and Hubbard, N. 1993: Changing interpretations of postglacial human subsistence and demography in Sahul. In M. A. Smith, M. Spriggs, and B. Fankhauser (eds), Sahul in Review: The Archaeology of Australia, New Guinea and Island Melanesia, Canberra: The Australian National University, 95–105. Osborn, G. and Bevis, K. 2001: Glaciation in the Great Basin of the Western United States. Quaternary Science Reviews, 20, 1377–410. Pachur, H. J. and Hoelzmann, P. 2000: Late Quaternary paleoecology and paleoclimates of the eastern Sahara. Journal of African Earth Sciences, 30 (4), 929–39. Pachur, H. J. and Wunnemann, B. 1996: Reconstruction of the paleoclimate along 308E in the eastern Sahara during the Pleistocene/Holocene transition. Palaeoecology of Africa, 24, 1–32. Pack, S. M., Miller, G. M., Fogel, M. L., and Spooner, N. A. 2003: Carbon isotopic evidence for increased aridity in northwestern Australia through the Quaternary. Quaternary Science Reviews, 22, 629–43. Pearce, R. H. and Barbetti, M. 1981: A 38,000 year old archaeological site at Upper Swan, Western Australia. Archaeology in Oceania, 16, 173–8. Petit-Maire, N. 1991: Recent Quaternary climatic change and man in the Sahara. Journal of African Earth Sciences, 12 (1/2), 125–32. Rech, J. A., Quade, J., and Betancourt, J. L. 2002: Late Quaternary paleohydrology of the central Atacama Desert (lat 228–248S), Chile. Geological Society of America Bulletin, 114 (3), 334–48. Richardson, N. 1992: Conjoin sets and stratigraphic integrity in a sandstone shelter: Kenniff Cave (Queensland, Australia). Antiquity, 66, 408–18. Rindos, D. and Webb, E. 1992: Modeling the initial human colonization of Australia: Perfect adaptation, cultural variability, and cultural change. Proceedings of the Australasian Society of Human Biology, 5, 441–54. Robbins, L. H., Murphy, M. L., Brook, G. A., Ivester, A. H., Campbell, A. C., Klein, R. G., Milo, R. G., Stewart, K. M., Downey, W. S., and Stevens, N. J. 2000: Archaeology, paleoenvironment, and chronology of the Tsodilo Hills White Paintings rock shelter, northwest Kalahari Desert, Botswana. Journal of Archaeological Science, 27, 1085–113. Robbins, L. H., Murphy, M. L., Stevens, N. J., Brook, G. A., Ivester, A. H., Haberyan, K. A., Klein, R. G., Milo, R., Stewart, K. M., Matthiesen, D. G., and Winkler, A. J. 1996: Paleoenvironment and archaeology of Drotsky’s Cave: Western Kalahari Desert, Botswana. Journal of Archaeological Science, 23, 7–22. Roberts, R., Flannery, T., Ayliffe, L., Yoshida, H., Olley, J., Prideaux, G., Laslett, G., Baynes, A., Smith, M., Jones, R., and Smith, B. 2001: New ages for the last Australian megafauna: Continent-wide extinction about 46,000 years ago. Science, 292, 1888–92. Roberts, R. G., Jones, R., and Smith, M. A. 1990: Thermoluminescence dating of a 50,000 year old human occupation site in northern Australia. Nature, 345, 153–6. Roberts, R. G., Jones, R., and Smith, M. A. 1993: Optical dating at Deaf Adder Gorge, Northern Territory, indicates human occupation between 53,000 and 60,000 years ago. Australian Archaeology, 37, 58–9. Roberts, R. G., Jones, R., Spooner, N. A., Head, M. J., Murray, A. S., and Smith, M. A. 1994: The human colonization of Australia: Optical dates of 53,000 and 60,000 years bracket

___________________________________________________________________________________

55

_______________________________________________________________________________

P e t e r H is c oc k a n d L y n l e y A . W a l l i s human arrival at Deaf Adder Gorge, Northern Territory. Quaternary Science Reviews, 13, 575–83. Roberts, R. G., Spooner, N. A., Jones, R., Cane, S., Olley, J. M., Murray, A. S., and Head, J. 1996: Preliminary luminescence dating for archaeological sediments on the Nullarbor Plain, South Australia. Australian Archaeology, 42, 7–16. Salzmann, U. 1996: Holocene vegetation history of the Sahelian zone of NE Nigeria: Preliminary results. Palaeoecology of Africa, 24, 103–14. Sandweiss, D. 2003: Terminal Pleistocene through mid-Holocene archaeological sites as paleoclimatic archives for the Peruvian coast. Palaeogeography, Palaeoclimatology, Palaeoecology, 194, 23–40. Scott, L. 1989: Climatic conditions in southern Africa since the Last Glacial Maximum, inferred from pollen analysis. Palaeogeography, Palaeoclimatology, Palaeoecology, 70, 345–53. Shaw, P. A. and Thomas, D. S. G. 1996: The Quaternary paleoenvironmental history of the Kalahari, southern Africa. Journal of Arid Environments, 32, 9–22. Singh, G. and Geissler, E. A. 1985: Late Cainozoic history of vegetation, ﬁre, lake levels, and climate at Lake George, New South Wales, Australia. Philosophical Transactions of the Royal Society of London, B311, 379–447. Smith, M. A. 1987: Pleistocene occupation in arid central Australia. Nature, 328, 710. Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24 (3), 93–105. Smith, M. A., Prescott, J. R., and Head, M. J. 1997: Comparison of 14C and luminescence chronologies at Puritjarra rock shelter, central Australia. Quaternary Science Reviews, 16, 299–320. Smith, M. A., Vellen, L., and Pask, J. 1995: Vegetation history from archaeological charcoals in central Australia: The late Quaternary record from Puritjarra rock shelter. Vegetation History and Archaeobotany, 4, 171–7. Thomas, D. H. 1982: An overview of central Great Basin prehistory. Society for American Archaeology Papers, 2, 156–71. Thomas, D. S. G. and Shaw, P. A. 2002: Late Quaternary environmental change in central southern Africa: New data, synthesis, issues, and prospects. Quaternary Science Reviews, 21, 783–97. Thomas, D. S. G., Brook, G., Shaw, P., Bateman, M., Haberyan, K., Appletone, C., Nash, D., McLaren, S., and Davies, F. 2003: Late Pleistocene wetting and drying in the NW Kalahari: An integrated study from the Tsodilo Hills, Botswana. Quaternary International, 104, 53–67. Thompson, R. S. 1990: Late Quaternary vegetation and climate in the Great Basin. In J. L. Betancourt, T. R. Van der Devender, and P. S. Martin (eds), Packrat Middens: The Last 40,000 Years of Biotic Change, Tucson: University of Arizona Press, 200–39. Thorley, P. B. 1998: Pleistocene settlement in the Australian arid zone: Occupation of an inland riverine landscape in the central Australian ranges. Antiquity, 72, 34–45. Thorley, P. B. 2001: Uncertain supplies: Water availability and regional archaeological structure in the Palmer River catchment, central Australia. Archaeology in Oceania, 36, 1–14. Thorne, A., Gru¨n, R., Mortimer, G., Spooner, N. A., Simpson, J. J., McCulloch, M., Taylor, L., and Curnoe, D. 1999: Australia’s oldest human remains: Age of the Lake Mungo 3 skeleton. Journal of Human Evolution, 36 (6), 591–612. Tonkinson, R. 1991: The Martu Aboriginies: Living the Dream of Australia’s Desert. Forth Worth, TX: Holt, Rinehart, and Winston.

___________________________________________________________________________________

56

_______________________________________________________________________________

P l e i s t o c en e S e t t l e m e n t o f D e s e r t s van der Kaars, S. A. 1991: Palynology of eastern Indonesian marine piston cores: A late Quaternary vegetational and climatic record for Australasia. Palaeogeography, Palaeoclimatology, Palaeoecology, 85, 239–302. Veth, P. 1989: Islands in the interior: A model for the colonization of Australia’s arid zone. Archaeology in Oceania, 24 (3), 81–92. Veth, P., O’Connor, S., and Wallis, L. A. 2000: Perspectives on ecological approaches in Australian archaeology. Australian Archaeology, 50, 54–66. Vogel, J. C. 1989: Evidence of past climate change in the Namib Desert. Palaeogeography, Palaeoclimatology, Palaeoecology, 70, 355–66. Wallis, L. A. 2001: Environmental history of northwest Australia based on phytolith analysis at Carpenter’s Gap 1. Quaternary International, 83–5, 103–17. Wang, X., van der Kaars, S., Kershaw, A. P., Bird, M., and Jansen, F. 1999: A record of ﬁre, vegetation, and climate through the last three glacial cycles from Lombok Ridge core G64, eastern Indian Ocean, Indonesia. Palaeogeography, Palaeoclimatology, Palaeoecology, 147, (3–4), 241–56. Warren, C. N. and Crabtree, R. H. 1986: Prehistory of the southwestern area. In W. L. d’Azevedo (ed.), Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 183–93. Wasson, R. J. 1984: Late Quaternary paleoenvironments in the desert duneﬁelds of Australia. In J. C. Vogel (ed.), Late Cainozoic Paleoclimates of the Southern Hemisphere, Rotterdam: A. A. Balkema, 419–32. Wells, P. V. 1976: Macrofossil analysis of wood rat (Neotoma) middens as a key to the Quaternary vegetational history of arid America. Quaternary Research, 6, 223–48. Woodcock, D. 1986: The late Pleistocene of Death Valley: A climatic reconstruction based on macrofossil data. Palaeogeography, Palaeoclimatology, Palaeoecology, 57, 272–83. Wroe, S., Field, J., and Fullagar, R. 2002: Lost giants. Nature Australia, 27 (5), 54–61. Yokoyama, Y., Purcell, A., Lambeck, K., and Johnston, P. 2001: Shoreline reconstruction around Australia during the Last Glacial Maximum and Late Glacial Stage. Quaternary International, 83–5, 9–18.

___________________________________________________________________________________

57

_______________________________________________________________________________

--------------------------------------------------------------------------------

4

----------------------------------------------------------------------------

Arid Paradises or Dangerous Landscapes A Review of Explanations for Paleolithic Assemblage Change in Arid Australia and Africa Peter Hiscock and Sue O’Connor

Introduction An inspection of archaeological studies from the arid and semi-arid regions of Australia and sub-Saharan Africa reveals very different images of the human occupation of each continent. Since the human history of each landmass is obviously distinct, many aspects of these images of the human past may reﬂect divergent cultural processes being played out in dissimilar arid environments. Such arguments fail to explain why the archaeological record in each continent displays parallel patterns, and discount fundamental questions about our characterization and explication of archaeological changes in the two continents. For example, in both Africa and Australia ancient humans employed a particular technology that involved backing of ﬂakes to produce distinctive and regularly shaped stone implements that we will call backed artifacts. (The term microlith has often been used for these forms but is applicable only to small size variants.) In both continents this technological behavior has been present for extended periods of time, but was emphasized in some periods but not in others. Rather than presume these parallel patterns must be unrelated or are uninteresting, this chapter examines the explanations that have been offered for the periodic emphasis on backed artifacts by archaeologists exploring the human history of these two continents. Our examination reveals that the formulation of seemingly similar themes of archaeological change has been radically different in each continent, and we employ that insight to both evaluate the viability of various kinds of explanations and to demonstrate that the inferred histories of humans in each landmass are founded on different images of human life in deserts.

___________________________________________________________________________________

58

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

Depicting the Archaeological Changes Our starting point in this exploration is to describe the broad chronological and spatial patterns of backed implements on each continent and to discuss the way archaeologists have described changes in lithic assemblages. Our review includes sites that are presently in arid and semi-arid environments, but is of course not limited to those places, since the location of environmental boundaries has shifted through time.

Assemblage Change in Australia Backed artifacts have been reliably dated to the early Holocene (9,000–6,000 bp) in Australia, but may have been present in the terminal Pleistocene (Hiscock and Attenbrow 1998). However, prior to the mid-Holocene backed artifact production was clearly occurring at low levels and may have been regionally restricted (Hiscock 2002). For at least some regions in eastern Australia these early examples of backed artifacts appear to be extremely similar to the size and morphological patterns found in later backed artifacts in each region, although the size and manufacturing strategies vary between regions. The relationship between these terminal Pleistocene/early Holocene specimens and later backed artifacts was unclear until Hiscock and Attenbrow (1998) emphasized the importance of distinguishing between the appearance of backed artifacts and subsequent periods of increased rates of production. The importance of this distinction is that low rates of backed artifact production may be hard to detect archaeologically and such a period might be mistaken for one without such implements, while a subsequent period in which backed artifacts were frequently manufactured will be highly visible in the archaeological record and may appear to signal the ‘‘introduction’’ of backed artifacts to a region when in fact it represents the ‘‘efﬂorescence’’ of production. This methodological issue was accentuated by the low discovery probabilities for backed artifacts in small assemblages that were common in the early Holocene levels of a number of key sites (Hiscock 2001). These considerations have removed doubts about the existence of early Holocene backed artifacts and clariﬁed the emergence of intensive backed artifact production following an extended period of low level backed artifact manufacture (Hiscock 2002). During the mid- and late Holocene the production rate for backed artifacts increased dramatically (Hiscock 2002). The timing of this proliferation appears to vary between regions, with some regions displaying marked increases in discarded backed artifacts around 4,000 bp, while other regions show the increased rate of backed artifact production at about 3,000 bp or less. This phase of high rates of backed artifact manufacture was brief in Australia, and in most localities was concluded by 1,000 bp. In some regions the chronology of these changes is not yet precisely known. The regionalized nature of these patterns is displayed on a continental scale. Backed artifacts are absent from the tropical north-central and northwestern

___________________________________________________________________________________

59

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

portions of the continent, landscapes in which bifacial points were manufactured. The island of Tasmania also has no recorded backed artifacts. Elsewhere on mainland Australia backed artifacts are known, although their relative and absolute abundance varies regionally. One broad scale spatial pattern involves the differential distribution of symmetrical and asymmetrical variants. Asymmetrical backed artifacts are very common on the east coast and in the southwest of the continent, while symmetrical crescent, triangle, and trapeze forms are abundant in the arid and semi-arid landscapes. The signiﬁcance of these geographic differences for explanations of assemblage variation is currently being debated.

Assemblage Change in Africa By comparison to the Australian situation the history of backed artifacts in subSaharan Africa is both long and complex. Large, thick backed artifacts have been present in very low quantities in at least some African assemblages during the initial phases of the Middle Stone Age (MSA) more than 300,000–200,000 years ago, although they were often infrequent. However, in central Africa the Lupemban Industry comprised assemblages dominated by backed artifacts dated to 300,000– 250,000 bp (Barham 2002; Clark 1988; Cornelissen 2002). This demonstrates that backing of small ﬂakes is a technical process that is present in sub-Saharan Africa from the very start of the MSA. In southern Africa backed artifacts are known from sites such as Klasies River during the last interglacial and are present in assemblages designated as Phases 1 and 2 of the MSA, labeled MSA 1 and MSA 2 (Wurz 2002). However, although present, backing is comparatively infrequent in MSA 1/2 assemblages of southern Africa. Nevertheless, backed artifact production proliferates approximately 80,000– 70,000 bp in a series of southern African assemblages. Termed the Howieson’s Poort Industry, these assemblages are best known for their numerous backed artifacts, although they also contain a range of other retouched artifacts including bifacially and unifacially ﬂaked points with bulb thinning (Deacon and Wurz 1996). As represented at sites such as Klasies River these backed objects were regular in shape and were often made from non-local, ﬁne-grained siliceous raw materials (Ambrose 2002; Wurz 2002). Following the Howieson’s Poort Industries the subsequent culture historical periods, Phases 3 and 4 of the MSA, labeled MSA 3 and MSA 4, have no recorded backed artifacts. Although the MSA 3/4 assemblages are rare and low density phenomena, creating questions of sample adequacy in assessing assemblage composition, the consensus is that backed artifacts are absent from the MSA 3 and MSA 4 assemblages that succeed Howieson’s Poort (Ambrose 2002). However, this interpretation is not universally supported by archaeological evidence and in eastern African assemblages backed artifacts are present in the MSA, from at least 65,000–50,000 bp (McBrearty and Brooks 2000). We therefore suggest that MSA 3/4 assemblages have very few backed specimens, but that it is not clear that such implements were absent from all sites during those phases.

___________________________________________________________________________________

60

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

In southern Africa archaeologists classify assemblages containing small backed artifacts, points, bipolar cores, and denticulates, often making extensive use of good quality ﬁne-grained raw material, as the Later Stone Age (LSA). The chronological and culture-historical coherence of early LSA assemblages is a matter of uncertainty, with most examples dating to less than 25,000 bp while a few sites display these characteristics at approximately 40,000 bp (Ambrose 2002). The early LSA assemblages have been classiﬁed into a variety of industries, including the Robberg and Albany, but the relevant feature is that although they typically contain backed artifacts these are infrequent elements in these assemblages (although see Mitchell 1988). Transitional and LSA industries in east Africa, such as the Nasera and Lemuta Industries of Tanzania and Kenya, also have backed artifacts and high quality non-local raw material imported a considerable distance (Ambrose 2002). In the early LSA industries of east Africa the proportional frequency of backed artifacts is highly variable. While backed specimens are extremely frequent in some assemblages, such as the Nasampolai as represented in Enkapune Ya Muto shelter, they are minor elements in many other east African LSA assemblages of the period prior to 25,000 bp. In both eastern and southern Africa during the LGM/terminal Pleistocene (20,000–13,000 bp) backed artifacts are less common in most industries than at earlier and later times (Ambrose 2002). Backed artifacts are again a dominant element in some Holocene industries, most notably that called the Wilton, which is concentrated in the more arid period of the mid-Holocene. Wilton assemblages have high frequencies of regular, usually small backed artifacts and ﬁne-grained raw materials. Volman (1984) and others have suggested that both the MSA and LSA backed artifacts were produced for hafting. Direct evidence is sparse, although a few examples have been found with mastic traces. Because of the relative difference in size it has been suggested that MSA backed artifacts were hafted in spears, whereas their LSA counterparts were inserted into arrows. Employing the distinction advocated by Hiscock and Attenbrow (1998) between periods representing a proliferation or efﬂorescence of backed artifacts and periods in which they are present but not emphasized, the sub-Saharan African sequence can be read as having multiple phases in which backed artifacts were produced at relatively high rates (measured as a proportion of retouched ﬂakes), interspersed with phases in which backed artifacts were either absent or produced in relatively lower frequencies. While the patchy nature of high quality assemblage information through time and space makes it difﬁcult to unambiguously specify the nature and duration of periods it is clear that we can talk of at least three or four efﬂorescence phases. The earliest is the early MSA backed artifact proliferation discussed by Barham (2002) for sites such as Twin Rivers and Kalambo Falls in central Africa. A distinct efﬂorescence is represented by Howieson’s Poort (ca. 70,000 bp), which is usually thought to be chronologically far earlier than LSA assemblages and therefore ‘‘is not considered to be transitional to the LSA’’ (Ambrose 2002: 12). A third period of intensiﬁed backed artifact production is found in some regional variants of the early LSA (50,000–30,000 bp). The fourth obvious efﬂorescence is represented by the Wilton and similar Holocene industries

___________________________________________________________________________________

61

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

(8,000–3,000 bp). It is not our intention to hypothesize that there are only four periods in which assemblages have large quantities of backed implements, that during a period of high backed artifact production rates every assemblage in a region will display that pattern, or that this sequence applies across all of subSaharan Africa. What is important here is the recognition of the reoccurrence of technological patterns emphasizing backed artifacts. In Africa, assemblages dominated by backed artifacts are neither randomly situated in time and space nor restricted to a single culture-historical horizon. It is not merely the existence of early backed artifacts in the African sequence that contrasts with Australia and requires explanation, but also this reoccurring pattern of backed artifact efﬂorescence through time.

Explanations for Backed Artifact Efﬂorescence We now turn to different explanations that archaeologists have offered for ancient backed artifacts in these two continents. Our concern is to characterize inﬂuential explanations and to evaluate their capacity to explicate the archaeological patterns as we have described them. Mitchell (1995) has presented a classiﬁcation of ‘‘approaches’’ to Paleolithic studies in south Africa, in which he describes four kinds of models: organizational, functional, sociological, and techno-environmental. Rather than engage in a classiﬁcatory exercise of this kind our critique proceeds by examining speciﬁc explanations that have been offered for changes in backed artifact frequency.

Cognitive Capacity as an Explanation One widely discussed explanation for the proliferation of backed artifacts in Africa is the emergence of modern human behavior or ‘‘cultural modernity’’ (Wadley 2001; see also Ambrose 2001). This explanation is based on the notion that earlier hominids had little or no capacity for language and symbolism and the social interactions that follow from those capacities. It is often argued that these capacities deﬁne modern human social life and that they must have emerged rapidly, as a revolution in human evolution (Wadley 2001). Authors such as Klein (2000: 27) have suggested a biological basis for the revolutionary shift in cultural capacities, arguing that we are observing a change in the ‘‘neural capacity for language or for ‘symboling.’ ’’ The archaeological signatures of such modern behavior are typically argued to be the manipulation and storage of symbolic information, represented by works of art and ornamentation, by campsite structures that reveal group interaction, by ritualized burials, and by stylistic expression in artifacts (e.g., Henshilwood et al. 2001a, 2002; Wadley 2001). It is this ﬁnal feature that has been invoked in explaining patterns of backed artifacts in Africa. Some researchers have argued that the nature of backing, perhaps being preparation for insertion in a composite tool, and the precise and repeated form of backed artifacts, makes them indicators of speciﬁcally designed and style-laden behavior indicative of

___________________________________________________________________________________

62

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

cultural modernity. For instance, Barham (2001, 2002) concludes that the existence of large numbers of backed artifacts in the Middle Pleistocene Lupemban Industry indicates a new suite of behaviors were present approximately 300,000 bp. Other archaeologists demand more stringent evidence and therefore place the transition to cultural modernity at more recent dates. For example, a number of archaeologists have hypothesized that the standardized and rather modern looking backed specimens common in the Howieson’s Poort Industry are a better signal that modern-like tool concepts and social networks are present, and therefore argue that the Howieson’s Poort represents the transition point to modern behavioral systems at approximately 70,000 bp (Wurz 1999, 2002). Even more conservative arguments are presented by archaeologists such as Wadley (2001), who require archaeological evidence for multiple forms of symbolic behavior as an indicator that the ‘‘package’’ of symbolic behaviors was present (but see Henshilwood et al. 2001b for claims of an early ‘‘package’’). Wadley (2001) rejects the Howieson’s Poort Industry because she concludes that it does not display the rapid turnover of styles she anticipates in modern behavioral systems, and instead argues that the LSA assemblages of sites such as Rose Cottage Cave and Sibudu Cave reveal the origins of cultural modernity in southern Africa about 40,000 years ago. Thackeray (1992) has likewise argued that the Howieson’s Poort backed artifacts (or any MSA examples) do not constitute evidence of modern behavior as they lack the standardization seen in the later LSA assemblages and what standardization as occurs is due to the functional requirements of hafting. Wurz (1999) has countered this with a metrical analysis that demonstrates that the backed artifacts of the Howieson’s Poort are as standardized as that of the Wilton and with the addition that the focus on exotic raw materials during the Howieson’s Poort is superﬂuous to hafting requirements. There are several dilemmas associated with these vigorous debates about the origins of modern behavior in sub-Saharan Africa and the explanation of backed artifact proliferation in terms of the transition to cultural modernity. One potential weakness of such an explanation is that it is not applicable to later episodes of backed artifact proliferation in Africa, since all later industries are presumed to have been created by behaviorally modern humans. Indeed, the bold Foley and Lahr (1997, 2003) argument that modern behavior is associated with Mode 3 technologies would imply that such an explanation is not even applicable to the ﬁrst episode of backed artifact proliferation. Reference to the onset of behavioral modernity as an explanation must also be invalid in the Australian situation where modern humans colonized the continent, and where backed artifact proliferation occurs in the Holocene. Consequently, only one of the multiple episodes of backed artifact proliferation can be understood with this explanation, and it follows that other causes must exist for the emphasis of such a technological strategy by a hunter-gatherer group. Torrence (2002) is correct in arguing that there may be multiple reasons for hunter-gatherer groups to emphasize backed artifacts, and hence the cause of an industry such as the Howieson’s Poort may be different to the cause of later backed artifact

___________________________________________________________________________________

63

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

proliferations. While this argument has merit it is nevertheless puzzling to see archaeologists invoke a major evolutionary shift to explain an archaeological phenomenon on one occasion, being obliged to seek more mundane explanations for the same technological shift on other seemingly similar occasions. Avoiding explanations of cognitive capacity by substituting models of demographic intensiﬁcation (Shennan 2001) offers the theoretical advantage that population densities may increase and decrease through time and therefore provide a mechanism for multiple proliferation events. Furthermore, it is curious that archaeologists have on the one hand advocated that backed artifacts acted as symbols in prehistoric African societies, but have not identiﬁed the ﬁrst proliferation of those implements as the point at which symbolic behavior was initiated (e.g., Wadley 2001; Wurz 1999, 2002). Unless archaeologists adopt the argument that the earliest episode of backed artifact proliferation marks the onset of modern behaviors, at more than 300,000 years ago, we are left with a scenario in which backing was a strategy employed prior to modern behavior, and therefore modern behavior/cognition is not a necessary explanatory condition of implement backing. On the other hand, if the Middle Pleistocene examples of backed artifact proliferation are taken as an indication of cultural modernity in the absence of other evidence of symbolic activity, one of two conclusions must be reached. The ﬁrst is that backed artifact production might be taking place in contexts without symbolic storage, in which case the conclusion that modern symbolic behavior is evidenced is unnecessary. The second possible conclusion is that backed artifacts alone demonstrate modern human symboling. If this inference is advocated the question that is raised is how to explain the subsequent appearance/proliferation of art and other symbolic paraphernalia if those phenomena do not emerge with the onset of cultural modernity but appear substantially later. This discussion reveals the fundamental limitation of this explanation citing cognitive capacity. The shortcoming demonstrated above is that in invoking the ‘‘capacity’’ for symboling such an explanation has obvious constraints. Such an explanation can be employed on only one instance in any region without hypothesizing the displacement of behaviorally modern populations by non-symboling ones. Furthermore, since the development of a capacity does not require it to be used, this model does not adequately explain why in some situations the capacity was employed to make large numbers of backed artifacts while on other occasions the ability to construct symbols was directed towards other ends. If backed artifacts were symbols, the capacity for symbolic expression is a precondition of their production, but not a sufﬁcient explanation for high rates of backed artifact production at any particular time. In other words, this model can be used as an explanation of the appearance/existence of these implements but not of their numerical proliferation or decline through time and space. As we have discussed, this model, by itself, cannot explain either the proliferation of backed artifacts in Australia or the pattern of multiple proliferations that occurred in Africa. For that reason archaeologists should seek additional or alternative explanatory models for these phenomena.

___________________________________________________________________________________

64

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

Ethnicity and Social Interaction as an Explanation A longstanding explanatory model that has been applied to variation in lithic assemblages is that implement forms denote not merely the capacity for symbolic behavior but the use of those symbols as markers of group identity. This kind of model explains the proliferation of speciﬁc implement types in terms of the immigration of a group into a region, the spread of a cultural system, or change in the intensity and nature of cultural contact between groups. In Australia it has been models of diffusion of socioeconomic organization, with accompanying symbols, that have commonly been discussed. Speciﬁcally, the Holocene proliferation of regular implement forms, including backed artifacts and unifacial and bifacial points, has been linked by many authors to either migrations of people or diffusion of symbolic systems through cultural contacts (e.g., Evans and Jones 1997; McConvell 1996). A dominant form of this argument has conventionally been based on a supposed coincidence in the timing of the arrival of the dingo and the appearance of these types of implement (Bowdler and O’Connor 1991; Flood 1983: 195). For example, Bowdler (1981) has argued at length that backed artifacts were a symbol of new religious and economic systems introduced with the dingo during the mid-Holocene. However, although this idea has been popular with Australian archaeologists, the notion that a mid-Holocene culture contact event explains the proliferation of backed artifacts in Australia is doubtful. For instance, the demonstration that the introduction of the dingo, economic change, and the ﬁrst production of backed artifacts are not contemporary, and that backed artifacts were occasionally produced for millennia prior to a widespread proliferation, has removed the empirical basis for a hypothesis of midHolocene contact as a viable explanation (Hiscock 1994, 2001; Hiscock and Attenbrow 1998). A further cause for skepticism about such diffusionist explanations arises because the mere existence of contact with different cultural systems and new forms of material culture does not explain their adoption across Australia; theories of external introduction still require an explanation for the acceptance of traits by recipient groups. One mechanism hypothesized to trigger the acceptance and/or spread of symbols in Holocene Australia is the reorganization of social units. For instance, David and Lourandos (1998) have argued that the emergence of compartmentalized cultural landscapes created a context in which symbols that assisted groups in the negotiation of resource rights would have been advantageous, and backed artifacts were one of the objects employed in more intensive social interactions. Ethnic or social symbolic explanations for the intensive use of backed artifacts have been much more extensively employed in African archaeology. In considering the criteria necessary to identify discrete social entities in the archaeological record, Clark (1988) contended that differences between social entities will be expressed in the types of material culture which groups carry and distribute. He advocated the view that distinctive styles embodying social identity would be applied to the items common to different groups. This view, long held in European archaeology, has also been explored in archaeological investigations of backed artifacts in Africa over

___________________________________________________________________________________

65

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

the last 30 years. For example, Singer and Wymer (1982: 107–9, 209–10) argued that the Howieson’s Poort, with its backed artifacts of non-local raw materials, was so unlike the preceding (MSA 2) and succeeding phases (MSA 3) of the MSA that it most likely represented the tool kit of an incoming or invading group whom they call the ‘‘Howieson’s Poort people’’ who, having occupied Klasies River Mouth for a relatively short period, moved on, taking their tool kit with them. The Indigenous population then reverted to producing the types of tools they had made during MSA 2. Singer and Wymer (1982: 107) explicitly rejected the notion that the technological emphasis on backing could have resulted as a response by the ‘‘Indigenous population . . . to a change of circumstances.’’ Less direct social explanations, such as those interpreting stylistic differences in terms of social communication and signaling, have become especially popular in the African literature since the Bleek and Lloyd historical accounts of !Kung were ‘‘rediscovered’’ in the 1980s. An example is Mazel’s (1989) argument that changes in the distribution of backed artifacts and steep scrapers across the Thukula Basin during the LSA from 9,000–2,000 bp reﬂect alterations in social regions and in the extent and complexity of regional alliance/exchange networks. Backed implements dominate the assemblages region wide in the early phase, beginning about 9,000 years ago and continuing until sometime after 5,000 bp, when three distinct regions emerge: those whose diagnostic component is dominated by backed segments, those dominated by backed points and blades but which also have segments, and those dominated by backed points and blades but which lack segments. The variations in assemblages through time and space were interpreted by Mazel as demarcating social regions. He proposes that through time a single social network disintegrated and separated into three distinct social regions and alliance networks. Mazel’s model invokes Clark’s proposition that backed artifacts function to mark material cultural boundaries and thus express and maintain difference between social entities. What is clear in Mazel’s argument is the detailed and extensive use of ethnographic images from historic Africa. Broadly similar models have also been advanced for the earlier proliferation events in southern Africa. For example, Deacon and Wurz (Deacon 1992; Deacon and Wurz 1996; Wurz 1999) have proposed that the backed artifacts of the MSA Howieson’s Poort assemblages were manufactured to be given as gifts in formal exchange systems and themselves functioned as symbols of mutual reliance and reciprocity in much the same way as the ethnographic projectiles produced and used by the San, discussed by Wiessner (1982, 1983). While such arguments are plausible in the historic and recent period, the basis for positing these social interactions in the distant past is insubstantial. Barham (1993) has argued that these models of social symbolic behavior are unjustiﬁed. In particular he argues that as expedient tools there is minimal capacity for these objects to store and convey social information. However, since the actual usage of these objects is unclear the conclusion that small expedient artifacts could not have performed social purposes should not be considered to have been established. A critique of such symbolic models can nevertheless be offered. To begin with, if they are to be comprehensive these models must also be capable of explaining the

___________________________________________________________________________________

66

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

decline in backed artifact production that has occurred in both continents. Hypotheses explaining backed artifact proliferation in terms of social interaction rarely suggest that subsequent declines indicate substantial decline in social connectivity, perhaps revealing a unilinear and progressivist image of cultural evolution. Symmetry in the logic of these models can be expected here; if the decline of backed artifact production is not indicative of reduced intensity of social networking, then the necessity of explaining backed artifact proliferation in these terms is diminished. The difﬁculty can be outlined as follows. If declining rate of backed artifact production reveals decreased social interaction or identity symboling, then the relevance of recent ethnographic accounts becomes a central question. On the other hand, if the nature and level of social information is undiminished when backed artifact production rates drop, the social value of high backed artifact production is called into question. These questions need to be explored if models of style are to be adopted. Added to these questions is an elementary issue of establishing the veracity of statements about ancient social interaction. The point about these models of social uses of symbolism in stone implements is the difﬁculty of providing independent tests of the hypothesized mechanisms for assemblage differences. In the absence of independent measures of group identity or the nature and intensity of social uses of symbolism these discussions have often relied only on untested assertions of the symbolic nature of archaeological implements.

Economics as an Explanation An alternative form of explanation is reference to the economic function of backed artifacts in foraging life. There are a large number of studies exploring this issue generally, and speciﬁc applications to both Australia and Africa. These explanations typically revolve around the technological responses by foragers to changing provisioning costs and risks. Costs are incurred because the rapid consumption of stone material creates a regular need to replenish the supply available to the individual or group, creating a provisioning cost. These provisioning needs may not only inﬂuence the structure of landscape use; they may also have involved a considerable expense in effort or required rescheduling of activities to enable resupply of materials (see Bamforth 1986; Hiscock 1986, in press; Jeske 1992; Torrence 1983, 1989a). In addition, the economic context of a foraging group involved the costs and returns for those tasks in which stone tools were involved. In response foragers employed technological strategies that lowered provisioning costs or facilitated greater returns. A distinctive form of economic explanation that has been employed to explain backed artifacts is the cultural response to ‘‘risk.’’ Archaeologists often think of risk as the probability and consequence of being unsuccessful in procuring resources (see Hiscock in press). For ancient foraging groups the magnitude of risk reﬂected the frequency or likelihood with which a group failed to procure adequate resources, representing the failure probability, and the consequences for the group of not obtaining adequate resources in an efﬁcient manner, representing

___________________________________________________________________________________

67

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

the failure cost (see Bamforth and Bleed 1997). The proposition advanced in risk models is that at least in some contexts ancient foraging groups may have adopted economic, social, and technological strategies that reduced the failure probabilities and costs (see Bamforth and Bleed 1997; Bousman 1993; Hiscock 1994; Kelly 1988; Nelson 1996; Torrence 1989b). Such models have been employed in both Africa and Australia for the emphasis of backed artifacts in tool kits, but with a different explanatory emphasis in each continent. In Australia the focus has been on models of technological responses to environmentally induced foraging risk. For example, Hiscock (1994) hypothesized that the movement of foragers into new landscapes created contexts in which resource availability was relatively unpredictable and foragers responded with an emphasis on reliable and maintainable multifunctional technologies, reﬂected archaeologically in the high production rates of standardized stone implements. His argument was that although backed artifacts could have had emblematic qualities their extensive use was probably caused by their capability as a technological device that reduced the level or consequence of economic risks. Subsequently, Hiscock (2002) argued that the onset of the ENSO controlled climate was one factor acting to increase foraging risk in the mid-Holocene, and that the increased production and use of backed artifacts, perhaps as part of a composite tool kit, was one response to the new context. More recently he has hypothesized that there were probably multiple environmental and social triggers to increased foraging risk, with the nature and timing of factors varying regionally and chronologically (Hiscock in press). These risk focused models typically cite the onset of heightened foraging risk as a key cause for the proliferation of backed artifacts and the reduction of foraging risk as allowing shifts to other technological strategies. In Australia less developed theories have proposed that social decisions may have been an integral part of both the construction of risk and the selection of responses to risk (see David and Lourandos 1998), although Hiscock (2002) has pointed out that environmental causes are embedded even in those theories. Dealing with Aboriginal occupation of the deserts of Western Australia, Gould (1977) inferred a multifaceted risk reduction system operating throughout the Holocene. He claimed technological strategies enhanced energy extraction from the landscape by permitting both high mobility and effective tool use, and that risk was additionally reduced through the maintenance of social networks displayed archaeologically in the circulation of raw material from distance sources. Gould did not explore possible connections between these risk minimizing strategies and the proliferation of backed artifacts in the Australian deserts, but explanatory models attempting to do that have been offered outside Australia. In the archaeological discussions of prehistoric sub-Saharan Africa, foraging risk has also been considered from a variety of perspectives ranging from ecological cost-beneﬁt models to those that emphasize social relations. For instance, Ambrose and Lorenz (1990: 25) view the assemblage changes from MSA 2 to Howieson’s Poort, and from Howieson’s Poort to MSA 3, as ‘‘a technological response to shifting patterns of raw material resource procurement’’ resulting from changes in territorial organization; most probably expansion and contraction of foraging

___________________________________________________________________________________

68

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

range. They argue that a change from warmer moister conditions, such as those that prevailed during MSA 1 and MSA 2, towards a cooler climate at the end of the last interglacial, would have resulted in more open environments or a mosaic of open and closed habitats which would have reduced the exploitable game biomass and thus ‘‘the spatio-temporal abundance and predictability of faunal and ﬂoral resources’’ (Ambrose and Lorenz 1990: 25). In response to the reduced game biomass and increased risk, people enlarged their territorial range, thereby gaining access to better quality raw materials, thus allowing them to make more ﬁnely retouched tools. Ambrose and Lorenz (1990) dismiss the possibility that exotic ﬁne-grained raw materials were acquired in order to produce backed artifacts because they argue that the archaeological record indicates that abrupt changes in lithic technology are imposed on more gradual changes in raw material, whereas the reverse would be expected if ﬁne-grained raw materials were acquired in order to produce hafted backed blades. There are several problems with the Ambrose and Lorenz model. Firstly, the Howieson’s Poort faunal assemblages provide only partial support for the model in that, while they contain small numbers of large grazing open grassland species, they are dominated by small to medium-sized non-gregarious species that prefer closed habitats. Secondly, the model fails to explain the relationship between climate/environment change, territorial expansion, and the proliferation of backed artifacts at this time, except to the extent that because better raw material was available people used it to make more backed artifacts. Where recent evidence has been presented backing technologies appear to be associated with a variety of environments and not explicable in terms of the existence of a single niche or environmental change (e.g., Cornelissen 2002). Thirdly, and perhaps most importantly, as Ambrose and Lorenz (1990) themselves acknowledge, the model fails to account for the demise of backed artifacts in the MSA 3. During this time conditions became colder and drier, and more open vegetation communities prevail. The model predicts that this ecological scenario should result in higher territorial and logistic mobility and therefore an increase in exotic raw material and backed artifacts. While the MSA 3 faunal assemblages do indicate an increasing reliance on gregarious nomadic ungulates, the lithic assemblage of the MSA 3 runs entirely counter to the prediction, with no or few backed artifacts and a signiﬁcant decline in the use of exotic material. Rather than question the relevance of the modern huntergatherer analogs they use to derive the model to the MSA, Ambrose and Lorenz interpret the non-compliance of MSA hunters with the predictive model (derived from modern hunter-gatherer behaviors) as proof of their lesser cognitive capabilities compared with their LSA counterparts or later modern humans. Recent and inﬂuential models propose that the solution to unpredictable and unproductive environments was not an increase in foraging range or an economically beneﬁcial technology, but a socially advantageous one. For instance, Deacon and Wurz (1996), while concurring with Clark (1992) that the backed blades of the Howieson’s Poort functioned as projectiles, have speciﬁcally argued against the use of the backed artifacts in composite tools to increase hunting efﬁciency. They follow Klein (2000) in proposing that any differences in prey species hunted during

___________________________________________________________________________________

69

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

this phase of the MSA can be ‘‘linked to ecological factors rather than the effectiveness of projectile equipment’’ (Deacon and Wurz 1996). Instead, they (Deacon 1992, 1995; Deacon and Wurz 1996) advance the proposition that backed artifacts in the Howieson’s Poort were symbolic items that were central to delayed reciprocal exchange systems that evolved into the ethnographically documented San hxaro gift system. The production of backed artifacts on non-local stone materials is seen as having no technological or practical advantage given backed artifacts were also made on local coarser materials; but as ‘‘value adding’’ to their use as exchange items. By exchanging ornaments the foragers are inferred to have promoted social solidarity that maintained these social networks during periods of heightened environmental risk, and that the continuation of these social interactions facilitated resource sharing, thereby reducing the consequences of resource failure for foraging groups. Wurz (1999), in a slightly different slant on this position, has demonstrated (contra Thackeray 1992: 423) that the MSA backed artifacts are as standardized as those of the LSA and argued that the degree of standardization is independent of functional requirements or the form/size of the raw material, but is due rather to social stylistic conventions intrinsically linked to symboling. Deacon (1992, 1995; Deacon and Wurz 1996), Ambrose (2002), and Wurz (1999) have hypothesized that such reciprocity-based risk minimizing systems, literally symbolized by backed artifact manufacture, were one of the key innovations in the evolution of modern human behavior and a contributing factor in the success of our species at the expense of other hominids. These ethnographically derived social models fail to explain variability seen in the archaeological record. Why, for example, during the most arid phase of the LSA age, the Last Glacial Maximum (LGM), when social reciprocity and reliance networks should have been most highly developed, is the Robberg Industry almost totally devoid of backed artifacts or indeed any standardized retouched tool form? Although citing foraging risk as the context in which these innovations emerge, such models are less about economic responses to environmental circumstances than they are economically situated variants of models emphasizing ethnicity and social interaction. For instance, by interpreting backed artifacts as functioning symbolically and identifying symbol use as a distinctively modern behavioral trait, archaeologists like Ambrose (2002: 21–2) have explicitly viewed backed artifacts as an archaeological marker of the movement and dispersal of behaviorally modern populations. Other researchers have seen long-term cultural continuity in southern Africa in the inferred emergence of the hxaro system from MSA occupation. These models raise the same questions (discussed above) of other symbolic interpretations of these stone artifacts: Why are images of behavioral modernity applied to one backed artifact proliferation event but not to the earliest one, and how do models of continuous symbolic activity explain declines in the production of, or the disappearance of, those symbols? By linking assemblage variation to changing economic contexts a mechanism for explaining increases and decreases in backed artifact production rates is created, but if backed artifacts are primarily symbols the interpretation of those changes in social terms requires further exploration.

___________________________________________________________________________________

70

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

Research Programs as Reﬂections of Desert Images We have described and evaluated a number of the models offered to explain the chronological and spatial changes in backed artifacts in Australia and Africa. Although we could have chosen a number of other traits we have focused on those which illustrate that the emphasis given to archaeological explanations has varied between these continents. Since these two regions are both dominated by arid and semi-arid lands, characterized by rich ethnographic records, and studied by schools of archaeological thought with strong historical links to the British paleoeconomy approach, we argue that these contrasts are worth considering further. Speciﬁcally, we suggest that these differences exemplify distinctive research programs (pace Lakatos 1978) that have built on divergent images of archaeological material and different images of deserts and human life in them, largely derived from historical and modern ethnographic sources. Perhaps most inﬂuential and formative were the images of African desert dwellers arising from the Man the Hunter conference held in 1966. Its ethnographic studies of the Kalahari !Kung San groups, emphasizing the afﬂuence of their subsistence base and the abundance of their leisure time (at least for the males), fostered a view of deserts as relatively benign landscapes for hunter-gatherers. In particular, the abundance of the wondrous, tasty, and highly nutritious plant food staple, the mongongo nut (Schinziophyton rautanenii) (Lee 1968, 1979), led to characterizations of the !Kung as enjoying ‘‘a kind of material plenty’’ (Marshall 1961: 243, as cited in Sahlins 1974: 9), a theme taken up and developed by Sahlins (1974: 1–39) in his essay on hunter-gatherer material existence entitled ‘‘The original afﬂuent society.’’ While such generalized forager models stressed notions of egalitarianism, neoMarxist anthropologists emphasized the way in which the social structures of hunter-gatherer groups were reinforced by delayed return systems such as gift exchange which acted to ensure reciprocal access to territory, acted as buffers against subsistence failure and thereby provided ‘‘each individual !Kung with reciprocal and guaranteed access to the material conditions of production’’ (Keenan 1981: 18). These images of African deserts as arid paradises have been powerful inﬂuences on archaeological thinking. As mentioned above, recent archaeological models for the context of backed artifact production in sub-Saharan Africa have been inspired by historic ethnographic images (e.g., Bleek and Lloyd n.d.), but have emphasized the structural rather than materialist position of social relations. Although they have stimulated archaeological analogies it is difﬁcult to see why these images of historical Kalahari Desert life have any relevance to the context under which backing took place in the MSA or for that matter most of the LSA in sub-Saharan Africa. Cultural continuity from MSA or LSA to historic cultures has not been demonstrated, and evidence independent of ethnographic analogy for the moderation of environmental risk by reciprocal social relationships is a moot point. Even the claim that backed artifact production is a social response to lower and less predictable environmental productivity is not clearly established in the southern African case. Indeed, simple

___________________________________________________________________________________

71

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r

functional associations between backed artifacts and a single environment or environmental productivity seem unfounded. For example, there is no strong association between production of backed artifacts and aridity. The archaeological assemblages indicate that when backed artifacts made of exotic material were ﬁrst produced in large numbers, hunter-gatherers were living in close proximity to the coast in relatively warm humid conditions, and exploiting the resources of predominantly closed vegetation communities. There is little evidence to suggest these environments were extremely unproductive. As noted earlier, backed artifacts are also absent from the Robberg lithic assemblages made during the heightened arid conditions of the LGM by LSA hunter-gatherers. This suggests that citing historic environmental conditions and cultural responses will not be an effective framework for exploring MSA patterns. Similar complexities exist in explaining patterns of Australian backed artifact production in simple environmental terms. Ethnoarchaeological studies of arid zone foragers in Australia, also carried out in the 1960s to 1980s, created an image diametrically opposite to the one developed in Africa. Australian researchers emphasized that desert life could be harsh and precarious and that social behavior and subsistence decisions were dominated by a variety of stress factors, primary among them being drought. Studies of Australian foragers typically adopted this image and described desert lifeways as marginal and conservative, stressing economic necessity rather than social advantages as the stimulus for reciprocity. These images of Australian deserts as dangerous landscapes have been powerful inﬂuences generally on archaeological models of Sahul. The dominant proponent of the dangerous desert explanation for prehistoric Australia, Richard Gould, argued these strategies were reﬂected in the long-term archaeological sequences from the Western Desert demonstrating conservative unchanging technologies, and emphasizing risk-minimizing economic strategies throughout the Holocene. Following this lead some subsistence models for Pleistocene hunter-gatherers assumed that desert dwelling was a late adaptation only possible during the Holocene and that earlier foragers would have avoided arid inland areas (Bowdler 1977), or that the arid center of Australia may have initially sustained human life but must have been abandoned with onset of aridity during the late Pleistocene (e.g., Horton 1981). Models of Holocene arid zone settlement and technology as consequences of harsh and unpredictable environments also reﬂect the image of Australian deserts as dangerous landscapes (Hughes and Hiscock in press; Ross et al. 1992; Veth 1995). Some recent research into the archaeology of Australian deserts has argued against inferences of conservative technologies and uniform economic systems (Hiscock and Veth 1991; Hiscock 1994; O’Connor and Veth 1996; O’Connor et al. 1998; Veth 1993), and has instead argued that both archaeological and historical evidence suggests that diverse and ﬂexible economic and social systems were employed. These revisionist models of Australian desert culture and cultural evolution were based more on archaeological patterns than ethnohistoric images, but most explanations for assemblage change in Australia remain based on ethnohistorical images whose relevance to the past is questionable.

___________________________________________________________________________________

72

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s

Having characterized the desert images embedded in archaeological models from each continent in this way it is necessary to pose the obvious question: Are the dissimilar pictures of African deserts of abundance and Australian deserts of danger reﬂections of genuine environmental differences? Without doubt the environments found on the two continents are different. As Gould (1991: 16) points out, differences in physiography and rainfall predictability in the two continents underlie divergent images of their deserts. For example, the Kalahari has some coordinated drainage networks and catchments, whereas the Western Desert of Australia has none and all water supplies depend entirely upon rain capture in impermanent surface and sub-surface catchments. More importantly, rainfall in the Western Desert is not only lower than that for the Kalahari, but also extremely unpredictable. Even the central deserts of Australia, which have limited catchments, display annual rainfall patterns that are less predictable than the Kalahari. There may therefore be an empirical basis for the images of deserts generated on each continent, although the complexity of socially mediated economic strategies makes any simple equation of environmental characteristics and social or economic hazard difﬁcult. However, if such an argument is sustained, and the Australian deserts are and have been inherently more difﬁcult for foragers, archaeologists will be compelled to return to the question with which we began this chapter, namely, why does the archaeological record in each continent display parallel patterns of backed artifact proliferation and decline? Comprehending in more detail the environmental and social contexts of technological change in each continent, and testing models of the prehistoric past against the archaeological rather than modern cultural evidence, will be one path forward in exploring not only the differences in ancient desert life between Australia and southern Africa but also the similarities. This quest should form one agenda of the research programs on each continent, and as we have demonstrated in this chapter archaeological model building will beneﬁt by moving away from local scales of explanation to a broad intercontinental comparative approach.

References Ambrose, S. H. 2001: Paleolithic technology and human evolution. Science, 291, 1748–53. Ambrose, S. H. 2002: Small things remembered: Origins of early microlithic industries in sub-Saharan Africa. In S. Kuhn and R. Elston (eds), Thinking Small: Global Perspectives on Microlithization, Washington, DC: Anthropological Papers of the American Anthropological Association (AP3A) no. 12, American Anthropological Association, 9–29. Ambrose, S. H. and Lorenz, K. G. 1990: Social and ecological models for the Middle Stone Age in southern Africa. In P. Mellers (ed.), The Emergence of Modern Humans, Edinburgh: Edinburgh University Press, 3–33. Bamforth, D. B. 1986: Technological efﬁciency and tool curation. American Antiquity, 51 (1), 38–50. Bamforth, D. B. and Bleed, P. 1997: Technology, ﬂaked stone technology, and risk. In C. M. Barton and G. A. Clark (eds), Rediscovering Darwin: Evolutionary Theory in Archaeology, Washington, DC: Archaeological Papers of the American Anthropological Association no. 7, American Anthropological Association, 109–40.

___________________________________________________________________________________

73

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r Barham, L. S. 1993: Comment: Reply to Wadley and Mazel. South African Archaeology Bulletin, 48, 51. Barham, L. S. 2001: Central Africa and the emergence of regional identity in the Middle Pleistocene. In L. S. Barham and K. Robson-Brown (eds), Human Roots: Africa and Asia in the Middle Pleistocene, Bristol: Western Academic and Specialist Press, 65–80. Barham, L. 2002: Backed tools in Middle Pleistocene central Africa and their evolutionary signiﬁcance. Journal of Human Evolution, 43, 585–603. Bleek, W. H. I. and Lloyd, L. n.d.: Unpublished notes. Capetown: Jagger Library, University of Cape Town and South African Public Library. Bousman, C. 1993: Hunter-gatherer adaptations, economic risk, and tool design. Lithic Technology, 18, 59–86. Bowdler, S. 1977: The coastal colonization of Australia. In J. Allen, J. Golson, and R. Jones (eds), Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia and Australia, London: Academic Press, 205–46. Bowdler, S. 1981: Hunters in the highlands: Aboriginal adaptations in the eastern Australian uplands. Archaeology in Oceania, 16, 99–111. Bowdler, S. and O’Connor, S. 1991: The dating of the Australian Small Tool Tradition, with new evidence from the Kimberley, W.A. Australian Aboriginal Studies, 1, 53–62. Clark, J. D. 1988: The Middle Stone Age in East Africa and the beginnings of regional identity. Journal of World Prehistory, 2, 235–305. Clark, J. D. 1992: African and Asian perspectives on the origins of modern humans. Philosophical Transactions of the Royal Society, B337, 201–15. Cornelissen, E. 2002: Human responses to changing environments in central Africa between 40,000 and 12,000 bp. Journal of World Prehistory, 16, 197–235. David, B. and Lourandos, H. 1998: Rock art and socio-demography in northeast Australian prehistory. World Archaeology, 30, 193–219. Deacon, H. J. 1992: Southern Africa and modern human origins. Philosophical Transactions of the Royal Society, B337, 177–83. Deacon, H. J. 1995: An unsolved mystery at the Howieson’s Poort name site. South African Archaeology Bulletin, 50, 110–20. Deacon, H. J. and Wurz, S. 1996: Klasies River main site, cave 2: A Howieson’s Poort occurrence. In G. Pwiti and R. Soper (eds), Aspects of African Archaeology: Papers from the 10th Congress of the PanAfrican Association for Prehistory and Related Studies, Harare: University of Zimbabwe Publications, 213–18. Evans, N. and Jones, R. 1997: The cradle of the Pama-Nyungans: Archaeologic and linguistic speculations. In N. Evans and P. McConvell (eds), Linguistics and Archaeology: Aboriginal Australia in Global Perspective, Oxford: Oxford University Press, 385–417. Flood, J. 1983: Archaeology of the Dreamtime. Sydney: Collins. Foley, R. and Lahr, M. 1997: Mode 3 technologies and the evolution of modern humans. Cambridge Archaeological Journal, 7, 3–36. Foley, R. and Lahr, M. 2003: On stony ground: Lithic technology, human evolution, and the emergence of culture. Evolutionary Anthropology, 12, 109–22. Gould, R. A. 1977: Puntutjarpa Rockshelter and the Australian Desert Culture. New York: Anthropological Papers of the American Museum of Natural History no. 54, American Museum of Natural History. Gould, R. A. 1991: Arid land foraging as seen from Australia: Adaptive models and behavioral realities. Oceania, 62, 12–33.

___________________________________________________________________________________

74

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s Henshilwood, C. S., d’Errico, F., Marean, C. W., Milo, R. G., and Yates, R. 2001a: An early bone tool industry from the Middle Stone Age at Blombos Cave, South Africa: Implications for the origins of modern human behavior, symbolism, and language. Journal of Human Evolution, 41, 631–78. Henshilwood, C. S., Sealy, J. C., Yates, R., Cruz-Uribe, K., Goldberg, P., Grine, F. E., Klein, R. G., Poggenpoel, C., van Niekerk, K., and Watts, I. 2001b: Blombos Cave, Southern Cape, South Africa: Preliminary report on the 1992–1999 excavations of the Middle Stone Age levels. Journal of Archaeological Science, 28, 421–48. Henshilwood, C. S., d’Errico, F., Yates, R., Jacobs, Z., Tribolo, C., Duller, G. A., Mercier, N., Sealy, J. C., Valladas, H., Watts, I., and Wintle, A. G. 2002: Emergence of modern human behavior: Middle Stone Age engravings from South Africa. Science, 295, 1278–80. Hiscock, P. 1986: Technological change in the Hunter River Valley and the interpretation of late Holocene change in Australia. Archaeology in Oceania, 21 (1), 40–50. Hiscock, P. 1994: Technological responses to risk in Holocene Australia. Journal of World Prehistory, 8 (3), 267–92. Hiscock, P. 2001: Sizing up prehistory: Sample size and composition of artifact assemblages. Australian Aboriginal Studies, 1, 48–62. Hiscock, P. 2002: Pattern and context in the Holocene proliferation of backed artifacts in Australia. In S. Kuhn and R. Elston (eds), Thinking Small: Global Perspectives on Microlithization, Washington, DC: Anthropological Papers of the American Anthropological Association (AP3A) no. 12, American Anthropological Association, 163–77. Hiscock, P. in press: Blunt and to the point: Changing technological strategies in Holocene Australia. In I. Lilley (ed.), Archaeology in Oceania: Australia and the Paciﬁc Islands. Oxford: Blackwell. Hiscock, P. and Attenbrow, V. 1998: Early Holocene backed artifacts from Australia. Archaeology in Oceania, 33, 49–63. Hiscock, P. and Veth, P. 1991: Change in the Australian Desert Culture: A reanalysis of tulas from Puntutjarpa. World Archaeology, 22 (3), 332–45. Horton, D. R. 1981: Water and woodland: The peopling of Australia. Australian Institute of Aboriginal Studies Newsletter, 16, 21–7. Hughes, P. and Hiscock, P. in press: The archaeology of the Lake Eyre South Area. In Aboriginal Occupation of the Lake Eyre South Region: Ethnography and Archaeology, Adelaide: Lake Eyre Monograph Series vol. 6, Royal Geographical Society of South Australia. Jeske, R. 1992: Energetic efﬁciency and lithic technology. American Antiquity, 57, 467–81. Keenan, J. 1981: The concept of the mode of production in hunter-gatherer societies. In J. S. Kahn and J. R. Llobera (eds), The Anthropology of Pre-Capitalist Societies, London: Macmillan Press, 2–21. Kelly, R. L. 1988: The three sides of a biface. American Antiquity, 53, 717–34. Klein, R. G. 2000: Archeology and the evolution of human behavior. Evolutionary Anthropology, 17, 18–36. Lakatos, I. 1978: Falsiﬁcation and the methodology of scientiﬁc research programmes. In I. Lakatos and A. Musgrave (eds), Criticism and the Growth of Knowledge, Cambridge: Cambridge University Press, 91–196. Lee, R. B 1968: What hunters do for a living, or, how to make out on scarce resources. In R. B. Lee and I. De Vore (eds), Man the Hunter, New York: Aldine, 30–48. Lee, R. B. 1979: The !Kung San. Cambridge: Cambridge University Press.

___________________________________________________________________________________

75

_______________________________________________________________________________

Pe t e r H i s c o c k a n d S ue O’ C o n n o r McBrearty, S. and Brooks, A. S. 2000: The revolution that wasn’t: A new interpretation of the origin of modern human behavior. Journal of Human Evolution, 39, 453–563. McConvell, P. 1996: Backtracking to Babel: The chronology of Pama-Nyungan expansion in Australia. Archaeology in Oceania, 31, 125–44. Mazel, A. D. 1989: Changing social relations in the Thukela Basin, Natal 7,000–2,000 bp. Southern African Archaeological Society Goodwin Series, 6, 33–41. Mitchell, P. J. 1988: The late Pleistocene early microlithic assemblages of southern Africa. World Archaeology, 20, 27–39. Mitchell, P. J. 1995: Stories in stone: A review of south African lithic research. Lithics Studies Society Occasional Paper, 5, 71–88. Nelson, M. 1996: Technological strategies responsive to subsistence stress. In J. Tainter and B. Tainter (eds), Evolving Complexity and Environmental Risk in the Prehistoric Southwest, New York: Addison-Wesley, 107–44. O’Connor, S. and Veth, P. 1996: A preliminary report on recent archaeological research in the semi-arid/arid belt of Western Australia. Australian Aboriginal Studies, 2, 42–50. O’Connor, S., Veth, P., and Campbell, C. 1998: Serpent’s Glen rock shelter: A Pleistoceneaged archaeological sequence from the Western Desert. Australian Archaeology, 46, 12–21. Ross, A., Donnelly, T., and Wasson, R. 1992: The peopling of the arid zone: Human– environment interactions. In J. Dodson (ed.), The Naive Lands, Melbourne: Longman Cheshire, 76–114. Sahlins, M. 1974: Stone Age Economics. London: Tavistock. Shennan, S. 2001: Demography and cultural innovation: A model and its implications for the emergence of modern human culture. Cambridge Archaeological Journal, 11, 5–16. Singer, R. and Wymer, J. 1982: The Middle Stone Age at Klasies River Mouth in South Africa. Chicago: University of Chicago Press. Thackeray, A. I. 1992: The Middle Stone Age south of the Limpopo. Journal of World Prehistory, 6, 385–440. Torrence, R. 1983: Time budgeting and hunter-gatherer technology. In G. Bailey (ed.), Hunter-Gatherer Economy in Prehistory, Cambridge: Cambridge University Press, 11–22. Torrence, R. 1989a: Tools as optimal solutions. In R. Torrence (ed.), Time, Energy, and Stone Tools, Cambridge: Cambridge University Press, 1–6. Torrence, R. 1989b: Retooling: Towards a behavioral theory of stone tools. In R. Torrence (ed.), Time, Energy, and Stone Tools, Cambridge: Cambridge University Press, 57–66. Torrence, R. 2002: Thinking big about small tools. In S. Kuhn and R. Elston (eds), Thinking Small: Global Perspectives on Microlithization, Washington, DC: Anthropological Papers of the American Anthropological Association (AP3A) no. 12, American Anthropological Association, 179–89. Veth, P. M. 1993: Islands in the Interior: The Dynamics of Prehistoric Adaptations Within the Arid Zone of Australia. International Monographs in Prehistory, Archaeology Series no. 3, Ann Arbor, MI. Veth, P. 1995: Marginal returns and fringe beneﬁts: Characterizing the prehistory of the lowland deserts of Australia. Australian Archaeology, 40, 32–8. Volman, T. P. 1984: Early prehistory in southern Africa. In R. G. Klein (ed.), Southern African Prehistory and Paleoenvironments, Rotterdam: A. A. Balkema, 169–220. Wadley, L. 2001: What is cultural modernity? A general view and a South African perspective from Rose Cottage Cave. Cambridge Archaeological Journal, 11, 201–21.

___________________________________________________________________________________

76

_______________________________________________________________________________

Ar i d P a r a d i s e s o r D a n g e r o u s L a n d s c a p e s Wiessner, P. 1982: Risk reciprocity and social inﬂuences on !Kung San economics. In E. Leacock and R. B. Lee (eds), Politics and History in Band Societies, Cambridge: Cambridge University Press, 61–94. Wiessner, P. 1983: Style and social information in Kalahari San projectile points. American Antiquity, 48, 253–76. Wurz, S. 1999: The Howieson’s Poort at Klasies River – an argument for symbolic behavior. South African Archaeological Bulletin, 54, 38–50. Wurz, S. 2002: Variability in the Middle Stone Age lithic sequence, 115,000–60,000 years ago at Klasies River, South Africa. Journal of Archaeological Science, 29, 1001–15.

___________________________________________________________________________________

77

_______________________________________________________________________________

----------------------------------------------------------------

Part II

--------------------------------------------------------------

Dynamics

The chapters in part two give an archaeological perspective on the long-term dynamics of desert societies. Both the regional and evidential focus of these chapters is varied, but each examines chronological change in human activities within deserts over a period of at least ten millennia. Douglas Bird and Rebecca Bliege Bird (Chapter 5) contrast the ecological strategies of North American and Australian hunter-gatherers. Pursuit of sophisticated understandings of life in arid lands is reﬂected in the modeling of forager response to environmental and social variability through time and space offered by both Bird and Bliege Bird (Chapter 5) and Veth (Chapter 6). Peter Veth’s chapter focuses on the key concepts of risk and marginality in desert life. Jo McDonald (Chapter 7) characterizes the nature and variability of rock art in the Australian deserts to explore the role of art and symbols in the adaptation of foragers to deserts. Luis Borrero (Chapter 8) examines Patagonian archaeology as an exemplar of hunter-gatherer adaptations to cold deserts. One of the signiﬁcant outcomes of these diverse investigations is the demonstration of the mechanisms by which humans have adapted to changing desert landscapes. For instance, each of the chapters in part two presents detailed arguments about how technological, foraging, and social strategies were employed by human groups to enhance their lives in desert environments. Each chapter also describes the dynamic nature of these desert adaptations, with regionalization and reorganization of social and subsistence patterns being widely observable in the archaeological record. Importantly, these chapters also reveal the complicated and historically contingent processes of cultural evolution in these desert landscapes. In particular, there are clear examples of the fact that change is not inevitably directional. Additionally, each of the chapters provides unambiguous evidence that ancient life in deserts was in some ways distinctively different from that observed in historic occupation of arid lands, reinforcing at a global level the inference that historically observed desert societies and adaptations are relatively

___________________________________________________________________________________

79

_______________________________________________________________________________

D yn amic s

recent consequences of evolving systems. These chapters build on that understanding to demonstrate how awareness of variability in desert life, and of the intricacy of employing ethnographic images of that life in archaeological reconstructions, can enhance our study of the long-term dynamics of human life in arid lands.

___________________________________________________________________________________

80

_______________________________________________________________________________

-------------------------------------------------------------------------------

5

----------------------------------------------------------------------------

Evolutionary and Ecological Understandings of the Economics of Desert Societies Comparing the Great Basin USA and the Australian Deserts Douglas W. Bird and Rebecca Bliege Bird

Introduction In this chapter we provide a comparative synthesis of evolutionary and ecological analyses of economic life among Indigenous peoples in the Great Basin of North America and the central deserts of Australia. We review the development of research in human–environmental dynamics and look at current approaches in these regions that draw on theory in ‘‘evolutionary ecology.’’ In keeping with a long tradition of research in both areas, we focus primarily on introducing readers to factors that inﬂuence variability in subsistence strategies, especially those related to sex-linked differences in hunting and gathering. The goal is to provide students and non-specialists with a broad understanding of the issues regarding subsistence strategies in two desert regions of the world that are very similar in some respects and very different in others.

Background Discussion The central theme of this chapter is that the economic lives of desert huntergatherers are arranged around variability in sex-linked foraging strategies. Twenty years ago, Thomas (1983: 439), an archaeologist working the Great Basin, wrote that despite their central economic role, ‘‘Female hunter-gatherers are almost invisible archaeologically.’’ While reﬂecting on this, James O’Connell (2003: pers. comm.), an anthropologist who has worked for many years in Australia and the Great Basin, commented: ‘‘If women are archaeologically invisible in the Great

___________________________________________________________________________________

81

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

Basin and Desert Australia, it is only because their economic presence is so large that we can’t see the forest through the trees.’’ The arid Great Basin of North America and the Western and Central Deserts of Australia provide a comparative setting for testing hypotheses about variability in this key component of desert hunter-gatherer socioeconomy (see Figures 5.1 and 5.2). Anthropologists have often noted a number of parallels among the Indigenous inhabitants of these two regions. One of the most obvious is the presence of a traditional subsistence economy based on a broad spectrum of resources acquired mostly by women. Ethnographic and archaeological studies that illustrate this pattern include Beck (1999), Cane (1987), Elston and Zeanah (2002), Fowler and Liljeblad (1986), Fowler and Fowler (1971), Gould (1969, 1991), Gould et al. (1972), Grayson (1993), I. Kelly (1932), R. Kelly (1995), Liljeblad and Fowler (1986), Meggitt (1957, 1964), O’Connell and Hawkes (1981, 1984), O’Connell et al. (1983), Peterson (1977, 1978), Service (1962), Steward (1933, 1936, 1938,

Hydrographic boundary Language boundary 200 km

old

r

Rive

Shoshone

alt

ts

L.

Western Shoshone

Washoe

Goshute

da M

eva

Wa

ra N

sat

Sier

San Francisco

tS

t

ea

mb

Gr

Hu

Mn

Northern Paiute

ch

40 N

nts

Ute

Mono Owens Valley Paiute

Panamint

Pacific Ocean

Southern Paiute

Colorado Plateau

Kawaiisu

N 120 W

Figure 5.1 The hydrographic Great Basin in North America indicating languages and groups discussed in the text.

___________________________________________________________________________________

82

_______________________________________________________________________________

The Economics of Desert Societies

Hydrographic boundary Language boundary

Alawa

Kimberley Plateau

400 km Karajarri Walpiri Gugadja

20 S

Alyawara

R. Pilbara Plateau

Ru

Central Desert

Pintupi

ll da

L. Mackay

Mardu

Ngatatjara

Macdonell Ranges

Luritja

L. Disappointment

Pitjantjara

Western Desert

S

d an

ov

e r R.

Arrernte Fi

nk

e

R. Lake Eyre Basin

L. Carnegie

Nyanganyatjara Kokatha

Indian Ocean Perth

N 130 E

Figure 5.2 The hydrographic Western and Central Deserts in Australia indicating languages and groups discussed in the text.

1941, 1955), Stewart (1941), Tonkinson (1991), Veth and Walsh (1988), and Walsh (1987, 1990). Studies such as these have also led to a greater appreciation of how ﬂexible socioeconomic life actually was, and in some cases still is, among foraging societies in these regions. In this chapter we explore the ecological underpinnings of such variability. At this point, a word of caution is in order. Comparing any two groups of people is problematic for at least two reasons. Firstly, by deﬁnition, we are analyzing different social contexts with different histories and environments. In comparing Great Basin and Australian subsistence strategies, we should not discount the inﬂuence of speciﬁc historical circumstances on social behavior. On the contrary, variability in these social and historical contexts is what we are trying to explain. Secondly, cross-cultural comparisons reﬂect our theoretical positions and the way we pose questions. The Great Basin and arid Australia invite comparison partly because they both have relatively long histories of research focused on subsistence

___________________________________________________________________________________

83

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

from a common theoretical approach. We need to keep in mind that this can be a potential source of bias. However, as long as we state our assumptions and methods clearly, working from a single theoretical approach (here, evolutionary ecology) can greatly facilitate comparative analyses.

Adaptation: A Key Concept Ecological questions about subsistence variability within groups, between groups, and through time in the Great Basin and arid Australia have often been framed by the concept of adaptation. Steward’s (1938) ethnographic synthesis of Numic speaking groups in the Great Basin was especially inﬂuential, and one of the ﬁrst to explicitly apply an adaptive model to hunter-gatherer social groupings (cf. Fowler 1986). He was particularly interested in moving away from attempts to deﬁne large culture areas to focus on the actual behavior of people. How do desert hunter-gatherers maintain social organization without large scale centralized institutions? Steward argued that in the Great Basin, ethnographic patterns of low population density, ﬂexible local organization and mobility, and a highly portable technology, were closely articulated with speciﬁc features of the natural environment, which in turn structured the nature of subsistence and kin-based social organization. Due to the scarcity of large game (the focus of men’s subsistence) in the Great Basin, people relied extensively on dependable wild seeds and roots collected by women. Groups of many families congregated when certain resources were abundant enough at particular times and places. But these larger groups were ephemeral and would soon disperse into small family units to wander across overlapping ranges exploiting a wide array of plants and small game. Thus, for Steward (1955), the ‘‘family’’ (a ‘‘patrilineal band’’) was a basic adaptive unit, organized around sex-linked differences in foraging and extensive food sharing. Steward (1938: 230–1) argued this sexual division of labor was an adaptation to efﬁciently provide food for the patrilineal band. In the 1950s Jennings began to develop the argument that the pattern of ‘‘adaptation’’ described by Steward had deep temporal roots and a large geographic distribution in western North America. Commonly referred to as the ‘‘Desert Culture’’ or ‘‘Desert Archaic’’ (Jennings 1957, 1964, 1974), Jennings posited that a pattern of low population density, hunting and gathering based on annual resource cycles, and mobile band social organization lasted for over ten millennia in the Great Basin. He proposed that this Desert Culture was a regional variant of a continent wide culture stage based on a foraging economy closely adapted to local resource abundance and variability (Jennings 1974). Traditional approaches to Australian prehistory and ethnography have also placed a heavy emphasis on the inﬂuence of local resource ecology (cf. Lourandos 1997). Cultural variability, especially subsistence and mobility, has often been seen as an adaptation to particular environmental circumstances. Stanner (1965) considered the ecological basis of Aboriginal territorial organization in terms of cultural adaptation to varying resource abundance. This approach was developed

___________________________________________________________________________________

84

_______________________________________________________________________________

The Economics of Desert Societies

further by Peterson (1976), who argued that large adaptive units of Aboriginal society were scaled to major continental drainage basins. The approach is analogous to Jennings’ Desert Culture concept: similarities in basic forms of subsistence and technology were adaptations to the abundance and variable distribution of key resources across large areas. With similar logic, Gould (1982, 1991) argued that many aspects of Ngatatjara Aboriginal society were adapted to ensure success in unpredictable environments. He proposed that open-ended kin-based social organization, highly ﬂexible band composition, and group mobility were adaptations to minimize the risks associated with unpredictable droughts. Periods of drought required social organization and mobility strategies that allowed either intensiﬁed use of reliable water sources within the ‘‘home’’ area, or complete abandonment of an estate for an extended period of time. This would require an extensive network of classiﬁcatory kin and sharing. Gould hypothesized that if sharing is an adaptation to reduce temporal and spatial ﬂuctuations in important resources, food sharing should be especially important in highly variable environments, where it can build social bonds that can be relied on in times of need. Among the Ngatatjara where resources are highly unpredictable and the environment especially austere, sharing is extensive, as opposed to other groups where resources are more reliable and sharing is restricted to a much smaller social radius.

Critiques of the Use of ‘‘Adaptation’’ from Evolutionary Ecology What do the above studies mean by ‘‘adaptation’’? The general argument is that groups of people in the Great Basin and desert Australia adapted to local environments to ensure some measure of success. But success in terms of what (nutritional needs, social bonds, survival, reproduction) and for whom (the individual, the group)? Do the studies simply conclude there is often a correlation between large scale environmental constraints and patterns of behavior; or, by adaptation do they mean speciﬁc traits that are designed to ensure survival and reproduction? If the former, why does the correlation take on the observed pattern as opposed to some alternative? If the latter, what is the adaptive unit: the ‘‘individual,’’ the ‘‘family,’’ the local ‘‘band,’’ or the ‘‘society’’? Units above the level of the individual are very difﬁcult to deﬁne, especially in the Great Basin and Australia, where local organization is ﬂuid, and groups are made up of individuals with varying degrees of competing interests. So by adaptation do we mean ‘‘shaped by natural selection’’ or rather that there is some observed beneﬁt in a speciﬁc behavior? A pattern of behavior may have a beneﬁcial outcome, but this does not necessarily mean that the behavior was designed for that purpose. These issues have been signiﬁcantly clariﬁed over the last 30 years with the development of ‘‘evolutionary ecology’’ and its application toward understanding human diversity (cf. Winterhalder and Smith 2000). Evolutionary ecology is the study of adaptive design in particular ecological contexts. In the framework of evolutionary biology, behavior is adaptive when it tracks environmental variability in ways than enhance an individual’s inclusive

___________________________________________________________________________________

85

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

ﬁtness, most generally deﬁned as an organism’s propensity to survive and reproduce (Williams 1966). Behavioral ecology is the subset of evolutionary ecology that studies ﬁtness-related decisions that individuals face in particular social and physical environments (Krebs and Davies 1997). The working assumption is that an individual’s success at surviving and reproducing depends on its behavior; thus natural selection will favor variants with the capacity to solve ﬁtness related tradeoffs efﬁciently. This capacity may involve various kinds of decision making, based on information obtained through individual effort or social learning; behavioral ecology does not rest on the claim that behavior is genetically determined. The optimal outcome of these tradeoffs is shaped by the opportunities and constraints of local environments (Krebs and Davies 1993: 4–22). The focus of behavioral ecology is on individuals: selection generally creates adaptive design through differential reproductive success of individuals within a population. This is simply because individuals reproduce at higher rates and with higher ﬁdelity than groups. As a result, individual level selection tends to swamp the effects of group selection (Williams 1966). Thus (except under special circumstances) traits are adaptive only when they are designed to contribute to the inclusive ﬁtness of individuals, regardless of their effect on the group. No trait is absolutely adaptive; individuals continually face ﬁtness related tradeoffs whose parameters change with variable ecological circumstances. For people, those circumstances are often deﬁned by an individual’s social and cultural situations (Cronk 1991). ‘‘Human behavioral ecology’’ (HBE) applies evolutionary ecology concepts to study human behavioral diversity through systematic hypothesis testing (Smith and Winterhalder 1992). HBE does this through the use of formal modeling tools that analyze behavior in terms of a stated goal in relation to (1) a decision (ﬁtness related tradeoff) of interest; (2) a currency (e.g., energy) in which to analyze the various choices; and (3) the proposed constraints on an individual’s choices (the intrinsic and extrinsic factors that deﬁne the relationship between the decision and currency) (Stephens and Krebs 1986: 5–12). Formal models provide a systematic guide for investigating the interaction of a limited number of variables (justiﬁed on theoretical grounds) and their predicted outcome relative to observable behavior; they clearly isolate mismatches between hypothetical expectations and observations. As we will discuss below, HBE analyses of desert socioeconomic life in Australia and the Great Basin have developed in tandem over the last 25 years, and have made important contributions to the way we think about the adaptive nature of human–environmental dynamics.

Subsistence Strategies Many researchers have noted broad social similarities between Great Basin groups and Australian desert Aborigines. Gould et al. (1972) presented a detailed account of parallels in the economic and colonial histories of the two regions. They noted that environmental constraints of these arid and unpredictable environments are

___________________________________________________________________________________

86

_______________________________________________________________________________

The Economics of Desert Societies

linked to similar social responses: low population density; high residential mobility with highly portable technology; ﬂexible local organization, with ﬂuid dispersal and coalescence of groups of people; and open-ended, widely extended social organization based on classiﬁcatory kinship. These similarities are especially striking in the economy where foraging is based on sex-linked differences in subsistence activities.

Links between Sex, Foraging, and Sharing Prior to European contact, socioeconomic life in the Great Basin and the Australian deserts revolved around collecting seeds from grasses and trees, seasonal fruit, insects, and small game. In arid Australia at least 92 plant species are identiﬁed as edible (Latz 1995; O’Connell et al. 1983; Veth and Walsh 1988; Walsh 1987), while in the Great Basin over 150 plants have been catalogued as edible (Fowler 1972). In both regions, plants and small game were mainly supplied by women. Men focused on the (infrequently encountered) larger game, which often gave very low returns given the time required for pursuit (Bliege Bird and Bird in press; Gould 1969; Gould et al. 1972). In both regions, the resources acquired by women were usually directed to a small group of dependants. Conversely, men’s efforts were often focused on providing more public goods. In the Great Basin, especially south of the Humboldt River, annual subsistence and mobility revolved around harvesting pine nuts in the fall. These provided staple resources stored through the winter, when family groups in some areas would congregate in ‘‘villages’’ of up to 200 people. Villages would break up in the spring when smaller groups would disperse to focus on roots, grass seeds, and small game. North of the Humboldt River, more abundant roots and berries took the place of pine nuts which were not locally available (see Fowler and Liljeblad 1986; Thomas et al. 1986). In his classic ethnography of Great Basin Paiute and Shoshone people, Steward (1938) stressed the daily importance of women’s activities, and the more public nature of men’s subsistence labor: Seeds [for Owens Valley Paiute] were the most important food (52) . . . A woman gathered seeds only for her immediate family (59) . . . Migrations during the year depended upon local abundance of seeds (60) . . . Animal foods were of secondary importance . . . Game procured by individual [male] hunters was divided among all members of the village, the hunter keeping only the hind quarter for himself and family (66) . . . [Among western Shoshone] seeds gathered were private property. Women shared them only with their husbands, children, and sometimes parents whom they supported . . . Large game, on the other hand, was shared communally with all village members, whether they were related or not (74) . . . Relative scarcity of animals made meat a minor importance . . . Hunting was relatively of greater importance during seed shortage, but considerable reliance was placed on rodents (83) . . . Large game was of secondary importance (90) . . . [In the northern Great Basin] there are more roots and berries [harvested by women] in proportion to seeds (103–4) . . . Big game hunting was carried out by individual hunters in accessible mountains. Meat was distributed freely to one’s neighbors (115) . . . [Among Gosiute

___________________________________________________________________________________

87

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d Shoshone] The importance of rodents [for women] was considerable in this region (138) . . . Communal hunts [of antelope and rabbits] were organized by a local chief, but he received no more meat than other participants [which included men, women, and children] (163) . . . It was customary if not obligatory for a hunter to share large game with his neighbors. (184)

Similar accounts are common for Australian Aborigines. In the cool-dry season (April–August) grass seeds, roots, Solanum fruit, and goannas were daily staples. During the late spring and summer (September–March), if water supplies were replenished with isolated rains, larger groups would congregate for religious purposes. Acacia tree seeds and goannas made up much of the diet during this time (Bliege Bird and Bird in press; Cane 1987; Gould 1969, 1991; Tonkinson 1991; Walsh 1987, 1990). Gould et al. (1972: 266) noted: the Pitjantjatjara [Western Desert] peoples subsisted mainly on vegetable foods, including eight staple varieties of fruits, berries, and seeds as well as a multitude of supplemental or occasional varieties. Nearly all these foods were collected by women. Except for after a successful hunt (a fairly rare event), vegetable foods comprised ‘‘ . . . 70 or 80 percent of the total food supply’’ (Meggitt 1964: 33). Protein was supplied primary by various lizards (mainly goannas [Varanus spp.]) and small ground-dwelling marsupials . . . Most small game was collected by women; men concentrated on larger game.

Gould (1982) also observed that sharing patterns among Western Desert peoples were similar to those recorded by Steward in the Great Basin. The results of women’s efforts were directed toward a narrow range of dependants, whereas men focused on game that was shared widely beyond a single hearth group.

Flexible Hunting and Sharing Strategies It is important to recognize that in both regions, sex-linked differences in subsistence were ﬂexible with variable seasonal, environmental, and social contexts. For example, among contemporary Martu of Australia’s Western Desert, during the cool-dry season (May–August) women spend a great deal of their foraging time hunting burrowed goannas and collecting Solanum fruit in the sandplains and dunes. In contrast, men spend almost all their foraging time hunting mobile game (bustard, kangaroo, and emu) across a range of habitats (Bliege Bird and Bird in press). During this season women’s resources are ‘‘owned’’ by the acquirer and shared with a relatively small group of dependants. Conversely, if successful, male hunters make no claim to the animal – the game is usually given to an elder (often to a spouse’s father) who takes responsibility for wide distributions (Tonkinson 1991). Primary sharing of kangaroo and emu is often dictated by kinship categories, but thereafter the appropriate recipient shares freely with anyone who desires a portion. The pattern of Martu subsistence is notably different during the hottest part of the year (January–April). During this time, sand goanna (Varanus gouldii) and perenti (V. giganteus) are more commonly encountered on the surface, and women

___________________________________________________________________________________

88

_______________________________________________________________________________

The Economics of Desert Societies

and men sometimes coordinate their hunting activities. If men have been unsuccessful in hunting larger game, they sometimes shift strategies and search for game resources targeted by women, something they very rarely do at other times of the year (Bliege Bird and Bird in press). Large game is encountered less often during this season, and sharing patterns for men during the hot season tend to be more narrowly focused. Ethnographers have noted similar variability in Great Basin hunting. During the fall, especially in areas south of the Humboldt River, large groups of men, women, and children would cooperate in game drives for both antelope and rabbits. Brush structures were constructed in a long V-shape and ﬂanked by people. Large crowds would then drive the game into a corral (for antelope) or nets (for rabbits) at the far end of the drive. The game acquired was shared among all participating families, with the organizer receiving no more than anyone else (Steward 1938: 163). Throughout the rest of the year, men independently hunted deer and mountain sheep, which when successful, were shared widely with all families in the local group and their neighbors. Outside of communal drives, women’s game acquisition was focused on trapping and snaring small mammals (especially ground squirrels) for immediate family consumption. Game drives were less important north of the Humboldt River. There men focused on deer and mountain sheep, but in the fall, men and women would sometimes cooperate in trapping and snaring small game (groundhogs, ground squirrels, rabbits), ﬁshing, and netting waterfowl (Kelly 1932; Steward 1938; Stewart 1941).

Hypotheses about Subsistence Variability Human behavioral ecology in Australia and the Great Basin is to some extent an attempt to put previous ecological approaches (such as those of Steward and Gould discussed above) on a solid theoretical foundation; in a sense, to make their assumptions explicit and hypotheses testable. As we noted previously, Steward (1938, 1955, 1968) had argued that many of the socioeconomic similarities discussed above were adaptations designed to beneﬁt a narrowly deﬁned afﬁnal and consanguineal corporation (a man, his wives, and their offspring). These were selfsufﬁcient units: without cooperation, individuals would not survive, and larger cooperative groups were rarely advantageous or feasible. Steward (1938: 230–7) surmised this was determined by the abundance and distribution of the subsistence resources discussed above. The ‘‘family organization of the local band’’ was based in a sexual division of labor, which served as a mini-economy of scale, by which men and women specialized in different kinds of subsistence to more effectively provision their common interests. Cooperative groups larger than this were ephemeral because they would quickly depress the local resource base. ‘‘The functional basis of band organization, then, was the habitual cooperation of its members in joint enterprises and its objective expression was the common name, chieftainship, and ownership of territory’’ (Steward 1938: 51).

___________________________________________________________________________________

89

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

O’Connell, Hawkes, and colleagues were the ﬁrst to formally apply HBE to questions about subsistence in the Great Basin and desert Australia (Hawkes and O’Connell 1981; Madsen and O’Connell 1982; O’Connell and Hawkes 1981, 1984; O’Connell et al. 1982). Their analyses, along with the more recent work discussed below, have brought into question two essential assumptions of previous ecological work in these areas: 1 2

That there is a straightforward relationship between resource abundance and subsistence/mobility patterns. That local groups are units of adaptation (i.e., that labor is designed to serve the common interests of the members of the group).

HBE in Desert Australia Initial HBE investigations in Australia were based on O’Connell’s measurements of the efﬁciency of hunting and gathering among Arandic-speaking Alyawara in the Central Desert in 1974–5. On the assumption that natural selection has shaped decision-making capacities, and that efﬁcient foraging will enhance individual ﬁtness, O’Connell and Hawkes (1981, 1984) generated speciﬁc predictions about the range of activities and prey types that would maximize foraging returns relative to the options available. They used two formal models from behavioral ecology to predict foraging decisions: the ‘‘prey choice model’’ (PCM) and the ‘‘patch utilization model’’ (PUM). The PCM assumes that the goal of foraging is to maximize energy (the currency) per unit time spent foraging (searching and handling). The decision it analyzes is whether a forager should handle an encountered resource, or pass it over to search for other resources. The most important constraint of the PCM is that it assumes that resources are distributed randomly; more accurately, that an encounter with an item of one type does not change the probability of encountering other items. The model ranks resources according to their ‘‘proﬁtability’’ – the ratio of energy gained relative to its expected post-encounter handling time (pursuit, capture, and processing). It assumes that foragers will take an encountered item only if its proﬁtability is greater than that expected by searching for and handling more proﬁtable resources. The main predictions of the PCM are: . The decision as to whether a resource will be handled when it is encountered is determined by the abundance of the more proﬁtable resource, rather than the abundance of the resource encountered per se. . Lower encounter rates with the most proﬁtable resources serve to expand dietary breadth to include less proﬁtable resources (Stephens and Krebs 1986: 17–24). The PUM relaxes the random distribution constraint of the PCM to consider situations where resources are patchy in time or space. The model ranks patches

___________________________________________________________________________________

90

_______________________________________________________________________________

The Economics of Desert Societies

according to their proﬁtability (yield for time spent foraging within a patch) and predicts the threshold at which a forager should exploit additional patches. Lower ranked patches will be exploited when the marginal gain of continuing to forage in a given patch drops its proﬁtability below the overall return rate of the habitat (yield for time spent traveling and foraging) (Stephens and Krebs 1986: 24–34). O’Connell and Hawkes (1981, 1984) have considered two questions about Alyawara foraging: 1 Why did women select a particular set of resources and ignore others that were abundant and formerly important in the diet? 2 Why did men pick particular hunting locales and ignore others that were at least equally accessible? Contrary to the conventional wisdom of cultural ecology, Alyawara socioeconomic patterns were not directly determined by sheer abundance or seasonality of subsistence resources. The difference in energy yields between the sandplain (3,200 kcal/hr searching and handling) and mulga woodland (652 kcal/hr) were such that when women traveled to these patches, if they chose to handle the abundant ripe acacia and grass seeds (<600 kcal/hr handling) of the latter, they would have reduced their foraging efﬁciency. The PCM predictions are consistent with the observation that women avoided these high-cost seeds. Cane (1987) found similar results among the Gugadja of the Western Desert: here too, seeds were costly, providing only 340 kcal per hour. These resources would be efﬁcient options only if encounter rates with resources like goanna and cossid larvae were signiﬁcantly reduced. O’Connell and Hawkes (1981) hypothesized that the availability of commercial ﬂour reduced pressure on the more proﬁtable resources, which in turn increased overall foraging efﬁciency and narrowed the range of resources exploited. As predicted by the PUM, Alyawara men did not sample potential hunting patches at random or simply as a function of proximity. When hunting on foot, men chose to travel to either the mulga or open woodland patches, and completely avoided the local ﬂoodplain. Average return rates from traveling to and hunting in the mulga and open woodland patches were not signiﬁcantly different (roughly 2 kg/hr traveling, searching, and pursuing). However, the efﬁciency of traveling to hunt in these patches is signiﬁcantly higher than that possible from hunting in the ﬂoodplain, where despite its accessibility, game was rare and inpatch return rates were predictably very low. Similar results were obtained from hunting trips in vehicles. As expected, the open grassland patch, which was visited most frequently, also had the highest average return rate. Traveling to this patch was signiﬁcantly more efﬁcient than the in-patch return rates from a wide range of locally available areas in the woodland, rocky ranges, and ﬂoodplain. Occasionally, hunters did use other patches, but usually en route to the open grassland, or when in the process of hunting in the grassland, certain conditions reduced their return rates to the point at which switching to an adjacent patch was a more efﬁcient option.

___________________________________________________________________________________

91

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

The Alyawara data are consistent with these hypotheses: . Women adjusted prey choice in ways consistent with maximizing efﬁciency while foraging within a given patch. . Men behaved as if their goal was to maximize their rate of energetic capture while hunting. However, the ways in which people did not meet the expectations of the models may be just as informative. Firstly, Alyawara men and women had very different foraging strategies. It is unlikely that both strategies were designed for the same purpose (e.g., reliably delivering the same currency to a household). In many cases we do not know whether foragers could have been more efﬁcient by exploiting resources or patches that they chose to pass over. Secondly, certain kinds of foods were more likely to be shared more widely than others, which (from the perspective of the acquirer) would signiﬁcantly change the expected efﬁciency of these resources. As mentioned above, traditional ecological explanations of sharing have almost always assumed it functions for the beneﬁt of the group. Behavioral ecologists tend to focus on the potential conﬂicts of interest between individuals and groups. If a hunter concentrates on acquiring goods that are shared widely and indiscriminately (surely a group beneﬁt), what keeps everyone else from free-riding off his labor, allowing them more time to focus on more reliable resources that are more narrowly shared? Over the last decade in HBE there has been a great deal of debate about the role of food sharing and sex-linked differences in foraging activities (see Bliege Bird 1999; Bliege Bird et al. 2003; Gurven et al. 2000; Hawkes 1993; Hill 2002; Kaplan et al. 2000; Winterhalder 1996). Some scholars have argued that specialized foraging and sharing activities are designed to provision highly dependant offspring. Accordingly, women tend to focus on activities that are compatible with childcare and resources that provide dependable daily income, while men specialize in skill-based activities that target large package resources whose acquisition is more variable, but widespread reciprocal sharing ensures efﬁcient provisioning (Kaplan et al. 2000). The problem is that ‘‘men’s resources’’ (e.g., large game) are not usually shared reciprocally. As noted above, the common pattern is one in which resources continually ﬂow from good hunters to others who do not return the favor, making large game hunting an unreliable provisioning strategy (e.g., Hawkes et al. 2001). This has led some researchers to suggest that men and women may not always have similar foraging goals (see review in Hawkes and Bliege Bird 2002). Along with Hawkes and Smith, we have recently argued that in some circumstances men may choose particular foraging activities and patches that maximize the efﬁciency of social display rather than some nutritional currency (e.g., Bliege Bird et al. 2001; Smith et al. 2003). Certain foraging activities, especially those that require a high degree of skill and cost, can be an important means of advertising underlying, difﬁcult to detect qualities of the forager (in HBE parlance, these activities serve as ‘‘costly signals’’, sensu Grafen (1990) and Zahavi and Zahavi

___________________________________________________________________________________

92

_______________________________________________________________________________

The Economics of Desert Societies

(1997)). Such displays are especially efﬁcient when they draw a large audience, by providing widely shared ‘‘public goods’’ for example. Both the signaler and the recipient can beneﬁt when the cost of the activity ensures the ‘‘honesty’’ of the information advertised in the display. We have suggested that certain aspects of men’s foraging among Martu Aborigines in the Western Desert may be consistent with a costly signaling hypothesis. During the cool-dry season men would increase their foraging return rates signiﬁcantly if they stopped their search for larger game (2,125 kcal/hr searching and handling) to dig for sand goanna and pick Solanum fruit (post-encounter returns ¼ 5,318 and 4,217 kcal/hr handling, respectively). Only on very rare occasions did we observe men discontinue their search to pursue these resources. We hypothesized that during this season the beneﬁts of displaying the skills of tracking mobile animals may make larger game especially attractive to certain men; and the fact that these game are shared widely ensures a broad audience (in spite of the extreme modesty of the hunters). In the hot season small game is more mobile and the opportunities to track them greatly increase. During this time men often interrupt their search for larger game to pursue sand goanna. We have hypothesized that men may prefer mobile game in all seasons because only tracking has the potential to discriminate skill levels among hunters. While Martu ascribe no political power to skilled trackers, they do make overt distinctions in skill for male and female trackers (Bliege Bird and Bird in press).

HBE in the Great Basin Rather than searching for something essential or universally typical of all huntergatherers, the arid Australian examples are especially instructive in drawing our attention to the causes of variability in foraging strategies. While the cases discussed above focused on variability within groups, applications of HBE in the Great Basin have generally investigated variability in subsistence through time. The range of HBE applications to Great Basin human prehistory is well beyond the scope of this chapter – syntheses can be found in Beck (1999), Bettinger (1991), Elston (1986), Elston and Zeanah (2002), Grayson (1993), Madsen and O’Connell (1982), and Simms (1987). Here we focus on an issue that we feel can be enriched by comparison with the Australian data discussed above: subsistence variability in the early to mid-Holocene. In the transition from the late Pleistocene/early Holocene (ca. 11,200–8,000 bp) to the mid-Holocene (ca. 8,000–4,000 bp), patterns in the Great Basin archaeological record shift dramatically (Beck and Jones 2001; Elston and Zeanah 2002; Grayson 1993; Simms 1987). Archaeological assemblages of technology prior to about 8,000 bp (commonly referred to as ‘‘Paleo-Indian’’) are dominated by large projectile points and formal ﬂaked stone tools. Grinding stones, which were historically used to process seeds, are rare until about 8,000 bp. While there continues to be debate about the diversity of the Paleo-Indian archaeological sites, a strong case can be made that, through the early Holocene, more extensive lakes and marshes were central to the subsistence economy of the people living in the Great Basin. When these pluvial lake systems disappeared,

___________________________________________________________________________________

93

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

so did the lowland sites with large projectile points. At roughly the same time, grinding stones became more common throughout the region. Many archaeologists have suggested that this may indicate a shift from an economy based primarily on hunting larger game to one based on a broad spectrum of plants and small animals, commonly referred to as the ‘‘Archaic.’’ Partly based on experience among the Alyawara, O’Connell et al. (1982) used the framework of the prey choice model (PCM, described above) to explain some of these archaeological changes. They hypothesized that if, in the late Pleistocene and early Holocene, encounters with large game were sufﬁciently common, overall foraging returns may have been high enough for foragers to bypass resources that would entail extensive processing with grinding stones (i.e., seeds). This hypothesis was based on the assumption that large game were the most proﬁtable (energy for time spent in pursuit and processing) option, and that seeds ranked lower than most other resources (small game, roots, berries, etc.). Simms (1987) tested this assumption with processing experiments on a wide array of Great Basin foods. Simms found that, indeed, grass seeds had the lowest proﬁtability of all major food resources – as in Australia, grass seeds from the Great Basin had on-encounter return rates usually less than 300 kcal/hr. Thus, based on the predictions of the PCM, it was reasonable to infer that seeds, and the technology required to process them, would only become important archaeologically with reductions in more proﬁtable prey, especially large and medium-sized game. This work has since inspired more detailed investigations into the nature of the Paleo-Indian/Archaic material record and our assumptions about the goals and efﬁciency of acquiring different kinds of resources. Two issues are especially relevant here. Firstly, there are very few Paleo-Indian sites in the Great Basin associated with large mammal remains, and we now have clear evidence that diets from this time included a wide array of small animals, ﬁsh, and seeds (see reviews in Elston and Zeanah 2002 and Grayson 1993). Secondly, large game is not always very proﬁtable. While each animal represents a large package of meat that requires relatively little processing, they commonly entail long pursuits and widespread sharing which can translate into very low returns for the hunter (e.g., Bliege Bird and Bird in press; Bliege Bird et al. 2001; Hawkes et al. 2001). As discussed above, hunting large game may often incorporate goals related to social display that may conﬂict with efﬁcient provisioning. If foragers in the Great Basin harvested a broad spectrum of subsistence resources through this long period of time, what accounts for the variability in the technological record of the Paleo-Indian/Archaic transition? Elston and Zeanah (2002) have recently proposed that the signiﬁcant changes in lithic technology reﬂect the change from an earlier condition characterized by low human population density, higher residential mobility, and men and women performing sex-speciﬁc tasks in a landscape that was very different from that known historically. Recent paleoenvironmental research indicates that the late Pleistocene was cooler and drier than today, while the early Holocene was warmer and wetter and possibly more variable. Based on a series of ecological simulations, Elston and Zeanah (2002) suggest that through the early Holocene, large animals

___________________________________________________________________________________

94

_______________________________________________________________________________

The Economics of Desert Societies

should have been relatively abundant in the low to mid-elevation foothills from fall to spring, and abundant wetlands would have provided reliable late winter–early spring foraging patches. Even if men and women may have had different foraging goals, basing camps in the lowlands would have provided local patches for women to acquire small game, ﬁsh, roots, and seeds, as well as opportunities for men to travel to hunting locales: The abundance of expansive wetlands, and lack of competition throughout the Great Basin, allowed Prearchaic foragers high mobility that maximized men’s encounters with large game without sacriﬁcing women’s foraging interests . . . our simulations suggest that the known Prearchaic record is biased toward sites positioned primarily to access women’s resources, but bearing a technology reﬂecting men’s [activities] . . . (Elston and Zeanah 2002: 120–1)

If this is so, the rarity of grinding stones in the early sites may be in part an epiphenomenon, the outcome of low population densities and high mobility (i.e., there would have been fewer chances for grinding stones to become part of the archaeological record). This case has also been made for the early Holocene record of Australia (Gorecki et al. 1997). Testing this hypothesis will require better experimental and archaeological data. For example, we still have much to learn about variability in the efﬁciency of different hunting techniques that may have been used in the past, return rates for collecting different kinds of vegetable foods in different settings, and changes in efﬁciency due to things like forager age, experience, and size.

Conclusion The long history of HBE research in arid Australia and the Great Basin demonstrates a close relationship between different aspects of the socioeconomic life and environmental circumstances of groups. Societies in both regions have much in common – ﬂuid and highly mobile local organization depending heavily on resources acquired by women. The parallels are not necessarily a description of something ‘‘essential’’ about arid hunter-gatherers; rather, they are similarities in the way that people vary their behavior under changing and sometimes unpredictable physical and social circumstances. Sex-linked differences in foraging strategies are especially instructive for exploring factors that inﬂuence this variability. In both regions, women focused on a broad spectrum of resources. A growing body of behavioral ecology originating from studies in the Great Basin and desert Australia suggests that women’s decisions about prey and patch choice closely track environmental variability in ways that increase their individual energy returns. Relative to this pattern, men’s decisions often seem idiosyncratic. Ethnographic accounts, quantitative data, and some archaeological patterns indicate that sometimes variability in men’s strategies may conﬂict with efﬁcient food provisioning. However, their strategies may be better understood if we take into account that men and women may have different goals in acquiring and sharing food.

___________________________________________________________________________________

95

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d

Human behavioral ecology is a useful framework for investigating aspects of such variability. It provides a coherent theory of human behavioral adaptation; not simply a description of how people respond to environmental variability, but a logical means by which we can generate and test hypotheses about why a pattern of behavior is maintained or varies in a particular way. Ethnographically, it provides a set of formal models, with a limited set of variables, that when tested against observable social behavior, can isolate variability that does not meet our theoretical expectations. For example, in this chapter we have demonstrated that the way in which men’s and women’s subsistence decisions fail to match the predictions of simple models promises to lead us in novel directions with new and very testable hypotheses. For issues about prehistoric subsistence, it can be argued that human behavioral ecology frees us from relying only on ethnographically known cases for models of the past. It does this by focusing on hypotheses designed to address issues of variability independent of speciﬁc ethnographic analogs (O’Connell 1995; Winterhalder and Smith 2000). In a sense, this is simply the process of doing good science related to human behavior; testing theoretically justiﬁable ideas against observations, with the goal of producing useful hypotheses that lead to new questions and new directions for exploring human–environmental dynamics.

References Beck, C. (ed.) 1999: Models for the Millennium: Great Basin Anthropology Today. Salt Lake City: University of Utah Press. Beck, C. and Jones, G. T. 2001: Lithic source use and foraging patterns in the Great Basin Paleoarchaic. New Orleans: Unpublished paper presented at the 66th Annual Meeting of the Society for American Archaeology. Bettinger, R. L. 1991: Hunter-Gatherers: Archaeological and Evolutionary Theory. New York: Plenum Press. Bird, D. W. and Bliege Bird, R. in press: Martu children’s hunting strategies in the Western Desert, Australia: Implications for the evolution of human life histories. In B. Hewlett (ed.), Hunter-Gatherer Children Senri Ethnological Studies: Papers from the 9th International Conference on Hunting and Gathering Societies. Bliege Bird, R. 1999: Cooperation and conﬂict: The behavioral ecology of the sexual division of labor. Evolutionary Anthropology, 8, 65–75. Bliege Bird, R. and Bird, D. W. in press: Human hunting seasonality: A case from Australia. In D. Brockman and C. van Shaik (eds), Primate Seasonality, Cambridge: Cambridge University Press. Bliege Bird, R., Bird, D. W., Smith, E. A., and Kushnick, G. C. 2003: Risk and reciprocity in Meriam food sharing. Evolution and Human Behavior, 23, 297–321. Bliege Bird, R., Smith, E. A., and Bird, D. W. 2001: The hunting handicap: Costly signaling in male foraging strategies. Behavioral Ecology and Sociobiology, 50, 9–19. Cane, S. 1987: Australian Aboriginal subsistence in the Western Desert. Human Ecology, 15, 391–434. Cronk, L. 1991: Human behavioral ecology. Annual Review of Anthropology, 20, 25–53. Elston, R. G. 1986: Prehistory of the western area. In W. L. d’Azevedo (ed.), The Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 135–48.

___________________________________________________________________________________

96

_______________________________________________________________________________

The Economics of Desert Societies Elston, R. G. and Zeanah, D. W. 2002: Thinking outside the box: A new perspective on diet breadth and sexual division of labor in the Prearchaic Great Basin. World Archaeology, 34, 103–30. Fowler, C. S. 1972: Comparative Numic ethnobiology. Unpublished PhD thesis, Pittsburgh, PA: University of Pittsburgh. Fowler, C. S. and Fowler, D. D. 1971: Notes on the history of the southern Paiutes and western Shoshonis. Utah Historical Quarterly, 39 (2), 95–113. Fowler, C. S. and Liljeblad, S. 1986: Northern Paiute. In W. L. d’Azevedo (ed.), The Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 435–48. Fowler, D. D. 1986: History of research. In W. L. d’Azevedo (ed.), The Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 15–30. Gorecki, P., Grant, M., O’Connor, S., and Veth, P. 1997: The morphology, function, and antiquity of grinding implements in northern Australia. Archaeology in Oceania, 32, 141–50. Gould, R. A. 1969: Subsistence behavior among the Western Desert Aborigines of Australia. Oceania, 39, 253–74. Gould, R. A. 1982: Comparative ecology of food sharing in Australia and northwest California. In R. S. O. Harding and G. Teleki (eds), Omnivorous Primates, New York: Columbia University Press, 344–421. Gould, R. A. 1991: Arid land foraging as seen from Australia: Adaptive models and behavioral realities. Oceania, 62, 12–33. Gould, R. A., Fowler, D. D., and Fowler, C. S. 1972: Diggers and doggers: Parallel failures in economic acculturation. Southwestern Journal of Anthropology, 28, 265–81. Grafen, A. 1990: Biological signals as handicaps. Journal of Theoretical Biology, 144, 517–46. Grayson, D. K. 1993: The Desert’s Past: A Natural Prehistory of the Great Basin. Washington, DC: Smithsonian Institution Press. Gurven, M., Hill, K., Hurtado, A. M., and Lyles, R. 2000: Food transfers among Hiwi foragers of Venezuela: Tests of reciprocity. Human Ecology, 28, 171–218. Hawkes, K. 1993: Why hunter-gatherers work: An ancient version of the problem of public goods. Current Anthropology, 34, 341–61. Hawkes, K. and Bliege Bird, R. 2002: Showing off, handicap signaling, and the evolution of men’s work. Evolutionary Anthropology, 11, 58–67. Hawkes, K. and O’Connell, J. F. 1981: Afﬂuent hunters? Some comments in light of the Alyawara case. American Anthropologist, 83, 622–6. Hawkes, K., O’Connell, J. F., and Blurton Jones, N. G. 2001: Hadza hunting and the evolution of nuclear families. Current Anthropology, 42, 681–709. Hill, K. 2002: Cooperative food acquisition by Ache foragers. Human Nature, 13, 105–28. Jennings, J. D. 1957: Danger Cave: A progress summary. El Palacio, 60, 179–213. Jennings, J. D. 1964: The Desert West. In J. D. Jennings and E. Norbeck (eds), Prehistoric Man in the New World, Chicago: University of Chicago Press, 149–74. Jennings, J. D. 1974: Prehistory of North America (2nd edn). New York: McGraw-Hill. Kaplan, H., Hill, K., Lancaster, J., and Hurtado, A. M. 2000: A theory of human life history evolution: Diet, intelligence, and longevity. Evolutionary Anthropology, 9, 156–85. Kelly, I. T. 1932: Ethnography of the Surprise Valley Paiute. Berkeley: University of California Press. Kelly, R. L. 1995: The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, DC: Smithsonian Institution Press.

___________________________________________________________________________________

97

_______________________________________________________________________________

D o u g l a s W. Bir d a nd R e b e c c a B l i e g e B i r d Krebs, J. R. and Davies, N. B. 1993: An Introduction to Behavioral Ecology (3rd edn). Oxford: Blackwell Scientiﬁc. Krebs, J. R. and Davies, N. B. (eds) 1997: Behavioral Ecology: An Evolutionary Approach (4th edn). Oxford: Blackwell Scientiﬁc. Latz, P. 1995: Bushﬁres and Bushtucker: Aboriginal Plant Use in Central Australia. Alice Springs: IAD Press. Liljeblad, S. and Fowler, C. S. 1986: Owens Valley Paiute. In W. L. d’Azevedo (ed.), The Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 414–34. Lourandos, H. 1997: Continent of Hunter-Gatherers: New Perspectives in Australian Prehistory. Cambridge: Cambridge University Press. Madsen, D. and O’Connell, J. F. 1982: Man and Environment in the Great Basin. Washington, DC: Society for American Archaeology. Meggitt, M. J. 1957: Notes on the vegetable foods of the Walbiri of central Australia. Oceania, 28 (2), 143–5. Meggitt, M. J. 1964: Aboriginal food gatherers of tropical Australia. In Proceedings and Papers of the IUCN 9th Technical Meeting, part 1, Nairobi, 30–7. O’Connell, J. F. 1995: Ethnoarchaeology needs a general theory of behavior. Journal of Archaeological Research, 3, 205–55. O’Connell, J. F. and Hawkes, K. 1981: Alyawara plant use and optimal foraging theory. In B. Winterhalder and E. A. Smith (eds), Hunter-Gatherer Foraging Strategies: Ethnographic and Archaeological Analyses, Chicago: University of Chicago Press, 99–125. O’Connell, J. F. and Hawkes, K. 1984: Food choice and foraging sites among the Alyawara. Journal of Anthropological Research, 40, 504–35. O’Connell, J. F., Jones, K. T., and Simms, S. R. 1982: Some thoughts on prehistoric archaeology in the Great Basin. In D. B. Madsen and J. F. O’Connell (eds), Man and Environment in the Great Basin, Washington, DC: Society for American Archaeology, 227–41. O’Connell, J. F., Latz, P., and Barnett, P. 1983: Traditional and modern plant use among the Alywara of central Australia. Economic Botany, 37, 80–109. Peterson, N. 1976: The natural and cultural areas of Aboriginal Australia: A preliminary analysis of population groupings with adaptive signiﬁcance. In N. Peterson (ed.), Tribes and Boundaries in Australia, Canberra: Australian Institute of Aboriginal Studies, 50–71. Peterson, N. 1977: Aboriginal uses of Australian Solanaceae. In J. G. Hawkes, R. N. Lester, and A. D. Skedling (eds), The Biology and Taxonomy of the Solanaceae, London: Linneaus Society, 171–88. Peterson, N. 1978: The traditional pattern of subsistence to 1975. In B. S. Hetzel and H. J. Frith (eds), The Nutrition of Aborigines in Relation to the Ecosystem of Central Australia, Melbourne: CSIRO, 25–35. Service, E. 1962: Primitive Social Organization: An Evolutionary Perspective. New York: Random House. Simms, S. R. 1987: Behavioral Ecology and Hunter-Gatherer Foraging: An Example from the Great Basin. Oxford: BAR International Monograph Series no. 381, Hadrian Books. Smith, E. A. and Winterhalder, B. (eds) 1992: Evolutionary Ecology and Human Behavior. Hawthorne: Aldine de Grutyer. Smith, E. A., Bliege Bird, R., and Bird, D. W. 2003: The beneﬁts of costly signaling: Meriam turtle hunters and spearﬁshers. Behavioral Ecology, 14, 116–26.

___________________________________________________________________________________

98

_______________________________________________________________________________

The Economics of Desert Societies Stanner, W. E. H. 1965: Aboriginal territorial organization: Estate, range, domain, and regime. Oceania, 36, 1–26. Stephens, D. W. and Krebs, J. R. 1986: Foraging Theory. Princeton, NJ: Princeton University Press. Steward, J. H. 1933: Ethnography of the Owens Valley Paiute. University of California Publications of American Archaeology and Ethnography, 33, 233–350. Steward, J. H. 1936: The economic and social basis of primitive bands. In Essays in Anthropology Presented to A. L. Krober, Berkeley: University of California Press, 331–50. Steward, J. H. 1938: Basin-Plateau Aboriginal Sociopolitical Groups. Washington, DC: Smithsonian Institution Bureau of American Ethnology. Steward, J. H. 1941: Culture element distributions: XXVIII: Nevada Shoshone. University of California Anthropological Records, 8, 263–392. Steward, J. H. 1955: Theory of Culture Change. Urbana: University of Illinois Press. Steward, J. H. 1968: Causal factors and processes in the evolution of pre-farming societies. In R. B. Lee and I. DeVore (eds), Man the Hunter, Chicago: Aldine, 321–34. Stewart, O. 1941: Culture element distributions: XIV: Northern Paiute. University of California Anthropological Records, 4, 361–446. Thomas, D. H. 1983: The Archaeology of Monitor Valley 2. Gatecliff Shelter. New York: Anthropological Papers of the American Museum of Natural History vol. 59, American Museum of Natural History. Thomas, D. H., Pendleton, L. S., and Cappannari, S. C. 1986: Western Shoshone. In W. L. d’Azevedo (ed.), The Handbook of North American Indians vol. 11, Washington, DC: Smithsonian Institution Press, 260–75. Tonkinson, R. 1991: The Martu Aborigines: Living the Dream of Australia’s Desert. Fort Worth, TX: Holt, Rinehart, and Winston. Veth, P. M. and Walsh, F. J. 1988: The concept of ‘‘staple’’ plant foods in the Western Desert region of Western Australia. Australian Aboriginal Studies, 2, 19–25. Walsh, F. J. 1987: The inﬂuence of spatial and temporal distribution of plant food resources on traditional Martujarra subsistence strategies. Australian Archaeology, 25, 88–101. Walsh, F. J. 1990: An ecological study of traditional Aboriginal use of ‘‘country’’: Martu in the Great and Little Sandy Deserts, Western Australia. Proceedings of the Ecological Society of Australia, 16, 23–37. Williams, G. C. 1966: Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton, NJ: Princeton University Press. Winterhalder, B. 1996: Social foraging and the behavioral ecology of intragroup resource transfers. Evolutionary Anthropology, 5, 46–57. Winterhalder, B. and Smith, E. A. 2000: Analysing adaptive strategies: Human behavioral ecology at twenty-ﬁve years. Evolutionary Anthropology, 9, 51–72. Zahavi, A. and Zahavi, A. 1997: The Handicap Principle: A Missing Piece of Darwin’s Puzzle. Oxford: Oxford University Press.

___________________________________________________________________________________

99

_______________________________________________________________________________

--------------------------------------------------------------------------------

6

----------------------------------------------------------------------------

Cycles of Aridity and Human Mobility Risk Minimization Among Late Pleistocene Foragers of the Western Desert, Australia Peter Veth Introduction Australia is the most arid continent to have been colonized by modern humans and the Western Desert is arguably the most marginal landscape within it. This is largely due to its high temperatures, lack of coordinated drainage, paucity of permanent waters, and patchy and unpredictable resources (cf. Mabbutt 1977; Veth 1993a). These factors have historically created conditions of high risk and stress for arid land foragers (cf. Yellen 1976). Remarkably, there are now ﬁrm dates for occupation of the Western Desert in excess of 30,000 bp (cf. Smith et al. 1997; Thorley 1998a) and increasing evidence for early and widespread use of most landforms. There is also mounting evidence that major changes in settlement and mobility patterns occurred with the onset of hyperarid conditions associated with the Last Glacial Maximum (LGM) (O’Connor et al. 1999; Veth in prep.). In this chapter I discuss the central role of mobility in the risk minimizing strategies employed by Western Desert foragers and examine assemblages from three sites dating to the ﬁrst phase of occupation from between 32,000 to 22,000 bp (Smith et al. 1997; Thorley 1998a; Veth 2000). I will examine eight different lines of archaeological evidence that may be used to infer residential mobility from these Pleistocene assemblages.

Risk Minimization and the Role of Mobility in the Australian Deserts There is increasing evidence from the Western Desert that conditions were more favorable for humans during the ﬁrst phase of occupation and then worsened

_____________________________________________________________________________________________________

100

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

signiﬁcantly towards the height of the LGM by approximately 18,000 years ago (Smith 1996; Thorley 1998a; Veth 2000). This period saw the likely loss of a considerable number of permanent water sources and food staples (Smith 1989). Increasing aridity associated with the LGM probably represented the ﬁrst test of human adjustment to truly arid conditions with a majority view that humans were unable to cope with the ensuing hyperarid conditions and restricted their territory in response (Hiscock 1988; Thorley 1998b; Veth 1995; see also Chapter 3, this volume). I propose that during the terminal Pleistocene Western Desert foragers engaged in a pattern of settlement dominated by high residential mobility to low logistical mobility (after Binford 1980) coupled with a ‘‘territorial’’ mobility strategy whereby adjacent localities were abandoned for varying periods of time (after Kelly 1992, 1995; Varien 1999). Mobility conﬁgurations likely had great dynamism given the ethnographically observed patterns for Western Desert Aborigines, whereby strategy switching occurred in response to drought. Strategy switching describes how groups responded by either withdrawing into their ‘‘home’’ territory or, alternatively, avoided areas altogether through temporary abandonment (cf. Gould 1991). This extreme ﬂexibility in mobility is seen to underpin a raft of risk minimizing strategies in the Western Desert (cf. Tonkinson 1991). The nature of Pleistocene arid zone adaptations may be examined using a resource structure model, such as that proposed by Ambrose and Lorenz (1990), where the archaeological correlates of hunter-gatherer behaviors in an unpredictable and scarce resource regime are deﬁned (see Table 6.1). The marginality of the Western Desert for humans is largely a function of poor rainfall reliability and generally impoverished species diversity in comparison to other desert regions of the world (cf. Gould 1991). The archaeological predictions of the resource structure model are examined in the following sections. Table 6.1 The behavioral and archaeological correlates of an unpredictable and scarce resource structure (after Ambrose and Lorenz 1990: 10). Behavioral Correlates

Resource Structure-Unpredictable and Scarce

Residential mobility Territorial strategy Information exchange Group size Population density Diet breadth

Very high, opportunistic Undefended, very permeable High Very small Very low Very high Archaeological Correlates

Occupation site intensity Macro-regional assemblage variability Raw material sources Intra-site spatial organization Faunal and ﬂoral diversity

_____________________________________________________________________________________________________

Very low at home base High stylistic uniformity Local and distant exotics Poorly structured Very high, mostly plant

101

______________________________________________________________________________

Pe t e r V eth

Pleistocene Sites of the Western Desert The sites of Serpent’s Glen (O’Connor et al. 1998), Puritjarra (Smith 1989, 1996), and Kulpi Mara (Thorley 1988a, 1988b) all have occupational deposits dating to before 23,000 bp (see Figure 6.1). While this sample is not large it does provide early sites from a range of different settings located within and on the margins of the Western Desert. All dates discussed here are uncalibrated. Serpent’s Glen rock shelter is located within quartz sandstone ranges among an expanse of low relief dunes. The uplands contain several springs that provide reliable and potable water (O’Connor et al. 1998). Excavation revealed three major stratigraphic layers; the lower layer has a minimum date of 23,550 bp and contains a sparse assemblage of stone artifacts, ocher, charcoal, and bone. The

Figure 6.1 Location of the Western Desert showing key sites mentioned in the text (after Gould 1980, 1985; Morse 1994; O’Connor et al. 1998; Smith 1989; Thorley 1998a; Veth 1993a, 1993b).

_____________________________________________________________________________________________________

102

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

middle layer is culturally sterile, while the upper layer dates from 4,710 bp to the modern period and contains over 98 percent of the site’s lithic materials, in addition to signiﬁcant quantities of ocher, charcoal, bone, and minor quantities of emu eggshell, bird eggshell, and macroﬂoral remains. Puritjarra rock shelter is located within a western outlier of the central Australian ranges, and has a major permanent water source nearby. Puritjarra has a Pleistocene layer that contains a sparse ﬂaked stone, charcoal, ocher, and bone assemblage dated to as early as 32,000 bp (Smith et al. 1998). While occupation at Puritjarra has been argued by Smith (1989) to span the subsequent period of 22,000 to 13,000 bp coinciding with peak glacial aridity, this has been challenged by Hiscock (1988) and Thorley (1998a, 1998b) on both theoretical and empirical grounds. Sparse assemblages are certainly registered by 12,000 bp and by approximately 6,500 bp the quantity of ﬂaked stone increases substantially with an efﬂorescence in both quantity and breadth of cultural materials dating to the last millennium. Kulpi Mara is a large sandstone rock shelter on the escarpment of the central Australian ranges and, in contrast to the two previous sites, has no permanent water source nearby (Thorley 1998a, 1998b). The excavation revealed three layers, the lower dating to approximately 24,000 to 30,000 bp. Cultural assemblages in this unit are typically sparse. A chronological hiatus is then recorded with the middle layer dating from approximately 12,000 bp, when an increase in site use is registered. The middle of the upper layer is dated to 2,500 bp with the upper spits containing abundant lithics, charcoal, and also ocher, wooden and resin artifacts, and spun ﬁbers. At all three sites the passage of time coinciding with peak glacial aridity is registered as either culturally sterile sediments or very slow rates of sediment and artifact accumulation, suggesting abandonment or, at most, several minor episodes of ephemeral occupation (cf. Thorley 1998b). This pattern is evident in other desert lowland sites surrounding the Western Desert, such as at Noala Cave which dates to 27,000 bp and Mandu Mandu rock shelter dating from 34,000 bp (Morse 1994; O’Connor et al. 1999; Veth 1993b, 1995; see Figure 6.1). As a ﬁrst step it is necessary to establish that some level of resource stress was experienced at these sites and ideally this should be independent from other categories of archaeological data. This is required to support assumptions that patchy and unpredictable resource structure in the terminal Pleistocene would have resulted in stress to Western Desert foragers.

An Independent Index of Stress It is possible to calculate an independent index of relative protein stress from selected Western Desert and central Australian sites. On the basis of ethnographic accounts indicating that Western Desert Aborigines pulverized whole game and extracted marrow to maximize protein yield, high levels of bone reduction in arid zone sites are seen to be an indicator of protein stress. Where these patterns can be reliably associated with processing and pulverizing of game, as opposed to

_____________________________________________________________________________________________________

103

______________________________________________________________________________

Pe t e r V eth

taphonomic factors, then it is possible to consider relative levels of stress (Gould 1978, 1996; Haley 1999; O’Connor et al. 1998). Indices of fragmentation (IoF) can be calculated for combined fractions of economic fauna from all spits (total bone weight/number of fragments) for the sites of Kaalpi, Serpent’s Glen, and the Holocene-aged sites of Puntutjarpa and Intitjikula (see Figure 6.1). Although Holocene data are included here, the sites show long-term consistencies in the degree of bone comminution (cf. Gould 1996: 74, 81). A related index of protein stress is found in the ratio of broken to whole macropod teeth. Breakage of these robust elements has been documented as a byproduct of the crushing and pulverizing of crania and mandibles (Gould 1996; O’Connor et al. 1998). Paleoenvironmental models for the Australian arid zone would predict for greater resource scarcity and unpredictability, and hence stress, in the lowlands of the Western Desert and less so on the periphery and within the central Australian ranges (e.g. Gould 1991; Ross et al. 1992; Veth 1993a). It follows that the degree of bone reduction should be greater within the lowlands and this should be registered in lower values for IoF and higher ratios of broken to whole macropod teeth. Some consideration must be made of relative species distributions between sites, although this is problematic as the levels of taxonomic assignment are low, ranging between 8.8 to 12 percent. There is some evidence that the proportion of larger macropods present on the lowland sites is lower than those of the ranges and this will likely serve to inﬂate the differences in IoFs between these sites (cf. Gould 1996). These predictions of relative protein stress are generally met when the faunal assemblages are analyzed. The sites of Serpent’s Glen and the nearby site of Kaalpi, both located within expansive duneﬁelds, register IoFs of 0.063 gm/fragment and 0.104 gm/fragment, respectively (Haley 1999; O’Connor et al. 1998). Over 98 percent of macropod teeth are broken. The site of Puntutjarpa to the east registers an average IoF of 0.530 gm/fragment with 100 percent of macropod teeth being fractured. In contrast, the site of Intirtekwerle within the central Australian ranges registers an IoF of 0.500 gm/fragment (similar to Puntutjarpa), but only 60 percent of macropod teeth have been broken (see Table 6.2). A number of conclusions can be drawn from this comparison. Firstly, conditions of protein stress have obtained throughout the history of occupation at all of these Table 6.2 Values for Index of Fragmentation and macropod tooth breakage from the sites of Serpent’s Glen, Kaalpi, Puntutjarpa, and Intirtekwerle (from Gould 1996; Haley 1999; O’Connor et al. 1998). Site name

Index of fragmentation

% macropod teeth broken

Serpent’s Glen Kaalpi Puntutjarpa Intirtekwerle

0.063 0.104 0.530 0.500

100 >98 100 60

_____________________________________________________________________________________________________

104

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

sites and are predicted to have been present also at the sites of Puritjarra and Kulpi Mara. Secondly, the cline in increasing protein stress away from the ranges lends support to models of marginality which predict that the highest levels of residential mobility will occur on the desert lowlands (cf. Smith 1989).

Archaeological Indices of High Residential Mobility The intensity of stone reduction On the assumption that occupation was by small groups and for a short duration it is expected that the intensity of reduction of locally available lithics will be relatively low. Locally available and abundant stone materials account for between 90 to 100 percent of the Pleistocene assemblages. The intensity of utilization of these materials is seen to be a product of occupation intensity and to reﬂect group mobility. Assuming that assemblage variability can be strongly determined by occupation intensity (after Rolland and Dibble 1990), this factor is examined by considering changes in artifact discard rates (Smith 1988), degree of artifact curation (cf. Shott 1989: 26), and the presence of economizing strategies, such as bipolar knapping (after Hiscock 1996). The number of artifacts clearly ascribed to the pre-glacial layers of the Western Desert sites is small and contrasts, in two of the three cases, markedly with numbers of artifacts from the Holocene units. Estimates of artifact discard rates for the three sites are provided in Table 6.3. Differences of this high order in themselves must indicate that different settlement patterns have operated through time. With reference to curation at Serpent’s Glen there are no retouched/utilized artifacts in the Pleistocene layers (O’Connor et al. 1998). At Puritjarra there are only amorphous retouched artifacts in Layer 2 and these are substantially larger than those in the upper Holocene layer (Smith 1988); see Table 6.4. Size reduction is evident in both local (silcrete/siliciﬁed sandstone) and exotic (chert/chalcedony) raw materials (see Table 6.5). There are only three retouched/utilized artifacts in the pre-LGM layer from Kulpi Mara (Square C) with Thorley (1998a: 235) concluding: ‘‘Use of local materials appears to have been predominantly opportunistic and the production Table 6.3 Estimated artifact discard rates from pre-LGM and Holocene layers from Pleistocene Western Desert sites. Site Serpent’s Glen Puritjarra Kulpi Mara

Pre-LGM number of artifacts/kyr

Late Holocene number of artifacts/kyr

5 6 140

1600 2087 250

_____________________________________________________________________________________________________

105

______________________________________________________________________________

Pe t e r V eth Table 6.4 Mean weight of amorphous retouched artifacts in grams, Puritjarra (after Smith 1988: 119). Pit Layer Ia Layer Ib Layer II

N9

N10

QR9

Z10

Total number

6.9 18.2 17.0

8.0 8.0 75.5

37.7 10.8 691.7

14.1 8.0 104.1

51 33 9

Table 6.5 Mean weights of total artifacts in grams, Puritjarra, Pit N19 (after Smith 1988: table 4.10). Pit

Local lithics

Exotic lithics

2.3 3.5 24.0

0.7 1.1 15.3

Layer Ia Layer Ib Layer II

of whole silcrete/quartzite ﬂakes . . . independent of [later] backed artifact manufacture and discard.’’ The combination of low numbers of artifacts and their low degree of modiﬁcation attests to a low intensity of site occupation with expedient use of locally available raw materials. It is possible that tools brought into these sites as part of a high residential mobility pattern were made from higher quality ‘‘exotics’’ and that these were extensively curated and transported until a preferred resupply zone was encountered, thereby increasing the expedient use of local lithics. Available data from Puritjarra and Kulpi Mara do not presently allow an assessment of reduction stages; however, the small assemblage from Serpent’s Glen is typical of other assemblages in the Western Desert dominated by early stages of reduction of local raw materials (see O’Connor et al. 1998; Veth 1993a). None of the site reports indicate any bipolar reduction in the Pleistocene, argued to be an indicator of lower residential mobility from other parts of northern Australia (Hiscock 1996).

The Diversity of Artifacts A number of studies have illustrated that there is an inverse relationship between assemblage diversity and level of residential mobility (cf. Andrefsky 1998: 204; Shott 1986: 25). This relationship is shown in Figure 6.2. Assemblage diversity equates with the number of artifact types (tools) recovered from an assemblage. A more sensitive index of technological diversity may be deﬁned by measuring the number of tool classes used during daily activities (Shott 1986: 23). Shott’s analysis was based on artifact data and mobility information from 12 ethnographic studies of hunter-gatherer groups. Base camps (after Binford 1980) where a more logistical strategy was practiced showed a positive correlation between artifact diversity and length of stay.

_____________________________________________________________________________________________________

106

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

Figure 6.2 Proposed relationship between artifact diversity and residential mobility (Andrefsky 1998: ﬁgure 8.5, adapted from Shott 1986: 25).

Andrefsky (1998: 206) contrasts this pattern with the low diversity expected at task speciﬁc sites and states: Although no ethnographic data on artifact diversity were available for special-taskoriented or ﬁeld camps, it would be logical to assume that special-task-oriented camps, such as hunt camps, plant collecting stations, or butchering sites, would have a relatively low diversity of artifacts. In other words, if a narrow range of activities were performed at a particular location, one would expect to ﬁnd a relatively low number of artifact types.

Such data are available, however, from ethnoarchaeological work carried out by the author from 1986 to 2004 with Martu Aborigines of the Western Desert (see Veth 1987, 1989, 1993a, 1996, unpublished notes). A range of artifact sites which had been used by Martu before contact with Europeans from the 1920s to the 1960s (cf. Tonkinson 1991) was analyzed. Because of these historical associations, localities could be reliably ascribed, at least for the historic period, as (a) major residential base camps which were used for aggregations; (b) smaller residential camps used predominantly in an ephemeral mode; and ﬁnally, (c) those which were clearly taskspeciﬁc or ﬁeld camps. A combination of high water permanency and high diversity of economic plant species predicted for greater permanency of occupation at sites and often by larger groups of people. Localities near ephemeral water sources and less diverse plant patches witnessed the highest residential mobility. Assumed aggregation sites located near permanent waters had the highest values for artifact diversity, the highest proportion of exotic materials, evidenced a greater proportion of artifacts in later reduction stages, contained more tools exhibiting high levels of curation and recycling, and had signiﬁcantly more basal grinding stones (Veth 1993a, 1996). The sites used by Martu custodians as ephemeral residences and which were referred to as ‘‘passing-through places’’ generally reﬂected the converse pattern.

_____________________________________________________________________________________________________

107

______________________________________________________________________________

Pe t e r V eth

I am not suggesting here that the late Holocene assemblages are directly comparable to those of the Pleistocene. Rather, I argue that the material outcomes of different mobility conﬁgurations in the Western Desert are tangible and do not support previous models arguing for the homogeneity of residential sites (Hayden 1976). Tools identiﬁed in sites which had been used by Martu in the historic period included retouched/utilized ﬂakes, amorphous grindstones, millstones, groundedge axes, backed blades, backed geometrics, tula adzes, and burren adzes. Artifact diversity values for these three classes of sites are shown in Table 6.6. It was expected that the Pleistocene assemblages of the three Western Desert sites would have diversity values closer to those of the ephemeral and task speciﬁc sites in contrast to the base camps. It must be noted, however, that the introduction of new standardized tools from the mid-Holocene, such as the adzes and backed pieces, will inﬂate the diversity values for the Holocene sites (cf. Hiscock and Veth 1991). Diversity values for the Pleistocene-aged sites are provided in Table 6.7. The tools located from these three sites may be incorporated into the category of retouched/utilized ﬂakes (referred to as ‘‘amorphous retouched’’ by Smith 1988: 126) where clear evidence of modiﬁcation to a blank is visible through retouch/ use-wear. These artifacts are made on a wide variety of blank morphologies and are consistent with the expedient tools recorded by Hayden (1979) from the Western Desert. Another category is also recognized here: primary ﬂakes that may have microscopic evidence for use-wear. This is simply a cautionary category that serves to inﬂate the artifact diversity values in lieu of use-wear analyses that are clearly in order. These values are clearly very low and stand in stark contrast to those recorded from the major aggregation sites, as shown in Table 6.6. It should be stressed that sample size effect and the known increase in formal tools by the mid-Holocene will serve to exaggerate this difference. It is reasonable to conclude on the basis of the

Table 6.6 Artifact diversity from Western Desert sites (from Veth 1993a, in prep.). Site type

Number of sites

Artifact diversity

9 34 15

8 4 2

Residential base camp Ephemeral residence Task speciﬁc

Table 6.7 Artifact diversity values for Pleistocene assemblages from Western Desert sites. Site

Layer

Artifact diversity

Serpent’s Glen Puritjarra Kulpi Mara

3 2 3

1 2 2

_____________________________________________________________________________________________________

108

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

available data, however, that there is a low diversity of tools in the Pleistocene assemblages.

The Presence of Local and Exotic Lithics It is expected that large foraging ranges associated with a high residential mobility strategy will result in the curation, maintenance, and occasional discard of lithics which may come from distant sources and hence be termed exotic – especially where suitable alternative sources are not readily available. The presence of abundant though highly variable quality silcretes and quartzites in the vicinity of all three Western Desert sites will have provided the range of blanks generally required for expedient maintenance tools in the Western Desert. The majority of artifacts from the three sites are from local materials (90–100 percent), with the presence of exotics likely a reﬂection of the scale of territorial range (Binford and Stone 1985 vs Gould 1985). Increases in the proportion of exotics are associated with the later appearance of standardized hafted tools, such as tula adzes and backed blades, by the midHolocene (e.g., Hiscock 1994; Hiscock and Veth 1991; Thorley 1998a: 235) and particularly during the subsequent emergence in the late Holocene of what are argued to be intensifying long distance exchange and ritual networks (Thorley 1998a). Major aggregation sites that served as centers of ‘‘production’’ for such regional networks evidence a signiﬁcant increase in the proportion of exotics. This increase in exotics is argued to be a reﬂection of longer distance exchange networks and the higher number of source areas from which participating groups traveled (e.g. Cane 1984; Thorley and Gunn 1996; Veth 1993a).

The Quantity of Grinding Material Pleistocene assemblages from arid Australia rarely contain seed grinders (e.g., Gorecki et al. 1997) and this may be linked to a range of factors, including the function and permanency of site use, assemblage size effect, and a general lack of intensive seed/food processing (see, however, Edwards and O’Connell 1995). Grindstones conforming to a wide range of morphologies are commonly found in stratiﬁed contexts from the arid and semi-arid zone from after approximately 3,500 bp (Smith 1986). This has been argued to represent one of the major transitions in the economy and settlement of arid land peoples as they shift towards the exploitation of unpredictable yet densely clustered resources, with consequent changes in mobility strategy (Thorley 1998a). At the site of Serpent’s Glen, Veth and O’Connor (1996) have recorded hundreds of whole and fragmented formal millstones, mullers, and amorphous grindstones on the surface and surrounding sandplains which are assumed to date to the late Holocene. In contrast, there are no ground artifacts in the Pleistocene layer of Serpent’s Glen. Equally, there are no grinding implements before the midHolocene in the sites of Puritjarra and Kulpi Mara (Thorley 1998a: 323). There is abundant ethnohistorical evidence (cf. Smith 1986) that seed grinding stations

_____________________________________________________________________________________________________

109

______________________________________________________________________________

Pe t e r V eth

were established at major aggregation sites and that these helped to underpin larger gatherings of people for extended periods of time. The mobility conﬁguration of such sites was clearly more towards a logistical strategy. The lack of any seed grinding materials in the Pleistocene layers of the three Western Desert sites is consistent with their use as ephemeral occupation sites. Data collected by Veth (1993a) from occupation sites adjacent to water sources of varying permanency show a clear correlation between water permanency and number of intact grindstones (see Table 6.8). Martu Aborigines noted that the sites adjacent to more reliable waters had acted as major aggregation venues and that groups from many linguistic afﬁliations had come to such places, sometimes traveling hundreds of kilometers in small family groups.

Stylistic Uniformity of Rock Art Patterning in art has been used by a range of researchers to indicate aggregation and dispersion patterns resting largely on interpretations of inside/outside (exclusive/inclusive) access to information networks (Conkey 1980; Galt-Smith 1997; Gould 1980). The high residential mobility scenario of the Pleistocene would predict for high stylistic uniformity and the predominant use of outside symbols. Both the Serpent’s Glen site, and the sites of Kaalpi and Durba Springs to the north in the Little Sandy Desert, have extensive rock art galleries comprising both petroglyphs and pictographs. Although dating the rock art is in its early stages, the majority of paintings are thought to date to the late Holocene, with ocher specimens from the site of Kaalpi dating from 1,300 bp and essentially absent from before this time (Haley 1999: 61). Recent dates from a shelter with paintings at Durba Springs also support a relatively recent origin for the paintings (Veth in prep.). In contrast, a large number of the engravings occur on quartz sandstone panels which have been signiﬁcantly weathered, chemically altered, and in some cases covered in very thick coatings of natural varnish and other crusts. It is quite likely that many of these engraved motifs date to the Pleistocene. Preliminary recording of motifs at these three sites suggests there are signiﬁcant differences in the style, spatial patterning, and frequency of different classes of motifs between the engravings and the paintings. It is suggested that these reﬂect a change in site function from the Pleistocene to the late Holocene (Veth et al. 2001). This is seen to be due to their changing nature as ephemeral residential sites in the Pleistocene to major aggregation sites in the late Holocene. The engravings include

Table 6.8 Mean number of whole grindstones located at sites adjacent to permanent, semi-permanent, and ephemeral water sources (from Veth 1993a). Water permanency Permanent Semi-permanent Ephemeral

Number of sites

Mean number of grindstones

8 23 16

16.625 4.739 1.188

_____________________________________________________________________________________________________

110

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

a range of tracks, circles, concentric circles, and arcs, in addition to a range of naturalistic motifs including macropods, marsupials, and various birds. Noticeable are some very large anthropomorphs which have extensive inﬁlling and the socalled archaic faces – disembodied faces which have a distribution through the arid zone from the coastal Pilbara to the Cleland Hills of central Australia, where Puritjarra is located (Dix 1977; see Chapter 7, this volume). The geometric motifs, archaic faces, and large anthropomorphs can be found in differing proportions at Kaalpi, Serpent’s Glen, and Durba Hills. The majority of the motifs appear to refer to foraging themes with the geometric motifs typical of those that have been documented as acting as mnemonic devices for mapping both water hole and ancestral routes at the regional level (after Gould 1980). This is assumed to reﬂect an outside or inclusive symbolic conﬁguration. In contrast, the paintings display greater assemblage diversity and contain a higher proportion of complex designs, although many of the fundamental geometric motifs are also present. Several Martu with whom I have worked had participated in ceremonies and gatherings at Kaalpi and Durba before contact with Europeans and were able to provide some mythological referants for the ﬁgurative paintings. Very detailed anthropomorphs are depicted with expanded headdresses, these only being recorded from this speciﬁc locality of the Western Desert – as such they may be signiﬁers of corporate identity. The use of inside/outside symbols is consistent with the increasing role of these sites for aggregation, rather than dispersion (cf. Galt-Smith 1997).

Changes in Sedimentary Records A shift from ephemeral occupation of a site under a high residential mobility regime to greater permanency may be registered in the matrix of a deposit. Such a process has been identiﬁed in the Pleistocene-aged occupation horizons at Puritjarra. Smith (1989: 99) concluded, a major change in the use of the shelter is clear from changes in the character of the sediments, with large amounts of ﬁnely divided charcoal and ﬁne rock fragments and a more open fabric.

Equally, short duration and low intensity occupation episodes are more likely to leave intact living surfaces due to longer periods of site abandonment and ongoing sedimentation processes and the decreased likelihood of disturbance during subsequent occupations (cf. O’Connor et al. 1993). Shorter site visits will also result in lower indices for artifact breakage due to treadage, as has been demonstrated by Hiscock (1988) for the Colless Creek sites.

Changes in Ocher Provenance The study of ocher distributions from Puritjarra dating to the terminal Pleistocene demonstrates changing provenance (Smith et al. 1998). From 32,000 to 13,000 bp

_____________________________________________________________________________________________________

111

______________________________________________________________________________

Pe t e r V eth

the red ocher source is located 125 km across major duneﬁelds of the Western Desert. The occupants of Puritjarra were either moving over large portions of the desert lowlands to obtain ocher or were in contact with groups that had access to this quarry. The pattern changes signiﬁcantly after 13,000 bp and particularly between 7,500 and 1,000 bp, with the ocher supply shifting to a much closer source located in the central Australian ranges. More intensive use of the rock shelter and increased quantities of ocher are registered after 7,500 bp. Smith et al. (1998: 287) synthesize both the changes in ocher source and other archaeological signatures of territoriality to conclude: If we see the production of pigment art . . . at the site as reﬂecting an increasing need to assert corporate rights and relationships to the site, and the more restricted catchment reﬂected in the ochers as reﬂecting an actual reduction in residential mobility, we get a picture of a shift from an open spatially extensive pattern of land use to a more closed system with smaller group territories after 13,000 bp.

Continuity of Occupation at Sites The three Western Desert sites all evidence a decrease, if not an absence, of occupational material during the LGM (O’Connor et al. 1998: 16; Smith 1988: table 4.4; Thorley 1998a: 239). At least two of the sites are located adjacent to permanent water sources, a rare commodity in the Western Desert at any time of its human occupation. The expectation for groups continuing to occupy a region as conditions became harsher and semi-permanent water sources are lost would be for greater tethering at permanent waters. This is not the case for any of the three sites. As Thorley (1998a: 323) concludes: As a site near permanent water, Puritjarra would be expected to show indications of more intensive occupation around the time ephemerally watered sites such as Kulpi Mara ceased to be occupied. This would be consistent with a model of populations falling back on more reliable ‘‘refuge’’ sites. Yet like Kulpi Mara, Puritjarra also shows a slowing of sediment and artifact deposition rates during the glacial maximum.

The late glacial record from all three sites sits more comfortably with groups who have restricted their territorial range rather than those who continue to occupy a landscape employing a high residential mobility strategy.

Implications of Mobility Indicators In this chapter I have examined eight different lines of archaeological evidence which may be used to infer residential mobility from Pleistocene assemblages in the Western Desert. Such assemblages are relatively intractable when compared to similar aged assemblages from tropical northern Australia. Clearly, multiple lines of evidence are preferable and can act as potential cross-checks on each other. The Pleistocene assemblages conform to the behavioral and archaeological predictions

_____________________________________________________________________________________________________

112

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y

of Ambrose and Lorenz (1990: 10), where the archaeological correlates of huntergatherer behaviors in an unpredictable and scarce resource regime are deﬁned. I have argued that the identiﬁcation of changes in conﬁgurations of residential mobility is central to understanding variability in the assemblages of these three Western Desert sites. Assumptions that early occupation was by small and highly mobile groups ﬁnds empirical support in this analysis. Independent lines of evidence for resource stress, such as the Index of Fragmentation, help to calibrate differences in resource scarcity and hence stress from different regions. Detailed analyses of the intensity of stone reduction, the provenance of stone artifacts and ocher, the diversity of tools and art, and sedimentary records all have the potential to inform on mobility strategies. It is clear from this review that ﬂexibility in mobility patterns, including territorial abandonment and strategy switching, will always have been a necessary part of risk minimization in Western Desert foraging behavior.

References Ambrose, S. H. and Lorenz, K. G. 1990: Social and ecological models of the Middle Stone Age in southern Africa. In P. Mellars (ed.), The Emergence of Modern Humans, Edinburgh: Edinburgh University Press, 3–33. Andrefsky, W. 1998: Lithics: Macroscopic Approaches to Analysis. Cambridge: Cambridge University Press. Binford, L. R. 1980: Willow smoke and dog’s tails: Hunter-gatherer settlement systems and archaeological site formation. American Antiquity, 45, 4–20. Binford, L. R. and Stone, N. M. 1985: ‘‘Righteous rocks’’ and Richard Gould: Some observations on misguided ‘‘debate.’’ American Antiquity, 50, 151–3. Cane, S. 1984: Desert camps: A case study of stone artifacts and Aboriginal behavior in the Western Desert. Unpublished PhD thesis, Canberra: The Australian National University. Conkey, M. W. 1980: The identiﬁcation of hunter-gatherer aggregation sites: The case of Altamira. Current Anthropology, 21, 609–30. Dix, W. 1977: Facial representations in Pilbara rock engravings. In P. J. Ucko (ed.), Form in Indigenous Art, London: Australian Institute of Aboriginal Studies Prehistory and Material Culture Series no. 13, Duckworth, 277–85. Edwards, D. A. and O’Connell, J. F. 1995: Broad spectrum diets in arid Australia. Antiquity, 69, 769–83. Galt-Smith, B. 1997: Motives for motifs: Identifying aggregation and dispersion settlement patterns in the rock art assemblages of central Australia. Unpublished BA (Hons) thesis, Armidale: University of New England. Gorecki, P., Grant, M., O’Connor, S., and Veth, P. 1997: The morphology, function, and antiquity of grinding implements in northern Australia. Archaeology in Oceania, 32, 141–50. Gould, R. A. 1978: Archaeological signatures of stress in the Australian Desert. Los Angeles: Unpublished paper presented at the 1978 Annual Meeting of the American Anthropological Association. Gould, R. A. 1980: Living Archaeology. Cambridge: Cambridge University Press. Gould, R. A. 1985: The empiricist strikes back: Reply to Binford. American Antiquity, 50 (3), 638–44.

_____________________________________________________________________________________________________

113

______________________________________________________________________________

Pe t e r V eth Gould, R. A. 1991: Arid land foraging as seen from Australia: Adaptive models and behavioral realities. Oceania, 62, 12–33. Gould, R. A. 1996: Faunal reduction at Puntutjarpa rock shelter, Warburton Ranges, Western Australia. Archaeology in Oceania, 31 (2), 72–86. Haley, M. 1999: Kaalpi: Investigation of archaeological assemblages from the Calvert Ranges Western Desert, Western Australia. Unpublished BA (Hons) thesis, Townsville: James Cook University. Hayden, B. 1976: Australian Western Desert lithic technology: An ethno-archaeological study of variability in material culture. Unpublished PhD thesis, Toronto: University of Toronto. Hayden, B. 1979: Paleolithic Reﬂections. Canberra: Australian Institute of Aboriginal Studies. Hiscock, P. 1988: Prehistoric settlement patterns and artifact manufacture at Lawn Hill, northwest Queensland. Unpublished PhD thesis, St Lucia: University of Queensland. Hiscock, P. 1994: Technological responses to risk in Holocene Australia. Journal of World Prehistory, 8 (3), 267–92. Hiscock, P. 1996: Mobility and technology in the Kakadu coastal wetlands. Indo-Paciﬁc Prehistory Association Bulletin, 15 (2), 151–7. Hiscock, P. and Veth, P. 1991: Change in the Australian Desert Culture: A reanalysis of tulas from Puntutjarpa. World Archaeology, 22, 332–45. Kelly, R. L. 1992: Mobility/sedentism: Concepts, archaeological measures, and effects. Annual Review of Anthropology, 21, 43–66. Kelly, R. L. 1995: The Foraging Spectrum. Washington, DC: Smithsonian Institution Press. Mabbutt, J. A. 1977: Desert Landforms vol. 2. Canberra: The Australian National University Press. Morse, K. 1994: West Side Story: Towards a prehistory of the Cape Range Peninsula. Unpublished PhD thesis, Perth: University of Western Australia. O’Connor, S., Veth, P., and Barham, A. 1999: Cultural versus natural explanations for lacunae in Aboriginal occupation deposits in northern Australia. Quaternary International, 59, 61–70. O’Connor, S., Veth, P., and Campbell, C. 1998: Serpent’s Glen rock shelter: Report of the ﬁrst Pleistocene-aged occupation sequence from the Western Desert. Australian Archaeology, 46, 12–21. O’Connor, S., Veth, P., and Hubbard, N. 1993: Changing interpretations of postglacial human subsistence and demography in Sahul. In M. A. Smith, M. Spriggs, and B. Fankhauser (eds), Sahul in Review: The Archaeology of Australia, New Guinea, and Island Melanesia, Canberra: The Australian National University, 95–105. Rolland, N. and Dibble, H. L. 1990: A new synthesis of Middle Paleolithic variability. American Antiquity, 55 (3), 480–99. Ross, A., Donnelly, T., and Wasson, R. 1992: The peopling of the arid zone: Human–environment interactions. In J. Dodson (ed.), The Naı¨ve Lands, London: Longman, 76–114. Shott, M. J. 1986: Technological organization and settlement mobility: An ethnographic examination. Journal of Anthropological Research, 42, 15–51. Shott, M. J. 1989: On tool class use lives and the formation of archaeological assemblages. American Antiquity, 54 (1), 9–30. Smith, M. A. 1986: The antiquity of seed grinding in central Australia. Archaeology in Oceania, 21, 29–39. Smith, M. A. 1988: The pattern and timing of prehistoric settlement in central Australia. Unpublished PhD thesis, Armidale: University of New England.

_____________________________________________________________________________________________________

114

______________________________________________________________________________

C y c l es o f A r i d i t y a n d H u m a n M o b i l i t y Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Smith, M. A. 1996: Prehistory and human ecology in central Australia: An archaeological perspective. In S. R. Morton and D. J. Mulvaney (eds), Exploring Central Australia: Society, Environment and the 1894 Horn Expedition, Chipping Norton: Surry Beatty and Sons, 61–73. Smith, M. A., Fankhauser, B., and Jercher, M. 1998: The changing provenance of red ocher at Puritjarra rock shelter, central Australia: Late Pleistocene to present. Proceedings of the Prehistoric Society, 64, 275–92. Smith, M. A., Prescott, J. R., and Head, M. J. 1997: Comparison of 14C and luminescence chronologies at Puritjarra rock shelter, central Australia. Quaternary Science Reviews, 16, 1–22. Thorley, P. B. 1998a: Shifting location, shifting scale: A regional landscape approach to the prehistoric archaeology of the Palmer River catchment, central Australia. Unpublished PhD thesis, Darwin: Northern Territory University. Thorley, P. B. 1998b: Pleistocene settlement in the Australian arid zone: Occupation of an inland riverine landscape in the central Australian ranges. Antiquity, 72, 34–45. Thorley, P. B. and Gunn, B. 1996: Archaeological research from the eastern border lands of the Western Desert. Canberra: Unpublished paper presented at the Western Desert Origins Workshop, Australian Linguistic Institute. Tonkinson, R. 1991: The Martu Aborigines: Living the Dream in Australia’s Desert. Fort Worth, TX: Holt, Rinehart, and Winston. Varien, M. D. 1999: Sedentism and Mobility in a Social Landscape. Tucson: University of Arizona Press. Veth, P. 1987: Martujarra prehistory: Variation in arid zone adaptations. Australian Archaeology, 25, 102–11. Veth, P. 1989: Islands in the interior: A model for the colonization of Australia’s arid zone. Archaeology in Oceania, 24, 81–92. Veth, P. 1993a: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory, Archaeology Series no. 3, Ann Arbor, MI. Veth, P. 1993b: The Aboriginal occupation of the Montebello Islands, northwest Australia. Australian Aboriginal Studies, 2, 39–50. Veth, P. 1995: Aridity and settlement in northwest Australia. Antiquity, 69, 733–46. Veth, P. 1996: Current archaeological evidence from the Little and Great Sandy Deserts. In P. Veth and P. Hiscock (eds), Archaeology of Northern Australia: Regional Perspectives, Tempus: Archaeology and Material Culture Studies in Anthropology no. 4, Anthropology Museum, University of Queensland, St Lucia, 50–65. Veth, P. 2000: Origins of the Western Desert language: Convergence in linguistic and archaeological space and time models. Archaeology in Oceania, 35 (1), 11–19. Veth, P. in prep.: Durba Springs and the emergence of the Wati language. Veth, P. and O’Connor, S. 1996: A preliminary analysis of basal grindstones from the Carnarvon Range, Little Sandy Desert. Australian Archaeology, 43, 20–2. Veth, P., Smith, M., and Haley, M. 2001: Kaalpi: The archaeology of a sandstone outlier in the Western Desert. Australian Archaeology, 52, 9–17. Yellen, J. E. 1976: Long-term hunter-gatherer adaptation to desert environments: A biogeographical perspective. World Archaeology, 8 (3), 262–74.

_____________________________________________________________________________________________________

115

______________________________________________________________________________

--------------------------------------------------------------------------------

7

----------------------------------------------------------------------------

Archaic Faces to Headdresses The Changing Role of Rock Art Across the Arid Zone Jo McDonald

Introduction The art systems of the Australian arid zone have long been thought of as culturally and stylistically homogeneous. Rock art, sand paintings, shields, body painting, and headdresses have been seen as different expressions of the same graphic systems from a region which is essentially culturally homogeneous (Tonkinson 1991). The rock art, in particular, is seen as a longstanding graphic tradition (Rosenfeld 2002). The Western Desert peoples engaged in a settlement pattern marked by high residential mobility across vast territories. This kind of mobility was dynamic. The environment required periodic spatial dispersal and aggregation. This necessitated maintaining relationships with people in one’s own residence group as well as with those from ‘‘far away’’ (Myers 2002). This extreme ﬂexibility in mobility is seen to underpin a raft of risk minimizing strategies in the Western Desert (Veth 2000a). It is generally understood that this type of mobility was accommodated by widely ramiﬁed open social networks with institutionalized kinship, marriage exchange, and ceremonial organization, and reciprocal rights to country. These different institutions of identity and exchange are the media for sustaining political relations over space and time in conditions that combine aggregation and dispersal (Myers 2002: 47). A common signal for this type of mediated social network would be the homogeneity of the arid zone art systems (Jochim 1983; Munn 1966; C. Smith 1989; Wiessner 1990). The central Australian and Western Deserts encompass lands which fall within the ‘‘Western Desert Cultural Bloc,’’ its people all speaking the Western Desert language (see Figure 7.1). The current linguistic model (McConvell 1996) sees this language moving from northwest to east across the arid zone possibly as recently as 1,500– 1,000 years ago (Veth 2000b). It has been argued (Veth and McDonald 2002) that

_____________________________________________________________________________________________________

116

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

alliance networks must have already been in place throughout the Western Desert to enable the rapid transmission of this language. By 5,000 years ago, then, the domestic and totemic landscapes would have been established across all landscapes and habitats of the Western Desert. Implicit in this proposition is that such long distance exchange networks were present well before the Holocene, and could have been in operation from the earliest period of occupation ca. 30,000 bp. Australia is the most arid continent on earth and the Western Desert arguably its most marginal populated landscape. A paucity of permanent water, absence of coordinated drainage, high temperatures and patchy resources are factors in this marginality. Despite this, we know that humans have lived in this region for at least 25,000 years. Key sites such as Serpent’s Glen, Puritjarra, Katampul, and further aﬁeld at Kulpi Mara and Allen’s Cave (see Figure 7.1) have evidence for occupation between 30,000–22,000 bp. These sites have varying records for subsequent occupation during the peak aridity of the Last Glacial Maximum (LGM). All sequences, however, register occupation from 5,000 years ago and at the majority of stratiﬁed

Jalijbang

tC

k.

KIMBERLEY PLATEAU

TANAMI DESERT

Forte

sue

ey R.

MACDONELL RANGES

R.

HAMMERSLEY RANGES Ashb urto nR i v er Gascoyne River

Lake Disappointment

Cleland Hills

Durba Springs Kaalpi

GIBSON DESERT

Puritjarra Kulpl Mara

Serpents Glen

Alice Springs

SIMPSON DESERT

Puntutjarpa

iver

GREAT VICTORIA DESERT

M

ur

ch iso nR

St

ur

De Gr

Burrup Penninsula

GREAT SANDY DESERT

N U L L A R B O R

P L A I N

PERTH

Approximate boundary of the Western Desert 0

1000 km

Figure 7.1 The Western Desert in the context of the Australian arid zone (after Veth 2000b).

_____________________________________________________________________________________________________

117

______________________________________________________________________________

Jo McD o n a l d

sites there are increases in discard rates and changes in lithic technology during the last 2,000 years through to the contact period. All arid zone researchers agree that Aboriginal culture has undergone signiﬁcant transformation during this recent period (O’Connor et al. 1998; Smith 1996; Thorley 1998; Veth 1993). Ongoing archaeological research in the Western Desert indicates an early date for occupation of all landscapes by groups employing a high residential mobility pattern. Dynamism in both residential mobility and the nature and size of territories is assumed for much of this history. While there is some evidence for long distance exchange from the terminal Pleistocene onwards (e.g., pearl and baler shell from the coast; see Ackerman and Stanton 1994), there are signs of increasing circumscription of territory by the early to mid-Holocene (see Table 7.1). These trends are supported by ocher distribution patterns (Smith et al. 1998) and the efﬂorescence of group identifying behavior as inferred from the in situ development of regionally unique localized pigment art styles (Galt-Smith 1997; Gunn 1995, 2000a). A number of workers have now argued for changes in the function and intensity of occupation in major habitation sites during the last 2,000 years. Across the Western Desert for much of the year, groups are highly nucleated and dispersed across vast areas of country, and there is evidence for this being longterm behavioral patterning. Major aggregation events occur during times of resource stress, when group territories contract into refugia (Veth 1993), or for major ceremonial activities during periods of resource abundance. At such times, it has been predicted that there will be a greater diversity of activities, both everyday economic, as well as religious and ceremonial. These types of sites occur where there are permanent water sources and where there is a range of other resources nearby. Recent work at one such location indicates a long period of art production and a changing graphic structure over the period of this area’s occupation. The art of the Calvert Ranges – an outlier in the midst of the vast sand dunes of the Little Sandy Desert – will be used as exemplar of the issues raised in this chapter (see Figure 7.1). Table 7.1 Occupation phases and key elements for the Western Desert (after Veth 2000b). Phase 1 2

3 4 5

Features

Timing

Early colonization of Western Desert; all landscapes used with high residential mobility Period of instability in population levels; subregional abandonment and/or more opportunistic use of marginal lands Onset of climatic amelioration; territorial expansion from ‘‘refuges’’ Occupation of all desert ecosystems; (re)establishment of information networks Increased intensity of occupation; accelerated ritual and ceremonial cycles

>25,000–22,000 bp

_____________________________________________________________________________________________________

118

22,000–13,000 bp

13,000–5,000 bp 5,000 bp to contact 1,500 bp to contact

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

In this chapter I am interested in tracking diachronic changes in the artistic systems of the arid zone. I will explore how the role of art has transmogriﬁed along with changes in stone tool technologies and subsistence strategies. The ways that art might be used to recognize changing mobility and territoriality are discussed. To demonstrate the changing graphic system over time and space, I will focus on a particular engraved ﬁgurative motif – the so-called ‘‘archaic face’’ – and look at its distribution, likely antiquity, and possible role within the Western Desert graphic system. The implications of a ‘‘shared’’ motif over a very extensive area and through many millennia will be examined. The way that this motif has been transformed, both into engraved anthropomorphic motifs and painted forms, is discussed. The development of headdresses as localized group identifying behavior across various arid desert and coastal style regions is shown to be part of the in situ development of local group identifying behaviors and a traceable factor in the changing graphic from the Panaramitee to a myriad of localized, mostly ﬁgurative styles.

The Arid Zone Art Graphic It is generally understood that Aboriginal painters from the arid zone paint their Dreamings (Gunn 1995, 2003; Kimber 1977; Layton 1992; Megaw 1982; Myers 1989, 2002; Spencer and Gillen 1899). Painters involved in the production of Western Desert acrylic art state that their paintings represent the Dreaming (Tjukurrpa) and their country (ngurra). They insist that the representations and images that they use ‘‘come from the Dreaming’’ (tjukurrtjana), and are more valuable than anything humans might invent (Myers 2002: 32–3). These paintings are representational and make use of a system of graphic signs: Munn’s ‘‘visual categories.’’ The iconic elements of this system are circles, arcs, lines, and meanders. In the Warlpiri graphic systems each visual element conveys a number of possible meanings. The ability to interpret the meaning depends on having the knowledge of relevant mythological information. It has been argued that this ambiguity provides a degree of this art’s aesthetic force (Morphy 1991). We have a relatively good appreciation of how rock art functioned in the recent past, based on solid anthropological and ethnographic evidence. The production of art is known to have taken place in a number of social contexts – from secular and casual, to sacred and ceremonial (Gunn 2000b). A number of researchers have noted the interconnectedness of art (or graphic) systems with other social institutions in this arid environment where the social networks were widely ramiﬁed. The interdependent relationship between the environment, belief systems, social structure, politics, and graphic system served to integrate and reinforce knowledge of the landscape into a holistic body of meanings; the graphic systems interweave the social, physiographic, and spiritual knowledge of society (Chewings 1909; Frederick 2000; Gillen 1968; Layton 1986, 1992; Myers 1986, 2002; C. Smith 1989). All members of society, regardless of levels of practice and cognizance of encoded meaning, have some involvement in the production and maintenance of the graphic system (Morphy 1989, 1991; Munn 1973).

_____________________________________________________________________________________________________

119

______________________________________________________________________________

Jo McD o n a l d

The graphic systems of the Western Desert were an important tool for promoting and controlling the exchange of information. They functioned at a multitude of levels, to identify and integrate as well as to demarcate social boundaries (David and Chant 1995; Dickens 1996; Frederick 2000; Galt-Smith 1997; Jestribek 1997; C. Smith 1989; and in other regions Lewis 1988; McDonald 1994, 1999). The application of a graphic system in the production of rock art served a number of functions and operated within a variety of social contexts (Frederick 2000). It was a way of marking place and concomitantly an individual’s afﬁliation to it; it had a signiﬁcant storytelling or instructive context; it was recorded as being used in initiation ceremonies (Spencer and Gillen 1899), and also often provided a physical manifestation of an ancestral being or event. Rock art was also a way of ‘‘marking country’’ by the identiﬁcation and naming of nodes in the social geography (Rosenfeld 2002).

Arid Zone Rock Art Research Rock art research in the Australian arid zone has had a recent efﬂorescence with a number of rock art specialists commencing or expanding research interests in this area (Frederick 2000; Galt-Smith 1997; Gunn 1995, 2000b; Rosenfeld 1993, 2002; Rosenfeld and Smith 2002; Ross 2003). There have also been a number of recent archaeological analyses of other arid zone graphic systems which have broadened our understanding of the social context of art production (Dickens 1996; Jestribek 1997; C. Smith 1989). Revealed by this increased research activity is the amount of localized stylistic variability apparent in rock art, particularly in the recent past, across the arid zone. This had not been documented previously, and like early views of Aboriginal occupation of this continent, people had identiﬁed only homogeneity and assumed a broad continuity and unchanging nature of rock art production across the arid zone (e.g., Dix 1977; Edwards 1968; Gould 1969; Maynard 1977; McCarthy 1979). Spencer and Gillen (1899) recorded that the primary difference between ‘‘ordinary and sacred’’ rock art was the location in which they were painted (Gunn 2000b, 2003). They commented: ‘‘motifs painted in sacred places were sacred: those painted elsewhere were not, even if visually the two motifs were identical’’ (Spencer and Gillen 1899: 618). Rosenfeld (2002) has pointed to a difference in design structure as indicative of which art was socially restricted. The Emily Gap (Ntherrke) painting site (which has recently had its restricted access lifted to allow public visitation) comprises a series of large formally structured bichrome designs, which Gunn (2003) calls Dreaming (or atywerrwenge) designs. Rosenfeld notes that striped designs such as these mark localities with powerful mythological signiﬁcance to informants (Stirling 1896: 67). Projecting the mythological signiﬁcance of motifs into the archaeological past is of course the nub of the problem (see Gunn 2003). This problem is of relevance when it comes to interpreting the social messaging potential of art in its social context (e.g., Wiessner 1990). Galt-Smith’s (1997) work on aggregation locales in central Australia demonstrates that the social context of the art’s production does affect social messaging.

_____________________________________________________________________________________________________

120

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

For instance, pigment art (and engravings) that occur in rock shelters (or open contexts) where there is other evidence for domestic habitation (stone tools, hearths, grinding stones or patches, and so on) was art largely produced in a public context. On the other hand, pigment and engraved art less tangibly connected with habitation sites may equate with a less public art medium (see McDonald 1999). Of course, through time, sites may have periods where they were closed access and other episodes of open context. The longevity of rock art sites means that these can demonstrate changing social contexts and record past social geographies (Rosenfeld 2002; Rosenfeld and Smith 2002). Both the archaeological evidence, from occupation deposits and an art graphic that varies over time, can demonstrate such changes.

Rock Art Chronology Determining the age of rock art across the arid zone has not been well researched at this time, and we have few dates to help in establishing a ﬁrm chronology. There are, however, a number of dates from surrounding regions, including a small number of direct dates on art motifs and several dated art contexts. Broad temporal trends can be deﬁned on the basis of the excavated evidence. Fairly recent developments have resulted in the successful use of Accelerator Mass Spectrometry (AMS) to date very small samples of carbon directly from rock art motifs (David et al. 1999; McDonald 2000; Watchman et al. 1997). Carbon is found within the actual art matrix (e.g., charcoal, organics in the binders) or as organics within skins which have formed on top of the art (e.g., oxalate crusts and desert varnish). Optically Stimulated Luminescence (OSL) has also been used in the Kimberley region to date mud-wasp nests which have been constructed on top of (Bradshaw style) pigment motifs (Roberts et al. 1997). We can assume that the pigment art of the arid zone is mostly relatively recent. Pigment and ochers generally do not have the same long-term survival potential presented by engravings as these materials fade and weather (although the probability of red pigment surviving is greater; see Cook et al. 1990; Roberts et al. 1997) and many rock shelter surfaces are unstable. Engravings, on the other hand, unless created on very soft friable surfaces, can survive for considerable periods of time. Most of the arid zone bedrock consists of case-hardened rocks. High levels of iron in the soils and extreme temperatures mean that the greatest threat to most engraved art comes from cracking of the matrix over time. At Puritjara, the use of ocher has been dated back to ca. 32,000 bp, and the earliest period of the site’s occupation. It has been argued that pigment use here was initially for body painting and that rock art production only commenced after 13,000 bp when a prevalence of ochers was located close to the back wall of the shelter (Smith et al. 1998; Rosenfeld and Smith 2002). While much of the arid zone pigment art is arguably recent, very old pigmented motifs are found in locations where there are good conditions for preservation. The Bradshaw ﬁgures in the Kimberley region have been dated to greater than 25,000 bp (Roberts et al. 1997)

_____________________________________________________________________________________________________

121

______________________________________________________________________________

Jo McD o n a l d

and pigment on buried rock fragments at the Carpenter’s Gap rock shelter has been dated to >40,000 bp (O’Connor 1995). Pecked and highly weathered Panaramitee style engravings (after Maynard 1977) are found throughout virtually the entire arid zone. This longstanding graphic tradition comprises circles, tracks, lines, and arcs, and the occasional ﬁgurative motif – the same visual categories identiﬁed by Munn (1973) in contemporary Warlpiri iconography and recent arid engraving traditions, albeit with the addition of a number of new motifs (Rosenfeld 2002). It has generally been considered that the pecked petroglyphs of the Panaramitee date to the Pleistocene. Engravings of this style have been dated (elsewhere) to 30,000 bp (Nobbs and Dorn 1988, 1993; although see Watchman 1992), and the weathering condition and patina of many of these sites indicate that they are indeed ‘‘ancient’’ in their production. Regional variants of this style are found elsewhere across the continent, with eastern variants being found in Laura (Queensland) and Sydney (New South Wales), where they have been found to predate the more recent regional art styles (McDonald 1998; Rosenfeld et al. 1981). It has been maintained that this multivalent and largely iconic art system connected groups over hundreds and indeed thousands of kilometers. Certainly, this graphic system crosses currently recognized geographic, language, and cultural boundaries. Increasingly apparent as a result of detailed current research is that across the arid zone recent rock engraving styles have also developed. These are largely ﬁgurative, but still contain track and geometric components. While not dated, it is assumed (on the basis of regional coastal trends) that these are a Holocene development. In those areas where there has been recent research (e.g., the Calvert Ranges, Burrup Peninsula, and the western portion of the central Australian ranges) these engraving styles can be seen as associated with developing localized pigment traditions. Not all recent pigment traditions, particularly those in central Australia, however, have associated highly developed engraved components (Ben Gunn, pers. comm.). This, too, would appear to be one of the traits of more recent localized artistic behavior – a mosaic of pigment and engraved components. While there is a paucity of direct dating, it is clear that art has probably been produced throughout the arid zone, from the time of its earliest occupation (at least ca. 30,000 bp), in both engraved and pigment forms. While there is little evidence of ancient pigment art surviving on rock surfaces across the arid zone proper (cf. the Bradshaws in the Kimberley region), the evidence from Carpenter’s Gap and Puritjarra indicates that pigments were being used on rock surfaces early and that this may have increased substantially by 13,000 bp. While we have no direct dates from the arid zone proper for engraved art, dates in excess of 30,000 bp for the oldest Panaramitee art style are posited.

Correlating Rock Art with Other Arid Zone Land Use Patterns So how can patterns in rock art production throughout the arid zone be correlated with archaeological models and linguistic evidence across the arid zone?

_____________________________________________________________________________________________________

122

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

‘‘Centralian’’ and Western Desert art are rarely linked (e.g., Gould 1969), although researchers recognize that the Western Desert Cultural Bloc, which describes the extent of this speciﬁc language area, transcends other geographic divisions (that is, the more marginal Western Desert compared with the better watered central uplands). Numerous researchers have identiﬁed close similarities in both the style and execution of Western Desert and Pilbara engravings and have documented mythological narratives which clearly link the two areas and their art (e.g., Dix 1977; Maynard 1977; Tonkinson 1991; Veth 2000b). While sharing a number of traits, distinctive styles are found on the Burrup Peninsula, within the central Pilbara uplands, and at several ranges along the Canning Stock Route. Engraved gracile anthropomorphic ﬁgures, elaborate headdresses, and a range of painted and engraved naturalistic motifs and geometrics would appear to link these areas in a stylistic sense. The Wandjina/Bradshaw styles of painting from the southern Kimberley region do not occur in the Western Desert, although it is notable that many of the early Wandjina motifs bear striking similarity to the archaic faces found across the arid zone (e.g., Crawford 1977; and see below). The mythological sagas of Wati Kutjarra (the two men), Papa (dingo), and Seven Sisters (Pillades), to name a few, connect speciﬁc sites (often with art) between the Pilbara and the Western Desert (cf. Tonkinson 1991; Peter Veth, pers. comm.). These narratives do not necessarily extend to the central Australian ranges or the Kimberley. While all the art regions being discussed still require reliable dating, it is reasonable to assume that the recent ‘‘maintained’’ art sites, and those forming an integral part of the mythological narrative, can be treated as recent, and connected. The distribution of these sites in the recent period gives us insight into the scale of the interrelatedness of sites throughout prehistory. While there are early connections evident throughout the northwest and arid zone proper, throughout time there has been a changing focus of contacts into and across the arid zone. Based on regional patterns in stratiﬁed archaeological sites and shared cultural attributes, Veth (2000b) has shown that the Western Desert language probably originated in the vicinity of the Pilbara uplands and that its spread occurred in the late Holocene (see also Chapter 12, this volume). This conclusion supported the linguistic models (e.g., McConvell 1996) which indicated that the homeland of Wati was located in or near the Hammersley (Karijini) Ranges and that the homeland of Western Desert was very close to the Calvert Ranges (see Figure 7.1). The linguists argue that proto-Wati moves out of the Hammersley Ranges after 3,000 bp, with initial Western Desert unity in the southern Pilbara dating from about 2,000 bp. It is proposed that with the spread of the Western Desert language came an increased intensity of site occupation, an accelerated ritual and ceremonial cycle, and an increase in long distance exchange. The proposed correlation of the linguistic and archaeological data gives a clear indication of how you might model the role of art within this changing social system (see Table 7.2). It is likely that the ramiﬁed social networks described by current anthropological research have functioned for at least 1,000–1,500 years and possibly for as much as 5,000 years (Veth 2000a, 2000b).

_____________________________________________________________________________________________________

123

______________________________________________________________________________

>22,000 bp

22,000–13,000 bp

13,000–5,000 bp

5,000–1,500 bp

1,500–500 bp

500 bp to contact

2

3

4

5

6

Timing

1

Occupation phase

Western Desert speakers encroach into central Australia

Spread of Western Desert languages; loan words from northern languages; some evidence of Arandic contact

Pama-Nyungan occupation of Western Desert

Non-Pama-Nyungan speakers

Linguistic correlations

Increased interaction with social networks in central Australia

Early colonization; all land systems in use; broadly based economy Changes in residential patterns; shifts in demography (LGM); lowlands used more opportunistically Climatic amelioration; marginal lands used more systematically Occupation of all desert ecosystems; reestablishment of regional exchange/information networks Increased intensity of site occupation; accelerated ritual and ceremonial cycle; increase in long distance exchange

Occupation model

Table 7.2 Occupation phases and language movements (after Veth 2000b) and likely art correlates.

Increased use of art to negotiate broad scale and local group identity with distinctive localized style regions evolving – art inﬂuences at this time come from the north As above – art inﬂuences also appear from the east and further aﬁeld

Art used to negotiate broad scale and local group identity

Broad scale social cohesion with perhaps increased localized identifying behavior – territorial tethering

Sporadic art production; widespread group cohesion

Likely art correlate

Ar c h a i c Fa c e s t o H e ad d r e s se s

It is also plausible that a tightening of social and territorial organization occurred just before the LGM following a long phase of high mobility and multivalent art. Groups could have become more tethered to uplands like the Calvert Ranges due to intensiﬁed aridity. A likely consequence of this would have been an inevitable intensifying process in those locations where occupation was maintained. Using this model, and information exchange theory, it is proposed that art was used to negotiate identity within this time frame. It is likely that networks functioning across the arid zone throughout the Pleistocene were more open and far reaching than observed at present, reﬂecting an extreme form of residential mobility (Veth 2000b). It is plausible that the initial occupation of the region was accompanied by the use of rock art. Low intensity, sporadic art production at this time would have demonstrated widespread group cohesion, resulting in a visually homogenous multivalent art graphic. The Panaramitee art ﬁts within this schema, supported by the various early dates around the continent. Extensive art networks functioned in the European Paleolithic at a similar time. Equally, the very early dates for the Bradshaw (pigment) art of the Kimberley region and art production at Carpenter’s Gap reinforce such a proposition. The addition of the so-called archaic face into the graphic system would appear to date to the earliest period of Panaramitee art. All examples are heavily weathered, highly patinated, and many are geologically altered. Since their earliest documentation, every rock art researcher who has observed these motifs has commented on their obvious antiquity (David et al. 1992; Dix 1977; Edwards 1968; McCarthy 1979; Terry 1932; Walsh 1988). One of the characteristics of the Panaramitee is the small proportion of ﬁgurative motifs which occur among the largely non-ﬁgurative and track repertoire. Interestingly, it is this minor ﬁgurative component of the style which provides insight into what could be classed ‘‘localized style regions’’ among this early art form. The distribution of archaic faces does not reﬂect the overall distribution of Panaramitee rock engravings. In other regions there are other unique forms which contribute to the small proportion of unusual motifs, such as the spoked concentric circles at Sturt’s Meadows (see Walsh 1988: 65). With the LGM and changes in residential patterning (Phases 2 and 3, from 22,000–13,000 bp), it is likely that there was broad scale social cohesion. Increased localized identifying behavior may well have resulted from territorial tethering during this period. Perhaps it was at this time that the large decorative inﬁlled motifs at the Calvert Ranges and localized variants of the archaic faces (many with bodies) were developed across the Western Desert. In the Calvert Ranges this type of motif is geologically old – the tops of the heads of one panel having weathered away. In this location, this style (on the basis of patination and weathering) appears at least as old as the other elements of classic Panaramitee art at this location. Three ancient-looking simple archaic faces (minus bodies) are also found among this assemblage. In other localities within the arid zone and its coastal margins, similar trends within developing local styles appear to have occurred. It is suggested that the use of rangeland systems such as the Calverts may have intensiﬁed with the onset of heightened aridity during the latter phase of the LGM. In coping with increased stress and the imperative to maintain ties to country,

_____________________________________________________________________________________________________

125

______________________________________________________________________________

Jo McD o n a l d

obvious responses would be to make shifts in residential mobility and to signal increasing boundedness as groups’ ranges were reconﬁgured. This is just the sort of catalyst that would create an artistic vocabulary signifying corporate identity (or identity of any kind) in the Western Desert during the Pleistocene (viz Rosenfeld 1993). Prior to the arrival of the Western Desert language, from around 5,000 years ago, it has been suggested that antecedent regional social networks would have developed (Veth 2000b). At this time, art would again have been used to negotiate both broad scale and local group identity. It would still be expected to have been relatively homogenous, given the requirements of ramiﬁed open social networks, although it is likely that there was an increased use of art to negotiate broad scale and local group identity with distinctive localized style regions evolving. Art inﬂuences at this time would be expected to have come from the northwest, coinciding with a range of archaeological and linguistic evidence. This pattern could have continued during the spread of the Western Desert language. In the period between 1,500 and 500 years ago one would expect that the regionally speciﬁc art provinces in operation at contact would have developed. Anthropological and archaeological evidence (e.g., Smith et al. 1998) supports a ﬂow of artistic (social) inﬂuences from the West to the East across the arid zone during the last 500 years. This set of correlates identiﬁes the likely triggers and social mechanisms that would have resulted in a complex graphic system, particularly one involving ﬁgurative motifs (cf. Rosenfeld 1993) developing in some areas before others. It also takes account of the ‘‘pulses’’ of artistic activity which are indicated by the (so far) limited archaeological evidence (e.g., Smith et al. 1998). This suggested patterning for diachronic change in the social context of art production and the likely manifestations of this in the graphic system, illustrating local versus regional patterns of land use, rest largely on interpretations of information exchange networks. The geographic scale at which these arid zone information networks are operating is colossal by world standards (see Gamble 1982), and the nature of the arid zone (with focalized rock art locations interspersed by vast tracts of sand dune country) provides us with a unique opportunity to investigate the dynamics of information exchange. The notion of style as social strategy (following Wobst 1977) is based on stylistic principles of communication processes. The major function of stylistic behavior is seen as linking members of a community who are not in constant verbal contact with each other, making their interaction more predictable and less stressful. ‘‘Style’’ here is deﬁned as the particular way of doing or producing material culture which signals the activity of a particular group of people who distinguish themselves from other, similarly constituted groups. Style is non-verbal communication which negotiates identity (Wiessner 1990). ‘‘Stylistic heterogeneity’’ is perceived here as being the variability in style which demonstrates widely dissimilar artistic components; the end result of doing or producing material culture which signals either a less culturally fettered activity by a certain group of people, or the activity of a certain group which has less rigorous

_____________________________________________________________________________________________________

126

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

stylistic rules. In either case the heterogeneity needs to be ‘‘relative’’; that is, it can only be deﬁned in comparison to more homogeneous stylistic activity. Previous rock art studies using information exchange theory have correlated stylistic variability with the nature of prehistoric social networks. Studies of art in fertile regions, where social networks are closed, have demonstrated a high degree of social information, and distinctive group identifying and bounding behavior (e.g., McDonald 1998). In the arid zone, where there are widely ramiﬁed open social networks, it has been argued that a stylistically homogeneous art form will demonstrate broad scale intergroup cohesion (C. Smith 1989). Current work indicates that not all arid zone art is homogeneous, but that information exchange theory can be used to explain the recent proliferation of localized style regions throughout the arid zone (Veth and McDonald 2002). The model presented here for the changing role of arid zone art throughout different phases of territorial mobility has also been based on these precepts. The rock art of the Calvert Ranges (Figure 7.2) encapsulates the changes in rock art production which have arguably occurred throughout the human occupation of the arid zone. The Calvert Ranges (or Kaalpi) are on the Canning Stock Route, south of Lake Disappointment. They comprise a series of well watered gorges in a small outlier of conglomerate quartz sandstone in the otherwise vast dune ﬁelds of the Little Sandy Desert. This location characterizes one of the more marginal landscapes to have been occupied by people in the Western Desert. The engraved art in this location is stylistically diverse, with time accounting for much of the variability inherent in the overall assemblage. The use of this art province extends back over a vast period of time. Here we ﬁnd intaglio/classic Panaramitee style art of mainly bird and macropod tracks, circles, arcs, and dots. There are cupules (deeply pecked hollows arranged across vertical – or horizontal – surfaces), archaic faces, and small anthropomorphic ﬁgures based around the theme of archaic faces, including decorated body and disembodied archaic face with headdress. There is a suite of very large (>3 m) pecked and decoratively inﬁlled anthropomorphic ﬁgures (these too appear to have considerable antiquity). And there are small inﬁlled terrestrial animals, battered ﬁgurative motifs, pecked outline and inﬁlled animals, and abraded outline ﬁgurative motifs. One site in the Calvert Ranges has been excavated previously (Veth et al. 2001). This decorated living site yielded an engraved plaque with a minimum age of 2,500 bp. At another site found during the current work an engraved motif was observed extending beneath the occupation deposit. A test probe next to the wall in this shelter revealed that the motif – an archaic face with headdress – did indeed extend beneath the deposit. Preliminary dates are equivocal, but one from ca. 20 cm below the current surface suggests that the upper portion of this buried motif was initially covered by deposit around 1,000 years ago (1,050110 bp, Wk-8807). This would suggest the actual production predated this date considerably (that is, enough time to allow at least 40 cm of deposit to accumulate). More extensive excavations are planned to explore the possible age and association of this motif.

_____________________________________________________________________________________________________

127

______________________________________________________________________________

Jo McD o n a l d

N 400 km

Jalijbang2

E

R

LE

Y

S

0

Burrup Penninsula

KI

M

B

Depuch Island Port Hedland P

I

Abydos L

Durba Hills

B

A

R

N. T.

Cleland Hills

Kaalpi

A

S. A.

Perth

Figure 7.2 Distribution of archaic faces across the arid zone.

Archaic Faces Archaic faces have a predominantly arid zone distribution (see Figure 7.2). They are found between the coastal and eastern Pilbara region to Wardaman country in the north, and between Cleland Hills in the east and the Calvert Ranges in the southwest. The site at which this motif type was originally recorded is some 200 km west of Alice Springs (Edwards 1968). The engravings here are found around a semipermanent rock hole in the Cleland Hills. The Cleland Hills site has the largest recorded concentration of this motif (n ¼ 16), although illustrations for only nine have been published (David et al. 1992; Dix 1972, 1977; Edwards 1968; Morwood 2002; Walsh 1988).

_____________________________________________________________________________________________________

128

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

Generally, these engravings have pecked eyes, comprised of dots surrounded by concentric circles, mouths (these are sometimes natural cavities on the rock surface), and some have noses and ears. The face is generally heart- (or pear-) shaped – and there are various examples which have no upper face outline above the eyes (see Figure 7.3). At Cleland Hills there are three examples which also have bodies. One of these comprises simple lines, while the other two are more complex designs. Several are found as part of more complex non-ﬁgurative (CXNF) engraved designs. It is the embodiment of the faces, and development of more complex design structures, which suggests the development of localized stylistic traits – a pattern which can be observed at other sites which also have the simple archaic face present. We do not know how old the archaic faces are, but the extensive geological alteration to the substrate on which they have been engraved and their generally

Figure 7.3 Examples of archaic faces used in the current analysis.

_____________________________________________________________________________________________________

129

______________________________________________________________________________

Jo McD o n a l d

Figure 7.4 Archaic face with body – the development of localized group identifying behavior (an example from the Calvert Ranges).

high degree of weathering is strongly suggestive of at least a terminal Pleistocene age (ca. 25,000–10,000 bp) for many of them. More than 1,000 km north of Cleland Hills is Wardaman country. Here, in a secluded sandstone gorge at a site known as Jalijbang 2, are three large complex pebraded (pecked and abraded) anthropomorphs on vertical sandstone walls. Nearby on large boulders are three pecked, highly weathered faces (David et al. 1992; Lewis and McClausland 1987). Nearby occupation evidence was found to be 4,000–3,000 years old, but none of the engravings was in a directly dateable context. It was concluded that ‘‘their precise antiquity remains a matter for speculation’’ (David et al. 1992: 74). An earlier analysis of the archaic face was undertaken using multidimensional scaling to compare the 21 motifs in the sample (David et al. 1992). This indicated strong regional clustering, although the Jalijbang 2 boulder and wall faces (located within 20 m of each other) are separated signiﬁcantly. Diachronic as well as synchronic variability would appear to account for these differences. For this chapter I analyzed a sample of 39 archaic faces from the Burrup Peninsula (n ¼ 18), Cleland Hills (n ¼ 10), the Calvert Ranges (n ¼ 4), Durba Hills (n ¼ 1), and Jalijbang 2 (n ¼ 6).1 A total of 19 variables were deﬁned and counted on this sample. Of these variables, 14 were binary (presence/absence) and 5 were metrical (see Table 7.3). Details of associated body, headdress, or other CXNF features were not measured, since these traits were not present on all motifs. Because the sample size is relatively small, a simple cluster analysis (using squared Euclidian distance on binary data) was undertaken to identify any grouping within the sample. The analysis found that the simple archaic faces could be classiﬁed on the basis of just six of the non-metrical traits (see Table 7.3). These traits demonstrated a clustering of the archaic faces from Cleland Hills, the Burrup Peninsula,

_____________________________________________________________________________________________________

130

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s Table 7.3 Variables used in the analysis of the archaic faces and codes used for the analyzed sample (see Figure 7.5). Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Codes

Pecked dot eye Concentric circle around eyes Head enclosed by outline? Pear/heart-shaped head Hair or headdress Nose Ears Distinctive mouth Associated body (simple) Intaglio elements Concentric facial elements Neck Associated body (complex) Part of complex-non-ﬁg? Length of head Width of head Distance between center eyes Number of concentric rings (left eye) Number of concentric rings (right eye) Burrup Peninsula Cleland Hills Calvert Ranges Durba Hills Jalijbang 2 wall and boulder

Binary

Metrical

B1-B12 CH1-CH10 CR1-CR4 DH1 JW1-JW3; JB1-JB3

and the Calvert Ranges. When a further two variables are included in the analysis a separation [is achieved] which broadly can be deﬁned on the basis of complexity of association and degree of embellishment (see Figure 7.5). There is a slight separation of the Burrup motifs, but generally there is a development of complexity across the sample area which creates patterning within the sample. And this cannot be explained purely in terms of geographic separation. It is considered signiﬁcant that this analysis did not demonstrate a geographic split and that all archaic faces (as generally deﬁned in the literature), regardless of provenance, clustered together. It is only when we consider the development of the face into full body ﬁgures and/or other associations that geographic emphasis appears to emerge. The absence of clear geographic divisions indicates that there are stylistic similarities between all areas analyzed. This is hardly surprising since the basis for selection within the sample was the presence of an archaic face! Regionalization is, however, present and the observed differences between the style provinces can be summarized as follows:

_____________________________________________________________________________________________________

131

______________________________________________________________________________

Jo McD o n a l d Label

0

5

10

15

20

25

CH7 B17 B18 CH1 JW1 DH1 JB2 B2 JW2 JW3 CH6 CR1 CR3 B8 CH3 JB1 CH2 CH9 CH10 B4 CR2 B11 B12 B10 B13 CH8 B14 CH4 CH5 B6 B7 B3 B5 B15 CR4 B16 JB3 B1 B9

Figure 7.5 Dendrogram showing results of analysis on archaic faces (refer to Table 7.3 for codes).

Burrup Peninsula: Small overall motif size, strong intaglio component, highly complex graphic with strong decorative inﬁll component, highly variable body shapes (human and animal), many faces have hair or headdress (Figures 7.6 and 7.7). Cleland Hills: Incorporation of one face into complex non-ﬁgurative engraved panel (see Figure 7.8), highly variable body shape (simple and more complex), in situ development of simple face (maybe an artifact of sample size).

_____________________________________________________________________________________________________

132

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

Figure 7.6 Burrup Peninsula. Fish motif (B17) with archaic face (courtesy Peter Kendrick).

Figure 7.7 Burrup Peninsula. Two archaic faces (B15 and B16) in complex non-ﬁgurative graphic (courtesy Peter Kendrick).

_____________________________________________________________________________________________________

133

______________________________________________________________________________

Jo McD o n a l d

Figure 7.8 Cleland Hills (CH1) archaic face (courtesy Mike Smith).

Calvert Ranges: Development of pecked complex body association (small and large; see Figure 7.4), incorporation of natural features (e.g., large quartz clasts) into motif design, complex decorative inﬁll, anthropomorphs have hands and feet (some with digits), some have a square head shape incorporating headdress, some have headdresses similar to the local painted varieties. Durba Hills: This single example looks more like several of the Burrup variants, but clusters most strongly with one of the boulder faces from Jalijbang 2. Wardaman Country: The simple pecked and highly weathered faces on the boulders are differentiated from those on the wall by both technique and traits. The three anthropomorphs on the wall are characterized by concentric inﬁll, the use of pebrading (pecking and abrading), and the addition of small koala-like ears; limbs are shown (some incomplete) and ﬁngers indicated. The embellishment of these faces in the various localities in which they are found raises the possibility that the faces are a common and shared theme over a huge area, acting initially to connect groups over long distances. The introduction and variations in body form and inﬁll and headdress style over time serve to signify a development of local identity and territoriality. So what are the likely circumstances that might have resulted in the development of individual identifying behavior within a social system which generally encour-

_____________________________________________________________________________________________________

134

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

Figure 7.9 Calvert Ranges. Panel of more recent anthropomorphs with pecked eyes and intaglio decorative inﬁll.

aged stylistic homogeneity? When does individual identity override the practice of group identity? Wiessner (1989) has modeled the different social situations that might stimulate a different identity emphasis. She proposed that group identity will be emphasized where there is real or potential intergroup competition and/or aggression, and a need for cooperation to achieve goals and impose political control. Individual identity, on the other hand, will dominate when there is real competition between individuals, when there are options for individual economic gain or status, or where there is a breakdown in the social order that would require individuals to seek solutions to their own problems. Similarly, Myers’ (2002: 48) work on contemporary Western Desert acrylic paintings concluded that artists elaborate certain representational templates motivated by the politics of personal and social identity. In a system of dynamic mobility, one might also suggest that periods of aggregation would result in the intense interaction of both groups and individuals seeking to establish themselves in the social order. Aggregation locales, then, are likely sites for stylistic innovation to occur. Across the arid zone the art that we ﬁnd at aggregation sites appears to record the interaction of neighboring (frequently dispersed) groups, but we should also expect to be able to identify innovation in stylistic traits (Conkey 1980). This is exactly what we appear to be seeing in the development of the more complex graphic based on the archaic face. Since their earliest usage, then, aggregation locales have allowed groups to assert identity and difference.

_____________________________________________________________________________________________________

135

______________________________________________________________________________

Jo McD o n a l d

The early graphic form, the archaic face, is generally found at the type of refugia generally assumed to have functioned as arid zone aggregation locales – Cleland Hills, Calvert Ranges, and Durba Hills. This would appear to have had a dual signaling purpose, to both connect and differentiate people. The distribution of this motif not only demonstrates that the ancient pecked Panaramitee art provided an essential underpinning for groups with extreme forms of residential and territorial mobility, but it also shows us that the likely range of this connection was over thousands of kilometers. The widespread presence of the archaic face – a clearly identiﬁable and relatively uniform graphic element – so early in the history of occupation of the Western Desert suggests groups were in contact over vast tracts of country. It also suggests that long distance arid zone information exchange networks appear to have considerable antiquity. Similar evidence occurs in the terminal Pleistocene from other archaeological evidence (e.g., the movement of ochers around the arid zone; Smith et al. 1998), in the late Holocene (e.g., exotic lithics such as Wingelina porphyry at Puntutjarpa; Gould 1977), and during the last 2,500 years with pearl shell and baler (from the Pilbara and/or Kimberley coastline) at rainmaking and ceremonial sites along the Percival Lakes and Canning Stock Route (Ackerman and Stanton 1994; Smith and Veth, 2004). The well watered gorges of the Calvert Ranges have evidence of extraordinary stylistic heterogeneity in the recent art (pigment and engraved) as well as demonstrating long-term diachronic change. The high degree of stylistic variability displayed here indicates that this place has acted as an aggregation locale over a considerable period of time [(Veth and McDonald 2002)]. We know that aggregation sites have served as important centers of ritual production, as well as facilitating the rapid exchange of language, material culture, and genes. During their periodic coalescences groups would renegotiate social contracts and reciprocity relations that set the necessary conditions for subsequent dispersal (Gibbs and Veth 2002). Aggregation locales in the Western Desert provide vital evidence for long-term patterns of mobility and exchange across the arid zone. They provide the foci in the landscape where changes in economy, intensity of site use, and the graphic systems signifying both local identity and long distance alliances are demonstrated. It is in these locations that changes in a complex and far reaching social system can be observed. Multiple phases of occupation at arid zone aggregation locales are most clearly demonstrated by multiphase rock art assemblages. This analysis of archaic faces has been based on their widespread distribution and assumption of great antiquity and the subsequent regional development of different stylistic signatures. It is suggested that there has been a change in the signifying role of rock art across arid zone social systems throughout its 30,000þ years of human occupation. The archaic face, developed out of a representational graphic into a simple but evocative ﬁgurative form, appears to have served a dual role. It initially connected people within a huge cultural bloc. Its ﬁgurative nature then allowed the development and demonstration of individual and local group identity. The development of individual traits and innovation within a representational graphic must have been triggered by the types of social factors that would be

_____________________________________________________________________________________________________

136

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s

expected in the ‘‘hot tub’’ environment of periodic aggregation. By viewing the art of the arid zone in terms of its information potential and likely role in mediating social situations, it is suggested that this is not a conservative and unchanging graphic. Indeed, the art of the arid zone has played a fundamental role in the success of extreme mobility and human adaptation to the most marginal landscapes of this most arid of continents.

Note 1 I have excluded two of the motifs used by David et al. (1992) in the current analysis. The Sturt Creek sample (Walsh 1988) was excluded on the basis of its non-comparable appearance, and the Yingalarri example because I could not ﬁnd a published scaled drawing. Sources for the current sample are attributable to Edwards (1968), David et al. (1992), and Dix (1972, 1977). Unpublished material was provided by Peter Kendrick and Peter Veth, as well as my own recording work.

References Ackerman, K. and Stanton, J. 1994: Riji and Jakoli: Kimberley Pearl Shell in Aboriginal Australia. Darwin: Northern Territory Museum of Arts and Science. Chewings, C. 1909: ‘‘About the blacks’’: Natives of central Australia. The Register, 1 (16), 1–11. Conkey, M. W. 1980: The identiﬁcation of hunter-gatherer aggregation sites – the case of Altimira. Current Anthropology, 21 (5), 609–30. Cook, N., Davidson, I., and Sutton, S. 1990: Why are so many ancient rock paintings red? Australian Aboriginal Studies, 1, 30–2. Crawford, I. M. 1977: The relationship of Bradshaws and Wandjina art in northwest Kimberley. In P. J. Ucko (ed.), Form in Indigenous Art, London: Duckworth, 357–69. David, B., Armitage, R. A., Hyman, M., Rowe, M. W., and Lawson, E. 1999: How old is north Queensland rock art? A review of the evidence with new AMS determinations. Archaeology in Oceania, 34, 103–20. David, B. and Chant, N. 1995: Rock art and regionalization in north Queensland prehistory. Memoirs of the Queensland Museum, 3 (2), 357–528. David, B., Chant, N., and Flood, J. 1992: Jalijbang 2 and the distribution of pecked faces in Australia. Memoirs of the Queensland Museum, 32 (1), 61–77. Dickens, J. 1996: Change and continuity in central Australian graphic systems. Journal of Anthropological Archaeology, 15, 20–40. Dix, W. 1972: Rock art along the Stock Route. Hemisphere, 16 (7), 18–25. Dix, W. 1977: Facial representations in Pilbara rock engravings. In P. J. Ucko (ed.), Form in Indigenous Art, London: Duckworth, 277–85. Edwards, R. 1968: Prehistoric rock engravings at Thomas Reservoir, Cleland Hills, western central Australia. Records of the South Australian Museum, 15 (4), 647–70. Frederick, U. 2000: Keeping the land alive: Changing social contexts of landscape and rock art production. In R. Torrence and A. Clarke (eds), The Archaeology of Difference: Negotiating Cross-Cultural Engagements in Oceania, London: Routledge, 300–30.

_____________________________________________________________________________________________________

137

______________________________________________________________________________

Jo McD o n a l d Galt-Smith, B. 1997: Motives for motifs: Identifying aggregation and dispersion settlement patterns in the rock art assemblages of central Australia. Unpublished BA (Hons) thesis, Armidale: University of New England. Gamble, C. S. 1982: Interaction and alliance in Palaeolithic society. Man, 17 (1), 92–107. Gibbs, M. and Veth, P. 2002: Ritual engines and the archaeology of territorial ascendancy. In S. Ulm, C. Westcott, J. Reid, A. Ross, I. Lilley, J. Prangnell, and L. Kirkwood (eds), Barriers, Borders, Boundaries: Proceedings of the 2001 Australian Archaeological Association Annual Conference, St Lucia: University of Queensland, 11–19. Gillen, F. J. 1968 [1901–2]: Gillen’s Diary: The Camp Jottings of F. J. Gillen on the Spencer and Gillen Expedition Across Australia. Adelaide: Libraries Board of South Australia. Gould, R. A. 1969: Yiwara: Foragers of the Australian Desert. London: Collins. Gould, R. A. 1977: Puntutjarpa Rockshelter and the Australian Desert Culture. New York: Anthropological Papers of the American Museum of Natural History no. 54, American Museum of Natural History. Gunn, R. G. 1995: Regional patterning in the Aboriginal rock art of central Australia: A preliminary report. Rock Art Research, 12, 117–27. Gunn, R. G. 2000a: Central Australian rock art: A second report. Rock Art Research, 17 (2), 111–26. Gunn, R. G. 2000b: Spencer and Gillen’s contribution to Australian rock art studies. Rock Art Research, 17 (1), 56–64. Gunn, R. G. 2003: Arrernte rock art: Interpreting physical permanence in a changing social landscape. Australian Aboriginal Studies, 1, 52–73. Jestribek, K. 1997: Different domains. An examination of gender relations and motif use in contemporary art from Balgo, Western Australia. Unpublished BA (Hons) thesis, Perth: University of Western Australia. Jochim, M. A. 1983: Paleolithic cave art in ecological perspective. In G. Bailey (ed.), HunterGatherer Economy in Prehistory: A European Perspective, Cambridge: Cambridge University Press, 212–19. Kimber, R. 1977: Mosaics you can move. Hemisphere, 21 (1), 2–7. Layton, R. 1986: Uluru: An Aboriginal History of Ayers Rock. Canberra: Australian Institute of Aboriginal Studies. Layton, R. 1992: Traditional and contemporary art of Aboriginal Australia: Two case studies. In J. Coote and A. Shelton (eds), Anthropology, Art and Aesthetics, Oxford: Clarendon Press, 137–59. Lewis, D. 1988: The Rock Paintings of Arnhem Land, Australia: Social, Ecological, and Material Culture Change in the Post-Glacial Period. Oxford: British Archaeological Reports International Series no. 415, Hadrian Books. Lewis, D. and McClausland, B. 1987: Engraved human ﬁgures and faces from Wardaman country, eastern Victoria River District, Northern Territory. Australian Aboriginal Studies, 1, 67–79. McCarthy, F. D. 1979: Australian Aboriginal Rock Art (4th edn). Sydney: Australian Museum. McConvell, P. 1996: Backtracking to Babel: The chronology of Pama-Nyungan expansion in Australia. Archaeology in Oceania, 31 (3), 125–44. McDonald, J. J. 1994: Dreamtime superhighway: An analysis of Sydney basin rock art and prehistoric information exchange. Unpublished PhD thesis, Canberra: The Australian National University.

_____________________________________________________________________________________________________

138

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s McDonald, J. J. 1998: Shelter rock art in the Sydney basin – a space-time continuum: Exploring different inﬂuences on stylistic change. In C. Chippendale and P. S. C. Tacon (eds), The Archaeology of Rock Art, Cambridge: Cambridge University Press, 319–35. McDonald, J. J. 1999: Bedrock notions and isochrestic choice: Evidence for localized stylistic patterning in the engravings of the Sydney region. Archaeology in Oceania, 34 (3), 145–60. McDonald, J .J. 2000: AMS dating charcoal drawings in the Sydney region: Results and issues. In G. K. Ward and C. Tuniz (eds), Advances in Dating Australian Rock Images: Papers from the First Australian Rock Picture Dating Workshop, Melbourne: AURA, 90–4. Maynard, L. 1977: Classiﬁcation and terminology in Australian rock art. In P. J. Ucko (ed.), Form in Indigenous Art, Canberra: Australian Institute of Aboriginal Studies, 385–402. Megaw, V. 1982: Western Desert acrylic painting – artifact or art? Art History, 5, 205–18. Morphy, H. (ed.) 1989: On representing ancestral beings. In H. Morphy (ed.), Animals into Art, London: Unwin Hyman, 144–60. Morphy, H. 1991: Ancestral Connections: Art and an Aboriginal System of Knowledge. Chicago: University of Chicago Press. Morwood, M. J. 2002: Visions from the Past: The Archaeology of Australian Rock Art. Sydney: Allen and Unwin. Munn, N. D. 1966: Visual categories: An approach to the study of representational systems. American Anthropologist, 66, 939–50. Munn, N. D. 1973: Walbiri Iconography: Graphic Representation and Cultural Symbolism in a Central Australian Society. Ithaca, NY: Cornell University Press. Myers, F. R. 1986: Always ask: Resource use and ownership among Pintupi Aborigines of the Australian Western Desert. In N. Williams and E. Hunn (eds), Resource Managers: North American and Australian Hunter-Gatherers, Canberra: Australian Institute of Aboriginal Studies, 173–96. Myers, F. R. 1989: Truth, beauty, and Pintupi painting. Visual Anthropology, 2, 163–95. Myers, F. R. 2002: Painting Culture: The Making of an Aboriginal High Art. Durham, NC: Duke University Press. Nobbs, M. F. and Dorn, R. I. 1988: Age determinations for rock varnish formation within petroglyphs. Rock Art Research, 5 (2), 108–46. Nobbs, M. F. and Dorn, R. I. 1993: New surface exposure ages for petroglyphs from the Olary Province, South Australia. Archaeology in Oceania, 27, 199–220. O’Connor, S. 1995: Carpenter’s Gap rock shelter 1: 40,000 years of Aboriginal occupation in the Napier Ranges, Kimberley, WA. Australian Archaeology, 40, 58–9. O’Connor, S., Veth, P., and Campbell, C. 1998: Serpent’s Glen rock shelter: Report on the ﬁrst Pleistocene-aged occupation sequence from the Western Desert. Australian Archaeology, 46, 12–22. Roberts, R., Walsh, G. L., Murray, A., Olley, J., Morwood, M., Tuniz, C., Lawson, E., Macphail, M., Bowdery, D., and Naumann, I. 1997: Luminescence dating of rock art and past environments using mud-wasp nests in northern Australia. Nature, 387, 696–9. Rosenfeld, A. 1993: A review of the evidence for the emergence of rock art in Australia. In M. A. Smith, M. Spriggs, and B. Fankhauser (eds), Sahul in Review: Pleistocene Archaeology in Australia, New Guinea, and Island Melanesia, Canberra: The Australian National University, 71–80. Rosenfeld, A. 2002: Rock art as an indicator of changing social geographies in central Australia. In B. David and M. Wilson (eds), Inscribed Landscapes: Marking and Making Places, Honolulu: University of Hawaii Press, 61–78.

_____________________________________________________________________________________________________

139

______________________________________________________________________________

Jo McD o n a l d Rosenfeld, A., Horton, D., and Winter, J. 1981: Early Man in North Queensland. Canberra: The Australian National University. Rosenfeld, A. and Smith, M. A. 2002: Rock-art and the History of Puritjarra Rock Shelter, Cleland Hills, Central Australia. Proceedings of the Prehistoric Society, 68, 103–24. Ross, J. 2003: Rock art, ritual, and relationships: An archaeological analysis of rock art from the central Australian arid zone. Unpublished PhD thesis, Armidale: University of New England. Smith, C. E. 1989: Designed dreaming: Assessing the relationship between style, social structure, and environment in Aboriginal Australia. Unpublished BA (Hons) thesis, Armidale: University of New England. Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Smith, M. A. 1996: Prehistory and human ecology in central Australia: An archaeological perspective. In S. R. Morton and D. J. Mulvaney (eds), Exploring Central Australia: Society, Environment, and the 1894 Horn Expedition, Chipping Norton: Surry Beatty and Sons, 61–73. Smith, M. A., Fankhauser, B., and Jercher, M. 1998: The changing provenance of red ochre at Puritjarra rock shelter, central Australia: Late Pleistocene to present. Proceedings of the Prehistoric Society, 64, 275–92. Smith, M. A. and Veth, P. M. 2004: Radiocarbon dates for baler shell in the Great Sandy Desert. Australian Archaeology, 58, 37–8. Spencer, B. and Gillen, F. J. 1899: The Native Tribes of Central Australia. London: Macmillan. Stirling, E. C. 1896: Anthropology. In B. Spencer (ed.), Report on the Work of the Horn Scientiﬁc Expedition to Central Australia: Part IV, Anthropology, Melbourne: Melville, Mullen and Slade, 1–157. Terry, M. 1932: Erldunda Station – Laverton, Letter book containing company correspondence. Manuscript, Adelaide: South Australian Museum. Thorley, P. B. 1998: Pleistocene settlement in the Australian arid zone: Occupation of an inland riverine landscape in the central Australian ranges. Antiquity, 72, 34–45. Tonkinson, R. 1991: The Martu Aborigines: Living the Dream in Australia’s Desert. Fort Worth, TX: Holt, Rinehart, and Winston. Veth, P. M. 1993: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory, Archaeology Series no. 3, Ann Arbor, MI. Veth, P. 2000a: Cycles of aridity and human mobility: Risk minimization amongst late Pleistocene foragers of the Western Desert, Australia. Philadelphia, PA: Unpublished paper presented at the Society of American Archaeology conference. Veth, P. 2000b: Origins of the Western Desert language: Convergence in linguistic and archaeological space and time models. Archaeology in Oceania, 35 (1), 11–19. Veth, P. and McDonald, J. J. 2002: The role of art in long-term cycles of aggregation in the Western Desert. Canberra: Unpublished paper presented to Arcling conference, National Museum of Australia. Veth, P., Smith, M., and Haley, M. 2001: Kaalpi: The archaeology of a sandstone outlier in the Western Desert. Australian Archaeology, 52, 9–17. Walsh, G. 1988: Australia’s Greatest Rock Art. Bathurst: E. J. Brill–Robert, Brown, and Associates. Watchman, A. 1992: Doubtful dates for Karolta engravings. Australian Aboriginal Studies, 2, 51–5.

_____________________________________________________________________________________________________

140

______________________________________________________________________________

Ar c h a i c Fa c e s t o H e ad d r e s se s Watchman, A., Walsh, G. L., Morwood, M. J., and Tuniz, C. 1997: AMS radiocarbon age estimates for early rock paintings in the Kimberley, NW Australia: Preliminary results. Rock Art Research, 14, 18–26. Wiessner, P. 1989: Style and changing relations between individual and society. In I. Hodder (ed.), The Meanings of Things: Material Culture and Symbolic Expression, London: Unwin Hyman, 56–63. Wiessner, P. 1990: Is there a unity to style? In M. Conkey and C. Hastorf (eds), The Uses of Style in Archaeology, Cambridge: Cambridge University Press, 105–12. Wobst, H. M. 1977: Stylistic behavior and information exchange. In C. E. Cleland (ed.), For the Director: Research Essays in Honour of J. B. Griffen, Anthropological Papers no. 61, University of Michigan, 317–42.

_____________________________________________________________________________________________________

141

______________________________________________________________________________

--------------------------------------------------------------------------------

8

----------------------------------------------------------------------------

The Archaeology of the Patagonian Deserts Hunter-Gatherers in a Cold Desert Luis Alberto Borrero

Introduction Most of this volume deals with hunter-gatherer groups located in hot deserts. However, in high latitudes there are signiﬁcant areas of cold desert where aridity is largely a function of low temperature. Do cold deserts pose different problems for hunter-gatherer populations? Do cold deserts have different histories? This chapter explores some of these issues using the early settlement of the Patagonian deserts as a case study (see Figure 8.1). Early Patagonian ethnographic sources describe the Tehuelche ; hunter-gatherers who relied on guanaco (Lama guanicoe) and n˜andu´ (Rhea americana, Pterocnemia pennata) hunting. The sources particularly noticed their use of bola stones to immobilize and capture these animals. This hunting device is characteristic of open spaces, and it is one of the more salient cultural responses to the exploitation of cold desert resources. Different ethnic groups displaying a variety of subsistence strategies, including a strong emphasis on the recollection of small package resources, were recognized in the northwest portion of Patagonia. This emphasis on a wide variety of low return resources is another characteristic of groups adapted to the Patagonian deserts.

Regional Background Patagonia and Tierra del Fuego extend today over more than 1,000,000 sq km, but at the end of the Pleistocene they constituted a single landmass that covered a much larger space, including a signiﬁcant area today ﬂooded by the sea. Large portions of Patagonia are included within the ‘‘arid diagonal’’ of South America

_____________________________________________________________________________________________________

142

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

Figure 8.1 Map of the Upper Limay River showing locations of archaeological sites mentioned in the text.

(Iglesias 1982), which is characterized by an arid climate not moderated by maritime air masses from the Atlantic or Paciﬁc Oceans. Patagonia includes a series of discontinuous desert regions. Deserts in the north of Patagonia, between 388 and 418S, are highly seasonal and very cold (seasonality increases with latitude and the south of Patagonia is more seasonal than the north). Available paleoclimatic and paleoecological data point to the existence of discontinuous arid conditions since at least the end of the Pleistocene. Late Pleistocene times were marked by the retreat of the cordilleran glaciers, a process that was well under way by ca. 14,000 bp (Clapperton 1993). This process was accompanied by an amelioration of climate, punctuated by short cold pulses (Clapperton 1993; Heusser 1993a; Markgraf 1993). Compared to northern North America the development of glaciers was less extensive in South America, with the more maritime late Pleistocene climate representing a less stressful habitat.

_____________________________________________________________________________________________________

143

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

The Colonization of Patagonian Deserts The human exploration and colonization of Fuego-Patagonia began in late Pleistocene times and took place under very cold and dry conditions. Most of the oldest available evidence comes from the cold deserts of the south, which is probably an artifact of the better archaeological visibility of sites in those regions. The colonization of new lands is a case of exogeny, in which new species are usually at an advantage. To start with, unless we are talking of niche expansion within a given biome, newly arrived species are usually generalists. In spite of several voices claiming that the ﬁrst human foragers to arrive in Patagonia were specialized, the consensus now is that they were not. Within this context it is relevant to ask: ‘‘How did humans adapt to new lands?’’ Several authors (Beaton 1991; Borrero 1989; Meltzer 2002; Rockman and Steele 2003) have explored this question. If we combine their insights with the evidence from Patagonia, we can conclude that considerable time was needed to colonize this huge area. Indeed, some time was required to establish a pattern of use in the use of a place and its surroundings, including acquiring critical knowledge about the location of the best lithic sources, the less costly paths for movement, and so on. The concept of ‘‘ecological perception’’ has been recently proposed to explain the kind of knowledge needed to interact with a new environment (Gutierrez and Martinez 2003). This would have been accompanied by the symbolic appropriation of territory (Politis 2003), a process which can take several generations to complete (Rockman 2003). What were the likely constraints on early settlement in these deserts? Did they include water, shelter, or food? Or were low temperatures a constraint? It must be noted that humans living in the northwest corner of Patagonia were not exposed to extreme cold. Thus physiological adaptations, such as those recorded for periArctic or Fuegian environments, were not required. Cultural responses easily compensated for observed climatic ﬂuctuations. Available evidence suggests that food was probably widespread and thus its distribution was not the organizing principle for the ﬁrst settlers. As for shelter, there is substantial support for the notion that the ﬁrst inhabitants of America possessed appropriate technology for constructing shelters and therefore were not dependant on the availability of caves for protection. Dillehay (1991) explored the role of diseases in colonizing populations, which may have been a retarding factor for the peopling of tropical America, but the constraints were probably not very important in cold Patagonia. For the more arid areas of Patagonia, the likely constraint was water, which was heterogeneously distributed (Belardi and Campan 1999). Very few major rivers, with an east–west axis, cross Patagonia and most of these rivers have their origin near the Cordillera at glacial lakes. Although there were changes in the availability of water through time, as well as regional differences, it is difﬁcult to visualize a Patagonian archaeological period when water was not a critical resource. In its extreme form a sequence from Cardiel Lake (498S) informs us that the lake was completely dry around 11,000 bp (Gilli et al. 2001). Initial human occupation at Cardiel is dated

_____________________________________________________________________________________________________

144

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

ca. 6,500 bp and appears to be related to ﬂuctuations in the lake stands (Gon˜i 2000–2). A similar situation was observed at Lago Argentino (Borrero and Franco 2000). If water was a constraining factor for colonizing human populations, then it is necessary to evaluate how difﬁcult it was to access it in northwest Patagonia – a vast and relatively ﬂat piece of land. It has been suggested that rivers were probably the main corridors for human circulation (Anderson and Gillam 2000; Kelly 2003). Others have suggested that it is opportunities rather than constraints that are more important for modeling the settlement of Patagonia. Following this line, it has been argued that the presence of a naive fauna unfamiliar with human predators would have been a factor underlying initial human movements into these regions (Martin 1973). The ecological process by which humans enter a new world can be analyzed as a series of stages, each of which involves different kinds of interactions between humans and other components of the environment. Exploration waves are likely to have been characterized by small groups of people that retained links with larger units, within which their social lives took place (e.g., Rogers 1990). For contrary views arguing for discontinuities in settlement see the ‘‘point and arrow pattern’’ proposed by Rockman (2003). Any human group entering a new landscape should have more than one social unit, in order to be successful. An adequate reproductive pool is a basic requirement to the survival of the social unit (MacDonald and Hewlett 1999; Moore and Moseley 2001; Wobst 1974). The rate of colonization remains a key issue, with archaeologists posing either models of fast moving (e.g., Martin 1973) or slow moving groups of people (e.g., Jaffe 1992). Most models infer the rate of spread from the chronological distribution of early sites (Hazelwood and Steele 2003), although it is often acknowledged that the rate need not have been constant (e.g., Dillehay 2000; Kelly 2003), and that the presence of barriers may retard dispersal of populations (Borrero 1994–5; Steele et al. 1998). In the case of human dispersion into Fuego-Patagonia, I argue that available spaces were only slowly ﬁlled, with settlement being discontinuous in time and space (Borrero 1989). A process of ﬁssioning of bands together with changes in hunting and collecting ranges is all that is required in order to explain this process of slow expansion (see also Surovell 2000). I argue that exploration is largely a logistic process. At least two scales of mobility were involved: one employed during the ‘‘exploration’’ phase, which is tied to the necessities of entering unknown territory; and the other during the ‘‘installation’’ phase, which is the multigenerational scale where ﬁnding mates and rearing children are considerations. This distinction between explorers and colonists was also highlighted by Mandryk (2003) and is also discussed in Chapter 1 of this volume. With some of these concepts in mind we can now explore the archaeological record of the initial occupation of Patagonia. With the important exception of Monte Verde, a site located west of the Andean Cordillera in an area not always included in Patagonia, the oldest radiocarbon dates available are early Holocene. These dates are late in comparison to those obtained farther south, which lie between 11,000–10,500 bp.

_____________________________________________________________________________________________________

145

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

Northern Patagonia The deserts of northwest Patagonia which lie between latitudes ca. 388S and 418S are characterized by an arid climate (Chiozza and Gonzalez van Domselaar 1958). Only near the Cordillera is precipitation high enough to support a forest, and here it only develops as a narrow band. The net result is that temperate rainforests occur mainly west of the Cordillera, while eastwards the Patagonian steppe stretches all the way to the Atlantic coast. The forest is evergreen and formed by several associations, including Anemone-Nothofagetum pumilionis, Fitzroyetum, and Pilgerodendronetum (Ramirez 1989). This forest is consistent with the resource zone that Beaton (1991) labeled a ‘‘megapatch’’; that is, a spatial unit of relative environmental homogeneity for foragers. In the eastern plains another megapatch occurs, comprising a xeric steppe dominated by coiro´n amargo (Stipa speciosa, S. humilis) and shrubs like neneo (Mulinum spinosum) and paramela (Adesmia campestris). The forest as a unit therefore differs markedly from the eastern desert environment. In the ﬁrst instance, water is plentiful in the forest, while it is scarce on the steppe. Additionally, the macrovertebrate faunas of the forest are dominated by two species of deer (Hippocamelus bisulcus and Pudu pudu), while in contrast a camelid – the guanaco (Lama guanicoe) – and ﬂightless birds (Pterocnemia pennata and Rhea americana) inhabit the steppe. From the point of view of foragers, these are two different worlds. The evidence for early human presence in north Patagonia west of the Andes (in what is known as the Chilean Lake Region) has been obtained at three sites: Monte Verde, Chinchihuapi, and Mariﬁlo 1. Monte Verde is a stratiﬁed open air site. It has provided a suite of radiocarbon dates falling around 12,500 bp (Dillehay and Pino 1997) for human occupations characterized by adaptation to a forest environment. Monte Verde has good preservation of organic remains, including bones, logs, plants, cordage, and hide. The variety of plants includes several species that are interpreted as important for human subsistence (Rossen and Dillehay 1997). Only a few lithic implements, such as lanceolate projectile points and bola stones, were found associated with these remains (Collins 1997). The Chinchihuapi site is located nearby, and is dated to 12,420130 bp (Beta-65842), making it almost contemporaneous with Monte Verde. Two ﬂakes, one spheroid, and three wooden artifacts were found at this site (Dillehay and Pino 1997). Finally, Mariﬁlo 1, which was occupied later than the previously described sites at about 8,42040 bp (Beta138919), is also associated with a forest environment. The pudu (Pudu pudu) is the most important resource used by the earliest inhabitants of this site (Vela´squez and Ada´n 2002), with small birds and mustelids also present. A hearth, lithic artifacts, and one artifact made on a pudu bone were also found. In sum, these three sites testify to the importance of the forest in early human adaptations within South America. Even though ca. 4,000 radiocarbon years separate Monte Verde/ Chinchihuapi from Mariﬁlo 1, the existence of the latter illustrates some continuity in the use of the forests. East of the Andes the available archaeological evidence comes from three caves, all of them located in the upper basin of the Limay River (see Figure 8.1) and

_____________________________________________________________________________________________________

146

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

chronologically the occupations are restricted to the early Holocene. Cueva Traful and Cueva Cuyı´n Manzano, which lie some 7 km from each other, are situated in the forest-steppe ecotone, near the headwaters of the basin. Cueva Epulla´n is on the steppe on the left side of the Limay River, some 100 km east of the ecotone. Available palynological data suggest that the environmental panorama at the beginning of the Holocene was similar to that of today (Heusser 1993b; Markgraf 1993; Prieto and Stutz 1996); however, some changes important for human expansion/occupation have occurred. For example, palynological research at Epulla´n Grande shows that from approximately 10,000–7,000 bp the availability of water near the cave was greater than today (Prieto and Stutz 1996). This is a signiﬁcant observation, given the fact that human residential patterns in this region appear to be controlled by water distribution (Borrero and Nami 1996).

Cueva Traful Cueva Traful is located on the right side of the Traful River near its conﬂuence with the Limay. The initial occupation of the site is dated to between 9,430230 bp (AC-2676) and 9,285105 bp (GX-1711AMS) (Crivelli Montero et al. 1993: 33). The intensity of occupation appears to be very low and testiﬁes to ephemeral human installation. The initial occupation took place on an irregular surface on the ﬂoor of the cave, with a hearth positioned directly on top of the basal rock. Several short-term occupation events have produced small, discrete assemblages. The lithic repertoire includes mostly unretouched ﬂakes, with only a few modiﬁed items. No bifacial artifacts were found; however, a few bifacial reduction ﬂakes attest to the existence of this technology (Cu´neo 1993: 165). Abundant fox (Dusicyon griseus and D. culpaeus) and low frequencies of guanaco (Lama guanicoe) bones characterize the faunal assemblage. Abundant remains of rodents were interpreted as the result of owl regurgitates (Pearson and Pearson 1993). The presence of Lagidium sp., birds, a small felid, and freshwater molluscs was also noted. It had been suggested that foxes were an important component of the diet of the initial inhabitants of the cave (Crivelli Montero et al. 1993: 38); however, the faunal analysis has not yet been fully published. Unfortunately, the lack of a taphonomic assessment of the faunal assemblages prevents this suggestion from being easily accepted (Martin 1998). Immediately after these initial occupations, the assemblages of the Component 1-Traful Unit, dated to 7,308285 bp (LP-8113) and 7,85070 bp (LP-5133), include well deﬁned hearths and lithic artifacts, including triangular projectile points. Guanaco remains are numerically more important, with the presence of Lagidium sp., fox, and small rodents also noted. With the evidence at hand it is not possible to assess the role of these animals in the human diet.

Cueva Cuyı´n Manzano Cueva Cuyı´n Manzano is a site located on a small tributary of the Traful River. The lower occupation, which while also ephemeral, seems to indicate a higher intensity

_____________________________________________________________________________________________________

147

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

of occupation than at Traful (Crivelli Montero et al. 1993: 39). The lowest layer is dated to 9,320240 bp (KN-1432) (Gonza´lez and Lagiglia 1973), with the ﬁrst occupation occurring in a natural depression (Ceballos 1982: 32). Bifacial artifacts are absent, but scrapers are a signiﬁcant component of the modiﬁed lithic repertoire. Two endscrapers and nine sidescrapers were described, as well as 12 retouched ﬂakes (Ceballos 1982). The faunal remains – which have not been published in any detail – include guanaco and appear not to be as abundant as fox and rodent remains. Ceballos (1982) considered that rodents (mainly Ctenomys sp.) were deposited by humans. However, the relationship of the faunal remains with human activities cannot be assigned with any certainty at this time. Generally speaking, consumption of rodents has proved difﬁcult to demonstrate in the absence of a detailed taphonomic analysis (Pardin˜as 1999). Even though the contribution of large mammals in the human diet is easier to evaluate, it still requires site formation information that is not yet available for Cuyin Manzano.

Cueva Epulla´n Grande Cueva Epulla´n is a site located on the xeric steppe. Its basal occupation includes small hearths, lithic artifacts, rock engravings, and probably one human skeleton. Lithic artifacts were manufactured from obsidian and basalt. Bifacial reduction debris makes up 39 percent of the assemblage, being more important in the initial occupations of Epulla´n than later in the sequence (Crivelli Montero et al. 1996: 200). Since obsidian is not locally available, this patterning argues for logistical mobility during initial occupation of the site. It appears that the site was only used on an intermittent basis from a distant base. Occupation events are dated to between 9,970100 bp (LP-213) and 7,550100 bp (Beta-47401). The faunal assemblage is composed of a variety of species. Guanaco remains are abundant, but several smaller species are also important. South American ostrich (cf. Pterocnemia pennata) bones are present and they display cut marks and burning. The presence of egg fragments, some of them also burnt, was also recorded. There are also cut marks on a skunk mandible (Conepatus sp.) and scraping marks on armadillo plaques (Chaetophractus villosus). Other small prey present at the site include red fox (Lycalopex culpaeus), grey fox (L. gymnocercus), a small Felidae, and rodents. A detailed taphonomic analysis of the latter clearly demonstrated that most of the rodent remains were deposited as bird regurgitates. Well preserved regurgitates were found in the upper, but not lower, layers (Pardin˜as 1999: 277). At Epulla´n the exploitation of cacti, especially Austrocactus aff. A. Bertinii, is documented starting at ca. 7,06090 bp (Beta-41622) (Crivelli Montero et al. 1996: 188). Later in the sequence there is abundant proof for the storage of this cacti and a generally higher diversity of species (Crivelli Montero et al. 1996: 206). This diversity is usually interpreted as a move towards a broader diet. The site was also used as a burial place with a total of four individuals recorded for the earlier period (Crivelli Montero et al. 1996: 214-16). Burial number 40 consists of human remains lying above a round platform made of rocks and plants. Redeposited charcoal found directly above the burial was dated to 7,90070 bp

_____________________________________________________________________________________________________

148

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

(Beta-44412), which represents a maximum date for the burial. The bones are badly preserved, stained with ocher, and dispersed. The burial was covered with stones in the same way as the historically known chenques, which are stone cairns marking the location of Tehuelche bodies. Burial number 82 is a juvenile also covered with stones which were deposited after ca. 7,900 bp. The same pattern was observed with Burial number 126, which is an individual of 5–7 years of age and for which a maximum date of 7,55070 bp (Beta-47401) has been attributed. Skeleton number 31 is a poorly preserved individual of more than 13 years of age, with its bones badly fragmented and dispersed, with only a few in anatomical position. It is dated by association to ca. 9,970100 bp. The skeleton was found in physical association with guanaco bones and the authors speculate that this may be a case of an individual dying at a locality and being subsequently covered by natural processes. The association of dates and burials appears to be secure, but direct dates on the bones would be highly desirable. With one exception, the skeletons were placed on beds of rocks or grass and ocher was used extensively in the burials. All the skeletons were located near the rear of the cave, especially the chenques (see also Mena and Reyes 1998, for a similar case at Ban˜o Nuevo but in a different setting). It would appear that one of the oldest modes of internment, which thus far has only been recorded in caves – possibly due to sampling bias – is to cover the body with rocks. In later times this mode became dominant in Patagonia. The main difference between the two areas is that in Patagonia many were open air cairns. Finally, it must be emphasized that panels of wall engravings, in the form of ﬁnger markings and incised lines, were discovered on the lower walls of Epulla´n Cave. They were covered by sediments dated between 9,970100 and 7,100 bp (Crivelli Montero et al. 1996: 190). These engravings testify to the use of the cave before the subsequent sediments were deposited. Although the sample of sites described above is small, their particular characteristics are relevant to an understanding of the process of initial human occupation, since in all three cases the ﬁrst occupation took place immediately above basal rock. This suggests that humans were important sedimentary agents, and that they selected these places for occupation during the early Holocene. Moreover, the abandonment of one unburied individual and the abundant presence of obsidian artifacts at Cueva Epulla´n Grande are consistent with an exploratory use of the region. The same can be said of the ephemeral character of the initial occupations as recorded at the three sites.

Late Holocene The late Holocene history of human occupation in northwest Patagonia is characterized by the repeated use of places located near water and the exploitation of a variety of resources, including small mammals, birds, molluscs, and plants (Barberena et al. in press; Borrero et al. 1996; Sanguinetti and Curzio 1996). Large vertebrates such as the guanaco were part of the human diet, but it appears that they

_____________________________________________________________________________________________________

149

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

did not play a central role, as guanaco remains are rarely abundant or dominant in faunal assemblages. When present, guanaco bones are highly fragmented, suggesting intensive exploitation. This latter pattern may be associated with the fact that many of these sites are dated around the Medieval Warm Period (Stine 1994; Villalba 1994), which was probably a period of water (and possibly protein) stress for humans inhabiting the warmest deserts of Patagonia (Borrero and Franco 2000). This diversity of resources is usually interpreted as an indication of greater dietary breadth (Borrero 1981; Crivelli Montero et al. 1996; Ferna´ndez 1988–9). However, in the absence of taphonomic analysis it is difﬁcult to accept that the presence of a wider variety of small mammals, birds, and other small package resources at several late Holocene sites necessarily indicates such a broadening in the diet. In the ﬁrst instance, the superior preservation of smaller resources at younger sites or layers must be taken into account. Their absence at some older sites or layers might be simply due to their lower potential for preservation. Secondly, even though these resources are present it is still necessary to demonstrate that they are actually related to human activities. For example, the hypothesis of human consumption of rodents was sustained for only a few of the many examples submitted to taphonomic scrutiny (Pardin˜as 1999). Finally, since the number of remains is usually greater in later occupations, the effect of sample size on diversity must be taken into account.

Discussion Subsistence Initial occupations of several sites in northern Patagonia, located both to the west and east of the Andes, show that guanaco – the most abundant vertebrate today – was not necessarily the most important prey (Ceballos 1982; Crivelli Montero et al. 1993; Dillehay 1997). Plants, perhaps foxes, and other small mammals appear to have been the most important resources. Since this evidence comes from both caves and open air sites, it invites researchers to (re)consider the existence of adaptations not centered on the guanaco. In this context, it is necessary to consider the case of megamammals, which are present at Monte Verde and Traful. At Traful the remains of ground sloth (Mylodontinae) were interpreted as probably being unassociated with initial human occupations (Crivelli Montero et al. 1993: 35). At Monte Verde, even where mastodon remains (Cuvieronius sensu Casamiquela) are present in the cultural layers (Casamiquela and Dillehay 1989), they do not appear to have been important for subsistence. Effectively, the mastodon bones were recovered with non-local embedded dirt (Karathanasis 1997), suggesting that they were collected as bones away from the site. It appears that in many regions of northern Patagonia, especially near the Andean Cordillera, the Pleistocene fauna had already become extinct when the ﬁrst humans arrived. Two radiocarbon dated paleontological locations indicate that ground

_____________________________________________________________________________________________________

150

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

sloths were available in the region between 14,665–13,750 bp (Hofreiter et al. 2003) and 12,600–10,800 bp (Nami 1996). We know that by ca. 12,500 bp humans were already installed west of the Cordillera, but there is no evidence of their presence east of the mountains. This is a pattern that can be traced to other sites located near the Cordillera, especially to the south (Borrero 2002), where all of the available evidence suggests that humans arrived in the area after the extinction of the megafauna. In comparison, small mammals appear to have sometimes made a higher dietary contribution. One example is the importance of fox remains in the lower layers at Traful and other sites. If their presence can be attributed to human activities, this evidence – together with that from Ban˜o Nuevo (Trejo and Jackson 1998) and the Pampas (Politis 1996) – illustrates the early importance of a resource which was then later ignored. The pattern observed for the Pleistocene–Holocene transition is one of dependence on small-sized prey from the forest, and on the guanaco – the largest modern vertebrate – on the steppe. Later in the sequences, there is evidence of increasing use of guanaco in the forest, and of a broad variety of resources on the steppe. The reality of these apparent patterns in terms of human dietary contribution must await further taphonomic analysis.

Mobility All of the early Holocene archaeological assemblages from northwest Patagonia appear to reﬂect ephemeral occupation and it is suggested that population levels were very low at this time. The existence of extensive home ranges is compatible with this scenario, but cannot however be assumed for all desert groups (see below). Ephemeral archaeological signatures are often taken to reﬂect the transient occupation of sites due to a system of high residential mobility (Binford 1980). In such cases, the low redundancy in the use of places results in a relatively high integrity of deposits. However, given that all of these early sites are caves, the functional requirements associated with the use of them may be shaping the record as we can now observe it. As the case of Monte Verde makes it abundantly clear, the different patterning in open sites may serve to change our perceptions of the mobility systems of the ﬁrst inhabitants. As water appears to have been permanent at only a few predictable locations, a pattern of land use relying on logistical forays is plausible. This is a variant of the system described as ‘‘tethered foraging’’ (cf. Kelly 1995: 126–7), whereby the known location of springs and other waterholes is used to organize movements to satellites which are still centrally focused on a permanent river. Given this suggested model of tethering, groups were probably not covering as much territory as would free wandering nucleated groups. The available archaeological evidence is consistent with this latter scenario. Here I argue that rivers were functioning as corridors into the arid zone. A corollary of this situation is that during drier periods there was probably a strong pressure on resources located close to the basin (see Kelly 1995: 127; Smith 1989).

_____________________________________________________________________________________________________

151

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

Technology There are also technological issues worthy of consideration. On the basis of the Traful data it was suggested that some of the early deposits found east of the Andes were based on a tradition of unifacial reduction (Crivelli Montero et al. 1993). However, further analysis of the lithic debitage instead supports the existence of bifacial reduction at the site. The same critique can be applied at several other early assemblages in other areas of Fuego-Patagonia that do not include projectile points (Nami 1993–4). The importance of these observations stems from the fact that a unifacial tradition has been proposed at this stage for the whole of South America (Dillehay 2000; Krieger 1964). It is necessary to stress that the samples recovered have been small, and that functional requirements within the context of initial peopling of the region, together with sampling error, may better explain the lack of visibility of bifacial reduction. The bifacial issue is also crucial in the discussion of southern cases. For that reason I will brieﬂy mention the ﬁndings at the lower layers of Los Toldos, El Ceibo, and other sites in the middle Deseado River located to the south of this region at ca. 488S (Cardich 1987). This region appears to constitute one of the ‘‘hotspots’’ or places of initial settlement. It appears that the quality and abundance of lithic sources, together with the presence of water sources and well protected canyons, constituted a ‘‘magnet’’ for human occupation. The earlier evidence at these sites is sometimes presented as proof of the existence of an industry named Level 11, which is characterized by large scrapers and unretouched ﬂakes. In order to maintain a separate identity for the Level 11 Industry, Cardich (1987) emphasized differences with another industry recognized by most other researchers working in the region, the Toldense. It seems to be a moot point whether there are one or two industries, unless we are asserting much more than techno-typological difference. Overall, faunal and lithic assemblages between the ‘‘industries’’ exhibit great similarity. The ﬁndings at the lower layers of the Arroyo Feo site are crucial here. They are similar to the so-called Level 11 Industry, but chronologically overlap with the Toldense at ca. 9,000 bp (Gradı´n et al. 1979). It appears that the absence of projectile points at some sites was taken as a marker that a separate industry existed. On reﬂection it would appear that the evidence for a unifacial tradition in Patagonia is not strong. If we now look at the wider picture of the peopling of America, it is clear that there is no need to invoke a unifacial industry belonging to these southern settlers. The production of bifacial points, the production of blades, and the use of bone and other so-called ‘‘sophisticated’’ technologies appear to have always been available to the initial explorers of Fuego-Patagonia, as exempliﬁed by recoveries from a number of early sites (e.g., Aguerre 1997; Nami 1993–4; Paunero 1993–4; Scheinsohn 1997; Yacobaccio and Guraieb 1994). During the earliest period of human adaptation to different habitats in Fuego-Patagonia, lithic artifacts were routinely and expediently made on rocks available in the immediate vicinity, with an emphasis on the transport of bifacial artifacts and/or preforms, adequate for situations of high mobility. Local raw material was used predominantly, with high quality and heavily curated exotic materials only present in very low frequencies.

_____________________________________________________________________________________________________

152

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts

In the cases where ‘‘exotics’’ are relatively abundant, such as at Epulla´n Grande, the patterns suggest that logistical mobility is the dominant strategy. On the other hand, if we review the archaeological record of the southern – and coldest – part of Patagonia, the chronological picture is much more consistent. In general, all of the sites on the middle Deseado River display only minimal evidence of use before the Pleistocene–Holocene transition, with dates of ca. 11,000 bp accepted as the oldest for initial human occupation (Miotti 1996; Paunero 1993–4). More intense and repetitive occupation occurs from ca. 10,000 bp. Remains of Pleistocene mammals are present in the earliest levels at some of the sites, but never in large numbers. Instead, guanaco is the dominant large vertebrate. Radiocarbon dates falling between ca. 11,000–10,500 bp are available from different areas in the south (Bird 1988; Nami and Nakamura 1995; Prieto 1991), including what today is referred to as Tierra del Fuego (Massone 1987). In all of these cases the stratigraphic evidence is backed with a reliable radiocarbon chronology, and the assemblages contain bifacial lithic artifacts, basin-shaped hearths, and butchered faunal remains. The archaeological evidence for the presence of humans in southern Patagonia near the end of the Pleistocene is clear and relatively abundant. This pattern indicates that some 2,000 years before the Pleistocene–Holocene transition there were human populations exploiting different landscapes within the region, and that around 11,000–10,500 bp most of the cold desert areas were already known to humans. The deserts of northwest Patagonia, in contrast, only appear to have been used for the ﬁrst time during the early Holocene. Should this be taken as evidence that this was a more difﬁcult and less attractive environment for humans? In answer I reiterate the fact that the northwest Patagonian deserts, being ﬂat and watercontrolled, would have been far less inviting environments for human occupation. While this may well be the case, the intensity of survey and sampling is still too low to make any ﬁrm conclusions at this stage about the exact timing of settlement.

Conclusion The deserts of northwest Patagonia appear to have been occupied later than the forested environments of Monte Verde and Chinchihuapi located to the west. The later chronology for human occupation from northwest Patagonia might imply that the costs of colonizing these deserts were higher. This would be consistent with the fact that detailed knowledge of plants is more difﬁcult to transfer to new and different lands, since they are usually highly speciﬁc to certain habitats (Rockman 2003: 19). These are exactly the kinds of ‘‘problems’’ envisioned by Beaton (1991) when he proposed that movement within the same environmental zone was probably faster, implying a less costly adaptation. On the basis of present data it appears that desert habitats were probably not at the top of the list of habitats selected for systematic human occupation. The presence of a biogeographical barrier between the forests and the deserts, the Andean Cordillera, may have added to the ‘‘costs’’ of colonizing the Patagonian deserts.

_____________________________________________________________________________________________________

153

______________________________________________________________________________

L u i s A lb er to Bo r r e r o

When we look at the settlement of northwest Patagonia within the larger regional context, it is clear that the process of human expansion toward the south was not one that required saturation of landscapes providing donors. From what is known of the archaeological record large areas remained sparsely populated or were not used at all. This appears similar to the spread of the North American populations as recorded through the distribution of projectile points (Anderson and Faught 2000). It seems that most landscapes near the Cordillera were only settled during the early Holocene and after the extinction of Pleistocene megafauna. The available evidence suggests that exploitation took the form of logistical forays from eastern settlement nodes (Borrero 2002; Franco 2002a, 2002b). One reason for this chronological pattern may be that these regions were at the same time too distant from the early circulation routes and equally too close to important biogeographical barriers, such as the Continental Ice Cap. In summary, the colonization of cold deserts required the selection of settlements, artifact production, and subsistence technologies that were organized to accommodate the patchy distribution of resources. This colonization process probably occurred at a multigenerational time scale, and began around the time of the Pleistocene–Holocene transition. The available evidence suggests that the drier cold deserts of Patagonia were colonized after other Patagonian habitats.

References Aguerre, A. M. 1997: Replanteo de la industria Toldense. Arqueologı´a de Patagonia CentroMeridional. Unpublished PhD thesis, Buenos Aires: Universidad de Buenos Aires. Anderson, D. G. and Faught, M. K. 2000: Paleoindian artifact distributions: Evidence and implications. Antiquity, 74, 507–13. Anderson, D. G. and Gillam, J. C. 2000: Paleoindian colonization of the Americas: Implications from an examination of physiography, demography, and artifact distribution. American Antiquity, 65, 43–66. Barberena, R., Manzi, L., and Campan, P. in press: Arqueologı´a de rescate en Piedra del Aguila, Neuque´n: Sitio cueva del Choique. Relaciones de la Sociedad Argentina de Antropologı´a. Beaton, J. 1991: Colonizing continents: Some problems from Australia and the Americas. In T. D. Dillehay and D. J. Meltzer (eds), The First Americans: Search and Research, Boca Raton, FL: CRC, 209–30. Belardi, J. B. and Campan, P. 1999: Estepa y bosque: La utilizacio´n de lagos y lagunas en la regio´n de Lago Aregentino, Provincia de Santa Cruz. In Soplando en el viento. Actas de las III Jornadas de Arqueologı´a de la Patagonia, Neuquen-Buenos Aires: Instituto Nacional de Antropologı´a y Pensamiento Latinoamericano, 25–41. Binford, L. R. 1980: Willow smoke and dog’s tails: Hunter-gatherer settlement systems and archaeological site formation. American Antiquity, 45, 4–21. Bird, J. 1988: Travels and Archaeology in South Chile. Iowa City: University of Iowa Press. Borrero, L. A. 1981: La economı´a prehisto´rica de los pobladores del Alero de los Sauces (Neuque´n, Argentina). Trabajos de Prehistoria, 1, 113–26. Borrero, L. A. 1989: Spatial heterogeneity in Fuego-Patagonia. In S. Shennan (ed.), Archaeological Approaches to Cultural Identity, London: Unwin Hyman, 258–66.

_____________________________________________________________________________________________________

154

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts Borrero, L. A. 1994–5: Arqueologı´a de la Patagonia. Palimpsesto. Revista de Arqueologı´a, 4, 9–69. Borrero, L. A. 2002: The archaeozoology of the Andean dead ends in Patagonia: Living near the Continental Ice Cap. Durham: Unpublished paper presented at the International Congress of the International Council for Archaeozoology. Borrero, L. A. and Franco, N. V. 2000: Cuenca superior del rı´o santa Cruz. Perspectivas temporales. Rı´o Gallegos: Desde el pais de los gigantes. Perspectivas arqueolo´gicas en Patagonia vol. 2, Universidad Nacional de la Patagonia Austral, 345–56. Borrero, L. A. and Nami, H. G. 1996: Arqueologı´a en Piedra del Aguila: Prospecciones y propuestas metodolo´gicas. III. Algunas hipo´tesis y propuestas de trabajo para una arqueologı´a regional. Praehistoria, 3, 35–42. Borrero, L. A., Campan, P., Martin, F. M., and Botella, F. 1996: La margen derecha del rı´o Limay entre el can˜ado´n del Potro y el can˜ado´n Mencue´. Informe preliminar. Praehistoria, 3, 173–84. Cardich, A. 1987: Arqueologı´a de Los Toldos y El Ceibo (Provincia de Santa Cruz, Argentina). Estudios Atacamen˜os, 8, 98–117. Casamiquela, R. and Dillehay, T. D. 1989: Vertebrate and invertebrate faunal analysis. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press, 205–10. Ceballos, R. 1982: El sitio Cuyı´n Manzano. Series y Documentos, 9, 1–66. Chiozza, E. and Gonzalez van Domselaar, Z. 1958: Clima. La Argentina. Suma de Geografı´a vol. 2, Buenos Aires: Peuser, 3–183. Clapperton, C. 1993: The Quaternary Geology and Geomorphology of South America. Amsterdam: Elsevier. Collins, M. 1997: The lithics from Monte Verde: A descriptive-morphological analysis. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press, 383–506. Crivelli Montero, E. A., Curzio, D., and Silveira, M. J. 1993: La estratigrafı´a de la cueva Traful 1 (provincia de Neuque´n). Praehistoria, 1, 9–160. Crivelli Montero, E. A., Pardin˜as, U. F. J., Ferna´ndez, M. M., Bogazzi, M., Chauvin, A., Ferna´ndez, V., and Lezcano, M. 1996: Cueva Epulla´n Grande (Provincia del Neuque´n): Informe de Avance. Praehistoria, 2, 185–265. Cu´neo, E. M. 1993: Ana´lisis de los desechos de talla de las ocupaciones iniciales y del Componente I de la Cueva Traful 1. Praehistoria, 1, 161–205. Dillehay, T. D. 1991: Disease ecology and initial human migration. In T. D. Dillehay and D. J. Meltzer (eds), The First Americans: Search and Research, Boca Raton, FL: CRC, 231–64. Dillehay, T. D. (ed.) 1997: Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press. Dillehay, T. D. 2000: The Settlement of the Americas: A New Prehistory. New York: Basic Books. Dillehay, T. D. and Pino, M. 1997: Radiocarbon chronology. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press, 41–52. Ferna´ndez, J. 1988–9: Ocupaciones de la cueva. Anales de Arqueologı´a y Etnologı´a, 43–45, 661–98. Franco, N. V. 2002a: Estrategias de utilizacio´n de recursos lı´ticos en la cuenca superior del rı´o Santa Cruz. Unpublished PhD thesis, Buenos Aires: Universidad de Buenos Aires.

_____________________________________________________________________________________________________

155

______________________________________________________________________________

L u i s A lb er to Bo r r e r o Franco, N. V. 2002b: (Es posible diferenciar los conjuntos lı´ticos atribuı´dos a la exploracio´n de un espacio de los correspondientes a otras etapas del poblamiento? El caso del extremo Sur de Patagonia. Werken, 3, 119–32. Gilli, A., Anselmetti, F. S., Ariztegui, D., Bradbury, J. P., Kelts, K. R., Markgraf, V., and McKenzie, J. A. 2001: Tracking abrupt climate change in the Southern Hemisphere: A seismic stratigraphic study of Lago Cardiel, Argentina (498S). Terra Nova, 13, 443–8. Gonzalez, A. R. and Lagiglia, H. 1973: Registro nacional de fechados radiocarbo´nicos. Necesidad de su creacio´n. Relaciones de la Sociedad Argentina de Antropologı´a, 7, 291–312. Gon˜i, R. 2000–2: Poblamiento humano, paleoambientes y cronologı´a en la cuenca de los lagos Cardiel y Strobel. Cuadernos del Instituto Nacional de Antropologı´a y Pensamiento Latinoamericano, 19, 669–71. Gon˜i, R. and Barrientos, G. 1998: Estudio de chenques en el lago Salitroso, provincia de Santa Cruz. Rı´o Gallegos: Desde el Paı`s de los Gigantes. Perspectivas Arqueolo´gicas en Patagonia vol. 1, Universidad Nacional de la Patagonia Austral, 161–75. Gradı´n, C., Aschero, C., and Aguerre, A. M. 1979: Arqueologı´a del Area Rı´o Pinturas. Relaciones de la Sociedad Argentina de Antropologı´a, 13, 187–227. Gutierrez, M. and Martinez, G. 2003: Human arrivals, ecological perception, and megamammal exploitation in the Pampean region of Argentina. Washington, DC: Unpublished paper presented at the Fifth World Archaeological Congress. Hazelwood, L. and Steele, J. 2003: Colonizing new landscapes: Archaeological detectability of the ﬁrst phase. In M. Rockman and J. Steele (eds), Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation, London: Routledge, 203–21. Heusser, C. J. 1993a: Late glacial of southern South America. Quaternary Science Reviews, 12, 345–50. Heusser, C. J. 1993b: Palinologı´a de la secuencia sedimentaria de la Cueva Traful 1. Praehistoria, 1, 206–10. Hofreiter, M., Betancourt, J. L., Pelliza Sbriller, A., Markgraf, V., and McDonald, G. 2003: Phylogeny, diet, and habitat of an extinct ground sloth from Cuchillo Cura´, Neuque´n Province, southwest Argentina. Quaternary Research, 59, 364–78. Iglesias, A. 1982: Tipos de clima. In E. Chiozza and R. Figueira (eds), Atlas Total de la Repu´blica Argentina, Buenos Aires: Tomo Fı´sico, 193–200. Jaffe, A. J. 1992: The First Immigrants from Asia: A Population History of the North American Indians. New York: Plenum Press. Karathanasis, A. D. 1997: X-ray diffraction, x-ray ﬂuorescence, and differential scanning calorimetry analysis of sediments from selected features and bones. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press, 817–24. Kelly, R. L. 1995: The Foraging Spectrum: Diversity in Hunter-Gatherer Lifeways. Washington, DC: Smithsonian Institution Press. Kelly, R. L. 2003: Colonization of new land by hunter-gatherers: Expectations and implications based on ethnographic data. In M. Rockman and J. Steele (eds), Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation, London: Routledge, 44–57. Krieger, A. 1964: Early man in the New World. In J. Jennings and E. Norbeck (eds), Prehistoric Man in the New World, Chicago: University of Chicago Press, 28–81. MacDonald, D. H. and Hewlett, B. S. 1999: Reproductive interests and forager mobility. Current Anthropology, 40, 501–23. Mandryk, C. 2003: Foreword. In M. Rockman and J. Steele (eds), Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation, London: Routledge, xiii–xv.

_____________________________________________________________________________________________________

156

______________________________________________________________________________

T he Ar c h a eo l og y o f t he Pa t a g o n i a n D eserts Markgraf, V. 1993: Paleoenvironments and paleoclimates in Tierra del Fuego and southernmost Patagonia, South America. Palaeogeography, Palaeoclimatology, Palaeoecology, 102, 53–68. Martin, F. M. 1998: Madrigueras, dormideros y letrinas: Aproximacio´n a la tafonomı´a de zorros. In L. A. Borrero (ed.), Arqueologı´a de la Patagonia Meridional (Proyecto ‘Magallania’), Concepcio´n del Uruguay: Ediciones Bu´squeda de Ayllu, 73–96. Martin, P. 1973: The discovery of America. Science, 179, 969–74. Massone, M. 1987: Los cazadores paleoindios de Tres Arroyos. Anales del Instituto de la Patagonia, 17, 47–60. Meltzer, D. J. 2002: What do you do when no one’s been there before? Thoughts on the exploration and colonization of the Americas. Memoirs of the California Academy of Sciences, 27, 27–58. Mena, F. and Reyes, O. 1998: Esqueletos humanos del Arcaico temprano en el margen occidental de la estepa centropatago´nica (Cueva Ban˜o Nuevo, XI Region). Boletı´n de la Sociedad Chilena de Arqueologı´a, 25, 19–24. Miotti, L. 1996: Piedra Museo (Santa Cruz): Nuevos datos para la ocupacio´n Pleistoce´nica en Patagonia. In J. Go´mez Otero (ed.), Arqueologı´a, So´lo Patagonia, Puerto Madryn: CENPAT-CONICET, 27–38. Moore, J. H. and Moseley, M. E. 2001: How many frogs does it take to leap around the Americas? Comments on Anderson and Gillam. American Antiquity, 66, 526–29. Nami, H. G. 1993–4: Observaciones sobre desechos de talla procedentes de las ocupaciones tempranas de Tres Arroyos (Tierra del Fuego, Chile). Anales del Instituto de la Patagonia, 22, 175–80. Nami, H. G. 1996: New assessments on early human occupations in the south core. In T. Akazawa and E. J. E. Szathma´ry (eds), Prehistoric Mongoloid Dispersals, Oxford: Oxford University Press, 254–69. Nami, H. G. and Nakamura, T. 1995: Cronologı´a radiocarbo´nica con AMS sobre muestras de hueso procedentes del sitio Cueva del Medio. Anales del Instituto de la Patagonia, 23, 125–33. Pardin˜as, U. F. J. 1999: Tafonomı´a de microvetebrados en yacimientos arqueolo´gicos de Patagonia (Argentina). Arqueologı´a, 9, 265–340. Paunero, R. S. 1993–4: El sitio Cueva 1 de la localidad arqueolo´gica Cerro Tres Tetas (Estancia San Rafael, Provincia de Santa Cruz, Argentina). Anales de Arqueologı´a y Etnologı´a, 48/49, 73–90. Pearson, O. and Pearson, A. 1993: La fauna de mamı´feros pequen˜os de Cueva Traful 1, Argentina: Pasado y presente. Praehistoria, 1, 211–24. Politis, G. 1996: A review of the late Pleistocene sites of Argentina. Fundamentos, 1, 153–70. Politis, G. 2003: The end of the trip: Homo sapiens expansion into the southern core of South America. Washington, DC: Unpublished paper presented at the Fifth World Archaeological Congress. Prieto, A. 1991: Cazadores tempranos y tardı´os en Cueva Lago Sofı´a 1. Anales del Instituto de la Patagonia, 20, 75–99. Prieto, A. and Stutz, S. 1996: Vegetacio´n del Holoceno en el Norte de la estepa patago´nica: palinologı´a de la Cueva Epulla´n Grande (Neuque´n). Praehistoria, 3, 267–78. Ramirez, C. 1989: Past and present landscape and land use. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 1, Washington, DC: Smithsonian Institution Press, 53–85.

_____________________________________________________________________________________________________

157

______________________________________________________________________________

L u i s A lb er to Bo r r e r o Rockman, M. 2003: Knowledge and learning in the archaeology of colonization. In M. Rockman and J. Steele (eds), Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation, London: Routledge, 3–24. Rockman, M. and Steele, J. (eds) 2003: Colonization of Unfamiliar Landscapes: The Archaeology of Adaptation. London: Routledge. Rogers, A. 1990: Group selection by selective emigration: The effects of migration and kin structure. American Naturalist, 135, 398–413. Rossen, J. and Dillehay, T. D. 1997: Modeling ancient plant procurement and use at Monte Verde. In T. D. Dillehay (ed.), Monte Verde: A Late Pleistocene Settlement in Chile vol. 2, Washington, DC: Smithsonian Institution Press, 331–50. Sanguinetti, A. C. and Curzio, D. 1996: Excavaciones arqueolo´gicas en el sitio Piedra del Aguila 11. Praehistoria, 3, 43–99. Scheinsohn, V. 1997: Explotacio´n de materias primas o´seas en la Isla Grande de Tierra del Fuego. Unpublished PhD thesis, Buenos Aires: Universidad de Buenos Aires. Smith, M. A. 1989: The case for a residential human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Steele, J., Adams, J., and Sluckin, T. 1998: Modeling Paleoindian dispersals. World Archaeology, 30, 286–305. Stine, S. 1994: Extreme and persistent drought in California and Patagonia during Medieval time. Nature, 369, 546–9. Surovell, T. A. 2000: Early Paleoindian women, children, mobility, and fertility. American Antiquity, 65, 493–508. Trejo, V. and Jackson, D. 1998: Ca´nidos patago´nicos: identiﬁcacio´n taxono´mica de mandı´˜ irihua, XI Region). Anales del Instituto de bulas y molares del sitio Ban˜o Nuevo 1 (Alto N la Patagonia, 26, 181–94. Vela´squez, H. and Ada´n, L. 2002: Evidencias arqueofaunı´sticas del sitio Mariﬁlo-1. Adaptacio´n a los bosques templados de los sistemas lacustres cordilleranos del centro sur de Chile. Boletı´n de la Sociedad Chilena de Arqueologı´a, 33/34, 27–35. Villalba, R. 1994: Tree-ring and glacial evidence from the Medieval Warm Epoch and the Little Ice Age in southern South America. Climatic Change, 26, 183–97. Wobst, M. 1974: Boundary conditions for Paleolithic social systems: A simulation approach. American Antiquity, 39, 147–78. Yacobaccio, H. D. and Guraieb, G. 1994: Tendencia temporal de contextos arqueolo´gicos: Area del Rı´o Pinturas y zonas vecinas. In C. J. Gradı´n and A. M. Aguerre (eds), Contribucio´n a la Arqueologı´a del Rı´o Pinturas, Concepcio´n del Uruguay: Editorial Bu´squeda de Ayllu,13–28.

_____________________________________________________________________________________________________

158

______________________________________________________________________________

--------------------------------------------------------------

Part III

-----------------------------------------------------------

Interactions

The chapters in part three explore factors shaping desert societies in addition to environmental conditions. The operation of social and political processes, and their possible roles in shaping desert societies, is examined for different places and times. The articulation of these processes with economic strategy and ecological context is an additional theme displayed in this section. Anne Thackeray (Chapter 9) describes changes in Later Stone Age settlement in the southern African deserts. Kathryn Przywolnik (Chapter 10) explicates long-term transitions in hunter-gatherers of arid coastal northwestern Australia. Karim Sadr (Chapter 11) explores the complex interactions between Kalahari hunter-gatherers and encroaching Iron Age farmers and pastoralists. Mike Smith (Chapter 12) considers how independent data sources from disciplines as different as archaeology and linguistics might be reconciled to create more robust depictions of change in desert life. Calogero Santoro and colleagues (Chapter 13) describe the consequences of interactions between hunter-gatherer groups in different environments in northern Chile. Andrew Smith (Chapter 14) focuses on the role of ideological systems in the contacts between hunter-gatherers and pastoral nomads in North Africa. Finally, Alistair Paterson (Chapter 15) examines issues of culture contact during the historic period in Australia, examining interactions of desert hunter-gatherers and European pastoralists in Australian drylands. These chapters share a number of features that build upon the perspectives established in parts one and two of the book. A primary theme reproduced by the authors contributing to part three is the importance of social conﬁgurations and cultural interactions for the nature of emergent desert adaptations. Social mechanisms revealed as important in shaping desert life in these chapters include ideological constructions and the role of identity in negotiating cross-cultural

_____________________________________________________________________________________________________

159

______________________________________________________________________________

In te r a c ti o n s

contacts. While these chapters primarily discuss such processes during the late Holocene, it is likely that their operation extends beyond recent times, as McDonald (Chapter 7) makes clear. Furthermore, these chapters are explicit in representing these social processes as acting in conjunction with the environment to create the context in which desert societies operated and evolved.

_____________________________________________________________________________________________________

160

______________________________________________________________________________

-------------------------------------------------------------------------------

9

----------------------------------------------------------------------------

Perspectives on Later Stone Age Hunter-Gatherer Archaeology in Arid Southern Africa Anne I. Thackeray

Introduction Arid southern Africa comprises the regions of the western subcontinent embraced by the 500 mm isohyet which cover most of Botswana, Namibia, and the Northern Cape Province of South Africa (see Figure 9.1). Nomadic Later Stone Age (LSA) hunter-gatherer bands lived throughout these regions, although not continuously in any given area, during the past 25,000 years or so. The LSA was originally deﬁned in the late 1920s as consisting of several stone artifact industries which replaced the large triangular ﬂakes and long ﬂake-blades of the preceding Middle Stone Age (MSA) with a package of innovations. This included the production of small microlithic stone tools less than 25 mm long – which were often shaped by retouch (secondary working) and hafted in wooden handles – items made from organic materials like ostrich eggshell beads, worked leather and bone, ornaments, decoration on artifacts, rock art, and burial in formal graves (see also Chapter 4, this volume). The LSA was considered to have been the legacy of physically modern people who originated in North Africa and became southern African huntergatherers (Goodwin and Van Riet Lowe 1929). However, it is now established that anatomically modern people were present in southern Africa at least 100,000 years ago, long before the emergence of the LSA, and many of the items in the package are now known to have ﬁrst appeared in MSA times, while others became common only during the LSA. This has generated vigorous and ongoing debate about how to deﬁne modern behavior, and whether MSA people or only LSA people were behaviorally modern (Klein 2001). In the 1970s, radiocarbon dating pushed the antiquity of the LSA back into the late Pleistocene, but the dating of its beginnings remains unclear. The earliest claimed LSA from South Africa is a 38,000 bp microlithic stone industry featuring

_____________________________________________________________________________________________________

161

______________________________________________________________________________

An n e I . T h a c k e r a y ANGOLA

Nos

28S

Haalenberg Pokkenbank

BOTSWANA

Zebrarivier

iv

R po po m

mm isohyet

er

NAMIBIA

MOZ AMBI QUE

Mirabib

5 00

Kwihabe KALAHARI Makgadikgadi Pane

Erongo Mountains

24S

ZIMBABWE

Li

Dâures

IB

OO AR

-K MA NA

M NA 20S

ZAMBIA

White Paintings Tsodilo⎧ ⎨ Depression Hills⎩ Okavango Delta Dobe ≠Gi Ju/'hoansi Toteng 7

SOUTHAFRICA SWAZILAND

Wonderwerk

Apollo II

r ive

al R

Va

/!Garib O r a n g e R i ve r

e Cap rn/Xam e h t or

Border Cave

28S

LESOTHO

N

Springbokoog Se

32S

ow ac l l e y

Blydefontein Va Elands Bay Driekoppen

/Xam Bushman groups Sites/localities International boundaries Simplified biomes 500 mm isohyet Perennial streams and pans

Ca

pe Fo

32S

ld Mountains

500 km 20E

24E

Figure 9.1 Selected Later Stone Age archaeological sites and the boundaries of simpliﬁed biomes in arid southern Africa (after Rutherford and Westfall 1986).

bladelets and tiny cores, together with ostrich eggshell beads and bone points from Border Cave in KwaZulu-Natal Province. However, some consider this and other South African industries dating to between 40,000–30,000 bp as transitional between the MSA and LSA, with elements of both stages present (Wadley 1993, 1997). In arid southern Africa, the youngest MSA dates to shortly before 25,500 bp at Apollo 11 in southern Namibia and a similar age at Driekoppen in the Northern Cape, while LSA microlithic technology may already have been present in northwest Botswana since 33,000 bp (Brooks et al. 1990). Worked ostrich eggshell is found there at 31,880 bp (Robbins 1999), while microlithic assemblages appear on the southern fringes of arid southern Africa in the period between 22,000–17,000 bp (Parkington 1990). The LSA became well established in the region in the period between 25,000–20,000 bp, though the origins, spread, and reasons for the widespread adoption of this new technology which ushered in the last stage of the Stone Age are unknown. About 2,000 years ago, LSA pastoralists with domesticated animals, pottery, temporary settlements, and social hierarchies began to move into the better watered parts of western southern Africa, and shortly thereafter metal was

_____________________________________________________________________________________________________

162

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

introduced to the subcontinent by Iron Age mixed farmers who were spreading from the north into eastern southern Africa (see Chapter 11, this volume). Europeans reached the southern African coast by ship and made contact with the indigenous inhabitants of arid southern Africa from the late ﬁfteenth century onwards. They initially traded with the pastoralists at the coast, but soon became colonists and moved inland, taking over the land and water resources from the hunter-gatherers. The latter resisted and in many areas waged guerilla war on the colonists, who greatly feared their poisoned arrows and hunted them down on the grounds that they were no better than vermin and required extermination. Those close to European settlements were killed or reduced to servitude, resulting in the destruction of hunter-gatherer cultures and languages in these areas by 1910. Others in more remote regions and areas lacking reliable supplies of surface water continued their way of life longer, despite interaction with other cultures, and survived in parts of Botswana and Namibia into the twentieth century. Today, all inhabitants of arid southern Africa are part of the so-called global village and only a handful still subsist by hunting and gathering.

Names for People The indigenous hunter-gatherers of southern Africa were nomads living in small bands consisting essentially of a few siblings and their children – bands which aggregated in times of plenty and dispersed in times of hardship. In general, women gathered the dietary staples of plant foods, while men hunted, trapped, and snared wild animals. They spoke a number of mutually unintelligible languages and did not have a name to refer to themselves in groupings larger than linguistic units; however, early European colonists needed a term of reference and used the name ‘‘Bushmen,’’ which appeared in written records from the late seventeenth century onwards. Unfortunately, it acquired derogatory connotations and in the 1960s was replaced in academic literature with San (pronounced ‘‘Saan’’), from the word used by the indigenous pastoralist Nama people for hunter-gatherers. This term was nevertheless subsequently also shown to have pejorative overtones since it referred to a person of low social status too poor to own livestock. Indigenous people without livestock living in the Western Cape Province of South Africa during the 1650s were in contemporary records called Sonqua or Soaqua (variously translated as ‘‘Bush-people’’ or ‘‘thief ’’). Other names include the Bantu language Tswana term used by the government of Botswana, BaSarwa (‘‘Sarwa’’ meaning ‘‘Bushmen’’ and the preﬁx ‘‘Ba’’ signifying ‘‘people’’), which is sometimes used to refer to southern African hunter-gatherers not only within but also outside Botswana. Some of the surviving groups in Botswana and Namibia prefer being called ‘‘Bushmen’’ while others choose BaSarwa. Another term initially applied by Europeans to all indigenous people living in the Western Cape Province and later used only for pastoralists was ‘‘Hottentot,’’ apparently derived from a European attempt to reproduce a word sung during a dance. This also came to have abusive connotations and has been replaced by

_____________________________________________________________________________________________________

163

______________________________________________________________________________

An n e I . T h a c k e r a y

Khoekhoen (previous spelling Khoikhoin, the initial syllable sounding like a breathy pronunciation of the ﬁrst three letters of ‘‘quest’’), a Nama self-appellation meaning ‘‘people’’ or ‘‘men of men.’’ Seventeenth-century records indicate that pastoralists living in the Western Cape at the time called themselves Quena (pronounced like Khoe with the sufﬁx ‘‘-na’’ added for the plural). A collective term for both indigenous hunter-gatherers and pastoralists, Khoisan, was coined by Leonhard Schultze in 1928 to indicate biological similarities between these people. It continues to be used as a convenient shorthand reference not only for biology but also for language and culture, although it is recognized that it is a construct of dubious validity. For the want of a generic name coined by, and acceptable to, southern African hunter-gatherers themselves, and where it is not possible to refer to individual groups by their own names for themselves, in this chapter the English term ‘‘Bushmen’’ will be used without any implied denigration.

Perceptions of Bushmen Modern-day Bushmen living in Botswana and Namibia are popularly believed to be one of the last remaining links with the hunting and gathering way of life that humans have pursued for most of their existence on earth, and are arguably the world’s most studied Aboriginal people. A considerable degree of our understanding of the lifeways of Stone Age people not only in arid southern Africa during LSA times, but elsewhere and during earlier times, has been inspired by sometimes uncritically used information about these people. Historical accounts of Bushmen by early European travelers, hunters, and naturalists unfortunately generally recorded only chance encounters with the hunter-gatherers, which provide tantalizing but ﬂawed glimpses of their way of life. These Europeans regarded them as curiosities from pre-civilized times and took some of them to Europe to be displayed at freak shows for the public to gawk at the ‘‘primitive savages,’’ and especially at perceived peculiarities of their anatomy. One of the most notorious such sad cases involved not a Bushman, but a Khoekhoen, the so-called ‘‘Hottentot Venus,’’ Sarah Baartman, who became a symbol of racist, colonialist, and sexist oppression, and whose remains were returned to South Africa and ceremoniously buried in 2002. Dignity was accorded southern African hunter-gatherers as a result of detailed ethnographic studies of contemporary hunter-gatherers in Botswana and Namibia conducted since the 1950s (e.g., Barnard 1992). Accounts describing their foraging lifestyle as successful and egalitarian transformed them from the barely human savages of colonial times into the ‘‘Harmless People.’’ Although they were not direct descendants of people who once lived throughout southern Africa (or elsewhere), but of people who had lived in Botswana and Namibia for at least thousands of years, the late twentieth-century accounts of their societies were nevertheless considered to epitomize a non-violent afﬂuent and isolated life that must have existed in Stone Age times everywhere in the world before the advent of agricul-

_____________________________________________________________________________________________________

164

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

ture. Archaeologists voraciously raided the accounts for data that could be used to interpret the archaeological record. However, the data concentrated only on selected societies, especially those of the !Kung-speaking Ju/’hoansi, and there is no direct link between the people in the ethnographic accounts and those who left Stone Age archaeological remains. Indeed, the idea that the ethnographic accounts of the Bushmen of Botswana and Namibia made in the 1950s and 1960s faithfully reﬂect one of our last connections with our Stone Age hunter-gatherer past has been challenged since the 1980s, resulting in a bitter argument dubbed the Great Kalahari Debate (Kent 1992). Revisionists argue that the accounts fail to show the hunter-gatherers as a dispossessed underclass that has been wracked by violence and poverty at least since the widespread introduction of borehole water wells from the 1950s allowed pastoralists to move into hunter-gatherer territories (Wilmsen 1989). Moreover, the presence of Stone Age artifacts in Iron Age sites and vice versa is claimed to show that the hunter-gatherers were once part of an Iron Age economy, and that such contact ﬁrst with Iron Age and later with European farmers and traders precludes using their lifestyle as a model for pristine Stone Age society. Others point out great diversity in social organization and economic strategies among groups of hunter-gatherers and suggest that the relationships between the huntergatherers and agriculturalists varied from place to place and time to time, and that some could well have remained autonomous hunter-gatherers until comparatively recently. Yet others downplay the signiﬁcance of a few potsherds, pieces of metal, and bones of domesticated animals in Stone Age sites and point out that the LSA archaeology of Botswana remains under researched (Sadr 1997).

Windows on the Past This is not to deny that critical use of historical and modern Bushmen records can provide windows through which archaeologists can glimpse the LSA world of arid southern Africa. Such an example is a remarkable collection of some 12,000 pages of verbatim accounts of daily life, rituals, and beliefs by a group of /Xam Bushmen from the Northern Cape Province of South Africa made by a German linguist, Dr Wilhelm Bleek, and his sister-in-law, Lucy Lloyd, in the 1870s and 1880s (Deacon and Dowson 1996). The Bushmen informants were convicts who came to the attention of Bleek, who devised a script for representing the click sounds of the /Xam language. He and Lucy Lloyd recorded the convicts’ accounts in notebooks, with the /Xam text on one side and the English translation on the other, an extraordinary record of a language and culture that are now extinct. The ethnographic studies of the Bushmen of Botswana and Namibia from the 1950s sparked renewed interest in Bushman records and led to the rediscovery of the Bleek and Lloyd notebooks, which provided Lewis-Williams (1981, 1983) with a key to interpreting Bushman rock art. There are many thousands of examples of Bushman rock art throughout arid southern Africa, from paintings on the walls of rock shelters and caves in

_____________________________________________________________________________________________________

165

______________________________________________________________________________

An n e I . T h a c k e r a y

mountainous areas like the Tsodilo Hills of northwest Botswana and Daˆures (Brandberg) massif of central Namibia, to engravings on boulders and exposed rocks in countless localities in the Northern Cape and Namibia, as well as small slabs of painted or engraved stones found in archaeological deposits in caves. The rock paintings vary from being monochrome (one color) to bichrome and polychrome, while the engravings can be incised, scraped, or pecked. The subjects depicted are often wild or domesticated animals, scenes with human ﬁgures, creatures with both animal and human features called therianthropes, hand prints, and designs including dots, zigzags, wavy lines, and grid patterns. There are also historical scenes which provide datable images and indicate that rock art continued until recent times, although such scenes may not necessarily depict actual events. The oldest known dated examples of rock art in arid southern Africa are painted, broken hand-sized slabs found in the Apollo 11 Cave in southern Namibia in 1969 and 1972 by Wendt (1976). The slabs are associated with artifacts from the end of the MSA and radiocarbon dated to 27,500 bp, indicating that they are of the same order of age as the oldest known rock art in Europe. However, some prefer to be cautious and date the slabs to at least 18,500 bp; the age of the base of the overlying layer. The oldest dated engraving in the region is also a small, portable broken slab of rock, this time showing a 10,200 year old ﬁne-line engraving of an unﬁnished mammal without a head. It was excavated in 1979 by Francis and Anne Thackeray from LSA layers which accumulated around a huge stalagmite near the entrance to the impressive Wonderwerk Cave (see Figure 9.2) in the Northern Cape Province of South Africa (Thackeray et al. 1981).

Figure 9.2 Interior of the Wonderwerk Cave, Northern Cape Province, South Africa, where radiocarbon dated engraved slabs were recovered by Francis and Anne Thackeray in 1979.

_____________________________________________________________________________________________________

166

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

Initial attempts to interpret the art dismissed it as ‘‘art for art’s sake,’’ a quaint record of the life of a hunter-gatherer, or ‘‘hunting magic.’’ From the 1970s, however, the insights into the beliefs of the Bushmen obtained especially from the Bleek records, and corroborated by other ethnographic and historical records, enabled Lewis-Williams to revolutionize southern African rock art studies by placing the art within a Bushman rather than a European worldview. This enabled him to propose that the art is essentially religious and related to the trance experiences of shamans (medicine men and women) (Lewis-Williams and Dowson 1988). He argued that many otherwise inexplicable depictions, such as eland bleeding from the nose, are metaphors for the trance experience, and that geometric designs and strange creatures are related to three stages of hallucination experienced by shamans. His shamanic interpretation has, however, recently been challenged on the grounds that many images of apparent shamanic activity could otherwise be interpreted in terms of mythology about creation, death, and the afterlife (Solomon 2000), while Helvenston and Bahn (2002) claim that his three stages of hallucination could not have been naturally induced, but brought about only by certain plants whose availability in southern Africa has not been demonstrated. While one cannot assume that hunter-gatherers living all over arid southern Africa throughout the LSA had exactly the same beliefs as the /Xam or other recent Bushmen, nor that these beliefs remained constant over time, the fact that there are noticeable similarities in the style of their rock art and stone tools in comparison with regions further north suggests that there were communication networks and a common belief system, at least on a regional scale. Moreover, although modern hunter-gatherers in Botswana and Namibia do not produce rock art and some do not even attribute it to their ancestors, their beliefs and rituals are very similar to those of the /Xam Bushmen from the Northern Cape, hence the Bleek and other ethnographic and historical records can justiﬁably be used cautiously as keys to an insider’s view of the art. Lewis-Williams’ approach also highlighted the fact that there were social reasons for the choices people made, rather than the 1960s and 1970s emphasis on largely environmental conditions (Deacon 1990). The ethnographic record has consequently inspired archaeologists to look at sites not only in terms of available resources, but also as part of seasonal rounds of aggregation or dispersal, associated with different patterns of ritual activity, exchange of goods, and gender relations, and how these changed over time. The Ju/’hoansi practice a form of sharing and delayed kin-based gift exchange called hxaro during times of aggregation. This reinforces social and economic networks and is a form of insurance for times of hardship, as the giver can exploit resources in areas where hxaro trading partners live during such times. Beads are one of the most important items for exchange. Some consequently suggest that quantities of ostrich eggshell beads in certain sites indicate this practice has an antiquity of at least 40,000 years or even earlier, while others (Wadley 1993) consider that it may go back some 12,000 years, when it aided the survival of people expanding into the arid southern African interior at this time by cementing

_____________________________________________________________________________________________________

167

______________________________________________________________________________

An n e I . T h a c k e r a y

their social relationships. However, it is often impossible to identify items as necessarily hxaro items rather than originating in other forms of procurement or acquisition, and Bushman groups other than the Ju/’hoansi have different exchange systems.

Environments – Past and Present Social ties were the glue which held the fabric of LSA societies together, but people’s stomachs were probably ﬁlled mainly with plant foods. Although meat hunted by men was socially important for twentieth-century southern African hunter-gatherers, some 80 percent of their diet was provided by plants gathered mainly by women. Plant food remains from southern African LSA archaeological sites indicate that both above ground plant foods such as melons, fruits, and nuts, as well as geophytes (bulbs, corms, tubers, rhizomes, and rootstocks) were an important source of food also during LSA times (Deacon 1993).

Modern Environments The dominant plants can be used to subdivide present-day arid southern Africa into three generalized biomes (see Figure 9.1).

The Kalahari The Kalahari of Botswana and eastern Namibia is an extensive basin some 1,000 m above sea level (asl) on the interior plateau of southern Africa, stretching from the Orange or !Garib River in South Africa to southern Angola, and from the mountains of central Namibia to the hills of eastern Botswana and western Zimbabwe. Landforms and deposits in this biome have accumulated mostly under semi-arid to arid conditions and are capped by windblown sands that form dune systems currently stabilized by vegetation. Although the area is known as the Kalahari Desert and there is no permanent water except in the inland drainage systems of the Okavango Delta and Makgadikgadi Pans of northern Botswana, the erratic summer rainfall creates temporary pans and allows a variable tree, bush, and grass cover. At least since historical times, this has supported large herds of animals, including springbok, gemsbok (oryx), blue wildebeest, Burchell’s zebra, and warthog, and, in the wetter areas, elephant and buffalo.

The Namib Extending some 2,000 km along the western coast from southern Namibia to southern Angola, and 100–200 km inland, this is the driest region of southern Africa. It varies from rocky plains to sand seas and dune areas, as well as gravel plains with occasional hills and inselbergs. The coast is cool and foggy and receives less than 20 mm of rain annually, while areas further inland receive no more than

_____________________________________________________________________________________________________

168

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

130 mm summer rainfall. Vegetation is very sparse with trees conﬁned to the major river valleys, although grasses appear after rains and provide grazing for herds of springbok and gemsbok. Dassies (rock hyrax), hares, steenbok, klipspringer, and mountain zebra are also found.

The Nama-Karoo This biome consists of vast semi-arid rocky plains with mesas ranging between the Cape Fold Mountains and the Orange River in South Africa and a belt through western Namibia. The annual rainfall of approximately 150–400 mm falls mainly in summer in the wetter northeastern parts, but mostly in winter, spring, or autumn in the drier western parts. The vegetation comprises small shrubs and succulents, with sparse grass between bare ground, and trees (mostly acacias) restricted to watercourses. Animals present in historical times included dassies, hares, springbok, gemsbok, steenbok, grey duiker, black wildebeest, and quagga (an extinct zebra).

Past Environments These present environments of arid southern Africa are continually subject to change, as evidenced by a wide variety of data, such as geomorphological evidence from dunes, caves, and lakes, and biological evidence from animal and plant remains, including fossil pollens and charcoal. These data are difﬁcult to quantify precisely and results are usually reported in terms of a comparison with the present or another time period, such as ‘‘cooler’’ or ‘‘drier.’’ It is also often difﬁcult to tease apart which and to what extent different animal and human agents contributed to and biased biological evidence. Short-term change includes local spells of wet and dry years (Tyson 1986), while on a global scale during the past 800,000 years there has been a series of warmer and colder periods, documented in cores from deep sea sediments and ice caps, known as oxygen isotope stages (OIS). Although it is difﬁcult to slot evidence from non-continuous land-based records into the global record, available data provide generalized glimpses of conditions in arid southern Africa during the last three oxygen isotope stages which are relevant here (Deacon and Lancaster 1988). OIS3 between about 60,000–25,000 bp saw mostly cooler but changeable conditions, with parts of arid southern Africa such as the Kalahari at times experiencing very wet conditions. OIS2 between about 25,000–10,000 bp had a mean annual temperature of 5–68C colder than today, with maximum aridity at about 18,000 years ago (during the Last Glacial Maximum (LGM)), although the northern Kalahari and southern Namib experienced a shorter period of aridity, and there was a shallow lake in the Tsodilo Hills between about 17,500–15,000 bp, as well as evidence for high lake and pan levels in the Kalahari until about 11,000 bp (Brook et al. 1992; Robbins et al. 1996). OIS1 dating to the last 10,000 years saw changeable but essentially modern conditions, with temperatures at their highest about 7,000 bp.

_____________________________________________________________________________________________________

169

______________________________________________________________________________

An n e I . T h a c k e r a y

Patterns in the LSA Archaeological Record of Arid Southern Africa The Earliest LSA The ﬁrst traces of the LSA in arid southern Africa are microlithic assemblages containing bladelets from northwest Botswana, which date back to before the LGM at Depression Cave in the Tsodilo Hills (Robbins 1990) and possibly as early as 33,000 bp at 6¼Gi, slightly further south (Brooks et al. 1990). Two worked ostrich eggshell pieces from White Paintings rock shelter in the Tsodilo Hills are directly radiocarbon dated to 31,880 bp and 26,460 bp (Robbins 1999). Further south, early LSA assemblages which lack bladelets are known from Apollo 11, Nos, Haalenberg, and Pockenbank in southern Namibia, but these are not well described (Vogel and Visser 1981; Wendt 1976). Those from Apollo 11 date to ca. 20,000 bp and have small cores crushed on both ends as a result of splitting on an anvil with a hammerstone – referred to as bipolar cores – and large scraper-like tools, as well as ostrich eggshell and bone beads. Few formal tools or bladelets are present at Elands Bay on the coast of the Western Cape Province of South Africa between about 22,000–17,000 bp (Parkington 1990). These tantalizing glimpses of the ﬁrst LSA technology in arid southern Africa provide no insights into the reasons for its adoption or apparent regional variability, nor how it spread through the subcontinent. It has been suggested that the development of the weapon synonymous with southern African hunter-gatherers – the bow and arrow – played an important role in the switch from MSA to LSA technology. Bone points suitable for arrow inserts are indeed present in the southern African archaeological record from even before LSA times, but they become frequent only after about 12,000 bp. Moreover, their presence does not explain the bladelets, which would not have been suitable arrow inserts, and which are characteristic of many early LSA assemblages (see extended arguments in Chapter 4, this volume). In the period between about 18,000 bp until perhaps as late as 10,000 bp, in the southern coastal regions of South Africa, perhaps as far west as Elands Bay and as far inland as the southern margin of the Kalahari, assemblages characterized by quantities of bladelets produced with a punch technique from small cores, scrapers, and bone tools are recorded in cave sites. These assemblages comprise a readily identiﬁable regional and temporal stone artifact pattern called the Robberg Industry. Although bladelets continued to be produced at this time in northwest Botswana at sites like White Paintings (Robbins 1990) and until 11,000 bp at Kwihabe (Robbins et al. 1996), they do not occur in the quantities found in the southern sites. Assemblages dating to this period from Namibia are not well described, but do not appear to record quantities of bladelets (Shackley 1985). The dichotomy in stone artifact patterning noted for the earliest LSA between the northern and southern regions of arid southern Africa thus appears to continue beyond the LGM. Very little is known about the subsistence strategies of LSA people in arid southern Africa during OIS2. This was a time of generally colder and drier conditions when places such as northwest Botswana and southern Namibia have

_____________________________________________________________________________________________________

170

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

archaeologically visible populations and are thought to have experienced favorable local conditions which ensured supplies of standing water. There is no evidence that people lived at least temporarily in large territorial groups that moved over the landscape after herds of migrating animals, as was once thought the Robberg people did. Rather, new data for the Robberg and evidence such as that for the exploitation of freshwater ﬁsh in northwest Botswana at this time (Robbins et al. 1994) suggest that the early LSA people of arid southern Africa made good use of speciﬁc locations and employed expedient subsistence strategies. However, it seems that they were not tied to these locations or at least had contact with people further aﬁeld: chert used for bladelets at Depression Cave in northwest Botswana and hornfels used for artifacts at Elands Bay Cave originated some 100 km away from these sites.

New Environments and New Tools The period from ca. 12,000–7,000 bp saw the establishment of broadly modern environmental conditions in southern Africa. Many areas in the interior of southern Africa were occupied by LSA people for the ﬁrst time, while on the coast previously exposed areas of the continental shelf were drowned by rising sea levels. A few animals became extinct in arid southern Africa, including the Cape horse (Equus capensis) and the giant hartebeest (Megalotragus priscus), last recorded at Wonderwerk Cave at about 10,000 bp and 7,500 bp, respectively. There are, however, no data to indicate to what extent, if any, expanding human populations played a role in these extinctions. In South Africa, the microlithic bladelet technology of the early LSA was replaced by the non-microlithic Oakhurst Complex, a tradition with a preference for coarse-grained rocks, rare bladelets, and few retouched tools, mostly large sideand D-shaped scrapers and sometimes naturally backed knives. Larger numbers of polished bone points like those used as arrowheads by Bushmen in historical times, as well as other bone tools, are also found; hence it has been suggested that the bow and poison-tipped arrow came into widespread use at this time. Regional industries in arid southern Africa include the Kuruman at Wonderwerk Cave (Humphreys and Thackeray 1983) and in parts of the Free State and Northern Cape Provinces of South Africa, the Lockshoek, which has few dated shelter sites (e.g., Blydefontein) but is highly visible. Some 829 sites have been reported in a 5,000 sq km area of the Seacow Valley, these usually being located close to water sources (Sampson 1985). It has been suggested that the large scrapers of the Oakhurst originated in southern Namibia, where they continued to be made from the earliest LSA times at sites like Apollo 11, which has one of the earliest dated Oakhurst occurrences at 12,000 bp. However, the large scraper pattern did not spread to northwest Botswana, where the bladelet tradition continued at sites like Depression Cave and Kwihabe. There is more evidence for the use of plant foods at this time. In central Namibia, people at Zebrarivier and Mirabib collected the protein- and fat-rich

_____________________________________________________________________________________________________

171

______________________________________________________________________________

An n e I . T h a c k e r a y

!nara melon (Acanthosicyos horrida) seeds, which are still an important food source for hunter-gatherers in this area. Along the coast at Elands Bay Cave, food sources that come in small packages, like shellﬁsh, ﬁsh, rock lobster, seabirds, ostrich eggs, tortoises, and newborn small antelope, were exploited more intensively. The greater archaeological visibility of Oakhurst populations, especially the increase in the number of open sites and occupation of new areas at this time, is considered to indicate a population expansion. It is thought that this would probably have been coupled with changes in social organization and that traces of the social relations such as the hxaro exchange system recorded among twentieth-century Bushmen may now arguably be seen in the archaeological record for the ﬁrst time (Wadley 1993).

Microlithic Assemblages Southern African LSA assemblages post-dating 7,000 bp (or earlier) are known as the Wilton Complex, consisting of a number of regional industries and characterized by the systematic production of small scrapers thought to have been used to prepare leather, especially ‘‘thumbnail’’-shaped ones, and backed microliths such as segments and backed blades. There are more formal tools, more frequent use of ﬁne-grained rocks, and a greater quantity and variety of bone and shell ornaments, decoration, and rock art, as well as burials with grave goods such as ostrich eggshell beads and shell pendants. Classic Wilton assemblages dominated by small scrapers date to between about 7,000–4,000 bp, and thereafter backed blades and sometimes adzes – presumed woodworking tools with steep retouch – become more frequent. The earliest dates for the Wilton Complex are from Apollo 11 at 10,420 bp from the very base of the layer and 9,430 bp slightly higher in the layer. One of the richest Wilton sites in arid southern Africa is Wonderwerk Cave (Humphreys and Thackeray 1983), where the Wilton Complex makes its appearance about 8,500 bp and records changes in style and frequencies of scrapers, backed microliths, and raw materials over the succeeding millennia, as well as unusual artifacts like stone pendants and rings, bone, shell, and wooden items, and the engraved stones mentioned above. Specularite, a greasy, silvery iron-ore available from only a few localities in the region and used in historical times as a cosmetic, occurs throughout the Wilton layers. Although large grazing animals were consumed by the occupants of the cave, small game such as porcupines, dassies, hares, and tortoises were an important part of their diet. Microfaunal remains indicate that conditions were dry during the ﬁrst few millennia of Wilton occupation, but that environments like those of the present developed after about 5,000 bp. The richest LSA archaeological record of the post-7,000 bp period in southern Africa is from the Cape Fold Mountains south of the arid areas. It is thought that high temperatures and extreme aridity played a role in limiting settlement in the arid areas, especially during the ﬁrst few millennia of this period. However, it also seems that people chose strategies to live sufﬁciently successfully in some such areas to leave archaeologically visible traces, and that the apparent non-occupation of some sites, such as the disuse of Elands Bay Cave between 7,900–4,300 bp, is

_____________________________________________________________________________________________________

172

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a

probably a consequence of local rather than regional conditions. The occupation of sites in northwest Botswana, such as the rich backed scraper occurrences between 3,650–3,200 bp and after 2,260 bp at Dobe, the hunting of lechwe, common reedbuck, and bushpig, and the use of barbed bone points to exploit aquatic resources at White Paintings, are indeed probably linked to wetter episodes. Nevertheless, increasing aridiﬁcation after about 3,200 bp at Toteng 7 in the southern Okavango Delta actually resulted in intensiﬁed ﬁshing probably because ﬁsh became easier to catch as the water became shallower (Robbins et al. 1998). The inselbergs of the Erongo Mountains in Namibia trapped water and were occupied from about 6,000 bp, as were shelters in the Daˆures. There is, however, no evidence for sustained occupation in vast tracts of the Nama-Karoo biome between some 7,000–4,500 bp. Southern Namibia seems to have been unoccupied between 5,100–2,300 bp, although the adjacent area of Bushmanland in the northwest Northern Cape Province of South Africa was occupied episodically by makers of the backed blades of the Springbokoog Industry between 4,500–4,300 bp and 2,600–2,300 bp. Wilton people in the Seacow Valley opted for sites with commanding views suitable for game spotting and were not as closely tied to water sources as their Lockshoek predecessors, perhaps because they made more use of storing water in ostrich eggshell water ﬂasks. It has also been suggested that people along the west coast had the best of two worlds by moving seasonally between the coast and the interior (Parkington 2001), though this is not supported by chemical studies of skeletal remains (Sealy and Van der Merwe 1988).

Concluding Perspectives Arid southern Africa is traditionally regarded as a marginal area, not only in present-day environmental terms, but also in terms of its archaeological record. LSA research in the subcontinent has focused on rich sites in the wetter southernmost areas, while those in the arid areas to the north tend to be seen in terms of ephemeral occupation pockets, such as northwest Botswana, or individual sites like Apollo 11 or Wonderwerk, where local conditions are thought to have been sufﬁciently favorable to permit settlement and the accumulation of archaeological deposits. Indeed, a 1980s map showing numerous signiﬁcant southern African LSA archaeological sites did not include a single one from Botswana (Deacon 1984: 230). Although the situation in some areas like northwest Botswana is improving, the LSA archaeology of arid southern Africa remains sidelined. This is unfortunate because many of the innovations in the LSA archaeological record of the subcontinent are ﬁrst recorded from sites located in arid southern Africa. Seasonal and short-term changes in the availability of plant foods have nutritional and demographic consequences for hunter-gatherers and long-term changes in the distribution and yield of these foods would have played a role in LSA demographic patterns and the choice of sites. However, archaeological studies of arid areas like the Seacow Valley show that people could nevertheless make choices about how to subsist and organize themselves, even in the harshest of

_____________________________________________________________________________________________________

173

______________________________________________________________________________

An n e I . T h a c k e r a y

environments. This contrasts with interpretations of change in the LSA record of the wetter areas, where changes in artifacts and food remains have often been tied to climatic and environmental changes. Even though there appears to be a time lag of the order of thousands of years between dates of marked environmental change and the appearance of new technology in these areas, the archaeological remains are still considered to be a delayed response to ‘‘social stress’’ caused by the changes. However, this does not explain why the artifacts took the form they did. There is indeed an increasing awareness that people make choices for personal, social, and historical reasons as much as for the need to feed themselves, and that they did not necessarily opt for what a twenty-ﬁrst century archaeologist would consider the most advantageous. Nevertheless, archaeologists still tend to write about the ‘‘limits’’ imposed by the environment and how people ‘‘coped,’’ rather than focusing on the possibilities and the choices made. Perhaps more insight into LSA lifeways could be gained by viewing the environment not as the prime mover in changing patterns in the LSA archaeological record, but as the scenery behind the stage on which the action took place. The makers of the remarkable LSA archaeological record could then become ﬂesh and blood characters, rather than shadowy ﬁgures jerked by the strings of changing environments, and their relationship to the indigenous hunter-gatherers of the historical and ethnographic accounts could become better understood. Indeed, although southern African museums no longer place displays of Indigenous hunter-gatherers in natural history sections, they still tend to be presented romantically and ahistorically as ‘‘living fossils’’ from the deep human past, a view that is encouraged by movie makers and the tourism industry. The LSA archaeology of arid southern Africa is thus uniquely charged to contribute to the story of both past and present communities.

References Barnard, A. 1992: Hunters and Herders of Southern Africa: A Comparative Ethnography of the Khoisan Peoples. Cambridge: Cambridge University Press. Brook, G. A., Haberyan, K. A., and De Filippis, S. 1992: Evidence of a shallow lake at Tsodilo Hills, Botswana, 17,500 to 15,000 yr bp: Further conﬁrmation of a widespread late Pleistocene humid period in the Kalahari Desert. Palaeoecology of Africa, 23, 165–75. Brooks, A. S., Hare, P. E., Kokis, J. E., Miller, G. H., Ernst, R. D., and Wendorf, F. 1990: Dating Pleistocene archaeological sites by protein diagenesis in ostrich eggshell. Science, 248, 60–4. Deacon, H. J. 1993: Planting an idea: An archaeology of Stone Age gatherers in South Africa. South African Archaeological Bulletin, 48, 86–93. Deacon, H. J. and Deacon, J. 1999: Human Beginnings in South Africa: Uncovering the Secrets of the Stone Age. Cape Town: David Philip. Deacon, J. 1984: Later Stone Age people and their descendants in southern Africa. In R. G. Klein (ed.), Southern African Paleoenvironments and Prehistory, Rotterdam: A. A. Balkema, 221–328. Deacon, J. 1990: Weaving the fabric of Stone Age research in southern Africa. In P. T. Robertshaw (ed.), A History of African Archaeology, London: James Currey, 39–58.

_____________________________________________________________________________________________________

174

______________________________________________________________________________

T h e L at er S t o n e A g e o f A r i d S ou thern A fr ic a Deacon, J. and Dowson, T. A. (eds) 1996: Voices from the Past: /Xam Bushmen and the Bleek and Lloyd Collection. Johannesburg: Witwatersrand University Press. Deacon, J. and Lancaster, N. 1988: Late Quaternary Paleoenvironments of Southern Africa. Oxford: Oxford University Press. Goodwin, A. J. H. and Van Riet Lowe, C. 1929: The Stone Age cultures of South Africa. Annals of the South African Museum, 27, 1–289. Helvenston, P. A. and Bahn, P. G. 2002: Desperately Seeking Trance Plants: Testing the ‘‘Three Stages of Trance’’ Model. New York: RJ Communications LLC. Humphreys, A. J. B. and Thackeray, A. I. 1983: Ghaap and Gariep: Later Stone Age Studies in the Northern Cape. Cape Town: South African Archaeological Society. Kent, S. 1992: The current forager controversy: Real versus ideal views of hunter-gatherers. Man, 27, 45–70. Klein, R. G. 2001: Southern Africa and modern human origins. Journal of Anthropological Research, 67, 1–16. Lewis-Williams, J. D. 1981: Believing and Seeing: Symbolic Meanings in Southern San rock Paintings. London: Academic Press. Lewis-Williams, J. D. 1983: The Rock Art of Southern Africa. Cambridge: Cambridge University Press. Lewis-Williams, J. D. and Dowson, T. A. 1988: Images of Power: Understanding Bushman Rock Art. Johannesburg: Southern Books. Mitchell, P. 2002: The Archaeology of Southern Africa. Cambridge: Cambridge University Press. Parkington, J. E. 1990: A view from the south: Southern Africa before, during, and after the Last Glacial Maximum. In C. S. Gamble and O. Soffer (eds), The World at 18,000 bp, Volume 2: Low Latitudes, London: Unwin Hyman, 214–28. Parkington, J. E. 2001: Mobility, seasonality, and southern African hunter-gatherers. South African Archaeological Bulletin, 56, 1–7. Robbins, L. H. 1990: The Depression site: A Stone Age sequence in the northwest Kalahari Desert, Botswana. National Geographic Research, 6, 329–38. Robbins, L. H. 1999: Direct dating of worked ostrich eggshell in the Kalahari. Nyame Akuma, 52, 11–16. Robbins, L. H., Murphy, M. L., Campbell, A. C., Brook, G. A., Reid, D. M., Haberyan, K. H., and Downey, W. S. 1998: Test excavation and reconnaissance paleoenvironmental work at Toteng, Botswana. South African Archaeological Bulletin, 53, 125–32. Robbins, L. H., Murphy, M. L., Stevens, N. J., Brook, G. A., Ivester, A. H., Haberyan, K. A., Klein, R. G., Milo, R., Stewart, K. M., Matthiesen, D. G., and Winkler, A. J. 1996: Paleoenvironment and archaeology of Drotsky’s Cave: Western Kalahari Desert, Botswana. Journal of Archaeological Science, 23, 7–22. Robbins, L. H., Murphy, M. L., Stewart, K. M., Campbell, A. C., and Brooks, G. A. 1994: Barbed bone points, paleoenvironment, and the antiquity of ﬁsh exploitation in the Kalahari Desert, Botswana. Journal of Field Archaeology, 21, 257–64. Rutherford, M. C. and Westfall, R. H. 1986: Biomes of Southern Africa – An objective classiﬁcation: Pretoria. Memoirs of the Botanical Survey of South Africa, 54, 1–98. Sadr, K. 1997: Kalahari archaeology and the Bushman debate. Current Anthropology, 38, 104–12. Sampson, C. G. 1985: Atlas of Stone Age settlement in the Seacow Valley. Memoirs of the National Museum Bloemfontein, 20, 1–116.

_____________________________________________________________________________________________________

175

______________________________________________________________________________

An n e I . T h a c k e r a y Sealy, J. C. and Van der Merwe, N. J. 1988: Social, spatial, and chronological patterning in marine food use as determined by d13C measurements of Holocene human skeletons from the southwestern Cape, South Africa. World Archaeology, 20, 87–102. Shackley, M. L. 1985: Palaeolithic archaeology of the central Namib Desert. Cimbebasia, 6, 1–84. Solomon, A. 2000: On different approaches to San rock art. South African Archaeological Bulletin, 55, 77–8. Thackeray, A. I., Thackeray, J. F., and Beaumont, P. B. 1981: Dated rock engravings from Wonderwerk Cave, South Africa. Science, 214, 64–7. Tyson, P. D. 1986: Climatic Change and Variability in Southern Africa. Cape Town: Oxford University Press. Vogel, J. C. and Visser, E. 1981: Pretoria radiocarbon dates II. Radiocarbon, 23, 43–80. Wadley, L. 1993: The Pleistocene Later Stone Age south of the Limpopo River. Journal of World Prehistory, 7, 243–96. Wadley, L. 1997: Rose Cottage Cave: Archaeological work 1987 to 1997. South African Journal of Science, 93, 439–44. Wendt, W. E. 1976: ‘‘Art mobilier’’ from the Apollo 11 Cave, South West Africa: Africa’s oldest dated works of art. South African Archaeological Bulletin, 31, 5–11. Wilmsen, E. N. 1989: Land Filled with Flies: A Political Economy of the Kalahari. Chicago: University of Chicago Press.

_____________________________________________________________________________________________________

176

______________________________________________________________________________

-----------------------------------------------------------------------------

10

--------------------------------------------------------------------------

Long-Term Transitions in Hunter-Gatherers of Coastal Northwestern Australia Kathryn Przywolnik

Introduction This chapter discusses transitions in patterns of hunter-gatherer occupation and use of coastal northwest Australia, with particular reference to archaeological research undertaken in northern Cape Range Peninsula as a case study (see Figure 10.1). Two rock shelter sites, Jansz and C99, have occupational sequences dating to 35,000 bp and 34,000 bp, respectively, which is the earliest evidence for human site use on the Western Australian coast. This area is of particular interest to studies of prehistoric hunter-gatherers worldwide, as there is a widespread assumption that long-term trends in hunter-gatherer economies are unidirectional, and therefore irreversible. In Australia, this metanarrative of unidirectional change in huntergatherer economies has been almost universally assumed to be one associated with an ‘‘intensiﬁcation’’ phase in the mid-to late Holocene, after which huntergatherers are understood to have developed similar economies to those that were recorded at the time of contact with Europeans during the eighteenth and nineteenth centuries. Long-term patterns in the archaeology of the study area demonstrate a complex series of changing patterns of land use in which hunter-gatherer economies become at ﬁrst more sedentary during the early Holocene, and after a mid-Holocene abandonment associated with mangrove environmental decline, ‘‘revert’’ to a more mobile specialist land use strategy. The data discussed here are considered to support Rowley-Conwy’s (2001) ‘‘adaptationist’’ view of longterm hunter-gatherer change, which perceives social change as non-directional and reversible, and a ﬂexible response to adaptive necessity rather than an inevitable move toward increasing complexity.

_____________________________________________________________________________________________________

177

_____________________________________________________________________________

Figure 10.1 Map of northwestern Australia and archaeological sites mentioned in the text. Inset: map of study area in northern Cape Range Peninsula.

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

Patterns of Hunter-Gatherer Change in Northern Cape Range Peninsula Cape Range Peninsula Environmental Context The case study area for this chapter is the northern tip of Cape Range Peninsula, an area of approximately 20 km by 10 km encompassing considerable environmental diversity (see Figure 10.1). Cape Range Peninsula is the westernmost extent of the Australian mainland, an 80 km long extension of rugged limestone range and coastal plain jutting into the Indian Ocean. Cape Range is the central limestone spine of the peninsula, which is accompanied on both the west and east sides by lower lying foothills and extensive coastal plains. The peninsula is ﬂanked by Ningaloo Reef to the west and Exmouth Gulf, a large, shallow, mangrove-edged embayment, to the east. Ningaloo Reef is a Holocene formation that is thought to have developed as the sea level rose following the Last Glacial Maximum (LGM) (Hatcher 1991). The reef is uniquely situated, being in very close proximity to both the mainland and the continental shelf, and is renowned for its marine biodiversity. The western coastline is dominated by dune systems and limestone reef, with alternating stretches of sandy bay interspersed along the predominantly rocky coast. On the eastern side of the peninsula, the coastline is less rocky, and the shallow waters of Exmouth Gulf are rimmed with sandy beaches and coastal dune systems. Cape Range Peninsula is also one of the closest points of the Australian mainland to the continental shelf, which is located only 12 km offshore. Cape Range is the westernmost extent of the semi-arid/arid zone of Western Australia, and is an extremely dry and harsh environment. It is also at the southern boundary of the tropics, and although the Cape Range area does not experience the strong seasonal monsoon characteristic of the north of Australia, there are distinct seasonal differences between the summer and winter months. The summer months between October and March are hot and dry, and are also the tropical cyclone season, when cyclonic high pressure systems originating from further north often descend to the peninsula, bringing heavy rains and strong winds. The winter months between April and September are cooler and on average wetter, while the months between September and December are typically dry.

Previous Archaeological Research in Cape Range Peninsula Cape Range Peninsula was archaeologically unknown until systematic archaeological research was ﬁrst carried out by Morse during the late 1980s and early 1990s (Morse 1993a). Her research focused on the analysis of excavations and surface collections from three stratiﬁed rock shelters and four unstratiﬁed coastal middens located within Cape Range National Park towards the middle and southern end of the range, and documented some of the oldest human occupation sites in Western Australia (Morse 1993b, 1993c, 1996, 1999). Morse’s study aimed primarily to establish a regional sequence for occupation of the region, and provide a model describing the subsistence strategies of past peoples in Cape Range Peninsula.

_____________________________________________________________________________________________________

179

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

Morse interpreted two main phases of occupation separated by a hiatus coinciding with the LGM, a pattern that was consistent between the three rock shelters. Dates obtained on marine shell show evidence of human occupation from at least 34,200 bp at Mandu Mandu Creek rock shelter, 17,410 bp at Pilgonaman Creek rock shelter (although dating at this site proved to be problematic), and 10,490 bp at Yardie Well rock shelter. The excavations uncovered well preserved archaeological assemblages, particularly in the Holocene layers, including artifactual stone, ocher, marine and terrestrial fauna, and an extremely rare example of body ornamentation in the form of shell beads dated to ca. 32,000 bp from the basal unit of an excavation square in Mandu Mandu Creek (Morse 1993d). Morse argued the archaeological evidence demonstrates a sophisticated and specialized utilization of marine resources from the earliest occupation in the Pleistocene when the coast was ca. 10–12 km away, which increased in intensity and frequency once the sea level ameliorated during the mid-to late Holocene (Morse 1993c). The midden sites investigated by Morse (1996: 22) were seen to provide supporting evidence to the overall importance of the coast to the human occupation, and ‘‘clearly demonstrate an ongoing and continued use of the coast and its resources from the beginning of the Holocene to recent times.’’

Northern Cape Range Archaeological Sites: Rock Shelters In 1997–8 I excavated a series of three rock shelter sites and made surface collections from eight shell middens in the study area at the northern tip of Cape Range Peninsula (Przywolnik 2002a). The rock shelters, Wobiri, C99, and Jansz, are three of twelve such sites in a small section of the foothills located approximately 5 km from Vlaming Head at the northern tip of Cape Range Peninsula (see Figure 10.1). The complex of shelters is just over 2 km from the current coast, in the lower to middle reaches of the western slopes of the range. Jansz and Wobiri are both small shelters, while C99 is a much larger, multichambered cave. The walls and ceiling of C99 are covered in ocher and charcoal rock art. Jansz and C99 share similar stratigraphic proﬁles: gray charcoal-rich, Holocene deposit overlaying deep red Pleistocene sands, containing between 70 and 90 cm of occupational deposit. The Wobiri deposit contains only the upper section of this proﬁle, and has gray Holocene sands only, followed by sterile decomposing limestone. The dates obtained from the three sites are summarized in Table 10.1.

Cape Range Pleistocene Occupation The combined archaeological evidence from Wobiri, C99, and Jansz paints an interesting picture of human occupation in Cape Range Peninsula. The regional sequence for the study area is summarized in Table 10.2. Humans ﬁrst became archaeologically visible in the northern Cape Range landscape approximately 35,000 years ago, when they sporadically sheltered in sites in the western foothills of the limestone slopes. The Pleistocene archaeological remains, particularly the faunal remains, from both Jansz and C99 are extremely fragmentary and small in

_____________________________________________________________________________________________________

180

_____________________________________________________________________________

Site

Koolan Shelter 2 Koolan Shelter 2 Koolan Shelter 2 Koolan Shelter 2 Koolan Shelter 2 Koolan Shelter 2 Widgingarri Shelter 1 Widgingarri Shelter 1 Widgingarri Shelter 1 Widgingarri Shelter 1 Widgingarri Shelter 1 Widgingarri Shelter 2 Widgingarri Shelter 2 Widgingarri Shelter 2 High Cliffy Island Shelter High Cliffy Island Shelter High Cliffy Island Shelter HC-3 HC-3 HC-3 HC-2

Location

South West Kimberely

Wk 1287 Wk 1372 Wk 1098 Wk 1099 Wk 1365 Wk 1366 Wk 110 Wk 1490 ANU/AMS 5–10 R11720 R11795 Wk 1397 Wk 1398 Wk 1101 Wk 1286 Wk 1257 Wk 1096 Wk 1863 Wk 2460 Wk 1864 Wk 1095

Lab Code 0–5 cm 22–27 cm 47–52 cm 48–53 cm 64–69 cm 64–69 cm 19–22 cm 30–34 cm ˜75 cm 80–89 cm 80–89 cm 15–20 cm 35–41 cm 45–50 cm 0–5 cm 5–10 cm 15–20 cm 35 60 70 0–1 cm

Context 630 3,710 23,900 10,550 26,500 14,400 320 1,700 18,900 modern 28,060 1,510 4,970 7,780 2,620 2,740 3,210 modern 650 6,690 370

Date (BP) charcoal marine shell charcoal marine shell marine shell charcoal charcoal marine shell marine shell charcoal marine shell marine shell marine shell marine shell marine shell marine shell marine shell charcoal charcoal charcoal marine shell

Sample rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter open site open site open site open site

Site type

(Continued )

O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a O’Connor 1999a

Reference

Table 10.1 Radiocarbon dates obtained from archaeological sites in northwestern Australia. Note that all dates shown are uncalibrated.

Noala 1 Noala 1 Noala 2 Haynes Cave Haynes Cave Haynes Cave Haynes Cave Haynes Cave Haynes Cave Haynes Cave Haynes Cave Haynes Cave North West Cape Midden 2 Babjarrimannos Midden 1 Babjarrimannos Midden 2 C99 SQ2 C99 SQ2 C99 SQ2 C99 SQ2 C99 SQ2 C99 SQ2 C99 SQ2

Monte Bello Islands

Northern Cape Range Peninsula

Site

Location

Table 10.1 (cont.)

Wk 2912 Wk 2913 WK 2905 Wk 2915 Wk 2910 Wk 2911 Wk 2906 Wk 2914 Wk 2907 Wk 2908 Wk 2909 Wk 2719 Wk 8468 Wk 8467 Wk 8466 Wk 8930 Wk 8928 Wk 8929 Wk 8927 NZA 10528 Wk 8926 Wk 8925

Lab Code Spit 1 Spit 6 Spit 7 Spit 6 Spit 1 Spit 7 Spit 1 Spit 3 Spit 5 Spit 8 Spit 12 Spit 13 surface surface surface 9 15 18 24 54 67 75

Context 8,730 10,030 27,220 7,930 7,810 8,240 7,460 8,090 7,630 7,890 7,820 7,560 4,830 960 5,620 2,110 650 1,980 8,010 28,050 31,570 33,930

Date (BP)

Terebralia Baler Baler Baler Baler Baler Baler Baler Baler Baler

Sample rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter Midden Midden Midden rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter

Site type

Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Veth 1995 Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a

Reference

C99 SQ3 C99 SQ3 C99 SQ3 C99 SQ3 C99 SQ3 C99 SQ3 C99 SQ3 Jansz SQ1 Jansz SQ2 Jansz SQ2 Jansz SQ2 Jansz SQ2 Jansz SQ2 Jansz SQ2 Jansz SQ2 Wobiri Giralia Road site 11

Giralia Road site 9

Mangrove Bay Coral Bay

Southern

Cape Range

Peninsula

Beta 26269 WK 1729

Wk 2859

Wk 8937 Wk 8936 Wk 8935 Wk 8934 Wk 8933 Wk 8932 Wk 8931 Wk 6321 Wk 8923 Wk 8924 Wk 8922 Wk 8920 NZA 10566 Wk 8921 Wk 8919 Wk 8611 Wk 2858

surface surface

surface

11 16 22 50 60 75 81 44 8 21 27 42 44 67 73 44 surface

1,092 5,620

1,640

1,510 900 2,430 7,210 21,110 29,450 33,480 2,360 710 920 9,900 10,730 31,360 34,490 35,230 5,160 1,528

marine shell marine shell

marine shell

Baler Baler Baler Terebralia Baler Baler Baler Baler Baler Baler Baler Terebralia Baler Baler Baler Baler marine shell

rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter surface scatter surface scatter midden midden (Continued )

Morse 1993a Morse 1993a

Morse 1993a

Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Przywolnik 2002a Morse 1993a

Location

WK 1430 WAIT 116 Wk 1512 WAIT 117 Wk 1511 SUA 2614 WK 1575 SUA 2354 Wk 1513 Wk 1576 Wk 1429 Wk 1474 R16098/2 Wk 2524 Wk 1520 R1698/1 R11879/1 ARL 245 WAIT 118

Creek Creek Creek Creek Creek Creek Creek Creek Creek

Low Point

Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mandu Mulanda Bluff

Pilgonaman Creek Pilgonaman Creek Pilgonaman Creek Pilgonaman Creek Pilgonaman Creek Pilgonaman Creek Tulki Well Turquoise Bay north

Lab Code Wk 1729

Site

Coral Bay

Table 10.1 (cont.)

30–35 cm 40–45 cm 50–55 cm 60–65 cm 60–65 cm 70–75 cm surface 25–35 cm

10 cm 25–30 cm 30–35 cm 35–40 cm 40–45 cm 65–70 cm 81 cm 85–90 cm 91 cm surface

surface

surface

Context

480 10,150 12,100 9,990 31,770 17,410 5,660 5,430

430 1,960 2,420 5,490 20,040 22,100 25,200 34,200 30,000 7,210

4,820

6,270

Date (BP) mangrove gastropod mangrove gastropod charcoal marine shell marine shell marine shell marine shell marine shell carbonate marine shell marine shell mangrove gastropod charcoal marine shell marine shell marine shell marine shell marine shell marine shell marine shell

Sample

rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter midden midden

rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter rock shelter midden

midden

midden

Site type

Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a

Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a Morse 1993a

Morse 1993a

Morse 1993a

Reference

Peron

Shark Bay:

Island Shark Bay: Useless Loop

Dirk Hartog

Wk 2755 Wk 2754 Wk 2439 Wk 2438 GS/CC 290 GS/CC 291 GS/CC 292 GS/CC 293 Beta 31499

Useless HK

Heirisson Prong 1 Heirisson Prong 2 Whale Well Whale Well Whale Well Whale Well Whale Well Eagle Bluff Well 2 ANU 7459

Wk 2562

Useless Inlet 4

Eagle Bluff

Wk 2996 Wk 2437 Wk 2561

WK 1476 Wk 1477 R11519/2 Wk 2997

Yardie Well Yardie Well Yardie Well Notch Point

Notch Point Useless Inlet 2 Useless Inlet 3

SUA 1735

Warroora

22 cm

surface surface 24 cm surface surface surface surface surface

surface

42–53 cm

surface

surface

30–35 cm 55–60 cm 95–100 cm surface

surface

4,690

5,980 4,890 1,820 1,240 1,510 190 190 3,970

7,400

6,620

5,160 2,300 7,020

5,690 7,290 10,490 7,460

7,810

mangrove gastropod marine shell marine shell marine shell mangrove gastropod marine shell Baler mangrove gastropod mangrove gastropod mangrove gastropod marine shell marine shell marine shell marine shell marine shell marine shell marine shell mangrove gastropod mangrove gastropod midden

midden midden midden midden midden midden midden midden

midden

midden

midden midden midden

rock shelter rock shelter rock shelter midden

midden

(Continued )

O’Connor 1999b

O’Connor 1999b O’Connor 1999b O’Connor 1999b O’Connor 1999b O’Connor 1999b O’Connor 1999b O’Connor 1999b O’Connor 1999b

O’Connor 1999b

O’Connor 1999b

O’Connor 1999b Bowdler pers.com. O’Connor 1999b

Morse 1993a Morse 1993a Morse 1993a O’Connor 1999b

Morse 1993a

Peninsula

Location

Beta 31498 Wk 7608 Beta 31497 Wk 6349 Wk 7609 ANU 7457 ANU 7456 Wk 2433 Wk 2434 Wk 2436 Wk 2435 ANU 7105 ANU 8223 ANU 7106

Eagle Bluff

Eagle Bluff Denham Site 9 Denham Burial 1 Silver Dollar Silver Dollar

Silver Dollar

Silver Dollar

Silver Dollar

Silver Dollar

Silver Dollar

Silver Dollar Silver Dollar Silver Dollar

Lab Code WAIT 141

Site

Eagle Bluff

Table 10.1 (cont.)

30–40 cm 73 cm 30–40 cm

9 cm

24 cm

9 cm

23 cm

0–1 cm

57 cm

surface surface

surface

surface

Context

18,730 20,160 19,750

7,360

7,290

7,260

7,240

6,950

4,880 3,460 430 4,940 6,640

4,870

4,690

Date (BP) mangrove gastropod mangrove gastropod marine shell marine shell Baler marine shell mangrove gastropod mangrove gastropod mangrove gastropod mangrove gastropod mangrove gastropod mangrove gastropod marine shell marine shell emu eggshell

Sample

open site open site open site

midden

midden

midden

midden

midden

midden midden open site midden midden

midden

midden

Site type

Bowdler 1999 Bowdler 1999 Bowdler 1999

Bowdler 1999

Bowdler 1999

Bowdler 1999

Bowdler 1999

Bowdler 1999

Bowdler pers.com. O’Connor 1999b Bowdler pers.com. Bowdler pers.com. Bowdler 1999

O’Connor 1999b

O’Connor 1999b

Reference

Dampier Archipelago

South of Shark Bay

ANU 8220 WATT 139 WAIT 140 Beta 22535 Beta 22534

Silver Dollar

Monkey Mia 1 Monkey Mia 2 Zuytdorp gully site

Zuytdorp clifftop site Wk 3460 Wk 3461 Wk 3462 Wk 3339 Wk 3337 Wk 3338 Wk 3345 Wk 3346 Wk 3347 Wk 3348 Wk 2647 Wk 2648 Wk 1780 Wk 2480 Wk 3210 Wk 3343 Wk 2650

ANU 8222

Silver Dollar

Wadjuru Rockpool Wadjuru Rockpool Wadjuru Rockpool Wadjuru Rockpool Clarke’s Cave Clarke’s Cave Anadara Mound Anadara Mound Anadara Mound Anadara Mound Anadara Shelter Anadara Shelter Nickol River Mound Nickol River Mound Nickol River Mound Nickol River Mound Not-so-Secret Shelter

ANU 7458

Silver Dollar

0.4 m 0.55 m 0.8 m 1.2 m 0.0 m 1.2 m 0.0 m 0.5 m 0.4 m 1.25 m 0.3 m 0.7 m 0.0 m 0.5 m 1.2 m 2.38 m 0.55 m

93 cm 60 cm

102 cm

86 cm

110 cm

6440 6530 6840 8520 1470 7140 2270 4270 6290 6510 4240 6380 1780 2480 3210 4090 6080

5,080

680 1,010 4,450

22,540

21,840

21,700

emu eggshell emu eggshell emu eggshell charcoal charcoal mangrove gastropod mangrove gastropod Terebralia Terebralia Terebralia Terebralia Anadara Terebralia Anadara Anadara Terebralia Terebralia Anadara Terebralia Anadara Anadara Anadara Anadara Terebralia open site open site open site open site rock shelter rock shelter open site open site open site open site rock shelter rock shelter open site open site open site open site rock shelter

midden

rock shelter rock shelter midden

open site

open site

open site

Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995 Bradshaw 1995

O’Connor 1999b

Bowdler 1995 Bowdler 1995 O’Connor 1999b

Bowdler 1999

Bowdler 1999

Bowdler 1999

16

13 14 15

10 11 12

4 5 6 7 8 9

1 2 3

0

Years bp 1000

Increasing

Decreasing Maximum

Modern conﬁguration

Mangrove distribution

Advancing

1–1.5 m asl

Modern

Sea level

Increasing humidity

Onset of monsoon

Monsoon in modern conﬁguration

Climate

Table 10.2 Northern Cape Range Peninsula regional sequence.

Earliest Terebralia

Increase in ocher, ﬁrst rock art phase, intensive occupational phase

Latest Terebralia Initial occupation

New rock art motifs, increase in turtle bone, ﬁne-grained stone, desert walnuts, intensive occupational phase

Cultural

Hiatus 9,900 bp 10,730 bp

2,360 bp

Jansz

Hiatus 7,210 bp 8,010 bp

2,430 bp

C99

5,160 bp

Wobiri

Retreating

19 20 21 22

32 33 34 35 36

23 24 25 26 27 28 29 30 31

120 m asl

17 18

Increasing aridity

Maximum aridity

Initial occupation Initial occupation

Pleistocene archaeological remains fragmentary, but indicate a mixed marine and terrestrial economy

35,230 bp

Hiatus 31,360 bp

33,930 bp

Hiatus 21,110 bp

K at h r y n P r z y w o l n i k

size; it appears that only the more robust pieces of bone and shell have been preserved. These earliest occupants utilized both marine and terrestrial resources, hunting macropods, turtles, and ﬁsh, as well as collecting shellﬁsh and emu eggs. The sites’ occupants ﬂaked coarse-grained, partially siliciﬁed limestone (PSL) and medium-grained silcrete that was quarried from a local outcrop located close to the sites, leaving behind amorphous ﬂakes, ﬂaking debris, and cores. Jansz and C99 were abandoned for several thousand years during the ﬁnal stages of the terminal Pleistocene. Jansz was abandoned between 31,000–11,000 bp, a span of some 20,000 years. C99 was abandoned considerably later than Jansz at ca. 21,000 bp, and reoccupation of the site also seems to have been slightly later at ca. 8,000 bp. While the large span of time could be in part attributed to the samples that were submitted for dating, sediment and artifact accumulation rates in both sites show that deposition either stopped completely or slowed dramatically during these times. This supports the conclusion that human use of the sites at this time all but ceased. The occupational hiatus coincides with the span of the LGM, a period that is associated in arid Australia with increased aridity and generally colder, drier conditions (Ross et al. 1992; see also Chapter 3, this volume). Maximum glaciation is thought to have been reached ca. 18,000 bp, and at this time sea levels worldwide were lowered as a result of large volumes of sea water being trapped in the enlarged polar icecaps, to an estimated maximum low stand of 130–120 m below the modern level (Chappell 1983, 2002; Fairbanks 1989; Lambeck and Chappell 2001). Unlike most parts of Australia, the lowered sea level would not have removed the coast more than 10–12 km from its current position in Cape Range Peninsula, due to the uniquely proximal continental shelf (Wyrwoll 1993). It is unclear whether people departed the study area to follow the retreating coast or relocated to a more ecologically attractive region to escape the increasing aridity and deteriorating local environment in Cape Range.

Cape Range Early to Mid-Holocene Occupation Evidence of mangroves appears for the ﬁrst time in northern Cape Range Peninsula at approximately 10,730 bp, when fragments of Terebralia palustra (a mangrove dwelling gastropod) appear in the Jansz deposit for the ﬁrst time. The appearance of mangroves coincides with the reoccupation of the study area by humans following the LGM. The presence of remains such as chiton and baler shell, ﬁsh bone, and turtle bone in the deposit occurring at the same time as mangrove species indicates that sections of rocky shore and sandy bay were interspersed with mangrove along the coast. The combination of mangrove, rocky shore, and sandy bay marine resources would have provided a rich economic environment for human occupation, as is demonstrated by the rates of artifact and sediment accumulation that in many cases suggests a more intensive occupation of the sites. During this early Holocene period, the ﬁrst phase of rock art in C99 emerged, and for the next few thousand years motifs applied with dry, white ocher crayons were drawn on the walls and ceiling of the site.

_____________________________________________________________________________________________________

190

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

C99 and Jansz were abandoned for a second time during the mid-Holocene. The intensive mangrove phase of habitation in the sites lasted for only a few thousand years. Jansz was reoccupied following the glacial phase abandonment at 11,000 bp and was used until 9,900 bp. In C99 the reoccupation occurred slightly later at 8,010 bp and lasted until 7,210 bp. In both sites the Holocene occupational hiatus lasted until ca. 2,500 bp. This period of abandonment occurred at the same time as the decline in mangrove distribution along the Cape Range Peninsula coast. The gradual disappearance of the mangroves in Cape Range Peninsula meant the loss of a signiﬁcant economic resource. The diversity and abundance of plants and animals in mangrove communities has been shown to be of value to Aboriginal people for a variety of reasons, only one of which is the procurement of food (Meehan 1982; White et al. 1990).

Cape Range Late Holocene Occupation Occupation of the rock shelters during the late Holocene following ca. 2,500 bp was more intensive than the previous two occupational phases during the early Holocene and the Pleistocene. Sediment and artifact accumulation rates are at the most rapid at this time, particularly in the last 1,000 years. A change in lithic raw material is also apparent in the late Holocene stone tool assemblage. Locally available stone featured prominently in the Pleistocene and early Holocene, particularly stone from a silcrete quarry located directly above Wobiri and C99, less than 30 m away. The proportion of the ‘‘local’’ silcrete and coarse-grained PSL decreased from the early to late Holocene, when increased amounts of exotic ﬁnegrained PSL were introduced into the sites. Coincident with the increase in ﬁne-grained stone was the introduction of a suite of new tool types characterized by hafted composite tools with secondary retouch such as tula adzes, burren adzes, and backed artifacts and points (e.g., Lourandos 1997; McCarthy 1976). The artifacts found in Jansz, Wobiri, and C99 were predominantly tula and burren adzes made using ﬁne-grained PSL, and show use-wear consistent with hard wood processing (Kamminga 1982). New and distinctive rock art motifs were introduced to the sites during the late Holocene. The older motifs were all drawn using white pigment applied dry to the walls. Red ocher motifs are consistently superimposed over the white pigment motifs, and are estimated to be late Holocene in age (a date obtained of ca. 1,980 bp using marine shell in C99 is associated with red ocher pigment in the shelter deposit, located immediately below red ocher motifs on the shelter ceiling). The red ocher motifs were applied much more thickly than the white pigment motifs, and appear more ‘‘bold’’ and visible in the sites. One of the ways in which archaeologists have attempted to delineate relationships between groups of people and landscapes in the past is through an analysis of the distribution of rock art styles in time and space. When particular styles of rock art are present in some areas and not in others, assumptions can be made about relationships between the groups of people who shared stylistic similarities, and those who were outside of stylistic culture boundaries. Analyses of changes in rock art stylistic regions over

_____________________________________________________________________________________________________

191

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

time have been interpreted as demonstrating fundamental shifts in land tenure and alliance systems elsewhere in Australia (e.g., David and Chant 1995; David et al. 1994; Harrison 2000; McDonald 1998). The presence of a broad two phase stratigraphic sequence of rock art styles and colors at the sites within the study area suggests that an examination of stylistic afﬁnities in the broader region may reveal further information regarding the cultural history of the region. The use of the sites in the late Holocene is characterized by an increased dependence on particular marine resources, predominantly turtle and ﬁsh. These resources feature prominently in the ethnohistorical descriptions of Aboriginal life in Cape Range Peninsula (Bates n.d.; Scurla 1996; Tindale 1974), and provided people with eggs, shell, and oil, in addition to edible ﬂesh. Turtles spend most of their lives in deep sea waters, but during the summer months female turtles come ashore to lay eggs. Turtle hatcheries occur along all the sandy beaches on the northern Cape Range coast, and also on the offshore islands to the northeast of North West Cape and in Exmouth Gulf. The increase in turtle remains occurs at the same time that the hafted and composite stone artifacts appear in the rock shelters. Ethnohistorically, Aboriginal people in the Cape Range region hunted turtles from wooden watercraft using nets and wooden composite ﬁshing spears (Scurla 1996). The appearance of desert walnuts (Owenia reticulata) as a food source in the rock shelter deposits also occurred during the late Holocene. Both desert walnuts and turtles are seasonally available resources. Desert walnuts must be roasted prior to consumption to breach the hard outer husk, and are ready for harvesting after ﬂowering in the late summer when the nut has dried (Latz 1995). This coincides with the seasonal increase in availability of turtles in the study area, and suggests a systematic exploitation of these resources focused on the warmer months from December to May. An important implication of this specialized seasonal adaptation is that people must have spent signiﬁcant amounts of their time during an annual cycle somewhere else, presumably taking advantage of different resource abundance in adjoining areas. Such a specialized mobile land and resource use strategy would require relatively open social alliance networks and relations with neighboring social groups, or a large territory to move within.

Northern Cape Range Archaeological Sites: Coastal Shell Middens More than 60 shell midden sites were located and recorded in the northern tip of the peninsula during ﬁeldwork, and detailed mapping and surface collections were undertaken at a sample of eight middens (Przywolnik 2002a). Shell middens are located in a wide spectrum of coastal and sub-coastal environments in Cape Range Peninsula, but are generally found in dune blowouts and swales behind the foredunes. The middens in the study area generally occurred in three geographical locations. Type 1 middens were found in sparsely vegetated dune blowouts extending from behind a usually low foredune to close to the high tide zone. Type 2 middens were located in swales or blowouts behind lightly to moderately vegetated, more substantial foredunes with the extent of the midden deposit

_____________________________________________________________________________________________________

192

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

located further from the high tide zone. Type 3 middens were generally furthest from the shore, situated in swales or blowouts well behind the current foredunes in more consolidated sands and surrounded by moderately to densely vegetated stable dunes. Midden deposits in Cape Range Peninsula vary considerably in size, composition, location, and density. All of the recorded shell middens in the Cape Range region are unstratiﬁed surface sites, ranging from dense surface accumulations of marine shell and artifactual stone to small sparse scatters of one or two species of shell (Morse 1996; Przywolnik 2002a, 2003). As argued elsewhere, the regular occurrence of severe weather in the form of tropical cyclones is a contributing factor to the appearance and condition of the sites (Przywolnik 2002b). Radiocarbon dates on samples of marine shell have been obtained from several midden sites in Cape Range Peninsula. Obviously, using archaeological material from an unstratiﬁed open site subject to all manner of turbation and climatic processes does not produce as reliable a sequence as would perhaps be obtained from other stratiﬁed sites. The unstratiﬁed nature of the middens precludes the use of familiar stratigraphic principles, such as the principle of superposition, therefore the dates obtained from these midden sites can only be interpreted as isolated dates for a single occupation episode at each site, rather than accurate estimates of initial, or even ﬁnal, occupation. Radiocarbon determinations were obtained using marine gastropod shell from three midden sites and are summarized in Table 10.1. All of the shell middens in this study are estimated to have been deposited during the mid- to late Holocene, and in general the middens located closer to the current shore tend to be younger, dating to the late Holocene, while middens situated further inland, Types 2 and 3 middens, tend to be older, dating to the midHolocene (Przywolnik 2002a). The composition of shellﬁsh species in midden sites is consistent with excavated stratiﬁed rock shelter deposits and geomorphological information for the Cape Range region.

Broader Patterns in the Archaeology of Hunter-Gatherers in Coastal Northwestern Australia Pleistocene Occupation of Northwestern Australia Pleistocene occupation in coastal sites in northwestern Australia was characteristically intermittent and infrequent. Although cultural remains dated to between 35,000 bp and 10,000 bp are uniformly sparse in all sites along the northwest coast, archaeological evidence from Cape Range Peninsula, the coastal Pilbara, and the western Kimberley all suggest a broad based, mixed marine and terrestrial economy (Morse 1999; O’Connor 1999a; Veth 1999). The Pleistocene economy of the Noala 1 site in the Monte Bello Islands includes both marine and terrestrial resources, but between initial occupation at 27,000 bp and 10,030 bp terrestrial remains dominate the faunal assemblage (Veth 1993a, 1999). Like the Kimberley and Cape Range sites, however, there is an indisputable marine component in the Monte Bello site that links the ﬁrst humans to use the site

_____________________________________________________________________________________________________

193

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

with the coast, and demonstrates the importance of marine foods in the Pleistocene diet. The Silver Dollar site in Shark Bay is the only site in the northwest where remains of marine resources were not recovered from the Pleistocene units (Bowdler 1999). The Silver Dollar Pleistocene faunal component consisted of macropod teeth, baler shell (Melo amphora), and emu eggshell (Bowdler 1990a). Bowdler suggests that this is a reﬂection of the distance of the site from the coast during the Pleistocene, between 50 km and at least 100 km, and concludes that Silver Dollar is evidence of successful early occupation of the arid interior (Bowdler 1999). Interestingly, fragments of marine shell were found in the Silver Dollar deposit dating to ca. 18,000 bp, which suggests a tenuous, if long distance, link with the coast even when the site itself was more than 100 km from the LGM coastline (Bowdler 1990a). The ‘‘coastal colonization’’ model proposed by Bowdler (1977) more than 20 years ago is still a relevant and debated issue in Australian archaeology. The model proposes that the Greater Australian land mass was colonized by humans following the coast, who entered the continental interior via estuarine and riverine environments where economic adaptation to unfamiliar environments was minimal. Bowdler (1977) argued that evidence for economic change in the Pleistocene was minimal, which she interpreted as evidence of a dominant stable, successful maritime economy. After a reevaluation of the archaeological evidence available some 10 years after her model was proposed, Bowdler only slightly adjusted her position, and while she reafﬁrmed the validity of coastal colonization, conceded that the model must allow for regional variation and that adaptation to nonaquatic resource use occurred earlier than was originally perceived (Bowdler 1990b). In contrast, O’Connor and Veth (2000) proposed that although the ﬁrst colonization of Greater Australia some 60,000 to 50,000 years ago required several oceanic water crossings, there is no evidence to suggest that the early colonists employed a maritime economy once arriving on the continent. They argued that despite a number of sites being located close to the coast and continental shelf in northwest Australia (particularly the southern Cape Range Peninsula sites investigated by Morse) and therefore strategically placed to provide a record of use by Aboriginal people in the past, the evidence of Pleistocene resource use gives ‘‘no hint of maritime dependence but rather suggests a terrestrial resource base with the ad hoc addition of coastal-fringe resources or a generalized mixed economy’’ (O’Connor and Veth 2000: 100). O’Connor and Veth (2000: 129) concluded that despite the necessity of maritime proﬁciency to gain entry into Greater Australia, after initial colonization ‘‘the inland water systems of the savanna interior were the main drawcard.’’ This conclusion is based on a number of lines of evidence. The oldest dates are currently from sites well inland in the north of Australia, coastal sites all post-date 35,000 bp, Pleistocene faunal assemblages are not marine speciﬁc, and sites show an increasing reliance on marine resources as conditions become increasingly arid over the LGM (O’Connor and Veth 2000). The archaeological evidence from Jansz and C99 points to an ephemeral and intermittent use of the sites during the Pleistocene. The signiﬁcant terrestrial faunal

_____________________________________________________________________________________________________

194

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

component indicates that the Pleistocene diet did not rely exclusively on resources procured from the sea, and that the early hunter-gatherers in Cape Range subsisted on a range of foods from all environmental contexts in the region. In this way, the Pleistocene economy in the study area is comparable with the economies described from sites with similar antiquity in the northwest, such as Mandu Mandu Creek, Widgingarri, and Noala 1. The archaeological evidence from Cape Range Peninsula demonstrates that Pleistocene humans were proﬁcient at utilizing a range of environments, which suggests a ﬂexible and versatile subsistence strategy, without a dependence on any one resource.

Glacial Transition Veth’s (1989, 1993b) model of arid zone colonization predicts ephemeral human use of arid, marginal areas during the Pleistocene, particularly over the period of maximum aridity between 18,000–15,000 bp. According to the model, occupation of arid environments until ca. 5,000 bp was limited by the lack of fully developed desert adaptive strategies, such as intensive seed grinding, hafting, and extended social networks, that enabled marginal environments to be permanently populated (Veth 1993b: 103–5). Veth (1993b) argued that occasional Pleistocene forays into marginal arid environments were uniformly intermittent and ﬂeeting. Drawing from a biogeographic framework, Veth’s model proposed that the arid interior of Australia housed several zones of ‘‘refuge,’’ ‘‘barrier,’’ and ‘‘corridor,’’ referring to the relative hospitality of the areas in terms of coordinated drainage and resource availability (Veth 1993b: 106). Barriers to human habitation were considered to be areas of sandridge desert, while three main zones of refuge were located in Western Australia: the southwest, inland western Pilbara, and western Kimberley (see Veth 1993b: 107). According to this model, Cape Range Peninsula is located in a corridor zone between the southwest and Pilbara refuges. The model predicts that while offering favorable occupational environments during the ‘‘lacustral’’ (i.e., pre-22,000 bp) Pleistocene period and Holocene, corridor zones would have acted as barriers to human occupation at the height of glaciation from 18,000–15,000 bp (Veth 1993b: 106). Based on Veth’s predictive model sites in corridor zones should therefore exhibit an occupational hiatus spanning the period of maximum aridity when environmental conditions deteriorated. As climatic amelioration increased the habitability of corridor regions, the model suggests that sites may be reoccupied as human groups repopulate and expand in response to the more favorable conditions. Periods of occupational hiatus spanning the LGM have been identiﬁed in both Jansz and C99. Sedimentation and artifact accumulation rates indicate that the sites were visited either extremely infrequently or not at all at the time of maximum glaciation, when deposition in the site abruptly diminished to a much slower rate. The timing of site abandonment differs somewhat between the two sites. Jansz, the smaller rock shelter of the two, was abandoned between 31,000 bp and 11,000 bp, a span of 20,000 years. The period of occupational hiatus in C99 was considerably shorter, and lasted from 21,000 bp until 8,000 bp.

_____________________________________________________________________________________________________

195

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

The glacial sequence of occupational abandonment in Jansz and C99 is in keeping with Veth’s arid zone model. The Mandu Mandu Creek site located further south in the region also shows a phase of occupational abandonment over the LGM, between 20,040 bp and 5,490 bp. Morse (1993c) links the abandonment of Mandu Mandu Creek with the retreating coastline, and speculates that the human population moved westward with the coastline, making only very occasional forays to the range itself. She comments that a single date from 17,000 bp in Pilgonaman Creek rock shelter located south of Mandu Mandu Creek is very likely to have been a result of geographical circumstance; Pilgonaman Creek is larger and more visible in its surrounds than Mandu Mandu Creek, making a more obvious shelter spot for hunter-gatherers venturing inland from the coast (Morse 1999: 76). Sites in the west Kimberley and Shark Bay also appear to have been abandoned around this time. In both regions the coast during the LGM would have been much further from its current position as the continental shelf lies a greater distance west. O’Connor (1999a: 122) argues that the occupational hiatus apparent in the Koolan and Widgingarri sites is not a response to a retreating coast, as Widgingarri was not abandoned until 19,000 bp when the coast was over 150 km distant from the site.

Early Holocene Transition: The Big Swamp When humans returned again to the northern tip of Cape Range Peninsula, the coastline was considerably different to the one they had left prior to glaciation. Mangrove forest now fringed parts of the coastline, and a large mangrove swamp dominated what is now the northern tip of the peninsula. The mangroves were a very rich environment, and provided a variety of economic resources that were exploited by the region’s occupants. Among the mangroves, and particularly along the southwestern edge of the study area, sections of rocky shore and sandy bay provided different foods to mangroves, such as chiton, turtle, and reef ﬁsh. By 4,000 bp the sea level stabilized at its present level, and mangroves were no longer extant at the northern tip of the peninsula, having retracted in distribution to the southern margin of the Exmouth Gulf and small pockets on the southern peninsula coast. This ecological drama is reﬂected in the archaeology of Cape Range, and is visible in both the midden sites and the rock shelters. The Jansz and C99 sites were used relatively intensively after they were reoccupied following the LGM. It is interesting to note the possible existence of stone ﬁsh trapping technology at this early date. The manufacture and use of ﬁxed facilities (such as ﬁsh and eel traps), which require ongoing maintenance for the gathering of food surpluses, have been linked elsewhere to phases of intensiﬁcation. Most facilities of this kind in Australia, although notoriously difﬁcult to date, are thought to have developed in the late Holocene. The implication of the use of such a ‘‘delayed return’’ system is that the early to mid-Holocene saw a shift to a more intensive and logistical land and resource use strategy which required complex socioeconomic relations. This is a signiﬁcant transition from the pre-LGM occupational pattern, and appears to be a

_____________________________________________________________________________________________________

196

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

localized response to changed environmental conditions resulting in ‘‘intensiﬁcation’’ of site and resource use in the area with evidence of more ‘‘closed’’ social networks. The early Holocene rock art phase in Cape Range, with its meandering lines and random cross hatching, ﬁnds few parallels in the rock art traditions of the broader region. Where rock art styles occur over relatively limited distributions with clear boundaries between stylistic-culture areas it has been interpreted as evidence for relatively closed, local alliance networks and localized group signifying behaviors. During the early to mid-Holocene, evidence from excavations and the presence of possible stone ﬁsh trapping structures suggests a less seasonal, more intensive occupation of the Cape. Wobst’s (1977) information theory of exchange posits a relationship between highly resource rich environments and relatively heterogeneous rock art styles (see also Smith 1989). The rich mangrove environments of ca. 10,000–4,000 bp appear to have encouraged more sedentary, intensive, and territorial occupation of the Cape than the more recent ethnohistorically derived land and resource use strategy. The occurrence of social boundary maintenance activity with evidence for more logistical settlement patterns suggests the occurrence of complex socioeconomic relations and more intensive land management during the early to mid-Holocene.

Mid-Holocene Transition There is considerable evidence that the study area was used on an extremely intermittent basis during the mid-Holocene. This is in stark contrast with views of mid-Holocene change and expansion current in most other parts of Australia (e.g., Lourandos 1997). No radiocarbon dates were obtained from any of the sites in this study that produced a result between 4,830 bp and 2,360 bp. Sedimentation and artifact accumulation rates from C99 and Jansz indicate that use of the shelter by humans reduced dramatically at this time, particularly in the latter where occupation of the site appears to have ceased. This site abandonment phase coincides with the decline and eventual disappearance of mangroves from the study area and Cape Range Peninsula region in general. As has already been demonstrated, the mangroves provided a rich economic environment for human habitation, and that resulted in a systematic and ‘‘high maintenance’’ exploitation of the coast. The increased distribution of mangroves during the early and mid-Holocene was not conﬁned to Cape Range Peninsula, but occurred right along the northwest coast from Shark Bay to the western Kimberley (O’Connor 1999b). The decline of the mangroves triggered a range of social responses in different areas, which is perhaps indicative of both the range of human economic and resource use strategies along the Western Australian coast and the large diversity of ecological conditions in the long, but single stretch of northwest coast. In contrast to the mid-Holocene hiatus in Cape Range, Veitch (1999) argues that a mid-Holocene change from mangrove species exploitation to a focus on Anadara granosa was a purely social strategy that enabled larger populations to subsist in formerly

_____________________________________________________________________________________________________

197

_____________________________________________________________________________

Years BP

Figure 10.2 Histogram of all radiocarbon determinations from coastal sites in northwestern Australia.

0

2

4

6

8

10

12

0-500 1000-1500 2000-2500 3000-3500 4000-4500 5000-5500 6000-6500 7000-7500 8000-8500 9000-9500 10000-10500 11000-11500 12000-12500 13000-13500 14000-14500 15000-15500 16000-16500 17000-17500 18000-18500 19000-19500 20000-20500 21000-21500 22000-22500 23000-23500 24000-24500 25000-25500 26000-26500 27000-27500 28000-28500 29000-29500 30000-30500 31000-31500 32000-32500 33000-33500 34000-34500 35000-35500

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

marginal environments or difﬁcult environments. Veitch (1999: 60) comments that this behavioral shift was not a response to environmental change, but an ‘‘elaboration of pre-existing themes’’ already discernible in Aboriginal society in that area, at that time. As with the northern Cape Range Peninsula sequence, no dates between 4,820– 2,420 bp were obtained from the archaeological sites in the central and southern Cape Range. As many radiocarbon sequences from coastal northwestern sites as possible were collated and the number of times radiocarbon determinations occurred within 500 year increments from the present to 36,000 bp was plotted as a histogram (see Figure 10.2), which shows that for sites located on the current coast between Shark Bay in the south and the southwest Kimberley in the north very few radiocarbon determinations produced results between 5,000–2,500 bp. From 50 sites and a total of 141 determinations (with more than 20 sites in the sample with a sequence of more than two dates and numerous sites with single dates) only twelve determinations were obtained between 2,500–4,500 bp, representing 8.5 percent of the sample. Between the present and 2,500 bp, 39 determinations were obtained (27.7 percent), and 30 determinations (21.2 percent) were obtained between 4,500–7,000 bp. These results clearly indicate that the use of sites in the coastal northwest region decreased dramatically in the mid-Holocene. The late Holocene section of the histogram – from the present to 2,500 bp – shows that sites were once again reoccupied with at least the same frequency as during the early Holocene. This pattern, however, may be inﬂuenced by several factors, such as taphonomic processes or differential preservation of archaeological remains due to localized conditions and sea level ﬂuctuation. The Silver Dollar and other Shark Bay sites such as Eagle Bluff have clear evidence of occupational abandonment during the mid-Holocene. Bowdler (1999) observes that the early Holocene economy in Shark Bay was highly dependant on mangroves. The apparent hiatus in site use coincides with the decline of mangroves in the Shark Bay region, and Bowdler (1999: 82) concludes ‘‘the local population diminished after ca. 3,500 bp, consequent on an environmentally determined decline in local mangrove populations.’’ This decline is site use from 6,000–2,500 bp has obvious connotations regarding population movement and relocation. The people who deserted the newly mangrove-impoverished regions of the northwest presumably moved elsewhere to utilize environments that were more economically valuable. This contrasts markedly with notions of steady population growth and expansion in most other parts of the continent (Beaton 1985; Lourandos 1997: 301; Smith 1986).

Late Holocene Transition The reoccupation of the study area following the mid-Holocene abandonment phase heralds a suite of new and different adaptive strategies employed by humans in Cape Range Peninsula. At this time a technological change occurs in the region characterized by the introduction of new tool types, primarily hafted and composite implements such as tulas, burrens, and backed artifacts. The new tool types are

_____________________________________________________________________________________________________

199

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

found in the uppermost stratigraphic layers in the rock shelters and on the midden sites, predominantly Type 1 middens which date to within the last 2,500 years. A wider range of stone raw materials was also introduced at this time, and the formal tools were made almost exclusively using ﬁne-grained raw materials. Prior to the late Holocene the dominant lithic raw materials tended to be medium- and coarse-grained locally available stone. The introduction of exotic ﬁne-grained stone suggests the installation of a more wide ranging set of social networks extending beyond the immediate conﬁnes of Cape Range, an expansion of the territory in which the people who inhabited the peninsula foraged, or a combination of these. The appearance of new rock art motifs during the Holocene with stylistic links to the Pilbara and Murchison regions suggests that long distance social networks with neighboring regions were established at this time. Archaeological assemblages from the rock shelter and midden sites in the study area indicate that the dominant land use strategy was seasonally based, and focused heavily on marine resources. This correlates well with the ethnohistoric model of human occupation of the region, which emphasized the high mobility and wide geographical range of the Cape Range occupants, who lived almost exclusively on marine resources, particularly turtle, while in the area. The maintenance of mobile and wide ranging resource procurement strategies in resource-rich environments is generally associated with ‘‘open’’ social networks with neighboring regions (Wobst 1977).

Was there Always ‘‘Intensiﬁcation’’ in the Mid- to Late Holocene? Challenging the Progressivist View of Hunter-Gatherer Change in Australia The most debated and discussed paradigm for explaining late Holocene change in Australian hunter-gatherer studies, particularly in southeastern Australia, is the ‘‘intensiﬁcation’’ model proposed by Lourandos (1983, 1997; see also Ross et al. 1992). Lourandos’ model is based on the premise that through time more complex and dynamic social systems are produced as a result of sociodemographic pressures and change. While he (Lourandos 1997: 21) speciﬁes that hunter-gatherer societies may shift either way along a spectrum of complexity that does not require a unidirectional trajectory of change, the model identiﬁes a general trend of increasing complexity in the mid- to late Holocene fueled by a range of factors, including environmental conditions, demographic variation, and sociocultural and technological change. The ‘‘more complex and dynamic’’ late Holocene social systems are characterized by a number of attributes visible in changes of land and resource use such as: . Use of smaller sized territories, and increased evidence of territoriality. . Sedentism and other changes in ‘‘logistical’’ strategies, such as earth mounds. . Use of ‘‘delayed-return’’ economic systems, such as ﬁsh traps.

_____________________________________________________________________________________________________

200

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia

. Resource specialization and economic intensiﬁcation (for example, exploiting resources such as grass seeds, roots, tubers, nuts, and large marine fauna that require specialized technology to access them). . Increased use of sites and increased rate of establishment of new sites. . Increased use of marginal environmental areas. . ‘‘Closed’’ societal networks (Lourandos 1997: 319). The late Holocene period in Cape Range Peninsula shares similarities with some of the predictions in the intensiﬁcation model, particularly evidence of increased frequency and intensity of site use and resource specialization. In the main, the attributes listed above are more descriptive of the early Holocene occupational phase than the mid- to late Holocene occupation of Cape Range and northwest Australia more generally. I have argued that the late Holocene in Cape Range was a period of ‘‘open’’ social networks maintained by highly mobile groups who made seasonal cycles of resource procurement within a wide ranging territory. Although the use of the landscape appears more ‘‘intensive,’’ with more midden sites dating to the late Holocene and increased rates of artifact accumulation in the rock shelter sites, this reﬂects an archaeological signature of an extremely specialized resource procurement system centered primarily on turtles, which is mobile and seasonal but still focused and systematic. This is in opposition to the view that late Holocene resource specialization encourages sedentism and territoriality (Lourandos 1997: 319). One of the major criticisms of subsequent interpretations of regional models of ‘‘intensiﬁcation’’ is the reliance on what are essentially unilinear models of sociocultural change, in which societies necessarily become more ‘‘complex’’ with time and change in a manner that is seen as ‘‘advancing’’ or ‘‘progressing.’’ This ‘‘progressivist’’ view assumes that social movement toward complexity is generally directional and irreversible. In contrast, the ‘‘adaptationist’’ view rejects traditional notions of progression, and perceives social change as non-directional and reversible, a ﬂexible response to adaptive necessity rather than an inevitable move toward increasing complexity (Rowley-Conwy 2001). The archaeological research presented here suggests a late Holocene adaptive strategy that is seemingly at odds with the prevailing notions of ‘‘closed,’’ ‘‘sedentary,’’ and ‘‘territorial’’ land use patterns that are seen to characterize this period in Australia. These attributes are in general more descriptive of the early Holocene pattern of occupation discussed above, and therefore counter-progressivist notions of social-demographic change directed toward ‘‘complexity.’’ Lourandos’ intensiﬁcation model is most applicable in those parts of Australia where environmental conditions are most amenable to providing for increased sedentism, ‘‘closed’’ social networks, territoriality, use of delayed-return economic systems, and long-term population growth, that is, the southeastern section of the continent. Although archaeological evidence from more arid regions in many cases supports some aspects of the intensiﬁcation model, it appears that in the Cape Range region the basic environmental infrastructure, access to reliable and permanent water, which would establish the long-term sociodemographic systems that are central to

_____________________________________________________________________________________________________

201

_____________________________________________________________________________

K at h r y n P r z y w o l n i k

Lourandos’ model, is lacking. This interpretation of archaeological evidence from the late Holocene Cape Range Peninsula provides a social framework for human use of the landscape that incorporates notions of ‘‘intensity’’ of site use that need not subscribe to the current views of ‘‘intensiﬁcation.’’ This does not refute Lourandos’ intensiﬁcation model, but provides an example of regional variation that occurs within a prevailing broader paradigm of late Holocene social change. The Cape Range case study stands in opposition to any notion of unidirectionality of change, as the late Holocene adaptive response by the hunter-gatherers of Cape Range peninsula does not demonstrate an inevitable progression toward sedentism or ‘‘intensiﬁcation’’ in the traditional sense, nor even a continuation and increase in complexity of an existing occupation pattern.

Conclusion Many archaeological studies of long-term hunter gatherer change have adopted a progressivist view, which sees societies necessarily becoming more complex with time and changing in a manner that is seen as advancing or progressing towards increased sedentism and incipient forms of agriculture. This has particularly been the case in Australia, where Lourandos’ model of sociodemographic intensiﬁcation in the late Holocene in southeastern Australia has often been assumed to describe long-term patterns of hunter-gatherer change throughout the continent. The archaeology of northern Cape Range Peninsula, and coastal northwest Australia in general, demonstrates that the rate and direction of hunter-gatherer change may be somewhat more complex than a simple trajectory towards increased sedentism and ‘‘complexity’’ through time. In the study area, a trend towards increased sedentism in the terminal Pleistocene and early Holocene is reversed after the depletion of favored mangrove environments, when hunter-gatherers developed a highly seasonal and mobile use of the study area with more in common with the Pleistocene occupation of the study area than the early Holocene. This case study supports Rowley-Conwy’s (2001) adaptationist view of long-term hunter-gatherer change, which perceives social change as non-directional and reversible, and a ﬂexible response to adaptive necessity rather than an inevitable move toward increasing complexity.

References Bates, D. n.d.: The tribal organization and geographical distribution: Western Australia. Canberra: Manuscript no. 92, Australian Institute of Aboriginal and Torres Strait Islander Studies Library. Beaton, J. M. 1985: Evidence for a coastal occupation time lag at Princess Charlotte Bay (north Queensland) and implications for coastal colonization and population growth theories for Aboriginal Australia. Archaeology in Oceania, 20, 1–20.

_____________________________________________________________________________________________________

202

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia Bowdler, S. 1977: The coastal colonization of Australia. In J. Allen, J. Golson, and R. Jones (eds), Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia, and Australia, London: Academic Press, 205–46. Bowdler, S. 1990a: The Silver Dollar site, Shark Bay: An interim report. Australian Aboriginal Studies, 2, 60–3. Bowdler, S. 1990b: Peopling Australia: The ‘‘coastal colonization’’ hypothesis re-examined. In P. Mellars (ed.), The Emergence of Modern Humans: An Archaeological Perspective, Edinburgh: Edinburgh University Press, 327–43. Bowdler, S. 1999: Research at Shark Bay, WA, and the nature of coastal adaptations in Australia. In J. Hall and I. J. McNiven (eds), Australian Coastal Archaeology, Canberra: The Australian National University, 79–86. Bradshaw, E. 1995: Dates from archaeological excavations on the Pilbara coastline and islands of the Dampier Archipelago, Western Australia. Australian Archaeology, 41, 37–8. Chappell, J. 1983: Sea level changes, 0 to 40 KA. In T. Donnelly and R. Wasson (eds), CLIMANZ 3 Proceedings of the Third Climanz Conference, Canberra: Department of Biogeography and Geomorphology, Research School of Paciﬁc Studies, Australian National University, 121–2. Chappell, J. 2002: Sea level changes forced ice breakouts in the Last Glacial cycle: New results from coral terraces. Quaternary Science Review, 21, 1229–40. David, B. and Chant, D. 1995: Rock art and regionalism in north Queensland prehistory. Memoirs of the Queensland Museum, 37 (2), 357–528. David, B., McNiven, I., Attenbrow, V., Flood, J., and Collins, J. 1994: Of Lightning Brothers and White Cockatoos: Dating the antiquity of signifying systems in the Northern Territory, Australia. Antiquity, 68, 241–51. Fairbanks, R. G. 1989: A 17,000 year glacioeustatic sea level record: Inﬂuence of glacial melting rates on the Younger Dryas event and deep ocean circulation. Nature, 342, 637–42. Harrison, R. 2000: Challenging the ‘‘authenticity’’ of antiquity: Contact archaeology and Native Title in Australia. In I. Lilley (ed.), Native Title and the Transformation of Archaeology in the Postcolonial World, Sydney: University of Sydney Press, 35–53. Hatcher, B. G. 1991: Coral reefs in the Leeuwin Current – an ecological perspective. Journal of the Royal Society of Western Australia, 74, 115–28. Kamminga, J. 1982: Over the Edge: Functional Analysis of Australian Stone Tools. St Lucia: Anthropology Museum, University of Queensland. Lambeck, K. and Chappell, J. 2001: Sea level change through the last glacial cycle. Science, 292, 679–86. Latz, P. 1995: Bushﬁres and Bushtucker: Aboriginal Plant Use in Central Australia. Alice Springs: IAD Press. Lourandos, H. 1983: Intensiﬁcation: A late Pleistocene–Holocene archaeological sequence from southwestern Victoria. Archaeology in Oceania, 18, 81–94. Lourandos, H. 1997: Continent of Hunter-Gatherers: New Perspectives on Australian Prehistory. Cambridge: Cambridge University Press. McCarthy, F. D. 1976: Australian Aboriginal Stone Implements (2nd edn). Sydney: Australian Museum Trust. McDonald, J. 1998: Shelter art in the Sydney Basin – a space-time continuum: Exploring different stylistic inﬂuences on stylistic change. In C. Chippindale and P. Tac¸on (eds), The Archaeology of Rock Art, Cambridge: Cambridge University Press, 319–35.

_____________________________________________________________________________________________________

203

_____________________________________________________________________________

K at h r y n P r z y w o l n i k Meehan, B. 1982: Shell Bed to Shell Midden. Canberra: Australian Institute of Aboriginal Studies Press. Morse, K. 1993a: West Side Story: Towards a prehistory of the Cape Range Peninsula, Western Australia. Unpublished PhD thesis, Perth: University of Western Australia. Morse, K. 1993b: New radiocarbon dates from North West Cape, Western Australia: A preliminary report. In M. A. Smith, M. Spriggs, and B. Fankhauser (eds), Sahul in Review: Pleistocene Archaeology in Australia, New Guinea, and Island Melanesia, Canberra: The Australian National University, 155–63. Morse, K. 1993c: Who can see the sea? Prehistoric Aboriginal occupation of the Cape Range peninsula. In W. F. Humphreys (ed.), The Biogeography of Cape Range, Western Australia. Records of the Western Australian Museum, Supplement no. 45, 227–42. Morse, K. 1993d: Shell beads from Mandu Mandu Creek rock shelter, Cape Range Peninsula, Western Australia, dated before 30,000 bp. Antiquity, 67, 877–83. Morse, K. 1996: Coastal shell middens, Cape Range Peninsula, Western Australia: An appraisal of the Holocene evidence. In P. Veth and P. Hiscock (eds), Archaeology of Northern Australia: Regional Perspectives, St Lucia: Anthropology Museum, University of Queensland, 9–25. Morse, K. 1999: Coastwatch: Pleistocene resource use on the Cape Range Peninsula. In J. Hall and I. J. McNiven (eds), Australian Coastal Archaeology, Canberra: The Australian National University, 73–8. O’Connor, S. 1999a: 30,000 Years of Aboriginal Occupation: Kimberley, North West Australia. Canberra: Department of Archaeology and Natural History and the Centre for Archaeological Research, The Australian National University, O’Connor, S. 1999b: A diversity of coastal economies: Shell mounds in the Kimberley region in the Holocene. In J. Hall and I. J. McNiven (eds), Australian Coastal Archaeology, Canberra: The Australian National University, 37–50. O’Connor, S. and Veth, P. 2000: The world’s ﬁrst mariners: Savanna dwellers in an island continent. In S. O’Connor and P. Veth (eds), East of Wallace’s Line: Studies of Past and Present Maritime Cultures of the Indo-Paciﬁc Region, Rotterdam: A. A. Balkema, 99–137. Przywolnik, K. 2002a: Patterns of occupation in Cape Range Peninsula (WA) over the last 36,000 years. Unpublished PhD thesis, Perth: University of Western Australia. Przywolnik, K. 2002b: Coastal sites and severe weather in Cape Range Peninsula, northwest Australia. Archaeology in Oceania, 37 (2), 137–52. Przywolnik, K. 2003: Shell artifacts from northern Cape Range Peninsula, northwest WA. Australian Archaeology, 56, 12–21. Ross, A., Donnelly, T., and Wasson, R. 1992: The peopling of the arid zone: Human– environment interactions. In J. Dodson (ed.), The Naive Lands, Melbourne: Longman Cheshire, 76–114. Rowley-Conwy, P. 2001: Time, change, and the archaeology of hunter-gatherers: How original is the ‘‘Original Afﬂuent Society’’? In C. Panter-Brick, R. H. Layton, and P. Rowley-Conwy (eds), Hunter-Gatherers: An Interdisciplinary Perspective, Cambridge: Cambridge University Press, 39–72. Scurla, S. 1996: The Stefano Castaways (2nd edn) [translated and edited by A. Sala]. Perth. [Original manuscript, I Naufraghi dello Stefano, 1876.] Smith, C. 1989: Designed dreaming: Assessing the relationship between style, social structure, and environment in Aboriginal Australia. Unpublished BA (Hons) thesis, Armidale: University of New England.

_____________________________________________________________________________________________________

204

_____________________________________________________________________________

H u n t e r - Ga t h e r e r s o f C o a s t al NW A u s tr a l ia Smith, M. A. 1986: The antiquity of seed grinding in Australia. Archaeology in Oceania, 21, 29–39. Tindale, N. 1974: The Aboriginal Tribes of Australia. Canberra: The Australian National University Press. Veitch, B. 1999: Shell middens on the Mitchell Plateau: A reﬂection of a wider phenomenon? In J. Hall and I. J. McNiven (eds), Australian Coastal Archaeology, Canberra: The Australian National University, 51–64. Veth, P. 1989 Islands in the Interior: A model for the colonization of Australia’s arid zone. Archaeology in Oceania, 24, 81–92. Veth, P. 1993a: The Aboriginal occupation of the Monte Bello Islands, northwest Australia. Australian Aboriginal Studies, 2, 39–47. Veth, P. 1993b: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory, Archaeology Series no. 3, Ann Arbor, MI. Veth, P. 1999: The occupation of arid coastlines during the terminal Pleistocene of Australia. In J. Hall and I. J. McNiven (eds), Australian Coastal Archaeology, Canberra: The Australian National University, 65–72. White, N., Meehan, B., Hiatt, L., and Jones, R. 1990: Demography of contemporary huntergatherers: Lessons from Arnhem Land. In B. Meehan and N. White (eds), HunterGatherer Demography: Past and Present, Sydney: University of Sydney Press, 171–85. Wobst, M. 1977: Stylistic behavior and information exchange. In C. E. Cleland (ed.), Papers for the Director: Research Essays in Honour of James B. Grifﬁn, Ann Arbor: University of Michigan Press, 317–42. Wyrwoll, K.-H. 1993: An outline of Late Cenozoic paleoclimatic events in the Cape Range region. In W. F. Humphreys (ed.), The Biogeography of Cape Range, Western Australia. Records of the Western Australian Museum, Supplement no. 45, 39–50.

_____________________________________________________________________________________________________

205

_____________________________________________________________________________

-----------------------------------------------------------------------------

11

--------------------------------------------------------------------------

Hunter-Gatherers and Herders of the Kalahari during the Late Holocene Karim Sadr

Introduction In popular imagination, the Kalahari Desert is the ‘‘land-that-time-forgot,’’ where Bushmen (also known as San or Basarwa; see Chapter 9, this volume) continue to hunt and gather in Paleolithic style. Contesting this image are scholars who claim that today’s Kalahari Bushmen are impoverished descendants of people who in the past two millennia had been cattle herders and suppliers of bush products to the Indian Ocean commercial network in medieval times. This has become known as the Kalahari Debate and is the focus of much anthropological and archaeological research. In this chapter three archaeological case studies are presented which suggest that through contact with farmers and herders, many Indigenous huntergatherers in and around the Kalahari were assimilated and did indeed turn to herding and trading, but at different times and under varying circumstances. In the deepest Kalahari, the evidence suggests that Bushmen managed to keep their traditional way of life until very recently. Still, this is not to say that the 1960s !Kung Bushmen, as recorded by the Harvard ethnographers around the Dobe waterhole, were indeed pristine hunter-gatherers. New interpretation of archaeological evidence suggests various groups of hunter-gatherers in different parts of the subcontinent herded livestock well before direct contact with Bantu-speaking farmers and herders (see also Chapter 2, this volume).

Environmental Background The Kalahari means different things to different people. At its most extensive, it is a sand covered plain that stretches from the Orange River to central Africa (Deacon

_____________________________________________________________________________________________________

206

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

and Lancaster 1988; Thomas and Shaw 1991). Here, I focus only on the southern half of this sandy expanse, from the Orange River to northern Botswana, and from the highlands of eastern Botswana and western Zimbabwe to the mountains of central Namibia (see Figure 11.1). In the northern part of the Kalahari annual

Angola

Zambia

e

iqu

b am

z

Mo

Zimbabwe Namibia Botswana Shaded area shows the Kalahari Desert

South Africa Cape Provinces

Za

m

be

zi

8

Ri

ve

r

4 2 9

1

6 7

Lim 10

po

po

Ri

ve

5 3 11

Ora

nge

Riv

er

r

1 Makgadikgadi Pans 2 Okavango Delta 3 Rietfon tein 4 Tsodilo Hills 5 Thamaga 6 Cae Cae 7 Toteng 8 Oruwanje 9 Geduld 10 Shashe Limpopo 11 Kwa Zult Natal

Figure 11.1 Map of southern Africa with places mentioned in the text.

_____________________________________________________________________________________________________

207

______________________________________________________________________________

Karim Sadr

rainfall is 500–800 mm, but is reduced to only 150–200 mm per year in the south. This means the Kalahari is not a desert proper but a ‘‘thirstland’’ where surface water is rare due to the permeable sands. Most drainage in the Kalahari is internal and the largest pans are at Makgadikgadi, which are the remnants of an ancient lake, the modern version of which is the Okavango Delta. This inland delta, fed by rainfall in the highlands of Angola, is now the only permanent surface water in the Kalahari (Cooke 1979). Recent research in the northern Kalahari has indicated several wetter episodes in the past, including one during the mid-Holocene and another about 1,500–1,000 bp (Robbins et al. 1996, 1998).

Archaeological Research in the Kalahari The most famous archaeological site in the Kalahari is not archaeological at all. In 1886 G. A. Farini, circus owner and tightrope walker, reported ﬁnding a lost city in the Kalahari (Farini 1973). The news inspired at least 25 later expeditions to relocate the site, which eventually turned out to be a natural rock formation near Rietfontein in South Africa (Clement 1967). Among other early but more conventional archaeological discoveries in the Kalahari were the descriptions of rock paintings in the Tsodilo Hills by Passarge (1904, 1907) during the late 1890s. Thereafter, many Europeans passing through the Kalahari collected stone tools (Campbell 1998). Until the 1960s, however, the lack of substantial sites encouraged relatively little archaeological interest in the Kalahari. It was only after ethnographers from Harvard University drew attention to the !Kung hunter-gatherers of the deepest Kalahari that interest in the region’s prehistory quickened (Lee and DeVore 1976; Marshall 1976). Yellen and Brooks (1989) carried out the ﬁrst professional excavations in Later Stone Age (LSA) sites of the Kalahari, starting in 1968 with eight sites in the Qangwa and Cae Cae (also known as /Xai /Xai) areas of western Ngamiland. In the late 1970s Wilmsen (1978, 1989a) also excavated at Cae Cae, and made it the cornerstone of what came to be known as the Kalahari Debate. In the early 1980s Robbins (1985, 1986) excavated rock shelters on the southeastern margin of the Kalahari near the village of Thamaga. In 1982 he and Campbell moved north and excavated rock shelters in the Tsodilo Hills. From there they have produced by far the richest body of information on the prehistory of the Kalahari (Robbins 1990; Robbins and Campbell 1989; Robbins et al. 1996). Aside from rock shelters, Iron Age villages have also been excavated in the Tsodilo Hills and elsewhere throughout northern and eastern Botswana by Denbow (1984, 1986, 1990) and Wilmsen (Wilmsen and Denbow 1990), among others. Since 1990, the local staff of the University of Botswana and the National Museum in Gaborone, as well as visiting researchers, have also been active in this ﬁeld (cf. van Waarden 1999). Some noteworthy examples are excavations at Toteng, which have provided new information about the earliest herders in the Kalahari (Campbell 1992; Huffman 1994). Other recent work in the Kalahari includes that by Walker (1994, 1995, 1998) on the origins of linguistic diversity

_____________________________________________________________________________________________________

208

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

among Bushmen in Botswana; excavations in Iron Age village sites in eastern Botswana by Reid and Segobye (2000); and ongoing research in northeast and northwestern Botswana by Tsheboeng, Dingalo, and van Waarden. Following in Robbins’ footsteps, in late 1996 I carried out excavations in LSA rock shelters near Thamaga in southeastern Botswana (Sadr 2002; Sadr and Plug 2001). Outside Botswana, Smith has been a prominent researcher. With Lee (Smith and Lee 1997) he investigated the antiquity of contact between Bushmen and farmers in the Namibian Kalahari. Further west, he and Jacobson excavated at Geduld (Smith and Jacobson 1995). To the south, Smith excavated several sites in the Orange River valley on the fringe of the South African Kalahari (Beaumont et al. 1995; Smith 1995). Other research on hunters and herders in the Namibian and the South African Kalahari and its fringes has been carried out by staff of the University of Cologne in northwestern Namibia, as well as by the Kinahans in western Namibia, and Beaumont, Humphreys, Thackeray, and Parsons in the northern Cape of South Africa (Albrecht et al. 2001; Beaumont and Vogel 1984; Humphreys and Thackeray 1983; Kinahan 1991, 2000; Parsons 2002). Despite the many active researchers, the archaeology of the Kalahari overall remains poorly understood. This is partly due to the lack of rich and informative sites. Small group sizes and the degree of mobility required to survive in a thirstland meant that few artifacts were left behind on small, short-term encampments. There are few rock outcrops and other natural features which would attract huntergatherers to camp repeatedly in the same spot – sites with signiﬁcant archaeological deposits are thus rare in the Kalahari. Another factor that may account for the low density of known archaeological sites is the size of the Kalahari and the difﬁcult logistics for archaeological work in such an inhospitable region. We are still far from a detailed understanding of late Holocene hunters and herders in the Kalahari, but a vague outline of their history can perhaps be discerned. The focus here is, ﬁrst, on the interactions between Bushman hunter-gatherers and Bantu farmers and, second, on the signiﬁcance and antiquity of livestock herding among Kalahari hunter-gatherer groups.

The Kalahari Debate Few, if any, Kalahari Bushmen now make a living by hunting and gathering. Since the 1950s, cordon fences designed to check the spread of foot and mouth disease have disrupted game migrations, while mechanically dug bore holes have allowed cattle posts to encroach deeper and deeper into the Kalahari, helping to drive away game while providing employment for Bushmen as cowboys (Campbell 1981; Hitchcock 1978; Williamson and Williamson 1981). The popular image of the Bushmen as the last true hunter-gatherers owes much to the publications of van der Post (1958) and the Marshall (1976) family, as well as to the Harvard ethnographers who worked in the Kalahari during the 1960s (Lee and DeVore 1976). In the 1980s and 1990s other scholars reacted against this isolationist view, and began to explore the history and recent archaeology of

_____________________________________________________________________________________________________

209

______________________________________________________________________________

Karim Sadr

Kalahari groups (e.g., Wolf 1982). Wolf, Wilmsen, and others suggested that the Kalahari Bushmen, far from being Paleolithic leftovers, had been client-herders of Bantu-speaking, iron-using, village farmers and herders since the ﬁrst few centuries ad, and had been supplying ivory to the Indian Ocean commercial network for at least 1,000 years (Wilmsen 1989b; Wilmsen and Denbow 1990). The stormy debate between the ‘‘isolationists’’ and the ‘‘revisionists’’ has calmed down and a synthesis now seems at hand (Sadr 1997). All Bushmen throughout southern Africa were clearly not major players in the mercantile world of the eleventh century ad as the revisionists indicated, but neither were they living fossils of Paleolithic hunting and gathering economies as the Harvard ethnographers had implied. Different groups of Bushmen in different parts of the subcontinent experienced contact and assimilation at different times and to different degrees (Kent 2002). Signiﬁcantly, archaeology now shows that across southern Africa many groups of Bushmen herded small stock even before the arrival of Bantuspeakers in the second century ad (Henshilwood 1996; Sealy and Yates 1994; Webley 1992). Clearly, some Bushmen were not ‘‘pristine’’ hunter-gatherers even 2,000 years ago. Three archaeological case studies are presented below to illustrate the variations in the timing and tempo of contact and assimilation among hunters and herders in and around the Kalahari. I start with the Thamaga area on the southeastern margin of the Kalahari, where Bushmen became absorbed into the herding and farming economy of Bantu-speakers in the nineteenth century. Then I discuss the sequence at the Shashe Limpopo conﬂuence, well off the eastern edge of the Kalahari, where the same thing happened as at Thamaga, but much earlier. Finally, I turn to the deepest Kalahari, where the process of Bushman assimilation only now seems to have reached its ﬁnal phase.

Contact between Bushmen and Bantu in the Kalahari It is believed that, beginning in the second century ad, Bantu-speaking, iron-using, village-dwelling farmers and herders began to spread southwards from the great lakes region of East Africa (Phillipson 1993). By the third century ad their villages were established on the coast of Kwa Zulu Natal Province in South Africa (Maggs 1984). By the seventh century ad, Iron Age villages had spread to the fringes of the Kalahari (Campbell et al. 1991, 1996; Denbow and Wilmsen 1983, 1986). The continuing inﬂux of farmers and herders from the north, the growth of populations already in southern Africa, and, beginning in the seventeenth century, the encroachment of European settlers, interacted to force some elements of the Bantu-speakers into the Kalahari proper (Boonzaier et al. 1996; Hall 1987; Huffman 1989). In close contact, it was generally (but not always) the Indigenous Bushmen who were culturally and economically assimilated into the Bantuspeakers’ world. The archaeological case study presented below, from the southeastern fringe of the Kalahari, illustrates the process.

_____________________________________________________________________________________________________

210

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

Contact at Thamaga Approximately 40 km west of Gaborone, the capital of Botswana, and only 50 km off the southeastern edge of the Kalahari sands, a long sequence of contact between hunter-gatherers and farmers is documented in two rock shelters – the Radiepolong and Ostrich shelters – near the village of Thamaga (Sadr 2002; Sadr and Plug 2001). Early Iron Age villages, presumably occupied by Bantu-speaking farmers and herders, appeared in this landscape ca. ad 600 and lasted until ca. ad 1000 (Campbell et al. 1996). The succeeding Middle Iron Age villages from ad 1000– 1500 contain a different style of ceramics known as Eiland Ware. In turn, the Late Iron Age villages from about ad 1500 to the early nineteenth century can be identiﬁed by yet another style of ceramics known as Moloko Ware. These Late Iron Age villages may represent settlements of early Tswana-speakers. Both Middle and Late Iron Age ceramics are found in the Thamaga rock shelters, indicating contact between hunter-gatherers and farmers since at least ad 1000. The early contact period is marked at Radiepolong rock shelter by a major peak in the quantities of artifacts, especially in formal stone tools such as arrowheads and skin scrapers, in the numbers of ostrich eggshell fragments and beads, in the numbers of animal bones, and other general occupational debris. This peak might merely represent a period when the shelter was inhabited for longer or by more people. It might, however, also represent an episode of increased production of traditional goods, such as beads and processed skins. Such a peak in production is evident in contact situations elsewhere too, in places as far apart as the Shashe Limpopo area, the Namib Desert, and the Australian interior (Hall and Smith 2000; Kinahan and Kimber pers. comm.). The peak in material debris may thus provide proxy evidence for heightened activity by hunter-gatherers, perhaps in response to the new economic climate created by proximity to a village of farmers and herders. It coincides with a regional economic boom created by the Indian Ocean trade network of that time. The Middle Iron Age village site of Moritsane, 20 km east of Thamaga, yielded a cache of glass beads and a cowrie shell, indicating its participation in the regional and international trade (Denbow and Wilmsen 1983: 406). It is possible that the Thamaga hunter-gatherers were also involved, however peripherally, in the Indian Ocean trade network of the time via Middle Iron Age villages such as Moritsane. Oral histories combined with archaeology suggest that the Middle Iron Age villagers of the Thamaga area, such as the occupants of Moritsane, were pushed into the Kalahari a few centuries ago by Tswana-speaking newcomers (Campbell et al. 1991; Okihiro 1976). The relationship between the Thamaga hunter-gatherers and the newly arrived Late Iron Age farmers was different. The late contact period in the Thamaga rock shelters is marked by layers which contain Late Iron Age ceramics in fairly high numbers, and much fewer of the traditional items such as stone tools and ostrich eggshell beads than in the early contact layers. The increase in the quantities of exotic, imported objects as seen in the rock shelters of Thamaga, and the concomitant decrease in the numbers of traditional objects manufactured by the hunter-gatherers, is perhaps an indication of the gradual

_____________________________________________________________________________________________________

211

______________________________________________________________________________

Karim Sadr

assimilation of the Thamaga Bushmen into the farmers’ world during the late contact period. There was an overall drop in the quantities of material debris in the late contact layers at the Thamaga shelters; this may reﬂect the economic marginalization of the Bushmen beginning about three centuries ago. At Ostrich shelter, occupied in the nineteenth century and located immediately adjacent to a contemporary Late Iron Age village, assimilation and marginalization had progressed so far that only a few stone tools and ostrich eggshell beads, along with now faded rock art, help us to identify the shelter occupants as the descendants of local hunter-gatherers. Their economy in the nineteenth century could no longer be described as hunting and gathering. A high proportion of bones of domestic stock and imported ceramic vessels found in Ostrich shelter indicate consumers of domesticated foods such as the meat of sheep and cattle and presumably the cereal products that were transported in ceramic pots. The domesticated foods may have been introduced into the shelter as rations from the neighboring Late Iron Age farmers in payment for services rendered, an arrangement which is common in Kalahari cattle posts (Hitchcock 1978; Silberbauer and Kuper 1966). The cultural and economic assimilation seen in the nineteenth-century layer at Ostrich shelter may represent a case of encapsulation in its original sense (Woodburn 1988); a condition when there was too little open land or game left for the pursuit of a traditional hunting and gathering life (see Sampson 1995 for a similar situation in the Karoo Desert of South Africa). Just 4 km away from Ostrich, the contemporary terminal occupation at Radiepolong shelter shows a lesser degree of material acculturation and less reliance on domestic animals. Even as short a distance as an hour long walk seems to have made a substantial difference in the effects of contact and its resultant pressures to assimilate. The same thing happened in the Shashe Limpopo area, but earlier.

Contact in the Shashe Limpopo Rock shelters in the South African part of the Shashe Limpopo show signiﬁcant similarities with those in Thamaga (Hall and Smith 2000). Prior to 2,000 bp, the Limpopo Basin was inhabited by LSA hunters and gatherers. From ad 350–600 contact at a distance is attested by the occasional potsherd found in the rock shelter deposits. At Salt Pan shelter, for example, the LSA hunter-gatherer levels are overlain by 30 cm of early pottery period occupation, which can be divided into two phases. The earlier phase contains so-called Bambata ceramics that may have been produced by the Indigenous hunter-gatherers themselves. The later part of the early pottery period includes potsherds of the so-called Happy Rest type, which are associated with Early Iron Age farmers in this region. The farmers’ villages, however, have so far only been found some 50 km south of the Limpopo Valley. The valley itself was perhaps too dry for farmers in the ﬁrst millennium, and may have formed a refuge for the hunter-gatherers of that time (Hall and Smith 2000). Echoing the ﬁnds at Thamaga, at a rock shelter in the Limpopo Valley called Little Muck ca. ad 1000 there was an increase in the number of scrapers with a

_____________________________________________________________________________________________________

212

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

concomitant decrease in arrowheads, more worked bone tools, ostrich eggshell, and land snail shells (for bead manufacture), ocher for painting, and animal bones from food waste. This peak in material remains at Little Muck, which recalls the peak in the Thamaga shelters, came at a time when the ﬁrst Iron Age chiefdom was emerging in the Limpopo Valley with its capital at the site of Schroda (Huffman 2000). The heightened activity at Little Muck during this period has been interpreted by the excavator as a reﬂection of the intensity of contact with the Iron Age chiefdom. The hunter-gatherers probably supplied hides and shell beads to the farmers who in turn were positioned at the terminus of the nascent Indian Ocean trade network. A short distance of 3–4 km from Little Muck, at the shelter called Balerno, contemporary hunter-gatherer occupations showed little sign of such ‘‘hyperactivity,’’ perhaps because even at that short a distance it lay outside the interaction sphere of the nearest farmer settlement at Leokwe Hill (Hall and Smith 2000). This ﬂorescence of contact and hunter-gatherer participation in the larger trade networks did not last long. By the time the Iron Age chiefdom had become more complex and a state society had crystallized in the Limpopo Valley with successive capitals at K2 and Mapungubwe (ca. ad 1100–1270), activity at Little Muck had subsided. There is hardly any debris in the rock shelters of this period and the excavator believes the proﬁtability of the Indian Ocean trade had led the lower classes of the farmer society to usurp the economic role played by the huntergatherers. Alternatively, the hunter-gatherers themselves may have become the lower classes. In either case, after ad 1350 there are no more traces of conventional hunter-gatherer LSA technology to be found in the Shashe Limpopo area, suggesting the total (material) assimilation of the local Bushmen into the farmer’s world.

The Deepest Kalahari Wilmsen (1989b), using evidence from his excavations at Cae Cae, claimed that in the deepest Kalahari the Bushmen had become encapsulated client herders of the Early Iron Age farmers by the sixth century ad. This would suggest intense, close contact and assimilation in the deepest Kalahari even earlier than in the Shashe Limpopo. The archaeological site at Cae Cae is an extensive scatter of LSA artifacts spread over an area of more than 3 sq km. This large area was sampled by Wilmsen (1978) with dozens of excavation pits. The ﬁnds included many typical LSA stone tools and other traditional items such as ostrich eggshell beads, plus a very few imported pieces of pottery and metals. The animal bones were mostly from wild game, but a few bones of cattle were also found in layers dated as early as ad 800. The numbers of imported objects are small, and the cow bones – assuming they were indeed correctly identiﬁed and associated with the dated levels (see Yellen 1990; Yellen and Brooks 1989, 1990) – account for a relatively small proportion of the entire mammalian faunal sample (Wilmsen 1989a). Rather than encapsulation of Bushmen at Cae Cae since the mid-ﬁrst millennium ad, these quantities of foreign objects, when compared with the evidence from Thamaga and the Shashe Limpopo area, suggest contact at a distance with Iron Age villages, presumably those at Nqoma and later Divuyu in the Tsodilo Hills over 100 km to the northeast

_____________________________________________________________________________________________________

213

______________________________________________________________________________

Karim Sadr

of Cae Cae. Like the remains from the period of contact at a distance in the Shashe Limpopo and Thamaga shelters, the hunter-gatherers of Cae Cae seem to have maintained their traditional diet, tool kit, and other technologies while sampling new materials such as ceramics and receiving some domestic animals such as cattle. Perhaps signiﬁcantly, the archaeological remains at Cae Cae show no sign of the heightened activity followed by the drop in traditional technologies, which characterize the sequence of contact and assimilation in the Shashe Limpopo and Thamaga areas. Recent archaeological and ethnographic work at Cho/ana on the Namibian side of the border supports this general picture: the hunter-gatherers in the deepest Kalahari maintained only contact at a distance with Bantu-speakers until the twentieth century (Lee 2002; Smith 2001; Smith and Lee 1997). Contact seems to have intensiﬁed since the second half of the twentieth century, when bore hole technology opened the desert to ranchers. Today’s Bushman craft producing village of D’Kar, near Ghanzi, from whence most of the ostrich eggshell bead necklaces and Bushman hunting kits found in southern African souvenir shops come from, might represent the period of heightened production of traditional items – the hyperactive phase – which in the Shashe Limpopo Valley and the Thamaga area happened in the ﬁrst and early second millennia ad, just before full assimilation of hunter-gatherers. In the Tsodilo Hills themselves, however, the situation was somewhat different. Here, Bushmen hunter-gatherers were apparently involved in mining and trading specularite (a crystalline form of hematite with a distinctive metallic luster) – widely used throughout southern Africa for sparkling body and hair decoration – with the Iron Age villagers who had settled at Tsodilo in the sixth century ad. The intensive mining activity documented at several places in the hills (Robbins et al. 1998) may represent the local version of a hyperactive phase of contact. Overall, the evidence from Cae Cae, far from showing early encapsulation of the Bushmen as Wilmsen claims, seems to conﬁrm that the Harvard ethnographers of the 1960s were indeed there to witness the end of the traditional Bushman way of life in the deepest Kalahari. Nevertheless, to say the deepest Kalahari was in effect the ﬁnal refuge of the traditional Bushman way of life is not to say it was the ﬁnal refuge of Paleolithic hunters and gatherers. As the next section attempts to illustrate, the Bushmen in the Kalahari had already ceased to be pristine huntergatherers a few centuries before direct contact with Iron Age farmers.

The First Herders in the Kalahari A more recent generation of ethnographers in the central Kalahari, such as Kent (1992, 1993) and Ikeya (1993), recognized and reported what the 1960s ethnographers such as Silberbauer (1965) and Tanaka (1976) ignored: Kalahari Bushmen hunter-gatherers kept small herds of goats and even tended melon gardens. Kent provides much detail about this form of low intensity food production among hunter-gatherers, pointing out that it consumed only about 10 percent of total time spent on subsistence activities and showed up in the camp’s refuse also as

_____________________________________________________________________________________________________

214

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

about 10 percent of the food debris. The Bushmen at Kutse, where Kent did her groundbreaking ethnographic work, considered themselves to be hunter-gatherers and strictly followed the traditions of sharing the meat of hunted game to maintain equality in their community. But they had found a way to own and herd domesticated small stock as a fallback option – a form of insurance policy – without negating their egalitarian ethos. Archaeology now suggests that this type of low intensity food production has considerable antiquity among the southern African hunter-gatherers, and predates the arrival of the Iron Age farmers by a few centuries. As described below, a few case studies from the Kalahari and its fringes provide an overview of this earliest herding economy in southern Africa.

Neolithic Herders at Toteng Just south of the Okavango Delta, near Lake Ngami, a series of sites was excavated around the village of Toteng by Campbell (1992). These open air sites in Kalahari sands preserved traces of a long sequence of occupation, beginning with LSA hunter-gatherers from the last few centuries bc. Unusually for the Kalahari, populations at Toteng had been attracted repeatedly to the same knolls that lay just above the reach of the ﬂooding Thamalakane River, thus creating archaeological sites with signiﬁcant depth to their deposits. The layers from the ﬁrst few centuries ad include pieces of thin-walled ceramic vessels of a type that have come to be known as Bambata Ware, named after the cave in the Matopos Hills of Zimbabwe where they were ﬁrst discovered (Walker 1983). As at Bambata Cave, at Toteng these thin-walled ceramics occur in an LSA context where many of the stone tools and ostrich eggshell beads show clear continuity with the pre-ceramic LSA occupations. These early ceramic layers at Toteng also include bones of domesticated sheep or goats and cattle, and show changes in the proportions of formal stone tools. Livestock at Toteng (and Bambata Cave) are only present in very small numbers: they constitute about a tenth of the mammalian bones found at these sites. Such proportions are comparable with the numbers generated by Kent at Kutse. The speciﬁc changes in the formal stone tools at Toteng are also illuminating. There was a shift away from the production of arrowheads towards the production of scrapers. The evidence suggests that at Toteng, nearly four or ﬁve centuries before the arrival of the Early Iron Age farmers, the LSA inhabitants had ceased to be pristine hunters and gatherers and had become low intensity herders. Further west, a similar situation is evident in Namibia.

Early Herders in Namibia At Geduld we can see the same sorts of changes as documented at Toteng, but with a local ﬂavor. Situated in north central Namibia, the site of Geduld is an LSA occupation and has pristine hunter-gatherer layers dated prior to 2,000 bp (Smith and Jacobson 1995). As at Toteng, in the ﬁrst few centuries ad thin-walled ceramics and small stock were introduced to Geduld. What is particularly interesting is the

_____________________________________________________________________________________________________

215

______________________________________________________________________________

Karim Sadr

style of ceramics at Geduld. Technologically, these sherds are of the thin-walled wares associated with early herding sites throughout southern Africa, but they were not decorated in quite the same style as the Bambata Ware from Toteng. The speciﬁc style of pottery decoration at Geduld suggests a local ceramic stylistic tradition. Oruwanje, further northwest in Namibia, also shows all the typical elements of an early ﬁrst millennium ad herding site, and its thin-walled vessels sport yet another distinct, local style of decoration (Albrecht et al. 2001; Keding and Vogelsang 2001). There are plenty of other such early ﬁrst millennium ad herding sites throughout southern Africa (e.g., Deacon et al. 1978; Schweitzer 1974; Webley 1992). Taken together, they point to the same fact: before the arrival of the Iron Age farmers and herders, Bushmen in southern Africa had ceased to be pristine hunters and gatherers, had adopted herding to various degrees, and had developed different styles of ceramics in different parts of southern Africa. Interesting questions remain: Did some groups of hunters and gatherers entirely reject food production, and if so why? Why did other groups, such as at Kasteelberg on the west coast of South Africa during the late ﬁrst millennium, go beyond low intensity herding to engage in intensive animal husbandry (Klein and Cruz-Uribe 1989)? And why did some of these groups eventually revert to low intensity herding or even to hunting and gathering without food production?

Hunter-Gatherers and Herders of the Kalahari The popular image of modern Kalahari Bushmen as the last of the Paleolithic hunter-gatherers is clearly myth. The archaeological record suggests that by about 2,000 bp many hunter-gatherer communities throughout southern Africa started to adopt small stock herding as a minor component of their subsistence activities. In parts of the central Kalahari such low intensity food production strategies could still be documented in the 1980s and 1990s. In other parts of the subcontinent, such as on the west coast, certain groups went further and established economies based on more intensive husbandry of small stock. Other Bushmen communities elsewhere in southern Africa may have gone to the other extreme and rejected animal husbandry. To make this patchwork of Neolithic adaptations in southern Africa even more complex, through time people changed their economic focus: intensive pastoralists became hunters-with-sheep, hunters-with-sheep went back to full time hunting and gathering, and full-time hunter-gatherers elsewhere decided to give intensive animal husbandry a go, at least for a while. The Neolithic gradually came to an end when, at some point in time, depending on where in the subcontinent they lived, Bushmen came into close contact with Iron Age, village-dwelling farmers and herders (or Europeans as the case may be). In the early phases of close contact and interaction, economic activity among the hunter-gatherers seems to have peaked, perhaps in response to new economic opportunities provided by the newcomers. Interestingly, such peaks in production have been observed elsewhere in Namibia and even Australia, in the close contact between Indigenous populations and Europeans (see Chapter 15, this volume).

_____________________________________________________________________________________________________

216

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i

Following this peak in production, the hyperactive phase of contact, many southern African LSA and Neolithic archaeological groups started to disappear from the record. Presumably, the majority were materially assimilated into the world of village farmers and lost their independent archaeological identity when they gave up their traditional crafts and ways of living. It is likely that this happened when the locals found themselves encapsulated in the farmer’s world, without the space and the means to continue their independent mobile, herding and foraging way of life. The evidence suggests that this happened ﬁrst around ad 1200 in the Shashe Limpopo, the heartland of the southern African economic boom during the Iron Age. In the economically more peripheral zones of southeastern Botswana, encapsulation took place later and was not complete until the nineteenth century. In the deepest Kalahari, it was perhaps only in the second half of the twentieth century that the traditional, Indigenous way of life became untenable with modern bore hole technology allowing the encroachment of cattle into the thirstland. Ultimately, geography, the environment, and the economic climate seem to have played the signiﬁcant roles in the timing and tempo of contact and assimilation between hunter-gatherers and herders in southern Africa.

References Albrecht, M., Berke, H., Eichhorn, B., Frank, T., Kuper, R., Prill, S., Vogelsang, R., and Wenzel, S. 2001: Oruwanje 95/1: A late Holocene stratigraphy in northwestern Namibia. Cimbebasia, 17, 1–22. Beaumont, P. B. and Vogel, J. C. 1984: Spatial patterning of the ceramic Later Stone Age in the Northern Cape Province, South Africa. In M. Hall, G. Avery, D. M. Avery, M. L. Wilson, and A. J. B. Humphreys (eds), Frontiers: South African Archaeology Today, Cambridge: Cambridge Monographs in African Archaeology no. 10, BAR International Series, 207, 80–95. Beaumont, P. B., Smith, A. B., and Vogel, J. C. 1995: Before the Einiqua: The archaeology of the frontier zone. In A. B. Smith (ed.), Einiqualand: Studies of the Orange River Frontier, Cape Town: University of Cape Town Press, 236–64. Boonzaier, E., Malherbe, C., Smith, A., and Berens, P. 1996: The Cape Herders: A History of the Khoikhoi of Southern Africa. Cape Town: David Philip. Campbell, A. C. 1981: A comment on Kalahari wildlife and the Khukhe fence. Botswana Notes and Records, 13, 111–18. Campbell, A. C. 1992: Southern Okavango Integrated Water Development Study. Archaeological Survey of the Proposed Main Reservoir. Gaborone: Department of Water Affairs. Campbell, A. C. 1998: Archaeology in Botswana: Origins and growth. In P. Lane, A. Reid, and A. Segobye (eds), Ditswa Mmung: The Archaeology of Botswana, Gaborone: Botswana Society and Pula Press, 24–49. Campbell, A. C., Holmberg, G., and van Waarden, C. 1991: A note on recent archaeological research around Gaborone. Botswana Notes and Records, 23, 288–90. Campbell, A. C., van Waarden, C., and Holmberg, G. 1996: Variation in the Early Iron Age of southeastern Botswana. Botswana Notes and Records, 28, 1–22. Clement, A. J. 1967: The Kalahari and its Lost City. Cape Town: Longman. Cooke, H. J. 1979: The origin of the Makgadikgadi Pans. Botswana Notes and Records, 11, 37–42.

_____________________________________________________________________________________________________

217

______________________________________________________________________________

Karim Sadr Deacon, H. J., Deacon, J., Brooker, M., and Wilson, M. 1978: The evidence for herding at Boomplaas Cave in the southern Cape, South Africa. South African Archaeological Bulletin, 33, 39–65. Deacon, J. and Lancaster, N. 1988: Late Quaternary Paleoenvironments of Southern Africa. Oxford: Clarendon Press. Denbow, J. R. 1984: Prehistoric herders and foragers of the Kalahari: The evidence for 1,500 years of interaction. In C. Schrire (ed.), Past and Present in Hunter-Gatherer Studies, Orlando, FL: Academic Press, 175–94. Denbow, J. R. 1986: A new look at the later prehistory of the Kalahari. Journal of African History, 27, 3–28. Denbow, J. R. 1990: Congo to Kalahari: Data and hypotheses about the political economy of the western stream of the Early Iron Age. African Archaeological Review, 8, 139–76. Denbow, J. R. and Wilmsen, E. N. 1983: Iron Age pastoralist settlements in Botswana. South African Journal of Science, 79, 405–8. Denbow, J. R. and Wilmsen, E. N. 1986: Advent and the course of pastoralism in the Kalahari. Science, 234, 1509–15. Farini, G. A. 1973: Through the Kalahari Desert. Cape Town: Struik. [Facsimile reprint of 1886 edition.] Hall, M. 1987: The Changing Past: Farmers, Kings, and Traders in Southern Africa 200–1860. Cape Town: David Philip. Hall, S. L. and Smith, B. W. 2000: Empowering places: Rock shelters and ritual control in farmer–forager interactions in the Northern Province. South African Archaeological Society Goodwin Series, 8, 30–46. Henshilwood, C. 1996: A revised chronology for pastoralism in southernmost Africa: New evidence for sheep at ca. 2,000 bp from Blombos Cave, South Africa. Antiquity, 70, 945–9. Hitchcock, R. K. 1978: Kalahari Cattle Posts. Gaborone: Ministry of Local Government and Lands, Republic of Botswana. Huffman, T. N. 1989: Iron Age Migrations. Johannesburg: University of Witwaterstrand Press. Huffman, T. N. 1994: Toteng pottery and the origins of Bambata. Southern African Field Archaeology, 3, 3–9. Huffman, T. N. 2000: Mapungubwe and the origins of the Zimbabwe Culture. South African Archaeological Society Goodwin Series, 8, 14–29. Humphreys, A. J. B. and Thackeray, A. I. 1983: Ghaap and Gariep: Later Stone Age Studies in the Northern Cape. Cape Town: South African Archaeological Society. Ikeya, K. 1993: Goat raising among the San in the central Kalahari. African Studies Monograph, 14 (1), 39–52. Keding, B. and Vogelsang, R. 2001: Vom Ja¨ger-Sammler zum Hirten: Wirtschaftswandel im nordo¨stlichen und su¨dwestlichen Afrika. In B. Gehlen, M. Heinen, and A. Tillmann (eds.), Zeit-Ra¨ume: Gedenkschrift fu¨r Wolfgang Taute, Bonn: Deutschen Gesellschaft fu¨r Ur-und Fru¨hgeschichte, 257–82. Kent, S. 1992: The current forager controversy: Real versus ideal views of hunter-gatherers. Man, 27, 45–70. Kent, S. 1993: Sharing in an egalitarian Kalahari community. Man, 28, 479–514. Kent, S. (ed.) 2002: Ethicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa. Washington, DC: Smithsonian Institution Press. Kinahan, J. 1991: Pastoral Nomads of the Central Namib Desert. Windhoek: Namibia Archaeological Trust.

_____________________________________________________________________________________________________

218

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i Kinahan, J. 2000: Cattle for Beads: The Archaeology of Historical Contact and Trade on the Namib Coast. Uppsala: Department of Archaeology and Ancient History. Klein, R. G. and Cruz-Uribe, K. 1989: Faunal evidence for prehistoric herder–forager activities at Kasteelberg, Western Cape Province, South Africa. South African Archaeological Bulletin, 44, 82–97. Lee, R. B. 2002: Solitude or servitude? Ju/’hoansi images of the colonial encounter. In S. Kent (ed.), Ethnicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 184–205. Lee, R. B. and DeVore, I. 1976: Kalahari Hunter-Gatherers. Cambridge, MA: Harvard University Press. Maggs, T. 1984: The Iron Age south of the Zambezi. In R. G. Klein (ed.), Southern African Prehistory and Paleoenvironments, Rotterdam: A. A. Balkema, 221–329. Marshall, L. J. 1976: The !Kung of Nyae Nyae. Cambridge, MA: Harvard University Press. Okihiro, G. Y. 1976: Hunters, Herders, Cultivators, and Traders: Interaction and Change in the Kgalagadi, Nineteenth Century. Ann Arbor, MI: University Microﬁlms International. Parsons, I. 2002: Later Stone Age lifeways in the Northern Cape, SA. Cape Town: Unpublished paper presented at the biennial conference of the South African Association of Archaeologists, 7–10 July. Passarge, S. 1904: Die Kalahari. Berlin: Dietrich Reimer. Passarge, S. 1907: Die Buschma¨nner der Kalahari. Berlin: Dietrich Reimer. Phillipson, D. W. 1993: African Archaeology. Cambridge: Cambridge University Press. Reid, A. and Segobye, A. 2000: Politics, society, and trade on the eastern margins of the Kalahari. South African Archaeological Society Goodwin Series, 8, 58–68. Robbins, L. 1985: The Manyana rock painting site. Botswana Notes and Records, 17, 1–15. Robbins, L. 1986: Recent archaeological research in southeastern Botswana: The Thamaga site. Botswana Notes and Records, 18, 1–13. Robbins, L. H. 1990: The Depression Site: A Stone Age sequence in the northwest Kalahari Desert, Botswana. National Geographic Research, 6, 329–38. Robbins, L. H. and Campbell, A. C. 1989: The Depression rock shelter, Tsodilo Hills. Botswana Notes and Records, 20, 1–3. Robbins, L. H., Murphy, M. L., Campbell, A. C., and Brook, G. A. 1998: Intensive mining of specular hematite in the Kalahari. Current Anthropology, 39, 144–50. Robbins, L. H., Murphy, M. L., Stevens, N. J., Brook, G. A., Ivester, A. H., Haberyan, K. A., Klein, R. G., Milo, R., Stewart, K. M., Matthiesen, D. G., and Winkler, A. J. 1996: Paleoenvironment and Archaeology of Drotsky’s Cave: Western Kalahari Desert, Botswana. Journal of Archaeological Science, 23, 7–22. Sadr, K. 1997: Kalahari archaeology and the Bushman debate. Current Anthropology, 38, 104–12. Sadr, K. 2002: Encapsulated Bushmen in the archaeology of Thamaga. In S. Kent (ed.), Ethnicity, Hunter-gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 28–47. Sadr, K. and Plug, I. 2001: Faunal remains in the transition from hunting to herding in southeastern Botswana. South African Archaeological Bulletin, 56, 76–82. Sampson, C. G. 1995: Acquisition of European livestock by Seacow River Bushmen between ad 1770–1890. South African Field Archaeology, 4, 30–6. Schweitzer, F. R. 1974: Archaeological evidence for sheep at the Cape. South African Archaeological Bulletin, 29, 75–82.

_____________________________________________________________________________________________________

219

______________________________________________________________________________

Karim Sadr Sealy, J. and Yates, R. 1994: The chronology of the introduction of pastoralism to the Cape, South Africa. Antiquity, 68, 58–67. Silberbauer, G. B. 1965: Report to the Government of Bechuanaland on the Bushman Survey. Gaborone: Bechuanaland Government Publications. Silberbauer, G. B. and Kuper, A. J. 1966: Kgalagari masters and Bushman serfs: Some observations. African Studies, 25, 171–9. Smith, A. B. 1995: Archaeological observations along the Orange River and its hinterland. In A. B. Smith (ed.), Einiqualand: Studies of the Orange Rive Frontier, Cape Town: University of Cape Town Press, 265–301. Smith, A. B. 2001: Ethnohistory and archaeology of the Ju/’hoansi Bushmen. African Study Monographs Supplement, 26, 15–25. Smith, A. B. and Jacobson, L. 1995: Excavations at Geduld and the appearance of early domestic stock in Namibia. South African Archaeological Bulletin, 50, 3–14. Smith, A. B. and Lee, R. B. 1997: Cho/ana: Archaeological and ethnohistorical evidence for hunter-gatherer/agropastoralist contact in northern Bushmanland, Namibia. South African Archaeological Bulletin, 52, 52–8. Tanaka, J. 1976: Subsistence ecology of central Kalahari San. In R. B. Lee and I. DeVore (eds.), Kalahari Hunter-Gatherers: Studies of the !Kung San and their Neighbors, Cambridge, MA: Harvard University Press, 98–120. Thomas, D. S. G. and Shaw, P. 1991: The Kalahari Environment. Cambridge: Cambridge University Press. van der Post, L. 1958: Lost World of the Kalahari. London: Hogarth. van Waarden, C. 1999: The Prehistory and Archaeology of Botswana: An Annotated Bibliography. Gaborone: Botswana Society. Walker, N. 1983: The signiﬁcance of an early date for pottery and sheep in Zimbabwe. South African Archaeological Bulletin, 38, 88–92. Walker, N. 1994: The Later Stone Age of Botswana: Some recent excavations. Botswana Notes and Records, 26, 1–35. Walker, N. 1995: The archaeology of the San. In A. J. G. M. Sanders (ed.), Speaking for the Bushmen, Gaborone: Botswana Society, 54–85. Walker, N. 1998: The Later Stone Age. In P. Lane, A. Reid, and A. Segobye (eds), Ditswa Mmung: The Archaeology of Botswana, Gaborone: Botswana Society and Pula Press, 65–80. Webley, L. 1992: Early evidence for sheep from Spoegrivier cave, Namaqualand. Southern African Field Archaeology, 1, 3–13. Williamson, D. T. and Williamson, J. E. 1981: An assessment of the impact of fences on large herbivore biomass in the Kalahari. Botswana Notes and Records, 13, 107–10. Wilmsen, E. N. 1978: Prehistoric and historic antecedents of a contemporary Ngamiland community. Botswana Notes and Records, 10, 5–18. Wilmsen, E. N. 1989a: The antecedents of contemporary pastoralism in western Ngamiland. Botswana Notes and Records, 20, 29–39. Wilmsen, E. N. 1989b: Land Filled with Flies. Chicago: University of Chicago Press. Wilmsen, E. N. and Denbow, J. R. 1990: Paradigmatic history of San-speaking peoples and current attempts at revision. Current Anthropology, 31, 489–524. Wolf, E. 1982: Europe and the People Without History. Berkeley: University of California Press. Woodburn, J. 1988: African hunter-gatherer social organization: Is it best understood as a product of encapsulation? In T. Ingold, D. Riches, and J. Woodburn (eds), Hunters and Gatherers, Volume 1: History, Evolution, and Social Change, Oxford: Berg, 31–65.

_____________________________________________________________________________________________________

220

______________________________________________________________________________

H u n t e r - Ga t h e r e r s a n d H e r d e r s o f t h e K a l a h a r i Yellen, J. E. 1990: Comment on: Paradigmatic history of San-speaking peoples and current attempts at revision, by E. N. Wilmsen and J. R. Denbow. Current Anthropology, 31, 516–17. Yellen, J. E. and Brooks, A. S. 1989: The Later Stone Age archaeology of the !Kangwa and /Xai /Xai Valleys, Ngamiland. Botswana Notes and Records, 20, 5–28. Yellen, J. E. and Brooks, A. S. 1990: The Later Stone Age archaeology in the /Xai /Xai region: A response to Wilmsen. Botswana Notes and Records, 22, 17–20.

_____________________________________________________________________________________________________

221

______________________________________________________________________________

-----------------------------------------------------------------------------

12

--------------------------------------------------------------------------

Desert Archaeology, Linguistic Stratigraphy, and the Spread of the Western Desert Language Mike Smith

Introduction Prehistory is written from many sources. When I began archaeological work in central Australia more than 20 years ago, my long-term objective was to use different strands of evidence to provide a ﬁne-grained reconstruction of the prehistory of the region. I hoped – eventually – to be able to compare a regional occupation sequence, obtained through archaeological excavations, with detailed information on changes in regional environments, rock art, patterns of trade and exchange between regions, and data on the genetic and linguistic history of central Australian Aboriginal groups. My premise was that an interdisciplinary approach and a wide range of data would be necessary before we could understand the development of the distinctive central Australian societies documented by ethnographers at the close of the nineteenth century. The task as I saw it was broadly historical, but concerned with cultural process and with environmental history, as much as with culture history. The problem is that it has taken much longer to build up this data-set than anyone imagined. Until recently, one major gap has been information relating to the linguistic history of central Australian groups. Despite a brief ﬂirtation in the late 1960s (e.g., Wurm 1972), linguists and archaeologists have largely worked in isolation over the last 30 years, each amassing larger data-sets for central Australia and the Western Desert and developing more detailed interpretations of linguistic relationships and regional prehistory, respectively. This period of independent growth has been beneﬁcial to both disciplines, but it is time to bring the two strands of research closer together, as McConvell (1990) has also strongly advocated. In the 1980s linguists were extraordinarily coy about the historical relationships and chronology of central Australian languages. This changed sharply with the ﬁrst Archaeology and

_____________________________________________________________________________________________________

222

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

Linguistics (ARCLING) conference in 1991 (McConvell and Evans 1997), followed in 1996 by McConvell’s pioneering synthesis of the linguistic prehistory of arid Australia, and in 1996 by a workshop on the origins of the Western Desert (Wati) language. In this chapter I provide a sketch of current archaeological knowledge of central Australia and the Western Desert and attempt to test the linguistic prehistory (as set out in McConvell 1996) against the archaeology of the region. A similar exercise was recently undertaken by Veth (2000), and both his paper and this chapter were originally written for the 1996 Western Desert workshop. While we reach similar conclusions about the apparent convergence of archaeological and linguistic data, in this chapter I want to draw attention to how we go about correlating the two data-sets.

Linguistic Prehistory Most Aboriginal people in the Western Desert today speak one or more of a series of closely related dialects (e.g., Pitjantjatjara or Pintupi) that are known collectively as Wati (see Figure 12.1). The expansion of the Wati language group to cover the entire western half of the Australian arid zone, >2,000,000 sq km, in recent

Figure 12.1 Map showing locations of Pama-Ngungan linguistic subgroups referred to in the text (from McConvell 1996: ﬁgure 1).

_____________________________________________________________________________________________________

223

_____________________________________________________________________________

M i k e S m it h

prehistoric times offers a clear focus for interdisciplinary research into desert prehistory. This pattern – involving the establishment of a common language and culture over a huge area – is precisely what we would expect to ﬁnd if the region had been recently colonized by people, but this raises many questions for the archaeology of the region.

The Spread of Wati In one of the most important contributions to desert prehistory over the last decade, McConvell (1996) outlined the linguistic prehistory of the region. He suggested that differentiation of languages within the Nyungic group began around 4,000–3,000 bp in central-western Queensland, with groups of Nyungic speakers moving both south (South Nyungic) and west (Northern Nyungic), bypassing existing non-Nyungic PN (Pama-Nyungan) groups, such as Arandic in central Australia. The subsequent break up of Northern Nyungic involved the movement of groups over the next millennium or so into the Tanami Desert, the Pilbara, the Gascoyne-Murchison region, and the western edge of the Western Desert. Over the next 2,000 years Warnman and Western Desert (Wati) languages differentiated and speakers of the latter moved east occupying the desert, reaching the western edge of the central Australian ranges by about 1,000 bp. Although the spread of a language may also occur through language shift, diffusion, and small scale population movements, McConvell (1996) opted strongly for migration as the principal mechanism of language spread through the Australian arid zone. The chronology for these developments is necessarily speculative. Historical linguistics provides only a relative sequence of linguistic differentiation, together with some poorly constrained estimates of the rate of change. If we assume uncertainties of at least 50 percent for these estimates, differentiation of the Nyungic group may have begun as early as 8,000 bp and the split between Wati and Warnman may have taken place sometime between 3,000–1,000 bp. In fact, McConvell’s estimate for the timing of Wati expansion largely relies on external factors – and here it is only weakly constrained by estimates of gene ﬂow between Western Desert and Arrernte populations and by tentative correlation with changes in the archaeological record ca. 1,000 bp in central Australia.

Implications for Archaeology Even a cursory reading of the archaeological literature will suggest the linguistic model presents a challenge to the interpretations of cultural continuity and internal development generally favored by archaeologists working in these regions. Archaeologists themselves have moved away from the model of long-term cultural stability originally proposed by Gould (1977), towards a more dynamic view of settlement and cultural change in desert systems. However, most would still lean towards reticulate models of cultural interaction and social change to explain observed trends in the record, rather than invoke widespread replacement of existing regional populations during the late Holocene (e.g., Evans and Jones

_____________________________________________________________________________________________________

224

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

1997; Thorley and Gunn 1996). Cultural and linguistic diffusion are likely to have been especially important factors in the Australian deserts because of the small, highly mobile population of these areas. Veth (pers. comm.) suggested that we could think of diffusion as operating as an earlier form of contemporary ritual and religious movements such as the ‘‘Desert Crusade’’ (cf. Evans and Jones 1997; Tonkinson 1974; Veth 2000). Distinguishing between cultural diffusion, language shift, and population movement is important for ideas about how desert cultural systems operated. Was Wati expansion part of a post-glacial recolonization of unoccupied land in the sandridge deserts (e.g., Smith 1993; Veth 1989); a correlate of new social networks in the Holocene (Veth 2000); or does it reﬂect displacement of an established desert population of hunter-gatherers by an outside group sometime in the late Holocene (McConvell 1996)? To discriminate between these scenarios we obviously need some internal chronological benchmarks for the linguistic sequence.

North American Deserts The North American experience shows how much is at stake here. In 1958, in an interesting parallel to McConvell’s study, Lamb (1958) argued that the historic distribution of Numic languages such as Shoshone, Paiute, and Ute in the desert regions of western North America was the result of a recent migration of Numicspeaking people across the region. However, after more than 30 years of research and despite a much richer archaeological record than in Australia there is still no general agreement on how the American linguistic and archaeological data ﬁt together (e.g., Bettinger and Baumhoff 1982; Grayson 1993: 258–76; Madsen and Rhode 1994). Some archaeologists allow Numic linguistic prehistory to drive their interpretation of archaeological data – effectively treating the linguistic model as the dominant paradigm. Others now doubt whether Numic prehistory can be reconstructed at all. In a recent review Madsen and Rhodes (1994: 3) commented: After a century, and despite the publication of scores of papers, about the only thing linguists, prehistorians, and ethnohistorians widely agree upon is the historic distribution of Numic languages.

The major obstacle to intercalibrating archaeological and linguistic sequences has been continuing uncertainty about Numic linguistic chronology. And because the archaeology of hunter-gatherers and historical linguistics provide such different structural perspectives on the past, any correlation of these invariably hinges on chronology. Without some chronological benchmarks for Numic languages, interdisciplinary analysis of these data-sets ﬂoundered. There are lessons here for Australian research. If we wish to use the different strengths of linguistics and archaeology to provide complementary perspectives on the prehistory of desert populations we will need a sound chronology for the linguistic sequence – not just a relative sequence of linguistic differentiation with some guesstimates of time depth. McConvell’s (1996) concept of ‘‘linguistic

_____________________________________________________________________________________________________

225

_____________________________________________________________________________

M i k e S m it h

stratigraphy’’ is an important breakthrough because it presents the possibility of correlations between linguistic and archaeological sequences by looking for items embedded within both sequences. Before looking at some of the opportunities for this sort of analysis in central Australia I want to examine some implications of the archaeological evidence for the linguistic prehistory.

Archaeological Background Initial Phase of Research The ﬁrst phase of systematic archaeological research in the Western Desert began with Gould’s (1977) excavation of Puntutjarpa rock shelter, near Warburton, between 1967 and 1970. The Puntutjarpa sequence was initially thought to be a more-or-less continuous cultural sequence spanning the last 10,000 years. Gould emphasized the lack of signiﬁcant change in the archaeological record and saw the site as exemplifying long-term cultural stability in the Western Desert. During the same period linguists such as Wurm (1972: 160–7) were beginning to draw connections between the spread of ‘‘Common Australian’’ and the late Holocene spread of new types of hafted artifacts. My impression is that the emerging archaeological picture of long-term cultural stability in the Western Desert cut short any attempt to pursue this line of thinking. In other parts of the world the existence of cultural and linguistic uniformity across a large geographic area is typically seen as evidence of recent population spread. In the Western Desert case it was seen as the product of cultural conservatism and open social networks (see, for example, Tonkinson 1978: 7), largely on the basis of the archaeological evidence.

Current Views Fieldwork in the arid zone since 1980 has revealed much greater regional and temporal variability in the archaeological record than previously allowed (e.g., Gunn 1995; Hiscock and Veth 1991; Smith 1988; Thorley 1998a; Thorley and Gunn 1996; Veth 1993). The new frameworks for desert prehistory that have emerged suggest the question of Wati origins should be reopened. This work also provides a framework against which any linguistic changes must have been played out. The late Pleistocene is not directly relevant to the period spanned by the linguistic prehistory but warrants brief comment here, followed by more detailed discussion of the last 5,000 years in the sections which follow. In doing this I shall try to restrict myself to those aspects where there is broad agreement among archaeologists working in the region about interpretation of the evidence. 1

Human groups were using rock shelter sites in various parts of the arid zone by 35,000–20,000 bp, including western central Australia (Puritjarra, Kulpi Mara), the Wiluna area (Serpent’s Glen), and in the Pilbara (O’Connor et al. 1998; Smith et al. 1997; Thorley 1998b; Veth 1995). Scale and resolution are perennial

_____________________________________________________________________________________________________

226

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

problems in interpreting the data, but in western central Australia the evidence appears to reﬂect occupation of the region rather than episodic visitation, with exploitation of regional ocher mines beginning by 32,000 bp (Smith et al. 1998). The same is probably also true of the Pilbara given the frequency of early evidence there. There is little evidence to indicate how widespread human occupation was in this period (Smith 1993). Interpretations vary according to how individual site records are related to broader patterns of adaptation; however, most researchers agree that occupation may have been patchy prior to 25,000 bp (cf. Veth 2000). 2 Paleoenvironmental reconstructions for the Last Glacial Maximum (LGM) suggest that the period of peak aridity, centered around 18,000 bp, would have created difﬁcult conditions for people in many parts of the arid zone (see also Chapter 3, this volume). This is likely to have led to some contraction of settlement in the arid zone, perhaps abandonment of some regions, but researchers differ on the extent, scale, and duration of any impact (cf. Smith 1989; Veth 1989). In western central Australia, people appear to have continued to visit Puritjarra rock shelter intermittently throughout the LGM, suggesting that human occupation of the central ranges and the sandy desert immediately west of the ranges continued throughout this period. Veth (1995) has summarized evidence for regional abandonment in northwestern Australia, based on links between patterns of rock shelter sedimentation and occupation. In the desert the clearest evidence for site abandonment is from Serpent’s Glen, where there is a sterile layer, dating somewhere between about 24,000 bp and 5,000 bp. The regional signiﬁcance of this gap in site use is yet to be determined. 3 By 12,000–10,000 bp there is evidence of occupation at several sites throughout the region, including Walga Rock in the Murchison region (Bordes et al. 1983), Puntutjarpa in the Warburton area, and at both Puritjarra and Kulpi Mara in western central Australia. At Puntutjarpa there is only a single hearth dated to ca. 10,200 bp, beneath rock fall and more recent occupation. At Puritjarra this time period is marked by an assemblage of distinctive large ﬂake implements. 4 Phytolith data and charcoal within the Puritjarra sediments indicate that the post-glacial period in central Australia saw a rapid amelioration of arid conditions with an increase in grasses and acacias by 13,000 bp, peaking around 7,000– 6,000 bp (Bowdery 1998; Smith et al. 1995). By 7,500–6,000 bp both Puntutjarpa and Puritjarra rock shelters have evidence for substantial occupation. At Puntutjarpa this is marked by a rich dark occupation soil, while at Puritjarra there is a major increase in the level of site use and evidence for a more localized site catchment. By 13,500–7,500 bp people may have begun painting the walls of the rock shelter and engraved boulders on the shelter ﬂoor with a series of plain circles, possibly an assertion of corporate rights in a site. Deeply patinated rock engravings at several locations in central Australia, the Pilbara, and in parts of the Western Desert (e.g., in the Calvert Range, south of Lake Disappointment) indicate systematic occupation of these regions by at least the early Holocene, though these could also relate to an earlier phase of occupation (perhaps preglacial). It is not clear when the major sandridge deserts, including the core of the

_____________________________________________________________________________________________________

227

_____________________________________________________________________________

M i k e S m it h

Western Desert, were ﬁrst occupied. Certainly, there is evidence for occupation by 5,000–3,000 bp (Smith and Clark 1993; Veth 1993). 5 On the western fringes of the Western Desert there is a series of excavated sites dating to the late Holocene. Sites in the Rudall River area have dates back to about 5,000–3,000 bp (Veth 1993). The upper unit at the Serpent’s Glen rock shelter dates from ca. 5,000 bp (O’Connor et al. 1998). Kaalpi rock shelter in the Calvert Range has occupation back to at least 2,300 bp (Veth et al. 2001) and there are indications that deeply weathered rock engravings in this range reﬂect an earlier occupation of this system. In a comparable environmental setting on the eastern fringe of the Western Desert, the Puritjarra shelter has an uninterrupted stratigraphic sequence with occupation over the last 32,000 years and with good evidence for occupation in the mid-to late Holocene. If we exclude sites other than those in the central part of the Western Desert we are left with few dated sites. Puntutjarpa (Main Cave) may have a broken sedimentary record. The upper occupation unit dates from ca. 400 bp and the lower about 7,000–6,000 bp. There may have been some use of the main rock shelter site at ca. 4,000 bp, but the presence of dense rock fall and a confused series of radiocarbon dates mean this cannot be determined with any certainty. At Puntutjarpa (West Cave) there is ﬁrmer evidence of occupation at this time – a hearth dated to ca. 3,800 bp. Winpuly, an open site situated 110 km east of Warburton, was occupied over the last 800 years, apparently with some occupation prior to this date (Gould 1982). 6 In the central Australian ranges, research by Thorley (1998a) has shown that there was also widespread occupation of the Palmer River catchment at 4,000–3,000 bp. The ﬁeld evidence suggests a discrete pulse of occupation rather than sustained use of sites at this time. However, the Palmer catchment is susceptible to ﬂuvial action which periodically reconﬁgures resource zones, and also affects sedimentary processes in these sites. As a result, many sites in the catchment show discontinuous stratigraphic and occupation sequences, suggesting that evidence from other parts of the ranges will be needed to characterize occupation at this time. 7 The last 1,000 years appears to have been a period of major change in the Western Desert and central Australia. There are indications of more sedentary occupation and signiﬁcant changes in levels of site use at about 1,000 bp (Smith 1988, 1996), possibly as part of a late Holocene demographic transition towards higher regional populations. A similar trend is evident in the Rudall River region of the Great Sandy Desert (Veth 1993). Phytolith data from Puritjarra show that at about 1,500 bp grasses stabilized at their highest level of representation in regional vegetation after a period of instability between 5,000–1,500 bp (see Bowdery 1998), suggesting that the modern summer rainfall regime became established shortly before these changes in settlement. Recent work in central Australia (e.g., Graham and Thorley 1996; Thorley and Gunn 1996) is beginning to reﬁne this picture, showing that even within this period there is evidence for cultural change at a range of geographic and time scales, particularly within the last 500–200 years. Other evidence indicates changes in rock art within the last 1,000 years (Layton 1992; Rosenfeld and Smith 2002).

_____________________________________________________________________________________________________

228

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

Implications for Linguistic Models The implication of these data for linguistic models is that the Western Desert and surrounding regions were occupied in some fashion prior to the postulated expansion of Wati languages 2,000 years ago. In western central Australia and in the Pilbara there have evidently been well established regional settlement systems since at least the terminal Pleistocene. Although archaeologists have explored the possibility of abandonment of parts of the desert during the LGM, it is unlikely that a signiﬁcant portion of the area now occupied by Wati languages was unoccupied in the late Holocene – immediately prior to Wati expansion. Even in the central parts of the Western Desert there is evidence for occupation prior to 2,000 bp, despite the fact that there has been very little ﬁeldwork in these areas, other than Gould’s early research. If Wati languages spread across the Western Desert through a combination of diffusion, language shift, and small scale population movements this could be readily accommodated within current archaeological interpretations. However, if the expansion of Wati languages involved migration – rather than language shift – it must have involved movement across territory occupied by other desert groups, and the displacement of those groups. So what evidence is there for actual population movements?

Population Movements in the Desert? Movements from the West? Is there any evidence for an incursion of Wati speaking groups into western central Australia about 1,500–1,000 bp, apart from that provided by historical linguistics? Major movements of Pintupi, Pitjantjatjara, and Yankunytjatjara people appear to have taken place in the ﬁrst few decades of the twentieth century as people migrated eastwards in response to severe drought conditions in their homelands (Goddard 1995; Myers 1986: 28–9; Tindale 1972). Today, Pintupi sometimes refer to themselves as yapurramalu kula, literally ‘‘he is from the west, indeed!’’ (see Myers 1986: 28 for further context). We need to ask whether these movements are a response, in some fashion or another, to the establishment of missions and ration depots in central Australia, or whether they reﬂect longer term historical processes. Some evidence indicates the latter. The Tingarri traditions, which involve mythological traveling groups of men and novitiates journeying west to east into the region (Myers 1986: 60–4), are suggestive of at least cultural diffusion from the west over a longer period. Similarities in rock engraving indicate more ancient west to east links (see below). However, the strongest evidence for population movement, rather than diffusion of Wati languages, is from genetics. The genetic evidence for movement of people in recent prehistory into western central Australia seems unequivocal. Birdsell (1993: Appendix F) drew attention to the sharp genetic divide coinciding with the linguistic boundary between Western

_____________________________________________________________________________________________________

229

_____________________________________________________________________________

M i k e S m it h

Desert and Arandic languages. Other genetic research has independently conﬁrmed the uniformity and distinctiveness of Western Desert groups (e.g., Balakrishnan et al. 1975; White 1997). This distinctiveness is particularly clear in serological data (where the highest frequencies of blood group A1 are in the Western Desert) and in the phenotype for tawny hair (which is characteristic of this population) (Birdsell 1993: ﬁgures B-1 and D-4). Given current rates of genetic exchange between the various groups, Birdsell estimated that the observed boundary between Arrernte and Western Desert groups could not have existed for more than 1,200–500 years. Birdsell’s view was that this had involved migration of the Arrernte into the region from an area south of the Gulf of Carpentaria, though in this view he was inﬂuenced by linguistic evidence for such a relationship (e.g., Wurm 1972). However, the remarkable genetic and linguistic homogeneity shown by the various Western Desert groups suggests it is more likely that it is this population that is a recent arrival in central Australia, rather than the Arrernte.

Archaeological Evidence Is there any archaeological evidence for an incursion of Wati speakers into central Australia? Exactly what the material correlates of such a movement would be is unclear, and the archaeology of hunter-gatherers only rarely affords the opportunity to address issues of ethnic identity directly (cf. Veth and McDonald 2002). If we take the obvious alternative route and attempt to correlate periods of major change in the archaeological record with the linguistic prehistory this inevitably takes us back to questions about the chronology proposed for Wati expansion. For instance, there are major changes at several points in the archaeological record: within the last 500 years; ca. 1,500–1,000 bp; 4,000–3,000 bp; and earlier at ca. 7,000 bp. Any one of these could be associated with population movements of the type speciﬁed by McConvell, but none inherently demand this sort of explanation. The Puritjarra sequence is an obvious starting point to review the archaeological evidence, as this site is strategically located, west of the MacDonnell Ranges, in an area that could be expected to have undergone a shift in cultural orientation with any movement of Wati speakers into western central Australia (cf. McConvell 1996: ﬁgures 4 and 5). 1

Recent changes in rock art and ochers. At Puritjarra, changes in the rock art – in particular in iconography and composition – indicate that geographic linkages towards the southeast weakened during the last millennium, and were replaced by stronger connections to the north (Rosenfeld and Smith in press). Other shifts in the rock art reﬂect changes in either the totemic referent of the site or the identity of the people occupying it. Changes in the provenance of red ocher reaching the site at this time also indicate a realignment of contacts, with a strengthening of links to the northwest of the Cleland Hills. Taken together, the evidence suggests a reorientation of cultural contacts at this rock shelter after ca. 1,000 bp, towards stronger connections with Western Desert and Walpiri groups to the north and northwest. While there is no strong evidence of

_____________________________________________________________________________________________________

230

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

population movement, these changes in cultural landscape may reﬂect the arrival of (new) Western Desert groups in the region. Thorley and Gunn (1996) summarize other evidence for changes in site use and rock art in western central Australia, within the last 500 years, that they suggest may indicate new or renewed associations with Western Desert groups. 2 Late Holocene changes in site use and population density. If the expansion of Wati languages involved population spread it must have been fueled to some extent by a growing population. It is relevant therefore that the best archaeological evidence for population growth – on both sides of the Western Desert – is dated ca. 1,500–1,000 bp. There are major changes in the archaeological record in central Australia ca. 1,000 bp, including a widespread increase in intensity of site use across the region, effectively registering a more sedentary pattern of occupation and possibly higher regional population densities than before (see Smith 1988, 1996 for details). However, as these changes also take place in parts of the central Australian ranges well away from the contact zone with Western Desert groups, there is no necessary connection with population movements from the west. It remains possible that both Wati expansion and higher regional populations are epiphenomena of some third factor, such as climate change. 3 Changes in technology. We might expect that the spread of desert adapted Wati speakers would be associated with the introduction of new items of material culture in these regions. At 4,000–3,000 bp a major change in tool kits occurs in central Australia with the appearance of an assemblage of small backed artifacts, hafted tula adzes, small semi-discoidal thumbnail scrapers, seed grinding implements, and (possibly) ground-edge axes (see Smith in press for description of the artifact sequence at Puritjarra). Some of these elements – such as hafted adzes and seed grinding implements – would be regarded as quintessential desert implements, raising the possibility that their appearance in central Australian assemblages marks the movement of Western Desert people into the region. If so this must have occurred somewhat earlier than the linguistic prehistory indicates. For other implement types – such as backed artifacts and thumbnail scrapers – current evidence indicates mainly a southern distribution (and southern origin). These implement types appear to have diffused across the southern half of the continent ca. 4,000–3,000 years ago, including the southern part of the Western Desert – though current research also shows that some backed artifacts occur in early Holocene contexts in eastern Australia (Hiscock and Attenbrow 1998). Ground-edge axes occur in late Pleistocene contexts in northern Australia and only spread across central and eastern Australia after 4,000 bp, but they appear to have been absent or rare in the southern sector of the Western Desert. These patterns of distribution cut across the geographic distribution of Wati languages and indicate there is no straightforward relationship between the new implement types and Wati expansion. 4 Population expansion in the early Holocene. Some interpretive problems posed by Wati expansion would be reduced if this coincided with initial colonization – or recolonization – of unoccupied desert lowlands in the post-glacial period. This process was probably in train by 13,000 bp but may not have been

_____________________________________________________________________________________________________

231

_____________________________________________________________________________

M i k e S m it h

completed until 7,000–6,000 bp. By ca. 7,000 bp there is evidence of substantial use of sites such as Puntutjarpa and Puritjarra, but Serpent’s Glen was not reoccupied until ca. 5,000 bp. If the latter indicates a widespread regional lag, then Wati expansion might correspond to the period when the western parts of the Western Desert are reoccupied. This would involve signiﬁcant recalibration of the linguistic prehistory. This illustrates a common difﬁculty: there is usually more than one plausible solution for matching up linguistic and archaeological sequences. Unless we can determine the relative position of technological, lexical, or other developments embedded in both sequences I doubt we can adequately discriminate between competing possibilities. Also, although existing archaeological evidence for this time period could be accommodated comfortably within the linguistic framework (e.g., migration ca. 1,500 bp), the archaeology alone would also support a number of other interpretations (e.g., cultural diffusion, demographic growth, new social networks). I would also be reluctant to rule out either longer (5,000–3,000 bp) or shorter chronologies (500 bp) for the spread of Western Desert people.

Movements from the East? Geneticists who have looked at population history in central Australia have also suggested that the Arrernte may be recent arrivals, from an area south of the Gulf of Carpentaria. This idea seems to have originated with Birdsell (1950) and was repeated in his 1993 magnum opus Microevolutionary Patterns in Aboriginal Australia. Balakrishnan et al. (1975) found that the Arrernte are genetically closer to some northern groups such as the Yolgnu (northeastern Arnhem Land) and people at Port Hedland, than to Western Desert groups. White (1997: 69) reached similar conclusions to Birdsell, but Cavalli-Sforza et al. (1994) found little evidence for migration. The evidence for migration of the Arrernte seems to rest on two observations: their genetic distance from neighboring Western Desert groups and their genetic similarity to populations to the north. The ﬁrst may simply reﬂect recent migration of Western Desert groups into central Australia (as discussed above). The second is difﬁcult to assess without some genetic data on populations north and east of the Arrernte. It may be simpler to see the Arrernte as one end of a genetic cline. White (1997) makes several important points here that indicate this is more likely than recent migration: 1 2

Any migration of the Arrernte must have occurred sufﬁciently long ago for skin pigmentation to adapt to local conditions. There is evidence for linear gene ﬂow into central Australia from the southwestern Cape York, which could result either from long-term intergroup marriages or from migration.

A further problem with the migration scenario is that it requires mass migration of people in sufﬁcient numbers to genetically or linguistically swamp an established

_____________________________________________________________________________________________________

232

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

population in the main central Australian ranges, one of the most densely populated parts of the arid zone. Given that any migration into the region would involve demographic expansion across less productive desert lowlands and semi-arid savanna, it seems unlikely that it would involve sufﬁcient numbers of people to physically displace an established population. My view is that contemporary Arrertne speaking people are most likely derived from late Pleistocene–early Holocene groups occupying the central Australian ranges. This gains some support from Cavalli-Sforza et al. (1994), who identify three major genetic trends in Australia: two are genetic clines across the continent, respectively from the northwest and northeast, while the third (Cavalli-Sforza et al. 1994: ﬁgure 7.6.2.C) delineates the central Australian ranges and is consistent with a long established population in this region.

Western Desert Origins Two main classes of archaeological data can be brought to bear on the question of Western Desert origins: stone tools and rock art (categories suggested as useful by Veth and McDonald 2002). Both illustrate the cultural links across the area occupied by Wati speakers, but neither yields a clear picture of likely origins.

Stone Tools Stone tool assemblages are not highly differentiated across the region and may therefore be of limited use in identifying donor areas. While central Australian assemblages have one or two types of implement, such as Pirri gravers and isolated ﬁnds of trimmed points, not found further west, these additional tool types are so rare in site assemblages – excavated or otherwise – as to negate their usefulness as markers. There are regional differences in the morphology and trimming of seed grinding implements which link these items in the Western Desert to those in the Pilbara and other areas in arid northwestern Australia, rather than to those in central Australia or western New South Wales, but this is an area that awaits further formal study. Backed artifacts in central Australia comprise mainly small geometric microliths made on ﬂakes (Smith 1988), whereas asymmetric backed artifacts made on long, thin blades are found in very low numbers in parts of the Western Desert (e.g., in the Rudall River area, Veth 1993: 16–19).

Rock Art Rock art on the western fringes of the Western Desert suggests early links with the Pilbara, replaced more recently by a new corpus of motifs with widespread afﬁnities across the region from the upper Gascoyne region to western central Australia. The most relevant recent analysis of continental patterning in Australian rock art is by Layton (1992). Using his classiﬁcation, recent rock paintings in the Durba

_____________________________________________________________________________________________________

233

_____________________________________________________________________________

M i k e S m it h

Hills, Calvert Range, and Great Sandy Desert are dominated by small silhouette motifs. All three of Layton’s ‘‘geometric’’ styles (Types A, B, and C) are also present. There are afﬁnities with rock art in parts of western central Australia with that in regions occupied by Wati speaking groups, where Type B and C geometric motifs and small silhouette motifs also occur. In the upper Gascoyne area, identiﬁed as the likely area of origin for Wati languages by McConvell (1996), Type A geometric motifs and small silhouettes are also common. In contrast, recent rock art in other parts of central Australia consists primarily of Type A geometric motifs (Panaramitee style), while in the Pilbara it is dominated by large silhouette motifs. Neither of these regions seems a likely origin for the corpus of recent rock art found in the Western Desert. However, this is not true of the earlier phase of rock art production, where regional linkages seem to have been quite different to those in the late Holocene. In the Calvert Range there are deeply weathered engravings which although undated must be of at least early Holocene age. The afﬁnities of the large silhouette and anthropomorphic ﬁgures represented in these engravings unequivocally point to connections with the Pilbara. The contrast between the older engraved motifs and the recent paintings suggests that signiﬁcant cultural changes took place in this region, although we presently do not have a clear picture of the chronology of these changes. The presence of engraved face motifs, similar to some of those in the Cleland Hills, 800 km to the east, also suggest movements – or at least strong contacts – from west to east across the Western Desert and into the western part of central Australia in the early Holocene or late Pleistocene (see Chapter 7, this volume). Although there is clear evidence of cultural links from western central Australia across the Western Desert, these data provide little indication of the ‘‘direction’’ of spread or ‘‘point of origin’’ for these traits. On present evidence we cannot rule out the Western Desert as a center of development for the Wati linguistic group – a conclusion which Veth (2000) independently reached. Recent archaeological work on chemically ﬁngerprinting red ochers from different sources (see Smith and Fankhauser 1996) may provide a means of directly testing some of these ideas. The famous red ocher mine of Wilgi Mia lies close to the putative homeland for Wati languages in the upper Gascoyne area. As Wati speakers occupied the Western Desert we could expect the new cultural networks that this dispersal created to facilitate widespread exchange of Wilgi Mia ocher across the region. Provided the linguistic prehistory has correctly identiﬁed the homeland of Wati languages we could expect to ﬁnd major changes in the distribution of this red ocher within the last 2,000 years.

Mechanisms We now need to return to the question of what processes would allow an expanding hunter-gatherer group to displace an established desert population over a wide area. If we accept the linguistic chronology, this replacement took

_____________________________________________________________________________________________________

234

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

place despite the fact that there was little difference in the technologies of the two groups: both used seed grinding implements and both had hafted adzes and other composite implements. If the desert was already occupied – and archaeological evidence indicates that it was – how then would Wati expansion have proceeded? One clue as to possible replacement mechanisms is provided by the inherent variability in these natural systems. It is clear that the ethnographic ‘‘pulse and reserve’’ model operated on a range of scales: as part of the annual cycle of population aggregation and dispersal in the Western Desert, as well as a response on a larger scale to cyclic drought. During major drought periods desert groups would ﬁssion and disperse in small groups and the bulk of the population would fall back on better watered country. It is not difﬁcult to see this process acting as a sort of ‘‘cultural pump,’’ drawing people into the desert during good seasons and forcing people out towards the margins of the region, or towards key waters within the desert, during drought periods. Tindale (1972: 219) provides a pertinent account of the process: During the almost country-wide drought of the years 1914–15, lack of water became so great a problem that the Pitjandjara were forced eastward. In desperation, and by reason of power of superior numbers, they drove the rightful inhabitants of the Eastern Musgrave ranges, the Jangkundjara, out of part of their territory, forcing them down Ofﬁcer Creek toward the southern limits of their territory, in the Everard Ranges . . . When the rains came again and the pressure was relieved, not all of the Pitjandjara returned within their former bounds. Instead, they remained in the Musgraves, usurping the country as far east as Ernabella.

Any systematic west to east differences in drought occurrence would accelerate the eastward movement of human populations. Such a scenario is obviously speculative, but it would account for the ability of Wati speakers to displace other hunter-gatherer groups in some relatively well watered ranges and uplands in western central Australia (e.g., George Gill Range, Musgrave Ranges). A corollary of the hypothesis is that the process must have operated throughout the Holocene. Wati expansion may then be simply the most recent example of a longer historical process of people being drawn into the desert and then forced out to its margins. The similarity of deeply weathered engravings on either side of the Western Desert (see above) provides suggestive evidence for earlier west to east movements across the desert.

Stratigraphic Correlation If the chronology for Wati expansion could be reliably calibrated, the linguistic prehistory would throw the archaeological changes, described above, into much sharper focus. There is great potential here for a ﬁner-grained reconstruction of recent desert prehistory than either discipline can deliver in isolation. But how do we establish some unequivocal points of connection between the linguistic stratigraphy and archaeological sequences in this region?

_____________________________________________________________________________________________________

235

_____________________________________________________________________________

M i k e S m it h

Any correlations will need to be more concrete than the arbitrary juxtaposition of linguistic and archaeological reconstructions attempted above. There are well known problems with the latter approach. First, there is the danger of constructing a circular argument, particularly as there are strong incentives for archaeologists to take the richly detailed linguistic ﬁndings into account in their interpretations of the more spartan archaeological record. Secondly, if we uncritically seek to align major points of change in the two sequences (e.g., the ‘‘Small Tool Tradition’’ with ‘‘Common Australian’’) we risk creating spurious event horizons. In doing this we would also lose the opportunity to explore more complex social processes shaping the prehistory of arid Australia, especially if the major linguistic and archaeological changes were in reality out of phase. McConvell (1996) has argued that there is a notional linguistic stratigraphy, produced by the overprinting of language divergence by speciﬁc semantic shifts, sound changes, and word borrowing. This concept is of fundamental importance for developing correlations between linguistic and archaeological prehistories. It opens the possibility of true stratigraphic correlations between the two sequences based on the relative position of technological or other developments embedded in both sequences. Ideally, these ‘‘benchmarks’’ will provide a dated terminus post quem or terminus ante quem for a given linguistic event. I point to several potential examples below, but the intense interdisciplinary conversations that this work requires have barely begun (for a rare example, see Evans and Jones 1997). As a preliminary observation I suggest there are two marker horizons – one at 1,000 bp, the other at 4,000–3,000 bp – which provide chronological benchmarks for linguistic prehistory in the Australian desert. What such ‘‘lexical strata’’ might consist of is described below.

Lexical Stratum at 1,000 bp Changes in the archaeological record at ca. 1,000 bp in central Australia reﬂect higher regional population densities, possibly accompanied by a trend away from open social networks towards greater emphasis on maintaining boundaries. What are the likely linguistic correlates of these changes? One possibility might be accelerated linguistic change. Rapid population growth would reduce the incentive to deal with people outside a particular cultural bloc, producing similar linguistic effects to isolation. If so, the archaeological changes may ﬁnd some linguistic expression in the series of radical sound changes (dropping ﬁrst the initial consonant, then the remainder of the initial syllable) which have transformed Arandic languages relatively recently, making them appear aberrant as PamaNyungan (PN) languages (Koch 1997). In Arandic, linguists should look for corresponding semantic changes in kinship terms, changes in terms for strangers, in the terminology for camp, country, heartland, or clan estate, or in the nomenclature associated with the ownership of sites or ceremonies (e.g., for ritual ‘‘manager’’ and ‘‘owner,’’ respectively kwertengerl and arrngertelengkwek-artwey in Alyawarr). If the age of the Arandic sound changes can be determined, the

_____________________________________________________________________________________________________

236

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

next step will be to ask whether or not there are any loan words from the Western Desert language into Arrernte that predate these changes.

Lexical Stratum at 4,000–3,000 bp The changes in stone tool technology at about 4,000–3,000 years ago essentially reﬂect a shift towards greater use of hafted composite implements and the use of thermoplastics such as Triodia and Plechtrachne resin as hafting cements. Many of the new tool types were made on higher quality raw materials such as cherts and chalcedony. Somewhere in the linguistic stratigraphy, therefore, there should be a series of semantic changes, or new terminology, relating to resins, hafts, composite implements, woodworking tools, and new stone materials. These changes might also represent a fundamental shift towards identiﬁable men’s and women’s tool kits (along the lines suggested by Hamilton 1980), with women continuing to use handheld, amorphous ﬂake implements of quartzite and silcrete and men appropriating the new technology. If such a linguistic stratum exists, it should predate Wati expansion but postdate initial break up of proto-PN. Another point of connection concerns implements used to process seeds. The most easily recognized in archaeological contexts are the large grooved millstones and small faceted topstones used for wet grinding grass and acacia seeds. These appear in archaeological sequences in the Western Desert and in central Australia around 4,000–3,000 bp. Some seed grinders appear much earlier at approximately 30,000 bp at Cuddie Springs in western New South Wales. There is an active debate about whether this means we have missed them elsewhere in the arid zone (Gorecki et al. 1997). Even if this is shown to be the case, the fact still remains that such implements do not become common across the arid zone until 4,000–3,000 bp. Irrespective of the uncertainties surrounding the earliest dating of this technology, 4,000–3,000 bp can be regarded as a useful minimum age. Somewhere in the linguistic stratigraphy there should be a stratum with a series of linguistic changes relating to grindstones and millstones, grass seeds and other seed foods, other seed processing gear and processes (winnowing and husking, parching grain), and terms for key grass species. It would be useful here to look at terms used for edible seeds as a major food group (such as ntange in Arrernte). The introduction of the dingo (Canis lupus dingo) to Australia ca. 3,500 bp (Corbett 1995) provides another potential chronological benchmark. There is little doubt that dingos would have spread rapidly across the continent and that Aboriginal people would have had to ﬁnd a term in their lexicon for what was an entirely new animal. There would presumably have been many different semantic and lexical solutions to this problem, resulting in a great variety of forms across the continent. The relative position of early words for dog in the linguistic stratigraphy should therefore provide a useful temporal marker as well as point to the origins of the Western Desert language. The last three of these markers should occupy the same relative position in the linguistic stratigraphy. Collectively, they allow an internal test of the chronology proposed for Wati expansion, at least to the extent that it should be possible to

_____________________________________________________________________________________________________

237

_____________________________________________________________________________

M i k e S m it h

determine whether the differentiation of Wati occurred before or after 3,000 bp. It is important that we attempt such tests before archaeologists and linguists begin to borrow heavily from each other’s interpretations of desert prehistory.

Conclusions Desert Prehistory A rapid colonization of the Western Desert within the last few thousand years, by a small founding population drawn from existing desert groups, is indicated by the remarkable cultural, linguistic, and genetic uniformity of the region. The problem is to determine just when this took place. Genetic data provide the strongest evidence for recent spread of people across the region, but do not provide ﬁrm age estimates for this, and merits further work by geneticists. In particular, Birdsell’s estimate of less than 1,200–500 years for gene ﬂow between Western Desert and Arandic populations urgently needs critical reassessment before it becomes the cornerstone of an interdisciplinary prehistory. Linguistic data provide evidence for the recent expansion of the Wati language group across the region, but leave open questions about both chronology and the mechanism for this language spread. Archaeological sequences for the Western Desert and central Australia provide only indirect evidence of these developments and offer a number of alternative possibilities for their timing within the last 7,000 years. Changes in the archaeological record ca. 1,000 bp reﬂect population growth in the region and, together with evidence for realignment of cultural linkages at this time, are likely to be connected with the spread of Western Desert populations. However, it is premature to rule out longer or shorter chronologies for this. For instance, we could plausibly correlate Wati expansion with reoccupation of desert sites such as Serpent’s Glen at ca. 5,000 bp and the subsequent spread of new types of implements at 4,000–3,000 bp. Or we might correlate Wati expansion with changes in the archaeological record around 500 bp. In such a view the arrival of (new) Western Desert people in central Australia might be one of a number of cultural, demographic, and environmental changes set in motion during the last millennium. If Wati expansion took place <5,000 bp it must have taken place across country already occupied by people. I have suggested that the high environmental variability of the Western Desert provides a plausible mechanism for this, periodically creating an ebb and ﬂow of desert populations. It is more likely to have been a series of small scale population movements rather than a major migration of people into the desert. And if so, it may well be a process that has operated throughout the history of this region: Wati may only be the latest of a series of population and cultural movements to unify the desert.

Methodological Issues In this chapter my major concern has been to argue that we need a better means of calibrating the chronology proposed for linguistic prehistory. Reconciling the

_____________________________________________________________________________________________________

238

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els

different perspectives provided by archaeology and historical linguistics requires a deep analytical engagement between the two disciplines that focuses on ways of correlating the chronologies provided by the former and the detailed culture history of the latter. Any correlation will need to be more concrete than an arbitrary juxtaposition of linguistic and archaeological sequences or the linguistic ‘‘paleontology’’ used to reconstruct putative homelands. The concept of linguistic stratigraphy is of fundamental importance for developing these correlations. The essential thing is to look for a stratum of lexical and semantic changes associated with economic, social, demographic, and technological changes identiﬁed in the archaeological record. It will not be sufﬁcient just to track terms for particular artifact types (e.g., seed grinders, backed artifacts, bifacial points, ground-edge axes, etc.) – we need to look for a ‘‘package’’ of lexical and semantic changes. There is a danger that the detailed picture provided by historical linguistics will simply become the dominant interpretive framework for desert prehistory – a new culture historical paradigm for the archaeology of the region. This would be to miss the opportunities that an interdisciplinary prehistory opens up to study not just culture history but also culture process. If this occurs, we would lose the opportunity to explore more complex social processes shaping the prehistory of arid Australia, especially if major linguistic and archaeological changes were in reality out of phase.

References Balakrishnan, V., Sanghvi, L. D., and Kirk, R. L. 1975: Genetic Diversity Among Australian Aborigines. Canberra: Australian Institute of Aboriginal Studies. Bettinger, R. L. and Baumhoff, M. A. 1982: The Numic spread: Great Basin cultures in competition. American Antiquity, 47, 485–503. Birdsell, J. B. 1950: Some implications of the genetic concept of race in terms of spatial analysis. Origin and Evolution of Man, Cold Spring Harbor Symposia on Quantitative Biology, 15, 259–314. Birdsell, J. B. 1993: Microevolutionary Patterns in Aboriginal Australia: A Gradient Analysis of Clines. Oxford: Oxford University Press. Bordes, F., Dortch, C., Thibault, C., Raynal, J.-P., and Bindon, P. 1983: Walga Rock and Billibilong Spring: Two Aboriginal archaeological sequences from the Murchison Basin, Western Australia. Australian Archaeology, 17, 1–26. Bowdery, D. E. 1998: Phytolith Analysis Applied to Pleistocene–Holocene Archaeological Sites in the Australian Arid Zone. Oxford: British Archaeological Reports Series no. 695, Hadrian Books. Cavalli-Sforza, L. L., Menozzi, P., and Piazza, A. 1994: The History and Geography of Human Genes. Princeton, NJ: Princeton University Press. Corbett, L. 1995: The Dingo in Australia and Asia. Sydney: University of New South Wales Press. Evans, N. and Jones, R. 1997: The cradle of the Pama-Nyungans: Archaeological and linguistic speculations. In P. McConvell and N. Evans (eds), Archaeology and Linguistics: Aboriginal Australia in Global Perspective, Oxford: Oxford University Press, 385–417.

_____________________________________________________________________________________________________

239

_____________________________________________________________________________

M i k e S m it h Goddard, C. 1995: Appendix 2: Comments on the Map by regional specialists. In P. Sutton (ed.), Country: Aboriginal Boundaries and Land Ownership in Australia, Canberra: Aboriginal History Monographs no. 3, 153. Gorecki, P., Grant, M., O’Connor, S., and Veth, P. 1997: The morphology, function, and antiquity of Australian grinding implements. Archaeology in Oceania, 32, 141–50. Gould, R. A. 1977: Puntutjarpa Rockshelter and the Australian Desert Culture. New York: Anthropological Papers of the American Museum of Natural History no. 54, American Museum of Natural History. Gould, R. A. 1982: ‘‘Indoor’’ vs ‘‘outdoor’’ living: Preliminary comparison of habitation residues in stratiﬁed open air and rock shelter sites in the Australian desert. In J. Siikala (ed.), Oceanic Studies: Essays in Honour of Aarne A. Koskinen, Helsinki: Finnish Anthropological Society, 241–52. Graham, R. and Thorley, P. 1996: Central Australian Aboriginal stone knives: Their cultural signiﬁcance, manufacture, and trade. In S. R. Morton and D. J. Mulvaney (eds), Exploring Central Australia: Society, the Environment, and the 1894 Horn Expedition, Chipping Norton: Surrey Beatty and Sons, 74–89. Grayson, D. K. 1993: The Desert’s Past: A Natural Prehistory of the Great Basin. Washington, DC: Smithsonian Institution Press. Gunn, R. G. 1995: Regional patterning in the Aboriginal rock art of central Australia: A preliminary report. Rock Art Research, 12, 117–28. Hamilton, A. 1980. Dual social systems: Technology, labor, and women’s secret rites in the eastern Western Desert of Australia. Oceania, 51, 4–19. Hiscock, P. and Attenbrow, V. 1998: Early Holocene backed artifacts from Australia. Archaeology in Oceania, 33, 49–62. Hiscock, P. and Veth, P. 1991: Change in the Australian desert culture: A reanalysis of tulas from Puntutjarpa. World Archaeology, 22, 332–45. Koch, H. 1997. Pama-Nyungan reﬂexes in the Arandic languages. In D. Tryon and M. Walsh (eds), Boundary Rider: Essays in Honour of Geoffrey O’Grady, Canberra: Paciﬁc Linguistics no. C-136, 271–302. Lamb, S. 1958: Linguistic prehistory in the Great Basin. International Journal of American Linguistics, 24, 95–100. Layton, R. 1992: Australian Rock Art: A New Synthesis. Melbourne: Cambridge University Press. McConvell, P. 1990: The linguistic prehistory of Australia: Opportunities for dialogue with archaeology. Australian Archaeology, 31, 3–27. McConvell, P. 1996: Backtracking to Babel: The chronology of Pama-Nyungan expansion in Australia. Archaeology in Oceania, 31, 125–44. McConvell, P. and Evans, N. 1997: Archaeology and Linguistics: Aboriginal Australia in Global Perspective. Oxford: Oxford University Press. Madsen, D. B. and Rhode, D. 1994: Across the West: Human Population Movement and the Expansion of the Numa. Salt Lake City: University of Utah Press. Myers, F. R. 1986: Pintupi Country, Pintupi Self: Sentiment, Place, and Politics among Western Desert Aborigines. Washington, DC: Smithsonian Institution Press. O’Connor, S., Veth, P., and Campbell, C. 1998: Serpent’s Glen rock shelter: Report of the ﬁrst Pleistocene-aged occupation sequence from the Western Desert. Australian Archaeology, 46, 12–22. Rosenfeld, A. and Smith, M. A. 2002: Rock art and the history of Puritjarra rock shelter, Cleland Hills, central Australia. Proceedings of the Prehistoric Society 68, 103–24.

_____________________________________________________________________________________________________

240

_____________________________________________________________________________

W e s t er n D e s e r t Ar c h a e o l o g y a n d L i n g u i s t i c M od els Smith, M. A. 1988: The pattern and timing of prehistoric settlement in central Australia. Unpublished PhD thesis, Armidale: University of New England. Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Smith, M. A. 1993: Biogeography, human ecology, and prehistory in the sandridge deserts. Australian Archaeology, 37, 35–50. Smith, M. A. 1996: Prehistory and human ecology in central Australia: An archaeological perspective. In S. R. Morton and D. J. Mulvaney (eds), Exploring Central Australia: Society, Environment, and the 1894 Horn Expedition, Chipping Norton: Surry Beatty and Sons, 61–73. Smith, M. A. in press: Characterizing late Pleistocene and Holocene stone artifact assemblages from Puritjarra rock shelter: A long sequence from the Australian desert. Records of the Australian Museum. Smith, M. A. and Clark, P. M. 1993: Radiocarbon dates for prehistoric occupation of the Simpson Desert. Records of the South Australian Museum, 26, 121–7. Smith, M. A. and Fankhauser, B. 1996: An archaeological perspective on the geochemistry of Australian red ochre deposits: Prospects for ﬁngerprinting major sources. Canberra: Unpublished report to the Australian Institute of Aboriginal and Torres Strait Islander Studies. Smith, M. A., Fankhauser, B., and Jercher, M. 1998: The changing provenance of red ochre at Puritjarra rock shelter, central Australia: Late Pleistocene to present. Proceedings of the Prehistoric Society, 64, 275–92. Smith, M. A., Prescott, J. R., and Head, M. J. 1997: Comparison of 14C and luminescence chronologies at Puritjarra rock shelter, central Australia. Quaternary Science Reviews, 16, 1–22. Smith, M. A., Vellen, L., and Pask, J. 1995: Vegetation history from archaeological charcoals in central Australia: The late Quaternary record from Puritjarra rock shelter. Vegetation History and Archaeobotany, 4, 171–7. Thorley, P. B. 1998a: Shifting location, shifting scale: A regional landscape approach to the prehistoric archaeology of the Palmer River catchment, central Australia. Unpublished PhD thesis, Darwin: Northern Territory University. Thorley, P. B. 1998b: Pleistocene settlement in the Australian arid zone: Occupation of an inland riverine landscape in the central Australian ranges. Antiquity, 72, 34–45. Thorley, P. B. and Gunn, R. 1996: Archaeological research from the eastern border lands of the Western Desert. Canberra: Unpublished paper presented at the Western Desert Origins workshop, July 1996, Australian Linguistic Institute. Tindale, N. B. 1972: The Pitjandjara. In M. G. Bicchieri (ed.), Hunters and Gatherers Today: A Socio-Economic Study of Eleven Such Cultures in the Twentieth Century, New York: Holt, Rinehart, and Winston, 217–68. Tonkinson. R. 1974: The Jigalong Mob: Aboriginal Victors of the Desert Crusade. Menlo Park: Cummings. Tonkinson. R. 1978: The Martu Aborigines: Living the Dream in Australia’s Desert. Fort Worth, TX: Holt, Rinehart, and Winston. Veth, P. 1989: Islands in the Interior: A model for the colonization of Australia’s arid zone. Archaeology in Oceania, 24, 81–92. Veth, P. 1993: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monograph in Prehistory, Archaeology Series no. 3, Ann Arbor, MI.

_____________________________________________________________________________________________________

241

_____________________________________________________________________________

M i k e S m it h Veth, P. 1995: Aridity and settlement in northwest Australia. Antiquity, 69, 733–46. Veth, P. 2000: Origins of the Western Desert language: Convergence in linguistic and archaeological space and time models. Archaeology in Oceania, 35, 11–19. Veth, P. and McDonald, J. 2002: Can archaeology be used to address the principle of exclusive possession in native title? In R. Harrison and C. Williamson (eds), After Captain Cook: The Archaeology of the Recent Indigenous Past in Australia, Sydney: Sydney University Archaeological Methods Series, no. 8, 121–9. Veth, P., Smith, M. A., and Haley, M. 2001: Kaalpi: The archaeology of an outlying range in the duneﬁelds of the Western Desert. Australian Archaeology, 56, 9–17. White, N. 1997: Genes, languages, and landscapes in Australia. In P. McConvell and N. Evans (eds.), Archaeology and Linguistics: Aboriginal Australia in Global Perspective, Oxford: Oxford University Press, 45–81. Wurm, S. A. 1972: Languages of Australia and Tasmania. The Hague: Mouton.

_____________________________________________________________________________________________________

242

_____________________________________________________________________________

-----------------------------------------------------------------------------

13

--------------------------------------------------------------------------

People of the Coastal Atacama Desert Living Between Sand Dunes and Waves of the Paciﬁc Ocean Calogero M. Santoro, Bernardo T. Arriaza, Vivien G. Standen, and Pablo A. Marquet

Introduction When Europeans colonized the coast of the Atacama Desert region (188–278S) in the sixteenth century they described the local people as skillful ﬁshermen. In particular, Bibar (1966 [1565]: 12) observed: Those who kill sea lions do not kill other ﬁsh, as we have said, and those who kill toninas [dolphin] do it routinely. In this way, each category of ﬁsherman kills the type of ﬁsh of his specialty and no other; and when they die, it is custom to place the skulls [of marine mammals] and all the ﬁshing tools such as nets, harpoons, and ﬁshhooks on top of the grave.

This specialized labor force was not linked to complex forms of social organization as depicted in prehistoric cultures of the coasts of central Peru and northwestern North America. For this reason, the Europeans were not able to appreciate the cultural richness of the native populations that were settled along the coast of the Atacama Desert. For the Spaniards, these Indians were dirty and poor, with a simple way of life and modest possessions (Lozano-Machuca 1885 [1581]); thus, the sixteenth century Hispanic records provide no information relating to other cultural traits of this Amerindian population, such as funerary procedures, body painting, or tool decoration, nor about their geoglyphs and petroglyphs. The coastal way of life has deep roots in the cultural history of the American continent, as it seems to be connected with the ﬁrst people who immigrated into South America, late in the Pleistocene epoch. Around 11,000–10,000 years ago,

_____________________________________________________________________________________________________

243

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

people focused on the coast of the Atacama territories and over several millennia they gradually transformed their culture. They created efﬁcient technologies and strategies to maximize the subsistence possibilities provided by the Paciﬁc Ocean and the desert coast (Llagostera 1992; Nu´n˜ez 1983; Rivera and Rothhammer 1991; Standen et al. in press). This allowed them to reduce risks of the uncertainties typical of a desert environment as they lived between diverse habitats within a large but narrow territory along the coast. Despite the fact that the hunter-gatherers and ﬁshermen of the coastal Atacama Desert inhabited a landscape with limited terrestrial resources, their descendants were able to create and develop a complex cultural system, with rather egalitarian political structures. However, the most fascinating creation of these early people was the development of sophisticated systems of artiﬁcial mummiﬁcation, the ideological and social principles of which were arguably crucial for the maintenance of social ties. As we show below, artiﬁcial mummiﬁcation, and its unknown cosmological world, was the most dynamic feature of the cultural history of these coastal desert people (Arriaza and Standen 2002). Lying adjacent to the littoral of the Paciﬁc Ocean in western South America, the Atacama Desert has a very limited range of edible plants and herded animals. This contrasts sharply with the Australian deserts or the Kalahari and Namib Deserts of southern Africa, which offer many (albeit patchy) terrestrial food resources for hunter-gatherers. These larger desert territories, however, lack the abundance of water seen in the Atacama in the form of permanent or seasonal streams originating in the high Andes, located around 100–150 km to the east (see Figure 13.1). This supporting network of water sources and the Paciﬁc Ocean, one of the richest and most diverse marine biomasses in the world, creates unusual conditions allowing for the permanent occupation of this area by hunter-gatherers and ﬁshermen. The ﬁrst known inhabitants who arrived on the Paciﬁc coastline of this region initiated a subsistence system that after 2,000–3,000 years of experimentation became strongly dependent on a wide variety of marine resources, hidden under the waves of the Paciﬁc (Bird 1943; Llagostera 1992; Nu´n˜ez 1983; Schiappacasse and Niemeyer 1984). They complemented their maritime economy with unpredictable and dispersed terrestrial resources, located along the quebradas (narrow deep canyons with permanent or intermittent courses of water) and oases touched by sand dunes swept by the desert winds. The Selknam, hunter-gatherers of Tierra del Fuego, can serve as an analogy for understanding early Atacama populations. They inhabited an extremely harsh environment, but also developed a complex ideological system expressed in sophisticated ceremonies, which included body painting, masks, and performances representing supernatural characters. These rituals helped them to maintain and pass on their ideology and promoted social and biological reproduction. Similarly, Australian desert Aborigines continue to exercise complex systems of thinking and performance through a rich iconographic rock art and curation of tjuringa (sacred boards with painted or incised motif), comprising ‘‘a semiotic system that mediates social and territorial relations’’ (Rosenfeld and Smith 2002: 112), and body painting with the ‘‘appearance of ancestral forces’’ (Morphy 1998: 23). In these

_____________________________________________________________________________________________________

244

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

P E R U 1 Ilo

Northern Atacama

2

3 4 Artica

5

6 7

I L

Pacific Ocean

H

9 Central Atacama

C

Iquique

8

E 10

Southern Atacama

Copiapó

Figure 13.1 Relief map of the Atacama Desert obtained by interferometric C-band radar from the space shuttle Endeavor (NASA/JPL/NIMA) showing principal hunter-gatherer archaeological sites discussed in the text: (1) Quebrada Jaguay; (2) Ring Site; (3) Quebrada Tacahuay; (4) Los Burros; (5) Acha and Chinchorro sites; (6) Camarones; (7) Punta Pichalo; (8) Tiliviche and Arago´n; (9) Patillo; (10) Las Conchas.

cases we are dealing with highly abstract thought processes and complex aesthetic expression which appears to have originated in the late Pleistocene (cf. Mulvaney and Kamminga 1999). In this chapter we present an analytical and comparative description of the origin and cultural evolution of people in the Atacama Desert, bringing together

_____________________________________________________________________________________________________

245

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

varied ecological and archaeological data to discuss the principal issues of the natural and cultural history of the region, from the late Pleistocene to the late Holocene (ca. 12,000–4,000 bp).

The Ecological Setting of the Atacama Desert The Atacama/Peruvian Coastal Desert is one of the driest places on earth. Hyperaridity has been present in the central Atacama for at least the last 15 million years, although recent studies have documented environmental changes (Betancourt et al. 2000; Grosjean and Nu´n˜ez 1994; Latorre et al. 2002). These paleoecological changes greatly affected the distribution of animals and plants as well as that of past human populations in this region. This desertic landscape, located on the west coast of South America, extends for more than 3,000 km along a narrow strip from northern Peru (Lat. 58S) to northern Chile (Lat. 278S). It owes its existence to the drying effect of the cold northward ﬂowing Humboldt Current, the existence of air masses associated with the sub-tropical high known as the South Paciﬁc Anticyclone, and to the rain-shadow effect of the Andean mountain ranges, which impedes the penetration of moisture carried by the eastern trade winds. Although this desert is continuous from Peru to Chile, it is usually broken into two main components. The Peruvian Coastal Desert extends from Tumbes (ca. 58S) to Tacna in southern Peru (ca. 188S), and the Atacama Desert from the area of Arica in northern Chile (ca. 188S) to Copiapo´ (ca. 278S) (Marquet et al. 2002; Rauh 1985). The northern part of the Peruvian Coastal Desert consists of a wide coastal plain with shifting sands known as the Sechura Desert, some 100–150 km wide. During the terminal Pleistocene and Holocene this coastal desert provided more favorable living conditions in the form of spring water and shallow lagoons with fresh water, as documented in Pampa de los Fo´siles in the Cupisnique Desert, of northern Peru. Further south in the Peruvian Coastal Desert, thick cloudbanks form over the desert lands as a result of the cooling effect of the Humboldt Current, as similarly occurs with the cold Bengela Current in the Namib Desert. When intercepted by isolated mountains or steep coastal slopes, this cool moist air gives rise to a fog zone known as garu´a in Peru and camanchaca in Chile. This moisture allows for the development of isolated and diverse vegetation formations called lomas (small hills), where plants are adapted to condense this humidity. The lomas, comprising communities of annual and perennial plants (Herbaceae and Gramineae) and cacti (Dillon 1997; Rauh 1985), grow vigorously during a short period in the winter of the Southern Hemisphere, and attract a wide variety of animals and birds (camelids, rodents, foxes). These ephemeral ‘‘fog oases’’ (Lavalle´e 2000) were occupied seasonally by hunter-gatherers whose base camps were located on the nearby coast (Benfer 1999; Quilter 1991). The Atacama Desert, located between 188S and 278S, can be divided in three sections: north, central, and south (see Figure 13.1). The latter, being the driest, has negligible archaeological data and thus is not described herein. Flanked to the east by the high Coastal Cordillera, the littoral offers discrete bays and beaches for

_____________________________________________________________________________________________________

246

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

human habitation, experiencing a mild climate (see Figure 13.2). In contrast to other Southern Hemisphere deserts (such as in Australia and southern Africa) average yearly precipitation is near zero. Thus the supply of fresh water depends on rain events outside the desert caused by convective air masses that cross the Andean crest bringing moisture laden air across the altiplano from the Amazon Basin. This phenomenon of the austral summer known as the invierno boliviano (the Bolivian or altiplanic winter) is responsible for the radical ﬂuctuation of superﬁcial runoff and groundwater that typically ﬂows from the western slopes of the Andes to the Paciﬁc. In the northern Atacama the runoff creates narrow and deep quebradas separated by 20–30 km of barren terrain with no vegetation at all. As rainfall gradually declines toward the south the quebradas do not reach the ocean, but rather discharge into inland basins, such as the Pampa del Tamarugal and Salar de Atacama in the central Atacama. The southern Atacama (248–278S) is a territory with no human habitation until recently, known as the despoblado (depopulate) de Atacama. In sum, although terrestrial biomass production is scanty and sparse, the existing water network, a distinctive feature of the Atacama Desert, offers a predictable resource for hunter-gatherers and ﬁshermen.

Figure 13.2 Typical landscape in the coast of the Atacama Desert ﬂanked to the east by the high Coastal Cordillera, leaving discrete bays and beaches for human habitation, with a rather mild climate. The arrow shows shell middens of a permanent Holocene camp, south of Camarones Valley.

_____________________________________________________________________________________________________

247

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

In contrast to limited terrestrial resources, the ocean offers a great variety and quantity of ﬁsh, molluscs, sea mammals, sea birds, and algae. This high productivity is triggered by a series of oceanographic, geographic, and climatic factors that produce very high levels of biomass (Llagostera 1992; Nu´n˜ez 1983; Schiappacasse and Niemeyer 1984; Yesner 1980) crucial to the establishment of relatively permanent human occupations along the coast of Peru and Chile appearing in the late Pleistocene (Lavalle´e 2000; Llagostera 1992; Nu´n˜ez 1983; Sandweiss et al. 1998, 1999; Schiappacasse and Niemeyer 1984). The Atacama Desert would have had a completely different cultural history if human populations had depended exclusively on terrestrial resources, as did many groups of the arid zone of interior Australia and the Kalahari (Hank 2002a). The combination of coastal and terrestrial resources is also observed in the Namib Desert, though the latter contains a much larger terrestrial biomass in terms of edible plants and animals (Hank 2002b). The northern section, known as the Fertile Coast, is characterized by ﬁve valleys or quebradas that traverse the desert (see Figures 13.1 and 13.3). In the Peruvian Desert to the north, more than 50 wider valleys reach the coast (Lavalle´e 2000). At the mouth and upstream of these valleys one ﬁnds localized habitats in the form of oases, which host fruit-bearing plants such as Prosopis (a mesquite-like tree found

Figure 13.3 Typical valley mouth in the northern Atacama Desert. Rich habitat for huntergatherers and ﬁshermen, who enjoyed marine and oasis resources and permanent, though brackish, water (a key feature). The shell midden in the foreground of the picture corresponds to Camerones 14, a permanent early Holocene camp.

_____________________________________________________________________________________________________

248

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

in patches), small game (such as Ctenomys, a gopher-like fosorial animal), and camelids (guanaco – Lama guanicoe), abundant sea birds, small quantities of freshwater ﬁsh and crayﬁsh, roots of reed and tomatillo (Lycopersicon chilense), and potable water. People also appear to have moved upstream to look for lithic raw materials, a resource which is noticeably absent on the coast. Conditions in the terminal Pleistocene appear to have been more lacustral than today, with ﬂuctuations of a magnitude that are seen to have had varying effects on human populations – an issue which is still under vigorous debate (Grosjean and Nu´n˜ez 1994; Latorre et al. in press; Nu´n˜ez et al. 2002; Usselmann et al. 1999). The central Atacama lacks the coastal oases, with fresh water restricted to brackish springs. The landscape is essentially barren and vegetation is restricted to lomas, the vegetation depending on the coastal fog which is reduced in extent along the coast and the Pampa del Tamarugal. This is a large inland basin connected to 15 to 20 quebradas, with Aroma, Tarapaca´, Chacarilla, Guatacondo, and Manı´ being the most important (Luis Briones pers. comm.), and a series of salares (ancient dry lakes) located behind the Coastal Cordillera. These interior basins experienced wetter conditions by the end of the Pleistocene and also possibly the Holocene (Latorre et al. in press).

The Peopling of the Region and Possible Migratory Routes Three human migratory routes have been proposed for the colonization of this desert zone by the end of the Pleistocene, linked to colonists entering South America through the Isthmus of Panama (ca. 13,000–12,000 bp). In the ﬁrst route, people are argued to have moved from the Amazonian Basin to the cold territories of the high Andes. On the coast of the Atacama Desert these late Pleistocene immigrants are argued to have become hunter-gatherers and ﬁshermen. Recent mitochondrial DNA analyses from late Holocene human remains of the Atacama (ca. 4,000 bp) illustrate a strong genetic link with current tropical forest populations (Moraga et al. 2002; Rivera and Rothhammer 1986, 1991). This does not necessarily mean, however, that all cultural traits found in the Paciﬁc were imported from the Amazon Basin. The second migratory route assumes that specialized hunter-gatherers, with sophisticated lithic industries, colonized the Andes from northern Colombia, next to the Isthmus of Panama (Rothhammer et al. 1984). According to Grosjean and Nu´n˜ez (1994), during the mid-Holocene a period of drought affected the highlands of the central Atacama Desert, causing the shrinking of lakes and the reduction of biotic resources with a subsequent loss of hunting opportunities (but see Latorre et al. 2002; in press, for a different paleoecological perspective). Settlements are seen to have been abandoned, creating a silencio arqueolo´gico (‘‘archaeological silence’’) as residential groups became more mobile or people concentrated in local refuges or moved to the coast (Nu´n˜ez 1983). Neither of these two hypothesized migratory scenarios is currently supported, as recent archaeological data show that the ﬁrst permanent settlements occurred

_____________________________________________________________________________________________________

249

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

slightly earlier on the coast (ca. 11,000 bp; Sandweiss et al. 1998) than within the highlands, where occupation is dated to ca. 10,800–10,600 bp (Nu´n˜ez et al. 2002). A third route proposes a migration along the coast from North America, with a population carrying an unspecialized lithic industry adjusted to coastal and marine environments (Llagostera 1992). Lavalle´e (2000) suggests that people used watercraft to cross the Paciﬁc Ocean from Siberia to Alaska by the end of the Pleistocene. This is plausible given that colonists of Australia used watercraft to cross a similar distance from Southeast Asia along the Wallacean Islands some 40,000 years earlier than the ﬁrst Americans (cf. Mulvaney and Kamminga 1999). Although this is an interesting and challenging hypothesis, it requires further archaeological testing using dated sequences from both the coast and interior.

The First Coastal Squatters (11,000–10,000 bp) Evidence for the earliest late Pleistocene inhabitants is difﬁcult to ﬁnd, since most of the sites are now located on drowned landscapes. Richardson (1981) has suggested investigating littoral zones with narrow continental shelves, where the rising ocean caused less horizontal displacement of the shoreline (see Chapter 10, this volume, for an extended discussion of this issue in the Australian context). It is precisely this kind of context that has allowed the preservation above modern sea level of the sites of Quebrada Jaguay, Quebrada Tacahuay, Ring Site, and Las Conchas (see Figure 13.1). For areas with less steeply declining continental shelves few archaeological sites are found along the coast before ca. 6,000 bp, when much of the current coastline formed and was settled. The same phenomenon is described for the coast of the Namib Desert (Hank 2002b) and accounts for the appearance of later occupation of the Australian coast, as well as the coast of south central Chile (Quiroz and Sanchez 2004). The earliest known occupation of the Atacama Desert took place at the end of the last glacial period, which was characterized by more humid conditions in the highlands between 13,000 and 9,500 bp, thereby providing the coast with higher discharges of freshwater. This process created exceptional estuarine habitats at the mouths of major valleys and small quebradas, including the formation of shallow freshwater lagoons, such as is found at the Ring Site (Sandweiss et al. 1989). The earliest archaeological sites are dated to around 11,000–10,000 bp and correspond to groups with ‘‘specialized ﬁsh-netting technology,’’ as recorded in Quebrada Jahuay, the Ring Site, and Quebrada Tacahuay (Keefer et al. 1998; Sandweiss et al. 1989, 1998). According to Sandweiss et al. (1998, 1999), site QJ 280 at Quebrada Jaguay is the only base camp dated to the late Pleistocene (ca. 11,105–11,088 bp), with evidence of post-holes representing an early form of architecture (Daniel Sandweiss, pers. comm.). It is interesting to note that people appear to have specialized their prey strategies in the capture of a single taxon of ﬁsh from the family Sciaenidae (calculated to be 17 cm in length, based on the size of the otoliths), and Mesodesma donacium, a species of mollusc bivalve normally found in sandy beaches (Sandweiss et al. 1998, 1999). Site QJ 280 also presents

_____________________________________________________________________________________________________

250

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

abundant lithic debitage from the production of bifacial and unifacial artifacts, including obsidian that was imported from outcrops located 130 km inland, at an altitude of 2,850 m above sea level (asl) (Sandweiss et al. 1998). This suggests that people operated within a large territorial range extending from the coast to the Andes. At Tacahuay, a logistical camp with dates beginning ca. 10,800 bp, there is also an extraordinary preference for small schooling ﬁsh of the same average size and taxon (anchovies – Engrundis ringens, anchoas – Anchoa spp., shad, herring), and sea birds such as cormorant and boobies (deFrance et al. 2001). The use of ﬁshing nets is inferred from the standardized size of the ﬁsh and the ﬁnding of a net fragment at Quebrada Jaguay dated to the Holocene (Keefer et al. 1998). In the Holocene there is evidence for people diving in the sub-tidal zone to procure molluscs using an extraction tool known as chope used to pry molluscs from rocks, for hunting with harpoons for large ﬁsh and marine mammals such as sea lions (Otaria juvata and O. ﬂavescens), for ﬁshing with ﬁshhook and ﬁshing lines, and for the collection of algae. The lack of similar evidence in the terminal Pleistocene assemblages suggests that during the late Pleistocene and early Holocene people employed different procurement strategies to access marine and terrestrial resources.

Gifted Holocene Hunter-Gatherers (10,000–4,000 bp) As deduced from several sites along the coast and the inland basin, by 10,000 years ago, human groups in both the northern and central sectors of the Atacama Desert employed a more specialized maritime economy, incorporating an integrated transhumance system (Nu´n˜ez 1983). The archaeological sites along the coast comprise extended and more permanent open camps with shell middens that cover several square kilometers and which have a depth of more than 4 m, such as the Punta Pichalo site excavated by Bird in the 1940s (Bird 1943) (see Figure 13.4). Occupants employed a sophisticated tool kit that included ﬁshhooks of vegetal thorns, weights and composite sinker hooks made of bone, ring-necked nets (chinguillo), and harpoon heads made of bone. They also manufactured a wide variety of lithic tools used as projectile points, knives, and scrapers. Obsidian implements and debitage are common at the coastal sites (Mun˜oz 1993), indicating that foraging included journeys to the interior highlands and/or that these and other materials were obtained through exchange networks with peoples of these interior areas. By the mid-Holocene (after ca. 7,000 bp) maritime technology evolved to include ﬁshhooks manufactured from Choromytilus chorus shell, awls, polishers, lanceolates and double points, burins, knife blades made on thick bifacial preforms, grinding stones, hand stones, choppers, multipurpose ﬂaked tools made out of chalcedony and basalt, wooden darts, bags stitched with ﬁber, cords, and reed mats. Later in the Holocene (ca. 4,000 bp) the manufacture and use of shell beads and pendants, fringed skirts, reed brushes, lithic artifacts with ocher, and pelican skins appear in the archaeological record, pointing to the beginning of a new cultural era.

_____________________________________________________________________________________________________

251

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

Figure 13.4 Shell midden at Punta Pichalo in central Atacama Desert, which demonstrates the permanent and intense occupation of the coast based on a hunting-gathering marine economy, during the Holocene (picture taken by Junius Birds in 1939; reproduced courtesy of the American Museum of Natural History). Grete Mostny, fomer Director of the Museo Nacional de Historia Natural, is shown in the right corner.

The lack of terrestrial animals appears to be ‘‘compensated’’ by the hunting of sea lions which were hunted with harpoons while asleep on rocky foreshores (Standen 2003). Molluscs were another source of protein, principally loco (Concholepas concholepas), lapas (Fissurella spp.), and chorito maico (Perumytilus purpuratus) (Luis Briones pers. comm.). The size of these species corresponds to individuals normally found in the intertidal and supratidal zones of rocky beaches. The evidence for watercraft at this time is still uncertain and seems to have been a later technological development. Sites in the inland basin of Pampa del Tamarugal such as Tiliviche and Arago´n (see Figure 13.1), located 40 km from the littoral and about 950 m asl, appear to have served as transitory logistical camps. It is thought that groups traveled there to gather raw material for lithic artifacts, to collect reed ﬁber, to gather scanty terrestrial plants, and to hunt small game. These camps were occupied from 9,000–4,500 bp when this region experienced wetter conditions (Latorre at al. in press). At Tiliviche there is evidence for domestic structures consisting of simple circular and shallow depressions surrounded by post-holes dug into the hard ground of the terrace (Nu´n˜ez 1983; see Figure 13.5). These kinds of wind shelters were possibly covered with reed mats, fur, and branches. Domestic refuse was dispersed around these enclosures, and after their abandonment, they were covered

_____________________________________________________________________________________________________

252

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

Figure 13.5 Open camp of Tiliviche, an inland oasis around 40 km from the coast. Postholes correspond to an early Holocene wind breaker (photo courtesy of Lautaro Nu´n˜ez). Note the shallow midden deposit compared with Punta Pichalo (Figure 13.3).

by refuse from nearby dwellings (see also Quiani 9, a middle Archaic camp, located in the south of Arica; cf. Mun˜oz 1993).

Social and Ideological Complexity in the Atacama Desert Coast According to Arriaza and Standen (2002), one of the most remarkable aspects of coastal Atacama Desert people were the processes they used to mummify their dead. This cultural trait is labeled ‘‘Chinchorro’’ after the name of a beach, north of Arica, where the ﬁrst evidence was found by Max Uhle, early in the twentieth century (Uhle 1919). Geographically this phenomenon has a rather narrow distribution from Ilo in southern Peru, to Patillo south of Iquique, approximately 300 km in length (see Figure 13.1); to date there is no evidence of artiﬁcial mummiﬁcation away from the coast. The earliest evidence to date for mummiﬁcation was located at the mouth of the valley of Camarones (see Figures 13.1 and 13.3), where Virgilio Schiappacasse y Hans Niemeyer established the association of domestic debris and Chinchorro inhumation. These ﬁndings in the Atacama represent the most ancient (ca. 8,000–7,000 bp and 3,500 bp) artiﬁcial mummiﬁcation known in the world, a typical trait of the northern and central Atacama

_____________________________________________________________________________________________________

253

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

Desert. South of Patillo, still in central Atacama and in the southern Atacama, artiﬁcial mummiﬁcation was not part of the Huentelauquen Tradition linked to the Las Conchas site (see Figure 13.1) in Antofagasta (Costa-Junqueira 2001; Llagostera et al. 2000). Although natural mummiﬁcation was likely known to the Chinchorro, they created their own unique artiﬁcial mummiﬁcation methods. These methods drastically transformed the bodies, halting decomposition, and were seen to allow the dead to ‘‘participate’’ in a new social context of unknown meanings within an egalitarian social system. It is important to note that the highland and coastal hunter-gatherers of western South America developed a wide variety of procedures to ‘‘manipulate’’ their dead. These practices become associated with ceremonial centers linked to the elite of highly stratiﬁed societies of the central Andes (Santoro et al. in press). Arriaza and Standen (2002: 23) point out ‘‘societies with centralized sociopolitical power and class differentiation would have richer and more sophisticated mortuary rites allocated for elite members than hunter-gatherer societies.’’ They created a ‘‘variety of techniques of artiﬁcial mummiﬁcation that transformed the bodies into gleaming black or red statues, a sharp contrast to the opaque shades of the desert’’ that rivals those of complex societies. Arriaza (1995) classiﬁed the body preparation procedures as black, red, bandaged, and mud-coated mummies. The black, red, and bandaged mummies correspond to the category known as complex preparation (Uhle 1919). This great variety of mummiﬁcation methods changed through time from black to red, and from red to mud-coated techniques. The ‘‘black mummies’’ were the most elaborate, requiring more time and knowledge of anatomy for their preparation. Examples of these mummies have been radiocarbon dated from 7,000 to 4,800 bp. According to Arriaza and Standen (2002), these bodies were completely skinned, disarticulated, and the organs and muscles were removed. Subsequently, the skeleton was rearticulated and the bones reinforced with sticks and reeds. The body cavities (head and trunk) were stuffed with ashes or dry grasses. Later, they modeled the entire body with an ash paste, recreating the original volume of the corpse. Following this process, the morticians reattached the skin, possibly the body’s own or that of an animal (seal or sea lion) and the head was adorned with a short, black haired wig. The Chinchorros also kept and modeled some of the facial features and genitalia. As a ﬁnal touch they painted the mummy with a thin layer of black-blue manganese paste; hence the reason for the title ‘‘black mummy’’ (Arriaza and Standen 2002: 29). The ‘‘red mummies’’ were simpler than the black ones in their preparations, characterized by incisions and sutures at the shoulders, groin, knees, and ankles with the objective of removing the organs and part of the muscles. The head was disconnected from the trunk and the brain removed. All the body cavities were dried with glowing coals and ashes, a custom known also in the coast of central Peru, where people added salt to treat their dead (Benfer 1999). Wooden poles were slipped under the skin through the arms, legs, and spine of the deceased to provide support and rigidity. The cavities were stuffed with feathers from sea birds (pelicans, seagulls), different types of soils, and wool from camelids (llamas (Lama glama), alpacas (Lama paco), guanacos, or vicun˜as (Vicugna vicugna), the

_____________________________________________________________________________________________________

254

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

other wild American camelid only found in the high Andes, >4,000 m asl). The head was treated similarly – they added a long wig made from locks of human hair and a paste of manganese paint for the face, modeling the facial characteristics, as well as securing the wig to the head. Afterwards, the entire body except the face and wig was painted with red ocher. This red style ﬁrst appeared approximately 4,000 years bp and continued for almost 500 years (Arriaza and Standen 2002). The ‘‘bandaged mummies’’ were basically a variation on the red mummies except the skin was reattached in the form of bandages. The ‘‘mud-coated mummies’’ were naturally dried bodies that were covered with a layer of mud, possibly prepared with a mixture of earth, water, and an adhesive of unknown origin (Arriaza and Standen 2002). Why did these groups expend so much energy and engage in such complex procedures to prepare their dead? Perhaps it was thought that the dead were going to face an unknown world of supernatural metaphysical forces or perhaps, instead, they would start a new life within their own community; therefore becoming a new category of social actor. It may be that their function was to mitigate the impact their death caused in their own social group. We do not know the reasons why different mummiﬁcation procedures were applied to members of the same group and if the members who received greater attention enjoyed greater social prestige during their lifetime and had certain privileges that they maintained even after death. These questions warrant further investigation. Arriaza and Standen (2002) suggest that through the act of mummiﬁcation groups produced inanimate beings with simultaneous spiritual bonds to the living, their ancestors, and to the gods. Particularly, it might be possible that post-mortem manipulation of the dead began due to beliefs that ‘‘the soul of the deceased runs the risk of losing its existence if the body decomposes, or the living would lose contact with the soul of the deceased should the body decompose’’ (Arriaza and Standen 2002: 40–1). Thus, these mortuary rituals may have been necessary to ensure the eternal existence of the body and the soul of the deceased, as well as the continuity of the group (Arriaza and Standen 2002: 40–1). What is certain, however, is that the Chinchorro invested a considerable amount of time and energy in mummifying their dead, which makes these mummies a unique phenomenon in the world, as much for their antiquity as for the social context in which they were produced. The technical knowledge and ideological concepts tied to mummiﬁcation, as well as the subsistence patterns which underpinned these, all required a system that allowed for the preservation of a historic memory for several millennia. To ensure the cultural continuity of such a legacy from one generation to the next, people adopted a course of action which included mummiﬁcation. This practice may have served to preserve internal cohesion, reinforce social integrity, and perpetuate a collective historic memory. The momentum created by the actual process of mummifying a newly deceased member of the community may have created a selfsustaining milieu for the Chinchorro people. Codes of social conduct could be instilled during the process of mummiﬁcation, as this practice may have created the necessary conditions for cultural transmission and reproduction.

_____________________________________________________________________________________________________

255

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t

Conclusion The cultural history of the people in the Atacama Desert can be divided into three major epochs, summarized as follows. The early epoch corresponds to immigrants arriving on the coast of the Atacama no later than 11,000 bp at Quebrada Jaguay and Quebrada Tacahuay, north of Arica. The available archaeological data show that these people were equipped with simple and non-specialized technologies and focused their procurement on a narrow spectrum of molluscs, sea birds, and ﬁsh. There is, however, evidence for the possible use of nets for ﬁshing, even at this early stage. Populations also occupied the high Andes at altitudes of 3,000–4,500 m asl on the margins of the desert from around 10,800 bp, but here they employed a more technologically sophisticated lithic industry characterized by highly formalized tools for the capture and butchering of large game (including camelids, deer, and occasionally the American horse), as well as small rodents and birds (Aldenderfer 1999; Nu´n˜ez et al. 2002; Santoro and Nu´n˜ez 1987). This different adaptation may be linked to the arrival of two cultural ‘‘traditions’’ to the Atacama Desert, which maintained their character through prehistory. While the modern descendents of these original coastal peoples are hard to distinguish among the current population of the coastal cities, the highland people have survived the pressure of European invasion and the acculturation trends of the modern republic and today they assert different forms of social and ethnic recognition. The ﬁrst populations detected archaeologically in the Atacama may be related to earlier immigrants into the Americas, from 13,000–12,000 bp (or even earlier). They may have moved rapidly through new lands, as Birdsell (1977) suggested for Australia, leaving behind parties that started the more serious process of colonization as seen in Quebrada Jaguay (site QJ 280) at around 11,000 years bp (Sandweiss et al. 1989). This fast movement of the colonists is consistent with the tight chronological sequence of the ﬁrst Americans, given that there is no clear chronological gap between the earliest known settlements in North vs South America. In fact, the earliest Paleo-Indian site in the South Americas is Monte Verde, south of Chile, dated between 13,500 and 12,500 bp, with other early sites from Patagonia (i.e., Cueva del Medio, Piedra Museo, and Los Toldos) dated to between 12,390 and 12,890 bp (Massone and Prieto 2004). The occurrence of later dates for the ﬁrst occupants of the Atacama Desert provides a challenge for future archaeologists to investigate whether a cline in settlement of this desert has occurred (a lag effect) or whether sites of comparable age have simply not been investigated yet or have not survived due to taphonomic processes. In the second epoch, dated to the Holocene, the coastal desert people show a clear shift in their subsistence technology, as well as their social and ideological strategies, which persevered, with only minor changes, for several millennia. The material expressions of this new ideology seem to have been quite dynamic, as seen in the wide variety of methods used for artiﬁcial mummiﬁcation. Specialized technological solutions for exploitation of marine resources were developed by

_____________________________________________________________________________________________________

256

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t

the mid-Holocene, and remained in place with only minor changes until late prehistory. A distinctive feature of the hunter-gatherer-ﬁshermen of the Atacama Desert was the opportunity to engage in agriculture, afforded by the presence of valleys and quebradas that cross-cut the desert in the northern section (see Figure 13.3). From ca. 4,000–3,000 bp the introduction of agriculture is seen to complement a maritime focused economy, thereby allowing desert people to establish permanent settlements in the form of small hamlets, with simple architecture (wooden posts, reed mats, and cane) of no more that 10 hectares in extension, as seen late in prehistory. This kind of infrastructure is commonly found along the quebradas and inland basins, such as Pampa del Tamarugal. Along the coast, people maintained more simple open camps until the sixteenth century ad; the time of the European invasion. This change in the way of life of the peoples of the coastal Atacama Desert can be accommodated by the concept of the ‘‘Neolithic Revolution,’’ as coined by the inﬂuential Australian archaeologist V. Gordon Childe, whose ideas were incorporated into the interpretation of Andean cultural history during the 1960s and 1970s. This ‘‘revolution’’ within the coastal economies of the Atacama did not result in substantial change in staple foods, however, as these were still provided by the ocean. Processes of social class differentiation and other cultural ‘‘sophistication’’ classiﬁed elsewhere as ‘‘civilization’’ took a different form here. The waves of the ocean and the sand dunes of the desert witness a different human social creation that we have attempted to shed some light on within this chapter.

References Adovasio, J. and Page, J. 2002: The First Americans: In Pursuit of Archaeology’s Greatest Mystery. New York: Random House. Aldenderfer, M. 1999: Montane Foragers: Asana and the South-Central Andean Archaic. Iowa: University of Iowa Press. Arriaza, B. 1995: Beyond Death: The Chinchorro Mummies of Ancient Chile. Washington, DC: Smithsonian Institution Press. Arriaza, B. and Standen, V. 2002: Death, Mummies, and Ancestral Rites: The Chinchorro Culture. Arica: Serie Patrimonio Cultural I Region de Tarapaca´, Universidad de Tarapaca´. Auderheide, A. 2003: The Scientiﬁc Study of Mummies. Cambridge: Cambridge University Press. Benfer, R. 1999: Proyecto de Excavaciones en Paloma, Valle de Chilca, Peru´. Boletı´n de Arqueologı´a PUCP, 3, 213–37. Betancourt, J., Latorre, C., Rech, J., Quade, J., and Rylander, K. 2000: A 22,000 year record of monsoonal precipitation from northern Chile’s Atacama Desert. Science, 289, 1546–50. Bibar, J. 1966 [1565]: Cro´nica y Relacio´n Copiosa y Verdadera de los Reinos de Chile. Santiago: Fondo Histo´rico y Bibliogra´ﬁco Jose´ Toribio Medina. Bird, J. 1943: Excavation in Northern Chile. New York: Anthropological Papers of the American Museum of Natural History 38 (4), American Museum of Natural History. Birdsell, J. 1977: The recalibration of a paradigm for the ﬁrst peopling of Australia. In J. Allen, J. Golson, and R. Jones (eds), Sunda and Sahul: Prehistoric Studies in Southeast Asia, Melanesia, and Australia, London: Academic Press, 113–67.

_____________________________________________________________________________________________________

257

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t Bryan, A. 1973: Paleoenvironments and cultural diversity in late Pleistocene South America. Journal of Quaternary Research, 3, 237–56. Chauchat, C. 1988: Early hunter-gatherers on the Peruvian coast. In R. Keatinge (ed.), Peruvian Prehistory, Cambridge: Cambridge University Press, 41–66. Costa-Junqueira, M. 2001: Modalidades de enterramientos humanos arcaicos en el norte de Chile. Chungara Revista de Antropologı´a Chilena, 33, 55–62. Dalton, R. 2003: The coast road. Nature, 422, 10–12. deFrance, S., Keefer, D., Richardson, J., and Umire, A. 2001: Late Paleoindian coastal foragers: Specialized extractive behavior at Quebrada Tacahuay, Peru. Latin American Antiquity, 12, 413–26. Dillehay, T. 2000: The Settlement of the Americas: A New Prehistory. New York: Basic Books. Dillon, M. 1997: Lomas Formations: Peru. In S. Davis, V. Heywood, O. Herrera-McBryde, J. Villa-Lobos, and A. Hamilton (eds), Centres of Plant Diversity: A Guide and Strategy for their Conservation, Oxford: WWF Information Press, 519–27. Gaete, N., Navarro, X., Constantinescu, F., Mera, C., Selles, D., Solari, M., Vargas, M., Oliva, D., and Duran, L. 2004: Una Mirada al Modo de Vida Canoero del Mar Interior, desde Piedra Azul. Chungara Revista de Antropologı´a Chilena, special volume, 333–46. Grosjean, M. and Nu´n˜ez, L. 1994: Late glacial, early, and middle Holocene environments, human occupations, and resource use in the Atacama (northern Chile). Geoarchaeology, 9, 271–86. Hank, J. 2002a: El Desierto de Kalahari. In R. Mittermeier, C. Goettsch Mittermeier, ´ reas Silvestres: P. Robles Gil, J. Pilgrim, G. Fonseca, T. Brooks, and W. Konstant (eds), A ´ ´ las Ultimas Regiones Vırgenes del Mundo, Mexico: CEMEX, 384–93. Hank, J. 2002b: The deserts of Namib. In R. Mittermeier, C. Goettsch Mittermeier, P. Robles ´ reas Silvestres: las U ´ ltimas Gil, J. Pilgrim, G. Fonseca, T. Brooks, and W. Konstant (eds), A ´ Regiones Vırgenes del Mundo, Mexico: CEMEX, 394–403. Hidalgo, J., Schiappacasse, V., Niemeyer, H., Aldunate, C., and Solimano, I. (eds) 1989: Culturas de Chile: Prehistoria desde sus Orı´genes hasta los Albores de la Conquista. Santiago: Editorial Andres Bello. Keefer, D., deFrance, S., Moseley, M., Richardson III, J., Satterlee, D., and Day-Lewis, A. 1998: Early maritime economy and El Nin˜o events at Quebrada Tacaguay, Peru. Science, 281, 1830–2. Latorre, C., Betancourt, J. L., Rylander, K. A., and Quade, J. 2002: Vegetation invasions into absolute desert: A 45,000 year rodent midden record from the Calam-Salar de Atacama basins, northern Chile (lat 228–248S). Geological Society of America Bulletin, 114 (3), 349–66. Latorre, C., Betancourt, J., Rech, J., Quade, J., Holmgren, C., Placzek, C., Maldonado, A., Vuille, M., and Rylander, K. in press: Late Quaternary history of the Atacama Desert, 238 south. In M. Smith and P. Hesse (eds), The Archaeology and Environmental History of the Southern Deserts, Canberra: National Museum of Australia. Latorre, C., Betancourt, J. L., Rylander, K. A., Quade, J., and Matthei, O. 2003: A vegetation history from the arid prepuna of northern Chile (22–23’S) over the last 13,500 years. Palaeogeography, Palaeoclimatology, Palaeoecology, 194, 223–46. Lavalle´e, D. 2000: The First South American: The Peopling of a Continent from the Earliest Evidence to High Culture (translated by P. G. Bahn). Salt Lake City: University of Utah Press. Llagostera, A. 1992: Early occupations and the emergence of ﬁshermen on the Paciﬁc Coast of South America. Andean Past, 3, 87–109.

_____________________________________________________________________________________________________

258

_____________________________________________________________________________

Pe o p l e o f t h e C o a s t a l At a c a m a D e s er t Llagostera, A., Weisner, R., Castillo, G., Cervellino, M., and Costa-Junqueira, M. 2000: El Complejo Huentelauque´n bajo una perspectiva macroespacial y multidisciplinaria. In Actas XIV Congreso Nacional de Arqueologı´a Chilena, Tomo I, Contribucio´n Arqueolo´gica no. 5, Atacama: Museo Regional de Atacama, 461–81. Lozano-Machuca, J. 1885 [1581]: Cartas del Factor de Potosı´ Juan Lozano Machuca al Virrey del Peru´, en donde se Describe la Provincia de Lipes. Madrid: Relaciones Geogra´ﬁcas de Indias, Tomo II. Marquet, P., Gonza´lez, H., Pinto, R., Santoro, C., Standen, V., and Zeballos, H. 2002: Desiertos Costertos del Peru´ y Chile. In R. Mittermeier, C. Goettsch Mittermeier, ´ reas Silvestres: P. Robles Gil, J. Pilgrim, G. Fonseca, T. Brooks, and W. Konstant (eds), A ´ las Ultimas Regiones Vı´rgenes del Mundo, Mexico: CEMEX, 364–73. Massone, M. and Prieto, A. 2004: Evaluacio´n de la Modalidad Cultural Fell 1 en Magallanes. Chungara Revista de Antropologı´a Chilena, special volume, 303–15. Meltzer, D. 1993: Search for the First Americans. Montreal: St Remy Press. Moraga, M., Aspillaga, E., Santoro, C., Standen, V., Carvallo, P., and Rothhammer, F. 2002: Ana´lisis de ADN Mitocondrial en Momias del Norte de Chile Avala Hipo´tesis de Origen Amazo´nico de Poblaciones Andinas. Revista Chilena de Historia Natural, 74, 719–26. Morphy, H. 1998: Aboriginal Art. Singapore: Phaidon Press. Mulvaney, J. and Kamminga, J. 1999: Prehistory of Australia. Sydney: Allen and Unwin. Mun˜oz, I. 1993: Spatial dimensions of complementary resource utilization at Acha-2 and San Lorenzo, northern Chile. In M. Aldenderfer (ed.), Domestic Architecture, Ethnicity, and Complementarity in the South-Central Andes. Iowa City: University of Iowa Press, 94–102. Nu´n˜ez, L. 1983: Paleoindian and Archaic Cultural Period in the arid and semi-arid region of northern Chile. Advances in World Archaeology, 2, 161–222. Nu´n˜ez, L. 1999: Chile. In T. Murray (ed.), Encyclopedia of Archaeology: History and Discoveries vol. 1, Santa Barbara, CA: ABC Clio, 301–15. Nu´n˜ez, L., Grosjean, M., and Cartajena, I. 2002: Human occupations and climate change in the Puna de Atacama, Chile. Science, 298, 821–4. Quilter, J. 1991: Late Preceramic Peru. Journal of World Prehistory, 5, 387–438. Quiroz, D. and Sanchez, M. 2004: Poblamientos Iniciales en la Costa Septentrional de la Araucanı´a [6500–2000 ap]. Chungara Revista de Antropologı´a Chilena, special volume, 289–302. Rauh, W. 1985: The Peruvian Chilean deserts. In M. Evenary, M. Noy-Meir, and D. Goodall (eds), Hot Deserts and Arid Shrublands, Amsterdam: Elsevier, 239–66. Richardson, J., III 1981: Modeling the development of early complex economies on the coast of Peru: A preliminary statement. Annals of Carnegie Museum, 50, 139–50. Rivera, M. and Rothhammer, F. 1986: Evaluacio´n Biolo´gica y Cultural de las Poblaciones Chinchorro. Nuevos Elementos para la Hipo´tesis de Contactos Transaltipla´nicos: Cuenca Amazonas-Costa Pacı´ﬁco. Chungara, 16–17, 295–306. Rivera, M. and Rothhammer, F. 1991: The Chinchorro people of northern Chile 5,000 bc– 500 bc: A review of their culture and relationships. International Journal of Anthropology, 6, 243–55. Rosenfeld, A. and Smith, M. 2002: Rock art and the history of Puritjarra rock shelter, Cleland Hills, central Australia. Proceedings of the Prehistoric Society, 68, 103–24. Rothhammer, F., Coccilovo, J., and Quevedo, S. 1984: El Poblamiento Temprano de Sudame´rica, Chungara, 18, 99–108.

_____________________________________________________________________________________________________

259

_____________________________________________________________________________

S a n t o r o , A r r i a z a , S t a n d e n , Ma r q u e t Sandweiss, D. 2003: Terminal Pleistocene through mid-Holocene archaeological sites as paleoclimatic archives for the Peruvian coast. Palaeogeography, Palaeoclimatology, Palaeoecology, 194, 23–40. Sandweiss, D., McInnis, H., Burger, R., Cano, A., Ojeda, B., Paredes, R., Sandweiss, M., and Glascock, M. 1998: Quebrada Jaguay: Early South American maritime adaptations. Science, 281, 1830–2. Sandweiss, D., Richardson, J., III, Reitz, E., Hsu, J., and Feldman, R. 1989: Early maritime adaptations in the Andes: Preliminary studies at the Ring Site, Peru. In D. Rice, C. Stanish, and P. Scarr (eds), Ecology, Settlement, and History in the Osmore Drainage, Peru, Oxford: British Archaeology International Series no. 545, 35–84. Sandweiss, D., Richardson, J., III, Reitz, E., Hsu, J., Feldman, R., Sandweiss, M., Cano, A., Ojeda, B., and Roque, J. 1999: Pescadores Paleoindios del Peru´. Investigacio´n y Ciencia, 277, 55–61. Santoro, C. 2001: Cultures of the Chilean Desert. Arica: Ediciones Universidad de Tarapaca´. Santoro, C. and Nu´n˜ez, L. 1987: Hunters of the dry and salt Puna. Andean Past, 1, 57–109. Santoro, C., Standen, V., Arriaza, B., and Dillehay, T. in press: Archaic funerary pattern or post-depositional alteration? The Patapatane burial in the highlands of south central Andes. Latin American Antiquity. Schiappacasse, V. and Niemeyer, H. 1984: Descripcio´n y Ana´lisis Interpretativo de un Sitio Arcaico Temprano en la Quebrada de Camarones. Santiago: Museo Nacional de Historia Natural. Standen, V. 2003: Chinchorro. Chungara Revista de Antropologı´a Chilena, 35, 175–207. Standen, V., Santoro, C., and Arriaza, B. 2004: Sı´ntesis y propuestas para el perı´odo Arcaico en el extremo norte de Chile. Chungara Revista de Antropologı´a Chilena, special volume, 201–12. Uhle, M. 1919: La Arqueologı´a de Arica y Tacna. Boletı´n de la Sociedad Ecuatoriana de Estudios Histo´ricos Americanos, 3 (7/8), 1–48. Usselmann, P., Fontugne, M., Lavalle´e, D., Julien, M., and Hatte´, C. 1999: Estabilidad y Rupturas Dina´micas en el Holoceno de la Costa Sudperuana: El Valle de la Quebrada de Los Burros (Departramento de Tacna). Bolletin Institute Francais Etudes Andines, 28, 1–11. Yesner, D. 1980: Maritime hunter-gatherers: Ecology and prehistory. Current Anthropology, 21, 727–50.

_____________________________________________________________________________________________________

260

_____________________________________________________________________________

-----------------------------------------------------------------------------

14

--------------------------------------------------------------------------

Desert Solitude The Evolution of Ideologies Among Pastoralists and Hunter-Gatherers in Arid North Africa Andrew B. Smith

Introduction: Religion Borne of Desert Solitude Then was Jesus led up of the Spirit into the wilderness . . . and when he had fasted forty days and forty nights, he was afterward anhungered. (Matthew 4: 1–2)

Jesus went on a vision quest into the desert solitude, where, after sensory deprivation of 40 days, he proved his holiness by refusing to be tempted by the Devil who appeared before him. In the third and fourth centuries ad, St. Anthony – the father of monasticism – had similar experiences in his desert retreat at Dayr al-Maymun in Egypt. His later hermitage at Dayr Mari Antonius, near the Red Sea, where he died in ad 356, still exists. In the early twentieth century, Charles de Foucauld built his cell at Assekrem in the Hoggar Mountains of southern Algeria, where he died at the hands of rebel Tuareg ﬁghting against France in 1916. It was no accident that St. Catherine’s monastery was set up in the Sinai Desert, where an experience similar to that of Jesus or St. Anthony would be possible by other monks. It is also close to the Red Sea, across which it is said that Moses led his people out of Egypt. In the Hebrew religion, Yahweh is recognized as a guide through the wilderness. He knows the safe trails, the rare wells, the spots where vegetation may be found for the cattle, and can guide His people without danger to themselves. This combination of mountain spirit, storm and volcanic deity, and wilderness guide, clearly goes back to the nomad period of Israel’s history. (Oesterley and Robinson 1966: 155)

_____________________________________________________________________________________________________

261

_____________________________________________________________________________

An d r e w B. S mith

Judeo-Christian-Islamic religions all come from desert nomadic traditions, and each still looks to the desert for solace. Colonel Gaddaﬁ, the Libyan leader and son of nomadic pastoral parents from the Fezzan, believes his strength and wisdom for his new world order comes from his asceticism in the desert. Desert solitude has given direction to philosophers and sages throughout the ages. The Greek word eremos, from which we derive ‘‘hermit’’ in English, means desert. Saharan nomads equally have traditions and beliefs about spirits of the lonely places, many of which would appear to predate the inﬂuence of Islam upon them.

Desert Nomads in Saharan Prehistory The following discussion considers the antiquity and development of pastoralism in northern Africa and likely scenarios for interaction between hunter-gatherers and these pastoralists. There is some dispute over whether early cattle bones found at Nabta Playa in the Western Desert of Egypt and dated to 10,000 years ago are wild or not. Wendorf and Schild (1994) used an ecological argument based on the fact that 90 percent of the fauna are hares and gazelles, and thus a reﬂection of arid conditions. They interpreted this as suggesting the environment at the time of occupation was dry, and therefore the cattle could not have lived so far from surface water without human intervention: thus, the cattle were domesticated. In contrast, the fragmentary condition of the specimens from Nabta led Clutton-Brock (1989) to question the domestic status. Smith (1992) suggested the environmental argument is weak, since there are no environments in Africa with just hares and gazelles, and that medium-size antelopes, like oryx, are missing from the faunal assemblage. In addition, the cattle bones are all large, comparable to the size of wild cattle from elsewhere in North Africa and Europe. It is a well known phenomenon that, with domestication, animals become smaller. By 7,400 bp, when the ﬁrst ovicaprids from the Near East appear in North Africa, we can be quite sure that these were domesticated. The cattle which appear at the same time are smaller than the wild forms, so we can also assume they were domesticated. Although there is evidence to support the idea that some of the indigenous wild cattle of North Africa were a source of genetic material found in modern African breeds (Hanotte et al. 2002), as yet there is no way of dating when this might have occurred, or whether it really was an independent domestication event. It is possible that wild African cattle were crossbred with incoming domesticates from the Near East approximately 7,400 years ago. In Di Lernia’s (1999a) Early Pastoral period of the central Sahara (7,400–6,400 bp) it would appear that herders lived alongside indigenous hunters for a while, as two different pottery types are evident. High lake levels indicate good rains and a savanna environment that provided the necessary grazing for keeping cattle. At the end of this period a dry spell lasting ca. 400 years made the central Sahara a difﬁcult environment for cattle herding, but this was followed by climatic amelioration improving the conditions once more.

_____________________________________________________________________________________________________

262

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a

The Middle Pastoral period (6,080–5,100 bp) was a time of pastoral consolidation, in which herding strategies were to keep cattle in the low lying areas and the small stock in the mountains. This may be the time of the black-faced bovidian rock art (Smith 1993), where people with African physical features are seen with detailed depictions of cattle, showing their many coat colors (Lajoux 1963). A complete cow skeleton, found at Adrar Bous in the central Sahara and dated to 5,800 bp, had the bowed appearance of an animal which had died from drought, with tightened skin (Clark et al. 1973), suggesting that even during this period conditions were not always favorable. Around 5,100 bp the climate of the Sahara began to ﬂuctuate (Petit-Maire and Riser 1983; Servant and Servant-Valdary 1980), with lakes becoming seasonal and possibly drying up completely in some years. This corresponds with the Late Pastoral period (5,100–3,900 bp) when the unpredictable conditions saw a greater reliance on small stock. By 4,000 years ago the environmental conditions began to approach those of the present in the central Sahara, where rainfall only occurs every few years and typically in single events. Accordingly, pastoralists had to make signiﬁcant adjustments to their lives, such as having to rely on underground water from wells, and perhaps only including the deep desert in their annual pastoral round during those years in which rain might fall (S. E. Smith 1980). Certainly, the mountainous areas, where sub-surface water could be tapped, would have been convenient refuges. Perhaps the greatest adjustment of pastoralists at this time was to begin controlling and manipulating the wild grains – sorghum and millet – and to begin growing these in gardens close to semi-permanent villages. Two such villages have been studied in Niger – Iwelen and Chin Taﬁdet – and both show similar dates of development (Paris 1990, 1992). Saharan nomads are notorious for their personal aversion to agriculture. They are, however, quite happy for other people to grow grains for them and to extract their tribute at harvest-time. It may well be that around 3,500 bp, with the ﬁrst appearance of domesticated grains, slaves were being brought from the south to work the ﬁelds for the nomads. If so, this would represent the ﬁrst signs of a hierarchical system where there was a social separation between the nomads and the agricultural Harratin of the desert oases. Pressure on agricultural people on the southern fringes of the Sahara may have been one signiﬁcant reason for the development of organized states, such as Ghana and Mali, for defense against pastoralist depredations (MacDonald 1998; Munson 1980). These villages coincided with the clustering of graves with well deﬁned mortuary architecture into cemeteries. These cemeteries are argued to represent more than just a collection of graves – they became a marker for social identity and a focus for ritual activity (cf. Pardoe 1988). Thus the cultural landscape of the central Sahara now included not only pastoralists, who tend to be hierarchical, but lower caste black slaves, whose job was to produce grains for their masters, a situation which still pertains in many places. The new villages could then be reoccupied for periods of the year, especially at the height of the dry season when Saharan nomads would be forced to stay around permanent water points.

_____________________________________________________________________________________________________

263

_____________________________________________________________________________

An d r e w B. S mith

Saharan Rock Art and Religious Life The drying up of the Sahara, ultimately forcing pastoralists to adopt a wider food base, also coincided with the appearance of later bovidian rock art in the Acacus Mountains and Tassili n’Ajjer of the central Sahara. This rich parietal art depicts cattle herders in social activities, as well as ritual observances. One such depiction is of a ﬁre ceremony (Kuper 1978: 426) (see Figure 14.1) where several men and a woman appear to be controlling ﬁre, watched by two horned animals with wavyline decoration on their coats. It is tempting to suggest that the wavy lines on the animals are related to the ﬂickering of the ﬁre. Several other paintings show this same unnatural decorative motif on cattle (Kuper 1978: 224–5, 430–1; Lajoux 1963: 110). Belief in ﬁre spirits was found among all pre-Islamic peoples of North Africa. The Arabs referred to them as madjus, meaning ﬁre worshippers (Mone`s 1988). Today, there are still remnant rituals of this belief among the Tuareg (Kel Tamasheq) of the Sahara (Rasmussen 1992). The artisans among the Tuareg (Inadan) are an endogamous caste who have a special relationship with the ﬁre djinns (spirits). These people are respected for their ability to control the ﬁre djinns, but are otherwise despised and considered of low status. These people are ritual specialists whose power allows them to be marriage brokers, and be politically active behind the scenes. The artisans both control the ﬁre djinns and also embody them. They practice curing in spirit possession, and because of their perceived liminal position within the society, can be seen as a ‘‘bridge’’ between nature and culture, between heaven and earth (Rasmussen 1992). Rock art is widely recognized as reﬂecting deep beliefs of the society of the artists. Often this is used as a metaphor for religious experiences, some of which may be shamanistic in form (Lewis-Williams 1981). In the pre-bovidian art of the Sahara we ﬁnd enigmatic human forms juxtaposed with depictions of Barbary sheep (Sansoni 1994). In one cave, Uan Afuda, in the Acacus Mountains of southwest Libya, a dung layer had formed (Di Lernia 1999a). From the excavation, it was apparent that this layer was created by Barbary sheep, whose bones

Figure 14.1 ‘‘Fire Ritual,’’ Uan Derbaouen, Tassili n’Ajjer from Colombel (in Kuper 1978).

_____________________________________________________________________________________________________

264

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a

comprised over 90 percent of the fauna from these layers. The archaeologists have deduced that the sheep were being penned inside the cave. Since these were wild animals, not easy to control in numbers, the occurrence of a poisonous plant among the ﬂoral remains can be suggested as having been used as a soporiﬁc to keep the animals acquiescent, much like Turkish herders do with willow leaves among their sheep today. The relationship between the hunters, the wild sheep, and the rock art can be inferred as one of control over the animal. Since rock art is not usually narrative, but is a metaphor for belief, the depiction of the animals on the rock walls of the cave suggest that these animals were not simply being used as a food source, but may have been kept on hand for ritual purposes some 10,000– 8,000 years ago. In this same period, further to the east in the Western Desert of Egypt, early remains of wild cattle have been found under conditions which suggest they may have been in the desert under human control. Certainly, by 6,400 bp cattle burials under stone tumuli are found in this area (Wendorf and Schild 1994). In Niger, charred cattle bones, also under a stone cairn, have been dated to 6,200 bp (Paris 2000). The combined evidence hints at a widespread cattle cult across the Sahara, even predating the cattle rituals which were central to religious belief in the predynastic Nile Valley of the Badarians. Cattle become central to religious belief in the Sahara from this time on. There are two distinct periods of cattle paintings in the rock art: one showing people of African genetic stock (Kuper 1978) (see Figure 14.2), while the later period shows people of Mediterranean ancestry (Kuper 1978) (see Figure 14.3). It is within paintings of the second group that there seems to be a relationship between decorated cattle and ﬁre. Decorated cattle paintings and engravings are widespread across the Sahara from Mali to the Nile Valley (Allard-Huard and Huard 1983). Even today, Fulani herders of Mali will paint some of their cattle during the annual migrations (Dupuy 1998).

Figure 14.2 ‘‘Black face’’people in cattle cleansing ritual, Uan Derbaouen, Tassili n’Ajjer from Colombel (in Kuper 1978).

_____________________________________________________________________________________________________

265

_____________________________________________________________________________

An d r e w B. S mith

Figure 14.3 ‘‘White face’’people in camp, Iheren, Tassili n’Ajjer from Colombel (in Kuper 1978).

Saharan Landscapes Today, the Sahara is a dry land, comprising a mixture of sand seas (ergs), stony plains (regs), and mountainous areas, some of which reach over 3,000 m above sea level. Herders are mostly restricted to the edges of the mountains, where water and pasture are available, although in the southern Sahara where rain sometimes falls a seasonal use of the open plains is possible. The environment was not always so harsh, and, as described above, before conditions reached those of the present, some 4,000 years ago, rainfall and evaporation regimes were such that much of the Sahara was a grassland biome. This meant that conditions were amenable to pastoral occupation in places that today would be impossible. Hunters and herders, although reliant on the landscape for their sustenance, have strong psychological and emotional ties to their exploitation territories. The space may be structured on cosmological principles, and camps/homesteads which are laid out according to belief in how the world is structured. Equally, the landscape comprises social spaces, and, as Jacobsohn (1995: 53) remarks of a group of nomadic herders from Namibia, the Himba, feel connected to their historical environment rather than to any built space . . . [the] perception of a broader spatial text contains a representation of the biographical past and the social and economic future. Although it is not built space, this text, with

_____________________________________________________________________________________________________

266

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a both ‘‘natural’’ and cultural properties, can be shown to be as implicated in the organization of the Kaoko social world as any onganda [homestead].

The landscape is equally a place of referents which are built into the oral histories, myth, and stories of the people. Even the most monotonous landscape to an outside observer is rich in events, the signiﬁcance of which are passed down through generations. Identity points may be natural or cultural, such as rock art or cemeteries. Tracks linking points are avenues along which people, animals, goods, and information are passed from person to person. As Tilley (1994: 27) notes: All locales and landscapes are . . . embedded in their social and individual times of memory. Their pasts as much as their spaces are crucially constitutive of their presents. Neither space nor time can be understood apart from social practices which serve to bind them together.

The landscape, therefore, has real meaning with respect to individual and group identity, and is more than just a backdrop for social and economic activities. The distribution of cattle paintings may well indicate social and ceremonial paths across the Sahara. Tracks linking identity points in the landscape are often built into stories and myths. These would have included important rock art sites, especially if indeed, as suggested here, that rock art is not simply narrative (i.e., depicting historical events), but used as a metaphor for religious experiences that could be understood within the cultural memory.

The Desert Solitude Today, virtually all North African nomads – Tuareg, Toubou, Fulani, and Maures – have come under the inﬂuence of Islam. Even though the Arab conquest of North Africa began in ad 670, we can recognize earlier beliefs, upon which Islam lays as a veneer. Among the Kel Ferwan Tuareg of Niger, esuf is the word for solitude or desert expanses. The desert winds that sweep across the Sahara are built into belief and stories. The spirits of the winds are called kel esuf, which are feared as potentially malevolent. The desert nomads will orient their tents so that the back is always into the wind (i.e., to the north), where the spirits lurk at night. The layout of the tent by the cardinal points will protect the owners from evil spirits: ‘‘If one lifts any of the lateral mats during the day, one must be very careful, when night falls, to drop those which face north, because a tent opened to the north will be exposed to the worst dangers’’ (Casajus 1987: 56). Of the cardinal points, the south is the least dangerous and the source of divine benediction. The south and north of the tent are thus spatial opposites which lead to other binary separations, such as benediction, danger, or ostentation (pride) (faste); bad luck, solemnity (nefaste). Inside the tent the bed is oriented east–west, with the head to the east.

_____________________________________________________________________________________________________

267

_____________________________________________________________________________

An d r e w B. S mith

The husband sleeps to the north, thus protecting his wife against the kel esuf. Such an orientation in turn leads to the idea of north (nefaste) being masculine, and the south (faste) feminine (Casajus 1987: 57). It is among Tuareg women that spirit possession is most likely to be found. They, then, can be seen to adhere more to pre-Islamic beliefs, as a foil against the men who are more liable to follow Islam, and will be the marabouts (priests). Tension thus exists between Islam (men) and spirit belief (women), even though the marabouts are not averse to falling back on spirit beliefs if the perceived need arises. This does not stop the more fundamental of the Islamicists bemoaning the lack of faith among the people as seen in the words of the Tuareg poet Al-Hajj Muhammad Abdullah bin Khalil al-Rahman al-Aghlali (as cited in Norris 1975: 191): Our Great Sahara has customs false and base which Contravene a Sunna proud and glorious. All we observe is the letter of the law and tribal customs We are like a people who have camped unceasingly in one wadi . . .

He goes on to say: Those lands which are inhabited save for our country Progressed. Their advance was rapid thinking with the mind. They attained a prominent state on lofty heights, While we were backward or retarded due to sterile custom Amidst a community which God’s book says, ‘‘Possessed what it deserved of treasures stored.’’ It evolved in habitations in the wilderness, but it grew up On deepening ignorance, denying us the power to think and to perceive. Many a youth of ours longs to awake and stay alert But the wasteland has made him simple, blurred his memory. By my father’s life, to travel, always on the move, is indeed a quickening, Reminding us of resurrection and the assembling on judgment day.

The Kabyle House The ideology behind the spatial orientation of the Tuareg tent is a nomadic analogue to the equally comprehensive and symbolic layout of the Kabyle house of the Atlas Mountains of Algeria and Morocco. The house is divided into ‘‘a dark part (the west and north sides) and a light part (the east and south sides)’’ (Bourdieu 1973: 109). According to the Kabyles, this reﬂects the division of the year into wet and dry seasons. Bourdieu orients his reader to the house by reference to the threshold on the light (east) side, across which one enters the abode. This is the entry that men or visitors (male) would be expected to use. The back, or west, door is a private portal used by women, and it is closest to the loom that is the domain of the woman of the house. As Bourdieu (1973: 110) states:

_____________________________________________________________________________________________________

268

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a

‘‘The orientation of the house is fundamentally deﬁned from the outside, from the point of view of men and, if one may say so, by men for men, as the place from which men come out.’’ Bourdieu, as a man, received the male perspective on Kabyle cosmology. If he had been a woman interviewing women perhaps a slightly greater emphasis would have been given to the world from the female side. In the 1977 translation of his book Outline of a Theory of Practice there is no mention of the woman’s door in the section dealing with the Kabyle house (Bourdieu 1977: 90–1). In the 1973 paper his reference to the door is almost in passing: ‘‘ . . . the low and narrow door, reserved for women, which opens in the direction of the garden, at the back of the house . . . ’’ (Bourdieu 1973: 106). The Tuareg tent is owned and made by a woman (often with her mother’s help before marriage). Thus the domain of women is the domestic sphere, while of men it is the herd and public life. Like the Kabyle house there is only one appropriate approach to the dwelling, but in the case of the Tuareg tent it is from the west where the door is located. Is there a connection between the two? If so, which inﬂuenced the other? The woman’s entry from the west might be a clue. In addition, both groups speak a Berber language. The archaeology of grain agriculture in the Mahgreb suggests that it arrived quite late, and like the rest of North Africa was preceded by nomadic pastoralism. As Camps (1974: 276) states: ‘‘There is no analytical element which will allow us to afﬁrm that agriculture was being practiced before the ﬁfth millennium bc on the African littoral.’’ Domestic animals were already in the Aure`s Mountains of Algeria in the sixth millennium bc (Roubet 1978). Thus fully sedentary communities like the Kabyles probably have their ancestry in mobile herding societies. If this is indeed the case, then the tents of the nomads, which existed before Islam, may have been the original structures on which the houses of the Kabyles were later modeled. Bourdieu (1977: 89) makes the point that Islam also affected the Kabyles where religion is male and magic female.

Funerary Monuments and Rituals Graves with ﬁnely ﬁnished funerary architecture are common in any of the mountainous areas of the Sahara where stone was available. Paris (1996), in his detailed study of funerary monuments in Niger, has shown there are several types which start appearing around 5,000 bp. There are two dominant styles. The ﬁrst are platesformes cylindriques a` gravillons (PCG) – a ﬂat-topped, gravel ﬁlled tumulus, with larger rocks holding the structure together, and the skeleton found in a hole dug into the parent soil – which exist over a period from 5,000–3,000 years ago, with the greatest number of such graves dating between 4,400–4,000 bp. The second type are known as tumulus tronconiques a` crete`re (TAC) – a conical tumulus, without the large retaining stones found in the above, with a depression in the middle – and overlap in time with the PCG style from 4,000–2,500 bp, but tend to dominate in the period 3,000–2,500 bp. There are several examples of a TAC grave being built upon an already existing PCG grave. While the details of construction of these monuments

_____________________________________________________________________________________________________

269

_____________________________________________________________________________

An d r e w B. S mith

differ signiﬁcantly, the common feature of several of the styles is a ﬂat top. Flattopped tombs are found across the Sahara to the Sudan (Sadr et al. 1994). Camps (1961) has suggested that ﬂat-topped cairns were used as a form of altar for rituals. He notes that the Nasemonians during the Roman period ca. 2,000 years ago would consult their ancestors by lying on their graves. The appearance of cemeteries with large numbers of graves in Niger after 4,000 bp is an indication that burial grounds were probably taking on a life of their own, where graves of respected elders were reference points for propitiation ceremonies. This idea is supported by the cultural material, such as pots, associated with the graves (Paris 1996: 303), presumably used as part of mortuary rituals. Certainly, there is evidence that during the pre-Islamic period in North Africa whole cemeteries grew up around respected people (Desanges 1981). As Parker (1999: 196) notes: ‘‘tombs are not just somewhere to put dead bodies: they are representations of power.’’ Such power symbols could be the rock art mentioned above, cemeteries, or any other humanly built structures. Tilley (1994: 109), in talking about megaliths in southwest Wales, says they were ‘‘created primarily as symbolic reference and ritually important ceremonial meeting-points on paths of movement, drawing attention to the relationship between local groups and the landscape.’’ Animal burials are often close to human graves, but analysis of cattle burials by Paris (2000) indicates no direct association with human burials, even though there appears to be no doubt that the animals were ritually slaughtered by slitting the throat (apparent from cut marks on the ﬁfth cervical vertebra of one individual). All were positioned in an east–west direction lying on their right side.

Hunter-Gatherers on the Periphery Prior to the advent of pastoralism ca. 7,500 bp, the Sahara was occupied by huntergatherers. These hunters have left signs of their passing in the caves of the central Sahara, where archaeological excavations have produced large quantities of the bones of the wild Barbary sheep, along with paintings of these animals on the rock walls (Di Lernia 1999a). For a short while these hunters apparently lived side by side with the new pastoral life in the Sahara (Di Lernia 1999b), but they quickly disappear from the archaeological record. Di Lernia (pers. comm.) believes that some of the small sites with slightly different pottery styles may have belonged to later hunters, but these people were most likely encapsulated or overrun by the more dominant herders, who would either incorporate them into their society as low status members, or keep them on the periphery as a convenient labor source. Certainly, this is the case for Saharan hunters, called Nemadi, who still survived by addax hunting until the 1950s. These people were socially marginalized by the pastoral Maures, and considered almost sub-human. This may have been due in part to their limited adherence to Islam, but more likely their status as non-stock owners would have placed them very much at the bottom of a social hierarchy which had slaves. To the Maures, the Nemadi were considered outcasts who were ignorant and

_____________________________________________________________________________________________________

270

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a

who would steal. Their status was such that it was shameful for a Maure to marry a Nemadi (Gabus 1952). Nevertheless, like hunters on the edge of herding societies elsewhere, the labor of the Nemadi and their knowledge of the desert were important to the dominant culture, among whom they would have patrons demanding tribute (Smith 1998).

Saharan Cultural Connections If the cosmology and cultural traditions of the Berber-speaking Tuareg and Kabyles have a common ancestry, with the nomads being potentially the original source, what is the history of the nomads? In their study of the genetic material of North Africans, Cavalli-Sforza et al. (1994) show a link between the Tuareg and the Beja, another desert pastoral group living between the Nile Valley and the Red Sea. What possible connection could there be between a Berber-speaking group from the central Sahara (Tuareg) and Cushitic speakers (Beja) from the Red Sea Hills? On the surface there would appear to be little in common, except that they are both camel herders. To seek some connection we need to go back to the beginning of the Holocene, some 10,000 years ago. I have already mentioned the hunters of wild Barbary sheep. There is a cultural connection between these hunters and people who were exploiting an aquatic habitat of the Saharan lakes and the Nile Valley using bone harpoons. These hunters and ﬁshers made a distinctive pottery type with a decorative motif known as Wavy-Line or Dotted Wavy-Line Ware which is found from Mali to the Gezira in Sudan, where the Blue and White Niles converge (A. B. Smith 1980). This is a vast area, but the pottery decoration is so similar, as are the radiocarbon dates, that we have to recognize a cultural connection. These linkages, which we can identify on the basis of cultural material, must also have been the result of connecting paths between sustainable water points for these nomadic hunters. Around 6,000 bp pottery of the pastoral Tenerian in Niger (A. B. Smith 1980) was decorated with a diamond rocker-stamped motif. This same motif is also to be found on pottery from around Khartoum in the Nile Valley, Sudan. Again, like the earlier period, it would appear that cultural connections can be postulated. Whether they were built upon the earlier pathways is difﬁcult to say, but the mobile aspect of ﬁrst nomadic hunters, and later pastoralists, would have used similar water points. The later connections appear to be with the focus on cattle, ﬁrst in the intentional burial of these animals, ca. 6,200 bp, and in the later bovidian paintings and engravings in the rock art ca. 4,500 bp. History has divided the Saharan nomads into somewhat smaller groups, but even if we look at the present day Tuareg and Fulani we see they are equally spread over vast areas of land. We might suggest that the linkages between the different groups of present day nomads, although broken today, might be connected by looking at the later inﬂuences upon them. For example, in the recent past the Tuareg have been arch-enemies of the Toubou, who live to the east beyond the Tenere Sand Sea. Both share a great deal in common, even though they do not look

_____________________________________________________________________________________________________

271

_____________________________________________________________________________

An d r e w B. S mith

alike, or speak similar languages. Not only are they camel herders, but they also have a hierarchical social organization of nobles, vassals, and slaves. Like the Tuareg, the Toubou have an endogamous caste of artisans who are both feared as the intermediaries with the spirit world and considered unclean and of low status. As noted above, the historical connection of the Tuareg appears to be with the Berbers of the North African littoral. With the Toubou, these Kanuric speaking people have more in common with the Sahel and Bornu to the south around Lake Chad. Both Tuareg and Toubou were well connected with trade routes to their respective neighbors, but we really have little information on the time depth of these connections. Earlier and more widespread connections are hinted at by a recent analysis of human genes from northern Cameroon, south of Lake Chad. One group has shown the distinctive gene markers only found among Asians, and this has been interpreted as a back migration to Africa from Asia (Cruciani et al. 2002). What this potentially means is that the almost inherent needs of nomadic pastoralism – pasture and water – across large tracts of open country from the Red Sea to the central Sahara resulted in complex movements of people and cultural material over considerable periods of time, and that they have left a sizable genetic imprint wherever they went. The period 4,000–3,000 years ago was a time of unstable climatic conditions that resulted in the drying out of freshwater lakes that had stood for millennia. This must have put tremendous pressure on the cattle nomads. The percentage of cattle remains from sites in this Later Pastoral period drop off in favor of small stock (Di Lernia 1999b). As noted above, this is also the time of domestication of African grain crops and the appearance of gardens attached to the semi-permanent villages of Niger (Paris 1992) and Mauretania (Holl 1993), with their associated cemeteries. Increase in rock art and possible use of ancestral graves for propitiation ceremonies are clues to a resurgence of religious belief resulting from these environmental pressures. Although cattle were central to religious and social life, they probably became psychologically more important as conditions deteriorated and animal numbers decreased. Their widespread depiction on the rock walls all across the Sahara suggests strong links over vast distances. These spaces were also inhabited by spirits who needed to be dealt with very carefully, or mishap might befall. Today, the Tuareg wear Koranic inscriptions wrapped in leather or encased in brass amulets (grisgris) to ward off the ‘‘evil-eye’’ or to protect the wearer against malevolent spirits. Thus we see in the present a blending of Islam with earlier spirit beliefs. The vast landscape of the Sahara allowed cultural contact from Mali to the Nile Valley as far back as 9,000 years ago. There appear to have been no barriers to movement, and even political entities which may have grown up did not deter culture transmission. An analogy might be made with Tuareg or Fulani herders today whose culture and languages are to be found over long distances. The environmental conditions allowed widespread appearance of lakes and rivers, permitting ease of movement, ﬁrst by hunter-ﬁshers and later by pastoralists. The distribution of wavy-line pottery ca. 9,000 bp is virtually the same as that of

_____________________________________________________________________________________________________

272

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a

diamond rocker-stamped wares ca. 6,000 bp, the former made by the lake and river dwelling hunters and the latter by pastoralists, such as the Tenerian people of Niger. The opening up of the Sahara for herding after 7,400 bp was made easier, not only by the convenient grassland niche, but also by lack of epizootic problems that later beset pastoralists as they tried to move southwards into tropical Africa. Tropical diseases such as Malignant Catarrhal Fever and East Coast Fever are transmitted by wild animal hosts endemic to the grasslands of eastern and southern Africa (Gifford-Gonzalez 2000). Thus, we can suggest that the tracks and water points opened up by the earlier hunter-ﬁshers were capitalized upon by later pastoral people. Every African landscape is redolent with myth and stories, passed on from generation to generation. It might be postulated that the geographical overlap between these two economic groups infers a transmission also of belief about the environment, and how to use it symbolically. These stories may well have been codiﬁed in the rock art. We are aware that the art is not purely narrative, but contains the deep meaning and belief systems of the artists. There is no reason to doubt that the same would have pertained with prehistoric Saharan rock art. The spread of decorated cattle paintings in the later bovidian period (after 5,000 bp) shows a focus on large stock that is supported by intentional burial of whole cattle under formal funerary architecture. Why should open desert environments spawn such belief systems? Anyone who has spent time in the desert can be totally overwhelmed by the experience, or be totally afraid of it. Either way, the open landscape holds the individual in thrall, and it is to desert spaces that ascetics go to commune with their gods or spirits. Africa has never been separated from the Arabian Peninsula and the Near East. There are suggestions based on work in northeastern Africa (Steve Brandt pers. comm.) that pre-Semitic languages and an African form of pastoralism were introduced to Arabia and the Near East 6,500 years ago. Both the Sahara and the Levant felt the increasing aridity after 4,500 bp (Hassan 2000). The proposition that drought conditions unsettled pastoral nomads in the Near East is suggested by the movements of the Patriarch Abraham from Ur of the Chaldees, in the Tigris– Euphrates Basin, to the Levant (Genesis 11: 31; 12: 5), then subsequently to Egypt (Genesis 12: 10). Is it purely a coincidence that the development of ideologies in the dry lands of the Sahara by nomadic people was also happening at the same time in the Levant? References Allard-Huard, L. and Huard, P. 1983: Les Gravures rupestres du Sahara et du Nil: Ii l’e`re pastorale. E´tudes scientiﬁques, 1983 (March–June). Bourdieu, P. 1973: The Berber House. In M. Douglas (ed.), Rules and Meanings, Harmondsworth: Penguin Books, 98–110. Bourdieu, P. 1977: Outline of a Theory of Practice. Cambridge: Cambridge University Press. Camps, G. 1961: Aux origines de la Berbe´rie: monuments at rites fune´raires protohistoriques. Paris: AMG.

_____________________________________________________________________________________________________

273

_____________________________________________________________________________

An d r e w B. S mith Camps, G. 1974: Les Civilisations pre´historiques de l’Afrique du Nord et du Sahara. Paris: Doin. Casajus, D. 1987: La Tente dans la solitude: la socie´te´ et les morts chez les Touaregs Kel Firwan. Cambridge: Cambridge University Press. Cavalli-Sforza, L. L., Menozzi, P., and Piazza, A. 1994: The History and Geography of Human Genes. Princeton, NJ: Princeton University Press. Clark, J. D., Williams, M. A. J., and Smith, A. B. 1973: The geomorphology and archaeology of Adrar Bous, central Sahara: A preliminary report. Quaternaria, 17, 245–97. Clutton-Brock, J. 1989: Cattle in ancient North Africa. In J. Clutton-Brock (ed.), The Walking Larder: Patterns of Domestication, Pastoralism, and Predation, London: Unwin Hyman, 200–6. Cruciani, F., Santolamazza, P., Shen, P., Macaulay, V., Moral, P., Olckers, A., Modiano, D., Holmes, S., Destro-Bisol, G., Coia, V., Wallace, D. C., Oefnet, P. J., Tottoni, A., CavalliSforza, L. L., Scozzari, R., and Underhill, P. A. 2002: A back migration from Asia to subSaharan Africa is supported by high resolution analysis of human Y-chromosome haplotypes. American Journal of Human Genetics, 70 (5), 1197–214. Desanges, J. 1981: The proto-Arabs. In G. Mokhtar (ed.), UNESCO General History of Africa vol. 2, London: Heinemann, 423–40. Di Lernia, S. 1999a: The Uan Afuda Cave: Hunter-Gatherer Societies of Central Sahara. Rome: Arid Zone Archaeology Monographs no. 1, Universita` degli Studi di Roma ‘‘La Sapienza.’’ Di Lernia, S. 1999b: Discussing pastoralism: The case of the Acacus and surroundings (Libyan Sahara). Sahara, 11, 7–20. Dupuy, C. 1998: Reﬂexion sur l’identite´ des guerriers repre´sente´s dans les gravures rupestres de l’Adrar des Iforas et de l’Aı¨r. Sahara, 10, 31–54. Gabus, J. 1952: Contribution a` l’e´tude des Ne´madi, chasseurs archaı¨ques du Djouf. Bulletin de la Socie´te´ suisse d’Anthropologie et d’Ethnologie, 52, 49–93. Gifford-Gonzalez, D. 2000: Animal disease challenges to the emergence of pastoralism in sub-Saharan Africa. African Archaeological Review, 17 (3), 95–139. Hanotte, O., Bradley, D. G., Ochieng, J. W., Verjee, Y., Hill, E. W., and Rege, J. E. O. 2002: African pastoralism: Genetic imprints of origins and migrations. Science, 296, 336–9. Hassan, F.A. 2000: Holocene environmental change and the origins and spread of food production in the Middle East. Adumatu, 1, 7–28. Holl, A. 1993: Late Neolithic cultural landscape in southeastern Mauretania: An essay in spatiometrics. In A. Holl and T. E. Levy (eds), Spatial Boundaries and Social Dynamics. International Monographs in Prehistory, Archaeology Series no. 2, Ann Arbor, MI, 95–133. Jacobsohn, M. 1995: Negotiating meaning and change in space and material culture: An ethnoarchaeological study among semi-nomadic Himba and Herero herders in northwestern Namibia. Unpublished PhD thesis, Rondebosch: University of Cape Town. Kuper, R. (ed.) 1978: Sahara: 10,000 Jahre Zwischen Weide und Wuste. Koln: Museen der Stadt. Lajoux, J.-D. 1963: The Rock Paintings of Tassili. London: Thames and Hudson. Lewis-Williams, J. D. 1981: Believing and Seeing: Symbolic Meanings in Southern San Rock Paintings. London: Academic Press. MacDonald, K. C. 1998: Before the Empire of Ghana: Pastoralism and the origins of cultural complexity in the Sahel. In G. Connah (ed.), Transformations in Africa: Essays on Africa’s Later Past, Leicester: Leicester University Press, 71–103.

_____________________________________________________________________________________________________

274

_____________________________________________________________________________

E v o l u t i o n o f Id e o l o g i e s i n A rid No r t h Afr ic a Mone`s, H. 1988: The conquest of North Africa and Berber resistance. In M. El Fasi and I. Hrbek (eds), UNESCO General History of Africa vol. 3, London: Heinemann, 224–45. Munson, P. J. 1980: Archaeology and the prehistoric origins of the Ghana Empire. Journal of African History, 21, 457–66. Norris, H. T. 1975: The Tuaregs: Their Islamic Legacy and its Diffusion in the Sahel. Warminster: Aris and Phillips. Oesterley, W. O. E. and Robinson, T. H. 1966: Hebrew Religion: Its Origins and Development. London: SPCK. Pardoe, C. 1988: The cemetery as symbol: The distribution of prehistoric Aboriginal burial grounds in southeastern Australia. Archaeology in Oceania, 23 (1), 1–16. Paris, F. 1990: Les Se´pultures monumentales d’Iwelen (Niger). Journal de la Socie´te´ des Africanistes, 60 (1), 47–74. Paris, F. 1992: Chin Taﬁdet village ne´olithique. Journal de la Socie´te´ des Africanistes, 62 (2), 33–53. Paris, F. 1996: Les Se´pultures du Sahara Nigerien a` l’Islamisation. Paris: ORSTOM. Paris, F. 2000: African livestock remains from Saharan mortuary contexts. In R. M. Blench and K. C. MacDonald (eds), The Origins and Development of African Livestock: Archaeology, Genetics, Linguistics, and Ethnography, London: University College of London Press, 111–26. Parker, M. P. 1999: The Archaeology of Death and Burial. College Station: Texas A&M University Press. Petit-Maire, N. and Riser, J. 1983: Sahara ou Sahel?: Quaternaire recent du Bassin de Taoudenni (Mali). Paris: CNRS. Rasmussen, S. J. 1992: Ritual specialists, ambiguity, and power in Tuareg society. Man, 27, 105–28. Roubet, C. 1978: Une E´conomie pastorale pre´-agricole en Alge´rie orientale: le Ne´olithique de tradition capsienne. L’Anthropologie, 82 (4), 583–6. Sadr, K., Castiglioni, A., Castiglioni, A., and Negro, G. 1994: Archaeology of the Nubian Desert. Sahara, 6, 69–75. Sansoni, U. 1994: Le Piu` Antiche Pitture del Sahara: L’arte delle Teste Rotonde. Milan: Jaca Books. Servant, M. and Servant-Valdary, S. 1980: L’Environnement quaternaire du basin du Tchad. In M. A. J. Williams and H. Faure (eds.), The Sahara and the Nile, Rotterdam: A. A. Balkema, 133–62. Smith, A. B. 1980: The Neolithic tradition in the Sahara. In M. A. J. Williams and H. Faure (eds.), The Sahara and the Nile, Rotterdam: A. A. Balkema, 451–65. Smith, A. B. 1992. Pastoralism in Africa: Origins and Development Ecology. London: Hurst. Smith, A. B. 1993: New approaches to Saharan rock art. Memoria della Societa` Italiana di Scienze Naturale, 26, 467–77. Smith, A. B. 1998: Keeping people on the periphery: The ideology of social hierarchies between hunters and herders. Journal of Anthropological Archaeology, 17 (2), 201–15. Smith, S. E. 1980: The environmental adaptation of nomads in the West African Sahel: A key to understanding prehistoric pastoralists. In M. A. J. Williams and H. Faure (eds.), The Sahara and the Nile, Rotterdam: A. A. Balkema, 467–87. Tilley, C. 1994: A Phenomenology of Landscape: Places, Paths, and Monuments. Berg: Oxford. Wendorf, F. and Schild, R. 1994: Are the early Holocene cattle in the eastern Sahara domestic or wild? Evolutionary Anthropology, 3 (4), 118–28.

_____________________________________________________________________________________________________

275

_____________________________________________________________________________

-----------------------------------------------------------------------------

15

--------------------------------------------------------------------------

Hunter-Gatherer Interactions with Sheep and Cattle Pastoralists from the Australian Arid Zone Alistair Paterson

Introduction Across the globe during the Holocene hunter-gatherers came into contact with farmers and herders. For desert dwellers this process was often delayed and took different forms due to ecological and environmental conditions speciﬁc to the world’s deserts. This chapter examines the interaction of Aboriginal people with sheep and cattle pastoralists in the Australian arid zone (also referred to as the drylands). In the continuum of these types of hunter-gatherer/pastoral interactions throughout the world the Australian example represents one of the most recent phases. It involved the extensive introduction of fully domesticated species (with no immediate wild ancestors) for commercial purposes and invoked massive environmental change. Pastoralism was linked to colonization and colonialism, and thus part of historical social and political processes. This chapter examines the arrival of pastoralists and the nature of Australian Aboriginal life at that time, beginning with a brief historical and geographical overview of sheep and cattle pastoralism. It considers interaction between Aboriginal people and settlers, and what happened to nomadic life and pre-contact settlement and subsistence. To explore processes of pastoralist and hunter-gatherer interaction, studies from archaeology, anthropology, and history are used to explore how Aboriginal people became involved (or not, as the case may be) in pastoral activities. This requires an understanding of how the different pastoral industries worked, and considering environmental, technical, and social parameters.

_____________________________________________________________________________________________________

276

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

Hunters and Foragers, Pastoralists, and Ranchers Firstly, it is necessary to consider the differences between hunting-foraging, pastoralism, and ranching. Here we are distinguishing between broad categories of human interactions with plants and animals, the models for which have developed from simple models to those accommodating more complicated dynamics (Harris and Hillman 1989). Australia was the last continent to be introduced to pastoralism; however, Aboriginal people have been present in this landscape for more than 40,000 years (Roberts et al. 1990; see also Chapters 3, 6, and 12, this volume). The ﬁrst domesticated animal was the dingo, having been introduced from island Southeast Asia during the mid-Holocene (Corbett 1985; Gollan 1984). No native species were domesticated by Australian Aborigines (Diamond 1998); however, there is evidence for complex human interactions with the environment through seed grinding, replanting of seeds and tubers, and the use of ﬁre (e.g., Jones 1980; Jones and Meehan 1989; Latz 1995; Singh et al. 1981; Smith 1989; Yen 1989). What do we mean by pastoralists? In Pastoralists at the Periphery: Herders in a Capitalist World, Chang and Koster (1994) review archaeological studies of modern and ancient pastoralists. The term pastoralist is often applied to ‘‘those who live by relying on the products of their herd animals in regions of Asia and Africa where a settled agricultural life is deemed too risky’’ (Chang and Koster 1994: 1). Yet in Australia (and other settler nations such as the Americas, New Zealand, and Canada) there is a more recent form of pastoralism, where Europeans established pastoral industries as capitalist and colonizing enterprises to provide commodities within a global network to industrialized states backed by international investment. This considerably broadens the notion of a pastoral mode of production. With this in mind, Chang and Koster (1994: xi) invited archaeologists to ‘‘investigate pastoralism in more recent periods such as the nineteenth century. Such a reorientation would allow archaeologists to use a rich empirical database to address important questions concerning pastoral adaptation.’’ Clutton-Brock (1988: 117) highlighted differences between sheep pastoralism and cattle ranching: The rancher loosely owns herds of animals for exploitation of meat and other resources that are often marketed. In origin the animals may be wild, feral, or domestic but they live as wild animals except that their territory is usually restricted.

Protection of sheep (and goats) from predators and during lambing was provided by shepherds or well-maintained fenced paddocks. In Australia, cattle required less care and may only come into contact with humans annually, or less often (McLaren 2000). Cattle were typically raised for their meat value, a process which required mustering, overland transport to market by droving (in the era before trucks), and butchery. The most common sheep product was wool ﬂeece, which required keeping the sheep alive, although during drought and price slumps mutton and tallow could also be proﬁtable. There was also local use of animals for meat

_____________________________________________________________________________________________________

277

_____________________________________________________________________________

Al i s t a i r P a t e r s o n

and dairy products, a distinction possibly apparent in a herd’s age and sex structure: ‘‘A meat herd will contain a high number of adolescent and young adult animals, whereas a dairy herd will consist mainly of adult females’’ (Renfrew and Bahn 1996: 278). This would be true, for example, for goats, which were present at some pastoral properties to provide dairy produce.

The Arrival of Pastoralists in Australia British settlement from 1788 heralded the introduction of farming. While the dominant animals introduced to Australia were sheep and cattle, other animals were important – namely goat, pig, camel, buffalo, donkey, and horse. The last four were largely used for transport and cartage. Additionally, European arrival saw the introduction of other species including rabbits, foxes, cats, and dogs. Sheep were well suited for semi-arid areas of Australia where other forms of crop farming were unsuitable, and cattle were particularly suited to the wetter northern regions. The history of the Australian pastoral cattle industry is well documented (Duncan 1967; McLaren 2000), as is the sheep industry (Barnard 1962; Bean 1910; Collier 1911; Culey 1970; Munz 1950). In brief, Australian drylands are away from the well watered coastal regions in southeastern Australia that were initially settled by Europeans (Heathcote 1987). In the interior is a zone of intermittent surface waters and interior deserts where surface waters are rare. Importantly for pastoralism, large artesian basins of subterranean water exist which were accessed from the 1880s onwards using bores. Sheep and cattle industries are restricted in their distribution (see Figure 15.1). This distribution would have changed over time as land suitable for crop production was taken up, as droughts pulled back the frontier, as grasslands changed, and as the limits of domesticated species and the carrying capacities of different environments were determined. There were great differences in grazing quality. The ﬁnest was where 0.4 hectares of pasture supported one sheep. In poor seasons and on more marginal land there existed poorer pastures where up to 8 hectares were required for each sheep. Seasonal variability impacted on pastoral industries. For example, in October 1859, at the Strangways Springs Station in the arid Lake Eyre Basin, Howitt (1859) wrote bitterly: ‘‘the country is perfectly worthless and would not feed a bandicoot.’’ During extensive droughts ﬁve years later the pastoral frontier retreated from arid regions; stations failed and sheep, cattle, Aboriginal people, and pastoralists fell back to dwindling waters. At Strangways Springs in 1865 the station manager, Jeffreys (Letter to Warren, July 6, 1865), wrote: ‘‘This terrible visitation has completely destroyed the prospects of our beautiful station.’’ Yet, in March the following year, he wrote (Letter to Warren, March 18, 1866): Now I have news to tell you. It is the most beautiful station we have, equal to any country I ever saw in Queensland, New South Wales, Victoria or South Australia. The pasture is so magniﬁcent . . . Where is another station in the colony where you can

_____________________________________________________________________________________________________

278

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

Figure 15.1 Distribution of sheep and cattle properties (after Cantrill 1963: 96; Heathcote 1987; O’Loghlen 1963: 295; Scott 1987). put nearly 4,000 sheep in one ﬂock? . . . In six weeks I will guarantee [the sheep] will be as fat as pigs.

It would take the introduction of artesian bores from the 1880s to provide more reliable water supplies, though not necessarily improved grazing. Pastoral expansion into the Australian drylands intensiﬁed from the 1840s, with overlanding of sheep, cattle, and horses into South Australia (SA) and Queensland (QLD). Pastoralists quickly followed on the heels of early explorers and surveyors, whetted by descriptions of inland grasslands. By the 1860s settlement had crossed from western QLD into the Gulf Country and onto the Barkley Tablelands via central Australia and QLD into the Northern Territory (NT) in the 1870s, by the mid-1850s into the Pilbara (WA), and into the Kimberley region in the 1880s. It should be noted that my focus is Australian drylands, and I shall not consider pastoralism elsewhere, such as in Tasmania and southeastern Australia. However, it should be remembered that there was access to a convict (indentured labor) workforce in these two instances, which was not the case for central and northern Australia.

_____________________________________________________________________________________________________

279

_____________________________________________________________________________

Al i s t a i r P a t e r s o n

The Europeans’ knowledge of resources and Australian conditions improved over time, changing introduced herding practices, and setting the scene for settlement of drylands. After the 1830s European settlers were accustomed to bush life. Stockmen were self-sufﬁcient, most life was spent in camps away from homesteads, supplies were delivered by wagon to outstations, and the techniques of controlling livestock without fences and permanent yards established. Knowledge of water is essential in Australian drylands, and pastoralists learned the value of saltbushes versus grasses and herbs, and the value of mulga (Acacia aneura) and kurrajong (Brachychiton) foliage during drought. Other plants proved poisonous, and restricted the distribution of animals.

Transitions Several transitions in the sheep and cattle industries are reﬂected in the archaeological record. First, over time the pastoral stations were built up from small ventures to more permanent industries. This is reﬂected in the material remains including fences; wells, tanks, dams, artesian bores, and drains; yards; head stations and outstations; wool scours, wool sheds, and shearing sheds; roads, tracks, and fords (ICOMOS 1995; Woodhouse 1993). Second, there were new technologies and techniques over time. After the 1870s local innovations and labor were replaced with imported materials, technologies, and labor (Heathcote 1987); however, there was variation in this general trend. Two examples of technology transforming the pastoral domain were the introduction of artesian bores and fencing wire. These allowed the establishment of large fenced paddocks, which changed the organization of pastoral work. There was much variation in the use of fencing; by 1900 most sheep properties had boundary fences, although there was not widespread use of internal fences until the 1960s. In the NT cattle country in the 1960s, 90 percent of properties were not fenced, and in the Gulf Country in 1979 only 11 percent of properties were fenced (Heathcote 1987). Third, shepherding declined over time. On sheep stations shepherding was practiced to protect against predators (dogs and Aboriginal hunters). However, over time, labor intensive shepherding stopped, a transition facilitated by a shift from open range to closer paddocking with surveillance by boundary riders. Fourth, there were differences in the distribution of sheep and cattle over time (see Figure 15.1). This distribution reﬂects the fact that sheep were found to be unsuited to what would become cattle country in northern Australia. For example, in the Kimberley region and NT in the late nineteenth century sheep had low breeding rates and were common victims of native dogs. Additionally, the expense of carting ﬂeece and difﬁculties with ﬁnding shepherds and shearers made sheep non-viable in northern regions. Nearly all sheep properties on the Barkley Tablelands and Kimberley were abandoned by the early twentieth century. Fifth, there were key differences between the handling of cattle and sheep. The cattle muster occurred only once or twice a year, a much lower rate of animal

_____________________________________________________________________________________________________

280

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

Figure 15.2 Aboriginal rock painting of pack horses and pastoralists on Wiluna Station (from Mulvaney 1989: plate 8).

handling than in other contexts such as North American rangelands (McLaren 2000; Osgood 1957), while sheep required constant shepherding unless in well fenced paddocks.

Interaction between Aboriginals and Pastoralists: Three Regional Studies In semi-arid and arid Australia the arrival of herders represented the ﬁrst sustained presence of Europeans, occurring 60 years after permanent European settlement in eastern Australia and preceded by diseases that may have disrupted Aboriginal health, population, and society (Reynolds 1972). The pastoral industries competed with Aboriginal people for access to land and resources, with economic, environmental, social, and political consequences. The eventual entry of new animals and people needed to be dealt with by Aboriginal people; Figure 15.2 shows animals herded across an inland station represented in a rock painting. New animals surprised Aboriginal people. The ﬁrst European to cross Australia, Stuart (as cited in Hardman 1964: 86–7), describes meeting Arabana people in Warriners Creek in 1858: [I] saw some natives walking along a valley . . . I hailed them and an old man came up to me. He was rather frightened, and trembled a great deal . . . I made [them]

_____________________________________________________________________________________________________

281

_____________________________________________________________________________

Al i s t a i r P a t e r s o n understand that I wanted water . . . I expected they were going to take us to some springs, and was disappointed when they showed us some rain water in a deep hole. They were quite surprised to see our horses drink it all.

New animals were to compete for resources with Aboriginal people. Aboriginal hunting caused dispute across Australia (Reynolds 1972), reducing stock and inducing panic; yet hunting decreased over time (Davidson 1988). In May 1865 Strangways Springs Station manager Jeffreys (Letter to Warren, May 28, 1865) wrote to his employer: ‘‘You will be glad to hear that there has been no sheep hunting for two months.’’ Violence played a role in the decrease in hunting. There are accounts of extreme violence on and near pastoral stations, some of the worst in the earliest days, but continuing into the twentieth century and constituting the colonial legacy. The historiography of European–Aboriginal interaction is contested (Attwood and Foster 2003), yet clearly the frontier was sometimes horriﬁc. To date, our understanding of Aboriginal involvement with European pastoralists has largely been based on historical sources and oral accounts. Oral histories assist our understanding of pastoralism, as some white and black pastoral workers survive to remember earlier days. This is less true for regions where pastoralism was introduced earlier, and much knowledge of the early days of pastoral settlements is gone. The pastoral industry often involved Aboriginal people as the earlier industries were labor intensive and the Europeans required knowledge of the environment and country. One Aboriginal veteran of the pastoral industry describes those times: All right, olden time people, whitefella, you know, for cattle, on the country, and no bore, nothing at all. Old blackfella, old blackfella like me, they used to bring out all whitefella mob of cattle and show rock hole, so we can give to cattle drink, or horses or nanny goats. Well that’s why olden time people, Aboriginal people used to show that rock hole, so he can bring is cattle, you know, give them a drink . . . You can’t walk. You can’t drive your car, only the horses and cattle. That’s all. Well, old blackfella know the rock hole. So, old blackfella used to work for whitefella, old whitefellas drive out the cattle. Well old blackfella used to show the rock hole. So they can give their cattle drink. Because no bore, nothing at all. No bore. (M. Kennedy 2002: pers. comm.).

An emerging research area is the archaeology of interaction between settlers and Indigenous Australians (see Murray 2002). The pastoral domain is an essential element of this topic because sustained interaction occurred on and near sheep and cattle stations, and many Aboriginal people became involved in these industries (McGrath 1987). Aboriginal people describe the pastoral industries as a central part of their life (e.g., Gill et al. in press; Rose 1991). However, there has been little archaeological research in arid contexts into white pastoralists’ interaction with Aboriginal people. This perhaps reﬂects the difﬁculty of the task: such studies require access to multiple sites within pastoral stations, most of which were – and some still are – very large. It is more manageable to focus more on localized sites, such as missions or townsites, in order to interpret forms of cultural interaction

_____________________________________________________________________________________________________

282

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

(e.g., Birmingham 2000). Important research into white–black interaction has occurred outside of the Australian drylands (and thus the topic of this chapter) but remains important for building method and theory more broadly. Three regions are explored here: the northern region of SA, the Kimberley region in northwest Australia, and the central NT. Some of the earliest pastoral properties in the central Australian drylands were in northern SA, where sheep stations were established after the 1850s. In one instance, at a set of artesian springs located between a stony plain and linear duneﬁelds, the Strangways Springs Station was established in 1862. This place had been used for millennia by Aboriginal people, who know it as Pangki Warruna. An analysis of archaeological sites across the station and written reports from station managers provided insight into the early years of the station (Paterson 2000, 2003). The nature of cultural interaction changed as the pastoralists adapted and transformed their economic and social behavior in response to the harsh environment and to economic and technical parameters related to the ﬂedgling pastoral industry. This historical archaeological project revealed two phases of interaction in the period ad 1862–1900. Phase 1 saw pastoral outstations established at artesian springs – these had been the main sites for Aboriginal occupation before and following European arrival. The outstations were simple huts and had brush yards to keep the ﬂocks at night. The pastoralists attempted to improve the ﬂow of water from the artesian springs, and were forced to rely on the pastures near the springs. In Phase 2, from the 1880s onwards, bores were used to access the artesian basin. Once that happened, it was possible to move the ﬂocks to new locations and to better use grasslands and herbage. Fencing decreased the need for shepherds. The evidence from campsites and worksites demonstrates differential Aboriginal involvement in the nineteenth-century pastoral domain. Some outstations indicated Aboriginal engagement (through the presence of artifacts, food processing equipment, and ocher). Over time, large camps were present at the head station, probably created by Aboriginal seasonal workers during the wool washing following shearing, and animal husbandry during the lambing season. These camps may have been rationed from the head station, supplied by both private and government rations. This research provided insight into the role of Aboriginal people on the sheep station. There was seasonal work (lambing, wool scouring) and more permanent work (animal herding and husbandry). The key unit of work was the family, and children and women were important workers. Some Aboriginal people were located within the pastoral domain, as seasonal or permanent workers. These people were described as receiving rations from the pastoralists on occasion. Other Aboriginal people were clearly ‘‘outside’’ the pastoral domain according to the pastoralists, who sometimes repelled ‘‘wild’’ blacks with force. Yet, on occasion, ‘‘outsiders’’ visited the station to involve the Aboriginal pastoral workers in ceremonial events. This case study shows that archaeological and historical evidence each provide different perspectives on past culture contact and human agency. Overall the evidence from Strangways Springs Station suggests Aboriginal ‘‘insiders’’ to the pastoral domain who were present at outstations and the head station with access

_____________________________________________________________________________________________________

283

_____________________________________________________________________________

Al i s t a i r P a t e r s o n

to ‘‘new’’ material culture and rationed items (e.g., food, tobacco). The impact of new foods such as ﬂour is indicated by a decrease in seed grinding equipment in Aboriginal camps, which may also suggest ‘‘freeing up’’ of time for pastoral work by women. Both men and women worked with the introduced animals, with some trained when young. Presumably there is an archaeological record of ‘‘outsiders’’ to the pastoral domain, although that was not revealed in this research, except perhaps by imported ocher found in shepherds’ huts. In the Kimberley region Fullagar and Head (1997, 1999), Smith (2002; Smith and Smith 1999), and Harrison (2002a, 2002b, 2002c) have used archaeological data to explore changes resulting from the presence of pastoralists, such as seen in settlement location, rock art, tool assemblages, plant uses, and diet. Changes seen in use of new materials, technologies, foods, and settlements are assumed to reﬂect social changes and continuities. These studies reveal that Aboriginal people were able to reafﬁrm cultural traditions and attachment to place. Overall, it becomes clear that the pastoral domain was created by both European and Aboriginal participants; within pastoral economies rationing of foods, tobacco, and clothes, rather than monetary payment, was widespread for Aboriginal workers. Elsewhere in northern Australia pastoralism was introduced later (Duncan 1967). A study of the pastoral industries in the Murchison and Davenport Ranges (northern central Australia) examines the late nineteenth and twentieth centuries (Gill et al. 2002, in press; Paterson et al. 2003). The project was supported by an Aboriginal elder and veteran of the pastoralist industry (M. Kennedy) and a geographer (Dr Nicholas Gill). In this research the analysis of archaeological sites, combined with oral histories, biographical mapping, and archival research, was used to recreate individual and group work histories across the landscape. This provided a ﬁne-grained insight into Aboriginal involvement with pastoralists and vice versa. The project revealed potential nodes for archaeological research within the pastoral domain, such as campsites used by Aboriginal people, head stations, and small camps run by non-pastoralists such as missionaries, miners, travelers, and butchers, to name but a few. There is material evidence for a range of past functions in this landscape, including ceremonial sites; religious places; yards for horses, cattle, pigs, or goats; huts and small shelters used for overnight camps; tracks for droving cattle; camel pads; horse or donkey trails; motor vehicle tracks; trucking yards; cattle dips; watering sites (rock holes, waterholes, rivers, wells, springs, and bores). These form a network of material evidence in the landscape to be compared with historical and oral accounts. In the Murchison Ranges, as elsewhere, pastoral activities ranged from larger mainstream pastoral ventures – which often still exist today – to smaller concerns. More marginal pastoralism existed on poor country such as in rocky rangeland and often involved Aboriginal people to a greater degree. It appears that in the twentieth century mixed descent men ran cattle and were granted grazing licenses, albeit rarely. This indicates that some Aboriginal people were running their own cattle ventures within one or two generations of the arrival of white pastoralists, while most pastoral stations relied to some extent on Aboriginal labor. Aboriginal work

_____________________________________________________________________________________________________

284

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

in the Murchison Ranges was diverse. Some camps represented the remains of Aboriginal people working with white pastoralists, such as those near yards for animals. Other camps networked with local towns, such as at Kelly Well, where Aboriginal people raised pigs and kept cattle for town butchers. Over time Aboriginal people moved around stations according to factors such as pay and rations, whether pastoralists were too violent, ceremonial and family obligations, the quality of station managers, and government policies. Women and men’s work differed. Generally, women were often responsible for the goat herds, while men worked with horses and cattle. Some women worked at the pastoral head stations as domestic laborers. Following the return of land, some Aboriginal people ran their own stock. Today, in the Murchison Ranges the Aboriginal people have some of the land and they run it with cattle and horses, although much of the interest in and knowledge of pastoral work is being lost as the older people die.

What Happened to Nomadic Life and Pre-Contact Settlement and Subsistence? Aboriginal hunter-foragers were inevitably altered by European settlement. This was immediate in some instances; however, Aboriginal subsistence continued in many cases, although restricted by some station owners. However, as late as 1951 in the Murchison Ranges, Patrol Ofﬁcer Sweeney wrote about Kurundi Station, on the edge of the Barkley Tablelands: ‘‘I consider natives on this station are still nomadic, the main boy being walkabout from Phillip Creek,’’ and station manager Driver wrote at the same time that not all Aboriginal people on the station were employees and ‘‘generally live a normal happy life hunting, ﬁshing etc. . . . and I expect more to come in . . . I don’t think they would be happy anywhere else’’ (F133 1963/488 Kurundi Station 1951–63). Aboriginal involvement with the pastoral properties depended on the availability of work, the weather, and what other places were available for them. In 1952 Driver (F133 1963/488 Kurundi Station 1951–63) wrote about the drought at Elkedra: ‘‘As a result of no tucker and no water the natives are crowding in the few available waters and looking for assistance from the station owner.’’ Another aspect of traditional life that changed was the ability to travel and conduct trade, potentially indicated by the presence of traded goods in archaeological sites. For example, in the Kimberleys the movement of artifacts is revealed by a speciﬁc type of artifact – the Kimberley point – a bifacially ﬂaked projectile point often made from bottle glass (Fullagar and Head 1997, 1999; Harrison 2002a). Another example is provided by ocher traded throughout the Lake Eyre Basin in a network with other items, such as stone grinders, pituri, and information (McBryde 1987, 2000). Pastoral properties became staging points in trade networks, as recalled at Strangways Springs by the station manager (Oastler 1908): The Aborigines had a custom of going periodically, in large and well-armed parties, usually 50 to 60 strong, to an Aboriginal red ocher mine, situated at Parachilna Creek.

_____________________________________________________________________________________________________

285

_____________________________________________________________________________

Al i s t a i r P a t e r s o n This ocher was to the Aborigines what gold is to the Whites. And consequently they would travel long distance and go through great hardships to obtain it. To illustrate this, I once weighed at Strangways Springs the loads carried by each member of those parties [returning] from Parachilna Creek, a distance of some 213 miles, and they weighed from 60 to 80 pounds.

For the Kimberley region, Fullagar and Head (1997) describe that during the wet season when there is little opportunity for pastoral work there was an increased occupation in rock shelters in remote locations. Seasonality should be evident for Aboriginal involvement with pastoralists in other regions. In southern central Australia the working year was punctuated by high demands for pastoral labor, such as for wool washing and during the lambing, which was met by Aboriginal people at Strangways Springs (and presumably elsewhere): ‘‘All my Black lambminders have come back to me as promised’’ (Oastler to Warren, March 4, 1868). Some Aboriginal people on sheep stations in Australian drylands became client herders. The shepherding work involved males and females, and children and adults. For example, at Strangways Springs European shepherds were commonly described as having been assisted by Aboriginal people. This is evidenced by the archaeological remains of outstations, where the assemblages in huts suggest both Aboriginal and European usage, as supported by historic accounts. Soon after the establishment of the pastoral station Aboriginal people were caring for animals – not just sheep, but also goats, horses, and bullocks. For example, according to letters there were Aboriginal girls shepherding at Strangways Springs in 1864, two years after the pastoralists ﬁrst arrived. In 1866 Jeffreys (March 5, 1866) described the detrimental effect of the drought on the bullocks and horses, which required treating with tar to help heal wounds caused by ﬂy bites: [Jeffreys] gave them in charge to Annie, Kalli Kalli’s little sister, one of the little shepherdesses. She looks after them with great care and musters them every day to a rain water hole three miles from here [Strangways Springs].

In the Murchison Ranges in the twentieth century Aboriginal people became proﬁcient with not just cattle, but also donkeys, camels, goats, and horses. For example, Marks (NTRS 226 TS 289, 1980) remembers Aboriginal women shepherding goats: The goats were taken away every morning. They were penned up at night and brought in about half a mile from the homestead and penned up there in log yards, railing type yards, because the dingoes would kill them otherwise. But every morning they were taken out and walked out to a waterhole about 5 miles away. They fed there all day with these goat shepherds and they walked back at night. The kids were left locked up in the yard and the goats were milked and the milk was brought up to the homestead.

The archaeological records of this herding process include outstations, camps, and yards; such sites need to be the focus of future archaeological research into

_____________________________________________________________________________________________________

286

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

Aboriginal care of animals and the gendered nature of client herding in Australia. Aboriginal people also assisted in the development of pastoral technologies; sites such as yards or droving camps potentially reﬂect Aboriginal work.

Approaches to Pastoral Domains as Settings for Cross-Cultural Interaction For southern Africa, Sadr (2002: 28) asked ‘‘how does one actually distinguish encapsulated or subjugated Bushmen from independent hunter-gatherers in the archaeological record?’’ Like Africa, the material signature of Australian Aboriginal peoples’ interactions with pastoralists is often difﬁcult to determine (Colley and Bickford 1996). One problem identiﬁed by Sadr, as well as by Colley and Bickford, is that the material signature of cross-cultural processes is often faint. However, often ‘‘new’’ material culture, or shifts in subsistence, are visible. For example, changes in diet may be indicated by the presence of new foods (for Aboriginal people, sheep and cattle remains) or tin cans, or by the absence of ‘‘traditional’’ food processing and hunting equipment. However, simple measures of the presence of European or settler material culture rely on ‘‘acculturation’’ as a framework for evaluating culture contact. Acculturation theories are diverse, yet have been criticized for being Eurocentric, one sided, and oriented towards trait lists; for treating cultures as singular ‘‘entities’’; for not emphasizing agency; and for not accommodating power structures inherent in colonial contexts (Cusick 1998). This is particularly true when relying on material culture in isolation from other forms of evidence, or when the taking up of European elements by Aboriginal people is viewed as a direct reﬂection of Aboriginal involvement in European social, political, and economic spheres. What does it mean when Aboriginal people do not use new material culture? It remains difﬁcult to disengage acculturation from assimilationist explanatory models. There are several other difﬁculties with acculturation. It has tended to be applied at the nodes or centers of settler activity, such as head stations, yet interaction occurs at many different locales as both pastoralists and Aboriginals move around the landscape. The availability of ‘‘new’’ material culture may be limited by external non-cultural factors such as the cost of goods, transport infrastructure, ﬁnancial resources, government expenditure, and policy concerning rations. Explanatory models also need to take into account site formation factors. For example, NT government records indicate that Aboriginal camps, particularly those close to pastoral head stations, were sometimes ‘‘cleaned’’ with bulldozers and ﬁre. Thus, the presence of new material culture is applicable as a rude indicator of contact, yet better approaches are required to interpret the nature of culture change and interactions. The concept that there is a gradual assimilation of settler lifeways by Aboriginal people seems overly simplistic, if not ﬂawed. Another important theoretical concept is ‘‘resistance.’’ For archaeologists, resistance refers to conscious human decisions to oppose the domination by another group (Orser 2002: 476). Archaeological evidence is effective in illustrating covert

_____________________________________________________________________________________________________

287

_____________________________________________________________________________

Al i s t a i r P a t e r s o n

resistance, as shown by Birmingham (1992) at Wybalenna Mission whose concept of ‘‘cultural accommodation’’ already built on existing Australian literature (such as Reynolds 1972). Resistance has the potential to link actions by different cultures. The frontier was sometimes a setting for resistance and violence, as it was in other countries. For example, this can be seen in the remains of fortiﬁed European outposts, such as the Barrow Creek Telegraph Station (NT). More recently, twentieth-century ‘‘walk-off’’ sites represent Aboriginal resistance to contemporary working conditions and power structures, and are subject to archaeological analysis (Deacon 2002). Other approaches have explored the evidence for social and economic stability in pastoral domains. For example, in the Kimberley region, Rowse (1987) identiﬁed how pastoralists provided rations, work, and patronage to Aboriginal people who then accommodated the needs of the pastoralists. This ‘‘cooperation’’ is tested by archaeological research, such as that cited above for the Kimberley, which explores the archaeological correlates for the pastoral domain, as well as testing longer term transformations in Aboriginal life extending from prior to the arrival of pastoralists (see Fullagar and Head 1997, 1999; and Harrison 2002c) in a similar way to Sadr (2002). Similarly, the term ‘‘shared histories’’ (Murray 1996) characterizes a range of studies which explore a previously understudied aspect of Australian history, namely Aboriginal contributions and experiences following European arrival. ‘‘Shared history’’ describes the intention to explore both Aboriginal and non-Aboriginal histories, yet requires explanatory frameworks to interpret cultural processes and methodology to link archaeological data to other evidence (e.g., historical sources, oral histories), as stated by Murray (2002). The concept of landscape is useful for archaeological studies of cross-cultural interaction, as it provides the potential to interpret the spatial organization of the environmental, economic, technical, and social settings for pastoralism. Additionally, mapping different Aboriginal and settler realms in ‘‘landscape formats’’ provides a potential to link different data (archaeological, oral, documentary, and environmental). GIS are potentially valuable in this regard in linking ﬁnegrained site-based data to regional data. Cross-cultural behavior at group and individual level may be linked into environmental histories – ‘‘ecohistories’’ (Hardesty and Fowler 2001) – to form ‘‘landscape biographies.’’ For example, in the Murchison Ranges project referred to above the material from archaeological sites is used with oral histories, biographical mapping, and archival research to recreate individual and group work histories across this landscape. In this, an Aboriginal elder’s life provides a microstudy mapping the ranges over time, and forms a ‘‘biography’’ of a historical landscape and social processes. Additionally, oral data assist our interpretation of the past uses and current distribution of material evidence. In conclusion, this chapter shows that the arrival of pastoralists in Australia resulted in extensive incursions into Aboriginal land, and was linked to a colonial society. The archaeological understanding of colonial enterprises and intentions is still relatively recent, although recent methodological and explanatory approaches described here may assist future interpretation.

_____________________________________________________________________________________________________

288

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a

References Attwood, B. and Foster, R. (eds) 2003: Frontier Conﬂict: The Australian Experience. Canberra: National Museum of Australia. Barnard, A. 1962: The Simple Fleece: Studies in the Australian Wool Industry. Melbourne: Melbourne University Press in association with The Australian National University. Bean, C. E. W. 1910: On the Wool Track. London: A. Rivers. Birmingham, J. 1992: Wybalenna: The Archaeology of Cultural Accommodation in Nineteenth Century Australia. Sydney: Australian Society for Historical Archaeology. Birmingham, J. 2000: Resistance, creolization, or optimal foraging at Killalpaninna Mission, South Australia. In R. Torrence and A. Clarke (eds), The Archaeology of Difference: Negotiating Cross-Cultural Engagements in Oceania, London: Routledge, 360–401. Cantrill, A. K. 1963: Sheep. In A. H. Chisholm (ed.), Australian Encyclopedia vol. 8., Sydney: Grolier Society of Australia, 89–98. Chang, C. and Koster, H. A. 1994: Pastoralists at the Periphery: Herders in a Capitalist World. Tucson: University of Arizona Press. Clutton-Brock, J. 1988: Introduction to pastoralism. In J. Clutton-Brock (ed.), The Walking Larder: Patterns of Domestication, Pastoralism, and Predation, London: Unwin Hyman, 115–18. Colley, S. and Bickford, A. 1996: ‘‘Real’’ Aboriginals and ‘‘real’’ archaeology: Aboriginal places and Australian historical archaeology. World Archaeological Bulletin, 7, 5–21. Collier, J. 1911: The Pastoral Age in Australasia. London: Whitcombe and Tombs. Corbett, L. K. 1985: Morphological comparisons of Australian and Thai dingoes: A reappraisal of dingo status, distribution, and ancestry. Proceedings of the Ecological Society of Australia, 13, 277–91. Culey, A. G. 1970: An Australian Bibliography on the Biology of the Sheep and the Sheep Industry, 1925–1967. Melbourne: Commonwealth Scientiﬁc and Industrial Research Organization. Cusick, J. G. 1998: Historiography of acculturation: An evaluation of concepts and their application in archaeology. In J. G. Cusick (ed.), Studies in Culture Contact: Interaction, Culture Change, and Archaeology, Carbondale: Center for Archaeological Investigations, Southern Illinois University, 126–45. Davidson, I. 1988: Escaped domestic animals and the introduction of agriculture to Spain. In J. Clutton-Brock (ed.), The Walking Larder: Patterns of Domestication, Pastoralism, and Predation, London: Unwin Hyman, 59–71. Deacon, J. 2002: A spatial analysis of Ngurrutji, an Aboriginal historic site in central Australia. Unpublished BA (Hons) thesis, Perth: University of Western Australia. Diamond, J. 1998: Guns, Germs, and Steel: A Short History of Everybody for the Last 13,000 Years. London: Vintage. Duncan, R. 1967: The Northern Territory Pastoral Industry, 1863–1910. Melbourne: Melbourne University Press and Monash University. Fullagar, R. and Head, L. 1997: Hunter-gatherer archaeology and pastoral contact: Perspectives from northwest Northern Territory, Australia. World Archaeology, 28 (3), 418–28. Fullagar, R. and Head, L. 1999: Exploring the prehistory of hunter-gatherer attachments to place: An example from the Keep River area, Northern Territory, Australia. In P. J. Ucko and R. Layton (eds), The Archaeology and Anthropology of Landscape: Shaping Your Landscape, London: Routledge, 322–35. Gill, N., Paterson, A., and Kennedy, M. 2002: Hidden landscapes. Cultural Survival Quarterly, 26 (2), 37–8.

_____________________________________________________________________________________________________

289

_____________________________________________________________________________

Al i s t a i r P a t e r s o n Gill, N., Paterson, A., and Kennedy, M. in press: ‘‘Murphy, do you want to delete this?’’ Hidden histories and hidden landscapes in the Murchison and Davenport Ranges. In G. Ward (ed.), Power of Knowledge and the Resonance of Tradition: AIATSIS Conference 2001, Canberra: Aboriginal Studies Press. Gollan, K. 1984: The Australian Dingo: In the shadow of man. In M. Archer and G. Clayton (eds), Vertebrate Zoogeography and Evolution in Australasia, Perth: Hesperian Press. Hardesty, D. L. and Fowler, D. D. 2001: Archaeology and environmental changes. In C. L. Crumley (ed.), New Directions in Anthropology and Environment: Intersections, Walnut Creek, CA: AltaMira, 72–89. Hardman, W. (ed.) 1964: Explorations in Australia: The Journals of John McDouall Stuart During the Years 1858, 1859, 1860, 1861, and 1862 When He Fixed the Centre for the Continent and Successfully Crossed it from Sea to Sea. London: Saunders Otley. Harris, D. R. and Hillman, G. C. (eds) 1989: Foraging and Farming: The Evolution of Plant Exploitation. London: Unwin Hyman. Harrison, R. 2002a: Archaeology and the colonial encounter: Kimberley spearpoints, cultural identity, and masculinity in the north of Australia. Journal of Social Archaeology, 2 (3), 352–77. Harrison, R. 2002b: Australia’s Iron Age: Aboriginal post-contact metal artifacts from Old Lamboo Station, southeast Kimberley, WA. Australasian Historical Archaeology, 20, 67–76. Harrison, R. 2002c: Shared histories and the archaeology of the pastoral industry in Australia. In R. Harrison and C. Williamson (eds), After Captain Cook: The Archaeology of the Recent Indigenous Past in Australia, Sydney: Sydney University Archaeological Methods Series vol. 8, Archaeology Computing Laboratory, 37–58. Heathcote, R. L. 1987: Pastoral Australia. In D. N. Jeans (ed.), Australia–A Geography, Vol. 2: Space and Society, Sydney: Sydney University Press, 259–300. Howitt, A. W. 1859: Diary kept by A. W. Howitt on journey to Davenport Range September 21, 1859–Oct. 31, 1859. Melbourne: La Trobe University Library. ICOMOS 1995: Pastoral Technology and the National Estate. Canberra: Australian Heritage Commission. Jones, R. (ed.) 1980: Northern Australia: Options and Implications. Canberra: The Australian National University. Jones, R. and Meehan, B. 1989: Plant foods of the Gidjingali: Ethnographic and archaeological perspectives from northern Australia on tuber and seed exploitation. In D. R. Harris and G. C. Hillman (eds), Foraging and Farming: The Evolution of Plant Exploitation, London: Unwin Hyman, 120–35. Latz, P. K. 1995: Bushﬁres and Bushtucker: Aboriginal Plant Use in Central Australia. Alice Springs: IAD Press. McBryde, I. 1987: Goods from another country: Exchange networks and the people of the Lake Eyre Basin. In D. J. Mulvaney and P. W. White (eds), Australians to 1788 vol. 1, Sydney: Fairfax, Syme, and Weldon, 253–73. McBryde, I. 2000: Travelers in storied landscapes: A case study in exchanges and heritage. Aboriginal History, 24, 152–74. McGrath, A. 1987: ‘‘Born in the Cattle’’: Aborigines in Cattle Country. Sydney: Allen and Unwin. McLaren, G. 2000: Big Mobs: The Story of Australian Cattlemen. South Fremantle: Fremantle Arts Centre Press.

_____________________________________________________________________________________________________

290

_____________________________________________________________________________

H un ter- G a t h e r e r – p a s t o r a l i s t c o n t a c t s i n A u s t r a l i a Mulvaney, D. J. 1989: Encounters in Place: Outsiders and Aboriginal Australians 1606–1985. St Lucia: University of Queensland Press. Munz, H. 1950: The Australian Wool Industry. Sydney: Angus and Robertson. Murray, T. 1996: Contact archaeology: Shared histories? Shared identities? In S. Hunt and J. Lydon (eds), SITES: Nailing the Debate: Interpretation in Museums, Sydney: Historical Houses Trust of New South Wales, 199–213. Murray, T. 2002: Epilogue: An archaeology of Indigenous/non-Indigenous Australia from 1788. In R. Harrison and C. Williamson (eds), After Captain Cook: The Archaeology of the Recent Indigenous Past in Australia, Sydney: Sydney University Methods Series vol. 8, Archaeology Computing Laboratory, 213–24. Oastler, J. 1908: Administration of justice in the back blocks. The Honorary Magistrate, July, 205–9. O’Loghlen, F. 1963: Cattle industry. In A. H. Chisholm (ed.), Australian Encyclopedia vol. 2., Sydney: Grolier Society of Australia, 292–9. Orser, C. E. 2002: Encyclopedia of Historical Archaeology. New York: Routledge. Osgood, E. S. 1957: The Day of the Cattleman. Chicago: University of Chicago Press. Paterson, A. 2000: Situating belonging: The role of historical archaeology in locating cultural relationships using examples from Australian pastoral domains. Balayi: Culture, Law, and Colonialism, 1 (2), 109–31. Paterson, A. 2003: The texture of agency: An example of culture contact in central Australia. Archaeology in Oceania, 38 (2), 53–67. Paterson, A., Gill, N., and Kennedy, M. 2003: Archaeology of historical realities? Two case studies of the short term. Australian Archaeology, 57, 82–8. Renfrew, C. and Bahn, P. G. 1996: Archaeology: Theories, Methods, and Practice (2nd edn.). London: Thames and Hudson. Reynolds, H. 1972: Aborigines and Settlers: The Australian Experience, 1788–1939. North Melbourne: Cassell Australia. Roberts, R., Jones, R., and Smith, M. 1990: Thermoluminescence dating of a 50,000 yearold human occupation site in northern Australia. Nature, 345, 153–6. Rose, D. B. 1991: Hidden Histories: Black stories from Victoria River Downs, Humbert River and Wave Hill Stations. Canberra: Aboriginal Studies Press. Rowse, T. 1987: ‘‘Were you ever savages?’’ Aboriginal insiders and pastoralists’ patronage. Oceania, 58 (1), 81–99. Sadr, K. 2002: Encapsulated Bushmen in the archaeology of Thamaga. In S. Kent (ed.), Ethnicity, Hunter-Gatherers, and the ‘‘Other’’: Association and Assimilation in Africa, Washington, DC: Smithsonian Institution Press, 28–47. Scott, P. 1987: Rural land use. In D. N. Jeans (ed.), Australia – A Geography, Vol. 2: Space and Society, Sydney: Sydney University Press, 201–36. Singh, G., Kershaw, A. P., and Clark, R. 1981: Quaternary vegetation and ﬁre history in Australia. In A. Gill, R. Groves, and I. Noble (eds), Fire and the Australian Biota, Canberra: Australian Academy of Science, 23–54. Smith, M. 1989: Seed gathering in inland Australia: Current evidence from seed grinders on the antiquity of the ethnohistorical pattern of exploitation. In D. R. Harris and G. C. Hillman (eds), Foraging and Farming: The Evolution of Plant Exploitation, London: Unwin Hyman, 305–17. Smith, P. A. 2002: Station camps: Legislation, labour relations, and rations on pastoral leases in the Kimberley region, Western Australia. Aboriginal History, 24, 75–97.

_____________________________________________________________________________________________________

291

_____________________________________________________________________________

Al i s t a i r P a t e r s o n Smith, P. A. and Smith, R. M. 1999: Diets in transition: Hunter-gatherer to station diet and station diet to the self select store diet. Human Ecology, 27 (1), 117–34. Woodhouse, M. C. A. 1993: Elements of a pastoral landscape: Holowiliena, South Australia, in 1888. Australasian Historical Archaeology, 11, 88–98. Yen, D. E. 1989: The domestication of environment. In D. R. Harris and G. C. Hillman (eds), Foraging and Farming: The Evolution of Plant Exploitation, London: Unwin Hyman, 55–75.

_____________________________________________________________________________________________________

292

_____________________________________________________________________________

-----------------------------------------------------------------------------

16

--------------------------------------------------------------------------

Conclusion Major Themes and Future Research Directions Peter Veth

This volume has for the ﬁrst time brought together regional studies of huntergatherers from deserts around the world. These different hunter-gatherer societies have experienced different evolutionary trajectories and have clearly adopted different organizational states through time. These organizational states include variations in settlement and mobility behaviors, territorial boundary making, technological strategies, and trade and exchange relations with neighboring groups. With respect to the interactions with these hunter-gather societies to adjacent groups, it is clear from all of the case studies that these were not simple demarcations with other hunter-gatherer, herder, pastoralist, or colonial settler societies. Desert hunter-gatherers, once established in new territories, established complex interactions with other land using groups. They might exercise rights to reciprocal access to other groups’ lands during times of resource scarcity, carry out herding on an opportunistic basis, or indeed swap into that mode of production for extended periods of time. Hunter-gatherers can also establish symbiotic production systems with colonial pastoralists and farmers or, somewhat alternatively, ‘‘devolve’’ into a hunting and gathering mode of production when faced with external competition or resource scarcity brought on by climate change. This collection of case studies from different deserts demonstrates overwhelmingly that arid zone hunter-gatherers are inﬁnitely ﬂexible and adaptable – both during their ‘‘colonizing’’ and ‘‘estate establishment’’ phases (see Chapter 1). Case studies have been presented from deserts that are located at both high and low altitudes, as well as on the coast and within the interior. These are both hot and cold deserts and landscapes that either concentrate or dissipate localized rainfall. Some of the desert systems, such as those of Australia, are enormous by global standards, while others, such as those of Patagonia, are quite circumscribed. To apply a formulaic set of predictions about what a desert hunter-gatherer society should look like, whether the analogue is from Australia, South Africa, or

_____________________________________________________________________________________________________

293

_____________________________________________________________________________

Pe t e r V eth

South America – and presumed to be in a ‘‘pristine’’ state – makes little sense in terms of the different biogeographic and historic settings of these groups. My own ethnoarchaeological and archaeological research with the Martu Aboriginal people of the Western Desert (a subset of members from one of the last societies in the world to come into contact with European colonialists in the 1960s; see Tonkinson 1991) aptly illustrates some of these points. Some members of this Western Desert group who were closer to better watered portions of northwest Australia came in – often under force or duress – from the desert in the early 1900s and focused their residency on government ration depots, missions, and pastoral and mining centers. While many Martu served as stockmen, maintenance teams, and domestic laborers, by the mid-1900s some had secured mining tenements or pastoral leases, and were actively engaged in the wider domestic economy. At most sedentary nodes, such as the pastoral stations, townships, and mining centers, there were (and still are) camps of transient groups who still align themselves with the more remote and essentially undeveloped interior desert lands. Today, the Common Law native title rights of Martu have been recognized through the Federal Court and this provides the right of both urban and remote constituents to engage in regional decision making, negotiate with land users, and establish appropriate governance structures. While members of the last Martu groups to have contact with Europeans in the 1960s were leading an unquestionably hunter-gatherer lifestyle, many of their countrymen had established alternative modes of existence and production around the fringes of the desert. The ability of these groups to engage in what was essentially a modern Western economy in such a compressed period of time, and the evidence for their active expansion, language spread, and territorial ascendancy in the millennium leading up to contact (see Chapters 2, 5, 6, 12, and 15), lend support to their inﬁnite adaptability to cope with new modes of production and land use. As highlighted in Chapter 1 and in the introductions to each of the three parts of the volume, there are a number of recurrent themes that are woven throughout the regional case studies whose comparative examination has only begun. These themes are worthy of consideration, as they offer researchers some of the most productive and outstanding questions worthy of future research in the world’s deserts. Rather than adopt a prescriptive theoretical approach here (e.g., an evolutionary ecological approach), I will aim to identify some of the core issues and then tease out areas of research likely to be productive for further comparative examination. It is expected that a range of interpretive and reconstructive approaches can be brought to bear on these problems. These core issues may be summarized as: 1 2 3

What kinds of environments qualify as deserts? Do groups need to be resident within such landscapes to qualify as desert societies, or might they use them in an opportunistic and episodic fashion? How can you discriminate from the archaeological record between early colonization and exploration of arid lands, as opposed to the establishment

_____________________________________________________________________________________________________

294

_____________________________________________________________________________

Conclusion

of continuous populations that assert territorial rights via domestic and totemic geographies? 4 Equally, given the apparent hyperﬂuidity of some of these groups, how can researchers understand mobility patterns that might involve episodes of local, regional, and even territorial abandonment – sometimes for considerable periods of time (cf. Kelly 1995; Veth 2003)? 5 What is the relationship between the perceptions and reality of marginality of resources in a desert landscape and the risk minimization measures adopted by different groups? 6 What are the different expressions of risk minimization adopted by these different groups? Themes that might be explored here include highly ﬂexible and responsive residential and logistical mobility patterns, standardization in the organization of technology, territorial ascendancy, and recruitment of allies on the margins to provide reciprocal access to lands. 7 Ramiﬁed kinship networks and long distance information exchange systems exist in one form or another in many desert societies; however, the ways they are expressed archaeologically will differ. The role of aggregation cycles, the nature of exchange goods, and the kinds of graphic systems employed both on portable objects and as rock art are worthy of further exploration. 8 The degree to which desert hunter-gatherers engage with other production systems, such as those of herders and pastoralists, and perhaps even swap between these strategies through time due to changing environmental and social circumstances, begs further comparative work. 9 The global timing for occupation of deserts by hunter-gatherers and how this may be linked to anatomically modern human behaviors (such as sea voyaging and the production of art) needs to be critically examined. Do the early sequences from Africa actually provide evidence for occupation of desertic landscapes before the emergence of Homo sapiens sapiens? 10 Finally, the implications of past human responses to climate change, and speciﬁcally the desertiﬁcation of lands that were once more lacustral, represents a topic of interest to all students of landscape evolution and human adaptation. These topics are expanded on further below, by way of seeding ideas and data-sets that might be taken up by the burgeoning number of archaeologists who now take it upon themselves to study deserts in a comparative fashion.

What Kind of Environments Qualify as Deserts? As Chapters 1 and 4 highlight, there is a clear need to accurately deﬁne what a desert is and then to focus on those elements of the desert landscape which are of consequence to hunter-gatherers. Is it appropriate to compare archaeological settlement patterns for hunter-gatherers from the Atacama and northwest Australian deserts – which front onto highly productive maritime provinces – with interior deserts such as Botswana or the Western Desert? Should hyperarid deserts

_____________________________________________________________________________________________________

295

_____________________________________________________________________________

Pe t e r V eth

(both dry and cold) be compared with those which in wetter cycles actually support livestock? Given that the boundaries between arid and adjacent semiarid areas are known to have moved through time – for example, in southeastern Australia where freshwater lakes become saline (see Johnston et al. 1998; Ross et al. 1992) – the relative richness, diversity, and abundance of key economic species are likely to have ﬂuctuated through time.

Do Groups Need to be Resident Within Such Landscapes to Qualify as Desert Societies, or Might They Use Them in an Opportunistic and Episodic Fashion? There is an ongoing debate in the Australianist literature about when it is that groups permanently map on to deserts (e.g., Smith 1989; Thorley 1998; Veth et al. 2001). Are the elements of a desert adaptation ﬁnally in place by the terminal Pleistocene or are early groups only using desert lowlands in an opportunistic fashion from better watered ranges and drainage systems (if you like, ‘‘refugia’’) and making the ﬁnal adaptation by the Holocene? The patterns appear to be complex and regionally variable.

What are the Expected Differences in the Archaeological Record between Early Colonization and Exploration as Opposed to the Establishment of Continuous Populations? The differentiation between colonization and establishment of groups in marginal settings is fraught with issues such as sample size, visibility, and taphonomic processes, let alone presumed changes in residential mobility. It appears that the majority of occupation sites in the Australian deserts that predate 20,000 bp are marked by small assemblages with low indices for permanency of occupation (see Chapter 6). Most authors agree that founding or early occupations were made up of small, highly mobile groups and that the processes may not be visible in the archaeological record – especially given the sandy matrices of most rock shelter sites. Where continuous records are available, there appears to be a quantum increase in both the size, breadth, and complexity of assemblages by the mid-to late Holocene – suggesting that a different kind of pattern of land use occurs – as well as signiﬁcant changes in population size.

Given the Hyperﬂuidity of Some Groups, How can Researchers Understand Mobility Patterns that Might Involve Episodes of Local, Regional, and Even Territorial Abandonment? Despite high concentrations of people in deserts during aggregation events, the general population densities across arid zone territories are generally low. Abandonment of country can occur for myriad reasons, including death, conﬂict, and prolonged drought (cf. Gould 1991; Veth 2003). Clearly, such patterns will only be discernable from regional studies which contain numerous dated sequences and

_____________________________________________________________________________________________________

296

_____________________________________________________________________________

Conclusion

even then there will be residual issues such as dating resolution compatible with seasonal/generational changes and climate phases.

What is the Relationship Between the Perceptions and Reality of Marginality of Resources in a Desert Landscape and the Risk Minimization Measures Adopted by Different Groups? If a range of mitigation measures is available to offset risks, then it is reasonable to pose the question: Do groups perceive that the relative paucity of resources and their patchiness across space and through time constitute a ‘‘marginal’’ environment? Ethnobotanical research and ecological modeling, and studies of prey choice and scheduling of companion foods, appear to render this concept somewhat redundant (cf. Veth 1993). This noted, it is fair to argue that the risks of deserts are real and the mitigation measures quite extreme on a global scale, including the exhaustive comminution of bone to extract protein (Gould et al. 2001) and the abandonment of entire territories during drought (after Kelly 1995).

What are the Different Expressions of Risk Minimization Adopted by Groups? Highly variable residential and logistical mobility patterns through time and in space are likely to be a hallmark of risk minimization (cf. Veth 2000), as is the standardization in the organization of technology, such as in the production of specialized implements (e.g., Hiscock 1994). Territorial expansion and the active ritual recruitment of ‘‘allies’’ in adjacent (and often more fertile) lands has also been argued to ensure reciprocal access to these lands (see Gibbs and Veth 2002). The role of aggregation cycles and locales in the exchange of ceremonial objects, durable goods, and genes is central to the active (re)negotiation of alliance and reciprocity networks (cf. Veth and McDonald 2002).

The Nature of Exchange Goods and the Kinds of Graphic Systems Employed at Aggregation Locales will Inform on Regional Alliances and Information Exchange It has been argued that the social contracts underpinning regional alliance networks are actively (re)negotiated at aggregation locales (Gibbs and Veth 2002; Veth and McDonald 2002). In the Australian deserts, group identifying behavior may be expressed through graphic systems, such as decoration on wooden artifacts and in pigment and engraved art. Graphic systems are seen to be either ‘‘inclusive’’ – such as geometric motifs (reafﬁrming regional alliance networks) – or ‘‘exclusive’’ – such as regionally distinctive anthropomorphs with unique headdresses – stressing territorial differences. The emergence of group identifying behaviors combining inclusive and exclusive graphic vocabularies in parallel is seen to be a key element in understanding the emergence of territoriality in desert societies (see Chapters 7 and 14).

_____________________________________________________________________________________________________

297

_____________________________________________________________________________

Pe t e r V eth

The Degree to Which Desert Hunter-Gatherers Engage with Other Production Systems, such as Herding and Pastoralism, Begs Further Comparative Work The archaeological interpretation of the complex relationships between huntergatherers and nearby herders and pastoralists has now begun in earnest (see Chapters 11 and 15). The comparative archaeology of the colonial occupation of lands fringing deserts occupied by hunter-gatherers has really only begun.

The Timing for Occupation of Global Deserts and How this may be Linked to Anatomically Modern Human (AMH) Behaviors such as Sea Voyaging and the Production of Art Archaeological evidence from Africa and Australia suggests that deserts were occupied at approximately the same time that Homo sapiens sapiens engaged in the systematic production of art, made personal adornments, and had the capacity to engage in successful sea voyaging ca. 60,000–40,000 years ago (see Chapters 1, 3, 4, and 6). Earlier evidence for the pre-AMH occupation of deserts in Africa, when they were better watered, remains contested. Recent arguments for pre-AMH sea crossings from Bali to Lombok (and then to Flores) have now been discounted (Mike Morwood pers. comm.).

Past Human Responses to Climate Change, and Speciﬁcally the Desertiﬁcation of Lands that were Once More Lacustral, Represents a Topic of Interest to all Researchers There are many questions that emerge for researchers interested in human responses to climate change when the global patterns for the evolution of desert societies are considered. Have modern humans always had the social systems to occupy deserts and cope with the onset of more arid conditions or is the permanent occupation of all desert habitats a more recent phenomenon – say of the midto late Holocene? If deserts represent such marginal landscapes, why were they occupied so early on? What was the role of population pressure in this process? Where occupation of more favorable landscapes was followed by more arid conditions, does persistence require a signiﬁcant change in economy, social organization, and territorial conﬁguration?

Prologue A Eurocentric perspective of deserts, as captured in the imagination of novelists, ﬁlmmakers, theologians, explorers, and perhaps even early anthropologists, has seen them as harsh and uncompromising landscapes. They are equated with extreme experiences, including revelations, religious epiphanies, treks through the ‘‘wilderness,’’ and privations of the most horrible kind.

_____________________________________________________________________________________________________

298

_____________________________________________________________________________

Conclusion

Yet archaeological evidence from around the world suggests that people actually chose to enter desert environments willingly or persisted in their occupation of landscapes that subsequently became arid. This may have occurred as early as 60,000–40,000 years ago. Persistence of modern humans in the glacial landscapes of Europe and Eurasia also occurred during this time, likely underpinned by complex information sharing behaviors – reﬂected in widespread sharing of iconographic objects such as the Venus ﬁgurines. Indeed, Gamble (2003) has noted parallels in regional occupation patterns in response to glacial cycles associated with the LGM from the Northern Hemisphere and the trends in occupation noted during LGM from the Southern Hemisphere (after Veth 1993). Occupation of marginal landscapes such as deserts appears to be dated as early as other modern behaviors, such as the development of artistic expression, and the ability to make ocean crossings. This creates the fascinating scenario in which modern humans appear to have colonized the world’s deserts at the same time that they developed other ‘‘modern’’ attributes underpinning a successful global diaspora, such as complex information exchange and maritime skills.

References Gamble, C. 2003: Keynote Address. Canberra: Unpublished paper presented at the 23 Degrees South: The Southern Deserts InterWorld Archaeology Conference, National Museum of Australia. Gibbs, M. and Veth, P. 2002: Ritual engines and the archaeology of territorial ascendancy. In S. Ulm, C. Westcott, J. Reid, A. Ross, I. Lilley, J. Prangnell, and L. Kirkwood (eds), Barriers, Borders, Boundaries: Proceedings of the 2001 Australian Archaeological Association Annual Conference, St Lucia: University of Queensland, 11–19. Gould, R. A. 1991: Arid land foraging as seen from Australia: Adaptive models and behavioral realities. Oceania, 62, 12–33. Gould, R. A., O’Connor, S., and Veth, P. 2001: Bones of contention: A reply to Walshe. Archaeology in Oceania, 37, 96–101. Hiscock, P. 1994: Technological responses to risk in Holocene Australia. Journal of World Prehistory, 8 (3), 267–92. Johnston, H., Clarke, P., and White, J. P. (eds) 1998: The Willandra Lakes: People and Palaeoenvironments. Sydney: Archaeology in Oceania, 33 (3), Oceania Publications. Kelly, R. L. 1995: The Foraging Spectrum. Washington, DC: Smithsonian Institution Press. Ross, A., Donnelly, T., and Wasson, R. 1992: The peopling of the arid zone: Human– environment interactions. In J. Dodson (ed.), The Naive Lands, London: Longman, 76–114. Smith, M. A. 1989: The case for a resident human population in the central Australian ranges during full glacial aridity. Archaeology in Oceania, 24, 93–105. Thorley, P. B. 1998: Pleistocene settlement in the Australian arid zone: Occupation of an inland riverine landscape in the central Australian ranges. Antiquity, 72, 34–45. Tonkinson, R. 1991: The Martu Aborigines: Living the Dream in Australia’s Desert. Fort Worth, TX: Holt, Rinehart, and Winston. Veth, P. M. 1993: Islands in the Interior: The Dynamics of Prehistoric Adaptations within the Arid Zone of Australia. International Monographs in Prehistory, Archaeology Series no. 3, Ann Arbor, MI.

_____________________________________________________________________________________________________

299

_____________________________________________________________________________

Pe t e r V eth Veth, P. 2000: Cycles of aridity and human mobility: Risk minimization amongst late Pleistocene foragers of the Western Desert, Australia. Philadelphia, PA: Unpublished paper presented at the Society of American Archaeology Conference. Veth, P. 2003: ‘‘Abandonment’’ or Maintenance of Country? A Critical Examination of Mobility Patterns and Implications for Native Title. Canberra: Native Title Research Unit, Australian Institute of Aboriginal and Torres Strait Islander Studies. Veth, P. and McDonald, J. 2002: Can archaeology be used to address the principle of exclusive possession in native title? In R. Harrison and C. Williamson (eds), After Captain Cook: The Archaeology of the Recent Indigenous Past in Australia, Sydney: Sydney University Archaeological Methods Series no. 8, 121–9. Veth, P., Smith, M., and Haley, M. 2001: Kaalpi: The archaeology of a sandstone outlier in the Western Desert. Australian Archaeology, 52, 9–17.

_____________________________________________________________________________________________________

300

_____________________________________________________________________________

General Index

Page numbers in bold refer to ﬁgures and tables abandonment 45–6, 85, 101, 103, 111–13, 118, 149, 188–9, 190–1, 195, 197, 199, 227, 229, 296–7 adaptationist approaches 2, 21–3, 26, 28, 34–5, 43–4, 48, 50, 79, 84–5, 89–90, 177, 194, 201–2, 216, 226, 256, 277, 294–6 agriculture 8, 24, 26–7, 162–5, 162, 166, 202, 257, 263, 269 alliance networks 66, 117–37, 192, 297 anatomically modern humans (AMH) 9, 37, 48, 62–4, 70, 161–74, 298 Arandic language, Australia 124, 223, 224, 230, 236, 238 Archaic Phase, North America 7, 94 classic approach 21 Clovis period, North America 7 coastal desert 3, 5, 48–9, 103, 177–202, 243–57, 295–6 coastal resources 9, 72, 172, 177–202, 244, 248, 247, 248, 250–2, 256–7 colonization 6–7, 9, 35–50, 72, 100–13, 142–55, 194, 231, 238, 243–4, 249–51, 256, 294, 296, 299, 347 cultural contact 19, 20, 24, 163, 206–17, 230, 261–73, 276–88 cultural ecology 1, 9, 65, 91

_____________________________________________________________________________________________________

demography 9, 21, 26, 42, 44, 87, 101, 118, 173, 199, 209, 227, 228, 231, 236 Desert Culture, North America 84–5 dingo 65, 237, 277, 280, 286 division of labor 87–9, 91–2, 94–5, 285 Dreaming, Australia 119, 120 drought 19, 72, 85, 101, 229, 234–5, 249, 263, 273, 277–8, 280, 285–6, 296–7 economy/economic strategy 6, 21, 35, 41–2, 46, 49, 65, 67–70, 72, 81, 105, 109, 118, 137, 167, 177, 194, 199, 251, 277, 288 evolutionary ecology 23, 81–96 ﬁlm 18–19, 25 foragers 1–2, 6, 20–1, 23–4, 26, 34–5, 41–6, 49, 67–8, 70–2, 79, 81–96, 100, 144, 146, 151, 200, 217 genetics 136, 229–30, 232, 238, 249, 271–2 group identity 65–7, 125–6, 130, 134–7, 163, 191, 263, 267, 297 Holocene 35, 41–3, 45–50, 59–61, 65, 68, 72, 93–5, 103–5, 108–11, 117–18, 122–3, 130, 142–54, 162–3, 172–3, 177, 179–80, 188, 190–3, 195–7, 199–202,

301

_____________________________________________________________________________

Index 206–17, 222–39, 246–57, 247, 248, 252, 253, 262–73, 276–7, 296, 298 human behavioral ecology 2, 23, 63–4, 70, 81–96, 295

reﬂecting paradigm 25–6 revisionist approaches 21–4, 28, 210 risk minimization 2, 10, 42, 67–70, 72, 85, 100–13, 116, 295, 297

indigenist approach 22–3, 29 Indigenous rights 22, 23, 25, 27, 294 inﬂecting paradigm 27–9 intensiﬁcation 200–2 Iron Age, Africa 159, 162–3, 165, 208–17

sea voyaging 9, 194, 250, 295, 298 sedentism 6, 87–8, 201–2, 228, 231, 257, 263, 269, 294 sediment accumulation 111, 113, 149, 169, 190–1, 195, 197, 227–8 settlement patterns 21, 37–8, 145, 147 social networks/organization 6, 10, 21, 34, 41, 47, 63, 65, 70–3, 79, 84–5, 87, 89, 95, 116–37, 124, 145, 162, 165, 167–8, 172, 192, 196, 200, 225, 243–4, 254–5, 266–7, 272, 276, 281, 283, 288, 295, 298 socioeconomic model 1, 6, 65, 82–3, 86–7, 89, 91, 95, 196 subsistence 35, 47, 79, 81–96, 103, 119, 142, 150–1, 154, 168, 170–2, 179–80, 193–5, 214–216, 244, 255–6, 262, 276, 285–7

Kalahari Debate 8, 18, 24, 165, 206, 208–9 kin/kinship 29, 84–8, 116, 236, 295 linguistics 8, 23, 110, 116–17, 122–7, 159, 163, 208–9, 222–39, 273 Man the Hunter conference 1, 26, 71 middens 10, 19, 178, 179–80, 190, 192–3, 196–7, 200–1, 247–8, 247, 248, 251–3, 252, 253 middle range theory 29 mobility 1, 2, 6, 9, 21, 27–8, 30, 44, 69, 84–5, 87, 90, 95, 100–13, 101, 107, 116–37, 118, 145, 151–3, 192, 201, 209, 225, 249, 272, 293, 295–7 optimal forager theory 26 Paleolithic (African) 60–2, 206, 210, 214, 216 pastoralists 20, 26–8, 159, 162–3, 165, 206–17, 261–73, 276–88, 281, 293–4, 298 patch utilization model (HBE) 90–5 Pleistocene 2, 7, 9, 15, 34–50, 59–63, 72, 93–4, 100–13, 118, 122, 125–6, 130, 136, 142–4, 150–1, 153–4, 161–74, 180–90, 188–9, 193–6, 202, 226–7, 229, 231, 233–4, 243, 245–6, 248–51, 296 prey choice model (cultural ecology) 90–5, 297

_____________________________________________________________________________________________________

taphonomy 36, 103–4, 147–8, 150–1, 199, 256, 296 technology/technological organization 10, 42, 46, 58, 63, 67–9, 72–3, 79, 84–6, 94–5, 106, 119, 144, 152–4, 162, 171, 174, 196, 199, 214, 217, 231–2, 236–7, 239, 244, 251–2, 255–6, 280, 284, 287–8, 293, 295, 297 territorial organization 1, 6, 34, 69, 71, 84–5, 101, 112, 118–19, 124, 125, 144, 170, 192, 200–1, 244, 293, 295, 297–8 water availability 3, 6, 34–5, 40–9, 73, 85, 93–4, 100–3, 110, 112, 117–18, 127, 144–7, 149, 151, 153, 163, 165, 168–71, 173, 201, 208, 214, 235, 244, 246–7, 249, 250, 262–3, 278–80

302

_____________________________________________________________________________

Index of Archaeological Features and Subjects

archaeological sites Africa Apollo 11, 162, 166, 170–3 Balerno 213 Bir Tarfawi 7 Blydefontein 171 Border Cave 161–2, 162 Cae Cae 207, 208, 213–14 Chin Taﬁdet 263 Cho/ana 214 Daures 162, 166, 173 Depression Cave 162, 170–1 Divuyu 213 Dobe 162, 173, 206 Driekoppen 162 Elands Bay 162, 170–3 Enkapune Ya Muto 61 Free State 171 Geduld 207, 209, 215–16 Gi 162, 170 Haalenberg 162, 170 Iwelen 263 K2 213 Kalambo Falls 61 Klasies River 60, 66 Kwihabe 170–1 Leokwe Hill 213 Little Muck Shelter 212–13 Mapungubwe 213 Mauretania 272

_____________________________________________________________________________________________________

Mirabib 171–2 Moritsane 211 Nos 162, 170 Nqoma 213 Oruwanje 216 Ostrich Shelter 211–12 Pockenbank 170 Radiepolong 211–12 Rose Cottage Cave 63 Salt Pan site 212 Schroda 213 Sibudu Cave 63 Thamaga sites 207, 208–14 Toteng 7, 162, 173, 207, 208, 215–16 Tsodilo Hills 47, 162, 166, 169–70, 207, 208, 213–14 Twin Rivers 61 Uan Afuda 264–5 White Paintings 162, 170, 173 Wonderwerk Cave 162, 166, 171–3 Zebrarivier 162, 171–2 Australia Allen’s Cave 37–8, 38, 117 Anadara mound 187 Babjarimanos 1 midden 178, 182 Babjarimanos 2 midden 178, 182 Barrow Creek Telegraph Station 288 Burrup Peninsula 117, 122–3, 128, 131–4, 131, 133, 134, 178

303

_____________________________________________________________________________

Index archaeological sites (Continued) C99 site 177, 178, 180–1, 182, 183, 188–9, 190–1, 194–7 Carpenter’s Gap 1, 37–8, 38, 40, 43–4, 121–2, 125 Clarke’s Cave 187 Colless Creek 111 Coral Bay midden 178, 183–4 Cuddie Springs 37–8, 38, 43–4, 237 Denham 186 Durba 102, 110–11, 117, 128, 131–4, 131, 136, 233–4 Eagle Bluff 178, 185–6, 199 Emily Gap 120 Giralia Road site 9 178, 183 Giralia Road site 11 178, 183 GRE8 37–8, 38 Haynes Cave 178, 182 Heirisson Prong sites 178, 185 High Cliffy Island 178, 181 Intirtekwerle 104 Intitjikula 104 Jalijbang 2 117, 128, 130–4, 131 Jansz 177, 178, 180, 183, 188–9, 190–1, 194–7 JSN 46 Kaalpi 102, 104, 110–11, 117, 127, 128, 228 Katampul 117 Koolan 178, 181, 196 Kulpi Mara 102, 102–6, 105, 108–9, 108, 112, 117, 226–7 Lake Mungo 9, 36–8, 38, 40, 45 Laura 122 Lawn Hill 46 Louie Creek Gorge 46 Low Point midden 178, 184 Mandu Mandu Creek 103, 178, 180, 184, 195–6 Mangrove Bay midden 178, 183 Monkey Mia site 178, 187 Mulanda Bluff 178, 184 Nickol River Mound 187 Noala Cave 103, 178, 181, 182, 193, 195 Notch Point 185 Not-So-Secret Shelter 187 NW Cape midden 2 178, 182 Pilgonaman Creek 178, 180, 184, 196

_____________________________________________________________________________________________________

Pilgramunna Bay midden 178 Puntutjarpa 102, 104, 117, 136, 226–8, 232 Puritjarra 37–8, 38, 44, 102, 102–6, 105, 106, 108–9, 108, 111–12, 117, 121–2, 226–8, 230–2 Riwi 37–8, 38 Rudall River sites 83, 228, 233 Serpent’s Glen 102–6, 102, 104, 105, 108–11, 108, 117, 226–8, 232, 238 Silver Dollar 178, 186, 194, 199 Strangways Springs Station sites 278, 282–3, 285–6 Sturt’s Meadows 125 Tulki Bay midden 178, 184 Turquoise Bay midden 178, 184 Ulungra Springs 43–4 Useless Inlet 178, 185 Wadjuru Rockpool 187 Walga Rock 227 Warroora midden 178 Whale Well 185 Widgingarri 178, 181, 195–6 Wilga Mia 234 Winpuly 228 Wobiri 178, 180, 183, 188–9, 191 Wybalenna Mission 288 Yardie Well 178, 180, 185 Zuytdorp sites 187 South America Aragon 245, 252 Arroyo Feo 152 Bano Nuevo 149, 151 Camarones 245, 248, 253 Chinchihuapi 146, 153 Cueva Cuyin Manzano 143, 147–8 Cueva del Medio 256 Cueva Epullan Grande 143, 147–9, 153 Cueva Traful 143, 147, 151–2 El Ceibo 152 Las Conchas 245, 250, 254 Los Burros 245 Los Toldos 152, 256 Mariﬁlo 1 146 Monte Verde 145–6, 150–1, 153, 256 Patillo 245, 253–4 Piedra Museo 256 Punta Pichalo 245, 251–3, 252

304

_____________________________________________________________________________

In d e x Quebrada Jaguay 245, 250–1, 256 Quebrada Tacahuay 245, 250–1, 256 Quiani 9, 253 Ring Site 245, 250 Tacahuay 251 Tiliviche 245, 252–3, 253 cultural groups Africa BaSarwa 163, 206 Beja 271 Berber 269, 271–2 Fulani 265, 267, 271, 272 Hadza 25 Hai//om 20, 28 Harratin 263 Himba 266 Ju/’hoansi 1, 19, 22, 162, 165, 167–8 Kabyle 268–9, 271 !Kung 2, 19, 66, 71, 165, 206, 208 Maures 267, 270–1 Nemadi 270–1 Quena 164 San 18–20, 22–4, 27–8, 66, 70–1, 162, 163–74, 206–17 Sonqua/Soaqua 163 Toubou 267, 271–2 Tuareg 261, 264, 267, 268, 269, 271, 272 Australia Alyawara (Alyawarr) 83, 90–4, 236 Arabana 281 Arrernte 1, 83, 224, 230, 232–3, 236–8 Gugadja 91 Jangkunjara 235 Martu/Mardu 83, 88, 93, 107–8, 110, 111, 294 Ngatatjara 83, 85 Pintubi 1, 83, 223, 229 Pitjantjatjara 1, 83, 88, 223, 229, 232–5 Wardaman 128, 130–4 Warlpiri 83, 119, 122, 230 Yankunytatjara 229 Yolngu 232 South America Selknam 244 Tehuelche 142, 149 North America Numic 84, 225

_____________________________________________________________________________________________________

Paiute 1, 2, 82, 87–8, 225 Shoshone 1, 2, 82, 87–8, 225 dating techniques 236 Accelerator Mass Spectometry (AMS) 121 ESR dating 36 luminescence dating 10, 36, 37, 121 radiocarbon dating 35, 37, 148, 181–7, 193, 198, 199 U-series dating 10, 36 faunal remains 10, 20, 42, 69, 94, 103–5, 105, 147–53, 169, 172, 180–90, 193–4, 211–13, 215, 250–2, 256, 262, 264–5, 272 Index of Fragmentation 104–5, 105, 113 megafauna 8, 42–3, 150–1, 154, 168–9, 171–3, 213 protein stress 103–5 ﬂoral remains 10, 103, 146, 171–2, 192 seeds 47, 84, 87–8, 91, 93–5, 172, 237 geographical locations Africa 1–3, 4–5, 6–7, 9, 19–20, 22, 24–5, 27, 47–8, 58–73, 161–74, 162, 206–7, 207, 261–73, 277, 287, 295, 298 Cape Fold Mountains 169, 172 Erongo Mountains 162, 173 Euphrates River 1 Kalahari Desert 3–4, 5, 7, 18, 20, 22, 24, 47, 71, 73, 159, 161–74, 162, 206–17, 207, 244, 248 Karoo Desert 212 Nama-Karoo 162, 163–4, 169, 173, 212 Namib Desert 3, 5, 7, 20, 47–8, 162, 168–9, 211, 244, 246, 248, 250 Northern Cape Province 161–74, 162, 209 Nyae Nyae 19 Okavango Delta 162, 168, 173, 207, 208, 215 Orange River 162, 168–9, 206–7, 207, 209 Qangwa 208 Sahara Desert 3, 4, 7, 48, 58, 60–1, 63, 68, 71, 262–7, 269–70, 272–3 Sahel Desert 48, 72, 272

305

_____________________________________________________________________________

Index geographical locations (Continued) Seacow Valley 162, 171, 173 Shashe Limpopo 207, 212–14, 216–17 Sinai Desert 261 Thamalakane River 215 Thukula Basin 66 Trafal River 147, 150–2 Atlantic Ocean 4, 143 Australia 1, 5, 6–10, 18–20, 22, 27, 34–50, 38, 39, 58–73, 81–96, 83, 100–13, 102, 116–37, 117, 128, 159, 177–202, 178, 198, 211, 216, 222–39, 244, 247–8, 256, 276–88, 279, 293, 295, 297–8 Alice Springs 102, 117, 128 Barkley Tablelands 37, 279, 280, 285 Calvert Ranges 118, 122–3, 125, 127–8, 130, 131–6, 131, 135, 227–8, 234 Canning Stock Route 123, 127, 136 Cape Range Peninsula 177–202 Cape York 46, 232, 246 Central Desert 9, 37, 40–1, 43–4, 73, 82, 83, 90–3, 102–4, 111–12, 122–3, 124, 222–39, 279, 283–4, 286 Cleland Hills 111, 128, 128–9, 130–6, 131, 135, 230, 234 Davenport Ranges 284 Everard Ranges 235 Finke River 41, 83 Gascoyne-Murchison region 102, 117, 224, 233–4 George Gill Ranges 235 Gibson Desert 102, 117 Great Sandy Desert 39–40, 102, 117, 228, 234 Great Victoria Desert 102, 117 Hammersley Ranges 102, 117, 123 Kimberley 20, 37, 39, 44, 46, 83, 102, 117, 128, 178, 193–4, 121–3, 125, 136, 193, 195–7, 199, 201, 227, 233, 279–80, 283–6, 288, 294–5 Lake Amadeus 41 Lake Disappointment 83, 102, 117, 127, 227 Lake Eyre 9, 39–40, 38, 39, 43, 83, 278–9, 283, 285 Lake Gregory 39–40 Lake Wood 40

_____________________________________________________________________________________________________

Little Sandy Desert 110, 118, 127 MacDonnell Ranges 37, 83, 102, 117, 230 Monte Bello Islands 178, 193 Murchison Ranges 102, 117, 200, 227, 284–6, 288 Musgrave Ranges 235 North West Cape 177–202, 178, 188–9, 198 Nullabor Plain 37, 102, 117 Palmer River 41, 228 Percival Lakes 136 Pilbara 83, 111, 123, 128, 136, 178, 193, 195, 200, 224, 226–7, 229, 233–4, 279 Simpson Desert 102, 117 Strzelecki Desert 46 Tanami Desert 102, 117, 224 Warburton 19, 227–8 Western Desert 3, 5, 18–19, 68, 72–3, 81–96, 83, 100–13, 108, 116–37, 117, 118, 124, 222–39, 223, 294, 295 Willandra Lakes 39–40, 39, 43, 45 Wiluna 226 Egypt 261, 273 Nabta Playa 262 Nile Valley 265, 271–2 Western Desert 262, 265 Europe 27, 65, 87, 111, 163–5, 167, 177, 208, 216, 243, 256–7, 277–9, 281–8, 299 Indian Ocean 4, 83, 179, 206, 210–11, 213 Indonesia 9, 250, 298 Levant 273 Near East 1, 8, 262, 273 Tigris River 1, 273 North America 1, 2, 3, 4, 7, 10, 20, 22, 48–50, 81–96, 82, 143, 154, 225–6, 243, 249, 256, 281 Great Basin 2, 3, 4, 48–50, 81–96, 82 Humboldt River 82, 87, 89 Mojave Desert 7 Paciﬁc Ocean 4–5, 82, 143, 244, 247, 249–50 South America 4, 6–7, 10, 47–50, 142–54, 143, 152, 243–57, 245, 247, 248, 294

306

_____________________________________________________________________________

In d e x Andes Cordillera 3, 145–6, 150, 152–3, 244, 246–7, 249, 251, 254–7 Atacama Desert 3, 4, 6, 49, 243–57, 245, 248, 295 Cardiel Lake 144–5 Chile 3, 143, 159, 246, 248, 250, 256 Colombia 249 Cupisnique Desert 246 Deseado River 152–3 Fertile Coast 248 Lago Argentina 145 Limay River 143, 146–7 Pampa de los Fosiles 246 Pampa del Tamarugal 247, 249, 252, 257 Pampas 151 Patagonia 3, 4, 49, 79, 142–54, 143, 256, 293 Peru 243, 246, 248, 253, 254 Sechura Desert 246 Tierra del Fuego 49, 142–54, 244 human skeletal remains 36, 148–9, 161, 172–3, 244, 253–5, 263, 269 cemeteries 263, 267, 269–70, 272–3 lithics 8–9, 44, 58–73, 93–4, 101, 103, 105–10, 105, 106, 108, 118, 144, 146–8, 152–3, 161–74, 190–1, 199–200, 211, 212, 215, 227, 231, 233, 237, 249–52, 256, 273, 285 Acheulian Industry 7, 48 adzes 47, 108–9, 191, 199, 231 Albany Industry 61 backed artifacts 58–73, 106, 108–9, 161–74, 191, 199, 231, 233 discard rates 42, 44, 59, 105–6, 118 grinding 34, 42, 46–7, 49, 84, 87, 91, 93–5, 107–10, 110, 195, 231, 233, 237, 277, 284–5 Howieson’s Poort Industry 60–1, 63, 66, 68–70 Kuruman Industry 171 Later Stone Age 24, 61, 63, 66, 69–72, 161–74, 162, 208–9, 213–15, 217, 295 Lemuta Industry 61 Levallois Industry 48 Level 11 Industry 152 Lockshoek Industry 171, 173

_____________________________________________________________________________________________________

Lupemban Industry 60, 63 Middle Stone Age 7, 60–1, 63, 66, 68–72, 161–2, 166, 170, 295 Mousterian Industry 48 Nasampolai Industry 61 Nasera Industry 61 Oakhurst Complex 171–2 Paleo-Indian assemblages 93–5, 256 reduction sequences 105–6, 152 Robberg Industry 61, 70, 72, 170–1 Springbokoog Industry 162, 173 style 67 Toldense Industry 152 use wear 108, 191 Wilton Industry 61, 63, 172–3 paleoenvironment 2, 6–7, 9–11, 34–50, 68–9, 94–5, 100–1, 104, 112, 118, 143–4, 147, 150, 168–9, 171, 174, 188–9, 190, 197, 208, 227, 231, 246, 249, 250, 262–3, 266, 272, 293–8 coral records 10 diatoms 10 foraminifera 10 hyperaridity 3, 6, 47–8, 100–1, 246, 296 ice cores 10, 169 isotopes 10 Last Glacial Maximum 35, 43, 45–8, 50, 70, 72, 100, 103, 105, 112, 117, 125, 169, 179–80, 190, 194–6, 227, 229, 299 marine cores 10, 39, 169, 298 mud wasp nests 10 phytoliths 10, 40, 227, 228 pollen 10, 147, 169 sea level change 43, 180, 190, 196, 250 speleothem 10 stick nest rat middens 10 tufa 10 people Dunlop, Ian 18–19, 25 Jennings, Jessie 84–5 Marshall, John 19, 25, 209 Sahlin, Marshall 21, 71 Steward, Julian 1–2, 23, 82, 84, 87–9 pottery 262 Bambata Cave ware 212, 215–16 Dotted Wavy-Line ware 271 Eiland ware 211

307

_____________________________________________________________________________

Index pottery (Continued) Geduld pottery 215–26 Happy Rest ware 212 Moloko ware 211 rocker stamped pottery 271, 272 style 216 Wavy-Line ware 271, 272 rock art 79, 110–12, 116–37, 124, 129, 130, 131, 132, 133, 134, 135, 148–9, 161, 165–7, 172, 180, 188, 190–2, 197, 200, 208, 212, 227, 228, 229, 230–1, 233–4, 244, 263–5, 267, 273, 281, 284, 295 Bradshaws 121–2, 125 Burrup Peninsula 117, 122–3, 128, 131–4, 131, 133, 134, 178 dating rock art 110, 121–2, 127, 166, 191, 234

_____________________________________________________________________________________________________

ocher 102–3, 110–13, 118, 121, 136, 149, 180, 188, 190–1, 213, 227, 230–1, 234, 251, 255, 283, 285–6 Panaramitee 119, 122, 125, 127, 136, 234 style 110, 116, 126–7, 133–7, 191, 230–1 symbolism 62–70, 79, 110–11, 116–37, 144 Wandjina 123 trade/exchange 20, 42, 61, 66, 109, 111–12, 117–18, 123, 124, 136–7, 163, 167–8, 172, 194, 206, 211, 213–14, 222, 251, 285–6, 293, 295, 297 exotic raw materials 70, 72, 101, 105–6, 109, 152–3, 191, 200, 211, 213, 251 hxaro system 70, 167–8, 172 reciprocity 34, 70–2, 116, 137, 293, 295, 297

308

_____________________________________________________________________________